Skip to main content

Full text of "Journal of agricultural research"

See other formats 


x^. 


J 


y,T"  ^'-^£ 


JOURNAL  OF 

AGRICULTURAL 

RESEARCH 


Volume  XII 

JANUARY  7— MARCH  25,  191 8 


BOTANICAL 


PUBLISHED  BY  AUTHORITY  OF  THE  SECRETARY  OF  AGRICULTURE 

WITH    THE    COOPERATION    OF    THE    ASSOCIATION    OF    AMERICAN 

AGRICULTURAL  COLLEGES  AND  EXPERIMENT  STATIONS 


WASHINGTON,  D.  C. 


V/.   \3l 


EDITORIAL  COMMITTEE  OF  THE 

UNITED  STATES  DEPARTMENT  OF  AGRICULTURE  AND 

THE  ASSOCIATION  OF  AMERICAN  AGRICULTURAL 

COLLEGES  AND  EXPERIMENT  STATIONS 


FOR  THE  DEPARTMENT 

KARL  F.  KELLERMAN,  Chairman 

Physiologist  and  Associate  Chief,  Bureau 
of  Plant  Industry 

EDWIN  W.  ALLEN 

Chief,  Office  of  Experiment  Stations 

CHARLES  L.  MARLATT 

Entomologist  and  Assistant  Chief,  Bureau 
of  Entomology 


FOR  THE  ASSOCIATION 

RAYMOND  PEARL* 

Biologist,  Maine  Agricultural  Experiment 
Station 

H.  P.  ARMSBY 

Director,  Institute  of  Animal  Nutrition,  The 
Pennsylvania  State  College 

E.  M.  FREEMAN 

Botanist,  Plant  Pathologist,  and  Assistant 
Dean,  Agricultural  ExperiTnent  Station  of 
the  University  of  Minnesota. 


All  correspondence  regarding  articles  from  the  Department  of  Agriculture  should  be 
addressed  to  Karl  F.  Kellerman,  Journal  of  Agricultural  Research,  Washington,  D.  C. 

*  Dr.  Pearl  has  undertaken  special  work  in  connection  with  the  war  emergency ; 
therefore,  until  ftirther  notice  all  correspondence  regarding  articles  from  State  Experi- 
ment Stations  should  be  addressed  to  H.  P.  Armsby,  Institute  of  Animal  Nutrition, 
State  College,  Pa. 
II 


CONTENTS 

Effect  of  Time  of  Digestion  on  the  Hydrolysis  of  Casein  in  the       Page 

Presence  of  Starch.     J.  S.  McHargue i 

Behavior  of  Sweet   Potatoes  in  the  Ground.     Heinrich  Has- 

SELBRING 9 

Studies  in  Soil  Reaction  as  Indicated  by  the  Hydrogen  Electrode. 

J.  K.  Plummer 19 

Pure  Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media.     W.  H. 

Long  and  R.  M.  Harsch 33 

Gossypol,  the  Toxic  Substance  in  Cottonseed.     W.  A.  Withers 

and  Frank  E.  Carruth 83 

Fruit-Fly  Parasitism  in  Hawaii  During  191 6.     C.  E.  PembERTOn 

and  H.  F.  Willard .• 103 

Irrigation     Experiments    on    Apple-Spot     Diseases.     Charles 

Brooks  and  D.  F.  Fisher 109 

Relation  of  Carbon  Dioxid  to  Soil  Reaction  as  Measured  by  the 

Hydrogen  Electrode.     D.  R.  Hoagland  and  L.  T.  Sharp.  .  .  .        139 
A  Study  of  the  Plow  Bottom  and  Its  Action  Upon  the  Furrow 

SHce.     E.  A.  White i49 

Influence  of  Nitrates  on  Nitrogen-Assimilating  Bacteria.     T.  L. 

Hills 183 

New-Place  Effect  in  Maize.     G.  N.  Collins 231 

Relation  of  the  Variability  of  Yields  of  Fruit  Trees  to  the  Accu- 
racy of  Field  Trials.     L.  D.  Batchelor  and  H.  S.  Reed 245 

Interrelations  of  Fruit-Fly  Parasites  in  Hawaii.     C.  E.  Pember- 

TON  and  H.  F.  Willard 285 

Water  Extractions  of  Soils  as  Criteria  of  their  Crop-Producing 

Power.     John  S.  Burd 297 

Effect  of  Season  and  Crop  Growth  in  Modifying  the  Soil  Extract. 

Guy  R.  Stewart 311 

The  Freezing-Point  Method  as  an  Index  of  Variations  in  the  Soil 

Solution  Due  to  Season  and  Crop  Growth.     D.  R.  Hoagland.  .       369 
Efficacy  of  Some  Anthelmintics.     Maurice  C.  Hall  and  Win- 

throp  D.  Foster 397 

Tobacco  Wildfire.     Frederick  A.  Wolf  and  A.  C.  Foster  ....       449 
Gipsy-Moth  Larvae  as  Agents  in  the  Dissemination  of  the  White- 
Pine  Blister-Rust.     G.  Flippo  Gravatt  and  G.  B.  Posey 459 

Influence  of  Carbonates  of  Magnesium  and  Calcium  on  Bacteria  of 
CC  Certain  Wisconsin  Soils.    H.  L.  Fulmer 463 


■< 


IV  Journal  of  Agricultural  Research  voi.xn 

Humus  in  Mulched  Basins,  Relation  of  Humus  Content  to  Orange 

Production,   and   Effect   of    Mulches  on   Orange    Production.       Page 
Charles  A.  Jensen 505 

Relation  of  Kinds  and  Varieties  of  Grain  to  Hessian-Fly  Injury. 
James  W.  McColloch  and  S.  C.  Salmon 519 

Wilt  Diseases  of  Okra  and  the  Verticillium-Wilt  Problem.     C.  W. 

Carpenter 529 

Winter  Cycle  of  Egg  Production  in  the  Rhode  Island  Red  Breed 

of  Domestic  Fowl.     H.  D.  GoodalE 547 

Digestion  of  Starch  by  the  Young  Calf.     R.  H.  Shaw,  T.  E.  Wood- 
ward, and  R.  P.  Norton 575 

Toxicity  of  Volatile  Organic  Compounds  to  Insect  Eggs.     Wil- 
liam Moore  and  Samuel  A.  Graham 579 

Corn-Stover  Silage.     J.  M.  Sherman  and  S.  L.  Bechdel 589 

Weevils  Which  Affect  the  Irish  Potato,  Sweet  Potato,  and  Yam. 
W.  Dwight  Pierce 601 

Sterility  in  the  Strawberry,     W.  D.  Valleau 613 

Effect  of   Nitrifying   Bacteria  on-  the  Solubility  of  Tricalcium 

Phosphate.     W.  P.  KellEY 671 

Respiration  of  Stored  Wheat.     C.  H.  Bailey  and  A.  M.  Gurjar.  .       685 

Effects  of  Mistletoe  on  Young  Conifers.     James  R.  Weir 715 

Determination  of  Fatty  Acids  in  Butter  Fat.     E.  B.  Holland  and 
J.  P.  Buckley,  jr 719 

Index 733 

ERRATA  AND  AUTHORS'  EMENDATIONS 

Page  60,  Table XIII,  heading,  " Folyporus  dryophilus"  should  read  " Polyporus  dryophilus." 

Page  105,  Table  I,  "  Peach  {Prunus  persica)"  should  read  "  Peach  {Amygdalus  persica)." 

Page  331,  line  8,  "clay"  should  read  "clay  loam." 

Page 335,  hne  5  from  bottom,  "(fig.  3-6)"  should  read  "(fig.  8-20)." 

Page  361,  line  18  from  bottom,  omit  "  10." 

Page  366,  citation  27,  "  1917"  should  read  "  1918." 

Pages  364-368,  in  citations  2,  16,  29,  30,  56,  and  57,  omit  "  Not  seen," 

Pages  394-395,  citations  3  and  8,  "  1917"  should  read  "  1918." 

Page  429,  Une  19  from  bottom,  "or  against  tapeworms"  should  read  "and  against  tapeworms." 

Page  430,  line  6,  "instability"  should  read  "irritability." 

Page455,  line  5,  "stroke"  should  read  "streak." 


ILLUSTRATIONS 

PLATES 

GossYPOL  THE  Toxic  Substance  in  Cottonseed 

Plate  I.  Effect  of  feeding  cottonseed  feeds  to  pigs:  A. — Pig  3,  showing  con-  page 
dition  on  the  ninety-fourth  day  on  a  feed  containing  ether-extracted  cot- 
tonseed kernels.  B . — Pig  3 ,  showing  condition  on  the  fiftieth  day.  C. — Pig 
4,  showing  condition  on  the  twenty-seventh  day  on  a  feed  containing  vita- 
mines.  D. — Pig  I,  showing  condition  on  the  fiftieth  day  on  a  feed  contain- 
ing cottonseed  meal.  See  figure  G.  E. — Pig  2,  showing  condition  on  the 
twenty-seventh  day  on  a  feed  containing  gossypol.  F. — Pig  3,  showing 
condition  on  the  twenty -seventh  day  on  a  feed  containing  ether-extracted 
cottonseed  kernels.  G.^Pig  i,  showing  condition  on  the  fiftieth  day  on  a 
feed  containing  cottonseed  meal 102 

Irrigation  Experiments  on  Apple-Spot  Diseases 

Plate  2.     A. — Early  stage  of  bitter-pit  on  Northern  Spy  apple.     B. — Cross 

section  of  the  apple  shown  in  A 138 

Plate  3.  A. — Late  stage  of  bitter-pit  on  Rhode  Island  Greening  apple.  B.— 
Internal  bro^\^ling  accompanying  bitter-pit.  C— Jonathan-spot  on  Jona- 
than apple.  D.— Early  stage  of  drouthspots  on  a  Winesap  apple  from 
Wenatchee,  Wash.  E. — Late  stage  of  drouthspots  on  a  Winesap  apple. 
F. — Cross  section  of  the  apple  sho\vn  in  E 138 

Plate  4.  A. — An  apple  orchard  showing  the  furrow  system  of  irrigation  em- 
ployed in  the  experimental  work  at  Wenatchee,  Wash.  B. — ^Jonathan 
apple  tree  showing  the  effects  of  drouth,  Wenatchee,  Wash 138 

Plate  5.  A. — Cork  on  Yellow  Newtown  apple  from  Hood  River,  Greg.  B. — 
Cross  section  of  the  apple  shown  in  A.  C. — White  Pearm.ain  apple  showing 
the  severity  of  the  1915  drouth  at  Wenatchee,  Wash.  D. — Cork,  or  "dr\-- 
rot, "  on  a  King  apple.  E. — Blister  on  an  Esopus  apple  from  Entiat,  Wash. 
F. — An  extreme  ca;se  of  Yorkspot  on  a  York  Imperial  apple.  G. — Cross 
section  of  the  apple  shown  in  F 138 

A  Study  of  the  Plow  Bottom  and  its  Action  Upon  the  Furrow  Slice 

Plate  6.  A. — A  plow  bottom  with  two  sets  of  straight  lines.  B. — A  plow  bot- 
tom, the  surface  of  which  is  composed  of  each  of  two  surfaces.  C. — A  plow 
bottom  similar  to  B,  but  vnth  the  surfaces  merging  into  each  other  farther 
back  on  the  moldboard.  D. — A  plow  bottom,  the  surface  of  which  does  not 
contain  an  infinite  set  of  straight  lines 182 

Plate  7.  A. — A  plow  bottom  with  a  convex  surface  which  has  two  sets  of 
straight  lines.  B. — Instrument  for  measuring  the  space  coordinates  of  any 
point  of  the  plow  bottom.  C. — A  sod  plow  showing  the  furrow  slice  turned 
by  it 182 

Plate  8.  A.— Rows  of  wooden  pins  driven  into  the  sod  for  estimating  the 
stretch  of  the  furrow  slice.  B. — Furrow  slice  showing  the  position  of  the 
pins  when  on  the  moldboard 182 

Plate  9.  A. — Plow  showing  attachment  used  to  obtain  the  x,  y,  and  2  coordi- 
nates of  points  in  the  furrow  slice.  B. — Moldboard  showing  the  paths  of 
five  soil  particles.  C. — Measurement  of  the  angle  Ny  by  use  of  a  protractor 
and  a  plumb  bob 182 

(V) 


VI  Journal  of  Agricultural  Research  voi.xii 

Interrelations  ok  Fruit-Fly  Parasites  in  Hawaii 

Plate  io.     Diachasvm  tryoni:  A. — Freshly  hatched  larva  with  its  mandibles     Page 
actually  embedded  in  the  body  of  a  newly  hatched  but  dead  larva  of  Opius 
humilis.     B. — Newly  hatched  larva  with  its  mandibles  closed,  showing 
ventral  serosal  material  surrounding  the  body  and  the  two  gill-like  appen- 
dages on  the  first  body  segment 296 

Plate  h.     Diachasma  tryoni:  A. — Lateral  view  of  larva  in  the  second  instar, 
showing  particularly  well  the  f atbody  of  the  host  recently  taken  in  as  food . 
B. — Lateral  view  of  a  2 -day-old  lai-va  engorged  with  food  and  about  to  molt, 
showing  the  enlarged  and  stiffened  body 296 

Plate  12.  Opiu?  humilis:  A,  B. — Dead  larva  in  first  instar;  killed  by  first- 
stage  larva  of  Diachasma  tryoni,  showing  cut  on  body  made  by  the  attacking 
larva  and  mandibles  extended  in  final  death  struggle.  C— Dead  larva  in 
first  instar;  killed  by  first-stage  larva  of  Diachasma  tryoni.  D. — Dead  larva 
in  first  instar;  badly  lacerated  and  distorted  by  attack  of  first-stage  larva  of 
Diachasma  tryoni 296 

Plate  13.  Opius  humilis:  A. — Dead  larva  in  first  instar,  with  body  shriveled 
and  twisted  through  attack  by  first-instar  larva  of  Diachasma  tryoni.  B. — 
Dead  larva  in  first  instar;  killed  by  first-instar  larva  of  Diachasma  tryoni. 
C. — Dead  larva  in  first  instar;  killed  by  first-stage  larva  of  Diachasvia  tryoni. 
D. — Healthy,  living  larva  in  first  instar.  E. — Healthy,  uninjured,  living 
larva  in  first  instar 296 

Effect  of  Season  and  Crop  Growth  in  Modifying  the  Soil  Extract 

Plate  14.  A. — General  views  of  soil  containers.  B. — Bins  for  storage  of  sur- 
plus soil 368 

Tobacco  Wildfire 

Plate  15.  Bacterium  tabacum:  A. — Tobacco  leaf,  four  days  after  artificial 
inoculation,  showing  chlorosis  and  lesions.  B. — Natural  infection  with 
brown  lesions  bordered  by  tissues  of  a  water-soaked  appearance 458 

Plate  16.  Bacterium  tabacum:  A. — Natural  infection.  Lesions  are  large  and 
concentrically  zonate.  B. ^Numerous  confluent  lesions  on  one  side  of 
the  midrib  have  resulted  in  distortion  of  the  leaf.  C. — Almost  the  entire 
leaf  is  involved  and  a  portion  of  the  rotted  tissues  have  fallen  out.  Natural 
infection 458 

Wilt  Diseases  of  Okra  and  the  Verticillium-Wilt  Problem 

Plate  A.  Fusarium  vasinfectum  on  vegetable  media:  1-3. — Growth  on 
steamed  potato.  Both  potato  cultures  show  pionnotes.  2,  4. — Growth  on 
rice.  Cultures  i  and  2  were  grown  in  a  strong  north  light;  3  and  4  in  a 
subdued  light 546 

Plate   17.     A-H. — Verticillium    albo-airum:    A. — Simple    conidiophores    and 
conidia.     B. — Same  showing,  respectively,  the  collection  of  the  conidia 
on  the  sterigma  in  irregular  aggregations  in  dry  air,  and  in  water  drops  in 
humid  air.     C. — Verticillate  conidiophores  bearing  one  and  three  whorls, 
or  virtels,  of  branches,  respectively.     D. — Verticillate  conidiophore  having 
conidial  heads,  from  humid  environment — that  is,  moisttu-e  drops  in  which 
the  conidia  float  as  in  figure  B.     E. — Mycelium  of  V.  albo-atrum,  in  the 
vascular  ducts  of  an  okra  plant  inoculated  with  this  fungus.     F,  H. — Ger- 
minating conidia.     G. — Swollen,  sclerotia-like  mycelium. 
I-M. — Fusarium  vasinfectum:  I. — Terminal,  intercalary  and  conidial  chlamy- 
dospores.     K. — Germinating  macroconidium.     L. — F.  vasinfectum  from 
okra-wilt.     M. — F.  vasinfectum  from  cotton-wilt 546 


Jan.  7-Mar.  25.  i9i8  I llustvationS  VII 

Plate  18.     Longitudinal  section  of  an  okra  plant  naturally  infected  with     Page 
V erticillium  albo-atrum,  showing  the  typical  appearance 546 

Plate  19.     Verticillium   albo-atrum:  Two-weeks-old   colony   on    potato  agar, 

showing  the  concentric  rings  of  black  sclerotial  bodies 546 

Plate  20.  Solanuvi  melongena,  showing  effect  of  wilt:  A. — Qjntrol  plant  of 
the  same  age  as  the  wilted  plant  (B).  B. — Wilted  plant  photographed  two 
months  after  inoculation  at  the  hypocotyl  with  Verticillium  albo-atrum 
isolated  from  wilted  eggplant 546 

Plate  21.  Abelmoschus  esculentus ,  showing  effect  of  wilt:  A. — Control  plant. 
B. — Wilted  plant  photographed  two  weeks  after  inoculation  at  the  hypo- 
cotyl with  a  pure  culture  of  Verticillium  albo-atrum 546 

Plate  22.  Abelmoschus  esculentus,  showing  effect  of  wilt:  A. — Wilted  plant 
inoculated  with  Verticillium,  albo-atrwrn.  B. — Control  plant  of  the  same 
age  as  wilted  plant.  Both  plants  were  photographed  two  months  after 
the  wilted  plant  had  been  inoculated 546 

Plate  23.     Abelmoschus  esculentus,  showing  the  effect  of  wilt  as  a  result  of 

inoculation  with  Fusarium  vasinfectum  isolated  from  okra- wilt 546 

Plate  24.     Gossypium  herbaceum  (Columbia  variety):  Control  plants  35  days 

old 546 

Plate  25.  Gossypium  herbaceum  (Columbia  variety),  showing  effect  of  wilt: 
Wilting  plants  photographed  15  days  after  inoculation  at  the  hypocotyl 
with  Fusarium  vasinfectum,  isolated  from  wilting  cotton  plants 546 

Plate  26.  Gossypium  herbaceum  (Columbia  variety),  showing  eft'ect  of  wilt: 
Wilting  plants  photographed  15  days  after  inoculation  at  the  hypocotyl 
with  Fusarium  vasinfectum,  isolated  from  wilting  okra 546 

Plate  27.     Abelmoschus  esculentus,  showing  the  characteristic  symptoms  of 

the  wilt  produced  by  V erticilliurn  albo-atrum 546 

Weevils  Which  Affect  Irish  Potato,  Sweet  Potato,  and  Yam 

Plate  28.  Trypopremnon  sanfordi:  Adult  from  Cuzco,  Peru.  A. — Dorsal 
view.  B. — Face  of  same.  C. — Side  view  of  thorax  and  head.  D. — Ven- 
tral view  of  adult 612 

Plate  29.  Trypopremnon  latithorax:  Larva  from  La  Paz,  Bolivia.  A. — Pro- 
thoracic  spiracle.  B. — ^Larva,  lateral  view.  C. — Lateral  view  of  head. 
D. — Right  side  view  of  apex  of  labium.  E. — Corresponding  hair  on  left 
side.     F. — Maxillary  palpiger  and  palpus,  lateral  view.     G. — Face 612 

Plate  30.  Trypopremnon  latithorax:  Pupa  from  La  Paz,  Bolivia.  A. — Dorsal 
vitfw.  B. — Ventral  vievv^.  C. — Enlarged  sketch  of  eighth,  ninth,  and 
tenth  abdominal  segments 612 

Plate  31.  Species  of  the  genus  Cylas:  A.—Cylasforviicarius  elegantulus  from 
Honolulu,  Hawaii,  side  view  of  head  and  thorax.  B. — Cylas  iurcipennis 
from  Sumatra,  side  view  of  head  and  thorax.  C. — Cylas  brunncus  from 
East  Africa,  dorsal  view  of  thorax.  D. — Cylas  brunneus,  side  view  of  head 
and  thorax.  E. — Cylas  brunneus  ventral  view  of  thorax.  F. — Cylas 
femoralis,  side  view  of  head  and  thorax 612 

Plate  32.  Sweet-potato  and  yam  weevils:  A. — Cylas  formicarius  elegantulus, 
female,  from, sweet  potatoes.  New  Orleans,  La.  B. — Same,  head  of  male. 
C. — Euscepes  batatae,  from  sweet  potatoes,  Hawaii.  D. — Same,  side  view 
of  head.  E.^ — Palaeopus  dioscoreae,  from  yams  {Dioscorea  batatas),  Jamaica. 
F.  Same,  side  view  of  head 612 

Plate  33.  Pups  of  sweet-potato  weevils:  A. — Euscepes  batatae,  Barbados, 
venter.  B. — Same,  latero-vcntral  view  of  fifth  to  tenth  segments.  C. — 
Same,  dorsal  view.  D. — Same,  venter  of  seventh  to  tenth  segments.  E. — 
Cylas  formicarius  elegantulus,  Victoria,  Texas,  ventral  view  of  sixth  to 
tenth  segments.  F. — Same,  ventral  view.  G. — Same,  latero- ventral 
view.     H. — Same,  dorsal  view 612 


VIII  Journal  of  Agricultural  Research  voi.xii 

Plate  34.     Larvae  of  su-eet-potato  weevils:  A. — Cylas  formicarius  elegantulus,     page 
Victoria,  Texas,  lateral  view.     B. — Same,  dorsum  of  head.     C. — Same, 
face.     D. — Same,  side  of  head.     E. — Eusccpes  batatce,  Barbados,  dorsum 
of  head.     F. — .Same,  face.     G. — Same,  side  of  head.     H. — Same,  lateral 
view  of  larva 612 

Steriutv  IX  THE  Strawberry 

Plate  B.  Minnesotii  3:  i. — Pollen  mother  cell  previous  to  synapsis.  2. — 
Presynapsis  in  the  pollen  mother  cell  showing  loops  extending  out  from 
synaptic  mass.  3. — Two  loops  and  portion  of  a  loop  extending  from  the 
presynaptic  mass.  Same  stage  as  figure  2.  4. — Synapsis  in  a  pollen 
mother  cell.  5. — A  postsynaptic  stage.  G.^Open  spireme  stage.  7. — A 
presegmentation  stage  of  the  spireme.  8. — A  portion  of  a  bivalent  spireme 
thread  of  the  same  stage  as  figure  7.  9. — Segmentation  of  the  bivalent 
spireme  into  chromosome  pairs.  10. — A  portion  of  the  bivalent  spireme 
during  segmentation.  11,  12. — Chromosome  pairs  during  the  contraction 
period  following  segmentation.  13. — Individual  chromosome  pairs  show- 
ing various  figures  commonly  formed  diu-ing  contraction.  14. — Diakenesis 
in  the  pollen  mother  cell.  15. — Multipolar  spindle  stage  of  pollen  mother 
cell.     16. — -Early  anaphase  of  the  heterotypic  division 670 

Plate  C.  i. — Late  anaphase  of  the  heterotypic  division.  2. — ^Chroraosomes 
on  the  equatorial  plate  of  the  homeotypic  division.  3. — A  portion  of  an 
anther  in  the  tetrad  stage,  showing  the  microspores  embedded  in  the 
gelatin-like  sheath.  4. — A  tetrad  at  the  same  stage  as  those  shown  in  figure 
3.  5. — A  microspore  shortly  after  liberation  from  the  tetrad.  6. — A 
liberated  microspore  in  which  growth  has  commenced.  7. — A  later  stage 
than  figure  6,  showing  the  slight  thickening  of  the  wall  and  the  irregularities 
due  to  grov^lh  of  the  wall.  8. — Microspore  growth  completed  previous  to 
division  of  the  microspore  nucleus.  9. — A  section  through  a  microspore 
nucleus  in  prophase  showing  the  continuous  univalent  spireme.  10. — 
Another  section  of  the  same  nucleus,  showing  the  first  stages  of  the  disappear- 
ance of  the  nucleolus.  11. — Metaphase  of  the  division  of  the  microspore 
nucleus.  12. — F.  virginiana.  Anaphase  in  the  division  of  the  microspore 
nucleus.  13. — Telophase  of  the  division  of  the  microspore  nucleus.  14. — 
A  later  stage  than  figure  13  in  which  the  generative  cell  has  been  definitely 
cut  off.  15. — A  young  pollen  grain  shortly  after  division,  showing  an  in- 
crease in  cytoplasm  content.  16. — End  view  of  a  pollen  grain  showing  the 
pattern  of  the  laminate  layers  shown  in  figure  15  and  Plate  D,  figures  i,  6, 
and  15 670 

Plate  D.  i. — Nearly  mature  pollen  grain.  2. — Mature  pollen  grain.  3,  4, 
5,  7. — Various  types  of  degenerate  microspores  from  anthers  bearing  micro- 
spores of  the  stage  shown  in  Plate  C,  figure  6.  6. — An  aborting  microspore 
from  an  anther  containing  half-grown  microspores.  8. — An  aborting  micro- 
spore of  the  same  type  as  that  shown  in  figure  6  from  an  anther  containing 
nearly  full-grown  microspores  as  in  Plate  C,  figure  8.  9,  11. — Microspores 
of  the  same  types  and  same  age  as  figures  6  and  8,  in  which  degeneration 
has  proceeded  farther.  10. — An  aborted  microspore  from  an  anther  con- 
taining microspores  of  the  stage  sho\\'Ti  in  Plate  C,  figure  8.  12. — An  early 
stage  of  degeneration  in  a  full-grown  i-nucleate  microspore.  13.— An 
early  stage  of  degeneration  in  a  full-grown  i-nucleate  microspore.  14. — 
An  aborting  microspore  containing  an  abnormally  small  amount  of  light 
staining  cytoplasm;  from  an  anther  containing  i-  and  2-nucleate  micro- 
spores. 15. — An  aborted  microspore  from  an  anther  containing  i-  and 
2-nucleate  microspores.  16. — An  aborted  microspore  containing  very 
scant  cytoplasm.  The  nucleus  has  completely  degenerated  and  degenera- 
tion of  the  cytoplasm  has  begun 670 


Jan.  7-Mar.  25,  1918  IllustratlOnS  IX 

Plate  E.  i. — A  slightly  more  advanced  stage  of  the  condition  shown  in  Page 
Plate  D,  figure  16.  2. — An  early  stage  in  the  abortion  of  a  full-grown  i- 
nulceate  microspore.  3. — An  early  stage  of  abortion  directly  following 
microspore  division.  4. — A  full-grown  i-nucleate  microspore  containing 
ver)'  scant  light-staining  cytoplasm;  from  an  anther  containing  i-  and 
2-nucleate  microspores.  5. — Another  type  of  degeneration  of  a  full-grown 
i-nucleate  microspore.  6. — An  aborted  microspore  found  among  i-  and 
2-nucleate  m.icrospores.  7. — A  later  stage  of  the  type  of  degeneration  shown 
in  Plate  D,  figure  13;  from  an  anther  containing  microspores  of  the  stage  of 
development  shown  by  Plate  C,  figure  15.  8. — Degeneration  of  the  gen- 
erative cell  shortly  after  division.  9,  10. — Common  types  of  aborted 
microspores  found  with  mature  pollen.  11. — An  aborted  microspore,  of 
the  same  type  as  that  shown  in  figure  7.  12. — A  pollen  grain  showing 
abortion  of  the  generative  cell  and  an  abnormal  vacuolate  condition  of  the 
cytoplasm.  13. — A  later  stage  of  the  type  of  degeneration  shown  in 
figure  8 670 

Plate  35.  A. — Tertiary  flower  of  the  pistillate  variety,  Minnesota  ioi7XPro- 
gressive— 13-40,  showing  prominent  staminodia.  B,  C. — Primary  and 
secondary  flowers  of  the  perfect  variety,  Minnesota  1017X  Progressive— 
9-24;  B  showing  intermediate  and  C  perfect  anthers.  D,  E,  F. — Two 
primary  and  a  secondary  flower  of  the  perfect  variety,  Minnesota  1017X 
Progressive  — 2-55,  showing  pistillate,  intermediate,  and  perfect  types  of 
flowers.  G,  H,  I,  J. — Flowers  from  the  perfect  variety,  Minnesota  1017X 
Progressive— 32-1 670 

Plate  36.  A,  B,  C,  D. — Cross  sections  of  two  loculi  of  staminodia  of  the  pistil- 
late varieties.  Crescent,  Columbia,  Minnesota  ioi7XProgressive— 11-59, 
and  Seedling  140,  respectively.  E. — Degeneration  of  the  tetrads  in  an  inter- 
mediate anther  of  Fragaria  virginiana.  F,  G. — Later  stages  of  the  con- 
dition shown  in  figure  E.  H. — A  portion  of  an  intermediate  anther  from 
the  first  flower  of  Minnesota  3 670 

Effects  of  Mistletoe  on  Young  Conifers 

Plate  37.  A. — Pseudotsuga  taxifolia  infected  with  Razoumofskya  douglasii. 
B. — Effect  of  an  inoculation  with  Razoumofskya  campylopoda  on  the  height 
growth  of  6-year-old  Pintis  jeffreyi 718 

38327°— 19 2 


TEXT  FIGURES 

Behavior  of  vSweet  Potatoes  in  the  Ground 

Fig.  I.  Graphs  showing  changes  in  composition  of  Big  Stem  sweet  potatoes     page 
during  the  latter  part  of  the  season,  from  September  18  to  November 
27,  and  the  minimum  temperatures  at  the  United  States  Weather 
Bureau  Observatory  at  Washington,  D.  C,  some  20  miles  distant, 
during  that  period 16 

GossYPOL,  the  Toxic  Substance  in  Cottonseed 

Fig.  I .  Graphs  of  the  growth  of  pigs qi 

2 .  Graphs  of  the  gains  per  week  of  pigs gi 

3 .  Graphs  of  the  growth  of  pigs 94 

Irrigation  Experiments  on  Apple-Spot  Diseases 

Fig.  I.  Diagram  showing  the  soil-moistvtre  conditions  in  irrigated  plats  of  Gano 

apples 113 

2.  Diagram  showing  the  soil-moisture  conditions  in  irrigated  plats  of 

Grimes  apples 114 

3.  Diagram  showing  the  amount  of  bitter-pit  on  Grimes  apples 116 

4.  Diagram  showing  soil-moisture  conditions  in  irrigated  plats  of  Grimes 

apples 117 

5.  Diagram  showing  the  amount  of  bitter-pit  on  Grimes  apples 118 

6.  Diagram  showing  the  relation  of  the  amount  of  bitter-pit  to  the  size  of 

apples 119 

7.  Diagram  showing  the  soil-moisture  conditions  in  irrigated  plots  of  Jon- 

athan apples  in  1915 121 

8.  Diagram  showing  the  amount  of  bitter-pit  on  Jonathan  apples  in  1915. .       123 

9.  Diagram  showing  the  soil-moistiu-e  conditions  on  plots  of  Jonathan 

apples  in  1916 124 

10.  Diagram  showing  the  amount  of  bitter-pit  on  Jonathan  apples  inigib.       125 

A  Study  of  the  Plow  Bottom  and  Its  Action  upon  the  Furrow  Slice 

Fig.  I.  Diagram  giving  the  generatrices,  directrices,  and  equations  of  surfaces 

of  historical  plow  bottoms 150 

2-10.  Graphs  of  the  development  of  plow  bottoms 152, 155, 157,  159, 160 

11-14.  Graphs  of  the  motion  of  soil  particles  in  plowing 163,  164,  166 

15.  JeflFerson's  plow  bottom 173 

16.  Lambruschini 's  plow  bottom 175 

17-18.  Small's  plow  bottom 175,  176 

19-2 1 .  Stephen 's  plow  bottom 177 

22-25.  Rahm's  plow  bottom 178,  179 

26.  Knox's  plow  bottom 179 

Relation  of  the  Variability  of  Yields  of  Fruit  Trees  to  the  Accuracy  of 

Field  Trials 

Fig.     I.  Diagram  showing  the  individual  tree  yield  of  the  navel-orange  grove .       252 
2.  Diagram  showing  the  individual  tree  yield  of  the  navel-orange  grove 

(Antelope  Heights) 254 

(XI) 


XII  Journal  of  Agricultural  Research  voi.xii 

Fig.    3.  Diagrani  showing  the  individual  tree  jdeld  of  the  Valencia  orange  page 

grove 255 

4.  Diagram  showing  the  individual  tree  yield  of  the  Eureka  lemon  grove .  255 

5.  Diagram  showing  the  individual  tree  yield  of  the  seedling  walnut 

orchard 256 

6.  Diagram  showing  the  individual  tree  yield  of  the  Jonathan  apple 

orchard 257 

7.  Graphs  of  tlie  reduction  of  the  coefficient  of  variability  by  increasing 

the  number  of  adjacent  trees  to  the  plot 261 

8.  Graphs  of  the  reduction  of  the  coefficient  of  variability  by  increasing 

the  number  of  trees  to  the  plot 262 

9.  Graphs  of  production,  32-tree  plot,  navel  oranges  (Arlington) 267 

10.  Curve  of  yields  of  individual  trees,  navel  orange  (Arlington) 268 

11.  Curve  of  yield  of  individual  trees,  navel  orange  (Antelope  Heights).  .       269 

Water  Extractions  of  Soils  as  Criteria  of  Their   Crop-Producing  Power 

Fig.  I.  Graphs  showing  soils  arranged  with  reference  to  yield  and  important 

characters 307 

Effect  of  Season  and  Crop  Growth  in  Modifying  the  Soil  Extract 

Fig.     I.  Design  of  soil  containers 323 

2.  Diagram  of  the  arrangement  of  the  soil  containers 324 

3.  Graphs  of  the  nutrients  extracted  from  soil  4  by  varying  the  ratios  of 

soil  to  water 336 

4.  Graphs  of  the  nutrients  extracted  from  soil  5  by  varying  the  ratios  of 

soil  to  water 336 

5.  Graphs  of  the  nutrients  extracted  from  soil  8  by  varying  the  ratios  of 

soil  to  water.     Calculated  to  parts  per  million  of  dry  soil 337 

6.  Graphs  of  the  nutrients  extracted  from  soil  10  by  var}-ing  the  ratios  of 

soil  to  water 337 

7.  Graphs  of  the  yield  of  grain  in  1915  and  1916,  expressed  as  a  percentage 

of  tlie  maximum  yield 342 

8.  Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  1,  Yolo  silty 

clay  loam 343 

9.  Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  2,  Yolo  silt 

clay  loam 344 

10.  Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  3 ,  Yolo  silty 

clay  loam 345 

11.  Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  4,  Yolo  silty 

clay  loam 346 

12.  Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  5,  Yolo  silty 

clay  loam 347 

13.  Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  6,  Yolo  clay 

loam 348 

14.  Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  7 ,  Hanford  fine 

sandy  loam 349 

15.  Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  8,  Fresno  fine 

sandy  loam 350 

16.  Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  9,  Kimball 

fine  sandy  loam 351 

17.  Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  10,  Tejunga 

fine  sandy  loam 352 

18.  Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  11,  jVIadera 

fine  sandy  loam 353 


Jail.  7-Mar.  25,  1918 


Illustrations  xiii 


Fig.  19.  Graphs  of  the  seasonal  studies  of  water  extract  of  soil  12,  Arnold  fine     Page 
sandy  loam 3  54 

20.  Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  14,  Standish 

fine  sandy  loam 355 

21.  Graphs  of  the  growth  of  crops  in  height,  season  of  1916 357 

22.  Graphs  of  the  daily  studies  of  the  water  extract  of  soils  lA  and  iB,  sea- 

son of  1916 359 

23.  Graphs  of  the  daily  studies  of  the  water  extract  of  soils  8A  and  8B, 

July,  1916 360 

24.  Graphs  of  the  daily  studies  of  the  water  extract  of  soils  8A  and  SB, 

August,  1916 360 

The  FreEzing-Point  Method  as  an  Index  of  Variations  in  the  Soil  Solution 
Due  to  Season  and  Crop  Growth 

Fig.  I.  Graphs  of  the  depressions  of  the  freezing  point  in  soils  i  and  2,  with 

and  without  crop 372 

2.  Graphs  of  the  depressions  of  the  freezing  point  in  soils  3  and  4,  with 

and  \\nthout  crop 373 

3.  Graphs  of  the  depressions  of  the  freezing  point  in  soil  5,  with  and  with- 

out crop 374 

4.  Graphs  of  the  depressions  of  the  freezing  point  in  soil  6,  with  and  with- 

out crop 375 

5.  Graphs  of  the  depressions  of  the  freezing  point  in  soils  7  and  8,  with  and 

without  crop 37° 

6.  Graphs  of  the  depressions  of  the  freezing  point  in  soils  9  and  10,  with  and 

without  crop 377 

7.  Graphs  of  the  depressions  of  the  freezing  point  in  soils  11  and  12,  with 

and  without  crop _ 37° 

8.  Graphs  of  the  depressions  of  the  freezing  point  in  soil  14,  with  and 

without  crop 37° 

9.  Graphs  showing  the  results  of  successive  extractions  of  soils  5  and  8 . .  .       392 

Efficacy  of  Some  Anthelmintics 

Fig.  I.  Apparatus,  with  control,  for  administering  copper-sulphate  solution  to 

sheep 407 

Tobacco  Wildfire 

Fig.  I.  Parenchyma  cells  from  the  margin  of  a  lesion  showing  Bacterium  taba- 

cum  in  the  intercellular  spaces  and  within  the  cells 454 

2.  a,  Flagellaof  Bacterium  tabacum  stained  by  Morrey's  method;  b,  Bad. 
tabacum  from  bouillon  stained  with  carbol-fuchsin,  showing  arrange- 
ment of  the  elements 454 

Influence  of  Carbonates  of  Magnesium  and  Calcium  on  Bacteria  of  Certain 

Wisconsin  Soils 

Fig.    I.  Diagram  showing  the  influence  of  calcium  carbonate  and  limestone 

on  the  number  of  bacteria  in  Colby  silt  loam 47^ 

2.  Diagram  showing  the  influence  of  calcium  carbonate  and  monocal- 

cium  phosphate  on  the  number  of  bacteria  in  Colby  silt  loam 472 

3.  Diagram  showing  the  influence  of  calcium  carbonate  and  limestone  on 

the  number  of  bacteria  in  Plainfield  sand 473 

4.  Diagram  showing  the  influence  of  calcium  carbonate  and  monocal- 

cium  phosphate  on  the  number  of  bacteria  in  Plainfield  sand 474 


XIV  Journal  of  Agricultural  Research  voi.xii 

Fig.    5.  Diagram  showing  the  influence  of  magnesium  carbonate  on  the  number     Page 
of  bacteria  in  Colby  silt  loam 475 

6.  Diagram  showing  the  influence  of  magnesium  carbonate  on  the  number 

of  bacteria  in  Plainfield  sand 477 

7.  Diagram  showing  the  influence  of  the  carbonates  and  chlorids  of  mag- 

nesium and  calcium  on  the  number  of  bacteria  in  Colby  silt  loam 

soil 481 

8.  Diagram  showing  the  influence  of  dibasic  phosphate,  monocalcium 

phosphate,  calcium  carbonate,  and  magnesium  carbonate  on  the 
number  of  bacteria  in  Colby  silt  loam 482 

9.  Diagram  showing  the  influence  of  calcium  carbonate,  magnesium  car- 

bonate, dibasic  magnesium  phosphate,  and  monocalcium  phosphate 

on  the  number  of  bacteria  in  Miami  silt  loam 483 

10.  Diagram  showing  the  influence  of  calcium  carbonate,  magnesium  car- 

bonate, limestone,  and  monocalcium  phosphate  on  nitrate  accumu- 
lation of  Colby  silt  loam 488 

1 1 .  Diagram  showing  the  influence  of  large  applications  of  magnesium  car- 

bonate on  Bacillus  azotobacter  in  sterile  Colby  silt  loam 494 

Digestion  of  Starch  by  the  Young  Calf 

Fig.  I.  Bag  for  receiving  feces  and  harness  for  supporting  it 577 

Sterility  in  the  Strawberry 

Fig.  I.  Diagram  showing  the  arrangement  of  flowers  of  the  strawberry  and  the 

order  of  blossoming 614 

2.  Flower  diagrams  of  Fragaria  spp.,  showing  stamen  arrangement  in  F. 

virginiana  and  F.  americana  and  many  found  in  cultivated  varieties.       61  c 

3.  Outline  camera-lucida  drawings  of  perfect  and  intermediate  anthers  and 

staminodia  of  strawberry 617 

4.  Graphs  showing  the  relation  between  sepal  number  and  flower  position 

in  the  seedling  varieties  Nos.  373,  968,  and  1006 620 

Respiration  of  Stored  Wheat 

Fig.  I.  Graph  showing  the  relation  of  the  moisture  content  of  wheat  to  the  rate 

of  respiration 692 

2.  Graphs  showing  the  comparative  rate  of  respiration  of  hard  spring,  soft 

red  winter,  and  soft  white  winter  wheat 694 

3.  Graphs  showing  the  rate  of  respiration  of  shriveled  wheat  and  of  plump 

wheat  of  the  same  class 698 

4.  Graphs  showing  the  rate  of  respiration  of  frosted  wheat  and  sound  wheat 

of  the  same  class 699 

5.  Graphs  showing  the  comparative  respiratory  activity  of  naturally  damp 

wheats  and  of  wheats  dampened  in  the  laboratory  three  days  before 
they  were  incubated 702 

6.  Graph  showing  the  relation  of  temperature  to  the  rate  of  respiration . .  .       705 

7.  Graph  showing  the  rate  of  respiration  during  successive  intervals  when 

the  respired  carbon  dioxid  was  permitted  to  accumulate  in  the  mass 

of  grain 707 

Determination  of  Fatty  Acids  in  Butter  Fat: 

Fig.  I.  Apparatus  employed  in  esterification 722 

2.  Apparatus  employed  in  fractionation 724 


Vol.  XII  JANUARY  7,  1918  No.  1 

JOURNAL  OF 

AGRICULTURAI/ 

RESEARCH 


CONXENXS 

Effect  of  Time  of  Digestion  on  the  Hydrolysis  of  Casein  in 
the  Presence  of  Starch   -       -       -       -*       •       -1 
J.  S.  McEARGUE 
( CooMlmtira  (zoffi  Xentvcky  AcrieaMonl  Bspetimeitf  Sti^ 

Behavior  of  Sweet  Potatoes  in  the  Ground  -       -       •       .       9 
HEINRICH  HASSELBRING 

(Coxrtribotlon  trotn  BuiMii  of  Plant  Indtistiy) 

Studies  in  Soil  Reaction  as  Indicated  by  the  Hydrogen 
Electrode        -       -       -       »       -       -       -       .       .      if) 

J.  K.  PLUMMER 
CoaWbBden  fn»i  nMb  CcMUaa  Acricnltona  Bxpwlia^ 


PUBLISHED  BY  AUTHORITY  OF  THE  SECRETARY  OF  AGRICULTURE, 

WITH  THE  COOPERATION  OF  THE  ASSOCUTION  OF  AMERICAN 

AGRICULTURAL  COLLEGES  AND  EXPERIMENT  STATIONS 


WASHINOTON,  D.  C. 


WMHINOTOM  !  OOVtHNMCNT  MIMTIMO  OTriOC  :  1618 


EDITORIAL  COMMITTEE  OF  THE 

UNITED   STATES   DEPARTMENT  OF  AGRICULTURE  AND 

THE  ASSOCIATION  OF  AMERICAN  AGRICULTURAL 

COLLEGES  AND  EXPERIMENT  STATIONS 


FOR  TBB  DEPARTBCSIIT 


FOR  TBB  ASSOCtATlOa 


KARL  F.  KELLERMAN,  Chahuian    RAYMOND  PEARL • 


Physiologist  and  Assodatt  Chief,  Buttam 
of  Plant  Industry 

EDWIN  W.  AIXEN 
Chief,  OfficecfExptfimemtStaAm 


CHARLES  L.  MARLATT 

Entomdooist  and  Assislam$  ChUi, 
of  Entomoloof 


Biologist,  Maine  AtricuUunI  E*$m1mml 
Station 

H.  p.  ARMSBY 

Director,  Institute  afAnimai  SitlrUititi  fit 
Pennsj4vania  State  Cotteg* 

E.  M.  FREEMAN 

Botanist,  Plant  Pathologist  tnd  Attislmit 
Dean,  Agricultural  Experimmi  Statiom  ef 
ihe  UnrversUp  ofMmmesoU 


All  correspcmdence  regarding  articles  ttom  tbe  Department  of  Apiculture  tfkould  bt 
addressed  to  Karl  F.  Kellerman,  Journal  of  Agricultural  Research,  Washington,  D.  C 

*  Dr.  Pearl  has  undertalsxn  special  work  in  connection  with  the  war  emeigeni^; 
therefore,  until  ftuther  notice  all  correspondence  regarding  articles  from  State  Ezpeti* 
ment  Stations  should  be  addressed  to  fi.  P.  Armsby,  Institute  oC  Animal  Nutdtioaj 
State  College,  Pa. 


JOMAL  OF  AGlilCimiRAL  RESEARCH 

Vol.  XII  Washington,  D.  C,  January  7,  191 8  No.  i 


EFFECT  OF  TIME  OF  DIGESTION  ON  THE  HYDROLYSIS 
OF  CASEIN  IN  THE  PRESENCE  OF  STARCH ' 

By  J.  S.  McHarguE, 
Chemist,  Laboratory  of  Chemical  Research,  Kentucky  Agricultural  Experiment 

Station 

The  Van  Slyke  ^  method  for  protein  analysis  was  worked  out  upon 
mixtures  of  relatively  pure  amino  acids  and  was  not  intended  to  be  applied     *-'-kt- 
directly  to  crude  sources  of  protein  contained  in  cereals  and  feeding 
stuffs. 

Notwithstanding  this  fact,  Grindley,  Slater,  et  al.,^  of  the  Illinois  Experi- 
ment Station,  published  in  191 5  the  results  of  the  determination  of  the 
amino  acids  contained  in  cottonseed  meal,  tankage,  and  alfalfa  hay, 
applying  the  Van  Slyke  method  directly  to  the  proteins  contained  in 
these  different  feeds. 

In  the  same  month  of  191 5  Nollau,*  of  this  Station,  published  his  results, 
obtained  by  the  Van  Slyke  method,  on  about  25  different  sources  of  crude 
protein  contained  in  various  seeds  and  feeding  stuffs. 

In  December,  1915,  Grindley,  Slater,  et  al.,^  published  a  second  paper 
on  the  amino-acid  content  of  various  feeds,  including  wheat,  oats,  barley, 
and  sov  beans,  a  number  of  which  had  been  analyzed  by  Nollau.  The 
summary  of  their  second  paper  in  part  is  as  follows : 

The  results  here  reported  confirm  the  conclusion  previously  drawn,  namely,  that 
the  Van  Slyke  method  for  the  determination  of  the  chemical  groups  characteristic 
of  the  amino  acids  of  proteins  can  be  applied  directly  to  the  quantitative  determina- 
tions of  the  amino  acids  of  feeding  stuffs  with  at  least  a  fair  degree  of  accuracy. 

Tne  results  which  we  have  obtained  for  the  quantitative  determinations  of  amino 
acids  in  feeding  stuffs,  on  the  whole,  do  not  agree  well  with  those  recently  published 
by  Nollau.  In  some  determinations  the  results  from  the  two  sources  are  quite  satis- 
factory, but  in  many  cases  the  agreement  is  far  from  satisfactory.     The  lack  of  con- 

•  Approved  for  publication  in  the  Journal  of  Agricultural  Research  by  A.  M.  Peter,  Acting  Director, 

Kentucky  Agricultural  Experiment  Station. 

^  Van  Slyke,  D.  D.  the  analysis  of  proteins  by  determination  of  the  chemical  groups  char- 
acteristic OF  THE  DIFFERENT  AMINO  ACIDS.    In  Jour.  Biol.  Chem.,  v.  lo,  no.  i,  p.  15-53.  2  fig.  1911. 

'  Grindley,  H.  S.,  Slater,  M.  E.,  et  al.  the  quantitative  determination  of  the  amino  acids  of 
FEEDING  STUFFS  BY  THE  VAN  SLYKE  METHOD.  In  J  OUT.  Amer.  Chem.  Soc,  V.  37,  no.  7,  p.  1778-1781; 
no.  12,  p.  2762-2769.  1915. 

*  Nollau,  E.  H.  the  AMiNO-AaD  content  of  certain  commeroal  feeding  stuffs  and  other 
SOURCES  of  protein.    In  Jour.  Biol.  Chem.,  v.  21,  no.  3,  p.  611-614.  1915. 

Journal  of  Agricultural  Research,  Vol.  XII,  No.  i 

Washington,  D.  C.  Jan.  7,  1918 

U  Key  No.  Ky.— 6 

(I) 


2  Journal  of  Agricultural  Research  voi.  xii,  no.  i 

cordant  results  is  probably  due  in  the  main  to  differences  in  the  details  of  procediure 
in  the  experimental  work. 

Hart  and  Bentley/  of  the  Wisconsin  Experiment  Station,  comment 
unfavorably  on  the  lack  of  agreement  between  the  results  obtained  by 
Grindley,  Slater,  et  al.,  and  those  obtained  by  Nollau  for  the  amount 
of  the  different  amino-acid  groups  contained  in  feeding  stuffs.  They 
state  that  whether  accurate  determinations  of  any  or  all  the  amino  acids 
can  be  secured  when  the  hydrolyzing  proteins  are  in  contact  with  hydro- 
lyzing  carbohydrates  must  first  be  determined  before  these  data  can  be 
accepted  as  final. 

Presumably  in  order  to  substantiate  the  theory  in  regard  to  the  effect 
of  hydrolyzing  carbohydrates  on  the  different  amino-acid  groups  in  pro- 
teins. Hart  and  Sure  ^  have  published  results  obtained  upon  the  hy- 
drolysis of  casein,  alone  and  in  the  presence  of  a  number  of  different 
carbohydrates.  In  one  of  their  experiments,  2.4  gm.  of  casein  and  12 
gm.  of  starch  were  hydrolyzed  by  boiling  in  20  per  cent  hydrochloric 
acid  for  a  period  of  48  hours.  The  result  obtained  for  lysin  in  this  ex- 
periment shows  that  approximately  50  per  cent  of  this  amino-acid 
group  has  been  changed  to  some  other  form  of  combination.  They 
summarize  their  results  in  part  as  follows : 

The  Van  Slyke  method  of  protein  analysis,  applied  to  casein,  hydrolyzed  in  the 
presence  of  various  carbohydrates,  brings  about  a  total  redistribution  of  the  amino- 
acids  varying  with  the  nature  of  the  carbohydrate  employed.  This  work  on  casein 
and  Gortner's  work  on  fibrin,  hydrolyzed  in  the  presence  of  cellulose,  definitely  show 
the  inapplicability  of  the  method  of  direct  hydrolysis  for  the  estimation  of  amino- 
acids  in  feeding  stuSs  by  Van  Slyke 's  method.  The  results  so  secured  will  be  inac- 
curate. 

Upon  the  publication  of  Hart  and  Sure's  results,  it  appeared  to  the 
writer  that  their  conclusions  were  much  broader  than  their  experiments 
justified.  In  fact,  Hart  and  Bentley^  make  statements  which  appear  to 
be  merely  forecastings  rather  than  conclusions  arrived  at  by  experi- 
mentation. In  order  to  be  able  to  say  positively  that  the  Van  Slyke 
method  for  protein  analysis  can  not  be  applied  directly  to  heterogeneous 
mixtures  of  protein  and  carbohydrate  requires  much  further  experi- 
mentation. It  is  by  no  means  to  be  taken  for  granted  that  results  obtain- 
ed on  a  48-hour  digestion  will  be  the  same  as  those  carried  on  for  a  shorter 
length  of  time. 

It  therefore  occurred  to  the  writer  that  a  duplication  of  the  experiment 
of  Hart  and  Sure  upon  the  effect  produced  on  the  hydrolysis  of  casein 
by  the  presence  of  starch,  in  which  the  time  of  digestion  varied,  would 
afford  more  conclusive  evidence  on  this  subject.  Accordingly,  five  ex- 
periments were  planned,  as  follows: 

'  Hart,  E.  B.,  and  Bentley,  W.  H.  the  character  OF  the  water-soluble  nitrogen  of  some  com- 
mon FEEDING  STUFFS.    In  Jour.  Biol.  Chem.,  v.  22,  no.  3,  p.  477-483.  1915. 

* and  Sure,  Barnett.    The  influence  of  carbohydrates  on  the  accuracy  of  the  van  slyke 

method  in  the  hydrolysis  of  casein.    In  Jour.  Biol.  Chem.,  v.  28,  no.  i,  p.  241-249.  1916. 

3  Hart,  E.  B.,  and  BentlEy,  W.  H.    Op.  cit. 


Jan.  7. 1918    Effect  of  Time  of  Digestion  on  Hydrolysis  of  Casein  3 

Five  lo-gm.  portions  of  Hammarsten's  casein  were  weighed  out  and 
transferred  to  five  i -liter  round -bottom  Jena  flasks.  Fifty  gm,  of 
cornstarch  were  then  weighed  out  and  added  to  each  of  the  flasks  except 
the  first,  which  contained  casein  alone.  Three  hundred  c.  c.  of  20  per 
cent  hydrochloric  acid,  specific  gravity  i.ii,  were  added  to  each  flask. 
All  the  flasks  were  then  heated  on  the  water  bath,  with  frequent  shak- 
ings, for  about  two  hours.  The  object  of  this  preliminary  heating  on 
the  water  bath  was  to  liquefy  the  starch-casein  mixtures,  which  had 
gelatinized  upon  the  addition  of  the  hydrochloric  acid.  After  the  starch 
had  become  liquid  all  the  flasks  were  removed  and  attached  to  reflux 
condensers  and  heated  to  a  gentle  boil. 

Experiments  i  and  2  were  allowed  to  digest  for  12  hours,  No.  3  for  15 
hours.  No.  4  for  24  hours,  and  No.  5  for  48  hours,  each  being  cut  out  at 
the  expiration  of  its  time  interval. 

After  each  of  the  experiments  had  stood  at  room  temperature  for  six 
or  eight  hours,  they  were  filtered  through  paper  on  a  Buchner  funnel 
and  washed  practically  free  of  chlorids  with  hot  water.  There  was  no 
insoluble  residue  remaining  on  the  filter  from  the  casein  digestion.  There 
were  rather  large  insoluble  carbonaceous  residues  remaining  from  each 
of  the  casein-starch  mixtures.  Each  of  these  was  dried  at  100°  C, 
bottled,  and  set  aside  for  further  investigation  as  to  their  nitrogen 
content. 

The  filtrates  in  each  of  the  experiments  were  concentrated  separately 
under  reduced  pressure  until  practically  all  of  the  excess  of  hydrochloric 
acid  was  removed.  The  residues  were  taken  up  in  water  and  run  through 
filters  into  separate  flasks  of  250-c.  c.  capacity.  After  the  filters  were 
washed  thoroughly,  the  contents  of  each  flask  were  brought  up  to  the 
mark  with  water,  and  duplicate  analyses  were  carried  out  by  the  Van 
Slyke  method  on  aliquots  from  each  of  these  hydrolyzed  solutions. 
The  results  obtained  are  shown  in  Table  I. 

From  the  data  in  Table  I  showing  the  average  results  obtained  upon 
casein  alone  and  upon  definite  mixtures  of  starch  and  casein  digested  at 
different  intervals  of  time  the  following  observations  may  be  made. 

In  all  of  the  experiments  there  is  but  slight  variation  in  the  ammonia 
determinations^ ;  the  maximum  result  is  obtained  in  the  48-hour  digestion. 
The  increase  in  this  case  is  in  all  probability  owing  to  the  change  of  some 
of  the  amino  groups  to  ammonia  compounds,  which  indicates  over- 
digestion. 

The  results  for  the  humin  determinations  show  a  diminution  in  the 
1 5-,  24-,  and  48-hour  digestions  over  those  of  the  12-hour  digestions. 
However,  the  humin  determination  in  the  12-hour  digestion  of  the 
starch-casein  mixture  agrees  well  with  the  humin  results  obtained  on 
casein  alone. 

'  Previous  to  the  amraonia  determinations  the  acidity  of  the  hydrolyte,  in  terms  of  the  calcium-hydrate 
suspension,  was  determined  by  titration,  with  phenolphthalein  as  the  indicator.  A  slight  excess  of  the 
calcium-hydrate  suspension  aoove  tne  amount  necessary  to  neutralize  the  acid  was  always  added. 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  I 


13 


•^ 


■^ 
S 
"« 


'Si 


fO 


•c  Saab's. 
><  o  o  «  u>a 

w 


'Sib'O 


d  a  u  "  "i  3 

attJ  rt  u.  o  J3 


oi-S 


■a  +  -«  8  !ii  S 
a  a  o  "  m  g 

g  CI  oJ  1-,  cjja 


S   0) 
>  bo 


12? 


b>  bo 

■5" 


■g  S  3  °'  S  S 
K  o  B  «  o>is 


a°fe1 

.p  4-  01  u  b 

•c  g  a  g  S 
t<  o  «  o>2 


>  bg 


O  f~*  O   "^*0   f^  »^  fO 
wr-OO    O    t-o^O 


o  o  o  OS  toQO  '^r  o 


r-  6  t>  o  lo  c^  t 


wiot*0\r*tH  woo 


ft   ^  c5   *  vi5  H   '  d>oo  6 


D   §  6   '  NO  M   *  6  dco 


r*w  COM  t^mc\^ 
K^OO  O\00  ^M  r-vo   O 

n    §  o^  *  wiO    '  CO  t^  d 


CL  S  Ov  w  tovO       00  r^  t 


NO  r»  t^co  ^-  O  00  o 

K-^^NTl-Ol-tl-tOO    -^PO 

(V  §  6  MNCod    '  00  t^  cj« 


t^O  00  «o  w  o 


■«t  O    HI    M  GO    C7» 

^  §  d  w  t^oo   *  00  o  d 


-5* 


H    r»  OnOO    Ov  OvO    !>■ 

"    ^dwoco    'oooooo 


a,S2 


W^Uct   fOO\0   M    coo    ■^ 
O     §    6    M   «^o6     '    OM-OO 


0\  M  O  ■"  t'  O  Oi« 

Q^  §  d  M  »iod   *  di  t^oo 


O   «   0(  ^  ^  ^00    t^ 

Q.  §  6  M  t^  t--   '  CJi  «-■<» 


i^.t;-3  Si;  °-t  2    « 


Jan.  7, 1918    Effect  of  Time  of  Digestion  on  Hydrolysis  of  Casein 


The  results  on  humin  represent  the  humin  in  solution  and  precipitated 
by  calcium-hydrate  suspension.  The  high  results  obtained  by  Grindley, 
Slater,  et  al.  and  Hart  and  Sure  for  humin  nitrogen  were  made  to  include 
the  total  nitrogen  in  the  insoluble  residue  and  also  the  humin  in  solution. 

It  has  been  the  experience  of  the  writer  that  in  cases  where  considerable 
insoluble  residue  was  included  in  the  total  volume  of  the  hydrolyte,  great 
difficulty  was  met  with  in  obtaining  uniform  aliquots  for  the  total 
nitrogen  in  the  solution  and  also  for  the  aliquot  for  determination.  This 
difficulty  is  avoided  by  filtering  out  and  washing  the  insoluble  residue. 
Then,  too,  the  question  arises,  Is  it  fair  to  consider  the  nitrogen  remaining 
in  the  insoluble  residue  as  humin  nitrogen? 

The  results  for  arginin  show  no  serious  loss  in  any  of  the  determina- 
tions, and  the  minimum  result  obtained  is  only  1.5  per  cent  below  Van 
Slyke's  result  for  arginin  on  casein  alone. 

The  histidin  results  are  practically  the  same  for  the  two  12-hour 
digestions,  on  casein  alone  and  on  the  casein-starch  mixture.  In  the 
15-hour  casein-starch  digestion  the  result  for  histidin  is  0.61  per  cent 
above  that  reported  in  Van  Slyke's  analysis.  In  the  24-  and  48-hour 
digestions  there  is  a  loss  in  histidin  nitrogen  of  considerably  more  than 
50  per  cent  of  that  found  in  the  15-hour  digestion.  Hence,  the  results 
for  histidin  in  the  two  last  experiments  are  very  significant,  indicating 
that  long  periods  of  digestion  of  starch  and  casein  bring  about  a  redistri- 
bution of  the  nitrogen  in  this  group.  It  is  to  be  borne  in  mind  that  Hart 
and  Sure^  obtained  similar  results  on  lysin.  These  writers  also  report 
7.30  per  cent  as  a  average  for  histidin  determinations  in  their  experiment. 

There  is  a  diminution  in  the  cystin  nitrogen  of  more  than  50  per  cent  in 
the  24-  and  48-hour  digestions.  Hart  and  Sure  state  that  their  results  for 
cystin  were  so  low  that  they  reported  the  results  obtained  by  Van  Slyke 
instead. 

The  results  for  lysin  agree  well  in  the  12-  and  15-hour  experiments.  In 
the  24-  and  48-hour  experiments  the  results  for  lysin  are  high.  Lysin 
nitrogen  is  obtained  by  deducting  the  sum  of  histidin,  arginin,  and  cystin 
nitrogen  from  the  total  nitrogen  in  the  bases;  therefore  any  diminution 
in  the  nitrogen  content  of  either  histidin,  arginin,  or  cystin  will  increase 
the  results  for  lysin  nitrogen  correspondingly. 

There  is  no  marked  difference  between  the  results  obtained  in  all  the 
experiments  for  the  amino-  and  nonamino-nitrogen  content  in  the 
filtrates  from  the  bases. 

In  the  footings  of  the  different  analyses  it  is  to  be  noted  that  the 
1 2 -hour  digestions  give  footings  more  than  2.5  per  cent  over  100.  In 
the  15-hour  digestion  the  footing  is  good,  while  in  the  24-  and  48-hour 
digestions  the  footings  are  2.75  per  cent  less  than  100,  thus  indicating 
that  the  12-hour  experiments  were  probably  not  completely  hydrolyzed; 

•  Hart,  E.  B.,  and  Sure,  Barnett.    Op.  cit. 


6  Journal  of  Agricultural  Research  voi.  xii,  no.  i 

whereas  the  15-hour  digestion  was  sufficient  to  bring  about  complete 
hydrolysis  and  the  24-  and  48-hour  experiments  were  overdigested  to 
the  extent  that  nitrogen  was  lost. 

The  insoluble  carbonaceous  residues  which  were  filtered  from  the 
hydrolyzed  solutions  were  dried  at  100°  C.  and  the  total  nitrogen  de- 
termined in  each. 

The  insoluble  residue  from  experiment  2,  or  the  12-hour  starch-casein 
digestion,  contained  1.30  per  cent  of  nitrogen.  That  from  the  15-hour 
digestion  contained  0.83  per  cent  of  nitrogen.  That  from  the  24-hour 
digestion  contained  0.80  per  cent  of  nitrogen  and  that  from  the  48-hour 
digestion  contained  0.855  per  cent  of  nitrogen.  The  results  show  that 
a  15-hour  digestion  removed  as  much  nitrogen  from  the  insoluble  residue 
as  the  24-  and  48-hour  digestions. 

Two  determinations  of  total  nitrogen  on  a  sample  of  the  dry  starch 
showed  an  average  nitrogen  content  of  0.05  per  cent.  The  small  amount 
of  nitrogen  contained  in  the  starch  and  the  comparatively  greater  amount 
found  in  the  insoluble  residues  indicate  that  some  nitrogen  compound 
was  absorbed  by  the  latter. 

Seven  gm.  of  the  dry-carbon  residue  were  weighed  out  and  transferred 
to  a  Claisen  flask,  60  c.  c.  of  a  10  per  cent  calcium-hydrate  suspension 
added,  together  with  250  c.  c.  of  distilled  water.  The  apparatus  was 
connected  up  as  in  an  ammonia  determination  and  distilled  under 
reduced  pressure  at  from  40°  to  45°  C.  for  30  minutes.  Nine-tenths 
c.  c.  of  N/io  hydrochloric  acid  was  neutralized  by  the  ammonia  evolved, 
which  shows  that  the  insoluble-carbon  residue  contained  only  a  trace 
of  ammonia  nitrogen.  The  insoluble-carbon  and  calcium-hydrate 
precipitate  remaining  in  the  Claisen  flask  was  filtered  and  washed 
thoroughly,  the  filtrate  made  acid  and  concentrated  under  reduced 
pressure  to  about  50  c.  c.  The  concentrate  was  transferred  to  a  Kjeldahl 
flask  and  the  total  nitrogen  determined  in  the  usual  way.  The  filtrate 
contained  0.0032  gm.  of  nitrogen  or  5.3  per  cent  of  the  total  nitrogen 
contained  in  the  insoluble  residue.  The  ammonia  nitrogen  was  2.1 
per  cent  of  the  total  nitrogen  in  the  carbon  residue.  It  is  therefore 
evident  that  a  very  small  percentage  of  the  total  nitrogen  contained  in 
the  insoluble  residue  is  affected  by  distilling  with  calcium-hydrate  sus- 
pension, which  indicates  that  the  nitrogen  remaining  in  the  insoluble- 
carbon  residue  after  digestion  and  washing  is  in  what  may  be  considered 
an  inert  form  and  should  not  be  included  in  the  humin  group. 

CONCLUSIONS 

From  the  data  contained  in  this  paper  the  following  conclusions  may 
be  drawn: 

(i)  The  Van  Slyke  method  for  protein  analysis,  when  applied  to 
mixtures  of  casein  and  starch  in  the  proportion  of  i  to  5,  and  hydrolyzed 


Jan.  7. 1918    Effect  of  Time  of  Digestion  on  Hydrolysis  of  Casein 


from  12  to  15  hours  with  20  per  cent  hydrochloric  acid  gives  results  for 
the  amino-acid  groups  that  are  comparable  with  those  obtained  by 
Van  Slyke  upon  casein  alone. 

(2)  A  digestion  period  of  more  than  15  hours  with  20  per  cent  hydro- 
chloric acid  on  a  casein-starch  mixture  brings  about  a  redistribution  of 
the  nitrogen  contained  in  the  histidin  and  cystin  groups. 

(3)  The  insoluble  residue  obtained  from  a  casein-starch  digestion 
after  being  thoroughly  washed  contains  nitrogen,  which  is  not  seriously 
affected  when  distilled  with  calcium-hydrate  suspension,  very  small 
amounts  being  split  off  as  ammonia  or  remaining  in  the  filtrate.  This 
indicates  that  the  nitrogen  is  in  an  inert  form  and  its  estimation  should 
not  be  included  in  the  humin  determination. 


BEHAVIOR   OF  SWEET  POTATOES    IN    THE   GROUND 

By  Heinrich  Hasselbring, 

Plant  Physiologist,  Plant  Physiological  and  Fermentation  Investigations, 
Bureau  of  Plant  Industry,  United  States  Department  of  Agriculture 

THE  PROBLEM 

In  the  course  of  former  investigations^  on  the  behavior  of  sweet 
potatoes  in  storage,  it  was  observed  that  the  percentage  of  starch  was 
always  highest  and  the  percentage  of  sugar  lowest  in  freshly  dug  pota- 
toes. This  observation  was  more  or  less  incidental,  having  been  made 
in  the  course  of  experiments  whose  object  was  the  solution  of  other  prob- 
lems. It  was  therefore  not  based  upon  a  systematic  study  of  the  roots 
throughout  the  latter  part  of  the  growing  season.  Nevertheless  the  con- 
stancy of  the  condition  seemed  to  justify  the  conclusion  that  in  the 
growing  sweet  potato  the  reserve  materials  exist  essentially  in  the  form 
of  starch,  and  that  the  appearance  of  sugar  in  considerable  quantities  is 
a  phenomenon  occuring  only  in  storage  or  after  the  destruction  of  the 
leaves. 

In  order  to  determine  whether  these  quantitative  relations  between 
the  starch  content  and  the  sugar  content  of  the  sweet  potato  remain 
constant  throughout  the  latter  part  of  the  growing  season,  and  to  what 
extent  they  are  changed  by  the  death  of  the  vines,  the  carbohydrate 
metabolism  in  Big  Stem  sweet  potatoes  was  followed  from  the  time  the 
roots  were  large  enough  to  furnish  the  requisite  samples  until  they  were 
seriously  damaged  by  frost. 

The  record  thus  obtained  of  the  condition  of  the  potatoes  during  this 
period  may  be  useful  as  an  aid  in  determining  the  time  for  harvesting 
the  crop;  for  it  is  evident  that,  if  marked  changes  occur  in  the  roots  during 
the  latter  part  of  the  season,  the  time  of  harvest  will  depend  upon  the 
purpose  for  which  they  are  destined,  whether  for  storage,  stock  feed,  silage, 
or,  as  Keitt  has  suggested,  for  the  manufacture  of  starch.  In  the  last 
case  it  is  evident  that  the  crop  should  be  harvested  when  the  starch 
content  is  greatest.  As  a  rule,  growers  are  advised  to  dig  sweet  potatoes 
when  they  are  fully  matured  or  after  they  have  thoroughly  ripened. 
While  these  phrases  imply  the  idea  that  the  roots  reach  a  more  or  less 
definite  state  of  ripeness,  the  characteristics  by  which  this  state  may  be 
recognized  are  not  precisely  defined.     On  this  matter  a  record  of  the 

1  Hasselbring,  Heinrich,  and  Hawkins,  L.  A.  physiologicai,  changes  in  sweet  potatoes  duiuno 
STORAGE.     In  Jour.  Agr.  Research,  v.  3,  no.  4,  p.  331-342.     191S. 

CARBOHYDRATE  TRANSFORMATIONS  IN  SWEET  POTATOES.    In  Jour.  Agr.  Research,  V.  s, 

no.  13,  p.  543-560.    1915. 

Journal  of  Agricultural  Research,  Vol.  XII,  No.  1 

Washington,  D.  C.  Jan.  7,  1918 

In  Key  No.  G— 131 

(9) 


lO 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  1 


changes  in  the  roots  during  the  latter  part  of  their  growth  may  also 

throw  some  light. 

PREVIOUS  INVESTIGATION 

The  only  systematic  examination  heretofore  published  on  the  behavior 
of  sweet  potatoes  in  the  ground  is  that  made  in  South  Carolina  by  Keitt/ 
who  investigated  the  behavior  of  four  varieties  during  1908  and  1909. 
Since  his  data  are  not  easily  summarized  his  tables  recalculated  on  the 
basis  of  dry  matter  are  given  here  for  purposes  of  comparison  with  sub- 
sequent data. 

Table  I. — Percentage  composition  of  sweet  potatoes  in  igo8,  according  to  Keitt 


Variety. 


Water. 

Starch. 

Glucose. 

Sucrose. 

69.23 

58.11 

1.98 

13-58 

73-  34 

55-  93 

5-93 

6.83 

69.08 

56-95 

6.08 

6.18 

68.70 

63.64 

7-35 

4-57 

68.75 

57.02 

4.90 

7-52 

70.  76 

52-94 

2.77 

14.  16 

73-85 

50-17 

6.  42 

10.7s 

77-89 

51.  II 

8.68 

9-23 

70.07 

51.69 

5-98 

1-37 

72-53 

54.06 

6-37 

7-35 

69.79 

60.31 

4.04 

13-31 

71-65 

54-14 

3-67 

14.04 

70.30 

52.86 

5.62 

9-87 

71.30 

63.62 

7-39 

5-64 

65-67 

56.22 

4.19 

6-35 

68.40 

54.08 

3-70 

15.98 

68.17 

52.69 

6-57 

13.10 

67.67 

54-38 

10.83 

6.  22 

66.  72 

63.91 

7-69 

1.86 

67.74 

58-03 

i.6i 

10.  42 

Total 
carbohy- 
drates. 


Aug.  28 
Sept.  7 
Sept.  18 
Sept.  29 
Nov.  18 

Aug.  28 
Sept.  7 
Sept.  18 
Sept.  29 
Nov.  18 

Aug.  28 
Sept.  7 
Sept.  18 
Sept.  29 
Nov.  18 

Aug.  28 
Sept.  7 
Sept.  18 
Sept.  29 
Nov.  18 


Nancy  Hall , 

do 

do 

do 

do 


Polo. 


.do. 
.do. 
.do. 
.do. 


Pvirple  yam 

do 

do 

do 

do 


Fulleton  yam 

do 

do 

do 

do 


73-67 
68.69 
69.  21 
75-56 
69-44 

69.87 

67-34 

69.  02 

59-04 
67.78 

77.  66 
71-85 

68.35 
76.65 
66.76 

73-76 
72.36 

71-43 
73-46 

70.  06 


'  Keitt,  T.  E.    the  formation  of  sugars  and  starch  in  the  sweet  potato.    S.  C.  Agr.  Exp.  Sta. 
Bui.  156,  14  p.    1911. 
SWEET  potato  investigation.    S.  C.  Agr.  Exp.  sta.  Bui.  165,  43  p.    1912. 


Jan.  7, 1918  Behavior  of  Sweet  Potatoes  in  the  Ground 


II 


Table  II. — Percentage  composition  of  sweet  potatoes  in  iQog,  according  to  Keitt 


Date. 


Aug.  31 
Sept.  10 
Sept.  21 

Oct.  10 

Oct.  26 

Aug.  31 

Sept.  10 

Sept.  21 

Oct.  2 

Oct.  10 

Oct.  26 

Aug.  31 

Sept.  10 

Sept.  21 

Oct.  2 

Oct.  10 

Oct.  26 


Aug.  31 
Sept.  10 
Sept.  21 
Oct.  10 
Oct.     26 


Variety. 


Pumpkin  yam. 

do 

do 

do 

....do 


Purple  yam. 

do 

....do 

....do 

....do 

....do 


Polo. 


.do. 
.do. 
.do. 
.do. 
.do. 


Brazilian. 
....do... 
....do... 
....do... 
....do... 


Water. 


72.45 
66.  12 
74.27 
68.34 
70.  20 

66.  97 
64.71 
62.  42 
61.59 

60.  91 
63-83 

66.81 
68.24 
69.44 
67.98 

61.  24 
66.08 

67-45 
64.31 

65-51 
66.  22 
69.97 


Starch . 


58.  22 
57.20 
55-03 
53-03 
59.60 

63-73 
61.  06 
69.  10 

69.  72 
56.46 
71.  00 

70.44 

61.  52 
68.52 

64-55 
58-38 

70.  70 

70.78 

62.  96 
69.99 

55-21 
65-77 


Glucose. 


13-03 

11.98 

6.18 

3-95 
4-5° 

8.27 
9.  18 
3-70 
6.77 
2.  76 
4.  06 


7.28 
8.21 
2.  61 

3-85 
4.86 


Sucrose. 


2.  50 

5-73 
8.94 

9-95 
14.70 


2.32 


5-74 
3-98 
7. 16 
7.70 
8.26 


Total 
carbohy- 
drates. 


73-75 
74.91 

70.15 
66.  93 
78.80 

77-51 
73-75 
77-56 
81.67 
66.  77 
81.56 

79-63 
77-39 
78.77 
71.79 

67-43 
81.64 

83.80 

75-15 
79.76 
66.76 
78.  89 


These  figures  exhibit  considerable  fluctuation.  In  1908  there  was  as 
a  rule  a  decrease  in  the  percentage  of  starch  from  August  28  to  September 
18,  then  a  sudden  increase  from  September  18  to  September  29,  and  an 
equally  sudden  fall  from  September  29  to  November  18.  A  frost  which 
killed  the  vines  occurred  on  November  6.  The  Polo  variety  does  not  con- 
form to  the  others  in  its  behavior.  During  the  same  period  all  the 
varieties  show  a  decrease  in  sucrose  up  to  September  29,  and  an  increase 
from  that  time  until  November  18.  The  glucose  shows  a  gradual  increase, 
reaching  a  maximum  in  the  different  varieties  between  September  18 
and  September  29,  after  which  there  is  a  loss  of  glucose.  During  the 
next  year  the  figures  show  even  greater  irregularity.  The  percentage  of 
starch  fell  and  rose  alternately  between  each  pair  of  successive  dates. 
It  is  notable  that  after  the  frost  which  killed  the  vines  on  October  13, 
the  starch  content  of  all  varieties  increased  considerately.  Hence,  there 
is  one  year  a  loss  of  starch  after  the  death  of  the  vines  and  in  the  following 
year  an  increase.  The  figures  representing  the  percentages  of  glucose 
and  sucrose  fluctuate  irregularly,  but  in  general  it  may  be  said  that  the 
glucose  fell  throughout  the  season,  while  the  sucrose  increased.  It  is  to 
be  noted,  however,  that  after  the  frost  both  sugars  increased.  Thus,  we 
have  the  remarkable  phenomenon  of  an  increase  in  the  percentage  of 
of  total  carbohydrates  in  the  roots  after  the  vines  had  been  killed. 


12  Journal  of  Agricultural  Research  voi.xii.no.  i 

EXPERIMENTAL  WORK 

In  the  present  investigation  Big  Stem  sweet  potatoes  were  used.  They 
were  grown  at  Bell  Station,  Maryland,  in  a  sandy  field  having  a  gentle 
slope  to  the  south.  Lots  of  15  to  20  kgm.  were  collected  each  week  from 
September  18  to  November  27.  The  collections  were  always  made  in 
the  afternoon.  The  roots  were  thoroughly  washed  and  stored  in  a 
covered  receptacle  in  the  laboratory  until  the  following  day,  when  the 
samples  were  prepared  for  analysis.  In  each  case  moisture,  starch,  cane 
sugar,  and  reducing  sugar  were  determined  in  duplicate  in  five  potatoes 
of  the  lot. 

At  the  time  of  the  first  digging,  September  18,  the  potatoes  were  still 
small,  so  that  it  was  difficult  to  get  roots  large  enough  to  furnish  the 
requisite  samples  for  analysis.  After  two  or  three  weeks  there  was  an 
abundance  of  large  roots..  The  vines  remained  green  and  in  a  healthy 
state  until  the  week  of  October  9.  Heavy  frosts  during  that  week, 
especially  on  October  15,  killed  the  leaves  but  not  the  stems.  By 
November  7  the  stems  which  remained  green  after  the  leaves  had  been 
killed  were  slowly  dying,  and  some  of  the  potatoes  showed  small  round 
injured  spots  on  the  portions  near  the  surface  of  the  ground  or  projecting 
above  it.  On  November  21  many  of  the  roots  showed  considerable 
injury  on  the  exposed  ends,  but  sound  potatoes  were  still  abundant.  At 
the  time  of  the  last  collection,  on  November  28,  the  roots  were  so  ex- 
tensively damaged  that  it  was  difficult  to  find  enough  sound  ones  for 
analysis.     After  that  date  the  experiment  was  discontinued. 

The  results  of  the  analyses  are  given  in  Table  III.  For  purposes  of 
discussion  the  analytical  data  have  been  reduced  to  the  basis  of  dry 
matter  in  the  roots.  These  results  are  given  in  the  left-hand  part  of 
the  table.  Since  it  may  be  desirable  to  have  for  reference  a  record  of 
the  actual  percentages  of  the  different  substances  contained  in  the  pota- 
toes, the  original  analytical  data  based  on  the  fresh  weight  of  the  roots 
are  given  in  the  right-hand  part  of  the  table. 


Jan.  7, 1918  Behavior  of  Sweet  Potatoes  in  the  Ground 


13 


Table  III. — Percentage  composition  of  Big  Stem  sweet  potatoes  in  the  ground  during 
the  latter  part  of  the  growing  season 


Date. 


Sept.    18 


Sept.  25 


Oct. 


Oct. 


Oct.     17 


Oct.     23 


Sweet  potato  No. 


13- 
14. 

US- 


fi6. 

17- 
18. 
19. 
20. 


23- 
24. 

Us- 


f26. 

27. 
28. 
29. 
130  ■ 


Average . 


Average . 


Average . 


Average .  . 


Average . 


Mois- 
ture. 


73 


Average . . . 


73 


73 


92 


24 


87 


77-43 


On  the  basis  of  dry  matter. 


Starch. 


71 


72 


70, 


39 


39 


90 


68.51 


Cane 
sugar. 


7-36 

"   70 

76 

17 
65 


93 


52 


8.86 


Reduc- 
ing 
sugar 

as  glu- 
cose. 


I.  20 
1.77 
I.  14 
1.78 
I- 31 


1-44 


I- 31 

.87 

I.  19 

I.  41 


1-49 
I.  62 
1.68 

1-45 
2.38 


I.  72 


1.  81 
2.31 
2.49 
2.23 

2.  29 

2.23 

2-49 
3-47 
2-99 
2.51 

2-59 
2.81 

2.30 
3.06 
2.  06 

3-69 
2.  91 

2.80 


Total 
carbo- 
hy- 
drates. 


80 


81 


76 


13 


43 


80.  18 


On  the  basis  of  fresh 
material. 


Starch. 


17 


19 


16 


63 


39 


56 


15-47 


Cane 
sugar. 


1.  96 

2.  II 
2.  04 
2.  16 
2.05 


2.  06 


2.30 
1.97 

1-77 

1.  92 

2.  07 


2.  01 


30 


1.99 

2-34 
2.  01 
2.  17 

*2.  00 


2.  02 
1.80 


2.  17 

1-97 

2.  07 
2.  II 
2.05 

1.  69 

2.  09 

2.  00 


Reduc- 
ing 
sugar 

as  glu- 
cose. 


37 


ii 


45 


44 
55 
61 

55 

59 
82 
70 
59 
63 

67 

53 
69 

49 
78 

Jl 

•63 


14 


Journal  of  Agricultural  Research 


Vol.  xir,  No.  I 


Table  III. — Percentage  composition  of  Big  Stem  sweet  potatoes  in  the  ground  during 
the  latter  part  of  the  growing  season — Continued 


Date. 


Oct.     30 


Nov. 


Nov.    13 


Nov.   20 


Nov.   27 


Sweet  potato  No. 


f3i 

32 

3Z 

34 

35 

Average 

36 

37 

38 

39 

40 

Average 

41 

42 

43 

44 

U5 

Average 

[46 

47 

48 

49 

^50 

Average 

5i---- 

52 

53 

54 

155 

Average 


Mois- 
ture. 


77 


75 


78 


77 


78 


99 


14 


78 


78.93 


On  the  basis  of  dry  matter. 


Starch. 


64, 


63 


55 


SO' 


65 


58 


27 


73 


46.  20 


Cane 
sugar. 


10.  65 
11.77 
10.  56 
14.47 
8.76 


II.  24 


14.65 

14-  15 
14.44 
11.82 
13-67 


13-75 


15.24 
18.56 
18.  20 
20.  08 
19-93 


lb.  40 


20.  99 

25-31 
19.72 
25.  16 
24-39 


23.  II 


Reduc- 
ing 
sugar 
as  glu- 
cose. 


2.87 
2.81 

2.  91 
4.24 
2.98 


3-  16 


3-40 
2.68 

2-95 
1.86 
2.58 


2.  69 


2-59 
2.99 

2-45 

3-40 
5-03 


3-29 


3-30 
3-  75 
2.79 

3-41 
2.  76 


3-  20 


28.  10 
26.59 
26.  61 

25-52 
23.48 


26.  06 


5.08 
3.60 
4.04 
3.61 
2.  91 


3-85 


Total 
carbo- 
hy- 
drates. 


79 


80, 


76 


77 


05 


19 


96 


03 


76.  II 


On  the  basis  of  fresh 
material. 


Starch. 


15-07 
14-37 
14-57 
12.  67 
15.22 


14.38        2 


Cane 
sugar. 


14.97 
15-51 
14.49 
16.44 

14.95 


15-27 


13.09 

11-95 
13.06 
10.  91 
11-54 


12.  06 
10.  04 
13-30 
9.  61 
11.56 


II.  31 


8.77 
10.  22 

9.  00 
10.  18 
10.58 


9-75 


49 


29 


5-48 


Reduc- 
ing 
sugar 

as  glu- 
cose. 


0.66 
.62 

•65 
.90 
.67 


70 


.81 
.64 
.69 

•47 
.61 


.64 


•57 
•65 
•56 
.72 
I.  09 


72 


.76 

•78 
.67 
•70 
•63 


71 


I.  02 

•77 
.82 
.78 
.64 


An  examination  of  Table  III  shows  that  the  sweet  potatoes  exhibit 
some  individual  fluctuation  in  composition,  but  these  variations  are  not 
sufficiently  great  to  obscure  the  seasonal  trend.  They  show,  however, 
that  small  deviations  in  the  general  contour  of  the  seasonal  changes  are 
to  be  expected  where  so  small  a  number  of  individuals  is  examined.  The 
significant  changes  lie  clearly  outside  the  limits  of  the  individual  fluctua- 
tions. The  seasonal  changes  in  the  percentage  of  the  various  constitu- 
ents of  the  sweet  potato  as  shown  by  the  table  are  briefly  described 
below. 


Jan.  7, 1918  Behavior  of  Sweet  Potatoes  in  the  Ground  1 5 

Moisture. — The  moisture  content  of  the  roots  remained  almost 
uniform  during  the  period  covered  by  the  first  three  collections,  from 
September  18  to  October  2.  After  that  time  there  was  a  gradual  increase 
in  the  percentage  of  moisture  until  the  end  of  the  season.  A  small  fluc- 
tuation appears  in  the  lot  collected  on  November  6.  The  changes  in 
water  content  of  the  roots  are  therefore  fairly  regular  and  uniform.  No 
such  marked  fluctuations  as  those  recorded  in  the  tables  of  Keitt  are 
evident  from  these  data.  The  beginning  of  the  accumulation  of  moisture 
in  the  roots  is  practically  coincident  with  the  destruction  of  the  leaves. 

Starch. — The  percentage  of  starch  in  the  dry  matter  of  the  roots 
shows  a  slight  decrease  from  September  18  until  October  23,  varying 
during  that  period  between  71.39  per  cent  and  68.51  per  cent.  On 
October  30,  the  date  of  the  next  collection,  the  starch  content  had  fallen 
to  64.65  per  cent.  From  that  date  the  starch  content  continued  to  fall 
until  the  end  of  the  season,  when  the  minimum  of  46.20  per  cent  was 
reached.  The  rapid  disappearance  of  starch  follows,  somewhat  delayed, 
upon  the  death  of  the  leaves. 

Cane  sugar. — The  cane-sugar  content  remains  practically  constant 
between  7.52  and  8.86  per  cent  until  the  time  when  the  percentage  of 
starch  begins  to  fall  rapidly.  With  the  decrease  in  starch  the  cane  sugar 
begins  to  increase  correspondingly  until  it  finally  represents  26.06  per 
cent  of  the  dry  matter  of  the  roots.  It  should  be  noted,  however,  that 
in  general  the  changes  in  cane  sugar  are  inaugurated  somewhat  later  than 
those  of  the  starch. 

Reducing  sugar. — ^The  reducing  sugar  content  remains  constant  at 
first  and  then  shows  a  gradual  rise  until  the  final  percentage  is  somewhat 
more  than  double  that  at  the  beginning.  The  increase  in  reducing  sugar 
antecedes  somewhat  the  rise  in  cane  sugar.  It  is  noteworthy  also  that 
under  these  conditions  reducing  sugar  apparently  does  not  accumulate 
to  the  same  extent  to  which  it  accumulates  in  sweet  potatoes  in  storage. 

ToTAi,  carbohydrates. — The  total  carbohydrate  content  undergoes 
very  little  change  until  toward  the  end  of  the  season,  when  the  roots  begin 
to  show  marked  injury  by  frost.  At  that  time  an  evident  loss  of  carbohy- 
drates becomes  apparent.  The  constancy  of  the  total  carbohydrate  con- 
tent of  the  sweet  potato  is  in  marked  contrast  to  the  fluctuations  observed 
by  Keitt. 

The  foregoing  facts  have  been  embodied  in  the  curves  in  figure  i.  The 
curves  are  based  on  the  averages  of  the  analyses  for  each  week. 

CONCLUSIONS 

It  may  be  concluded  from  this  investigatiou  that  the  changes  occurring 
in  sweet  potatoes  in  the  ground  during  the  later  part  of  the  growing 
season  proceed  in  a  regular  and  orderly  manner.  During  the  later  part 
of  the  period  of  growth  the  composition  of  the  roots  remains  remarkably 
uniform,  and  presents  no  striking  or  irregular  fluctuations.     During  this 


i6 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  I 


period  the  root  is  characterized  by  a  high  starch  content,  and  a  low  sugar 
content.  The  changes  which  occur  later  are  associated  with  the  death 
of  the  vines.  Prominent  among  these  changes  is  the  accumulation  of 
water  in  the  roots  as  a  result  of  the  cessation  of  transpiration  in  conse- 


13/6.  SEPT. 


OCTOBER 


NOVEMBER 


Fig.  I. — Graphs  showing  changes  in  composition  of  Big  Stem  sweet  potatoes  during  the  latter  part  of  the 
season,  from  September  i8  to  November  27,  and  the  minimum  temperatures  at  the  United  States  Weather 
Bureau  Observatory  at  Washington,  D.  C,  some  20  miles  distant,  during  that  period.  The  ordinates 
indicate  percentages  in  the  one  case  and  degrees  Fahrenheit  in  the  other. 

quence  of  the  destruction  of  the  leaves.  With  the  termination  of  the 
flow  of  materials  from  the  vines  the  carbohydrate  transformations 
characteristic  of  sweet  potatoes  in  storage  are  inaugurated.  These 
changes  consist  in  the  transformation  of  starch  into  sugars.     In  point 


Jan.  7. 1918  Behavior  of  Sweet  Potatoes  in  the  Ground  1 7 

of  time  the  decrease  in  starch  and  the  increase  in  reducing  sugar  precede 
somewhat  the  increase  in  cane  sugar.  It  appears,  therefore,  that  reducing 
sugar  is  formed  first  as  an  intermediate  step  in  the  change  from  starch 
to  cane  sugar.  The  loss  caused  by  respiration,  which  is  considerable 
during  the  curing  process  and  in  storage,  is  apparently  slight  in  sweet 
potatoes  in  the  ground.  Appreciable  destruction  of  carbohydrates 
appears  not  to  occur  under  these  conditions  imtil  late  in  the  season  when 
the  roots  have  been  injured  by  frosts. 

The  changes  here  described  have  a  practical  bearing  on  the  question 
of  maturation  of  sweet  potatoes  and  on  the  choice  of  the  time  of  harvest. 
Since  the  carbohydrate  relations  of  the  roots  in  the  ground  remain  practi- 
cally unchanged  while  the  vines  are  uninjured,  the  roots  can  not  be  said 
to  undergo  a  definite  process  of  ripening,  in  the  sense  of  a  progressive 
transformation  of  one  reserve  substance  into  another,  such  as  the  change 
of  insoluble  pectin  into  soluble  pectin  derivatives  in  the  peach,  or  the 
transformation  of  starch  into  cane  sugar  and  invert  sugar  in  the  apple. 
Under  ordinary  conditions  the  potatoes  continue  to  grow  until  frost 
without  reaching  any  definite  state  of  maturity  recognizable  by  pro- 
gressive changes  in  the  reserve  materials  which  they  contain.  The 
changes  in  storage,  which  may  perhaps  be  regarded  as  a  process  of  matura- 
tion, do  not  come  in  for  consideration  here,  since  statements  relating  to 
the  degree  of  maturity  of  sweet  potatoes  always  refer  to  the  growing 
roots.  It  is  evident  from  these  considerations  that  the  choice  of  time  of 
harvest  is  not  a  matter  of  maturity  of  the  roots,  but  is  governed  by  other 
factors.  The  potatoes  may  safely  be  kept  in  the  ground  until  the  leaves 
have  been  injured  by  frost. 

Of  the  changes  which  occur  after  the  destruction  of  the  leaves, 
the  accumulation  of  water  in  the  roots  deserves  foremost  considera- 
tion. It  can  scarcely  be  doubted  that  this  increased  water  con- 
tent is  detrimental  to  the  successful  storage  of  the  roots,  and  causes 
them  to  be  more  subject  to  decay  than  roots  of  normal  water  content. 
One  of  the  objects  of  the  rather  expensive  operation  of  curing  is  to  elimi- 
nate a  part  of  the  water  contained  in  the  roots.  As  a  rule  observers  agree 
that  cured  sweet  potatoes  keep  better  than  uncured  ones.  Only  occasion- 
ally a  statement  to  the  contrary  is  found.  It  may  therefore  safely  be 
assumed  that  the  increase  in  the  relative  proportion  of  water  in  the  roots 
will  be  detrimental  to  storage.  On  this  account  it  is  of  utmost  importance 
that  the  harvesting  of  sweet  potatoes  be  not  long  delayed  after  the  leaves 
have  been  killed  by  frost.  The  other  changes  occurring  in  sweet  potatoes 
in  the  ground  are  essentially  the  same  as  the  changes  occurring  in  storage. 
These  changes  are  therefore  in  no  way  detrimental  to  the  crop,  since  no 
appreciable  loss  of  carbohydrates  occurs  until  the  roots  have  been  so 
severely  injured  that  they  have  lost  their  market  value. 
27804°— 18 2 


STUDIES   IN  SOIL  REACTION  AS   INDICATED  BY  THE 
HYDROGEN  ELECTRODE 

By  J.  K.  Plummer, 
Soil  Chemist,  Division  of  Agronomy,  North  Carolina  Agricultural  Experiment  Station 

INTRODUCTION 

There  has  been  developed  in  the  past  few  years  a  rather  voluminous 
literature  dealing  with  the  subject  of  soil  reaction.  One  has  only  to  scan 
this  literature  in  order  to  find  wide  variations  of  opinion  between  inves- 
tigators as  to  the  cause  and  nature  of  soil  acidity. 

Until  recently  the  various  qualitative  and  quantitative  methods  in 
vogue  for  indicating  soil  acidity  or  lime  requirement  have  not  been  suffi- 
ciently delicate  to  draw  definite  conclusions  as  to  the  "true  reaction''  of 
soils.  The  lack  of  uniformity  and  accuracy  of  methods  has  undoubtedly 
caused  such  confusion  on  this  subject.  By  adopting  modem  methods 
for  measuring  soil  reaction^many  of  the  contentions  should  be  obliterated. 

Though  the  hydrogen  electrode  has  been  used  for  some  time  in  indi- 
cating changes  in  reaction,  Gillespie  (7)  ^  was  the  first  to  use  it  on  an 
extended  scale  as  an  indicator  of  reaction  in  soils.  Sharp  and  Hoag- 
land  (12)  have  since  measured  the  H-ion  concentration  of  a  number  of 
oils  in  suspension,  and  have  extended  this  method  to  studying  other 
soil  phenomena. 

The  significance  of  the  terms  "true  acidity,"  "true  alkalinity,"  and 
"true  neutrality"  need  not  be  defined  here,  except  in  so  far  as  an  expla- 
nation of  the  method  adopted  in  reporting  results  obtained.  Pure  water 
dissociates  into  H  and  OH  ions  in  equal  concentration.  The  product  of 
the  concentration  of  these  ions  in  a  solution  is  a  constant,  approximately 
I  X  10— 14.  When  the  H  ions  are  present  in  a  concentration  greater 
than  I  X  10—7,  the  solution  is  acid;  the  presence  of  OH  ions  in  greater 
concentrations  than  1X10—7  results  in  an  alkaline  solution.  For  a 
more  detailed  discussion  of  this  subject,  the  reader  is  referred  to  texts 
on  electrochemistry. 

The  investigations  herein  reported  were  begun  about  the  time  of  the 
appearance  of  Sharp  and  Hoagland's  paper,  to  ascertain  if  appreciable 
differences  occurred  in  the  H-ion  concentration  of  soils  of  humid  regions, 
especially  from  those  of  the  Southern  States.  It  would  appear  that  with 
the  excessive  rainfall  of  this  region  an  accumulation  of  soluble  acids  in 
the  soil  would  be  almost  impossible. 

1  Reference  is  made  by  number  (italic)  to  "Literature  cited,"  p.  30-31. 

Journal  of  Agricultural  Research,  Vol.  XII,  No.  i 

Washington,  D.  C.  Jan.  7,  1918 

lo  -  Key  No.  N.  C— 7 

(19) 


20  Journal  of  Agricultural  Research  voi.xii,  no.  i 

The  method  has  been  used  for  indicating  the  reaction  of  the  soil  film 
water.  The  effects  of  several  fertilizer  materials  have  been  studied  on 
the  H-ion  concentration  of  various  field  soils,  receiving  applications  of 
such  fertilizers  for  a  number  of  years.  Lastly,  the  effect  of  ammonium 
sulphate  and  monocalcium  phosphate  on  the  reaction  of  soil  film  water 
have  been  investigated. 

METHODS  OF  INVESTIGATION 

The  apparatus  for  measuring  the  H-ion  •  concentration  of  soil  suspen- 
sions and  extracts  was  essentially  that  described  by  Hildebrand  {8),  and 
modified  by  Sharp  and  Hoagland.  Palladium  was  substituted  for  plati- 
num as  the  electrode,  which  was  treated  in  a  manner  similar  to  that 
described  by  Findlay  (6)  for  coating  electrodes.  The  supply  of  hydrogen 
was  obtained  by  electrolyzing  water,  with  potassium  hydro xid  as  the 
electrolyte.  All  the  precautions  of  rigidity  of  connections,  insulation  of 
apparatus,  time  (which  often  varied)  for  estabhshment  of  equihbrium 
between  soil  and  solution,  coating  electrodes,  and  prevention  of  loss  of 
CO2  were  strictly  observed. 

It  might  be  said  in  passing  that  no  difficulties  were  encountered  due 
to  the  reduction  of  nitrates  to  ammonia,  as  has  been  suggested. 

In  the  preparation  of  soil  suspensions,  unless  otherwise  stated,  the 
arbitrary  ratio  of  10  gm.  of  air-dried  soil  to  100  c.  c.  of  as  pure  distilled 
water  as  obtainable  was  used.  In  the  case  of  field  soils  their  content  was 
determined  as  soon  after  sampling  as  practicable,  and  the  same  ratio  of 
soil  to  water  maintained.  All  samples  except  those  taken  directly  from 
the  field  were  screened  through  a  2-mm.  sieve. 

In  all  cases  the  readings  became  constant  in  a  few  minutes.  Dupli- 
cate readings  on  the  same  sample  of  soil  could  easily  be  read  to  0.02  volt. 
However,  to  be  certain 'that  equilibrium  had  been  established  the  elec- 
trode was  allowed  to  remain  in  contact  with  solution  for  30  minutes. 
It  was  almost  impossible  to  get  such  closely  agreeing  results  as  0.02 
volt  with  different  samples  of  the  same  soil.  This  can  be  accounted 
for  in  the  lack  of  uniformity  of  mixing. 

The  results  are  reported  in  the  usual  manner  for  such  measurements, 
units  of  gram-molecules  of  H  ion  per  liter.  The  tables  of  Schmidt  {11) 
were  used  in  securing  the  H-ion  concentration  from  the  voltmeter  read- 
ings. 

RESULTS  OBTAINED 

In  Table  I  will  be  found  the  results  of  measurements  of  the  H-ion 
concentrations  of  68  samples  of  untreated  soils,  including  subsoils,  which 
represent  a  wide  range  in  types  of  five  series.  Included  in  this  table 
are  results  derived  from  five  samples  of  treated  soil.  The  samples  are 
taken  as  a  fair  representation  of  the  soils  common  to  the  area  of  the 
southeastern  portion  of  the  United  States,  which  extends  from  and 
including  the  Appalachian  Mountains  to  the  Atlantic  Ocean. 


Jan.  7,  1918 


Studies  in  Soil  Reaction 


21 


Table  I. — Hydrogen-ion  concentration  of  soil  suspensions 


Sample 

No. 


Soil  type.a 


Depth. 


Volt- 
meter 
readmgs. 


H-ion  concen- 
tration. 


1565 
1566 
1067 
1068 
1501 
1502 
1287 
1288 
1397 
1398 
1483 
1484 

1344 

1345 

991 

992 

1552 
1029 
1030 

1519 
1520 
1499 
1500 
1352 
1353 
1287 
1288 
I125 
I126 
1261 
1262 

1337 
1205 

1333 
1322 

1323 
1435 
1436 
1121 
1122 
1344 
1345 
1328 
1329 

1257 
1258 

395 
396 
397 
398 
639 
640 
6652 
653 


Norfolk  sand  (sand-hill  phase). 

....do 

....do 

....do 

Norfolk  coarse  sandy  loam .... 

....do 

....do 

....do 

Norfolk  sandy  loam 

....do 

....do 

....do 

Norfolk  line  sandy  loam 

....do 

....do 

....do 

Norfolk  very  fine  sandy  loam. 

do 

do 

do 

Norfolk  silt  loam 

do 

do 

do 

Cecil  coarse  sandy  loam 

do 

do 

do 

Cecil  fine  sandy  loam'. 

do 

do 

do 

Cecil  clay  loam 

do 

do 

do 

Cecil  clay 

do 

do 

do 

Iredell  fine  sandy  loam 

do 

do 

do 

Iredell  loam 

do 

do 

do 

Porter's  sand 

do 

do 

do 

Porter 's  sandy  loam 

do 

do 

do 

Porter's  loam 


Inches. 
0-4 
4-21 
0-4 
4-18 
0-6 
6-28 
0.6 
6-26 
0-7 
7-28 
0-7 
7-28 
0-6 
6-36 
0-6 
6-36 
0-6 
6-24 
0-8 
8-30 
0-6 
6-28 
0-6 
6-32 
c^7 
7-20 
c^6 
6-30 
0-8 
8-24 
0-8 
8-26 
0-6 
6-30 
0-6 
6-36 
0-7 
7-20 
0-6 
6-36 
0-6 
6-36 
0-8 
8-24 
0-8 
8-24 
0-6 
6-26 
0-6 
6-20 
0-6 
6-20 
0-7 
7-28 
0-7 

7-30 
0-8 


700 
684 
712 
706 
684 
684 
696 
681 
674 
668 
662 
674 
660 
671 

645 
640 

657 
660 
614 
583 
574 
568 
560 

555 
712 
699 
706 
700 
702 
681 
699 

673 
652 
666 
659 
643 
641 
582 
646 
607 
732 
738 
711 
704 
692 

714 
684 
679 

673 
668 
640 
638 
648 
641 
656 
649 
644 


Gram  mole- 
cules per  liter. 

o.  4X10"^ 

9X10-6 

3  X 10-6 
3X10-6 

9X10  6 
9X10-8 
5X1O-S 

oX  10-6 

I X 10-5 

I X 10-5 
2  X 10-5 
I X 10-5 

2  X  10-5 

I X 10" 
4x10-5 

5x10-5 
2x10-5 

2  X  10-5 

I X 10-4 

5x10-4 
7x10-4 

9Xio~4 
I X  lo-a 
1X10-* 
3X10-8- 
5X10-8 
3X10-6 
4X10-6 
4X10-8 

oXlO-8 

4x10-6 

I X 10-5 
3X10-5- 

I X 10-5 

2  X 10-5 
5X10-5 
5X10-5 
5X10-4 

4X10    5 

I X 10-4 

I X 10-6 

I X 10-8 

10-8 

10-6 

10-6 


3X1 

4Xj 
6X: 


2  X 10-6 
9X10-6 

I X 10-5 
I X 10-5 
I X 10-5 
5x10-5 
5x10-5 

3X10    5 

5x10-5 

2  X  10-5 

3x10-5 
4x10-5 


1  The  writer  is  indebted  to  Dr.  W.  H.  Mclntire,  of  the  Tennessee  Agricultural  Experiment  Station,  for 
the  samples  of  the  treated  soils. 


22 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  I 


Table  I. — Hydrogen-ion  concentration  of  soil  suspensions — Continued 


Sample 
No. 


Soil  type. 


Depth. 


Volt- 
meter 
readings. 


H-ion  concen- 
tration. 


I202 
1372 

1373 
674 

684 

68s 
1007 
1212 
1077 
2526 


Porter's  loam 

do 

do 

Porter's  clay 

do 

do 

do 

Muck 

do 

do 

do 

Chickamauga  limestone  soil«. 

Chickamauga  limestone  soil  &. 

Cumberland  loam  « 

Cumberland  loam  & 

Tillico  sandy  loam  a 


Inches. 
8-30 

o-l 

7-24 
0-7 
7-24 
0-6 

6-20 

0-18 

0-24 

0-36 
0-36 


•  639 

.647 

•  639 

•577 
.566 

•594 
•587 
•483 
.  462 

•536 
.427 
.864 
.824 

•893 
.856 
.866 


Gram  rtiole- 
cules  per  liter. 
.5X10-5 
.3X10-5 
.5X10-5 
.6X10-* 
.1X10-3 
.3X10-* 
.4X10-* 
.2X10-2 
.6X10-2 
.3X10-3 
.2X10-1 
.7X10-9 
.3X10-8 
.2X10-9 
.9X10-9 
.6X10-9 


o Treated  in  1912  with  16,000  pounds  of  calcium  carbonate  per  acre  in  excess  of  Vietch  indication. 

6  Treated  in  1912  with  MgCOsOCaCOs  at  rate  of  16,000  pounds  per  acre  in  excess  of  Vietch  indication. 

The  results  shown  in  Table  I  indicate  wide  variations  in  the  H-ion 
concentrations  of  the  untreated  soils  under  experiment.  This  vaijies 
from  very  nearly  neutral  in  some  of  those  of  the  Iredell  series  to  rather 
excessive  acidity  in  the  Norfolk  silt  loam  and  mucks.  This  is  in  accord 
with  what  would  be  expected.  The  Iredell  soils  are  of  residual  origin 
formed  from  basic  eruptions,  mainly  diorite.  The  amount  of  basic 
elements  supphed  this  soil  is  greatly  in  excess  of  those  commonly  found 
in  the  area  included  in  this  study.  The  Norfolk  silt  loam  being  a  trans- 
ported soil,  was  formed  under  conditions  through  which  the  basic  ele- 
ments have  been  removed.  It  also  contains  rather  high  amounts  of 
partially  decomposed  organic  matter,  and  should  show  a  high  concen- 
tration of  H  over  OH  ions.  Field  and  pot  tests  with  various  crops  have 
shown  indications  of  excessive  acidity  on  some  of  the  muck  soils.  Indeed 
on  some  of  the  muck  fields  from  which  the  samples  were  taken  little  or 
none  of  the  common  agricultural  crops  will  thrive  until  the  land  has 
been  limed. 

No  definite  relationship  appears  to  exist  between  the  H-ion  concen- 
tration and  types  of  different  texture.  It  can  be  noticed  that  there  is  a 
tendency  for  the  H  ion  to  increase  as  the  number  of  fine  particles  increase 
in  a  series.  However,  there  are  exceptions.  The  clay  loams  and  loams 
of  the  Iredell  and  Porter's  series  show  greater  H-ion  concentration  than 
do  those  of  coarser  texture.  It  may  be  observed  that  in  many  cases  the 
subsoil  shows  a  greater  degree  of  acidity  than  the  corresponding  surface 
soil.  The  general  practice  in  farming  these  soils  has  been  the  removal 
from  the  land  of  all  crops.  Probably  there  is  a  tendency  for  plant  roots 
to  remove  more  bases  from  lower  depths  than  from  the  surface. 


Jan.  7, 1918  Studies  in  Soil  Reaction  23 

The  samples  of  soil  which  have  been  treated  witii  excessive  amounts  of 
calcium  and  magnesium  carbonate  show  a  greater  concentration  of  OH 
than  H  ions  in  solution.  Dr.  Mclntire  informed  the  writer  that  the  car- 
bonates have  long  since  disappeared  in  these  soils.  Obviously  the  new 
compounds  of  calcium  and  magnesium,  whatever  they  may  be,  give  a 
strong  basic  reaction. 

These  results  bear  out  very  forcibly  the  contentions  of  Gillespie  (7)  and 
Sharp  and  Hoagland  {12)  that  there  is  a  preponderance  of  H  over  OH  ions 
in  the  liquid  phase  of  many  soil  suspensions.  Whether  the  "  true  acidity  " 
as  indicated  from  the  results  of  Table  I  is  developed  from  organic  acids 
or  acid  silicates  can  not  be  stated.  In  some  of  the  sandy  soils  the  organic 
matter  is  quite  low,  yet  indications  lead  to  the  conclusion  that  there  is  a 
greater  concentration  of  H  ions  than  found  in  neutral  solutions.  With 
the  muck  samples  the  organic -matter  content  is  very  high,  being  90  per 
cent  or  more,  and  most  marked  acidity  is  shown. 

H-ION  CONCENTRATION  OF  SOIL  FILM  OR  CAPILLARY  WATER 

The  question  has  often  been  raised,  "Is  the  water  held  as  a  liquid  film 
around  the  soil  particles  of  different  reaction  from  that  in  the  free  state  ? " 
The  difficulty  encountered  in  securing  any  workable  amount  of  the  film 
water  has  militated  against  any  direct  study  of  this  question.  Recently, 
Morgan  (9)  has  devised  a  workable  modification  of  the  "oil  displacement " 
method  for  obtaining  the  soil  solution  in  an  unaltered  form.  A  study  of 
this  water  should  throw  much  light  on  some  of  the  obscure  problems  of 
soils.  One  serious  difficulty  yet  remains  with  the  method,  which  is  its 
inability  to  get  back  all  of  water  held  as  thin  films  around  the  soil  grains. 
However,  a  study  of  the  reaction  of  that  portion  of  the  film  water  obtain- 
able should  give  indications  of  value  as  to  the  conditions  of  the  soil 
solution. 

The  Morgan  apparatus,  with  a  few  unimportant  modifications,  has 
been  used  in  obtaining  that  portion  of  the  capillary  water  displaced  by 
paraffin  oil  under  high  pressure.  The  oil  used  in  this  work  was  the 
purest  obtainable.  As  far  as  could  be  noticed,  it  gave  a  neutral  reaction 
with  the  hydrogen-electrode  apparatus  and  by  titrating  against  standard 
alkali  with  methyl  orange,  methyl  red,  and  phenolpthalein  as  indicators. 

Some  of  the  soil  samples  used  were  taken  directly  from  the  filed; 
others  had  been  stored  in  the  laboratory  for  a  number  of  years.  The 
portion  which  passed  a  2-mm.  sieve  was  thoroughly  mixed,  and  the  water 
content  determined.  Sufficient  distilled  water  was  added  to  bring  the 
moisture  content  up  to  about  what  is  considered  the  optimum  for  that 
soil.  Water  lost  through  evaporation  was  replaced  at  frequent  intervals, 
after  which  the  soils  were  thoroughly  packed  in  the  cylinder  and  treated 
with  oil  under  high  pressure. 

The  results  of  this  investigation  are  shown  in  Table  II 


24 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  I 


Table  II. — Reaction  of  soil-film  water 


Film  water. 

Free  water. 

5 

Soil  type. 

1 

h 

m   ll>   2 

'o 

0 

a 

"3 

.1 

5 
1 

"3 

1 

1 
0 

0 

a 
8 

"o 

.1 

a 

'•3 
« 

1 

H-ion  concentratic 

Cecil  clay  loam 

Lbs. 

lO 
lO 

P.d. 

25 

25 

p.d. 

7-2 

6.9 

4 

4 

C..C 

5° 

50 

G»w. 

C.c. 

Daj'J. 

Gram-mole- 
cules per 

liter. 
0. 8+ 10-* 

Do 

S88 
679 
593 
604 
684 
614 
624 
661 
666 
654 
694 
929 
877 
976 
806 

.  4+  IC-* 

Do 

5 

so 

4 

.1  +  10-* 

Porter's  loam 

lO 
lO 

25 
25 

6.  4 
4-9 

,- 

.3  +  IO-* 

Do 

4  1       50 

.2  +  10-* 

Do 

5 

50 

4 

.9+10-8 

Norfolk  fine  sandy  loam 

lO 
lO 

20 
20 

15-3 
13-3 

3 
3 

50 
SO 

.i  +  io-< 

Do 

.1  +  10-* 

Do : 

S 

50 

3 

.  2  +  IQ-* 

Norfolk  sand 

lO 
lO 

15 
15 

18.4 
20.  6 

2 
2 

50 
50 

.1  +  10-5 

Do 

.3  +  10-5 

Do 

S 

50 

2 

.  6+ 10-5 

Tellico  sandy  loam  (^ 

5 

20 

12.8 

3 

25 

.S  +  iQ-iO 

Do 

z-S 

25 

3 

.4+10-9 

5 

25 

7-  7 

3 

25 

.8+10-11 

Do 

2-5 

25 

3 

.1  +  10-9 

o  Sufficient  soil  not  available  for  duplicate  extractions. 

The  data  presented  in  Table  II  show  rather  conclusively  that  the 
soil-film  water  has  the  same  reaction  as  the  free  water.  The  difference 
is  only  in  degree  of  intensity.  Those  soils  which  showed  an  acid  reaction 
in  suspensions  gave  a  greater  concentration  of  H  ions  in  the  film  water. 
Conversely,  those  which  indicated  a  greater  OH-ion  concentration  than 
H  in  suspension  gave  a  greater  intensity  of  OH  ions  in  the  solution. 
These  results  are  not  in  keeping  with  those  of  Sharp  and  Hoagland  (12), 
who  found  no  appreciable  change  in  H  or  OH  ions  in  varying  the  pro- 
portions of  soil  to  water  in  making  suspensions. 

With  the  freezing-point  method  for  measuring  the  concentration  of 
the  soil  solution  Bouyoucos  and  McCool  (j)  show  that  as  the  amount  of 
water  decreases  in  an  arithmetic  progression  the  lowering  of  the  freez- 
ing point  (increase  in  concentration)  increases  in  a  geometric  progression. 
This  apparently  is  due  to  the  rendering  inactive  of  a  portion  of  the  water 
in  the  soils,  hence  this  water  does  not  take  part  in  dissolving  the  solutes 
of  the  soil.  That  portion  left  free  or  tmcombined  becomes  highly  charged 
with  soluble  salts  and  gives  phenomenal  increases  in  the  freezing-point 
lowering.  The  same  line  of  reasoning  may  be  applied  to  the  increase  in 
intensity  of  reaction  of  film  water  when  compared  to  that  of  soil  sus- 
pensions. 

With  the  heavier  types  of  soil  only  a  small  percentage  of  the  added 
water  was  recovered.  It  would  be  of  extreme  interest  to  note  the 
intensity  of  reaction  of  the  thinnest  moisture  film  which  could  remain 
in  contact  with  the  soil  grains.  In  other  words,  all  of  the  liquid  added 
to  the  soil  should  be  recovered  and  studied. 


Jan.  7,  1918 


Studies  in  Soil  Reaction 


25 


EFFECT  OF  FERTILIZER  MATERIALS  ON  SOIL  REACTION 

It  has  frequently  been  stated  that  certain  fertilizer  materials,  more 
especially  ammonium  sulphate  and  acid  phosphate,  increase  the  acidity 
of  soils.  To  obtain  additional  evidence,  the  H-ion  concentration  of 
soil  suspensions  have  been  measured  in  samples  taken  from  the  ferti- 
lizer plots  of  the  North  Carolina  Experiment  Station.  Some  of  the 
soils  have  received  annual  applications  of  these  fertilizers  for  as  many 
as  15  years.  All  fields  are  located  on  well-defined  soil  types  for  each 
area  of  the  State.     Three  distinct  types  have  been  studied. 

The  samples  of  both  soil  and  subsoil  were  collected  as  carefully  as 
possible.  Borings  were  made  at  several  points  on  the  plots  for  both 
surface  and  subsoil  samples.  These  were  taken  to  the  laboratory  as 
quickly  as  possible  and  thoroughly  mixed,  after  which  water  was  deter- 
mined in  each  composite  sample.  The  ratio  of  water  to  soil  was  kept 
the  same  as  with  the  air-dried  soils. 

Table  III  gives  a  compilation  of  the  total  amount  of  each  fertilizer 
material  and  lime  which  has  been  applied  to  the  plots  at  the  different 
branch  stations. 


Table  III. 


-Total  quantity  {in  pounds  per  acre)   of  fertilizer  materials  applied  to 
experimental  plots 


Branch  station. 

Ammo- 
niiun  sul- 
phate (20 

per  cent 
nitrogen). 

Sodium 

nitrate 

(14  per  cent 

nitrogen). 

Dried  blood 
(14  per  cent 
nitrogen). 

Acid 
phosphate 
(16  per  cent 
phosphorus 
pentoxid). 

Potassium 

sulphate, 

(40  per  cent 

potassium 

oxid). 

Lime 
(90  per  cent 

calcium 
carbonate). 

Buncombe 

4,826 
1,288 

"■i,'638' 

1,846 
3>332 
3,237 
3,372 

I,  100 
540 
307 

457 

6,  000 

Iredell 

577 
577 
577 

675 
675 
67s 

4,  000 
4,  000 
4,  000 

Central 

Edgecombe 

The  data  secured  from  the  measurements  of  the  H-ion  concentration 
of  plots  fertilized  with  ammonium  sulphate  and  sodium  nitrate  are  given 
in  Table  IV. 

Marked  increases  in  acidity  can  be  noticed  in  samples  taken  from  plots 
fertilized  with  ammonium  sulphate.  In  every  case  pronounced  increase 
of  H  ions  is  evident,  and  extends  to  the  subsoil  in  all  of  the  fields  studied. 
On  the  Norfolk  fine  sandy  loam  more  acidity  is  noted  in  the  subsoil  than 
surface. 

There  does  not  appear  to  be  any  relationship  between  texture  and 
development  of  acidity  by  ammonium  sulphate.  This  is  in  agreement 
with  the  work  of  Allison  and  Cook  (/). 

The  results  secured  from  the  sodium-nitrate  plots  are  in  harmony  with 
the  accepted  theory  regarding  its  effect  on  soil  reaction.  A  reduction  of 
acidity  is  apparent. 


26 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  r 


Table  IV. — Effect  of  ammonium  sulphate  and  nitrate  of  soda  on  the  H-ion  concentration 

of  field-soil  suspensions 


Branch  station. 

Year  of 
begin- 
ning 

experi- 
ment. 

Soil  type. 

Depth. 

Treatment. 

Volt- 
meter 
read- 
ings. 

H-ion  con- 
centration. 

Iredell     

1907 
1907 
1907 
1907 
1907 
1907 
1907 
1907 
1907 
1907 
1907 
1907 
1907 
1907 
1907 
1907 

Cecil  clay  loam 

Inches. 
0-6 
6-24 
0-6 
6-24 
0-6 
6-24 
0-8 
8-24 
0-8 
8-24 
0-8 
8-24 
0-6 
6-24 
0-6 
6-24 
0-6 
6-24 

Ammonium  sulphate . .  , 
do 

0  ';42 

Gram-mole' 
cules  per 
liter. 
0.2X10-' 

Do 

do 

568 
694 
678 
656 
649 
548 
582 
671 
656 
613 
603 
556 
548 
666 
648 
646 
6?8 

.9X10-* 

Do  

do 

Sodium  nitrate 

.6Xio-« 

Do 

.   ...do 

do 

.iXic-* 

Do 

do 

Untreated 

do 

Ammonium  sulphate .  . . 
do 

.2X1C-* 

Do 

do     

.3X1C-* 

Central 

Do 

Durham  sandy  loam . . . 
do              

.2Xio-» 
•  sXic-* 

Do 

...    do 

Sodium  nitrate 

.1X10-* 

Do 

do 

do 

.2X10-5 

Do 

do 

Untreated 

.iXio-< 

Do     ■ 

do 

do 

.2XlO-« 

Edgecombe .... 
Do 

Norfolk  fine  sandy  loam 
. .     do     

Ammonimn  sulphate .  . . 
.do... 

.1x10-8 
.2x10-8 

'  Do 

do 

.iXio-s 

Do 

do   

do 

.3X10-5 

Do 

do 

.4X10-S 

Do 

1907 

do 

do 

.  8X  lo-s 

EFFECT  OF  POTASSIUM  SULPHATE  ON  SOIL  REACTION 
Skinner  and  Beattie  {13)  and  others  have  observed  that  potassium 
sulphate  increased  the  lime  requirement  of  soils.  Measurements  have 
been  made  of  the  H-ion  concentration  of  soil  suspensions  from  plots  to 
which  have  been  added  different  amounts  of  this  salt.  These  data  are 
given  in  Table  V. 

Table  V. — Effect  of  potassium  sulphate  on  the  H-ion  concentration  of  field-soil  sus- 
pensions 


Branch  sta- 
tion. 


Buncombe . 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Iredell 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Edgecombe 

Do 

Do 

Do 

Do 

Do 

Do 

Do 


Year 

of 
begin- 
ning 
ejcperi- 
ment. 


Soil  type. 


191 1 
1911 
191 1 

1911 
1911 
1911 
1911 
1911 
1903 
1903 
1903 

1903 
1903 
1903 
1903 
1903 
1903 
1903 
1903 

1903 
1903 
1903 
1903 
1903 


Porter's  loam. 

....do 

....do 


do 

do 

do 

do 

do 

Cecil  clay  loam. 

do 

do 


.do. 
.do. 
.do. 
.do. 
do. 


Norfolk  fine  sandy  loam 

do 

do 


.do. 
.do. 
.do. 
.do. 
.do. 


Depth. 


Inches. 

0-& 

8-24 


8-24 

0-8 

8-24 

0-8 

S-24 

0-8 

8-36 

0-8 

8-36 

0-8 

8-36 

0-8 

0-36 

0-8 

8-24 

0-8 

8-24 

0-8 

8-24 

0-8 

8-24 


Treatment. 


Dried  blood 

do 

Dried  blood,  potassium 

sulphate. 

do 

Potassium  sulphate 

do 

Untreated 

....do 

Dried  blood 

....do 

Dried  blood,  potassium 

sulphate. 

do 

Potassium  sulphate 

do 

Untreated 

....do 

Dried  blood 

...do 

Dried  blood,  potassiimi 

sulphate. 

....do 

Potassium  sulphate 

....do 

Untreated 

....do 


Volt- 
meter 
read- 
ings. 


0.656 
.654 
.623 

.623 
.  609 
•  596 
.648 
.651 
.682 
.  671 
.630 

.618 
.621 
.608 
.676 
.  672 
.632 
.  620 
.  602 

•S90 
•596 
.  600 
•634 
.628 


H-ion  con- 
centration. 


Gram- 

nwlecules 
per  liter. 
o.  2X10-5 
.3X10-5 
.iXio-< 

.iXio-< 
.1X10-* 
.3Xio-< 
.3X10-6 
.3X10-5 
. I X 10-8 
.1X10-5 
.8X10-5 

.1X10- 
.1X10- 
.1X10- 
.1X10-' 
.iXio- 
.7X10- 
.iXio- 
.2X10-' 

.3X10- 

.3X10 

.2X10- 

.6X10-5 

.8X10-5 


Jan.  7,  1918 


Studies  in  Soil  Reaction 


27 


Slight  increase  in  the  H-ion  concentration  was  obtained  from  the 
plots  to  which  potassium  sulphate  had  been  applied.  However,  nothing 
like  as  marked  an  effect  in  producing  "true  acidity"  is  found  on  these 
plots  as  those  to  which  have  been  added  ammonium  sulphate.  Why 
this  should  be  is  not  clear,  unless  it  has  been  caused  by  nitric  acid  devel- 
oped by  the  soil  organisms. 

EFFECT  OF  ACID  PHOSPHATE  AND  LIME  ON  SOIL  REACTION 

Table  VI  gives  the  results  from  the  plots  which  have  received  annual 

applications  of  acid  phosphate  and  lime. 

Table  VI. — Effect  of  acid  phosphate  and  lime  on  the  H-ion  concentration  of  field-soil 

suspensions 


Branch  sta- 
tion. 


Buncombe . 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Iredell 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Edgecombe 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 


Year 

of 
begin 

ning_ 
experi- 
ment. 


191 1 
19 1 1 
1911 
1911 
1911 

1911 
1911 


1911 
1911 


1911 
1911 
1911 
1911 
19H 
1903 
1903 
1903 
1903 
1903 

1903 
1903 


1903 
1903 


1903 
1903 
1903 
1903 
1903 
1903 
1903 
1903 
1903 
1903 

1903 
1903 


1903 
1903 


1903 
1903 
1903 
1903 
1903 


Soil  type. 


Porter's  loam. 

...do 

....do 

....do 

....do 


.do. 
.do. 


do. 
do. 


do 

do 

do 

do 

...do 

Cecil  clay  loam. 

...do 

....do 

do 

....do 


.do. 
.do. 


.do. 
-do. 


do. 
.do. 
.do. 
.do. 

do. 


Norfolk  fine  sandy  loam. 

do 

do 

do 

do 


do. 
.do. 


.do. 
.do. 


do. 
.do. 
.do. 

do. 
.do. 


Depth. 


Inches. 
0-8 
8-30 
0-8 
8-30 
0-8 

8-30 
0-8 


8-30 
0-8 


8-30 

0-8 

8-30 

0-8 

8-30 

0-6 

6-30 

0-6 

6-30 

0-6 

6-30 
0-6 


6-30 
0-6 


6-30 

0-6 

6-30 

0-6 

&-30 

0-6 

6-24 

0-6 

6-24 

0-6 

6-24 
0-6 


6-24 
0-6 


6-24 

0-6 

6-24 

0-6 

6-24 


Treatment. 


Dried  blood 

.....do 

Acid  phosphate 

.....do 

Dried  blood,  add  phos- 
phate. 

.....do 

Dried  blood,  acid  phos- 
phate, potassimn  sul- 
phate. 

do 

Dried  blood,  acid  phos- 
phate, potassium  sul- 
phate, lime. 

do 

Lime 

do 

Untreated 

.....do 

Dried  blood 

.....do 

Acid  phosphate 

.....do 

Dried  blood,  acid  phos- 
phate. 

do 

Dried  blood,  acid  phos- 
phate, potassium  sul- 
phate. 

do 

Dried  blood,  acid  phos- 
phate, potassium  sul- 
phate, lime. 

do 

Lime 

do 

Untreated 

...do 

Dried  blood 

...do 

Acid  phosphate 

...do 

Dried  blood,  acid  phos- 
phate. 

...do 

Dried  blood,  acid  phos- 
phate, potassium  sul- 
phate. 

...do 

Dried  blood,  acid  phos- 
phate, potassium  sul- 
phate, lime. 

....do 

Lime 

....do 

Untreated 

....do 


Volt- 

meter 

H-ion  con- 

read- 

centration. 

mgs. 

Gram- 

mclecuUs 

per  liter. 

0.656 

0.2X10-5 

•654 

•  3X10-5 

.662 

.  2X10-^ 

•6SS 

.2X10-* 

.652 

.3X10-5 

.660 

.2X10-^ 

.612 

.1X10-* 

.  610 

.1X10-6 

.702 

.4Xio-« 

.716 

.2X10-* 

.728 

.1X10-8 

•736 

.iXio-« 

.648 

.3X10-6 

■651 

.3X10-6 

.682 

i.oXio-6 

.671 

.1X10-6 

.674 

.1X10-6 

.679 

.1X10-6 

.684 

.9X10-6 

.666 

.1X10-6 

.613 

.1X10-* 

.609 

.iXio-< 

.712 

.3X10-6 

•732 

.1X10-6 

.746 

.  7X10-' 

•754 

.5X10-' 

.666 

.1X10-6 

.6si 

.3X10-6 

.632 

.7X10-^ 

.620 

.iXio-< 

.  624 

.1X10-* 

.624 

.iXio-* 

.  614 

.iXio-« 

.609 

.iXio-< 

.600 

.2Xio-< 

•S92 

•  3Xio-« 

.712 

•  3X10-6 

.716 

.2X10-6 

.726 

.iXio-« 

•744 

.8X10-' 

.638 

.SXio-6 

.628 

.8X10-6 

28  Journal  of  Agricultural  Research  voi.xii,  no.  i 

The  results  from  the  plots  to  which  acid  phosphate  has  been  added 
do  not  show  any  greater  H-ion  concentration  than  the  ones  used  as 
controls.  These  plots  have  received  rather  heavy  annual  applications 
of  this  fertilizer  for  the  past  15  years,  the  total  amount  applied  being 
over  3,000  pounds  per  acre.  The  fine  sandy  loam  may  be  some  excep- 
tion. In  this  case  the  readings  are  so  nearly  the  same  as  from  those 
plots  which  have  received  no  fertilizer  that  the  differences  are  within 
the  range  of  experimental  error.  Indeed  there  is  as  much  evidence  in 
indicating  an  increased  basicity  from  the  use  of  acid  phosphate  in  the 
clay  loam  and  loam  as  from  an  increase  in  acidity  in  the  sandy  loam. 
This  is  in  agreement  with  the  more  recent  work  of  Conner  (5),  Brooks  (4), 
and  Bear  and  Salter  (2). 

Additions  of  lime  alone  or  lime  in  combinations  with  the  fertilizer 
materials  have  materially  reduced  the  acidity  in  all  plots.  This  is  often 
more  marked  in  the  subsoil  than  in  the  surface.  With  the  exception  of 
the  Cecil  clay  loam,  lime  has  not  been  used  in  sufficient  amounts  to  pro- 
duce basicity. 

EFFECT  OF  AMMONIUM  SULPHATE  AND  MONOCALCIUM    PHOSPHATE 
ON  H-ION  CONCENTRATION  OF  SOIL-FILM  WATER 

The  results  heretofore  reported  in  this  paper  with  ammonium  sulphate 
and  acid  phosphate  have  been  derived  from  soil  suspensions.  The  ques- 
tion arises.  Is  the  film  water  of  the  soil  affected  in  the  same  or  different 
manner  from  the  free  water?  To  secure  data  on  this  question  three 
soils  were  treated  with  the  ammonium  sulphate  and  monocalcium  phos- 
phate at  optimum  moisture  conditions  and  extractions  made  with  the 
Morgan  apparatus.  The  materials  were  applied  in  solution  as  a  fine 
spray  over  the  soils  in  order  to  get  as  good  distribution  as  possible. 
The  soils  were  well  worked  after  the  additions  to  get  a  uniform  mass. 
Monocalcium  phosphate  was  substituted  for  acid  phosphate  on  account 
of  its  complete  solubility.  Table  VII  gives  the  data  derived  from 
treatment  with  ammonium  sulphate. 

The  film  water  is  shown  to  be  more  strongly  acid  from  the  treatment 
with  ammonium  sulphate  than  that  developed  when  the  same  amount 
of  salt  is  applied  in  suspension.  The  indications  from  this  are  that 
methods  for  estimating  soil  reaction  or  lime  requirement  based  on  treat- 
ing the  soil  with  a  neutral  solution  do  not  give  the  total  acidity  in  the 
filtered  extract. 

The  mechanism  of  this  reaction  has  been  the  subject  of  much  conten- 
tion. The  explanation  which  has  been  offered  that  the  basic  radicle  has 
been  absorbed  by  celloidal  material;  and  the  acidity  developed  from 
the  combination  of  SO4  with  2H  of  the  slightly  ionized  HjO  leaves  an 
unbalanced  equation.  Parker  {10)  contends  that  the  fine  soil  particles 
catalyze  the  reaction  (NHj2S04  +  2HOH  =  2NH,OH-}-H2S04  with 
the  removal  of  the  entire  base  from  solution  by  selective  adsorption 


Jan.  7,  1918 


Studies  in  Soil  Reaction 


29 


phenomena.  The  contention  that  the  base  has  been  removed  by  com- 
bination with  the  difficultly  soluble  acids  more  nearly  agrees  with  the 
results  obtained.  The  measurements  showing  the  effect  of  monocal- 
cium  phosphate  on  the  H-ion  concentration  of  soil-film  water  are  given 
in  Table  VIII. 

Table  VII. — Effect  of  ammonium  sulphate  on  the  H-ion  concentration  of  soil-film 

water 


Film  water 

Free  water. 

■3 

c2 
3  2 

> 

B 

•6 

a! 

4-» 

1 

.3 

3 
3 

CD 

3 

Soil  type. 

^^t 

O'm 

u 

—  ij 

0 

"o 

oit 

8 

>. 

.^fi 

0  a 

.1 

-1 

SI 

8.i 

.^i 

>. 

J -a 

fe  1 

It 

^ 

n 

3  2 

"0 

'0 

"3 

1 

3 

II 

3  2 

3 

a 
3 

as 
1 

§ 

a 

OH 

;^ 

s 

> 

> 

s 

a 

Oti 

a 

> 

a 

Gram-mole- 

Grantrmole- 

ciUes  per 

cules  per 

Lbs. 

Gm. 

p.  a. 

p:ct. 

C.c. 

liter. 

Gwi. 

Gm. 

c.c. 

c.c. 

liter. 

Cecil  day  loam .... 

10 

10 

25 

6.0 

so 

0.523 

0.5X10-3 

5 

0.  II 

5° 

so 

0.  644 

0. 4X  io-« 

Do 

10 

20 

25 

4.8 

SO 

.518 

.7X10-'' 

5 

.  22 

50 

50 

.  612 

.iXio-< 

Do             

so 

•  496 

.1X10-2 

5 

•  33 

50 

50 

•  591 

■  3Xio-« 

Do 

S 

5° 

50 

•  674 

.iXio-s 

Porter's  loam 

10 

10 

25 

13.6 

SO 

■532 

.4X10-3 

S 

.  11 

SO 

50 

.  602 

.2X10-* 

Do 

10 

20 

25 

14. 1 

SO 

•S03 

.1X10-2 

S 

.  22 

50 

so 

.588 

.4Xio-« 

Do 

10 

30 

25 

II.  2 

SO 

.  461 

.6X10-2 

5 

•33 

SO 

so 

•572 

.8Xio-< 

Do 

s 

SO 

50 

.686 

.9Xio-« 

Norfolk  fine  sandy 

loam 

10 

10 

20 

10.  6 

SO 

•  540 

.2X10-3 

s 

.  II 

SO 

5° 

•S84 

•  SXio-< 

Do 

10 

20 
30 

20 

II.  9 
13-2 

SO 

.528 

.4X10-3 
i.oXio-3 

s 

5 

.  22 

•  1? 

50 
50 

SO 

S° 

•578 
.566 

.6Xio-< 

Do. 

.1X10-3 

Do 

S 

SO 

SO 

.670 

.1X10-5 

Table  VIII. — Effect  of  monocalcium  phosphate  on  the  H-ion  concentration  of  soil- 
film  luater 


Soil  type. 


Quan- 
tity of 
soil. 


Quan- 
tity of 
mono- 
calcium 
phos- 
phate 


Mois- 
ture 
con- 
tent, 
dry 
basis. 


Mois- 
ture 
recov- 
ered. 


Vol- 
ume of 

film- 
water. 


Volt- 
meter 
read- 
ings. 


Per 

cent. 

C.c. 

7-4 

50    0 

(^•3 

50 

6.1 

50 

II. 8 

50 

10.  9 

50 

14.4 

50 

12.  2 

50 

II.  I 

50 

II.  9 

50 

H-ion  concen- 
tration. 


Cecil  clay  loam 

Do 

Do 

Porter 's  loam 

Do 

Do 

Norfolk  fine  sandy  loam 

Do 

Do 


Pounds. 
10 
10 
ID 
10 
10 


Cm. 
10 
20 

30 
10 
20 
30 
10 
20 
30 


Per 
cent. 

25 
25 
25 
25 
25 
25 
20 
20 
20 


562 

598 
602 
562 
602 
591 


Gramrmolecules 
per  liter. 
0.4X10-4 
.3X10-* 
■3X10-^ 
.2X10-* 
. 2  X IO-* 
.1X10-3 
.2X10-* 
•3X10-" 
.4X10-* 


By  comparing  the  data  of  Table  II  with  those  derived  from  this  experi- 
ment it  is  apparent  that  only  excessive  applications  of  monocalcium 
phosphate  have  increased  the  H-ion  concentration.  The  20-gm.  appli- 
cations of  the  salt  do  not  show  any  increase  in  "true  acidity"  with  any 
soil  used.     The  clay  loam  and  loam  give  a  higher  H-ion  concentration 


30  Journal  of  Agricultural  Research  voi.  xii,  No.  x 

with  the  30-gm.  addition.  The  iine  sandy  loam  shows  more  "true 
acidity"  with  a  20-gm.  application  than  when  10  gm.  are  added,  and 
still  more  when  30  gm.  have  been  applied. 

The  fixation  or  removal  from  solution  of  phosphates  is  supposed  to 
be  done  by  the  bases,  such  as  iron,  aluminum,  calcium,  etc.,  in  the  soil. 
The  clay  loam  and  loam  soils  are  well  supplied  with  very  fine  particles 
of  iron  and  aluminum  compounds.  They  therefore  have  the.  capacity  of 
fixing  more  soluble  phosphate  than  the  fine  sandy  loam,  which  has  a 
relatively  low  content  of  bases.  These  data  are  in  accord  with  those 
obtained  by  Conner  on  soils  of  Indiana. 

SUMMARY 

The  hydrogen  electrode  has  been  used  for  indicating  soil  reaction  on 
a  number  of  untreated  soils  in  suspension.  The  soils  experimented  with 
represent  a  wide  range  in  texture  of  those  common  to  the  area  of  the 
southeastern  portion  of  the  United  States,  extending  from  and  including 
the  Appalachian  Mountains  to  the  Atlantic  Ocean.  The  H-ion  concen- 
tration varies  from  almost  "true  neutrality"  to  rather  excessive  "true 
acidity"  in  the  soils. 

With  the  Morgan  apparatus  for  extracting  film  water  from  soils,  it 
is  shown  that  its  reaction  is  the  same  as  the  free  water,  differing  only  in 
intensity. 

The  effects  of  certain  fertilizers  on  the  H-ioh  concentration  of  long- 
time-treated plots  of  three  soils  have  been  measured,  with  the  following 
results:  (i)  Ammonia  sulphate  has  materially  increased  the  H-ion 
concentration  of  all  plots  which  have  received  applications  of  this  material. 
The  acidity  thus  developed  extends  often  to  the  subsoil.  (2)  Sodium 
nitrate  has  slightly  reduced  the  acidity  of  the  plots  to  which  it  has  been 
applied.  (3)  Potassium  sulphate  increases  the  "true  acidity"  when 
applied  to  soils,  though  not  as  greatly  as  ammonium  sulphate.  (4)  Acid 
phosphate  does  not  appear  to  have  affected  in  either  direction  the  H-ion 
concentrations  of  field  soils.  (5)  Lime  materially  increases  the  OH-ion 
concentration  of  field  plots  to  which  it  has  been  added. 

The  acidity  developed  from  ammonium  sulphate  is  more  intense  in  the 
film  than  in  the  free  water  of  three  soils. 

Monocalcium  phosphate  does  not  change  in  any  way  the  soil-film  water 
until  excessive  amounts  are  added. 

LITERATURE  CITED 
(i)  AxLisoN,  F.  E.,  and  Cook,  R.  C. 

191 7.   THE  EFFECT  OF  AMMONIUM  SULPHATE  ON  SOIL  ACIDITY.      In  Soil  vScicnCC, 

V.  3,  no.  6,  p.  507-512,  I  fig. 
(2)  Bear,  F.  E.,  and  Salter,  R.  M. 

1916.  THE  residual  effects  OF  FERTILIZERS.    W.  Va.  Agr.  Exp.  Sta.  Bui. 
160,  26  p.,  2  diagr. 


Jan.  7.  i9i8  Studies  in  Soil  Reaction  31 

(3)  BouYOUCOS,  G.  J.,  and  McCooL,  M.  M. 

1916.  FURTHER  STUDIES  ON  THE  FREEZING  POINT  LOWERING  OF  SOILS.  Mich. 
Agr.  Exp.  Sta.  Tech.  Bui.  31,  51  p.,  i  fig. 

(4)  Brooks,  W.  P. 

1915.  PHOSPHATES  IN  MASSACHUSETTS  AGRICULTURE;  IMPORTANCE,  SELECTION 

AND  USE.     Mass.  Agr.  Exp.  Sta.  Bui.  162,  p.  131-167,  2  pi. 

(5)  Conner,  S.  D. 

^  1916.  acid  soils  and  the  effect  of  acid  phosphate  and  other  fertilizers 
UPON  THEM.  In  Jour.  Indus,  and  Engin.  Chem.,  v.  8,  no.  i,  p.  35-40, 
2  fig. 

(6)  FiNDLAY,  Alexander. 

1906.  PRACTICAL  PHYSICAL  CHEMISTRY.  282  p.,  92  fig.  London,  New  York 
and  Bombay. 

(7)  GaLESPiE,  L.  J. 

1916.  THE  REACTION  OF  SOIL  AND  MEASUREMENTS  OF  HYDROGEN-ION  CONCEN- 

TRATION.    In  Joiu-.  Wash.  Acad.  Sci.,  v.  6,  no.  i,  p.  7-16,  2  fig. 

(8)  HiLDEBRAND,  J.  H. 

I913.  SOME  APPLICATIONS  OF  THE  HYDROGEN  ELECTRODE  IN  ANALYSIS,  RE- 
SEARCH AND  TE.^CHiNG.  In  Joiu.  Amer.  Chem.  Soc,  v.  35,  no.  7, 
p.  847-871,  15  fig. 

(9)  Morgan,  J.  F. 

191 7.  THE  soil  SOLUTION  OBTAINED  BY  THE  OIL  PRESSURE  METHOD.      In  Soil 

Science,  v.  3,  no.  6,  p.  531-546,  i  pi.     Literature  cited,  p.  544-545. 

(10)  Parker,  E.  G. 

1913.  SELECTIVE  ABSORPTION  BY  SOILS.  In  Jour.  Agr.  Research,  v.  i,  no.  3, 
p.  179-188,  2  fig. 

(11)  Schmidt,  C.  L.  A. 

1909.  TABLE  OF  H  -f-  AND  OH  —  CONCENTRATIONS  CORRESPONDING  TO  ELECTRO- 
MOTIVE FORCES  DETERMINED  IN  GAS-CHAIN  MEASUREMENTS.      In  Univ. 

Cal.  Pub.  Phys.,  v.  3,  no.  15,  p.  101-113. 

(12)  Sharp,  L.  T.,  and  Hoagland,  D.  R. 

1916.  ACIDITY    AND    ADSORPTION    IN    SOILS    AS    MEASURED    BY    THE    HYDROGEN 

ELECTRODE.     In  Jour.  Agr.  Research,  v.  7.  no.  3,  p.  123-145,  i  fig. 
Literature  cited,  p.  143-145. 

(13)  Skinner,  J.  J.,  and  BeattiE,  J.  H. 

1917.  INFLUENCE    OF   FERTILIZERS    AND   SOIL    AMENDMENTS    ON    SOIL   ACIDITY. 

In  Jour.  Amer.  Soc.  Agron.,  v.  9,  no.  i,  p.  25-35. 


ADDITIONAL  COPIES 

OP  THIS  PXTBI.ICATION  MAY  BE  PROCURED  FROM 

THE  SUPERINTENDENT  OF  DOCUMENTS 

GOVERNMENT  PRINTING  OFFICE 

WASHINGTON,  D.  C. 

AT 

10  CENTS  PER  COPY 

Subscription  Price,  S3.00  Per  Year 

A 


CONXKNXS 

Pure  Ctiltxires  of  Wood-Rotting  Fucgi  on  Artificial  Media 
W.  H.  lONG  and  R.  M.  HAJRSCH 

(  Contribution  from  Bureau  ol  Plant  lurtustr;.'  ' 

Gossypoi,  tbe  Toxic  Substance  in  Cottonseed 

W.  A.  WITHERS  aad  FRAKK  E.  CAIIRUTH 

( ;  .-•ntribaiion  Ii'cta  North  Carolina  AKricotWral  'KxDerhjrient  Station 

Frait-FIj  Parasitism  La  Hawaii  Dimng  19  Hj       -        -        - 
C.  E.  PEMBERTON  and  H,  F.  WILLARD 
( CoatdbuUan  from  Bureau  ot  Entamolosy) 


S3 


103, 


PDBUSHEft  BY  AUTHORm  OF  THE  SECRETARY  OF  AGRICULTURE. 

WITH  THE  COOPERATION  OF  THE  ASSOCIATION  OF  AAIERICAN 

AGRirriTTT'RAf   COIJ.VCVP,  XMW  FTPFRIMET^T  STATIONS 


WASMINGT01>i,  r>.  C 


w/iMNOTOM  '.  aOVSR^'Mew 


EDITORIAL  COMMITTEE  OF  THE 

ONITED    STATES    DEPARTMENT    OF    AGRICULTURE    AND 

THE  ASSOCIATION  OF  AMERICAN  AGRICULTURAL 

COLLEGES  AND  EXPERIMENT  STATIONS 


yOR  THK  OBPAJRriVlBM  I 

5CARL  F,  KRLLERMAN,  Chaikman 

Pkysiologisl  and  Associate  Chief,  Buretm 
of  Plant  Industry 

KDWINW.  ALLEN 

Chief,  OfJicf.  of  Exi>eriment  Statumi, 

CHARLES  L.  MARLATT 

BntomolQgisl  and  Asiixiani  Chief,:  B«utti>M 
of  Rntomotoov 


fOR  THK  ASSOCIATION 

RAYMOND  PEARL* 
Biologitf,  Afaiue  Aoricuihual  Ktpfrtmm*.' 
Sla^ion 

H.  P,  ARAISBV 

Director,  Imlitute  of  AntftitJ  Ntttriitm.  Tkti 
Pennsylvania  Slate  ColU'i' 

8.  M.  FREEMAN 

Bfftnnist.  plant  Paihotoyift  and  Attutant 
Oean,  Afriatlturet  Experiment  Stmi^m  oi 
ibeUntversiiyofMinntsota 


All  correspondence  regarding  articles  from  the  Dep^rtmeatof  Agricultimi  should  txe 
addressed  to  Karl  F,  Kellerman,  Jourtxal  of  Agricultural  Research,  Washington,  D.  C. 

*  Dr.  Pearljias  undertaken  special  work  in  connection  with  the  war  emei^ncy; 
therefore,  until  further  notice  •all  correspondence  regarding  articles  from  State  Experi- 
ment Stations  should  be  addressed  to  H.  P.  Arniaby,  Institute  of  Animal  Nutrition . 
State  College.  Pa, 


JOraALOFAGRKlTIIALlSEARCe 


Vol.  XII 


Washington,  D.  C,  January  14,  191 8 


No. 


PURE  CULTURES  OF  WOOD-ROTTING  FUNGI  ON 
ARTIFICIAL  MEDIA 

By  W.  H.  Long,  Forest  Pathologist,  and  R.  M.  Harsch,  Assistant  in  Forest  Pathology, 
Bureau  of  Plant  Industry,  United  States  Department  of  Agriculture  ' 

INTRODUCTION 

The  study  of  wood-rotting  fungi  by  means  of  cultures  on  artificial  media 
has  been  very  meager  in  the  past  compared  to  the  almost  universal  use  of 
cultural  methods  by  bacteriologists  and  workers  with  strictly  parasitic 
fungi.  A  critical  study  of  the  existing  literature  on  cultures  of  wood- 
rotting  fungi  develops  the  fact  that  much  of  this  work  was  either  not  done 
under  proper  control  conditions  where  the  purity  of  the  organism  under 
investigation  was  guaranteed  or  the  media  used  by  many  of  the  workers 
consisted  of  pieces  of  wood,  bread,  dung  decoctions,  etc.,  and  not  artificial 
media  of  such  a  character  that  others  could  reproduce  the  media,  growth 
conditions,  etc.,  and  thus  repeat  and  verify  the  experiments. 

Brefeld  (i-sY,  Falck  (5-7),  Humphrey  and  Fleming  (8),  Lyman  (11), 
Rumbold  (12),  and  Zeller  (75, 14)  are  some  of  the  workers  who  have  made 
cultural  studies  of  a  number  of  hymenomycetous  fungi  on  a  rather  exten- 
sive scale.  However,  the  line  of  investigation  followed  by  most  of  them 
has  been  more  along  the  lines  of  polymorphism  in  spore  forms,  enzymic 
action,  the  rot  caused  by  each  fungus,  and  the  prevention  or  control  of 
these  fungi  in  the  rotting  of  structural  timber  rather  than  a  critical  study 
of  their  cultural  characters  on  artificial  media.  Probably  the  most  seri- 
ous drawback  to  investigators  in  working  with  wood-rotting  fungi,  espe- 
cially the  Polyporaceae,  has  been  the  fact  that  it  was  not  possible  under 
conditions  used  by  them  to  obtain  with  any  degree  of  certainty  the  sporo- 
phores  of  the  various  fungi  on  artificial  media. 

This  paper  deals  with  two  lines  of  investigation  of  fungous  activity 
when  grown  in  pure  cultures :  (i)  A  method  by  which  various  wood- rotting 
fungi  can  be  differentiated  from  each  other  by  their  cultural  characters 
alone  when  grown  upon  artificial  media;  and  (2)  a  method  by  which  the 

1  The  writers  are  under  obligations  to  Dr.  E.  A.  Burt  for  assistance  in  identifying  the  various  species 
of  the  Thelephoraceae,  and  to  Dr.  W.  A.  Murrill  for  identifying  the  more  di£5cult  species  of  the  Polyporaceae 
discussed  in  this  paper. 

2  Reference  is  made  by  number  (italic)  to  "  Literature  cited,"  pp.  81-82. 


Journal  of  Agriculture  Research, 

Washington,  D.  C. 

Ip 


(33) 


Vol.  XII.  No.  2 
Jan.   14,  1918 
Key  No.  G — 132 


34  Journal  of  Agricultural  Research  voi.  xii.  no.  2 

fruiting  bodies  or  sporophores  of  wood-rotting  fungi  can  be  produced  from 
pure  cultures  on  artificial  media.  In  this  paper  the  writers  have  only 
given  in  a  general  way  the  results  of  somewhat  extended  investigations 
on  many  species  of  wood-rotting  fungi,  leaving  for  a  later  article  a  large 
amount  of  detail  and  the  discussion  of  special  methods  and  culture  media 
which  they  have  found  very  valuable  in  working  wdth  this  group  of 
organisms. 

GENERAL  METHODS  OF  EXPERIMENTATION 

ORIGIN   OF  CULTURES 

The  initial  cultures  of  all  of  the  wood-rotting  fungi  under  investigation 
by  the  writers  have  been  obtained  from  the  three  following  sources: 
(i)  Small  pieces  of  diseased  wood,  (2)  small  pieces  of  sporophores,  and 
(3)  spores.  Pieces  of  inoculum  4  or  5  mm.  in  diameter  have  been  found 
to  be  better  than  smaller  pieces.  The  old  idea  that  the  smaller  the  piece 
the  freer  it  is  from  contamination  is  good  reasoning  theoretically,  but  in 
actual  practice  pieces  of  the  size  mentioned  above  have  been  found  more 
viable  than  small  bits  and  as  free  from  contamination.  The  larger  the 
piece  the  greater  are  the  chances  of  viable  mycelium  being  present.  The 
pieces  should  be  inserted  endwise  into  the  middle  of  the  agar  slant  until 
about  one-half  of  the  wood  is  buried  in  the  agar.  Care  should  be  taken 
to  avoid,  as  far  as  possible,  burying  the  wood  in  the  agar,  since  the  cover- 
ing of  agar  excludes  the  air  and  either  retards  or  prevents  entirely  the 
fungus  from  starting  on  the  culture  medium.  A  pair  of  long-handled 
scissors  or  forceps  are  especial!}'  suitable  for  this  work. 

In  initial  cultures  the  writers  have  found  it  very  convenient  to  use  a 
series  of  10  tubes,  including  2  tubes  each  of  carrot,  malt,  commeal, 
prune,  and  parsnip  agars.  These  agars  have  been  found  to  give  a  fairly 
good  growth  of  mycelium,  and  at  the  same  time  indications  of  the  fruiting, 
cultural  characters,  etc.  of  the  organism  may  be  obtained  even  from 
these  initial  cultures. 

The  writers  have  made  approximately  10,000  cultures  of  wood-rotting 
fungi  in  their  preliminary  studies  here  reported.  All  of  the  inoculations, 
both  initial  and  subcultures,  have  been  made  in  an  open  room  without 
the  use  of  any  special  inoculating  chamber.  The  percentage  of  pure 
subcultures  obtained  when  the  original  tube  was  uncontaminated  has 
been  very  high.  For  instance,  out  of  1,000  transfers  recently  made  only 
7  contaminated  tubes  were  found. 

METHODS   USED   IN    MAKING    SUBCULTURES 

The  writers  desire  to  describe  here  a  method  which  they  have  found 
very  useful  in  making  transfers  of  fungus  cultures  when  10  or  more 
transfers  are  to  be  made  from  the  same  tube.  The  instruments  used  in 
these  transfers  are  a  pair  of  long-handled  scissors  made  by  lengthening 
the  handles  of  a  pair  of  dissecting  scissors,  a  small  square  glass  jar  with  a 


Jan.  14, 1918     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      35 

triangular  section  cut  out  of  the  aluminum  screw  top  large  enough  to 
hold  the  plug  from  the  mother  tube,  and  a  salt-mouth  bottle  (holding  300 
or  400  c.  c.)  filled  about  two-thirds  full  with  95  per  cent  alcohol.  The 
blades  of  the  scissors  used  in  making  the  transfers  are  kept  in  this  95 
per  cent  alcohol  when  not  in  actual  use.  The  glass  jar  and  top  are  cleaned 
by  washing  in  hot  water  and  then  dried  before  using.  The  opening  in  the 
top  is  thoroughly  flamed  over  an  alcohol  lamp  or  Bunsen  burner  and  the 
jar  is  placed  on  its  side  with  the  triangular  opening  toward  the  operator. 
During  the  actual  inoculation  the  cotton  plug  from  the  mother  tube  is 
placed  in  the  triangular  opening  with  the  lower  end  of  the  plug  inside  the 
jar  in  such  a  manner  that  only  the  sharp  edges  of  the  top  of  the  jar  come 
in  contact  with  the  cotton  plug.  In  this  position  the  plug  is  protected 
from  outside  contamination  and  at  the  same  time  the  hands  of  the 
operator  are  left  free  to  handle  the  scissors,  two  culture  tubes,  and  the 
cotton  plug  from  the  tube  to  which  the  transfer  is  being  made. 

In  making  the  transfers  of  certain  standardized  series  it  was  necessary 
to  obtain  small  inocula  as  near  the  same  size  for  each  transfer  as  possible. 
The  ordinary  inoculating  needles  and  loops  made  either  of  platinum  or  of 
iridio-platinum  are  too  soft  and  in  other  ways  unsuited  for  making 
transfers  of  fungus  mycelium.  The  writers  therefore  adopted  the  use  of 
the  scissors  for  such  work,  since  by  using  them  the  mycelial  layer  on  the 
surface  of  the  agar  in  the  culture  tubes  can  be  readily  cut  and  any  desired 
size  of  inoculum  transferred  without  loss  of  time  and  with  a  minimum  of 
outside  contamination. 

VEGETATIVE  CULTURAL  CHARACTERS  ON  ARTIFICIAL  MEDIA 

MEDIA   USED 

In  studying  the  cultural  characters  of  the  various  fungi  as  outlined 
under  No.  i  of  the  introduction,  the  following  general  system  was  adopted : 
A  series  of  10  different  culture  media  in  agar  was  used  for  each  fungus. 
These  10  media  were  (i)  1.5  and  2  per  cent  carrot  agar,  +3.5  to  +5.0; 
(2)  1.5  and  2  per  cent  malt  agar,  +7.0;  (3)  1.5  per  cent  beet  agar,  +2.5 
and  +3.0;  (4)  1.5  per  cent  celery  agar,  +9.5  to  +15.5;  (5)  i-5  per  cent 
bean  agar,  +  i.o  to  +1.5;  (6)  1.5  and  2  per  cent  corn-meal  agar,  +0.25; 
(7)  1.5  and  2  per  cent  prune  agar,  +1.0  and  +1.5;  (8)  1.5  and  2  per 
cent  alfalfa  agar,  +13.5  to  +15.5;  (9)  i-5  per  cent  parsnip  agar,  4-9.0 
to  +  13.5;  and  (10)  1.5  and  2  per  cent  potato  agar,  +2.0  and  +3-5-  The 
acidity  of  the  media  here  given  is  based  on  Fuller's  scale  and  is  the  actual 
acidity  of  the  media  after  tubing  and  as  used  in  the  cultures. 

In  any  series  of  a  given  fungus  each  corresponding  agar  for  each  strain 
had  the  same  percentage  and  the  same  acidity.  For  instance,  there  were 
nine  strains  of  Trametes  pint  compared.  The  carrot  agar  used  for  each  of 
these  nine  sets  was  2  per  cent  and  had  an  acidity  of  +3.5. 

The  writers  selected  the  10  media  for  the  study  of  the  cultural  characters 
of  the  different  fungi  not  with  a  view  to  obtaining  vigorous  growth  but 


36  Journal  of  Agricultural  Research  voi.xii,  no. 


to  get  media  on  which  the  growth  on  each  would  be  different  for  the  same 
organism.  In  other  words,  the  media  used  were  not  intended  to  develop 
general  characters  but  specific  ones  which  might  differentiate  the  fungus 
under  investigation  from  other  closely  related  species. 

NUMBER   OF  TUBES   OF   EACH   MEDIUM   INOCUL.^TED 

One  tube  of  each  of  these  10  media  was  used  in  the  series  for  any  given 
fungus.  Better  results  would  probably  have  been  obtained  by  using 
three  or  more  tubes  of  each  medium  rather  than  one,  but  the  writers  could 
not  do  this  in  their  preliminary  work  for  lack  of  sufficient  equipment. 
However,  in  a  great  majority  of  cases  it  is  believed  that  accurate  results 
were  obtained  with  these  series  of  10,  since  many  of  them  have  been 
repeated  to  the  fourth  and  fifth  subcultures  with  different  batches  of 
media,  and  the  resulting  characters  when  grown  under  the  conditions 
described  below  were  practically  identical  for  each  subculture  of  the 
fungus  for  each  medium. 

POSITION   OF  CULTURE   TUBES  IN   REFERENCE  TO   GRAVITY 

After  inoculating  the  series  of  10  tubes,  they  were  placed  in  a  horizon- 
tal position,  side  by  side,  in  shallow  boxes  with  the  surface  of  the  agar 
slant  uppermost.  The  boxes  were  from  2  to  4  cm.  deep  and  about  14 
to  14.5  cm.  wide.  The  culture  tubes  (150  mm.  long  or  longer)  had  their 
tops  resting  on  the  upper  edges  of  the  boxes  and  were  therefore  tilted  at 
a  slight  angle.  These  boxes  were  then  placed  on  shelves  in  front  of 
windows  with  a  western  exposure  where  they  received  all  of  the  diffused 
light  which  came  through  and  during  the  afternoon  received  the  direct 
rays  of  the  sun  from  one  to  four  hours  daily.  Under  these  conditions 
the  agar  in  the  tubes  gradually  dried  and  the  upper  portion  of  it  separated 
from  the  glass,  leaving  a  space  of  varying  depth  between  the  agar  and  the 
glass,  on  which  the  aerial  mycelium  could  grow  even  to  the  bottom  of 
the  tube. 

AMOUNT   OF   DIRECT   SUNLIGHT   CULTURES   RECEIVED 

The  earlier  cultures  of  the  writers  received  only  one  or  two  hours  of 
direct  sunlight.  As  the  season  advanced,  the  quantity  of  direct  sunlight 
received  became  greater,  until  finally  the  amount  received  was  judged 
to  be  too  great  and  the  intensity  of  the  direct  sunlight  was  decreased, 
first,  by  a  single  screen  of  cheesecloth  tacked  over  the  front  of  the  frames 
holding  the  culture  tubes.  Later,  a  second  piece  of  cheesecloth  was 
tacked  over  the  first  one.  The  general  effect  of  the  sunlight  on  the  cul- 
tures in  the  tubes  thus  exposed  was  to  check  the  growth  of  the  fungi, 
compared  to  similar  tubes  when  placed  in  very  weak,  diffused  light  or 
absolute  darkness.  The  sunlight  also  seems  to  intensify  the  colors  of 
the  aerial  mycelium  when  it  is  normally  other  than  white. 


Jan.  14, 191S     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      37 

TEMPERATURE   RECORDS 

During  the  entire  time  the  cultures  were  under  observation,  two 
thermographs  were  run  continuously.  One  was  placed  on  the  shelf 
with  the  fungi  exposed  to  sunlight.  A  soil  thermograph  was  used  to 
record  the  temperature  of  the  tubes  kept  in  the  dark. 

DEFINITION   OF  TERMS   USED 

It  was  foimd  early  in  the  study  of  the  cultural  characters  of  the  fungi 
under  consideration  that  a  set  of  descriptive  terms  especially  adapted 
to  the  cultures  of  fungi  grown  under  the  conditions  here  described  would 
have  to  be  used.  The  terms  employed  in  the  tables  in  this  paper  and 
in  the  body  of  the  text  are  those  usually  found  in  ordinary  botanical  litera- 
ture, but  they  have  been  modified  somewhat  to  fit  the  conditions  ob- 
taining for  fungus  growth.  The  fungus  growth  on  artificial  media  is 
divided  by  the  writers  into  two  general  classes,  aerial  and  submerged. 
The  aerial  mycelium  consists  of  that  which  is  on  or  above  the  surface  of 
the  agar;  the  submerged  mycelium  includes  all  that  is  beneath  the  surface 
of  the  agar. 

Great  difficulty  was  found  in  obtaining  appropriate  terms  which 
would  express  the  character  of  growth  of  the  aerial  mycelium.  In  de- 
scribing this  aerial  growth  terms  which  are  usually  used  in  describing 
the  pubescence  of  leaf  surfaces  have  been  employed.  In  other  words, 
the  surface  of  the  agar  is  considered  as  the  surface  of  a  leaf  and  the 
character  of  the  mycelium  growing  on  this  surface  is  discussed  in  terms 
of  leaf  pubescence  with  some  minor  modifications  made  necessary  by 
the  character  of  the  organism  under  discussion.  The  following  terms  as 
used  by  the  writers  require  special  definition,  since  they  depart  in  some 
instances  from  the  usually  accepted  definitions  of  these  terms: 

Appressed:  Mycelium  which  is  prostrate  on  ttie  surface  of  the  agar.  This  with 
•many  fungi  is  the  first  stage  in  the  aerial  growth  of  the  mycelium.  Later  this  appressed 
mycelium  may  give  place  to  other  forms. 

Cobwebby:  Long,  weak,  intertangled  hairs  which  are  not  thick  enough  to  be  either 
woolly  or  felty  and  are  not  short  enough  to  be  considered  as  downy. 

Cottony:  Erect,  rather  long  (3  to  5  mm.)  mycelium  spreading  in  all  directions. 

Downy:  Short,  fine  hairs,  loosely  scattered  over  the  surface  of  the  mycelium,  giving 
it  a  downy  appearance. 

Felty:  Matted  with  intertwined  hairs,  resembling  felt. 

Floccose :  Scattered  patches  of  short  mycelium. 

Plumose:  Tufts  of  mycelium  with  a  central  axis  from  which  short  hyphae  radiate. 

Silky:  Long  parallel  threads  of  mycelium,  more  or  less  prostrate,  like  combed  silk. 

Sodden:  Mycelium  having  a  water-soaked  appearance;  usually  such  myceliiun  is 
appressed. 

Subfelty:  A  thin  layer  of  mycelium  consisting  of  short  intertwined  hairs. 

Velvety:  Layer  of  mycelium  with  distinct,  dense,  straight,  short  hairs  like  pile  of 
velvet. 

Woolly:  A  dense  mass  of  mycelium  consisting  of  long,  tortuous,  matted  hairs. 
Cottony  and  woolly  may  both  later  become  felty  by  the  long  hairs  becoming  matted 
and  prostrate. 


38  Journal  of  Agricultural  Research  voi.  xii,  no.  2 

Hyphenated  compound  words,  like  "appressed-downy,"  "felty- 
woolly,"  indicate  a  condition  intermediate  between  the  two  names,  while 
"downy  to  appressed"  means  that  the  older  portions  are  downy,  while 
the  yomiger  portions  are  appressed. 

In  all  of  the  tubes  the  growth  of  the  fungus  is  both  lateral  and  longi- 
tudinal. Of  course,  the  lateral  extension  is  very  limited,  since  the  inside 
diameter  of  the  tubes  is  only  about  20  mm.,  while  the  length  varies  with 
the  length  of  the  agar  in  the  tube.  The  first  record  for  growth  shown  in 
all  the  tables  is  always  that  of  the  lateral  growth.  For  instance,  a  record 
showing  20  by  30  mm.  means  that  the  lateral  growth  was  20  mm.  and  the 
longitudinal  30  mm.  When  the  same  set  of  figures  are  repeated  for  two 
intervals  of  time,  like  60  days,  20  by  80  mm.,  and  80  days,  20  by  80  mm., 
this  indicates  that  the  growth  of  the  fungus  had  reached  the  bottom  of 
the  tube  at  the  first  record  given  and  would  therefore  be  the  same  for  the 
second  interval  of  time. 

IMPORTANT  DIFFERENTIAL  CRITERIA 

The  following  criteria  have  been  found  of  value  in  the  dififerentiation 
of  the  various  species:  (i)  Macroscopic  characters,  including  rapidity 
of  growth,  color  of  aerial  and  submerged  mycelium,  character  of  aerial 
mycelium  such  as  to  texture,  etc.,  staining  of  the  agar,  decoloration  of 
the  agar,  the  comparative  rate  of  growth  between  the  aerial  and  sub- 
merged mycelium,  especially  when  the  submerged  mycelium  is  colored 
and  markedly  in  advance  of  the  aerial;  (2)  microscopic  characters,  such 
as  septation,  branching,  size  and  color  of  hyphae,  clamp  connections, 
polymorphism  in  spore  formation,  etc.  A  few  of  the  species  of  Polypora- 
ceae  examined  by  the  writers  have  in  addition  to  the  usual  basidiospores 
other  spore  forms  variously  known  as  conidia,  oidia,  chlamydospores,  etc. 
These  various  nonbasidiosporic  forms  may  be  divided  into  two  general 
groups,  spores  which  are  borne  on  the  aerial  hyphae  and  the  so-called 
spores  which  are  borne  on  the  submerged  hyphae,  often  referred  to  as 
chlamydospores.  The  latter  have  been  found  by  the  writers  to  be  more 
widely  distributed  in  the  Polyporaceae  than  the  aerial  spores  and  their 
presence  and  characters  as  well  as  those  of  the  aerial  should  always  be 
noted,  since  they  are  of  great  diagnostic  value. 

It  will  be  noted  from  the  tables  that  certain  fungi  have  colorless  or 
white  aerial  mycelium  throughout  on  certain  agars,  while  others  have 
colored  depending  upon  the  agar  used.  Such  color  differences  are  very 
important,  since  they  are  usually  constant  for  a  given  species  on  a  given 
agar. 

In  some  species  of  the  fungi  examined  the  submerged  mycelium  in 
certain  media  is  constantly  colored,  while  in  other  species,  whether  the 
submerged  mycelium  is  colored  or  colorless,  seems  to  depend  upon  certain 
environmental  factors,  such  as  the  amount  of  moisture  present  in  the 
medium  or  the  acidity  or  alkalinity  of  the  medium. 


Jan.  14, 1918     Cultures  of  Wood-Rotting  Fungi  on  A  riificial  Media      39 

Some  of  the  most  important  criteria  for  distinguishing  different  but 
closely  related  fungi  are  found  in  the  first  10  or  15  days  of  the  growth  of 
the  subcultures,  such  as  rapidity  of  growth,  color  changes  in  the  myce- 
lium, staining  of  the  agar,  decoloration  of  media,  etc.  Important  char- 
acters which  are  sharply  defined  at  one  stage  of  growth  often  disappear 
or  are  obscured  by  the  later  mycelial  development;  and  for  this  reason 
the  cultural  data  in  the  tables  have  been  given  for  several  periods  of 
time  in  the  growth  of  the  cultures,  say  at  10,  20,  and  30  day  intervals. 

INFlyUENCE  OF   SUNUGHT  ON   CULTURAI,  CHARACTERS 

One  of  the  special  benefits  which  seems  to  be  derived  from  exposing 
cultures  to  the  sunlight  is  the  accentuating  of  the  color  characteristics 
and  toning  down  of  the  mycelial  growth  of  the  fungus,  thereby  making 
it  more  characteristic  and  uniform  for  a  given  species  than  when  placed 
under  similar  conditions  in  the  darkness. 

The  differentiation  of  the  characters  of  the  mycelium  produced,  both 
as  to  texture  and  color  of  the  aerial  mycelium,  is  very  much  more  marked 
when  the  cultures  are  grown  in  the  presence  of  light  at  ordinary  room 
temperatures  than  when  grown  in  incubators  at  the  optimum  and  con- 
stant temperature  for  the  mycelial  growth  of  the  fungus  under  considera- 
tion. This  probably  explains  why  no  one  up  to  the  present  time  has 
seriously  attempted  to  differentiate  the  various  species  of  wood-rotting 
fungi  by  means  of  cultural  characteristics  alone. 

Furthermore,  the  cultures  when  grown  in  darkness  and  at  a  more 
or  less  constant  and  high  temperature  overrun  very  rapidly  the  surface 
of  the  agar  in  the  tube,  thus  obscuring  the  real  growth  of  the  fungus  as 
observed  in  the  cultures  subject  to  daylight  conditions. 

GROWTH   OF   WOOD-ROTTING   FUNGI    ON    AGARS 

Texture. — In  the  growth  of  wood-rotting  fungi  on  agars  the  fungus 
as  it  spreads  from  the  inoculum  on  to  the  surface  of  the  slant  proper 
assumes  certain  well-defined  stages  in  its  growth,  which  may  be  roughly 
divided  into  two  general  divisions:  (i)  Fungi  whose  advancing  young 
mycelial  zone  is  appressed  and  (2)  fungi  whose  advancing  zone  is  downy, 
felty,  woolly,  etc.  There  is  but  little  real  difference  between  these 
two  methods  of  growth,  since  as  a  rule  the  character  of  the  mycelium 
first  to  appear  is  appressed.  If  the  true  aerial  mycelium,  in  contra- 
distinction to  that  which  is  strictly  prostrate  on  the  surface  of  the  agar 
keeps  pace  in  its  growth  with  the  appressed  mycelium,  the  zone  of  growth 
will  be  downy,  felty,  woolly,  etc.  If,  on  the  other  hand,  the  growth  of 
the  strictly  aerial  mycelium  is  much  retarded,  the  appressed  mycelium 
will  present  a  well-defined  zone  from  one  to  several  millimeters  across. 

The  appressed  mycelium  is  usually  either  colorless  or  both  colorless 
and  sodden,  and  from  this  the  true  aerial  mycelium  usually  develops. 
The  cottony  mycelium  as  a  rule  does  not  persist  in  this  condition  for 


40  Journal  of  Agricultural  Research  voi.  xii,  no.  2 

any  great  length  of  time.  The  long,  divergent,  aerial  strands  usually 
become  more  or  less  compact  and  finally  felty.  The  usual  steps  in  the 
growth  of  the  mycelium  of  a  fungus  are  first  appressed,  then  downy, 
then  felty,  woolly,  etc.  Cottony  mycelium  usually  develops  the  cottony 
stage  immediately  from  the  appressed  condition.  Many  of  the  fungi 
pass  so  rapidly  from  the  downy  to  the  felty  or  woolly  stage  that  it  is 
unnecessary  in  the  description  to  indicate  that  there  is  an  intermediate 
downy  stage. 

Colors. — The  colors  of  the  fungus  as  a  rule  follow  certain  definite 
changes.  Excluding  the  color  of  the  mycelium  on  the  inoculum,  the 
first  color  which  usually  appears  in  the  early  stages  of  the  fungus  will 
be  either  colorless  or  white,  depending  to  a  considerable  extent  on 
whether  the  young  mycelium  is  appressed  or  downy.  The  next  step  in 
the  color  changes  will  be  for  the  older  whitish  areas  to  become  light 
buff  \  warm  buff,  antimony  yellow,  etc.,  if  the  fungus  happens  to  belong 
to  some  of  the  brown  polyporaceae.  As  the  culture  ages,  the  color  of 
the  mycelium  on  the  older  areas  will  assume  a  deeper  and  deeper  tone 
until  finally  a  color  is  reached  beyond  which  no  appreciable  change  is 
observed.  In  the  large  majority  of  cases  the  color  of  the  older  mycelium 
constitutes  a  rather  extended  area  compared  to  that  of  the  5^ounger 
zone.  In  practically  every  instance  the  cultures  obtained  from  any 
given  fungus  on  at  least  several  of  the  culture  media  will  approach  very 
closely  the  color  of  the  sporophores  as  they  appear  in  nature.  For 
instance,  if  one  is  attempting  to  grow  cultures  of  Polyporus  dryophilus , 
Fomes  texanus,  or  other  brown  fungi,  one  would  expect  to  have  at  least 
several  of  the  culture  tubes  with  brown  mycelium  similar  to  that  of  the 
fungus. 

TABLES  SHOWING  CULTURAL  CHARACTERS 

In  this  preliminary  report  only  a  few  tables  are  given  out  of  a  large 
number  which  the  writers  have  complied  on  the  cultural  characters 
of  certain  species  of  wood-rotting  fungi.  The  following  species  having 
brown  sporophores  are  given  to  illustrate  the  close  resemblance  in  colors, 
texture,  etc.,  of  the  same  fungus  on  different  hosts:  Four  strains  of 
Fomes  texanus  (Tables  I-IV)  and  two  strains  of  Polyporus  farlowii 
(Tables  V-VI). 

'  The  colors  used  in  this  paper  are  according  to  the  following  standards: 

RiDGWAY,  Robert,    color  standards  and  color  nomenclature.    43  p.,  53  col.  pi.    Washington, 
D.C.  1912. 


Jan.  14. 1918     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media  '   41 


^  :- 


^  a 


O       in 


•^ 


"^ 


« 

a 

s 

a 

t.1 

s 

u 

1— 1 

3 

w 

0 

n 

J 

3 

n 

r/; 

< 

' — ' 

H 

>r,  0   "1  N    i/v 

g    ^  i~  ^  •*  to 

0  m  0  -^  0   , 

00    r^  -^  to 

d 

d 

>. 

^XXXXXXXXXX 

;§o  000000000 

I 

•0 

0 

d 

5 

.    -3  2    -S      : 

2 

Si 

5 

20inuii/l>o000"0 

•o 

d 

0        CJ2       -T 

.w^                           0 

?:  «*^i/^r-ofow  ■<r'^t^'o^ 

0 

2 

Q 

U^ 

0     -S  12    -a 

•3     Q 

ixxxxxxxxxx 

^0000000000>o 

■o 

_C3 

^^     -2 

s 

•0 

5 

S 

oS   pq 

5 

CD 

00 

s 

S3 

1        2 

"o 

a 

g"    <H    K-.CO    «    rj    0    <o  U1  1/100 

Igxxxxxxxxxx 

^oovr>./ioooooio>n 

1 

C4 

0 

a 

1 

0 

:  a       : 

S            i2 
•Q            3 

"   1 

pq       u 

•0 

Xi 

J3 

II     1     ti 

>> 

E 

> 

2 

■j3 
5 

d 

ca 
•0 

jj       :  vO       ■      -    M    to  rO  t     •    « 

^  -.x  :  ;xxxj^  ;x 

/Si 

a 

5 

c 

a 

d         : 

3 

1 

&      ■ 

0 

1 

"o 

"3 
0 

c) 

•0 

1 

3 

a 

V 

u 

a 
< 

1 
3 

a      : 
>.     : 

a    d 

g    -a 
^      : 

2       £ 

■■^5       5 
§        1       ': 

•0 

s8 

;  ;  I'd  ;T3  :  :  :  : 
:  :  :^  :^  :  :  :  : 
:  :  -.Z  -a  :  :  ■  ■■ 

;;;§:§;;;: 

.    .    .  c3    •  a    . 

:  :  :ot30  :  :  :  : 

>>  :  :  >.a  >,>,  :  :  : 

r 

r 

0      •      ■  «   &  (U   CI      •     ■     • 

a 

■a 

w 

>   :   :><>>  : 

ui 

tn     •        cX 

" 

0 

d 
•0 

3                ; 

1 

1 

"o 

■35     "3 

"o 

0 

i 

"0 

00 

;   :   ••d 

:  :  :  0. 
■  :  ■  0. 

'■d    •    ■    •    • 
:  (u  :  :  :  ; 

.  i,   :  :  .   . 

•  0.  ■  ■  •  • 

•  &  :   :  :  : 

l5 

0 

1  i 

•0 
0  ' 

d  0 
0 

2? 

did 

^3 
0  C 

a 
^      : 

0 

tn-Oii 

q      *J  0 
■w       31., 

a         • 
^       1 

e 
5 

c 

X 

1 

a 

.     .     .  C3     •  0!     .     .     .     . 

:  :  :°-do  :  :  :  : 

~°     1     3  3  OJ4   „ 

E* 

'^.c 

15  «  0  »     ja*" 

3       -s       « 

■3    ■    ■'S 
>   :   :>« 

2.i;'C  •  •  • 
<>>  :  :  : 

-'i 

m 

0 

<)    u<    u    - 

I)      0      t/ 

0 

0^ 

OJ. 

■4J    0) 

5 

1 

3  3  III  ^^ 

a 
3 

ni 
00 

0    "^ 

0  >. 

3 

S  . 

"o 
0 
0 

-U           -4. 

3      : 

q    .  c 

1    • 
1 

01 

•0 

...   4, 
.      .      .   u 

...  a 
:  :  :  c 

■•d  :  :  :  : 

:g  :  :  :  : 

■  a  :  :  :  : 

a 

51 

< 

ll§ 

^-6 
•Si 

a 

.    .    .  ta 

•  •  ■5"' 
>   :  :>■ 

•  a   ...   . 

g2  ;  ;  ;  ; 

^  HJ   V     •     •     ' 

<>>  ;  :  : 

a 

>. 

~F 

1 

0     p 

0    g-S 

•       9       ' 

3  a 

i    0 

"o 

M 

ijj 

"3 

t- 

a  .Ss" 

."    ^^ 

^       "^    t 
0       00: 

a        "S  a     ( 

a             in  0 

2a 

■3 
0 

3 

J3 

8 
a 

n 

-  .e-s 

a  O-Soj 
•f  >'3-S  c 

-  3^32 

3  0*11 

■•d  ;  :  :  : 

0 

< 

c 

)         W     OM     0 

55      rt      c 

3 

•0 

.  i_  .  .  .  . 
.  a  .  •   •   • 
•  a  :  :  :  : 

:  ^ 

d 

:_o 

^ 

•«.... 

^ 

•"3 
.  0 

0 

J2 

^ 

:>•:'. 

?'2  :  ; ; ; 
s >>>>  •  ■  >> 

•0 

3  ■ 

•:  I 

•0     >§ 

a 
0 

J5 

:       ^ 

a 
3 

2 

■  V    ■    ■ 

:_>   : 

^ 

ta 

4-* 

3 

; 

3 

:>  :  i 

<5>>  :  ;> 

^           ' 

^ 

;   Qpi   ps 

m 

H 

4 

tit 

bt 

w 

a 
3 
5 

i 

n 
0 

P. 

e 

q 

5   f 

1< 

i    ■ 

<      p 

I 

M 
t 

IS, 

:«:::: 

pquP^  <PuPi 

42 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  2 


1^   « 


fv. 

'~* 

oo 

H 

•d 

<N) 

a. 

•a 

-  i 

§    S 
s     3 

^    3 
g     ^ 

I"! 

s    is 

S       bo 
S5       nl 


^    .= 


pq 

5h 


Wo 


S     -a 


3-0 


m    Pi      5 


3.^ 


o  H 

pq:s 


!SS 


O  oj      --*        O 


rt    o 


■SI  ill 


j2  a 

iH     ""      O     t^ 

<    o< 


a  o 


^M 


a  o 

Is*- 


a  3 

B)    "    &    0 

<        u 


^  a  .  a 


C  t-  i* 

ni  V-  >> 
*J.D  a 

on 


it 


^s^ 


2^o 


Sogg-s^ 


^5 

(JO 


^  S        O       1-1 


•5  s 


aS 


o    mo 


■a  2 


_.  o  w  in  N  o  mo  o  O  m 

■Ixxxxxxxxxx 

BjOOOOOOOOON 


;   O   O   O   "o  O   "1  O   "1  O   M 

JXXXXXXXXXX 

'loOOOOOOOOOO 


-icOWWOOOOOONO 

;2xxxxxxxxxx 

~<   U-,  o   >noO   O  00   O  O  00   O 


xxxx 


•g  o  o  o  aj  tl  t)  °  o  o 
•^    •    •    •  o.'oJ'qJ 

>  :   :  ■■<» 


ti  °  °  °  Jitit!  «  wjj 

*S     ■    •    ■  Q.'oJ'S     -    -    - 
>    ■■     ■    ■<» 


•do 


■SooofJ-ScJooo 


<J» 


•y  O  O  1>  o  \„   ^ 


0Sffl0K0fl,.^|l<fL, 


Jan.  14. 1918     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      43 


s   ^ 
II 

S    3 

o     — . 


K    8 

i-t     ^ 

< 


Q    "I 


n    S 


5  a    .J.     0  >. 

go     §^£§ 

■^  P  3  ^  S     fe  « 
^  3  o-g  o.^     =  & 


Pi    (3 


«  >  S 

V  a  SiJ 
O     O 


o<J 


5S& 

So 

■w  o  b 


^ 


u<    o    w 


<    m 


<1        UhI    o    w 


«    o 


^  s 


2        J     -J 


^ 


On-t 


*i  a 

E    -^ 


a 


I    si 


■a 


ran  Qj  Olol-''^  MO 


;XXXXXXXXXX 

=  0000000000 


CO    loO    «    lAOoO    O    i-^.O 
.    PO^N    TON    •*rO'^'*)M 

SXXXXXXXXXX 

^jCOOOO'^OOOOOO 


XXXXXXX 


tloooii-sjooo-tj 


<5> 


,01-*-' 
■C^.2  o  i;  ai  t;  °  o 


•a  o 


a"" 


.   '■  6. , 

i:S-a 


S  a  it's  (u  Q  CSa  c«  o 


44 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  a 


-^     2 


v8 

S 

H 

«5 

a. 

A 
1 

k 

a 

g 

w 

en 

s 

■*i- 

m 

•>  -a 


1 

■** 

^ 

s 

a 

e 

*»^ 


^  1 

Si        o 


Y 


>     -3 


n 

< 


^  J' =■ 


•  3  ■*      JS 


S3 


o  a 


eg 


•aM 


•2^ 

r 


a^  a 


-*^   •  Sj=  9  2 

(u  g  a  S      43  »j 
H  P  en  a;  .^  .-^ 


Pw       og       M.S  C  SS  1=^  en  S.«„ 


3^>.j 


CJ 


^  i!)  3       3.fl-Q  3 
3  OJ3— '.a-~  3X1 


Si 

s 

o 

O 

3 

n 


a  2 
feja 

2gta 


O  ^  - 

oja  o  „  _ 


g|E2 


5      P  o 


p-a 


B  "^  'S  -^  _,  y 


fc     -s 


» S     rt 


2      n   omo   fw    <:p4 


i^5RvS5 


oooooooooo 


SOOO'Nio>/iO»^"Ooo 

^gxxxxxxxxxx 

^oooo>oooooo 


grtu^OOOOOnOv) 

:2xxxxxxxxxx 


'S'S 


o.  a  (u 
«> 


•SjoddSSJ-gd  p_^ 


"S  •  •  •  o,  a'oJ  ■  ■  - 
>  :  ••  ■■«> 


rt  «  S'S  8  p  Sag  5  o 


Jan.  14. 1918 


Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      45 


s 
•«i. 


^  ^    i 

S   .   .  <"£  Sn   . 

V  6  o  ax)  u  o 

•3  *jt«  o 

O         WO 


6  6  6 

PQO 


•  n  w   ■  • 
5  :  :  «5  K  : 

i  o  o  0"^^  o 
>   ■   -w    o   • 


■5    a 


^    .2 


e      3 


.ir>0>nOQOOOOv> 

iOO    OvCO    00000    Q\  O^  0\  t^ 

ixxxxxxxxxx 
•0000000000 


:  00  acO  0>  t^OO  CC  CT»  Ov  t^ 

:xxxxxxxxxx 

"ooooaooooo 


~ioOOOOOOO>"0 

^xxxxxxxxxx 

^0000000000 


pS« 


:^n 


c  c  c 
So  £ 

J3  OX! 

c  &  c 


It 


:-°  9 


ic 


O    iJ    4J      i- 


aj2  1- 


;X  :xx 


•  *  •  fi  ^  ■  >>  ^  * 

:  ".  :  &  c  :  ^  c  : 

-  •  ■  o  <y  -^2  <u  ■ 

;  *.  : TD "O  'WO 


'o'o'o'o  ^''o^'o'oS 

OOOO^SO-jO^O-g 


•  >,  ■ 
:xi  : 

•X3     • 
.   c*      • 

;  &  : 

•J3     ■ 

■   O   >. 

'■  0*3 


a  o 

Tl    '^ 


>.8^ 


^^c^S  ti  o  S  lu  S 
•o-o2o&>-a.2£S 

^^u^O  :oOo 


■  <o 


s  0^2-52 


46 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  a 


*<. 


U 


<        - 


4'  i_ 


^    a 


<> 


2        -2 


>      5        -3 


i 

vn         0          Q    m'r, 
t^        00         00    t^  t^ 

3         w  m        >n        0        1 

J,          NO     I^          00            t^ 

SX     X     XXX     X     XX     X     X     1 

g 

:§s 

0 

>  0  0 

3         0    0         0         0 

3 

31A10       0        mio       10       0       1 

CS 

>> 

^X     X     XXX     X     XX     X     X 

*o 

•a 
•0 

CI 

•§  8 

D 

DOO          0          00          0          0 

w 

m 

.    0        00 

^00    0          > 

0     :,.    0     .0 

>> 

g     N            M 

-<               •     N             «            H 

« 

4;X     X     XXX     X       :x     X     X 

•a 

^8 

> 

3    11  0 

n          .  VI        10        1^1 

■* 

^""     »      :«     M     «    1 

J 

;  i'O 

:      :  >> 

Ji 

S 

.     .  lU 

>, 

"O 

•  ■  s  . 

0 

•0 

•3     ^    •o-aara-3     >>>•     >>    "o 

0    ^       ■   ■  *      ■ 

.      .   U.  ^    0        ^^  T*        •" 

n 
.3 

>    ^ 

; 

•    •  ^  tn  ^y       0)  a»       oj 

(U 

^Si-a     S 

:  >> 

a 

•c 

1 

•T3 

a 

'■      c 

+■ 

3  c  ij     c 

;     :2 

>. 

rt 
^ 

NO 

f  ^ 

3  0  a  ^  c 

^^    ^    1 

^ 

r' 

:^<5    P 

.     fefe      p:,      ^ 

3 

1 

0 

•  i-o 

•  0 

H 

en 

>. 

:  ■  01 

:  >» 

^    s 

•>     '.  V      ■   ^ 

,    :«  •>; 

•0 

°  >>t 

i 

>-§ag| 

:2>>a     0    1 

Tf 

z^l 

c 

::::a2    -q    1 

O'tn  4 

c 

'  :^"^ 

■  o'«-g 

0      > 

C 

:<i)    c 

;U       0 

a 

>. 

S 

rt 

V3 

•C 

0     c 

c 

00     c 

00         0         0 

V, 

0 

"o     •= 

•c 

"o/q    t: 

»oxi    -a    -a    1 

^ 

tS 

a 

"9 

ui 

t/i 

o 

t-. 

>> 

01 

■a 

0 
"S      c 

c 

00     0 

00     0 

c 

1 

NO 

•3   •= 

•c 

•o-q    "c 

•a -a     -c     -u     1 

a 

3 

in 

O 

tn 

S 

b 

>. 

*— < 

_o 

•a 

0    0 

c 

00     0 

000     0     1 

"c 

a  " 

•c 

•W      "C 

•o-o     "0      -o      1 

(J 

^ 

a 

:  :     fl 

:  :     fl 

u 

.  .     f-i 

.    .       u 

0 

:  ■     0 

.  :  ;    ,^  .^' 

>. 

J3    . 

•  :    J 

rl 

+j  c3 

•  jn       ,j_, 

c  :  :    -^  S43 

T3 

^&  0 

0 

d.^     ^ 

^06     •«  S  e 
o-o.y      <-<  ^  S 

vg 

OgT3 

•a 

•Og       0 

3.a 

'  '3      "^ 

.D   -''^     ai^ci 

g 

m 

:t3    « 

:^     pq     ?: 

.2 

"v 

a 

■  •     « 

•  •     a 

a 

>> 

.    .       ;-< 

.  .     1- 

>. 

0 

•    •       0 

'■    '■      M 

.  : ;  -  .?■ 

"rt 

CO 

-4-1    Q      ' 

6M    -^ 

a   :  •    -^^  C.C 

'O 

^  &  d 

0 

e  0  (u    -^  ?  B 

« 

NO 

^s-o 

•0 

•a.  s     " 

0  73  ,+^        0  C  g 

■3 

3X!     • 

pq 

:c3    « 

5     -t?        ^.C^ 

, 

•  •    J.       ■ 

U 

"3 

•    •     "3 

:  :     >. 

CO 

•0 

1     « 

g  1 
1  ^ 

1 

e-S    a 

^  0    g 

?^3  ^ 

:  :     >>    d 

•  :     g     3 

On 
1      ll^ 

B 

_3 

^ 

:    !     "a 

•5 

1 

•4-) 

« 

CS3       CO       0 

U 

s 

n 

u«    0 

PL,<       fL, 

(k 

1 

Jan.  14. 1918     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      47 

DISCUSSION    OF   CUIvTURAI^  TABIvES 

Fames  texanus.  The  cultural  characters  for  all  of  the  four  strains 
show  remarkable  uniformity  for  each  of  the  corresponding  agars  in  the 
series  of  ten.  This  fungus  seems  to  be  more  susceptible  to  differences 
in  the  amount  of  sunlight  it  received  than  any  species  thus  far  inves- 
tigated. When  the  intensity  of  the  sunlight  was  diminished  by  the 
cheesecloth  screens  previously  mentioned,  this  species  on  several  of  the 
agars  immediately  responded  to  the  decrease  in  Ught  by  making  a  more 
vigorous  growth  and  turning  a  lighter  shade  of  brown. 

Polyporus  jarlowii.  This  fungus  is  also  very  uniform  in  growth  on 
each  of  the  10  agars  as  the  tables  show.  When  grown  in  the  dark,  the 
strain  from  Acer  negundo  developed  a  submerged  mycelium  on  prune 
and  corn-meal  agars  which  was  Mars  brown  in  place  of  colorless.  Whether 
this  change  in  colors  of  the  mycelium  was  due  to  the  darkness  is  doubtful 
since  the  other  strain  from  Populus  iialica  when  grown  in  darkness  still 
retained  its  colorless  mycelium  the  same  as  when  grown  in  the  light. 

Series  were  grown  of  nine  strains  of  Trametes  pini  obtained  from 
material  collected  in  four  States  and  growing  on  seven  hosts — ^viz,  Pinus 
echinata,  P.  fiexilis,  P.  ponderosa,  Picea  engelmannii,  Pseudotstiga  taxi- 
folia,  Abies  arizonica,  and  A.  lasiocarpa.  This  series  represented  sub- 
cultures ranging  from  i  to  7.  While  there  was  some  slight  variation, 
especially  in  the  colors  of  the  mycelium  of  the  various  strains,  the  dif- 
ferences were  not  so  marked  as  to  constitute  real  specific  characters. 
There  was  practically  no  difference  between  the  cultural  characters 
obtained  from  the  different  subcultures.  This  would  indicate  that  the 
general  fundamental  characters  of  the  fungus  are  not  materially  changed 
through  successive  subcultures,  at  least  in  this  instance  to  the  seventh 
subculture. 

As  will  be  noted  from  the  temperature  record,  there  was  rather  a  wide 
variation  in  temperature  during  the  time  the  various  series  of  cultures 
were  growing.  Nevertheless  the  general  cultural  characters  as  shown  in 
tables  are  practically  identical.  This  identity  of  characters  for  a  given 
organism  on  a  given  culture  medium  is  still  more  marked  when  a  series 
of  from  three  to  six  tubes  of  the  same  agar  for  the  same  strain  is  made 
at  the  same  time  and  then  compared  as  the  growth  progresses. 

CULTURAL  CHARACTERS  FOR  DIFFERENT  STRAINS  OF  SAME  FUNGUS 

The  sources  of  the  initial  cultures  from  the  nine  strains  of  Trametes 
pini  were  five  from  sporophore  tissue  and  four  from  infected  wood,  rep- 
resenting seven  hosts,  while  the  cultures  from  Fomes  texanus  all  were 
from  tissue  but  from  two  hosts.  The  Polyporus  farlowii  cultures  were 
both  from  tissue  but  from  different  hosts.  A  careful  comparison  of  the 
cultural  characters  of  the  various  strains  of  each  of  these  fungi  shows 
no  appreciable  differences  between  cultures  of  a  given  fungus  whether 
obtained  from  infected  wood  or  from  sporophores;  neither  do  the  hosts 


48  Journal  of  Agricultural  Research  voi.  xii.  no.  2 

of  the  fungus  seem  to  make  any  marked  changes  in  the  fundamental 
cultural  characters,  as  is  clearly  shown  in  the  various  tables  when  strains 
from  different  hosts  are  compared.  There  may  be  minor  differences 
due  to  the  host  from  which  the  strain  came,  but  nothing  more. 

The  comparison  of  the  cultural  characters  of  many  species  of  parasitic 
fungi  has  long  been  recognized  as  a  reliable  index  to  the  identity  of  the 
fungus  under  investigation,  and  there  is  no  reason  why  the  cultural 
characters  of  wood-rotting  fungi  which  are  just  as  uniform  and  depend- 
able should  not  be  used  for  identification  purposes. 

The  writers  have  purposely  avoided  going  into  a  discussion  of  the 
results  of  the  use  of  synthetic  agars  and  of  other  special  media,  as  these 
will  be  taken  up  in  a  later  article.  They  have  presented  here  the  results 
obtained  from  agars  easily  made  and  apparently  of  a  uniform  enough 
composition  for  similar  cultural  characters  to  appear  on  different  batches 
of  the  same  agar  even  when  the  acidity,  alkalinity,  and  water  content 
vary  considerably.  Just  how  great  a  variation  in  these  factors  must 
occur  to  produce  a  decided  change  in  the  cultural  characters  is  a  problem 
for  future  investigation. 

EXAMPLES  OF  THE  DIAGNOSTIC  VALUE  OF  CULTURAL  CHARACTERS 
IN  SPECIES  DETERMINATION 

In  comparing  the  cultural  characters  of  closely  related  but  really  dis- 
tinct species  marked  and  constant  differences  in  the  character  of  the 
mycelium  will  be  found  on  certain  corresponding  agars  in  the  series  of 
cultures  representing  the  two  species,  while  if  the  two  fungi  are  really 
the  same  species  no  constant  differences  of  specific  rank  will  be  found. 
The  following  fungi  will  illustrate  the  diagnostic  value  of  the  cultural 
characters  in  determining  the  real  position  of  the  species. 


Jan.  14, 1918     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      49 


t-. 


00        C8 
C\J       g 


^   "5 


s 


^  a 


u  a 


Is 


ils     8.2     sis 

^   ai^   ai^   a 

*T3.2  iJf-d.S  "-0.2 

^"sl^a-i^sa- 


at      ^ 


u 


s  6 


0*^ 

To 


u    ^ 


•la 

a 


:-a  E  °  "s  a 

3  1    C  3   I    "2 

^  2.  ^1j  2  bs  i* 
.  o  ■*  >>  o  "  "^  'J 

acn-°a«-^a^ 

SI'S  n  c«  o  »  "o 

y       o       o 


i-i     O     S     U 


•?       a-  -rfl.S-r 


U     < 


X    j2  ">  a 


^  n       a 


•o    'd    'O    -d 


o   •      o  !5      O.J, 


.2    J3 

8     S 


fl-H      H.S 

Sg   eg 


ca  O  3 


'3& 


>.  o 


u 

a  o 


a     ° 


o  .^ 


o   •  *-  u  >  -,  d 
5^.cSJ2.Q>?a 


IS 


.2f  a 


rt  O  3 

u    n 


as 


cc  u  a 


fi  a 


J.  O     .       _!, 


o    '"Ma 

rtXl  aJ3i3.Sf.2 

0    <       >-i^ 


1  1 


-I 
i4 


O        CBi'm>--^j3 
«     .  O  C   >■  (U   O  r;  J3     . 

>>"'o  o-yo  o  o  in  !; 
rt03-wc«03-S.?0 

U      «      **      "^      "^ 


;^    « 


■•r    ^ 


.2     0 

O    ."o     • 


"8 

1 


e    8 


1-30 

■>-'    j5  ^  Bii 

do-     .2f      3  -J  *  o 

o      :    a    pq    ^ 


(3   ;^      n 


•a  2 


(1<      <!      PM      P^ 


27805°— 18- 


Jxxxxxxxxxx 

.^0000000000 


.OOw.  mOOOO*^  iy>vO 

.:;xxxxxxxxxx 

»e^OOOOOOOOOO 


.00    O   O   vnti^^oo    «   »^to 

ixxxxxxxxxx 

;0000000000 


•a     ;-0     ; 

<u  M  <u  S! 

i-o 

ao  aS 

Sg-rt  a 

: « 

13 

.2"  "'.2'  nl 

•  c  ^ 

S? 

•i|'-32&«ga 

O  g  O  (U-C  ^  cs  o 


ir  1-1  5«     •'^ 

*    nl    £•   OJ    $•      :    2irrH    ^ 

■XI  o  a  oj'o^  «  o 


Q  C.  O.  t  -.  u 

>.C  ^■d  oS  g'g^ 
fl-s  d"S  ♦J  o  ^  «i  a 

n  o  0.  ■*-'  ^  c3  o 
•i!  >>5  >.'>,S'S  °  >> 

•31;  o^  o  03^1;  o 


•§•§£■§§1  :?^| 


t*::  <uii  !»  "  3.iS  b  ■« 
OSWy«U(l<<fr<P< 


50 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  a 


I        § 

CI,  M 

o 
a  .9 
s      « 


•S  ^ 

^  1 

<  e 

.1  3 


">^ 


m 

< 


o      a  o 


r 


U        Q        U 


as 


9    .  in '^3  g    .  w 

">—   !?  ??  -^  <" -^   f^ 

So      So 


c;  >,  -^  (U  >,      a;  >.    . 
!     I-     Is    Is 


"3  d 


"  o  o  S  «  >-  o  o 
rt  0-3  a  a  ca  o-S 

So       So 


_o        -*  a  S "         ** 


o    < 


.£f  a 


— T-S  M  _  a 

Isssia 
p  «  p 


;-a    is       iJ 


-     2*^     as    ^"5 
o    <       n      IS 


3    -^   i 


fi   'S 


-■9 


.^      ^      4-»  a» 
-  "o       "      •S.'C 

5        «     hJ 


■s 


fiii    *§ 


>.a'3  o 


^-  ^' 


3l  J 
.2*9 
i4 


^    «    S        pq    a 


tt^ 


ca 


a^  o 

2  btTn 


btTo  ■    a 
s  Sin     '^ 


a=3ig 
t;  o  >.  S 
a-w  «ja 
O     O 


o    S 


^    o    a 


a    -!. 


oS  3 


3  -"^.Sf  o 


Jan.  14, 1918     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      51 


10  I- 

r>  r» 

0 

invO 

« 

0 « -^  1 

ij   i>*  r^  fo  (>■  m  10        r>-vO   r*  fo 

.fxxxxxx    xxxx 

p^oooooo      0000 

>, 

CD 

•0 

0 

-r 

.  1/1  1000  000        "1  0  t^  n 

gr^ot^r^i^io      o^or*^ 

sgxxxxxx    xxxx 

i^oooooo       0000 

i 

T3 

u 

0 

a 

to 

"S 

.00   1^.  ftO"lM         >ni/iioO 

N 

S 

a 

i XXX XXX     XXXX 
^000000      0000 

•0 
0 

'nmSSmm     SmHS 

ixxxxxx    xxxx 

,^OODOi/-.00         0«i/li« 

& 

•c 

0 

i'S^i'S  >"     ': 

is^sa    ; 

•0 

U    ?)    V,— 

u      .      . 

:  aSi  Gen 

iti    .    ; 

•  a  a  c.  a  ^-   • 

•  c3  a  M  «^   • 

Qj    ;    • 

(0 

:>.«>.-«  : 

a  :  • 

>> 

tnaiul 
ainly 
mainl 
ty   to 
ppres 
3wny 
toap 

felty. 

•0 

0 

•t 

ytoi 
tom 
ytoi 
y-fel 
ny  a 
tod< 
felty 

y  to 

own 
elty 
own 
own 
dow 
elty 
Iky- 
elty 
own 

fc 

IPtoOQ     PhW^Q     1 

•■0    ••^  >, 

.  <u-a  (J ^ 

■i^^a 

•0  ; 

■      t-l     y^      ^4'-^ 

0 

:  a  u  an 

\   -J. 

:&g^a^ 

l--\ 

in 

:>.«>.-,« 

a  •  w 

^ 

>.'a  >.'5  S  0  >.  -r  >v  5    1 
a  .2 —  .9      "rc^ea 

s 

•0 
0 

to  dowi 
y  to  ma 
to  main 
y  toma 
y-felty 
■ny  app 
to  dowi 
felty  to 
to  dowi 
.y  to  ap 

3 
1 

a 

"a 
1 

fefaCfeOQ     toiofefi 

•d'St 

*Vi 

0  s  « 

0 

"  S^ 

•d  :  • 

v 

u  .  : 

^ 

l-o&g.S 

?!  :-d 

1 

i 

ny..  . 
pressc 
ily  ap 
inly  a 
to  ma 

owny 
appn 
•essed 
presse 

Oj 

•a 
0 

to  dowi 
y  toap 
to  mair 
y  to  ma 
y-felty 
scd. 
Ity  to  d 
felty  to 
to  appi 
,y  to  ap 

elty, 
elty 
own 
city 
own 
own 
pres 
Jbfe! 
Iky- 
elty 
own 

p:,{i,G'^aQ     wjjfaQ 

L 

a 

>!>•'■.'■.■ 

1 

•S     1  :  ■••a' 

1 
1 

>. 

i  l^;i 

M 

& 

2    2:^  :& 

•0 
0 

elty 

elty 
to  si 
[elty 
oap 

>»  >>  y)   >.  .^ 

•v-2-v>.T*' 

4-»+J{/i^-»>l?^W^♦.•,^?^ 

ou  im  a  53  ^  go  ^  g 

k. 

to 

M 

i 

•3 

3 

5 

*j  4- 

M  ( 

'■■I 

a   , 
s 

0 

2s3  «  0 

SP 

il 

>p: 

.S 

p 

Lc<pHfl4 

52 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  2 


k 


">^ 


"*i. 


^ 


^     3 

S      .2 

§     '3 


a  I 


HI 


2     °i  .5?i  . 

n  m      <u  ^      a/  u 

t-cT!  a;    ■  H  V    ■  n  v 
^.2'S  B'rt"  B'a'S 


S      Ri    -gi 


2cj    &8    SI 


8   I   |-S   8 

<2    .Oo<n'2:30    ."' 

8sa-oS§8£3!3. 
Sua      S    u 


o  P  fl 

e2a 

i-Sla 


o  '§  o 

1-  ■  n  v>  • 

a.a  ^S  2.2 

ss  'Ss  ss  . 

5o  .S-g  tS-3  6 

""  -a^  ss 

3  « 


43  T) 


"  O 


^   g   ! 

2    5° 
o    -a   .3 

^j2  a  ^.sfo 


•as 

n 


^    ^' 


:2  s   ^8.2 
3      « 


+3  o 

o  o 


O       J, 

*^      o 
u 

•^^^^ 
g2|s 


I'll 

Is  5 

2   MO     • 

■5  o  ^-S 

flii  rtXI 

<    u 


2,       *§ 


(3.0 


1^      u      13 


J3 

0 

0 

0 

^;s^ 

>, 

■S  >>3 

cn 

u 

0 

2    -w    J 
53     *     <i> 

(J  ;s  « 


^  o 


^        a     ^ 


(!<      «< 


Jan.  14.  i9i8     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      53 


;xxxxxxxxxx 

JOOOOOOOOOO 


•xxxxxxxxxx 

;oooooooooo 


.r«g«OOt/iOO>n<n 
jXXXXXXXXXX 

;0000000000 


Jxxxxxxxxxx 

1^  o  o  "oo  >o<»  0000 


rtS^-a  -3! 
>>S  92pS 

.  3  aHii  o  a 
'  d  cQ  o  <i>  "^  cQ 


>>'0  ' 


o  rt  a  H  *i  a 

"O  d  cJ  o  m  rt 


■•a  >.; 


>''a  ?!  !;  o  0.  M  p  >? 
oHa^'iioao" 

ot"  o  .-."i  ^o  -^f 


:  O- 
:  a 


"a  o  Qa  00    -30 


„   .   .  ^     •  •  oj  ^  '5  2 
E-aJfirt  E  I  «  t5 


54 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  2 


^      T^ 


2     - 


"^  s. 


< 


s     z. 


S    Z 


i 

-e 

ri 

a 

S 

N 

^ 

a 

n 

U 

11 

^   1 


■"tj  _;  ot3   •  >n-a 


•5         Si  o 


iS     »  i;     «  S     w  t/  >.  o 
">-  a". 2"  c.S"^  B.2  '>.•«  4) 

u      u      o      u    o 


00  oj  .  a 

J,    n 

sea      w 

•2,.sa|g 


a!2 


CJ 


o  0 


"n3foaf'ci"<C^ 

01  =i     D  i^'      (u  !J  °  o  o 
«^  gmXl  ?»!£  g'o  '^  " 

4!  .5  s  .5  S3  .£  u^2 

o  n  «  o  H  CS  o  R  CJ.2  Ho 

(J        O        U        U     O 


S^a 

|og| 
.2  5^  a  .2 
o        o 


'S 


m-o  a  o-a  a  "^-a 
7agiag7c 

M>,o.  i-'>>a)i-">. 

T  fl.a'n  a.S~  a. 
-ia^^.i.ai'— <a 
CJ       o       u 


V   01 

o  o 

u  u 

1^ 

a-s 
o 


m"^  S 


-0.2.2 


^  *i  ^  o 


O  J,       o 


-S    2 
3S     3 


r5^.s-a 


d^H^     — 


^5 


O        O       -O  ™ 

o      u   .  o  o 

8  •    -S5  o 

«^«l32 

o    o    m 


ll 

S2 


o  a     o  rt     o  o 

go    toS    m  w 

gja     30     s  .  3 

n  ^^  2  M  2        M-3  M^ 

...  o  —  - 


o 

8 

o     o 


O       B 


i.nj3.-  a 


►4         h-t 


bo's         60  O 

.-<  o     —  to 
h4 


S     u     u   •     t-*J3     ia 
o     o     ^  li!     o         3 


0   o    m 


'•3   ^ 

j^ai^s 


3  3 

■4J-S  *J   1 

jao  ja;; 

Mo  M ; 


& 


^    S5 

J-  ti^.- 


^  3 

«  s  n  =»  a 


U     h4 


^-ti 


J3  5      *. 

^4"   '3 


3  .2 
a;j=  o  "in 


O     CU     <J 


Jan.  14,  i9i8     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      55 


^ 

aXXX     XXXXXXX 
sooo      0000000 

CS 

"0 
0 

^««<,       n«c««c« 

* 

000         Oln'nOu-.Oul 

i 

.   r^O    ^       «^po»^*^»^t^(*l 

SXXX    xxxxxx 

i 

•0 
0 

itMOoo      0000000 

"3 

.. 

>n  0  f 

0  0  w  0  «  0  0 

1 

i 

,0»ot*5         loro-*On^<*5 

gXXX     XXXXXXX 

•a 
0 

v^ooo      0000000 

wo  C 

000     >/^CC     ■/!    0 

S5 

«5    Ci     «     M     «     «     N 

^ 

SXXX     XXXXXXX     1 

•0 
0 

M 

^88« 

0    "noo    0000 

0. 

i 

c 

1 

> 

1 

■*  1- 

* 

a 

0  4|  a  a 

0 

c 

a 

0  a  0  " 

1 

c 

owny 

owny  to 
Ho 

c 

■a 

nbfelty  t 
elty  to  a 
elty  to  d 
owny  to 

m 

C 

Q 

codtP^P 

6. 

a 

•?     •= 

•c 

_c 

1 

>.T3 

^   1 

i 

>.£ 

>  cn  pt  ki 

eg 

c 

0. 

0  ji  a  a 

•0 

&  , 

a 

•0  0.&  0, 

a 

0 

0  = 

en 

0  a  0  * 

3 

>. 

a 

fO 

> 

"33 

elty  to  d 
owny     1 
pressed, 
owny  to 
do. ... 

c 

abfelty  t 
elty  to  a 
elty  to  d 
owny  to 

fetoC 

P 

c/^bfaP 

6. 

cd 

0 

01 

Z 

>" 

3 

•a 

•B      1 

s 

7 

01 

>.T3  >.S 

in 

*? 

s 

C  S   C  M 

H 

>. 

>.£ 

£ 

&  s  s  g 

a 

a 

a 

0  J*  0  a. 

"d 

&  , 

a 

■a  Q-o  a 

0 

o.£ 

a 

0  a  0  " 

•d 
0  .. 

•dS 

*J    >>  Ul    >.    _ 

*j  *5  *j  :>. 

•53^  0 

0 
•0 

Ilbfel 
elty 
Ilbfel 
own 

fefeS 

P 

wfewP 

•2' 

en 

&i 

•d 

•d 

•0 

>>s 

^ 

V 

>> 

=S  rt 

0 

S 

> 

S 

•^  0 

k« 

^ 

•3 

2  >. 

a 
a 

a 
a 

a 
a 

0 

+J 

a 

B 

0 

et 

•^ 

.&~7 

0 

0 

S-0  0 

*i-ii    Oi 

tr>. 

^  >.r 

>>>.>._ 

&&  0 

.  ftmu  1 

3  0  0 

li  000 

•3  a  0 
W<!P 

wQO 

ppp 

3 

:« 

* 

*» 

• 

I 

0 
fc 

a 

■4-1 

Si 

2 

t 

^ 

2 

?3 

M   0 

0 

S 

n 

c 

n 

U 

(^ 

< 

Oi 

!k 

56 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  3 


g 


">-^ 


\    I 
^    1 


< 


O  I-. 
.10!  i 


I     M     .    O 


a^  "3 


O        U    (J    u        So 


o  -^ja 

"3    >■!- 

u  a  <u 
t.^2^ 


^°      o 

2  8     " 

.•la  l2s.-s| 

H-t      n      i-t 


S-3 

•a  3 


a 
Sta 


o    XI  ^  o;S 
u  </;  S  ot^r 

o    ^     eq 


I    O    o 


S    ri 


S  ?:  «  o 


g  s 


tn-S 


P  p 


s-a 
0.3 

Si  S 


^    S  o  sS 

>>S  u  o  o  o 

O     PQ     « 


o  O 

5o 


D    <      i-f      w      '-' 


is       "" 
O        O  o 

u    W 


^  2  3 


.2         M 

>>3  q 
0:0.2 

aa^so 

*3  o  >■  5 


.iS'-"        O  — ' 


0 


d     -i 


aa^ 
5  &  a  4; 

"  Q  b  b 


-a     'a 


a  s 


•g  11         .2 


U     Ph 


(11    (X« 


Jan.  14, 1918     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      57 


tA  0    l^          VI 

0        VO    I^  »nOO    fO 

i 

gXXX     XX     XXXXX     1 
sooo      00      00000 

(S 

•a 
0 

g«««      «< 

M           «    «    «    N    « 

•* 

i 

SXXX     XX     XXXXX 

s 

•0 

gooo      00      00000 

(^ 

0 

CO 

^ 

0   u^  0         OT  I 

n       n  i/i  N  0  0 

v 

.  t^vo   'n        10  r 

in 

gXXX     XX     XXXXX     1 

■o 

Sooo      00      00000 

0 

S   M   «   «        «  C 

0  >o  u>     00  >/>     N  lAoo  00  0 

a 

.    •*  «)  w          «    c 

gXXX     XX     XXXXX 

•a 
0 

Soom     00      oo>nooo 

^ 

•«    -2 

t  -s'l  i 

T 

' 

» 

m 

s? 

a   0.8   0  ^  ^  >."  1 

iS 

a 

■w  0  «   C   B. 

T3 

'S>.. 

^■7   >i  0  0  '^ 

J>.cn-H>>i?iS^-i-'<-'-t-'>>      1 

'3 

SS29£5ii>.>.H 

0  0.-^  0  O.o-q'Su  0 

\ii 

«     wC 

Cc/3pt,feQ 

6. 

6.   ^• 

0 

g 

•a  p   g> 

>•□ 

"E 

CO 

M       «  n 

^  av 

»    1 

C3 

^       1^ 

5ii&& 

a 

•0 

0 

2    S  = 

^lll^ 

-.3 

vX. 

^•V  >,o  o-^ 

?< 

•>,ffi-w>>in>-w+j-w>.      1 

^■a  9  22  g 

£  52  >-->.g 

a 

*C/ 

0  C-b  c 

0     tflC 

*rt 

fe 

pMt.feQ 

1 

i 

"O 

•z 

CI 

u 

3 

•a    -a 

•a 

'4-1 

a     4) 

a  >>'0 

en 

t/i 

«       V! 

en  a  <" 

» 

H 

>. 

>.a    s 

0  E:  S  >.K 

Id 

c        a 

■S  0  «  c  a 

•a 

'3        % 

■0    -;  0? 

>.'a  a^  g- 

0 

il  0  ao  "> 

^  -w  COT)  0 

":     Td  +j  4j 

0  .    cu       ^ 

**•*  ^  0  0  t^ 

«  >.S  >.  ir 

>,*j  -w  -i;  >. 

±i±i  i>  t:  i>  c 
"S'S  0  a  0  c 

1,      0  ^  (u  0  0 

%U^P>      P< 

•s 

•s 

(A 

tfl 

ss 

s 

s.  -g 

g- 

"S 

a    w 

a 

u 

ui 

«     S 

<A 

>. 

S 

>> 

0     C 

0 

M 

t 

B 

■>-'       Q. 

a 

T) 

>>     a 

i 

^  >.d. 

0 

^        M 

oil« 

^ 

s     o-a 

"o 

iijj  0 

^1  •*-' 
lu  V  a 
a^  0 

> 

"a 

2|   i§ 

"1° 
.  0 

p^wQ     P<^ 

fi 

w>Q 

3 

a 

'^ 

I 

^ 

1 

4J 

PC 

> 

"a 
0 

S 
pq 

1 
a 

1 

J 

a 

0 

S 

58 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  a 


I    -5 


s 


►J 

PQ 

< 


<     3 


i    ; 


<, 


o  o  [H  o  1-  o 
0*3  n  cs  M  o' 


•3     I 


'  oa  '  0;:=  '  oa 

iJ>>o  o>>o  S>>0 

"5  S'h'5  o'Cb  h'C  b 
ogsoaaiOHiio 


£    .  '  °l^ 

S  rn   O  -    4>  fl 

8*0   K  _•   M       "^ 

CD  OT*  ""  — 


O 


ot3  d  """d 


•""d 


©a  '  oa  '  o-A 
^  <u  o  .^  11  o  K^  ai 

<  ^  >.      l»  >.      4)  >< 

; B'Cb u'cn a-j:"5    s^'cB'Cm 
)      o      o      o   u    u      S 


2" 
00 


>.        >.        So         3 
Eg    S 2 

5  a  .^a 


^3. 

M   y   rt         «   . 
^        O       P 


O       o  m 

O  ^xi.g 

|o.ra 


'I  ^s 


o      «     o 


rtX!  3 


g  8    XI  "    5 


^; 


^"  3 


1^     X 

:§ia 


-.2  ^ 

XI  a  t    -a  f 


|a  :1a  Js^  Js 


™  8jj    ^8 

POP 


Is-- a 


5     o 


t«  J3  3  < 


•°i  ^X  -"i 

|8  :1a  |8 

p       i3      P 


^^Isim     3  8    -a a    .3  8  :    .«! 


2    2 


W      hi 


S  So   . 


S 


2 
1 


as  :S 


2    2 

o  «  o*j 
_-3j3qJ3 

CO   o   M  <>   bfl 

^    a   3 


S.2*-'.S 

as-as 

•J  o  ^  o 
■5     hJ 


Jan.  14,  i9i8     Cultuves  of  Wood-RoUiug  Fungi  on  Artificial  Media      59 


6o 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  2 


^        <u 

S    "I 


1&       ca 


<3 


^ 


X    s 

l-t     « 


•o  u;5 


u  0 


6  6 


t23 


>>  1-   lU    >> 

a  o-r;  a 
&'o  o  & 

01  u  o  (« 

H     OH 


a  ot;  a 


.Si     S'.s     o 


o    .tiS' 


o  o 


•05 

:3 


■^.a    .g. 


5    5 


o 


in  K 
3  2 

.CJ3 


.ti  n 
^  2 


.ti  fl 

&  p 


6  >■ 


-O  rt  g  3  >,3 


o  o-j;  a    t; 

&  o  o  fc     ^ 
cil  u  o  =«      o 


H     U 


55 


•^     .ti 


a 

a  a 


»  9     -S 


^xxxxxxxxxx 

Soooooooooo 


O  >^  i^.^O  <r)  1/1  o  o   o  00 
.  00  r^  wi  >Avo  r^oo  ^  00  « 

SXXXXXXXXXX 
s.,0000000000 


JXXXXXXXXXX 

(OOOOOOOOOO 


SXXXXXXXXXX 

VmOOcCOOOOOOO 


J  o  g  o  G£oi::£a 
5-0  0-0  0-3 -o-o •35 


s  o 


^a-o 


CI  o 


a  o  a  a.i'.i'.r^a 

*j   :  4->  4J  o  o  o  ■»-> 
o  ;  o  o>>>  o 


_>. 


>,>,>,>,  ^  .^  _^ 

a  a  a  a^^^a 

000  o-^TJ-rr  q 


a;  o  a  o  o  o^.^.S?a 
o  o  o*i 


Jan.  14. 1918     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      61 


r^  10  u^  0 
0   i^  r.  Co 

"  ° 

S 

?:R 

>. 

-■xxxx 

XX 

X 

X 

XX 

ca 

s  0  0  0  0 

00  0 

0 

0 

0  0 

ui 

•0 
0 

;^«... 

A  " 

0  « 

^ 

0 

"&      6      6 

odd      6 

606 

ui 

0   u^  «   ui 

.  «0  0    CI    't 

00  w 
0 

-o' 

■^ 

^S 

^ 

1    C     C 
0 

u 

oaa   0 

POO 

>> 

C3 

SXXXX 

XX 

X 

X 

XX 

ca 

■a 
0 

*2  0  0  0  0 

^     C<     «     C4     C< 

00  0 

C3    " 

0 

0 

0  0 

0   0  «  r. 

CO  CO 

0 

w 

t  0 

1 

SXXXX 

XX 

X 

X 

XX 

a 

_3 

>. 

•a 
0 

^    0    0   "1  0 

00  0 

0 

0 

0  0 

Ti 

•a 

tn'          •          • 

<^ 

13 

0 

.^      d      d 

6  6  6     0 

d      d      d 

«i 

0  CO    CI    C 

«8 

0 

0 

00  >n 

a 

S   "^    ■^. 

•o-c-a     TS 

73      •«      -o 

>> 

ca 

SXXXX 

XX 

X 

X 

XX 

"T; 

•g 

3      : 

•0 
0 

^  0  0  00  c 

CO   "• 

e   " 

0 

0  0 

^ 

Ov 

t:£ 



1 

H 

11 

0 

a 

a 

U3 

v 

(U 

bl 

01 

a 

3 

1      • 

(A 

s 

E 

1 

8 

a 

a, 

S 

"o 

0 

"5 

tn 

5 

M  : 

<3 

0 

0 

3 

•0 
0 

.2     0 

0 

•0 

0 

z 

d      0 

•C      T3 

0 

•a 

0 
•a 

0 
•a 

ca 

> 

at  6 

2  a 

a » 

> 

a 
0 

•  > 
Z  a 
•a  & 

°2 

a 

■3      ; 

2 
1 

0 

■0 
% 

2  Mg.^ 

n 

.>;2 

01  a 

■0 
0 

11 

3-0 

•'°'2 

<>> 

°w  ^5 

ii  0  > 

1^ 

3^ 

5  >;>. 

u 

^Hii. 

b) 

"S.^ 

s 

Q   (D    tu    0 

*J  0  a* 

■w  0 

■w    ly    D 

■■i- 

cjfcfefe 

uto 

U 

fe 

;^(:4 

0 

•a 

S 

: 

ft, 

1 

0 

.200 

000      c 

000 

s   '«.   ■« 

•q-CT)     t: 

73     T)     -q 

> 

a 

"3 

E 

0 

s 

0 

0 

0 

.V 

1 

1 
1 
3 

>. 

■0 

0 

^    1   'S 

0 
0 

0 

0       c 

■2   g  .a 

g    .  ca    . 

^    «    u 

1 

>. 

ca 
•a 
0 

0% 

II 

0 

0 

a  a 
0  0 

0 
> 

V              > 

C 

> 

> 

0 

d 

3 

'3 

•* 

a 

0  0 

0  C 

0 

0 

.s 

rtj2  3S.2 

1-0  rt.Sf «     « 

C8 

00 

OCJ 

0 

0 

if 

u    PQ    M 

Oh}(J      c 

•g 

— 



— |- 

S 

a 

OS 

-M       a       ^ 

6 

■u       a 

0         K        -M 

ca 

s 

ca 

•a  1  ^ 

0       J3  3  c 
0    ..g  wX 

ia 

.2=     X 

-    2    % 

"0 

^ 

> 

a 

a  • 
0  : 

0  • 

^    ; 

b  ■ 

0 

§ 

i 

0 

0 

2    \ 

^3     0    2 

«5o      0    . 

1 

1 

> 

CJ 

0 
•0 
0 

0 

3 

y 

rtJ3  3-W.2 
0     W     1- 

8^  o(ci  <: 

a-3  s   oiQ 

-a 

5  8 

> 

>. 

0  ■ 

-0 

iS 

>> 

a 

>,fi  >.  3  > 

^  P  ca      asi 
^00 

0 

4< 

^-" 

0    ; 

"i 

s 

^ 

0     CO  M  ca^a  CI 
t     Oh4U     C 

5  " 

0> 

^     >.     > 

K 

s 

III! 

11 

J 

c 

!•§ 

s 

•H 

■"     a 

li 

■!->         4 

0      «i      -y 

?i 

•a  I 

=     2d 

0  X  3 

.3       ■*■ 

■u     -y     j: 

•^  *■ 

4- 

3 

0 

•is 

t 

3 

3 

C9 

a 

IS 

.  1  1  f 

0  0  0  c 
OUOO 

0  c 
00 

c 

a 

S  i 

1 

0  .i:     c 

> 

X 

0 

0 

•a 
0 

0      o-y 
0    W    P 

>         •  0    ^ 

5   ap   8« 

s 

a 

^cj.M  tax  0 
•     Onto      C 

J      ^  f  a       asi 
>     ^     U     U 

i 
•0 

3 

a 

<LI 

S 

E 

< 

c 
E 

< 

ja 

■t-j 

■  0    4. 

:    5    5 

H 

is 

;  1 

:^    -i 

'    ."^ 

0 

0 

"c 

1. 

1 

- 

•0 
0 

0 
0 

8d4 

j     1 

:    5 

^1 

••a  .  < 

'di 
-52 

:    2    •§ 
0 

i   '^w  a^s 

l-S 

2 

>•> 

a  c 
2i 

OS 

H 

0  e 

■    -i 

c 

c 

d 
•q 

06  ta)"^  '^  c 

3    fS    D    ts 

1- 
t 

C! 

E 

1 

1 

'"rt 

i     J 

i  1 
3  p 

;     i 
5     C 

J  i 

;P 

;a  , 

5(3     P 

.(    0 

c 

i 

5      < 

< 

1 
P 

< 

2 

c 
)    P 

E 

1 

< 

c 
■E 

P 

'  d 
la 
0 
P< 

S  :    a 


62  Journal  of  Agricultural  Research  voi.  xii,  no.  » 

Tables  VII  to  XII  show  the  cultural  characters  on  the  lo  agars  for 
Fomes  rimosus  (two  strains)  from  Acacia  roemeriana  (catclaw),  F. 
rimosus  from  Siderocarpos  flexicaulis  (Texas  ebony),  F.  rimosus  from 
Prosopis  juli flora  (mesquite),  F.  robiniae  from  Robinia  neomexicana  and 
a  species  of  Fomes  from  Juglans  rupestris  (?).  A  study  of  these  six 
tables  fails  to  show  any  marked  and  constant  differences  in  cultural 
characters  for  any  of  the  lo  agars  for  the  six  strains  given,  although  four 
of  the  six  strains  represent  Fomes  rimosus,  one  F.  robiniae  and  one  a 
Fomes  on  Juglans  rupestris  ( ?)  which  the  writers  have  been  referring  to 
F.  everhartii.  Some  of  the  cultures  on  carrot,  malt,  and  parsnip  show 
slight  color  differences,  but  none  great  enough  to  be  of  any  specific  value. 
On  parsnip  and  malt,  the  submerged  mycelium  for  all  of  these  strains  is 
colored,  while  carrot  has  colored  to  colorless  mycelium.  The  dominant 
characteristics  of  Fomes  rimosus  is  seen  on  beet,  celery,  bean,  and  alfalfa 
agars.  On  each  of  these  agars  there  is  but  little  really  aerial  mycelium 
which  is  downy  to  mainly  appressed,  while  the  submerged  mycelium 
extends  far  beyond  the  aerial  mycelium,  giving  a  peculiar  and  very 
characteristic  glassy  appearance  to  the  surface  of  the  agar.  This  peculi- 
arity of  growth  of  the  submerged  mycelium  on  these  four  agars  is  seen 
in  all  of  the  six  strains  and  indicates  that  they  are  one  and  the  same 
species — viz,  that  Fomes  robiniae  is  only  a  form  of  F.  rimosus,  as  many 
scientists  have  always  believed,  and  that  the  specimen  of  Fomes  sp. 
supposed  to  have  been  collected  on  Juglans  rupestris  also  belongs  to  this 
species.  These  six  examples  show  how  valuable  the  cultural  characters 
can  become  in  determining  the  specific  position  of  closely  related  or 
identical  species.  Only  one  viable  sporophore  of  the  Fomes  supposed  to 
have  been  collected  on  /.  rupestris  was  available  for  culture  work,  and 
since  this  specimen  may  have  been  wrongly  labeled  as  to  host,  the 
writers  do  not  wish  to  place  the  Fomes  so  common  on  /.  rupestris  through- 
out the  southwestern  United  States  as  belonging  positively  to  Fomes 
rimosus  until  further  cultures  with  specimens  absolutely  known  to  have 
grown  on  this  host  have  been  made. 

The  cultural  characters  of  Fomes  everhartii,  F.  arctostaphyli  (lo),  F. 
igniarius  from  Populus  tremuloides,  and  a  species  of  Fomes  from  Alnus 
sp.,  sent  to  the  writers  as  F.  igniarius,  when  compared  showed  that  the 
Foines  from  Alnus  is  more  closely  related  to  F.  arstostaphyli  than  to 
F.  igniarius  from  P.  tremuloides,  but  that  it  is  apparently  neither  of 
these  two  species;  nor  did  it  have  any  of  the  cultural  characters  of  F. 
everhartii.  In  this  instance  the  sporophore  from  Alnus  sp.  was  very 
similar  in  all  its  characters  to  the  sporophore  of  F.  igniarius  from  P. 
tremuloides,  yet  the  cultural  characters  instantly  showed  the  two  were 
not  the  same  species. 

Polyporus  farlowii,  P.  dryophilus,  and  P.  texanus  are  three  closely 
related  species  which  differ  but  little  in  their  general  sporophore  charac- 
ters.    Tables  V,  VI,  XIII,  and  XIV  show  the  cultural  characters  of  each 


Jan.  14. 1918     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      63 

of  these  three  fungi.  The  cultural  differences  between  P.  texanus  and 
P.  dryophilus  are  much  more  marked  than  those  between  P.  farlowii 
and  P.  dryophilus.  A  study  of  the  tables  giving  the  characters  of  these 
three  fungi  at  once  shows  marked  and  constant  differences  between  each 
.  of  the  three  species.  It  will  be  seen  that  general  color  resemblances  are 
not  as  distinctive  in  differentiating  specific  characters  as  are  certain 
other  factors.  Many  of  these  species  have  a  buckthorn  brown  color  on 
several  of  the  agars,  which  indicates  their  general  relationship  while 
specific  characters  must  be  sought  in  the  differences  of  growth  on  certain 
agars. 

The  difference  in  the  fruiting  of  related  species  on  the  various  agars  is 
also  of  much  value  in  differentiating  species.  For  instance,  P.  farlowii 
fruits  vigorously  on  several  of  the  ten  culture  media  producing  perfect 
and  typical  pores,  P.  dryophilus  very  rarely  fruits  on  the  culture  media 
here  given  unless  the  inoculum  is  fresh  sporophore  tissue,  while  P.  texanus 
has  so  far  fruited  on  only  two  of  them. 

IDENTIFICATION   OF   UNKNOWN   ROTS   BY  CULTURAL  CHARACTERS 

The  cultural  character  method  here  given  can  be  used  to  determine 
what  fungus  produces  a  given  rot.  In  a  large  majority  of  cases  rots, 
both  heart  and  saprophytic,  will  be  found  without  any  sporophores 
being  present  to  indicate  what  fungus  produced  the  rot.  If  careful 
inoculations  are  made  from  such  infected  wood,  it  is  a  comparatively 
easy  matter,  as  a  rule,  to  obtain  pure  cultures  of  the  causative  organisms 
and  later  grow  them  again  on  the  10  media  given,  thus  determining  their 
cultural  characters  and  from  them  the  fungi  producing  the  rots. 

The  number  of  species  of  fungi  producing  heartrots  in  living  trees  is 
not  very  great.  A  few  of  them  produce  rots  which  can  usually  be 
identified  by  the  character  of  the  rot  alone,  such  as  Trametes  pini,  but 
the  great  majority  of  them  can  not  be  certainly  determined  by  the  rot. 
For  instance,  the  rot  produced  in  conifers  by  Polyporus  schweiniizii, 
P.  stUphureus,  and  Fonies  laricis  are  so  similar  that  no  one  can  be  certain 
by  examining  the  rot  alone  which  of  these  three  fungi  was  the  cause  of 
the  rot  in  question. 

The  rots  of  structural  timbers  are  more  numerous  than  the  true 
heartrots  and  the  causative  organism  producing  each  rot  is  more  difficult 
to  determine  from  the  rot  alone,  since  the  majority  of  the  structural- 
timber  rots  are  very  similar  in  general  appearance.  Most  of  these  rots 
are  of  the  carbonizing  type,  where  a  brownish,  brittle  rot  is  produced 
which  on  dessication  breaks  up  into  little  cubical  blocks  of  varying  sizes. 
Pure  cultures  from  these  structural  timber  rots  will  differentiate  them 
as  to  their  causative  organisms  at  once.  The  value  of  such  means  of 
determining  the  causative  organism  from  the  rot  in  the  absence  of  any 
sporophore  is  very  evident  to  anyone  who  has  had  to  deal  with  such  rots. 


64  Journal  of  Agricultural  Research  voi.  xii.  No.  a 

EXAMPLES    OF    IDENTIFICATION    OF    UNKNOWN     ROTS    BY    CULTURAI, 

CHARACTERS 

The  following  examples  illustrate  how  the  cultural  characters  here 
described  may  be  used  to  determine  the  causative  organism  of  unknown 
rots: 

(a)  There  are  two  common  heartrots  found  in  conifers  in  the  western 
United  States.  One  is  the  redrot  caused  by  Trameies  pint  and  the  other 
is  what  the  senior  author  has  previously  called  western  redrot  (9) .  In 
their  early  stages  of  growth  these  two  rots  resemble  each  other  very 
much.  Pure  cultures  from  each  of  them  differentiate  the  two  immedi- 
ately, since  the  pure  cultures  of  the  fungus  which  causes  western  redrot, 
called  previously  Polyporus  ellisianus,  produces  entirely  white  cultures 
on  all  of  the  10  media,  while  the  cultures  of  Trameies  pini  are  varying 
shades  of  brown. 

(h)  On  a  recent  field  trip  in  eastern  Texas,  when  the  senior  writer  was 
studying  the  rots  of  bridge  timbers  and  railroad  ties,  he  often  found  a 
certain  rot  in  driven  bridge  piling  made  of  creosoted  longleaf  pine  (Pinus 
palustris).  At  that  time  in  the  year  there  were  no  fruiting  bodies  present 
on  the  rotting  piling  and  the  question  at  once  arose  as  to  what  fungus 
was  the  cause  of  this  serious  rot.  From  the  peculiar  odor  of  the  freshly 
opened  wood  as  well  as  the  character  of  the  rot  it  was  believed  that  it 
was  caused  by  Leniinus  lepideus,  but  nothing  could  be  determined 
definitely  from  an  examination  of  the  rot  alone.  Specimens  of  the 
rotted  piling  were  forwarded  to  the  laboratory  and  cultures  made  from 
them.  Pure  cultures  were  obtained  showing  all  of  the  characters  of 
L.  lepideus  when  grown  on  the  10  cultural  media.  Later  many  of  these 
cultures  developed  the  typical  sporophores  of  L.  lepideus.  Cultures  of 
this  fungus  can  usually  be  recognized  by  the  presence  of  abortive  sporo- 
phores which  develop  on  the  surface  of  the  agar  and  also  by  the  presence 
of  large,  thick-walled,  obovate  to  subglobose  spores  in  the  submerged 
mycelium. 

(c)  While  studying  the  rots  of  cypress  (Taxodium  disiichum)  ties  in 
wet  locations  in  eastern  Texas  two  unknov^Ti  rots  were  found  in  this 
wood.  Pieces  of  the  rotting  ties  were  sent  to  the  laboratory  and  pure 
cultures  of  two  different  fungi  were  obtained.  One  set  of  these  cultures 
showed  all  the  cultural  characters  of  Leniinus  lepideus,  while  the  other 
series  produced  a  sporophore  of  an  unknown  species  of  Poria. 

{d)  A  rot  in  the  heartwood  of  Quercus  gamhelii,  collected  in  New 
Mexico,  was  determined  in  the  field  as  caused  by  Polyporus  dryophilus. 
Cultures  of  the  diseased  wood  gave  all  of  the  cultural  characters  of 
Fames  everhartii  and  none  of  those  of  P.  dryophilus. 

(e)  A  specimen  of  rot  in  Pinus  ponderosa,  which  was  supposed  to  be 
caused  by  Polyporus  sulphureus,  was  received  from  Oregon.  Cultures 
from  the  wood  showed  that  the  rot  was  unquestionably  caused  by 


Jan.  14, 1918     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      65 

Lentinus  lepideus.  Since  the  rot  was  sent  in  as  a  specimen  of  heartrot 
found  in  the  western  yellow  pine  in  Oregon,  the  writers  are  wondering 
whether  L.  lepideus  produces  a  real  heartrot  of  living  pine  in  that  State 
or  whether  this  rot  came  from  a  dead  area  on  a  living  or  from  a  dead  tree 
which  had  been  attacked  by  this  saprophytic  fungus. 

Many  other  instances  could  be  cited  of  the  determination  of  the  causa- 
tive organism  of  a  rot  by  use  of  the  cultural  methods  here  outlined. 
Usually  it  is  not  even  necessary  that  the  sporophore  stage  should  be 
developed,  since  the  vegative  cultural  characters  on  the  10  special  media 
will  usually  determine  the  identity  of  the  fungus.  The  practical  impor- 
tance of  such  a  method  of  determination  is  of  great  value  and  is  easily 
recognized  by  anyone  who  has  worked  for  any  length  of  time  with  organ- 
isms of  this  character.  In  fact,  one  of  the  worst  stumbling  blocks  to  a 
successful  study  of  the  various  rots  of  wood,  both  saprophytic  and 
heart  rots,  has  been  the  lack  of  methods  by  which  the  organisms  producing 
these  rots  could  be  grown  and  identified  in  pure  cultures  on  artificial 

media. 

SPOROPHORE  PRODUCTION 

The  fact  has  long  been  known  that  the  production  of  sporophores  in 
nature  in  many  of  the  Hymenomycetes  was  more  or  less  dependent 
upon  light.  This  fact  has  also  been  demonstrated  for  a  few  species  mainly 
by  gross  cultures  on  dung  decoctions,  pieces  of  wood,  bread,  etc. 
Buller  {4)  has  shown  among  other  things  how  the  light  influences  sporo- 
phore production  for  a  few  species,  mainly  Agaricaceae,  but  none  of  his 
experiments  were  made  with  pure  cultures  on  artificial  media.  It  has 
been  taken  for  granted  that  light  was  essential  to  the  formation  of 
sporophores  of  the  wood-rotting  fungi,  including  the  Polyporaceae,  but 
such  had  never  been  proved  with  pure  cultures  under  control  conditions  on 
artificial  media  with  a  sufficiently  large  number  of  species  to  determine 
the  actual  influence  of  light  as  well  as  other  factors  on  sporophore  pro- 
duction. 

The  studies  here  made  indicate  that  there  are  many  Polyporaceae 
which  fruit  in  diffused  light  of  varying  degrees  of  intensity  and  others 
apparently  require  the  direct  rays  of  the  sun  to  produce  perfect  spore- 
bearing  sporophores,  while  some  can  form  sporophores  in  absolute 
darkness.  Since  this  is  only  a  preliminary  report,  no  attempt  is  made  to 
determine  a  large  number  of  factors  which  should  be  ascertained  in  a 
complete  study  of  this  phase  of  fungus  life.  For  instance,  there  must 
be  a  minimum,  optimum,  and  maximum  condition  as  to  light,  heat,  and 
moisture  under  which  a  given  fungus  will  produce  sporophores. 

It  will  be  seen  by  consulting  Table  XVI  that  only  three  species, 
Polyporus  farlowii,  Trametes  serialis,  and  P.  cinnabarinus ,  were  able  to 
develop  sporophores  in  absolute  darkness.  Only  the  first  two  of  these 
fungi  produced  both  sporophores  and  spores.  P.  cinnabarinus  pro- 
duced fairly  typical  pores,  but  no  spores  were  found. 
27805°— 18 3 


66  Journal  of  Agricultural  Research  voi.xii.No.a 

METHODS   OF    EXPERIMENTATION 

In  the  earlier  sporophore  study  the  tubes  containing  the  cultures 
were  kept  both  horizontally  and  vertically.  If  the  species  under  investi- 
gation produced  sporophores  at  all,  they  were  able  to  produce  them  in 
either  position.  However,  the  practice  was  soon  abandoned  of  placing 
the  tubes  in  a  vertical  position,  since  it  was  found  very  difficult  to  obtain 
uniformity  in  the  proper  Ughting  of  the  cultures  and  to  get  spore  prints 
from  the  sporophores  produced  in  such  a  position. 

The  general  method  followed  in  the  sporophore  studies  was  to  take  a 
series  of  tubes  on  different  agars  and  place  them  in  the  same  general  posi- 
tion that  was  described  in  the  study  of  the  cultural  characters.  It  was 
found  very  important  early  in  the  study  that  the  slant  side  of  the  tube 
should  be  kept  uppermost  and  that  the  relative  position  of  the  tube 
in  reference  to  gravity  and  sunlight  should  be  the  same  throughout  the 
experiment.  The  tubes  were  so  placed  that  the  cotton  plugs  were 
away  from  the  simUght  and  the  bottom  of  the  tubes  faced  the  light. 
As  soon  as  there  was  any  indication  of  a  hymenium  forming,  the  tubes 
were  placed  with  the  slanting  surface  downward.  After  this  was  done, 
the  sporophores  usually  continued  to  develop  normally,  and  in  due  season 
spores  would  be  formed  and  discharged  against  the  side  of  the  tube  op- 
posite the  hymenium. 

The  first  sets  of  fungi  which  were  kept  in  darkness  were  placed  in 
pasteboard  boxes  in  a  horizontal  position  and  these  boxes  inclosed  in 
other  pasteboard  boxes.  It  was  found,  however,  that  some  diffused 
light  reached  the  tubes  in  spite  of  the  double-box  arrangement.  Inside 
one  of  these  boxes  the  recording  tube  of  the  soil  thermograph  was 
placed,  while  the  registering  portion  of  the  instrument  was  kept  on  a 
shelf  outside  the  boxes.  These  boxes  were  kept  on  the  shelf  beside  the 
other  bcJxes  containing  the  tubes  which  were  exposed  to  the  direct  rays 
of  the  sun.  This  was  done  in  order  to  obtain  as  near  as  possible  the 
same  environment  for  the  tubes  kept  in  the  sunlight  and  those  kept  in 
the  darkness,  except  for  the  single  factor  of  Ught. 

In  a  later  series  of  experiments,  the  ones  which  are  recorded  in  detail 
in  this  paper,  the  cultures  were  kept  in  pasteboard  boxes  in  a  horizontal 
position  and  these  boxes  were  placed  in  a  photographic  dark  room  from 
which  all  light  was  excluded  by  means  of  sheets  of  cardboard  being  placed 
over  the  ruby  lights.  By  this  arrangement  absolute  darkness  was  ob- 
tained for  the  cultures.  The  soil  thermograph  was  arranged  in  the  same 
manner  in  the  dark  room  as  when  on  the  outside.  Table  XV  shows  the 
highest  maximum  and  minimum,  the  lowest  maximum  and  minimum, 
the  highest  and  the  lowest  mean,  and  the  average  of  the  daily  maximum, 
minimum,  and  mean  temperatures  for  each  month  as  recorded  by  the 
two  thermographs. 


Jan.  14. 1918     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      67 

Table  XV. — Temperature  records  (°F.)  for  cultures  grown  in  daylight  and  in  darkness 


Date  and  degree  recorded. 

In  light. 

In  dark. 

Maximum. 

Minimum. 

Mean. 

Maximum. 

Minimum. 

Mean. 

December  18-31,  1916: 

Highest 

Lowest 

80 
72 

66 

55 

73 
65 

75 
66 

65 

57 

69 
64 

Average 

76 

61 

68 

71 

61 

66 

January,  1917: 

Highest 

Lowest 

89 
65 

70 
52 

77 
58 

78 
63 

73 
58 

74 
60 

Average 

79 

60 

69 

71 

68 

69 

February,  191 7: 
Highest 

97 
74 

67 
56 

81 
66 

86 
68 

72 
59 

78 
65 

Lowest 

Average 

87 

60 

73 

77 

65 

71 

March,  1917: 

Highest . 

96 
71 

72 
50 

80 
62 

88 
73 

80 
62 

83 
68 

Lowest 

Average 

86.5 

59 

72.7 

79 

68 

73 

April,  1917: 
Highest 

96 
69 

66 
46 

77 
65 

86 
68 

80 
65 

80 

Lowest 

67 

Average 

82 

59 

70-5 

78.7 

70.8 

74-7 

May,  1917: 
Highest 

100 
61 

67 
49 

83 
56 

77 
60 

74 
57 

75 
59 

Lowest 

Average 

So 

60 

70 

71 

68 

69-5 

June,  191 7: 

Highest 

no 

82 

78 
50 

93 
68 

92 
70 

88 
69 

90 
70 

Lowest 

Average 

96 

■     70 

83 

84 

80 

82 

EXPLANATION  OP  TERMS   USED  IN  SPOROPHORE  TABLE  XVI. 

In  Table  XVI  most  of  the  headings  are  self-explanatory;  however,  there  are  a  few 
that  need  special  explanation.  The  heading  for  the  sixth  column  is  "  Inoculum  and 
date  of  inoculation."  Under  this  heading  the  character  of  the  inoculum  used  when 
sphorophores  were  produced  is  given,  as  well  as  the  date  or  dates  when  each  tube 
which  produced  sphorophores  was  inoculated.  "Wood"  means  the  inoculum  used 
was  infected  wood;  "tissue"  means  pieces  of  sphorophore  tissue,  while  "malt," 
"potato,"  etc.,  signifies  that  mycelium,  was  used  as  an  inoculum  and  that  it  was 
taken  from  a  cultiu-e  on  malt  agar,  potato  agar,  etc.  "  Inoculum  as  used  means 
the  piece  of  wood,  tissue,  etc.,  used  to  inoculate  the  media.  "  Development  period  of 
sporophores"  means  the  number  of  days  from  the  time  the  inoculation  was  made  to 
the  first  evidence  of  the  formation  of  a  hymenium. 


68  Journal  of  Agricultural  Research  voi.  xii,  no.  2 

Table  XVI  shows  that  the  authors  have  obtained  under  the  cultural 
methods  described  629  sporophores  in  the  light  and  11  in  darkness,  640 
in  all,  representing  4  families  (Agaricaceae,  Polyporaceae,  Thelepho- 
raceae,  and  Tremellaceae),  16  genera  (Coprinus,  Daedalea,  Exidea,  Fomes, 
Ganoderma,  Irpex,  Lentinus,  Lenzites,  MeruUus,  Panus  (?)  Pleurotus, 
Polyporus,  Polystictus,  Poria,  Stereum,  and  Trametes),  42  species,  and 
97  strains  from  65  host  species  collected  in  11  States.  If  the  genera 
given  in  North  American  Flora  are  used,  there  would  be  24  genera,  since 
the  following  8  genera  would  be  added :  Coriollelus,  Elfvingia,  Inonotus, 
Laetiporus,  Pycnoporus,  Pyropolyporus,  Spongipeilis,  and  Tyromyces. 


Jan.  14. 1918     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      69 


M  O  H 


ti  u"  B  u 
+^"  s,  S  a  S 


o  a_.- 


Ph      «JJ 


,  CT.  'J  Ov  i^  Cv  ij  O  I 


i  M  M    '^  w 

K  'i  !iS     <''  S>  "> 
4j  -  "  *;      .  "  . 


o  "    - 

»   3   " 


'^cB 


0.T) 


o>  >>  >,  >■ 


iSSS 


^"      -    ?   ^  K" 


>.>.>■ 


>'.P 


jQ 


8  a?     E-S- 


so"or?o"oS-"oc"SSHSoS~"SS3oSJ5dSdSaoi5dogK*'S 


fi  3. 3;  a  4;  8^3-  vj  a.^"  «j"j3'^.  34 


•-  a 


S^ 


f-;S 


^H 


^H 


o  o  o—  o 

•0T3T)   g-O 


8 
•  Si 


ji  a 


.-^  0  c  a  ~  -^  2  -^ , 


o 
s  S 


!°-S.9-$.8 


aS-S-al'^  ssis   "so 


*.2-§  °  O  o.^:-^ 


2.2.*. 


5<^ 


o  S  S  S  o  >~-Si^  «  •?  <j  ;;ra  ;j  o       o  si-S: 

a,'>'>.oa,'<a,'§>-ja,'<a,i/osai    0,0,0, 


a  :'^ 


acfo,    Q, 


I 

:!■§• 


a  & 

O   o 


Ocni^i^mo'da^      6  6  6  6  6  6  §!S  d  d 

:S:Sgg  :  :•£!      ::::::«;  g  :  : 
■fHS^H   :  ■.<iz      :  :  :  :  :  :o^   •  • 


3  §  S 


.s.S- 


:*.t:-~> 


s  2  S  i-  00^  S' 
ooa'HSSooS    -5 


It,     f*. 


^   ^  s* 


►2  ::§ 


8-3 


0,0, 


8  S 


00  00  >Soo  oo  o  i^t^oo^ooof'^ooo       wiowTj-wrroooo^oo      ^o      oor^t^t^"^      *ooo  r^      jj-oo 


70 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  2 


o 

*i    C 

§ 

.9  3 

o 

"o  g 

S 

41    O 

o 

t-> 

1 

1 

*« 

SJ 

1-.'^ 

p 

t^ 

n  a 

f,< 

•3ii' 

n 

S 

u 

«^ 

o? 

U 

<  <   ^ 


Ph     W         O     W     Ph      Pk     O 


0>M 

<  aa 

^"2  2 


■u*     aT     o   •  iT 


t^  t^  p, 

Oi  o> .. 


o>    oi  2'  2'  o\  I 


tCi-1  M  *">  o!  ti  «  N  M  n"  N  '^  IT'S  "  '^  §  9  '^  f^  a 


^ ;? 


■©, 


I  1 


fell 


a   s  is 


a,    0,^0, 


h    a   a 


a,    a 


> 

< 


3  5  =3 


fe    <J    :2; 


;2;3 


S  d  g 


:q,    q. 


0\  Oi  r*       00 


O   O    C*       CO 


Jan.  14. 1918     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      71 


72 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  a 


o  3  te 


5     5 


J2 

a 
o 


.^•i 


H  s  O 


PQPhM         PiW     cw 


o  o  u 


>  l;  S 


O-  .000000--  •■  ■  •        o 


00   r^  »^00    ro  t^  «  00  CO 


SO   fO«c«    f^Ooo   O 


3   O  J5 


o    ■        -o-ooooo----  o 


S     "3  o  K  E 


SB^B 


rt  1  t: 


a"!- 


o  o. 


u  o 

set:    .  „ 

C        c«        p  m 

S  fl  <->     o  J5 

0.  b  _-    .  .-•  n, 

g  «  a  rt 


Sa 


^•50  J 


ii  c  H  a  S  gifi"  n  d  3  p  g  ^J-a  dpSi{|HS«|dSag5iiSg  o  -  j 


3  •-  S  >-'h~  rt 

TrSje  "S  «  s  2 


III 

s  SS 


28 


S^: 


O  O  Qf^  o  o 


.2  2     ■  wis  I*  K     ■     -.2  ^     • 


6  6  6  6  6  ui'^ 

'O'O'd'd'O  ca  t» 


H^H 


■2.8 
11 

!•- 

.§■§ 


o  d  6  *^  ^  6 


qtij 


2< 

a  S 


^JJ^Kj 


f^  O  t^  in  foco  M  00  't  w  00  M  ro 
00  O  t^  r^OO  "^OO  O  t^O  OD  w  >0 
HMrtMHMMWMMlHOlO> 


ovO   O   On  t/lvO 


Jan.  14, 1918     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      73 


2  a! 

Bo 


g 


.S  6 


^      5 

M  3 

£it:i  «^ 

a  0) 

ngUsh 
dine  re 
renadiii 
scarlet. 
Do. 

So 

w   0 

P^ 

5« 


50.5 


A    -i 


i.    i 


iij  1^        so. 


rt  ffl  O  » 


>^s^^« 


oH  V  O 


\A   3 


"S    -^^'^-St!"  8  S  p  § 
•^■5t5-^5ca^-">;n 

*-'   u   -    O  ±^   rt   J=^ 


•-  O   9  S   < 


5^ 


13  a  ;3     „- 

o         -o  S    S    o    u 


-oi 
aa 


s  a  .  a 

aa  oS 


ggiioSaBgS§S2s^--g 

ga«aafeSaj5  5SS8a«„- 
U     S        OU     OP4UCJ        Sfu 


<.>  o 


a   a 


a   a   a. 


So 

<5i 


1 

1 

.2  ^ 


3  o 


t,     < 


o  o  ^ja 


2- 


ci.a, 


00  00         f*l*0   fO  N 


74 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  2 


s|.^   . 

h  R  o     _:  o 

0 

0  0 

Q 

PQ 

•3  «    .Ef^ 

AhU     h4 

-0 


S  S  J' 
>  S  g 


h2 


'^1-6 
•o  0..2 
►g;  a  a 


as 


Sh«d 


sy 


^13 


O  o 

a  CO 


5    '«'3    "m 


16  „ 


^ad 


S  "  ao  a  a 


0  o 

8  H. 


^■-     E 


i'^'d! 


c.  3 
S  a 


-      .  5    1)    O    H    >v 

i'>--wr<-»!3<j-i-rrr!-w'''?<-i-'t.!3C^'-22i-i 


3  aaggj 


-2  "3  • 


■2        -5 


■S      S 


•o-a  ^ 


gS 


a; 


^;    <'A 


o  <u 


3T3 


!2;   H 


§2 


0,0,   :o. 


£5  20 


O   O    Oi       00 


t^  O    O    O       00 


00    O^        OtOO 


Jan.  14. 1918     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      75 


i  «±;  01  o 

"0  t-  'o  +j  -*j 

J3  CI  rt  «  a 


o  o  d.3  oj-H      4)  ca  g  a  2  ^  •»      ' 


«     0 


ca  rt  3  — 


Au 


o  caja  o 
OOP 


2 


.hi 


t>.  o» 

00 

t^ 

000 

i. 

in 

k 

J," 

1^00      ^a 


i  a 


ca  o  cij 


t!     S  d    dg  fc  .0. 


•ast:at5S 


■H  b^fl     hrS" 


oH 


flS«      §  art  O-a  «  n  Qg  rt  &iH  H  w^  Ubj  H  o  art 


G  a~ 


(-.a 


a,    a, 


-idS 


^3i  ' 


i 


xo  do  t^  o  o\ 


76  Journal  of  Agricultural  Research  Voi.  xii.  No. » 

Sporophores  were  produced  when  various  kinds  of  inocula  were  used, 
such  as  wood,  tissue,  spores,  or  the  myceUum  from  other  cultures  and  on 
a  great  variety  of  artificial  culture  media.  The  wide  range  of  genera 
and  species  covered  by  this  investigation,  together  with  the  large  number 
of  sporophores  produced,  clearly  proves  that  the  sporophore  production 
here  reported  is  not  an  accident,  but  is  a  constant  and  permanent  per- 
formance determined  in  the  great  majority  of  cases  by  the  presence  of 
light  and  to  a  very  limited  extent  by  the  character  of  the  substratum. 
Aeration  and  humidity  are  two  other  factors  which  also  enter  into  sporo- 
phores production  on  artificial  media. 

INFLUENCE    OP   SUBSTRATUM    ON    CHARACTER   OF    HYMENIUM 

The  influence  of  the  host  or  substratum  on  the  character  of  the  spore- 
bearing  surface  is  well  illustrated  in  the  several  strains  of  Polysticius 
hirsutus  when  grown  in  artificial  cultures.  In  all  of  these  strains  the  best 
developed  and  well-defined  pore  surfaces  are  produced  on  carrot,  malt, 
and  parsnip  agars.  On  these  the  pore  surface  is  fairly  typical  of  that 
produced  in  nature,  both  as  to  size  and  color  of  pores.  In  the  prune  agar 
the  pore  surface  is  usually  reduced  in  the  majority  of  the  tubes  to  a  few 
scattering  irpiciform  spines.  On  some  of  the  prune-agar  tubes  the  spines 
are  not  flattened  like  those  of  an  Irpex,  but  are  round  like  those  of  a 
Hydnum.  In  the  corn-meal  tubes  the  hymenium  in  a  large  majority  of 
cases  is  very  similar  to  that  produced  in  various  species  of  the  Thele- 
phoraceae,  being  reduced  to  a  smooth  or  slightly  granular  surface  in  which 
there  are  no  definite  pores.  In  all  of  the  tubes  (carrot,  malt,  prune,  corn- 
meal,  and  parsnip  agars)  there  is  an  enormous  production  of  spores  irre- 
spective of  the  character  of  the  hymenium.  In  fact,  in  some  of  the  tubes 
the  first  evidence  of  any  sporophore  production  is  the  deposit  of  spores  on 
the  opposite  side  of  the  culture  tube  when  even  a  careful  examination 
with  a  hand  lens  fails  to  show  any  signs  of  pores  or  spines. 

The  size,  shape,  and  coloring  of  the  pores  and  tubes  produced  in  artifi- 
cial cultures  on  many  of  the  agars  are  practically  identical  with  those 
found  in  nature  for  a  given  species.  However,  it  often  happens  that  in 
abortive  sporophores  on  certain  agars  the  coloring  is  not  as  pronounced 
as  on  agars  where  the  sporophores  reach  their  full  development.  In 
Polyporus  cinnabarimis,  for  instance,  the  sporophores  produced  on 
potato  agar  are  nearly  white,  while  the  same  strain  will  produce  the 
typical  grenadine- red  to  flame-scarlet  pores  on  malt. 

TRUE   PILEI    IN    ARTIFICIAI,   CUI.TURES 

One  of  the  most  interesting  facts  brought  out  in  this  investigation  was 
that  in  all  of  the  thousands  of  cultures  made  with  the  hundreds  of  sporo- 
phores produced  not  a  single  one  had  a  typical  pileus,  unless  the  fungus 
was  a  gill-bearing  form  ( Agaricaceae) ,  when  the  culture  tubes  were  so 
placed  that  the  slant  faced  the  light  in  such  a  manner  that  its  rays  were 


Jan.  14, 1918     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      77 

more  or  less  at  right  angles  to  the  agar  surface  where  the  hymenium  was 
being  developed. 

At  first  it  was  believed  that  the  absence  of  pilei  was  probably  due  to 
the  fact  that  the  tubes  were  left  in  a  horizontal  position  and  therefore 
pilei  had  no  chance  to  develop.  The  same  results  were  obtained  when 
the  tubes  were  placed  vertically  with  the  slant  side  facing  the  light  as 
when  placed  horizontally.  Very  recently,  however,  the  writers  have 
devised  a  method  by  which  small  but  otherwise  typical  pilei  have  been 
grown  on  artificial  media  in  the  test  tubes.  The  method  was  as  follows : 
In  place  of  arranging  the  culture  tubes  so  that  the  light  fell  directly  on 
the  slanted  surface  the  tubes  were  placed  in  one  of  two  positions  (i) 
vertical  in  opaque  boxes  in  such  a  manner  that  the  rays  of  light  would 
fall  on  the  tops  of  and  parallel  to  the  tubes  and  none  on  the  sides  or 
bottom,  and  (2)  nearly  horizontal,  with  slanted  surface  of  the  agar 
turned  downward  but  ^^dth  the  light  again  falling  only  on  the  tops  and 
parallel  with  the  tubes.  Typical  sporophores  were  produced  by  this 
means  for  Polyporus  dryophilus ,  P.  hirsutus,  and  Fomes  rimosus,  the 
only  species  tried  so  far.  Whether  this  method  will  produce  pilei  with 
all  species  is  not  known.  There  are  objections  to  both  methods  where 
one  desires  to  obtain  spores  for  plating  in  order  to  obtain  individual  spore 
colonies — viz,  when  the  culture  tubes  are  kept  vertically  the  discharged 
spores  fall  on  the  agar  rather  than  on  the  inner  surface  of  the  culture 
tubes  and  when  the  culture  tubes  are  kept  nearly  horizontal  with  the 
slanted  surface  downward  from  the  first  the  mycelium  grows  around  the 
edges  of  the  agar  onto  the  glass,  thereby  covering  the  surface  of  the  tubes 
where  the  spores  will  fall.  There  is  also  this  further  objection  to  both 
methods — viz,  that  the  agar  in  drying  separates  from  the  glass  tubes  in 
a  very  irregular  manner  in  place  of  from  only  the  top  surface  as  it  does 
when  the  tubes  are  kept  horizontal  with  the  slanted  surface  uppermost. 

INFLUENCE   OF   LIGHT  ON   THE  FORMATION   OF  PILEI 

The  pilei  of  the  Polyporaceae  always  developed  in  such  a  manner  that 
their  tops  were  directly  toward  the  sunlight.  Also  the  pilei  of  the 
Agaricaceae  when  grown  in  cultures  were  strongly  proheliotropic  from 
the  very  beginning  of  their  formation.  This  positive  heliotropism  was 
especially  marked  when  the  sporophores  of  Lentinus  lepideus  and  Pleuro- 
ius  ostreatus  were  developing.  The  writers  tried  P.  osireatus  in  three 
different  positions:  (i)  Culture  tubes  placed  vertically  but  with  the 
lower  part  so  shaded  that  the  light  entered  at  the  top  of  the  tube;  (2) 
tubes  placed  horizontally  and  covered  with  black  paper  so  that  the  light 
entered  only  at  the  top  of  the  tube  and  this  placed  toward  the  light; 
(3)  the  third  experiment  was  made  in  a  flask  in  which  the  medium  was 
slanted  on  the  side  of  the  flask.  The  flask  was  placed  upright  and  the 
development  of  the  sporophores  from  this  more  or  less  slanting  surface 
was  observed.     In  every  instance  from  the  earliest  development  of  the 


78  Journal  of  Agricultural  Research  voi.  xii,  No.  2 

sporophore  to  the  complete  expansion  of  the  pileus,  the  sporophores 
always  pointed  directly  toward  the  light.  In  the  case  of  the  vertical 
tubes  it  was  to  be  expected  that  normally  these  sporophores  would  grow 
vertically,  since  it  was  presumed  that  gravity  as  well  as  sunlight  might 
be  a  factor  in  the  upward  growth  of  the  sporophore.  When  the  tube 
was  placed  horizontally,  a  totally  dififerent  condition  existed.  In  this 
case  the  sporophore  developed  directly  toward  the  sunlight  but  at  right 
angles  to  the  force  of  gravity.  In  the  experiment  with  the  flask  the 
upper  two-thirds  of  the  flask  was  covered  with  dark  paper.  Ten  or 
fifteen  sporophores  developed  on  the  slanting  surface  of  the  agar  in  the 
flask,  all  of  them  without  exception  pointing  downward  toward  the  source 
of  light. 

Lentinus  lepideus  was  also  grown  on  a  more  or  less  vertical  slant  with 
the  face  of  the  slant  turned  towards  the  light.  Perfect  sporophores 
which  developed  under  these  conditions  always  turned  directly  toward 
the  sunUght. 

INFIyUENCE  OF  GRAVITY  ON  THE  FORMATION   OF  PORES 

In  all  of  the  experiments  conducted  with  the  Polyporaceae  the  formation 
of  the  pores  was  always  parallel  to  the  action  of  gravity.  If  the  tubes 
were  left  in  their  original  horizontal  position  with  the  slant  side  upward, 
pores  developed  on  the  slant  with  their  mouths  pointing  upward,  thus  mak- 
ing them  parallel  to  gravity.  If,  on  the  other  hand,  the  culture  tubes  were 
placed  in  a  vertical  position,  the  pores  were  formed  on  the  sides  of  the 
slant  in  such  a  manner  that  their  mouths  pointed  more  or  less  downward, 
or  both  upward  and  downward  in  some  cases  and  parallel  to  the  force  of 
gravity.  In  fact,  the  position  of  the  pores  of  a  fungus  which  produces  a 
vigorous  sporophore  in  artificial  cultures  could  be  governed  at  will  by 
simply  changing  the  position  of  the  tube  in  reference  to  gravity.  It  seems 
therefore  that  while  light  is  usually  the  main  factor  governing  the  initia- 
tion of  a  hymenium,  gravity  is  the  dominant  force  which  determines  in 
what  direction  the  pores  will  point,  irrespective  of  the  incident  light  or 
whether  the  spores  when  discharged  will  fall  onto  the  walls  of  the  porec 
or  not. 

INFLUENCE  OF  SUBSTRATUM  ON  PORE  FORMATION 

Although  it  is  a  well-known  fact  that  cultures  of  strictly  parasitic 
fungi  can  fruit  on  a  wide  range  of  media  irrespective  of  the  special  host 
on  which  the  parasite  usually  thrives,  the  idea  seems  to  have  been  gener- 
ally accepted  that  some  special  culture  media  would  have  to  be  used  in 
order  to  obtain  sporophores  of  the  Polyporaceae  on  artificial  media  in  test 
tubes,  flasks,  etc.  The  experiments  here  given  show  that  the  same  gen- 
eral rule  as  to  food  material  appUes  to  the  wood-rotting  fungi  as  a  whole — 
viz,  that  no  special  decoction  is  necessary  for  each  species  in  order  to 
obtain  sporophores;  in  fact,  the  wide  range  of  artificial  media  upon  which 
many  of  the  Polyporaceae  have  fruited  is  rather  remarkable  when  the 
nature  of  these  fungi  is  taken  into  consideration. 


Jan,  14, 1918     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      79 

Table  XVI  shows  that  there  were  56  strains  of  wood-rotting  fungi 
which  produced  sporophores  on  malt  agar,  52  on  corn  meal,  48  on  prune, 
44  on  carrot,  27  on  parsnip,  14  on  potato,  11  on  celery,  11  on  beet,  10  on 
bean,  and  only  i  on  alfalfa  agar,  while  there  are  four  of  these  agars  pre- 
eminently suitable  for  sporophore  production — viz,  malt,  com  meal, 
prune,  and  carrot,  in  the  order  given. 

POSITION   OF  SPOROPHORES  ON   MEDIA 

One  most  interesting  fact  in  reference  to  sporophore  production  on 
artificial  media  is  the  fact  that  a  very  large  percentage,  probably  more 
than  95  per  cent,  of  the  sporophores  were  developed  on  the  upper  half  of 
the  slant.  No  perfect  sporophores  have  ever  been  observed  by  the  writers 
other  than  on  the  slant,  although  the  agar  in  the  tubes  often  dries  out 
sufficiently  to  leave  ample  room  for  the  sporophores  to  develop  from  the 
slant  to  the  bottom  of  the  tube  over  a  distance  of  from  60  to  80  mm. 
Whether  this  means  that  the  formation  of  the  sporophore  was  dependent 
upon  a  small  amount  of  moisture  or  whether  it  needed  the  greater  aeration 
which  the  upper  end  of  the  tube  afforded  is  not  known.  In  many  in- 
stances, especially  with  the  Stereums,  the  fruiting  surface  was  formed  in 
a  narrow  zone  at  the  extreme  upper  limit  of  the  agar  slant.  The  studies 
so  far  made  indicate  that  an  important  factor  in  this  case  is  probably 
the  drier  condition  of  the  agar  at  the  upper  end  of  the  slant.  It  would 
seem  that  if  the  drying  of  the  agar  was  the  only  requisite  for  the  forma- 
tion of  sporophores  that  as  the  agar  dries  in  the  lower  portion  of  the 
tube  a  condition  would  be  reached  which  would  normally  produce  sporo- 
phores; yet  such  is  not  the  case.  The  by-products  produced  by  the 
fungus  may  be  more  or  less  deleterious  to  the  formation  of  sporophores; 
and,  since  such  by-products  would  be  more  abundant  in  the  lower  portion 
of  the  tube  than  under  the  slant,  it  follows  that  few,  if  any,  sporophores 
would  develop  there. 

DENSITY   OP   MYCELIUM   AND  SPOROPHORE  PRODUCTION 

Another  very  peculiar  fact  develops  that  the  sporophore  production 
usually  occurs  on  that  portion  of  the  agar  slant  where  the  aerial  growth 
of  mycehum  is  the  least.  If  a  dense  mass  of  mycelium  forms  over  the 
entire  agar  slant,  the  chance  for  the  formation  of  pores  on  such  a  surface 
is  materially  lessened.  This  may  explain  to  some  extent  why  sunlight 
plays  an  important  part  in  the  production  of  sporophores,  since  cultures 
kept  in  the  dark  usually  develop  a  denser  mass  of  mycelium  on  the  surface 
of  the  slant  than  corresponding  cultures  in  the  light.  There  are  many 
species  of  fungi  which  have  been  tried  in  both  light  and  darkness  in 
which  the  aerial  growth  is  very  limited,  even  when  grown  in  the  dark; 
and  still  no  sporophores  were  produced  when  in  darkness,  but  were  pro- 
duced in  the  light. 


8o  Journal  of  Agricultural  Research  voi.  xii,  no.  2 

INFLUENCE   OF   INOCUI.UM    ON    SPOROPHORE    DEVELOPMENT 

In  a  few  of  the  species  of  fungi  the  interesting  fact  developed  that 
when  tissue  from  sporophores  was  used  for  the  inocula  the  presence  of 
this  tissue  materially  shortened  the  development  period  for  sporophores 
on  the  agar.  For  instance,  in  Polyporus  alhidus  from  Pinus  ponderosa 
(FP  21875),  when  tissue  was  used  as  an  inoculum  on  corn-meal  agar, 
the  development  period  of  the  sporophore  was  5  days;  when  mycelium 
from  potato  agar  was  used  the  development  period  was  1 1  days ;  for 
P.  anceps  from  Thuja  plicata  (FP  21 801),  when  tissue  was  used  on  malt 
the  development  period  for  the  sporophore  was  5  days;  when  mycelium 
from  potato  agar  was  used  for  the  inoculum  on  malt  agar  the  development 
period  was  12  days.  This  rapid  formation  of  the  sporophore  when  tissue 
was  used  as  an  inoculum  was  characteristic  of  many  of  the  strains  of 
this  group  of  fungi. 

This  shortening  of  the  development  period  of  the  sporophores  is  also 
found  occasionally  in  other  species  of  fungi  than  this  group.  For  instance, 
it  was  rather  marked  in  Polyporus  dryophilus  when  fresh  sporophore 
tissue  was  used,  in  various  strains  of  Trametes  peckii,  and  to  a  slight 
extent  in  Fomes  roseus.  In  several  cases  not  only  was  the  development 
period  for  sporophores  shortened  by  the  presence  of  pieces  of  sporophore 
used  as  the  inoculum,  but  cultures  made  from  infected  wood,  mycelium 
or  spores  produced  sporophores  only  on  one  or  two  agars,  while  the  same 
species  would  produce  sporophores  on  several  media  when  the  inocula 
were  pieces  of  sporophores.  In  those  cases  where  the  sporophore  tissue 
shortened  the  development  period,  the  pores  usually  but  not  always 
start  directly  on  the  tissue  inoculum  and  then  spread  rapidly  often  over 
the  entire  agar  slant.  In  no  case  was  the  development  of  the  pores 
limited  to  the  pieces  of  inoculum,  while  in  many  instances  the  pores 
would  start  on  areas  not  immediately  adjacent  to  the  inoculum. 

SUMMARY 

(i)  The  following  criteria  were  found  of  value  in  the  differentiation  of 
the  various  species:  (a)  Macroscopic  characters,  including  rapidity  of 
growth,  color  of  aerial  and  submerged  mycelium,  character  of  the  aerial 
mycelium  as  to  texture,  etc.,  staining  of  the  agar,  decoloration  of  the 
agar,  the  comparative  rate  of  growth  between  the  aerial  and  submerged 
mycelium,  etc.  (b)  Microscopic  characters,  such  as  septation,  branching, 
size  and  color  of  hyphge,  clamp  connections,  polymorphism  in  spore  for- 
mation,' etc. 

(2)  The  sunlight  was  found  to  accentuate  the  colors  and  tone  down  the 
mycelial  growth  of  the  fungus,  thereby  making  it  more  characteristic  and 
uniform  for  a  given  species  than  when  placed  under  similar  conditions  in 
the  darkness  or  in  weak  diffused  light. 

(3)  The  cultural  characters  of  vegetative  development  of  the  various 
strains  of  a  given  species  of  fungus  show  no  appreciable  difference  between 
cultures  of  this  fungus  whether  obtained  from  infected  wood  or  from 


Jan.  14. 19^8     Cultures  of  Wood-Rotting  Fungi  on  Artificial  Media      8i 

sporophores;  neither  do  the  hosts  of  the  fungus  seem  to  make  any 
marked  changes  in  the  fundamental  cultural  characters  when  strains 
from  different  hosts  are  compared.  There  may  be  minor  differences  due 
to  the  host  from  which  the  strain  came  but  nothing  more. 

(4)  When  the  cultural  characters  of  closely  related  but  really  distinct 
species  are  compared,  marked  and  constant  differences  in  the  character 
of  the  mycelium  will  be  found  on  certain  corresponding  agars  in  the 
series  of  cultures  representing  the  two  species,  while  if  the  two  fungi  are 
really  the  same  species,  no  constant  differences  of  specific  rank  will 
occur.  Unknown  rots  can  also  be  identified  by  making  pure  cultures  of 
the  causative  organism  from  the  diseased  wood  and  determining  from  the 
cultural  characters  of  the  fungus  thus  isolated  its  identity. 

(5)  The  presence  of  light  is  essential  to  the  production  of  sporophores 
when  grown  on  artifical  media  in  the  great  majority  of  fungi  here  inves- 
tigated, while  the  character  of  the  substratum  plays  only  a  very  minor 
roll  in  sporophore  initiation. 

(6)  The  medium  on  which  the  fungus  is  grown  often  governs  to  some 
extent  at  least  the  form  of  the  hymenium  which  develops. 

(7)  The  size,  shape,  and  color  of  the  pores  and  tubes  produced  in 
artificial  cultures  on  many  of  the  agars  are  practically  identical  with 
those  found  in  nature  for  a  given  species. 

(8)  The  pilei  of  both  the  Poplyporaceae  and  the  Agaricaceae  when 
grown  in  pure  cultures  on  artificial  media  are  from  the  very  beginning 
of  their  formation  strongly  proheliotropic,  while  the  formation  of  the 
pore  tubes  in  the  Polyporaceae  is  always  such  that  they  are  placed 
parallel  to  the  action  of  gravity. 

(9)  In  a  few  species  of  fungi  the  presence  of  tissue  as  the  inoculum 
shortened  the  period  of  sporophore  development  from  one  to  several  days, 

(10)  Workers  with  wood- rotting  fungi  now  have  the  following  means 
for  determining  the  identity  of  a  given  fungus  or  the  causative  organism 
of  a  given  rot :  (a)  The  sporophore  characters  as  usually  found  in  nature, 
(b)  the  characters  of  the  rot  produced,  (c)  the  vegetative  characters 
developed  when  grown  in  pure  cultures  on  artificial  media  when  exposed 
to  light,  and  (d)  the  characters  of  the  sporophores  and  various  spore 
forms  when  produced  on  artificial  media. 

LITERATURE  CITED 
(i)  BrEFELd,  Oscar. 

1877.    BOTANISCHE     UNTERSUCHUNGEN      USER      SCHIMMELPILZE.      III.    BASIDIO- 

MYCETEN  I.     226  p.,  II  pi.     Leipzig. 

(2) 

1888.  UNTERSUCHUNGEN       AUS       DEM       GESAMMTGEBIETE       DER       MYKOLOGIE. 

VII.  BaSIDIOMYCETEN.      II.     PROTOBASIDIOMYCETEN.      178    p.,    II   pi. 

Leipzig. 
(3)  

1889.  UNTERSUCHUNGEN       AUS       DEM       GESAMMTGEBIETE       DER        MYKOLOGIE. 

VIII.  BASIDIOMYCETEN.     III.     AUTOBASIDIOMYCETEN.     305   p.,    12    pi. 

Leipzig. 
27805°— 18 4 


82  Journal  of  Agricultural  Research  voi.  xii.no.  a 

(4)  BuLLER,  A.  H.  R. 

1909.  RESEARCHES  ON  PUNGi.    287  p.,  83  fig.,  5  pi.     London,  New  York,  [etc.] 

(5)  Falck,  Richard. 

1902.   DIE    CULTUR     DER    OIDIEN     UND  IHRE    RUCKFUHRUNG    IN     DIE     HOHERE 

FRUCHTFORM  BEi  DEN  BASiDiOMYCETEN.    In     BeitT.     Biol.     Pflanz., 
Bd.  8,  Heft  3,  p.  307-346,  pi.  12-17. 
(6) 

(7) 


1909.   DIE     LENZITES-FAULE     DES     CONIFERENHOLZES.     In    MoUer,    A.     Haus- 
schwamm-Forschungen.     Heft  3,  234  p.,  24  fig.,  7  pi.    Jena. 


1912.  DIE  MERULius-FAULE  DES  BAUHOLZEs.    In  MoUer,  A.    Hausschwamm- 
Forschungen.    Heft  6,  405  p.,  73  fig.,  17  pi.    Jena. 

(8)  Humphrey,  C.  J.,  and  Fleming,  Ruth  M. 

1915.  THE   TOXICITY    TO   FUNGI    OF    VARIOUS    OILS    AND    SALTS,    PARTICULARLY 
THOSE  USED  IN  WOOD  PRESERVATION.      U.    S.    Dept.    AgT.     Bul.    227 

38  p.,  4  pi.     Bibliography,  p.  37-38. 

(9)  Long,  W.  H. 

191 7.  a  preliminary  report  on  the  occurrence  op  western    red-rot 

IN  PINUS  PONDEROSA.     U.  S.  Dept.  AgT,  Bul.  490,  8  p. 

(10)  

1917.  THREE  UNDESCRiBED  SPECIES  OP  POLYPORES.    /»  PapcTS  New  Mexico 
Chapter  Phi  Kappa  Phi,  v.  i,  no.  i,  p.  1-3. 

(11)  Lyman,  G.  R. 

1907.  culture  studies  on  polymorphism  of  hymenomycetes.    in  proc. 

Boston  Soc.  Nat.  Hist.,  v.  33,  no.  4,  p.  125-209,  pi.  18-26.    Literature, 
p.  203-209. 

(12)  RUMBOLD,  C. 

1908.  BEITRAGE    ZUR    KENNTNIS    DER    BIOLOGIE    HOLZZERSTOrENDER    PILZE. 

In  Naturvv.  Zt«chr.  Forst.  u.  Landw.,  Jahrg.  6,  Heft  2,  p.  81-140, 
6  fig.,  I  pi.     Literatiu-,  p.  139-140. 

(13)  Zeller,  S.  M. 

1916.  studies  in  the  physiology  op  the  pungi.    ii.  lenzites  saepiaria 

FRIES,   WITH     SPECIAL     REFERENCE     TO     ENZYME     ACTIVITY.      In  Ann. 

Mo.  Bot.  Gard.,  v.  3,  no.  4,  p.  439-512,  pi.  8-9.     Bibliography,  p.  504- 

509- 
(14) 

1917.  STUDIES  IN    THE  PHYSIOLOGY   OP    THE    FUNGI.     III.     PHYSICAL  PROPER- 

TIES    OP     WOOD     IN     RELATION     TO     DECAY     INDUCED     BY    LENZITES 

SAEPIARIA  FRIES.     In  Ann.  Mo.  Bot.  Gard.,  v.  4,  no.  2,  p.  93-164, 
pi.  9-13,  II  charts.     Bibliography,  p.  154-155- 


GOSSYPOL.  THE  TOXIC  SUBSTANCE  IN  COTTONSEED 

By  W.  A.  Withers  and  Frank  E.  Carruth, 

Chemical  Division,  North  Carolina  Agricultural  Experiment  Station 

REVIEW  OF  PREVIOUS  WORK 

Since  our  previous  publication  (77)^  on  this  subject,  several  articles 
have  appeared  in  which  other  explanations  of  cottonseed-meal  poison- 
ing have  been  offered. 

Thus,  Rommel  and  Vedder  (14)  have  suggested  that  poisoning  by 
cottonseed  meal  is  similar  to  beriberi,  and  is  caused  by  deficient  diets. 
This  view  was  based  on  the  similarity  of  post-mortem  symptoms  noted 
in  pigs  fed  on  rice  and  tankage. 

Wells  and  Ewing  (15)  have  concluded  that  cottonseed-meal  injury  is 
due  in  large  part  to  incomplete  diets. 

Richardson  and  Green  (11)  fed  white  rats  and  concluded  that  cotton- 
seed meal  and  flour  are  not  actively  toxic,  but  contain  insufficient 
minerals  and  possibly  inadequate  amounts  of  the  fat-soluble  growth- 
promoting  substance. 

Osborne  and  Mendel  (10)  have  secured  results  similar  to  those  of  the 
last-named  authors  with  cottonseed  meal  and  flour,  but  on  subsequently 
feeding  raw  cottonseed  kernels  supplied  by  us,  they  have  corroborated 
the  results  which  we  had  obtained  with  the  kernels.  They  admit  the 
presence  of  a  deleterious  substance  in  raw  cottonseed,  but  apparently 
still  hold  the  view  that  cottonseed  meal,  the  product  resulting  from 
cooking  the  kernels  and  pressing  out  the  oil,  is  nontoxic,  at  least  for  rats 
and  chickens  (9). 

Inasmuch  as  no  comparative  experiments  with  an  isolated  and  purified 
substance  have  been  reported,  we  present  the  results  of  additional 
experiments  with  various  animals  to  supplement  those  given  in  our 
previous  experiments,  in  which  rabbits  and  fowls  were  used. 

The  toxic  efltect  of  an  ether  extract  of  raw  cottonseed  has  been  well 
shown  in  the  rat-feeding  experiments  described  by  McCollum  and  co- 
workers (6)  and  by  Osborne  and  Mendel  (10).  This  extract  contains 
about  2  per  cent  of  gossypol,  which  is  equivalent  to  about  0.6  per  cent 
of  the  weight  of  the  kernels  from  which  the  extract  is  obtained.  Our 
rat  diet,  containing  20  per  cent  of  this  extracted  oil,  caused  prompt 
decline  in  grown  rats.     Osborne  and  Mendel  (10)  used  as  little  as  i  per 

1  Reference  is  made  by  ntunber  (italic)  to  "I,iterature  cited,"  pp.  loo-ioi. 

Journal  of  Agricultural  Research,  Vol.  XII,  No.  2 

Washington,  D.  C.  Jan.  14.  1918 

Im  Key  No.  N.  C— 8 

(83) 


84  Journal  of  Agricultural  Research  voi.xii.  No.  a 

cent  of  the  extract  (equivalent  to  about  0.02  per  cent  of  gossypol  in  the 
diet)  and  found  that  the  growth  of  the  rats  was  greatly  retarded.  Our 
experiments  have  led  to  the  conclusion  that  raw  cottonseed  kernels  are 
highly  toxic  to  rats,  but  that  cooked  cottonseed  is  only  slightly  toxic. 
Whether  cottonseed  meal  made  from  cottonseed  sufficiently  cooked 
with  moist  heat  is  toxic  to  rats  seems  to  depend  on  the  diet  in  which  it 
is  fed.  In  a  short  feeding  experiment  a  diet  such  as  was  used  by  Rich- 
ardson and  Green  (ii)  (45  per  cent  of  cottonseed  meal,  17  per  cent  of 
whole  milk  powder,  10  per  cent  of  starch,  and  28  per  cent  of  lard)  has 
shown  no  definite  toxic  effect  ^  on  our  rats  even  when  it  contains  a 
short-cooked  (28  minutes)  meal.  When  the  meal  is  the  sole  source  of 
vitamines,  protein,  and  minerals,  we  have  rarely  had  such  favorable 
growth  as  is  reported  by  Richardson  and  Green  (//)  and  by  Osborne 
and  Mendel  (/o). 

If  compared  with  ether-extracted  raw  cottonseed  or  with  soybean 
meal,  the  rate  of  growth  has  been  very  small.  The  explanation  of  this, 
according  to  Osborne  and  Mendel,  might  be  that  the  diet  was  unpalatable 
and  that  consequently  less  food  was  ingested.  Unpublished  experi- 
ments indicate  to  us  that  there  still  remains  something  toxic  in  long- 
cooked  cottonseed  meals  which  in  restricted  diets  is  objectionable  to 
rats,  causing  a  lower  food  intake,  but  the  effect  of  which  is  overcome  in 
supplemented  diets. 

This  seems  to  be  the  same  phenomenon  discussed  by  McCollum  (7), 
who  finds  that  in  diets  very  well  supplemented  the  toxic  effect  of  the 
fat  of  wheat  embryo  and  other  slightly  toxic  substances  is  overcome. 
The  evidence  of  a  toxic  factor  of  moderate  intensity  for  rabbits  and 
pigs  is  also  furnished  in  our  rabbit  and  pig  experiments. 

TOXICITY   OF   RAW   COTTONSEED   KERNELS 

In  a  large  number  of  experiments  with  rats  we  have  found  that  the 
effect  of  cottonseed  products  on  rats  can  be  predicted  accurately  when 
chemical  tests  indicating  the  presence  or  absence  of  gossypol  have  been 
made  (Table  I).  In  order  to  appreciate  the  significance  of  some  of  the 
experiments  we  may  briefly  describe  the  properties  of  gossypol.  A 
yellow  plant  pigment  having  the  apparent  formula  C30  Hjg  O9  (molecu- 
lar weight  532),  not  soluble  in  and  not  extracted  by  petroleum  ether, 
readily  soluble  in  acetone  and  ether,  moderately  soluble  in  alcohol, 
benzene,  and  chloroform,  dissolves  readily  in  sodium  hydroxid  (NaOH) 
and  sodium  carbonate  (NajCOg)  and  is  slowly  soluble  in  sodium  bicar- 
bonate (NaHCOg).  It  may  be  titrated  as  a  dibasic  acid  with  aqueous 
alkalies.  It  crystallizes  well  from  a  mixture  of  ether  and  acetic  acid  as 
a  sparingly  soluble  substance  containing  10.  i  per  cent  (i  molecule)  of 

'Data  ia  article  not  yet  published. 


Jan.  14, 1918  Gossypol,  the  Toxic  Substance  in  Cottonseed 


85 


acetic  acid.  When  dissolved  in  ether  or  oil  and  treated  with  anilin,  it 
forms  a  bright-yellow  insoluble  substance  which  is  apparently  a  di- 
anilin  salt  of  gossypol  (i  molecule  of  gossypol  to  2  molecules  of  anilin) 
CgoHjgOg,  2C6H5NH2.  This  substance  is  very  insoluble  in  most  solvents, 
even  aqueous  sodium  hydroxid. 


Table  I. — Results  of  feeding  raw  cottonseed  kernels  in  milk  diets  to  rats 


Rat 
No. 

Weight. 

Period. 

Remarks. 

Initial. 

Final. 

Change. 

Diet  366  (59  per  cent  of  ker- 
nels, 17  per  cent  of  whole- 
milk  powder,  10  per  cent  of 
starch ,  1 4  per  cent  of  lard ) . 
Do 

43 

44 

45 

46 
71 

72 
74 

75 
58 

59 
60 

60 

47 

48 
45 

46 

Gm. 

^33 

217 
156 

139 

154 

186 
130 

117 
160 

170 
131 

105 
112 

184 
128 

104 

Gm. 
no 

179 
128 

109 

100 

112 

87 

80 
107 

112 
103 

122 
185 

188 
177 

159 

Percent. 
-17 

-17 
-18 

-25 
-32 

-40 
-33 

-31 
-33 

-34 
—  21 

+  16 

+  65 

+    2 
+  38 

+  .S3 

Days. 
7 

7 
7 

7 
18 

21 
23 

19 
61 

61 
59 

20 
85 

85 
14 

Died  nth  day. 
Discontinued. 

Do 

Recovered  on  ether- 

Do 

extracted  kernels. 
Do. 

Diet  377  (30  per  cent  of  ker- 
nels, 38  per  cent  of  milk- 
powder,     II    per    cent    of 
starch,  14  per  cent  of  lard,  7 
per  cent  of  butter). 

Do 

Died. 
Do. 

Diet  378  (28  per  cent  of  dry- 
heated    (no"    C.)  kernels, 
equivalent  to  30  per  cent  of 
raw,  40  per  cent  of  milk- 
powder;  otherwise  like  diet 

377)- 

Do 

Do. 
Do. 

Diet  379  (10  per  cent  of  ker- 
nels, 50  per  cent  of  milk- 
powder,    12    per    cent    of 
starch,  28  per  cent  of  lard). 
Do 

Do. 
Do. 

Do 

Discontinued. 

Control  diet  364  (60  per  cent  of 
milk-powder,  12  per  cent  of 
starch,  28  per  cent  of  lard). 

Diet  367  (39  per  cent  of  ether- 
extracted   kernels,    17   per 
cent  of  milk-powder,  10  per 
cent  of  starch,  34  per  cent  of 
lard). 

Do 

Do. 

Subsequently   de- 
clined   from    dis- 
ease. 

Do. 

Do 

Were  on  unextracted 

Do 

kernels  for  i  week 
previous. 
Do. 

86 


Journal  of  Agricultural  Research 


Vol.  xn,  No.  2 


TabIvE  I- — Results  of  feeding  raw  cottonseed  kernels  in  milk  diets  to  rats — Continued 


Rat 

No. 


Weight. 

Period. 

Initial. 

Final. 

Change. 

Gm. 

Gm. 
112 

Per  cent. 
-19 

Days. 

6 

"'I 
200 

lOI 

185 
118 

-19 

—  10 

—  12 

8 
8 
8 

140 

107 

-23 

8 

113 
162 

133 

94 
144 
106 

-16 

—  II 

—  20 

4 
4 
7 

141 

102 

-28 

112 

80 
81 

71 

70 

—  II 
~^3 

108 

112 

102 

138 

+36 

21 

Remarks. 


Diet  368  (20  per  cent  of  ether- 
extract(equivalent  to  about 
60  per  centof  raw  kernels), 
replacing  20  per  cent  of 
lard  in  control  diet  364). 

Do 

Do 

Do 

Diet  369  (0.4  per  cent  gossypol 
(equivalent  to  about  20  per 
cent  of  ether  extract),  re- 
placing 0.4  per  cent  of 
starch  in  control  diet) . 

Do 

Do 

Do 

Diet  373  (o.i  per  cent  of  gos- 
sypol added  to  control  diet). 

Do . 

Do 

The  survivor,  rat  58,  was  then 
put  on  the  same  diet  minus 
gossypol. 


51 


30 
31 
32 

58 

59 
60 


Died. 


Discontinued. 
Do. 
Do. 

Do. 


Do. 
Do. 
Died. 

Discontinued. 

Died. 
Do. 

Discontinued. 


The  sparingly  soluble  compounds  of  gossypol  w^ith  acetic  acid  and 
with  anilin  have  been  used  to  estimate  the  amount  of  gossypol  present 
in  cottonseed  kernels.  Both  methods  have  given  results  which  show 
that  gossypol  exists  in  cottonseed  kernels  to  the  extent  of  approxi- 
mately 0.6  per  cent. 

In  order  to  explain  the  change  in  toxicity  in  cottonseed  after  being 
cooked  in  the  mill,  we  offer  the  following  hypothesis:  Under  the  action 
of  moist  heat  the  gossypol  streams  from  the  glands  and  is  spread  over  the 
seed  tissue.  Part  is  oxidized  to  a  less  toxic  substance  which  we  may  for 
convenience  call  "  D-gossypol; "  part  is  left  in  combination  with  the  bases 
or  protein  as  a  salt  of  gossypol;  and  part  is  expressed  in  the  oil.  The 
degree  to  which  these  changes  take  place  is  dependent  on  the  method  of 
cooking  and  the  condition  of  the  seed.  In  dry  heating  to  100°  C.  there 
is  practically  no  decomposition  of  the  gossypol.  In  very  dry  seeds  the 
gossypol  may  not  spread  over  the  seed  tissue  and  be  changed  unless  much 
moisture  is  added  and  the  cooking  prolonged. 

Some  quantitative  data  on  the  amount  of  gossypol  left  in  the  seed 
have  been  obtained  from  samples  of  kernels  cooked  various  lengths  of 
time — 5,  10,  20,  and  28  minutes.  The  percentage  of  gossypol  extracted 
by  ether  in  these  cases  was,  respectively,  0.62,  0.24,  o.io,  and  0.07. 


Jan.  14, 1918        Gossypol,  the  Toxic  Substance  in  Cottonseed 


87 


METHOD  OF  REMOVING  THE  TOXIC  SUBSTANCE  FROM  THE  ETHER 

EXTRACT 

By  treatment  of  the  ether  extract  of  raw  kernels  with  an  excess  of 
anUin  the  gossypol  is  practically  quantitatively  precipitated.  The 
dianilin  salt  produced  is  extremely  insoluble  in  most  solvents  except 
hot  anilin  and  alcoholic  potassium  hydroxid  (KOH).  The  substance 
itself  is  not  toxic  because  of  its  insolubility.  It  passes  through  the 
alimentary  canal  unchanged,  as  can  be  seen  by  a  glance  at  the  feces. 
One-half  gm.  doses  of  this  anilin  compound  were  fed  for  seven  con- 
secutive days  to  a  rabbit  without  result,  and  it  was  also  given  to  rats 
in  a  milk  diet  (0.3  per  cent).  The  food  intake  of  the  rats  was  not  dimin- 
ished; nor  were  the  rats  affected  perceptibly 

Gossypol  "acetate"  was  then  prepared  from  this  compound  as  follows: 
The  substance  was  decomposed  by  means  of  an  alcoholic  alkali.  The 
anilin  was  steamed  off,  and  the  gossypol  was  extracted  with  ether  and 
crystallized  as  the  "acetate"  by  the  addition  of  acetic  acid.  This  was 
fed  in  amounts  (0.25  per  cent)  equivalent  to  the  anilin  compound  (0.3 
per  cent)  fed  previously.  The  rats  which  had  not  been  affected  by  the 
anilin  compound  were  promptly  affected  and  consumed  but  little  food 
(see  Table  II). 

By  passing  steam  through  the  extract  from  which  the  anilin  com- 
pound had  separated  the  excess  anilin  was  removed,  and  the  resultant 
oil  did  not  prove  toxic  to  rats.  The  result  of  this  experiment  has  led 
us  to  believe  that  gossypol  is  the  only  substance  in  raw  cottonseed 
possessing  marked  toxic  properties.  This  conclusion  was  indicated 
in  our  previous  experiments  {17),  wherein  we  found  that  the  gossypol 
extract  freed  from  gossypol  was  not  toxic. 

Table  II. — Results  of  feeding  gossypol  "acetate"  to  rats 


Rat 

No. 

Weight. 

Period. 

Remarks. 

Initial. 

Final. 

Change. 

Diet  436  (gossypol   "acetate"  (0.25 
per  cent)  prepared  from  decompo- 
sition of   the    insoluble   nontoxic 
gossypol-anilin  compound,   added 
to  the  control  diet). 

Do 

201 

202 
203 
160 

171 

175 
94 
96 
90 

Grams. 
166 

122 
128 
117 

155 

95 
124 
122 

Grams. 
119 

91 

79 

190 

190 
118 
167 
137 

Per  cent. 

-28 

-25 
-26 

—  32 

+  27 

+  25 
+  24 

+35 

+  12 

Days. 
12 

12 
14 
14 

39 

39 
39 
39 
39 

Died. 
Do. 

Do 

Do. 

Do 

Do. 

Diet  448  (12.5  per  cent  ether  extract 

freed  from  gossypol  by  treatment 

with    anilin,    replacing    12.5    per 

cent  of  lard  in  control  diet). 

Do 

Alive. 
Do. 

Do 

Do. 

Do 

Do. 

Do 

Do. 

88 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  3 


TOXICITY    OF    GOSSYPOL    TO    RABBITS 

In  our  previous  paper  (17)  most  of  our  feeding  experiments  with 
gossypol  were  with  the  "acetate,"  a  crystalline  substance  containing 
acetic  acid  in  its  composition.  This  product  had  the  same  toxic  action 
as  the  product  precipitated  by  petroleum  ether;  therefore  we  inferred 
there  was  no  change  wrought  by  crystallization.  We  have  recently 
fed  gossypol  in  amounts  equivalent  to  a  toxic  weight  of  cottonseed 
kernels  and  have  found  it  to  produce  serious  results  in  every  case.  Where 
gossypol  itself  is  added  to  a  diet  in  appreciable  amounts,  the  toxic 
effect  is  marked. 

Gossypol  was  mixed  with  the  feed  in  four  forms:  (i)  Precipitated  (by 
petroleum  ether),  (2)  recrystaUized  "acetate"  (lo.i  per  cent  of  acetic 
acid),  (3)  "free"  gossypol,  a  very  pure  product,  and  (4)  as  the  sodium 
salt  of  gossypol  (gossypol  "acetate"  neutralized  with  three  molecules  of 
sodium  hydroxid,  10.5  c.  c.  of  N/2  alkaU  to  i  gm.  of  substance).  The 
results  are  summarized  in  Table  III. 

Table  III. — Results  of  feeding  gossypol  to  rabbits 


Rabbit 
No. 

Weight. 

Quan- 
tity of 
gossy- 
pol 
eaten. 

Equi- 
valent 
in  ex- 
tracted 
ker- 
nels. 

Feed- 
ing 
period. 

Diet. 

Initial. 

Final. 

Gain 

or 

loss. 

Result. 

Gossypol   precipitated   from 
ether  solution  by  petroleum 
ether. 

Do     

990 

996 
998 

994 

1,001 

21 

22 

17 
18 

19 

Grams. 
1,490 

2,440 

i»o7S 

2,330 

975 

2,000 

1,500 

950 
850 

700 

Grams. 

1,205 

2,10s 
850 

2, 180 

820 

1,900 

1,440 

800 
750 

590 

Grams. 

-28s 

-335 

-225 
-150 

-155 

—  100 

—  60 

-150 

—  100 

—  no 

Grams. 
1. 81 

2.61 
.68 

.87 

•47 

0.4 

•33 

•433 
•333 

-153 

Grams. 

200 

290 

75 

97 

53 

Days. 

24 

26 
14 

14 

8 

4 

4 

Died. 
Do. 

Do 

Do. 

RecrystaUized  gossypol  "ace- 
tate." 
Do                  

Do. 
Do. 

Gossypol  free  from  acetic  acid 

(o.i  gm.  daily  per  animal, 

equivalent  to  about  17  gm. 

of-raw  cottonseed  kernels). 

Do     

Died  isth  day. 
Died. 

Do. 

Do 

Died  nth  day. 

Sodium  salt  of  gossypol 

Died  13th  day. 

A  peculiar  feature  about  the  effect  of  gossypol  and  oftentimes  of 
cottonseed  kernels  is  that  the  animals  may  eat  these  substances  for 
several  days  without  being  affected,  then  they  may  suddenly  cease  eating, 
waste  away,  and  finally  die.  This  was  the  case  with  rabbits  21  and  22 
in  this  experiment. 

In  our  pre\dous  paper  (77)  we  described  the  nontoxic  product  obtained 
by  oxidation  of  gossypol  by  action  of  air  on  its  alkaUne  solution.  This 
oxidation  product  may  also  be  formed  to  some  extent  in  the  cooking  of 
cottonseed,  but  there  is  no  evidence  of  it.     The  meal  still  contains  con- 


Jan.  14, 19x8        Gossypol,  the  Toxic  Substance  in  Cottonseed  89 

siderable  amou  nts  (about  i  per  cent)  of  a  substance  which  we  have  called 
"D-gossypol." 

"D-gossypol"  is  very  slightly  soluble  in  ether.  For  rabbits  ether 
extraction  does  not  render  the  meal  nontoxic.  But  where  we  find  in  the 
meal  after  6  hours'  extraction  with  ether  considerable  amounts  of  a 
substance  giving  color  reactions  for  gossypol  or  "D-gossypol,"  it  would 
seem  that  the  substance  is  bound  in  some  way.  To  explain  this  we  have 
assumed  that  it  may  be  combined  with  the  protein  or  some  other  con- 
stituent. There  is  some  evidence  of  this  in  the  properties  of  these  sub- 
stances. We  have  mentioned  that  gossypol  combines  with  anilin  and 
with  acetic  acid  to  form  less-soluble  compounds.  Possibly  similar  com- 
bination may  take  place  with  free  amino  and  free  carboxyl  groups  in  the 
protein  molecule.  Marchlewski  (<?)  mentioned  the  fact  that  gossypol 
behaves  like  tannin  toward  basic  dyes.  Tannin  also  precipitates  pro- 
teins as  insoluble  compounds.  A  similar  combination  of  gossypol  with 
protein  may  occur  in  the  cooking  of  cottonseed. 

Confirmatory  evidence  that  these  substances  may  be  the  cause  of 
cottonseed-meal  poisoning  is  given  in  the  previous  publications  of  this 
Station.  Thus,  when  cottonseed  meal  is  treated  with  an  alcoholic 
alkali  (jp),  the  meal  is  rendered  nontoxic  to  rabbits.  When  the  meal 
is  fed  with  iron  salts  to  pigs  {18)  and  rabbits  {16) ,  the  toxic  effect  is  greatly 
diminished.  Pigs  fed  on  cottonseed  meal  and  corn  meal  (1:3)  with 
ferrous  sulphate  (copperas)  did  not  die  in  180  days,  whereas  without 
ferrous  sulphate  all  the  animals  died.  Rabbits  were  fed  106  days  with 
ferric  ammonium  citrate  without  harmful  results.  We  have  explained 
this  by  assuming  that  the  alkali  treatment  promoted  oxidation  of  the 
gossypol  and  by  assuming  that  the  iron  salt  formed  an  insoluble  pre- 
cipitate with  the  gossypol,  or  possibly  assisted  the  organism  to  oxidize  it. 

The  experiments  referred  to  in  this  article  support  our  previous  view 
that  gossypol  is  toxic  and  that  it  is  the  only  toxic  substance  in  the  raw 
kernels.^  Extensive  experiments  with  various  meals  with  rats,  rabbits, 
fowls,  and  swine  show  that  there  still  remains,  even  in  thoroughly  cooked 
meals,  an  injurious  factor.  Such  thoroughly  cooked  meals  are  harmful 
to  rabbits  and  swine,  but  seem  to  have  little  effect  on  rats  and  fowls  when 
fed  on  adequate  diets. 

PRELIMINARY  EXPERIMENTS  WITH   PIGS 

In  order  further  to  test  the  correctness  of  our  view  that  cottonseed-meal 
injury  is  due  to  a  toxic  substance  rather  than  to  dietary  deficiencies,  we 
have  conducted  a  few  preliminary  experiments  with  small  pigs. 

It  seemed  desirable,  in  view  of  the  extreme  position  taken  by  Rommel 
and  Vedder  (14)  to  ascertain  (i)  whether  gossypol  is  toxic  to  pigs;  (2) 

'  The  results  of  other  experiments  showing  that  cooking  exerts  a  profound  influence  on  the  toxicity  will 
be  published  at  an  early  date. 


90 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  2 


whether  extraction  of  gossypol  from  cottonseed  by  a  solvent  renders  the 
residue  nontoxic;  (3)  whether  by  the  addition  of  vitamine-containing 
feeds,  cottonseed-meal  poisoning  can  be  averted. 

Four  small  Duroc-Jersey  pigs  were  confined  in  small  pens  about  3.5  by 
8  feet.  The  pens  had  a  concrete  floor  which  was  bedded  with  pine 
shavings.  The  water  used  was  secured  from  the  city  mains.  The  pigs 
were  fed  the  diets  given  in  Table  IV. 


Table  IV. — Percentage  composition  of  diets  for  pigs 


Pig  I. 

Pig  2. 

Pig3- 

Pig4. 

Feed. 

Period  i 

(ist-28th 

day). 

Period  2 

(29th-38th 

day). 

Period  3 

(39th-45th 

day). 

Cottonseed  meal 

25 

75 

21.  7 

30 

45-5 
45-5 

Com  meal 

49-  75 
0.  25 

50 

75 

Gossypol 

Wheat  bran 

25 

65 

60 

Ether-extracted  cottonseed  kernels 

Milk  (solids)        

13 
(a) 

ID 

(a) 

A 

Green  feeds 

a  About  one-half  pound  daily. 

Pig  2  received  at  the  start  1.22  gm.  of  gossypol  daily.  This  figure  is 
based  on  yields  (about  i  per  cent)  of  crude  crystalline  gossypol  acetate, 
obtained  from  "oil-free"  cottonseed  kernels.  Gossypol  "acetate" 
equivalent  to  the  required  amount  of  gossypol  was  dissolved  in  ether;  the 
acetic  acid  present  was  removed  from  the  ether  solution  by  agitation  with 
water.  The  ether  solution  of  gossypol  was  then  spread  over  a  part  of 
the  corn  meal  and  the  ether  evaporated.  This  was  no  doubt  an  unneces- 
sary procedure,  as  we  have  found  no  difference  in  the  action  of  the 
"acetate"  and  the  "free"  gossypol. 

Pig  3  was  fed  on  cottonseed  kernels  from  which  practically  all  the 
gossypol  had  been  removed  by  percolation  with  ether. 

Pig  4  was  fed  with  a  view  to  supplying  any  deficiency  of  vitamines  in 
the  cottonseed  meal  by  wheat  bran,  whole  milk,  and  some  green  food, 
chiefly  leguminous.  It  is  not  possible  to  give  the  exact  composition  of 
the  ration  of  this  pig.  The  green  feed  (about  J/z  pound  daily)  was  not 
always  consumed.  By  disregarding  the  green  feed  eaten  and  by  assuming 
that  the  milk  contained  12  per  cent  of  solids,  the  composition  of  the  diet 
was  approximately  as  given  above. 

Figures  i  and  2  and  Tables  V  to  VII  summarize  the  important  data  of 
the  experiment. 


Jan.  14.  i9i8  Gossypoly  the  Toxic  Substance  in  Cottonseed 


91 


Fig.  I. — Graphs  of  the  growth  of  pigs  i,  2,  and  3.    "C.  S.  K."=^  cottonseed  kernels. 


•^^ 


Fig.  2.— Graphs  of  the  gains  per  week  of  pigs  i,  2,  and  3.    VC.  S.  M."=  cottonseed  meal. 


92 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  3 


Table  V. — Results  of  feeding  various  diets  to  pigs 


Pig 

No. 


Feed. 


Cottonseed  meal 

(jossypol 

Ether-extracted  cottonseed  kernels 
(Cottonseed  meal,  milk,  etc 


Weight. 


Initial. 


Pounds. 

29-5 
28.  25 
20.7s 

19-5 


Final. 


Pounds. 

31-5 
27 

«  44-  75 
35 


Pounds. 

2 
-1.25 
24 

15-5 


Result. 


Died  50th  day. 
Died  48th  day. 
Lived. 
Died  45th  day. 


»  Weight  of  pig  3  on  the  50th  day.    This  pig  weighed  128  pounds  when  the  experiment  was  discontinued 
on  the  isstb  day,  a  daily  gain  of  0.69  pound. 

Table  VI. — Weight  of  cottonseed  feed  consumed,  by  weeks  « 


Week. 

Pig  I 
(cotton- 
seed 
meal). 

Pig  2  (gossypol).& 

(extracted 
cotton- 
seed ker- 
nels). 

Pig  4 
(cotton- 
seed 
meal). 

1            

Pounds. 
1.68 

1-7 

1.63 

1.58 

1.36 

1.27 

I.  04 

Gm. 

8.37(1-82) 
9.  76  (2.  15) 
8.04(1.77) 
6.  76  (i.  50) 
4.  06  (  .  90) 
5-56(1.22) 
3-5o(  -77) 

Pounds. 
1-3 

1-5 
1-75 
2.  18 
2.  18 
2.63 
2-53 

Pounds. 

1-3 
I.  52 

1-43 
I.  62 

c    

2.  14 

6 

2.8 

7 

I.  I 

Total  cottonseed  feed  eaten 

Total  feed  eaten 

10.  26 
41.  0 

49.  95  (lo-  13) 
40.  5 

14.07 
56-3 

II.  91 
46.  0 

o  Maximum  estimates  are  given  for  pigs  i,  2,  and  4. 

6  Figures  in  parentheses  give  the  weight  (in  pounds)  of  oil-free  kernels,  which  correspond  to  the  gossypol 
eaten. 

Table  VII. — Comparison  of  post-mortem  notes  on  pigs  i,  2,  and  4<^ 


Organ,  etc. 

Pig  I. 

Pig  2. 

Pig4- 

Lungs 

Congested,     edema- 
tous. 

Thrombus 

Congested,     edema- 
tous. 

Extremely     edema- 

Heart 

tous,    with    some 
congestion. 
Thrombus. 

Chest  cavity .... 

Abdominal  cav- 
ity   

Small  intestines  . 

Penis 

2  to  3  ounces  of  serous 
fluid. 

Slight  excess  of  fluid . 

Considerable    injec- 
tion of  blood  ves- 
sels. 

Sheath  swollen 

Poor 

4  ounces  of  fluid 

Slight  excess  of  fluid . 
Deeply  injected 

Sheath   swollen,   or- 
gan paralyzed,  and 
protruding. 

Very  poor 

About  16  ounces  of 
fluid. 

Slight  excess  of  fluid. 
Inflamed  areas. 

Sheath  swollen. 

Nutrition. .   . 

Ck)od. 

o  These  pigs  were  examined  by  Dr.  G.  A.  Roberts,  Veterinarian  of  this  Station. 


Jan.  14, 1918        Gossypol,  the  Toxic  Substance  in  Cottonseed  93 

GENERAL  DISCUSSION 

For  the  first  few  days  of  the  experiment  all  the  pigs  ate  well,  and  all 
gained  in  weight.  Pigs  i  and  2  occasionally  left  part  of  their  feed.  In 
two  weeks'  time  all  except  pig  3  began  to  show  loss  of  appetite  and  reg- 
ularly left  a  portion  one-half  to  one-fourth  of  their  feed.  On  the 
twenty-fifth  day  the  pig  2  (fed  gossypol)  was  quite  sick  and  not  able  to 
walk  well.  At  this  time  pig  3  was  the  thriftiest  of  the  four,  while  the 
rations  of  No.  2  and  3  were  reduced  on  account  of  refusal  to  eat.  On 
the  twenty-ninth  day  the  ration  of  pig  4  was  changed  to  i  part  of  cotton- 
seed meal  and  2  parts  of  bran.  On  the  thirty-second  day  the  feed  of 
pig  2  was  changed.  The  wheat  bran  was  replaced  by  middlings,  for  which 
the  pig  had  a  better  appetite.  On  the  thirty-ninth  day  the  wheat  bran 
in  the  ration  of  pig  4  was  replaced  by  corn  meal,  the  pig  getting  equal 
parts  of  cottonseed  meal  with  corn  meal.  At  that  time  this  pig  was 
leaving  one-half  to  two-thirds  of  the  wheat-bran  mixture.  For  three  or 
four  days  he  ate  the  new  mixture  with  much  better  appetite,  but  then 
refused  a  large  part  and  died  on  the  forty-fifth  day.  Pig  3  maintained 
perfect  appetite  up  to  the  forty-ninth  day,  when  she  did  not  clean  up 
the  last  trace  of  feed  as  usual.  When  removed  to  the  yard  to  be  photo- 
graphed, it  was  noted  that  she  had  an  abnormal  gait  in  walking,  the 
forelegs  showing  a  tendency  to  double  under  her.  This  animal  seemed 
to  have  a  great  desire  to  eat  dirt,  manure,  etc.  However,  on  being 
allowed  the  freedom  of  a  large  lot,  the  animal  soon  recovered.  She 
was  given  a  little  ferrous-sulphate  solution,  chalk,  and  milk  on  the 
fiftieth  day.  Her  normal  appetite  returned,  and  in  three  or  four  days 
she  was  able  to  trot.  At  no  time  did  this  pig  show  the  rough  coat  and 
lack  of  appetite  that  characterized  the  others.  On  the  fifty-second  day 
the  feed  of  cottonseed  was  increased  slightly,  and  the  pig  received  about 
one-half  pint  of  milk  daily  for  the  following  nine  days.  Whether  this 
pig  was  suffering  from  deficiency  of  some  sort  in  the  ration,  from  lack  of 
exercise,  or  from  the  daily  intake  of  a  small  amount  of  gossypol  in  the 
kernels,  we  are  not  able  to  say.  Even  granting  that  this  pig  had  a  slight 
attack  of  beriberi,  we  can  reasonably  conclude  from  the  experiment 
that  the  deficiency  factor  is  one  quite  secondary  to  the  toxicity  factor. 
Plate  I  shows  the  condition  of  these  pigs  at  various  periods  of  the  ex- 
periment. 

This  preliminary  experiment  was  originally  planned  to  run  for  a  short 
period,  but  as  the  pig  on  the  extracted  kernels  seemed  to  be  in  a  path- 
ological condition,  in  that  the  forelegs  tended  to  double  under  her  when 
she  attempted  to  run,  it  seemed  desirable  to  continue  the  animal  on  this 
diet.  At  this  time,  when  pig  i,  which  had  been  given  the  diet  most 
closely  resembling  that  of  pig  3  in  chemical  composition,  had  died  in  a 
rather  emaciated  condition,  pig  3  was  a  plump,  very-well  nourished 
animal  (PI.  i,  B).     It  is  quite  possible  that  the  above-mentioned  condi- 


94 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  a 


tion  was  due  both  to  a  slight  toxicity  of  the  extracted  kernels  and  to  the 
restricted  diet.  This  condition  was  remedied,  as  previously  described. 
The  animal  was  continued  on  the  outdoor  turf  lot  for  35  days  and  then 
removed  to  a  small  indoor  pen,  where  for  70  days  longer  she  gained 
steadily. 

Throughout  the  experiment  the  daily  feed  was  maintained  at  i  per 
cent  (of  body  weight)  of  extracted  kernels  plus  3  per  cent  of  corn  supple- 
ment. 


ysc? 


A/O 


\ 

i.  <so 


\ 


'^O 


^o 


: 

•' 

'   y 

/v 

/ 

y^ 

y 

y' 

lO 

,  "S 

s-... 

^ 
^ 

"X" 

.'' 

^  £> 

'ifi? 

4 

y 

/ 

[1 

e 

/' 

A 

^ 

p 

«/■ 

^' 

/ 

•c 

<< 

• 

y 

i^- 

/ 

4- 

r 

4 

) 

.ft 

> 

,fi' 

A 

y 

f  r 

,.5- 

i-/ 

'' 

.(? 

U 

7 

u^ 

?'-' 

■ih 

^ 

/ 

c 

\0 ,  •■• 

V'' 

t^ 

V 

V 

r 

■~/ 

ix 

r 

..•o 

^ 

^ 

^  ^    e*  3  /o 


Fig.  3. — Graphs  of  the  growth  of  pigs  3,  s,6,7,and8.    "  C.S.  M."=cottonseedmeal:  "EE.C.S.K."= 
ether-extracted  cottonseed  kernels. 

Two  Other  small  pigs,  No.  7  and  8,  were  also  fed  on  this  diet  of  ex- 
tracted kernels  and  corn  meal  on  the  same  basis  as  in  the  previous  case. 
On  a  few  days  it  was  necessary  to  use  whole  or  cracked  corn  in  place  of 
corn  meal,  owing  to  a  lack  of  meal.  The  results  in  general  were  the  same. 
There  was  fair  growth,  but  after  six  or  eight  weeks  the  pigs  lost  their 
keen  appetites  and  developed  a  tendency  to  squat  on  their  hindquarters 
and  to  walk  stiffly.  The  growth  curves  in  figure  3  show  the  general 
results.  The  gain  was  fair,  0.581  pound  per  day  for  the  female  and 
0.556  pound  for  the  male  pig.  The  experiment  was  discontinued  on 
the  eighty-eighth  day. 


Jan.  14, 1918  Gossypol,  the  Toxic  Substance  in  Cottonseed 


95 


The  female  pig  was  observed  for  a  few  weeks  after  the  experiment. 
The  animal  continued  to  increase  in  weight,  but  still  retained  the  squat 
ting  tendency.     Addition  of  small  amounts  of  milk  and  outdoor  exer- 
cise did  not  eliminate  this  condition.     The  nutrition  of  the  animal  was 
excellent  throughout. 

In  view  of  the  slight  toxicity  of  the  extracted  kernels,  as  shown  in  our 
rabbit  experiments,  it  is  possible  that  this  slightly  pathological  condition 
may  be  due  to  a  toxic  factor,  although  with  such  very  young  pigs  it  may 
be  due  to  the  limitations  of  this  food  mixture. 

No  doubt  if  these  pigs  had  been  continued  on  this  diet  confined  to 
pens,  they  would  have  ultimately  failed,  as  do  swine  fed  on  restricted 
diet  of  cereal  grains  (see  Hart  and  McCollum,  2).  This  phenomenon, 
however,  should  not  be  confused  with  what  is  commonly  understood  as 
cottonseed-meal  "injury"  or  poisoning. 

We  do  not  claim,  however,  that  this  diet  is  an  adequate  one,  and 
it  is  quite  possible  that  the  condition  described  was  due  in  large  part 
to  the  inadequacy  of  certain  dietary  factors.  Just  what  factors  are  in- 
sufficient in  this  particular  diet  is  not  at  present  apparent.  In  consid- 
ering this  question  we  have  taken  the  view,  tentatively,  that  the  supply 
of  vitamines  in  cottonseed  is  similar  to  that  of  other  seeds  and  that 
the  mineral  content  is  very  much  better. 

The  fact  that  cottonseed  is  cooked,  and  subsequently  pressed,  raises 
the  question.  Are  the  vitamines  thus  rendered  partially  inactive  or  re- 
moved by  the  crude  oil  ?  In  answer  to  this,  we  may  point  to  the  excel- 
lent growth  of  rats  reported  by  Richardson  and  Green  (//,  12,  13,)  and 
Osborne  and  Mendel  {10)  as  evidence  that  even  the  cooked  meal  is  as 
well  supplied  with  vitamines  as  any  similar  vegetable  food. 

With  the  dietary  factors  more  favorable,  as  in  the  cottonseed-meal, 
experiment  to  be  described,  probably  the  pigs  would  not  have  manifested 
this  stiflFness  of  gait  and  squatting  tendency.  It  is  also  of  interest  to 
compare  the  mineral  content  of  a  diet  of  cottonseed  meal  and  com  meaJ 
(i  :3)  with  the  ash  content  of  a  diet  found  successful  for  growth  in  rat? 
(vSee  Table  VIII.) 

Table  VIII. — Average  mineral  content  of  a  diet  containing  I  part  of  cottotiseed  meal  ard 
J  parts  of  corn  meal;  also  the  analysis  of  m,inerals  of  other  materials 


Feed. 


Cottonseed  meal 

GDm  meal 

Average  mixttire  1:3 
Pig  and  rat  diet  A  (2) 

Diet  BO) 

Dry  skim  milk  (/).... 


Pro- 
tein. 


36 

87 
15-5 


Ash. 


Sodi- 
um. 


O.  26 
113 
14 
022 
029 


Potas- 
sium. 


1.66 

•  19 

•  55 
•335 
.  076 

I.  27 


Calci- 

lltQ. 


o.  27 

.015 
.08 
.266 
.080 

1-34 


Magne- 
sium. 


0-55 
.  122 
.  22 

•275 
.  oog 
.  146 


Chlorin 


04 
07 

055 
041 

057 

935 


Phos- 
phorus. 


51 


35 
264 

248 
186 
979 


S'lV 
phur. 


o.  49 
.  20 
.  20 
.089 
.  141 
•357 


96  Journal  of  Agricultural  Research  voi.xii.No.a 

Rat  diet  A  represents  the  mineral  content  of  a  diet  which  was  found 
successful  for  growth  of  pigs  on  artificial  rations  by  Hart  and  McCollum 

(2). 

Rat  diet  B  represents  the  mineral  content  of  a  diet  which  did  not  pro- 
duce pronounced  stunting  in  rats  (4).  The  diet  was  satisfactory  in  other 
factors. 

The  recent  work  of  Hogan  (5)  and  of  McCollum,  Simonds,  and  Pitz  (5) 
indicates  that  the  deficiencies  of  corn  lie  in  the  poor  ash  content,  in  poor 
proteins  and  low  amounts  of  protein,  and  inadequate  amounts  of  fat 
soluble  A. 

Since  our  pigs  on  ether-extracted  kernels  plus  corn  meal  made  much 
better  gains  than  could  be  expected  on  whole  com  alone  under  these 
conditions,  our  opinion  is  confirmed  that  the  addition  of  this  cottonseed 
feed  to  a  corn  diet  furnishes  a  greatly  improved  protein  and  a  mineral 
basis  for  nutrition.  It  is  then  evident  that,  where  death  ensues  or  poor 
growth  is  manifested,  this  is  a  result  of  an  injurious  substance  rather 
than  of  dietary  deficiencies. 

FURTHER  EXPERIMENTS  WITH  GOSSYPOL 

Gossypol  was  fed  to  two  other  pigs  of  approximately  50  pounds'  weight. 
In  one  case  a  rather  large  amount  was  given  in  the  ration.  Pig  5  ate  a 
slop  made  of  i  pound  of  a  mixture  of  corn  meal  and  soybean  meal  (1:3) 
containing  4.5  gm.  of  gossypol.  The  pig  showed  a  poor  appetite  for  the 
same  amount  given  the  next  morning,  but  ate  it  slowly  after  a  pint  of 
milk  was  poured  into  it.  Next,  a  half  dose  (containing  2.3  gm.)  was 
offered,  one-half  of  which  was  refused.  For  the  next  five  days  the  pig 
was  offered  smaller  doses,  but  refused  it  all  or  in  part,  even  when  tempted 
to  eat  by  using  bran  and  milk  in  the  feed.  The  pig  showed  a  good  appe- 
tite for  other  feed  that  did  not  contain  the  gossypol  mixture.  On  the 
fifth,  sixth,  and  seventh  days  the  pig  showed  no  appetite  at  feed  time 
but  seemed  sluggish  and  showed  a  desire  to  lie  down.  The  pig  lost  in 
weight.  The  experiment  was  then  stopped.  A  control  pig,  No.  6,  fed 
on  a  similar  ration  without  gossypol  showed  a  good  appetite  and  devel- 
oped no  such  symptoms. 

After  a  rest  of  six  days,  the  previous  control  pig.  No.  6,  was  used 
for  a  gossypol  experiment,  and  No.  5  served  as  the  control.  Gossypol 
was  fed  in  amounts  equivalent  to  that  in  0.5  to  0.6  pound  of  oil-free 
cottonseed  kernels,  approximately  2.2  gm.  daily.  An  ether  solution  of 
gossypol  was  dried  on  corn  meal  and  this  was  mixed  with  more  corn 
meal  and  wheat  bran.  The  other  pig  received  the  same  ration  without 
gossypol.  The  gossypol  pig  ate  practically  all  its  feed  for  a  week,  and 
then  began  to  show  a  poor  appetite  for  it  and  refused  part.  On  the 
eighth  day  one-half  the  feed  was  left.  On  the  ninth  and  tenth  days  the 
pig  ate  scarcely  any  feed.  From  the  eleventh  to  the  fifteenth  day  the 
animal  was  given  dry  recrystallized  gossypol  acetate  (10  per  cent  acetic 


Jan.  14,  i9i8  Gossypol,  the  Toxic  Substance  in  Cottonseed  97 

acid)  as  a  finely  crystalline  powder  mixed  v/ith  the  feed  in  place  of  gossypol 
evaporated  on  corn  meal.  The  pig  ate  it  for  one  day,  and  then  refused  it 
as  before.  By  using  com  meal  only  the  animal  ate  one  more  feed,  but 
refused  it  when  repeated.  During  the  last  few  days  of  the  experiment 
the  goss3^1  animal  seemed  to  be  growing  weaker  and  very  preceptibly 
thinner.  In  the  last  week  the  animal  lost  3  pounds^  while  the  control 
gained  2  pounds.  This  experiment  was  discontinued  on  the  fifteenth 
day  because  of  the  refusal  of  the  animal  to  eat. 

In  both  of  these  cases  of  feeding  gossypol  it  was  very  evident  that  the 
animals  were  physiologically  afifected  at  an  early  date.  All  told  not 
over  15  gm.  of  gossypol  were  eaten  in  the  first  case  and  not  over  22  gm. 
in  the  second  case.  These  amounts  of  gossypol  are  equivalent  to  about 
4  and  6  pounds  of  oil-free  cottonseed  kernels,  respectively. 

Besides  the  direct  proof  of  the  existence  of  a  toxic  substance  in  cot- 
tonseed meal,  we  have  a  strong  argument  against  the  deficiency  theory 
in  the  results  of  feeding  rabbits  on  cottonseed  meal  treated  with  boiling 
alcoholic  alkali.  This  treatment,  which  would  be  expected  to  destroy 
the  natural  vitamines  present,  so  changes  the  meal  that  it  becomes  non- 
toxic to  rabbits.  This  change  in  toxicity  we  have  shown  is  explained 
by  the  ease  with  which  gossypol  undergoes  oxidation  in  alkaline  solution. 
This  was  confirmed  by  feeding  to  rabbits  the  products  formed  by  oxida- 
tion of  gossypol  in  alkaline  solution  by  air. 

CAN  COTTONSEED-MEAL  POISONING  BE  OVERCOME  IN  A  FAVORABLE 
DIET  UNDER  FARM  CONDITIONS? 

The  two  largest  and  oldest  pigs  (No.  5  and  6)  were  fed  in  a  turf  lot 
about  50  feet  square.  Grass  was  abundant,  and  a  good  part  of  the  time 
there  was  water  in  the  lot  from  frequent  rains  which  also  kept  the  turf 
soft.  These  pigs  were  fed  on  a  mixture  of  equal  parts  of  cottonseed 
meal,  corn  meal  or  com,  and  wheat  bran,  with  a  pint  of  milk  apiece  each 
day.  The  cottonseed  meal  fed  each  day  was  about  1.33  per  cent  of  body 
weight,  rather  higher  than  has  been  the  practice  at  this  Station.  It  was 
thought  that  for  a  while,  about  the  fiftieth  day,  the  pigs  acted  somewhat 
suspiciously.  One  showed  a  lack  of  appetite  for  the  mixture.  Its  eyes 
seemed  partly  closed  and  somewhat  watery.  The  pigs  also  seemed  rather 
short-winded.  They  retained,  however,  perfect  control  of  their  limbs 
and  were  able  to  mn  very  well  at  all  times.  Finally,  after  160  days  of 
high  feeding  of  cottonseed  meal,  these  pigs  were  put  under  cover  in  a 
small  pen.  The  pigs  were  soon  eating  sparingly  and  losing  weight.  The 
pig  which  had  acted  somewhat  queerly  at  times  in  the  experiment 
became  sick  and  died  on  the  one  hundred  and  ninety-eighth  day,  show- 
ing typical  symptoms,  although  there  was  also  a  pneumonic  appearance 
of  the  lungs.  The  experiment  with  the  other  pig  was  then  discontinued. 
On  changing  this  pig's  feed  to  corn,  he  began  to  regain  weight.  This 
animal  was  later  fed  corn,  wheat  bran,  meat  scraps,  etc.  (Table  IX.) 
27805°— 18 5 


98 


Journal  of  Agricultural  Research  voi. xii, no.  « 


Table  IX. — Results  of  feeding  cottonseed  meal  to  pigs  5  and  6  under  favorable  conditions. 


Weight. 

Time  of  weighing. 

Pig  6. 

Pigs. 

Gain  per  day. 

Pig  6. 

Pigs- 

Day, 

Pounds. 

55-5 
129 

158 

150 
142 

Pounds. 
50-5 

112.  5 

153-5 
145 
145 
(^) 
158 
C177 

Pound. 

Pound. 

0.8 
.64 

0.67 

9^ 

160 

.65 

'■n 

^■^b  •'o" 

a  Dead. 


6  Discontinued  feeding  cottonseed  meal. 


c  Slaughtered. 


As  long  as  these  pigs  were  kept  in  the  outdoor  lot,  no  marked  symptoms 
of  cottonseed-meal  poisoning  were  noted.  It  was  noted  that  when  kept 
itJkdoors  they  lost  both  appetite  and  weight. 

It  is  evident  from  our  indoor  and  outdoor  experiments  that  the  effect 
of  cottonseed  meal  is  more  severe  on  pigs  kept  in  pens,  a  fact  that  has 
long  been  known ;  however,  past  records  show  that  typical  sudden  deaths 
from  "acute  cottonseed-meal  poisoning"  may  also  occur  among  pigs 
receiving  cottonseed  meal  when  on  pasture.  Such  deaths  may  follow 
excellent  gains  and  may  be  without  previous  sickness,  often  occurring 
when  the  animals  are  exercised  violently.  Consequently,  a  conclusion 
that  the  meal  was  without  effect  during  the  outdoor  experiment  can  not 
be  drawn.  Certainly  the  subsequent  loss  in  weight  after  removal  from 
the  turf  is  suspicious.  It  is  plausible  to  suppose  that  the  outdoor  con- 
ditions stimulate  metabolism  so  that  the  animal  is  enabled  to  overcome 
or  resist  the  injurious  factor.  Possibly  the  difference  in  effect  of  cotton- 
seed meal  on  rats  and  pigs  may  be  in  part  explained  by  the  more  vigorous 
metabolic  activity  of  the  smaller  animal. 

In  a  series  of  three  recent  articles,  Richardson  and  Green  {11,  12,  13) 
have  well  shown  the  high  nutritive  efl&ciency  of  cottonseed  meal  and  flour 
for  rats,  indicating  the  economic  value  of  this  substance.  In  some  points, 
however,  we  believe  that  they  have  misinterpreted  facts.  Thus,  they 
speak  of  the  flour  as  a  "highly  milled"  or  "refined"  product  and  account 
for  the  apparent  slight  nutritive  superiority  of  the  unbolted  meal  over 
the  flour  by  stating  that — 

This  suggests  a  greater  amotmt  of  the  growth-promoting  substance  associated  with 
certain  fats  in  the  less  highly  milled  product. 

It  seems  hardly  possible  in  the  case  of  cottonseed  meal  to  effect  such  a 
change  in  nutritive  value  by  mechanical  means.  We  would  suggest 
that  the  difference  was  accountable  on  the  basis  of  different  conditions 


Jan.  14, 1918  Gossypol,  the  Toxic  Substance  in  Cottonseed 


99 


in  the  cooking  of  the  products,  as  we  have  found  these  to  be  the  greatest 
cause  of  variation  in  toxicity  of  meals. 

Richardson  and  Green  (u,  p.  316),  state  in  conclusion: 

Our  results  indicate  that  cottonseed  meal  does  not  contain  sufficient  minerals  for 
growth,  is  not  actively  toxic,  contains  efficient  protein  and  perhaps  fat-soluble,  growth- 
promoting  substances,  similar  to  those  of  butter  fat,  but  in  less  adequate  quantities. 

Our  own  extensive  unpublished  experiments  on  the  toxicity  of  cotton- 
seed products  indicate  that  the  toxicity  of  cottonseed  meals  varies  with 
the  conditions  of  cooking  the  raw  seed.  While  we  find  that  the  flour 
and  thoroughly  cooked  meals  have  no  apparent  toxicity  for  rats  when 
fed  in  diets  supplemented  by  milk  powder,  these  same  products  fed  in 
unsupplemented  diets  are  inferior  to  ether-extracted  cottonseed  kernels. 
Even  thoroughly  cooked  cottonseed  meals  are  definitely  injurious  to 
rats  and  pigs. 

The  ash  analysis  of  cottonseed  flour  given  by  Richardson  and  Green 
(ii,  12,  ij)  in  each  of  these  three  articles  differs  radically  in  some  respects 
from  that  given  by  Forbes  (i)  (Table  X). 

Table  X. — Ash  analyses  of  cottonseed  flour 


Constituent. 


Inorganic  salts 

Silicic  oxid  (Si02) 

Chlorin 

Sulphur  trioxid  (SO3) 

Phosphorus  pentoxid  (P2O5) 

Potassium  oxid  (KgO) 

Calcium  oxid  (CaO) 

Magnesium  oxid  (MgO) 

Sodium  oxid  (NajO) 


Analysis 
of  cotton- 
seed flour 
according 
to  Rich- 
ardson 

and 
Green. 


Per  cent. 

5-5° 

o.  14 

None. 

.06 

2-57 

2.  01 

.26 

•25 
None. 


Constituent. 


Ash 

Chlorin. . . . 
Sulphur. . . . 
Phosphorus 
Potassium . . 

Calcium 

Magnesium . 
Sodium . . . . 


Analysis 
of  cotton- 
seed meal 
according 
to  Forbes. 


Per  cent. 
7.  629 

.  042 
•536 
1.479 
I.  81 
.  291 

•599 
.283 


Richardson  and  Green's  data  are  from  an  analysis  of  material  left  after 
ignition,  which,  as  is  well  known,  causes  loss  of  elements,  such  as  sulphur 
and  chlorin.  While  the  elements  chlorin  and  sodium  are  not  necessary 
for  plant  growth  and  the  amounts  present  may  vary,  it  seems  hardly 
possible  that  they  are  entirely  lacking  in  cottonseed  flour.  They  are 
certainly  essential  to  animals.  It  may  be  noted  that  the  rats  of  Richard- 
son and  Green  "have  grown  and  maintained  body  weight  for  135  days" 
on  a  diet  containing  cottonseed  flour  as  the  sole  source  of  minerals. 

In  their  second  article  Richardson  and  Green  (12)  have  attempted  to 
repeat  some  of  our  work  with  extracts  of  cottonseed.  Instead  of  using 
unheated  cottonseed  kernels,  as  we  did,  they  used  kernels  heated  to 
120°  C.     Thus,  they  fail  to  find  toxic  the  ether  extract  of  petroleum- 


loo  Journal  of  Agricultural  Research  voi.  xii,  no.  2 

ether-extracted  kernels.  Their  results  may  be  due  to  the  influence  of 
previously  heating  the  kernels  to  120°  C,  thus  possibly  decomposing 
some  of  the  gossypol,  and  to  incomplete  extraction  so  that  the  remain- 
der was  left  in  the  three  fractions,  the  extracts  and  the  residue.  They 
also  assume  that  the  ethyl-ether  extract  of  petroleum-extracted  kernels 
is  always  2  per  cent  of  the  weight  of  the  kernels.  This  is  the  case  only 
with  long-continued  extractions. 

SUMMARY 

Raw  cottonseed  kernels  contain  about  0.6  per  cent  of  gossypol  and  are 
highly  toxic  to  rats.  Ether  extraction  renders  the  material  nontoxic 
and  gives  a  highly  toxic  extract  containing  about  2  per  cent  of  gossypol. 
Gossypol  fed  in  milk  diets  in  amounts  equivalent  to  those  contained  in 
the  raw  cottonseed  diets  has  proved  as  toxic  as  raw  cottonseed.  Gossy- 
pol may  be  quantitatively  removed  from  the  ether  extract  by  precipita- 
tion as  its  insoluble  anilin  compound.  The  extract  is  thus  rendered  non- 
toxic to  rats.  The  insoluble  anilin  compound  of  gossypol  is  not  toxic 
because  of  its  insolubility.  Gossypol  prepared  from  this  compound 
possesses  its  original  toxic  properties. 

Cottonseed  meal  is  much  less  toxic  than  raw  cottonseed,  owing  mainly 
to  the  oxidation  of  gossypol  during  cooking. 

Cottonseed  meal,  ether-extracted  cottonseed,  and  gossypol  have  been 
fed  to  small  pigs  in  pens  under  comparable  conditions.  Cottonseed 
meal  has  been  found  definitely  injurious,  while  the  ether-extracted  raw 
seed  does  not  appear  to  cause  cottonseed-meal  poisoning.  Gossypol  has 
been  found  toxic  to  pigs. 

If  the  presence  of  an  injurious  substance  in  the  meal  is  disregarded,  a 
diet  of  cottonseed  meal  and  corn  meal  has  nutritive  limitations  which 
may,  under  restricted  conditions  of  living,  lead  to  failure  of  pigs  to  thrive. 
Such  failure  is  a  phenomenon  distinct  from  cottonseed-meal  poisoning. 

Outdoor  exercise,  access  to  forage  and  soil,  and  improved  diets  tend 
to  postpone  or  avert  cottonseed-meal  poisoning  of  swine.  The  defi- 
ciency hypothesis  that  cottonseed-meal  poisoning  of  swine  is  similar  to 
beriberi  is  untenable. 

LITERATURE  CITED 

(i)  Forbes,  E.  B.,  Beegle,  F.  M.,  and  Mensching,  J.  E. 

1913.  MINERAL  AND  ORGANIC  ANALYSES  OF  POODS.      Ohio  Agr.  Exp.  Sta.  Bul. 

255,  p.  2II-23I. 

(2)  Hart,  E.  B.,  and  McCollum,  E.  V. 

1914.  INFLUENCE    ON     GROWTH    OF    RATIONS     RESTRICTED    TO     THE    CORN    OR 

WHEAT  GRAIN.     In  Jour.  Biol.  Chem.,  v.  19,  no.  3,  p.  373-395,   11 
charts,  i  pi. 

(3)  HOGAN,  A.  G. 

1916.  THE  NUTRITIVE   PROPERTIES  OF  CORN.     In  Jour.   Biol.   Chem.,   v.   27, 
no.  I,  p.  193-208.     Bibliography,  p.  208. 


Jan.  14, 1918  Gossypol,  the  Toxic  Substance  in  Cottonseed  loi 

(4)  McCoLLUM,  E.  v.,  and  Davis,  Marguerite. 

191 5.  THE   INFLUENCE    OF    THE    COMPOSITION    AND    AMOUNT   OP   THE    MINERAL 

CONTENT  OP  THE  RATION  ON  GROWTH  AND  REPRODUCTION.      In  Jour. 

Biol.  Chem.,  v.  21,  no.  3,  p.  615-643,  11  charts. 

(5)  SiMMONDS,  Nina,  and  PiTz,  Walter. 

1916.  THE    DIETARY    DEFICIENCIES    OP   THE    MAIZE    KERNEL.     In   Jour.    Biol. 

Chem.,  V.  28,  no.  i,  p.  153-165,  10  charts. 
(6) 


I916.    THE    DISTRIBUTION    IN    PLANTS    OF    THE    FAT    SOLUBLE    A,    THE    DIETARY 

ESSENTIAL  OP  BUTTER  FAT.     In  Amer.  Jour.  Physiol.,  v.  41,  no.  3, 
p.  361-375,  2  fig.,  II  charts. 


(7)  

1916.    THE  NATURE  OP  THE  DIETARY  DEFICIENCIES  OF  THE  WHEAT  EMBRYO.      In 

Jour.  Biol.  Chem.,  v.  25,  no.  i,  p.  105-131,  19  charts. 

(8)  MarchlEwski,  L. 

1899.  gossypol,  Ein  bestandteil  der  baumwollsamEn.     In  Jour.  Prakt. 
Chem.  N.  F.,  Bd.  60,  Heft  1/2,  p.  84-90. 

(9)  Osborne,  T.  B.,  and  Mendel,  L.  B. 

1916.  THE  effect  of  THE  AMINO-ACID  CONTENT  OP  THE  DIET  ON  THE  GROWTH 

OF  CHICKENS.     In  Jour.  Biol.  Chem.,  v.  26,  no.  2,  p.  293-300,  i  pi. 
(10) 

1917.  THE  USE  OF  COTTON  SEED  AS  FOOD.     In  Jour.  Biol.  Chem.,  v.  29,  no.  2, 

p.  289-317,  5  charts. 
(11)  Richardson,  Anna  E.,  and  GrEEn,  Helen  S. 

1916.  nutrition  INVESTIGATIONS  UPON  COTTONSEED  MEAL.      I.      In  Jour.  Biol. 

Chem.,  V.  25,  no.  2,  p.  307-318,  5  charts. 
(12) 

1917.  NUTRITION     INVESTIGATIONS     UPON     COTTONSEED     MEAL.      II.      In    JOUT. 

Biol.  Chem.,  v.  3c,  no.  2,  p.  243-258,  13  charts. 
(13) 

I917.    NUTRITION    INVESTIGATIONS     UPON    COTTONSEED    MEAL.      III.      COTTON- 
SEED   PLOUR.      the    NATURE    OF    ITS    GROWTH-PROMOTING    SUBSTANCE 

AND  A  STUDY  IN  PROTEIN  MINIMUM.     In  Jour.  Biol.  Chem.,  V.  31, 
no.  2,  p.  379-388,  4  charts. 

(14)  Rommel,  G.  M.,  and  Vedder,  E.  B. 

1915.  beriberi   AND  cottonseed  POISONING  IN  PIGS.      (PRELIMINARY   NOTE.) 

In  Jour.  Agr.  Research,  v.  5,  no.  11,  p.  489-493. 

(15)  Wells,  C.  A.,  and  Ewing,  P.  V. 

1916.  COTTONSEED  MEAL  AS  AN  INCOMPLETE  FOOD.     In  Jour.  Biol.  Chem.,  V. 

27,  no.  I,  p.  15-25.     References,  p.  24-25. 

(16)  Withers,  W.  A.,  and  Brewster,  J.  F. 

I913.    studies  ON  COTTONSEED    MEAL    TOXICITY.      II.      IRON  AS    AN  ANTIDOTE. 

In  Jour.  Biol.  Chem.,  v.  15,  no.  i,  p.  161-166. 

(17)  and  Carruth,  F.  E. 

191 5.    GOSSYPOL,  the  TOXIC  SUBSTANCE  IN  COTTONSEED  MEAL.      In  Jour.   AgT. 

Research,  v.  5,  no.  7,  p.  261-288.     Literature  cited,  p.  287-288. 
(18) 

1917.  IRON   AS    AN   ANTIDOTE    TO   COTTONSEED   MEAL   INJURY.      In   Jour.  Biol. 

chem.,  V.  32,  no.  2,  p.  245-257,  4  charts. 
(19)  and  Ray,  B.  J. 

1912.   A  method  FOR  the  REMOVAL  OF  THE  TOXIC  PROPERTIES  PROM  COTTON- 
SEED MEAL.     A  PRELIMINARY  REPORT.     In  Scicncc,  n.  s.,  V.  36,  no. 

914,  p.  31-32. 


PLATE  I 
Effect  of  feeding  cottonseed  feeds  to  pigs: 

A. — Pig  3,  showing  condition  on  the  ninety-fourth  day  on  a  feed  containing  ether- 
extracted  cottonseed  kernels. 

B. — Pig  3,  showing  condition  on  the  fiftieth  day. 

C. — Pig  4,  showing  condition  on  the  twenty-seventh  day  on  a  feed  containing 
vitamines. 

D. — Pig  I,  showing  condition  on  the  fiftieth  day  on  a  feed  containing  cottonseed 
meal.    See  figiire  G. 

E. — Pig  2,  showing  condition  on  the  twenty -seventh  day  on  a  feed  containing 
gossypol. 

F. — Pig  3,  showing  condition  on  the  twenty-seventh  day  on  a  feed  containing 
ether-extracted  cottonseed  kernels. 

G. — Pig  I,  showing  condition  on  the  fiftieth  day  on  a  feed  containing  cottonseed 
meal. 

(I02) 


Gossypol,  the  Toxic  Substance  in  Cottonseed 


Plate  1 


Journal  of  Aericultural  Research 


Vol.  XM.  No.  2 


FRUIT-FLY  PARASITISM  IN  HAWAII  DURING  1916 

By  C.  E.  Pemberton,  Assistant  Entomologist,  and  H.  F.  Willard,  Fruit-Fly  Qtiar- 
antine  Inspector,  Mediterranean  Fruit-Fly  Investigations,  Bureau  of  Entomology, 
United  States  Department  of  Agriculture 

Since  the  introduction  of  parasites  of  the  Mediterranean  fruit  fly 
{Ceratitis  capitata  Wiedemann)  into  the  Territory  of  Hawaii  in  191 3,  by 
the  Territorial  Board  of  Agriculture  and  Forestry,  more  or  less  continuous 
notes  have  been  kept,  from  month  to  month,  indicating  the  extent  of 
parasitism  exerted  upon  the  larvae  of  the  fruit  fly  by  these  parasites  and 
by  other  species  subsequently  brought  in.  Papers  presenting  sum- 
maries of  these  records,  separately  for  the  years  1914^  and  1915,'  have 
already  been  published.  Opportunity  for  special  investigations  of  fruit- 
fly  parasites  in  Hawaii  in  191 6  has  made  possible  the  accumulation  of 
much  careful  data  on  fruit-fly  parasitism  during  this  year,  of  much  the 
same  nature  as  that  given  in  the  publications  just  cited,  and  it  is  the  pur- 
pose of  this  paper  to  give,  possibly  more  in  detail,  some  results  of  the  work 
in  1916. 

It  is  felt  that  a  separate  record  of  the  conditions  of  parasitism  as  existing 
in  Hawaii  in  191 6,  three  years  after  the  first  and  probably  the  most 
important  of  these  introductions,  will  be  of  no  little  vaTue  and  interest  to 
entomologists,  by  way  of  comparison  with  the  parasitism  in  1914  and 
1 91 5,  for  interesting  developments  in  connection  with  the  question  of 
general  parasite  introductions  and  as  a  necessary  contribution  to  the  history 
of  fruit-fly  parasitism  in  Hawaii. 

The  tabulation  during  most  of  the  year  of  the  exact  degree  of  infesta- 
tion of  large  quantities  of  host  fruits  of  the  fruit  fly,  from  many  localities, 
has  been  an  important  part  of  this  work  (Tables  I-III).  All  fruits  col- 
lected for  parasitism  records  on  their  contained  maggots  have  been 
accurately  counted  and  placed  in  separate  boxes  over  sand.  The  fruit 
is  then  kept  in  the  boxes  until  practically  all  the  contained  fruit-fly 
larvae  have  developed,  emerged,  and  entered  the  sand  below  for  pupation. 
A  record  of  the  total  number  of  larvae  thus  developing  and  pupating  is 
secured.  The  larvae  quickly  pupate  after  leaving  the  fruit.  The  pupae 
are  placed  in  vials  and  later  carefully  counted  in  determining  the  degree 
to  which  they  have  been  parasitized.  The  filuig  of  exact  data  of  this 
nature  from  year  to  year  is  necessarily  the  most  reliable  and  positive 
method  of  ascertaining  the  actual  degree  to  which  the  parasites  now 

'  Back.  E.  A.,  and  Pemberton,  C.  E.  parasitism  among  the  larv.E  of  the  mediterranean  fruit 
FLY  (c.  capitata)  IN  HAWAU  IN  1914.  In  Bien.  Rpt.  Bd.  Comrs.  Agr.  and  Forestry  Hawaii.  1913/14.  P- 
ISJ-161.     1915. 

S parasitism  among  the  I^ARViB  op  the  mediterranean  fruit  fly  (c.  capitata)  in  HAWAU 

IN  1915.     /n  Jour.  Econ.  Ent..  V.  9.  no.  3,  p.  306-311.     1916. 


Journal  of  Agricultural  Research,  Vol.  XII,  No.  a 

Washington,  D.  C.  J*°-  ■■*.  1918 

Is  Key  No.  K— 59 

(103) 


I04  Journal  of  Agricultural  Research  voi.  xn.  no.  a 

established  are  contributing  toward  a  control  of  the  fruit  fly.  Such  data 
not  only  indicate  the  extent  of  fruit  infestation  from  various  localities, 
but  also  the  amount  of  parasitism  among  the  larvae  from  month  to  month 
and  the  seasonal  efficiency  of  each  parasite. 

Seasonal  diflferences  in  the  value  and  prolificness  of  certain  species  of 
the  introduced  parasites  have  been  most  striking.  This  is  suggested  by 
an  examination  of  any  of  the  parasite  notes  from  almost  any  locality 
and  by  a  comparison  of  emergences  of  different  species  for  each  month 
of  the  year.  But  most  convincing  proof  that  seasonal  differences  exist 
is  obtained  by  the  inspection  of  records  from  fruit  collected  from  the  same 
localities  month  by  month.  Some  species  of  trees  in  Hawaii  bear  fruit, 
normally  a  host  of  the  fruit  fly,  almost  continuously  throughout  the  year. 
The  systematic  collection  of  fruit  from  such  trees  and  the  filing  of  exact 
data  bearing  on  the  extent  of  parasitism  of  fruit-fly  larvae  secured  from 
such  individual  trees  throughout  the  year  have  thrown  most  light  upon 
the  seasonal  values  of  the  different  parasites. 

Mention  of  possible  fluctuations  in  the  abundance  of  different  species 
has  already  been  made.^  The  work  of  191 6,  wherein  it  has  been  possible 
to  concentrate  parasites  in  fruit  collections  from  individual  trees,,  has 
most  impressively  shown  the  rise  of  the  parasite  Diachasma  tryoni 
Cameron  in  the  summer  and  fall  of  the  year  and  its  certain  decline  during 
the  winter  and  particularly  the  spring  months.  Changes  in  temperature, 
of  no  great  magnitude,  alone  seem  responsible  for  this.  The  parasite 
Opius  humilis  Silvestri,  more  hardy  and  prolific  than  any  of  the  other 
introduced  species,  has  been  overshadowed  by  the  other  species,  partic- 
ularly by  D.  tryoni,  and  has  had  its  seasonal  rise  and  fall  directly  the 
reverse  and  entirely  dependent  upon  the  rise  and  fall  of  this  species  of 
Diachasrha.  The  slight  seasonal  changes  have  little  visible  effect  upon 
the  activities  of  Opius  humilis,  however,  for  in  the  winter  and  spring, 
with  the  decrease  in  abundance  of  D.  tryoni,  it  rapidly  ascends  and  be- 
comes the  most  effective  check  upon  the  fruit  fly  (Table  III).  These 
interrelations  are  treated  elsewhere  by  the  writers. 

The  problem  of  control  of  the  fruit  fly  in  Hawaii  through  parasites  is 
only  partially  solved.  The  four  species  already  established  are  accom- 
plishing a  certain  control,  particularly  in  the  coffee  districts,  but  a  casual 
survey  of  the  extent  of  infestation  of  most  host  fruits  as  shown  in  Table  I 
will  convince  one  of  the  continued  destructiveness  of  this  pest  in  Hawaii. 
An  average  parasitism  of  40  per  cent  of  all  of  the  larvae  developing  is, 
numerically  considered,  of  much  importance;  but  from  the  standpoint 
of  the  practical  needs  of  the  horticulturist  it  brings  little  relief. 

>  Back,  E.  A.,  and  Pemberton,  C.  E.  parasitism  among  the  larv.b  of  the  mediterranean  fruit 
Fi,y  (c.  capitata)  in  Hawaii  in  1915.     /«  Jour.  Econ.  Ent.,  v.  9,  no.  2,  p.  306-311.     1916. 


Jan.  14.  1918      Fruit-Fly  Parasitism  in  Hawaii  during  19 16 


105 


Table  I. — Extent  of  infestation  of  host  fruits  by  larvcB  of  Ceratitis  capatata  in  Hawaii 

during  igi6 


Host  fruit. 


Number  of 
fruits  col- 
lected. 


Number  of 

larvsE  of  C. 

capitala 

emerging. 


Average 

number  of 

larvpp  per 

•fruit. 


Kamani  ( Terminalia  catappa) 

Mango  (Mangifera  indica) 

Coffee  {Coffea  arabica) 

Strawberry  guava  {Psidium  cattleianum) 
Black  myrobalan  {Terminalia  chebula). . 

Peach  (Prunus  persica) 

Rose-apple  (Eugenia  jambos) 

Chrysophyllum  monopyrenum 

Brazilian  plum  (Eugenia  braziliensis) .  .. 

French  cherry  (Eugenia  uniflora) 

Mimusops  elengi 

Ochrosia  elliptica 

Kamani  (Calophyllum  inophyllum) 

BestiU  (Thevetia  neriifolia) 

Averrhoa  carambola 

Chinese  orange  (Citrus  japvnica) 

Noronhia  emarginata 

Guava  (Psidiu^n  guajava) 


15.723 
1.317 

41,  605 

13.825 
6,615 
669 
1,258 
1.956 
4,398 
7,  627 

11,883 

77 
218 

1.532 
159 
1.588 
5.296 
I.  791 


149.415 

2,291 

21, 224 

22, 017 

46,  639 

13, 738 
7,001 

4.034 

3,808 

6,  617 

63,017 

240 

737 

5,540 

214 

4, 843 

9.032 

12,  248 


Table  I  shows  few  fruits  that  have  a  yearly  average  infestation  of  less 
than  two  larvae  per  fruit.  Considering  the  large  quantity  of  fruits  col- 
lected during  the  year,  from  which  the  records  have  been  made,  an  aver- 
age of  two  larvae  per  fruit  is  high.  It  means  that  great  numbers  of  fruits 
from  all  localities  are  nearly  always  heavily  infested.  The  fruits  col- 
lected for  such  a  record  are  not  selected  with  the  purpose  of  obtaining  only 
heavily  infested  fruits  or  only  sound  fruits.  All  available  fruits  that  have 
matured  are  gathered  and  brought  in  whether  infested  or  not.  In  this 
manner  the  exact  average  condition  of  fruit-fly  abundance,  injury,  and 
parasitism  is  obtained. 

To  refer  again  to  the  tables  particular  attention  should  be  called  to  the 
mango  (Mangifera  indica),  guava  (Psidium  gtiajava),  Mimusops  elengi, 
Noronhia  emarginata,  and  the  Chinese  orange  (Citrus  japonica).  Great 
numbers  of  fruit-fly  larvae  develop  in  these  fruits  and  are  but  slightly 
parasitized,  as  shown  in  the  total  column  for  these  fruits  in  Table  11. 
Certain  characters  of  these  fruits  prevent  the  parasites  from  reaching  the 
larvae  within.  This  in  part  accounts  for  the  constant  presence  of  this 
pest,  in  spite  of  the  establishment  of  parasites  well  adapted  to  the  condi- 
tions of  the  country  and  of  great  prolificness. 


io6 


Journal  of  Agricultural  Research  voi.  xir,  no.  a 


Table  II. — Percentage  of  larval  parasitism  of  Ceratitis  capitata  in  Hawaii,  1916'^ 


Host  fruit. 


Month  of  col- 
lect iou. 


Num- 
ber of 
larvae 
emerg- 
ing 
during 
first 
2  to  6 
days. 


Percentage  of  parasitism. 


Opius 

humi- 

lis. 


Dia- 
chasma 
tryoni. 


Dia- 
chasma 
fulla- 

■wayi. 


Teiras- 
iichus 
giffar- 

dianus. 


Kamani  (Terminalia  catappa) 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Mango 

Do 

Do 

Do 

Do 

Coffee  b 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Strawberry  guava 

Do 

Do 

Do 

Do 

Do 

Black  myrobalan 

Do 

Do 

Peach 

Do : 

Rose-apple 

Do 

Do 

Chrysophyllum  mo nopyremmt 

Do 

Do 

Do 

Do 

Brazilian  plum 

Do 

Do 

French  cherry 

Do 

Do 

Do 


January. ..  . 

March 

April 

May 

June 

July 

August .  . .  . 
September. 
October. ..  . 
November. 
December. 
June 

July 

August .... 
September . 
October.  . .  . 
February.  . 

March 

April 

May 

June 

August .  .  .  . 
October  .  . . 
November. . 
December.  . 

April 

May 

June 

July 

October  .  . . 
November. . 
October  .  . . 
November.. 
December. . 

March 

April 

May 

June 

July 

January.  . .  . 
February.. . 

March 

April 

May 

June 

November.. 
December.  . 

January 

March 

April 

May 


115 
2,792 

9.558 

I.  391 

3.094 

3.569 

4,017 

3.526 

3.403 

2,299 

1,408 

283 

299 

47 

53 

39 

390 

62 

I,  621 

105 

131 

308 

288 

1,192 

643 

2,705 

48 

979 

1,013 

189 

34 

3,081 

2,518 

1,319 

2,3" 

951 

170 

1,089 

14 

996 

702 

378 
78 

634 
1,306 
78 
5Z 
41 
95 
25 

Ii4> 


22.  6 

30.5 

46.  I 

24.  7 

6.8 

2.4 

8.4 

11.  8 

IS- 5 
8.3 
9- 
9- 

9-7 
6.3 
I. 

5- 
60.  7 

4 
60.3 

57-1 
85-4 
17.  2 

59- o 
22.3 

12.  2 
38.1 
27.  I 
10.  2 
10.3 

3-  I 


27. 
3- 

15- 
4- 


6-5 


74-3 
57- o 

19-5 
7.6 
6.0 

34- ij 

8.4 
64.0! 
10.5! 


0.6 
I.  I 

16.5 

7-3 
27.7 

53-9 
58.  5 
51-5 
49 
3 
12.  o 

4' 

5-6 

12.8 

6.4 


•07 

•4 

•9 


6.7 

2,-Z 


29.7 

5-7 


66.  I 

4-3 

•9 


52.9 
10.  4 

II- 3 
14.  6 


9-4 
12.  I 
27.9 
17.  I 

2-7 


6.7 
46.  o 

14.8 
41. 1 

3-8 
8.6 


6.0 

3-5 


-5 
37-6 
62.  4 

57-1 
I.  I 


5-8 


16.6 

16.  7 

3-0 


1-5 

-5 

I.  o 

12.8 

-9 
22.3 
29.4 
21.  2 
14.  6 


32.  4j     9.  6 


0-5 
-03 


1-7 

2-3 

1.  2 
-3 

1-3 

2.  I 


•4 

-5 


•3 

I.  I 

12.8 


a  Most  of  the  fruits  represented  in  this  table  were  collected  about  Honolulu  at  low  elevations;  the  cofTee, 
however,  was  collected  on  the  island  of  Hawaii,  in  addition  to  localities  in  Honolulu,  and  much  of  it  came 
from  points  1,000  to  2,000  feet  above  sea  level. 

b  The  June  collection  of  coffee  came  from  the  Waianae  Mountains,  where  only  Opius  humilis  was 
established. 


Jan.  14. 1918      Fruit-Fly  Parasitism  in  Hawaii  during  igi6 


107 


Table  II. — Percentage  of  larval  parasitism  of  Ceratitis  capitata  in  Hawaii,  IQI'^ — Con. 


Month  of  col- 
lection. 

Num- 
ber of 
larvse 
emerg- 
ing 
during 
first 
2  to  6 
days. 

Percentage  of  parasitism. 

Host  fruit. 

Optus 

humi- 

lis. 

Dia- 

chasma 
tryoni. 

Dia- 
chasma 
fulla- 
wayi. 

Teiras- 
tichus 
Qiffar- 
dianus. 

Total. 

French  cherry 

June.  . . . 

479 

60 

862 

181 

314 

976 

10,  514 

2,535 

103 

18 

819 

210 

167 

116 

18 

22 

85 

173 

274 

392 

84 

159 

32 

84 

143 

319 

39 

258 

76 

1,767 

853 
403 
781 
248 
1,670 
496 

4.8 
1.6 

3-4 

4.4 

.6 

3-5 

8.4 

10.  0 

IS- 5 
16.6 

.8 
1.9 

4-7 
.8 

5-5 

6.2 

1.6 

10.7 

II.  4 

48.3 

8.4 

•5 

Do 

Tulv  . 

22.  4 
51.5 

Do 

December.  . .  . 

January 

February 

March 

April 

Mimusops  elengi 

22.  5 

Do 

4.  9 
6 

Do 

3-5 

8     A1 

Do 

•3 
36.8 

5-5 

.  01 
•3 
23-3 
22.  2 

.  2 

.02 
.  I 

.  2 

Do 

May 

"•  43 
10.7 

75-6 

44-3 

I.  2 
1.9 

4.7 
8 

Ochrosia  elliptica 

May 

Do 

June 

Kamani  {Calophyllum  ino- 
phyllum) 

January 

March 

April 

Do 

Do 

Do 

December. ..  . 

March 

May 

Bestill 

16.6 

Do 

Do 

June 

7.0 

10.  9 

3-6 

3-3 

9.4 

3-4 

2-5 

3-3 

9.4 

2-3 

22.  6 
19.  I 
39-2 
16.  9 

^•7 
S8 
9.6 

7-1 
13^8 

25.8 
18.3 
34-5 
35-3 
46.3 
31-9 
40.  6 

13-0 

4.8 

.  2 

Do 

August 

September . .  . 

October 

November. . .  . 

December 

September .  .  . 

October 

January 

March 

Tulv 

Do 

Do 

Do 

Do 

I.  2 
40.  6 

13.0 
4.  I 

Averrhoa  carambola 

Do 

•7 
.  2 

5-1 
2.7 

Do 

Do 

"3-'8" 

23.6 

.  2 

•4 

3-4 

2.  0 

2-5 

5-2 

4.6 

7-7 
1-5 

12  8 

Do 

August 

September . .  . 
June 

8  0 

Do 

23.6 
1.7? 

Noronhia  emarginata 

1-5 
1.6 

•4 

3-6 
2.  I 

•OS 

Do 

July 

2.  0 

Guava 

Mav 

3-8 
2.  2 

Do 

June 

Do 

Tulv 

•4 
•OS 



6^5 

7-35 
4.6 

Do 

September . .  . 
October 

Do 

The  data  in  Tables  I  and  II  covering  the  guava  are  of  unusual  interest. 
This  shrub  grows  wild  and  luxuriantly  over  most  of  the  uncultivated 
portions  of  the  islands  up  to  an  elevation  of  1,500  to  2,000  feet,  and  fruits 
throughout  the  year.  It  is  not  generally  considered  by  the  layman  of 
Hawaii  as  a  favored  host  of  the  fruit  fly,  though  of  the  18  host  fruits 
given  in  Table  I,  it  stands  fourth  in  degree  of  infestation,  showing,  from 
1,791  fruits  collected  during  the  year,  an  average  infestation  of  6.8  larvae 
per  fruit.  Infestation  of  this  fruit  is  not  easily  detected  until  it  has 
decayed.  The  larvae  are  small  and  nearly  all  inconspicuous  at  the  time 
the  fruits  are  picked  and  eaten  or  converted  into  preserves  or  jelly. 
This  fruit,  though  heavily  infested  in  most  localities,  protects  the  larvae 


io8 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  2 


from  parasite  attack  and  thus  constantly  liberates  great  numbers  of 
flies  throughout  the  year  and  serves  but  in  a  small  measure  toward  the 
building  up  of  favorable  quantities  of  parasites. 

Table  III. — Total  parasitism  by  month  of  all  larvcB  of  Ceratitis  capitata  collectedin 

Hawaii  during  igi6 


Month. 


Number  of 
larvse. 


Percentage  of  parasitism. 


Opius 
humilis. 


Diachasvta 
iryoni. 


Diackasvia 
fuUawayi. 


Teirasti- 

chus  gif- 

fardianus. 


Total. 


January.... 
February.. 

March 

April 

May 

June 

July 

August. .. . 
September 
October. . . 
November. 
December. 


2,295 

1,  406 

7,161 

21,  619 

5,525 
10,  013 

6,134 
4,803 

5,631 
7,972 
6,  205 
4,540 


0.4 

1-7 


2 
6 

15 

13 

27 

34 

27 

25.2 

20.  9 


0.08 


.  2 
.04 
.09 
.  009 
.02 

•9 

•4 
1.4 
1.4 
4.6 


6.98 

19- S 
14.7 

37-64 

26.  69 

27.  809 
18.52 

37-5 
45-2 
44-3 
44-3 
44.1 


A  comparison  of  fruit-fly  parasitism  data  secured  during  the  years 
1914,  1915,  and  1 91 6  would  indicate  that  the  parasites  now  present  in 
the  Territory  have  reached  their  maximum  degree  of  development  and 
can  hardly  be  expected  to  attain  a  greater  control  of  the  fruit  fly  than 
that  evidenced  in  191 6.  There  has  been  some  variation  during  the  past 
three  years  in  the  activities  of  the  different  species  introduced,  as  already 
noted  in  regard  to  the  fluctuations  in  abundance  of  Diachasma  iryoni  and 
Opius  humilis;  but  the  check  upon  the  work  of  this  pest  by  the  present 
parasites  can  hardly  exceed  its  present  limits.  Some  hope,  however,  is 
yet  felt  for  the  parasite  Tetrastichus  gifjardianus  Silvestri.  It  has  grad- 
ually increased  in  numbers  about  Honolulu  since  its  establishment  late 
in  1 914.  Certain  valuable  points  in  its  favor  may  enable  it,  after  further 
acclimatization  and  general  adaptation  to  new  environment,  to  exceed 
the  work  of  the  braconids  and  thus  increase  the  total  average  parasitism. 


ADDITIONAL  COPIES 

OP  THIS  PTTBLICATION  MAY  BE  PROCURED  FROM 

THE  SUPERINTENDENT  OF  DOCUMENTS 

GOVERNMENT  PRINTING  OFFICE 

WASHINGTON,  D.  C. 

AT 

20  CENTS  PER  COPY 

Subscription  Price,  $3.00  Pee  Year 

V 


Vol.  XH  JANUARY  21,   1918  No.  3 

JOURNAL  OF 

AGRICULTURAL 
RESEARCH 


CONXKNTS 

Page 

Irrigation  Experiments  on  Apple-Spot  Diseases  -        -        -    109 
CHARLES  BROOKS  and  D.  F.  FISHER 

( Contributkin  from  Bureau  of  Plant  Industry ) 

Relation  of  Carbon  Dioxid  to  Soil  Reaction  as  Measured 

by  the  Hydrogen  Electrode     -        -        -        -        -        -     139 

D.  R.  EOAGLAND  and  L.  T.  SHARP 

( Contribution  from  California  Agricultural  Experiment  Station) 


PUBUSHED  BY  AUTHORITY  OF  THE  SECRETARY  OF  AGRICULTURE. 

WITH  THE  COOPERATION  OF  THE  ASSOCIATION  OF  AMERICAN 

AGRICULTURAL  COLLEGES  AND  EXPERIMENT  STATIONS 


WASWINOXOM,  E).  C. 


WASHmOTON  ■■  OOveRNMtHT  piflNTlNO  OKFIC" 


EDITORIAL  COMMITTEE  OF  THE 

UNITED    STATES    DEPARTMENT    QF    AGRICULTURE    AND 

THE  ASSOCIATION  OF  AMERICAN  AGRICULTURAL 

COLLEGES  AND  EXPERIMENT  STATIONS 


FOR  THE  DEPARTMENT 

KARL  F.  KELLERMAN,  Chairman 

Physiologist  and  Associate  Chief,  Bureau 
of  Plant  Industry 

EDWIN  W.  ALLEN 

Chief,  Office  of  Experiment  Siatiorts 

CHARLES  L.  MARLATT 

Entomologist  and  Assistant  Chief,  Bureau 
of  Entomology 


FOR  THE  ASSOCIATION 

RAYMOND  PEARL* 

Biologist,  Maine  Agricultural  Experitneni 
Station 

H.  P.  ARMSBY 

Director,  Institute  of  Animal  Nulrition,  The 
Pennsylvania  State  College 

E.  M.  FREEMAN 

Botanist,  Plant  Pallwlogist  and  Assistant 
Dean,  Agricultural  Experiment  Station  of 
the  Universily  of  Minnesota 


All  correspondence  regarding  articles  from  the  Department  of  Agriculture  should  be 
addressed  to  Karl  F.  Kellerman,  Journal  of  Agricultural  Research,  Washington,  D.  C. 

*Dr.  Pearl  has  imdertaken  special  work  in  connection  with  the  war  emergency; 
therefore,  until  further  notice  all  correspondence  regarding  articles  from  State  Experi- 
ment Stations  should  be  addressed  to  H.  P.  Annsby,  institute  of  Animal  Nutrition, 
State  College,  Pa. 


JOIML  OF  ACMIMIRAL  RESEARCH 

Vol.  XII  Washington,  D.  C,  January  21,  191 8  No.  3 


IRRIGATION  EXPERIMENTS  ON  APPLE-SPOT  DISEASES^ 

By  Charles  Brooks,  Pathologist,  and  D.  F.  Fisher,  Assistant  Pathologist,  Bureau 
of  Plant  Industry,  United  States  Department  of  Agriculture 

THE  PROBLEM 

The  present  paper  deals  with  the  effects  of  soil-water  supply  upon 
bitter-pit,  Jonathan-spot,  and  certain  other  nonparasitic  spot  diseases 
of  the  apple  (Mains  sylvestris).  It  also  includes  notes  upon  the  relation 
of  the  time  of  picking  to  the  development  of  apple-spots  in  storage. 

BITTER-PIT 

HISTORICAL  REVIEW 

Bitter-pit  was  first  described  by  Wortmann  {22)  ^  under  the  name 
"Siippen."  It  has  been  frequently  discussed  in  the  publications  of  the 
State  Experiment  Stations  under  the  name  "Bald  win -spot"  and  was 
referred  to  in  the  Nineteenth  and  Twentieth  Reports  of  the  New  Hamp- 
shire Experiment  Station  as  "fruitpit."  Lewis  (8)  and  Allen  (7)  appar- 
ently used  the  term  "fruitpit"  to  refer  to  the  troubles  discussed  later 
in  this  paper  under  the  name  "cork,"  and  McAlpine  (9-12)  apparently 
included  cork  and  also  drouthspot  under  the  name  "bitter-pit." 

Various  explanations  have  been  offered  as  to  the  cause  and  nature  of 
bitter-pit.  Wortmann  (22)  reported  that  the  disease  was  due  to  abnor- 
mal transpiration  conditions  and  that  varieties  of  apples  in  which  the 
water  was  conducted  most  readily  from  the  deeply  seated  cells  to  replace 
that  lost  by  transpiration  were  least  susceptible. 

Sorauer  (19,  p.  80)  thought  that  the  pits  were  produced  by  rupturing 
of  the  cells  during  the  process  of  swelling.  In  a  later  publication  {20,  p. 
116-169)  he  stated  that  the  disease  was  worst  on  porous  dry  soils  and 
suggested  that  the  pits  were  produced  by  an  overrapid  maturing  of 
certain  cell  groups  resulting  from  the  checking  effect  of  drouth  upon  the 
accumulation  of  organic  material. 

Evans  (5)  reported  that  the  disease  was  due  to  a  bursting  of  cells  in 
the  apple  tissue  that  resulted  from  the  sudden  checking  of  transpiration 
at  night  while  the  root  action  of  the  tree  remained  vigorous. 

'  Studies  on  Fruit  Rots  and  Spots:  III. 

2 Reference  is  made  by  number  (italic)  to  "Literature  cited,"  p.  136-137. 

Journal  of  Agricultural  Research,  Vol.  XII,  No.  3 

Washington,  D.  C.  Jan.  21,  1918 

Iv  Key  No.  G-133 

(109) 


no  Journal  of  Agricultural  Research  voi.  xii.  no.  3 

Ewart  (6,  7)  concluded  that  the  disease  was  the  result  of  local  poison- 
ing and  mentioned  spray  materials  and  the  toxic  salts  of  the  soil  as 
possible  causes. 

White  {21)  considered  that  the  disease  was  the  result  of  the  poisoning 
effects  of  arsenical  compounds  and  other  spray  materials. 

McAlpine  (9-12)  thought  that  the  disease  was  produced  by  a  shortage 
of  water  in  the  affected  tissue  and  that  the  condition  might  be  brought 
about  either  by  transpiration  exceeding  the  water  supply  or  by  the  growth 
of  the  pulp  tissue  being  too  rapid  to  allow  time  for  the  formation  of  the 
new  vascular  tips  needed  to  supply  it  with  water.  He  found  that  there 
was  slightly  less  of  the  disease  on  trees  receiving  two  irrigations  than  on 
those  receiving  one. 

DESCRIPTION   OF   BiTTER-PiT 

Bitter-pit  makes  its  first  appearance  as  water-soaked  bruiselike 
spots  on  the  surface  of  the  apple.  The  epidermal  tissue  is  at  first  entirely 
normal,  the  spotted  effect  being  due  to  the  breaking  down  of  cells  in  the 
subepidermal  region.  The  spots  soon  become  depressed  into  rather 
definite  pits,  2  to  6  mm.  in  diameter,  hemispherical  in  shape,  and  fairly 
regular  in  outline.  They  develop  a  higher  color  than  the  surrounding 
surface  of  the  apple,  becoming  a  deeper  red  than  the  adjacent  tissue 
when  occurring  on  the  colored  portion  of  the  fruit  and  a  darker  green 
when  on  the  lighter  parts  (PI.  2,  A).  As  the  disease  advances  farther, 
the  spots  take  on  a  brownish  color  owing  to  the  dead  pulp  cells  beneath 
the  epidermal  layers,  and  in  late  stages  of  the  trouble  the  affected  area 
may  entirely  lose  its  normal  color,  becoming  a  deep  brown  (PI.  3,  A). 
The  diseased  tissue  is  dry  and  spongy,  the  cells  are  collapsed  but  still 
full  of  starch,  and  the  cell  walls  show  no  sign  of  thickening  or  disinte- 
gration. The  affected  tissue  often  has  rather  a  bitter  taste,  and  this 
together  with  the  sunken  nature  of  the  spots  has  given  rise  to  the  term 
'' bitter-pit." 

The  pits  are  usually  associated  with  the  terminal  branches  of  the 
vascular  bundles,  and  the  surface  spotting  is  often  accompanied  by  a 
browning  of  the  vascular  tissue  deeper  in  the  fruit,  giving  the  appear- 
ance of  numerous  brown  spots  in  the  flesh  when  the  apple  is  cut  (PI.  2,  B). 
This  internal  browning  is  especially  common  in  the  tissue  within  a  centi- 
meter of  the  surface  of  the  apple.  While  the  internal  browning  and  sur- 
face pitting  are  commonly  associated,  either  may  occur  without  the  other. 

Bitter-pit  is  often  confined  to  the  calyx  half  of  the  apple.  Baldwin, 
Northern  Spy,  Grimes,  Jonathan,  and  Yellow  Bellflower  are  especially 
susceptible  to  the  disease;  and  Rome  Beauty  and  Winesap  are  fairly 
resistant;  but  almost  all  varieties  are  sometimes  affected. 

Bitter-pit  is  very  similar  in  appearante  to  rosy-aphis  stigmonose,  but 
the  latter  disease  is  not  accompanied  by  a  browning  of  the  vasculars 
and  the  subepidermal  tissue  has  a  firmer  texture  and  a  darker  color  than 
is  the  case  with  bitter-pit.     Stigmonose  is  found  only  on  limbs  that 


Jan.  21, 1918     Irrigation  Experiments  on  Apple-Spot  Diseases  iii 

were  infested  with  aphids  earlier  in  the  year,  and  the  spots  usually 
appear  several  weeks  before  picking  time,  while  bitter-pit  is  rather 
evenly  distributed  over  the  tree  and  is  found  only  on  mature  or  nearly 
mature  fruit.  Bitter-pit  can  be  distinguished  from  fruitspot  by  the 
fact  that  with  the  latter  disease  there  is  an  almost  entire  absence  of 
subepidermal  browning,  and  the  spots  have  an  irregular  outline  and  a 
flecked  or  speckled  appearance. 

EXPERIMENTAI,   WORK 

The  writers  were  convinced  by  earlier  investigations  that  bitter-pit 
was  not  due  to  fungi  or  bacteria.  They  had  frequently  seen  unsprayed 
fruit  that  was  seriously  affected  with  the  disease,  thus  making  the 
theory  that  spray  materials  were  responsible  for  the  trouble  seem  en- 
tirely untenable.  Drouth  had  frequently  been  mentioned  as  a  cause  of 
bitter-pit,  and  several  writers  had  suggested  an  excessive  or  uneven 
water  supply  as  a  possible  cause,  but  little  experimental  data  had  ever 
been  furnished  in  support  of  any  of  these  theories.  The  question  of  the 
influence  of  soil-water  supply  seemed  to  the  writers  to  be  an  extremely 
important  one,  and  a  series  of  experiments  were  started  to  determine 
the  effect  of  irrigation  upon  the  disease. 

The  work  has  been  located  at  Wenatchee,  Wash.  The  climate  of  the 
section  is  arid,  but  little  precipitation  occurring  from  April  till  October, 
thus  making  the  trees  almost  entirely  dependent  upon  irrigation  for  their 
soil-water  supply  during  the  growing  season.  Except  where  otherwise 
mentioned,  the  water  was  applied  by  the  furrow  method  (PI.  4,  A). 
The  contrasts  in  the  amount  of  water  on  the  different  plats  were  secured 
by  varying  the  frequency  and  duration  of  the  irrigations  and,  in  some 
cases,  by  differences  in  the  number  of  furrows  supplying  water  to  the 
row  and  by  variations  in  the  head  of  water  at  the  flume. 

The  amount  of  water  in  the  soil  was  determined  by  means  of  samples 
taken  with  a  soil  auger.  In  the  beginning  of  the  work  samples  were 
taken  at  depths  of  6,  18,  30,  42,  and  54  inches — that  is,  from  the  middle 
of  each  of  the  first  5  successive  feet  of  soil — but  in  the  final  experiments, 
as  reported  later,  samples  were  taken  only  from  the  one  or  two  depths 
that  seemed  most  important  in  determining  the  condition  under  which 
the  tree  roots  were  working  in  the  particular  orchard. 

Some  difficulty  was  found  in  securing  samples  that  represented  the 
average  moisture  conditions  of  the  tree  row.  The  lateral  movement  of 
soil  water  is  very  slow,  resulting  in  considerable  contrast  between  the 
amount  of  moisture  beneath  the  irrigation  furrow  and  a  few  feet  from  it, 
especially  in  the  upper  layers  of  the  soil.  Samples  were  usually  taken 
at  a  distance  from  the  furrow  equal  to  one-fourth  the  space  between 
the  furrows,  thus  securing  soil  from  a  point  midway  between  the  wettest 
and  dryest  areas.  The  plan  of  sampling  was  always  the  same  for  the 
different  plots  of  a  particular  orchard.  Samples  were  taken  at  intervals 
of  7  to  10  days,  and  usually  just  before  and  i  or  2  days  after  an  irriga- 
tion, thus  obtaining  a  record  of  the  extremes  in  soil-water  conditions. 


112 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  3 


As  soon  as  obtained,  the  samples  were  transferred  to  tin  cans  with 
tightly  fitting  lids,  and  the  cans  immediately  closed.  The  weight  of 
the  fresh  soil  sample  was  determined  and  a  second  weighing  made  after 
the  soil  had  been  reduced  to  constant  weight  in  a  drying  oven,  the 
difference  between  the  two  weighings  being  taken  as  the  moisture  con- 
tent of  the  sample.  The  percentage  of  saturation  was  determined  by 
comparing  the  moisture  content  of  the  sample  with  the  total  water- 
holding  capacity  of  the  soil.  In  the  experiments  of  1914  and  191 5  the 
latter  was  secured  by  taking  the  average  water-absorbing  capacity  of 
a  large  number  of  samples,  but  in  191 6  saturation  tests  were  made  on 
each  soil  sample. 

Notes  were  taken  on  the  amount  of  bitter-pit  at  picking  time,  and 
later  notes  were  taken  to  determine  the  increase  in  storage.  The  apples 
were  cut  open  at  the  time  of  the  last  note-taking  and  a  record  made  of 
the  amount  of  internal  browning.  An  apple  was  counted  as  affected 
with  bitter-pit  if  it  had  either  internal  or  external  evidence  of  the  dis- 
ease, but  very  few  apples  showed  internal  browning  that  did  not  also 
have  the  external  pitting. 

EXPERIMENTS  ON  GANG  APPLES  IN  1914 

The  irrigation  experiments  were  begun  in  191 3.  The  results  of  the 
first  season  were  of  little  value,  since  the  main  trouble  in  the  experi- 
mental orchards  was  found  to  be  stigmonose  instead  of  bitter-pit.  In 
1 91 4,  the  data  from  the  most  promising  orchard  were  lost  on  account 
of  mistakes  of  the  picking  crew,  but  some  interesting  contrasts  were 
obtained  in  an  orchard  of  Gano  apples.  The  trees  in  the  latter  orchard 
were  11  years  old  and  thrifty;  the  soil  was  a  volcanic  ash,  uniform  in 
texture  to  a  depth  of  6  feet.  The  orchard  had  been  under  clean  culti- 
vation but  at  the  time  of  the  experiments  was  sown  to  vetch.  There 
were  four  trees  in  each  plat.  The  soil-moisture  condition  for  the  season 
is  shown  in  figure  i.  All  of  the  plats  became  quite  dry  the  middle  of 
August  on  account  of  trouble  with  the  irrigation  canals. 

Table  I. — Percentage  of  bitter-pit  on  Gano  apples  in  igi4 


Irrigation  treatment. 

Total 
num- 
ber 
of  ap- 
ples. 

Percentage  of  apples  of  following  sizes: 

Per- 

Plat 
No. 

3K 

to  4 

inches. 

3J4 
t0  3K 
inches. 

3K 

to  3'^ 
inches. 

3  to 
inches. 

2H 

■  ^V 

inches. 

ageof 

bitter- 

pit- 

I 
2 
3 
4 

Heavy  throughout  season 

Medium  throughout  season 

Light  throughout  season 

129 

152 
176 
144 

2.7 
.8 
.  0 
.6 

25.  2 

18.8 

7.0 

21.  9 

27.  0 
30.  6 
30.  6 
23-9 

36.6 
41.9 

39-9 
41.  0 

8-5 

7-9 

22.  s 

12.  6 

7.0 

2.6 
2-3 

.  0 

Medium  till  Aug.  i,  then  light. . . 

The  crop  was  quite  heavy,  averaging  about  25  bushels  per  tree.  There 
was  no  bitter-pit  on  the  fruit  at  picking  time.  Five  boxes  of  apples 
from  each  plat  were  placed  in  cold  storage  and  held  for  three  months. 
Tab  e  I  gives  the  results  of  notes  taken  at  the  end  of  this  storage  period. 


Jan.  21, 1918     Irrigation  Experiments  on  Apple-Spot  Diseases 


113 


The  results  make  it  evident  that  heavy  irrigation  favored  bitter-pit, 
and  also  increased  the  size  of  the  apples.     It  can  be  seen  that  there  is 


60^ 


GO^o 


r? 

■^  Co 


I 


^O  "^  *0  ^  "^ 
N. 

lis 


I. 


xis 


^S 


^  ^  ^  ^  '^  '^  <Q 


^0  O  ^  'v  Itn  "^  "O 
N.  N^  <  »0  <  N)  n; 

I    I    I    I     I    »     V 

N.  <0  "v  <6  V.  (o  "^ 


I   I 


•O  >x  »0    >*   ^ 

<  K>  <  fO  > 

^  Co  N.  <o  ^ 


^  ^>  $  §  ^  ^  J*J 


<  no  N.  fo 
I    I     I    I 

N.  CO  "^  <0 

^^^^ 


^ 
K 
^ 

^ 


10  u 

(3:  q;  ^  ^ 

Fig.  I. — Diagram  showing  the  soil-moisture  conditions  in  irrigated  plats  of  Gano  apples  in  1914.  The 
results  show  the  average  percentage  of  saturation  for  each  half  month  based  on  the  average  of  soil  samples 
at  depths  of  24  and  36  inches. 

a  very  close  relation  between  the  size  of  the  apple  and  the  amount  of 
the  disease,  but  there  is  hardly  sufficient  parallelism  to  justify  the  con- 
clusion that  the  increase  in  bitter-pit  is  entirely  due  to  increase  in  size. 


114 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  3 


EXPERIMENTS  ON  GRIMES  APPLES  IN   I915 

Similar  irrigation  experiments  were  carried  out  in  191 5  on  Grimes 
apples.  The  experimental  plats  were  located  on  bottom  land  along  the 
Wenatchee  River.     The  soil  was  of  an  alluvial  nature  composed  of 


so^ 


60% 


*^)   >v  Wv  >j  U 


^  ^  s^'  to  q: 


^  i>  "^  >  "O 

k 


Fig.  2. — Diagram  showing  the  soil-moisture  conditions  in  irrigated  plats  of  Grimes  apples  in  1915.  The 
vertical  bars  show  the  average  percentage  of  saturation  for  each  half  month  based  on  the  average  of  soil 
samples  at  depths  of  24  and  36  inches. 

medium  heavy  sandy  loam  with  considerable  clay  and  was  uniform  to 
a  depth  of  3  feet.  It  was  kept  under  cultivation  in  the  tree  row.  The 
trees  were  5  years  old  and  making  a  vigorous  growth.  They  were  quite 
uniform  in  size  and  vigor  and  satisfactory  in  every  respect  for  compara- 
tive experiments.    Three  trees  were  used  in  each  plat.     The  fruit  was 


Jan.  21, 1918     Irrigation  Experiments  on  Apple-Spot  Diseases 


115 


kept  practically  free  from  stigmonose  by  means  of  a  late  dormant  spray 
of  lime-sulphur  and  nicotin  sulphate. 

The  irrigation  of  the  orchard  was  not  satisfactory  because  of  a  short- 
age of  water  resulting  from  trouble  with  the  canals.  A  cloudburst  on 
July  26  gave  an  indiscriminate  watering  to  all  the  plats.  Irrigations 
were  made  according  to  plan  on  August  24  and  September  12.  The  soil 
moisture  conditions  for  the  latter  part  of  the  season  are  shown  in  figure  2. 

It  will  be  noted  that  in  spite  of  the  unfavorable  conditions  a  decided 
contrast  in  soil  moisture  was  secured  on  the  different  plats.  Plat  i 
was  given  heavy  irrigation  throughout  the  season;  plat  2,  medium; 
plat  3,  light;  and  plat  4,  medium,  followed  by  heavy.  The  percentages 
of  soil  saturation  given  do  not  indicate  any  decided  contrast  between 
plats  2  and  3,  but  the  condition  of  the  trees  in  the  two  plats  made  it 
very  evident  that  a  distinct  contrast  in  soil-water  conditions  had  been 
secured. 

The  yield  in  the  orchard  was  light,  being  about  a  bushel  to  the  tree. 
The  fruit  was  gathered  on  September  22,  about  10  days  later  than  the 
average  commercial  picking  of  Grimes  apples  in  that  section.  It  was 
placed  immediately  in  cellar  storage  at  a  temperature  of  about  50°  F. 
Notes  were  taken  on  bitter-pit  seven  days  later.  The  results  are  shown 
in  Table  II. 

Table  II. — Percentage  of  bitter-pit  on  Grimes  apples.     September  2g,  igi§ 


Plat 
No. 


Irrigation  treatment. 


Heavy 

Medium 

Light 

Medium  till  Aug.  24,  then  heavy 


Total 
num- 
ber 
of 
apples. 


299 
222 
156 
175 


Percentage   of   apples 
affected  with  bitter-pit. 


Wind- 
falls. 


90 
36 

77 


Picked 
fruit. 


43 
17 
14 
49 


Total. 


56 
25 
23 
59 


The  contrasts  are  quite  striking  and  make  it  evident  that  heavy 
irrigation  tended  to  increase  the  amount  of  bitter-pit.  It  is  interesting 
to  note  that  plat  4,  which  was  heavily  irrigated  late  in  the  season, 
showed  a  greater  percentage  of  the  disease  on  the  picked  fruit  than 
plat  I,  which  was  heavily  irrigated  early  as  well  as  late. 

All  of  the  fruit  that  was  apparently  free  from  bitter-pit  was  returned 
to  cellar  storage  and  notes  were  taken  again  on  November  9.  The 
results  are  given  in  Table  III.  All  of  the  percentages  but  those  in  the 
last  column  are  based  on  the  number  of  apples  returned  to  storage 
and  not  on  the  number  in  the  original  yield  from  the  plats. 


ii6 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  3 


Table  III. — Percentage  of  bitter-pit  on  Grimes  apples  in  storage.     November  g,  igis 


Plot 
No. 


Irrigation  treatment. 


Total 
num- 
ber 
of 
apples. 


Percentage  of 
apples  of  the 
following  sizes. 


Larger 
than 


inches. 


2Vi 

inches 

and 
smaller. 


Percentage  of  apples 
that  developed  bitter- 
pit  from  Sept.  29  to 
Nov.  9. 


Apples 

larger 

than 

2^ 

inches. 


Apples 

inches 

and 
smaller. 


Total. 


Total 
per- 
cent- 
age 
of 
bitter- 
pit 
de- 
vel- 
oped 
by 
Nov. 


Heavy 

Medium 

Light 

Medium    till    Aug.    24,    then 
heavy 


132 
166 
106 

72 


92.  2 
83.0 

08.6 


18.  2 

7.8 

17.  o 

1.4 


63- 9 
48.4 

33-7 
63-4 


4.2 
.  o 

II.  o 


79 
59 

52 


62.  5 


P£P  CENT  OF Brrrep-PJT 


\  PEO  CENT  or  B/rrEQ-Pir  om  s£pr.  29. 

\p£»,CENr  OF  Birrc/f-piT  dcvclopfo bcfween S£Pr.39ANO nov.s. 

\P£a  CE/VT  OF  BITTEQ-PIT  DEVELOPED  BFTWEEr^  N0V.3  AND  UAN.^*. 

Fig.  3.— Diagram  showing  the  amount  of  bitter-pit  on  Grimes  apples  in  1915.  The  black  portions  of 
the  bars  indicate  the  percentage  of  apples  affected  with  bitter-pit  one  week  after  picking;  the  shaded  por- 
tions, the  amount  developed  between  September  29  and  November  9;  and  the  white  portion,  the  amount 
between  November  9  and  January  4.  All  of  the  percentages  are  based  on  the  number  of  apples  at  the 
beginning  of  the  experiment.    See  figure  2  for  soil-moisture  conditions. 

A  study  of  Table  III  shows  that  nearly  all  of  the  bitter-pit  occurred 
on  the  apples  that  were  larger  than  2^  inches.  The  percentages  in 
the  next  to  the  last  column  show  that  the  contrasts  in  bitter-pit  on 
the  stored  samples  were  similar  to  those  found  a  week  after  picking, 
and  indicate  the  importance  of  orchard  conditions  in  determining  the 
susceptibility  of  the  fruit  in  storage.  These  percentages  are  estimated 
on  the  basis  of  the  sound  apples  left  on  September  29.  If  the  original 
number  of  apples  were  taken  as  a  base  in  estimating  percentages,  these 
contrasts  would  partially  disappear,  as  is  shown  in  figure  3;  but  this 
would  be  an  unfair  comparison,  so  far  as  determining  behavior  in 
storage  is  concerned,  as  a  large  number  of  the  apples  had  already  been 
eliminated  from  the  experiment.  The  last  column  in  Table  III  shows 
the  total  amount  of  bitter-pit  to  November  9,  estimated  on  the  basis  of 
the  original  number  of  apples. 

The  sound  fruit  from  the  above  experiment  was  returned  to  cellar 
storage  and  a  third  set  of  notes  taken  on  January  4,  the  fruit  being  cut 
open  at  this  time  to  determine  the  amount  of  internal  streaking  or 
browning.  But  very  few  specimens  of  bitter-pit  were  found,  and  these 
gave  but  little  contrast  between  the  fruit  from  the  diflferent  irrigation 
plats. 


Jan.  21. 1918     Irrigation  Experiments  on  Apple-Spot  Diseases 


117 


The  results  obtained  on  Grimes  apples  in  191 5  are  shown  in  graphic 
manner  in  figure  3.  The  contrasts  for  the  season  are  similar  to  those 
obtained  on  September  29. 

EXPERIMENTS  ON  GRIMES  APPLES  IN   1916 

In  1 91 6  the  experiments  were  continued  in  the  Grimes  orchard  de- 
scribed above.     Five  trees  were  included  in  each  plat.     It  was  possible 


ll 


I 


60%   ■ 


I 


^  r>  <  o  N  »o  < 
t   I    I   I   I   I   I 


<  10  V  ^0  >n'  (o  V 

I    I   I    I    I    I    t 
^  «o  •>.  to  "^  ©  >. 

■>v        v>       ■>» 

•M 
K. 


II    I   I    I    I.  I 
N.  to  ■«>  to  •»>  to  ■^ 


<o  O  "o  -v  "O  V  ^0 

V  *5  X  lo  V  »o  s, 
I     I     I     I     I     I     I 

V  to  ^  to  ^  to  ^ 

N        >^        >. 

•<!  Uj  ^    X  ,  •     ■  •< 


"o  O  "0  ^  'O  X  "o 

.  <  10  >.  »0  V  f^  < 

I    I  I   I    I    I  I 

X  to  •>»  to  ■>■  to  V 

N.  N  S. 

So  ijj  i.  X       .  k- 

>  s  vj  vi  to  o  Q 


s  "0  s  "0  ^  •>  >: 
I  I  I  I  I  I  I 
X  b  V.  to  ■>.  to  ■"^ 


5  $  y  SI 


Fig.  4.— Diagram  showing  soil-moisture  conditions  in  irrigated  plats  of  Grimes  apples  in  1916.  The  results 
show  the  average  percentage  of  saturation  for  each  half  month  based  on  the  average  of  soil  samples  at 
depths  of  24  and  36  inches.  Plat  i  received  heavy  irrigation  throughout  the  season;  plat  2,  medium; 
plat  3,  light;  plat  4,  medium  till  late  in  July,  then  heavy;  plat  5,  medium  in  June,  heavy  in  July,  and 
light  in  August  and  September;  and  plat  6,  heavy  throughout,  with  the  exception  of  a  sudden  drop  to 
medium  in  July. 

to  carry  out  the  irrigation  schedule  much  more  satisfactorily  than  in 
the  preceding  year.  The  soil-moisture  conditions  for  the  season  are 
shown  in  figure  4. 

The  fruit  on  plat  3  was  noticeably  smaller  than  that  on  the  other  plats 
as  early  as  August  i ,  and  by  the  close  of  the  season  the  effects  of  irriga- 
tion were  quite  evident  in  the  size  of  the  fruit  from  the  various  plats.  At 
picking  time  the  fruit  on  plat  3  was  found  to  be  somewhat  riper  and  more 
highly  colored  than  that  on  the  other  plats.     The  apples  were  picked 


ii8 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  3 


on  September  i6.  The  crop  was  uniform  and  quite  heavy,  making  it 
possible  to  secure  approximately  2  bushels  from  each  tree  for  storage. 
The  fruit  was  placed  in  cellar  storage  in  open  packages.  Hygrother- 
mograph  records  showed  that  from  September  19  to  October  18  the 
temperature  of  the  cellar  averaged  55°  F.,  and  the  relative  humidity 
approximately  55  per  cent;  that  from  October  18  to  November  9  the 
average  temperature  was  48°  F.  and  the  average  relative  humidity  68 
per  cent,  and  that  from  November  9  to  March  20  the  temperature  was 
fairly  constant  at  35°  F.,  the  relative  humidity  averaging  80  per  cent. 
Notes  w&re  taken  on  September  19,  when  the  fruit  was  picked,  and  on 
October  18,  November  9,  and  March  20.  At  the  time  of  the  last  note 
taking  the  apples  were  cut  open  and  a  record  made  of  the  internal  brown- 
ing as  well  as  the  bitter-pit  spots.  The  vascular  bundles  of  about  half 
the  pitted  apples  were  browned,  but  the  apples  that  showed  no  external 

Pea  CENtoFBirreR-P/r 
o  s  10  IS  \?o  2s  30  35  ^o 


MIIIIIIIHA  pea  CENTOF BITT£fi-PlT/>T PICK/NG  TIME. 

^^BB  •°^*  CCNTOF BITT£Q-PIT DEvetOPED  BETWEEN  SEPT.  1641^0  OCT  10. 
\  I  PEO  CENT  OF  BlTT£!i-Pir  DEVEL  OPED  BETWEEN  OCT.  IB  AND  N0\/.  9. 

I  I  PEQ  CENT  OF  BITTE/I'PIT  OEVELOPEO  BET»V££NN0V.9ANDf1AOC/iS0. 


Fig.  s.— Diagram  showing  the  amount  of  bitter-pit  on  Grimes  apples  in  1916.  The  diagonally  shaded 
portions  of  the  bars  indicate  the  percentage  of  apples  having  bitter-pit  at  picking  time;  the  solid  por- 
tions, the  percentage  developed  between  September  19  and  October  18;  the  horizontally  shaded  portions, 
the  amount  developed  between  October  18  and  November  9;  the  white  portions,  the  amotmt  developed 
between  November  9  and  March  20.    See  figure  4  for  soil-moisture  conditions. 

evidence  of  bitter-pit  were  free  from  internal  browning.  The  bitter-pit 
results  are  given  in  Table  IV.  The  percentages  in  the  first  and  last 
columns  are  based  on  the  total  number  of  apples,  those  in  the  other  col- 
umns on  the  number  of  sound  apples  at  the  previous  note  taking. 

Table  IV. — Percentage  of  bitter-pit  on  Grimes  apples 


Percentage  of  apples  affected  with  bitter-pit. 

Plat  No. 

At  picking 
time,  Sept.  19. 

Developed  be- 
tween Sept.  19 
and  Oct.  18. 

Developed  be- 
tween Oct.  18 
and  Nov.  9. 

Developed  be- 
tween Nov .  9 
and  Mar.  20. 

Total. 

I 

I- 5 
I.  0 

•7 
1-3 

•  4 
4.  I 

19.  8 
9.8 

12.  I 

28.  7 

5-3 
35-2 

0.8 

1.  0 

2.  0 

1-5 
1.6 

5-5 

2>-Z 
2.  0 
4.1 
2.  0 
.6 
3-2 

23.0 

12.  7 

17.  6 

31- I 

7.6 

40.8 

2 

-J 

4 

c 

6 

Jan.  21, 1918     Irrigation  Experiments  on  Apple-Spot  Diseases  119 

A  study  of  the  table  shows  that  nearly  all  of  the  disease  developed 
during  the  first  month  of  storage.  The  contrast  between  the  plats, 
however,  makes  it  evident  that  the  development  of  the  disease  was 
largely  determined  by  orchard  conditions.  The  apples  from  the  heavily 
irrigated  plats  were  in  all  cases  more  susceptible  to  bitter-pit  than  those 
from  the  lightly  irrigated  ones.  The  amount  of  disease  was  much  greater 
on  plats  4  and  6,  which  were  irrigated  heavily  only  late  in  the  season, 
than  on  plat  i,  which  was  heavily  irrigated  throughout  the  season.  It 
was  less  on  plat  5,  which  had  heavy  irrigation  followed  by  light,  than  it 
was  on  plat  3,  which  received  light  irrigation  throughout  the  season,  or 

k 
StZ£OF/lPPL£S  J  PEP  CENT  OF B/TTEQ'PIT 

/O 20  JO  ^O 


6\ 

4\ 

2  3/4 /M  AND     / 

SMALLER         Z 

2 

6 

6 

/ 

2%  TO  3/M     J 

2 

Jl 

6 
4^ 
Z/N.AND  / 

LARGER  J 
2 
J 


J 


Fig.  6. — Diagram  showing  the  relation  of  the  amount  of  bitter-pit  to  the  size  of  apples.  The  bars  show 
the  amount  of  disease  on  the  different  plats  and  are  grouped  according  to  size  of  apples.  It  will  be  noted 
that  heavy  irrigation  increased  the  disease  as  much  on  the  small  fruit  as  on  the  large.  For  the  irrigation 
of  the  different  plats  see  figure  4. 

on  plat  2,  which  received  medium  irrigation  throughout  the  season. 
The  results  indicate  that  the  character  of  the  irrigation  during  the  last 
weeks  in  which  the  apples  are  on  the  trees  largely  determines  the  amount 
of  bitter-pit  developed  in  storage. 

The  total  amount  of  bitter-pit  for  the  season  is  shown  in  graphic 
manner  in  figure  5.  All  of  the  percentages  are  based  on  the  original 
number  of  apples. 

In  the  note  taking  of  October  18  the  apples  were  graded  according  to 
size,  and  the  record  on  bitter-pit  made  accordingly.  The  results  are 
given  in  Table  V  and  figure  6. 


I20 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  3 


Table   V. — Percentage,   according   to   size,    of  Grimes  apples  affected   with  bitter-pit. 

October  i8,  IQ16 


Total 
num- 
ber of 
ap-  ., 
pies. 

Percentage  of  apples  of  various  sizes. 

Percentage  of  apples  of  various  sizes  af- 
fected with  bitter-pit. 

Total 
per- 

Plat 
No. 

3'A 
to 

inches. 

3 '4  to 

3^ 
inches. 

3% 
inches. 

2?4 

to  3 
inches. 

2K 
inches. 

and 
smaller. 

sJ^to 

3K 
inches. 

3%  to 

3^^ 
inches. 

3  to 
inches. 

2K 

to  3 

inches. 

2K 

inches 

and 
smaller. 

cent- 
age of 
bitter- 
pit. 

I 

2 

3 
4 
5 
6 

776 

775 
879 
560 
690 
715 

0.4 

.  I 

•4 

10.  8 
6.4 

10.3 

12.5 
6.7 

20.  7 

30.8 
26.  5 
28.3 
38.6 
24.8 
31.2 

37-9 
50.  2 
48.8 
39-  I 
45-1 
40.  8 

20.  I 
16.  9 

12.5 
9.8 

23-4 
6.9 

100.  0 

100.  0 


100.  0 

52-4 
36.  0 

14-5 

54-3 

7.0 

31-9 

20.  9 
13.6 
26.  7 
31.0 
15.2 
50.0 

13-6 
6.9 
9.6 

23-3 

5- 1 

31-5 

10.  9 

2-3 

3-6 

9.1 

I.  2 

24-5 

19-8 
9-8 

12. 1 

28.7 
5-3 

35-2 

The  large  apples  were  much  more  susceptible  to  bitter-pit  than  the 
small  ones,  but  evidently  size  can  not  be  taken  as  a  measure  of  sus- 
ceptibility, since  the  small  apples  on  the  heavily  irrigated  plats  often 
developed  more  disease  than  the  large  ones  on  the  lightly  irrigated  ones. 
(Table  V;  fig.  6.)  A  study  of  the  table  shows,  however,  that  the  same 
soil  conditions  that  favored  bitter-pit  also  tended  to  increase  the  size 
of  the  fruit,  the  plats  standing  in  practically  the  same  order  as  to 
percentage  of  apples  larger  than  3X  inches  as  they  do  in  percentage  of 
bitter-pit. 

EXPERIMENTS   ON  JONATHAN  APPLES   IN    1915 

Irrigation  experiments  were  made  on  Jonathan  apples  similar  to  those 
already  reported  on  Grimes.  The  work  was  carried  out  in  an  orchard 
at  Wenatchee,  Wash.  The  soil  was  a  rich  gravelly  loam,  with  a  con- 
siderable percentage  of  clay,  underlain  at  a  depth  of  i6  inches  with 
a  layer  of  medium  fine  gravel.  For  several  years  previous  to  the 
beginning  of  the  experiments  the  orchard  had  been  heavily  manured  with 
slaughterhouse  refuse,  and  during  the  time  of  the  experiments  it  was 
kept  in  alfalfa.  The  trees  were  6  years  old,  and  there  were  5  trees  in 
each  plat.  The  experiments  were  begun  in  191 5.  Breaks  in  the  irri- 
gation canals  at  various  times  and  a  rainstorm  on  July  26  made  it  impos- 
sible to  secure  much  contrast  in  the  different  plats  before  the  first  of 
August.  All  the  trees  but  those  of  plat  i  were  extremely  dry  the  latter 
part  of  June  and  the  first  half  of  July.  A  further  report  of  this  condition 
is  given  later  in  this  paper  under  the  head  "Drouthspot."  There  was 
a  shortage  of  water  several  times  in  August,  plat  5  suffering  severely 
from  drouth  at  this  time  and  finally  losing  more  than  75  per  cent  of  its 
foliage  and  considerable  of  its  fruit  (PI.  4,  B).  Plat  3  suffered  from 
drouth  the  latter  part  of  August,  but  no  defoliation  occurred.  Plat  2 
was  practically  as  wet  as  plat  i  during  the  latter  part  of  July  and  first  of 
August,  but  became  quite  dry  about  the  middle  of  August.  The  moisture 
conditions  for  the  season  are  given  in  figure  7. 


Jan.  21, 191S     Irrigation  Experiments  on  Apple-Spot  Diseases 


121 


The  first  picking  was  made  on  September  3,  when  the  apples  were 
rather  green,  a  second  on  September  17,  when  they  were  right  for 
commercial  picking;  and  a  third,  October  i  when  the  fruit  was  dead 
ripe.  In  most  cases  a  bushel  of  apples  was  secured  from  each  tree  at 
each  picking.  There  was  practically  no  bitter-pit  on  the  fruit  at  picking 
time.  The  apples  were  placed  in  cellar  storage  at  an  average  tempera- 
ture of  about  47°  F.,  and  notes  were  taken  November  10.  The  results 
are  given  in  Table  VI. 

All  apples  more  than  2%  inches  in  diameter  were  counted  as  large, 
and  the  others  as  small.     There  was  little  contrast  as  to  size  in  the  fruit 


•     40^ 

40^' 

il 

.1 

^ 

, 

1 

B 

S  ^  J^  "^  ■>  '^ 
^  ?  ^  >  V  > 

?>?^^^7^ 

^i5J55iQ?5lQ 

^  Us   >>  <^  >v   Irs 

•^  n:  "^  X  n  vi 

?!Q^5^^5!Q 

<§<50  <  !fc< 

^  <  i$  «!.  (§  < 

S$  <  S§  <  J§< 

<$<  SQ<  i$  -1. 

io  <  !§<  (5  < 

lls^^l" 

l^llll 

IsIlP' 

|S3^^^ 

^  ^  ^  ^  i$  ^* 
5  5  5  ^  ^  "Si 

'^  ^  '^  ^  "^  to 

^  "5  ^  ^  "^  ^ 

^  ''j  *>  "^  '^  <0 

^  ^  ^  ^^  to 

*^^  ^  ^^  «0 

>» 

\ 

^ 

^^ 

«0 

K 

% 

k 

k 

K 

5 

5 

*i. 

^ 

*^ 

Q. 

^ 

Flo.  7. — Diagram  shor/ing  the  soil-moisture  conditious  iu  irrigated  plats  of  Jonathan  apples  in  1915.  The 
average  percentage  of  saturation  is  given  for  each  half  month  and  is  based  on  soil  samples  taken  at 
a  depth  of  16  inches.  Plat  i  was  to  receive  heavy  irrigation  throughout  the  season;  plat  2,  medium; 
plat  3,  light;  plat  4,  mediiun  in  August,  then  heavy;  and  plat  5,  heavy  till  August  i,  then  light.  The 
schedule  was  followed  as  closely  as  the  water  supply  ^v■ould  allow. 

of  the  different  pickings,  and  all  three  were  combined  to  obtain  the  data 
given  on  size. 

The  large  apples  again  have  much  more  bitter-pit  than  the  small  ones. 
The  apples  of  the  first  picking  had  more  than  twice  as  much  bitter-pit  as 
those  of  the  second  and  those  of  the  second  several  times  more  than  those 
of  the  third.  It  might  be  suggested  that  a  part  of  this  contrast  should  be 
attributed  to  the  fact  that  the  earlier  pickings  had  been  in  storage  longer, 
but  the  later  development  of  the  disease  in  storage  gives  no  support  for 
this  hypothesis.     The  more  mature  fruit  was  apparently  much  less  sus- 


122 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  3 


ceptible  to  the  disease.  A  study  of  the  total  bitter-pit  as  given  in  the 
last  column  of  Table  VI  shows  effects  from  irrigation  similar  to  those 
obtained  on  Grimes  apples.  The  fruit  from  the  trees  receiving  medium 
irrigation  followed  by  heavy  irrigation  late  in  the  season  had  the  most 
bitter-pit,  and  that  from  the  trees  irrigated  heavy  both  early  and  late 
the  next  in  amount.  As  has  already  been  mentioned,  the  contrast 
between  plats  i  and  2  in  the  amount  of  irrigation  was  not  as  great  as 
intended;  the  latter,  however,  received  less  water  and  had  less  pit  than 
the  former.  Plats  2  and  5  had  but  little  bitter-pit,  even  on  the  large 
apples.  The  fruit  from  plat  2,  however,  was  of  an  inferior  quality  on 
account  of  the  sunscald  that  resulted  from  the  defoliation  of  the  trees. 

Table  VI. — Percentage  of  Jonathan  apples  affected  with  bitter-pit.     November  10,  IQI5 


Plat 
No. 


Irrigation  treatment. 


Percentage 

of  apples  of 

following 

sizes. 


Larger 
than 

inches 


inches 


small- 
er. 


Percentage  of  apples  affected  with  bitter-pit. 


First 
picking. 


Ap- 
ples 
larger 
than 
.  iVi 
inches 


Ajj- 
ples 


inches 

or 
small- 


er. 


Second 
picking. 


Ap- 
ples 

larger 
than 
2H 

inches 


Ap- 
ples 

inches 
or 

small- 
er. 


Third 
picking. 


Ap- 
ples 
larger 
than 
2H 
inches 


Ap- 
ples 

inches 
or 

small- 
er. 


Total. 


Ap- 
ples 
larger 
than 
2H 
inches 


Ap- 
ples 

inches 

or 
small- 


Large 
and 
small 


Heavy 

Medium 

Light 

Medium,  followed  by  heavy 

Heavy,  followed  by  severe 

drouth 


91.9 

82.4 
66.9 
92.9 

49.8 


8.1 
17.6 

33-1 
7-x 

so.  2 


32-9 
32- S 
13-7 
44.4 

4-S 


22.7 
S-9 
8.2 

16.7 

4.4 


15- o 

XI- 7 

S-4 

23.0 


16.7 

2.0 
1.4 
8.3 


4.0 

1-3 


18.5 

16.1 

5-9 

25.8 

3-4 


22.0 
2-3 
3-8 
9.1 

3-3 


18.6 
13-6 

S-7 
24-6 

3-4 


The  above  fruit  was  held  in  cellar  storage  and  a  second  examination 
made  on  February  7.  At  this  time  the  apples  were  cut  open,  and  any 
that  had  either  browning  of  the  vascular  tissue  or  surface  pitting  were 
counted  as  affected  with  bitter-pit.  The  results  are  given  in  Table  VII, 
the  percentages  being  computed  on  the  number  of  apples  that  were 
free  from  bitter-pit  at  the  time  of  the  last  note-taking.  There  was 
little  contrast  in  the  amounts  of  disease  on  the  different  pickings,  and 
the  three  are  considered  together. 

Table  VII. — Percentage  of  bitter-pit  on  Jonathan  apples.     February  7,  igi6 


Plat 
No. 


Irrigation  treatment. 


Heavy 

Medium 

Light 

Medium,  then  heavy.  . 
Heavy,  then  very  light 


Percentage  of  apples 

that  de- 

veloped  bitter-pit  in  storage 

from  November  lo  to 

February  7. 

Large 

Small 

L,arge  and 

Apples. 

Apples. 

apples. 

7.6 

5-5 

7-4 

2.  I 

.6 

1-3 

3-6 

1.9 

3-3 

1.9 

.  0 

1.6 

•9 

I.  2 

I.  0 

Total 
percent- 
age for 
season. 


24.  6 

14.7 
8.8 

25.8 
4-3 


Jan.  21, 1918     Irrigation  Experiments  on  Apple-Spot  Diseases 


123 


But  little  bitter-pit  had  developed  on  any  of  the  apples  during  the 
three  months  of  cellar  storage.  This  may  have  been  because  the  sus- 
ceptible apples  had  already  been  eliminated,  or  may  have  been  due 
to  the  fact  that  the  apples  were  in  an  open  package  and  finally  became 
slightly  shriveled.  The  relative  amounts  of  disease  on  the  apples  from 
the  various  irrigation  plats  is  little  different  from  that  given  in  Table  I. 
The  results  for  the  season  are  shown  in  the  last  column  of  Table  VII 
and  also  in  figure  8. 

EXPERIMENTS  ON  JONATHAN  APPLES  IN   I916 

The  bitter-pit  experiments  were  continued  in  191 6  in  the  Jonathan 
orchard  already  described.     The  irrigation  conditions  were  much  more 


yIPPLSS  PICk£0  ^SPTEMBEff  \S 


APPLES  PICKED  OCTOBER  I. 


mi 


■  PEO  CENT  OF  BITTER-Pir  DEVELOPED  BY  NOV. /O. 

3  PEP:  C^ENT  OF  BiTTEP-PtT  DEVELOPED  BETWEE^t  NOV.IOANO FEB-T 


c 


Fig.  8.— Diagram  showing  the  amount  of  bitter-pit  on  Jonathan  apples  in  1915.  The  solid  portions  of  the 
bars  indicate  the  percentage  of  apples  affected  with  bitter-pit  on  November  10.  the  white  portions  the 
percentage  developed  between  November  10  and  February  7.  All  of  the  percentages  are  based  on  the 
number  of  apples  at  the  beginning  of  the  experiment.    See  figure  7  for  soil-moisture  conditions. 

satisfactory  than  in  191 5.     The   percentages  of   soil  saturation  main- 
tained on  the  different  plats  are  shown  in  figure  9. 

All  of  the  trees  were  in  vigorous  condition  except  those  of  plat  3, 
which  were  apparently  suffering  from  the  effects  of  the  drouth  of  191 5. 
The  apples  of  this  plat  were  very  highly  colored,  while  those  of  plats 
5  and  7  were  rather  low  in  color.  The  first  picking  was  made  on  Septem- 
ber 22  and  a  second  on  October  2.  The  apples  of  the  first  picking  were 
undercolored  and  immature,  while  those  of  the  second  were  well  colored 
and  suited  for  commercial  picking.  Approximately  3  bushels  of  apples 
were  saved  from  each  plat  in  the  first  picking,  and  approximately  2 
bushels  from  each  in  the  second,  and  placed  in  cellar  storage.  There 
was  no  bitter-pit  on  the  apples  at  picking  time  and  none  had  developed 
by  October  24.  The  results  obtained  from  notes  taken  on  November  14 
27806°— 18 2 


124 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  3 


and  on  March  i8  are  given  in  Table  VIII.  From  the  time  of  the  storage 
of  the  fruit  till  November  14  the  average  temperature  of  the  cellar  was 
approximately  50°  F.,  and  the  average  relative  humidity  about  61  per 
cent.  From  November  14  to  March  18  the  temperature  averaged  38°  F., 
and  the  relative  humidity  80  per  cent. 

Table  VIII. — Percentage  of  Jonathan  apples  affected  with  hitter-pit  in  igi6 


Plat 
No. 


Irrigation  treatment. 


Heavy 

Medium 

Light 

Medium,  followed  by  heavy 

Heavy,  followed  by  light 

Alternating,  heavy,  medium,  heavy. 

Alternating,  heavy,  medium,  heavy, 

medium 


Percentage  of  apples  affected  with  bitter-pit. 


Nov.  14. 


First 
pick- 
ing. 


2.9 


Second 
pick- 
ing. 


D.  O 

•7 
.  o 
.  o 

•4 
•4 


Total. 


1.6 


Mar.  18. 


First 
pick- 
ing. 


4.9 

2.9 

2.6 

II.  I 

1.8 
3-0 

S-8 


Second 
pick- 
ing. 


2-3 

1-7 

.  o 

2.  2 

•3 
5-2 


Total 

for 

year. 


3-5 

2-3 
2.  2 

7-5 
1.4 

3-7 
4.2 


60;t 


I    t     I     I     I     I     I        I    I    I     I     t    I     I 

V  <o  ^  "o  V  <o  >.    N  <o  ^  <o  V  to  ^ 


^  •O  ^  "0  ^  *0  ^ 

I  I  1  I  I  I  I 

"^.  to  ^  to  •'^  <o  ^ 


V  *\  >.  to  >  »0  ^ 

I  r  I  I  I  I  I 

V  to  ^  to  "^  <<>  ^ 
iy  !y  ^  >> ..;  L<  J^ 

3( 


I  r 


<  I  I 


■  to  V  to  ^  tos 


60^ 


Fig.  9. — Diagram  showing  the  soil-moisture  conditions  on  plats  of  Jonathan  apples  in  1916.  The  average 
percentage  of  saturation  is  given  for  each  half  month  and  is  based  on  soil  samples  taken  at  a  depth  of  16 
inches.  Plat  i  received  heavy  irrigation  throughout  the  season;  plat  2,  medium;  plat  3,  Ught;  plat  4, 
medium  followed  by  heavy;  plat  s,  heavy  followed  by  light;  plat  6,  heavy  in  June,  medium  in  July,  and 
heavy  in  August  and  September;  plat  7,  heavy  till  the  middle  of  July,  medium  till  August,  heavy  the 
first  half  of  August,  and  medium  the  remainder  of  the  season. 


Jan.  21. 191S     Irrigation  Experiments  on  Apple-Spot  Diseases 


125 


The  relative  susceptibility  to  bitter-pit  of  the  apples  from  the  dif- 
ferent plats  was  the  same  as  in  previous  experiments,  the  fruit  from 
the  trees  receiving  heavy  irrigation  late  in  the  season  having  the  largest 
amount  of  disease,  that  from  those  heavily  irrigated  throughout  the 
season  the  next,  and  that  from  those  receiving  heavy  irrigation  followed 
by  light  having  the  least  (fig.  10).  As  was  found  in  the  experiments 
of  1 91 5,  the  apples  from  the  early  picking  showed  much  greater  sus- 
ceptibility to  bitter-pit  than  those  of  the  late  picking. 

The  size  of  the  apples  from  the  various  plats  and  the  relative  suscep- 
tibility of  the  different  sizes  to  bitter-pit  is  shown  in  Table  IX. 


P£Q  CENT  OF  B/Tr5P'PfT 

3  4  S 


77t 


P£P  CENT  OF  BiTTER-PtT  D£V£LOPeD  BY  NOV.  I-*. 


c 


J  pea  CENT  OF  BITTER-PIT  DEVELOPED  BETWEEN  NOV  14.  AND  MARCH  IB. 


Fig.  10. — Diagram  showing  the  amount  of  bitter-pit  on  Jonathan  apples  in  1916.  The  black  portions  of 
the  bars  indicate  the  percentage  of  apples  affected  with  bitter-pit  on  November  14  and  the  white  portions 
the  percentage  developed  between  November  14  and  March  18.    See  figure  9  for  scil-moisture  conditions. 

Table  IX. — Percentage,  according  to  size,  of  Jonathan  apples  affected  with  bitter-pit. 

March  18,  igi6 


Total 
num- 
ber 
of 
ap- 
ples. 

Percentage  of  apples  of  various  sizes. 

Percentage  of  apples  of  various  sizes 
afiected  with  bitter-pit. 

Total 
per- 
cent- 
age 
of 
bitter- 
pit. 

Plat. 

3'Ato 

.f3K 

mches. 

3Kt0 
inches. 

3  to 

3K 

inches. 

2j<tO 
3 

inches. 

inches 

and 
smaller. 

i'Ato 

.3K 

mches. 

3Kto 
mches. 

mches. 

23^  to 

3 
mches. 

inches 

and 
smaller. 

I 
2 

3 
4 
S 
6 

7 

579 
593 
445 
440 
946 
566 
542 

0-5 
•4 

5-7 
.8 

•7 

17.7 

1-3 
9.6 

4.  I 

3P-9 
22.  5 
1. 1 
42.8 
16.3 
43-3 
32-4 

59-2 
65- 9 

40.5 
34.5 
68.4 

43-3 
59-6 

4.2 
10.8 
57-7 

4-5 
14.  0 

3-4 
3-9 

50.0 
.  0 

18.2 
.  0 
.  0 

23.  I 

8-3 

20.  0 

9.1 

5-° 
3-8 

20.  0 
6.4 
3-9 
2.9 

6.3 

1-3 
2.  2 
4.4 
1-3 
•9 
I.  2 

3-  I 

0.  0 

1-5 
•  4 
.  0 
.  0 
.  0 

.  0 

3-5 
2-3 

2.  2 

75 
1-4 
3-7 
4.  2 

The  plats  receiving  heavy  irrigation  late  in  the  season  had  more 
large  apples  than  the  others.     The  amount  of  bitter-pit  on  the  fruit  of 


126  Journal  of  Agricultural  Research  voi.  xii,  no.  3 

a  particular  size  was  hardly  sufficient  to  form  a  basis  for  conclusions, 
but  it  is  evident  that  the  disease  was  worse  on  the  large  apples  than  on 
the  small,  and  that  with  the  exception  of  one  or  two  cases  where  there 
were  but  few  apples  heavy  irrigation  increased  the  amount  of  disease 
on  the  medium-sized  as  well  as  on  the  large  fruit. 

DISCUSSION   OF  RESULTS  OF   BiTTER-PiT   EXPERIMENTS 

The  results  of  the  various  experiments  have  been  uniformly  consistent 
in  showing  that  heavy  irrigation  favors  the  development  of  bitter-pit. 
Heavy  irrigation  throughout  the  season  has  given  less  of  the  disease 
than  medium  irrigation  followed  by  heavy,  and  light  irrigation  through- 
out the  season  has  resulted  in  more  bitter-pit  than  heavy  irrigation 
followed  by  light.  Heavy  irrigation  the  first  half  of  the  season  caused 
the  trees  to  develop  a  more  luxuriant  foliage  and  probably  produced  a 
lower  concentration  of  cell  sap  in  the  apples,  both  of  which  facts  would 
tend  to  make  the  fruit  less  susceptible  to  the  forcing  efifects  of  late  irriga- 
tion. The  amount  of  irrigation  in  August  and  September  has  apparently 
largely  determined  the  amount  of  disease. 

Sudden  changes  in  the  amount  of  soil  water  do  not  appear  to  have 
had  any  effect  upon  the  amount  of  disease.  No  evidence  has  been 
found  that  bitter-pit  is  brought  about  by  a  rupture  or  bursting  of  the 
cells. 

Large  apples  have  been  more  susceptible  to  bitter-pit  than  small 
ones,  but  the  increase  in  the  disease  from  heavy  irrigation  has  been 
almost  as  great  on  the  small  and  medium  sized  fruit  as  on  the  large. 
This  fact  is  brought  out  in  Tables  I,  V,  VI,  and  IX,  and  in  a  particularly 
striking  manner  in  figure  6.  Apparently  apples  are  not  susceptible  to 
bitter-pit  merely  because  they  are  large,  but  rather  because  of  condi- 
tions that  may  sometimes  accompany  an  increased  growth. 

The  results  as  a  whole  point  to  the  harmful  effects  of  heavy  late  irri- 
gation regardless  of  the  size  of  the  fruit.  In  looking  for  the  final  cause 
of  the  disease  not  only  the  direct  growth -forcing  effects  of  the  water 
should  be  considered  but  also  the  effects  of  the  excess  water  upon  the 
soil  flora  and  soil  solutes.  This  subject  will  be  more  fully  discussed  in  a 
later  publication  upon  the  effects  of  fertilizers. 

JONATHAN-SPOT 
HISTORICAL  REVIEW 

Jonathan-spot  was  first  reported  by  Scott  (17).  He  suggested  the 
possibility  that  the  trouble  might  be  due  to  the  effects  of  arsenate  of 
lead.  Later  Scott  and  Roberts  {18)  gave  a  fuller  report  on  the  disease, 
showing  that  it  could  not  be  due  to  the  effects  of  spraying  and  that 
while  fungi  were  sometimes  present  in  the  spots  they  could  not  be  taken 


Jan.  21. 1918     Irrigation  Experiments  on  Apple-Spot  Diseases  127 

as  the  C3,usal  agency.  They  considered  the  disease  of  a  physiological 
nature  and  found  that  it  could  be  partially  prevented  by  early  picking, 
prompt  cold  storage,  and  early  consumption  after  removal  from  storage. 

Norton  (15)  reported  that  spots  practically  identical  in  appearance 
with  the  Jonathan-spot  could  be  produced  by  the  gases  of  ammonia  and 
folmaldehyde. 

Cook  and  Martin  (5)  considered  Jonathan-spot  to  be  a  form  of  rot 
caused  by  a  species  of  Alternaria.  In  a  later  report  (4)  they  made  a 
distinction  between  the  small,  nearly  black,  typical  Jonathan-spots  that 
were  more  commonly  confined  to  the  dark  area  of  the  skin,  and  the 
larger  light-brown  "Alternaria"  spots  that  were  more  common  on  the 
lightest  area  of  the  skin.  They  reported  that  they  were  able  to  reduce 
the  amount  of  the  disease  by  keeping  the  apples  covered  with  glassine 
bags  during  the  latter  part  of  the  summer,  and  considered  that  this 
fact  furnished  further  evidence  that  the  spots  were  of  fungus  origin. 

DESCRIPTION   OF  JONATHAN-SPOT 

"Jonathan-spot"  is  the  term  applied  to  superficial  black  or  brown 
spots  that  are  especially  common  on  Jonathan  apples.  The  trouble  is 
also  found  on  Esopus,  Yellow  Newtown,  Stayman  Winesap,  and  other 
varieties.  In  the  early  stages  of  the  disease  only  the  surface  color-bear- 
ing cells  are  involved  and  the  spots  are  seldom  more  than  2  mm.  in  diame- 
ter, but  later  the  spots  may  enlarge  to  a  diameter  of  3  to  5  mm.,  become 
slightly  sunken  and  spread  down  into  the  tissue  of  the  apple  to  a  con- 
siderable depth.  In  this  later  stage  of  the  disease  rot  fungi  are  often 
present,  Alternaria  being  particularly  common. 

EXPERIMENTAL  WORK 

The  Jonathan-spot  experiments  were  carried  out  in  the  same  Jonathan 
orchard  and  on  the  same  apples  as  the  bitter-pit  experiments,  and  the 
details  in  regard  to  soil,  irrigation,  time  of  picking,  and  condition  of  stor- 
age have  already  been  given. 

In  1 91 5,  plat  5  suffered  severely  from  drouth  the  latter  part  of  the 
season,  the  trees  finally  losing  more  than  three-fourths  of  their  foliage 
and  the  fruit  becoming  badly  bronzed  by  the  sun.  Plats  2  and  3  also 
became  very  dry  in  August  but  there  was  no  defoliation.  The  soil  moist- 
ure conditions  for  the  season  are  given  in  figure  7.  The  first  picking  was 
made  September  3.  The  fruit  at  this  time  lacked  fully  10  days  of  being 
at  its  best  stage  of  maturity  for  picking.  A  second  picking  was  made 
on  September  1 7  and  a  third  picking  on  October  i .  The  fruit  of  the  last 
picking  was  highly  colored  and  dead  ripe. 

There  was  no  Jonathan-spot  at  picking  time.  The  results  of  notes 
taken  on  November  10  and  February  i  are  given  in  Table  X. 


128 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  3 


Table  X. — Percentage  of  Jonathan  apples  affected  with  Jonathan-spot  in  iqi^ 


Percentage  of  apples  afff ected  with  Jonathan- 

Percentage  of 

spot. 

apples  of  follow- 

ing sizes. 

Plat 
No. 

Total 

November 

10. 

February 

I. 

Date  of  picking. 

ber  of 

apples. 

Larger 

2H 

Ap- 
ples 

Apples 

inches 

and 
smaller. 

Ap- 
ples 

Apples 

inches 

and 
smaller. 

than 

inches 
and 

larger 
than 

Total. 

larger 
than 

Total. 

inches. 

smaller. 

2H 
inches. 

2H 
inches. 

I 

281 

84 

16 

24 

II 

22 

52 

50 

51 

2 

237 

86 

34 

18 

0 

IS 

45 

32 

41 

Sept.  X 

3 
4 

338 
250 

62 

38 
7 

5 

13 

3 
6 

4 
12 

54 
38 

72 
28 

62 

'^^r^    •  O 

93 

38 

I     S 

179 

49 

51 

2 

3 

3 

22 

18 

20 

I 

260 

98 

2 

34 

0 

33 

67 

17 

65 

2 

291 

81 

19 

18 

4 

15 

48 

45 

47 

Sept.  17 

3 
4 

354 
195 

79 
94 

2  I 

23 
17 

16 

2 1 

77 
93 

65 
67 

76 
92 

•^'^r  *"    */ 

6 

8 

16 

I     S 

56 

52 

48 

3 

0 

2 

38 

15 

27 

I 

183 

95 

5 

17 

22 

18 

79 

53 

77 

2 

198 

78 

22 

3 

I 

2 

72 

30 

63 

Oct.  I 

3 

267 

60 

40 

11 

4 

8 

79 

35 

61 

4 
I     S 

I 

173 

92 

8 

4 

0 

2 

62 

79 

64 

724 

92 

8 

26 

12 

25 

64 

47 

62 

Total     for     all 

2 

726 

82 

18 

14 

2 

12 

52 

37 

49 

pickings 

3 

959 

67 

ZZ 

14 

6 

II 

71 

58 

67 

4 

618 

93 

7 

12 

5 

II 

62 

54 

62 

I     5 

235 

50 

50 

3 

3 

3 

26 

18 

22 

There  was  more  Jonathan-spot  on  the  large  apples  than  on  the  small 
ones,  and  at  the  time  of  the  first  note-taking  there  was  more  on  the  fruit 
of  the  first  and  second  pickings  than  on  that  of  the  third.  Irrigation 
apparently  had  but  little  effect  upon  the  disease.  The  apples  from  plat 
5  had  the  least  Jonathan-spot;  but,  as  already  mentioned,  these  were 
badly  sunburned  and  therefore  not  suitable  for  use  in  comparison  with 
those  of  the  other  plats. 

In  1 91 6  the  experiments  were  continued  in  the  same  orchard.  All  the 
trees  were  in  a  healthy,  vigorous  condition  except  those  of  plat  3.  These 
were  the  same  as  used  in  plat  5  the  preceding  season  and  showed  the 
effects  of  the  previous  year's  drouth  in  their  thin  foliage  and  short  twig 
growth.  The  soil-moisture  conditions  for  the  various  plats  are  given  in 
figure  9.  Pickings  were  made  on  September  22  and  October  2.  The 
fruit  from  plats  5  and  7  was  rather  poorly  colored,  while  that  from  plat 
3  was  very  highly  colored.  The  conditions  of  storage  have  already  been 
given  in  connection  with  the  notes  on  bitter-pit.  The  results  for  the 
season  are  shown  in  Tables  XI  and  XII. 


Jan.  21. 1918     Irrigation  Experiments  on  Apple-Spot  Diseases 


129 


Table  XI. — Percentage  of  Jonathan  apples  affected  with  Jonathan-spot  in  igi6 


Plat 
No. 


Irrigation  treatment. 


Percentage  of  apples  affected  with  Jonathan-spot. 


November  14. 


First 
pick- 
ing. 


Second 
pick- 
ing. 


Total. 


March  18. 


First 
pick- 
ing. 


Second 
pick- 
ing. 


Total 

for 
year. 


Heavy 

Medium 

Light 

Medium,  followed  by  heavy 

Heavy,  followed  by  light 

Alternating,  heavy,  medium,  heavy 

Alternating,  heavy,  medium,  heavy, 

medium 


1-3 
3-8 


0-3 
o.  o 
0.0 

.6 
o.  o 

•4 


0.8 
2.  I 
II.  8 
3-0 
•3 
2.  I 


55-8 
51-2 
72.  2 
66.3 
18.  2 
53-3 


82.8 


16.  2 


33-9 


65-4 
54-  2 
66.5 
69.  I 
24.  o 
64.7 

23.8 


Table  XII. — Percentage,  according  to  size,  of  Jonathan  apples  affected  with  Jonathan- 
spot  in  igi6 


Total 
nimi- 
ber  of 
apples. 

Percentage  of  apples  of  various  sizes. 

Percentage  of  apples  of  various  sizes  affected 
with  Jonathan-spot. 

Plat 
No. 

3Hto 

3K 
mches. 

3KtO 

mches. 

3  to 
mches. 

inches. 

inches 

and 
smaller. 

3Mto 

3K 

mches. 

3J<to 
mches. 

3  to 

.    3K 

mches. 

2j^tO 
3 

mches. 

.  2H 
mches 

and 
smaller. 

Total. 

I 
2 

3 
4 
5 
6 

7 

579 

593 
445 
440 
946 
566 
542 

0-5 
•4 

5-7 
.8 

•7 
17.7 

1-3 
9.6 

4.1 

30-9 

22.  5 

I.  I 

42.8 

16.3 

43-3 
32-4 

59-2 
65- 9 
40-5 
34-5 
68.4 
43-3 
59-6 

4.2 
10.8 

57-7 

4-5 

14.  0 

3-4 
3-9 

72.8 
60.  0 

61.5 
53-4 
60.  0 
68.7 
18.8 

65-3 
22.  7 

66.8 
52- 9 
58.3 
77-7 
24.7 
69.8 
25- I 

66.6 
62.5 
73-2 
70.  0 
26.6 
63.2 
23-4 

65-4 
54-2 
66.5 
69.  I 
24.  0 
64.7 
23.8 

100.  0 

55- 0 

33- i 
38.2 
13.6 

On  November  14  the  Jonathan-spot  was  worse  on  the  apples  of  the 
first  picking  than  on  those  of  the  second,  but  by  March  18  this  condi- 
tion had  in  most  cases  been  reversed.  There  was  little  contrast  be- 
tween the  amount  of  disease  on  the  fruit  of  different  sizes.  The  con- 
trasts between  the  irrigation  plats  were  not  very  consistent,  but  in 
general  indicated  that  heavy  irrigation  favored  the  disease.  Plats  5 
and  7,  on  which  the  fruit  was  lowest  in  color,  had  least  of  the  disease. 

DISCUSSION    OF    RE;SULTS    on    JONATHAN-SPOT 

The  experiments  on  Jonathan-spot  have  furnished  little  in  the  way 
of  consistent  positive  results.  In  both  191 5  and  1916  the  apples  of 
the  early  picking  had  more  of  the  disease  than  those  of  the  late  picking. 
In  1 91 5  the  large  apples  developed  more  Jonathan-spot  than  the  small 
ones,  but  this  did  not  hold  in  19 16.  The  results  of  both  years  gave 
some  evidence  that  heavy  irrigation  was  more  favorable  to  the  disease 
than  light  irrigation,  but  there  was  nothing  to  indicate  that  the  amount 
of  soil  moisture  was  an  important  factor  in  determining  the  amount 
of  Jonathan-spot. 


130  Journal  of  Agricultural  Research  voi.  xii.  no.  3 

OTHER  PHYSIOLOGICAL  SPOT  DISEASES  OF  THE  APPLE 
DROUTHSPOT 

The  term  "drouthspot"  (2)  has  been  applied  to  certain  fairly  large 
areas  of  dead  brown  tissue  usually  occurring  near  the  surface  of  the  apple, 
but  sometimes  found  deeper  in  the  flesh.  The  disease  may  appear 
at  almost  any  stage  in  the  development  of  the  apple,  but  the  fruit 
appears  to  be  more  susceptible  after  it  is  one-third  grown.  The  spots 
are  usually  located  on  the  blossom  half  of  the  fruit.  In  typical  cases 
the  trouble  first  appears  as  large,  irregular,  water-soaked  spots  that 
often  have  a  reddish  margin  and  are  usually  covered  with  drops  of  a 
yellowish,  sticky  ooze  that  is  sweetish  to  the  taste,  and  later  hardens 
into  a  brittle,  crystalline-like  deposit  (PI.  3,  D).  At  this  stage  the  spots 
resemble  fireblight  infection  (caused  by  Bacillus  amylovorus)  and  are 
sometimes  mistaken  for  it.  Upon  cutting  the  apple  open  a  very  shal- 
low layer  of  dead  brown  tissue  is  found  in  the  region  of  the  vascular 
network  just  beneath  the  skin.  Occasionally  brown  streaks  follow  the 
vascular  bundles  deeper  into  the  apple  pulp.  The  afifected  tissue  is 
very  bitter  to  the  taste.  The  skin  of  the  apple  over  the  diseased  area 
finally  regains  its  normal  appearance;  but  growth  is  arrested  at  this 
point,  and  the  enlargement  of  the  surrounding  tissue  soon  gives  rise 
to  a  much  misshapen  apple  (PI.  3,  E,  F).  On  account  of  its  manner 
of  development,  the  disease  has  sometimes  been  referred  to  as  "spot- 
necrosis"  {13).  Mix  {14)  has  given  a  full  discussion  of  the  characters 
of  the  disease  as  it  occurred  in  the  Champlain  Valley  of  New  York. 

The  above  description  applies  particularly  to  the  trouble  as  it  has 
been  observed  on  Winesap  and  Stayman  Winesap  apples  in  the  irrigated 
sections  of  the  West.  It  was  first  produced  experimentally  at  Wenatchee, 
Wash.,  in  191 3,  by  subjecting  Winesap  trees  to  a  sudden  and  severe 
drouth  when  the  fruit  was  about  i  inch  in  diameter.  At  about  the 
same  time  it  was  observed  at  Peshastin,  Wash.,  on  Ben  Davis  trees  that 
had  suffered  from  a  similar  drouth.  It  occurred  again  at  Wenatchee  in 
1 91 4  and  in  191 5,  always  on  trees  that  had  been  subjected  to  a  sudden 
and  severe  drouth  and  that  had  been  making  a  normal  or  vigorous 
growth  earlier  in  the  season.  The  drouth  periods  resulted  from  trouble 
with  the  irrigation  canals.  The  affected  trees  were  usually  located  on 
shallow  soils  or  on  soils  underlain  with  coarse  gravel  at  a  slight  depth, 
thus  making  them  peculiarly  susceptible  to  drouth. 

In  1 91 5  a  series  of  drouth  periods  occurred,  the  first  and  most  severe 
coming  the  latter  part  of  June  and  the  first  of  July,  the  second  the  latter 
part  of  July,  and  the  third  about  the  middle  of  August.  At  the  time  of 
the  first  drouth  even  the  trees  on  deep  soil  began  to  suffer,  and  those  on 
shallow  soil  lost  a  large  percentage  of  both  their  foliage  and  fruit.  The 
fruit  that  remained  on  the  trees  was  much  shrunken  in  size,  sometimes 
being  reduced  to  two-thirds  its  normal  diameter.     White  Pearmain  ap- 


Jan.  21  191 8     Irrigation  Experiments  on  Apple-Spot  Diseases  131 

pies  became  very  badly  shriveled  and  wrinkled  (PI.  5,  C),  and  Jonathan 
and  Delicious  apples  showed  slightly  less  serious  effects;  but  with  the 
return  of  irrigation  water  all  of  these  regained  their  turgor  without  spot- 
ting. The  Winesap  and  Stayman  Winesap  apples  did  not  become  as 
badly  shriveled  as  the  White  Pearmain,  but  they  developed  typical 
drouthspots  before  they  became  shriveled.  It  was  also  observed  that 
the  oozing  of  the  fruit  sap,  as  well  as  the  spotting  of  the  fruit,  preceded 
the  renewal  of  irrigation.  The  apples  subjected  to  the  early  drouth  were 
also  involved  in  the  later  ones,  and  the  result  was  a  series  of  spots  on  the 
same  apples  that  could  be  distinguished  as  to  time  of  formation  by  the 
color  of  the  skin  and  the  depth  of  the  pitting. 

On  September  3  samples  of  fruit  were  obtained  from  the  Jonathan 
trees  that  had  suffered  most  severely  from  drouth,  and  on  October  13 
similar  samples  were  secured  from  the  Winesap,  Stayman  Winesap,  and 
White  Pearmain  trees.  All  of  the  apples  were  placed  in  cellar  storage 
until  January  13  and  were  then  cut  open  and  examined.  The  Jonathan 
and  White  Pearmain  apples  had  developed  no  spots,  but  their  flavor  was 
decidedly  poor.  With  the  Winesap  and  Stayman  Winesap  apples  the 
spots  had  not  enlarged,  and  there  was  but  little  brown  tissue  beneath 
the  skin  (PI.  3,  F).  The  flavor  of  the  affected  tissue  was  bitter  and 
acrid,  but  that  of  the  rest  of  the  apple  was  normal. 

The  above  trees  that  had  suffered  from  drouth  appeared  to  recover 
largely  before  the  close  of  the  season  and  their  leaves  came  out  normally 
the  following  spring;  but  a  number  of  them  died  a  few  months  later,  and 
the  remainder  showed  a  lack  of  vigor  throughout  the  summer.  Their 
foliage  was  thin  and  they  appeared  to  suffer  from  drouth  even  with  a 
slight  decrease  in  the  percentage  of  soil  moisture.  The  usual  number  of 
irrigations  were  made,  and  there  were  no  real  drouth  periods;  yet  more 
than  half  of  the  apples  on  some  of  the  Winesap  and  Stayman  Winesap 
trees  developed  typical  drouthspots.  The  weak  condition  of  these  trees 
and  the  death  of  others  earlier  in  the  summer  probably  resulted  from  the 
destruction  of  some  of  the  smaller  roots  during  the  drouth  of  the  preced- 
ing season. 

CORK 

The  disease  or  group  of  diseases  called  "cork"  may  be  similar  to 
drouthspot  in  cause,  but  is  distinctly  different  in  many  of  its  gross  char- 
acteristics. Instead  of  being  subepidermal,  the  spots  are  located  in  the 
pulp  of  the  apple,  often  quite  deeply  seated  and  often  closely  associated 
with  the  larger  vascular  bundles  (PI.  5,  B).  The  patches  of  dead,  brown 
tissue  are  usually  much  larger  than  in  the  case  of  bitter-pit  and  much 
deeper  than  in  drouthspot.  They  resemble  the  internal  browning  of  the 
former  disease,  but  are  firmer  in  texture,  more  corky,  and  less  spongy. 
Affected  apples  are  often  slightly  less  firm  than  others,  and  usually  have 
a  cheesy  consistency  when  cut.     When  the  spotting  occurs  near  the  core 


132  Journal  of  Agricultural  Research  voi.  xii,  No.  3 

only,  there  is  usually  no  external  marking  to  indicate  the  disease;  but 
when  the  outer  pulp  tissue  is  affected,  depressions  occur  over  the  dead 
spot,  and  the  apple  becomes  more  or  less  roughened  or  corrugated  (PI. 
5,  A,  B),  The  development  of  the  disease  in  the  case  of  these  corrugated 
apples  is  similar  to  that  of  drouthspot  in  many  respects.  It  appears  first 
as  reddish  stains  on  the  surface  of  the  apples,  and  these  stained  areas 
may  gradually  become  water-soaked  and  covered  with  a  sticky  yellow 
ooze.  Later  the  skin  regains  its  normal  color,  but  large  areas  of  dead, 
brown  tissue  are  left  in  the  pulp. 

Apples  affected  with  cork  are  sometimes  also  affected  with  a  condition 
known  locally  as  "apple-blister."  The  trouble  first  appears  as  slightly 
raised  brown  or  reddish  spots  on  the  skin  of  the  apples  (PI.  5,  E).  The 
center  of  the  raised  portions  is  very  hard  and  corky,  but  only  the  outer 
epidermal  layers  are  involved.  As  the  apple  develops,  the  blisters  crack 
and  scale  off,  exposing  a  rough  corky  layer  that  has  formed  beneath. 
The  later  stages  of  blister  have  usually  been  found  on  apples  that  were 
also  affected  with  cork,  but  blister  appears 'early  in  the  spring,  very 
often  becoming  evident  as  soon  as  the  petals  have  fallen. 

Troubles  identical  with  cork,  or  very  similar  to  it,  are  quite  widely 
distributed.  They  have  been  observed  by  the  writers  in  the  Wenatchee, 
Entiat,  Spokane,  Okanogan,  and  White  Salmon  districts  of  Washington, 
in  the  Willamette  and  Hood  River  Valleys  of  Oregon,  in  the  Okanogan 
district  of  British  Columbia,  in  the  Champlain  Valley  of  New  York,  and 
in  various  apple  sections  of  Virginia  and  West  Virginia.  It  is  evident 
from  McAlpine's  reports  {9-12)  that  the  disease  is  of  considerable 
importance  in  Australia. 

McAlpine's  {9-12)  photographs  indicate  that  he  included  the  disease 
under  the  name  "bitter-pit."  Lewis  {8)  included  "  corerot "  and  "  dry  rot " 
as  forms  of  fruitpit  or  bitter-pit.  Allen  (j)  referred  to  the  disease  as 
"fruitpit."  Mix  {14)  has  very  carefully  distinguished  between  cork  and 
bitter-pit.  In  British  Columbia  the  disease  is  known  as  "  malformation  " 
and  in  Washington  as  "dryrot."  A  trouble  known  in  Virginia  as  "York- 
spot,"  or  "punky  disease"  {16),  and  in  California  as  "hollow-apple" 
are  apparently  very  closely  related  to  cork. 

The  losses  from  the  disease  are  usually  local,  but  sometimes  severe. 
At  Entiat,  Wash.,  in  1916,  two  carloads  of  apples  from  one  20-acre 
orchard  were  rendered  worthless  on  account  of  cork.  On  the  lower  flats 
of  the  Okanogan  Valley  in  British  Columbia  it  is  regarded  as  the  most 
serious  of  all  apple  troubles,  and  in  certain  sections  of  the  Hood  River 
Valley,  Oreg.,  it  was  the  cause  of  considerable  annual  loss  prior  to  the 
introduction  of  systematic  irrigation. 

The  cause  of  cork  is  not  known.  Allen  (j)  has  reported  that  fruitpit 
is  worse  on  trees  in  a  dry  soil  or  in  a  soil  lacking  in  organic  matter. 
The  disease  is  apparently  not  produced  by  fungi  or  insects.  The 
writers  have  made  repeated  attempts  to  isolate  an  organism  from  the 


Jan.  ax,  is»i8     Irrigation  Experiments  ou  AppleSpot  Diseascs  133 

affected  tissue,  but  with  negative  results.  Close  observations  have  been 
made  on  the  work  of  insects  in  orchards  where  the  disease  was  serious, 
but  no  evidence  has  been  secured  to  indicate  the  association  of  any  insect 
with  the  production  of  the  disease.  Orchards  affected  with  rosette  are 
sometimes  also  affected  with  cork,  but  the  latter  disease  occurs  in  orchards 
that  are  free  from  the  former.  In  nearly  every  case  where  the  disease  has 
been  observed  either  in  the  East  or  West,  its  occurrence  in  the  orchard 
has  been  closely  correlated  with  certain  peculiar  soil  conditions;  some- 
times an  excess  of  alkali  or  an  outcropping  of  slate,  but  more  often  a 
shallowness  or  openness  of  the  soil.  In  most  sections  cork  has  been  most 
serious  when  there  was  a  shortage  in  soil-water  supply,  either  resulting 
from  light  rainfall  or  a  lack  of  irrigation. 

An  orchard  at  Entiat,  Wash.,  that  has  been  seriously  affected  with 
cork  has  been  under  close  observation  for  the  past  three  years.  The 
orchard  is  located  on  a  low  bench  near  the  Columbia  River,  and  has  had 
a  permanent  cover  crop  of  alfalfa.  Soil  samples  from  the  orchard  showed 
that  in  the  sections  where  spotting  had  been  most  prevalent  the  surface 
soil  was  only  about  3  inches  deep  and  was  composed  of  a  coarse  sand 
with  only  a  small  percentage  of  humus.  The  subsoil,  which  was  more 
than  6  feet  deep,  differed  from  the  surface  soil  only  in  the  absence  of  the 
humus  and  was  underlain  with  coarse  gravel.  In  sections  of  the  orchard 
where  spotting  had  been  less  prevalent,  the  soil  was  found  to  be  a  much 
finer  sand,  and  in  sections  where  no  spotting  had  occurred  it  was  a  typical 
volcanic  ash,  very  fine  in  texture,  closely  compacted  when  wet,  and  very 
retentive  of  moisture.  Soil-moisture  determinations  made  soon  after 
the  spring  rains  showed  that  while  the  surface  soils  in  the  different  orchard 
sections  retained  their  moisture  fairly  well,  the  subsoil  in  the  first  section 
dried  out  quickly  and  that  in  the  last  section  was  very  retentive  of  its 
moisture.  It  will  be  seen  that  the  occurrence  of  the  disease  varied  with 
the  character  of  the  soil,  particularly  with  the  water-holding  capacity  of 
the  subsoil. 

The  irrigation  of  the  orchard  was  inadequate.  The  furrows  were  5 
feet  from  the  tree  rows,  and  alfalfa  growing  near  the  trees  and  in  the  tree 
rows  was  yellow,  frequently  wilted,  and  very  evidently  suffering  from 
lack  of  water.  The  trees  suffered  from  drouth,  especially  in  the  spring, 
before  the  irrigations  were  begun.  In  191 6  the  first  irrigation  was  made 
several  weeks  later  than  usual  and  the  trees  became  very  dry.  Later 
the  apples  developed  an  unusually  high  percentage  of  cork,  the  disease 
first  appearing  in  blister  form  soon  after  the  petals  had  fallen.  The  con- 
ditions in  the  orchard  indicated  that  the  soil-water  supply  was  at  least 
one  important  factor  in  determining  the  amount  of  disease. 

The  circumstances  under  which  cork  and  drouthspot  have  occurred 
in  the  Champlain  Valley  have  been  quite  fully  described  by  Mix  {14). 

A  special  form  of  cork  known  in  certain  sections  as  "Yorkspot"  and 
in  others  as  "hollow-apple"  has  been  found  particularly  common  on 


134  Journal  of  Agricultural  Research  voi.  xii,  no.  3 

York  Imperial  apples  and  has  also  occurred  on  the  Gano  and  the  Esopus 
varieties.  The  disease  has  been  under  close  observation  for  several  years 
at  Wenatchee,  Wash.,  and  in  the  summer  of  191 6  a  careful  study  was 
made  of  it  in  orchards  at  Staunton,  Va.  In  the  latter  case  the  disease 
v\ras  found  only  on  York  Imperial  apples.  It  could  not  be  correlated 
with  any  peculiar  soil  conditions,  but  was  found  decidedly  worse  on  trees 
that  were  lightly  loaded  than  on  those  with  a  medium  or  heavy  crop.  It 
was  much  worse  on  the  south  side  of  the  tree  than  on  the  north  side  and 
slightly  worse  on  the  east  side  than  on  the  west.  It  occurred  almost 
exclusively  on  apples  well  exposed  to  sunlight,  always  on  the  blush  side 
of  the  fruit,  and  always  on  fruit  surfaces  that  would  receive  the  oblique 
rather  than  the  direct  rays  of  light.  The  spots  were  similar  in  appear- 
ance to  cork,  but,  instead  of  being  scattered  over  the  apple,  were  often 
located  in  a  crescent-shaped  line  at  the  edge  of  the  blush  surface  of  the 
fruit.  In  some  cases  there  was  a  definite  ring  almost  entirely  surrounding 
the  point  which  received  the  most  direct  sunlight  (PI.  5,  F,  G).  The 
skin  of  the  apple  was  alwa3^s  normal,  and  the  corky  tissue  beneath  was 
usually  indicated  by  surface  depressions.  While  it  seems  probable  that 
Yorkspot  is  in  part  an  effect  of  drouth,  its  occurrence  is  undoubtedly 
greatly  influenced  by  sunlight  and  possibly  by  soil  conditions  and  other 
agencies. 

The  observations  reported  above  seem  to  indicate  that  cork  is  a  form 
of  drouth  injury;  yet  the  disease  appears  to  differ  from  typical  drouth- 
spot,  both  in  characteristics  and  conditions  of  occurrences.  With  certain 
varieties  of  apples  drouthspot  can  apparently  be  produced  on  any  soil 
under  conditions  of  sudden  and  extreme  drouth.  Cork  seems  to  be  the 
result  of  a  less  severe  but  more  chronic  drouth  on  trees  located  on  certain 
peculiar  soils,  especially  on  soils  that  are  lacking  in  humus  and  are  not 
retentive  of  moisture.  Blister  is  closely  associated  with  cork  and  is 
probably  produced  by  the  same  agencies. 

It  should  be  noted  in  this  connection  that  the  harmful  effects  of 
drouth  are  not  always  in  proportion  to  the  degree  of  desiccation.  Other 
factors  must  be  considered  in  a  study  of  drouth  troubles,  and  among 
these  are  the  percentage  of  harmful  substances  in  the  soil  water  and  the 
general  growth  condition  of  the  plant. 

SUMMARY 

(i)  Bitter-pit  and  Jonathan-spot  are  distinguished  from  rosy-aphis 
stigmonose,  drouthspot,  cork,  and  blister.  Bitter-pit  usually  appears 
first  as  spots  of  dead,  brown  tissue  in  the  subepidermal  portion  of  the 
apple.  These  spots  are  associated  with  the  terminal  branches  of  the 
vascular  bundles  and  in  later  stages  of  the  disease  the  browning  often 
follows  the  vasculars  deep  into  the  flesh  of  the  apple.  Rosy-aphis  stig- 
monose is  characterized  by  similar  brown  spots  in  the  subepidermal  region 


jaii.21,  i9i8     Irrigation  Experiments  on  Apple-Spot  Diseases  135 

but  the  affected  tissue  is  firmer  than  in  the  case  of  bitter-pit  and  there 
is  no  association  with  the  vascular  bundles.  The  early  stages  of  Jona- 
than-spot are  confined  to  the  color-bearing  cells  of  the  skin  of  the  apple. 
Drouthspot  is  characterized  by  the  checking  of  the  growth  at  certain 
points  on  the  apple  without  the  production  of  any  large  quantity  of  corky 
tissue.  Cork  differs  from  the  drouthspot  in  the  presence  of  compara- 
tively large  spots  of  brown  corky  tissue  and  in  the  fact  that  these  are 
usually  rather  deeply  seated  in  the  flesh  of  the  apple.  Blister  is  a  super- 
ficial lesion  associated  with  cork  and  characterized  by  its  blister-like 
appearance. 

(2)  Drouthspot  has  been  produced  by  sudden  and  extreme  drouth. 
It  has  occurred  on  trees  that  were  favorably  located  as  well  as  on  those 
that  were  growing  under  rather  unfavorable  soil  conditions.  Cork  is 
apparently  also  a  drouth  effect,  but  it  differs  from  drouthspot  in  the 
fact  that  its  occurrence  is  usually  associated  with  certain  peculiar  soil 
types. 

(3)  Experiments  have  shown  that  there  is  a  close  relationship  between 
the  soil-water  supply  of  the  orchard  and  the  development  of  bitter-pit 
in  storage.  Heavy  irrigation  has  greatly  increased  the  disease,  but  not 
so  much  as  medium  irrigation  followed  by  heavy  irrigation.  Light  irri- 
gation has  greatly  reduced  it,  but  heavy  irrigation  followed  by  light  has 
resulted  in  the  lowest  percentage  of  the  disease.  Sudden  changes  in  the 
amount  of  soil  water  have  apparently  not  increased  the  disease. 

(4)  Heavy  irrigation  may  have  been  slightly  favorable  to  the  develop- 
ment of  Jonathan-spot,  but  the  contrasts  have  been  too  slight  to  justify 
definite  conclusions. 

(5)  Large  apples  have  shown  greater  susceptibility  to  bitter-pit  than 
small  ones,  but  with  Jonathan  apples  heavy  irrigation  increased  the 
disease  on  the  medium-sized  fruit  as  well  as  on  the  large,  and  with 
Grimes  the  percentage  of  increase  from  heavy  irrigation  has  been  even 
greater  on  small  apples  than  on  large  ones.  Apparently,  large  apples 
are  not  susceptible  to  bitter-pit  merely  because  they  are  large,  but 
rather  because  of  certain  conditions  under  which  they  become  large. 

(6)  In  1 91 5  there  was  more  Jonathan-spot  on  the  large  apples  than  on 
the  small  ones,  but  in  191 6  there  seemed  to  be  no  correlation  between 
size  of  fruit  and  severity  of  disease. 

(7)  During  the  first  weeks  of  cellar  storage  there  was  always  more 
Jonathan-spot  developed  on  apples  that  were  picked  early  than  on  apples 
that  were  picked  late,  but  with  longer  periods  of  storage  these  contrasts 
seemed  to  largely  disappear.  The  results  indicate,  however,  a  greater 
susceptibility  in  the  early-picked  fruit. 

(8)  Bitter-pit  was  worse  on  the  Jonathan  apples  that  were  picked  early 
than  on  those  that  were  picked  late. 


136  Journal  of  Agricultural  Research  voi.  xii,  N0.3 

LITERATURE  CITED 
(i)  Allen,  R.  W. 

191 5.  CONDITION  OF  ROOT  SYSTEM  OF  APPLE  TREES  IN  THE  HOOD  RIVER  DISTRICT. 

In  Oreg.  Agr.  Exp.  Sta.  Rpt.  Hood  River  Branch,  1914-15,  p.  20-24, 
fig.  7-8. 

(2)  Brooks,  Charles,  and  Fisher,  D.  F. 

1916.  spot  diseases  of  the  apple.    In  Proc.  i2t±i  Ann.  Meeting  Wash.  State 

Hort.  Assoc,  1915,  p.  46-51,  i  fig. 

(3)  Cook,  M.  T.,  and  Martin,  G.  W. 

1913,  THE  JONATHAN  SPOT  ROT.    In  Phytopathology,  v.  3,  no.  2,  p.  119-120. 

(4) 

1914.  THE  JONATHAN  SPOT  ROT.    In  Phytopathology,  v.  4,  no.  2,  p.  102-105. 

(5)  Evans,  I.  B.  P. 

1909.  BITTER-PIT  OF  THE  APPLE.    Transvaal  Dept.  Agr.  Tech.  Bui.  i,  18  p., 
5  pi.     Bibliography,  p.  16. 

(6)  EwART,  A.  J. 

1913.  ON  BITTER  PIT  AND  THE  SENSITIVITY  OF  APPLES  TO  POISON.      (2nd.  Paper) 

In  Proc.  Roy.  Soc.  Victoria,  n.s.,  v.  26,  pt.  i,  p.  12-44,  pl-  3-5- 

(7) 

1914.  ON  BITTER  PIT  AND  SENSITIVITY  TO  POISONS.    (3rd  Paper)    In  Proc.  Roy. 

Soc.  Victoria,  n.s.,  v.  26,  pt.  2,  p.  228-242,  pi.  23. 

(8)  LEWIS,  C.  I. 

1915.  FRUIT-PIT  STUDIES  IN  THE  WILLAMETTE  VALLEY.      In  Ore.  AgT.  Exp.  Sta. 

2nd  Bienn.  Crop  Pest  and  Hort.  Rpt.,  1913-14,  p.  35-37,  fig.  8. 

(9)  McAlpinE,  D. 

i911-12.  bitter  pit  investigation.  the  past  history  and  present  posi- 
tion of  the  bitter  pit  question.  first  progress  report.  197 
p.,  34  pi.    Melbotune.     Literature,    p.  111-117. 


(10) 


(II) 


(12)- 


1912-13.  BITTER  PIT  INVESTIGATION.  THE  CAUSE  OF  BITTER  PIT:  ITS  CONTRIBU- 
UTING  FACTORS,  TOGETHER  WITH  AN  INVESTIGATION  OF  SUSCEP- 
TIBILITY AND  IMMUNITY  IN  APPLE  VARIETIES.  SECOND  PROGRESS 
REPORT.     224  p.,  61  pi.,  map.     Melbourne.     Literature,  p.  96. 


I913-14.   BITTER  PIT  INVESTIGATION.      THE  CONTROL  OF  BITTER  PIT  IN  THE  GROW- 
ING  FRUIT.      THIRD    PROGRESS    REPORT.      1 76    p.,    38   pi.,    5   maps. 

Melboiune.     Literature,  p.  96. 


1914-15.  BITTER  PIT  INVESTIGATION.  THE  EXPERIMENTAL  RESULTS  IN  THEIR 
RELATION  TO  BITTER  PIT,  AND  A  GENERAL  SUMMARY  OF  THE  INVES- 
TIGATION. FOURTH  REPORT.  178  p.,  41  pi.  Melbourne.  Litera- 
ture, p.  84. 

(13)  MELANDER,  a.  L.,  AND  Heald,   F.  D. 

1916.  THE  CONTROL  OF  FRUIT  PESTS  AND  DISEASES.    In  Wash.  Agr.  Exp.  Sta. 
Pop.  Bui.  100,  61  p. 

(14)  Mix,  a.  J. 

1916.   CORK,    DROUTH    SPOT    AND    RELATED    DISEASES    OF    THE    APPLE.      N.    Y. 

Geneva  Agr.  Exp.  Sta.  Bui.  426,  p.  473-522,  12  pi.     Literature  cited, 
p.  520. 

(15)  Norton,  J.  B.  S. 

1913.  JONATHAN  FRUIT  SPOT.     In  Phythopathology,  v.  3,  no.  2,  p.  99-100. 

(16)  Reed,  H.  S.,  and  Crabill,  C.  H. 

1915.  NOTES  ON  plant  DISEASES  IN   VIRGINIA  OBSERVED  IN   1913   AND   1914. 

Va.  Agr.  Exp.  Sta.  Tech.  Bui.  2,  p.  37-58,  17  fig. 


Jan.  21, 1918     Irrigation  Experiments  on  Apple-Spot  Diseases  137 

(17)  Scott,  W.  M. 

191 1.  A  NEW  FRUIT  SPOT  OP  APPLE.     In  Phjrtopathology,  v.  i,  no.  i,  p.  32-34. 

(18) AND  Roberts,  J.  W. 

1913.  THE  JONATHAN  FRUIT-SPOT.     In  U.  S,  Dept.  Agr.  Bur,  Plant   Indus. 
Circ.  112,  p.  11-16,  2  fig. 
(19)  SORAUER,  Paul. 

1900.  scHUTz  DER  OBSTBAUME  gegEn  krankheitEN.    238  p.,  iio  fig.    Stutt- 
gart. 

(20) 

1909.  handbuch  DER  pflanzenkrankheiten.    Aufl.  3,  Bd.  I.     Berlin. 

(21)  White,  Jean. 

1911.  bitter  pit  in  apples.    In  Proc.  Roy.  Soc.  Victoria,  n.  s.,  v.  24,  pt.  i, 
p.  1-19,  9  pi. 

(22)  Wortmann,  Julius. 

1892.  UEBER  die  sogenannte  "stippen"  der  aepfel.    In  Landw.  Jahrb., 
Bd.  21,  p.  663-675. 


PLATE  2 

A. — Early  stage  of  bitter-pit  on  Northern  Spy  apple  from  Westminister,  Vt.,  No. 
vember  i6, 1916. 

B. — Cross  section  of  the  apple  shown  in  A.  Brown  spots  are  evident  just  beneath 
the  skin,  and  a  few  others  can  be  seen  deeper  in  the  flesh  of  the  apple. 

(138) 


Irrigation  Experiments  on  Apple-Spot  Diseases 


Plate   2 


/ 


Journal  of  Agricultural  Research 


Irrigation  Experiments  on  Apple-Spot  Diseases 


Plate  3 


*v  . 


#■ 


¥ 


'  i 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  3 


PLATE  3 

A. — Late  stage  of  bitter-pit  on  Rhode  Island  Greening  apple. 

B. — Internal  browning  accompanying  bitter-pit. 

C. — Jonathan-spot  on  Jonathan  apple. 

D. — Early  stage  of  drouthspots  on  a  Winesap  apple  from  Wenatchee,  Wash.     The 
drops  of  exudate  can  be  seen  on  the  surface  of  the  apple. 

E. — Late  stage  of  drouthspots  on  a  Winesap  apple.     Note  the  deep  depressions 
scattered  over  the  surface  of  the  apple . 

F. — Cross  section  of  the  apple  shown  in  E.     Note  the  almost  entire  absence  of  brown 
corky  tissue. 

27806°— 18 3 


PLATE  4 

A. — An  apple  orchard  showing  the  furrow  system  of  irrigation  employed  in  the  ex- 
perimental work  at  Wenatchee,  Wash. 

B. — ^Jonathan  apple  tree  showing  the  effects  of  drouth,  Wenatchee,  Wash.  Photo- 
graphed on  September  i,  191 5. 


Irrigation  Experiments  on  Apple-Spot  Diseases 


Plate  4 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  3 


Irrigation  Experiments  on  Apple-Spot  Diseases 


Plate  5 


Journal  ot  Agricultural  Research 


PLATE  5 

« 

A. — Cork  on  Yellow  Newtown  apple  from  Hood  River,  Oreg.  Note  the  roughene, 
appearance. 

B. — Cross  section  of  the  apple  shown  in  A.     Note  the  area  of  brown  corky  tissue. 

C. — "White  Pearmain  apple  showing  the  severity  of  the  1915  drouth  at  Wenatchee, 
Wash.     No  drouthspots  were  developed  on  such  apples. 

D. — Cork,  or  "  dryrot ",  on  a  King  apple.  Note  the  brown  corky  tissue  near  the  core. 
In  surface  view  such  an  apple  appears  normal. 

E. — Blister  on  an  Esopus  apple  from  Entiat,  Wash. 

F. — ^An  extreme  case  of  Yorkspot  on  a  York  Imperial  apple.  Note  the  circular 
nature  of  the  injury. 

G. — Cross  section  of  the  apple  shown  in  F.  Note  the  pockets  and  the  brown  corky 
tissue  beneath  the  surface  depression. 


RELATION   OF  CARBON    DIOXID   TO    SOIL  REACTION 
AS  MEASURED  BY  THE  HYDROGEN  ELECTRODE  ^ 

By  D.  R.  HoAGtAND  and  L.  T.  Sharp,' 
Assistant  Chemists,  Agricultural  Experiment  Station  of  the  University  of  California 

INTRODUCTION 

In  a  previous  article  (ii) '  the  authors  have  presented  data  con- 
cerning the  question  of  soil  reaction  as  determined  by  the  hydrogen 
electrode.  Since  this  work  did  not  include  direct  measurements  of  the 
effect  of  carbon  dioxid  on  the  reaction  of  soils,  it  was  thought  desirable 
to  carry  out  further  experiments  on  this  point.  Before  discussing  the 
data  obtained  in  these  additional  experiments  it  will  be  well  to  emphasize 
again  the  fundamental  principles  upon  which  the  conclusions  of  our 
first  paper  were  based. 

The  present  tendency  to  advance  involved  explanations  of  the  nature 
of  soil  acidity  seems  to  be  unnecessary,  for  the  simple  conception  of  the 
relations  of  H-  and  OH-ion  concentrations  are  in  accord  with  the  facts 
so  far  ascertained  and  are  warranted  by  the  accepted  teachings  of  chemis- 
try. The  lack  of  agreement  in  the  literature  appears  to  be  due  to  the 
attempt  to  use  interchangeably  the  terms  "lime  requirement"  and 
"soil  acidity."  In  the  methods  of  determining  the  lime  requirement  it 
is  proposed  to  measure  the  amount  of  lime  required  to  bring  the  soil  to  an 
end  point  dependent  upon  arbitrarily  selected  conditions.  These  methods 
are  in  themselves  so  varied  and  the  final  measurement  of  reaction  so  dif- 
ficult, that  the  discordant  results  which  appear  in  the  literature  are  wholly 
to  be  expected. 

On  the  other  hand,  the  term  "soil  acidity"  has  a  definite  and  precise 
meaning — ^namely,  that  condition  of  the  soil  in  which  its  aqueous  solu- 
tion contains  H  ions  in  excess  of  OH  ions.  In  our  opinion  it  would  be 
preferable  to  refer  to  soil  acidity,  soil  neutrality,  and  soil  alkalinity  as 
those  phases  of  soil  reaction  in  which  the  H-ion  concentration  is  respec- 
tively greater  than,  equal  to,  or  less  than  the  OH-ion  concentration. 
These  H-  ion  concentrations  may  be  definitely  determined  by  measure- 
ments with  the  hydrogen  electrode. 

The  lime  requirement,  in  so  far  as  it  is  related  to  soil  acidity,  would 
consist  of  that  amount  of  lime  necessary  to  bring  an  acid  soil  to  the  neutral 
point  as  ascertained  by  the  above-mentioned  procedure.  Such  a  lime 
requirement  implies  that  the  dissolved  and  total  undissolved  soil  acids 
have  been  neutralized.     To  put  this  procedure  into  practice  may  involve 

1  From  the  Divisions  of  Agricultural  Chemistry  and  Soil  Chemistry  and  Bacteriology  in  equal  cooper- 
ation. 

2  Reference  is  made  by  number  (italic)  to  "Literature  cited,"  p.  147-148. 


Journal  of  Agricultural  Research,  Vol.  XII,  No.  3 

Washington,  D.  C.  Jan.  21,  1918 

Iq  Key  No.  Cal.— 13 

(139) 


140  Journal  of  Agricultural  Research  voi.  xii,  N0.3 

certain  inherent  difficulties.  The  reaction  is  so  prolonged  either  by  the 
slow  rate  of  solution  of  the  soil  acids  or  their  slow  diffusion  through  the 
soil  particles  that  the  point  of  neutrality  may  not  be  easy  to  establish 
and  maintain  permanently.  Thus  an  apparent  state  of  equilibrium  at 
the  neutral  point  may  be  attained  with  a  subsequent  slow  return  to  an 
acid  condition,  owing  to  the  solution  and  diffusion  of  the  soil  acids. 
But  once  the  total  soil  acids  have  been  neutralized,  a  further  return  to 
an  acid  condition  can  come  about  only  through  leaching  processes  or, 
possibly,  in  a  few  cases  through  decomposition  of  organic  matter. 

As  pointed  out,  some  soils  whose  solutions  are  neutral  or  alkaline 
remove  considerable  quantities  of  calcium  hydroxid  from  the  solution 
without  materially  increasing  the  OH-ion  concentration  of  the  soil 
suspension.  In  certain  cases  this  reaction  might  be  erroneously  attrib- 
uted to  the  soil  acids.  Although  the  term  "lime  requirement"  is  in 
common  usage  in  agricultural  literature,  it  has  been  variously  interpreted 
by  different  investigators.  At  present  the  term  is  devoid  of  scientific 
significance.  In  distinction  thereto  soil  reaction  whether  acid,  neutral, 
or  alkaline,  is  capable  of  precise  definition  and  determination. 

EFFECT  OF  CARBON  DIOXID  ON  SOIL  REACTION 

In  considering  the  effect  of  carbon  dioxid  on  soil  reaction  Maclntire 
(8)  states  that  many  acid  soils  when  extracted  with  water  saturated 
with  carbon  dioxid  yield  alkaline  extracts.  He  also  makes  the  following 
conclusion : 

Since  we  admit  that  the  soil  solution  is  the  medium  through  which  a  plant  absorbs 
its  mineral  supply,  we  are  compelled  to  conclude  that  a  plant's  sotu-ce  of  nutrition 
is  almost  always  alkaline,  but  of  varying  degrees  of  alkalinity. 

From  the  nature  of  the  chemical  equilibria  involved  {2,  9,  4),  we  have 
been  unable  to  reach  the  conclusion  that  a  solution  existing  in  contact 
with  an  acid  soil  can  ever  become  alkaline  owing  to  any  change  in  the 
partial  pressure  of  carbon  dioxid.  An  increase  in  carbon-dioxid  tension 
would  either  be  without  effect  upon  the  H-ion  concentration,  or  else 
would  increase  it,  depending  upon  the  relative  dissociation  constants  of 
carbonic  acid  and  the  soil  acids.  It  is  conceivable  that  certain  acid  soils 
when  extracted  with  water  saturated  with  carbon  dioxid  might  yield 
filtrates  which  would  give  an  alkaline  reaction  after  their  carbon-dioxid 
content  had  come  into  equilibrium  with  the  partial  pressure  of  atmos- 
pheric carbon  dioxid,  but  the  equilibria  governing  the  reaction  of  the  soil 
solution  in  contact  with  the  soil  are  by  no  means  identical  with  those 
regulating  the  reaction  of  the  filtrate  obtained  from  such  a  soil. 

In  order  to  obtain  direct  evidence  on  the  influence  of  carbon  dioxid  on 
soil  reaction,  the  experiments  reported  in  this  paper  were  undertaken. 
The  method  of  procedure  was  similar  to  that  previously  described  by  the 
authors,  with  the  addition  of  an  arrangement  for  controlling  the  partial 
pressure  of  carbon  dioxid  in  the  atmosphere  above  the  soil.     By  adopting 


Jan.  21,  i9i8        Relation  of  Carbon  Dioxid  to  Soil  Reaction  141 

a  simplified  form  of  the  apparatus  used  by  McClendon  and  Magoon  (7) 
and  McClendon  (6)  for  investigating  the  H-ion  concentration  of  sea  water, 
we  have  been  able  to  obtain  the  desired  data. 

DESCRIPTION  OF  APPARATUS  AND  RESULTS  OBTAINED 

The  hydrogen-electrode  apparatus  was  the  same  as  that  previovisly 
described  in  this  journal.  To  provide  a  chamber  for  mixing  the  hydrogen 
and  carbon  dioxid  a  graduated  i  ,000-c.  c.  cylinder,  the  base  of  which  had 
been  cut  off,  was  immersed  in  a  larger  cylinder  filled  with  mercury.  The 
upper  end  of  the  inner  cylinder  was  tightly  stoppered  and  contained  two 
capillary  stopcocks  for  admission  and  outlet  of  the  gases.  A  definite 
quantity  of  purified  hydrogen,  electrolytically  generated,  was  admitted  to 
this  cylinder  through  one  stopcock.  Through  the  other  stopcock  there 
was  admitted  from  a  gas  burette  a  known  quantity  of  pure  carbon  dioxid. 
Both  gases  were  measured  at  atmospheric  pressure.  A  sufiicient  time 
„was  then  allowed  for  the  thorough  diffusion  of  the  gases,  which  was  aided 
by  raising  and  lowering  the  inner  cylinder.  The  reservoir  of  mixed  gases 
was  then  connected  to  the  hydrogen-electrode  chamber  which  contained 
the  soil  suspension.  Forty  to  seventy  c.  c.  of  the  gas  mixture  were  forced 
into  the  space  above  the  soil  suspension,  adjusted  to  atmospheric  pressure, 
and  the  hydrogen-electrode  cell  was  then  closed.  Equilibrium  was 
hastened  by  the  shaking  method,  and  the  voltmeter  readings  were  noted. 
This  procedure  was  repeated  with  new  portions  of  the  gas  mixture  until 
the  voltmeter  readings  were  constant  to  within  0.005  volt.  The  experi- 
mental details  and  results  are  recorded  in  Table  I. 

The  term  "With  previous  car  bona  tion"  signifies  that  carbon  dioxid 
has  been  passed  through  the  soil  suspension  for  a  period  of  >^  to  2  hours 
previous  to  the  determination  of  the  H-ion  concentration.  The  purpose 
of  this  step  was  to  ascertain  the  effect  of  thoroughly  saturating  the  soil 
with  carbon  dioxid  upon  its  subsequent  reaction.  This  also  proved  to 
be  necessary  in  the  case  of  some  of  the  alkaline  soils  to  insure  complete 
saturation  of  the  carbonates  present,  thus  making  it  possible  to  attain 
the  final  equilibrium  when  using  the  smaller  percentages  of  carbon 
dioxid,  without  the  preparation  of  excessive  quantities  of  the  gas  mix- 
ture. By  satisfying  in  this  manner  the  capacity  of  the  soil  to  combine 
with  carbon  dioxid,  it  is  possible  to  reach  a  partial  pressure  of  carbon 
dioxid  above  the  suspension  in  the  hydrogen-electrode  cell  comparable 
to  that  in  the  mixing  cylinder. 

As  expressed  in  Table  I  the  o  per  cent  of  carbon  dioxid  means  that 
several  portions  of  pure  hydrogen  were  used  to  obtain  equihbrium. 
With  such  a  technic  the  loss  of  carbon  dioxid  from  the  soil  is  minimized . 
Thus,  the  atmosphere  above  the  suspension  will  undoubtedly  contain  a 
small  percentage  of  carbon  dioxid. 


142 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  3 


Table  I. — Effect  of  carbon  dioxid  on   the  reaction  of  soil  suspensions 


Description  of  soils. 

Quan- 
tity 
of 
soil. 

Quan- 
tity 
of 
water. 

Per- 
cent- 
age of 
carbon 
dioxid 
in  gas 
mix- 
ture by 
volume. 

Reaction  without 
previous  carbonation. 

Reaction  with 
previous  carbonation. 

Soil 
No. 

Read- 
ings on 

volt- 
meter." 

H-ion  con- 
centration. 

Read- 
ings on 

volt- 
meter." 

H-ion  con- 
centration. 

I 

Fine  sandy  loam  (California) 
do 

Gm. 

10 

C.c. 

so 

0.  OO 
•42 

1.90 
9.  00 
.00 
.  22 
.42 

•49 

1.90 

4.  80 

9.  00 

.00 

.42 

1.90 

4.80 

9.  00 

.  00 

.  22 

1.90 

9.  00 

.00 

4.80 

.  00 

1.90 

.00 

1.90 

.  00 

1.90 

.00 

.  20 

2.00 

.00 

.  20 

.42 

1.90 

4.90 

.00 

•50 

1.  90 
4.90 

0.589 
•590 

Gram-mole- 
cules per  liter . 
0. 40  X 10-^ 
.4oXio-< 

0.592 

Gratn-mole- 
cules  per  liter. 
0.37X10-' 

I. . . . 

do 

•  S89 

do 

•  SSo 
•783 

.S9Xio-< 
.17X10-' 

2 

3 

Silty  clay  loam  (California).. 
do 

lO 

SO 

•  763 
.742 

.38X10-' 
.92X10-' 

do 

•747 

.73X10-' 

do 

•744 
•  726 

.SsXio-' 
.  18X 10"* 

do 

.  726 
.709 
.688 
.896 

.18X10-6 
.34X10-' 
.8oXio-« 
.  20X  lO"* 

do 

do 

3 

Clay  loam  (California) 

do 

10 

5° 

3 

•773 
.768 

.  26X  lO"' 

do 

•  33  Xlo-» 

do 

•7S3 
.  762 
.  700 
.687 

.59X10-' 
.42X10-' 
.49Xio-« 
.84Xl0-« 

do 

•7S6 
.712 

•53X10^ 
.3iXio-« 

4 

4. . . . 

Clay  loam  (Pennsylvania)... 
do 

lO 

SO 

do 

.685 

.9oXro-« 

do 

•  66s 
.667 

•  644 
.658 

.20X10-* 
.18X10-* 
.46X10-5 
.26X10-6 

s.... 

Clay  loam  (  Louisiana) 

do 

10 

50 

.66s 
•647 
.663 
•653 
.685 
.664 
.602 

.  20X10-* 
.  40X 10-* 
.21X10-* 
.32X10-6 
.  90X  io-« 

.  2lXlO-* 

6.... 
6.... 

Silty  clay  loam  (Wisconsin). 
do 

lO 

SO 

7 

7 

Clay  loam  (Louisiana) 

do 

10 

SO 

.684 

.94Xl0-« 

8.... 
8.... 

Silty  loam  (California) 

do 

10 

SO 

.  604 
•  603 
.792 

.23X10-^ 
.24X10-^ 
.13X10-' 

.25X10-* 

do 

10 

SO 

.789 
.770 
.726 

.14X10-' 

.  30X I0-' 
.  I8XI0-* 

do 

9. . . . 

do 

10 

Silty  clay  loam  (California).. 

lO 

SO 

•7^0 

.30X10-' 

.764 
•7S8 
.724 
.  700 

.38X10-' 

.  49X IO-' 
.i9Xio-« 
.49X10-8 

do 

10. . . . 

do 

do 

do 

lO 

so 

.782 

.19X10-' 

II 

do 

•75S 
•735 
•  718 

.55X10-' 
.  12X10-* 

do 

do 

.24X10-0 

o  Mercury  cell,  with  mercuric  chlorid  and  potassium  chlorid  in  Nlio  concentration 

Since  the  measurement  of  the  H-ion  concentration  is  based  on  hydrogen 
at  atmospheric  pressure,  any  diminution  in  this  pressure  caused  by  the 
admixture  of  carbon  dioxid  would  result  in  a  certain  lowering  of  the 
electromotive  force.  Loomis  and  Acree  (5)  have  determined  the  changes 
in  electromotive  force  resulting  from  the  diminution  of  the  partial  pres- 
sure of  hydrogen  within  certain  ranges.  Their  data  indicate  that  these 
changes  are  so  slight  as  to  be  without  significance  in  the  present  investi- 
gation, hence  no  corrections  for  this  factor  have  been  made.  Most  of  the 
determinations  reported  above  are  the  averages  of  duplicates.  The 
agreement  of  these  duplicates  was  in  nearly  all  cases  within  0.005  volt. 

The  data  in  the  above  table  have  to  do  with  the  efifect  of  carbon  dioxid 
on  three  general  types  of  soil  reaction.     The  acid  type  of  reaction  is 


Jan.  at.  1918         Relation  of  Carbon  Dioxid  to  Soil  Reaction  143 

represented  by  soils  1,4,  5,  6,  7,  and  8,  the  neutral  or  slightly  alkaline 
type  by  soils  2,  9,  10,  and  11,  and  the  strongly  alkaline  reaction  by  soil  3. 

Considering  first  the  case  of  the  acid  soils,  we  note  that  the  increase  in 
H-ion  concentration  of  the  soil  suspensions,  in  contact  with  even  the 
higher  percentages  of  carbon  dioxid,  is  scarcely  greater  than  the  errors 
of  observation.  In  soils  i  and  8,  whose  suspensions  give  H-ion  concen- 
trations of  the  magnitude  io~*,  there  has  been  no  perceptible  change  in 
reaction  due  to  the  presence  of  carbon  dioxid.  The  other  acid  soils, 
which  have  a  smaller  concentration  of  H  ion,  show  some  slight  increase 
in  acidity  when  in  contact  with  the  mixture  of  gases  containing  carbon 
dioxid.  This  might  be  expected  from  a  consideration  of  the  relation  of 
the  dissociation  of  carbonic  acid  to  that  of  the  soil  acids.  Those  soils 
whose  acids  dissociate  comparably  to  carbonic  acid  are  not  measureably 
affected  by  the  partial  pressures  of  carbon  dioxid  used  in  these  experi- 
ments, while  soils  containing  less  dissociated  acids  have  their  H-ion  con- 
centration increased  to  a  sHght  extent  by  the  carbon  dioxid. 

The  H-ion  concentration  of  suspensions  of  the  slightly  alkaline  soils  is 
appreciably  increased  by  increasing  the  partial  pressure  of  the  carbon 
dioxid.  The  degree  of  increase  in  acidity  seems  to  be  dependent  upon 
the  proportion  of  carbon  dioxid  in  the  gas  mixture.  In  fact,  the  higher 
percentages  of  carbon  dioxid  brought  about  a  slightly  acid  reaction  in 
the  suspensions  of  these  soils. 

In  case  of  soil  3,  a  so-called  "alkali"  soil,  which  had  a  very  low  H-ion 
concentration,  a  large  increase  was  caused  by  the  introduction  of  carbon 
dioxid  into  the  hydrogen-electrode  cell. 

DISCUSSION   OF   RESULTS 

The  results  recorded  in  this  paper  show  that  the  effect  of  carbon  dioxid 
on  the  H-ion  concentration  of  soil  suspensions  is  not  an  insurmountable 
difiSculty  in  obtaining  the  reaction  of  soils  under  various  conditions  by 
means  of  the  hydrogen  electrode.  The  technic  followed  in  the  former 
investigation  in  which  the  loss  of  carbon  dioxid  is  minimized,  although 
not  entirely  avoided,  evidently  gives  results  of  the  same  order  of  magni- 
tude as  would  be  obtained  if  there  were  no  loss  of  carbon  dioxid.  The 
adoption  of  such  a  view  is  warranted  by  the  fact  that  the  maintenance 
of  a  small  partial  pressure  of  carbon  dioxid  above  the  soil  suspension  in 
the  hydrogen-electrode  cell,  only  altered  the  H-ion  concentration  by 
less  than  a  magnitude.  It  should  be  remarked,  however,  that  in  soils 
containing  alkali  carbonates,  and  having  a  high  OH-ion  concentration, 
the  partial  pressure  of  the  carbon  dioxid  exercises  a  very  pronotmced 
effect  upon  the  reaction,  as  instanced  by  soil  3.  For  a  clear  exposition 
of  equilibria  governing  such  systems  the  reader  is  referred  to  Johnston  {4) . 

So  far  we  have  not  considered  the  question  of  H-ion  concentration  in 
soils  under  field  conditions.  By  determining  the  carbon-dioxid  content 
of  the  soil  atmosphere  under  field  conditions  and  then  duplicating  the 


144  Journal  of  Agricultural  Research  Voi.xii.  No.  3 

partial  pressure  of  carbon  dioxid  in  the  manner  suggested  in  this  paper 
it  should  be  entirely  practicable  to  obtain  a  measurement  of  the  H-ion 
concentration  identical  with  that  of  the  soil  in  the  field.  Russell  and 
Appleyard  {10)  have  found  that  the  carbon-dioxid  content  of  the  soil 
atmosphere  under  different  conditions  varied  between  0.02  and  2  per  cent, 
the  general  mean  of  arable  soils  being  0.25  per  cent.  If  these  percentages 
of  carbon  dioxid  are  to  be  regarded  as  typical  for  field  soils,  then  in  view 
of  the  present  experiments  the  changes  in  carbon-dioxid  content  during 
sampling  and  laboratory  manipulation  would  not  invalidate  our  inferences 
with  regard  to  the  reaction  of  soils  under  natural  conditions. 

The  a  priori  considerations  already  presented  in  the  first  portion  of  this 
paper  with  respect  to  the  effect  of  carbon  dioxid  on  the  reaction  of  soils 
are  entirely  substantiated  by  the  experimental  data,  which  lead  to  con- 
clusions at  variance  with  those  of  Maclntire  {8)  on  this  point.  Even 
saturating  the  soil  suspension  with  carbon  dioxid  previous  to  measuring 
the  reaction  did  not  decrease  the  H-ion  concentration.  Therefore  it 
follows  that  an  acid  soil  would  never  present  to  the  plant  a  soil  solution 
of  alkaline  reaction,  notwithstanding  any  increase  in  the  partial  pressure 
of  carbon  dioxid.  It  should  be  recalled  that  the  criteria  heretofore 
used  for  judging  the  reaction  of  soils  do  not  always  permit  of  an  accurate 
distinction  between  soils  of  different  reactions.  Some  soils  may  be 
judged  as  acid  from  the  standpoint  of  certain  lime  requirement  methods, 
when  in  reality  their  reaction  may  be  alkaline.  For  this  reason  it  may 
be  doubted  whether  some  of  the  soils  reported  as  yielding  alkaline 
extracts  were  in  fact  acid.  Although  in  certain  instances  the  application 
of  lime  may  be  followed  by  an  increased  crop  yield,  this  result  may  not 
be  dependent  upon  any  change  in  the  reaction  of  the  soil  due  to  the  addi- 
tion of  lime.  The  more  accurate  interpretation  of  liming  experiments 
demands  that  an  attempt  be  made  to  differentiate  lime  as  a  neutralizer 
of  acidity  and  the  other  directly  or  indirectly  beneficial  effects  of  lime 
on  the  soil  or  the  plant. 

The  statement  frequently  met  in  the  literature  that  the  extracts  from 
soils  considered  acid  have  a  neutral  or  alkaline  reaction  has  led  to  the 
conclusion  that  water-soluble  acids  are  not  found  in  acid  soils.  We 
desire  to  emphasize  again  the  point  that  extracts  of  acid  soils,  especially 
those  prepared  with  carbonated  water,  might  become  neutral  or  alkaline 
after  the  loss  of  carbon  dioxid  from  the  extract.  Moreover  the  reported 
results  on  the  extracts  are  likely  to  be  misinterpreted,  for  they  do  not 
take  into  account  the  H-ion  concentration  but  are  based  upon  titrations, 
using  indicators  whose  end  points  may  be  removed  from  neutrality  by 
several  magnitudes.  Indeed  Gillespie  {3)  has  found  that  soils  determined 
to  be  acid  by  the  hydrogen  electrode,  yield  extracts  whose  reactions, 
colorimetrically  estimated,  were  in  close  agreement  with  the  hydrogen- 
electrode  measurements. 


Jan.  21. 1918         Relation  of  Carbon  Dioxid  to  Soil  Reaction  145 

One  of  Bouyoucos's  (i)  conclusions  from  his  valuable  and  ingenious 
application  of  the  freezing-point  method  to  soil  investigations  is  as  fol- 
lows: 

Since  no  mineral  soil  out  of  a  great  number  tested  gave  an  acid  curve  but  only  an 
absorption  curve,  and  inasmucb  as  the  free  acid,  and  acid  salt  produced  in  these  soils 
when  they  were  treated  with  neutral  salts,  or  acid  and  acid  salts,  were  carried  away  by 
washing  and  the  soils  then  gave  an  absorption  ctu^ve,  the  conclusion  seems  to  be  that 
the  presence  of  soluble  acids,  or  acid  salts,  in  the  mineral  soils  under  favorable  natural 
conditions  is  only  temporary,  if  ever  present,  and  never  permanent.  The  acidity 
or  lime  requirement  of  soils,  therefore,  seems  to  be  due  mainly  to  the  insoluble  acids 
of  the  soil,  the  silicic  acid,  silica,  acid  alumino-silicates,  and  perhaps  to  the  insoluble 
organic  matter.  There  appears  to  be  then  practically  no  active  acidity  in  the  mineral 
soils,  but  only  negative.  Exceptions  to  these  general  statements  are  probably  very 
few. 

Contrary  to  the  above  conclusion,  the  data  presented  in  this  and  other 
papers  {3,  11),  show  that  the  solution  in  equilibrium  with  the  soil  par- 
ticles of  certain  soils  contains  H  ions  in  excess  of  OH  ions.  Such  soils 
are  therefore  necessarily  acid  and  they  include  various  types,  many  of 
which  would  be  called  mineral  soils.  Furthermore,  it  is  well  to  bear  in 
mind  that  the  hydrogen  electrode  is  capable  of  measuring  specifically  the 
H-ion  concentration,  while  the  freezing-point  method  is  unsuited  to  this 
purpose.  This  is  especially  true  in  dealing  with  such  heterogeneous 
systems  as  those  of  the  soil  mass.  These  statements  are  not  to  be  con- 
strued as  denying  the  possible  value  of  the  freezing-point  method  in 
estimating  the  total  "lime  requirement." 

The  soil  acids  are  frequently  referred  to  as  insoluble,  but  such  insolu- 
bility is,  as  a  matter  of  fact,  only  relative,  for  complete  insolubility  is 
practically  unknown  and  the  soil  acids  must  therefore  have  a  definite 
solubility  although  it  may  be  slight.  The  important  consideration  is  that 
the  solution  of  an  acid  soil  must  be  continuously  acid  in  just  the  same 
way  that  a  solution  in  contact  with  silicic  acid  or  other  slightly  soluble 
acids  is  always  acid.  It  is  quite  true  that  when  the  soil  is  treated  with  a 
base  in  order  to  bring  about  a  condition  of  alkalinity,  by  far  the  greater 
part  of  the  base  is  used  in  combining  with  acids  which  at  any  given 
moment  were  not  in  the  solution.  But  this  is  in  no  way  opposed  to  the 
conclusion  that  the  solution  in  contact  with  an  acid  soil  is  an  acid  solution 
and  would  accordingly  offer  an  acid  medium  for  plant  growth.  The 
effect  of  such  a  medium  would  be  related  to  its  H-ion  concentration  and 
not  to  the  total  quantity  of  base  required  to  neutralize  all  the  soil  acids 
present.  In  other  words,  the  reaction  of  a  soil  is  concerned  with  the 
dissolved  fraction,  which  is  in  equilibrium  with  the  undissolved  soil  mass. 
If  any  added  substance — for  example,  calcium  carbonate — disturbs  this 
equilibrium,  then  it  is  clear  that  the  undissolved  portion  of  the  soil  will 
enter  into  the  reaction  in  accordance  with  the  laws  of  mass  action. 


146 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  3 


EFFECT  OF   POTASSIUM  CHLORID  ON   THE   H-ION   CONCENTRATION 

OF  SOILS 

As  previously  shown,  neutral  salts  added  to  suspensions  of  certain 
soils  considerably  increased  their  H-ion  concentration.  This  fact  has 
brought  up  the  question  as  to  the  effect  of  the  diffusion  of  potassium 
chlorid  from  the  agar  connecting  tube  into  the  soil  suspension.  In 
order  to  determine  the  magnitude  of  the  possible  changes  induced  in  the 
reaction  of  soil  suspensions  by  the  escape  of  potassium  chlorid  into  them, 
several  experiments  were  undertaken.  For  this  purpose  it  was  necessary 
to  eliminate  as  far  as  possible  the  diffusion  of  potassium  chlorid  into  the 
soil  suspension.  As  shown  in  Table  II,  one  procedure  consisted  in 
bringing  the  soil  suspension  into  equilibrium  with  hydrogen  without 
any  possible  chance  for  contamination  with  potassium  chlorid,  then  the 
electromotive  force  was  read  just  as  the  agar  tube  touched  the  suspension, 
thus  reducing  the  diffusion  of  the  potassium  chlorid  to  a  minimum. 
This  reading  was  then  compared  with  subsequent  readings  made  in  the 
manner  heretofore  described.  In  addition,  some  agar  tubes  prepared 
with  soil  extracts  were  substituted  for  the  potassium  chlorid  tubes. 
The  sensitivity  of  our  galvanometer  did  not  allow  of  a  greater  accuracy 
than  0.02  volt  when  the  soil-extract  tubes  were  used. 


Table  II. — Effect  of  potassium  chlorid 

on  the  reaction 

7/  soil  sus 

pensions 

Laboratory 
number  of 

With  first  contact  of 
potassium  chlorid  tube. 

After  many  contacts  of 
potassium  chlorid  tube. 

Soil-extract  tube. 

soil. 

Readings 
on  volt- 
meter. 

H-ion  concentra- 
tion. 

Readings 
on  volt- 
meter. 

H-ion  concentra- 
tion. 

Readings 
on  volt- 
meter. 

H-ion  concentra- 
tion. 

7 

0.681 

Gram-molectUes 
per  liter. 
0.  IlXlO"5 

0.  664 
.764 
•589 
•675 
.651 

•675 
.699 
.608 
•738 
.660 

•653 

Gram^nwleciUes 
per  liter. 
0.  21  X 10-5 
.38X10-^ 
.40X10-" 
.13X10-5 
•35X10-5 
.13X10-5 

.5iXio-« 
.  19X10-" 

.  IlXlO-8 

.24x10-5 
.33x10-5 

0.  661 

.771 

Gram-molecules 
per  liter. 

0.  23X10-5 

.  29X10   ^ 

2C 

I 

•595 
.677 

•653 
.678 
.  702 
.  610 

■739 
.660 

•653 

•33X10-" 
.12X10-5 
.32X10-5 
.12X10-5 
.46Xio-» 
.  18X10-" 
I.  oXio-7 
.24X10-5 
•33X10-5 

26 

27 

28 

2Q 

^I 

7.2 

■i-x 

?4 

It  is  evident  from  the  data  in  Table  II  that  the  slight  diffusion  of 
potassium  chlorid  from  the  agar  tube  has  a  tendency  to  increase  the 
H-ion  concentration  of  the  soil  suspension.  In  almost  all  cases  this 
increase  corresponds  to  less  than  0.005  volt.  For  most  agricultural 
purposes  this  difference  has  no  significance.  By  bringing  the  entire 
system  into  equilibrium  with  hydrogen  before  immersing  the  agar  tube 
and  by  keeping  the  tube  out  of  the  suspension  except  momentarily 
at  the  time  of  reading  the  electromotive  force,  it  is  believed  that  the 
error  will  be  entirely  negligible. 


Jan.  21. 1918         Relation  of  Carbon  Dioxid  to  Soil  Reaction  147 

SUMMARY 

(i)  The  H-ion  concentrations  of  soil  suspensions  have  been  measured 
under  various  partial  pressures  of  carbon  dioxid. 

(2)  The  H-ion  concentration  of  suspensions  of  acid  soils  is  not 
markedly  affected  by  increasing  the  content  of  carbon  dioxid  up  to 
10  per  cent.  The  H-ion  concentration  of  slightly  alkaline  soils  is  slightly 
increased  by  such  treatments.  A  notable  increase  in  H-ion  concen- 
tration is  observed  when  soils  containing  ^Ikali  carbonates  are  similarly 
treated. 

(3)  It  has  not  been  found  that  any  treatment  with  carbon  dioxid  can 
produce  an  alkaline  reaction  in  the  suspension  of  an  acid  soil. 

(4)  When  the  original  conditions  are  restored,  no  permanent  change 
in  soil  reaction  could  be  attributed  to  the  carbon  dioxid. 

(5)  Further  experiments  with  the  hydrogen  electrode  have  confirmed 
the  point  of  view  that  solutions  in  equilibrium  with  acid  soils  contain 
H  ion  in  excess  of  OH  ion. 

LITERATURE  CITED 
(i)  BouYOUcos,  G.  J. 

I916.  THE  FREEZING  POINT  METHOD  AS  A  NEW  MEANS  OF  DETERMINING  THE 
NATURE   OF    ACIDITY    AND   LIME   REQUIREMENT   OF   SOILS.      Mich.    Agt. 

Exp.  Sta.  Tech.  Bui.  27,  56  p.,  18  fig. 

(2)  Cameron,  F.  K.,  and  Bell,  J.  M. 

1907.   THE  ACTION  OF  WATER  AND  AQUEOUS  SOLUTIONS  UPON  SOIL  CARBONATES. 

U.  S.  Dept.  Agr.  Bur.  Soils  Bui.  49,  64,  p.,  5  fig. 

(3)  Gillespie,  L.  J. 

I916.  THE  REACTION  OF  SOIL  AND  MEASUREMENTS  OF  HYDROGEN-ION  CON- 
CENTRATION.    In  Jour.  Wash.  Acad.  Sci.,  v.  6,  no.  i,  p.  7-16,  2  fig. 

(4)  Johnston,  John. 

I916.  THE  DETERMINATION  OF  CARBONIC  ACID,  COMBINED  AND  FREE,  IN  SOLU- 
TION, PARTICULARLY  IN  NATURAL  WATERS.  In  Jotir.  Amcf.  Chem. 
Soc,  V.  38,  no.  5,  p.  947-975- 

(5)  LooMis,  N.  E.,  and  Agree,  S.  F. 

I916.     THE    EFFECT    OF    PRESSURE    UPON    THE    POTENTIAL    OF    THE    HYDROGEN 

ELECTRODE.     7«  JouT.  Amcf.  Chem.  Soc,  v.  38,  no.  11,  p. 2391-2396. 

(6)  McClendon,  J.  F. 

I916.  THE  COMPOSITION,  ESPECIALLY  THE  HYDROGEN  ION  CONCENTRATION 
OF  SEA   WATER   IN   RELATION   TO   MARINE   ORGANISMS.      In   JotU".    Biol. 

Chem.,  V.  28,  no.  i,  p.  135-152,  2  fig.     Bibliography,  p.  152. 

(7)  and  M.'VGOON,  C.  A. 

I916.  AN  IMPROVED  HASSELBALCH  HYDROGEN  ELECTRODE  AND  A  COMBINED 
TONOMETER  AND  HYDROGEN  ELECTRODE,  TOGETHER  WITH  RAPID 
METHODS    OF    DETERMINING   THE    BUFFER    VALUE   OF   BLOOD.      In   JoUT. 

Biol,  Chem.,  v.  25,  no.  3,  p.  669-681,  3  fig.     References,  p.  681. 

(8)  MacIntire,  W.  H. 

i916.  factors  influencing  the  lime  and  magnesia  requirements  of  soils. 
Tenn.  Agr.  Exp.  Sta.  Bui.  115,  48  p. 

(9)  MiCHAELis,  Leonor. 

1914.    DIE     W.VSSERSTOFFIONEN-KONZENTRATION.      2IO       p.,       41       fig.      Berlin. 

Literatiirverzeichnis,  p.  196-207. 


148  Journal  of  Agricultural  Research  voi.  xii,  N0.3 

(10)  RusSEL,  E.  J.,  and  ApplEyard,  Alfred. 

I915.   THE    ATMOSPHERE    OF    THE    SOIL:  ITS    COMPOSITION    AND   THE   CAUSES    OP 

VARIATION.     In  Jour.  Agr.  Sci.,  v.  7,  pt.  i,  p.  1-48,  17  fig. 
11)  Sharp,  L.  T.,  and  Hoagland,  D.  R. 

1916.    ACIDITY    AND   ADSORPTION   IN    SOILS    AS    MEASURED    BY    THE   HYDROGEN 

ELECTRODE.    In  Jour.  Agr.  Research,  v.  7,  no,  3,  p.  123-145,  i  fig. 
Literature  cited,  p.  143-145. 


ADDITIONAL  COPIES 

OF  THIS  PUBLICATION  MAY  BE  PEOCtJEED  FEOM 

THE  SUPEEINTENDENT  OF  DOCUMENTS 

GOVEENMENT  FEINTING  OFFICE 

WASHINGTON,  D.  C. 

AT 

15  CENTS  PER  COPY 
Subscription  Peice,  S3.00  Per  Year 

A 


Vol.  XII  JANUARY  2S,  1918  No.  4 

JOURNAL  OP 

AGRICULTURAL 
RESEARCH 


COISTXE^NTS 

Pag* 

A  Study  of  the  Plow  Bottom  and  Its  Action  Upon  the 

Furrow  Slice       -----        _        _        -      149 

E.  A.  WHITE 

( Contribution  from  Cornell  University  Agricultural  Experiment  Station ) 

Influence  of  Nitrates  on  Nitrogen-Assimilating  Bacteria   -      183 

T.  L.  HXLLS 

(Contribution  bom  Wisconsin  Agricultural  Experiment  Station) 


PUBLISHED  BY  AUTHORITY  OF  THE  SECRETARY  OF  AGRICULTURE. 

WITH  THE  COOPERATION  OF  THE  ASSOCIATION  OF  AMERICAN 

AGRICULTURAL  COLLEGES  AND  EXPERIMENT  STATIONS 


WASHINOXON,  r>.  C. 


WASHINQTON  :  GOVERNMENT  PRINTINQ  OFFICE  :  Itit 


EDITORIAL  COMMITTEE  OF  THE 

UNITED    STATES    DEPARTMENT    OF    AGRICULTURE    AND 

THE  ASSOCIATION  OF  AMERICAN  AGRICULTURAL 

COLLEGES  AND  EXPERIMENT  STATIONS 


FOR  THE  DEPARTMENT 

ICARL  F.  KELLERMAN,  Chairman 

Physiologist  and  Assodai-e  Chief,  Bureau 
of  Plant  Industry 

EDWIN  W.  ALLEN 

Chief,  Office  of  Exherimeni  Sfatians 


CHARLEvS  L.  MARLATT 

Entomologist  and  Assistant  Chief,  Bureau 
of  Entomology 


FOR  THE  ASSOCIATION 

RAYMOND  PEARL* 

Biologist,  Maine  Agricullural  Experiment 
Station 

H.  P.  ARMSBY 

Director,  Institute  of  Animal  Nutrition,  The 
Pennsylvania  State  College 

E.  M.  FREEMAN 

Botanist,  Plant  Pathologist  and  Assistant 
Dean,  Agricultural  Experiment  Station  of 
the  University  of  Minnesota 


All  correspondence  regarding  articles  from  the  Department  of  Agriculture  should  be 
addressed  to  Karl  F.  Kellerman,  Journal  of  Agricultural  Research,  Washington,  D.  C. 

*Dr.  Pearl  has  undertaken  special  work  in  connection  with  the  war  emergency; 
therefore,  imtil  further  notice  all  correspondence  regarding  articles  from  State  Experi- 
ment Stations  should  be  addressed  to  H.  P.  Armsby,  Institute  of  Animal  Nutrition, 
State  College,  Pa. 


JOINALOFAGRKMTIALISEMCH 

Vol.  XII  Washington,  D.  C,  January  28,  1918  No.  4 

A   STUDY   OF   THE    PLOW   BOTTOM   AND    ITS  ACTION 
UPON  THE  FURROW  SLICE  ^ 

By  E.  A.  White,  ^ 

Assistant  Professor  of  Farm  Mechanics 

College  of  Agriculture  of  the  University  of  Illinois 

INTRODUCTION 

The  most  ancient  records  show  that  from  a  very  remote  period  man 
has  used  the  plow,  in  one  form  or  another,  to  assist  him  in  stimulating 
the  earth  to  bring  forth  a  more  bountiful  harvest.  As  has  been  the 
case  in  many  other  lines  of  endeavor,  theory  has  trailed  far  behind 
observation  and  experience  in  developing  this  implement.  In  fact,  as 
far  as  can  be  ascertained,  it  was  not  until  the  last  half  of  the  eighteenth 
century  that  any  serious  attempt  was  made  to  develop  a  plow  bottom 
from  a  theoretical  standpoint,  and  even  then  the  productions  of  Jefifer- 
son,  Lambruschini,  Small,  Rham,  and  others  can  not  be  considered  as 
thoroughly  grounded  upon  well-developed  theories;  rather  their  works 
should  be  looked  upon  as  hypotheses  (fig.  i).  Experience  in  the  field 
generally  proved  that  the  machines  designed  by  these  men  were  not  all 
that  could  be  desired — ^for  example,  it  is  reported  ^  that  when  Lam- 
bruschini's  helicoidal  moldboard  was  taken  into  the  field  for  trial  the 
driver  of  the  draft  animals  immediately  observed  that  the  force  required 
to  move  this  plow  was  too  great  for  the  results  obtained.  To  be  sure, 
geometrically  exact  moldboards  furnished  the  basis  in  many  instances 
for  more  perfect  developments,  but  the  results  obtained  by  empirical 
plow  designers  who  worked  in  the  field  were  so  far  superior  to  the  results 
obtained  by  the  men  who  worked  in  the  laboratory  that  the  theorists 
were  soon  completely  outstripped  and  even  held  up  to  ridicule  by  the 
men  who  developed  their  machines  in  the  hard  school  of  experience, 
until  at  the  present  time  we  find  special  types  of  plow  bottoms  designed 

1  Approved  for  publication  in  the  Journal  of  Agricultural  Research  by  the  Director,  Cornell  University 
Agricultural  Experiment  Station. 

*  The  experimental  work  for  this  paper  was  done  under  the  direction  of  Prof.  H.  W.  Riley,  of  the  De- 
partment of  Rural  Engineering,  Cornell  University,  and  the  mathematioal  developments  were  prepared 
under  the  supervision  of  Prof.  F.  R.  Sharpe,  of  the  Department  of  Mathematics.  In  addition  to  the  above, 
grateful  acknowledgments  are  given  to  the  following:  To  Prols.  James  AIcMahon  and  Virgil  Snyder,  of 
the  Department  of  Mathematics,  for  their  most  timely  and  helpful  suggestions;  to  Mr.  J.  E.  Reyna, 
Instructor  in  Drawing,  College  of  Agriculture,  Cornell  University;  and  to  Mr.  L.  S.  Baldwin,  Instructor 
in  General  Engineering  Drawing,  University  of  IlUnois,  for  making  the  drawings. 

3  Lambruschini,  R.    d'ijn  nxjovo  orbcchio  da  coltri.    In  Gior.  Agr.  Toscano,  v.  6,  p.  37-80.    1832. 

Journal  of  Agricultural  Research,  Vol.  XII,  No.  4 

Washington,  D.  C.  Jan.  :8,  1918 

It  Key  No.  N.  Y.  (Cornell)— 3 

(149) 


I50 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  4 


to  meet  certain  field  conditions;  but  no  well-developed  theory  is  avail- 
able to  serve  as  a  guide  in  this  work. 

This  paper  is  an  attempt  to  begin  a  fundamental  analysis  of  the  plow 
bottom  and  its  work,  in  the  hope  that  some  light  may  be  thrown  upon  the 
theory  of  this  humble  but  perplexing  machine,  and  other  attempts  stimu- 
lated to  delve  further  into  the  secrets  which  are  still  to  be  revealed 
regarding  the  theory  of  this  important  implement.  Empirical  methods 
have  given  the  world  plow  bottoms  which  work  well.  It  is  still  to  be 
hoped  that  scientific  investigation  can  refine  and  further  perfect,  supple- 
ment as  it  were,  the  productions  of  experience. 

The  work  undertaken  by  the  writer  can  be  naturally  divided  into  three 
parts:  (i)  A  study  of  the  forms  of  plow  bottoms;  (2)  an  attempt  to 


Date 

Name 

Generatrix 

Directrices 

Eiquation  ofSurfaa 

Small 

Stnjiight  line 

Sfmigiit  line  E,  Catena/y 

Stephens 

5tra/gbt  b'ne 

Straight  line  and 
arc  ofCirc/e 

^  =  tan[fMJ 

1768 

JefTerson 

5tra/ghtb'ne 

Straight  lines 

l^//-^-- 

1616 

Da^/s 

/\nc  ofCin:Ie 

/ires  of  Circle 

1632 

Lamb/vschini 

5tra/gbt  Line 

Straight  Line  and 
tierir 

^  =  fanfaz) 

1639 

MtherviY 
andP/erce 

Arc  ofCyc/o/c/ 

Arcs  ofCycloiaf 

1640 

Rham 

Straight  line 

Straight  l/nes 

f-^=2nz 

f340 

Rham 

Stnsight  line 

Curves 

/852 

Knox 

Stroig/it  Line 

/Ires  of  Circles 

f?o/ed  sunbce  of 
e/ghth  order 

1854 

Gibbs 

Straight  line 

Arcs  of  Circles 

r.-f,-^-o 

/863 

l\deoa/ 

Straight  Line 

Arcs  of  Circles 

/867 

tio/b/vo/r 

Stroightline 

Straight  l/nes 

1634 

Jacobs 

/t  port/on  from  each  of  2  Sur/iices;  each  surface 
haWng  S  sets  ofsfraighf  //ne  generators. 

Pig.  I. — Diagram  giving  the  generatrices,  directrices,  and  equations  of  surfaces  of  historical  plow  bottoms . 

analyze  the  motion  of  the  soil  particles  as  they  pass  over  the  surface, 
and  (3)  a  mathematical  analysis  of  the  surfaces  of  the  most  important 
historical  plow  bottoms  which  were  designed  to  be  geometrically  exact. 
It  was,  and  still  is,  hoped  that  a  knowledge  of  just  what  the  plow  bottom 
is  and  how  it  performs  its  work  will  be  of  material  assistance  in  developing 
a  theory  which  will  furnish  a  very  definite  basis  for  the  proper  design  of 
this  fundamental  implement  of  tillage. 

FORMS  OF  THE  PLOW  BOTTOM 

A  study  of  modern  American-manufactured  plow  bottoms  reveals 
the  fact  that  a  large  number  of  these  are  so  constructed  that  their  surfaces 
contain  sets  of  straight  lines,  each  set  consisting  of  an  infinite  number  of 
straight  lines,  so  related  that  an  equation  or  equations  satisfied  by  the 
coordinates  of  points  on  the  surface  can  be  found. 


Jan.  25. 1918  Study  of  Plow  Bottoms  151 

Plate  6,  A,  represents  a  bottom  with  two  sets  of  straight  lines.  The 
few  lines  shown  in  the  illustration  indicate  that  through  every  point  of 
the  surface  two  straight  lines  can  be  drawn  which  lie  wholly  on  the  surface 
until  they  pass  off  the  edges  of  the  bottom.  These  straight  Unes  furnish 
the  basis  for  the  proof  that  such  a  surface  is  a  portion  of  an  hyperboloid 
of  one  sheet  (for  the  form  of  this  surface  see  fig.  3  to  7)  whose  equation 
can  be  developed  and  studied  with  mathematical  exactness.  The 
method  of  developing  this  equation  will  be  given  later,  but  at  present  we 
are  mainly  interested  in  the  fact  that  there  is  a  classs  of  plow  bottoms  on 
whose  surfaces  lie  sets  of  straight  lines,  and,  further,  that  one  equation  can 
be  developed  which  Vv'ill  approximately  represent  the  working  surface  of 
such  a  bottom. 

Further  study  shows  that  the  surfaces  of  other  plow  bottoms  contain 
sets  of  straight  lines,  but  that  one  equation  will  not  completely  describe 
such  a  surface.  In  Plate  6,  B,  a  bottom  is  shown  whose  surface  is  com- 
posed of  a  portion  of  each  of  two  surfaces.  Plate  6,  C,  shows  a  similar 
bottom,  but  in  this  case  the  two  surfaces  merge  into  each  other  farther 
back  upon  the  moldboard. 

In  Plate  6,  D,  a  class  of  bottoms  is  represented  whose  entire  surfaces 
do  not  contain  an  inlinite  set  of  straight  lines.  It  is  true  that  the  share 
and  back  end  of  the  moldboard  exhibit  the  same  characteristics  that  the 
first  two  classes  have  shown,  but  the  lines  do  not  continue  to  the  fore  part 
of  the  moldboard. 

Plate  7,  A,  shows  a  plow  bottom  with  a  convex  surface  which  has  two 
sets  of  straight  lines. 

The  American-manufactured  plow  bottoms  studied  can  thus  be 
divided  into  three  general  classes:  (i)  A  portion  of  one  quadric  surface; 
(2)  a  portion  of  each  of  tv/o  quadric  surfaces,  and  (3)  nonquadric  sur- 
faces. Nearly  all  forged  bottoms  belong  to  classes  i  and  2  with  the 
majority  falling  into  class  2,  while  most  of  the  cast  bottoms  belong  to 
class  3.  It  should  be  noted,  however,  that  some  recently  designed  cast 
bottoms  depart  from  the  general  characteristics  of  class  3  and  show 
clearly  the  two  quadric  surfaces  of  class  2.  The  lines  running  in  the 
general  direction,  front  to  rear,  marked  "/,"  (PI.  6,  A)  will  be  called 
longitudinal  lines,  and  those  running  in  the  general  direction,  top  to 
bottom,  marked  "t"  (PI.  6,  A)  will  be  called  transverse  lines. 

For  the  purpose  of  studying  the  forms  of  the  various  surfaces  under 
consideration,  a  machine,  illustrated  in  Plate  7,  B,  was  designed  and  built 
for  measuring  the  space  coordinates  of  any  desired  point. ^  By  means 
of  slots  and  a  system  of  pulleys  attached  to  the  drafting  board  the  cross- 
bar can  be  kept  horizontal  and  be  moved  both  laterally  and  vertically, 
while  the  drafting  board  is  attached  to  a  frame  which  can  be  moved 

'  Similar  machines  are  described  in  the  following  publications:  Gould,  J.  S.,  et  al.    report  on  trials 
OP  PLOWS.    /»  Trans.  N.  Y.  state  Agr.  Soc,  V.  27,  pt.  I,  1867,  p.  426.    1868. 
Giordano,  Fedetigo.    le  ricerche  sperimentali  di  meccanica  agrarl\.    p.  no.    Milanu,  1906. 


152 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  4 


backward  and  forward  upon  guides  so  marked  that  the  board  in  all  posi- 
tions will  be  squarely  across  the  guides.  When  a  plow  bottom  is  properly 
placed  upon  the  platform  the  x,  y,  and  z  coordinates  of  any  point  upon 

the  surface  can  thus  be  recorded  upon  coor- 
dinate paper  fastened  upon  the  drafting 
board. 


/ 
/ 

/■>^ey6'26y 


a'^zVs^e 


Y 


y 


y 


DBVEI<OPMENT  OF  THE   EQUATION 


.yCxy^' 


Q 

Fig. 


From  a  mathematical  standpoint  the  sur- 
face  shown    in    Plate    6,    A,   presents  the 
fj      problem  of  finding  the  equation  of  a  surface, 
x,yA     given  two  sets  of  straight-line    generators. 
This  can  be  done  if    the   equations   of  any 
three    lines   in   the    same    set   are   known. 
Select    three   Unes    ah,  cd,  and   ef   (fig.   2). 
Let  %!,  yi,  Zi,  and  X2,  y^y  z^  be  the  coordinates 
of  two  points  upon  line  ab;  x^,  y^,  z^,  and  x^,  y^,  z^  of  two  points  upon  line 
cd;  and  x^,  y^,  z^,  and  Xq,  y^,  z^  of  two  points  upon  line  ef. 
The  equations  of  the  lines  ab,  cd,  and  ef,  are 


(I) 
(2) 

(3) 


and 


X- 

-X, 

^2- 

-yi 
-yi 

z- 

-2l 

*2- 

-X, 

-2l 

X.- 

z3. 

y- 

Ji- 

2  - 

-Z3 

*4- 

-*3 

y^- 

-ys 

24- 

-23 

X- 

-^5 

-Iz 

-ys 

z  - 

3 

■^5        ^6-^5        2^5- 


From  (2)  the  following  equation  for  a  plane  perpendicular  to  the  XY- 
plane  and  containing  the  line  cd  is  obtained : 


W4=(«-  X3)  O4 -  ys)  -(y-ys)  (x^ -  X3) = O. 


(4) 


Similarly  from  (2)  the  equation  of  a  plane  perpendicular  to  the   YZ- 
plane  and  containing  the  line  cd  is 


^5=(y-y3)  (24- 23) -(^-23)  iyi-y3)=o. 


(5) 


From  (3),  the  equation  of  a  plane  perpendicular  to  the  XV-plane  and 
containing  the  line  ef  is 


i(x-x,)(y^-y^)-(y-y,)(x,-x,)  =  0. 


(6) 


Similarly,  from  (3)  the  equation  of  a  plane  perpendicular  to  the  VZ-plane 
and  containing  the  line  ef  is 


Uj^(y-y^){z^-z,)-  (z-z,)(y^-y^)  =  0. 


(7) 


Jan.  28,  i9i8  Study  of  Plow  Bottoms  153 

Consider 

Ui  =  Au^.  (8) 

where  A  is  a  constant.     This  is  the  equation  of  a  plane  which  contains 
the  intersection  of  planes  (4)  and  (5) ;  hence  it  contains  the  line  cd. 
Similarly 

u^  =  Bu^  (9) 

where  5  is  a  constant,  is  the  equation  of  a  plane  which  contains  the  line  ej. 
If  A  and  B  have  such  values  that  the  point  %',  y' ,  z'  is  on  (8),  (9),  and 
(i),  the  line  of  intersection  of  (8)  and  (9)  meets  (i)  and  is  a  generator 
(see  fig.  2).     Hence, 

_  {x'-x^)  {y-  y^)  -  {y'-y^){x-  x^) . 

{y'-y,){z-^,)-{z'-^^)iy.-yz)'  ^   ^ 

_  {x '  -  Xs)  (ye  -  ^5)  -  (:^ '  -  yo)  K  -  x^  . 

iy'-y.)i?.-^.)-i?'-z,)iy,-y,)'  ^"^ 


and 

x'-x^    y'-yi    z'-Zi 


=  K;  (12) 


«2-«i   y2-yi   22-2i 

where  7v  is  a  constant. 
From  equations  (12) 

x'  =  K(x2-Xi)+Xi  (13) 

y'  =  K(y^-y,)+yi  (14) 

z'=K(z2-Zi)+Zi  (15) 

From  equations  (10),  (13),  (14),  and  (15) 

i[K{x.2-Xi)+x^-X3](y^-y3))-([K{yr.-yj)+y^-  y3](x^-Xs))^  ,^. 

ilK(y2-yr)+yy-y3]{Zi-h))-i[K{^2-2i)+^i-h]iy*-y3))'  ^    ' 


and  from  equation  (8) 

J^JKy■-^■^{y^-y^-{y-y■^{'^^-'>^. 

(y- J'3)(24 -%)  -  (2-23)(j'4-3'3) 

From  equations  (11),  (13),  (14),  and  (15) 


(17) 


(\K{Xi-x^-\-x^-x^{y^-y^y)-{\K{y.2-y;)-\-y^-y^{:x^-x^)^  ,  ^. 

^~  ^\K(y.,-y,)\y,-y^{z^-z-^)-{\K(z^-z,nz,-z^\y^-y^)r  ^'""^ 

and  from  equation  (9) 

jj_^(^-^5)(y6-y5)-(y-?'5)(^6-^5)  t    \ 

{y-yC){2t-z^)-iz-zi){y^-y^)  ^' 

Eliminating  A^  B,  and  K  from  (16),  (17),  (18),  and  (19),  we  have  the 
equation  of  a  surface  through  the  lines  ab,  cd,  and  r/.  The  equations  are 
left  in  this  form  because  numerical  substitutions  are  more  easily  made 


154  Journal  of  Agricultural  Research  voi.  xii,  No.  4 

at  this  point  than  would  be  the  case  if  the  indicated  operations  were  first 
performed  with  the  symbols/  The  general  form  of  the  equation  resulting 
from  the  previous  operations  is 

ax'  +  by^  -{-cz^  +  2fyz  +  2gxz  +  2hxy  +  2lx  +  2my  +  2nz+d=^  O.         (20) 

To  reduce  equation  (20)  to  its  simplest  form  the  axes  must  be  trans- 
lated and  rotated. 

TRANSLATION    OF    AXKS" 

The  origin  of  equation  (20)  is  translated  to  the  center  by  putting 

x=x'+x„,       y-y'+y'o,       z+z'+z^;  (21) 

he  values  of  x„,  yo,  and  Zg  being  obtained  from  the  following: 

ax„  +  hyo+gzo+l  =  0  (22) 

hxo  +  byo+lzo  +  ^n  =  0  (23) 

gXo+fyo+cz„  +  n  =  0.  (24) 

These  substitutions  give,  after  dropping  the  accents  from  x',  y' ,  and  z' , 
an  equation  of  the  following  form : 

ax^  +  fer^  +  cz'  +  2fyz  +  2gxz  +  2hxy  +  (7  =  0;  (25) 

where  G  =  /.ro  +  wj,,  +  w2o  +  rf.  (25a) 

ROTATION    OF    axes' 

Equation  (25)  can  be  further  reduced  by  a  rotation  of  the  axes.     This 
is  accomplished  by  means  of  a  cubic  equation 

k:^-{a-\-b-\-c)k-'  +  {ab+ac  +  bc-p-g--h-)k-D  =  0;  (26) 


where  D  = 


a         h         g 
h         b  / 

n         j         c 


(26a) 


Let  the  roots  of  (26)  be  A',,  k.^,  and  k^.     The  desired  equation,  after  trans- 
lating and  rotating  the  axes  is 

k,x'+k,y!'  +  k^-^^^^^j^=0:'  (27) 

'  A  numerical  problem  is  ilcvclopeJ  by  this  motho<l  upon  pages  i,i6  lo  i6o. 

'  Snydkr,  Virgil,  and  Sis/vm,  C.  H.    analytic  gbombtry  oP  spacu,.    p.  77.    Now  York,  1914. 

'Idem,  p.  79. 

Mdcm,  p.  86. 


Jan.  28,  191S 


Study  of  Plow  Bottoms 


155 


SKeleton.  Hyperboloid  of  One  Sheet 
Fig.  3. 


Section  z=o.  Fig  3. 
Fig.  4. 


Section  y=o,  Fig.  3. 
Fio.  s. 


o        y 


Section   K'O,  Fig  3 

Fig.  6. 


Hyperboloid  of  One  Sheet,  showing  Lines 
upon  the  Sorfoce 


Pio.  7. 


156  Journal  of  Agricultural  Research  voi.xii.No.4 

where  A  =  DG.  (27a) 

The  direction  cosines  X,  m,  v,  of  the  angles  which  the  new  X-axis  makes 
with  the  original  axes  are  obtained  from  the  following: 

(a—ki)\+hij.+gv  =  0  (28) 

h-\  +  {b-k,)ti+jv  =  0  (29) 

9X+/M  +  (c-fei>  =  0  (30) 

2X+m2+v2=i.  (31) 

Similarly,  the  direction  cosines  of  the  angles  which  the  Y-  and  Z-axe 
make,  after  rotation,  with  the  original  axes  are  found  by  substituting 
k^  and  ^3,  respectively,  for  k^  in  equations  (28),  (29),  (30),  and  (31). 

When  equation  (27)  was  developed  from  the  surface  of  a  plow  bottom 
having  two  sets  of  straight-line  generators,  it  had  the  following  general 
form: 

This  is  the  equation  of  an  hyperboloid  of  one  sheet,  a  vase-shaped  figure, 
the  skeleton  of  a  section  of  which  is  shown   in  figure  3.     When  2  =  0 

equation   (32)   becomes  -^  +  r2=i,  and  the  cross  section  through  the 

plane  z=0   (fig.  4)  is  an  elUpse.     When  y  =  0,\.\\e  equation  becomes 

-^  —  -3=  I,  and  the  section  through  the  plane  y  =  0  (fig.  5)  is  a  hyperbola. 

Similarly,  when  %  =  0,  ^--3=  i  (fig.  6).     Figure  7  indicates  the  two  sets 

of  straight-line  generators  which  lie  on  the  surface  of  an  hyperboloid  of 
one  sheet  .-^ 

APPLICATION    OF   THB    DEVELOPMENT   TO    A    PROBLEM 

In  order  to  develop  the  equation  which  will  describe  the  surface  of  a 
plow  bottom,  it  is  necessary  to  obtain  the  data  called  for  in  equations 
(16),  (17),  (18),  and  (19).  This  application  of  the  development  will  be 
carried  through  for  the  bottom  represented  in  Plate  6,  A,  which  bottom 
was  placed  upon  the  machine  shown  in  Plate  7,  B,  so  that  the  origin  of 

JThe  constants  a,  b,  and  c  of  this  equation  do  not  necessarily  have  the  same  numerical  values  as  in 
previous  equations. 

*  The  method  for  obtaining  the  equations  of  any  line  on  the  surface  is  given  in  Snyder,  Virgil,  and 
SiSAM,  C.  H.    op.  cit.,  p.  93. 


Jan.  28,  J918 


Study  of  Plow  Bottoms 


157 


coordinates  came  at  O,  figure  8.  The  plane  y=0  contains  the  points  O, 
m,  and  n;  and  the  plane  x=0  contains  the  points  O  and  m  and  is  per- 
pendicular to  the  plane  y  =  0.  The  plane  2  =  0  is  perpendicular  to  both 
the  planes  y=0  and  x=0.  The  axes  are  considered  to  be  positive  in  the 
directions  indicated  by  the  arrowheads  (fig.  8).  Three  transverse  lines, 
ab,  cd,  and  ef  (fig.  8),  were  selected  and  the  following  data  obtained: 


Fig.  8. 

TABtE  I. — Values  (in  inches)  developed  for  the  surface  of  the  plow  bottom  shown  in 

Plate  6,  A 


Xi=  2.04 
J'i=  5-7 

2i  =  l6.0 
*4=    8.54 

74=  6.43 

24=23.0 


X2=  7.42 
J2=  3-78 
22=19.0 

X5=  9-7 

)'5=  10.88 
2t  =  26.0 


X3=   4-42 

J'3=    8.74 
23  =  20.0 

0:6=12.58 

>'6=    7-65 
2fi=28.0 


When   the   above  values  are  substituted   in   the  equations  already 
developed, 
From  (16) 


-2.28i<:  +  i3-83 


(33) 


158 


Journal  of  Agricultural  Research  voi.xii,  No.  4 


From  (17) 


From  (33)  and  (34) 


K  = 


A  = 


—  y— 1.78^  +  20 
i.299)/+3- 31.35' 


From  (18) 

From  (19) 

From  (36)  and  (37) 
K 


-  15.73;  +  io.04y-  13.832+ 1 1.95 
x-i.i77y-  2.282+51.45       ' 

1.584K  +  6.335, 


B  = 


B  = 


K-7.29 

—  %— .892}/  +  i9.4 
.619^+2-32.74  ' 


7.29^  +  2.58^-6.3352  +  65.85 
x+. 0887-1.542  +  32.46 


(34) 


(35) 


(36) 


(37) 


(38) 


By  eliminating  K  from  equations  (35)  and  (38)  the  following  equation 
for  the  surface  of  the  plow  bottom  is  obtained : 


3.9*2 +/  +  3.4522- 7.53:^/2- 7.28x2  +  6.79%)/ 

+  87.i«+i2o.75>'- 75.052  + 227.25  =  (9. 


(39) 


Table  II  is  compiled  for  purposes  of  checking  the  values  computed  from 
equation  (39)  with  those  obtained  by  measuring. 

Table  II. — Values  {in  inches)  for  the  surface  of  the  plow  bottom  shown  in  Plate  6,  A, 

obtained  by  measurement 


X  computed 

z 

y 

X 

from 
equation  (39) 

Difference. 

10 

2 

2.9 

2.  27 

0.63 

15 

6 

1-53 

1.56 

-    -03 

IS 

4 

3-58 

3-77 

-    .19 

15 

2 

6.9 

6.32 

•58 

20 

10 

3-72 

^•^ 

-    .08 

20 

8 

4.73 

4.76 

-    -03 

20 

4 

7-83 

7-94 

-    -13 

25 

12 

8.22 

8.12 

.  I 

25 

9 

9.07 

9.2 

-    •  13 

25 

6 

10.43 

10.  46 

-    -03 

30 

10 

14 

13.86 

•  14 

32 

9 

16.5 

16.  I 

•4 

Jan.  28,  1918 


Study  of  Plow  Bottoms 


159 


To  find  the  geometric  center,  substitute  the  coefficients  from  equation 
(39)  into  equations  (22),  (23),  and  (24).     Solving,  we  find 

X(,=  — 1.405  inches. 
yo=  6.52  inches. 
2o=      16.4    inches. 

This  translation  of  axes  is  shown  in  figure  9.  From  equation  (25a) 
G=— 57.3.  From  (25)  the  equation  of  the  surface  referred  to  parallel 
axes  through  the  center  is 

3.9XH/+ 3-452"- 7-5372- 7-28«2+ 6.79x^-57.3  =  0.  (40) 

y 


Fig.  9. 

To  find  the  equation  of  the  surface  referred  to  the  principal  axes  through 
the  center,  substitute  the  coefficients  from  (39)  into  (26),  and  we  have 

fe'-8.35fe2_  20.17^+2.45  =  0.  (41) 

On  solving  by  Horner's  method 

ki=  10.27 
^2=   0.128 
^3=    -2.05 

Substituting  the  values  just  found  for  k^,  k^,  k^,  D,  and  G  in  equation 

(27),  we  find 


or 


10.27*^+. 128^^—2.052-=  57. 3 


^  y^ 


I. 


(42) 


(2.36)2     (21.2)2     (5.29)^ 

The  direction  cosines  of  the  angles  which  the  axes  make  after  rotation 
with  the  original  axes  are  obtained  by  making  the  proper  substitutions  in 
equations  (28),  (29),  (30),  and  (31). 


i6o 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  4 


For  the  X-axis 


For  the  V-axis 


For  the  Z-axis 


7=  T0.6136 
H=  =Fo.48 
v=  ±0.627. 

7=  ±0.7515 
M==Fo.i437 
u=  ±0.6445. 


T0.1415 

±0.828 

±0.5425- 


Figure  10  shows  the  axes  after  translation  and  rotation  and  the  por- 
tion of  the  hyperboloid  of  one  sheet  which   is  a  close  approximation 

to    the   surface   of   this    plow 
i^  bottom . 


SURFACES  ONE  PORTION  FROM 
EACH  OF  TWO  QUADRIC 
SURFACES 


Fig,  10. 

were    obtained  from  the  share 
board. 


By  the  use  of  the  method 
which  has  just  been  em- 
ployed to  develop  the  equation 
of  the  surface  of  the  plow 
bottom  shown  in  Plate  6, 
A,  two  equations  can  be 
developed  which  will  approx- 
imately represent  the  surface 
of  the  bottom  shown  in 
Plate  6,  B.  By  taking  the 
origin  as  at  O,  figure  8,  the 
data  of  Tables  III  and  IV 
and    the  front   portion   of  the    mold- 


Table  III. — Values  {in  inches)  developed  for  the  surface  of  the  share  and  front  portion 
of  the  moldboard  of  the  plow  bottom  shown  in  Plate  6,  B 


Xi=  3.92 


yi= 

2i  = 


^4= 

3^4  = 


.8 
8.0 

=  6.78 

=  1-75 
=  16.0 


X2=    7-4 

y2=     -75 
£2=12.0 

.1:5=  2.36 

r5=  4-05 
25=15-0 


••^3=   1-73 

)'3=    2.67 

23=12.0 

•1^6=    5-87 

}•&=  2.7 


0.25%' +  2.34^^  +  0.462^— 3.2572— o.77a!;2:  + 2. 66:^^ 

+  6.88a;  +  32.3>/- 


5.812-4.4  =  0        (43) 


Jan.  28,  igiS 


Study  of  Plow  Bottoms 


161 


Table  IV. — Values  (in  inches)  for  the  surface  of  the  share  and  front  part  of  the  moldboard 
of  the  plow  bottom  shown  in  Plate  6,  B,  obtained  by  measuring 


X  computed 

2 

y 

X 

from  equa- 
tion (43) 

Difference 

10 

I 

4-75 

4-75 

0.  00 

10 

2 

1-75 

1.54 

.  21 

15 

I 

8.37 

9.  00 

-   .63 

15 

2 

5-47 

5-64 

-   .17 

15 

3 

3-77 

3.82 

-   -05 

15 

4 

I.  I 

1-3 

—    .  2 

From  the  remaining  surface  of  the  moldboard  the  following  data  of 
Tables  V  and  VI  were  obtained : 


Table  V. — Values  {in  inches)  of  rest  of  surf  ace  of  moldboard  shown  in  Plate  6,  B 


x^=  8.67 

yi=  4-95 

21=24.0 
x^=  9.08 

J'4=    9.0 

24=27.0 


X2=    4.96 

y2=  8.64 
22=22.0 

3:5=13.62 

^5=    6.23 
25  =  33-0 


^3=11-73 
yz=  4-8i 
23=29.0 

3tg=I2.24 

j'6=ii-89 

26=3 1 -O 


i.o73c^—  Loyy^+z^—  S-ggyz—  1.53:2-1-16.37x7 

-f6o.55x-f  1 25. ay- 48.52+ 109.5  =  0        (44) 

Table  VI. — Values  (in  inches)  for  the  rest  of  the  Tnoldboard  surface  shown  in  Plate  6,  B, 

obtained  by  measurement 


2 

y 

X 

I  computed 

from 
equation  (44) 

Difference. 

20 

2 

8.85 

8.68 

0.17 

20 

4 

6.67 

6.78 

—    .  II 

20 

6 

4.9 

4-95 

-    -05 

20 

8 

3-4 

3-5 

—    .  I 

25 

3 

10.  6 

10.  5 

.  I 

25 

5 

9-3 

9.2 

.  I 

25 

7 

8.23 

8.  12 

.  II 

25 
25 

9 
II 

7-4 
6.82 

7-34 
6.77 

.06 
•05 

30 

5 

12.  2 

12.  I 

.  I 

30 

7 

II.  7 

11.63 

.07 

30 

9 

11.38 

"•35 

•03 

30 

II 

"•3 

II.  24 

.06 

30 

13 

II.  4 

"•3 

.  I 

35 
35 

5 
7 

14.65 
14.72 

U-53 
14.66 

.  12 
.06 

35 

9 

15 

14-93 

.07 

35 
35 

II 
13 

15-45 
16.  I 

15-32 
15-85 

•13 

•25 

40 

8 

17-57 

17.  62 

-    -05 

40 

10 

18.5 

18.  52 

—    .  02 

i62  Journal  of  Agricultural  Research  vo1.xii,no.4 

From  a  study  of  Tables  II,  IV,  and  VI  it  is  evident  that  the  share  can 
not  be  as  accurately  described  by  mathematical  equations  as  can  the 
moldboard.  However,  the  differences  even  upon  the  share  are  not  very 
great.  It  must  be  remembered  that  these  surfaces  have  been  developed 
empirically;  experience  and  an  extensive  knowledge  of  the  conditions  to 
be  met  have  been  the  chief  guides.  Yet  this  implement  produced  in  the 
school  of  experience  has  a  surface  approximately  mathematically  exact 
in  form.  Further,  the  surfaces  of  cast  bottoms,  which,  because  of  the 
difficulty  of  manufacture,  are  not  changed  unless  necessity  demands, 
consist  in  some  cases  approximately  of  a  portion  from  each  of  two  quadric 
surfaces.  It  will  be  shown  later  in  discussing  the  history  of  the  plow  that 
the  surfaces  of  the  Holbrook  bottoms  were  designed  to  be  portions  of 
hyperboloids  of  one  sheet.  In  the  Utica  (N.  Y.)  plow  trials  these 
machines  received  many  first  awards  and  much  commendation  from  the 
judges  for  the  excellence  of  their  work.  In  addition  to  this,  Mr.  J.  J. 
Washburn,  of  Batavia,  N.  Y.,  who  knew  Mr.  Holbrook  and  was 
present  at  the  Utica  plow  trials,  stated  that  the  Holbrook  plows  did  as 
good  work  as  any  that  it  has  ever  been  his  pleasure  to  witness.  Thus, 
there  is  considerable  evidence,  based  upon  field  experience,  which  indi- 
cates that  a  portion  of  a  hyperboloid  of  one  sheet  is  the  proper  form  for 
the  surface  of  a  plow  bottom.  So  far  as  is  known,  this  hypothesis  awaits 
definite  proof. 

MOTION  OF  THE  SOIL  PARTICLES  IN  PLOWING 

For  the  purpose  of  studying  the  motion  of  the  soil  particles  in  plowing, 
the  work  was  limited  to  sod  ground  available  in  the  vicinity  of  Ithaca, 
N.  Y.  From  observations  on  a  sod  plow  at  work  in  the  field  (PI.  7,  C), 
the  following  general  facts  regarding  the  furrow  slice  were  noted : 

The  lower  outside  ^  edge  of  the  furrow  slice  did  not  appear  to  be  either 
stretched  or  compressed. 

The  upper  outside  edge  of  the  furrow  slice  appeared  to  be  compressed. 

The  inside  of  the  furrow  slice  was  stretched,  the  lower  edge  more  than 
the  upper  edge. 

As  the  furrow  slice  passed  over  the  moldboard  the  cracks,  which  had 
formed  on  the  inside  in  traveling  over  the  share  and  the  front  portion 
of  the  moldboard,  closed  up  as  the  soil  passed  over  the  rear  of  the  plow 
bottom,  indicating  a  point  of  maximum  stretching. 

The  above  considerations  made  it  evident  that  a  more  detailed  study 
of  the  behavior  of  the  furrow  slice  was  desirable.  For  this  purpose  rows 
of  pins  were  set  in  the  unplowed  ground,  the  pins  being  driven  in  the 
ground  to  the  estimated  depth  of  plowing,  as  shown  in  Plate  4,  A.  The 
longitudinal  rows  are  parallel  to  the  line  of  motion  of  the  plow,  which  is 
also  parallel  to  the  Z-axis  (fig.  8)  and  the  transverse  rows  perpendicular 

»The  portion  of  the  furrow  slice  immediately  adjacent  to  the  furrow  is  called  the  "outside." 


Jan.  28, 1918  Study  of  Plow  Bottoms  1 63 

to  this  same  line  of  motion.  The  longitudinal  rows  are  numbered  from 
II  to  VI  (Row  I  was  omitted  because  the  colter  upset  the  pins),  and  the 
pins  in  each  row  numbered  from  i  to  10,  as  shown  in  figure  11.  When 
the  part  of  the  furrow  slice  in  which  the  pins  were  set  was  upon  the  mold- 
board,  it  took  the  form  shown  in  Plate  8,  B.  In  order  to  obtain  the 
X,  y,  and  z  coordinates  of  points  in  the  furrow  slice  upon  the  moldboard, 
the  apparatus  shown  in  Plate  9,  A,  was  used.  In  this  apparatus  the 
axes  have  the  same  relation  to  the  plow  bottom  as  those  shown  in  figure 
8.  This  more  detailed  study  of  the  furrow  slice  upon  the  moldboard 
revealed  the  following: 

The  length  of  Row  II,  pins  i  to  10,  on  top  of  the   furrow   slice   was 
greater  than  the  length  before  the  soil  had  passed  upon  the  moldboard, 
indicating  that  this  por- 
tion of  the  furrow  slice  yt 
had  been  stretched. 

The  length  of  Row 
II,  pins  I  to  10,  was 
greater  upon  the  bot- 
tom of  the  furrow  slice 
than  its  length  before 
the  soil  passed  upon 
the  moldboard. 

The  length  of  Row 
VI,  pins  I  to  10,  on  top 
of  the  furrow  slice  was 
less  than  its  length  be- 
fore the  soil  passed  upon  the  moldboard,  indicating  that  this  portion 
of  the  furrow  slice  had  been  compressed. 

The  length  of  Row  VI,  pins  i  to  10,  on  the  bottom  of  the  furrow  slice 
was  greater  than  its  length  before  the  soil  passed  upon  the  moldboard. 

The  lengths  of  Rows  IV  and  V,  pins  i  to  10,  on  top  of  the  furrow  slice 
were  approximately  the  same  as  their  lengths  before  the  soil  passed  upon 
the  plow  bottom,  indicating  neither  compression  nor  stretching. 

The  lengths  of  Rows  IV  and  V,  pins  i  to  10,  on  the  bottom  of  the 
furrow  slice  was  greater  than  their  lengths  before  the  soil  had  passed 
upon  the  plow  bottom. 

The  z  distances  of  pin  10  on  top  of  the  furrow  slice  were  approximately 
the  same  for  each  row,  but  less  than  the  distance  which  the  plow  had 
moved  forward. 

The  z  distances  of  pin  10  on  the  bottom  of  the  furrow  slice  were 
approximately  the  same  for  each  row  and  equal  to  the  distance  which 
the  plow  had  moved  forward.  (The  coordinates  of  the  pins  at  the 
bottom  of  the  furrow  slice  were  measured  by  cutting  away  a  portion 
of  the  soil  but  leaving  the  pins  in  place.) 


Fig. 


164 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  4 


Fig.  h. 


These  observations  reveal,  first,  that  when  a  cross  section  of  the  furrow 
slice  is  considered  (fig.  12)  the  portion  marked  "A"  is  compressed  in 
plowing  and  the  portion  marked  "B"  is  stretched,  while  the  soil  in  the 
position  of  line  /;  is  neither  compressed  nor  stretched;  and,  second,  that 
there  is  a  definite  relation  between  the  z  coordinate  of  a  soil  particle  and 

the  distance  the  plow  has  moved  for- 
ward. This  relation  is  developed  on 
pages  164  to  167. 

The  next  step  was  to  analyze  in  detail 
the  motion  of  the  soil  particles.  This 
study  was  limited  to  the  soil  particles 
upon  the  bottom  of  the  furrow  slice,  but 
the  methods  developed  are  applicable  to 
other  portions.  The  paths  of  the  soil  particles  upon  the  bottom  of  the 
furrow  slice  can  be  very  accurately  traced  from  the  scratches  which 
they  make  upon  the  moldboard.  Plate  9,  B,  shows  the  paths  of  five  soil 
particles.  By  taking  the  axes  as  shown  in  figure  8,  a  projection  of  these 
paths  upon  the  plane  z-0  showed  a  very  uniform  set  of  curves.  Each 
of  these  curves  (fig.  13)  can  be  very  accurately  described  by  equations 
of  the  general  form 

aoc^  +  b'f  +  lx  +  'my  +  d  =  0.  (45) 

When  these  same  paths  are  projected  upon  the  plane  y  =  0,Q.  set  of  curves 
resulted  (fig.  14),  each  of  which  could  be  very  accurately  described  by 
equations  having  the  following  general  form : 

ax'^-\-hz''  +  lxz  +  inx  +  nz  +  d  =  0,  (46) 

From  equation  (45) 

-~-  and  -jZ  the  veloc- 
dt  df 

ity  and  acceleration, 
respectively,  of  a  soil 
particle  in  the  y  direc- 
tion  can  be  found  if 

dx  d^x 

-jr  and  -j^  are  known. 

dt  df 


dx 

The  values  of  —r, 

dt 


(Px 
df 


and 


-  I  I  I  I  I  r+y  t-'r 


/S      I*      13      12      //      10 


can  be  found  from 

dz 
equation  (46)  if  -7-  and 

d^z 

-Ta  are  known.     Thus, 

to  analyse  the  velocity  and  acceleration  of  any  soil  particle  whose  path 
upon  the  surface  of  the  plow  bottom  is  known,  an  equation  must  be 
found  between  z  and  time  (/). 


Fig. 


-Projectiou  of  tlie  paths  shown  in  Plate  9,  A,  upon  plane 

2  =  0. 


Jan.  28, 1918 


Study  of  Plow  Bottoms 


165 


This  was  accomplished  by  comparing  the  z  coordinates  of  the  bottom 
ends  of  the  pins  with  the  distance  which  the  plow  had  moved  forward. 
The  distance  which  the  plow  moved  forward  is  designated  by  s,  so  that 

s  =  vi,  (47) 

where  v  =  velocity  of  the  plow,  and  i  =  time. 

By  the  use  of  the  apparatus  illustrated  in  Plate  9,  A,  the  data  given  in 
Table  VII  were  obtained  for  the  soil  particles  upon  the  bottom  of  the 
furrow  slice  whose  paths  are  shown  in  Plate  9,  B.  These  data  are  typical 
of  12  sets  of  observations. 

Table  VII. — Values  (in  inches)  of  points  in  the  furrow  slice 


Row  II. 

Row  III. 

Row  IV. 

Row  V. 

z 

J 

z—s 

z 

s 

z—s 

z 

s 

z—s 

1 
z    1    s 

\ 

z—s 

i6i 

is! 

\ 

16 

iSf 

\ 

15^ 

iSf 

* 

i5i 

I  Si 

0 

2C4 

iqI 

i 

20| 

iQi 

^ 

19* 

i9i 

A 

2oi 

iqi 

i 

24 

2,a 

X 
4 

23^ 

23^ 

* 

24i 

23! 

i 

23i 

23i 

0 

2^\ 

27f 

0 

27^ 

27i 

-* 

27-1 

27i 

-i 

27i 

27i 

-i 

32i 

.^I* 

h 

32* 

3ii 

i 

3% 

3ii 

i 

31* 

3ii 

* 

35I 

35i 

-* 

3  si 

3  Si 

0 

35i 

35f 

* 

35i 

3Si 

-i 

39I 

Z9l 

-I 

Z9h 

39i 

-i 

39f 

39i 

-i 

39i 

39i 

-i 

Unfortunately  the  soil  available  in  the  vicinity  of  Ithaca  was  not  well 
adapted  for  taking  observations  of  the  kind  reported  in  Table  VII.  This 
soil  is  not  uniform  in  texture,  contains  many  stones,  cracks  much  more 
readily  than  it  stretches,  and  the  surface  is  not  as  level  as  could  be  desired 
for  this  work.  At  times  it  was  difficult  to  drive  the  pins  straight  into 
the  ground.  The  data  of  Table  VII  show,  however,  a  distinct  tendency 
for  the  difference  between  z  and  s  to  reach  a  maximum  value  and  then 
decrease  again  to  zero;  and  also  a  slight  tendency  for  this  maximum 
difference  to  decrease  from  Row  I  to  Row  V.  When  the  work  was  begun, 
it  was  hoped  that  sufficiently  accurate  data  could  be  obtained  from  which 
a  law  between  z  and  s  could  be  developed,  but  on  account  of  the  difficul- 
ties already  explained  this  was  impossible.  Consequently,  in  order  to 
develop  a  method  for  future  work,  a  set  of  conditions  were  assumed  which 
agreed  qualitatively  with  the  observed  facts.  It  should  always  be  kept 
in  mind  that  this  was  done  simply  as  an  hypothesis  whose  exactness  should 
be  thoroughly  tested  upon  a  soil  better  adapted  to  this  work.  The 
conditions  assumed  for  the  relations  between  z  and  s  are  as  follows : 

(A)  That,  for  each  path,  when  2  =  40,  5  =  40. 

(B)  That  there  was  no  stretching  or  compression  in  the  outside  bottom 
edge  of  the  furrow  slice  up  to  the  point  z  =  40. 

(C)  That  the  maximum  difference,  z—s,  for  Path  I  was  1.05  inches. 
27807°— 18 2 


i66 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  4 


\ 

• 

1 

^. 

^ 

1 

V 

vv 

\ 

V^ 

^> 

v\ 

^ 

\ 

\ 

^ 

\\ 

\\ 

\ 

\ 

\N 

X\ 

\ 

vC 

\ 

\^ 

A^ 

\\ 

\ 

\\ 

y 

\ 

\ 

\ 

\\ 

\ 

I  \ 

\^ 

\\ 

\ 

\\ 

V 

\ 

\\ 

\ 

\ 

\ 

\ 

i  \ 

\ 

■■i 

\ 

\  \ 

A 

\ 

\ 

\  \ 

\ 

\ 

\  \ 

I 

\ 

1 

\  ^ 

\ 

\ 

\ 

\ 

\ 

\ 

\ 

' 

\ 

\ 

\ 

c 

\ 

\ 

\ 

1 

Fig.  14. — Projection  of  the  paths  shown  in    Plate   9. 
the  plane  j'=0. 


A,    upon 


(D)  That  the  maxi- 
mum difference,  z—s, 
for  each  path  decreased 
uniformly  across  the 
furrow  slice.  Thus,  for 
Row  I,  ji(;= 0.85  inch,  the 
maximum  z—s=i.o5 
inches,  and  when 
3^=13.6  inches,  the 
width  of  the  furrow- 
slice,  the  maximum 
z—s=0;  so  when 
x=7.$  inches,  the  max- 
imum z—s  for  Row  V 
is  0.45  inch. 

(E)  That  the  stretch- 
ing in  each  row  took 
place  uniformly  up  to 
the  maximum  point 
and  then  decreased 
uniformly  until  it  was 
zero  when  z=s  =  40. 

(P)  That  the  maxi- 
mum stretching 
occurred  midway  be- 
tween the  point  where 
the  soil  particle  passed 
upon  the  plow  bottom 
and  the  point  ^  =  40. 
Thus,  for  Path  I  where 
the  soil  particle  passed 
upon  the  moldboard  at 
the  point  ^  =  0.6: 

40  —0.6  =  39.4  inches. 
39.4 -^  2  =  19.7  inches. 
19.7  +  0.6  =  20.3  inches, 

4  For  Path  I  the  point 
of   maximum    stretch- 

2  ing  was  at  ^•  =  20.3 
inches. 

^  The  computations 
below  show  that  for 
Path  V,  where  the  soil 


4iP 


36 


34 


32 


30 


28 


26 


24 


22 


ZO 


16 


10 


Jan  28,1918  Study  of  Plow  Bottoms  167 

particle    passed  upon  the    share  at    the    point  i^=ii.6,  the    point  of 
maximum  stretching  occurs  at  5^  =  25.8  inches. 

40  — 11.6  =  28.4  inches. 
28.4^2  =14.2  inches. 
14.2+ 11.6  =  25.8  inches. 

The  following  is  the  simplest  form  of  a  function  which  meets  the 
requirements  imposed  by  the  above  conditions  and,  when  the  constants 
are  determined,  will  describe  the  relations  between  2  and  ^  for  a  soil 
particle  on  the  bottom  of  the  furrow  slice  as  it  passes  over  the  surface  of 
the  plow  bottom: 

z  —  s  =  a{r  +  hs+cf  (48) 

From  equations  (47)  and  (48) 

z-vt=a[{vtf  +  hvt+cY;  (49) 

dz  d^z 

From  (49)  -j7  and  -^ ,  the  velocity  and  acceleration,  respectively,  of  a 

soil  particle  in  the  z  direction  can  be  obtained. 
From  equation  (46)  by  differentiation  we  have 

dx  dz 

{2ax+lz+fn)-^+ {2bz+lx+n)-j-  =  0;  (50) 

and 

+  (26^+/x+«)y,+(.6^  +  /^j^-0.  (51) 

Similarly  from  equation  (45)  we  find 


and 


(2a^  +  0j+(26:v+m)^=O;  (52) 

(2a.  +  //^+  -(!>+  (^^y+-)g^+  ^<|>=  O.  (53) 


.  dx   dy 

From  equations  (50),  (51),  (52),  and  (53)  the  velocities  -j.,  ^-,  and  the 

d  X  d/^ 
accelerations  -yw,  -r^-  of  a  soil  particle  on  the  bottom  of  the  furrow  slice 

dz  d/Z 

can  be  obtained  when  -n  and  -j^  are  known. 
at  dr 

In  this  problem,  however,  we  are  interested  in  the  accelerations  in  the 

directions  of  the  normal  to  the  surface,  designated  by  "A^,"  the  tangent 

to  the  soil  path  "  T,"  and  the  perpendicular  to  the  plane  formed  by  the 

normal  and  the  tangent  "7?." 


1 68  Journal  of  Agricultural  Research  voi.  xii.  no.  4 

We  can  find  Xi,  /Xi,  Wj,  the  direction  cosines  of  the  angles  which  N  makes 
with  the  X-,  Y-,  and  Z-axis  in  either  of  the  following  ways : 

If  (20)  (the  equation  of  the  surface  of  the  plow  bottom)  is  known,  we 
have  by  differentiation 

Xi ^^1 

ax^  +  byo  +  gzo  +  l     hxg  +  byo  +  fzo  +  m 

=  ^1  _  (rA) 

gxo+fyo+czo+^i 
I 


v 


l(aXf,  +  by^  +  gz^  +  lf+  {hx^  +  by^  +  fz^  +  mf' 

or  if  the  paths  of  the  soil  particles  are  known  but  the  equation  of  the 
surface  is  unknown  the  angle  A^^-  can  be  measured  by  means  of  a  pro- 
tractor and  plumb  bob,  as  shown  in  Plate  9,  C.  The  direction  cosines 
Xi  and  Ui  can  then  be  computed  from  the  following : 

(Xi)'+(mi)^+ (1^1^=1  (55) 

^  dx  ^      dy  ^     dz     ^  .  ^, 

where  the  values  for  ^,  J-,  and  -y.  can  be  obtained  from  (49),  (50),  and 

(52). 

,       dx  dy        ,  dz 
The  direction  cosines  of  T  (X^,  yUj,  1^2)  are  proportional  to  ^,  ^-,  and  ^' 

Hence 


dx    dy    dz        //^Y  ,  /^V  +  /^— V  (57) 

dt     dt     dt     '\\dt)'^\dt)^\dt) 

The  direction  cosines  of  T  (X3,  ^3,  v^  can  be  computed  from  the  follow- 
ing:^ 

(X3)'+(M3)'+(i'3)'=I-  (58) 

^' ^ ^^— =±i.  (59) 


M3l^2  ~  l^3M2    /^3«^1  —  l^sMl    Ml  1^2  —  l'lM2 

The  components  in  the  directions  N,  T,  and  R  of  the  forces  acting  on  a 
soil  element  of    mass  M,   moving  with    the  component  accelerations 
d^x  d^y       J  d^z 
W  df  ^^^  dt^  ^'^ 

d^x        d^v       d^z 
Fn  =  M(\^^+^,J  +  v,-i^,)  (60) 

FT-MCX.g^+^f+v.lf)  (6.) 

F.  =  M(X^  +  ,,^,  +  v,^).  (6a) 

'  Snyder,  Virgil,  and  Sisam,  C.  H.    Op.  cit.,  p.  40. 


Jan.  28, 1918  Study  of  Plow  Bottoms  1 69 

EVALUATING   THE   CONSTANTS   IN    EQUATIONS  (48),   (46),  AND  (45) 

The  methods  of  evaluating  the  constants  in  equations  (48),  (46),  and 
(45)  for  a  given  soil  path  will  now  be  considered.  For  this  purpose 
Path  V  (PI.  9,  A)  will  be  taken.     The  general  form  of  equation  (48)  is 

z-s  =  a(s^  +  bs  +  cy.  (48) 

From  the  assumptions  that  have  already  been  made  (p.  164  to  168)  the 
following  data  for  this  curve  are  obtained : 


5- 

2 

II.6 

11.6 

25-8 

26.25 

40.0 

40.0 

On  substituting  the  above  values  for  s  and  z  in  equation  (48),  three 
equations  are  obtained  from  which  it  is  found  that 

a  =  o.ooooiii4 
&=-5i-6 
0  =  464 
giving 

z— s  =  0.00001  II  4{s^—  ^i. 6s +4.64)^  (63) 

To  determine  the  values  of  the  constants  in 

ax^  +  bz^  +  lxz  +  mx+nz  +  d^O,  (46) 

the  origin  is  moved  to  ^=7.65,  2=11.6.     For  this  point  as  origin  an 
equation  of  the  following  form  describes  the  curve ; 

a(x'y  +  b(zy  +  l^x'z'  +  m,x'^0.  (64) 

Taking  a=i,  only  three  constants,  b,  l^,  and  m^,  remain  to  be  evaluated. 
From  the  trace  of  Path  V  on  the  surface  of  the  plow  bottom  the  following 


data  were  obtained : 

x' 

I 

3 
6 

z' 

13-55 
20.05 

27-15 

Substituting  these  values  for  x'  and  z'  in  equation  (64)  gives 
tions  from  which 

three 

equa- 

b- 

=  —0.019 

(x')^- 

0.  oig(z'y  —  ( 

=  -0.453 

=      8.63 

D.453x'z'  +  8.63x'=6>. 

(65J 

Translating  the  axes 

back  to  the 

x'  = 

z'  = 

original  origin, 

X-    7.65 
2—  II.  6 

gives 

rt^- 0.0192^-0.453x2-  I.  4501+3.  912-49.  92  =  0.  (66) 


lyo  Journal  of  Agricultural  Research  voi.  xit.  No.  4 

To  determine  the  values  of  the  constants  in 

ax^  +  by'^  +  lx  +  my+d=0,  (45) 

the  origin  is  moved  to,  x  =  7.  65,  y  =  o.  2.     This  changes  the  form  of  the 
equation  to 

a{x'f  +  h{y'f^l^x'-\-my^O.  (67) 

Taking  a=i,  three  constants  remain  to  be  evaluated.     From  the  trace 
of  Path  V  upon  the  surface  of  the  plow  bottom, 

x'  y' 

I  3.1 

4  5.45 

7  6.68 

Substituting  these  values  of  x'  and  y'  in  equation  (67)  gives 

a=        I 
6=       4.29 
/i=-3o.85 
Wi=-   3.67 

(x')'  +  4.29(/)'-3o.85%'-3.67/  =  0.  (68) 

The  axes  are  translated  back  to  the  original  origin  by  substituting 

x  =  x'-7.6s 
y  =  y'—o.  2 

in  equation  (68),  which  gives 

ac2  +  4.  29^-46. 15a;- 5.  397+ 295. 45  =  0.  (69) 

Numerical  Example 
The  surface  of  a  plow  bottom  is  represented  by  the  equation 

o.54x:2—  1.527^+ 1. 123^—3.6972;—  1.62x2  i- 2. o^xy 
+  53.63X+  ii^-goy-  46.42+  49.4  =  0. 

The  motion  of  a  soil  particle  which  passes  upon  this  bottom  at  the  point 
x=6.g,  y=o.2,  2  =  9.5  is  described  by  the  following  equations: 

2  =  0.00001622(^2  — 45. 55'+ 342)2  +  5'  (70) 

—  0.II922—  1.126^2+ 20.78:^+  10.032;— 201.63  =  0  (71) 

x^  +  1. Sy^ - 42. 4XX- 1. 5y+ 245.25  =  0  (72) 

s  =  vt.  (47) 


Jan.  38, 191S 


Study  of  Plow  Bottoms 


171 


From  equations  (70),  (71),  (72),  and  (47)  the  following  are  obtained; 
Table  VIII. — Values  (in  inches)  for — 


s 

c 

z 

y 

18 
27 

36 

18.4 
27.4 
36.0 

7-55 
II- 5 
19-5 

3-6 

8.25 
II.  0 

dz 

■^=o.oooo2244[{vH^-4S.5vt+s42){2vH-4S-5'v)]+'o 


d^z 

^=o.oooos244[{vH^-4S.5vt+s42){2V')+(2vH-45.svy] 

^^     (.2382+1. i26x-io.o3)^ 
dt  2x  — 1.1262+20.78 

(,.,38,+..,.6,-,o.03)g-.(g)'+o..38(g)-+.../(g)(g) 
2a:— 1.1262+20.78 

dy     (-^^+4^-41)^ 
dt 


(73) 
(74) 

(75) 
(76) 

(77) 

(78) 

(79) 


3-6>'-i.5 

^^     (-2X+42.4x)g-2(g)^-3.6(g)' 
dt^  S.6y-i.s 

The  plow  moved  forward  with  a  velocity  of  36  inches  per  second,  giving 

.^=36^  (80) 

From  equations  (74),  (75),  (76),  {77),  (78),  (79),  and  (80)  the  values 
listed  in  Table  IX  are  computed. 

Table  IX. — Values  for — 


di 

rf»i 

dy 

£^'y 

d2 

d^z 

di 

dt^ 

dt 

dfi' 

dt 

dt^ 

Sec. 

18 

% 

7.09 

53-6 

16.  9 

28.4 

37-7 

-   9-07 

27 

'A 

25-15 

47-75 

17-32 

—  50.  0 

34-44 

—  10.  21 

3& 

I 

38-4 

41.  6 

3-44 

-74.8 

36 

29.52 

By  making  the  proper  substitutions  from  (80),  (74),  (76),  and  (78)  in 
equations  (54),  (57),  (58),  and  (59)  the  values  of  the  direction  cosines 
for  the  normals  A^  the  tangents  to  the  path  T,  and  the  perpendiculars 
to  the  planes  formed  by  the  normals  and  tangents  R  for  three  points 
are  computed  and  listed  in  Table  X. 


172 


Journal  of  Agricultural  Research 


Vol.  xn.  No.  4 


Table  X. — Values  of  the  direction  cosines  for  normals,  tangents  to  the  path,  and  per- 
pendiculars io  the  planes 


«=7-55    5^3-6    2=  18.4 


COS  N^= 

cos  iYy= 

COS  iVj= 

0.549 

.716 

-  .429 

cos  rx=  0. 169 

COS  Ty=   .  4025 
cos  Ta=   .  9 

cos  i?x= 

cos  Ry  = 

cos  R^= 

0.817 

-  .564 
.0977 

1=11.5  y=S.25    2=27-4 

COS  N.^= 

cos  A^y= 

COS  N^= 

0.728 

.  229 

-  .646 

COS  T^=o.  546 
cos  Ty=  .  376 
COS  Tj=  .  749 

cos  R^= 

cos  Ry= 
cos  i?8= 

0. 4145 

-  .897  • 

.149 

X=l9-5  ^=11  2=36 

COS  A^x= 

COS  A^y  = 

COS  N^=^ 

0.698 

-  .215 

-  .683 

cos  Tx=o.  728 
COS  T'y=  .  065 
cos  T,=  .  683 

cos  R^= 

cos  i?y  = 

cos  /?2= 

0.  102 

-  -975 
.  2 

For  the  purpose  of  computing  the  forces  a  block  of  soil  2  inches  wide, 
I  inch  long,  and  )4  inch  thick  is  taken.     The  mass  of  this  soil  is 


M= 


(2. 1. 5)62. 5p    0.0362P 


(81) 


1728.32.2. 12     32.2.12 
p=  density. 


By  the  proper  substitutions  from  Tables  IX  and  X  into  equations  (60), 
(61),  and  (62)  the  forces  necessary  to  produce  the  accelerations  are  com- 
puted and  listed  in  Table  XI. 

Table  XI. — Forces  necessary  io  produce  acceleration  in  soil  particles 


x=  7-55 
Fn=     .00S03P 

x=ii.S 
Fii=     .00281P 

x=ig.s 
Fn=     .00234P 


y=  3-6 
Ft=     .001  i6p 

y=  8.25 

Ft=     .000248P 

y=ii 
Ft=     .00428P 


2=18.4 
Fe=     .002  52P 
2=27.9 

^E=       •OO592P 
2=36 

Fb=     .00778P 


A  soil  particle  in  passing  over  the  surface  of  the  plow  bottom  will  be 
acted  upon  by  the  following : 

(a)  A  force  from  the  surface  of  the  bottom  acting  in  the  direction  of 
the  normal. 

(b)  Gravity. 

(c)  Pressure  from  the  weight  of  the  soil  above  the  particle. 

(d)  Friction  between  the  particle  and  the  surface. 


Jan.  28, 1918  Study  of  Plow  Bottoms  1 73 

ie)  Shearing,  stretching,  or  compression  on  each  of  the  remaining  five 
sides  of  the  particle,  due  to  its  contact  with  other  soil  particles. 

The  force  which  produces  the  movement  of  a  soil  particle  in  any  direc- 
tion will  be  the  resultant  of  the  components  of  the  above-listed  forces 
which  act  in  the  direction  of  the  movement. 

The  preceding  analysis  of  the  motion  which  certain  soil  particles 
have  in  the  operation  of  plowing  has  not  been  developed  from- as  refined 
methods  nor  as  uniform  data  in  all  cases  as  could  be  desired,  but  the  re- 
sults obtained  furnish  abundant  evidence  that  the  problem  here  at- 
tempted is  by  no  means  hopeless.  The  study  should  be  continued  upon 
a  tough  sod,  which  would  stretch  more  uniformly,  and  some  apparatus 
which  would  remove  the  necessity  of  certain  soil  particles  remaining  iu 
line  with  each  other  should  be  substituted  for  the  pins. 

HISTORY  OF  THE  DEVELOPMENT  OF  PLOW  BOTTOMS 

The  Annual  Report  of  the  New  York  State  Agricultural  Society  for 
1867  contains  an  ex- 
cellent treatise  giving 
the  geometrical  con- 
struction of  the  sur- 
faces of  many  histori- 
cal plow  bottoms,  but 
no  attempt  has  been 
made  in  that  report 
to  classify  these  sur- 
faces upon  the  basis 
of  their  mathematical 
forms.  Using  the  above-mentioned  work  as  a  basis,  the  author  has 
attempted  to  work  out  the  mathematical  forms  of  the  most  important 
of  these  historical  surfaces  with  a  view  to  making  fundamental  compari- 
sons with  present-day  plow  bottoms. 

JEFFERSON'S  PLOW  BOTTOM 

In  1788  Thomas  Jefferson,  while  making  a  tour  in  Germany,  devel- 
oped what  appears  to  be  one  of  the  first  methods  recorded  for  making  the 
surface  of  the  moldboard  geometrically  exact  in  form.*  He  argued  that 
the  offices  of  the  moldboard  were  to  receive  the  soil  from  the  share  and 
invert  it  with  the  least  possible  resistance.  In  order  to  do  this,  Jeffer- 
son developed  a  surface  which  he  considered  best  adapted  for  the  work 
of  plowing,  but  attention  should  be  called  to  the  fact  that  no  evidence" 
is  offered  to  prove  the  assertion.  Figure  15  shows  the  framework  for 
generating  the  Jefferson  moldboard,  in  which  lines  em  and  oh  are  the 
directrices.     To  generate  the  surface  a  straightedge  is  laid  upon  eo  and 

1  GouiD,  J.  S.,  et  al.    Op.  cit.,  p.  403. 


from  Report  N Y.  Stole /Igric.Soc.  1867 

Fig.  is. 


I  y4  Journal  of  Agricultural  Research  voi.  xii.  No.  4 

moved   backward,    the   straightedge   remaining    parallel    to    the    plane 
z=0.     By  taking  the  point  o  as  the  origin,  the  equation  of  the  surface  is 

T,hyz—2dxz—2hly  +  2hdz=0'^  (82) 

h  =  breadth  of  furrow 
c^=  depth  of  furrow 
/  =  length  of  moldboard. 

On  rotating  the  XV-axes  through  tan-^=  2^/36,  the  equation  is 


{()h^'  +  J[(P)y'z-^hdlx'-6bHy'  +  2hd^Jghi'+^(Pz=0.  (83) 

On  rotating  the  yZ-axes  through  tan-^>^V2,  the  equation  13(96^ +4^^) 
{iy"f-{z'Y\-?>hdlx' 

^-2{hd^|^WT^-2>hH^fi)\y"  +  z']=-0,  (84) 

Translating  the  axes  to  the  points 

y  =>'    +yo 

z'  =  z"  +Zo 

where  yo  has  such  a  value  that 

2{9f^  +  4(P)yo+  2[bd^iSb^+Sd^-  3m-yf^]=0,  (85) 

and  Zo  has  such  a  value  that 

-2(9b^  +  4d^)Zo+2[bd-yJi8b'  +  8d^-3m^]=0,  (86) 

gives 

(9b^+4d^)[(j"'y-  (z'y]-Sbdlx'+  (>'o'-2o')(9&'  +  4cP) 

+  (yo+Zo)  {2bd^iSb^+Sd'-  2,m4i)  =0.  (87) 

Letting  the  constant  terms  in  (87)  equal  C  gives 

(962  +  4d2)[(y'")2-  {z"y]-2>bdlx'  +  C=0.  (88) 

Translating  the  axes  to  the  point  x'  =  x"-\-Xo  where  Xq  has  such  a  value 

that 

-8bdlXo-VC=0 
gives 

(962+4<P)[C|/'")'-  {z"f]  =  8bdlx".  (89) 

This  is  the  equation  of  a  hyperbolic  paraboloid.^ 

IvAMBRUSCHINl'S  PI.OW   BOTTOM 

Lambruschini,^  an  Italian,  describes  a  method  for  generating  the  sur- 
face of  a  plow  bottom  which  he  considered  to  be  more  efficient  than 
the  surface  developed  by  the  Jefferson  method.     Lambruschini  proposed 

»  The  method  of  developing  the  equation  for  this  surface  is  given  upon  pages  150  to  156. 
*Snyder,  Virgil,  and  Sisam,  C.  H.    Op.  dt.,  p.  73. 

»  lyAMBRUSCHINl,  R.      Op.  cit.,  p.  37-80.      1832. 


Jan.  28,  1918 


Study  of  Plow  Bottoms 


175 


a  helacoid  generated  as  follows :  Lay  out  a  rectangle  opan  (fig.  1 6)  twice 
the  desired  width  of  the  furrow  and  of  an  empirically  determined  length. 
Take  the  point  m  midway  between  points  0  and  p  and  draw  the  line  mm 
parallel  to  pq.  A  straightedge  laid  upon  mo  and  moved  backward  along 
the  line  mm,^  being  kept 
parallel  to  the  plane 
z  =  0,  and  with  an 
angular  rotation  pro- 
portional to  the  move- 
ment toward  m^,  gen- 
erates the  surface  of 
the  Lambruschini  bot- 
tom. The  point  of  the  straightedge  which  was  at  0  will  describe  the 
helix  00i7  (fig.  1 6) .     The  equation  of  this  surface  is 

-=tane, 

X  ' 

where  9  has  uniformly  increasing  values  as  2  increases. 
Then  d=f  (z),  when  0  =  90°=—  radians, 


Fig.  16. 


/  =  length  of  line  mnii 


^  =  2' 


2        2 


n 


Hence, 


f=Kr> 


(90) 


small's  plow  bottom  ^ 


About  1760,  a  Scotchman,  James  Small,  established  a  factory  in 
Scotland  for  the  manufacture  of  plows.  The  surface  of  Small's  mold- 
board  is  obtained  by 
laying  a  straightedge 
upon  op  (fig.  17)  and 
moving  it  backward 
parallel  to  the  plane 
2=0,  with  the  line  pvi 
and  the  curve  oh  as 
directrices.  The  equa- 
tion of  the  curve,  a  half 
catenary,  is  obtained  by 
drawing  a  line  og  (fig.  18)  the  length  of  line  og  (fig.  17).  At  o  erect  a  line 
ooj  perpendicular  to  line  og  and  equal  in  length  to  line  gh  (fig.   17). 

'  Gould,  J.  S.,  et  al.    Op.  at.,  p.  415. 


rrom  Report  of  N.Y.Sfote Agric50C.  I3S7 

Fig.   17. 


176 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  4 


Through  point  Oj  (fig.  18)  draw  a  line  oji  parallel  and  equal  to  line  og. 
With  h  and  0  as  points  of  suspension  describe  a  catenary  with  its  lowest 
point  at  O.  Taking  the  point  O  (fig.  18)  as  origin,  the  equation  of  the 
catenary  is 


y  _  _  /g21z  /3ba  ^  g-21z  y3ba\  ^ 

a  =  Og. 
Transferring  the  origin  to  the  point  0  gives 

yz=  —  (g21z  /3ba  ^  g-21z  /3ba\  _  ^j 


(91) 


(92) 


as  the  equation  of  the  catenary  oh  (fig.  17).     The  equations  of  line  pm 
(fig.  17)  are 


Fig.  18. 

Any  plane  parallel  to  the  plane  2  =  0  is  given  by  2==c,  and  this  plane  cuts 
the  line  pm  at  the  point 

z^  =  c. 
It  also  cuts  the  catenary  oh  at  the  point 

^^~2l^ 

J2  =  /(C) 
Z2  =  C. 

The  equation  of  the  fine  in  the  plane  z  =  c  which  cuts  the  line  pm  and  the 

catenary  oh  (fig.  17)  is 

x—h  _  y—0 

(93) 


or 


3&       ,     }{c)-0 


(94) 


Jan.  28, 1918 


Study  of  Plow  Bottoms 


177 


As  this  line  is  always  parallel  to  the  plane  2;  =  O,  it  follows  that  c  =  2  and 

f(c)=nz). 

From  equations  (92)  and  (94)  then, 

(x  -  &)[  J  (e^iz  /3ba  +  g-2iz  /3ba)  _  ^1  _  y  ?^^  -b\  =  0,  (95) 

which  is  the  equation  of  Small's  moldboard. 

STEPHEN'S  PLOW   BOTTOM  * 

About  the  same  time  that  Small  brought  out  his  moldboard  another 
Scotchman  named  Stephens  developed  a  method  for  forming  the  surface 


f/x>m  Report  ofNYSfoHAgric  SoC- 1867 
Fig.  19. 


From  Report  of  N  Y  Star-^  Aqric  Soe  /36T 
Fig.  21. 


V^A 


Fig.  20. 


of  a  moldboard  the  general  plan  of  which  is  shown  in  figure  19.  The 
generator  for  this  surface  is  a  straightedge  laid  upon  op  (fig.  19)  and  moves 
backward  parallel  to  the  plane  z=0  with  the  line  on  and  the  curve  ph  as 
directrices.  Stephen  designed  his  surface  by  taking  a  quarter  cylinder 
opmnhg  and  laying  out  p^vti  (fig.  20)  equal  in  length  to  pm  (fig.  19). 
Perpendicular  to  line  pj^m^^  draw  mjii  equal  to  the  length  of  arc  mh  (fig. 
19).  Through  points  p,  h^  (fig.  20)  pass  a  circle  of  radius  2nb.  The  plane 
figure  p^m^hyh^  (fig.  20)  is  then  laid  upon  the  quarter  cylinder  (fig.  19) 
so  that  /?!  falls  upon  p,  m^  upon  m,  and  h^  upon  h.  This  will  locate  the 
curve  ph  (fig.  19),  leaving  a  figure  as  shown  in  figure  21.     It  will  be 

'GoutD.  J.  S.,  ct  al.    Op.  cit.,  p.  431. 


178 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  4 


observed  in  figure  21  that -  =  tan  6  where  d  has  gradually  increased  values 

from  O  at  2=0  to  90°  at  z  =  L     Further,  ^  =  ?.  radians  where  7  represents 

the  lengths  of  arcs  11',  22',  etc.;  then  |  =  tan  QY     From  figure  20  the 
equation  of  the  circle  with  its  center  at  O,  taking  p^  as  the  origin  is 


In  figure  20 


F=2nb  cos  <l> 


(y-Fy+(z+Gy  =  4n^b^ 

<t>+   7=   90°; 
B+B'=   90°; 
a  +  B'+   7=180°; 
<l)  =  a-B; 


(96) 
(97) 


G=2nb  sin  ^ 


= zzzzz — +" 


(98) 


I 

2n6* 


(99) 


Substituting  the  values  for  F  from  equation  (98)  and  for  G  from  equation 
(99)  gives 

|  =  tan[/(2)],  .     (100) 

which  is  the  equation  of  the  surface. 

rahm's  plow  bottom  ^ 

In   1846   Rev.  W.  L.  Rham,  an    Englishman,  brought    forward    the 
theory  that   the  lines   of   the   moldboard    running   in   the  longitudinal 

direction      should      be 
T— jL..^^^^  y         straight,  but   that   the 

section  of  the  mold- 
board  formed  by  any 
plane  z  =  c  (fig.  22) 
should  be  a  straight 
line  or  a  curve,  ac- 
cording to  the  phys- 
ical characteristics  of  the  soil  to  be  worked.  Mr.  Rham  agreed  that 
for  medium,   mellow   soils  the  surface   of  the    moldboard    should    be 

'  Gould,  J.  S.,  et  al.    Op.  cit.,  p.  442. 


^<J 


From  Report  of  N.  Y.  State  Agric.  Soc.  1S67. 
Fig.  22. 


Jan.  28, 1918 


Study  of  Plow  Bottoms 


179 


generated  by  laying  a  straightedge  upon  oe  and  moving  it  backward 
parallel  to  the  plane  z=0  with  the  lines  e^h  and  em  as  directrices. 
This  surface  will  be  a  portion  of  a  hyperbolic  paraboloid,  the  same 
general  type  as  the  surface  which  Jefferson  proposed.  The  orthogonal 
projection  of  the  generator  in  various  positions  upon  the  plane  z=0 
will  look  as  shown  in  figure  23.     For  stiff,  clay  soils  the  lines  (fig.  24) 


e  o 

From  Report  of  N.  Y.  State 

Agric.  Soc.  1867 

Fig.  23. 


e   O 

From  Report  of  N.  V.  State 

Agric.  Soc.  1867 

Fig.  24 


e  O 

From  Report  of  N.  Y.  State 

Agric.  Soc.  1867 

Fig.  25. 


are  made  concave  and  for  loose,  sandy  soils  (fig.  25)  they  are  made 
convex.  As  no  exact  description  was  given  regarding  the  shape  of 
the  curves  (fig.  24,  25),  it  has  not  been  possible  to  develop  equations 
for  the  surfaces.  However,  as  it  is  known  that  these  surfaces  have 
straight  lines  in  one  direction  and  can  not  be  described  by  an  equation 
of  the  second  order,  they  are  of  the  fourth  order  or  higher. 

KNOX'S    PLOW    BOTTOM* 

In  1852  Samuel  A.  Knox,  of  Worcester,  Mass.,  applied  for  a  patent  upon 
the  surface  of  a  plow  bottom  which  was  certainly  unique.     The  skeleton 

of  this  surface  is  shown 
in  figure  26.  The  seg- 
ments of  circles  I, 
II,  and  III  are  placed 
in  parallel  planes  12 
inches  apart,  so  that  a 
series  of  straight  lines 
will  cut  the  three  cir- 
cles. Circles  I  and  III 
have  equal  diameters 
and  the  diameter  of 
circle  II  is  one-half 
that  of  circles  I  and 
III.  As  the  equation  of  this  surface  is  of  the  eighth  order,  it  will  not  be 
worked  out  in  detail,  but  a  development  will  be  given  to  show  how  the 
equation  could  be  obtained. 

Let  the  equation  of  the  three  circles  be  ^ 


from  ifepon  of r*Y  stole  Agric  Soc  /e67 


Fig.   26. 


Z=0, 


{x-aY+iy 

z=k 


H'^' 


'  Gouij>,  J.  S.,  et  al.    Op.  cit.,  p.  49-. 

*  This  development  is  the  work  of  Virgil  Snyder,  Professor  of  Mathematics,  Cornell  University. 


i8o  Journal  of  Agricultural  Research  voi  xii.  No.  4 

and  (x-c)-+{y-dy-  =  R' 

Z=2k. 

Draw  the  line  from  a  point  {x^,  y^,  O)  on  the  first  circle  to  a  point  {x^,  y^,  2k) 
on  the  third.     Its  equations  are 


from  which 


x-x^_  y-yi_  z  ^ 

^2~%   y2~yi   2fe 


2k(x—X,)  +  z(x,—c) 

-  =  x^  —  c. 


Z 

2k(y-y^)  +  z(yi-d) 


■■y^-d. 


Since 


{x,-cy+{y,-df  =  R\ 
we  have,  after  simplifying, 

^\{x-x,f+{y-y,y]+Akz[{x-x,){x,-c) 

+  {y-yd{y-d)]+z\{x-cf+{y-dy-R']=o.      (loi) 

This  is  the  equation  of  a  cone  with  vertex  at  {x^,  y^,  O)  and  passing  through 
the  third  circle. 

In  the  same  way,  find  the  equations  of  the  line  from  (x^,  y^  O,)  to 
(«3.  yst  k)  on  the  middle  circle 

x-x^^  y-yi_z ^ 
^3-^1    yz-yi    k' 

k(x—x^)  +  z(x^  —  a) 


z 

Hy-yi)+z(yi-b). 


■  ^'3 — <i> 
-ys-^- 


Since 

(x,-ari-(y,-b)2  =  (^fj' 
we  have,  after  simplifying, 

k\ix-x,y+(y-y,r]+2kz[(x-x,)(x,-a) 

+  {y,-b)(y-y,)]+2'^(x,-ar+(y,-by-~j=0.     (102) 

When  equations  (loi)  and  (102)  are  multiplied  out,  it  will  be  seen  that 
x\,  fi  always  enter  in  the  form  x\  +  fj^  =  R\  By  substituting  R-  for 
0^1+ y^i  in  each,  the  equations  are  of  the  forrr 

Ax,+  By,==C, 
A%+B'y,  =  C'. 


Jan.  28,  i9i8  Study  of  Plow  Bottoms  i8i 


Solve  these  equations  for  x^,  y^  and  put  their  values  in 

A  =  [4^2  (x+  c)  —  2cz^  —  2>xhr], 

B  =  {^kzly  -\-d)-  adz"  -  8yk% 

C=  [4R~k^-4k^(x^+y^)  -  4kxz-4kzy-  4.kRh-hz^(c^  +  d?)] 
A '  =  [2kz(x+d)  —  2xk'^—  2az^], 
B'=[2kz(y+b)-2yk^-2bz^], 

a  =  [R'k^  +  k\x^  +  f)-4kz(ax+by-R^)i-z\a^  +  b^  +  ^R^)]. 

_B'C-BC' 
^'~AB'-A'B' 

_C'A-CA' 
^'     AB'-A'B' 
hence  {B'C-BCy+{C'A-CAy  =  R\AB'-A'B)\  (103) 

CYLINDRICAL   PLOW    BOTTOMS 

In  1854  an  American,  Joshua  Gibbs/  patented  a  plow  bottom  the 
surface  of  which  is  a  portion  of  a  circular  cylinder.  Taking  a  point  upon 
the  axis  of  the  cylinder  as  the  origin,  the  equation  of  this  surface  is 

x^     y"^ 

-2  +  p-I=C>  (104) 

In  some  foreign  countries,  notably  Germany,  the  hyperbolic  cylinder  has 
been  suggested  as  suitable  for  forming  the  surface  of  the  moldboard.  In 
this  cotmection  it  is  interesting  to  note  that  any  cylindrical  surface  can  be 
described  by  an  equation  of  the  general  form. 

^,±^±.=0  (105)= 

mead's  plow  bottom  ^ 

In  1863  a  Mr.  Mead,  of  New  Haven,  Conn.,  patented  a  plow  bottom, 
the  surface  of  which  conformed  exactly  to  a  portion  of  a  frustrum  of  a 
cone.     The  general  equation  of  this  surface  is 

^         ,^2         ^2 

a'  +  p-3-«  COS) 

holbrook's  plow  bottom 

The  Report  of  the  New  York  State  Agricultural  Society  for  1867  con- 
tains a  very  complete  report  of  the  plow  trials  held  at  Utica,  N.  Y., 
in  1867,  at  which  trials  a  line  of  plows  designed  by  F.  F.  Holbrook,  of 
Boston,   Mass.,   showed   general   superiority   to   all   other   makes.     The 

»GouLD.  J.  S.,  etal.    Op.  cit,  p.  502.  »  GotJiD.  J.  S.,  et  al.    Op.  cit..  p.  S05. 

*  Snyder.  Virgil,  and  Sisam,  C.  H.    Op.  dt.,  p.  82. 

27807°— 18 3 


i82  Journal  of  Agricultural  Research  voi.xii,  no.  4 

following  quotation  gives  a  very  good  description  of  the  Holbrook  sur- 
faces : 

We  1  were  interested  in  the  most  minute  details  of  these  plows  by  Gov.  Holbrook 
and  the  trials  at  Utica  and  subsequently  at  Brattleboro,  Vt.,  showed  very  clearly  the 
influence  of  the  warped  surface  which  is  generated  by  his  method  upon  the  texture 
of  the  soil.  Gov.  Holbrook  is  as  yet  unprotected  by  a  patent  on  his  method,  and  we 
are  therefore  most  reluctantly  compelled  to  withhold  a  description  of  it  but  we  have 
no  hesitation  in  saying  that  it  is  the  best  system  for  generating  the  true  cvirve  of  the 
moldboard  which  has  been  brought  to  our  knowledge.  This  method  is  applicable  to 
the  most  diversified  forms  of  plows,  to  long  or  short,  to  broad  or  narrow,  to  high  or 
low,  no  matter  what  the  form  may  be,  this  method  will  impress  a  family  likeness 
upon  them  all.  There  will  be  straight  lines  in  each  running  from  the  front  to  the 
rear  and  from  the  sole  to  the  upper  parts  of  the  share  and  moldboard.  None  of  these 
lines  will  be  parallel  to  each  other,  nor  will  any  of  them  be  radii  from  a  common  cen- 
ter. The  angle  formed  by  any  two  of  them  will  be  tmlike  the  angle  formed  by  any 
other  two;  a  change  in  the  angle  formed  by  any  transverse  lines  will  produce 
a  corresponding  change  in  the  vertical  lines,  and  there  will  always,  in  every 
form  of  this  plow,  be  a  reciprocal  relation  between  the  transverse  and  vertical  ^  lines. 
Plows  made  upon  this  plan  may  appear  to  the  eye  to  be  as  widely  different  as  it  is 
possible  to  make  them,  and  yet,  on  the  application  of  the  straightedge  and  protractor, 
it  will  be  found  that  they  agree  precisely  in  their  fundamental  character.  The 
siuface  of  the  moldboard  is  always  such  that  the  different  parts  of  the  furrow  slice  will 
move  over  it  with  unequal  velocities. 

From  the  above  description  it  is  evident  that  the  surfaces  of  the  Hol- 
brook plows  are  portions  of  a  hyperboloid  of  one  sheet  whose  general 
equation  is 

-2+52  ^-i 

MISCElvIyANEOUS   PLOW   BOTTOMS 

In  addition  to  the  surfaces  already  described  there  remain  at  least 
three  which  show  unique  characteristics,  but  data  were  not  available  for 
developing  the  equations. 

In  1 81 8  Gideon  Davis,^  of  Maryland,  patented  the  surface  of  a  plow 
bottom  which  was  obtained  by  using  the  segment  of  a  circle  as  a  gen- 
erator and  two  segments  of  another  circle  as  directrices.  Somewhat  later, 
1834,  James  Jacobs,*  another  American,  brought  out  a  plow  bottom  the 
surface  of  which  was  a  combination  of  two  mathematical  surfaces,  each 
of  which  had  sets  of  straight  lines  in  two  directions. 

In  1839  Samuel  Witherow,  of  Gettysburg,  Pa.,  and  David  Pierce,  of 
Philadelphia,  Pa.,  brought  out  a  plow  bottom  whose  surface  was  gen- 
erated by  the  most  ingenious  use  of  the  arc  of  a  cycloid.  A  more  detailed 
description  of  this  plow  can  be  found  in  the  Report  of  the  New  York 
State  Agricultural  Society  for  1867,^ 

*  GoTJXD,  J.  S.,  et  al.    Op.  dt.,  p.  586. 

*  It  should  be  noted  that  the  hnes  here  called  transverse  are  designated  as  longitudinal  (PI.  2,  A),  and  the 
]ines  called  vertical  are  designated  as  transverse. 

^GoxJiD,  J.  S.,  etal.    Op.  dt.,  p.  452. 

*  Idem,  p.  486. 
*Idem,  p.  491. 


PLATE  6 

A. — A  plow  bottom  with  two  sets  of  straight  lines. 

B. — A  plow  bottom,  the  surface  of  which  is  composed  of  each  of  two  surfaces. 

C— A  plow  bottom  similar  to  B,  but  with  the  smfaces  merging  into  each  other 
farther  back  on  the  moldboard. 

D. — A  plow  bottom,  the  surface  of  which  does  not  contain  an  infinite  set  of  straight 
lines.    ' 


study  of  Plow  Bottoms 


Plate  6 


Journal  of  Agricultural  Research 


Vol. XII,  No.  4 


study  of  Plow  Bottoms 


Plate  7 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  4 


PLATE  7 

A. — A  plow  bottom  with  a  convex  surface  whicli  has  two  sets  of  straight  lines. 
B. — Instrument  for  measuring  the  space  coordinates  of  any  point  of  the  plow  bottom. 
C. — A  sod  plow  showing  the  furrow  slice  tiuned  by  it. 


PLATE  8 

A. — Rows  of  wooden  pins  driven  into  the  sod  for  estimating  the  stretch  of  the  fur- 
tow  slice. 

B. — Furrow  slice  showing  the  position  of  the  pins  when  on  the  moldboard. 


study  of  Plow  Bottoms 


Plate  8 


^Mj^mfs^^ 


5s.^   "•.  't'jfr  yr  x 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  4 


study  of  Plow  Bottoms 


Plate  9 


Journal  of  Agricultural  Research 


Vol. XII,  No.4 


PLATE  9 

A. — Plow  showing  attachment  used  to  obtain  the  x,  y,  and  2  coordinates  of  points  in 
the  furrow  slice. 

B. — Moldboard  showing  the  paths  of  five  soil  particles. 

C. — Meastirement  of  the  angle  Ny  by  use  of  a  protractor  and  a  plumb  bob. 


INFLUENCE  OF  NITRATES   ON   NITROGEN-ASSIMILAT- 
ING BACTERIA^ 

By  T.  L.  Hn^ivS,- 
Research  Bacteriologist,  Idaho  Agricultural  Experiment  Station 

INTRODUCTION 

REIvATlON    Olf    NITRATES   TO    VARIOUS   FORMS    OF    PLANT    LIFE 

The  importance  of  nitrogen  to  plant  life  can  not  be  overestimated. 
It  is  one  of  several  elements  essential  to  plant  growth,  one,  moreover, 
which  is  apt  to  be  deficient  in  arable  soils.  These  facts  are  well  brought 
out  by  the  almost  innumerable  investigations  which  have  been  made 
concerning  the  source  of  nitrogen  for  plants. 

The  influence  of  nitrate  nitrogen  on  various  plants  has  been  the  con- 
trolling idea  in  many  of  these  experiments.  Very  little  attention  has 
been  placed  on  the  effect  of  nitrates  on  the  lower  plants,  especially  the 
bacteria.  Because  of  the  relation  that  exists  between  higher  plants  and 
bacteria  it  seems  advisable  to  consider  the  effect  of  nitrates  on  the  soil 
bacteria.  Indeed,  progress  in  the  knowledge  of  nitrogenous  fertilizers 
depends  on  a  study  of  the  effect  of  the  fertilizer  on  the  soil  organisms  as 
well  as  on  the  higher  plants.  The  action  of  fertilizers  on  the  different 
groups  of  soil  organisms,  the  relation  of  these  organisms  to  higher  plants, 
and  the  separation  of  the  important  from  the  unimportant  groups  are 
some  of  the  factors  involved  in  the  problem  of  soil  fertility. 

REVIEW  OF  LITERATURE 

The  relation  of  nitrates  to  the  germination  of  seeds  has  been  studied 
by  De  Chalmot  {iiy,  who  found  that  corn  germinated  in  solutions  con- 
taining nitrate  was  more  robust  than  com  germinated  under  similiar 
conditions  without  nitrate.  He  also  noted  that  if  too  concentrated 
solutions  of  nitrate  were  used  germination  was  retarded  rather  than 
hastened.  The  presence  of  nitrate  also  increased  the  amount  of  al- 
buminous material  in  the  seed. 

The  direct  influence  of  nitrate  nitrogen  on  the  growing  plant  is  too 
well  known  to  justify  any  lengthy  discussion  here.  Jost  {26,  p.  134) 
gives  the  results  of  experiments  made  by  Boussingault,  who  grew  the 
sunflower   {Helianthus  argophyllus')  in  sand  with  and  without  nitrate. 

'Major  portion  ot  a  paper  submitted  in  partial  fulfillment  of  the  requirements  for  the  degree  of  doctor  of 
philosophy  in  bacteriology  in  the  Graduate  School  of  the  University  of  Wisconsin,  December,  J916. 

2  The  \vriter  wishes  to  acknowledge  his  appreciation  oi  the  sugsiestions  and  criticisms  obtained  through- 
out the  progress  ot  this  work  from  Prof.  E.  B.  Fred  and  E.  G.  Hastings,  of  the  University  of  Wisconsin. 

'  Reference  is  made  by  number  (italic)  to  "Literature  cited."  pp.  227-230. 


Journal  of  Agricultural  Research,  .  Vol,  XH,  No.  4 

Washington,  D.  C.  J^"   '^-  ^^iS 

1-  Key  No.  Wis. — 10 

(X83) 


i84 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  4 


During  the  three  months'  growth  of  the  plants  1.40  gm.  of  potassium 
nitrate  were  added.  At  the  end  of  the  period  the  dry  weight  of  the 
plant  supplied  with  nitrate  was  nearly  60  times  greater  than  that  of  the 
plant  where  no  nitrate  was  added.  The  relation  between  the  growth  of 
nonleguminous  plants  and  the  amount  of  nitrate  nitrogen  supplied  is 
shown  in  a  very  striking  manner  in  the  following  table  taken  from  Hell- 
riegel  and  Wilfarth  {21,  p.  53-54)'. 


Nitrogen  as  Ca  (N03)2  added  to  pots, 
gm 


Dry  weight  of  oats  (grain  and  straw) . 
gin. . 


None 


0.3605 
.4191 


0.056 

5.9024 

5-8510 
5-  2867 


10.  9814 
10.  9413 


21.  273a 
21.4409 


o-33<5 


But  little  work  has  been  done  on  the  direct  influence  of  nitrates  on  the 
development  of  the  Eumycetes.  Some  investigations  have  been  made  as 
to  the  ability  of  certain  fungi  to  assimilate  nitrate  nitrogen  directly. 
Ritter  {42)  studied  many  species  and  found  that  some  forms  would 
assimilate  nitrate  directly,  while  others  reduced  it  to  nitrite  and  am- 
monia. He  found  some  forms  which  failed  to  grow  on  media  containing 
nitrate.  Kossowicz  {28)  found  that  various  fungi  utilized  nitrates  and 
that  nitrite  and  ammonia  were  produced. 

Miinter  {36)  studied  the  influence  of  inorganic  salts  on  the  growth  of 
various  Actinomycetes.  He  found  that  potassium  and  sodium  nitrates 
in  quantities  equivalent  to  5  per  cent  permitted  good  growth  of  the 
organisms  but  retarded  spore  formation.  Calcium,  barium,  and  stron- 
tium nitrates  in  small  quantities  affected  some  species  but  not  others. 
Small  quantities  of  these  nitrates  did  not  affect  growth  to  any  extent, 
but  larger  quantities  were  detrimental  to  growth  and  spore  formation. 
Silver  nitrate  in  all  amounts  studied  almost  entirely  prohibited  growth. 

Nitrates  appear  to  exert  some  influence  on  the  yeasts.  Drabble  and 
Scott  {13)  studied  the  effect  of  sodium  nitrate  on  these  organisms.  They 
found  that  the  greatest  reproduction  took  place  in  solutions  containing 
0.2  gram-molecule  of  the  nitrate.  Increasing  amounts  of  the  salt  led 
to  a  decrease  in  reproductive  activity  until  with  0.7  gram-molecule 
present  no  reproduction  took  place.  From  their  results  it  is  evident 
that  small  quantities  of  nitrate  stimulated  reproduction,  whereas  larger 
amounts  proved  detrimental.  Kayser  (27)  studied  the  effect  of  man- 
ganese nitrate  on  yeasts.  He  found  that  the  amount  which  produced 
the  maximum  increase  in  the  alcoholic  fermentation  of  sugar  varied  with 
the  strain  of  yeast  employed.  He  likewise  found  that  manganese  nitrate 
produced  greater  increase  than  did  the  same  quantity  of  potassium  ni- 
trate. Fembach  and  Lanzenberg  {14)  concluded  that  nitrates  hindered 
the  rapidity  of  cell  multiplication  of  yeasts  but  greatly  accelerated  the 
action  of  the  zymase.  More  alcohol  was  formed  in  the  presence  than  in 
the  absence  of  nitrate.  According  to  Kossowicz  {28),  nitrates  are  not  a 
suitable  source  of  nitrogen  for  yeasts. 


Jan.  28,  i9i8  Nitrogen- Assimilating  Bacteria  185 

The  direct  influence  of  nitrates  on  bacteria  has  been  studied  to  a  limited 
extent.  The  influence  of  various  nitrates  on  soil  bacteria  has  been 
studied  by  Greaves  (19).  He  added  sodium,  potassium,  calcium,  mag- 
nesium, manganous  and  ferric  nitrates  to  soil  in  varying  quantities. 
The  amount  added  to  the  soil  was  such  that  in  each  case  equivalent 
quantities  of  the  anion  (NO3)  in  the  various  forms  were  added.  The 
effect  of  these  salts  on  the  bacteria  was  determined  by  using  ammonifi- 
cation  as  an  index  of  the  bacterial  activity.  He  found  that  sodium- 
potassium,  manganous  and  ferric  nitrates  in  small  amounts,  approxi- 
mately 0.97  to  5.5  mgm.  of  nitrate  in  100  gm.  of  soil,  slightly  stimulated 
ammonification.  Greater  concentrations  of  these  salts  proved  toxic  as 
evidenced  by  a  decrease  in  the  amount  of  ammonia  formed.  Sodium 
nitrate  was  much  more  beneficial  to  ammonification  than  potassium^ 
nitrate.  From  his  results  as  a  whole  Greaves  concludes  that  it  is  the 
electronegative  ion  which  stimulates  bacterial  activity.  Calcium  and 
magnesium  nitrates  proved  toxic  in  all  concentrations  studied. 

However,  a  majority  of  the  investigations  have  been  directed  toward 
a  determination  of  the  effect  of  the  bacteria  on  the  nitrates.  But  little 
work  appears  to  have  been  done  on  the  direct  action  of  nitrates  on 
bacteria.  Pfeffer  (j8,  p.  351)  cites  some  experiments  showing  the 
repellant  action  of  potassium  nitrate  toward  certain  bacteria.  Spir- 
illum undula  was  repelled  by  a  solution  of  potassium  nitrate  having  an 
osmotic  concentration  equivalent  to  0.5  to  i.o  per  cent.  With  Spirillum 
voluians  a  much  higher  concentration  was  necessary  to  bring  about  the 
same  reaction.  It  was  found  that  different  organisms  required  different 
quantities  of  the  same  nitrate  to  repel  them. 

It  can  be  readily  seen  that  by  far  the  greatest  amount  of  work  on  the 
relation  of  nitrates  to  plant  growth  has  been  done  in  the  realm  of  the 
higher  plants.  Obviously  further  investigations  should  be  made  in 
respect  to  the  effect  of  nitrates  on  the  lower  forms  of  plant  life,  especially 
the  bacteria.  In  this  paper  an  attempt  is  made  to  set  forth  the  results 
secured  in  a  study  of  the  influence  which  nitrates  exert  on  certain  groups 
of  soil  bacteria,  including  not  only  their  reproduction  but  also  some  of 
their  physiological  properties. 

EXPERIMENTAL  WORK 

OUTLINE  OF  PROBLEM 

The  results  of  much  careful  experimentation  show  that  nitrate  nitro- 
gen is  most  readily  assimilated  by  higher  plants.  As  a  rule  it  seems  to 
stimulate  the  plant  to  increased  activity.  In  some  cases  this  is  un- 
doubtedly due  to  increased  nutrition,  while  in  others  it  is  a  result  of 
nuclear  stimulation  with  a  consequent  cell  multiplication.  No  sharp 
line  can  be  drawn  between  these  two  effects.  Probably  one  overlaps 
the  other,  and  the  increased  growth  of  the  organism  can  be  attributed 
to  a  combination  of  the  two  actions. 


1 86  Journal  of  Agricultural  Research  voi.  xii,  no.  4 

From  a  practical  standpoint  the  relation  of  nitrates  to  the  nitrogen- 
assimilating  organisms  of  the  soil  is  of  importance.  Hence,  it  was 
arranged  to  study  the  effect  of  nitrates  on  soil  bacteria,  especially  those 
forms  concerned  with  the  fixation  of  atmospheric  nitrogen.  The  work 
naturally  falls  into  two  rather  distinct  lines  of  investigation.  First,  the 
influence  of  nitrates  on  Azotobacter  was  determined.  Here  studies  were 
made  of  the  effect  of  nitrates  on  the  growth  of  the  organism  in  soil  and 
also  the  effect  of  these  salts  on  the  nitrogen-fixing  property  of  these 
bacteria.  The  action  of  Azotobacter  on  nitrates  in  solution,  the  relation 
of  nitrates  to  pigment  production  and  to  the  formation  of  volutin  bodies 
were  studied.  Second,  the  influence  of  nitrates  on  the  growth  of  Bacillus 
radicicola  in  soil  was  studied.  The  action  of  B.  radicicola  on  nitrates  in 
solution  and  the  possible  nitrogen-assimilating  properties  of  the  legume 
in  the  presence  of  nitrates  were  investigated.  Also  the  influence  of 
nitrates  on  gum  production  was  determined.  The  latter  part  of  the 
investigations  included  a  study  of  the  relation  of  nitrates  to  nodule 
formation  on  alfalfa. 

METHODS   USED  IN   EXPERIMENTS 

Nitrates  were  determined  by  the  reduction  method  with  Devarda's 
alloy  and  also  by  the  phenolsulphonic  acid  (colorimetric)  method. 

The  total  nitrogen  content  of  all  samples  was  determined  by  the 
modified  Kjeldahl  method  with  sulphuric  acid,  salycilic  acid,  sodium 
thiosulphate,  and  copper  sulphate.  Where  nitrate  nitrogen  was  present, 
50  c.  c.  of  concentrated  sulphuric-salycilic  acid  (25  c.  c.  of  concentrated 
acid  plus  25  c.  c.  of  distilled  water)  were  added  to  the  cultures  slowly 
and  with  constant  stirring.  This  acid  was  allowed  to  react  for  a  few 
days,  after  which  the  usual  procedure  was  carried  out.  Digestion  was 
continued  for  five  to  six  hours  subsequent  to  the  clarification  of  the 
liquid. 

The  amount  of  ammonia  was  determined  by  dist'ilation  with  steam 
in  the  presence  of  magnesium  oxid. 

Nitrites  (qualitative  test)  were  tested  for  with  Trommsdorf 's  reagent. 

In  all  distillations  NI14  acid  and  alkali  were  used. 

In  the  preparation  of  agar  cultures  of  alfalfa  seedlings  the  seeds  were 
treated  with  a  0.25  per  cent  solution  of  mercuric  chlorid  and  rinsed  in 
sterile  distilled  water.  Three  bacteria-free  seeds  were  transferred  to  the 
surface  of  soft  mannit  agar  (0.7  per  cent  agar)  in  each  tube. 

The  nitrates  were  added  in  solution  to  all  cultures.  Gram-molecular 
quantities  of  potassium,  sodium,  calcium,  and  ammoniun  nitrates 
(Merck's)  were  weighed  into  sterile  distilled  water.  These  solutions 
were  prepared  in  such  a  manner  that  5  c.  c.  contained  450  mgm.  of 
nitrate.  In  all  nitrate  solutions  the  nitrate  radical,  or  anion,  was 
present  in  the  same  quantities,  while  the  cation,  or  metal,  was  present 
in  varying  quantities,  depending  upon  the  particular  salt. 


Jan.  28,  i9i8  Nitrogen-Assimiloting  BacteHa  187 


Plate  counts  of  all  soil  cultures  were  made  by  weighing  20  gm.  (dry 
weight)  of  the  soil  into  a  200-c.  c.  water  blank.  From  this  suspension 
all  subsequent  dilutions  were  made.  Mannit  agar  ^  was  used  for  the 
plate  counts  in  the  cultures  of  Azotobacter  and  B.  radicicola.  Duplicate 
plates  were  made  for  each  dilution  poured. 

SOIL   USED 

Only  one  type  of  soil  was  employed,  Miami  silt  loam  obtained  from 
the  Experiment  Station  farm.  No  chemical  analyses  of  the  soil  were 
made  other  than  an  estimation  of  its  organic  matter  content,  which  was 
approximately  2.75  per  cent.  The  soil  was  neutral  in  reaction  and  its 
nitrate  content  was  approximately  1.5  mgm.  of  nitrogen  as  nitrate  in 
100  gm.  of  the  dry  soil. 

ISOLATION   OF   AZOTOBACTER   AND   BACILLUS   RADICICOLA 

Azotobacter. — (i)  Strain  A  was  isolated  from  a  silt  loam  soil.  This 
strain  grew  well  on  mannit  agar,  but  produced  no  pigment  after  three 
weeks'  growth.  (2)  Strain  B  was  isolated  from  a  sandy  loam  soil. 
This  strain  grew  equally  well  on  mannit  agar  and  produced  a  brownish 
black  pigment  within  one  week's  growth.  Both  strains  assimilated 
practically  the  same  amount  of  atmospheric  nitrogen  under  laboratory 
conditions. 

BACiiyLUS  RADICICOLA. — A  stock  laboratory  culture  of  B.  radicicola 
was  replated  twice  before  taking  the  final  culture.  The  nodule  produc- 
ing power  of  the  organism  was  determined  by  inoculating  bacteria-free 
alfalfa  seedlings  (in  soft  agar).  After  sufficient  incubation  nodules  were 
produced  in  abundance. 

influence;  of  nitrates  on  azotobacter 

INFLUENCE   OF   NITRATES   ON  THE  GROWTH   AND  REPRODUCTION   OF   AZOTOBACTER   IN 

STERILIZED  SOIL 

What  effect  do  nitrates  have  on  pure  cultures  of  Azotobacter  in  ster- 
ilized soil  ?  Do  these  salts  cause  a  decrease  in  the  numbers  of  the  organ- 
isms? Do  they  cause  an  increase  in  numbers?  Or  do  they  exert  no 
particular  influence  one  way  or  the  other  ?  It  is  difficult  to  believe  that 
the  latter  could  be  true,  inasmuch  as  nitrates  have  such  a  profound 
effect  on  higher  forms  of  plant  life.  Such  readily  soluble  and  assimilable 
substances  as  nitrates  could  hardly  remain  without  affecting  either  an 
increase  or  a  decrease  in  the  number  of  organisms  existing  in  their 
presence. 

With  the  idea  of  determining  what  effect  nitrates  might  have  on 
Azotobacter  when  grown  in  sterilized  soil,  the  following  experiments  were 
planned.     In  this  work  both  strains  of  the  Azotobacter  (described  on 

*  Fred,  E.  B.   a  i^aboratory  manual  of  son,  bacteriology,   p.  io8.   Philadelphia  and  London,  1916. 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  4 


p.  187)  were  employed  and  conditions  governing  the  preparation  and 
incubation  of  the  cultures  were  similar  in  the  case  of  each  strain.  The 
only  variation  was  the  periods  used  in  incubating  the  cultures.  Counts 
were  made  after  one  and  two  weeks'  incubation  with  strain  A  and  after 
one,  two,  and  three  weeks  with  strain  B. 


Table  I. 


-Influence  of  potassuim  nitrate  on  the  growth  of  Azotohacter  {strain  A)  in 
sterilized  soil 


Treatment 
(nitrate 
in  100 
gm.  of 

dry  soil). 

Number  of  organisms  in  i  gm.  of  dry  soil. 

Culture  No. 

At  begin- 
ning. 

After  I  week. 

Relative. 

After  2  weeks. 

Relative. 

^ 

Mgm. 

0 

0 

10 

10 

25 

25 

50 

50 

100 

100 

150 

150 

200 

200 

300 

300 

15,600 
15, 600 
15,600 
15,  600 
15, 600 
15,  600 
15,  600 
15,  600 
15,  600 
15,  600 
15,  000 
15,  600 
15,  600 
15,  600 
15,  600 

IS,  600 

825,  000 

935)000 

I,  500,  000 

Per  cent. 
\             100 

}             170 

I             523 

}        2, 233 

}        I)  295 

I             179 

}               27 

r            ° 

r           315)000 
L           360, 000 
f        I,  175,  000 

Per  cent. 

>             100 

}             348 

4 

c         

4,  200,  000 

5,  000,  000 
20,  400,  000 
18,  900,  000 
II,  000,  000 
II,  820,000 

r  12, 350, 000 

I    10, 750, 000 
1"   27,750,000 

}       3, 418 

6 

1        8,210 

3 

r    9, 000, 000 

I       9)  150,  000 

r       25, 000 
\       55, 000 
f             0 
I             0 
f             0 
I             0 

}        2,  685 

12   

I)  575)  000 

225,  000 

250,  000 

0 

0 

1                 ^^ 

1                  ° 

le 

16   

1                  ° 

Table  II. — Influence  of  sodium  nitrate  on  the  growth  of  Azotobacter  {strain  A)  in 

sterilized  soil 


Treatment 
(nitrate 
in  100 
gm.  of 

dry  soil). 

Number  of  organisms  in  i  gm.  of  dry  soil. 

Culture  No. 

At  begin- 
ning. 

After  I  week. 

Relative. 

After  2  weeks. 

Relative. 

Mgm,. 

0 

0 

10 

10 

25 

25 

50 

50 

100 

100 

150 

150 

200 

200 

300 

300 

13, 800 
13, 800 
13, 800 
13, 800 
13,  800 
13, 800 
13,  800 
13, 800 
13,  800 
13, 800 
13, 800 
13, 800 
13,  800 
13,  800 
13, 800 
13, 800 

310,  000 
225,  000 

575,  000 

430,  000 

2,  850, 000 

5,  800,  000 

15,  200,  000 
12,  750,  000 
17,750,000 

16,  200,  000 
550,  000 
400,  000 

0 
0 
0 
0 

Per  cent. 
>             100 

1             188 
}         1,615 

}     s.. 

}       6, 335 

}      ■" 
}        » 
}        " 

r             425,  000 
\             490,  000 
/             875,  000 

Per  cent. 

2 

(■             100 

\ 

1             191 

c 

/          2,  250,  000 

1 

6 

1             492 

7 

f       15,  500,  000 

I     13)  300,  000 
/      9)  850,  000 
I     15)  750,  000 
f          690, 000 
I          375)  000 
/                      0 
I                     0 

/                      ° 
I                     0 

\ 

8  

1      3, 150 

0 

1       2,800 

10 

I 

12 

1           117 

1-2 

\ 

14 

1 

IC 

\ 

16 

1 

Jan.  28,  1918 


Nitrogen- A  ssimilaiing  Bacteria 


189 


Table  III. — Influence  of  calcium,  nitrate  on  the  growth  of  Azotobacter  {strain  A)  in 

sterilized  soil 


Treatment 

(nitrate 

in  100 

gtn.  of 

dry  soil). 

Number  of  organisms  in  i  gm.  of  dry  soil. 

Culture  No. 

At  begin- 
ning. 

After  I  week. 

Relative. 

After  2  weeks. 

Relative. 

1 

Mgm. 
0 
0 
10 
10 
25 
25 
50 

50 
100 

100 
ISO 
150 
200 
200 
300 
300 

p  p  p  p  p  p  p  p  p  p  p  p  p  p  p  p 
OOOOOOOOOOOOOOOO 

260,  000 

330,  000 

5,  800,  000 

Per  cent. 
>             100 

1        I,  966 

}      3, 440 

1             310,000 
\             260,  000 

/          975, 000 

\        I,  090,  000 
f        9,  200,  000 
\      8,  600,  000 

r   13, 200, 000 
\   12, 600, 000 

f      8,  750,  000 
\      8,  000,  000 
f       2,  000,  000 
I       2,  350,  000 
/                      0 

1                      0 
f                      0 
I                      0 

Per  cent. 
>            100 

2 

•2 

I             362 

5 

6 

10,  700,  000 
0, 600,  000 

1        3> 122 

7 

13,250,000    \ 
II,  600,  000  iJ        ^'      ^ 
6,  600,  000   1 

6;  050;  000 1  ^'^44 

^,  qoo-  coo  |1 

8 

1       4,  526 

0 

10 

}          2,  938 

II 

12 

13 

14 

IS 

16 

3, 900, 000 
0 
0 
0 
0 

1       1,254 

}       ° 
}        = 

}                763 
1                      ° 

One  hundred  and  fifty  gm.  of  soil  (dry  weight)  were  weighed  into  500- 
c.  c.  Erlenmeyer  flasks  and  the  nitrates  added  in  solution,  as  indicated  in 
the  following  tables.  At  the  same  time  i  per  cent  of  mannit  was  added 
in  solution  and  the  moisture  content  was  raised  to  approximately  18  per 
cent.  The  flasks  were  allowed  to  remain  at  room  temperature  for  one 
day,  when  the  contents  were  thoroughly  mixed.  The  flasks  and  contents 
v/ere  then  sterilized  at  15  pounds'  pressure  for  three  hours.  Upon  cooling 
they  were  inoculated  with  5  c.  c.  of  a  suspension  of  the  organisms  in  sterile 
distilled  water.  The  cultures  were  incubated  at  28°  C.  and  counts  made 
at  the  intervals  already  indicated.  Mannit  agar  was  used  in  pouring  the 
plates.  Each  number  in  the  following  tables  represents  an  average  of 
duplicate  plates.  Tables  I,  II,  and  III  show  the  results  of  the  work  vdth 
strain  A  and  Tables  V,  VI,  and  VII  the  results  with  strain  B. 

It  will  be  seen  at  a  glance  that  all  three  nitrates  exerted  an  enormous 
influence  on  the  growth  of  the  Azotobacter.  The  smallest  concentration 
did  not  appear  to  exert  much  influence  either  in  increasing  or  decreasing 
the  number  of  Azotobacter.  There  was  a  slight  gain,  but  it  was  not  so 
marked  as  that  brought  about  by  higher  concentrations  of  nitrates. 
When  25,  50,  and  100  mgm.  of  nitrate  were  present  in  100  gm.  of  soil, 
very  large  increases  were  obtained  in  practically  all  instances.  In  one 
instance  sodium  nitrate  caused  the  greatest  relative  gain,  but  the  most 
consistent  increase  was  produced  by  calcium  nitrate.  Beginning  with 
150  mgm.  the  number  of  Azotobacter  began  to  decrease.  This  decrease 
was  especially  noticeable  in  the  cultures  containing  potassium  and 
sodium  nitrates.     At  the  end  of  the  first  week,  Azotobacter  organisms 


190 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  4 


were  still  found  in  the  potassium-nitrate  cultures  where  200  mgm.  were 
present.  However,  at  the  end  of  the  second  week  the  organisms  were 
dead.  The  same  concentration  of  sodium  and  calcium  nitrates  proved 
even  more  toxic.  No  evidences  were  secured,  indicating  that  these 
organisms  can  resist  concentrations  in  excess  of  300  mgm.  of  nitrate  per 
100  gm.  of  soil. 

The  question  may  be  raised  in  regard  to  the  influence  of  sterilization 
on  the  nitrate  present  in  the  soil.  Does  the  prolonged  heating  in  the 
presence  of  soil  organic  matter  reduce  the  nitrate?  In  order  to  study 
this  point,  a  few  cultures  were  prepared  similar  to  those  already  described. 
They  were  subjected  to  sterilization  under  pressure  of  15  pounds  for  two, 
three,  and  five  hours.  Nitrate  determinations  at  the  end  of  these  periods 
failed  to  show  any  reduction.  In  the  presence  of  i  per  cent  of  mannit 
the  nitrate  content  remained  unchanged  during  sterilization. 

From  these  results  it  is  evident  that  small  amounts  of  nitrate  up  to 
150  mgm.  of  nitrate  in  100  gm.  of  soil  greatly  increased  the  reproduction 
of  Azotobacter.  In  regard  to  the  toxicity  of  higher  concentrations, 
sodium  nitrate  appeared  to  exert  the  greatest  influence  in  this  direction, 
followed  by  calcium  and  potassium  nitrates  in  the  order  named.  The 
results  of  the  experiment  are  recorded  in  Table  IV. 

Table  IV. — Inflttence  of  ammonium  nitrate  on  the  growth  of  Azotobacter  {strain  A)  in 

sterilized  soil 


Treatment 

(nitrate 

in  100 

gm.  OL 

dry  soil). 

Nirmber  of  organisms  in  i  gm.  of  dry  soil. 

Culture  No. 

At  begin- 
ning. 

After  I  week. 

Relative. 

After  2  weeks. 

Relative. 

I 

Mgm, 
0 
0 
25 

25 
100 
100 
200 
200 

18,  500 
18,  500 
18,  500 
18,  500 
18,  500 
18,  500 
18,  500 
18, 500 

I,  400,  000 

1,  050,  000 
5,  600,  000 
4,  goo,  000 

2,  900,  000 
2,  600,  000 
I,  100,  000 

950,  000 

Per  cent. 

'•             100 

}             427 
}             223 
}                84 

/            975. 000 
\        I,  100,  000 
f        5,  000,  000 
\       3,  900,  000 

/      3. 95°.  000 
\      4, 100, 000 
/          875, 000 
I          915)  000 

Per  cent. 
\ 

2.  . . 

>             100 

■2 

"I 

A 

I             430 

e. .. 

}             388 

6 

7 

I                      Q^ 

8 

>           86 

That  the  nitrate  radical  and  not  the  combined  metal  was  the  causal 
agent  in  the  increase  in  the  number  of  Azotobacter  was  indicated  from 
the  results  of  the  next  test.     Here  ammonium  nitrate  was  used. 

It  will  be  seen  from  the  data  of  this  experiment  that  ammonium  nitrate 
caused  an  increase  in  the  number  of  Azotobacter  when  present  in  small 
amounts.  However,  the  increase  in  the  presence  of  ammonium  nitrate 
was  less  marked  than  when  equal  quantities  of  the  other  nitrates  were 
used.  Since  the  experiments  with  ammonium  nitrate  were  not  made  at 
the  same  time  as  the  preceding  experiments  (discussed  on  pp.  189-190),  it 
is  possible  that  conditions  varied  sufficiently  to  account  for  the  less  pro- 
nounced results.     When  200  mgm.  of  nitrate  were  present  in  100  gm.  of 


Jan.  28,  1918 


Nitrogen- A ssiviilating  Bacteria 


191 


soil  the  number  of  Azotobacter  showed  a  decrease.  Apparently  ammo- 
nium nitrate  is  more  toxic  than  potassium,  sodium,  and  calcium  nitrate. 
However,  the  main  point  at  issue  seems  fairly  well  established — namely, 
that  the  increase  in  the  number  of  Azotobacter  is  caused  by  the  nitrate 
radical  and  not  by  the  combined  metal. 


Table  V. 


-Influence  of  potassiuin  tiitrate  on  the  growth  of  Azotobacter  [strain  B)  in 
sterilized  soil 


Treat- 
ment 

(nitrate 
in 

100  gm. 
of  dry 
soil). 

Number  of  organisms  in  i  gm.  of  dry  soil. 

Culture 
No. 

At  begin- 
ning. 

After  I  week. 

Relative. 

After  2  weeks. 

Relative. 

After  3  weeks. 

Rela- 
tive. 

I.  .  . 

2 

Mgm. 
0 
0 
10 
10 
25 
25 

SO 
50 
100 
100 
150 
150 
200 
200 
300 
300 

12,  600 
12,  600 
12,  600 
12,  600 
12,  600 
12,  600 
12,  600 
12,  600 
12,  600 
12,  600 
12,  600 
12,  600 
12,  600 
12,  600 
12,  600 
12,  600 

235.  000 

Per  cent. 
\      100 

|l>  510 
|2,  436 

}i,34o 

|l,320 

}     851 
}     373 
1         ° 

f         112,  500 
\       no,  500 
f  2,  100,  000 
\  2,  250,  000 
1  I,  575-000 
\  1,950,000 
1  3,  250,  000 
\  4,  900,  000 

r  4, 000, 000 
\  3, 500, 000 
r  2, 000, 000 
\  2, 100, 000 
r    800, 000 
I    750, 000 
f            0 
\ 

Per  cent. 

\      100 
J 

}i,  9SO 
}i>58i 
I3.65S 

}i,838 
}     695 

\        0 

J 

/        116,000 
l^        117,000 

r    875,  oco 
\  1, 260, 000 
r  1, 700, 000 
1 1,325,000 
/  3, 525. 000 
\  2, 960, 000 
r  2, 500, 000 
\  2, 900, 000 
r  1, 500, 000 
\  2, 000, 000 
r    650, 000 
\    700, 000 

(      I 

Per  cl. 
\      100 

f     9^^ 
!-l,300 

y,  783 

12,317 
|i,  502 
}      580 
1           ° 

3-  •  • 
4.  .  . 

5--  • 
6.  .  . 
7..  . 
8..  . 
9.  .  . 

10.  .  . 

11 .  .  . 

12.  .  . 

13.  .  . 

14.  .  . 
15..  . 
16... 

3, 750,  000 

3,  300,  000 

5,  750,  000 

5,  700,  000 

3,  100,  000 

3,  200,  000 

3,  200,  000 

3, 000,  000 

2,  ZOO,  000 

I,  900,  000 

875,  000 

880,  000 

0 

0 

Table  VI. — Influence  of  sodium  nitrate  on  the  growth  of  Azotobacter  {strain  B)  in 

sterilized  soil 


Treat- 
ment 
(nitrate 

in 
100  gm. 
of  dry 
soil). 

Number  of  organisms  in  i  gm.  of  dry  soil. 

Cultu 
No. 

At  begin- 
ning. 

After  I  week. 

Relative. 

After  2  weeks. 

Relative. 

After  3  weeks. 

Rela- 
tive. 

1.  .  . 

2.  .  . 

3--  • 
A- 
5--  • 
6.. 

7-- 
8.. 
9.  . 

10.  . 

11.  . 

12.  . 

13- • 
14.  . 

15- • 
16.. 

Mgm. 

0 

0 

10 

10 

25 

25 

50 

50 

100 

100 

•  150 

•  150 
200 
200 
300 

.       300 

15,  600 
15,  600 
15,  600 
15,  600 
15,  600 
15,  600 
15,600 
15,600 
15,  600 
15,  600 
15,  600 
15,  600 
15,600 
15,600 
15,600 
15,  600 

158,000 

149,  000 

I,  250,  000 

990,  000 

I,  765,  000 

I,  825,  000 

1,  875,  000 
,2,  250,  000 

2,  200,  000 
I,  950,  000 

165,  000 
170,  000 
0 
0 
0 
0 

Per  cent. 
1 

>■      100 
}      727 
}l.  165 
}i,338 

}i,35o 
I     108 

1         ° 
0 

f        110,500 
\        126, 000 
1   1,750,000 
I   1,350,000 
{  6,  600,  000 
1  5>  300,  000 

r  2, 025, 000 
1 3, 040, 000 
r  2,775,000 
\  3, 200, 000 
/    530, 000 
\    785, 000 

/            ° 
1            0 

/            ° 
I            0 

Per  cent. 
\      100 

(■I.  310 

(-5. 029 

|-2,  141 

1-2,525 

\  556 

}  ° 
}  » 

/        112,500 
\        115,000 

r  5, 000, 000 

\  6,  600,  000 

r  9, 150, 000 
1 7, 150, 000 
(15,950,000 

\i4,  600,000 
{  5,  Soo,  000 
1   5,  250,000 
1  3, 100,  000 
\  2,  750,000 
f                   0 
I 

{          I 

Per  cl. 
>        100 

}   5,097 
}  7. 161 
}i3,  423 
}  4,860 
}  2,  573 
1           ° 

}     » 

27807°— IS- 


192 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  4 


Table  VII. 


-Influence  of  calcium  nitrate  on  the  growth  of  Azotobacter  {strain  B)  in 
sterilized  soil 


Treat- 
ment 
(nitrate 

in 
100  gm. 
of  dry 
soil). 

Number  of  organisms  in  i  gm.  of  dry  soil. 

Culture 
No. 

At  begin- 
ning. 

After  I  week. 

Relative. 

After  2  weeks. 

Relative. 

After  3  weeks. 

Rela- 
tive. 

I 

2 

3 
4 

5 
6 

7 
8 

9 

10 
II 

Mgm. 

0 

0 

10 

10 

25 

25 

50 

50 

100 

100 

150 

150 

200 

200 

300 

300 

22,  000 
22,  000 
22,  000 
22,  000 
22,  000 
22,  000 
22,  000 
22, 000 
22,  000 
22,  000 
22,  000 
22,  000 
22,  000 
22,  000 
22,  000 
22,  000 

90s,  000 
860,  000 
23, 200,  000 
19,  600,  000 
17,  200,  000 
19,  600,  000 
11,800,  000 
14,  000,  000 
7,  500,  000 
II,  000,  000 

2,  550,  000 

3,  500,  ooo 
107,  500 

87,  500 
0 
0 

Per  cent. 
>      100 

|2,  423 

|-2,  084 

|l,  461 

}i>o53 
}     342 

}    " 
}    ° 

1   1,475,000 
\   I,  460,  000 
("28,  000,  000 
136,  000,  000 
(52,  000,  000 
l43>  500,  000 
("22,  500,  000 

\20,  000,  000 
112,  000,  000 

{    5,300,000 

\  6,  500,  000 
f  2,  750,  000 
I  3.225,000 

{          0 

Per  cent. 
\      100 

U  iSl 

13.255 
li,448 
1     818 
\     402 
1     203 
1          ° 

f  I,  130,  000 
I  1,157,500 
/34,  050,  000 
(34,  600,  000 
I29,  750,  000 

\22,  250,  000 
(■30,  400,  000 

\29,  850,  000 

r2i,  750, 000 

\i8,  950,  000 
\  4,  800,  000 
\ 

Per  ct. 
>      100 

|3,  002 

y,  273 

|2,  633 
}i,  780 
•     420 

1         ^^ 
f          ° 

13 
14 

15 
16 

f  130, 000 
\       120, 000 

{                  c 

A  glance  at  the  figures  of  Tables  V,  VI^  and  VII  shows  that  the  small- 
est concentration  of  nitrate  used  produced  a  much  more  marked  relative 
increase  in  numbers  with  strain  B  than  it  did  with  strain  A.  On  the 
other  hand,  the  greater  resistance  of  this  strain  to  the  higher  nitrate 
concentrations  is  clearly  evident.  In  the  potassium-  and  calcium- 
nitrate  cultures  the  organisms  were  present  in  an  active  state  where 
the  nitrate  was  added  in  amounts  equivalent  to  200  mgm.  of  nitrate 
in  100  gm.  of  soil.  However,  this  same  concentration  of  sodium  nitrate 
prevented  the  development  of  the  Azotobacter.  The  first  five  concen- 
trations of  all  three  nitrates  caused  a  very  large  increase  in  the  number 
of  Azotobacter  when  compared  with  control  cultures  where  no  nitrate 
was  added.  In  one  instance  an  enormous  increase  was  noted  after  three 
weeks'  incubation  in  the  presence  of  50  mgm.  of  nitrate  as  sodium  nitrate. 
This  increase  far  excelled  that  noted  with  other  concentrations  of  the 
same  salt.     The  writer  can  offer  no  conjecture  as  to  this  occurrence. 

Similar  results  were  obtained  by  the  writer  in  19 14  (25)  with  a  strain 
of  Azotobacter  isolated -from  a  silt  loam  soil  at  the  Pennsylvania  Experi- 
ment Station.  It  was  found  that  soil  and  liquid  cultures  containing 
small  amounts  of  potassium,  sodium,  and  calcium  nitrates  caused  an 
increase  in  the  number  of  Azotobacter  in  pure  culture  compared  with 
control  cultures  containing  no  nitrate.  An  increasing  concentration  of 
the  nitrates  continued  favorable  to  the  growth  of  the  organism  up  to  a 
certain  limit,  but  higher  concentrations  retarded  its  growth.  Finally 
a  nitrate  concentration  was  attained  at  which  Azotobacter  growth 
altogether  ceased. 


Jan.  2S,  i9i8  Nitrogen- Assimilating  Bacteria  193 

The  results  of  the  study  of  nitrates  and  their  influence  on  Azotobacter 
in  sterilized  soil  show  very  clearly  that  small  amounts  of  nitrate  cause 
a  great  increase  in  the  number  of  Azotobacter  cells.  Higher  concentra- 
tions are  not  so  favorable  to  the  growth  of  the  organisms,  and  the  highest 
concentrations  studied  prevented  the  development  of  the  Azotobacter 
in  sterilized  soil. 

From  a  study  of  the  results  of  these  experiments,  it  seems  that  the 
increase  in  number  of  Azotobacter  in  the  presence  of  small  amounts 
of  nitrate  is  a  direct  result  of  nuclear  stimulation.  Later  studies  to 
be  cited  (pp.  205-208)  show  that  nitrates  exerted  considerable  influence 
on  the  internal  structure  of  the  Azotobacter  cell.  It  appears  reasonable 
to  expect  that  the  nitrate  affected  the  nuclear  structure  in  such  a  manner 
that  an  increase  in  cell  multiplication  resulted.  It  seems  probable  that 
the  action  of  nitrate  as  a  simple  nutrient  would  be  shown  by  a  slower 
increase  in  cell  multiplication. 

INFLUENCE   OF   NITRATES   ON   THE  FIXATION   OF   NITROGEN   BY  .\ZOTOBACTER 

It  has  been  shown  in  the  preceding  paragraphs  that  the  presence  of 
small  quantities  of  nitrate  in  sterilized  soil  bring  about  a  large  increase 
in  the  number  of  Azotobacter.  This  increase  was  noted  in  the  case 
of  both  strains  of  Azotobacter.  It  would  be  of  interest  to  know  whether 
the  increase  in  bacterial  numbers  was  accompanied  by  a  corresponding 
increase  in  the  amount  of  nitrogen  assimilated. 

The  results  secured  by  a  few  investigators  indicate  that  in  the  presence 
of  combined  nitrogen  as  nitrates  the  nonsymbiotic  nitrogen-fixing 
organisms  will  not  fix  atmospheric  nitrogen.  Stoklasa  {44,  p.  492-50J) 
studied  the  influence  of  Azotobacter  on  sodium  nitrate  in  aerobic  and 
anaerobic  liquid  cultures.  He  found  only  a  small  gain  in  organic  nitro- 
gen and  from  these  results  he  concluded  that  in  the  presence  of  nitrates 
Azotobacter  could  not  assimilate  atmospheric  nitrogen.  It  has  been 
shown  by  Hanzawa  {20)  that  in  a  hquid  culture  containing  12  mgm, 
of  nitrate  (from  potassium  nitrate)  in  100  c.  c.  of  medium,  a  mixed 
culture  of  Azotobacter  fixed  5.25  mgm.  of  nitrogen.  Under  the  same 
conditions  with  60  mgm.  of  nitrate  present  in  100  c.  c.  of  medium  he 
found  but  5.35  mgm.  of  nitrogen  fixed.  He  concluded  that  nitrates 
remained,  as  far  as  small  quantities  were  concerned,  almost  without 
influence  on  the  amount  of  atmospheric  nitrogen  fixed  by  Azotobacter. 

Some  studies  have  been  carried  on  with  respect  to  the  influence  of  ni- 
trates on  the  nonsymbiotic  anaerobic  nitrogen-assimilating  organism, 
Clostridium  spp.  Bredemann  (9)  showed  that  ammonium  nitrate  in 
solution  caused  a  decrease  in  the  amount  of  nitrogen  fixed  by  species 
of  Clostridium.  Pringsheim  (40)  grew  cultures  of  C.  atnericamim  in 
solutions  containing  potassium  nitrate.  He  found  that  in  the  presence 
of  available  energy  the  organism  fixed  some  nitrogen  when  nitrate  was 


194 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  4 


present  but  to  a  less  extent  than  did  control  cultures  containing  no 
nitrate. 

From  these  results  it  appears  that  nitrates  do  not  stimulate  the  nitro- 
gen-assimilation of  the  nonsymbiotic  nitrogen-fixing  bacteria. 

Inasmuch  as  nitrates  in  small  amounts  caused  such  an  increase  in 
the  number  of  Azotobacter  in  sterilized  soil,  it  was  thought  advisable 
to  determine  just  what  influence  these  salts  exert  on  nitrogen  fixation 
by  Azotobacter.  Accordingly,  experiments  were  carried  out  with 
Azotobacter  on  agar  films,  in  soil  cultures  and  in  solution. 

Agar-film  cultures. — In  this  work  both  strains  of  the  Azotobacter 
were  used.  One  hundred  c.  c.  of  mannit  agar  were  placed  in  liter 
Erlenmeyer  flasks  and  nitrates  of  potassium,  sodium  and  calcium 
added  in  varying  quantities.  The  flasks  and  contents  were  sterilized 
at  ID  pounds'  pressure  for  25  minutes,  cooled,  and  inoculated  with  10 
c.  c.  of  a  suspension  of  the  organism  in  sterile  distilled  water.  The 
flasks  were  incubated  at  28°  C.  for  three  weeks.  The  weight  of  both 
inoculated  and  uninoculated  flasks  was  maintained  throughout  the 
experiment  by  the  addition  of  sterile  distilled  water.  At  the  end  of  the 
incubation  period  total  nitrogen  analyses  were  made.  Because  of  the 
high  nitrate  content  dilute  sulphuric-salycilic  acid  was  added  slowly 
and  carefully  to  prevent  loss  of  nitrogen  by  the  evolution  of  gaseous 
oxids  of  nitrogen.  The  acid  was  allowed  to  react  for  a  few  days  before 
continuing  the  total  nitrogen  determination.  The  results  of  the  experi- 
ments are  presented  in  Tables  VIII  and  IX. 

Table  VIII. — Influence  of  nitrates  on  the  fixation  of  nitrogen  by  Azotobacter  {strain  A) 

on  agar  films 


Cul- 
ture 
No. 


3 
4 
5 
6 

7 
8 

9 

10 


13 
14 
IS 


Treatment  (nitrate  in  looc.  c.  of  medium). 


O 

o 

o 

50  mgm.  of  NO3  potassium  nitrate . 

do 

100  mgm.  of  NO3  potassium  nitrate 

. . .  .do 

50  mgm.  of  NO3  sodium  nitrate. . . 

....do 

100  mgm.  of  NO3  sodium  nitrate. 

do 

50  mgm.  of  NO3  calcium  nitrate. . 

....do 

100  mgm.  of  NO3  calcium  nitrate . 
....do 


Nitrogen  contained  in  loo  c.  c.  of  medium. 


Inoculated. 


Found.      Average. 


Mgm. 
13.00 

12.  70 
12.  60 
18.50 
18.  40 
27.  60 

27-75 
18.65 
18.30 
27.  CO 
27.65 

13-75 
13-70 
18.80 

19-15 


Mgm. 
12.  80 

}  18.  45 
I  27.  70 
}  18.  50 
}  27-35 
}  13-  75 
}  18.  95 


Uninoculated. 


Found.      Average. 


Mgm. 
4.0 
4.0 
4-  I 
7.  00 

7.  20 
16.80 
15-70 

7-50 

7-30 

15.00 

15.  20 

8.00 

8.  50 
14.  50 
14-30 


Mgm. 
4-05 

I     7.10 
I   16.25 

}'- 

}    8-25 
14.40 


Nitrogen 
fixed. 


Mgm. 
8 

II 
II 
II 
12 

5 
4 


Jan.  2S,  1918 


Nitrogen- A  ssimilaiing  Bacteria 


195 


Table  IX. — Influence  of  nitrates  on  the  fixation  of  nitrogen    by  Azotobacter  {strain  B) 

on  agar  films 


Cul- 
ture 
No. 


Treatment  (nitrate  in  looc.c.  of  medium). 


Nitrogen  contained  in  100  c.  c.  of  medium. 


Inoculated. 


Found.      Average 


Uninoculated. 


Found.      Average. 


Nitrogen 
fixed. 


75  mgm.  of  NO3  as  potassixxm  ni- 
trate   

...do 

150  mgm.  of  NO3  as  potassium  ni- 
trate   

....do 

75  mgm.  of  NO3  as  sodium  nitrate 

do 

150  mgm.  of  NO3  as  sodium  nitrate 

do 

75  mgm.  of  NO3  as  calcium  nitrate 

do 

150  mgm.  of  NO3  as  calcium  nitrate 
do 


15-50 
15-70 
15.60 

25.  20 
25.40 

36.40 
36.90 
25.  60 
25.  70 
37.60 
37.  20 
20.  10 
19.  60 
32.80 
33-30 


Mgm. 


15.60 


25-30 


Mgm,. 


}36 

I  3i 


Mgm,. 


6.  40 


}  13-85 


I  23.60 
I  13-00 
}  25.  80 
I  12.3s 
}  24.  8s 


Mgm. 
9.  20 

11-45 

13-05 

12.65 

II.  60 

7-50 

8.  20 


A  glance  at  the  results  (Tables  VIII  and  IX)  shows  that  an  increase  in 
nitrogen  fixation  occurred  where  potassium  and  sodium  nitrates  were 
present,  whereas  a  marked  decrease  in  the  total  nitrogen  content  was 
observed  where  calcium  nitrate  was  used.  Whether  the  calcium  itself 
is  detrimental  to  an  increase  in  organic  nitrogen  or  whether  it  is  the  com- 
bination of  calcium  with  nitrate  can  not  be  stated.  It  is  significant, 
however,  that  this  decrease  in  fixation  of  nitrogen  was  noted  throughout 
all  the  experiments  where  calcium  nitrate  was  employed.  It  is  very- 
evident  that  calcium  nitrate  exerts  some  detrimental  effect  on  the  nitro- 
gen assimilating  properties  of  the  organism. 

There  seems  to  be  but  a  slight  difference  in  the  nitrogen-fixing  ability 
of  the  two  strains  studied.  In  the  absence  of  nitrates  the  amount  fixed 
varies  but  little.  Also  in  the  presence  of  potassium  and  sodium  nitrates 
the  relative  increase  in  amount  of  nitrogen  fixed  remains  about  the 
same.  Calcium  nitrate  offers  an  exception  where  it  is  employed.  The 
detrimental  effect  seems  to  be  more  marked  in  the  case  of  strain  A  than 
with  strain  B.  Strain  A  under  normal  conditions  fixed  slightly  less  nitro- 
gen than  strain  B,  so  it  may  be  possible  that  this  strain  is  weaker. 

The  formation  of  pigment  by  the  Azotobacter  in  the  presence  of  the 
nitrates  is  of  interest.  Strain  A  normally  produced  no  pigment  by  the 
end  of  three  weeks'  incubation.  But  when  grown  on  the  agar  films  in 
the  presence  of  nitrate  a  most  marked  pigment  production  appeared. 
This  pigment  was  especially  noticeable  in  the  presence  of  the  calcium 


196  Journal  of  Agricultural  Research  voi.  xii,  no.  4 

salt.  Since  strain  B  normally  produces  a  good  pigment,  the  influence 
of  nitrate  on  this  strain  was  not  very  marked.  The  relation  of  nitrates 
to  pigment  formation  will  be  taken  up  later  (pp.  203-205). 

From  the  results  of  the  experiments  with  agar  films  containing  various 
amounts  of  nitrate,  it  seems  apparent  that  potassium  and  sodium  nitrates 
in  amounts  of  50  and  100  mgm.  of  nitrate  in  100  c.  c.  of  medium  cause  a 
small  increase  in  the  amount  of  nitrogen  fixed.  However,  this  increase 
in  fixation  is  not  at  all  parallel  with  the  increase  in  number  of  Azoto- 
bacter  caused  by  nitrates  in  sterilized  soil. 

It  may  be  concluded  that  an  increase  in  the  number  of  Azotobacter  in 
sterilized  soil  as  a  result  of  nitrate  stimulation  does  not  mean  a  corre- 
sponding increase  in  nitrogen  fixation  on  agar  films. 

Soil  cultures. — ^The  conditions  obtaining  in  these  experiments  were 
strictly  comparable  with  those  heretofore  cited  dealing  with  the  influence 
of  nitrates  on  Azotobacter  in  sterilized  soil  (pp.  187-193). 

The  fixation  of  nitrogen  was  studied  in  pure  culture  in  sterilized  soil 
and  in  unsterilized  soil.  One  hundred  and  fifty  gm.  of  soil  (dry  weight) 
were  weighed  into  i -liter  Erlenmeyer  flasks,  nitrates  were  added  in  vary- 
ing amounts  from  10  to  200  mgm.,  and  i  per  cent  of  mannit  was  also 
added.  Triplicate  flasks  were  prepared  for  each  amount  of  nitrate  studied. 
The  moisture  content  was  raised  to  approximately  18  per  cent  and  the 
flasks  allowed  to  remain  at  room  temperature  for  one  day.  The  con- 
tents were  then  thoroughly  mixed  and  a  fine  crumb  structure  produced. 
The  flasks  for  the  experiments  with  pure  cultures  in  sterilized  soil  were 
immediately  sterilized  at  15  pounds'  pressure  for  three  hours.  After 
cooling,  two  of  each  set  were  inoculated  with  5  c.  c.  of  a  suspension  of 
Azotobacter  (strain  A)  in  sterile  distilled  water.  The  remaining  flask  of 
each  set  was  not  inoculated,  but  was  incubated  at  28°  C.  with  the  inocu- 
lated flasks.  The  moisture  lost  by  evaporation  was  replaced  from  time 
to  time  by  the  addition  of  sterile  distilled  water.  At  the  end  of  the  incu- 
bation period  the  soil  was  removed  and  spread  out  in  thin  layers  and 
allowed  to  dry.  It  was  then  thoroughly  ground  in  a  porcelain-ball  mill 
for  one  hour.  At  the  end  of  this  time  all  of  the  soil  passed  through  a 
loo-mesh  sieve. 

Soil  cultures  used  in  the  study  of  the  effect  of  nitrates  on  nitrogen 
fixation  in  unsterilized  soil  were  prepared  in  a  similar  manner,  except 
that  the  flasks  were  not  sterilized.  Previous  to  incubation  a  small 
inoculum  of  Azotobacter  (strain  A)  was  added  to  insure  the  presence  of 
the  nitrogen-fixing  organism  in  the  soil  cultures.  The  proper  moisture 
content  was  maintained  in  the  same  manner  as  in  the  case  of  the  pure 
cultures  in  sterilized  soil  and  the  incubation  period  was  the  same  for  both. 
The  results  are  given  in  Tables  X,  XI,  XII,  and  XIII. 


Jan.  28,  1918 


Nitrogen- A  ssimilating  Bacteria 


197 


Table  X. 


-Influence  of  sodium  nitrate  on  the  fixation  of  nitrogen  by  Azotobacter  in 
sterilized  soil 


Treatment 
(nitrate  in 
100  gm.  of 
dry  soil). 

Total  nitrogen  in  100  gm.  of  dry  soil. 

Nitrogen 

Culture  No. 

Inoculated. 

Uninoculated. 

fixed  in 
100  gni.  of 

Found. 

Average. 

Found. 

Average. 

dry  soil. 

I 

Mgm. 

0 

0 

0 

0 

0 

0 

10 

10 

10 

10 

10 

10 

50 

50 

SO 

50 

50 

50 

150 

150 

150 

150 

150 

150 

Mgm. 
135-0 
134.  0 
132.  0 
133-0 
137-0 

Mg7n. 

\     133-  7 

[     135-  0 

136.6 

[     137-  0 
[     149-  0 
[     149-  2 
\     162. 3 
[     162.  5 

Mgm. 
132.0 

131- 5 
131-0 

Mgm. 

\     131- 5 

Mgm. 

I 

2-  7 

I 

2 

I 

2 

1 

3-5 

::::::::::: 

7, 

137-0 
136.0 
137-0 
136-5 
137-5 
137-0 
149.0 
149.0 
149.  0 
148.5 
149-5 
149-5 
163.0 
162.  0 
162.  0 
162.  5 
163.0 
162.  0 

i     134-  0 

133-  5 

[     134-0 

[     133-  7 

■3, 

2.  9 

7, 

a. 

1 

A. 

f::::::::::: 

[           3-3 

A 

[..: 

c 

f     140. 0 

137-  0 

i     138-  5 

f 

}     138-5 

e 

10.  5 

e 

6 

] 

6 

\         10.7 

6 



[ 

7 

7 

f     152-0 
\     150-  0 
I     152. 5 

151-5 

10.8 

7 

8 

1 

8 

(::::::::::: 

>         II.  0 

8 

1 

J 

Table  XI. — Influence  of  sodium  nitrate  on  the  fixation  of  nitrogen  by  Azotobacter  in 

unsterilized  soil 


Treat- 
ment (ni- 
trate in 
100  gm.  ol 
dry  soil). 

Total  nitrogen  in  100  gm.  of  dry  soil. 

Nitrogen 

Culture  No. 

Inoculated. 

Uninoculated. 

fixed  in 
100  gm.  of 

Foimd. 

Average. 

Found. 

Average. 

dry  soil. 

I 

Mgm. 

0 

0 

0 

0 

0 

0 

10 

10 

10 

10 

10 

10 

50 

50 

SO 

SO 

50 

50 

150 

150 

150 

150 

150 

150 

Mgm,. 
132.0 
135-0 
13s- 0 
132.0 
134.0 
134.0 

137-5 
138.8 
138.8 
137-5 
137-  5 
138.0 
150.0 
151.  0 
150.0 
149.0 
149-5 
150.  5 
169.  0 

167.  0 

168.  0 

167-5 
168.0 
168.5 

Mgrit. 
134.  0 

•  133-  3 

•  137-  8 
137-  7 

•  150-  3 
■       149-  7 

168.  0 
168.  0 

Mgm^. 
[        130-  0 

133-  5 
I      132-  0 

Mgm. 
1        131-8 

Mgm. 

X 

2.  2 

I 

2 

1 

2 

1      '•' 

•2 

f      134-  0 
\       ^33-  0 
I       ^33-  0 

\       ^33-  3 

5 

4-5 

•2 

] 

J 



[         4-4 

A      .  .            .                       

C 

f       140. 0 
{       140. 5 
[       142. 0 

i       140. 8 

e 

9-5 

c 

6 .    .    . . 

1 

6 

, 

8.9 

6 .    .  . 

7 

f       148. 0 
{       154-  0 
I      153-  5 

1       151- 8 

7 

16.  2 

7 

8 

1 

8 

\       16.2 

8 

J 

198 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  4 


TABtE  XII. 


-Influence  of  calcium  nitrate  on  the  fiocation  of  nitrogen  by  Azotobacter  in 
sterilized  soil 


Treat- 
ment (ni- 
trate in 
100  gm.  of 
dry  soil). 

Total  nitrogen  in  100  gm.  of  dry  soil. 

Nitrogen 

Culture  No. 

Inoculated. 

Uninoculated. 

fixed  in 
ICO  gm.  of 

Fotmd. 

Average. 

Found. 

Average. 

dry  soil. 

1 

Mgm. 

0 

0 

0 

0 

0 

0 

10 

10 

10 

10 

10 

10 

50 

50 

50 

50 

50 

50 

200 

200 

200 

200 

200 

200 

Mgm. 
133- 0 
133-6 
133-3 
133-0 
134.2 

133-4 
137.0 
137-0 
136.5 
136.5 
137.0 

137-5 
148.  0 
148.5 
149.0 
148.5 
149.0 
148.  0 
173.0 
173-0 
174.0 

173-5 
173-0 
174.0 

Mgm. 
^33-  3 

133-5 

136.  8 

137.  0 
148.  5 
148.  5 
173-  7 
173-  5 

Mgm.. 
131.0 

131.  0 

132.  0 

Mgm.. 

1       ^3^-3 

Mgm. 

I 

2.  0 

I 

2 

1 

2 

\             2.  2 

2 

J 

•J 

135-  0 
134.0 

135-  0 

\       134-  7 

2 

2.  I 

2 

1 

1       '' 

A 

C 

140.  5 

141.  0 

140.  5 

i       140.  7 

e 

7.8 

e 

6 

1    ■• 

6 

6 

7 

163.  0 

164.  0 
164.  5 

163.8 

7 

9.9 

7 

8 

1 

8 

1        9-7 

8 

Table  XIII. 


-Influence  of  calcium  nitrate  on  the  fixation  of  nitrogen  by  Azotobacter  in 
unsterilized  soil 


Treat- _ 
ment  (ni- 
trate in 
100  gm.  of 
dry  soil). 

Total  nitrogen  in  100  gm.  of  dry  soil. 

Nitrogen 

Culture  No. 

Inoculated. 

Uninoculated. 

fixed  in 
100  gm.  of 

Found. 

Average. 

Found. 

Average. 

dry  soil. 

I 

Mgm.. 

0 

0 

0 

0 

0 

0 

10 

10 

10 

10 

10 

10 

50 

50 

50 

50 

50 

50 

200 

200 

200 

200 

200 

200 

Mgm. 

134-5 
136.0 

136-5 
135-0 
135-5 
135-5 
138.5 
138.0 
139.0 
138.0 
137.5 
138.5 
151- 5 
152.  0 
151.  0 

150.  0 

151-5 

151.  0 
177.0 
178.0 
176.  0 

176.5 
177.0 
178.  0 

Mgm. 

\       135-  7 

[       135-  3 

138.  5 

138.  0 

[  151.5 
150.  8 
177.0 

[       177-  2 

Mgm,. 
[         134.  0 

1       ^33-  5 
I       132. 0 

Algm. 

\       ^33-  2 

Mgm. 

I 

2-5 

1 

2 

1 

2     . .  .                .                     ... 

2.1 

2 

J 

•2 

[       133-  5 
133-0 
^33-  0 

\       ^33-  2 

? 

5-  3 

■t 

4 

1 

A 

\         4-8 

A 

J 

C 

[       140.  5 

141.  0 

L       141.  5 

i       141.0 

C 

10.  5 

c 

6 

1 

6 

9-8 

6 

7 

f       164. 0 

165.  0 

[       164. 0 

[       164. 3 

7 

12.  7 

7 

8 

1 

8 

[       12.9 

8. 

Jan.  28,  i9i8  Nitrogen- Assimilating  Bacteria  199 

It  will  be  seen  at  a  glance  that  a  greater  relative  increase  in  nitrogen 
fixation  in  the  presence  of  nitrates  occurred  in  the  soil  cultures  than  on 
the  agar  films.  But  in  the  latter  instance  the  amount  of  nitrogen  as- 
similated in  the  absence  of  mistakes  is  far  in  excess  of  that  assimilated 
in  the  soil  cultures  under  similar  conditions.  The  amount  of  nitrogen 
fixed  in  the  soil  cultures  is  surprisingly  low,  but  as  relative  increases  or 
decreases  are  desired  this  does  not  materially  influence  the  results. 

The  influence  of  sodium  nitrate  on  the  fixation  of  nitrogen  by  pure 
cultures  of  Azotobacter  in  sterilized  and  unsterilized  soil  is  brought  out 
very  clearly  in  the  figures  of  Tables  X  and  XL  In  both  cases,  where 
no  nitrate  was  added,  an  equal  fixation  of  nitrogen  occurred.  Where  10 
mgm.  of  nitrate  were  added  to  100  gm.  of  soil,  slightly  more  nitrogen  was 
assimilated  in  the  unsterilized  soil  than  in  sterilized.  The  reverse  seemed 
to  be  true  when  50  mgm.  of  nitrate  were  added.  But  in  the  presence  of 
150  mgm.  of  nitrate,  the  fixation  by  the  pure  culture  in  sterilized  soil 
did  not  increase  materially  in  comparison  with  that  which  occurred  in 
the  50  mgm.  of  nitrate  concentration.  Evidently  the  maximum  fixation 
under  these  conditions  had  been  reached.  The  gain  in  the  unsterilized 
soil  at  the  highest  concentration  of  nitrate  studied  almost  doubled  the 
amount  fixed  in  the  pure  culture.  It  appears  evident  that  the  presence 
of  sodium  nitrate  causes  a  greater  fixation  of  nitrogen  in  unsterilized  soil 
than  it  does  under  similar  conditions  in  sterilized  soil  inoculated  with 
Azotobacter. 

In  the  case  of  calcium  nitrate,  somewhat  comparable  results  were  ob- 
tained. The  fixation  where  no  nitrate  was  added  was  equivalent  to 
that  obtained  in  the  controls  for  the  sodium  nitrate.  Where  nitrate 
was  added  in  amounts  equal  to  10  mgm.  of  nitrate  in  100  gm.  of  soil,  an 
increased  fixation  was  obtained  in  the  unsterilized  soil,  but  practically 
no  increase  occured  in  the  pure  culture  in  sterilized  soil.  Fifty  mgm. 
of  nitrate  in  100  gm.  of  soil  produced  an  increase  in  fixation.  In  the 
highest  concentration  of  calcium  nitrate  the  difference  in  nitrogen  fixed 
between  the  pure  culture  in  sterilized  soil  and  unsterilized  soil  was  not  so 
great  as  in  the  case  where  sodium  nitrate  was  used. 

In  the  sterilized  soil  where  the  two  nitrates  were  present  in  equal 
amounts  it  can  be  seen  that  more  fixation  took  place  in  the  presence  of 
sodium  nitrate.  The  diilerence  is  not  marked,  but  it  exists  neverthe- 
less. It  will  be  remembered  that  calcium  nitrate  had  a  detrimental 
effect  on  nitrogen  fixation  by  Azotobacter  on  agar  films.  However,  in 
soil  cultures  this  same  nitrate  stimulated  Azotobacter  to  an  increased 
assimilation  of  nitrogen.  This  difference  is  not  suprising  as  it  has  been 
shown  repeatedly  that  bacterial  activities  in  soil  and  in  artificial  cultures 
are  not  always  comparable. 

From  the  results  of  the  experiments  performed  with  reference  to  the 
influence  of  nitrates  in  soil  on  the  fixation  of  nitrogen  therein,  it  appears 


200  Journal  of  Agricultural  Research  voi.  xii.  no.  4 

evident  that  in  pure  cultures  both  sodium  and  calcium  nitrates  in  the 
amounts  studied  produced  an  increase  in  the  amount  of  nitrogen  fixed. 
The  sodium  salt  stimulated  this  process  to  a  slightly  greater  extent  than 
did  the  calcium  salt.  In  unsterilized  soil  nitrates  exerted  the  same 
action  but  to  a  more  marked  extent.  The  amount  of  nitrogen  fixed 
under  these  conditions  was  generally  in  excess  of  that  fixed  under  similar 
conditions  in  sterilized  soil  inoculated  with  a  pure  culture  of  Azotobacter. 

Such  large  relative  increases  in  total  nitrogen  in  the  soil  in  the  presence 
of  nitrates  would  not  normally  take  place  under  field  conditions  for  here 
no  accumulations  of  nitrate  occur  in  quantities  sufficiently  large  enough  to 
influence  this  process. 

Summing  up  all  the  experiments  performed  in  relation  to  the  influence 
of  nitrates  on  the  fixation  of  atmospheric  nitrogen  by  Azotobacter,  it 
appears  that  the  increase  in  total  nitrogen  in  the  presence  of  these  salts 
is  by  no  means  comparable  to  the  increase  in  the  number  of  organisms 
in  sterilized  soil  under  the  same  conditions.  An  increase  in  the  number  of 
Azotobacter  does  not  mean  a  parallel  increase  in  the  amount  of  nitrogen 
fixed. 

INFLUENCE   OF   AZOTOBACTER  ON   NITRATES   IN   SOLUTION 

Attention  has  been  thus  far  directed  toward  the  influence  exerted 
by  nitrates  on  the  growth  and  nitrogen-assimilating  power  of  Azotobacter. 
The  following  points  are  now  to  be  considered:  Do  the  nitrogen-fixing 
bacteria  reduce  nitrates  to  nitrites  and  ammonia  ?  Is  there  an  increase 
or  decrease  in  the  amount  of  organic  nitrogen  as  a  result  of  the  presence 
of  nitrate  in  the  medium? 

Beijerinck  and  Van  Delden  (5)  found  that  Azotobacter  ckroococcum 
reduced  nitrate  directly  to  ammonia.  Stoklasa  (44,  p.  4^2-503) 
studied  the  changes  in  a  nutrient  solution  containing  0.2  per  cent 
of  sodium  nitrate  inoculated  with  Azotobacter.  He  found  under  an- 
aerobic conditions  that  the  nitrate  was  largely  reduced  to  nitrite  and 
ammonia  and  that  a  very  small  amount  of  organic  nitrogen  was  formed. 
Under  aerobic  conditions  there  v/as  more  nitrite  formed  than  under 
anaerobic  conditions  and  very  little  ammonia  or  oganic  nitrogen.  He 
concluded,  therefore,  that  Azotobacter  did  not  fix  atmospheric  nitrogen 
in  the  presence  of  nitrates. 

The  following  experiments  were  performed  in  an  endeavor  to  answer 
the  questions  raised  in  the  initial  paragraph  of  this  section.  To  Erlen- 
meyer  flasks  of  500-c.  c.  capacity,  containing  loo-c.  c.  portions  of  mannit 
solution,  sodium  and  ammonium  nitrates  were  added  in  amounts  equiva- 
lent to  150  mgm.  of  nitrate  in  100  c.  c.  of  the  solution.  Nine  flasks  were 
prepared  for  each  nitrate  and  the  same  number  for  the  controls  containing 
no  nitrate.  The  flasks  and  contents  were  sterilized  at  10  pounds  pres- 
sure for  30  minutes.     After  cooling,  six  of  each  set  were  inoculated,  three 


Jan.  28,  1918 


Nitrogen-A  ssimilating  Bacteria 


201 


with  strain  A  and  three  with  strain  B,  and  the  remaining  three  were  left 
uninoculated  to  serve  as  controls.  The  flasks  were  incubated  at  28°  C. 
for  21  days.  The  total  weight  was  maintained  throughout  the  experi- 
ment by  the  addition  of  sterile  distilled  water  from  time  to  time.  At  the 
end  of  three  weeks  the  contents  of  each  set  of  triplicate  flasks  were  poured 
together  and  50-c.  c.  samples  drawn  for  analysis.  Nitrate  ammonia  and 
total  nitrogen  were  determined  as  given  under  "Methods."  The  results 
are  shown  in  Tables  XIV,  XV,  and  XVI. 

Table  ^1\ .^Influence  of  Azotobacter  on  nitrates  in  solution,  giving  the  quantity  of 

nitrate  lost 


Treatment  (ni- 
trate   in    100 
c.   c.   of   me- 
dium). 

Nitrate  in  loo  c.  c.  of  medium. 

Cul- 

Strain A. 

Strain  B. 

ture 
No. 

Inoculated. 

Uninoculated. 

Nitrate 
lost. 

Inoculated. 

Uninoculated. 

Found. 

Aver- 
age. 

Found. 

Aver- 
age. 

Found. 

Aver- 
age. 

Found. 

Aver- 
age. 

lost. 

Mgm.. 

0.00 

.00 

80.9 
80.6 

100.3 
102. 1 

Mgm. 
>  0. 00 

}8o.7S 

>IOI.2 

Mgm,. 
f    0.00 
\      .00 

/iSO-4 
liSi-3 

/i49- 6 

Uso.  0 

Mgm. 
?•  0.00 

}i50.8 

}i49-8 

Mgm. 
0.00 

0—70.03 

a— 48.60 

Mgm. 
{    0.00 
\      .00 

f  los-  6 
\i05- 2 

/1311 
I130- 7 

Mgm. 
>  0.00 

}to5.4 
|i30.9 

Mgtn. 
(    0.00 
\      .00 

/150.4 
USl-3 

/149. 6 
Uso.o 

Mgm. 
!•  0.00 

}i5o.8 

}i49.8 

Mgm. 

0.00 

10-18 

isogm.of  NO3 
as  sodium  ni- 
trate  

10-18 

do 

i>— 45-40 

19-27 

iSomgm.ofNOs 
as  ammonium 

nitrate 

do 

«— 18.  90 

1  strong  NO2  r,eaction. 


6  Medium  NO2  reaction. 


cSlightNO:  reaction. 


Table  XV. — Influence  of  Azotobacter  on  nitrates  in  solution,  giving  the  quantity  of 

ammonia  produced 


Treatment  (ni- 
trate   in    100 
c.    c.   of  me- 
dium). 

Nitrogen  as  ammonia  in  100  c.  c.  of  medium. 

Cul- 

Strain A. 

Strain  B. 

ture 
No. 

Inoculated. 

Uninoculated. 

Am- 
monia 

pro- 
duced. 

Inoculated. 

Uninoculated. 

Am- 

Found. 

Aver- 
age. 

Found. 

Aver- 
age. 

Found. 

Aver- 
age. 

Found. 

Aver- 
age. 

pro- 
duced. 

1-9 

1-9 

10-18 

Mgm. 

0.  20 

.  10 

2.  CO 

1.80 

13-90 
13- 9S 

Mgm,. 
}  0.  IS 

}  1-90 

}l3-97 

Mgm. 
/    0.00 
\      .00 

|-.IO 

I      .  20 

/  13-90 
I  13-90 

Mgm,- 
\  0.00 

|i3-90 

Mgm. 
0.  IS 

1. 85 

-■07 

Mgm. 
f    0".  00 
\     '.20 

f    2.20 

\      2-40 

/ 13-80 
\  13-80 

Mgm. 
>  0. 10 

I  2.30 

}i3-8o 

Mgm. 
f    0.00 
\      .00 

f-.io 
\      .20 

/  13-90 
I  13-90 

Mgm. 
}■  0. 00 

|i3-90 

Mgm. 

0. 10 

ISomgm.ofNOs 
as  sodium  ni- 
trate  

.  do 

2.25 

19-27 
19-27 

ISomgm.ofNOs 
asammonium 

nitrate 

do 

.  10 

202 


Journal  of  Agriculiiiral  Research 


Vol.  XII,  No.  4 


Table  XVI. — Influence  of  Azoiobacter  on  nitrates  in  solution,  giving  the  quantity  of 

nitrogen  fixed 


Treatment  (ni- 
trate   in    100 
c.    c.   of  me- 
dium). 

Total  nitrogen  in  100  c.  c.  of  medium. 

Cul- 

Strain A. 

Strain  B. 

ture 

No. 

Inoculated. 

Uninoculated. 

Nitrogen 
fixed. 

Inoculated. 

Uninoculated. 

Nitro- 

Found. 

Aver- 
age. 

Fotmd. 

Aver- 
age. 

Found. 

Aver- 
age. 

Found. 

Aver- 
age. 

gen 
fixed. 

1-9 

Mgm. 
5.00 
5- 00 

22.50 
22.60 

47-00 
46.90 

Mgm. 
I  S-oo 

I22.SS 

I46. 95 

Mgm.. 
f    2.60 
I    2.70 

f  14.00 
I  14.20 

/  43-20 
I  42-90 

Mgm,. 
}  2-65 

|i4. 10 

}43-05 

Mgm. 
2.3s 

8-45 

3- 90 

Mg-rtt. 
/    5-00 
I    S-io 

r  28.00 
1 27.80 

/  48. 10 
\  48. 20 

Mgm,. 
}  5- 05 

I27.90 

}48- IS 

Mgm. 
f    2. 60 
\    2.70 

f  14.00 
\  14.  20 

/  43-  20 
I  42-90 

Mgm- 
}  2.65 

|i4. 10 
|43-0S 

Mgm. 

2.40 

lO-iS 
10-18 

isomgm.of  NOa 
as  sodium  ni- 
trate   

..  do 

13-80 

19-27 
19-27 

isomgm.of  NO3 
asammonium 

nitrate 

.....do 

5-10 

Table  XIV  showing  the  effect  on  the  total  nitrate  content  will  be 
discussed  first.  Strain  A  differed  widely  from  strain  B  in  its  ability  to 
reduce  nitrates.  It  will  be  noted  that  strain  A  reduces  nitrate  more 
readily  than  strain  B  in  the  presence  of  both  sodium  and  ammonium 
nitrate.  In  order  to  determine  the  nature  of  the  reduction  of  the  nitrates, 
qualitative  and  quantitative  tests  were  made.  The  reduction  of  nitrates 
by  Azotobacter  takes  place  with  the  formation  of  nitrites  as  shown  in 
Table  XIV.  Strain  A  effected  a  strong  reduction  of  nitrate  to  nitrite 
with  both  sodium  and  ammonium  nitrate.  Strain  B  also  reduced  nitrate 
to  nitrite,  but  to  a  lesser  degree  than  did  strain  A. 

An  inspection  of  the  data  in  Table  XV  indicates  that  the  reduction  of 
nitrates  ceased  with  the  formation  of  nitrite,  since  no  appreciable  amounts 
of  ammonia  were  produced  by  either  strain  of  Azotobacter. 

In  regard  to  the  fixation  of  atmospheric  nitrogen  by  these  strains  of 
Azotobacter,  it  was  found  that  nitrogen  was  assimilated  both  in  the 
presence  and  absence  of  nitrate.  In  the  presence  of  nitrate  there  was  a 
large  increase  in  the  total  organic  nitrogen.  Sodium  nitrate  stimulated 
both  strains,  although  strain  B  fixed  the  larger  amount.  Similar  results 
were  obtained  when  the  fixation  of  nitrogen  on  agar  films  was  studied. 
In  the  presence  of  ammonium  nitrate  the  amount  of  nitrogen  fixed  was 
considerably  decreased,  but  the  amount  fixed  was  in  excess  of  the  control 
cultures  containing  no  nitrate.  It  seems  evident  that  sodium  and 
ammonium  nitrate  in  the  amounts  studied  did  not  prevent  the  fixation 
of  atmospheric  nitrogen.  In  fact,  the  presence  of  these  salts  seemed  to 
stimulate  the  process. 

Under  aerobic  conditions  both  strains  of  Azotobacter  studied  caused  a 
reduction  in  the  total  amount  of  nitrate  present  in  the  solution.  This 
reduction  may  be  accounted  for  in  two  ways:  (i)  The  reduction  of 
nitrate  to  nitrite  and  (2)  the  assimilation  of  nitrate  by  the  organisms. 
Practically  no  ammonia  was  formed  under  the  conditions  of  these  experi- 
ments.    These  results  agree  with  those  of  Stoklasa.     However,  in  con- 


Jan.  28,  i9i8  Nitrogen-Assimilating  BacteHa  203 

trast  to  the  work  of  Stoklasa,  both  strains  of  Azotobacter  assimilated 
more  atmospheric  nitrogen  in  the  presence  of  nitrates  in  solution  than 
in  the  absence  of  these  salts. 

INFLUENCE  OF  NITRATES  ON  THE  PRODUCTION  OP  PIGMENT  BY  AZOTOBACTER 

It  has  already  been  noted  in  the  experiments  dealing  with  the  effect 
of  nitrates  on  the  fixation  of  atmospheric  nitrogen  on  agar  films  that 
nitrates  favor  pigment  production.  This  was  true  in  the  case  of  both 
strains  of  the  Azotobacter. 

Moreover,  it  has  been  observed  by  other  investigators  that  Azotobacter 
when  grown  in  the  presence  of  nitrate  will  produce  a  darker  pigment 
than  when  grown  in  its  absence.  Beijerinck  (4,  p.  575)  states  that 
Azotobacter  in  pure  culture  will  form  a  dark-brown  pigment  in  the 
presence  of  glucose  and  a  small  amount  of  nitrate.  Sackett  {43)  found 
that  nitrate  caused  an  increase  in  pigment  production  by  Azotobacter. 
In  media  without  the  nitrate  the  pigment  formation  was  materially 
decreased  and  in  some  cases  practically  eliminated.  He  also  noted  that 
the  amount  of  nitrate  present  has  a  direct  influence  on  the  intensity  of 
the  pigment  formation.  He  found  that  when  sodium  nitrate  was  added 
to  a  suitable  medium  to  give  a  content  of  0.0,  o.oi,  0.03,  0.05,0.08,  o.i, 
0.3,  and  0.5  per  cent,  with  glucose  used  as  the  source  of  energy,  the 
organisms  produced  pigment.  Streak  inoculations  were  made,  and 
after  14  days'  incubation  he  found  that  the  maximum  of  color  was 
obtained  at  0.05  to  0.08  per  cent  and  that  greater  concentrations  did  not 
increase  the  intensity  of  the  brown-black  pigment.  From  his  results  it 
is  evident  that  sodium  nitrate  caused  an  increase  in  pigment  formation 
by  azotobacter. 

In  order  to  determine  the  possible  effect  of  potassium,  sodium,  and  cal- 
cium nitrate  on  pigment  formation  with  strains  A  and  B,  the  following 
experiment  was  performed. 

Under  normal  conditions  on  mannit  agar  free  from  nitrate  strain  A 
produced  little  or  no  pigment  even  after  three  weeks'  growth.  At  the 
end  of  this  time  dirty-yellow  streaks  occurred  throughout  the  growth, 
but  no  brown  pigment  was  produced.  However,  with  strain  B  at  the 
end  of  two  or  three  weeks  a  decided  brown  to  brown-black  pigment  was 
produced  in  the  absence  of  nitrate. 

Agar  slope  cultures  containing  increasing  amounts  of  potassium,  so- 
dium, and  calcium  nitrate,  as  indicated  in  Table  XVII,  were  prepared. 
These  were  inoculated  with  both  strains  of  Azotobacter  and  incubated 
at  28°  C.  for  10  days.  Daily  observations  were  made  for  first  evidences 
of  pigment  formation.  In  some  of  the  cultures  of  strain  A  growing  on 
media  containing  calcium  nitrate  this  pigmentation  was  observed  as 
early  as  48  hours  subsequent  to  inoculation.  The  following  day  pig- 
mentation developed  in  strain  B.  The  cultures  on  the  potassium  and 
sodium-nitrate  media  began  to  show  evidence  of  pigmentation  in  four  to 
six  days.  The  final  results,  obtained  after  10  days'  incubation,  are  found 
in  Table  XVII. 


204 


Journal  of  Agricultural  Research 


VoK  XII,  No.  4 


1 

0 

0 

,i4 

'a 

S    H    Ci                 HO 

a  2  fl   •   •  2-^   • 

^pg          Qpq 

ii 

PQ 

til 

0 

0 

"S 

IH 

H 

,Q 

1 

g 

2 

W 

'd 

30000  0^0 
T^       •    •    •    -"S    ■ 

^ 

<u 

(33 

^ 

« 

d 

2 

d 

d 

■+-» 

K* 

s 

+3 

M 

a 

'a 

a 
2 

1 

0 

g    ; c 

S  0  0  0  0  0  o-Q 

1 

E 

0) 

c3 

s 

« 

a 

I 

(U 

en 

5h 

.a 

'3 

2 

PI 

c 

2 

« 

0  S  ^  OJ 

2 

^>3§ 

fi 

pq 

fi 

S 

I^ 

d 

<j 

2 

d 

0 

a 

■3 

+-> 

^ 

i-i 

42 

'3 

d 

^ 

0 

m 

t:) 

■^  c 

Sc 

c 

0 

ai  ^-d  -e'O.S'a'd     1 

OT 

0    OJ 

P 

V 

2 

^ 

•  d 

'r3 

2 

•   :  ^ 

s 

3 

^  p   : 

1 

■^-^  0  S  c 

c 

>  c 

) 

tu  •^■y  "Td  .s  "0 13  "Td     i 

(S 

^ 

s 

i!  ^'o 

•^ 

. 00  ij^  00000 

n 

e       Mwiooiooo 

«                                      H     M     C^     CO 

2  ° 

,2 

1 

g.a 

-2 

^ 

^ 

O     o 

a 

1    q;     . 
"3  v«   O 

H    C)     CO  ''t  lO^O    t^CO 

55^ 

1 

Jan.  28.  i9i8  Nitrogen-Assimiloting  BacteHa  205 

A  general  idea  may  be  gained  from  Table  XVII  concerning  the  relative 
increase  in  pigment  formation  in  the  presence  of  the  nitrates.  A  study 
of  the  table  gives  a  fair  idea  of  the  relative  differences  in  pigment  pro- 
duction. 

Very  interesting  results  were  obtained  with  strain  A.  It  will  be  seen 
from  Table  XVII  that  no  pigment  was  produced  in  the  control  culture 
after  10  days,  while  in  the  presence  of  nitrates  pigmentation  was  noted. 
The  intensity  of  the  pigment  varied  with  the  increase  of  nitrate  up  to  150 
mgm.  Beyond  150  mgm.  there  was  no  increase.  Potassium  and  sodium 
nitrate  did  not  exert  such  a  decided  influence  on  pigment  production  as 
calcium  nitrate.  The  latter  salt  produced  an  intense  dark -brown  to 
brownish-black  pigment. 

In  the  case  of  strain  B  the  influence  of  nitrate  was  not  so  pronounced 
since  this  strain  normally  produced  considerable  pigment  in  the  absence 
of  nitrates.  Potassium  and  sodium  nitrate  caused  a  slight  increase  in 
pigment  formation.  Here,  again,  the  calcium  salt  brought  about  most 
pronounced  increase.  However,  the  relative  increase  in  pigment  forma- 
tion in  strain  B  was  not  so  pronounced  as  in  strain  A. 

Where  the  nitrate  was  present,  a  much  more  spreading  growth  was 
obtained.  A  heavy  bacterial  growth  accumulated  at  the  base  of  the  slope 
except  in  the  two  cultures  in  which  the  highest  concentrations  were  used. 
In  the  latter  instances  the  accumulation  was  less  than  those  in  cultures 
growing  en  media  containing  no  nitrate.  Although  the  original  inocula- 
tion could  not  be  made  absolutely  uniform,  so  far  as  number  of  organ- 
isms was  concerned;  nevertheless  it  was  evident  that  on  those  slopes 
containing  10,  25,  50,  and  100  mgm.  of  nitrate  in  100  c.  c.  of  the  medium 
a  much  more  abundant  growth  was  obtained  than  on  those  slopes  free 
from  nitrate.  Here,  again,  it  is  seen,  in  a  rough,  comparative  way,  that 
the  smaller  amounts  of  nitrates  caused  an  increase  in  the  number  of 
Azotobacter. 

The  results  of  this  work  on  pigment  production  are  quite  in  accord  with 
those  of  Sackett.  Potassium,  sodium,  and  especially  calcium,  nitrates 
in  varying  amounts  increase  pigment  formation  by  Azotobacter  with  an 
increase  in  nitrate  concentration.  This  effect  is  especially  marked  in 
strain  A,  which  under  normal  conditions  does  not  produce  any  pigment. 

INFLUENCE   OF    NITRATES    ON    THE   FORMATION   OF    VGLrUTIN    BODIES   IN    AZOTOBACTER 

The  presence  of  volutin  bodies,  or  metachromatic  granules,  in  Azoto- 
bacter has  been  shown  by  Bonazzi  (7).  These  substances,  according  to 
Meyer  {34,  p.  238),  are  reserve  food  materials  other  than  fat  droplets, 
glycogen,  and  similar  substances  reacting  with  iodin  stain  which  occur 
in  the  cytoplasm  of  the  cells  of  various  bacteria.  With  Millon's  reagent 
they  give  no  reaction.  He  believes  that  these  bodies  are  composed  of 
nucleic-acid  compounds,  but  are  not  nuclear  proteids. 


2o6 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  4 


In  connection  with  the  foregoing  investigations  concerning  the  influ- 
ence of  nitrates  on  pigment  formation  by  Azotobacter,  it  was  thought 
that  some  results  of  cytological  interest  might  be  obtained  in  regard  to 
the  effect  of  varying  amounts  of  nitrates  on  the  volutin  bodies. 

Slope  cultures  of  mannit  agar  were  prepared  containing  the  different 
nitrates  as  indicated  in  Table  XVIII.  These  slopes  were  inoculated 
with  both  strains  of  Azotobacter  and  incubated  at  28°  C.  for  10  days. 
At  the  end  of  this  time  each  culture  was  stained  and  examined  micro- 
scopically. The  following  method  was  used  for  demonstrating  the 
presence  of  the  volutin  bodies.  The  organisms  to  be  examined  were 
air  dried  on  a  glass  slide  and  then  fixed  in  the  flame  of  a  Bunsen  burner. 
The  preparation  was  then  flooded  with  a  i  to  10  aqueous  solution  of 
methylene  blue  (Merck's)  prepared  by  adding  10  c.  c.  of  a  saturated 
aqueous  solution  of  methylene  blue  to  90  c.  c.  of  distilled  water.  The 
stain  was  washed  off  after  five  minutes  with  a  i  per  cent  solution  of 
sulphuric  acid  and  immediately  rinsed  in  distilled  water.  The  prepa- 
ration was  dried  and  examined  with  the  oil-immersion  objective.  The 
volutin  bodies  appeared  within  the  cytoplasm  as  very  dark  blue  dots, 
the  outline  of  the  cell  wall  was  a  lighter  blue,  while  the  cell  net  work 
was  stained  a  very  light  blue. 

Guignard's  stain  ^  was  also  used  to  demonstrate  the  presence  of  the 
volutin  bodies.  Fresh  smears  on  a  glass  slide  were  fixed  over  10  per 
cent  osmic  acid  for  three  minutes.  The  preparation  was  then  air-dried 
and  fixed  to  the  slide  by  rapidly  passing  the  latter  a  few  times  through 
a  Bunsen  burner.  The  preparation  was  covered  with  the  stain  which 
was  allowed  to  react  for  five  minutes.  The  stain  was  then  washed  off 
with  distilled  water,  dried,  and  examined  with  the  oil-immersion  objec- 
tive. The  outline  of  the  cell  as  well  as  the  net  work  wdthin-was  stained 
light  purple.  The  granules  within  the  cytoplasm  were  a  reddish  purple. 
The  results  are  given  in  Table  XVIII. 

Tabls  XVIII. — Influence  of  nitrates  on  the  formation  of  volutin  bodies  in  Azotobacter 

in  10  days 


Treat- 
ment 
(nitrate 
in  100 
c.  c.  of 

me- 
dium). 

Strain  A. 

Strain  B. 

Culture 
No. 

Potassium 
nitrate. 

Sodium 
nitrate. 

Calcium 
nitrate. 

Potassium 
nitrate. 

Sodium 
nitrate. 

Calcium 
nitrate. 

Mgm. 
0 
10 

25 

so 
100 
ISO 
200 
300 

Present." — 

do." 

do.o 

do  a 

Doubtful.... 
Present". . . . 

do.a 

do  6     . 

Doubtful 

Present". . . . 

Present" 

do." 

Doubtful.... 

Present" 

do.6 

Present." 

Do." 

do." 

do." 

Do." 

Doubtful 

do." 

do." 

Do." 

do  0 

do. 6 

do. ft 

do. 6 

Do." 

6.. 

do  " 

do  b 

do" 

....  do.b 

do.fc 

D0.6 

...     do.h  ... 

do.6 

do.b 

do.6 

do.6 

Do." 

8 

do.ft 

do.6 

do." 

do.6 

do.b 

D0.6 

"Representing  an  approximate  average  of  two  volutin  bodies  per  cell. 

b  Representing  an  approximate  average  of  four  volutin  bodies  per  cell. 

*  Guignard'sstain.  Fifty  c.  c.  of  2  per  cent  fuchsin  in  i  per  cent  acetic  acid;  40  c,  c.  of  0.2  per  cent  methyt 
green  in  i  per  cent  acetic  acid;  i  c.  c.  of  glacial-acetic  acid.  Distilled  water  was  used  in  making  the  i  per 
cent  acetic-acid  solution. 


Jan.  28,  1918 


Nitrogen- A  ssimilating  Bacteria 


207 


It  will  be  seen  that  all  three  nitrates  exerted  considerable  influence 
on  the  formation  of  volutin  bodies.  Not  only  was  the  number  of  bodies 
increased,  but  also  the  size.  The  relative  increase  in  size  of  the  granules 
was  much  more  marked  than  was  the  numerical  increase.  In  Azoto- 
bacter  cells  grown  on  mannit  agar  containing  no  nitrate  the  number  of 
volutin  bodies  in  each  cell  averaged  about  two ;  in  the  presence  of  nitrate 
four  to  five  volutin  granules  were  found.  The  greatest  increase  in  num- 
ber, as  well  as  size,  occurred  where  the  nitrate  concentration  was  highest. 
With  both  strains  sodium  nitrate  apparently  caused  the  greatest  increase. 
This  was  true  in  the  lower  as  well  as  in  the  higher  concentrations.  The 
volutin  bodies  in  strain  B  seemed  to  respond  to  the  presence  of  nitrates 
more  noticeably  than  did  those  of  strain  A,  especially  in  the  presence  of 
potassium  nitrate.  It  is  evident  that  nitrates  of  potassium,  sodium, 
and  calcium  cause  an  increase  in  the  number  and  size  of  volutin  bodies 
in  Azotobacter  cells. 

Do  these  salts  tend  to  hasten  the  appearance  of  these  bodies,  or  do 
they  at  first  retard  their  development?  The  following  experiment  was 
carried  out  in  an  endeavor  to  determine  this  point.  Only  sodium  nitrate 
was  used,  since  this  particular  salt  proved  most  beneficial  to  the  forma- 
tion of  volutin  bodies.  Agar  slopes  were  prepared  containing  the  different 
amounts  of  nitrate  as  indicated  in  Table  XIX.  The  cultures  were  incu- 
bated at  28°  C.  and  examined  daily  for  the  presence  of  volutin  bodies. 
The  methylene  blue — i  per  cent  sulphuric  acid — method  of  staining  was 
employed.     The  results  of  the  experiment  are  given  in  Table  XIX. 

Table  XIX. — Influence  of  sodium  nitrate  on  the  rate  of  formation  of  volutin  bodies. in 

Azotobacter 


Time. 

Nitrate  in  100  c. 

c.  of  mediiun. 

Strain  A. 

strain  B. 

0  Mgm. 

25  Mgm. 

100  Mgm. 

300  Mgm. 

0  Mgm. 

2S  Mgm. 

100  Mgm. 

300 
Mgm. 

Day. 

Absent... 

Present". 
...do.".... 
...do.".... 

Absent.. . 

Present". 
...do.".... 
...do.".... 

Doubtful. 

Present". 
...do.".... 
...do.b... 

Doubtful. 
...do.b... 
...do.b... 
...do.b... 

Absent.. . 
...do 

Absent... 
Present". 

Doubtful. 
...do.b... 
...do.b... 
...do.b... 

Dbtful. 

Do.b. 

3 

4 

Present". 
...do.".... 

...do.b... 
...do.b... 

Dob. 
Do.b. 

"  Representing  an  approximate  average  of  two  volutin  bodies  per  cell. 
b  Representing  an  approximate  average  of  four  volutin  bodies  per  cell. 

A  study  of  Table  XIX  shows  that  it  is  rather  doubtful  whether  the 
nitrate  present  tended  to  hasten  the  appearance  of  the  volutin  bodies. 
No  convincing  evidence  has  been  presented  for  or  against  this  statement. 
No  granules  were  seen  in  the  first  day's  growth  of  strain  A,  although 
the  next  day  they  were  present  in  all  four  cultures.  In  strain  B  more 
convincing  proof  is  furnished  that  the  sodium  nitrate  hastened  the 
appearance  of  these  reserve  food  substances.  The  volutin  bodies  were 
not  present  in  the  control  and  lowest  nitrate  concentration  cultures 
the  first  day,  but  they  were  very  noticeable  in  the  culture  containing 
the  highest  concentration  of  nitrate  and  doubtful  in  the  remaining  one. 
On  the  second  day  volutin  bodies  were  present  in  all  cultures  grown  on 
27807°— 18 5 


208 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  4 


nitrate  media,  while  the  control  culture  was  still  free  from  them.  The 
third  day  showed  the  presence  of  volutin  bodies  in  all  four  cultures. 
Strain  B  offers  the  better  proof  that  sodium  nitrate  tends  to  hasten  the 
appearance  of  volutin  bodies  in  the  cells  of  Azotobacter.  Further 
experiments  were  not  made  in  an  endeavor  to  determine  what  influence 
nitrates  might  have  on  the  cytology  of  the  Azotobacter  cell.  The  brief 
studies  reported  here  were  made  in  connection  with  the  pigment  forma- 
tion experiments,  but  do  not  bear  any  particular  relation  to  them.  The 
increase  in  number  and  size  of  volutin  bodies  may  bear  some  relation  to 
the  increased  amount  of  nitrogen  fixed  or  assimilated  by  Azotobacter  in 
the  presence  of  nitrates. 

INIflvUENCE   OF    NITRATES   ON    BACILLUS   RADICICOLA 

INFLUENCE  OP  NITRATES  ON  THE  GROWTH  AND  REPRODUCTION  OF  BACILLUS  RADICICOLA 

IN  STERILIZED  SOIL 

One  hundred  and  fifty  gm.  (dry  weight)  of  the  soil  were  weighed 
into  500-c.  c.  E^rlenmeyer  flasks  and  the  nitrates  added  as  indicated 
in  Tables  XX-XXII.  Duplicate  cultures  for  each  amount  of  nitrate 
were  prepared.  One  per  cent  of  mannit  (in  5  c.  c.  of  distilled  water) 
was  also  added.  The  flasks  were  kept  at  room  temperature  for  one 
day  and  the  contents  then  thoroughly  mixed.  The  flasks  were  steri- 
lized at  15  pounds'  pressure  for  three  hours.  Upon  cooling  they  were 
inoculated  with  5  c.  c.  of  a  suspension  of  Bacillus  radicicola  in  sterile 
distilled  water.  The  number  of  bacteria  in  the  inoculum  was  deter- 
mined. The  moisture  content  was  then  approximately  18  to  20  per 
cent.  The  flasks  were  incubated  at  28°  to  30°  C.  and  mannit-agar 
plates  poured  at  the  end  of  one  and  two  weeks.  The  results  of  these 
experiments  are  given  in  Tables  XX,  XXI,  and  XXII,  in  which  each 
figure  represents  an  average  of  duplicate  plates. 

Table  XX. — Influence  of  potassium  nitrate  on  Bacillus  radicicola  in  sterilized  soil 


a  Contamination, 


Jan.  38, 1918 


Nitrogen- A  ssimilating  Bacteria 


209 


Table  XXI. — Influence  of  sodium  nitrate  on  Bacillus  radicicola  in  sterilized  soil 


Treat- 
ment (ni- 
trate in 
100  gm.  of 
dry  soil). 

Nimiber  of  organisms  in  i  gm.  of  dry  soil. 

Ctdture  No. 

At  begin- 
ning. 

After  I  week. 

Relative. 

After  2  weeks. 

Relative. 

I 

Mgm. 

0 

0 

10 

10 

25 
2C 

15,  500 
15.  500 
i5»  500 
i5>  500 
15.  500 

TC      coo 

I,  500,  000 

1,  250,  000 

2,  560,  000 

3,  000,  000 
6,  150,  000 
5,  375,  000 

4,  850,  000 

5,  570,  000 
2,  000,  000 
1, 850,  000 
1, 060,  000 

835, 000 
760,  000 

725,  OOO 

250,  000 
365,  000 

Per  cent, 
f             100 

>  201 
1             418 
}             378 

>  140 
}               69 

}                ^^ 

>  22 

r        6,  750,  000 

I      5.  950,  000 

(      10,  000,  000 
\       12,  500,  000 
/       14,  650,  000 
1       15.  700,  000 
/               («) 

\      8,  500,  000 
f       I,  520,  000 
\       I,  650,  000 
f          850, 000 
\          940, 000 

r       500, 000 
\      620, 000 
f       150,000 
\       210,000 

Per  cent. 

2 

\      100 

5 

4. 

I      177 

r 

6 

(      240 

7 

50              15.  500 

50              15,  500 

100            T  c.  con 

8 

}      134 

0 

1 

10 

100 
150 
150 
200 
200 
300 
300 

-J7    J- 

i5>  500 
15,  500 
15.  500 
i5»  500 
15.  500 
15.  500 
■  15,  500 

(         ^5 

II    

12 

I        ^"^ 

XT. 

14 

1         8.8 

le 

16 

1         2.8 

a  Contamination. 
Table  XXII. — Influence  0/  calcium  nitrate  on  Bacillus  radicicola  in  sterilized  soil 


Treat- 
ment (ni- 
trate in 
100  gm.of 
dry  soil). 

Number  of  organisms  in  i  gm.  of  dry  soil. 

Culture  No. 

At  begin- 
ning. 

After  I  week. 

Relative. 

After  2  weeks. 

Relative. 

I 

Mgm. 

0 

0 

10 

10 

25 

25 

50 

50 

100 

100 

'     150 

150 

200 

200 

300 

300 

p  p  p  p  p  p  p  p  p  p  p  p  p  p  p  p 
OOOOOOOOOOOOOOOO 

960,  000 
850,  000 

3,  650,  000 
3,940,  000 

5,  500,  000 

6,  700,  000 

4,  000,  000 
3,  500,  000 

1,  200,  000 

2,  050,  000 
865, 000 

I,  050,  000 

375)000 

260,  000 

35.  000 

47,  000 

Per  cent. 

>  100 

I        419 
1        674 
1        414 
}        180 

>  106 

I        35 
I          4-5 

f        4,  675,  000 

Per  cent. 
1 

2 

>    100 

7 

A 

I    5,450^000  ■/  ^^4 

c 

6 

I   14,700,000  /   '^ 
/    9.350,000  \\ 
I    8,670,000  /  ^95 
f    1,500,000  \ 

I          1,750,000    ij       >55 

/                765,000,;! 

\               800    000    i  '           ' 

7 

8 

0 

10 

12 

13 

14 

/               350,  000 
\               300,  000 
f                  25,000 
\                  40,  000 

}    '■» 

15 

16 

I          .70 

An  inspection  of  all  three  tables  reveals  two  marked  differences  from 
the  results  obtained  in  similar  woric  with  Azotobacter.  First,  it  will  be 
noted  that  nitrates  do  not  appear  to  exert  such  a  marked  stimulating 
effect  with  B.  radicicola  as  with  Azotobacter.  The  numerical  increase 
due  to  the  presence  of  the  nitrate  is  clearly  shown  in  the  percentage 
columns.  Second,  it  will  be  noted  that  B.  radicicola  does  not  seem  to  be 
so  sensitive  to  higher  concentrations  of  nitrates  as  does  Azotobacter. 
In  all  instances  at  concentrations  equivalent  to  300  mgm.  of  nitrate  in 


2IO 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  4 


lOO  gm.  of  soil  the  legume  organisms  were  still  alive,  although  present 
in  numbers  far  below  those  of  the  control  cultures.  In  all  Azotobacter 
cultures  no  organisms  survived  this  concentration. 

No  one  nitrate  produced  an  excessive  stimulation  in  comparison  with 
the  others.  The  calcium  salt  present  as  150  mgm.  of  nitrate  in  100  gm. 
of  soil  at  the  end  of  the  first  week  gave  the  greatest  stimulation  for  con- 
centrations of  that  amount.  However,  at  the  end  of  the  second  week 
this  concentration  had  caused  a  marked  decrease  in  the  number  of  or- 
ganisms. In  the  case  of  all  three  nitrates  the  concentration  representing 
25  mgm.  of  nitrate  in  100  gm.  of  soil  produced  the  greatest  stimulation. 
This  resulting  stimulation  also  held  true  throughout  the  second  vv^eek. 
The  decrease  in  number  below  those  of  the  control  cultures,  due  to  in- 
creasing concentrations  of  nitrate,  began  first  in  the  presence  of  potassium 
nitrate  at  100  mgm.  of  nitrate  per  100  gm.  of  soil,  then  with  sodium 
nitrate  at  150  mgm.,  and  lastly  with  calcium  nitrate  at  200  mgm.  But 
the  number  of  organisms  present  in  the  soil  cultures  containing  sodium 
nitrate  in  amounts  equivalent  to  100  mgm.  and  calcium  nitrate  at  100 
mgm.  at  the  end  of  the  second  week  was  below  those  of  the  control 
cultures. 

It  therefore  appears  from  these  results  that  small  amounts  of  potassium, 
sodium,  and  calcium  nitrate  stimulate  the  reproductive  activity  of  B. 
radicicola.  Concentrations  of  nitrates  greater  than  those  amounts  which 
produced  maximum  stimulation  cause  a  decrease  in  the  number  of  or- 
ganisms. The  highest  concentration  of  nitrate  studied  did  not  entirely 
prevent  the  growth  of  the  bacteria,  but  it  reduced  the  number  of  organ- 
isms far  below  those  contained  in  control  cultures  where  no  nitrates 
were  added. 

Ammonium  nitrate  was  also  employed.  The  soil  cultures  were  pre- 
pared as  already  described  and  inoculated  with  B.  radicicola.  The  cul- 
tures were  incubated  at  28°  to  30°  C.  and  counts  were  made  at  the  end 
of  one  and  two  weeks'  time.  The  results  of  the  study  with  ammonium 
nitrate  are  given  in  Table  XXIII. 

Table  XXIII. — Influence  of  ammonium  nitrate  on  Bacillus  radicicola  in  sterilized  soil 


Treat- 
ment (ni- 
trate in 
100  gm.  of 
dry  soil). 

Number  of  organisms  in  i  gm.  of  dry  soil. 

Culture  No. 

At  begin- 
ning. 

After  I  week. 

Relative. 

After  2  weeks. 

Relative. 

I 

Mgm. 

0 

0 

25 

25 

100 

100 

200 

200 

S  p  p  p  p  p  p  p 
OOOOOOOO 

850,  000 
765,  000 

2,  500,  000 

3,  050,  000 
I,  350,  000 
I,  050,  000 

700,  000 
655,  000 

Per  cent. 
>              100 

}             343 
}             148 
}               84 

f        I,  365,  000 
\        I,  400,  000 

r     5, 060, 000 
\    4, 320, 000 
r     1, 030, 000 
\       950, 000 
r       63  s,  000 
\       605, 000 

Per  cent. 

1               TOO 

2        

5 

}               338 

4 

c 

}          - 

6     

7 

}         « 

8 

jaM.  28.  i9i8  Nitrogen-Assimilating  Bacteria  211 

From  the  results  as  a  whole  it  appears  that  it  is  the  nitrate  radical 
and  not  the  combined  salt  which  causes  the  increase  in  the  number  of 
B.  radicicola  when  small  amounts  of  nitrates  are  present.  A  stimulation 
occurred,  resulting  in  an  increase  in  number  which  is  quite  comparable 
to  that  obtained  with  potassium,  sodium,  and  calcium  nitrates.  The 
highest  concentration  of  ammonium  nitrate  used  did  not  appear  to  have 
such  an  inhibiting  effect  as  did  the  corresponding  concentrations  of  the 
three  other  salts. 

Throughout  the  work  with  B.  radicicola  in  sterilized  soil  compara- 
tively low  numbers  of  these  organisms  were  found.  Whether  or  not 
this  depression  was  due  to  toxic  substances  formed  as  a  result  of  steriliza- 
tion can  not  be  stated.  If  this  decrease  in  numbers  as  a  result  of  the 
presence  of  toxic  substances  is  true,  it  is  very  evident  that  the  detrimental 
effect  had  not  become  materially  lessened  at  the  end  of  the  incubation 
period.  However,  in  any  event  the  validity  of  the  outcome  is  not  im- 
paired, since  comparative  and  not  absolute  data  are  of  importance  and 
since  in  all  probability  the  same  conditions  obtained  throughout  the 
cultures. 

It  seems  certain  from  the  results  of  these  studies  on  the  effect  of 
potassium,  sodium,  calcium,  and  ammonium  nitrates  on  the  growth  of  B. 
radicicola  in  sterilized  soil  that  small  amounts  of  nitrate  stimulate  the 
growth  of  the  organisms.  It  is  also  shown  that  B.  radicicola  is  much 
more  resistant  than  Azotobacter  to  higher  concentrations  of  potassium, 
sodium,  calcium,  and  ammonium  nitrates. 

INFLUENCE   OP   BACILLUS   RADICICOLA  ON   NITRATES  IN  SOLUTION 

The  series  of  soil  culture  experiments  just  discussed  served  to  give  an 
idea  concerning  the  effect  of  nitrates  on  the  legume  organism.  It  was 
found  that  in  small  amounts  nitrates  stimulated  the  bacteria  to  increased 
reproduction.  But  no  study  was  made  as  to  the  effect  of  Bacillus  radicicola 
on  the  nitrate.  Does  the  organism  break  up  the  nitrate,  reducing  it  to 
nitrite  or  ammonia  ?  Does  it  cause  any  loss  in  nitrate  when  grown  in  a 
solution  containing  that  salt?  Beijerinck  {2,  p.  J62)  as  a  result  of 
physiological  experiments  with  B.  radicicola,  states  that  the  organism  does 
not  reduce  nitrate.  Prucha  {41)  also  states  that  B.  radicicola  does  not 
reduce  nitrates.  However,  Zipfel  (49)  found  that  B.  radicicola  will  reduce 
nitrates  to  nitrites  but  not  to  ammonia. 

The  following  experiments,  somewhat  similar  to  those  already  cited  in 
relation  to  Azotobacter,  were  carried  out  in  an  endeavor  to  answer  these 
questions. 

To  twenty  500-c.  c.  Erienmeyer  flasks  containing  200  c.  c.  of  mannit 
solution,  potassium,  sodium,  calcium,  and  ammonium  nitrates  were 
added  as  indicated  in  Tables  XXIV,  XXV,  and  XXVI.  Quadruplicate 
flasks  were  prepared  for  each  concentration  of  nitrate  and  for  the  control 
cultures  without  nitrate.     The  flasks  and  contents  were  sterilized  at 


212 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  4 


lo  pounds'  pressure  for  30  minutes.  After  cooling,  two  of  each  set  of  four 
flasks  were  inoculated  with  5  c.  c.  of  a  suspension  of  B.  radicicola  in 
sterile  distilled  water.  The  remaining  two  flasks  of  each  set  (uninocu- 
lated)  served  as  controls.  The  flasks  were  incubated  at  28°  C.  for  2 1  days. 
The  total  weight  of  the  flasks  was  maintained  throughout  the  incuba- 
tion period  by  the  addition  of  sterile  distilled  water  from  time  to  time. 
At  the  expiration  of  the  period  of  incubation  the  nitrate,  ammonia,  and 
total  nitrogen  contents  were  determined  as  given  under  "Methods  used 
in  experiments."  The  contents  of  the  duplicate  inoculated  flasks  were 
poured  together  and  50  c.  c.  samples  drawn  for  analysis.  The  same 
procedure  was  followed  in  the  case  of  the  uninoculated  flasks.  The 
results  are  given  in  Tables  XXIV,  XXV,  and  XXVI. 

Table  XXIV. — Influence  of  Bacillus  radicicola  on  nitrates  in  solution  giving  the  quantity 

of  nitrate  lost 


Treatment  (nitrate  in  loo  c.  c.  of  medium). 

Nitrate  in  loo  c.  c.  of  medium. 

Cul- 
ture 
No. 

Uninoculated. 

Inoculated. 

Nitrate 
lost. 

Found. 

Average. 

Fotmd. 

Average. 

None 

Mgm. 
0.  00 
.  00 

151- 4 
151.  0 
148.8 
148.8 
154.8 
155-6 
151- 4 
151.  6 

Mgm. 
>      0.  00 

}i5i-  2 
|i48.  8 
}i55-2 
}i5i-5 

Mgm.. 
(      0.  00' 
\        .  00 

/117-O 
\li7-0 
[ii4-4 
\li4.  0 
/  76.6 
I  76.8 
fi42.  6 
\i42.  6 

Mgm.. 
>      0.  00 

>ii7.  0 

V114.  2 

}  76.7 
>i42.  6 

Mgm. 

..     do 

0.  00 

3 
4 

5 

150  mgm.  of  NO3  as  potassium  nitrate . 
do          

-34-2 

150  mgm.  of  NO3  as  sodium  nitrate .  . 
.  .  do 

-34-6 

7 
8 

150  mgm.  of  NO3  as  calcium  nitrate . 
do 

-78.5 

9 

150  mgm .  of  NO3  as  ammonium  nitrate 
do  

-8.9 

Table  XXV. — Influe?tce  of  Bacillus  radicicola  on  nitrates  in  solution  giving  the  quantity 
of  nitrogen  as  ammonia  fanned 


Treatment  (nitrate  in  loo  c.  c.  of  medium). 

Nitrogen  as  ammonia  in  loo  c.  c.  of  medium. 

Nitro- 

Cul- 
ture 
No. 

Uninoculated. 

Inoculated. 

gen  as 
ammo- 
nia 

Found. 

Average. 

Found. 

Average. 

formed. 

None 

Mgm. 
0.  20 
.  10 
.  10 
.  20 
.  20 
.  20 
.40 
•30 
13.90 

13-95 

Mgm.. 
}      0.  15 

}        ■  ^^ 
>        .  20 

}      -35 
}  13-92 

Mgm. 
j      0.  10 
\        .  20 
J        .  00 
\        .  10 
f         .  20 
I        .30 
/        -30 
\        .  10 
f    13.80 
I   13-  85 

Mgm. 
}      0-15 

}         '°^ 

}         '^^ 
>        .  20 

}  13- 82 

Mgm. 

...do 

0.  00 

3 
4 

5 
6 

150  mgm.  of  NO3  as  potassium  nitrate . 
....  do 

— .  10 

150  mgm.  of  NO3  as  sodium  nitrate . . 
do 

+.05 

7 
8 

150  mgm.  of  NO3  as  calcium  nitrate. 
do 

--IS 

9 

JO 

150  mgm.  of  NO3  as  ammonium  nitrate 
do 

-f.io 

Jan.  38, 1918 


Nitrogen- A ssimilating  Bacteria 


213 


Table  XXVI. — Influence  of  Bacillus  radicicola  on  nitrates  in  solution  giving  the  quantity 

of  nitrogen  fixed 


Cul- 
ture 
No. 


2 
3 
4 
5 
6 

7 
8 

9 
10 


Treatment  (nitrate  in  100  c.  c.  of  medium). 


None 

....do 

150  mgm.  of  NO3  potassium  nitrate . . 

....do 

150  mgm.  of  NO3  sodium  nitrate.  .  .  . 

....do 

150  mgm.  of  NO3  calcium  nitrate .  .  . 

....do 

150  mgm.  of  NO3  ammonium  nitrate . 
....do 


Total  nitrogen  in  100  c.  c.  of  medium. 


Uninoculated. 


Inoculated. 


Found.      Average.      Found.      Average. 


MgTti. 
2.  40 
2.50 
18.  00 
17.90 
16.80 
17.  CO 

14.  00 

13.80 

40.50 

41.  20 


Mgm. 

45 


13 


40, 


95 


90 


Mgm. 
3-3° 
3-5° 

18.  70 

19.  00 
19.30 
19.  20 
14.  60 
14.  70 
41.30 
41.  70 


Mgm. 
}  3- 40 

}  18.  8s 

}  19-25 

}  14-65 

\   41-  50 


Nitro- 
gen 
fixed. 


Mgm. 

0-95 

.90 

2-35 
•  75 
•65 


The  data  in  Table  XXIV  show  that  a  rather  large  reduction  in  the 
total  nitrate  content  took  place.  This  reduction  varied  rather  markedly 
among  the  four  different  nitrates  studied.  The  greatest  reduction  oc- 
curred where  calcium  nitrate  was  used.  Potassium  and  sodium  were 
next  in  order;  the  loss  was  almost  the  same  for  both  salts.  Ammonium 
nitrate  was  last  with  but  a  very  small  comparative  reduction  in  total 
nitrate. 

The  question  arises  as  to  whether  the  nitrate  is  reduced  to  nitrite, 
ammonia,  or  elemental  nitrogen  or  whether  the  reduction  in  amount  is 
due  to  a  natural  assimilation  of  the  nitrate  by  the  organisms.  The  first 
possibility  was  precluded  when  qualitative  tests  for  nitrites  were  made 
and  none  found.  Table  XXV  reveals  the  fact  that  no  ammonia  was 
produced.  Table  XXVI  shows  no  loss  in  total  nitrogen.  Therefore  it 
seems  obvious  that  reduction  in  total  amount  of  nitrate  present  is  brought 
about  by  the  assimilation  of  those  compounds  by  the  organisms. 

An  inspection  of  Table  XXVI,  which  gives  the  results  of  the  total 
nitrogen  determinations,  shows  that  a  slight  fixation  of  atmospheric 
nitrogen  took  place.  This  fixation  is  entirely  possible,  as  will  be  shown 
later  when  the  influence  of  nitrates  on  the  fixation  of  nitrogen  is  taken 
up.  In  the  presence  of  potassium,  sodium,  and  ammonium  nitrates  the 
amount  of  nitrogen  assimilated  is  somewhat  decreased.  But  in  the 
case  of  sodium  nitrate  a  large  increase  in  the  amount  of  total  nitrogen 
seems  to  have  taken  place.  This  is  interesting  in  the  light  of  results 
to  be  presented  later. 

From  the  results  of  the  work  on  the  effect  of  B.  radicicola  on  nitrates 
it  may  be  concluded  that  the  organisms  do  not  reduce  the  nitrates  to 
nitrite  or  ammonia  or  elemental  nitrogen  under  aerobic  conditions. 


214  Journal  of  Agricultural  Research  voi.  xii.  No.  4 

INFLUENCE  OF   NITRATES   ON  THE  FIXATION  OF  ATMOSPHERIC   NITROGEN   BY   BACILLUS 

RADICXCOLA 

The  ability  of  B.  radicicola  to  fix  atmospheric  nitrogen  in  the  absence 
of  the  host  plant  has  been  studied  by  numerous  investigators.  From 
the  results  of  their  work  it  seems  fairly  probable  that  the  legume  organ- 
ism can  fix  nitrogen  to  a  slight  extent  when  growing  in  a  nonsymbiotic 
state.  Beijerinck  (j)  was  one  of  the  earliest  to  make  a  study  of  the 
possible  fixation  of  atmospheric  nitrogen  by  B.  radicicola  under  these 
conditions.  He  found  that  a  small  quantity,  0.91  to  1.82  mgm.  of  nitro- 
gen was  fixed  per  100  c.  c.  of  the  medium,*  Prasmowski  (jp,  p.  55)  and 
Berthelot  (6)  concluded  as  a  result  of  their  experiments  that  when  the 
organism  was  grown  outside  the  host  plant  the  gain  in  nitrogen  was 
small.  The  greatest  gain  in  nitrogen  was  found  by  Maze  (52)  who  re- 
ported an  increase  of  23.4  mgm.  of  nitrogen  per  100  c.  c.  of  the  medium  in 
1 6  days.  lyewis  and  Nicholson  {30)  found  by  incubating  the  cultures  for 
a  considerable  length  of  time  that  a  large  increase  in  fixation  occurred. 
Bottomley  {8)  found  that  a  pure  culture  of  B.  radicicola  fixed  approxi- 
mately I  mgm.  of  nitrogen  in  15  days.  Fred  (77)  made  a  study  of  the 
possible  fixation  of  nitrogen  by  the  legume  organism  and  found  that  it 
fixed  approximately  1.2  mgm.  of  nitrogen  in  100  c.  c.  of  the  medium. 
He  found  that  on  agar  films  a  greater  fixation  occurred  than  when  the 
organisms  were  grown  in  a  liquid  medium. 

A  few  investigators,  however,  found  that  no  increase»r  in  nitrogen 
occurred  when  B.  radicicola  was  grown  outside  the  host  plant.  Frank 
{16)  states  that  in  a  nitrogen-free  medium  the  legume  organisms  did  not 
fix  enough  nitrogen  to  be  accurately  measured.  Immendorf  (25)  also 
found  no  increase  in  nitrogen  when  pure  cultures  of  B.  radicicola  were 
grown  in  soil  containing  a  nitrogen-free  solution. 

It  will  be  seen  that  the  majority  of  investigators,  especially  the  more 
recent  ones,  found  that  a  slight  amount  of  atmospheric  nitrogen  was 
fixed  or  assimilated  by  B.  radieicola  when  grown  outside  the  host  plant 
and  on  a  medium  suitable  for  its  development. 

It  has  already  been  shown  that  nitrates  cause  an  increase  in  the  num- 
ber of  B.  radicicola  when  grown  in  pure  culture  in  sterilized  soil.  Does 
such  an  increase  in  the  number  of  organisms  necessarily  mean  an  in- 
creased fixation  of  nitrogen?  Three  experiments  using  agar  films  were 
carried  out  in  order  to  determine  this  point.  Erlenmeyer  flasks  of 
I -liter  capacity  containing^  100  c.  c.  of  mannit  agar  were  used.  Before 
the  medium  solidified,  the  nitrates  were  added  in  the  proportions  indi- 
cated in  Table  XXVI I.  Six  flasks  for  each  different  quantity  of  nitrate 
were  prepared,  except  in  one  case,  as  shown  in  Experiment  II.  The 
flasks  were  plugged  with  nonabsorbent  cotton  and  sterilized  at  10  pounds' 
pressure  for  30  minutes.  After  cooling,  three  of  each  set  were  inoculated 
with  5  c.  c.  of  a  suspension  of  B.  radicicola  in  sterile  distilled  water. 
The  organisms  had  been  growing  on  mannit  agar  at  28°  C.  for  six  days. 
The  flasks  in  Experiments  I  and  HI  (Table  XXVII)  were  incubated  at 


Jan.  2S,  1918 


Nitrogen- A  ssimilating  Bacteria 


21 


28°  C.  for  three  weeks  and  those  in  Experiment  II  for  two  weeks.  The 
moisture  lost  by  evaporation  in  both  inoculated  and  uninoculated  flasks 
was  replaced  from  time  to  time  by  the  addition  of  sterile  distilled  water. 
At  the  expiration  of  the  incubation  period  the  total  nitrogen  was  deter- 
mined as  given  under  "Methods  used  in  experiments."  The  results  of 
the  experiments  are  given  in  Table  XXVII. 

An  inspection  of  the  data  reveals  the  fact  that  B.  radicicola  in  pure 
culture  fixed  a  small  amount  of  nitrogen  when  growing  in  a  nonsymbiotic 
state  with  no  nitrate  present.  In  the  presence  of  nitrates  there  was  an 
increased  fixation.  Although  the  increase  in  total  nitrogen  is  small, 
because  of  the  number  of  determinations  made,  it  may  be  considered 
as  positive.  The  potassium  and  sodium  salts  seemed  to  be  more  effective 
than  the  calcium  nitrate,  with  one  exception  (Table  XXVII,  Experiment 
I).  It  will  be  remembered  that  the  latter  salt  appeared  to  depress  nitro- 
gen fixation  by  Azotobacter  and  the  two  former  somewhat  to  favor  it  (p. 
194-195). 

Table  XXVII. — Influence  of  nitrates  on  the  fixation  of  nitrogen  by  Bacillus  radicicola, 
giving  the  increase  in  nitrogen 

EXPERIMENT  I 


Culture 
No. 


Treatment  (nitrate  in  100  c.  c.  of  medium.) 


None 

do 

do 

75  mgm.  of  NO3  as  sodium  nitrate  . .  . 

do 

do 

150  mgm.  of  NO3  as  sodium  nitrate .  . 

do 

do 

75  mgm.  of  NO3  as  calcium  nitrate  .  . 

do 

do 

150  mgm.  of  NO^as  calcium  nitrate  . . 

do 

do 


Total  nitrogen  in  100  c.  c.  of  medium. 

Uninoculated. 

Inoculated. 

Foimd. 

Average. 

Found. 

Average. 

Mgm. 

Mgm. 

Mgm. 

Mgm. 

4-5 

1      4.7 

4.4 

\  4-45 

4-6 

\  4.  60 

4.4 

I     4-5 

8.7 

f  II.  9 

8.7 

i  8.  70 

\    II.  8 

["•75 

8.6 

I  11.6 

12.  5 

r  14.9 

12.7 

>I2.  60 

14.6 

I  14-7 
f  12.3 

[14-  70 

8.8 

8.9 

[  8.90 

12.8 

>I2.  40 

9.0 

[  12.  I 

13-3 

f  14-5 

13-1 

|i3-  20 

^4-o 

[14.  10 

13.2 

[   13-8 

1 

Nitrogen 
increase. 


Mgm. 
o.  15 

3-03 

2.  10 

3-  30 
o.  90 


EXPERIMENT  11 


None 

do 

75  mgm.  of  NO3  as  sodium  nitrate  . 

do 

150  mgm.  of  NO3  as  sodium  nitrate. . 

do 

75  mgm.  of  NO3  as  calcium  nitrate   . 

do 

150  mgm.  of  NO3  as  calcium  nitrate . 
do 


4.90 
4.90 
8.  70 
8.50 

13-30 
13.00 
II.  IS 
II.  10 

14.  70 
(a) 


}5-07S 
I  9-  50 
}u.35 
}ii.65 

}i5-25 


o-  175 

0.  90 

1.  20 
o-  525 
o-  550 


a  Lost  by  breakage  during  sterilization. 


2l6 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  4 


Table  XXVII. — Influence  of  nitrates  on  the  fixation  of  nitrogen  by  Bacillus  radicicola, 
giving  the  increase  in  nitrogen — Continued 


EXPERIMENT  III 


Culture 

No. 


9 

lO 

II 

12 
13 
14 
15 
16 

17 
18 

19 
20 


Treatment  (nitrate  in  100  c.  c.  of  medium): 


None 

....do 

....do 

75  mgm.  of  NO3  as  potassium  ni- 
trate   

....do 

....do 

150  mgm.  of  NO3  as  potassium  ni- 
trate   

.  .  .  .do 

....do 

75  mgm.  of  NO3  as  sodium  nitrate 

....do 

....do 

150  mgm.  of  NO3  as  sodium  nitrate 

....do 

....do 

75  mgm.  of  NO3  as  calcium  nitrate 

....do 

....do 

150  mgm.  of  N03as  calcium  nitrate 

....do 

....do 


Total  nitrogen  in 

100  c.  c.  of  medium. 

Uninoculated. 

Inoculated. 

Found. 

Average. 

Found. 

Average. 

Mgm. 

Mgm. 

Mgm. 

Mgm. 

5.10 

5-50 

5-  10 

5-07 

j     5- 40 

5-50 

5.00 

I     5-45 

9-35 

f  10.85 

9-50 

9-37 

<   10.  90 

>    10.  90 

9-25 

I  10.95 

14.50 

f  15-65 

14.  20 

[14.28 

15-30 

[    15-45 

14-  15 

I  15-  40 

8.50 

9-85 

8.30 

8.38 

\     9- 90 

9-83 

«•  3.'5 

i     9-70 

12.35 

f  12.95 

12.  40 

[12.33 

j    13-  10 

[    13-03 

12.  20 

I   13-05 

8.95 

f     985 

9.  10 

\  9-  °i 

\     9- 90 

[     9-93 

9.  00 

I   10.05 

13.90 

f  14-40 

13.80 

13.80 

\    14-50 

[  14-  42 

13.70 

I   14-35 

Nitrogen 
increase. 


Mgm. 

0.  43 

1-53 

1.  17 

I- 45 
o.  70 
o.  92 
o.  62 


It  has  been  shown  that,  when  nitrates  are  added  in  varying  quantities 
to  sterilized  soil,  the  number  of  B.  radicicola  are  increased.  Provided  the 
the  organism  can  fix  a  small  amount  of  nitrogen  in  the  absence  of  nitrate 
nitrogen,  is  it  not  possible  that  this  increase  in  nitrogen  fixation  may  be 
due  merely  to  the  increase  in  the  number  of  cells  ?  It  seems  that  this  is 
true  according  to  the  results  in  Table  XXVII.  It  appears  probable  that 
the  increase  in  nitrogen  fixed  in  the  presence  of  nitrates  is  very  likely 
because  of  an  increase  in  the  number  of  bacterial  cells  and  not  to  any 
physiological  change  brought  about  in  the  organism  itself. 

There  was  a  marked  increase  in  bacterial  growth  on  the  media  con- 
taining the  nitrate  compared  with  the  same  media  free  from  nitrate. 
The  growth  on  the  latter  medium  exhibited  a  normal,  whitish  watery 
appearance,  characteristic  of  this  organism.  On  the  cultures  containing 
nitrates  a  much  more  profuse  growth  occurred.  In  many  instances  a 
pinkish  tint  was  observed.  This  pigment  production  was  especially 
marked  in  the  case  of  the  culture  containing  the  sodium  salt.  After  the 
first  experiment  had  been  completed,  it  was  thought  that  possibly  this 
pigmentation  was  due  to  an  impurity  in  the  culture.  Therefore  the  two 
remaining  experiments  were  made,  using  a  subculture  from  the  original. 


Jan.  28,  i9i8  Nitrogen- Assimilating  Bacteria  217 

This  culture  was  plated  three  times,  each  plating  being  made  from  a  well- 
isolated  colony.  The  final  subculture  was  taken  from  a  similar  well- 
isolated  colony.  However,  pigment  formation  in  the  presence  of  nitrate 
persisted  in  the  two  final  experiments,  showing  clearly  that  some  reaction 
took  place  between  the  nitrate  and  the  organism  grown  on  the  medium. 
It  is  of  interest  to  note  that  the  pigment  formation  in  the  presence  of 
nitrate  was  observed  in  later  work  where  the  influence  of  nitrates  on 
nodule  formation  was  investigated.  Prucha  (41)  found  that  on  agar 
slopes  of  medium  containing  0.5  per  cent  of  potassium  or  calcium 
nitrate,  the  growth  of  B.  radicicola  became  opaque  and  that  an  iridescent 
tint  was  produced. 

Although  the  results  of  these  experiments  may  vary  somewhat  among 
themselves,  taken  as  a  whole  it  appears  evident  that  B.  radicicola  may 
fix  a  small  amount  of  atmospheric  nitrogen  when  grown  without  the 
host  plant  and  on  a  suitable  medium.  The  addition  of  various  amounts 
of  nitrates  as  indicated  increased  somewhat  the  amount  of  nitrogen 
assimilated  by  B.  radicicola. 

INFIvUENCE  OP  NITRATES  ON  THE  PRODUCTION  OP  GUM  BY  BACILLUS  RADICICOLA 

Since  nitrates,  especially  in  smaller  amounts,  cause  an  increase  in  the 
number  of  B.  radicicola  in  pure  culture,  it  was  thought  advisable  to 
determine  what  influence  these  salts  have  on  the  production  of  gum.  In 
culture  media  favorable  to  the  growth  of  B.  radicicola  these  bacteria  will 
produce  a  gelatinous  substance  which  is  readily  precipitated  with  95  per 
cent  alcohol  or  acetone.  Upon  the  addition  of  either  of  these  precipi- 
tants  a  fairly  heavy,  water-white,  frothy  gelatinous  mass  is  formed 
which  soon  rises  to  the  surface  of  the  liquid.  Upon  standing,  this  mass 
contracts  somewhat,  and  portions  of  it  may  fall  to  the  bottom  of  the 
liquid  from  which  it  has  been  precipitated. 

Chemical  analyses,  according  to  Buchanan  (jo),  have  shown  that  this 
gum  is  a  carbohydrate.  Upon  hydrolysis  with  2  per  cent  sulphuric  acid 
and  15  pounds'  pressure  for  one  hour,  Fehling's  solution  is  reduced, 
showing  the  presence  of  a  sugar.  The  gum  does  not  give  proteid  reac- 
tions with  the  Millon,  biuret,  or  xanthoproteic  tests.  Hence,  the  gum 
is  not  protein  in  character;  nor  does  it  contain  nitrogen  in  the  combined 
form.     Clearly  it  is  a  nonnitrogenous  body. 

In  the  experiment  undertaken  to  determine  whether  nitrates  influence 
the  fonnation  of  gum  only  relative  dififerences  are  noted.  No  attempt 
was  made  to  obtain  quantitative  results. 

Erlenmeyer  flasks  of  i -liter  capacity  containing  200  c.  c.  of  mannit 
solution  were  used.  The  cultures  contained  various  quantities  of  nitrate 
as  indicated  in  Table  XXVIII.  Triplicate  flasks  for  each  amount  of 
nitrate  were  prepared.  In  this  table  these  three  flasks  are  represented 
as  "a,"  "b,"  and  "c."  After  sterilization  at  15  pounds'  pressure  for  25 
minutes  the  flasks  were  cooled  and  inoculated  with  5  c.  c.  of  a  suspension 


2l8 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  4 


of  B.   radicicola  in  sterile  distilled    water.     The    cultures    were    then 
incubated  at  room  temperature  (approximately  25°  C.)  for  eight  weeks.^ 

At  the  expiration  of  the  incubation  period  the  contents  of  the  flasks 
were  poured  into  hydrometer  cylinders  of  equal  depth  and  diameter. 
One  hundred  and  fifty  c.  c.  of  acetone  were  added  to  precipitate  the  gum. 
After  careful  shaking,  the  cylinders  were  covered  with  inverted  petri 
dishes  to  prevent  evaporation.  At  the  end  of  24  hours  the  amount  ol 
gum  precipitated  was  observed.  The  relative  amounts  are  recorded  in 
Table  XXVIII. 

Table  XXVIII. — Influence  of  nitrates  on  the  production  of  gum  by  Bacillus  radicicola 


Cul- 
ture 

Treatment  (nitrate  in  loo  c.  c.  of  medium). 

Relative  production  of  gtun — precipitated  by 
acetone. 

No. 

Flask  a. 

Flask  b. 

Flask  c. 

None 

Large 

Very  large . 

Large 

Very  large . 
Large 

Large 

Very  large . 

Large 

...do 

...do 

Large. 
Very  large. 
Large. 
Very  large. 
Large. 
Do. 

2 

3 
4 

5 
6 

75  mgm.  of  NO3  as  potassium  nitrate  . 
450  mgm.  of  NO3  as  potassium  nitrate . 
75  mgm.  of  NO3  as  sodium  nitrate  .... 
450  mgm.  of  NO3  as  sodium  nitrate. .  . 
75  mgm.  of  NO3  as  calcium  nitrate. .  . 

. .  .do 

...do 

7 

450  mgm.  of  NO3  as  calcium  nitrate. . 

...do 

C  0  nsider- 
able. 

Con  sider- 

able. 

From  the  results  it  is  certain  that  the  nitrates,  especially  in  the  smaller 
of  the  two  concentrates,  caused  a  very  considerable  increase  in  the  amount 
of  gum  produced  by  B.  radicicola.  The  nitrates  of  potassium  and  sodium 
caused  a  production  of  more  gum  than  did  the  calcium  salt.  It  will  be 
remembered  that  in  the  experiments  v^rhere  the  influence  of  nitrates  on 
the  fixation  of  atmospheric  nitrogen  by  B.  radicicola  was  studied,  less 
nitrogen  was  fixed  in  the  presence  of  calcium  nitrate  than  in  the  pres- 
ence of  the  other  two  salts.  Here  again  the  greater  stimulative  action 
of  potassium  and  sodium  nitrates  is  emphasized. 

Buchanan  in  his  investigations  on  the  formation  of  gum  by  B.  radici- 
cola has  found  that  varying  amounts  of  potassium  nitrate  in  a  2  per 
cent  saccharose  solution  or  in  a  2  per  cent  saccharose-clover-extract 
solution  caused  a  slight  increase  in  growth  and  in  gum  production. 

It  seems  probable  that  the  increased  gum  production  in  the  nitrate 
cultures  is  caused  not  only  by  an  increase  in  bacterial  cells  but  also 
perhaps  by  an  increased  stimulation  in  the  formation  of  gum  by  the  cells 
themselves.  The  relative  increase  in  the  amount  of  gum  produced  in 
the  presence  of  nitrates  seems  to  be  greater  than  the  actual  increase  in 
number  of  organisms  brought  about  by  the  stimulating  effect  of  the 
nitrate.  In  the  latter  instance  this  stimulating  effect  has  been  deter- 
mined in  soil  cultures  only  and  so  a  fair  basis  of  comparison  can  not  be 


Jan.  28, 1918  Nitrogen- Assimilating  Bacteria  219 

found.  Had  the  influence  of  nitrates  on  the  growth  and  reproduction 
of  B.  radicicola  been  determined  in  liquid  culture,  as  well  as  in  soil 
cultures,  then  a  comparison  could  have  been  made.  Furthermore,  the 
divergencies  in  the  time  element,  eight  weeks'  incubation  in  the  liquid 
cultures  and  three  weeks  in  the  soil  cultures,  are  such  as  to  render  futile 
any  attempt  at  correlation.  It  may  be  that  the  large  formation  of  gum 
was  due  to  the  prolonged  incubation.  A  shorter  period  of  three  weeks 
undoubtedly  would  show  a  relatively  smaller  amount  of  gum  produced 
as  a  result  of  the  presence  of  the  nitrate. 

However,  from  the  results  of  the  experiment  it  is  certain  that  potas- 
sium, sodium,  and -calcium  nitrate  influence  the  formation  of  gum  by  B. 
radicicola.  The  three  nitrates  studied  caused  a  large  increase  in  the 
amount  of  gum  obtained  by  precipitation  with  acetone.  Calcium  nitrate 
caused  the  least  stimulation,  but  the  difference  was  not  large. 

INFLUENCE   OF  NITRATES  ON   NODULE  FORMATION 

The  results  of  numerous  investigations  have  shown  that  nitrates 
retard  and  oftentimes  entirely  prevent  the  formation  of  nodules  on 
leguminous  plants  when  grown  in  soil  or  liquid  cultures.  Vines  (45), 
working  with  the  horse  bean,  found  that  the  use  of  large  amounts  of 
nitrate  in  the  form  of  potassium  nitrate  retarded  nodule  formation.  He 
concluded  that  a  decrease  in  the  amount  of  nitrates  meant  an  increase 
in  the  number  of  nodules.  Woods  (48)  found  that  leguminous  plants 
assimilated  more  nitrogen  when  they  were  grown  in  the  absence  of 
potassium  and  calcium  nitrate  than  when  thus  supplied.  His  results 
seem  to  indicate  that  nodule  development  was  retarded  somewhat  by 
these  salts.  Similar  results  were  obtained  by  Frank  (16).  Nobbe  and 
Richter  (37)  in  1 902  grew  soybeans  in  a  rich  garden  soil  and  found  upon 
inoculation  that  a  gain  of  74.7  per  cent  of  nitrogen  occurred.  However, 
upon  the  addition  of  nitrates  this  gain  was  considerably  reduced,  the  extent 
of  the  reduction  corresponding  to  the  amount  of  nitrate  added.  About 
this  same  time,  Wohltmann  and  Bergen6  (47)  using  many  different 
types  of  soils,  found  that  nodules  were  not  formed  on  the  roots  of  peas 
when  ammonium  nitrate  was  added.  Creydt  (12)  found  that  sodium 
nitrate  retarded  nodule  formation  on  yellow  lupines  when  these  legumes 
were  grown  in  soil.  Fred  and  Graul  (j^")  found  that  very  small  amounts 
of  nitrates  did  not  appreciably  decrease  nodule  formation,  but  that 
larger  amounts  proved  detrimental  and  finally  prohibited  entirely  the 
development  of  nodules. 

The  presence  of  nitrates  in  culture  solutions  has  also  been  found  to 
reduce  and  oftentimes  to  inhibit  the  formation  of  nodules  on  leguminous 
plants.  Marchal  (31)  concluded  that  alkaline  nitrates  in  concentrations 
of  I  to  10,000  in  liquid  cultures  prevented  the  formation  of  nodules  on 
peas.     Flamand  (13)  grew  vetch  and  beans  in  a  nutrient  solution  and 


220  Journal  of  Agricultural  Research  voi.  xii.  no.  4 

found  that  nitrates  in  the  following  amounts  prevented  nodule  forma- 
tions: potassium  nitrate,  i  to  10,000,  sodium  nitrate  i  to  2,000,  ammo- 
nium nitrate  i  to  2,000,  and  calcium  nitrate  i  to  2,000  and  i  to  10.000. 
Hiltner's  {24)  experiments  showed  that  5  mgm.  of  nitrogen  as  potassium 
nitrate  per  liter  prevented  nodule  formation  on  peas. 

In  contrast  to  these  experiments  Bassler  (i)  claimed  that  results 
obtained  from  his  work  indicated  that  no  effect  was  noticed  by  adding 
nitrates  to  lupines  growing  in  quartz  sand. 

The  question  naturally  arises  whether  this  condition  is  due  to  the 
weakening  of  the  organism  brought  about  by  growth  in  a  nitrated 
environment  and  to  a  consequent  impairment  or  entire  loss  of  its  infect- 
ing power,  or  whether  it  is  caused  by  some  interreaction  between  the 
salt  and  the  plant  root,  tending  to  increase  the  latter's  resistance  to 
the  attack  of  this  particular  organism. 

INFLUENCE  OP  NITRATES   ON   THE   INFECTING  POWER  OP   BACILLUS   RADICICOLA 

Some  investigations  have  been  carried  out  to  determine  what  effect 
nitrates  have  on  the  legume  organisms  themselves.  Wilson  {46)  showed 
that  although  nitrates  inhibit  the  formation  of  nodules,  the  organisms 
capable  of  producing  nodules  did  not  lose  their  vitality  or  nodule-pro- 
ducing power  when  grown  in  the  presence  of  nitrates.  The  results  of 
Prucha  {41)  are  in  accord  with  those  of  Wilson.  He  found  that  B.  radi- 
cicola  does  not  seem  to  lose  its  infecting  power  when  grown  on  media 
containing  nitrate.  During  the  course  of  his  work  he  found  that  potas- 
sium and  sodium  nitrates  inhibited  the  formation  of  nodules.  Further 
evidence  that  the  organisms  appear  to  retain  their  vitality  in  the  pres- 
ence of  nitrates  is  produced  by  the  results  of  Maze  (jj,  p.  ly-i"/),  who 
showed  that  legume  bacteria  were  able  to  fix  a  slight  amount  of  nitrogen 
when  grown  in  a  soil  extract  solution  containing  i  per  cent  sodium  nitrate. 
Herke  (22)  states  that  potassium  nitrate  favors  the  growth  of  nodule 
bacteria. 

However,  other  investigators  state  that  nitrates  have  a  harmful 
effect  on  B.  radicicola.  Laurent  (29,  p.  134)  found  that  legume  organ- 
isms failed  to  grow  in  a  pea  or  lupine  decoction  containing  nitrate  in  the 
form  of  potassium  and  sodium  salts  in  amounts  equivalent  to  i  to  500 
and  I  to  1,000.  Moore  {35)  in  his  experiments  demonstrated  that  nitrates 
at  I  to  10,000  were  sufficient  to  prevent  nodule  formation.  He  states 
that  B.  radicicola  loses  its  power  of  infection  when  grown  in  a  medium 
containing  nitrates. 

From  the  results  cited  it  can  be  seen  that  there  is  some  disagreement 
as  to  the  influence  exerted  by  nitrates  on  B.  radilcicola.  In  some  cases 
the  organism  seems  to  retain  its  vitality  in  the  presence  of  nitrates, 
while  in  others  it  appears  to  have  become  weakened.     It  must  be  ad- 


Jan.  28,  191S 


Nitroge  n-A  ssimilating  Bacteria 


221 


mitted,  however,  that  the  evidence  seems  to  favor  the  former  contention — 
namel)?-,  that  nitrates  do  not  cause  the  bacteria  to  lose  their  nodule- 
producing  power. 

In  order  to  determine  whether  or  not  nitrates  weaken  these  organisms, 
the  following  experiments  were  made:  Slopes  of  mannit  agar  (in  test 
tubes)  containing  various  amountsof  sodium  and  calcium  nitrates  as  indi- 
cated in  Table  XXIX  were  inoculated  with  B.  radicicola.  These  cul- 
tures were  incubated  at  28°  C.  for  one  week,  when  transfers  were  made 
to  fresh  nitrate  media  and  incubated  at  28°  C.  for  another  week.  At 
the  expiration  of  this  time,  three  4-day-old  seedlings  of  alfalfa  were 
inoculated  with  three  drops  of  a  suspension  of  the  organism  in  5  c.  c. 
of  sterile  distilled  water.  The  same  slope  cultures  were  incubated  at 
28°  C.  and  used  for  all  subsequent  inoculations  in  this  experiment. 
The  inoculated  seedlings  were  placed  in  the  greenhouse  under  cheese- 
cloth covering.  The  temperature  here  during  the  daytime  averaged 
approximately  30°  C.  The  seedlings  were  examined  for  the  first  appear- 
ance of  nodules  and  in  no  case  did  they  appear  before  18  to  20  days. 
A  total  count  of  nodules  on  all  plants  was  made  at  the  end  of  45  days. 
Three  subsequent  inoculations  were  made  under  the  same  conditions. 
In  this  way  organisms  in  contact  with  nitrate  for  varying  lengths  of 
time  could  be  used.  The  results  of  the  inoculation  experiments  are 
given  in  Table  XXIX. 

Table  XXIX. — Influence  of  nitrates  on  the  infecting  power  of  Bacillus  radicicola 


Cul- 

Treatment (nitrate  in  loo  c.  c.  of  medium). 

Number  of  nodules  after  45  days. 

ture 
No. 

Inoculated    Inoculated 
June  3.         June  15. 

Inoculated 
July  II. 

Inoculated 
July  17. 

I 

None 

7 
5 

IS 
3 
4 

5 
6 

7 

9 

8 

II 

9 
4 
6 

3 
5 
5 

II 
8 
7 
3 
8 

4 
6 
8 
9 
3 
3 

2 

3 
4 
5 
6 

7 
8 

15  mgm.  of  NO3  as  sodium  nitrate .  .  . 
37  mgm.  of  NO3  as  sodium  nitrate  .  . . 
75  mgm.  of  NO3  as  sodium  nitrate  .  . . 
150  mgm.  of  NO3  as  sodium  nitrate  .  . 
225  mgm.  of  NO3  as  sodium  nitrate . . 
450  mgm.  of  NO3  as  sodium  nitrate  . . 
None 

3 
5 

7 
4 
5 
6 
2 
7 
4 
9 
5 
7 
5 
6 
0 
0 

4 
6 

5 
8 

5 
4 
8 

4 
4 
5 

\ 

8 

9 
10 
II 
12 

13 
14 

15 
16 

15  mgm.  of  NO3  as  calcium  nitrate .  . 
37  mgm.  of  NO3  as  calcium  nitrate. . . 
75  mgm.  of  NO3  as  calcium  nitrate  .  . 
150  mgm.  of  NO3  as  calcium  nitrate.  . 
225  mgm.  of  NO3  as  calcium  nitrate  . 
450  mgm.  of  NO3  as  calcium  nitrate . 
Uninoculated 

do 

From  the  results  given  in  Table  XXIX  it  is  very  evident  that  under 
the  conditions  of  the  experiment  the  legume  organisms  did  not  lose 
their  power  of  producing  nodules  when  grown  on  a  medium  containing 


222  Journal  of  Agricultural  Research  voi.  xii.  no.  4 

varying  amounts  of  sodium  and  calcium  nitrates.  The  numbers  of 
nodules  produced  on  the  alfalfa  plants  by  organisms  grown  on  media 
containing  nitrate  do  not  vary  widely  from  those  on  the  plants  inocu- 
lated v»^ith  organisms  grown  on  media  containing  no  nitrate.  Not  only 
did  the  organisms  fail  to  lose  their  nodule-producing  power,  but  from 
all  appearances  their  infecting  power  did  not  seem  to  be  materially 
weakened. 

It  therefore  seems  apparent  that  an  explanation  for  the  failure  of 
nodules  to  develop  on  leguminous  plants  in  the  presence  of  nitrates  is 
not  found  in  the  theory  that  the  organisms  producing  these  nodules  are 
weakened  when  grown  in  the  presence  of  nitrates. 

INPLUBNCB  OP  NITRATES  ON  ALFALFA  ROOTS  AND  NODULE  FORMATION 

The  next  step  taken  would  naturally  be  in  the  direction  of  a  study  of 
the  influence  of  the  nitrates  on  the  plant  roots  themselves  in  order  to 
determine  whether  or  not  they  thus  are  made  more  resistant  to  the 
attack  of  these  organisms. 

A  review  of  the  literature  shows  that  almost  nothing  has  been  done 
touching  this  phase  of  the  question.  Wilson  {46),  studying  the  effect  of 
certain  salts  on  nodule  production,  states  that  possibly  the  salt  has  some 
effect  on  the  root,  making  it  resistant  to  the  attack  of  the  organism. 
Maze  (55,  p.  15-17),  who  also  concluded  that  nitrates  did  not  cause  B. 
radicicola  to  lose  its  infecting  power,  says  that  nodules  do  not  develop 
on  roots  of  legumes  when  nitrates  are  present  because  the  carbohydrate 
in  the  roots  is  changed  into  protein  material  by  the  absorption  of  the 
nitrate. 

Alfalfa  seedlings  (Medicago  sativa)  growing  in  soft  agar  containing 
potassium,  sodium,  and  calcium  nitrates,  as  indicated  in  Table  XXX, 
were  used  in  this  study.  Quadruplicate  tubes  were  prepared  for  each 
amount  of  nitrate.  The  higher  concentrations  of  the  nitrate  were  not 
used,  since  it  was  found  that  germination  and  subsequent  growth  were 
considerably  impaired  in  the  presence  of  such  large  amounts.  The  tubes 
with  the  mannit  agar  and  nitrate  were  sterilized  at  1 5  pounds'  pressure 
for  30  minutes.  These  were  cooled  and  sterilized  alfalfa  seeds  planted 
as  given  under  "Methods  used  in  experiments."  The  tubes  were 
then  placed  in  the  greenhouse  under  cheesecloth  covering  and  the  seeds 
allowed  to  germinate.  Germination  took  place  in  all  instances,  although 
it  was  retarded  somewhat  by  the  presence  of  the  nitrate.  At  the  end  of 
five  days  the  first  tube  of  each  set  was  inoculated  with  three  drops  of  a 
suspension  of  JB.  radicicola  in  sterile  distilled  water.  Subsequent  inocu- 
lations were  made  as  indicated  in  Table  XXX.  These  were  made  at 
different  intervals  in  order  to  allow  the  roots  of  the  seedlings  to  remain 
for  a  longer  time  in  contact  with  the  media.  It  was  hoped  that  in  this 
way  an  idea  might  be  obtained  as  to  the  time  when  the  root  first  became 
resistant.     The  results  are  given  in  Table  XXX. 


Jan.  2S,  1918 


Nitrogen- A  ssimilating  Bacteria 


223 


Tabl,E  XXX. — Influence  of  nitrates  on  alfalfa  roots  and  nodule  formation 


Culture 

Treatment  (nitrate  in  100  c.  c.  of  medium). 

Total  number  of  nodules  in  each  tube  of 
seedlings  inoculated  after — 

No. 

S  days' 
growth. 

10  days' 
growth. 

18  days' 
growth. 

22  days' 
growth. 

I 

None 

3 
0 
0 
0 
0 
0 
0 
0 
0 

(^) 

(^) 
I 
0 
0 
0 

3 
I 
0 
0 
0 
0 
I 
0 
0 
0 
0 

3 
0 
0 

0 

5 
2 
0 
0 
0 

3 
0 
0 
0 
0 
I 
0 
0 
0 
0 

4 

2 

3 
4 
5 
6 

7 
8 

9 
10 
II 
12 
13 
14 
15 
16 

10  mgm.  of  NO3  as  potassium  nitrate 

25  mgm.  of  NO3  as  potassium  nitrate 

50  mgm.  of  NO3  as  potassium  nitrate 

100  mgm.  of  NO3  as  potassium  nitrate  .... 
150  mgm.  of  NO3  as  potassium  nitrate  .  . . . 

10  mgm.  of  NO3  as  sodium  nitrate 

25  mgm.  of  NO3  as  sodium  nitrate 

50  mgm.  of  NO3  as  sodium  nitrate 

100  mgm.  of  NO3  as  sodium  nitrate 

150  mgm.  of  NO3  as  sodium  nitrate 

10  mgm.  of  NO3  as  calcium  nitrate 

25  mgm.  of  NO3  as  calcium  nitrate 

50  mgm.  of  NO3  as  calcium  nitrate 

100  mgm.  of  NO3  as  calcium  nitrate 

150  mgm.  of  NO3  as  calcium  nitrate 

0 
0 

0 
0 

0 
0 
0 
0 
0 
0 
0 
0 
0 

a  Fungus  contamination. 


t>  Plant  died  after  few  days'  growth. 


It  will  be  seen  that  in  a  few  instances  where  a  high  concentration  of 
nitrates  occurred  the  development  of  the  seedlings  subsequent  to  ger- 
mination ceased.  This  condition  may  have  been  due  to  too  high  a  con- 
centration of  soluble  salts  or  to  inferior  seed.  However,  losses  were  not 
sufficiently  serious  materially  to  affect  the  outcome  of  the  experiment. 

In  all  cases  the  seedlings  grown  in  agar  without  nitrate  produced 
nodules  when  inoculated  with  B.  radicicola.  A  few  nodules  appeared  on 
seedlings  in  cultures  containing  the  loAvest  concentration  of  all  three 
nitrates.  The  number  of  nodules  in  these  cases  was  less  than  in  the 
control  cultures.  No  nodules  whatever  developed  in  any  concentration 
above  lo  mgm.  of  nitrate  in  loo  c.  c.  of  medium.  Under  normal  condi- 
tions in  test-tube  cultures  the  nodules  make  their  appearance  at  about 
18  to  20  days  after  inoculation.  The  incubation  of  all  cultures  was 
extended  40  days  after  inoculation  in  order  to  make  certain  that  no 
further  nodule  development  would  take  place. 

The  nonproduction  of  nodules  was  not  due  to  the  failure  of  the  inoc- 
ulum. In  all  cases  an  excellent  inoculum  growth  was  obtained,  espe- 
cially in  the  case  where  nitrate  was  present  in  the  medium.  Indeed,  it 
was  so  luxuriant  that  in  many  cases  the  organism  grew  in  considerable 
quantity  far  down  into  the  root  zone.  In  many  cases  where  nitrates 
were  present  the  same  pink  coloration  was  produced  that  was  discussed 
under  another  caption,  on  page  216. 

As  has  been  already  stated,  seedlings  of  varying  ages  were  inoculated 
for  the  reason  that  it  was  thought  that  a  more  or  less  prolonged  contact 
of  the  roots  with  the  nitrate  in  the  medium  might  serve  as  an  index  to 
27807°— 18 6 


224  Journal  of  Agricultural  Research  voi.  xii,  no.  4 

the  time  in  the  growth  of  the  seedling  when  permanent  resistance  to 
"'attack  of  the  organisms  was  established.  The  results  obtained  do  not 
■  seem  to  indicate  that  seedling  roots  18  to  20  days'  old  are  any  more 
resistant  to  the  attack  of  the  organisms  than  are  those  that  are  younger. 
Evidently  if  any  reaction  takes  place  between  the  nitrate  and  the  plant 
root  it  occurs  during  the  very  early  stages  in  the  development  of  the 
plant. 

These  results  seem  to  point  to  the  conclusions  that  the  nonformation 
of  nodules  in  the  presence  of  nitrates  is  due  not  to  a  weakening  of  the 
vitality  of  the  organism,  but  to  some  reaction  between  the  plant  root 
and  nitrate.  One  naturally  queries  whether  the  plant  root  cells  are 
made  more  resistant  to  the  bacteria  seeking  to  gain  entrance  there  or 
whether  the  naturally  occurring  carbohydrate  food  supply  to  be  used 
by  the  organisms  after  gaining  entrance  is  diminished  by  its  conversion 
into  protein  material  by  the  absorption  of  nitrate?  Further  studies 
were  not  made  in  an  endeavor  to  solve  this  question. 

INFLUENCE   OF   NITRATES   IN   SOIL   ON    ALFALFA   NODULES    AND   ON   THE   REFORMATION 

OF    NODULES 

Additional  studies  were  made  with  nitrates  in  relation  to  their  influ- 
ence on  nodules  already  formed  and  on  the  redevelopment  of  nodules 
once  removed  from  alfalfa  plants.  The  experiments  were  carried  out 
in  an  endeavor  to  determine  whether  nitrates  prevented  an  increase  in 
the  number  of  nodules  on  plants  possessing  nodules  and  whether  they 
prevented  the  reformation  of  nodules  when  removed.  Experiments 
revealed  clearly  that  removed  nodules  were  replaced  by  new  ones  pro- 
vided the  plant  was  carefully  replaced  in  the  soil  (soil  with  normal  low 
nitrate  content)  and  the  proper  amount  of  moisture  maintained. 

In  these  experiments  i -gallon  earthenware  jars  were  used.  These 
were  filled  to  within  an  inch  of  the  top  with  1,800  gm.  of  soil  of  a  low 
nitrate  content.  Different  amounts  of  the  nitrates  to  be  studied  were 
added  in  the  quantities  indicated  in  Table  XXXI.  Concentrations  of 
100  and  300  mgm.  of  nitrate  in  100  gm.  of  soil  were  also  used,  but  the 
transplanted  alfalfa  seedlings  were  unable  to  withstand  such  excessive 
concentration,  with  the  result  that  all  died  within  a  week  or  ten  days 
after  transplanting.  Quadruplicate  pots  were  prepared  for  each  con- 
centration of  nitrate.  The  nitrates  in  solution  were  mixed  with  the 
proper  amount  of  distilled  water  which,  when  added  to  the  pots,  brought 
the  moisture  content  to  approximately  20  per  cent.  The  pots  were  then 
allowed  to  remain  undisturbed  for  one  day  at  room  temperature  to 
allow  the  water  containing  the  nitrate  to  become  well  diffused  through- 
out the  soil  mass.  Into  two  pots  of  each  set  were  transplanted  young 
alfalfa  plants  from  which  the  nodules  had  been  removed.  The  two 
remaining  pots  contained  transplanted  alfalfa  plants  with  the  nodules 
left  on  and  their  location  noted.     The  plants  used  in  this  experiment 


Jan.  28,  igif 


Nitrogen- A  ssimilating  Bacteria 


225 


were  removed  from 'an  alfalfa  plot,  the  soil  of  which  was  a  sandy  loam. 
Previous  to  transplanting  the  roots  of  the  young  plants  were  carefully 
washed  in  running  water  and  immediately  transplanted.  The  pots 
were  kept  well  watered,  and  after  two  or  three  days  they  were  removed 
to  the  greenhouse.  Here  they  were  watered  when  necessary.  Trans- 
plantations were  made  on  the  27th  of  June  and  the  experiment  termi- 
nated on  the  3d  of  August.  The  plants  were  removed  from  the  pots, 
the  roots  carefully  washed  and  examined  for  the  presence  of  nodules. 
The  results  are  presented  in  Table  XXXI. 

Table  XXXI. — Influence  of  nitrates  in  soil  on  alfalfa  nodules  and  on  the  reformation  of 

nodules 


Nitrate  in  loo  gm.  of  dry  soil. 

Treatment  of 
nodules. 

Number  of 
nodules — 

Pot  No. 

At 
begin- 
ning. 

At 
end. 

A  I 

None   ....       

Removed 

do 

0 
0 
4 
3 
0 
0 

4 
8 
0 
0 

4 
I 
0 
0 
4 

5 
0 
0 
2 
6 
0 
0 

4 
0 
0 

4 
2 

3 

A 

A  2 

.do      

A  ^ 

do 

Not  removed . .  . 
do 

8 

A4 

do 

7 
0 

(a) 

3 
5 
0 
0 

B  I 

B  2 

25  mgm.  of  NO3  as  potassium  nitrate  .... 
.    . . .do 

Removed 

do 

B  7. 

.  do    

Not  removed . . . 
do 

B  4 

do 

Ci 

C2 

50  mgm.  of  NO3  as  potassium  nitrate 

do 

Removed 

do 

C  z 

..do 

Not  removed . . . 
do 

2 

ci 

do 

I 

D  I 

D  2  . 

25  mgm.  of  NO3  as  sodium  nitrate 

do 

Removed 

do 

0 
0 

D  ^  . 

....do 

Not  removed .  .  . 
do 

2 

D  4 

.do 

I 

r:  ^ 

E  I 

E2 

50  mgm.  of  NO3  as  sodium  nitrateft 

do 

Removed 

do 

0 

(a) 

I 

E^ 

do 

Not  removed . . . 
do 

eI:.  . . 

do 

3 
0 
0 

Fi 

F  2 

25  mgm.  of  NO3  as  calcium  nitrate 

do 

Removed 

do 

F  7 

do 

Not  removed . . . 
do 

2 

F4 

....do 

3 
0 
0 

G  I 

G  2 

50  mgm.  of  NO3  as  calcium  nitrate 

.    .   do 

Removed 

do 

G   T. 

.    .   do 

Not  removed . . . 
do 

7. 

G  4 

.do 

I 

"  Plants  died. 

It  will  be  seen  in  the  control  pots,  where  no  nitrate  was  present  (ex- 
cept the  small  amount  normally  present  in  the  soil  at  the  beginning  of 
the  experiment),  that  if  the  nodules  were  removed,  new  ones  formed. 
The  location  of  the  nodules  before  their  removal  was  noted,  and  the 
new  ones  were  found  to  occupy  the  same  place.  However,  when  nitrates 
were  added  to  the  soil  no  new  nodules  were  formed.  This  statement 
holds  true  for  both  concentrations  of  all  three  salts  in  all  experiments. 


226  Journal  of  Agricultural  Research  voi.  xii.  no.  4 

Some  interesting  results  were  obtained  where  the  nodules  were  not 
removed.  In  the  control  pots  an  increase  in  nodule  formation  took  place. 
It  can  not  be  stated  definitely  whether  the  new  nodules  appeared  as 
a  result  of  inoculation  from  the  soil  or  whether  the  organisms  had  already 
gained  entrance  to  the  roots  before  the  plants  were  removed  from  the 
field  soil  previous  to  transplanting.  Nevertheless,  it  is  shown  that  the 
number  of  nodules  increased  as  compared  with  the  number  present  at 
the  time  of  transplanting.  But  where  nitrates  were  added  a  reduction 
in  number  occurred  rather  regularly  throughout  all  the  pots.  In  two 
instances  the  number  remained  constant,  in  10  it  was  reduced,  and  in  none 
was  it  increased.  The  calcium  salt  appeared  to  effect  the  least  reduction 
in  number  of  nodules.  Conclusions  concerning  the  comparative  in- 
fluence of  the  three  salts  in  this  regard  can  not  be  drawn  because  of  the 
small  number  of  determinations  made.  It  is  sufficient  to  note  that 
nitrates  present  in  amounts  equal  to  25  and  50  mgm.  of  nitrate  in  100 
gm.  of  soil  did  not  permit  an  increase  in  number  of  nodules,  but  rather 
caused  a  decrease. 

The  conclusions  drawn  from  the  experiments  relative  to  the  influence 
of  nitrates  on  nodule  formation  are:  (a)  the  presence  of  nitrates  is 
detrimental  to  the  formation  of  nodules  by  alfalfa;  (b)  the  nonformation 
of  nodules  is  not  due  to  a  weakening  of  B.  radicicola  yvhen  grov/n  in  the 
presence  of  nitrates;  (c)  some  reaction  takes  place  between  the  nitrates 
and  the  plant  root,  thus  preventing  nodule  formation;  (d)  nitrates  in  the 
soil  prevent  the  re-formation  of  nodules  once  removed  and  also  cause 
a  decrease  in  the  number  of  those  already  present. 

SUMMARY 

(i)  Small  quantities  of  potassium,  sodium,  and  calcium  nitrates 
caused  a  great  increase  in  the  number  of  Azotobacter  in  sterilized  soil. 
Ammonium  nitrate  in  the  same  quantities  caused  a  less  marked  in- 
crease. Higher  concentrations  were  not  so  favorable  to  the  growth 
of  the  organisms. 

(2)  Potassium  and  sodium  nitrates  in  the  concentrationj  studied 
caused  an  increase  in  the  amount  of  nitrogen  assimilated  by  Azoto- 
bacter on  agar  films.  Calcium  nitrate  in  the  same  amounts  brought 
about  a  decrease  in  the  amount  of  nitrogen  fixed  to  a  point  even  below 

'  that  representing  the  amount  assimilated  in  the  absence  of  nitrates.  In 
soil  cultures  nitrates  of  sodium  and  calcium  caused  an  increase  in  total  ni- 
trogen, which  was  more  marked  in  the  unsterilized  cultures  than  in  those 
cultures  sterilized  and  inoculated  with  a  pure  culture  of  Azotobacter. 
However,  the  increase  in  total  nitrogen  is  not  commensurate  with  the 
increase  in  the  number  of  Azotobacter  noted  under  the  same  conditions. 

(3)  Under  aerobic  conditions  Azotobacter  in  liquid  cultures  reduced 
nitrate  to  nitrite,  but  not  to  ammonia.  More  atmospheric  nitrogen  was 
assimilated  in  the  presence  of  nitrate  than  in  the  absence  of  this  salt. 


Jan.  2S,  igis  Nitroge7i- Assimilating  Bacteria  227 

(4)  Pigmentation  occurred  when  potassium  and  sodium  nitrates, 
and  especially  calcium  nitrate,  were  used  with  Azotobacter,  the  colora- 
tion increasing  with  the  concentration  of  the  salt.  This  effect  was  more 
marked  in  Azotobacter  strains  which  produce  little  or  no  pigment  in  the 
absence  of  nitrates. 

(5)  All  three  nitrates  studied  caused  an  increase  in  the  number  and 
size  of  volutin  bodies  in  Azotobacter  cells.  From  all  appearances 
these  salts  also  tended  to  hasten  the  development  of  these  bodies. 

(6)  The  number  of  Bacillus  radicicola  in  sterilized  soil  was  increased 
by  the  addition  of  small  quantities  of  potassium,  sodium,  ammonium, 
and  calcium  nitrates.  This  increase  was  not  so  marked  as  in  the  Azoto- 
bacter cultures.  B.  radicicola  appeared  to  be  much  more  resistant  to 
higher  concentrations  of  nitrates  than  Azotobacter. 

(7)  B.  radicicola  under  aerobic  conditions  did  not  reduce  nitrates 
in  solution  to  nitrite,  ammonia,  or  elemental  nitrogen.  The  presence 
of  nitrates  did  not  materially  influence  the  small  amount  of  atmos- 
pheric nitrogen  fixed  under  these  conditions. 

(8)  When  grown  on  agar  films,  B.  radicicola  fixed  a  small  amount 
of  nitrogen,  varying  from  0.15  to  0.43  mgm.  of  nitrogen  in  100  c.  c.  of 
the  medium.  The  addition  of  various  amounts  of  potassium,  sodium, 
and  calcium  nitrates  increased  to  a  slight  extent  the  amount  of  nitrogen 
assimilated. 

(9)  In  liquid  cultures  all  three  nitrates  caused  a  large  increase  in  the 
amount  of  gum  obtained  by  precipitation  with  acetone. 

(10)  The  presence  of  large  amounts  of  potassium,  sodium,  and  cal- 
cium nitrates  proved  detrimental  to  the  formation  of  nodules  on  alfalfa. 
B.  radicicola  did  not  appear  to  lose  its  infecting  power  when  grown  on 
media  containing  varying  amounts  of  sodium  and  calcium  nitrates. 
Alfalfa  seedlings  grown  in  the  presence  of  large  amounts  of  nitrate  did 
not  produce  nodules  when  inoculated  wdth  a  viable  culture  of  B.  radicicola. 
Nitrates  in  soil  cultures  prevented  the  re-formation  of  nodules  once 
removed  and  also  caused  a  decrease  in  the  number  of  nodules  already 
present. 

LITERATURE  CITED 
(i)  Bassler,  p. 

1895.    SANDKULTURVERSUCHE  UBER  DIE  STICKSTOFFASSIMILATION  DER  GELBEN 
LUPINE    IM    STERIUSIERTEN    UND    GEIMPFTEN     BODE.V     BEX     DARGEBOT 

WECHSELNDER  MENGEN  VON  SALPETERSAUREN  SALZEN.     In  Jahresbef. 
Agr.  Chein.,  n.  F.,  Bd.  18  fjahrg.  38),  p.  131. 

(2)  BEIJERINCK,  M.  W. 

1888.    DIE    BACTERIEN    DER   PAPILIONACEEN-KNOLIXHEN.      In    Bot.    Ztg.,    Jahfg. 

46,  No.  48,  p.  758-771.     (Continued  article.) 

(3)  

1891.    OVER   OPHOOPING    VAN    ATMOSPHERISCHE    STICKSTOP   IN   CULTUREN    VAN 

BACILLUS  RADICICOLA.     In  Verslag.  en  Meded.  K.  Akad.  Weteiisch. 
[Amsterdam],  r.  3,  deel  8,  p.  460-475. 


228  Journal  of  Agricultural  Research  voi.  xn,  No.  4 

(4)  Beijerinck,  M.  W. 

1901.  UEBER  OLiGONiTROPHiLE  MiKROBEN.     In  Centbl.  Bakt.  [etc.],  Abt.  2, 

Bd.  7,  No.  16,  p.  561-582,  I  pi. 

(5)  and  Delden,  A.  van. 

1902.  UEBER  DIE  ASSIMILATION  DES  FREIEN  STICKSTOFFS  DURCH  BAKTERIEN.      In 

Centbl.  Bakt.  [etc.],  Abt.  2,  Bd.  9,  no.  1/2,  p.  3-43. 

(6)  Berthelot,  M.  p.  E. 

1893.  RECHERCHES   NOUVELLES   SUR   LES    MICRORGANISMES    FIXATEURS    DE    L' 

AZOTE.     In  Compt.  Rend.  Acad.  Sci.  [Paris],  t.  116,  no.  17,  p.  842-849. 

(7)  BoNAZzi,  Augusto. 

1915.   CYTOLOGICAL   STUDIES    OF   AZOTOBACTER  CHROOCOCCUM.      In  JoUT.    Agr. 

Research,  v.  4,  no.  3,  p.  225-239.     Literature  cited,  p.  238-239. 

(8)  BOTTOMLEY,  W.  B. 

1909.  SOME  EFFECTS  OF  NITROGEN-FIXING  BACTERIA  ON  THE  GROWTH  OF  NON- 
LEGUMINOUS  PLANTS.  In  Proc.  Roy.  Soc.  London,  s.  B,  v.  81,  no. 
548,  p.  287-289. 

(9)  Bredemann,  G. 

1909.  bacillus  amylobacter  a.  mt  et  bredemann  in  morphologischer, 
PHYSiOLOGiscHER  UND  SYSTEMATiscHER  BEziEHUNG.  In  Centbl.  Bakt 
[etc.],  Abt.  2,  Bd.  23,  no.  14/20,  p.  385-568,  13  fig.,  2  pi.  Literatur- 
verzeichnis,  p.  559-566. 

(10)  Buchanan,  R.  E. 

1909.  THE  GUM  PRODUCED  BY  BACILUS   RADICICOLA.     In  Centbl.    Bakt.    [etc.], 

Abt.  2,  Bd.  22,  No.  11/13,  p.  371-396.     Citations,  p.  395-396. 

(11)  Chalmot,  G.  de 

1894.  THE  INFLUENCE  OF  NITRATES  ON  GERMINATING  SEEDS.    In  Agr.  Science, 

V.  8,  no.  lo/ii,  p.  463-465. 

(12)  CrEydt,  Bodo. 

1915.  UNTERSUCHUNGEN    UBER    DIE    KALKEMPFINDLICHKEIT    DER   LUPINE    UND 

IHRE  BEKAMPFUNG.     In  Jour.  Landw.,  Bd.  63,  Heft  2,  p.  125-191,  6 
pi. 

(13)  Drabble,  Eric,  and  Scott,  Daisy  G. 

1907.  on  the  effect  of  ACIDS,  ALKALIS,  AND  NEUTRAL  SALTS  ON  THE  FER- 
MENTATIVE ACTIVITY  AND  ON  THE  RATE  OF  MULTIPLICATION  OP  YEAST 

CELLS,     hi  Biochem.  Jour.,  v.  2,  no.  7/8,  p.  340-349,  i  fig.     Literature 

P-  349- 

(14)  Fernbach,  a.,  and  Lanzenberg,  A. 

1910.  DE   L'aCTION    DES    NITRATES    DANS    LA   FERMENTATION    ALCOOLIQUE.      In 

Compt.  Rend.  Acad.  Sci.  [Paris],  t.  151,  no.  17,  p.  726-729. 

(15)  Flamand,  Henri. 

1905.  XJBER  den  EINFLUSS  DER  ERNAHRUNG  AUP  DIE  ENTWICICLUNG  DER 
KNOLLCHEN  DER  LEGUMiNOSEN.  In  Centbl.  Agr.  Chem.,  Jahrg.  34, 
Heft  II,  p.  738-740. 

(16)  Frank,  B. 

1892.  DIE  ASSIMILATION  FREIEN  STICKSTOFFS  BEI  DEN  PFLANZEN  IN  IHRER 
ABHANGIGKEIT   VON  SPECIES,    VON    ERNAHRUNGSVERHALTNISSEN   UND 

VON  BODENARTEN.     In  Landw.  Jahrb.,  Bd.  21,  p.  1-44. 

(17)  Fred,  E.  B. 

1913.   A  PHYSIOLOGICAL  STUDY  OF  THE  LEGUME  BACTERIA.      In  Va.  Agr.   Exp. 

Sta.,  Ann.  Rpt.,  1911/12,  p.  145-173,  fig.  34.     Literature,  p.  172-173. 

(18)  and  Graul,  E.  J. 

1916.  THE  EFFECT  OF  SOLUBLE   NITROGENOUS   SALTS   ON   NODULE  FORMATION. 

In  Jour.  Amer.  Soc.  Agron.,  v.  8,  no.  5,  p.  316-328.     Literature  cited, 
p.  327-328. 


Jan.  28,  i9i8  Nitrogen-Assimilating  Bacteria  229 

(19)  Greaves,  J.  E. 

1916.   THE  INFLUENCE  OP  SALTS  ON  THE  BACTERIAL  ACTIVITIES  OP  THE  SOIL.      In 

Soil  Science,  v.  2,  no.  5,  p.  443-480,  4  fig. 

(20)  Hanzawa,  J. 

I914.  EINIGE  beobachtungen  UBER  STICKSTOFF-BINDUNG  durch  azotobac- 
TER   IN   STICKSTOFFARMEN    UND     IN   STICKSTOFFREICHEN   SUBSTRATEN. 

In  Centbl.  Bakt.  [etc.],  Abt.  2,  Bd.  41,  No.  18/23,  p.  573-576. 

(21)  HellriEgel,  H.,  and  Wilfarth,  H. 

1888.    UNTERSUCHUNGEN  UBER  die  STICKSTOFFNAHRUNG  DER  GRAMINEEN  UND 

LEGUMiNOSEN.     In    Ztschr.    Ver.    Riibenzuckerindus.,    Beilageheft, 
Nov.,  234  p.,  6  pi. 

(22)  Herke,  S. 

i913.  contributions  on  nitrogen  fixation  and  nutrition  op  bacillus 
radicicola  and  on  bacterial  tests  op  nitragin  and  azotogen. 
(Abstract.)  In  Exp.  Sta.  Rec,  v.  29,  no.  8,  p.  733.  (Original  article 
in  Kiserlet.  Kozlem.,  v.  16,  no.  3,  p.  311-322,  1913.     Not  seen.) 

(23)  Hills,  T.  L. 

the  INFLUENCE  OF  NITRATES  ON  THE  GROWTH  OP  AZOTOBACTER.  Un- 
published. Offered  for  publication  in  the  Ann.  Rept.  of  the  Pa.  Agr. 
Exp.  Sta. 

(24)  HiLTNER,  L. 

1900.  UEBER    DIE    URSACHEN,     WELCHE    DIE    GROSSE,     ZAHL,     STELLUNG    UND 

WIRKUNG  DER  WURZELKNOLLCHEN  DER  LEGUMINOSEN  BEDINGEN.      In 

Arb.  K.  Gsndhtsamt.,  Biol.  Abt.,  Bd.  i.  Heft  2,  p.  177-222,  pi.  3. 

(25)  Immendorfp,  H. 

1892.    BEITRAGE    ZUR    LOSUNG     DER     "  STICKSTOFFRAGE. "      In    Landw.    Jahxb. 

Bd.  21,  p.  281-339. 

(26)  JosT,  Ludwig. 

1907.    LECTURES  ON  PLANT  PHYSIOLOGY.     Translated  by  R.  J.  H.  Gibson. 
464  p.,  172  fig.     Oxford.     Bibliography  at  the  end  of  each  lecture. 

(27)  Kayser,  E. 

I910.      INFLUENCE  DES  NITRATES  SUR  LES  FERMENTS  ALCOOLIQUES.      /nCompt. 

Rend.  Acad.  Sci.  [Paris],  t.  151,  no.  19,  p.  816-817. 

(28)  Kossowicz,  Alexander. 

I914.      UBER  DAS  VERHALTEN  VON  HEFEX  UND  SCHIMMELPILZEN  ZU  NITRATEN. 

In  Biochem.  Ztschr.,  Bd.  67,  Heft  4/5,  p.  400-419. 

(29)  IvAURENT,  Emile. 

1891.    RECHERCHEs  SUR  LES  nodosite;3  radicales  DES  l:6gumineuses.    In 
Ann.  Inst.  Pasteur,  annee  5,  p.  105-139,  3  fig. 

(30)  Lewis,  L.  L.,  and  Nicholson,  J.  F. 

1905.      soil  INOCULATION.      TUBERCLE-FORMING  BACTERIA  OP  LEGUMES.      Okla. 

Agr.  Exp.  Sta.  Bui.  68,  30  p.,  8  fig. 

(31)  Marchal,  Emile 

1901.  INFLUENCE    DES    SELS    MIN^RAUX    NUTRITIFS    SUR   LA   PRODUCTION    DES 

NODOSIT^S  CHEZ   LE  POis.     In  Coinpt.    Rend.   Acad.    Sci.    [Paris], 
t.  133,  no.  24,  p.  1032-1033. 

(32)  Maz6,  p. 

1897.  FIXATION    DE    L 'azote    LIBRE    PAR    LE     BACILLE     DES     NODOSITIES    DES 

l6guminEuses.     In  Ann.  Inst.  Pasteur,  annee  11,  no.  i,  p.  44-54. 

{zz) 

1898.  LES  MICROBES   DES   nodosit6s   DES  l6guminEuses.     In  Ann.    Inst. 

Pasteur,  ann.  12,  no.  i,  p.  1-25,  i  fig. 
(34)  Meyer,  Arthur. 

1912.     DIE  ZELLE  DER  BAKTERiEN.    285  p.,  34  fig.,  I  col.pl.  Jena.     Literatur, 
p.  267-282. 


230  Journal  0}  Agricultural  Research  voi.  xii.  no.  4 

(35)  Moore,  G.  T. 

1905.  soil,  inoculation  for  legumes;  wit?i  reports  upon  the  successful 
use  of  artificial  cultures  by  practical  farmers.  u.  s.  dept. 
Agr.  Bur.  Plant  Indus.  Bui.  71,  72  p.,  10  pi. 

(36)  MUNTER,  F. 

I916.      tJBER  DEN  EINFLUSS  ANORGANISCHER  SALZE  AUF   DAS   WACHSTUM   DER 

ACTiNOMYCETEN.  III.     In  Centbl.  Bakt.  [etc.],  Abt.  2,  Bd.  44,  No. 
24/25,  p.  673-695,  9  fig. 

(37)  NoBBE,  F.,  and  Richter,  L. 

1902.      UBER    den    EINFLUSS    DES    NITRATSTICKSTOFFS    UND    DER    HUMUSSUBS- 

stanzen  AUF  DEN  iMPFUNGSERFOLG  BEi  lEguminosen.     In  Landw. 
Vers.  Stat.,  Bd.  56,  Heft  5/6,  p.  441-448. 

(38)  Pfeffer,  W. 

igo6.  THE  physiology  op  PLANTS,  cd.  2.  Translated  by  A.  J.  Ewart.  v.  3. 
Oxford. 

(39)  Prazmowski,  Adam. 

1891.  die  WURZELKNOLLCHEN  DER  ERBSE:   II.  TEIL.  DIE  BIOLOGISCHE  BEDEU- 

TUNG  DER  WURZELKNOLLCHEN.     In  Landw.  Vers.  Stat.,  Bd.  38,  p. 

5-56. 

(40)  PringshEim,  Hans. 

1914.  ZUR     STICKSTOFFASSIMILATION     IN     GEGENWART     VON     SALPETER.      In 

Centbl.  Bakt.  [etc.],  Abt.  2,  Bd.  40,  no.  1/8,  p.  21-23. 

(41)  Prucha,  M.  J. 

1915.  PHYSIOLOGICAL  STUDIES  OF  BACILLUS  RADICICOLA  OP  CANADA  FIELD  PEA. 

N.  Y.  Cornell  Agr.  Exp.  Sta.  Mem.  5,  83  p.     Bibliography,  p.  79-83. 

(42)  RiTTER,  G. 

1909.  AMMONIAK  UND  NITRATE  ALS  STICKSTOFFQUELLE  PUR  SCHIMMELPILZE. 
In  Ber.  Deut.  Bot.  Gesell.,  Bd.  27,  Heft  10,  p.  582-588. 

(43)  SackETT,  W.  G. 

1915.  THE  PIGMENT  OP  azotobacter  chroococcum.  In  Proc.  35th  Ann. 
Meeting  Soc.  Prom.  Agri.  Sci.,  1914,  p.  80-88,  2  col.  pi. 

(44)  Stoklasa,  Julius. 

1908.  BEITRAG  ZUR  KENNTNIS  DER  CHEMISCHEN  VORGANGE  BEI  DER  ASSIMI- 
LATION   DES    ELEMENTAREN    STICKSTOFFS    DURCH    AZOTOBACTER    UND 

radiobacter.     In  Centbl.  Bakt.  [etc.],  Abt.  2,  Bd.  21,  no.  15/16,  p. 
484-511. 

(45)  Vines,  S.  H. 

1888.  on  the  relation  between  the  formation  of  tubercles  on  the 
roots  op  leguminosae  and  the  presence  op  nitrogen  in  the 
SOIL.     In  Ann.  Bot.,  v.  2,  no.  7,  p.  386-389. 

(46)  Wilson,  J.  K. 

I915.  PHYSIOLOGICAL  STUDIES  OF  BACILLUS  RADICICOLA  OP  SOY  BEAN.  (Ab- 
stract.)    In  Science,  n.s.,  v.  41,  no.  1048,  p.  180. 

(47)  wohltmann,  Ferdinand,  and  bergen^, 

1902.      DIE  KNOLLCHEN-BAKTERIEN  IN  IHRER  ABHANGIGKEIT  VON  BODEN  UND 

DXJNGUNG.     In  Jour.  Landw.,  Bd.  50,  Heft  4,  p.  377-395. 

(48)  Woods,  C.  D. 

1892.  the  acquisition  op  atmospheric  nitrogen  by  growing  plants.    in 

Conn.  Storrs  Agr.  Exp.  Sta.,  4th  Ann.  Rpt.,  1891,  p.  17-28. 

(49)  ZiPFEL,  Hugo. 

I911.      BEITRAGE  ZUR  MORPHOLOGIE  UND  BIOLOGIE  DER  KNOLLCHENBAKTERIEN 

DER  LEGUMINOSEN.     In  Centbl.  Bakt.  [etc.],  Abt.  2,   Bd.  32,  No. 
3/5'  P-  97-137-     Literature,  p.  136-137. 


Vol.  XII  FEBRUARY  ^,  1918  No.  5 

JOURNAL  OF 

AGRICULTURAL 
RESEARCH 


CONXEMStXS 

Page 

New-Place  Effect  in  Maize       ------      231 

G.  N.  COLLINS 

(Contribution  from  Bureau  of  Plant  Industry) 

Relation  of  the  Variability  of  Yields  of  Fruit  Trees  to  the 

Accuracy  of  Field  Trials        -       -       -       -       -       -      245 

L.  D.  BATCHELOR  and  H.  S.  REED 

<  Contribution  from  California  Agrjcaltuial  Experiment  Station) 

Interrelations  of  Fruit-Fly  Parasites  in  Hawaii        -        -      285 
C.  E.  PEMBERTON  AND  H.  F.  WILI^RD 

( Contribution  from  Buneau  of  Entomology) 


PDBUSHED  BY  AUTHORITY  OF  THE  SECRETARY  OF  AGRICCITDRE, 

WITH  THE  COOPERATION  OF  THE  ASSOCIATION  OF  AMERICAN 

AGRICULTURAL  COLLEGES  AND  EXPERIMENT  STATIONS 


WASI^IISIGTON,  D.  C. 


WASHINOTON  :  OOVERNMENT  PnillTlNO  OTFICe  Mei8 


EDITORIAL  COMMITTEE  OF  THE 

UNITED    STATES    DEPARTMENT    OF   AGRICULTURE    AND 

THE  ASSOCIATION  OF  AMERICAN  AGRICULTURAL 

COLLEGES  AND  EXPERIMENT  STATIONS 


FOR  THE  DEPARTMENT 

E:ARL  F.  KELLERMAN,  Chairman 

Physiologist  and  Associate  Chief,  Bureau 
of  Plant  Industry 

EDWIN  W.  ALLEN 

Chief,  Offite  of  Experiment  Stations 

CHARLES  L.  MARLATT 

Entomologist  and  Assistant  Chief,  Bureau 
of  Entomotogy 


FOR  THE  ASSOCIATION 
RAYMOND  PEARL* 

Biologist,  Maine  Agricultural  Experinuiil 
Station 

H.  P.  ARMSBY 

Director,  Instiiutt  of  Animal  Nutrition,  Thf 
Pennsylvania  State  College 

E.  M.  FREEMAN 

Botanist,  Plant  Pathologist  and  AssistattI 
Dean,  Agricultural  Experiment  Station  of 
the  University  of  Minnesota 


All  correspondence  regarding  articles  from  the  Department  of  Agriculture  should  be 
addressed  to  Karl  F.  Kellermaa,  Journal  of  Agricultural  Research,  Washington,  D.  C. 

*Dr.  Pearl  has  undertaken  special  work  in  connection  with  the.  war  emergency; 
therefore,  until  further  notice  all  correspondence  regarding  articles  from  State  Experi- 
ment Stations  should  be  addressed  to  H.  P.  Armsby,  Institute  of  Animal  Nutrition, 
State  College,  Pa. 


J01N£  OF  AGRIOmm  RESEARCH 

Voiv.  XII  Washington,  D.  C,  February  4,   1918  No.  5 


NEW-PLACE  EFFECT  IN  MAIZE 

By  G.  N.  Collins,  Botanist,  Office  of  Acclimatization  and  Adaptation  of  Crop  Plant?  and 
Cotton  Breeding,  Bureau  of  Plant  Industry, United  States  Department  of  Agriculture 

INTRODUCTION 

Widely  divergent  opinions  have  been  expressed  regarding  the  ad- 
visability of  transferring  seed  from  one  region  to  another.  With  maize 
(Zea  mays)  the  transfer  of  seed  is  generally  held  to  be  disadvantageous. 
Numerous  experiments  have  shown  that  when  seed  of  the  same  original 
variety,  but  grown  at  two  places,  is  planted  side  by  side  at  one  of  the 
places  the  results  are  in  favor  of  the  local  seed. 

The  relative  superiority  of  the  locally  selected  seed  has  been  so  pro- 
nounced that  the  securing  of  seed  from  distant  localities,  except  where 
grown  for  forage,  has  been  discouraged.  The  natural  result  is  to  confine 
the  utilization  of  carefully  selected  strains  to  the  localities  where  the 
breeding  is  done.  Caution  in  transferring  seed  is  certainly  desirable,  but 
the  results  of  the  experiments  here  reported  indicate  that  two  opposing 
factors  are  involved,  the  relative  importance  of  which  must  be  determined 
before  generalizations  are  made. 

The  differences  shown  between  the  geographically  separated  lines  of 
the  same  variety  when  brought  together  in  the  locality  where  one  of  the 
lines  has  been  grown  may  be  ascribed  to  three  general  causes: 

(i)  Cross-pollination.  Varieties  removed  to  different  localities  may 
become  crossed  with  other  sorts,  with  the  result  that  their  character- 
istics are  changed  in  a  greater  or  less  degree. 

(2)  Different  standards  of  selections.  Changed  conditions  in  con- 
nection with  the  diversity  that  exists  in  all  varieties  bring  slightly  different 
types  of  plants  into  prominence,  and  selection,  either  conscious  or  un- 
conscious, results  in  a  changed  type. 

(3)  New-place  effects.  The  more  or  less  temporary  changes  that  fol- 
low a  transfer  to  new  conditions,  caused  by  the  novelty  or  "shock"  of  the 
new  environment  without  special  reference  to  the  nature  of  the  change. 

So  far  as  known,  new-place  effect,  which  the  experiments  here  re- 
ported show  to  be  a  significant  factor  in  the  transfer  of  seed,  has  not 
previously  been  considered  with  regard  to  maize. 

Journal  of  Agricultural  Research,  Vol.  XII,  No.  s 

Washington,  D.  C.  Feb.  4,  1918 

lu  Key  No.  G— 134 

(231) 


232  Journal  of  Agricultural  Research  voi.  xii.  no.  5 

These  experiments  were  originally  planned  with  the  idea  that  the  seed 
of  a  first-generation  hybrid  might  be  transferred  to  a  distant  locality 
without  showing  the  reduced  yields  thought  to  follow  the  transfer  of 
pure  strains. 

It  was  believed  that  with  first-generation  hybrid  seed  it  might  be  a 
matter  of  indifference  where  the  hybrid  was  made,  provided  the  parents 
were  suited  to  the  localities  where  the  crop  was  to  be  grown.*  If  found 
to  be  the  case,  this  would  constitute  an  important  addition  to  the  ad- 
vantages that  follow  the  use  of  first-generation  hybrid  seed,  making 
possible  a  wider  application  of  the  results  secured  through  the  work 
of  skilled  breeders. 

The  results  indicate  that  hybrids  made  in  one  locality  and  grown  in 
another  are  not  only  at  no  disadvantage  compared  with  the  same  hybrids 
produced  at  the  locality  where  the  comparison  is  made  but  that  the 
introduced  hybrid  may  even  be  superior.  It  further  appears  that  the 
same  is  true  of  pure  strains  and  that,  after  the  effects  of  cross-pollination 
and  selection  have  been  eliminated  from  the  problem,  there  is  a  residual 
effect  of  the  transfer  to  the  new  place  that  tends  to  increase  rather  than 
to  reduce  the  A^gor  and  yield  of  the  plants. 

NATURE  OF  THE  EXPERIMENTS 

It  was  planned  to  conduct  the  experiments  at  the  following  places: 
Stockton,  Kans. ;  Victoria,  Tex. ;  and  Lanham,  Md.  These  places 
represent  a  wide  range  of  soil  and  climate.  At  Stockton,  which  is 
in  the  eastern  part  of  the  semiarid  Great  Plains  area,  there  is  a  fertile 
friable  soil,  low  rainfall,  low  atmospheric  humidity,  and  a  prevalence  of 
high  winds.  Victoria,  located  in  the  Gulf  region  of  Texas,  has  a  stiff 
black  clay  soil  of  good  fertility.  The  total  rainfall  is  usually  large  and 
the  humidity  high,  though  severe  drouths  are  not  uncommon.  At 
Lanham,  a  few  miles  north  of  Washington,  D.  C,  the  soil  is  sandy, 
acid,  and  relatively  infertile.  The  rainfall  and  humidity  are  usually 
neither  deficient  nor  excessive.  In  the  discussion  of  results  these  locali- 
ties will  be  referred  to  as  Kansas,  Texas,  and  Maryland. 
Four  varieties  of  maize  were  chosen  for  the  experiment,  as  follows: 
(i)  Stockton,  a  white  dent  variety  developed  at  Stockton,  Kans. 

(2)  Strawberry,  a  large-eared  Texas  variety  with  red  and  white  varie- 
gated dent  seed,  well  adapted  to  the  conditions  at  Victoria,  Tex. 

(3)  Hickory  King,  a  strain  of  this  variety  grown  in  Virginia  and  well 
suited  to  the  conditions  at  Lanham,  Md. 

(4)  Boone,  a  strain  of  "Boone  County  White,"  the  seed  of  which  was 
obtained  from  Illinois. 

In  the  spring  of  191 2  seed  of  the  four  varieties  were  planted  at  each 
of  the  three  places  named.     The  precaution  was  taken  to  mix  the  seed 

1  COLUNS,  G.  N.     THE  VAI,ue  OF   FIRST-GENERATION  HYBRIDS  IN    CORN.   U.  S.   Dept.  AgT.  Bur.  Plant 

Indus.  Bui.  191,  p.  33.    1910. 


Feb.  4. 1918  New-Place  Effect  in  Maize  233 

of  each  of  the  kinds  so  that  the  portions  sent  to  the  several  localities 
should  be  as  nearly  alike  as  possible.  The  order  of  planting  was  the  same 
at  all  the  places,  as  follows:  Every  alternate  row  was  planted  to  the 
Boone  variety,  which  was  used  as  the  male  parent  in  making  hybrids. 
The  seed  of  each  of  the  four  varieties,  including  the  Boone  itself,  was 
planted  in  the  rows  alternating  with  the  Boone. 

The  Boone  plants  standing  in  the  alternate  rows  throughout  the  field 
were  allowed  to  shed  pollen.  All  others,  including  the  interplanted 
Boone,  were  detasseled,  so  that  the  only  pollen  shed  was  from  the  Boone 
variety.  Seed  was  saved  from  the  detasseled  rows  only.  This  was  of 
four  kinds,  (i)  Stockton  X  Boone,  (2)  Strawberry  X  Boone,  (3)  Hickory 
King  X  Boone,  and  (4)  cross-pollinated  seed  of  the  Boone.  At  each 
locality  the  experiment  was  placed  at  a  distance  from  all  other  corn. 

In  1 91 3  it  was  planned  to  compare  the  behavior  of  the  plants  raised 
from  the  seed  produced  at  the  three  localities  when  grown  at  each  of 
these  places.  Although  grown  in  the  same  field,  no  attempt  was  made  to 
compare  the  relative  merits  of  the  several  hybrids,  each  hybrid  together 
with  the  pure-seed  Boone  constituting  a  separate  experiment  involving 
only  the  comparison  of  the  yield  from  the  seed  of  the  three  localities. 
Thus,  there  were  four  experiments  to  be  made  at  each  of  the  three 
places.  Since  the  arrangement  was  the  same  at  all  places,  one 
description  will  suffice. 

To  compare  the  Stockton  X  Boone  hybrid  from  the  three  localities, 
the  seed  from  the  several  places  were  planted  in  adjoining  rows,  the 
first  row  from  the  Kansas  seed,  the  second  from  the  Texas  seed,  and  the 
third  from  the  Maryland  seed.  The  series  was  repeated  10  times,  making 
10  distinct  comparisons.  A  similar  procedure  was  followed  with  the 
three  other  hybrids  and  with  the  pure-seed  Boone. 

At  Stockton,  Kans.,  excessive  drouth  destroyed  the  entire  corn  crop. 
Since  no  results  were  secured  from  Stockton,  the  behavior  of  the  Kansas- 
grown  seed  will  be  eliminated  from  the  discussion  of  the  results,  which 
will  be  confined,  therefore,  to  the  experiments  conducted  in  Texas  and 
Maryland. 

At  Victoria,  Tex.,  the  rows  were  100  feet  long.  The  seed  was  drilled, 
and  the  plants  were  thinned  to  about  2  feet  in  the  row.  When  harvested, 
a  weighed  sample  of  20  pounds  of  ears  was  saved  from  each  row.  This 
sample  was  thoroughly  air-dried,  after  which  it  was  again  weighed  to 
determine  the  loss  of  water.  The  percentage  of  grain  to  cob  was  also 
determined.  No  significant  differences  in  water  content  or  percentage  of 
grain  were  found  in  the  crops  from  the  seed  from  different  localities,  and 
these  determinations  are,  therefore,  not  discussed. 

At  Lanham,  Md.,  the  seed  was  planted  in  hills  3  feet  apart,  in  rows 
132  feet  long.  The  plants  were  thinned  to  one  stalk  per  hill.  The 
method  of  harvesting  was  similar  to  that  of  Victoria  except  that  no 
determinations  of  dry  weight  were  made. 


234 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  s 


At  each  locality  the  com  from  all  the  experiments  was  harvested  the 
same  day;  and  the  weight  of  ears,  together  with  the  number  of  plants, 
was  recorded  for  each  row. 

To  avoid,  so  far  as  possible,  differences  due  to  inequalities  of  soil  and 
to  obtain  reliable  averages,  each  pair  of  rows  consisting  of  one  row  each 
of  Maryland-  and  Texas-grown  seed  was  considered  a  separate  test.  The 
relative  behavior  of  the  plants  from  the  two  sources  of  seed  was  deter- 
mined by  an  average  of  all  the  comparisons,  usually  lo  in  number. 

In  Table  I  are  given  the  yield  in  pounds  per  row  and  the  yield  per  plant 
from  the  several  rows.  Yields  which  stand  opposite  in  the  table  are 
from  adjoining  rows  in  the  field. 

Table   I. — Behavior  of  plants  from  Maryland-   and   Texas-grown    seed  subseqtiently 
planted  in  Maryland  and  Texas 

STOCKTON      X      BOONE 


Compared  at  Lanham,  Md. 

Compared  at  Victoria,  Tex. 

Variety  and  factor. 

Yield 
from 
seed  pro- 
duced at 
Lanham. 

Vield 
from 
seed  pro- 
duced at 
Victoria. 

Yield  from  seed 
produced  at 
Lanham  ex- 
pressed as  a 
percentage  of 
the  mean. 

Yield 
from 
seed  pro- 
duced at 
Lanham. 

Yield 
from 
seed  pro- 
duced at 
Victoria. 

Yield  from  seed 
produced  at 
Lanham  ex- 
pressed as  a 
percentage  of 
the  mean. 

Yield  per  plant 

Do 

Pou 

O 

ids. 
90 
90 
92 
93 
89 
60 
26 
58 
78 
67 

Pounds. 

I- IS 
I.  06 

•93 
.81 

•74 
•52 
.61 

•56 
•78 
•71 

Per  cent. 

87.8 
91.8 
99.4 

106.  9 
109.  2 

107.  I 
59-8 

loi.  7 

100.  0 

97.1 

Pounds. 
0.  65 

.65 
.62 

■63 
•65 
.62 
.64 

•59 
.64 
.64 

Pounds. 
0.  65 

.66 
•56 
.81 

•63 
.69 
.66 
.62 
.64 
.66 

Per  cent. 

100.  0 

99.2 

105.  I 

87.  5 
loi.  5 
94.6 
98.5 
97^5 
100. 0 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

98.5 

' 

Average .... 

•743 

.787 

96.  o8±2.  75 

•633 

.658 

98.  25 ±0.86 

Yield  per  row 

Do 

30-5 
32-5 
33^o 
25.  0 

29-5 
9.0 

5-5 
18.0 

23-5 
22.  0 

31.0 
41.5 
32-5 
29.  0 

18.5 
II.  0 

19-5 
18.5 
26.  5 
22.  0 

99.1 

87.8 

100.8 

92.  6 
122.  9 
90.  0 
44.0 
98.6 
94.0 
100.  0 

32-5 
31.8 
32.0 
32.8 
34- 0 
28.5 
34-5 
31.0 
34-8 
38.0 

33- 0 

32.5 
32.  0 
38.0 
31-5 
36.5 
30-5 
36-5 
34-5 
33- 0 

99.2 

98.9 

100.  0 

Do 

Do 

92.6 
103.8 

87.7 
106.  2 

Do 

Do 

Do 

Do 

91.9 
100.  4 

Do 

Do 

107.  0 

Average . . . . 

22 

•85 

25.  0 

92.  99±3.  25 

32-97 

33- 80 

98.  76±i.  64 

STRAWBERRY      X      BOONE 


Yield  per  plant . 

Do 

Do 

Do 


1.80 
1. 81 

I.  92 

1. 61 

I.  92 

1.89 

1.74 

I.  64 

96.8 

105.9 
104.9 
107.  I 


0.  70 

0.51 

.72 

.72 

•  78 

.84 

.80 

.84 

II5-7 
100.  o 

96- 3 
97.6 


Feb.  4,  1918 


New-Place  Effect  in  Maize 


235 


Tablet. — Behavior  of  plants  from  Maryland-   and   Texas-grown   seed  subsequently 
planted  in  Maryland  and  Texas — Continued 

STRAWBERRY  X  BOONE — Continued 


Compared  at  Lanham,  Md. 

Compared  at  Victoria,  Tex. 

Variety  and  factor. 

Yield 
from 
seed  pro- 
duced at 
Lanham. 

Yield 
from 
seed  pro- 
duced at 
Victoria. 

Yield  from  seed 
produced  at 
Lanham  ex- 
pressed as  a 
percentage  of 
the  mean. 

Yield 
from 
seed  pro- 
duced at 
Lanham. 

Yield 
from 
seed  pro- 
duced at 
Victoria. 

Yield  from  seed 
produced  at 
Lanham  ex- 
pressed as  a 
percentage  of 
the  mean. 

Yield  per  plant. . .  . 

Do 

Do 

Pounds. 
1-83 
I.  72 

Pounds. 
1.65 

1-38 

Per  cent. 

105.  2 
III.  0 

Pounds. 
0.74 

•74 
•77 
.69 

-71 
.70 

Pounds. 

0.82 

•85 

-74 

.78 

•65 
.72 

Per  cent. 

94.9 
93-1 

Do 

93-9 

104.4 

98.6 

Do 

Do 

Average .... 

1.83 

1.66 

105.  15  ±1.09 

•735 

•747 

99- 65  ±1.32 

Yield  per  row 

Do 

75-5 
63-5 
71.0 
68.0 
62.  0 
67.  0 

71.0 

64- 5 

61.  0 

62.  5 
61.0 
47.0 

103.  07 
99.  22 

107.  58 

104.  21 
100.  81 

"7-  54 

39.00 
42.  50 
41.50 
39.00 
37.00 
36.80 
39-30 
37.00 
37.00 
38.00 

27.30 
30-30 
32.00 
34-50 
37-50 
37-50 
33-30 
39.80 
30.50 
38.00 

117.  7 
116  8 

Do 

Do 

112.  9 
106  I 

Do 

Do 

Do 

99-3 
99.0 

108.3 
96.4 

109.  6 

Do 

Do 

Do 

Average .... 

67.8 

61.  2      105.  401b  I.  80 

38.70 

34-05 

106.  61  ±  I.  82 

HICKORY   KING    X    BOONE 


Yield  per  plant .  .  . . 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

A.verage .  . . 

Yield  per  row 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Average . . . 


1.  016 


21.5 
17-5 
17-5 
32.0 
46.  o 
50.  o 

51-5 
52.0 
47.0 
51.0 

38.6 


.60 
.60 
•97 
^•15 
1.08 
1.50 
1.24 
1.36 
1.30 


1.038 


18-5 
18.0 
40.  5 
38.0 
40.  o 

58-5 
46.  o 

49.0 
52.0 

38-15 


104.  I 

91.9 

93^8 

94.0 

loi.  3 

107.3 

92-5 

105.  o 

93-8 
loi.  5 


22±I.  42 


lOI.  2 
97.2 
98.6 
88.3 

109s 
III.  I 

93^6 
106.  I 

97^9 
99.0 

100.  25 ±1.43 


.606 


31.00 

29.  00 
28.  00 
32.30 
31.30 
28.  00 
28.80 
32.80 
31-50 
28.  50 

30.  10 


545 


no.  9 
loi.  7 
100.  o 
103-9 
104.5 
104.  2 
113. 1 
99.1 
no.  5 
105.2 


105.  3i±i.  o 


21.  00 
23.00 

25.  00 
27-30 
27.50 

26.  00 
23.00 
24.  00 
25-50 
26.  50 

24-87 


119.  2 
III.  5 

105-7 
108.  4 
ic6.  4 
103-7 
III.  I 

II5-4 
no.  5 
103.6 

109.  55  ±  I.  13 


236 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  5 


Table  I. — Behavior  of  plants  from  Maryland-  and   Texas-grown  seed  subsequently 
planted  in  Maryland  and  Texas — Continued 


BOONE 


Compared  at  Lanham,  Md. 

Compared  at  Victoria,  Tex. 

Variety  and  factor. 

Yield 
from 
seed  pro- 
duced at 
Lanham. 

Yield 
from 
seed  pro- 
duced at 
Victoria. 

Yield  from  seed 
produced  at 
Lanham  ex- 
pressed as  a 
percentage  of 
the  mean. 

Yield 
from 
seed  pro- 
duced at 
Laaham. 

Yield 
from 
seed  pro- 
duced at 
Victoria. 

Yield  from  seed 
produced  at 
Lanham  ex- 
pressed as  a 

percentage  of 
the  mean. 

Yield  per  plant 

Do 

Pounds. 
1.38 

I-  15 
1.08 
1.30 
I.  10 

1-35 
I.  22 
1.30 
I.  16 
I.  14 

Pounds. 
1.36 
I.  21 
1.36 
I.  26 

I-  13 
I.  II 

I-  51 

1-35 
I.  22 
1.24 

Per  cent. 

100.  7 

97-5 
88.5 

loi.  6 
98.7 

109.  8 
89.4 
98.  I 

97-5 
95-8 

Pounds. 
0.56 
.60 
■58 
•63 
.60 
.69 
.60 
.62 

Pounds. 
0.78 

•73 
.68 

.67 
.64 
.62 

.68 
■63 

Per  cent. 

83.6 
90.  2 
92.  I 
96.9 
96.8 

105- 3 
93^8 
99.2 

Do 

Do 

Do            

Do               

Do 

Do 

Do      

Do 

Average  — 

I.  218 

1-275 

97-76±o.9i 

.  610 

.678 

94.  74±i-S4 

Yield  per  row 

Do 

41-5 
42.5 
45-5 
53-5 
39-5 
44-5 
47-5 
48.0 
44.0 
41.  0 

45- 0 
46.0 

47-5 
48.0 

43- 0 
41.0 

S3-0 
46.  0 

47- S 
49-5 

95-95 
96.05 

97-85 
105.  42 

95-76 
104.  09 

94-53 

102.  13 

96.17 

90.  61 

30-  50 
29.50 
33-50 
34.00 
34.50 
32.00 
32-50 
31-25 

27.  00 
32-30 
32.00 
30.  CO 
30.00 
29-30 

28.  00 

32-00 

106.  I 

95-6 

102.3 

106.  7 

Do 

Do 

Do 

107.0 

104.7 

107.4 

98.8 



Do      

Do 

Do 

Do 

Do        

Average. .. . 

44-75 

46.65 

97.  87  ±1.02 

32.22 

30.06 

103.  5±i.  50 

The  yield  of  the  plants  from  the  Maryland-  and  Texas-grown  seed  is 
made  comparable  by  expressing  the  yield  of  the  former  as  a  percentage 
of  the  mean  yield  of  both.  For  example,  in  Table  I  where  the  behavior 
of  the  Stockton  X  Boone  hybrid  is  considered,  the  yield  per  plant  of  the 
first  row  at  Lanham,  Md.,  which  was  from  Maryland-grown  seed,  is  shown 
as  0.90  pound;  the  adjoining  row  from  Texas-grown  seed  yielded  1.15 
pounds  per  plant.  The  mean  of  the  yield  of  these  two  rows  is  102.5 
pounds,  and  the  yield  from  the  Maryland-grown  seed  is  87.8  per  cent  of 
this  mean,  the  value  given  in  the  fourth  column.  The  average  of  the  10 
comparisons  is,  in  this  case,  96.08  per  cent — that  is,  the  yield  of  the 
plants  from  the  Maryland-grown  seed  averages  3.92  per  cent  below  the 
mean  yield  of  this  strain  grown  in  Maryland.  This  expression  for  the 
relative  behavior  in  Maryland  of  the  plants  from  the  Maryland-grown  seed 
is  to  be  compared  with  the  results  of  the  similar  comparison  made  in 
Texas,  given  in  the  next  three  columns.  From  these  it  is  seen  that  in 
Texas  the  yield  per  plant  of  the  plants  from  Maryland-grown  seed  was 
98.25  per  cent  of  the  mean  yield  of  the  strain. 


Feb.  4,  1918 


New-Place  Effect  in  Maize 


237 


At  both  localties  the  Maryland-grown  seed  of  this  cross  is  inferior  to 
that  produced  in  Texas,  but  the  point  to  which  attention  is  now  directed 
is  that  the  inferiority  is  greater  in  Maryland  than  in  Texas. 

Table  II  contrasts  the  average  behavior  of  each  of  the  three  hybrids 
and  the  pure-seed  Boone  at  the  two  localities.  For  example,  when  com- 
pared in  Maryland,  the  yield  per  row  of  plants  from  the  Maryland-grown 
seed  of  Stockton  X  Boone  averaged  92.99  per  cent  of  the  mean  yield 
of  the  cross.  In  Texas  the  same  comparison  showed  the  average  yield 
of  plants  from  Maryland-grown  seed  to  be  98.76  per  cent  of  the  mean. 
Thus,  the  plants  from  Maryland-grown  seed  averaged  5.77  per  cent 
higher  in  yield  in  Texas  than  in  Maryland. 

Table  II. — Average  behavior  of  Maryland-grown  seed  expressed  as  a  percentage  of  the 
mean  of  Maryland-  and  Texas-grown  seed,  igij 


Factor  and  kind  of  seed. 

Compared  in 
Maryland. 

Compared  in 
Texas. 

Comparative  in- 

increase  of 

Maryland-grown 

seed  in  Texas. 

Increase 
divided 
by  prob- 
able error. 

Yield  per  plant: 
Stockton  X  Boone 

Per  cent. 

96.  08 ±2.  75 
105.  15  ±1.09 

98.  22  ±  I.  42 

97-76±o.  91 

92.  99±3-25 
105.  40±i.  80 
100.  25 ±  I.  43 

97.  87  ±  I.  02 

Per  cent. 
98.  25  ±0.86 
99- 65  ±1-32 

105.  3i±i.  05 

94-74±i-54 

98.  76  ±1.64 

106.  6lit:l.  82 

109.  55±i.  13 
103.  5o±i.  50 

Per  cent. 

2.  I7±2.  88 
-5.  5o±i.  71 

7.  09±i.  76 
—3.  02  — I.  69 

5.77±3-64 

I.  2I±2.  56 

10.  01  ±  I.  58 

5.  63  ±1.81 

Per  cent. 

0-75 
-3.22 

4-03 
—  I.  69 

1-59 

•47 

6.34 

311 

Strawberry  X  Boone 

Hickory  King  X  Boone .... 

Boone . . 

Yield  per  row: 
Stockton  X  Boone 

Strawberry  X  Boone 

Hickory  King  X  Boone.  . .  . 
Boone 

In  comparing  the  yields  both  the  yield  per  row  and  the  yield  per  plant 
were  considered.  In  Texas  the  yield  per  row  was  much  more  dependent 
on  the  number  of  plants  in  the  row  than  in  Maryland.  In  Texas,  as  the 
number  of  plants  increased,  there  was  a  pronounced  tendency  for  the  yield 
per  row  to  increase  and  the  yield  per  plant  to  decrease,  while  in  Maryland 
an  increase  in  the  number  of  plants  in  the  row  was  accompanied  by  only 
a  small  increase  in  the  yield  per  row,  and  there  was  an  almost  correspond- 
ing increase  in  the  yield  per  plant. 

This  difference  in  behavior  at  the  two  localties  is  not  difficult  to  under- 
stand. In  Maryland  the  failure  to  secure  a  perfect  stand  was  largely 
the  result  of  infertile  spots  in  the  field,  and  the  same  unfavorable  condi- 
tions which  reduced  the  number  of  plants  also  reduced  the  yield  of  those 
that  survived.  In  Texas,  the  loss  of  plants  was  more  the  result  of  acci- 
dental factors,  which  influenced  the  yield  of  the  remaining  plants  only 
by  permitting  them  to  take  advantage  of  the  increased  space  with  a 
consequent  slight  increase  in  the  yield  per  plant.  The  method  of  planting 
in  Texas  accentuated  this  difference,  for  with  the  plants  close  together 
in  the  row  the  additional  space  resulting  from  a  missing  plant  was 
utilized  more  by  the  neighboring  plants  in  the  same  row  than  by  those 
in  adjoining  rows.  In  Maryland,  on  the  other  hand,  where  the  plants 
were  spaced  equally  in  both  directions,  half  the  space  made  available 
by  a  missing  hill  would  be  appropriated  by  the  adjoining  rows. 


238 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  s 


Table  III. — Stand  of  plants  secured  from  Maryland  and  Texas  grown  seed  compared  in 

Maryland  and  Texas 


Variety. 


Compared  in  Maryland. 


Number 
of  plants 
from 
Mary- 
land- 
grown 
seed. 


Number 
of  plants 
from 
Texas- 
grown 
seed. 


Stockton  X  Boone. 
Do 


Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 


Average . 


Strawberry  X  Boone. 
Do 


Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 


Average . 


Hickory  King  X  Boone. 
Do 


Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 


Average . 


Boone. . 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 


Average. 


34 
36 
36 
27 

33 
15 
21 

31 
30 
2,2, 


Stand  of  Maryland- 
grown  seed  ex- 
pressed as  percent- 
age of  mean. 


27 
39 
35 
3(^ 
25 
21 

32 
33 
34 
31 


III.  5 

96.  o 

loi.  4 

85-7 
113.  8 

83-3 
79.2 
96.9 
93-8 
103. 1 


42 

37 

35 

40 

37 

35 

3& 

38 

34 

37 

39 

34 

34 
34 

37 
39 
40 
40 
38 
39 
38 


30 
37 
42 

41 
36 
Z2> 
39 
37 
38 
36 


36 
31 
30 
42 
?,3 
37 
39 
37 
36 
40 


33 
38 
35 
38 
38 
37 
35 
34 
39 
40 


96.  5  ±2.  50 


106.3 

93-3 
102.  8 

97-3 

95-8 

106.8 


Compared  in  Texas. 


Number 
of  plants 
from 
Mary- 
land- 
grown 
seed. 


50 
49 
52 
52 
52 
46 
54 
53 
54 
59 


100.  4±i.  86 


97.1 
104.  6 
104.  8 


93- 
108. 
103. 

lOI. 
lOI. 

104.  o 
97-4 


loi.  6zb. 


95-2 

98.7 

109.  I 

103.8 


97- 

94. 
105. 
104. 

98. 

94. 


100.  i±i.  24 


56 

59 
53 
49 
50 
50 
51 
54 
52 
54 


Number 
of  plants 
from 
Texas- 
grown 
seed. 


44 
48 
49 
52 
54 
45 
51 
58 
50 
47 


54 
49 
58 
54 
57 
46 

54 
50 


51 
49 
57 
47 
50 
53 
46 
59 
54 
50 


53 
42 
38 
41 
46 

44 
45 
51 
47 
53 


Stand  of  Mary- 
land-grown 
seed  expressed 
as  percentage 
of  mean. 


99.0 
100.  o 

95-4 
105.  I 
102 


92. 
108. 

94. 
100. 
108.3 


100.  5±i.  20 


102.  8 
116.8 
116.  5 
108.8 

104.  2 
106.  4 
106.3 
102.  9 

105.  I 


100.  9 


37 
40 

44 
48 

52 
46 

54 
42 

50 
48 


36 

44 
47 
45 
47 
47 
41 
51 


107.  i±i.  iJ 


108.6 
109.  I 
105.4 
104.  o 
loi.  9 
98.9 
97.1 
116.  o 
100.  o 

98.9 


104.  o±i.3i 


120.0 
105.4 
no.  5 
109.  I 
109.  6 
98.9 

113- 7 
99.0 


108.  3  ±  I.  72 


Feb.  4,  1918 


New-Place  Effect  in  Maize 


239 


Ability  to  produce  a  stand  may  legitimately  be  considered  one  of  the 
manifestations  of  greater  vigor.  That  it  is  a  definite  and  positive  factor 
is  shown  in  Tables  III  and  IV,  in  which  the  relative  stands  are  compared. 
With  all  four  kinds  a  comparison  of  the  relative  stand  at  the  two  local- 
ities is  in  favor  of  the  transferred  seed.  In  the  Boone  variety  the  transfer 
of  seed  has  resulted  in  an  8  per  cent  increase  of  stand,  a  difference 
nearly  four  times  the  probable  error.  Since  the  analysis  of  the  compar- 
ative stand  of  local  and  transferred  seed  shows  that  the  differences  are 
not  accidental,  but  are  consistently  in  favor  of  the  transferred  seed,  it 
would  seem  that  yield  per  row  is  a  more  reliable  measure  of  comparative 
vigor  than  yield  per  plant.  Yield  per  row  is  the  measure  of  the  practical 
results,  and  from  this  standpoint  it  is  seen  that  all  four  strains  showed 
an  increase  in  yield  as  a  result  of  transfer  of  seed.  In  Texas,  where 
there  was  a  definite  tendency  for  an  increased  number  of  plants  in  a  row 
to  reduce  the  yield  per  plant,  yield  per  plant  is  obviously  ill  calculated 
to  bring  out  the  real  difference  in  vigor. 

Table  IV. — Average  stand  of  plants  secured  from  Maryland-grown  seed  expressed  as  a 
percentage  of  the  mean  of  Maryland  and  Texas-grown  seed 


Kind  of  seed. 

Compared  in 
Maryland. 

Compared  in 
Texas. 

Increased 
stand  of  Mary- 
land-grown 
seed  in  Texas. 

Increase 
divided 
by  prob- 
able error. 

Stockton  X  Boone 

Per  cent. 
96.  5  ±2.  50 

100.  4±i.  86 
loi.  6±o.  98 
100.  i±i.  24 

Per  cent. 
100.  5±i.  20 

107.  i±i.  18 
104.  o±i.  31 

108.  3  ±  I.  72 

Per  cent. 
4-o±2.  77 
6.  7±2.  20 
2.  4±i-63 

8.  2±2.  12 

Per  cent. 
1.44 
3- 04 
1.47 
3-87 

Strawberry  X  Boone 

Hickory  King  X  Boone 

Boone 

EXPERIMENTS  IN  1915  AND  1916 

The  results  of  the  191 2  and  191 3  experiments  were  so  at  variance  with 
current  belief  that  it  was  thought  best  to  obtain  additional  evidence 
before  publishing.  A  somewhat  similar  experiment  was  therefore 
planned  and  carried  out  during  the  years  191 5  and  191 6.  The  same 
varieties  were  used  as  in  the  previous  experiment,  but  the  localities 
were  changed  by  substituting  Greenville,  Tex.,  and  Sacaton,  Ariz.,  for 
Victoria,  Tex.,  and  Stockton,  Kans. 

Crop  failure  at  Greenville  again  limited  the  experiment  to  two  locali- 
ties: Lanham,  Md.,  and  Sacaton,  Ariz.  At  Sacaton  the  temperatures 
are  high,  and  there  is  practically  no  rainfall  during  the  growing  season, 
the  crop  being  grown  by  means  of  irrigation. 

To  eliminate  differences  due  to  irregularities  in  the  stand  of  plants,  a 
different  system  of  planting  was  adopted.  Seed  from  both  localities 
were  planted  in  each  hill,  the  seed  from  the  two  sources  being  identified 
by  their  positions  in  the  hill.  At  harvest  the  measurements  were  con- 
fined to  the  hills  which  contained  plants  from  both  Maryland-  and  Ari- 


240 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  s 


zona-grown  seed.  For  all  such  hills  the  height  of  each  plant  was  re- 
corded with  the  total  length  of  the  ear  or  ears.  In  this  way  each  hill 
constituted  a  unit  of  comparison.  The  height  of  each  plant  was  ex- 
pressed as  a  percentage  of  the  mean  height  of  the  two  plants  in  the 
same  hill.  These  determinations  were  then  averaged  to  secure  an 
expression  of  the  mean  behavior  of  the  plants  from  each  source  of  seed. 
Length  of  ear  was  treated  in  the  same  way.  Table  V  gives  the  results. 
Unfavorable  conditions  so  reduced  the  yields  at  Lanham,  Md.,  that  length 
of  ear  was  recorded  for  only  three  of  the  strains,  and  even  for  these 
there  was  so  much  variation  that  the  results  are  of  doubtful  significance. 
They  serve,  however,  to  supplement  the  results  on  the  height,  with  which 
they  are  in  accord. 

Table  V. — Average  behavior  of  Maryland-grown  seed  expressed  as  a  percentage  of  the 
mean  of  Maryland-  and  Arizona-grown  seed,  igi6 


Kind  of  seed. 

Compared 

in 
Maryland. 

Compared 

in 
Arizona. 

Increase  of 
Maryland- 
grown  seed 
in 
Arizona. 

In- 
crease 

Factor. 

Produced  in  Mary- 
land, 1915. 

Produced  in  Arizona, 

1915- 

divided 

by 
prob- 
able 
error. 

Height  of 
plants. 
Do... 
Do.... 
Do... 
Do.... 

Do 

Stockton  X  Boone 

Boone  X  Stockton 

Strawberry X Boone.  . 
BooneXStrawberry.  . 
Hickory       King    X 

Boone. 
BooneXSelf 

StocktoaXBoone. . . . 

do 

Strawberry  X  Boone . 

do 

Hickory      King    X 

Boone. 
BooneX  Boone 

StocktonX  Boone.... 

Hickory      King   X 

Boone. 
BooneX  Boone 

Per  cent. 

92.9±I-2I 

9a.7io.85 
89-2±i.74 
9S-7ii-44 
103.  s±o.  53 

108. 1  i  0.90 

88.8i7-6 

iia.4i2. 2 

146. 3  ±6.0 

Per  cent. 
98. 9±  0.38 

99- 7  i  0.43 
97- 1  ±0-58 
IOC.  9  ±0.46 
97- si 0.46 

109.  o±o.  s6 

I03.9i2.  7 

I05.ii3.3 

iS9-4i2-9 

Per  cent. 
6.  oil.  27 

7-oio.9S 

7-9ii-83 

5-2±i.50 

— 6.  o±o.  70 

o.9±i.o6 

IS- lis.  1 

— 7-3i4-o 

13- lis- 7 

Per  ct. 
4- 7a 

7-37 
4- 3a 

3-47 
-8.57 

0.85 

Length  of 
ear. 
Do.... 

Do 

BooneXStockton 

Hickory       King    X 

Boone. 
BooneXSelf 

1.9 
1.8 
2.3 

At  Sacaton,  Ariz.,  in  191 5  reciprocal  crosses  were  made  with  Stock- 
ton X  Boone  and  Strawberry  X  Boone,  and  these  reciprocals  were  sepa- 
rately compared  with  the  seed  grown  at  Lanham,  where  Boone  was 
used  only  as  the  male  parent.  In  all  of  the  six  comparisons  except  one, 
transferring  the  seed  resulted  in  increased  height ;  and  in  all  but  one  the 
difference  is  almost  certainly  not  the  result  of  chance. 

In  1915  the  crosses  in  both  Maryland  and  Arizona  were  made  by  hand 
instead  of  by  detasseling  alternate  rows  as  in  191 2.  In  gathering  pollen 
an  effort  was  made  to  obtain  pollen  from  as  many  plants  as  possible  and, 
so  nearly  as  might  be,  in  equal  amounts  from  each  plant.  In  like  manner 
selection  of  female  parents  was  avoided  so  far  as  possible.  In  spite  of 
these  precautions,  it  is  evident  that  there  would  still  be  a  measure  of 
selection.  Some  plants  produce  virtually  no  pollen,  and  many  plants 
fail  to  develop  an  ear.  Furthermore,  since  the  plants  were  thinned  to  a 
stand  of  one  in  a  hill  from  each  locality,  obviously  weak  plants  being 
removed,  it  would  seem  that  here  too  there  would  be  a  tendency  to  retain 


Feb.  4.1918  New-Place  Effect  in  Maize  241 

the  types  of  plants  best  adapted  to  the  conditions  where  the  experiment 
was  tried.  The  entire  effect  of  selection  would  be  to  favor  the  home- 
grown seed,  and  that  the  transferred  seed  was  not  superior  to  the  home- 
grown in  every  instance  may  not  be  held  to  vitiate  the  cases  in  which 
significant  differences  in  favor  of  the  transferred  seed  were  observed. 

The  results  indicate,  however,  that  the  stimulation  is  more  pronounced 
in  some  stocks  than  in  others.  Thus,  in  the  191 6  comparisons  Boone  X 
Hickory  King  stands  out  as  a  conspicuous  exception.  In  all  other  stocks 
the  transferred  seed  produced  taller  plants  than  the  home-grown  seed; 
but  with  Boone  X  Hickory  King,  the  home-grown  seed  exceeded  the 
transferred  by  6.2  per  cent,  a  difference  not  to  be  ascribed  to  chance, 
being  more  than  eight  times  the  probable  error.  Of  the  three  stocks 
in  which  the  yield  was  taken,  Boone  X  Hickory  King  is  also  the  only 
one  to  show  superiority  for  the  home-grown  seed.  Taken  alone,  the  dif- 
ferences in  yield  could  not  be  considered  significant,  but  the  agreement 
with  the  results  for  height  confirms  the  reliability  of  these  results. 

The  insignificant  increase  in  the  case  of  Boone  may  be  explained  by 
the  fact  that  the  Arizona-grown  seed  was  more  closely  selected  to  fit 
the  Arizona  conditions  than  were  the  other  kinds.  At  Sacaton  the  pure- 
seed  Boone  was  obtained  by  selling.  This  procedure  would  restrict  the 
plants  from  which  seed  was  secured  to  those  able  to  produce  both  ears 
and  pollen  under  Arizona  conditions.  With  cross-pollinated  seed  and 
hybrids,  plants  that  produced  no  ears  would  be  represented  in  the  progeny 

as  male  parents. 

DISCUSSION  OF  RESULTS 

Three  classes  or  degrees  of  new-place  effects  ("neotopism")  have  been 
recognized  by  Cook  S  chiefly  with  reference  to  cotton:  (i)  Those  in  which 
there  is  merely  a  stimulation  of  growth;  (2)  those  in  which  there  is  also 
a  definite  general  change  of  the  hereditary  characteristics  of  the  variety; 
and  (3)  those  in  which  the  new  conditions  call  forth  a  promiscuous 
mutative  diversity. 

The  results  here  reported  give  evidence  under  the  first  of  these  cate- 
gories only.  With  respect  to  the  more  pronounced  changes  that  fol- 
low the  transfer  of  varieties  from  the  Tropics  to  a  temperate  climate 
it  may  be  said  that  many  such  changes  do  occur  in  maize,  some  of  which 
at  least  are  inherited.  For  conclusive  qualitative  evidence  on  this  point, 
however,  there  is  lacking  definite  information  regarding  the  behavior 
of  the  introduced  varieties  in  their  native  countries. 

Roberts-  has  pointed  out  that  the  striking  effects  which  have  been 
ascribed  to  acclimatization  in  maize  are  to  be  referred  either  to  cross- 
pollination  with  native  varieties  or  to  the  results  of  selection.  The  effect 
of  cross-pollination,  which  misled  early  investigators,  has  presumably 

'  Cook,  O.  F.    aspects  op  kinetic  evolution.    In  Proc.  Wash.  Acad.  Sci..  v.  8.  1906,  p.  336.    1907. 
2  Roberts.  H.  F.    acclimatization  with  reference  to  corn  breeding.    In  ist  Aon.  Rpt.  Kans. 
Com  Breeders'  Assoc,  [i9os]/o6,  p.  60-64.    1906. 


242  Journal  of  Agricultural  Research  voi.  xii,  no.  s 

been  eliminated  from  recent  experiments,  but  the  effects  of  selection 
are  so  pronounced  and  speedy  that  in  experiments  hitherto  reported 
any  direct  effect  of  the  environment  on  the  characters  of  the  plants 
would  be  completely  masked.  The  characteristics  of  a  maize  variety 
are  altered  readily  by  selection.  When  grown  in  a  new  locality  for  a 
few  years,  even  without  conscious  selection,  the  type  may  change  rapidly; 
and  when  brought  back  to  the  original  locality,  it  is  in  reality  a  differ- 
ent variety.  The  characters  brought  into  prominence  in  the  new  locality 
may  render  the  stock  less  suited  to  the  old  conditions,  though  better 
adapted  to  the  new. 

It  would  be  very  difficult,  if  not  impossible,  to  eliminate  completely 
all  selective  action.  Even  when  all  seed  is  saved,  those  individuals  or 
types  of  plants  which  are  best  adapted  to  the  conditions  under  which 
they  are  grown  will  produce  a  greater  proportion  of  the  seed  than  will 
the  types  which  are  less  well  adapted,  and  those  least  adapted  may 
produce  no  seed  at  all.  In  two  localities  where  different  conditions 
prevail,  the  highest  yielding  plants — hence  those  contributing  the  largest 
proportion  of  the  seed — would  presumably  be  of  different  types;  and 
when  brought  together  and  compared,  we  shou^ld  expect  to  find  a  slight 
advantage  for  the  locally  grown  seed.  Yet  the  results  of  the  present 
experiments  indicate  that  the  effect  of  selection  during  a  single  season 
may  be  so  slight  as  not  to  mask  completely  the  opposing  new-place  effect. 

Since  new-place  effect  in  maize  seems  to  operate  as  a  stimulus,  it  would 
tend  to  obscure  any  lack  of  adaptation  in  newly  introduced  varieties. 
The  recognition  of  new-place  effect  may  be  said,  therefore,  to  increase 
rather  than  diminish  the  importance  that  must  be  assigned  to  adaptation. 

As  a  result  of  the  stimulation  due  to  new-place  effect,  the  cultivation 
of  an  inferior  strain  might  be  extended  as  a  result  of  its  satisfactory  per- 
formance the  first  year  following  its  introduction. 

The  stimulus  that  followed  the  transfer  of  seed  in  these  experiments  is 
doubtless  similar  to  the  increased  vigor  imparted  to  many  vegetables 
by  growing  the  crop  in  localities  remote  from  the  place  where  the  seed 
was  produced.  The  economic  utilization  of  increased  vigor  secured  in 
this  way  is  usually  confined  to  crops  which  are  grown  for  the  sake  of  some 
part  other  than  the  seed.  In  cotton,  for  example,  the  increase  of  vigor 
in  the  plants  following  a  transfer  of  seed  is  often  very  pronounced, 
although  the  crop  of  seed  and  fiber  may  be  reduced.  In  maize,  as  a 
result  of  the  determinate  habit  of  the  plant,  vegetative  vigor  and  seed 
production  are  more  closely  associated,  so  that  the  possibility  of  prac- 
tical utilization  seems  greater. 

CONCLUSIONS 

Hybrids  between  the  same  pairs  of  varieties  made  at  different  local- 
ities showed  no  decrease  in  yield  as  a  result  of  transferring  the  first- 
generation  seed  to  a  new  locality.     On  the  contrary,  the  change  of 


Feb.  4, 1918  New-Place  Effect  in  Maize  243 

environment  seemed  to  act  as  a  stimulus,  with  the  result  that  the  yields 
were  increased  in  all  but  one  of  the  hybrids  tested.  One  unhybridized 
variety  was  included  in  the  experiment,  and  this  also  gave  slightly 
increased  yields  as  a  result  of  being  transferred  to  a  new  environment. 

In  6  of  the  10  ^comparisons  the  increase  is  too  large  to  be  ascribed  to 
experimental  error  and  indicates  that  new-place  effect  should  be  taken 
into  consideration  as  a  factor  of  production. 

That  significant  increases  may  be  secured  by  taking  advantage  of 
new-place  effect  in  maize  should  not  be  used  as  an  argument  in  favor  of 
the  general  transfer  of  seed.  There  is  no  evidence  that  the  importance 
of  using  acclimatized  seed  has  been  overestimated.  On  the  contrary, 
the  experiments  show  that  new-place  effect  may  often  obscure  the  differ- 
ences between  acclimatized  and  unacclimatized  seed  when  first  com- 
pared, and  thus  interfere  with  a  full  appreciation  of  the  value  of  adap- 
tation. 

The  investigations  show  the  existence  of  a  hitherto-neglected  factor 
in  maize  production,  but  much  more  extensive  experiments  are  needed 
to  ascertain  the  extent  and  practical  importance  of  this  factor.  The 
existence  of  one  definite  exception  indicates  that  the  tendency  to  increased 
vigor  following  a  transfer  of  seed  is  not  universal.  The  results  also 
indicate  that  adaptation  in  maize  comes  about  through  selection  rather 
than  as  a  direct  reaction  to  the  environmental  conditions. 


RELATION  OF  THE  VARIABILITY  OF  YIELDS  OF  FRUIT 
TREES  TO  THE  ACCURACY  OF  FIELD  TRIALS^ 

By  L.  D.  Batchelor,  Professor  of  Plant  Breeding,  and  H.  S.  ReED,  Professor  of  Plant 
Physiology,  University  of  California,  Citrus  Experiment  Station.- 

INTRODUCTION 

The  value  of  the  outcome  of  any  trial  depends  upon  the  probability 
that  a  similar  result  will  be  obtained  if  the  trial  is  repeated.  In  recent 
years  the  agricultural  experiment  stations  of  all  countries  have  greatly 
increased  the  number  and  size  of  their  field  trials.  A  casual  examination 
of  such  trials  usually  shows  a  wide  range  in  the  reliability  of  the  results. 
The  yields  of  control  plots  in  different  parts  of  the  same  tract  will  often 
differ  as  much  among  themselves  as  the  yields  of  fertilized  and  un- 
fertilized plots  differ  from  each  other.  The  purpose  of  this  paper  is  to 
present  the  results  of  a  study  of  the  variation  in  recorded  yields  of  fruit 
trees  taken  singly  and  in  groups  of  various  sizes,  and  especially  to  deter- 
mine the  effect  upon  variability  of  various  combinations  and  repetitions 
of  unit  plots. 

It  is  recognized  that  the  results  of  a  single  experiment  are  often 
untrustworthy ;  yet  experimenters  have  published  single  results  and  have 
based  practical  advice  upon  them.  However  well  planned  field  trials 
may  be,  the  interpretations  of  the  results  can  hardly  be  considered  of 
infinite  reliability — that  is,  results  which  will  invariably  be  obtained 
when  the  trials  are  repeated.  The  best  that  can  be  done  is  to  construct 
the  most  probable  results  from  the  more  or  less  varying  observed  results 
of  individual  trials.  If  it  is  impossible  to  obtain  perfect  accuracy,  it  is 
not  impossible  to  fix  the  limits  of  error  and  thus  to  determine  whether  the 
differences  obtained  are  due  to  the  treatments  appUed  or  to  unavoid- 
able errors — that  is,  whether  the  differences  are  significant.  A  number 
of  trials  are  necessary  before  a  reasonably  reliable  result  can  be  obtained. 
The  average  result  of  a  series  does  not  always  represent  the  truth. 
When  averages  are  used,  they  should  always  be  accompanied  by  their 
probable  errors,  which  are  a  measure  of  their  reliability. 

Before  attempting  to  interpret  the  results  of  a  plot  experiment  with 
fertiUzers,  it  is  necessary  to  know  whether  the  differences  observed  are 
any  greater  than  those  which  might  have  occurred  had  none  of  the  plots 
been  fertilized.  The  scientific  method  is  to  recognize  the  inevitable  error 
and,  while  reducing  it  by  every  possible  precaution,  at  the  same  time  to 

>  Paper  44,  University  of  California,  Citrus  Experiment  Station,  Riverside.  Cal. 

'  The  writers  wish  to  make  acknowledgment  of  their  indebtedness  for  aid  and  criticism  to  Prof.  H.  H. 
Love,  of  Cornell  University;  Prof.  E.  B.  Babcock,  Dr.  H.  B.  Frost,  and  other  members  of  the  University 
of  California. 

Journal  of  Agricultural  Research,  Vol.   XII,  No.  5 

Washington,  D.  C.  Feb.  4,  1918 

Iw  Key  No.  Cal.— 14 

(24s) 


246  Journal  of  Agricultural  Research  voi.  xn,  No.  $ 

measure  its  probable  amount  so  as  to  make  sure  it  is  not  likely  to  vitiate 
our  conclusions. 

One  of  the  chief  difficulties  in  obtaining  reliable  results  in  field  trials 
is  the  natural  variability  of  the  material  with  which  we  are  deaUng. 
Crops  are  living  organisms  with  inherent  tendencies  to  vary,  even  though 
it  were  possible  to  make  their  environmental  conditions  identical. 

In  agronomic  experiments  the  number  of  plants  taken  is  usually  so  large 
that  inherent  variability  ceases  to  be  a  factor  of  importance.  In  horti- 
cultural experiments,  however,  where  fruit  trees  are  under  observation, 
the  limited  number  of  trees  possible  to  include  in  a  plot  may  make  the 
factor  of  inherent  variability  an  important  consideration. 

Further  variation  is  induced  as  a  result  of  the  many  factors  of  the 
environment  which  are  beyond  the  control  and  possibly  the  recognition 
of  the  experimenter.  Some  of  these  factors  are  independent;  others 
react  upon  one  another.  In  designing  a  set  of  field  trials,  we  try  to 
avoid,  as  far  as  possible,  all  secondary  factors  which  may  exert  a 
disturbing  influence. 

Lack  of  uniformity  in  both  the  physical  and  chemical  characteristics 
of  the  soil  is  one  of  the  foremost  factors  causing  variation  in  productivity 
of  plants.  Apparently  uniform  surface  soils  may  be  underlain  with  a 
heterogeneous  subsoil.  Differences  also  occur  which  are  not  evident  on 
a  careful  inspection  of  both  the  soil  and  the  crops,  but  which  are  easily 
measured  by  weighing  the  yields.  In  other  words,  the  weighing  machine 
is  more  sensitive  than  the  eye  and  reveals  differences  that  mere  inspection 
can  not  detect. 

The  past  treatment  of  the  soil  brings  in  variables  the  significance  of 
which  may  not  be  comprehended  at  the  time  a  field  trial  is  begun.  The 
persistent  effects  left  by  the  application  of  stable  manure  on  some  of  the 
Rothamsted  plots  show  how  large  a  part  is  played  by  the  past  history  of 
the  field.  Plots  which  for  40  years  have  had  identical  treatment  still 
give  different  crop  yields  because  of  the  effect  of  dressings  of  barnyard 
manure  applied  at  an  earlier  period. 

Unequal  prevalence  of  diseases  and  insects  may  bring  about  further 
error  in  the  results. 

Besides  the  above  sources  of  variation  and  possibly  outweighing  them 
at  times  is  the  effect  of  season.  No  season  is  entirely  normal;  and  it  is 
only  when  the  experiment  has  been  repeated  for  several  years,  or,  in  other 
words,  until  "a  fair  sample  of  seasons"  has  been  made,  that  any  sort  of 
allowance  can  be  made  for  seasonal  effects. 

As  is  shown  by  observations  at  Rot  hamsted,  it  is  not  possible  to  estab- 
lish a  schedule  of  relative  jdelds  for  a  series  of  plots,  even  after  several 
years'  comparison.  In  this  case  two  grass  plots  were  treated  alike  for 
50  years;  by  taking  the  yield  of  one  plot  as  the  standard,  the  yield  on 
the  other  in  the  same  season  has  been  as  low  as  90  per  cent  and  as  high  as 
196  per  cent. 


Feb.4,i9i8     Variability  of  Yields  of  Fruit  Trees  and  Field  Trials     247 

Recent  investigations,  which  are  reviewed  in  succeeding  pages,  have 
thrown  light  upon  the  nature  of  the  variability  inherent  in  experiments 
conducted  with  groups  of  plants  or  of  animals.  So  far  as  the  writers 
know,  however,  few  studies  have  yet  been  published  upon  the  uncer- 
tainty with  which  one  deals  in  attempting  to  carry  on  plot  trials  with 
orchard  trees.  In  connection  with  plans  for  extending  the  trial  plots  of 
this  Experiment  Station,  the  writers  have  attempted  to  study  the  ques- 
tion of  variability  of  tree  yields  and  to  formulate  some  plan  to  determine 
the  probable  reliability  of  the  results.  It  is  highly  desirable  to  make 
the  tree  plots  as  small  as  possible  without  sacrificing  too  much  accuracy 
on  account  of  the  large  amount  of  land  required  for  each  individual. 

A  fruit  tree  is  possibly  more  affected  by  environmental  conditions  than 
an  annual  crop  growing  from  seed  to  maturity  in  one  season.  The  tree 
roots  penetrate  more  deeply  into  the  soil  and  may  be  affected  by  soil 
differences  to  a  considerable  depth.  Climatic  conditions  during  the  rest- 
ing period  may  have  marked  influence  on  the  crop  production  of  the  trees. 

Pruning  also  introduces  variation.  Uniform  pruning  is  desirable; 
yet  different  trees  need  different  types  of  pruning,  and  more  uniform 
results  will  be  obtained  if  this  is  recognized. 

Individual  trees  have  apparent  idiosyncrasies  in  fruiting  habits.  A 
tree  may  yield  large  crops  of  fruit  in  alternate  years,  with  very  light 
crops  in  intervening  years.  The  character  of  the  stock  upon  which  the 
particular  variety  was  budded  to  form  the  tree  may  profoundly  influence 
the  type,  habit,  and  productivity  of  the  adult  tree.  Furthermore,  we 
can  not  overlook  the  possibility  of  errors  in  yields  due  to  predacious 
animals  which  devour  or  otherwise  destroy  more  or  less  fruit. 
Fruit  trees  therefore  present  opportunity  for  more  variability  than 
would  be  expected  in  the  growth  of  annual  plants. 

PREVIOUS  STUDIES 

The  varying  productivity  of  fruit  trees  has  been  called  to  our  attention 
by  the  writings  of  Macoun  (1904),^  Munson  (ipoy),  Shamel  (1912), 
Hedrick  (1912),  Stewart  (191J,  p.  552-554),  Fletcher  {1913),  Coit  (Jpjo), 
Whitten  (19 15),  Lewis  and  Vickers  (19 15,  p.  30-31),  Barre  {1915), 
Gourley  {19 15,  p-  72-73),  Kraus  (19 16),  and  others.  The  above  observa- 
tions have  been  largely  made  in  connection  with  plant  breeding  and 
orchard  management  problems,  and  were  not  made  to  bear  necessarily 
on  the  accuracy  of  plot  trials. 

Holtsmark  and  Larsen  {1906)  were  among  the  first  to  call  attention  to 
the  errors  of  field  trials.  They  recognized  the  inevitable  variation  of 
field  results,  and  showed  how  it  may  be  estimated  by  the  use  of  the 
standard  deviation  and  the  coeflficient  of  variability.     They  also  showed 

1  Bibliographic  citations  in  parentheses  refer  to  "Literature  cited",  pp.  382-283. 
27808°— 18 2 


248  Journal  of  Agricultural  Research  voi.  xii.  No.  s 

that  the  coefficient  of  variability  decreases  as  the  plots  are  enlarged,  but 
not  proportionally  to  the  size  of  the  plot. 

The  limitations  of  field  experiments  were  discussed  by  Carleton  (ic^og), 
who  called  the  attention  of  experimenters  to  the  various  uses  of  control 
plots,  and  to  the  general  precautions  necessary  to  obtain  reliable  results. 

Hall  {igog),  Mercer  and  Hall  (igii),  and  Hall  and  Russell  (igii) 
recorded  extensive  studies  of  the  soil  variations  in  experimental  grounds 
and  the  influence  of  size  and  repetition  of  plots  upon  accuracy.  This 
work  was  done  largely  with  the  yields  of  wheat,  mangel,  and  hay  crops. 
The  conclusions  from  the  above  work  are  that  the  error  in  field  trials 
diminishes  as  the  size  of  the  plot  increases,  but  that  the  reduction  is 
small  when  the  plot  is  enlarged  to  a  size  greater  than  one-fortieth  of  an 
acre.  The  error  may  be  further  diminished  by  increasing  the  number  of 
plots  similarly  treated  and  scattering  them  about  the  area  under  experi- 
ment; but  there  is  not  much  to  be  gained  by  increasing  the  number  of 
plots  above  five. 

Wood  and  Stratton  (igio)  sounded  notes  of  caution  concerning  the 
interpretation  of  experimental  results.  Frequency  distribution  is  dis- 
cussed from  the  point  of  view  of  its  bearing  on  the  reliability  of  averag- 
ing results.  The  applications  of  the  probable-error  methods  to  questions 
of  sampling  for  analysis,  to  field  experiments,  and  to  feeding  experi- 
ments are  illustrated.  The  probable  error  of  field  experiments  was 
investigated  by  two  independent  methods  and  found  to  be  about  5 
per  cent  of  the  mean  yield.  Tables  are  given  showing  the  number  of 
duplicate  plots  or  number  of  animals  in  a  feeding  trial  which  must  be 
employed  to  give  any  desired  precision  in  the  result.  It  is  shown  that 
more  accurate  results  may  be  obtained  by  employing  large  numbers  of 
small  scattered  plots  than  by  using  one  large  plot. 

The  estimation  of  errors  in  field-plot  tests  has  been  given  consid- 
erable attention  by  Lyon  {igi2)  and  coworkers.  It  was  shown  that  it 
is  not  possible  to  establish  a  schedule  of  relative  yields  for  a  series  of 
plots,  even  after  several  years'  comparison.  Also,  there  seems  to  be 
little  gain  by  using  plots  larger  than  one-fiftieth  of  an  acre  in  size  when 
the  comparative  yield  of  the  crops  is  made  the  criterion.  An  area  of 
one-twenty-fifth  of  an  acre  of  land  distributed  in  four  widely  separated 
plots,  devoted  to  any  one  test,  secures  a  much  greater  degree  of  accuracy 
than  the  same  area  of  land  in  one  body.  The  probable  error  was  reduced 
from  4.5  to  2  per  cent  by  such  distribution. 

Pickering  (igii),  from  studies  on  apples  and  pears,  concluded  that 
experimental  plots  should  include  6  to  12  fruit  trees.  Precautionary 
advice  was  also  given  concerning  the  measurement  of  results  b)^  crop 
production,  foliage,  and  tree  characteristics.  In  comparing  the  results 
on  the  treated  plots  with  the  controls,  instead  of  taking  the  average  of 
the  controls,  he  prefers  to  plot  these  results  out  and  to  draw  a  smoothed 


Feb.4.i9i8     Variability  of  Yields  of  Fruit  Trees  and  Field  Trials     249 

curv^e  through  them,  and  then  to  compare  the  results  of  the  experimental 
plots  with  readings  taken  at  corresponding  points  of  this  curve. 

Wood  {191 1)  showed  how  the  degree  of  reliance  can  be  determined  for 
any  set  of  experimental  results  by  the  use  of  the  probable  error.  The 
use  of  this  constant  was  demonstrated  in  interpreting  laboratory  analy- 
ses, as  well  as  both  plot  and  feeding  experiments.  Working  with  mangel 
yields,  the  author  calculates  the  number  and  size  of  plots  required  to  at- 
tain any  desired  precision,  and  working  with  the  probabl  eerror  of  live- 
weight  increase  of  sheep,  tables  are  given  showing  the  number  of  animals 
required  in  an  experiment  to  attain  various  degrees  of  reliability. 

Several  papers  by  Harris  {1912,  1913a,  19131),  and  1915)  have  drawn 
our  attention  to  several  phases  of  the  experimental  error  in  field  tests. 
A  measure  of  the  variability  of  the  soil  productivity  is  obtained  by 
determining  the  correlation  between  the  yields  of  ultimate  small  plots 
and  the  yields  of  various  groups  of  adjacent  plots.  The  more  nearly 
this  correlation  approaches  zero,  the  more  homogeneous  the  soil.  This 
method  of  measurement  does  not  seem  to  provide  as  definite  a  means 
of  obtaining  a  corrective  term  as  the  use  of  the  coefficient  of  variability 
and  the  probable  error  as  used  by  Wood,  Wood  and  Stratton,  Mercer 
and  Hall,  etc.,  or  the  contingency  method  of  correction  as  used  by 
Surface  and  Pearl  (1916). 

Montgomery  (1912)  has  also  discussed  the  comparative  variabiHty 
resulting  from  increasing  the  size  of  the  plot  and  from  distributing  small 
ultimate  plots  over  the  area.  The  latter  method  was  found  to  be  more 
accurate.  In  a  subsequent  paper  {1913)  the  relative  reliabihty  of  yields 
of  wheat  planted  in  rows  and  in  square  blocks  is  discussed. 

An  exhaustive  and  discriminating  discussion  of  the  nature  and  magni- 
tude of  variability  in  the  results  of  feeding  experiments  has  been  given 
by  Mitchell  and  Grindley  {191 3).  Much  of  their  discussion  is  equally 
applicable  to  experimentation  with  plants. 

Olmstead  (19 14)  applied  the  method  of  least  squares  in  calculating 
the  reliability  of  the  yields  of  the  mangel  and  wheat  crop  records  of 
Mercer  and  Hall,  the  potato  records  of  Lyon,  and  the  wheat  yields  of 
Montgomery.     The  conclusions  from  this  series  of  observations  are: 

The  estimation  of  the  probable  error  of  a  large  number  of  small  duplicate  plots  well 
distributed  in  the  area  devoted  to  a  field  experiment  indicates  that  the  precision  of 
agricultural  experiments  can  be  increased  by  replicating  the  experiments  on  small 
plats. 

Coombs  and  Grantham  (1916)  have  studied  the  variation  in  the 
yields  of  rice  and  coconuts  for  one  year,  and  discussed  the  range  and 
interpretation  of  the  probable  error.  They  showed  that  the  yields  from 
any  two  single  plots  could  only  be  significant  when  the  difference 
amounted  to  22.8  per  cent  of  the  mean.  They  also  introduced  calcula- 
tions to  show  the  odds  that  any  increase  is  a  real  increase  and  not  a 
probable  error. 


250  Journal  of  Agricultural  Research  voi.  xii,  No.  s 

The  use  of  controls  and  repeated  plantings  in  varietal  tests  was  studied 
by  Pritchard  {1916)  in  breeding  work  with  sugar  beets.  His  studies  lead 
to  the  conclusion  that  the  practice  of  dispensing  with  control  rows  and 
using  the  mean  of  all  progeny  rows  as  a  standard  of  comparison  appears 
to  be  less  accurate  than  the  employment  of  frequent  controls.  However, 
the  employment  of  every  alternate  row  as  a  control  was  not  sufficient  to 
offset  the  variability  in  yield  arising  from  irregularities  of  soil. 

Stockberger  {1916)  discussed  the  value  of  a  number  of  the  common 
methods  for  determining  the  normal  yield  of  treated  plots  based  upon 
the  yields  of  hops.  Normal  yields  for  various  plots  varied  widely  accord- 
ing to  the  method  of  computation,  the  values  in  some  cases  differing  from 
the  actual  yield  by  as  much  as  40  per  cent.  Repetition  brought  about 
a  very  marked  reduction  in  variability,  although  with  only  five  repetitions 
the  error  is  still  relatively  large. 

The  work  of  Surface  and  Pearl  {1916)  shows  an  advance  in  the  refine- 
ment of  methods  of  conducting  field  trials.  With  the  realization  that 
the  use  of  frequent  control  plots  often  produces  results  far  from  satis- 
factory, these  workers  have  calculated  by  the  contingency  method  the 
probable  yield  of  each  plot  of  ground  in  their  grain-testing  series.  This 
calculated  yield  represents  the  most  probable  yield  of  each  plot  on  the 
supposition  that  they  have  all  been  planted  with  a  h)'pothetical  variety 
whose  mean  yield  is  the  same  as  the  observed  mean  of  the  field.  This 
"calculated"  yield  may  then  be  used  as  a  basis  for  determining  a  cor- 
rection factor,  whereby  each  area  must  be  given  a  handicap  plus  or 
minus  the  actual  yield,  depending  upon  whether  the  plot  in  question  is 
calculated  to  be  a  low-  or  a  high-producing  area.  This  method  of  cor- 
recting the  soil  variation  is  combined  with  four  systematically  repeated 
plots  of  one-fortieth  acre  of  each  variety,  and  gives  a  high  degree  of 

accuracy. 

MATERIAL  USED  FOR  STUDY 

The  studies  to  be  reported  in  this  paper  deal  with  the  variability  of 
fruit-tree  yields.  They  are  based  upon  the  individual  tree  yields  of 
oranges  {Citrus  sitiensis),  lemons  (Citrus  limonia) ,  walnuts  (Juglans  regia), 
and  apples  {Malus  sylvestris)  from  orchards  which  had  received  uniform 
treatments  for  a  number  of  years — indeed,  so  far  as  known,  from  the 
time  of  planting  the  trees.  The  orchards  were  carefully  examined,  and 
the  records  for  all  trees  which  were  known  to  be  abnormal  from  disease 
or  other  apparent  causes  were  eUminated.  In  place  of  the  records  of 
trees  thus  eHminated  the  average  yield  of  the  eight  surrounding  trees 
was  substituted  (assuming  that  the  tree  stood  at  the  center  of  a  square 
block  of  nine  trees). 

This  substitution  is  not  entirely  satisfactory,  yet  it  was  felt  that  it 
was  necessary  in  order  to  compute  plots  of  homologous  size  and  syste- 
matic arrangement.  The  writers  found,  as  a  matter  of  fact,  that  there 
is  a  very  high  degree  of  correlation  in  these  orchards  between  the  yields 


Feb.4,i9is     Variability  of  Yields  of  Fruit  Trees  and  Field  Trials     251 

of  individual  trees  and  the  average  of  surrounding  trees.  For  example, 
they  found  coefficients  of  correlation  as  high  as  0.652  ±0.065  for  the 
Eureka  lemons  and  0.628  ±0.060  for  the  Arlington  navels.  According 
to  the  formula  given  by  Harris  (1915),  the  correlation  between  individual 
trees  and  eight  adjacent  trees  in  a  plot  of  the  Arlington  navels  is 
0.576  ±0.04.  In  view  of  these  results,  the  writers  felt  justified  in  using 
this  method  of  substituted  values. 

The  fruit  plantations  herein  discussed,  to  judge  by  the  surface  soil, 
size,  and  condition  of  the  trees,  as  well  as  their  apparent  fruitfulness, 
appeal  to  the  observer  as  uncommonly  uniform.  All  the  orchards 
studied  are  situated  in  semiarid  regions  and  are  artificially  irrigated 
during  the  summer  months.  This  fact  is  believed  to  be  a  distinct  advan- 
tage for  the  purpose  of  reducing  the  variability  of  one  year's  yield  com- 
pared with  another,  since  it  insures  a  fairly  uniform  water  supply  for  the 
soil  and  reduces  one  of  the  variants  inevitable  in  nonirrigated  localities. 

All  yields  of  the  several  fruit  and  nut  plantations  are  given  in  pounds 
per  tree  of  the  ungraded  product. 

DESCRIPTION   OP  THE  PLANTATIONS 

Navel  orange  (Arlington). — These  records  were  of  the  191 5-1 6 
yields  of  one  thousand  24-year-old  navel-orange  trees  near  Arlington 
station,  Riverside,  Cal.  The  individual  tree  production  is  shown  by 
figure  I. 

The  grove  consists  of  20  rows  of  trees  from  north  to  south,  with  50 
trees  in  a  row,  planted  22  by  22  feet.  A  study  of  the  records  shows 
certain  distinct  high-  and  low- yielding  areas.  The  northeast  comer  and 
the  south  end  contain  notably  high-yielding  trees.  The  north  two-thirds 
of  the  west  side  contains  a  large  number  of  low-yielding  trees.  These 
areas  are  apparently  correlated  with  soil  variation.  Variations  from 
tree  to  tree  also  occur,  the  cause  of  which  is  not  evident.  These  varia- 
tions, which  are  present  in  every  orchard,  bring  uncertainty  into  the 
results  of  field  experiments. 

In  making  their  calculations  this  grove  was  divided  by  the  writers  into 
imaginary  plots  of  any  size  and  shape  desired.  The  yields  of  these  plots 
were  then  compared  with  one  another  and  their  variability  ascertained. 
The  distribution  of  both  the  theoretical  and  actual  yields  of  this  grove 
is  shown  in  figure  10.  The  yields  of  the  individual  trees  when  plotted 
according  to  their  frequency  give  a  skew  curve  of  Pearson's  Type  I,  since 
the  critical  function 

4(4^2-3^x)(2i82-3/3i-6)         °-^5- 

The  distribution  of  the  actual  yields  is  shown  on  the  figure  by  small 
circles.  The  points  for  the  theoretical  curve  were  calculated  by  the 
formula  \    1  /         x    2 


252 


Journal  of  Agricultural  Research  voi.  xii,  no  s 


I 


I 


04 

1 

.a 

w 

H 

•^ 

o 

■(-> 

4) 

^ 

IH 

t*. 

<u 

'lU 

eu 

Si 

•d 

-«-> 

s 

> 

'rt 

;> 

TJ 

a 

bo 

.a 

I 

a 
I 


Total 
yield 
per 
row. 

fOOvO    «    O    t^OO    «/^0    O>r0N    N    ^H    miooo    MSO    C>N    0^O»t^vO    C^O*N^O    TOO    «0    inr*^0 
lOOO    M    0»  w    t^O    w    MOO    --tTtiriM    r^  OnO    r-  »r(  -ft  «    O    O    ro  O^^O    »At-N)HMro^^ioMr^Ol 
Ov  <o  t^  looo    ■<tV3  00    ■*  O  N    ro  w    o  -.tvo  \0    ■*  QO    M    N    roOO    H    WM    ■*'<fOitt^r-t^rOt-t^O    M 

M«N«M««NNHN«NC.    «NC.«««««««««««M«««Nt'jn<-5«tO 

o 

rOM   r>.M   N    O'O   «   «tnw   r^OswvO   W   O   0\r>.r*   OS  (*5C0   «   0*cOO»^u^O   t^w   O   t>  O^  ^ri  t^  Q 
00   t^r^O   'too   >Oi-i    r^l>.i^NO   u-^Tl■0^0^<^*   O   OO   <n   «0   w^O  COOO         t>.O«o*ON^^O00 

» 

O    O    N    lOOO    ■^•OOO    tO«    t^  \ri  '^  t^  O    Osw    O    <^0<S    f<C0O    Ov^f^i^ioroM    t^O    N    tJ-moO    tT 
t^oo   0   »AMQOOO   row   cow   r^  ^O   u^urO    fO'^w    <n   uic^oovim   m   o   «   O   0\  fOCO   i^  »n  lo  ro  w 

00 

u^w    O-OvO    O^OO    0»»0i>«    low    M\0    r^O    0    m  OsOO   O    «    -^t^roOvr^a  »nOO  \0  »0    »0  ro  »0  m    ^ 
WM«MWWMMMMMtHMWMMMWI-t                    tH            M            WHW                            MMMMHdM 

t^ 

S|5|^5>|5  5^^K|S>o'|^a5Sv?M"&^<g^^SS8«S5^^5Saf: 

■o 

^^S^2.S^|K^^^o^§,'£|8,gg5"?3;KS8^3>S'S<<SSSJ§Sga 

^ 

2  5<2  ?  ^  ?  S  S  ^  2  2  -2  S  S  ^^  It-  B^  S -2  ^  H  ^  3  J  5  ^^  ^^  S  a  H  5 

T 

? 

°S22?H???SS^?'2t'lS"sS2S2°S^^2S  =  SSH^?^ 

^ 

0000i>0O>Ot>Ot~0wiH<7>00>Ot^t^00-*MM0'O'<t(N0i00000000\w00    0i0i«-* 
O    OvO    r-OlO    ■-■    «00    -^JJO    t^t^'^iJt^OO    -"tc-^OOOO    -^tl    OvOPl    O    OVOOCO    •»«<»    ^00    «w 

:: 

N   I-  r»  to  mO   M   "   nvO   O   O   h   h   i/tO    n    «00   «Oii^,  <»)t~mo»r<OoO'*>nO   fO   O   O   w  00 

o 

HTrr^00\>O00<^<^000«0tN-*f0r)-00>0N>ni-iCT.>0i^05.0OC0Mt»N«Q'*t^ 

'-22's:?°>:r;?2S'2?s^28?3:i'^^^:?s°°»^?'SJj??'Ss?s2;?s 

o 

c2s:c3K2^^||'ssg&s;^s>|s;s§«'^|^5st:^^?,'|^sSv^a>s 

00 

Tf  M    roOO    O    ■*«    ^oc^O    w    lAOQO    -^^    rol^iy-)t^«    MvO    f^jQ    CO    M    O    »^M    wiin  l^vO  O    -^t  It 

r- 

^5S?^^SS^^HS2'^°^E'S52«H^l2  2?HS2H^8SS^?«^H 

■c 

5  2  c^l  ^  §,  §,^  ;5c^  i^  ^v?  s  5  S.S  v&<^  j3  ^1  ir=^%  ^ « ^  g5  s  s- ?  ^^  a  ^5 

- 

<X'5u^S!v&a3^^s;'&^<g&?f;jiifi?:s?^a^5o5^s,2g|225^ 

■>r 

« 

S^S^cSRScSS^aS^&S.SKS'S'SScZ'cSScS'^&S^^gvS^^^-Sgg 

„ 

S  ff^  ^  tl5  5  5'<2  3  5,5  <2  S  ?  8;  S"§«  S  &<g  ^^' ^^  v^"^  K<S  ^  5^  ^  ^  8  ;^  ?  ^ 

" 

<ssss^'^Rs;fi^5§s^^§a;5  5K5K^^2;§^^s.s^ffsi:?:?f7cg^ 

Row. 

»  «  .;,  4  .Ao  r;ad  c^  o  M  2  ;j  j  ;j-.-o  ^«  a  0  M  j;  jj  4  ^o  rj-cc  »  o  H  j^  ^ 

Feb.4,i9i8     Variability  of  Yields  of  Fridt  Trees  and  Field  Trials     253 


^ 

„ 

r^t 

^ 

r4 

r» 

i^ 

in 

r- 

■«•  0 

t-» 

w 

0\  -X 

N 

ex 

m 

m 

t^ 

fO 

**: 

n 

to  t  to  ■*  ■* 

1^ 

„ 

t~cO 

at 

M 

" 

" 

^ 

^ 

10 

1/1 

„ 

in 

m 

„ 

^ 

CO 

r^ 

" 

" 

^ 

" 

*^ 

w 

« 

CH 

i~. 

r^ 

„ 

in 

r^oo  00 

m 

■n 

•t 

^ 

■<a 

" 

" 

■^ 

*^ 

M 

u 

" 

CI 

^* 

" 

r* 

0 

„ 

„ 

I^vO 

•<t  TtoO 

t^ 

„ 

■*  •* 

to 

00 

Oi 

H 

>o 

" 

" 

" 

H 

t^ 

„ 

r^ 

„ 

M 

M 

m 

TtOO  00 

m 

0  00 

s- 

" 

" 

w 

^ 

CI 

" 

CI 

r>. 

M 

in 

^ 

^ 

^ 

<-) 

„ 

0 

M 

Oc 

0 

„ 

0 

•<t 

(^ 

PI 

n 

t^ 

in 

CO 

1/ 

r* 

8 

r~-o 

0 

0 

0 

„ 

in 

r- 

0>  01  OvOO 

r* 

t 

" 

•^ 

^ 

t-t 

M 

*^ 

^* 

C^ 

fH 

„ 

•s 

in 

in 

on 

^ 

-n 

„ 

0 

m 

to 

00  -o 

■D 

0 

" 

M 

1.1 

^ 

^ 

C< 

"' 

CI 

r^ 

fl 

in 

•fl- 

„ 

IH 

0 

0 

0 

0 

r* 

t  to  t 

to 

M 

" 

w 

"^ 

Ct 

ct 

c* 

1^ 

ir, 

1/1 

CO 

•fl- 

„ 

r^ 

■* 

m 

-o 

<n 

W 

" 

" 

*^ 

11 

" 

c* 

"' 

" 

CI 

^ 

00  %o 

(-) 

00 

„ 

tn 

f7\ 

in 

0 

„ 

too 

0 

■* 

" 

« 

"^ 

" 

"^ 

^ 

"J 

" 

t^ 

^ 

(10 

? 

fn 

Oi 

„ 

OV 

>o 

„ 

•* 

„ 

0 

r^ 

w 

^ 

*^ 

*^ 

M 

M 

*^ 

c-* 

*^ 

■S 

m 

0 

Tf  0 

Ov  TfOO 

Ol 

„ 

m 

0 

^ 

00 

■^ 

w 

w 

w 

w 

CI 

(.■-1 

t^ 

^ 

M 

in 

„ 

r^ 

ro 

0 

■0- 

o\  0 

M 

■TJ- 

■* 

" 

" 

W 

" 

** 

C* 

CI 

« 

CI 

w 

t^ 

0  -o 

„ 

O 

0 

Ol 

„ 

t^ 

„ 

0 

«, 

o< 

^ 

fH 

c 

to 

NO 

n, 

in 

m 

„ 

M 

M 

„ 

0 

„ 

" 

" 

M 

" 

CI 

u 

CI 

w 

0 

ct 

CI 

o 

^ 

ir, 

^ 

n 

nn 

•<J-00 

„ 

00-0 

^ 

r> 

10 

r^ 

f) 

■0 

" 

W 

w 

^ 

^ 

*^ 

-o 

„ 

rt 

m 

00 

„ 

0 

0 

Oc 

0 

00 

» 

ao 

^ 

^ 

i.A 

^ 

"^ 

u 

^ 

CI 

"^ 

"2 

rv 

tn 

„ 

r^ 

m 

„ 

r^ 

„ 

0  0  NO  *0 

" 

*^ 

^ 

" 

" 

"^ 

CI 

" 

w 

w 

^ 

„ 

m 

0-0 

m 

„ 

w 

Oi 

t 

S:"" 

0 

u 

a 

•a 

>. 

3 

H 

: 

^ 

: 

3 

: 

r^co 

■* 

^ 

•^ 

<» 

10 

254 


Journal  of  Agricultural  Research 


Vol.  XII  No.s 


Navel  orange  (ANTEiyOPE  Heights)  . — The  navel-orange  grove  later 
referred  to  as  the  Antelope  Heights  navels  is  a  plantation  of  480  ten-year- 
old  trees  planted  22  by  22  feet,  located  at  Naranjo,  Cal.  The  individual 
tree  records  of  this  planting  are  shown  in  figure  2  and  are  the  yields 
obtained  in  1916.  The  general  appearance  of  the  trees  gives  a  visual 
impression  of  uniformity  greater  than  a  comparison  of  the  individual 
tree  production  substantiates;  however,  the  distribution  of  the  yields 
approximates  closely  the  normal  curve  of  errors,  having  a  skewness  of 
only  o.ooi  ±0.037,  ^s  shown  by  figure  11. 

Fig.  2. — Diagram  showing  the  individual  tree  yield  (in  pounds)  of  the  navel-orange 

grove  (Antelope  Heights). 

North 

Yield  per  tree  {pounds) 


Row. 


Total 
yield 
per 
row. 


4 

5 

6 

7 

8 

9 

10 

II 

12 

13 

14 

IS 

16 

17 

18 

19 

20 

ai 

22 

23 

24 

25 

26 

27 

28 

29 

30 

31 

32 

33 

Total  yield 
per  row . . 


130 
120 
200 
ISS 

250 
300 
22s 

ISO 
300 
250 
200 
ISO 
no 
190 
200 
210 
140 
los 
120 
130 
ISO 
I9S 
100 
100 
los 
100 
100 
13  S 
24s 
200 

22s 
200 
250 


200 
ISO 
ISO 
20s 
225 
200 
2  SO 
325 

ISO 

i8s 
100 
200 
150 
200 
ISO 
ISO 
100 
100 
ISO 
120 
100 
105 
no 
ISO 
100 
100 
ISO 
i8s 
255 
170 
225 
200 
150 


250 
230 
300 
295 
300 
240 
225 
22s 
290 
17s 
170 
230 
135 
i6s 
130 
100 
IIS 
200 
ISO 
120 
200 
ISO 
15s 
200 
ISO 
270 
230 
280 
250 
SO 
250 
200 
ISO 


130 
270 
235 
26s 
200 
200 
200 
210 
200 
I3S 
175 
ISO 
160 
135 
100 
no 
135 
ISO 
100 
150 
175 
185 
265 
150 
1 75 
100 
ISO 
250 
300 
3  SO 
IIS 
21S 
170 


300 
275 
280 
320 
290 
210 
275 
225 
100 
170 
200 
200 
180 
150 
18s 
ISO 
160 
200 
200 
200 
275 
200 
ISO 
175 
200 
150 
ISO 
115 
260 
240 
235 
240 
155 


260 
300 
200 
250 
250 
250 
250 
200 
225 
200 
200 
250 
250 
200 
220 
150 
250 
200 
200 
ISO 
160 
190 
200 
175 
175 
175 
150 
180 
i8s 
240 
225 
200 
220 


28s 
340 
170 
130 
185 
220 
200 
250 
225 
225 
210 

215 

220 
190 
150 
225 
200 
185 
22s 
225 
200 
i5o 
140 

125 
ISO 

250 
270 
175 
150 
200 
200 
200 
250 


2SS 

195 

250 
200 
ISS 

150 

ISO 
150 
22s 
200 
300 
200 
200 
200 
ISO 

150 

150 

I2S 

200 

ISO 

100 
ISO 
ISO 
100 
175 
175 
200 
200 
250 
210 
270 
250 
250 


225 
250 
ISO 
210 
i6s 
I3S 
170 
205 
250 
I6S 
190 
160 
225 
300 
200 

I2S 
225 
150 
125 
200 
200 
22s 

ISO 
240 
ISO 
150 
200 
i8s 
200 
250 
200 
250 
250 


ii3SO 
>45S 
.285 
,285 
.475 
,  240 
.250 
>370 
,  190 
.84s 
>82S 
.775 
.835 
,870 
.64s 
.345 
.320 
,  200 
,28s 
.145 
I  210 

.215 

.075 
.125 
,280 
,420 
.730 
.635 
.005 
.  170 
,110 
.025 

.125 


5. 460 


5,385 


6,  580  6,010  6,815 


6, 930  6,  745 


6,  23s  6,  27s  6,475  5.  70016, 175  6,  260 


Valencia  orange. — The  Valencia  orange  grove  is  composed  of  240 
15-year-old  trees,  planted  21  feet  6  inches  by  22  feet  6  inches,  located  at 
Villa  Park,  Cal.  Figure  3  represents  this  planting  and  the  individual 
tree  yields  which  were  obtained  in  1916. 

Eureka  lemon, — The  lemon  yields  were  obtained  from  a  grove  of 
364  23-year-old  trees,  located  at  Upland,  Cal.  Figure  4  represents  the 
individual  tree  yields  of  this  planting.  The  records  extend  from  October 
I,  1915,  to  October  i,  1916.  The  grove  consists  of  14  rows  of  23-year-old 
trees,  extending  north  and  south,  with  26  trees  in  a  row,  planted  24  by 


Feb.4,i9i8     Variability  of  Yields  of  Fruit  Trees  and  Field  Trials     255 

Fig.  3. — Diagram  showing  the  individual  tree  yield  (in  pounds)  of  the  \'alencia 

orange  grove. 

Northwest 

Yield  per  tree  {pounds) 


Row. 


3 
4 
5 
6 
7 
8 
9 
10 
II 
12 
13 
14 
IS 
16 
17 
18 
19 


100 

375 

37 

63 

75 

100 

SO 

300 

100 

100 

75 

225 

425 

SO 

437 

400 

3SO 

175 

3SO 

400 


Total  yield  , 
per  row 14.087 


375 
325 

75 
175 

75 
212 
136 
200 
I7S 
ISO 
125 
22s 
200 
350 
238 
163 
275 
325 
350 
350 


250 
275 
175 
250 
175 
125 
100 
100 
30 
100 

125 

350 
225 
200 
225 

ISO 
400 

250 
400 
450 


275 
375 
260 
350 
300 
25 
75 
75 
300 
150 
275 
250 
325 
125 
200 
300 
300 
350 
250 
375 


275 
200 
150 
300 
300 
200 
250 
150 
263 
150 
125 
300 
75 
225 
350 
4SO 
462 
350 
275 
200 


ISO 
375 
300 
275 
325 
275 
100 
200 
363 
262 
275 
150 
275 
150 
375 
300 
288 
3SO 
275 
300 


57363 


200 
225 

25 

325 
300 
325 
250 
200 

350 
250 
150 
300 

425 

225 
225 
400 
412 
100 
212 
350 


175 
230 
275 
300 
275 
250 
100 
275 
300 
275 
225 
300 

275 
250 
225 
350 
363 
300 


S,3o6 


225 
275 
450 
250 
375 
250 
190 
200 
337 
363 
312 
262 
375 
250 
300 
350 
375 
300 
350 
275 


6,064 


27S 
225 
250 
275 
275 
100 
150 
200 
338 
337 
188 
265 
300 
300 
400 
250 
300 
200 
250 
350 


S.318 


200 
150 

225 

ISO 
375 
3SO 
225 
225 
25 
300 
375 
325 
350 
175 
150 
325 


125 
262 
250 


75 
225 
250 
150 
250 
ISO 
225 
225 
175 
350 

225 
lOO 

175 

100 
100 


3>962 


Total 
yield 
per 
row 


7625 
.367 
7S72 
.113 
.950 
,087 
,851 
,  200 
,  101 
-737 
,250 
>077 
.150 
,600 
,700 
,663 
•975 
■  140 
-ISO 
,850 


59) 158 


Fig.  4. — Diagram  showing  the  individual  tree  yield  (in  pounds)  of  the  Eureka 

lemon  grove. 

North 

Yield  per  tree  {pounds) 


Total 
yield 
per 
row. 


3 

4 

S 

6 

7 

8 

9 

10 

II 

12 

13 

14 

IS 

16 

17 

18 

19 

20 

21 

22 

23 

24 

2S 

26 

Total  yield  per 
row 


259 
204 

475 
383 
Sio 
421 
425 
218 
301 
409 
152 

79 
287 
348 

81 
250 
III 
195 
333 
291 
III 
361 
210 
140 
126 
241 


2S8 
157 
235 
292 
250 
188 
228 
274 
257 
264 
165 
1x8 
196 
258 
199 
•146 
165 
162 
156 
172 
189 
114 
245 
197 
229 
263 


227 
305 
274 
190 
277 
167 
206 
176 
164 
123 
150 
170 
161 
144 
106 
223 
13s 
170 
127 
181 
127 


251 
314 
183 
263 
202 
338 
195 
133 
276 
188 
194 
120 
264 
267 
180 
282 
259 
241 
154 
203 
255 
296 
315 
207 
i8s 
220 


379 
381 
412 
288 
313 
369 
343 
234 
230 

221 

276 

258 

292 
345 
268 
253 
227 
275 
357 
104 
332 
326 
414 
350 
275 
397 


325 
264 
251 
356 
300 
346 

341 
312 
319 
336 
222 
241 
345 
278 
290 
361 
291 
325 
239 
328 
252 
294 
388 
297 
399 


201 
283 
210 
260 
157 
286 
202 
211 
167 
234 
278 
261 
2.55 
237 
213 
245 
278 
256 
186 
222 
235 
396 
201 
343 
412 
341 


232 
211 
271 
172 
169 
140 
167 
169 
206 
149 
177 
245 
159 
216 
209 
218 
116 
122 
80 

112 
176 
158 
128 
289 
241 

335 


349 
292 
443 
327 
327 
382 
334 
386 
180 
439 
314 
268 
341 
380 
336 
360 
304 
259 
226 
364 
298 
282 
416 
410 
328 
438 


237 
329 
348 
322 
399 
346 
377 
429 
471 
354 
441 
308 
399 
353 
312 
368 
315 
363 
335 
371 
398 
325 
313 
302 
262 
266 


6,921  , 


'  4j  825 


5.985' 


)8,o63 


6, 550  4, 867 


8,  783  ! 


296 
273 
378 
345 
281 
233 
312 
298 
342 
248 
283 
278 
317 
365 
299 
322 
338 
385 
369 
349 
329 
294 
326 
341 
333 
413 


i 8,347 


264 
218 
298 
295 
304 
24S 
300 
281 
238 
279 
281 
322 
307 
301 
279 
338 
318 
3S8 
358 
402 
307 
322 
410 
283 
280 
237 


7,82s 


241 
236 
310 
279 
321 
313 
308 
277 
339 
293 
282 
305 
298 
200 
223 
344 
326 
356 
223 
276 
204 
2S7 
222 
253 
I3t 
328 


7,176 


104 
222 

222 

375 

280 

333 
266 
100 
342 
287 
279 
222 
295 
285 
248 
245 
3" 
294 
288 
338 
244 
236 
276 
292 
277 
316 


6.977 


,584 
,611 
,341 
,231 
.003 
,217 
,987 
•557 
,837 
,848 
.581 
•157 
,801 
,061 
,269 
,767 
,653 
,69a 
,560 
-S70 
-587 
.776 
.954 
.040 
.592 
■  383 


98,658 


256 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  s 


24  feet  apart.  This  grove  presents  the  most  uniform  appearance  of  any 
under  consideration.  The  land  is  practically  level,  and  the  soil  is  appar- 
ently uniform  in  texture.  The  records  show  a  grouping  of  several  low- 
yielding  trees;  yet  a  field  observation  gives  one  the  impression  that  the 
grove  as  a  whole  is  remarkably  uniform. 

SUEDiyiNG  WALNUTS. — The  walnut-tree  records  used  in  the  following 
calculations  were  obtained  during  the  seasons  of  191 5  and  191 6  from  a 
24-year-old  Santa  Barbara  softshell  seedling  grove,  located  atWhittier, 
Cal.  The  planting  is  laid  out  10  trees  wide  and  32  trees  long,  entirely 
surrounded  by  additional  walnut  plantings,  except  on  a  part  of  one 

Fig.  5. — Diagram  showing  the  individual  tree  yield  (in  pounds)  of  the  seedling 

walnut  orchard. 

North 

Yield  per  tree  {pounds) 


30. 

31- 

32. 


Row. 


172 

164 

74 

122 

120 

160 

170 

95 

6S 

85 

78 

SI 

17 

7S 

117 


83 
44 
86 
71 
87 
41 
86 
169 
112 


III 
135 
89 
SO 

lOI 

106 

34 
104 

69 
179 
188 
122 
144 
155 

83 
176 

SO 
III 
136 
158 
ISO 
100 


135 
96 
80 

103 
94 

128 
71 
61 

100 


73 
123 

88 
103 

69 
170 
198 
125 
151 
140 

74 

95 
151 
166 


Total  yield  per  row 2,664  3-052  3-037  2,936  2,625  2,849  2)7SS  2,904  2,545  2)689 


113 
III 
40 
80 
108 
74 
135 
82 
66 
195 
100 
112 
71 
95 
68 
35 
105 
180 
124 
32 
192 
130 
58 
170 
54 
80 
138 


113 
35 
207 
78 
112 
129 
121 
93 
61 
89 
72 
76 
82 
57 
6 
47 
65 
120 
48 
142 
113 
147 
112 
123 
115 
75 
148 
39 


167 
3i 
94 
57 

118 
40 

131 
93 

134 
45 
63 
86 
99 
76 
35 


165 
197 
97 
76 
119 
71 
96 
99 
178 
151 
192 


87 
67 
79 

104 
80 

123 
84 
93 
61 
61 

133 
15 
15 
96 
90 

104 
91 

144 

122 

lOI 

81 
210 
104 
160 
130 
90 
73 
157 


207 
136 
104 
144 
40 
126 
72 
70 
43 
14 
113 
81 
69 
80 
8s 
31 
106 
77 
160 
131 
113 
94 
109 
87 
"5 
94 
215 


164 
164 
195 

S4 

52 
107 

30 
8 

60 

25 

73 
S3 
28 
25 
94 
42 
72 
85 
117 
90 
65 
149 
113 
25 
166 
ISO 
166 
160 


164 
164 
97 
62 
58 


96 
64 
43 
loi 
83 
50 
65 
13 
75 
97 
105 
83 
87 
39 
131 
126 
148 
129 
128 
235 
201 


Total 
yield 
per 
row. 


1,277 

1, 116 

1,246 

811 

896 

828 

1,016 

970 

771 

677 

842 

815 

612 

629 

603 

858 

914 

1.252 

1,080 

1,117 

1,022 

1.386 

950 

1, 169 

1. 031 

1. 15s 

1.475 

1.538 


28,056 


side  which  is  adjacent  to  an  orange  grove.  The  trees  are  planted  on 
the  square  system,  50  feet  apart.  Figure  5  gives  the  yield  and  arrange- 
ment of  these  trees. 

Jonathan  apples. — ^The  apple  records  ^  were  obtained  from  a  lo-year 
old  Jonathan  apple  orchard  located  at  Providence,  Utah.  The  surface 
soil  of  this  orchard  is  very  uniform  to  all  appearances  except  on  the 
extreme  eastern  edge,  where  the  percentage  of  gravel  increases  slightly. 
The  trees  are  planted  16  feet  apart,  east  and  west,  and  30  feet  apart 
north  and  south.     Figure  6  gives  the  yield  and  arrangement  of  these  trees. 

1  The  authors  wish  hereby  to  express  their  appredation  of  the  kindness  of  the  Utah  Bxperiment  Station 
in  furnishing  these  records. 


Feb. 4, 1918     Variability  of  yields  of  Fruit  Trees  and  Field  Trials      257 

RANGE   OF   INDIVIDUAIv   TREE   YIELDS 

The  extremes  of  individual  tree  productivity  are  shown  in  Table  I. 
This  indicates  a  wide  range  of  variation  even  in  oranges,  lemons,  and 
apples,  which  are  clonal  varieties.  The  greatest  range,  however,  is  in 
the  yield  of  the  seedling  walnuts.  The  variations  set  forth  in  this  table 
may  seem  excessive  to  workers  with  annual  crops.  To  those  familiar 
with  the  variation  in  tree  crops,  however,  this  will  be  recognized  as  only 
the  normal  variation  which  occurs  in  most  fruit  plantations  growing  on 
apparently  uniform  soil,  as  mentioned  in  a  previous  section.  Wide 
variation  in  the  vigor  of  the  rootstock,  as  well  as  variation  in  soil  pro- 

FlG.  6. — Diagram  showing  the  individual  tree  yield  (in  pounds)  of  the  Johnathan 

apple  orchard. 

Yield  per  tree  (pounds) 


Row. 


Total  yield  per  row 7, 082  8, 070  8, 469 


387 
187 
337 
42s 
42s 
100 
325 

I2S 

375 
300 

512 

162 

275 
175 

312 

I2S 

300 
250 
350 
187 

300 
200 
250 

187 

112 
SO 
187 
162 


250 
200 
112 

437 
450 
212 
437 
350 
400 
450 
487 
312 
375 
462 
387 
400 
375 
200 
175 
15° 
300 
187 
100 
200 
287 
125 
100 
ISO 


550 
175 
575 
400 
425 
350 
337 
262 
350 
312 
212 
250 

50 
362 
312 
375 
250 
250 
450 

5° 
262 
387 
337 
337 
175 
200 
287 
187 


362 
400 
S50 
450 
375 
575 
375 
437 
412 
250 
325 

37 
286 
387 
350 
ISO 
312 
250 
261 
350 
375 
362 
362 
337 
32s 
375 

SO 
200 


212 
350 
300 
250 
52s 

337 
SCO 
212 
550 
525 
425 
537 
47S 
187 
262 
375 
1 75 
7S 
287 
375 
325 
42s 
312 
450 
32s 
437 
87 
SO 


437 
100 
550 
362 
375 
137 
S50 
32s 
450 
337 
500 
300 
275 
337 
337 
375 
375 
300 

87 
100 
42s 
200 
100 
300 
200 
275 
ISO 

7S 


8,334 


337 
500 
475 
287 
437 
337 
350 
475 
437 
437 
437 
375 
412 
437 
500 
50 
450 
262 
312 
275 
350 
312 
225 
200 
150 
275 
325 
100 


437 
375 
300 
362 
375 
350 
337 
300 
250 
225 

212 

475 
300 
337 
75 
100 
425 
225 
375 
187 
225 
150 

125 

287 
250 


Total 
yield 
per 
row. 


2,747 
2,324 
3,161 
3,048 
3,387 
2,348 
3,236 
2,561 
3,324 


2,  223 
2,087 
2,559 
2,935 
2,150 
2.S74 
1,662 
2,548 
1. 912 
2,562 
2,448 
1,873 
2,236 
1,724 
1,862 
1,473 
1,174 


68. 284 


ductivity,  may  have  been  instrumental  in  causing  such  a  variation  in 
yield. 

Table  I. — Range  of  variability  in  crop  production  0/ fruit  and  nut  trees 
[Extremes  and  range  expressed  as  percentages  of  the  mean  yields  of  the  respective  plantations] 


Kind  of  fruit. 


Naval  oranges  (Arlington) .  . 
Naval   oranges    (Antelope 

Heights) 

Valencia  oranges 

Eureka  lemons 

Seedling  walnuts 

Jonathan  apples 


Mean  yield. 


Pounds. 
i37±i-6 

i86±i.  7 
246 ±4.  3 

270±2.  9 

86±i.  6 
304  ±5-  6 


Extreme  yield. 


Per  cent. 
9.  I  to  246.  3 


.  4  to  193.  8 
.  I  to  188.  o 
.  4  to  180.  3 
.  4  to  276.  I 
-  3  to  193.  2 


Range. 

Mode. 

Percent 
237.2 

Pounds. 
129.8 

160.  4 
182.9 

186.  I 
270.8 

147.9 
261    7 
180.  9 

299.9 

75-6 

345-1 

Skewness. 


o.  143  ±0.025 


.  001  ± 
•  25o± 

•359± 

. 269 ±  .  049 

•329±  -055 


037 
053 
043 


258 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  5 


The  biometrical  constants  for  the  several  plantations  are  given  in 
Table  II.  The  oranges,  lemons,  and  apples,  as  might  be  expected,  show 
less  variability  than  the  seedling  walnuts.  The  coefficient  of  variability 
of  the  clonal  varieties  ranges  from  29.72  to  41.23  per  cent.  This  total 
range  of  only  11. 51  per  cent  shows  a  marked  similarity  of  the  extent  of 
variation. 

Table  II. — Variability  in  yield  of  the  different  individual  fruit  trees 


Kind  of  fruit. 


Naval  orange  (Arling- 
ton  

Naval  orange  (Ante- 
lope Heights) 

Valencia  orange 

Eureka  lemon 

Seedling  walnut 

Seedling  walnut 

Seedling  walnut,  aver- 
age  

Jonathan  apple 


Acre- 

Total 

Crop. 

age 

num- 

Mean yield 

per 

ber  of 

per  tree. 

tree. 

tree^ 

Pounds. 

191S 

0.011 

1,000 

137. 6±i.  2 

1916 

.on 

495 

186.  2  ±1.7 

191S 

.  on 

240 

246. 3  ±4.  3 

191S 

.013 

364 

270.  7  ±2-  9 

191S 

•OS7 

280 

99-8±i.9 

1916 

.057 

280 

77-6±l.  7 

(1915 

1 

]and 

i  -057 

280 

86.4±i-6 

[1916 

1 

1914 

.oil 

224 

303.9±s-6 

Standard 
deviation. 


Pounds. 
54- 42  ±  0.8a 

55- 33  ±  I-  19 
97-  84±3-  01 
81.38i2.03 
47-77±l-36 
41-94±I-I9 

40.  io±  1. 14 

125.  30±4.  CO 


Coefficient  of 
variability. 


39- 55  ±o- 68 

29.  72±o.  69 
39- 72  ±  I- 40 

30.  o6±o.  8i 
47.  86±  I.  64 
53- 91  ±1-92 

46.  41  ±1-58 

4i-23±l.  S2 


Probable  error. 


Pounds 
per  tree. 


Percent- 
age of 
mean. 


26.67 

20.05 
26.79 
20.  28 
32-28 

31-30 
27-81 


The  probable  error,  expressed  in  pounds  of  fruit  per  tree,  is  the  greatest 

in  case  of  the  Jonathan  apple,  amounting  to  85  pounds,  while  the  Valencia 

orange  and  the  Eureka  lemon  fall  to  66  and  55  pounds,  respectively. 

Such  probable  errors,  expressed  in  pounds  per  tree,  are  not  comparable, 

however,  unless  the  mean  yields  are  approximately  the  same.    The 

probable  error  expressed  as  a  percentage  of  the  mean  is  therefore  added 

to  Table  II  to  make  it  more  easily  compared  with  tables  of  other  writers 

who  have  seen  fit  to  use  this  constant  rather  than  the  coefficient  of 

variability. 

METHODS  OF  CALCULATING  VARIABILITY 

The  yields  of  the  various  fruit  plantations  have  been  studied,  with 
trees  singly  and  combined  into  plots  of  various  sizes.  The  coefficient  of 
variability  and  probable  error  have  been  used  as  the  basis  of  comparison 
in  most  cases. 

Plots  of  different  sizes  necessarily  have  varying  mean  yields  per  plot; 
therefore  the  coefficient  of  variability  is  more  readily  interpreted  than 
the  standard  deviation.  The  probable  error  may  only  be  used  with 
accuracy  in  cases  where  the  number  of  variants  is  relatively  large  and 
their  distributions  normal. 

VARIABILITY  OF  ORCHARD   PLOT  YIELDS 

EFFECT  OF  INCREASING  THE  NUMBER  OF  ADJACENT  TREES  PER  PLOT 

The  first  point  studied  was  the  effect  of  increasing  the  number  of 
adjacent  trees  per  plot,  measured  by  the  coefficient  of  variabihty.  Based 
on  the  theory  of  random  samphng  of  variables,  the  average  production  of  10, 
or  even  5,  trees  should  be  a  more  typical  sample  of  the  orchard  than  that 


Feb.4.i9i8     Variability  of  Yields  of  Fruit  Trees  and  Field  Trials     259 

of  I  tree.  The  reduction  of  the  coefficient  of  variability  by  combining  a 
number  of  adjacent  trees  in  a  plot  would,  however,  be  expected  to  fall 
short  of  the  theoretical  reduction,  because  such  a  combination  may  have 
a  tendency  to  group  trees  of  similar  productivity  together.  Gradual  soil 
variation  from  one  side  of  the  plantation  to  the  other,  or  irregularities  of 
the  field  which  are  larger  than  the  area  taken  up  by  a  single  tree,  will  tend 
to  bring  about  a  correlation  between  the  yields  of  adjacent  trees. 

For  practical  purposes  the  two  more  or  less  antagonistic  sources  of 
variation  between  plots  may  be  arranged  in  two  groups : 

1.  Those  which  may  cause  the  variations  to  become  greater  as  the  size 
of  the  plot  increases — ^for  example,  variation  in  soil  productivity. 

2.  Those  which  may  cause  the  variations  to  become  less  as  the  size  of 
the  plot  increases — for  example,  variations  in  inherent  producti\-eness 
of  the  trees.  This  may  depend  to  no  small  degree  on  the  variation  in 
vigor  and  character  of  growth  of  the  rootstock.  Measured  by  crop  pro- 
duction, it  may  be  practically  impossible  with  grafted  or  budded  trees  to 
separate  by  mere  inspection  the  variation  which  may  be  caused  in 
inherent  qualities  of  the  bud  from  those  of  the  rootstock  on  which  it  is 
propagated. 

The  reduction  of  the  coefficient  of  variability  in  the  several  plantations 
as  a  result  of  increasing  the  number  of  adjacent  trees  per  plot  is  shown  in 
Table  III.  The  acreage  per  plot  is  recorded  for  sake  of  comparison  with 
similar  work  by  agronomists,  where  the  size  of  the  plots  studied  has  been 
dependent  entirely  on  acreage  rather  than  number  of  plants  to  the  plot. 
Other  biometrical  constants  are  likewise  included  for  ease  of  comparison 
with  above-mentioned  studies. 

The  effect  of  increasing  the  number  of  adjacent  trees  per  plot  on 
reducing  the  coefficient  of  variability  between  the  plots  of  all  the  fruit 
crops  studied  is  shown  in  the  summary  of  Table  III,  and  figure  7  shows 
the  same  thing  graphically.  The  curves  show  a  marked  similarity  be- 
tween the  varieties  of  fruits  and  agree  quite  closely  in  demonstrating 
that  there  is  little  to  be  gained  in  including  more  than  eight  adjacent 
trees  in  a  plot.  As  a  rule,  there  is  a  rapid  reduction  in  the  coefficient, 
as  progress  is  made  from  a  i-tree  to  an  8-tree  plot.  Increasing  the  plot 
above  eight  .adjacent  trees  shows  only  a  comparatively  small  reduction 
of  the  coefficient  of  variability.  In  fact,  the  reduction  is  not  significant 
when  the  probable  errors  ^  are  considered.  The  Antelope  navels  and 
apples  show  a  reduction  slightly  less  than  the  probable  error  between 
a  4-tree  and  an  8-tree  plot.  Again,  the  lemons  show  an  apparently 
exceptional  reduction  when  the  i6-tree  plot  is  compared  with  the  24-tree 
plot.  The  same  is  true  of  the  8-  and  i6-tree  plots  of  apple  trees.  These 
exceptions  are  in  part  explained,  where  they  concern  the  larger  plots, 

'  The  probable  error  of  the  difference  between  two  averages  A  i  and  A  2,  of  which  the  probable  errors 
El  and  £2  are  known,  is  the  square  root  of  the  sum  of  the  squared  probable  errors;  or  probable  difference 

of /li-/l2=±y£i+£j 


26o 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  s 


by  the  fact  that  in  a  given  area,  as  the  size  of  the  plot  increases  the 
number  of  plots  necessarily  decreases,  and  thus  lessens  the  reliability 
of  the  comparisons.  Thus,  these  exceptions  may  be  due  to  chance  in  a 
small  population  and  might  not  hold  true  with  a  larger  number  of  variants. 

Table  III. — Effect  of  increasing  the  number  of  adjacent  trees  per  plot 


Kind  of  tree. 


Navel  oranges  (Ar- 
lington)  

Do 

Do 

Do 

Do 

Do 


Navel  oranges  (An- 
telope Heights) 

Do 

Do 

Do 

Do 

Do 


Valencia  oranges 

Do 

Do 

Do 

Do 

Do 


Eureka  lemons. 

Do 

Do 

Do 

Do 

Do 


Num- 
ber of 
trees 
per 
plot. 


Seedling  walnuts. 

Do 

Do 

Do 

Do 

Do 


Jonathan  apples. . 

Do 

Do 

Do 

Do 

Do 


Num- 
ber of 
plots. 


,000 
500 
250 
125 
60 
40 


495 

247 

125 

61 

30 

20 

240 
120 
60 
30 


364 


280 
140 


224 

112 

56 

28 

14 


Acre- 
age per 
plot. 


o.  on 
.  022 

.044 
.088 
.176 

.  264 


.  on 
.  022 
.044 
.088 
.176 
.  264 


.011 
.  022 
.044 


Mean  yield 
per  plot. 


Pounds. 

137- 6i  1.2 

275. 6i  2.9 

551. 6i  7.4 

1, 100.  8i  19.  2 

2,  220.  oi  56.  o 

3.343-8i  92.5 


186.  2i  1.  7 

373- oi  4-2 

742.  7i  10.  o 

1,460.  7±  26.  7 

2,973.  3  ±  68.4 

4,450.  oi  121.  6 


246.3  i  4.3 
487.  92  i  10.  O 
991-  7  i  23.  9 
1,966.  7  i  59-4 
176  3,813.3  ii58.6 
264    5,880.0  i243.  o 


•013 
.  026 


.  211 

•317 

.057 
.  114 
.  228 
.456 
.  912 
1.368 

.oil 
.  022 
.044 
.088 
.  176 
.  264 


270.  7  i  2.  9 
544.  3  ±  6.  8 
1,081.  o  ±  17.  3 
2, 172.  o  i  46.  7 
4,395-o  ii2i.  7 
6,692.0  ii53.  5 

86.4  i      1.6 

178.  7   i     3.  6 

353-  6  i     8.  6 

717.  I   i    20.  7 

1,409.  4   i   58.  O 

2,154.6  il03.3 

303-9  ±     S-6 

6og.  8  i  12.  3 

I,  210.  7  i  28.  8 

2,414.3  i  76.9 

4,864.3  ii75-S 

7,277.8  i3l3-9 


Standard 
deviation. 


Pounds. 

54.  42i  0.82 

95.  6oi  2.  03 

173-  89i  5-  24 

319.  ooi  13.60 

642.  6oi  39.  58 

867.  70i  65.43 


S5-33±  I- 19 
97.  2ii  2.95 
164. 37±  7.07 
309.  05  i  18.90 
555- 55 i  48.38 
806.  54i  86.01 

97-  84i  3-  01 
161.  8oi  7-  04 
275.  2oi  16.  92 
482.  6oi  42.  02 
910.  58iii2.  13 
i,l40.ooil7i.90 

81.  38i  2.  03 
135-  40i  4-  78 
245.  35  i  12.  27 
448.  50i  33.  01 
827.  ooi  86.  07 
820.  30i  108.  50 

40.  io±  1. 14 

62.  40i  2.  52 

106.  40  i  6.  06 

181.  24i  14.  60 

354- 40i  41.00 

508.  ooi  73.05 

125.  30i  4-  00 
193.  ooi  10.  44 
319.  40  i  20.  40 
603.  38i  54.38 
973.  68  i  124.  II 
1,396. ooi  221.90 


CoelBcient 

of 
variability. 


39.55±o.  68 
34.68io.  82 
31.  52ii.o5 
28.  98ii.30 
28.  95  i  I.  92 
2S-95±2.oS 


29.  72  ±0.69 

26.  o6±o.  84 

22.  I3il.  00 

21.  i6ii.  35 
18.  68ii.68 
18.  ioil.91 

39- 72  ±1.40 
33-  i6ii.6i 

27.  75ii.84 

24.  54i2.  26 

23.  88i3.  10 
I9-39±3.03 

30.  06 io.  81 

24.  88  i  0.93 

22.  70il.  19 
20.  65  i  I.  57 

18.  82  i  2.  02 
12.  26ii.  65 

46.  41  i  I.  58 
34-92ii.  57 
30.  09ii.  86 

25.  27i2.  16 

25.  I5i3.  07 
23-  S8i3.  53 

41.  23il.  52 
31.65i1.87 

26.  38il.  80 
24.  99i2.  39 
20.  02  i  2.  65 

19.  i8i3.  16 


Probable  error. 


Pounds.     f^J- 
'  centage 
of 
mean. 


per 
tree. 


26.67 

23-39 
21.  26 
19.55 
19.53 
17- SO 


20.05 
17-58 
14-93 
14.27 
12.60 
12.  21 

26.79 
22.37 
18.72 
16.5s 
16.  II 
13-08 

20.  28 
16.78 
IS- 31 
13-93 
12.69 
8.27 

31.30 

23- ss 
20.30 
17.04 

16.96 

15.90 
27.81 

21- 3S 
17.79 
16.86 
13.50 
12.94 


SUMMARY 


Coefficient  of  variability. 

A^-era 
tiou  ol 

^'e  reduc- 
coetiicient 

Ntmi- 
ber  of 
trees 

Navel  oranges. 

Valencia 
oranges. 

Eureka 
lemons. 

Seedling 
walnuts. 

Jonathan 
apples. 

Average. 

of  variability  by 
increasing  num- 
ber of  adjacent 
trees  per  plot. 

per 
plot. 

Arlington. 

Antelope 
Heights. 

In- 
crease 
from — 

Average 
reduc- 
tion. 

39.  55 io.  68 
34.68io.82 
31.52i1.05 
28.98i1.30 
28.9si1.92 
25-9S±2-o8 

29.  72 io.  69 
26.  ooi  0.84 
22.  13  i  I.  00 
21.l6il.35 
i8.68ii.68 
18.  ioii.91 

39-72ii-40 
33,.  l6ii.6i 
27.  75ii.84 
24.  54i2.  26 
23.88i3. 10 
i9.39i3-03 

30.  o6io.  81 
24.88io.93 
22.  7oii.  19 
20.65i1.57 
18.  82  i  2.  02 
12.26i1.6s 

46.  4iii.58 
34.92i1.57 
30.  09ii.  86 
25.  27i2.  16 
25-l5±3-07 
23-s8i3-S3 

4l-23il.S2 
31.65il.87 
26.38il.80 
24.99i2.  39 
20.  02  i  2.  65 
19.  l8i3.  16 

37-78io.  52 
30.  89io.  55 
26.  76 io.  62 
24.  27io.  77 
22.  58ii.oi 
19.  74ii.o8 

2 

4 

8. 

16 

24 

1  to  2 

2  to  4 
4  to  8 
8  to  16 

16  to  24 

6. 89io.  76 
4.  13  i  0.83 
2.  49io.99 
i.69ii.27 
2.84±I<4S 

Feb.4,i9i8     Variability  of  Yields  of  Fruit  Trees  and  Field  Trials     261 

The  averages  of  all  six  fruit  plantations  show  that  there  is  a  rapid 
reduction  of  the  coefficient  of  variability  until  the  8-tree  plot  is  reached, 
but  from  then  on  the  reduction  is  less  in  comparison  with  the  probable 
error.     (See  Table  III,  summary,  and  figure  8.) 


'^2 
"M 
38 
36 
3^ 
32 
3C 
23 

2e 

\22 

K 

\36\ 

\7. 

\30 
\2S 

V'' 

V)  22 

\^20 

\  /6 

^3S 

3^^ 
32 
30 
23 
26 
2^ 
22 
20 
/3 


\^o 


^2 

38 
36 
3^ 
32 
30 
28 
26 
2^ 


c/OA^^r///^//  ,^^/:>/.£' 


^£'£'£>z//^(^  /y^^/v<yr 


/2^ 


S  /6  2^^ 


/^•^       ^  /6  2^^ 


Fjg.  7. — Graphs  of  the  reduction  of  the  coefficient  of  variability  by  increasing  the  number  of  adjacent  trees 

to  the  plot. 

Although  the  average  reduction  of  the  coefficient  between  the  8-and 
i6-tree  plots  is  doubtful  when  compared  with  the  probable  error,  this 
reduction  in  all  six  cases  is  constant — that  is,  the  variation  occurs  in 


26: 


Journal  of  Agricultural  Research 


Vcl.  XII,  No.  s 


one  direction  only,  and  therefore  has  more  significance  than  is  indicated 
by  a  mere  comparison  of  the  averages. 

EFFECT  OF  SYSTEMATIC   DISTRIBUTION  OF  PI.OTS  OVER  THE  AREA  STUDIED 

The  importance  of  distributing  plots  over  the  experimental  area  is 
more  or  less  obvious,  and  has  been  dwelt  upon  by  many  writers.  Its 
value  arises  from  the  fact  that  the  soil  varies  over  the  area,  and  it  is 
better  to  have  similar-sized  plots  on  both  high-  and  low-yielding  areas 

than  to  have  them 
solely  on  one  or  the 
other  kind  of  soil. 
The  method  should  be 
of  special  value  on 
areas  which  vary 
rather  uniformly  in 
one  direction. 

Increasing  the  num- 
ber of  trees  to  the  plot 
k^^"  \    \'^'\^  in  scattered  units  of 

^^y-         ,s>  T«L   _  either  four   or   eight 

trees  gives  a  more 
typical  sample  of  the 
productivity  of  the 
total  planting  than 
the  same  number  of 
adjacent  trees.  In 
scattering  the  plots 
throughout  the  area 
studied,  they  were  sys- 
tematically repeated. 
For  example,  if  there 
were  loo  plots  in  all  to 
be  grouped  in  pairs, 
the  first  and  fifty-first, 
and  the  second  and  fifty-second  were  united,  and  so  on  through  the  series. 
If  a'  quadruple  series  was  desired,  the  first,  twenty-sixth,  fifty-first,  and 
seventy-sixth  plots  were  combined. 

Table  IV  shows  the  results  of  scattering  4-  and  8-tree  plots,  respectively, 
in  the  plantations  studied.  Figure  8  illustrates  the  reduction  of  the 
coefficient  of  variability  by  increasing  the  number  of  trees  to  the  plot  in  *. 
both  4-  and  8-tree  scattered  units,  compared  with  the  average  coefficient 
of  variability  for  the  several  fruits  by  increasing  the  size  of  a  plot  from  i 
to  24  adjacent  trees,  together  with  the  theoretical  curve  calculated  from 
the  mean  coefficient  of  variability  of  all  the  i-tree  units. 

A  comparison  of  the  curve  for  adjacent  trees  and  those  for  scattered 
units  shows  at  once  the  marked  decrease  in  favor  of  the  scattered  units. 


Fig.  8. 


—Graphs  of  the  reduction  of  the  coefficient  of  variability  by 
increasing  the  number  of  trees  to  the  plot. 


Feb.4,i9i8     Variability  of  Yields  of  Fruit  Trees  and  Field  Trials     263 


Table  IV. — Effect  of  increasing  number  of  trees  per  plot  in  scattered  units 
FOUR  TREES  IN  A   UNIT 


Total 
num- 
ber of 
trees 
per 
plot. 

Coefficient  of  variability. 

Unit 
No. 

Navel  oranges. 

Valencia 
oranges. 

Eureka 
lemons. 

Seedling 
walnuts. 

Jonathan 
apples. 

Arlington. 

Antelope 
Heights. 

Average. 

I. . . . 
2. . . . 
3...- 
4.... 

S 

6.... 

4 
8 
12 
16 

20 
24 

31.52i1.00 
18.  59±o.  82 
15.  ilio.  81 

10.  76  io.  66 

11.  isio.  76 
9.  89  i  0.74 

22.  I3il.  00 
l3.o6io.  85 
li.07±o.88 
9.  S2±o.87 
S-  74±o.  58 
6.  38±o.  72 

27.75±i-8o 
16.  04ii.  43 
16.  48ii.  81 
10.  39±i.  29 
9.28±I.  28 
13.  16  i  2. 02 

22.  7oii.  10 
14.  76il.  07 
I2.6iii.  16 
8.  Olio.  81 
9. 48ii.  07 
S-38io.66 

30.09ii.  8s 
12.36i1.01 
14.  I7il.  41 
8.  59  i  0.99 
6.  97  i  0.89 
10.  OS  ±  1.4s 

26.  38  i  I.  80 
15.  89±i.47 
12.04i1.37 
8.  44ii.  08 
7.  7S±i-i3 
6.  i2±o.  97 

26.  76  i  0.60 
15.  I2io.  47 
13.  58io.  S3 
9.  29io.  40 
8.  40  i  0.40 
8.  49  i  0.49 

EIGHT  TREES  IN  A  UNIT 


28.  98ii.30 
17.  49±I09 
II.  78io.  89 
14- 85  i  I.  30 
10.  O9io.  96 


21.  i6±i.  35 

24.  54i2.  26 

20.  6sii.  57 

9.  2iio.  83 

15.  22il.  92 

10.  98ii.  17 

7.  44io.  84 

X4.34±2.  21 

II.  7oii.  44 

8.  87ii.  14 

8.72±i.57 

7-56ii.  16 

8.  6oi  I.  24 

5-  24ii.  02 

7.  8i±i.32 

25.  27i2.  16 

10.  94il.  27 

11.  3S±I.  63 
6.68ii.  13 

II.  oii2.  01 


24.  99i2.  39 
13.  20il.  69 
10.  99il.  7S 
10.  54il.  90 
4-94±I-0S 


24.  27io.  77 
12.  84io.  s6 
II.  27io.  63 
9.  S4±o.  57 
7-9S±o.S4 


The  4-tree  unit  apparently  gives  a  greater  degree  of  accuracy  than  the 
8-tree  unit  with  the  same  total  number  of  trees.  This  point  is  clearly 
shown  by  the  curves.  In  considering  a  total  number  of  either  8,  16, 
or  24  trees  the  curve  for  the  4-tree  unit  approaches  more  nearly  the 
theoretical  curve  than  the  curve  of  the  8-tree  unit.  With  a  total  num- 
ber of  24  trees,  for  example,  the  4-tree  units  would  be  scattered  about 
regularly  in  six  different  places,  while  the  8-tree  units  would  be  located 
in  three  different  places.  The  larger  number  of  unit  plots  thus  gives  a 
more  typical  sample  of  the  area  than  half  as  many  units  with  twice  the 
number  of  trees  in  a  unit. 

In  combining  both  the  4-  and  the  8-tree  units  in  the  regular  scattering 
of  ultimate  plots,  an  attempt  was  made  to  throw  both  high-  and  low- 
yielding  small  plots  into  a  combination,  although  a  systematic  distri- 
bution was  maintained.  The  fact  that  the  curve  of  the  4-tree  units 
drops  below  the  theoretical  in  one  place  indicates  that  this  attempt 
was  successful.  An  ordinary  regular  scattering  of  the  ultimate  plots 
might  not  approach  the  theoretical  curve  of  random  sampling  so  closely 
if  a  knowledge  of  the  relative  productivity  of  the  soil  was  not  available 
before  arranging  the  plots.  A  4-tree  unit  might  not  be  practical  in 
tests  of  cultural  methods,  fertilizer,  or  irrigation  trials.  In  such  trials 
an  8-tree  plot  is  usually  the  smallest  practical  unit.  In  the  case  of 
walnuts,  however,  which  should  be  planted  at  least  50  by  50  feet,  a 
4-tree  unit  in  a  single  row  would  allow  for  a  space  50  feet  wide  by  200 
feet  long,  and,  if  guard  rows  were  planted  between  the  experimental 
trees,  the  plots  would  be  100  by  250  feet,  a  very  practical  size  upon 
which  to  handle  orchard  machinery.  For  the  trial  of  rootstocks,  prun- 
ing experiments,  variety  tests,  etc.,  the  4-tree  plot  is  a  practical-sized 
27808°--18 3 


264  Journal  of  Agricultural  Research  voi.  xii,  no.  s 

unit  and  could  be  expected  to  give  more  reliable  results  if  repeated  at 
four  regularly-placed  intervals  than  either  two  8-tree  units,  or  16  ad- 
jacent trees — that  is,  such  a  regular  scattering  of  the  several  units 
which  make  up  the  combination  plot  reduces  the  error  of  the  final  com- 
parisons which  is  caused  by  the  variation  in  soil  productivity. 

The  fact  that  marked  soil  variations  occur  which  tend  to  make  ad- 
jacent trees  or  adjacent  plots  yield  alike,  even  on  soils  which  were 
chosen  because  of  their  apparent  uniformity,  is  well  shown  by  the  work 
of  Harris  (19 13).  The  criterion  for  the  measurement  of  such  variability 
proposed  by  this  author  is  the  coefficient  of  correlation  between  neigh- 
boring plots  of  the  field.*  Applying  this  to  the  Arlington  navel  oranges, 
the  writers  have  calculated  the  correlation  between  the  yield  of  the 
8-tree  plot  as  the  ultimate  unit,  and  the  yield  of  the  combination  of 
four  such  adjacent  plots  and  it  was  found  that 

r=  +0.533  ±0.085. 

This  shows  a  marked  correlation,  indicating  a  pronounced  hetero- 
geneity in  the  soil  of  this  grove,  influencing  fruit  production. 

However,  when  we  calculate  the  correlation  between  the  8-tree  plot  as 
the  ultimate  unit  and  the  yield  of  the  combination  of  four  such  system- 
atically scattered  plots,  it  is  found  that — 

r=  -f  o.i37±o.i20 

This  coefficient  is  practically  equal  to  its  probable  error  and  can  be 
regarded  as  significantly  zero.  This  is  merely  another  means  of  calcu- 
lating the  value  of  scattering  a  32-tree  plot  in  four  ultimate  plots  of  8 
trees  each  rather  than  selecting  32  adjacent  trees. 

DEGREE    OF    ACCURACY    EXPECTED   WITH    A    PLOT    OF    A    GIVEN    SIZE 

Assuming,  for  example,  that  experimental  plots  have  been  laid  out 
in  the  navel  oranges  (Arlington)  with  a  total  of  32  trees  to  the  plot  in 
four  scattered  units  of  eight  trees  each,  the  question  might  logically  be 
asked,  "What  differences  in  the  yields  of  such  plots  can  safely  be  at- 
tributed to  differential  treatment  as  different  methods  of  irrigation  or 
fertilization,  and  what  may  probably  be  due  to  mere  chance  because  of 
soil  heterogeneity  and  the  fluctuating  variation  of  the  trees?" 

Table  IV  shows  a  coefficient  of  variability  of  14.85  ±1.30  in  this 
plantation  laid  out  in  32-tree  plots  of  four  scattered  units  of  eight  trees 
each.  The  probable  error,^  then,  in  this  example,  that  such  a  plot  of 
32  trees  is  typical  of  the  area  in  question,  is  14.85X0.6745=  ±  10.02 

•  The  formula  used  is  ,  2 

r_    .      <[S(C  P)-S(P^)]lm[n(n-j)]}-p 

where  />==  yield  of  an  individual  plot;  ot=  number  of  larger  plots,  each  made  up  of  n  contiguous  ultimate 
units;  Cp=  yield  of  the  larger  combination  plots;  S=  summation  of  the  yields  of  all  the  ultimate  or  com- 
bination plots  of  the  field. 

*  The  probable  error  of  a  single  variant  of  a  population  may  be  defined  as  that  departure  from  the  mean 
on  either  side,  within  which  exactly  one-half  of  the  variants  are  found.  Expressed  as  a  percentage  of 
the  mean,  it  is  determined  by  multiplying  the  coefficient  of  variabihty  by  0.6745. 


Feb. 4, 1918      Variability  of  Yields  of  Fruit  Trees  and  Field  Trials     265 

per  cent  of  the  mean  production.  That  is,  the  chances  are  even  that  any 
plot  as  described,  of  32  trees,  will  fall  within  ±10.02  per  cent  above  or 
below  the  true  mean.  The  chances  are  equally  as  good  that  such  a  plot 
will  not  fall  within  the  accuracy  of  ±  10.02  per  cent  of  the  mean.  In 
comparing  two  such  plots,  both  with  the  same  probable  error  of  ±  10.02 
per  cent,  the  probable  error  of  such  a  comparison  will  be  greater  than 
the  probable  error  of  one — that  is,  it  will  be  equal  to  ±10.02  per  cent 
xV2=±  14-17  per  cent.  Therefore,  if  plots  undergoing  differential 
treatment  vary  from  each  other  by  only  ±14.17  per  cent  of  the  mean 
of  the  plantation,  half  the  time  such  differences  in  yield  may  be  due  to 
the  treatment,  and  half  the  time  they  may  be  due  to  casual  variation. 

It  is  clear,  then,  if  fertihzer  or  irrigation  experiments  laid  out  in  such 
plots  differ  from  each  other  by  only  14.17  per  cent  of  the  mean  produc- 
tion of  the  total  area,  we  are  not  assured  beyond  an  even  chance  that  the 
difference  is  a  real  difference  due  to  the  factors  which  are  being  experi- 
mented upon.  So  slight  an  assurance  can  hardly  be  expected  to  be 
sufficiently  reliable  to  prompt  a  farmer  to  purchase  fertilizer,  to  change 
his  method  of  irrigation,  or  to  undertake  any  new  business;  much  less 
will  this  assurance  justify  an  experimenter  in  drawing  conclusions  from 
the  result  of  a  field  trial. 

Our  present  knowledge  of  orchard  fertilization  in  the  arid  West  will 
hardly  justify  any  assumption  on  our  part  more  reliable  than  an  even 
chance  that  one  fertilizer  will  produce  an  increased  yield  of  fruit  com- 
pared with  another,  or  even  cause  an  increase  over  an  untreated  plot. 
The  same  thing  may  be  said  in  comparing  different  methods  of  irrigation. 
In  most  cultural  trials  we  would  therefore  be  comparing  two  results 
where  the  difference  may  occur  in  either  direction.     (Tables  V  and  VI. )^ 

Table  V. — Table  of  odds  for  differences  which  may  occur  in  cither  direction 


Difference  be- 

Difference 

tween  two  re- 

Odds against  such 

from  the  mean 

sults  in  terms 

difference  occurring 

in  terms  of 

of  probable 

under  uniforoi  condi- 

probable error. 

error  of  each 
result. 

tions. 

I.  00 

I.  41 

I  to  1 

1-25 

1.76 

3  to  2 

1.44 

2.03 

2  to  I 

I.  71 

2.41 

3  to  I 

I.  90 

2.68 

4  to  I 

2.  00 

2.83 

9  to  2 

2.05 

2.87 

5  to  I 

2.  50 

3-53 

10  to  I 

2-93 

4-  13 

20  to  I 

3.00 

4.  24 

22  to  I 

3.20 

4-51 

30  to  I 

4.  00 

5.66 

140  to  I 

4.90 

6-93 

1 ,00c  to  I 

5.  00 

7.07 

1,350  to  I 

'  Tables  V  and  VI  are  taken  from  the  writings  of  Wood  (igii)  who  in  turn  adopted  thexn  "from  one  of 
the  standard  reference  books  on  astronomy. " 


266 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  s 


Tabi<E  VI. — Table  of  odds  for  differences  which  may  occur  in  one  direction  only 


Difference  be- 

Difference 

tween  two  Re- 

Odds against  such 

from  the  mean 

sults  in  terms 

difference  occurring 

in  terms  of 

of  probable 

under  uniform  condi- 

probable error. 

error  of  each 
result. 

tions. 

I.  OO 

1.41 

3  to  I 

I.  25 

1.76 

4  to  I 

1.44 

2.03 

5  to  I 

1.58 

2.23 

6  to  I 

I.  71 

2.41 

7  to  I 

I.  81 

2.55 

8  to  I 

I.  90 

2.68 

9  to  I 

2.  00 

2.83 

10  to  I 

2.48 

3-50 

20  to  I 

2.  70 

3.81 

30  to  I 

2.89 

4.07 

40  to  I 

3.00 

4.24 

44  to  I 

3-03 

4.  28 

50  to  I 

3-44 

4-85 

100  to  I 

4.  00 

5.66 

290  to  I 

5.00 

7.07 

2 ,  700  to  I 

On  assuming  that  a  lo-to-i  chance  is  a  reasonable  assurance,  the  ques- 
tion logically  arises,  "What  difference  between  any  two  plots  must  be 
manifested  for  this  degree  of  confidence  that  the  difference  is  due  to 
treatments  applied?"  By  referring  to  columns  2  and  3,  Table  V,  one 
may  find  the  difference  in  terms  of  the  probable  error  which  is  necessary 
between  two  results  to  obtain  this  degree  of  reliability.  Here  it  can 
be  seen  in  column  2  that  there  must  be  a  difference  3.53  times  the  prob- 
able error  to  give  the  odds  of  10  to  i  in  column  3  against  such  a  difference 
occurring  under  uniform  conditions.  Thus  we  find  in  this  example  that 
the  difference  between  two  32 -tree  plots  in  four  scattered  units  must  be 
at  least  14.17  per  cent  X  3.53  =  50.02  per  cent  of  the  mean  production, 
to  give  the  assurance  of  a  lo-to-i  chance  that  the  difference  is  due  to 
fertilizer,  irrigation,  or  whatever  factors  are  under  consideration.^ 
Even  with  this  difference,  conclusions  based  on  such  results  obtained  in 
this  navel  orange  (Arlington)  grove  may  be  correct  10  times  out  of  11 
and  wrong  once  out  of  11  times. 

On  turning  now  to  the  32-tree  plot  of  adjacent  trees  with  navel  oranges 
(Arlington),  it  is  seen  the  probable  error  is  16.42  per  cent.  To  pro- 
ceed as  before,  16.42  per  cent,  the  probable  error  of  one  plot,  X  V^  = 
23.22  per  cent,  the  probable  error  of  the  difference;  23.22X3.53  =  81.97 
per  cent.  Therefore  a  difference  between  two  such  plots  of  81 .97  per  cent 
of  the  mean  of  the  total  area  would  be  necessary  to  give  the  assurance 
that  such  differences  are  real  10  times  out  of  11  and  due  to  pure  chance 

1  If  comparisons  were  being  made  between  two  radically  different  treatments  which  were  known  to  pro- 
duce different  effects  in  fruit  production,  such  as  irrigation  compared  with  dry  farming,  or  the  use  of  large 
quantities  of  stable  manure  on  light  soils,  compared  with  no  manuring,  then  reference  should  have  been 
made  to  columns  2  and  3  in  Table  VI,  which  are  based  on  the  differences  occurring  in  one  direction  only., 


Feb.  4, 191 8     Variability  of  Yields  of  Fruit  Trees  and  Field  Trials     267 


only  once  out  of  11  times;  yet  this  grove  was  chosen  because  of  its  ap- 
parent regularity,  for  it  has  been  judged  sufficiently  uniform  for  plot  trials. 

The  point  might  be  justly  raised  that  the  small  number  of  plots  in- 
volved in  the  above  calculations  are  not  sufficient  to  give  the  laws  of 
chance  a  fair  opportunity  of  asserting  themselves.  On  laying  the  area  out 
in  plots  of  adjacent  trees  there  were  30  plots,  while  made  up  of  scattered 
units  there  were  31  plots.  Figure  9  shows  the  distribution  of  these  plots, 
together  with  the  theoretical  cur\^e,  which  was  caclulated  for  the  scat- 
tered-unit curve.  The  scattered-unit  curve  closely  approaches  the  theo- 
retical normal  curve 
of  errors,  and  there-  I     '     ■" 

fore  reliance  can  be 
placed  upon  its 
probable  error. 

In  the  case  of  the 
plots  of  adjacent 
trees,  however,  the 
30  units  are  not  suf- 
ficient to  give  the 
laws  of  chance  fair 
play.  Table  VII 
sums  up  the  results 
of  the  foregoing  cal- 
culations, adding  ex- 
treme and  mean 
yields  of  the  two  dif- 
ferent types  of  plots 
and  the  theoretical 
probable  error.  The  theoretical  probable  error  based  on  the  theory  of 
random  sampling  for  a  hypothetical  32-tree  plot  is  the  probable  error  of 

one  tree  26.67-^ -^32  =4-71  P^'^  cent.  This  is  readily  calculated  from 
the  distribution  of  the  yields  on  a  one-tree  unit,  the  curve  of  which  is 
shown  by  figure  10.  The  large  number  of  trees  involved,  even  though 
the  distribution  is  not  normal,  justifies  the  use  of  the  probable  error  as  a 
minimum  probable  error.  Based  upon  the  theory  of  random  sampling, 
two  hypothetical  32-tree  plots  with  a  probable  error  of  4.71  per  cent 
should  show  a  minimum  difference  of  (4.71  X  V  2  X  3.53)  =  23.51  per  cent 
to  give  an  assurance  of  a  lo-to-i  chance  that  such  a  difference  is  real 
and  not  due  to  casual  variation.  Therefore,  if  the  calculations  in  Table 
VII  based  on  adjacent  trees  can  not  be  fully  relied  upon  because  of  the 
small  number  (30)  in  the  population  and  because  their  distribution  is  not 
normal,  we  may  at  least  reasonably  expect  that  the  necessary  difference 
between  two  such  plots  will  fall  between  the  theoretical  23.51  per  cent 
and  81.97  with  a  practical  certainty  that  it  will  be  greater  than  50.02  per 
cent  of  the  mean. 


2      ^       -F       ^       e       7-      e 

Fig.  9. — Graphs  of  production,  32-tree  plot,  navel  oranges  (Arlington). 

Scattered  in  four  8-tree  units. 

Adjacent»32-tree  units. 


268 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  s 


Tabl^  VII. — Comparison  of  the  reliability  of  a  plot  of  32  adjacent  trees  with  that  of  a  J2- 
tree  plot  of  four  scattered  units  of  8  trees  each.     Navel  oranges  {Arlington) 


E.Ktreme  yields 
per  plot. 

Mean  yield 
per  plot. 

Standard 
deviation. 

Probable 
error  per 
plot  (per- 
centage 
of  mean). 

Difference  necessary 
to  give  lo-to-i  as- 
surance. 

Plot. 

Percent- 
age of 
the  mean. 

Observed 
yield. 

Pounds. 
2,500  to  7,500 
3,000  to  6,000 

Pounds. 
4,367±I3I 
4,484±  81 

Pounds. 
i,o63±93 
666±57 

±16.  42 
±  10.  02 

±81.97 
±50.02 

Pounds. 
±3.s8o 

±2.243 

Theoretical  (based  on  ran- 

±4-  71 

±23-51 

± 1 1 03s 

On  turning  now  to  the  navel  oranges  (Antelope  Heights),  which  the 
calculations  show  to  be  the  most  uniform  planting  of  any  observed,  the 

question  might  arise 
as  to  what  degree  of 
reliability  may  be  ex- 
pected in  comparing 
two  i6-tree  plots. 
Table  VIII  shows  a 
comparison  between 
such  plots  made  up  of 
adjacent  trees  and 
plots  composed  of 
scattered  units  of  four 
trees  each  together 
with  the  theoretical. 
A  much  greater  range 
is  found  with  the  plot 
of  adjacent  trees,  while 
the  necessary  differ- 
ence between  two 
plots  for  a  10- to- 1  as- 
surance is  62.94  per 
cent.  With  plots 
made  up  of  scattered 
units  the  necessary 
difference  is  32.05  per 
cent.     The     hetero- 


!0  ^  10 
f\j  W^  K 


\ 


Fig.  10.— Curve  of  yields  of  individual  trees,  navel  orange  (Arlington). 

geneity  of  this  soil  is  clearly  shown  by  both  the  increase  in  range 
and  the  increased  probable  error  when  plots  are  composed  of  adja- 
cent trees.  This  block  of  trees,  however,  appeals  to  the  observer  as 
unusually  uniform  and  would  be  considered  desirable  for  plot  experimen- 
tation, the  fluctuation  in  the  productivity  of  the  trees  approaching  closely 


Feb.  4, 1918     Variability  of  Yields  of  Fruit  Trees  and  Field  Trials     269 


the  normal  curve  (see  fig.  11);  the  mean  production  per  tree  is  186.2  ± 
1.7  pounds,  standard  deviation  55-33±ii9,  coefficient  of  variability 
29.72  ±0.69  per  cent,  probable  error  20.05  per  cent  of  the  mean,  skewTiess 

o.ooi  ±0.037.    Never-  

^  r 


theless,  if  devoted  to 
plot  experiments,  a 
difference  between 
two  plots  of  16  adja- 
cent trees  each,  of 
even  62.94  per  cent  of 
the  mean  production, 
might  be  due  to  differ- 
ential treatment  10 
times  out  of  11  and 
due  to  casual  varia- 
tions of  soil  and  trees 
once  out  of  11.  The 
calculations  sound  a 
note  of  warning 
against  drawing  con- 


FiG.  II.— Curve  of  yield  of  individual  trees,  navel  orange  (Antelope 
Heights). 


elusions  between  such  plots  if  the  differences  are  less  than  50  per  cent  of 
the  mean  production  of  the  plantation,  provided  we  wish  to  have  such 
conclusions  as  dependable  as  a  lo-to-i  chance. 

Table  VIII. — Comparison  of  the  reliability  of  a  plot  of  i6  adjacent  trees  with  that  of  a 
plot  of  16  trees  of  four  scattered  units  of  4  trees  each.     Navel  oranges  {A  ntelope  Heights) 


Plot 

Extreme  yields 
per  plot. 

Mean  yield 
per  plot. 

Standard 
deviation. 

Probable 
error  per 
plot  (per- 
centage 
of  mean). 

Difference  necessary 
to      give      lo-to-i 
assurance. 

1 
Percent-    ^,            , 
age  of      0^?:"'^'l 
the  mean.       y^'-^- 

1 

Pounds. 
1,800  to  4,000 
2,400  to  3,600 

Pounds. 
2,973±68 
2,96o±36 

Pounds. 

556±48 

282±25 

±12.61 
±  6.42 

±62.  94 
±32-05 

Pounds. 

±1.871 
±     949 

±  S-oi 

±-'5-03 

±     746 



An  example  taken  from  the  Jonathan  apple  orchard  may  well  be  con- 
sidered. Suppose  it  is  desired  to  know  the  necessary  difference  which 
must  exist  between  plots  of  16  trees  each  to  give  us  the  reliance  of  a  lo- 
to-i  chance  that  it  is  due  to  differential  treatment  (a)  when  the  plots  are 
made  up  of  adjacent  trees,  (b)  when  the  plots  are  made  up  of  two  scattered 
units  of  eight  trees  each,  and  (c)  when  the  plots  are  made  up  of  four 
scattered  units  of  four  trees  each. 

The  probable  errors  in  the  above  cases,  a,  b,  and  c,  are  ±  13.49,  ±8.90, 
and  ±5.69  per  cent,  respectively.  By  proceeding  as  before  it  is  found 
that   the   necessary   difference   for  a    lo-to-i    chance   is   as  follows:  (a) 


270 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  s 


±67.35  per  cent;  (b)  ±44.44  per  cent;  (c)  ±28.42  per  cent.  Table  IX 
summarizes  the  results.  It  seems  probable,  therefore,  that  a  difference 
between  two  16  adjacent  tree  plots  of  less  than  50  per  cent  of  the  mean 
production  should  be  considered  with  caution  before  attributing  it  to 
differential  treatment.  The  scattering  of  the  units  of  the  plots  increased 
the  accuracy  very  decidedly,  four  units  giving  more  accurate  comparison 
than  two.  Even  with  this  scattering,  differences  of  less  than  30  or  40 
per  cent  are  well  within  the  realm  of  chance.  The  apparent  cause  for  the 
4-unit  plot  having  a  probable  error  less  than  the  theoretical  is  accounted 
for  by  the  fact  that  the  variation  in  productivity  of  the  soil  was  known 
when  the  distribution  of  the  units  was  made.  The  results  might  not  have 
approached  so  closely  to  the  theoretical  if  the  distribution  had  been 
decided  upon  before  harvest. 

Table  IX. — Comparison  of  the  reliability  of  a  plot  of  16  adjacent  trees  with  that  of  two 
units  of  8  trees  each  and  of  four  units  of  4  trees  each.     Jonathan  apples 


Extreme  yields 
per  plot- 

Mean  yield 
per  plot. 

Standard 
deviation. 

Probable 
error  per 
plot  (per- 
centage 
ofmean)- 

Difference  necessary 
to      give      lo-to-i 
assurance. 

Percent- 
age of 
the  mean. 

Observed 
yield. 

Pounds. 
3,000  to  6,500 
3,900  to  6,100 
3,800  to  5,400 

Pounds. 

4,864±i76 
4,857±ii5 
4,87i±   74 

Pounds. 

974±I24 
641 ±    82 
4II±    52 

±13-49 
±  8-90 
±  5-69 

±67-35 

±44-44 
±28.42 

Pounds. 

±2,158 

± I , 384 

Theoretical    (based    on 

±  6-9S 

±34-  70 

±1,687 

The  large  probable  errors  which  are  apparently  always  present  with 
plot  trials  of  fruit  trees  emphasize  the  importance  and  value  of  obtaining 
individual  tree  records  of  experimental  orchards  before  differential 
treatment  is  started.  The  probable  errors  will  likely  differ  somewhat 
from  year  to  year,  and  possibly  be  further  influenced  by  the  advanced 
age  of  the  trees.  Nevertheless,  if  but  one  or  two  years '  records  of  mature 
trees  are  available  before  differential  treatment  is  commenced,  at  least 
some  idea  can  be  obtained  of  the  casual  variation  of  the  plots — that  is, 
limitations  can  be  placed  beyond  which  observed  differences  in  plots 
may  be  due  to  chance  rather  than  to  the  factors  under  experimentation. 
In  the  absence  of  such  previous  records,  the  employment  of  frequent 
controls  or  standard  treatments  may  be  indicative  of  the  probable  error 
of  the  entire  area. 

REIyATlON    OF    THE    SHAPE    OF    THE    PLOT    TO    THE    VARIABIUTY    OF  THE 

COMPARATIVE   YIELDS 

The  shape  of  small  plots  may  be  of  great  importance  when  cultural 
operations  are  considered. 

Lyon  {igi2)  found  no  satisfactory  evidence  that  long  and  narrow 
plots  are  less  likely  to  error  than  square  plots  when  no  control  plots  are 


Feb.4,i9i8     Variability  of  Yields  of  Fruit  Trees  and  Field  Trials     271 

used,  but  when  controls  are  placed  every  second  or  third  plot  in  the 
series,  the  evidence  is  in  favor  of  the  long  and  narrow  plots.  The  use  of 
every  second  or  third  plot  in  the  series  as  a  control  plot  obtains  greater 
accuracy  than  when  no  controls  are  used  and  the  average  of  the  field  is 
considered  the  normal  yield  for  all  plots. 

A  long  plot  is  much  more  economical  of  time  and  labor  than  a  square 
plot  containing  the  same  number  of  trees.  If  the  orchard  must  be  irri- 
gated, a  square  plot  containing  nine  trees  will  require  three  standpipes 
instead  of  one  and  also  three  times  as  many  irrigation  furrows. 

Arguments  have  been  advanced  from  time  to  time  in  favor  of  both 
the  linear  and  the  square  plot.  The  advocates  of  the  former  have 
urged  its  reliability  on  two  points:  first,  if  the  soil  or  other  conditions 
change  in  a  direction  parallel  to  the  plot,  it  will  contain  both  high-  and 
low-yielding  areas,  and  the  average  will  correct  one  or  the  other  error; 
second,  if  conditions  change  in  a  direction  more  or  less  perpendicular  to 
the  plots,  each  plot  will  vary  from  its  adjoining  plot  because  of  its  shape 
by  such  a  minimum  that  intercomparisons  are  more  reliable.  On  the 
other  hand,  the  advocates  of  the  square  plot  have  claimed  that  the 
arguments  apply  equally  well  to  their  case,  provided  the  plot  be  made 
small  enough. 

In  view  of  the  importance  of  the  shape  of  the  plot  for  cultural  opera- 
tions, the  writers  have  investigated  the  variability  of  plots  of  various 
sizes  and  shapes  in  three  of  the  experimental  orchards.  The  results 
(Table  X)  are  for  unrepeated  plots  and  for  one  year's  yield  of  fruit, 
except  the  apples,  which  are  for  two  years.  Considerable  interest  lies 
in  the  computation  on  the  9-tree  plot  in  the  total  navel-orange  grove, 
because  it  is  based  on  a  large  population  (i,cxx)  trees)  and  because  the 
soil  is  known  to  vary  in  a  general  way  from  northeast  to  southwest. 

Comparisons  were  made  between  square  plots  containing  9  trees  and 
linear  plots  consisting  of  a  single  row  of  9  trees  extending  in  the 
north-south  and  in  the  east-west  direction.  It  is  interesting  to  see  how 
closely  the  coefficients  of  variability  coincide  in  the  three  cases.  The 
differences  between  the  coefficients  are  only  a  fraction  of  the  probable 
errors.  Therefore  on  this  lot  of  i  ,000  orange  trees  there  is  no  difference 
between  a  square  and  a  linear  plot  of  9  trees,  so  far  as  the  reliability 
of  comparative  yields  is  concerned.  (The  deviations  were  taken  from 
the  mean  of  all  the  plots  of  the  grove.) 

Similar  comparisons  were  made  on  a  selected  block  of  256  trees  in  this 
same  grove.  The  coefficient  of  variability  of  the  trees  taken  singly  is 
34.47  ±1.14,  which  indicates  that  these  trees  were  a  fairly  typical  sample 
of  the  entire  grove.  The  block  was  divided  into  plots  of  various  shapes 
containing  16  trees.  With  one  exception  there  is  little  difference  in  the 
variability  of  the  plots,  regardless  of  their  shape.  This  exception  is 
found  in  the  linear  plot  of  i  by  16  trees  extending  from  east  to  west, 
which  has  a  lower  coefficient  than  any  other  arrangement.     In  the  block 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  5 


of  trees  chosen  there  is  a  gradual  variation  of  soil  in  this  direction,  and 
these  plots  therefore  include  both  high-  and  low-yielding  trees  in  about 
equal  proportions,  the  difference  from  plot  to  plot  being  small,  as  shown 
in  Table  X. 

TabIvE  X. — Effect  on  variability  of  changing  the  shape  of  plots 


Kind  of  trees. 


Total  navel-orange  grove  (Arlington) 

Do 

Do 

Portion  of  navel-orange  grove  (Arlington), 
256  trees 

Do 

Do 

Do 

Do 

Do 

Walnuts,  1915 

Do 

Do 

Do 

Do 

Do 

Do 

Apples,  1914-15 

Do 

Do 

Do 

Do 

Do 

Do 


Number 
of  trees 
per  plot. 


16 
16 
16 
16 
16 

4 
4 

9 
9 
9 

24 

24 


16 
16 

24 
24 


Shape  of  plots. 


3  by  3 

I  by  9  N.-S.  . 
I  by  9  E.-W. 

1  by  I 

4  by  4 

2  by  8  N.-S.  . 
2  by  8  E.-W. 
I  by  16  N.-S. 

1  by  16  E.-W 

2  by  2 

I  by  4 

3  by  3 

I  by  9  N.-S. . 
I  by  9  E.-W . 

4  by  6 

1  by  24 

2  by  4 

I  by  8  N.-S. . 

1  by  8  E.-W. 

4  by  4 

2  by  8  N.-S. . 

4  by  6 

I  by  24 


Coefficient  of 
variability. 


28.  i8±l.  48 
28.  92  ±  I.  46 
28.  36±i.  43 

34- 47  ±  I-  14 

16.  77  ±2.  05 

17.  i7±2.  10 
iS.39±i.88 
16.  36 ±2.  00 
io.09dbi.  31 

31.  09±i.  93 
30.40±i.  89 

23.  94±2.  32 

26.  23±2.  57 
22.  46±2.  16 

21.  62±3.  25 
6.  56±i.  00 

19.  32±I.  81 

18.  7o±i.  74 
21.  35±2.  18 

15-  i3±i-97 
16.  4i±2.  15 

i3-75±2.36 
12.  8i±2.  20 


Changing  the  shape  of  plots  of  walnut  trees  had  little  effect  upon  the 
coefficient  of  variability  except  in  the  case  of  the  larger  plots.  The 
coefficients  vary  only  slightly  from  one  another,  regardless  of  shape,  in 
the  case  of  the  4-  and  9-tree  plots.  In  the  24-tree  plot,  however,  there 
is  a  great  difference  between  the  4-  by  6-  tree  plot  and  the  i-  by  24-  tree 
plot  in  favor  of  the  latter.  It  should  be  borne  in  mind,  nevertheless, 
that  a  plot  of  24  walnut  trees  is  an  abnormally  large  plot,  and  in  the 
example  just  referred  to  it  is  a  plot  250  feet  by  350  feet,  compared  with 
one  100  feet  by  1,250  feet.  This  may  take  into  consideration  marked 
variation  in  productivity,  even  on  apparently  uniform  soil. 

In  the  case  of  the  apple  trees  there  is  little  difference  between  the 
variability  of  plots  of  equal  size  but  different  shapes,  even  in  the  case 
of  the  24-tree  plots.  Regarding  the  question  of  the  shape  of  the  plot, 
it  therefore  appears  that  in  most  cases  there  is  no  difference  in  the  varia- 
bility of  a  linear  or  a  square  plot. 


Feb.4,i9i8     Variability  of  Yields  of  Fruit  Trees  and  Field  Trials     273 

proportional  to  their  distance  from  the  plot.  This  would  obviously  be 
In  the  case  of  a  large  number  of  trees,  where  a  linear  plot  is  long 
enough  to  include  both  high-  and  low-yielding  areas  in  each  plot,  there 
appears  to  be  an  advantage  in  favor  of  the  hnear  plot.  The  great  advan- 
tage of  the  hnear  plot  in  most  cases  is  the  economy  of  cultural  operations. 
Another  advantage  will  be  discussed  in  the  following  section  on  the 
interpolation  of  control  plots. 

A  carefully  conducted  fertilizer  experiment  with  orchard  trees  requires 
a  guard  row  between  adjacent  plots  in  order  to  eliminate  the  possibility 
of  the  fertilizer's  affecting  the  margin  of  the  next  plot,  since  there  are 
many  observations  which  show  that  tree  roots  extend  considerable  dis- 
tances and  often  reach  the  lines  of  adjoining  rows  (Hedrick,  1914;  Ballan- 
tyne,  1916).  It  is  obvious  that  it  is  more  reliable  to  employ  guard  rows 
rather  than  to  divide  cultural  treatments  midway  between  tree  rows. 
The  number  of  guard  trees  required  for  square  plots  is  smaller  than  for 
a  linear  plot  containing  the  same  number  of  trees.  A  9-tree  plot  in  the 
form  of  a  square  requires  7  guard  trees;  in  the  form  of  a  linear  plot  it  ■ 
requires  1 1  guard  trees.  In  the  former  case  seven-sixteenths  of  all  trees 
are  in  the  guard  rows,  in  the  latter  eleven-twentieths.  As  the  size  of  the 
plot  increases,  the  difference  becomes  greater.  It  thus  becomes  a  ques- 
tion of  the  extent  to  which  one  is  willing  to  go  in  enlarging  the  size  of  the 
linear  plot  before  the  increase  in  the  number  of  trees  in  guard  rows  offsets 
the  economy  in  cultural  operations  thus  obtained. 

USE  OF  CONTROL  PLOTS 
INTERPOLATED  CONTROL  PLOT 

Agronomists  are  in  the  habit  of  using  every  third,  fourth,  or  fifth  row 
or  plot  in  the  experimental  tract  as  a  standard  from  which  the  normal 
yield  of  any  intervening  treated  plot  may  be  calculated.  The  nature  of 
the  crops  and  the  cultural  methods  used  commend  this  system  of  arrange- 
ment. 

This  method,  like  others,  has  its  advantages  and  disadvantages.  After 
the  results  are  obtained  there  is  still  a  need  for  a  proper  method  of  com- 
parison. There  are  several  different  methods  of  estimating  the  ' '  normal ' ' 
yield  of  any  plot.     The  "normal"  (A^)  may  be  estimated  by  the  formalu 

Q-fQ-f-    ...Cn 
n 

which  is  simply  the  mean  of  all  the  control  plots  in  the  area.  If  the 
soil  of  the  area  were  uniform  and  all  variations  in  the  yield  of  the  controls 
were  purely  chance  variations,  this  method  would  give  a  satisfactory 
result.  Again  N  may  be  estimated  from  the  yields  of  the  two  nearest 
control  plots.  For  example,  if  every  third  plot  is  a  control  and  the  ar- 
rangement is  Cj,  A,  B,  Cj,  and  so  on,  the  normal  for  A  would  be 
yiC^  +  y^C^-  In  this  way  the  yields  of  the  controls  receive  weights  in- 
versely proportional  to  their  distance  from  the  plot.     This  would  ob- 


274  Journal  of  Agricultural  Research  voi.  xii.  No.  s 

viously  be  satisfactory,  provided  there  was  no  difference  in  the  amount 
of  variation  between  the  plants  on  different  plots  and  that  the  soil  varied 
uniformly  in  one  direction. 

The  first  method  mentioned  computes  the  ' '  normal ' '  (N)  for  the  whole 
area;  the  second,  for  a  locus  on  that  area.  These  normals  may  be  com- 
bined to  represent  the  resultant  of  both  general  and  local  conditions. 
Thus,  the  formula  N=}4  (C+  ^Q  +  J^Cj)  indicates  that  N  is  the  mean  of 
the  values  for  N  computed  by  the  two  preceding  formulas.  This  assumes 
an  equal  value  for  the  adjacent  control  plots  and  the  mean  of  all  control 
plots.  Its  use  as  N  brings  up  the  calculated  yields  of  plots  on  low- 
yielding  areas  and  reduces  the  same  on  high-yielding  areas.  In  the  case 
of  cereals  there  is  usually  small  chance  for  difference  in  the  yielding  powers 
of  a  plot  and  its  nearest  control  on  account  of  their  proximity;  but  in 
the  case  of  orchard  trees  situated  some  distance  apart  there  may  be 
greater  soil  changes  between  adjacent  plots,  and  consequently  a  marked 
difference  in  yield,  aside  from  the  effect  of  treatment,  between  a  plot  and 
its  nearest  control  plots.  The  introduction  of  the  mean  of  all  control 
plots  might  be  expected  to  introduce  a  stabilizing  factor.  In  the  formula 
^2  (C+^Ci  +  J4C2)  the  mean  of  all  control  plots  has  equal  weight  with  the 
normal  derived  from  the  nearest  controls.  Since  the  soil  over  any  but 
very  small  areas  may  not  be  uniformly  variable,  it  might  seem  more 
logical  to  weight  the  normal  derived  from  the  nearest  controls  more 
heavily  than  that  derived  from  the  mean  of  all  control  plots,  and  to 
combine  the  two.     This  has  been  done  by  Olmstead  (1914)  and  others, 

making  the  formula  ^^^ 1  L  h —  ^^  which  p^  and  p2  are  constants 

Pl'p2 

arbitrarily  chosen.  Stockberger  {1916)  found  satisfactory  results  by 
assigning  the  values  pi=i  and  p2=3- 

The  method  used  by  the  Office  of  Cereal  Investigations,  of  the  Bureau 
of  Plant  Industry,  is  K(c+Q).  which  employs  half  the  sum  of  the  mean 
of  all  control  plots  and  the  yield  of  the  nearest  control  plot  as  the  normal 
for  any  given  plot. 

Stockberger  {1916)  compared  the  relative  precision  of  these  formulas 
in  computing  the  normal  yields  of  plots  of  hops.  Using  the  yields  of 
six  years  he  obtained  the  greatest  precision  from  the  formula 

P,  +  P2 

though  no  formula  maintained  the  same  relative  rank  throughout  the 
six  years.  It  would  appear  that  there  is  no  way  of  determining  in 
advance  the  formula  best  suited  to  any  particular  case,  at  least  not 
until  more  applications  of  the  different  formulas  have  been  studied. 

The  five  formulas  above  stated  have  been  tested  on  the  Arlington 
grove  of  navel  oranges.  The  grove  was  parceled  into  linear  plots  of 
10  trees  each.     Each  alternate  plot  was  designated  as  a  guard  row  and 


Feb.4.i9i8     Variability  of  Yields  of  Fruit  Trees  and  Field  Trials     275 


discarded  from  the  calculation.  This  left  32  "treated"  plots  whose 
"normal"  yields  were  computed.  The  arrangement  of  these  plots  and 
their  yields  are  given  in  Table  XI. 

Table  XI. — Arrangement    and    yields    of   lo-tree    plots    of    the    navel-orange   grove 

{Arlington) 

[Yield  expressed  in  tens  of  pounds] 

108G  84* 

150*  88B 

112  control  119* 
124*  146  control 

121F  151* 

159*  174C 

142E  132* 

148*  182 D 

139  control  166* 
153*  174  control 

177D  180* 

179*  169E 

176C  131* 

188*  162F 

183  control  183* 
178*  147  control 

203B  188* 

182*  168G 

206A  223* 

199*  201H 

219  control  191* 
191*  228  control 


80  control 

116* 

105* 

152  control 

88H 

184* 

115* 

155A 

117G 

164* 

94* 

154B 

109  control 

160* 

118* 

163  control 

94F 

147* 

76* 

122C 

83E 

142* 

97* 

137D 

104  control 

III* 

lOI* 

131  control 

99D 

148* 

117* 

154E 

looC 

165* 

112* 

134F 

74  control 

III* 

lOI* 

114  control 

112B 

109* 

107* 

123G 

98A 

78* 

108* 

97H 

80  control 

80* 

108* 

106  control 

108H 

137* 

no* 

135A 

Computing  the  normal  yields  of  these  plots  by  the  five  formulas  above 
described  gives  the  results  shown  in  Table  XII. 

Table  XII. — Value  of  different  formulas  for  computing  the  coefficient  of  variability  and 
probable  error  of  yields  of  oranges  in  lO-tree  plots  in  Arlington  grove 


Formula  used. 

standard 
deviation. 

CoelEcient  of 
variability. 

Probable 

error  ( per- 

centaKC  of 

mean). 

I    C,4-Co                     +Cn. 

Pounds. 
3(>3±3^ 

I7i±i4 
2i9±i8 
i8o±is 

233±20 

26.  S5±2.  39 

12.  52±I.  07 

16.  or±i.  38 
12.  93  ±  I.  11 

16.  78±i.46 

17.91 
8.44 

n 
2.  2^(7  j_izc 

,    l4(C+^4C,  +  ]/iC^)  

10.80 

.    i>,C-\-t>o(-AC,  +  i4C-,)  

9.  00 

11.32 

c    U(C-\-C,)                     

*  Guard  row. 


276  Journal  of  Agricultural  Research  voi.  xii.  No.  s 

These  figures  show  that  there  is  less  error  in  this  case  by  the  use  of 
formulas  (2)  and  (4)  and  that  there  was  very  little  difiference  between 
these  two. 

This  question  of  a  "normal"  yield  of  a  plot  depends  obviously  upon 
the  portion  of  the  population  chosen  as  a  standard  and  upon  the  method 
of  calculation.     Investigators  differ  in  the  choice  of  both  of  these  factors. 

An  illustration  may  be  given  to  show  the  results  of  calculating  the 
normal  from  different  standards.  The  coefficient  of  variability  of  the 
lo-tree  plot  of  Arlington  navel  oranges  was  computed,  taking  the  devia- 
tions from  (a)  the  mean  of  all  plots,  (b)  the  mean  of  all  control  plots, 
and  (c)  the  "normal"  calculated  by  formula  2.  The  results  are  shown 
in  Table  XIII. 

TabIvE  XIII. — Deviation  of  yields  from  mean  of  area  compared  with  deviation  from  lo-iree 

linear  control  plot 


For- 
mula 
No. 

Deviation. 

Coefficient  of 
variability. 

Deviation  taken  from  mean  of  all  plots 

26.  00±2.  30 

26.  55 ±2.  39 
12.  5±i.  10 

Deviation  taken  from  mean  of  control  plots      .   .     .         .... 

3 

Deviation  taken  from  normal  calculated  from  nearest  control 

Since  no  differential  treatments  had  been  given  these  plots,  the  mean 
of  all  plots  differs  little  from  the  mean  of  all  control  plots,  the  respective 
means  being  1,376  and  1,367  pounds.  The  variability  of  plots  calcu- 
lated from  these  two  standards  is  not  significantly  different.  In  the 
case  of  the  deviation  taken  from  the  normal  yield,  however,  there  is  a 
very  significant  decrease  in  variability,  since  there  are  discontinuous  soil 
variations  not  recognizable  in  the  general  average  of  the  area  which  are 
taken  into  account  by  this  formula. 

CHANCE   ARRANGEMENT   OF  CONTROL  PLOTS 

A  weakness  of  this  system  of  comparison  with  interpolated  plots  lies 
in  the  possibility  that  the  series  of  control  plots  may  not  be  representa- 
tive of  the  area.  The  plots  chosen  for  controls  may  be  on  soil  superior 
or  inferior  to  that  of  the  intervening  plots.  There  are  indications  that 
this  possibility  may  be  more  real  than  one  would  expect  from  purely 
random  sampling. 

A  few  computations  will  show  the  extent  to  which  the  different  methods 
of  choosing  control  plots  may  affect  the  results.  The  Arlington  grove 
records  were  recomputed,  shifting  the  control  plots  back  one  row.  The 
yield  of  the  first  control  plot  (see  Table  XI)  was  1,051  pounds  instead  of 
800,  and  the  last  was  1,160  instead  of  1,520  pounds.  The  first  arrange- 
ment will  be  termed  "arrangement  A,"  the  second  "arrangement  B." 
The  mean  yield  of  all  control  plots  is  not  greatly  changed  by  the  different 


Feb.4,i9i8     Variability  of  Yields  of  Fruit  Trees  and  Field  Trials     277 

arrangement.  The  mean  of  the  "A"  control  plots  is  1,367  pounds,  that 
of  the  "B  "  control  plots  is  1,389  pounds,  yet,  as  shown  in  Table  XIV,  the 
coefficient  of  variability  is  increased  from  12.5^1.1  in  "arrangement  A" 
to  18.2 ±1.6  in  "arrangement  B"  for  single  plots,  with  similar  increases 
for  the  repeated  plots. 

Table  XIV. — Effect  of  two  methods  of  arranging  control  plots 


Coefi&cient  of  variability. 

Arrangement  of  plots  of  lo  trees. 

Repeated  once. 

Repeated  twice. 

Repeated 
three  times. 

Repeated 
four  times. 

Arrangement  A  ^Arlington). . . 
Arrangement  B  (Arlington). . . 

12.  52  ±1.07 

18.  2S±I.  59 

7.  53  ±0.89 
16.  o7±i.  97 

7.  68±i.  17 
12.  46±i.  91 

4-33±o-73 
8.  7S±i.5o 

The  records  of  Stockberger's  hop  yields  have  been  computed  in  the 
same  manner.  The  coefficient  of  variability  of  the  plot  yields  in  his 
arrangement  is  16.71  ±3.86.  Moving  the  control  plots  down  one — that 
is,  using  plot  A  as  c^  and  so  on — produces  a  coefficient  of  variability  of 
1 5.93  ±3.55.     There  is  obviously  no  difference  between  these  two  values. 

The  problem  was  also  investigated  by  the  use  of  the  8-tree  plot  lemon 
records.  In  this  case  three  possibilities  were  tried,  with  the  first,  second, 
and  third  plots  in  turn  as  c^  and  every  subsequent  third  plot  as  a  control 
plot.  The  coefficients  of  variability  for  the  different  arrangements  were 
21. 7±2. 21,  22.5±2.29,  and  23.8i2.82.  In  view  of  the  probable  errors 
of  the  coefficients,  there  seems  to  be  no  real  difference  in  the  result  of  the 
different  arrangements  in  this  grove. 

Since  a  decided  difference  was  found  in  one  case  out  of  the  three  studied, 
it  would  seem  that  there  is  a  rather  high  probability  that  significant 
differences  may  result  from  different  arrangements  of  control  plots. 

VARIABIIvlTY  IN  THIS  YIELDS  OF  MORE  THAN  ONE  YEAR 

It  is  often  assumed  that  the  mean  vield  of  two  or  more  years  is  less 
variable  than  the  yield  of  one  year. 

We  have  had  opportunity  to  study  the  variability  in  the  yield  of  60 
navel-orange  trees  over  a  period  of  several  years.  The  data  were  kindly 
furnished  us  by  Mr.  A.  D.  Shamel,  of  the  Bureau  of  Plant  Industry. 
The  trees  in  question  had  been  selected  for  their  uniformly  good  produc- 
tion and  the  individual  yields  recorded  for  six  years. 

The  figures  presented  in  Table  XV  show  that  the  variability  of  yields 
fell  off  distinctly  after  one  year,  but  the  reduction  was  negligible  after 
the  yields  of  two  years  were  combined.  It  will  be  noted  that  the  coeffi- 
cient of  variability  of  the  single  trees  even  for  one  year  is  notably  low. 
Considering  the  records  of  single  trees,  the  average  of  six  years'  records  is 
not  less  variable  than  the  average  of  two. 


278 


Journal  of  Agricultural  Research 


Vol.  Xn,  No.  s 


Table  XV. — Comparative  variability  of  yields  of  navel-orange  trees  through  a  period  of 

six  years 

Coefficient  of  variability  of  the  yield  of — 

Class. 

I  year. 

2  years. 

3  years. 

4  years. 

5  years. 

6  years. 

Individual  trees 

Plots  of  10  trees  each 

22.  8±I.  5 

3.9±o.8 

15.  o±o.  9 
2.  9±o.  6 

i4-7±0-9 
2.  7  ±0.5 

13.  3  ±0.8 
2.  6±o.  5 

13.  7  ±0.9 
3-4±o.  7 

14.  6±i.  0 
3.9±o-8 

The  coefficient  of  variability  for  the  lo-tree  plot  is  remarkably  low 
for  the  reason  that  the  trees  themselves  are  so  uniform  and  only  a  small 
area  of  ground  is  involved.  It  is  no  surprise,  therefore,  to  find  that  the 
coefficients  of  variabiUty  are  nearly  equal,  calculated  from  one  to  six 
years.  The  probable  errors  are  relatively  large  and  it  is  difficult  to 
assert  that  there  is  any  real  difference. 

The  walnut  yields  can  be  used  as  additional  data  for  the  study  of  this 
question.  Table  XVI  shows  the  coefficient  of  variabihty  of  the  191 5 
and  1 91 6  yields  and  their  total,  with  both  i-  and  8-tree  units.  By  con- 
sidering the  individual  tree  as  a  unit,  the  total  yield  for  the  two  years 
was  less  variable  than  the  191 6  yield,  the  difference  equaling  three  times 
the  probable  error.  The  variability  of  the  191 5  yield  is  practically 
equal  to  that  of  the  total.  On  considering  an  8-tree  plot  as  a  unit,  there 
is  a  difference  between  the  coefficients  for  191 5,  191 6,  and  the  total  of 
the  two  years  respectively ;  however,  the  observed  difference  is  less  than 
three  times  the  probable  error  and  its  significance  may  be  somewhat 
doubted.  Apparently  the  mean  of  two  years'  yields  in  this  case  is  less 
variable  than  one  year's  yield. 

Table  XVI. — Comparative  variability  of  yields  of  seedling  walnut  trees   through  a 

period  of  two  years 


Class. 


Individual  trees 

Plots  of  eight  trees  each. 


Coefficient  of  variability  of  the  yield. 


191S  yield. 


47-  9±i-6 
30.  0±2.  6 


1916  yield. 


53-9±i-9 


Mean  of 

191S  and  1916 

yields. 


46.  4±i.  6 
25.  3±2.  2 


Further  studies  on  the  comparison  of  the  variability  of  yields  through 
several  years  were  made,  from  data  published  by  Hedrick  (igii).  Table 
IV  of  the  bulletin  cited  gives  the  yearly  yields  of  individual  apple  trees 
from  1902  to  1 910,  inclusive,  upon  which  our  computations  are  based. 

Four  differential  fertilizer  treatments  have  been  given  to  eight  plots 
of  five  trees  each,  each  treatment  being  duplicated  on  nonadjacent  plots. 
There  are  three  nontreated  plots  serving  as  controls. 

The  variability  of  the  individual  trees  was  computed  on  the  1 5  control 
trees. 


Feb.4.i9i8     Variability  of  Vields  of  Fruit  Trees  and  Field  Trials     279 

The  variability  of  plots  is  based  on  hypothetical  plots  made  up  of  one 
tree  from  each  of  the  treated  and  one  from  each  of  the  untreated  plots. 
These  hypothetical  plots  are  therefore  made  up  of  an  equal  number  of 
trees  having  similar  treatments.  The  variability  of  the  10  plots  thus 
obtained  was  computed  for  the  single  year  1910,  and  for  the  sum  of  two, 
three,  four,  and  seven  years.  Table  XVII  gives  the  coefficients  of  varia- 
bility for  single  trees  and  for  5-tree  plots. 

Table  XVII. — Comparative  variability  of  yields  of  Baldwin  apple  trees  through  a  period 

of  seven  years  « 


Coefficient  of  variability  of  the  yield. 

Class. 

I  year  (1910). 

2  years  (1909 
and  1910). 

3  years  (1908- 
1910). 

4  years  (1906, 
1908-1910). 

7  years  (1902, 

1903.  190S.  1906, 

1908-1910). 

Individual  trees .  . 

Plots  of  five  trees 

each 

37-3±5-2 
18.  5±2.9 

33.  6 ±4.  6 
16.  8±2.  6 

32- 5  ±4- 4 

18.  I  ±2.  8 

32.  2  ±4.  4 
18.  2  ±2.  8 

34-  o±4-  7 
2i.7±3-4 

"Records  taken  from  Hedrick  (rpir,  p.  172-174). 

It  is  interesting  to  note  how  slightly  the  variability  of  the  yields  is 
decreased  by  combining  two  or  more  years.  If  regard  is  paid  to  the 
probable  errors,  it  can  not  be  said  that  there  is  any  real  difference.  In 
other  words,  one  year's  records  of  the  yields  of  these  apple  trees  seem  to 
be  as  reliable  for  variation  studies  as  those  for  several  years. 

It  seems,  therefore,  that  the  continuation  through  several  seasons 
may  not  so  materially  decrease  the  variability  of  tree  5aelds  as  one  might 
expect.  This  has  a  direct  bearing  on  the  reliability  of  the  major  portion 
of  the  calculations  of  this  paper  which  are  based  on  the  variability  of  the 
yields  of  one  season.  It  should  be  kept  in  mind,  however,  that  these 
studies  do  not  concern  the  relative  yield  of  one  plot  compared  with  another, 
but  rather  deal  with  the  total  variation  from  the  mean  of  the  yields  of 
all  the  plots.  As  cited  before  from  the  work  of  several  experimenters, 
the  relative  productivity  of  a  group  of  plots  may  not  be  fully  determined 
even  after  a  period  of  years,  whereas  the  tree  yields  from  Hedrick,  Shamel, 
and  the  data  of  the  writers  indicate  that  a  measure  of  the  variability  for 
one  year  of  a  group  of  trees  divided  into  plots,  may  be  very  representative 
of  the  mean  variability  for  several  years. 

SUMMARY 

(i)  The  present  paper  is  the  result  of  a  study  of  the  nature  and  extent 
of  the  casual  variability  of  yields  of  fruit  trees  under  field  conditions  and 
its  bearing  on  the  reliability  of  plot  trials. 

(2)  Studies  have  been  made  upon  the  variability  of  the  yields  of  orange, 
lemon,  apple,  and  walnut  trees.     The  orchards  studied  were  selected  on 

27808°— 18 4 


28o  Journal  of  Agricultural  Research  voi.  xii,  no.  s 

account  of  uniformity  of  treatment  and  appearance,  yet  the  variability 
in  productivity  was  considerable.  The  coefficient  of  variability  for  the 
yield  of  individual  trees  of  the  clonal  varieties  ranged  from  29.27 ±0.69 
to  41 .23  ±  1 .52  per  cent,  but  for  the  individual  seedling  walnuts,  the  coeffi- 
cient was  somewhat  higher,  reaching  53.91^:1.92  per  cent.  The  varia- 
bility of  these  tree  yields  approaches  the  normal  curve  of  errors.  This 
variability  may  be  assumed  to  be  the  result  of  "casual"  factors  which  are 
beyond  the  control  and  possibly  the  recognition  of  a  careful  experimenter. 

(3)  The  effect  upon  variability  of  combining  trees  into  plots  of  various 
sizes  and  shapes  has  been  investigated.  As  the  number  of  trees  per 
plot  is  increased,  the  coefficient  of  variability  decreases.  The  coeffi- 
cient of  variability  does  not  decrease,  however,  in  proportion  to  the  in- 
creased number  of  trees  per  plot.  In  most  cases  there  is  little  gained  in 
accuracy  by  increasing  the  plot  to  include  more  than  eight  adjacent  trees. 

(4)  One  of  the  great  causes  of  variability  in  yields  appears  to  be  the 
heterogeneity  of  apparently  uniform  soil.  While  a  combination  of  a 
sufficient  number  of  adjacent  trees  into  a  plot  will  overcome  largely  the 
fluctuations  of  individuals,  nevertheless  the  plots  may  not  sufficiently 
include  both  high-  and  low-yielding  areas  to  give  a  typical  average. 
Greater  reliability  may  be  secured  by  a  systematic  repetition  and  dis- 
tribution of  plots  through  the  experimental  area.  A  consistent  gain  in 
reliability  resulting  from  this  method  of  repetition  is  shown  by  the  use 
of  several  different  methods  of  computing  the  variability. 

The  coefficient  of  variabiHty  for  an  average  plot  of  16  adjacent  trees 
was  22. 58 ±1.01,  while  16  trees  in  four  scattered  ultimate  plots  each  of 
four  trees  have  a  coefficient  of  variability  of  9.29^0.40.  The  larger  the 
number  of  units  in  a  combination  plot  the  more  typical  is  the  sample  of 
the  area  obtained.  A  i6-tree  plot  can  be  expected  to  give  more  reliable 
results  if  divided  into  four  equal  plots  and  repeated  at  four  regularly 
placed  intervals  than  can  either  two  8-tree  plots,  or  6  adjacent  trees. 
The  same  principle  holds  true  for  larger  units.  A  given  number  of  unit 
plots  will  give  a  greater  accuracy  than  half  the  number  of  units  with 
twice  as  many  trees  per  unit. 

Four  repetitions  of  an  ultimate  plot  reduced  the  coefficient  of  variability 
to  a  point  considered  practical  for  cultural  operations.  Further  repeti- 
tions, though  reducing  the  coefficient  in  less  degree,  did  not  appear  to 
justify  the  additional  number  of  trees  required.  A  minimum  of  8  to  10 
trees  is  required  for  plots  involving  cultural  experiments.  In  the  case  of 
rootstock,  pruning,  or  variety  trials,  twice  as  many  plots  each  contain- 
ing half  as  many  trees  might  be  used  to  obtain  greater  accuracy. 

The  fact  that  marked  soil  variations  occur  which  tend  to  make  adjacent 
trees  or  adjacent  plots  yield  alike,  even  on  soils  which  were  chosen 
because  of  their  apparent  uniformity,  is  well  shown  by  applying  the 
formula  proposed  by  Harris  {191 5)  for  measuring  the  coefficient  of  correla- 
tion between  neighboring  plots  of  the  field.    Applying  this  to  the  Arlington 


Feb.4,i9i8     Variability  of  Yields  of  Fruit  Trees  and  Field  Trials     281 

navel  oranges,  the  writers  have  calculated  the  correlation  between  the 
yield  of  the  8-tree  plot  as  the  ultimate  unit  and  the  yield  of  the  com- 
bination of  four  such  adjacent  plots,  and  it  was  found  that 

r=  +0.533  ±0.085. 

This  result  shows  a  marked  correlation,  indicating  a  pronounced 
heterogeneity  in  the  soil  of  this  grove  influencing  fruit  production. 

However,  when  the  correlation  between  the  8-tree  plot  as  the  ultimate 
unit  and  the  yield  of  the  combination  of  four  such  systematically  scattered 
plots  was  calculated  it  was  found  that 

r=  +o.i37±o.i20. 

This  coefficient  is  practically  equal  to  its  probable  error  and  can  be 
regarded  as  significantly  zero. 

(5)  In  the  computations  made  by  the  writers  emphasis  is  also  laid 
upon  the  nature  and  magnitude  of  the  probable  error.  It  is  shown  in 
several  cases  that  the  probable  error  of  comparison  between  plots  may  be 
so  large  that  relatively  large  differences  must  be  evident  between  treated 
and  untreated  plots  for  a  reasonable  assurance  that  it  is  due  to  the 
factors  being  experimented  upon.  With  the  plots  of  16  to  32  adjacent 
trees  which  were  studied,  a  difference  of  from  62.94  to  81.97  per  cent  of 
the  mean  production  would  be  necessary  in  order  to  obtain  chances  of  10 
to  I  that  the  results  were  due  to  differential  treatment  and  not  to  casual 
variation  in  the  productivity  of  the  trees.  With  the  same  number  of  trees 
in  scattered  units,  a  difference  of  28.42  to  50.02  per  cent  would  be  necessary 
for  the  same  odds.  It  seems  probable,  therefore,  that  a  difference  between 
two  tree  plots  of  less  than  50  per  cent  of  the  mean  production  should 
be  considered  with  caution  before  attributing  it  to  differential  treatment. 

(6)  The  relation  between  the  shape  of  a  plot  and  its  variability  was  in- 
vestigated by  making  comparisons  betvveen  square  plots  and  linear  plots 
containing  the  same  number  of  trees.  Except  in  the  case  of  large  plots,  the 
difference  in  the  variability  of  plots  of  different  shapes  was  insignificant. 

(7)  In  any  method  of  field  experimentation  where  a  standard  of  com- 
parison is  desired  the  theoretical  or  "normal"  yield  of  a  plot  is  a  question 
of  importance.  By  the  use  of  certain  formulas  the  "normal"  yield  may 
be  computed  from  control  plots.  As  a  standard,  one  may  use  the  average 
yields  of  the  control  plots  of  the  entire  area,  or  of  the  nearest  control 
plots,  or  a  combination  of  the  two.  In  cases  studied,  the  coefficient  of 
variability  was  reduced  50  per  cent  by  calculating  the  normal  yield  from 
the  nearest  controls  in  place  of  using  the  mean  of  the  entire  area.  The 
employment  of  every  alternate  row  as  a  control  plot  was  not  sufficient 
to  offset  the  variability  due  to  soil  heterogeneity. 

(8)  Computations  made  on  the  yields  of  orange,  walnut,  and  apple 
trees  for  several  consecutive  years  showed  little  annual  fluctuation  in 
their  variability.  One  or  two  crops  may  not  show  greater  variability 
than  the  average  of  six  or  seven  crops. 


282  Journal  of  Agricultural  Research  voi.  xii.  No.  s 

LITERATURE  CITED 
Ballantyne,  a.  B. 

1916.  FRUIT  TREE  ROOT  SYSTEMS.     Utah  AgT.  Exp.  Sta.  Bui.  143,  15  p.,  5  fig. 
BarrE,  H.  W. 

1915.  APPLE  breeding.     In  S.  C.  Agr.  Exp.  Sta.  28th  Ann.  Rpt.  [i9i4]/i5,  p.  13-17. 
Carleton,  M.  a. 

1909.  LIMITATIONS   IN  FIELD   EXPERIMENTS.     In   Proc.   30th   Ann.    Meeting   Soc. 

Prom.  Agr.  Sci.,  p.  55-61. 
CoiT,  J.  E. 

1910.  THE  RELATION  OF  ASEXUAL  OR  BUD-MUTATION  TO  THE  DECADENCE  OF    CALI- 

FORNIA CITRUS   ORCHARDS.     In   Proc.   37th   Fruit  Growers'  Conv.    Cal., 

P-  32-39- 
Coombs,  G.  E.,  and  Grantham,  J. 

1916.  FIELD  experiments  AND  THE  INTERPRETATION  OP  THEIR  RESULTS.       In  Agr. 

Bul.  Fed.  Malay  States,  v.  4,  no.  7,  p.  206-216,  i  fig. 
Fletcher,  S.  W. 
1913.  eliminating  unprofitable  trees  from  the  apple  orchard.     in  proc.  33d 

Bienn.  Session  Amer.  Pomol.  Soc,  p.  138-145. 
GOURLEY,  J.  H. 

1915.  STUDIES  IN  FRUIT  BUD  FORMATION.     In  N.  H.  Agr.  Exp.  Sta.  Tech.  Bul.  9, 
79  p.,  5  fig.,  4  col.  pi.,  13  charts.     Bibliography,  p.  77-79. 
Hall,  A.  D. 

1909.  THE  EXPERIMENTAL  ERROR  IN  FIELD  TRIALS.     In  JouT.  Bd.  Agr.  [London], 
V.  16,  no.  5,  p.  365-370- 
and  Russell,  E.  J. 

1911.  FIELD  TRIALS  AND  THEIR  INTERPRETATION.     In  Jour.  Bd.  Agr.  [London], 

Sup.  7,  p.  5-14,  2  fig. 
Harris,  J.  A. 

1912.  ON  THE  SIGNIFICANCE  OP  VARIETY  TESTS.     In  Science,  n.  s.,  v.  36,  no.  923, 

p-  318-320. 

1913a.   AN    ILLUSTRATION    OF     THE     INFLUENCE    OF    SUBSTRATUM    HETEROGENEITY 

UPON  EXPERIMENTAL  RESULTS.     In  Science,  n.  s.,  v.  38,  no.  975,  p.  345-346, 
I  fig. 


1913b.   SUPPLEMENTARY     NOTE    ON     THE     SIGNIFICANCE    OP    VARIETY     TESTS.       In 

Science,  n.  s.,  v.  37,  no.  952,  p.  493-494. 


I915.   ON    A    CRITERION    OP    SUBSTRATUM    HOMOGENEITY    (OR    HETEROGENEITY)    IN 

FIELD  EXPERIMENTS.     In  Amer.  Nat.,  v.  49,  no.  583,  p.  430-454. 
Hedrick,  U.  p. 

1911.  IS  IT  NECESSARY  TO  FERTILIZE   AN   APPLE  ORCHARD?       N.  Y.  State   Agr.  Exp. 

Sta.  Bul.  339,  p.  153-195.  7  fig-.  4  pl- 

1912.  pedigreed   NURSERY  STOCK.      N.  Y.  State  Agr.  Exp.   Sta.   Circ.    18,  8  p. 


1914.  A  COMPARISON  OF  TILLAGE  AND  SOD  MULCH  IN  AN  APPLE  ORCHARD.       In  N.  Y. 

State  Agr.  Exp.  Sta.  Bul.  383,  p.  249-281,  i  fig.,  6  pl. 
HoLTSMARK,  G.,  and  Larsen,  B.  R. 

1906.   tJBER  DIE  FEHLER,  WELCHE  BEI  FELDVERSUCHEN  DURCH  DIE    UNGLEICHARTIG- 
KEiT  DES   BODENS   BEDINGT  wERDEN.     In  Landw.   Vers.   Stat.,   Bd.   65, 
Heft  1/2,  p.  1-22. 
Kraus,  E.  J. 

1916.  SOMATIC  SEGREGATION.     In  Jour.  Hered.,  v.  7,  no.  i,  p.  3-8,  i  fig. 
Lewis,  C.  I  ,  and  Vickers,  H.  A. 

1915.  ECONOMICS  OP  APPLE  ORCHARDING.     In  Ore.  Agr.  Exp.  Sta.  Bul.  132,  104  p., 

14  fig. 


Feb.  4, 1918     Variability  of  Yields  of  Fruit  Trees  and  Field  Trials      283 

YON,  T.   L. 

1912.  SOME   EXPERIMENTS  TO   ESTIMATE   ERRORS  IN  FIELD  PLOT  TESTS.       In  ProC 

Amer.  Soc.  Agron.,  v.  3,  1911,  p.  89-114,  5  fig. 
Macoun,  W.  T. 

1904.  INDIVIDUALITY  OF  FRUITS.     In  Exp.  Farms  [Canada]  Rpt.,  1903,  p.  102-104. 
Mercer,  W.  B.,  and  Hall,  A.  D. 

1911.  THE  EXPERIMENTAL  ERROR  OF  FIELD  TRIALS.      Ill  Jour.  Agr.  Sci.,  V.  4,  pt.  2, 
p.  107-132,  10  fig. 
Mitchell,  H.  H.,  and  Grindley,  H.  S. 

1913.  THE   ELEMENT  OF  UNCERTAINTY  IN  THE  INTERPRETATION  OF  FEEDING  EXPERI- 

MENTS.    In  111.  Agr.  Exp.  Sta.  Bui.  165,  p.  459-579,  8  fig.  Bibliography, 

P-  578-579- 
Montgomery,  E.  G. 

i912.  variation  in  yield  and  method  of  arranging  plats  to  secure  com- 
PARATIVE RESLTLTs.  In  Nebr.  Agr.  Exp.  Sta.  25tli  Anai.  Rpt.,  [i9ii]/i2, 
p.  164-180,  4  fig. 


I913.    EXPERIMENTS  IN  WHEAT  BREEDING:    EXPERIMENTAL  ERROR  IN  THE   NURSERY 

AND  VARIATION  IN  NITROGEN  AND  YIELD.     U.  S.  Dept.  Agr.  Bur.  Plant 
Indus.  Bui.  269,  61  p.,  22  fig.,  4  pi. 
Munson,  W.  M. 

1907.  ORCHARD  RENOVATION.     In  Maine  Agr.  Exp.  Sta.  Bui.  139,  p.  53-56. 
Olmstead,  L.  B. 
i914.  some  applications  of  the  method  of  least  squares  to  agricultural 
EXPERIMENTS.     In  ]ouT.  Amer.  Soc.  Agron.,  v.  6,  no.  4/5,  p.  190-204,  i  fig- 
Pickering,  S.  U. 

1911.  EXPERIMENTAL  ERROR  IN  HORTICULTURAL  WORK.  In  Jour.  Bd.  Agr.  [lyondon]. 

Sup.  7,  p.  38-47. 
Pritchard,  F.  J. 

I916.   THE     USE    OF    checks    AND     REPEATED    PLANTINGS    IN    VARIETAL    TESTS.      In 

Jour.  Amer.  Soc.  Agron.,  v.  8,  no.  2,  p.  65-81,  3  fig. 
Shamel,  a.  D. 

1912.  BUD  SELECTION  as  A  MEANS    OF  IMPROVING  CITRUS   AND    OTHER  FRUITS.       In 

Ann.  Rpt.  Amer.  Breeders'  Assoc,  v.  8,  1911,  p.  497-503. 
Stewart,  J.  P. 

1913.  FERTILIZATION     AND   CULTURE    METHODS    IN    APPLE    ORCH.\RDS.      In    Pa.    Agr. 

Exp.  Sta.  Ann.  Rpt.  1911/12,  p.  497-563. 
Stockberger,  W.  W. 
1916.  relative  precision  op  formula  for  calculating  normal  plot  yields. 
In  Jour.  Amer.  Soc.  Agron.,  v.  8,  no.  3,  p.  167-175. 
Surface,  F.  M.,  and  Pearl,  Raymond. 

1916.   A  METHOD  OF  CORRECTING  FOR  SOIL  HETEROGENEITY  IN  VARIETY  TESTS.       In 

Jour.  Agr.  Research,  v.  5,  no.  22,  p.  1039-1050,  4  fig.  Literature  cited, 
p.  1050. 
Whitten,  J.  C. 

1915.  BUD  SELECTION  FOR  INCREASING  YIELDS.      In  Mo.  Agr.  Exp.  Sta.  Bul.  131, 

p.  479-480. 
Wood,  T.  B. 

191 1.  THE  INTERPRETATION  OF  EXPERIMENTAL  RESULTS.     In  Jour.  Bd.  Agr.  [Lon- 
don], Sup.  7,  p.  15-37,  2  fig. 

and  Stratton,  F.  J.  M. 

1910.  THE  interpretation  OF  EXPERIMENTAL  RESULTS.      In  JoUT.  Agr.  Sci.,  V.  3, 
pt.  4,  p.  417-440,  10  fig. 


INTERRELATIONS  OF  FRUIT-FLY  PARASITES  IN  HAWAII 

By  C.  E.  Pemberton,  Assistant  Entomologist,  and  H.  F.  Willard,  Fruit-Fly  Quar- 
antine Inspector,  Mediterranean  Fruit-Fly  Investigations,  Bureau  of  Entomology, 
United  States  Department  of  Agriculture 

INTRODUCTION 

The  introduction  and  ultimate  establishment  of  four  species  of  larval 
parasites  of  the  Mediterranean  fruit  fly  {Ceratitis  capitata  Wied.)  in 
Hawaii,  wherein  exist  ideal  conditions  for  the  rapid  and  unchecked 
development  of  the  host  throughout  the  year,  has  opened  an  exceptional 
opportunity  for  an  investigation  of  all  phases  of  the  work  of  these  para- 
sites, not  only  in  their  relations  to  the  host  but  also  to  one  another,  and 
has  made  possible  the  determination  of  many  points  of  unusual  interest  to 
biological  students  of  insect  parasitism  and  of  particular  value  to  ento- 
mologists dealing  with  considerations  relating  to  general  parasitic  control 
of  insect  pests. ^ 

The  information  herein  presented  is  given,  not  as  a  final  decision  or 
positive  argument  against  the  introduction  of  many  parasites  of  a  single 
pest,  but  to  reveal  the  actual  need  for  careful  biological  studies  of  para- 
sites, especially  in  their  interactions  upon  one  another,  before  general 
ntroductions  or  liberations  can  be  intelligently  undertaken. 

Entomologists,  detailed  in  the  past  for  researches  in  insect  parasitism 
in  foreign  countries,  have  customarily  adopted  the  policy  of  assembling 
all  available  species  of  primary  parasites  of  the  insects  under  investiga- 
tion, with  the  final  intention  of  conveying  them  all  to  the  home  country 
for  propagation  and  liberation.  The  chief  caution  of  these  workers  has 
usually  been  the  elimination  of  all  secondary  parasites  from  the  material 
prepared  for  shipment.  Admirable  results  have  often  been  achieved. 
However,  after  the  introduction  of  several  species,  through  one  or  suc- 
cessive importations,  few,  if  any,  considerations  have  been  given  to  the 
possibility  of  detrimental  results  arising  from  interference  with  the  action 
of  one  parasite  of  primary  importance  and  great  prolificness  by  another 
of  less  value  and  proved  inferiority. 

The  important  bearing  that  a  preliminary  and  detailed  knowledge  of 
parasite  habits  may  have  upon  the  general  question  of  parasite  importa- 
tions has  already  been  well  directed  to  the  attention  of  entomologists  by 
Dr.  L.  O.  Howard.     He  states  ^  that— 

It  is  unwise  and  most  unpromising  to  attempt  heterogeneous  and  miscellaneous 
importations  of  parasites  without  careful  study  of  the  host  insect  on  its  home  ground 

1  For  a  history  of  these  parasitic  introductions  and  a  discussion  of  climatic  and  host  relationships  favor- 
jng  parasitic  increase,  see  Back,  E.  A.,  and  Pemberton,  C.  E.,  the  mediterranean  fruit  fly  in 
HAWAII.     U.  S.  Dept.  Agr.  Bull.  536,  119  p.,  21  pi.,  24  fig.     1918. 

s  Howard,  L.  O.  the  practical  use  op  the  insect  enemies  op  injurious  insects.  In  U.  S.  Dept. 
Agr.  Yearbook  1916,  p.  282.    1917. 

Journal  of  Agricultural  Research,  Vol.  XII,  No.  5 

Washington,  D.  C.  Feb.  4,  1918 

jr  Key  No.  K— 60 

(28s) 


286  Journal  of  Agricultural  Research  voi.  xii,  No.  s 

and  in  its  natural  environment  throughout  the  whole  range  of  its  existence  and  a 
similar  biological  study  of  its  parasites  and  natural  enemies  under  such  conditions. 

Some  results  of  the  recent  work  in  Hawaii  most  strongly  bear  this  out. 
Sufficient  evidence  has  been  obtained  to  throw  serious  doubt  upon  the 
assumption,  often  accepted,  that  the  greater  the  number  of  species  of 
parasites  associated  with  the  host,  the  greater  the  chances  for  its 
control.  It  is  felt  that  the  following  data  quite  definitely  indicate,  at 
least  in  some  cases,  that  better  results  may  be  obtained  by  a  method  of 
judicious  selection  of  desirable  species  for  introduction  rather  than  by  a 
wholesale  and  indiscriminate  procedure.  If  90  per  cent  of  all  individuals 
of  an  insect  pest  are  destroyed  by  a  single  species  of  parasite,  is  it  wise 
to  attempt  further  control  by  bringing  in  other  species,  until  it  is  known 
by  positive  and  careful  experimentation  in  the  laboratory  and  field  that 
these  new  species  will  not  interfere  with  or  check  the  normal  activities 
of  the  first  species  ? 

In  May,  191 3,  Prof.  F.  Silvestri  succeeded  in  bringing  two  species  of 
opiine  parasites  of  the  fruit  fly  into  Hawaii.  One,  Opius  humilis  Sil- 
vestri, he  brought  from  South  Africa,  and  the  other,  Diachasma  tryoni 
Cameron,  was  secured  in  Australia.  Both  were  soon  established  in  the 
Kona  coffee  district  of  the  island  of  Hawaii.  By  191 5  it  had  become 
clearly  evident  that  O.  humilis  was  often  parasitizing  from  60  to  90  per 
cent  of  all  of  the  fruit-fly  larvae  developing  in  the  coffee  cherries.  D. 
tryoni  steadily  but  slowly  increased  and  in  time  exhibited  a  capacity  for 
occasionally  parasitizing  50  per  cent  or  more  of  the  host  larvae.  Here 
it  is  obvious  that  overlapping  or  duplication  in  parasitism  was  occurring. 
Clearly  some  fly  larvae  were  being  stung  by  both  species  of  parasites  and 
frequently  to  a  very  considerable  extent. 

CANNIBALISM  AMONG  THE  PARASITES 

Early  in  1916,  Dr.  E.  A.  Back,  of  the  Bureau  of  Entomology,  while 
examining  the  contents  of  some  parasitized  fruit-fly  material  from  the 
field,  observed  under  the  microscope  a  larva  of  the  parasite  Diachasma 
tryoni  attacking  one  of  the  parasite  Opius  humilis.  A  suspicion  of  dis- 
advantageous consequences  arising  from  complications  attending  the 
interactions  of  these  parasites  led  Mr.  C.  L.  Marlatt  to  assign  to  the 
writers  an  investigation  of  this  subject. 

Careful  microscopical  examinations  of  large  numbers  of  fruit-fly  larvae 
and  pupae,  collected  from  localities  where  both  species  of  parasites  were 
known  to  be  well  established  and  actively  working  together,  soon  revealed 
one  striking  fact.  In  the  majority  of  cases  where  fruit-fly  larvae  had  been 
parasitized  by  both  Diachasma  tryoni  and  Opius  humilis  the  latter  was 
killed  and  the  former  developed  to  maturity.  O.  humilis  is  killed  purely 
by  wounds  and  lacerations  inflicted  upon  it  by  the  long,  curved,  sickle- 
like mandibles  of  the  newly  hatched  larva  of  D.   tryoni.     This  larva 


Feb.  4, 1918     Interrelations  of  Fruit-Fly  Parasites  in  Hawaii  287 

attacks  the  young  of  O.  humUis,  usually  at  a  point  a  few  segments  back 
of  the  head,  the  point  of  contact  being  clearly  seen  in  Plate  12. 
Its  mandibles  open  wide  and  snap  into  the  body  of  the  attacked  larva 
spasmodically  and  with  remarkable  quickness.  Often  the  entire  opera- 
tion of  broadly  opening  and  closing  the  mandibles  may  be  almost  imper- 
ceptible to  the  eye.  It  may  move  its  entire  body  quickly,  the  caudal 
tip  may  be  curled  beneath  the  body  and  extended  again  suddenly,  and 
the  mandibles  may  be  repeatedly  opened  and  closed  until  successfully 
grasping  the  Opius  larva.  Besides  possessing  unusual  powers  for  inflict- 
ing injury  to  other  parasitic  larvae  about  it,  it  may  avoid  counterattack 
through  ability  to  move  quickly  and  through  the  protection  afforded 
the  entire  ventral  surface  of  the  body  by  a  thick  mass  of  serosal,  cellular 
material  that  accompanies  the  larva  when  it  emerges  from  the  egg,  and 
which  remains  with  it  during  its  entire  life  in  the  first  instar.  This  mass 
of  cells  may  be  seen  clinging  to  Diachasma  larvae  of  the  first  instar  in 
Plate  10. 

The  newly  hatched  larva  of  Opius  humilis  possesses  mandibles  which 
are  also  long  and  pointed,  as  shown  in  Plate  12,  A,  B.  These  may 
be  used  to  good  advantage  when  the  larva  is  successful  in  bringing  them 
in  contact  with  individuals  of  its  own  or  of  other  species  of  parasites. 
The  larva,  however,  is  sluggish,  moves  much  less  quickly  than  that  of 
Diachasma  tryoni,  is  protected  ventrally  by  a  much  thinner,  less  adhesive 
mass  of  serosal  cells,  is  much  less  capable  of  quick  and  powerful  move- 
ment of  the  mandibles,  and  usually  holds  the  body  in  a  somewhat  hori- 
zontal and  exposed  position.  These  deficiencies  seem  to  explain  its 
inability  to  avoid  destruction  by  larvs  of  D.  tryoni  or  to  offer  successful 
counterattack  when  the  two  are  lodged  within  the  same  host  larva. 

During  the  examination  of  nearly  3,000  fruit-fly  larvae  or  pupae,  para- 
sitized in  each  case  by  the  two  species  of  opiines,  the  dead  or  dying  and 
often  struggling  Opius  larvae  were  frequently  dissected  from  the  tissues 
of  the  host  while  still  tightly  clasped  in  the  mandibles  of  the  Diachasma 
larvae.  Plate  10,  A,  is  reproduced  from  a  photomicrograph  of  a  larva 
of  O.  humilis  actually  within  the  grasp  of  an  active,  living  larva  of 
D.  tryoni.  In  this  particular  case  the  operation  of  removing  the  two 
larvae  from  the  host,  placing  them  upon  a  microscope  slide  in  water, 
and  covering  them  with  a  cover  glass  failed  to  separate  them,  and  the 
Diachasma  larva  ultimately  died  with  its  mandibles  deeply  embedded 
in  the  body  of  the  dead  Opius  larva  in  the  exact  position  as  shown. 
Plates  12  and  13,  A,  B,  C,  show  larvae  of  O.  humilis  in  various  stages  of 
laceration  and  distortion  just  as  they  were  removed  from  fruit-fly  larvae 
or  pupae  in  which  were  also  one  or  more  larvae  of  D.  tryoni. 

Extensive  laboratory  experiments  have  exactly  confirmed  the  results 
of  the  first  series  of  field  observations.  The  aggressive,  cannibalistic 
period  of  activity  of  the  larvae  of  these  parasites  is  during  existence  in 


Journal  of  Agricultural  Research  voi.  xii,  no.  s 


the  first  instar,  and  particularly  during  the  early  period  of  this  stage, 
before  the  body  becomes  engorged  and  swollen  with  food.  Plate  ii,  B, 
which  shows  a  mature  first-instar  larva  of  the  parasite  Diachasma  tryoni, 
is  interesting  in  this  connection  when  compared  with  the  other  illustra- 
tions of  newly  hatched  Diachasma  larvse  of  this  instar.  The  head  is 
unchanged  in  size,  but  the  body  is  greatly  distended  after  two  days  of 
feeding  preparatory  to  molting  to  the  second  instar.  In  all  of  the  illus- 
trations the  enlargement  is  the  same.  It  is  obvious  that  the  enlarged 
and  somewhat  rigid  body  of  the  well-fed  larva  permits  much  less  free- 
dom of  movement  than  is  possible  shortly  after  hatching.  The  body  is 
first  elastic,  flexible,  and  capable  of  quick  and  effective  action.  This 
change  in  the  size  of  the  body  may  occur  in  from  two  to  two  and  one-half 
days. 

The  molt  to  the  second  instar  still  further  incapacitates  the  opiine 
larva  for  carrying  on  further  cannibalistic  action.  The  mandibles  are 
then  small,  soft,  and  are  almost  imperceptible,  even  under  high  magnifi- 
cation, because  of  their  transparency,  being  wholly  unfitted  for  active 
use  except  in  the  separation  of  the  semiliquid  media  in  which  the  larva 
lies  after  the  host  has  formed  the  puparium.  The  strong  and  heavily 
chitinized  head  of  the  opiine  larva  in  the  first  instar  is  thus  entirely 
discarded  upon  the  molt  to  the  second  instar,  and  all  further  cannibalism 
ceases.  The  helpless  condition  after  the  first  molt  is  well  suggested  in 
Plate  1 1 ,  A,  which  shows  a  larva  of  Diachasma  tryoni  in  the  second  instar. 
During  191 6  a  microscopical  examination  of  the  contents  of  a  total 
of  2,925  parasitized  fruit-fly  larvse  and  pupae  definitely  showed  that  the 
parasites  readily  oviposit  in  the  same  host  larva  more  than  once  and 
exhibit  no  discernible  instinct  of  selection  of  parasitized  or  unparasitized 
larvse.  All  eggs  so  deposited  hatch.  Thus,  fly  larvse  commonly  opened 
were  found  to  contain  several  eggs  or  larvae  of  the  two  species  of  para- 
sites and  sometimes  a  third  species  (Diachasma  jullawayi) ,  a  later  intro- 
duction. 

Here  w^as  certain  evidence  of  a  complicated  overlapping  or  duplication 
of  parasitism.  In  no  single  instance  were  two  parasite  larvse  ever 
observed  to  develop  to  maturity  in  the  same  host  larva,  except  in  the 
case  of  the  chalcid  Tetrastichus  gifjardianus  (p.  292).  All  but  one  were 
killed  while  still  in  the  first  instar,  or  occasionally  before  hatching.  At 
times  from  8  to  10  opiine  larvae  were  found  within  the  host  larva.  Only 
one  would  mature  and,  as  a  rule,  if  a  larva  of  Diachasma  were  one  of  the 
number,  it  survived  all  others. 

If  several  eggs  of  Diachasmu  tryoni  or  D.  jullawayi  alone  are  deposited 
in  an  individual  host  larva  over  a  period  of  two  or  three  days,  the  last 
parasite  larva  to  hatch  stands  the  best  chance  for  destroying  the  others 
and  maturing.  The  case  is  identical  with  the  Opius  humilis  However, 
a  well-grown  and  fully-fed  larva  of  D.  tryoni  or  D.  jullawayi,  if  still  in  the 


Feb.  4.  i9i8     Interrelations  of  Fruit-Fly  Parasites  in  Hawaii  289 

first  instar,  seems  entirely  able  to  destroy  larva  of  O.  humilis  that  are 
newly  hatched  and  unencumbered  with  a  body  engorged  with  food. 

All  opiine  larvae  within  an  individual  fruit-fly  larva  or  pupa  may  some- 
times be  fatally  wounded  and  no  parasite  develop.  This  is  not  frequent. 
Often  the  deposition  of  8  or  10  parasite  eggs  into  a  single  fly  larva  causes 
its  death.  In  such  cases  the  parasite  eggs  usually  hatch  and  the  resulting 
larvae  die  within  a  short  time. 

Cases  occurred  in  which  as  many  as  10  dead  larvae  of  Opius  humilis  were 
dissected  from  a  fruit-fly  larva,  together  with  a  single,  vigorous,  active 
larva  of  Diachasma  tryoni.  This  does  not  necessarily  mean  that  all  of 
the  10  larvae  were  killed  by  the  latter.  No  doubt  some  of  the  former 
larvae  destroyed  each  other,  but  it  clearly  shows  the  superior  aggressive 
and  defensive  power  of  the  larva  of  D.  tryoni.  Many  cases  have  been 
observed  in  which  a  larva  of  O.  humilis  was  badly  cut  and  distorted  from 
attack  by  a  larva  of  D.  tryoni.  Some  such  cases  are  shown  in  Plates 
10,  12,  and  13.  Occasionally  the  body  may  be  found  entirely  severed  from 
the  head.  These  extreme  cases  are  no  doubt  caused  by  reattack  upon  the 
larva  a  day  or  more  after  it  has  died  and  has  become  somewhat  softened. 
In  most  cases  the  death  of  the  larva  seems  to  be  caused  by  the  first  grasp 
or  pinch  of  the  attacking  larva.  A  single  perforation  in  the  body  wall 
should  be  sufficient  to  cause  the  death  of  the  larva  in  a  short  time. 

Occasionally  an  opiine  larva  will  destroy  eggs  of  its  own  kind  when  it 
occurs  in  the  same  individual  host  with  the  eggs.  In  this  manner  mature 
embryos  are  sometimes  very  much  distorted  and  almost  unrecognizable. 
This  is  not  frequently  seen. 

Cool  weather  materially  retards  the  development  of  the  opiine  &gg, 
particularly  in  the  case  of  Diachasma  tryoni  and  D.  fullawayi.  At  such 
times  many  cases  have  been  observed  in  which  larvae  of  Opius  humilis 
have  developed  to  the  second  instar  before  an  egg  of  a  species  of  Dia- 
chasma, which  had  been  deposited  into  the  same  host  larva  harboring 
O.  humilis,  had  hatched.  Upon  hatching  the  small,  active  Diachasma 
larv^a  quickly  destroyed  the  large,  bulky  Opius  larva.  This  unusual  con- 
dition has  been  observed  only  in  January  in  the  cool,  elevated,  coffee 
districts  on  the  Island  of  Hawaii. 

From  over  2,900  cases  where  parasitized  fruit-fly  puparia  have  been 
opened,  in  no  single  instance  has  a  case  been  observed  in  which  the  host 
pupa  was  fofmed.  A  host  lar\'a  once  parasitized  quite  readily  forms 
into  a  normal  puparium  when  sufficiently  developed,  but  the  presence 
of  a  single  small  opiine  egg  within  its  body  invariably  prevents  any 
further  development.  The  puparium  is  formed,  the  histolysis  of  the 
tissues  is  completed,  and  here  all  development  of  the  fruit  fly  ceases. 
The  broken-down  and  liquid  medium  thus  prepared  within  the  puparium, 
in  which  the  parasite  larvae  may  move  about  and  feed,  enables  them  to 
reach  all  portions  within  and  easily  to  come  in  contact  with  any  other 
parasitic  individuals  that  may  occur  there  with  them. 


290  Journal  of  Agricultural  Research  voi.  xii.  No.  s 

Thus,  the  inevitable  tendency  of  every  individual  opiine  lan^a  upon 
hatching  is  to  destroy  every  other  parasitic  larva  in  its  domain,  whether 
it  be  one  of  its  own  kind  or  of  another  species.  This  would  appear  to  be 
an  infallible  instinct  and  one  of  great  consequence  in  Hawaii  in  the 
development  of  the  opiine  parasites  now  present. 

SUPPRESSION   OF   OPIUS  HUMILIS   BY  DIACHASMA  TRYONI   AND   D. 

FULLAWAYI    • 

The  parasites  Diachasma  tryoni  and  the  closely  related  Diachasma 
fullawayi,  by  virtue  of  this  larv^al  instinct,  coupled  with  further  en- 
dowed superior  body  characters,  have  been  responsible  for  the  great 
suppression  of  the  parasite  Opius  humilis.  The  last-named  species  is 
more  prolific  and  hardy  than  either  of  the  two  others  and  is  more  gener- 
ally efficient  than  both  combined.  By  their  association  with  the  0.  humilis, 
they  have  worked  a  detriment  by  reducing  the  total  extent  of  parasitism 
to  a  point  below  that  to  which  it  is  capable  of  exerting  alone.  The  evi- 
dence of  such  suppression,  gained  from  microscopical  examinations  of 
fruit-fly  larvae  and  pupae  secured  from  various  fruits  in  Hawaii  during 
191 6  and  191 7,  may  be  expressed  as  follows: 

From  April  16  to  May  10,  1916,  a  dissection  was  made  of  757  fruit-fly 
pupae,  freshly  secured  from  coffee  collected  in  the  Kona  coffee  district  of 
the  Island  of  Hawaii.  From  this  total,  345  were  parasitized  by  only 
Opius  humilis,  90  contained  living  larvae  of  Diachasma  tryoni  together 
with  dead  larvae  of  O.  humilis,  9  contained  living  larvae  of  O.  humilis 
together  with  dead  larvae  of  D.  tryoni,  i  was  parasitized  by  only  D. 
fullawayi,  5  contained  living  larvae  of  D.  fullawayi  together  with  dead 
larvae  of  O.  humilis,  2  contained  living  Opius  larvae  together  with  dead 
larvae  of  D.  fullawayi,  57  contained  living  larvae  of  only  D.  tryoni,  and 
248  were  not  parasitized.  Here  it  is  seen  that  from  the  total  of  757  pu- 
paria,  106  cases  of  duplication  in  parasitism  occurred  in  which  a  species 
of  Diachasma  was  found  in  the  same  puparium  with  one  or  more  indi- 
vidual larvae  of  O.  humilis,  and  that  in  95  of  these  the  latter  was  killed 
and  the  former  survived. 

A  further  series  of  microscopical  examinations  of  the  contents  of  fruit- 
fly  puparia,  freshly  secured  from  coffee  fruits  in  this  same  district  in 
January,  191 7,  during  the  coolest  part  of  the  year,  strongly  confirmed 
the  results  of  the  previous  year.  Six  hundred  and  twenty-seven  puparia 
were  opened  and  examined.  Of  these  a  total  of  343  were  parasitized  by 
Opius  humilis  alone,  67  were  parasitized  by  only  Diachasma  tryoni,  2  con- 
tained only  larvae  of  Diachasma  fullawayi,  129  contained  living  larv'ae  of 
D.  tryoni  together  with  dead  larvae  of  O.  humilis,  8  contained  dead  larvae 
of  D.  tryoni  together  with  living  larvae  of  O.  humihs,  4  contained  living  larvae 
of  D.  fullawayi  together  with  dead  larvae  of  O.  humilis,  2  contained  dead 
larvae  of  D.  fullawayi  together  with  living  larvae  of  O.  humilis,  and  72 
puparia  were  unparasitized.     These  results  are  again  very  significant. 


Feb.  4, 1918     Interrelations  of  Fruit-Fly  Parasites  in  Hawaii 


291 


From  the  total  of  627  puparia  143  cases  are  noted  in  which  an  overlapping 
in  parasitism  occurred,  wherein  the  puparia  in  each  case  contained  larvae 
of  O.  humilis  in  combination  with  larvae  of  a  species  of  Diachasma,  and 
in  133  of  these  cases  the  former  was  destroyed. 

The  collection  of  extensive  data  on  the  percentage  of  parasitism  ex- 
erted by  the  fruit-fly  parasites  in  Hawaii  over  a  period  of  three  years 
gave  abundant  proof  that  the  parasite  Diachasma  iryoni  was  most  active 
during  the  warmer  months  of  the  year.  This  increase  in  activity  and 
abundance  paralleled  a  reciprocal  decrease  in  the  abundance  of  Opius 
humilis.  The  reverse  was  true  during  the  remainder  of  the  year,  when 
the  former  species  rapidly  decreased  and  the  latter  ascended  to  a  place 
of  first  importance.  The  data  presented  in  Table  I  most  positively  reveal 
the  extent  of  fluctuation  in  the  comparative  abundance  of  O.  humilis  and 
D.  iryoni,  the  effectiveness  of  O.  humilis  being  clearly  at  its  maximum 
during  the  spring,  when  the  abundance  of  D.  iryoni  is  at  its  lowest, 
owing  to  the  accumulated  effect  of  the  cool  winter  months. 

Table  I. — Comparison  of  seasonal  abundance  of  Optus  humilis  and  Diachasma  tryoni 


Locality. 


Honolulu,  Oahu 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Kona  District,  Hawaii 

Do 

Do 

Do 

Do 

Do 


Number  of 

Number  of 

Percentage 

Date  of  collec- 

Diachasma 

Opius  hu- 

of 

tion  of  host,  a 

tryoni 

milis  emerg- 

Diachasma 

emerging. 

mg. 

tryoni. 

Mar.,   19 1 6 

30 

I  000 

2.9 

Apr.,   1916 

1,200 

9,778 

10.  9 

May,    1916 

499 

2,127 

19.  0 

June,  1916 

2,303 

998 

69.8 

July,   1916 

1,786 

549 

76.  s 

Aug.,  1916 

2,286 

649 

77-9 

Sept.,1916 

4,514 

1,139 

79-9 

Oct.,    1916 

6,772 

2,  061 

76.7 

Nov.,  1916 

4,451 

I,  206 

78.9 

Dec,  1916 

2,605 

1,602 

61.  9 

Jan.,    19 1 7 

1,406 

679 

67.4 

Feb.,  1917 

558 

I,  lOI 

33-6 

Sept.,  1915 

330 

118 

73-6 

Dec,  1915 

210 

274 

43-3 

Mar.,   1916 

85 

440 

16.  I 

Apr.,   1916 

756 

3,031 

19.9 

Aug.,  1916 

271 

77 

77-9 

Jan.,    19 1 7 

558 

4,749 

10.5 

Percentage 
of  Opius 
humilis. 


97.  I 
89.  I 
81.0 
30.2 

23-5 
22.  I 

20.  I 

23-3 

21.  I 

38.1 
32.6 
66.4 

26.  4 
56.7 

83- 9 
80.  I 


a  For  similar  data  for  1915,  see  Back,  E.  A.,  and  Pemberton,  C.  K.,  the  mediterranean  fruit 
FLY  IN  HAWAII.    U.  S.  Dept.  Agr.  Bui.  536,  119  p.,  21  pi.,  24  fig.     1918. 

From  September  to  December,  1916,  a  microscopical  examination  was 
made  of  the  contents  of  618  fruit-fly  puparia  obtained  about  Honolulu 
from  coffee,  guavas,  and  the  winged  kamani  (Terminalia  catappa).  Of 
this  total,  55  puparia  were  parasitized  by  only  Opius  humilis,  331  were 
parasitized  by  only  Diachasma  iryoni,  35  contained  only  larvae  of  D.  fulla- 
wayi,  96  contained  living  larvae  of  D.  tryoni  together  with  dead  larvae  of 
O.  humilis,  4  contained  dead  larvae  of  D.  tryoni  together  with  living  larvae 


292  Journal  of  Agricultural  Research  voi.  xii,  no.  5 

of  O.  humilis,  6  contained  living  larvae  of  D.  jullawayi  together  with  dead 
larvae  of  O.  humilis,  i  contained  a  dead  larva  of  D.  jullawayi  together  with 
a  living  larv^a  of  O.  humilis,  and  90  were  unparasitized.  Here  again  is 
striking  evidence  of  positive  suppression  of  the  parasite  O.  humilis  by  the 
other  parasites,  particularly  by  D.  tryoni.  Of  the  618  puparia  examined, 
the  O.  humilis  occurred  107  times  in  combination  with  a  species  of  Dia- 
chasma  and  won  the  struggle  for  existence  in  only  5  cases. 

The  foregoing  would  tend  to  explain  the  very  noticeable  fluctuations 
in  the  comparative  abundance  of  the  two  principal  species  of  parasites 
at  differenct  seasons  of  the  year,  as  shown  in  Table  I.  It  is  seen  that  the 
extent  of  parasitism  by  Opius  humilis  is  greatly  influenced  by  the  abun- 
dance or  scarcity  of  the  parasite  Diachasma  tryoni,  and,  as  elsewhere 
discussed,  the  abundance  of  this  latter  parasite  in  Hawaii  is  very  much 
dependent  upon  seasonal  conditions.  This  causes  a  seasonal  rise  and 
fall  in  the  effectiveness  of  0.  humilis.  In  the  summer  and  autumn  of 
the  year  the  ascendancy  of  the  D.  tryoni  causes  a  great  reduction  in  the 
abundance  of  O.  humilis.  During  the  winter  and  spring  seasons  the 
reduced  activity  of  the  Diachasma  permits  a  rapid  increase  in  parasitism 
by  O.  humilis.  This  is  particularly  true  in  the  elevated  Kona  coffee 
district,  where  the  winter  temperatures  are  somewhat  below  those  about 
Honolulu. 

The  results  of  the  first  series  of  dissections  of  fruit-fly  larvae  and  puparia 
on  the  island  of  Hawaii  in  January,  1916,  led  to  further  laboratory  ex- 
periments in  Honolulu  on  a  comprehensive  scale,  duplicating  as  closely 
as  possible  the  field  conditions.  Unparasitized  fruit-fly  lars^ae  were  ex- 
posed within  fruit  to  the  attack  of  both  Opius  humilis  and  one  or  both  of 
the  species  of  Diachasma  for  a  few  hours.  The  fruit-fly  puparia  thus 
obtained  were  usually  all  parasitized,  and  no  loss  in  time  resulted  from 
examinations  of  unparasitized  material.  In  this  manner  393  cases  were 
obtained  in  which  fruit-fly  larvae  were  parasitized  by  both  O.  humilis  and 
D.  tryoni.  In  387  of  these  the  Opius  larvae  were  all  killed  and  the  Dia- 
chasm.a  survived.  In  only  5  of  the  puparia  did  the  former  succeed  in 
overcoming  the  latter  and  developing. 

Out  of  yj  cases  in  which  fruit-fly  larvae  were  parasitized  by  both  Opius 
humilis  and  Diachasma  jullawayi,  in  only  2  did  the  Opius  develop. 
In  the  remaining  75  the  Opius  larvae  were  all  killed  by  the  Diachasma 
larvae. 

SUPPRESSION  OF  OPIUS  HUMILIS  BY  TETRASTICHUS  GIFFARDIANUS 

The  fruit-fly  parasite  Tetrastichus  gifjardianus  Silvestri,  a  late  introduc- 
tion into  Hawaii,  has  proved  decidedly  destructive  to  any  of  the  opiines 
when  occurring  in  the  same  fly  larvae  or  puparia  with  them.  This 
chalcid,  seemingly  of  importance,  has,  after  a  two  years'  trial  in  Hawaii, 
given  but  small  promise  of  accomplishing  any  perceptible  control  of  the 


Feb.  4, 1918       Interrelations  of  Fruit-Fly  Parasites  in  Hawaii  293 

fruit  fly.  It  is  occasionally  bred  out  from  fruit-fly  material  secured  in 
the  field.  Several  times  during  dissections  of  puparia  from  the  field  its 
larvae  have  been  found  in  combination  with  larvae  of  Opius  humilis  or  of 
one  of  the  species  of  Diachasma.  Though  soft,  sluggish,  and  armed  with 
small,  inconspicuous,  blunt  mandibles,  it  nevertheless  survives  the  opiine 
larvae  by  sheer  force  of  numbers  and  consequent  rapid  absorption  of 
food.  When  ovipositing,  this  chalcid  usually  places  about  10  eggs  in 
the  host  at  one  insertion  of  the  ovipositor.  The  opiines  deposit  but  one 
egg,  remove  the  ovipositor,  and  look  for  further  larvae  to  attack.  A 
single  oviposition,  then,  by  a  T.  giffardianus  into  a  host  larva  already 
parasitized  by  an  opiine,  places  about  10  individuals  of  the  chalcid  with 
I  of  the  opiine.  If  the  host  larva  has  already  received  several  opiine  ovi- 
positions,  usually  only  one  individual  is  alive,  as  already  shown.  The  larvae 
of  T.  giffardianus  exhibit  no  cannibalistic  tendencies,  and  do  not  destroy 
each  other.  Many  of  the  chalcid  larvae  are  then  killed  by  the  opiine 
larvae,  but  seldom  all  of  them.  The  opiines  ultimately  die,  and  one  or 
more  of  the  Tetrastichus  larvae  develop.  The  death  of  the  Opius  or  Dia- 
chasma larvae  results  usually  from  starvation  or  suflfocation  or  possibly 
by  the  absorption  of  toxic  excretions  of  the  Tetrastichus  larvae.  Certainly 
the  chalcid  larvae  inflict  no  visible  bodily  injury  on  the  opiine  larv^. 

In  view  of  the  demonstrated  ineffectiveness  of  this  chalcid  and  of  its 
capability  for  surviving  the  Opius  humilis  when  occurring  in  fruit-fly 
larvae  with  it,  it  is  here  considered  a  detrimental  introduction  because  of 
interference  with  the  work  of  the  latter. 

OPIUS  HUMILIS,  THE  MOST  PROLIFIC  OF  THE  INTRODUCED 

PARASITES 

From  the  foregoing  summaries  of  the  data  it  is  obvious  that  the 
parasite  Opius  humilis  is  killed  in  the  larva  stage  in  almost  every  instance 
in  which  its  larvae  become  associated  in  a  host  larva  with  any  one  of  the 
three  other  species  of  introduced  fruit-fly  parasites  and  that  the  percent- 
age of  cases  of  such  duplication  is  large.  Biological  studies  of  all  of  these 
parasites,  so  far  as  two  and  three  years'  adaptation  in  a  new  country  may 
show,  have  indicated  quite  clearly  that  under  Hawaiian  conditions 
O.  humilis  is  the  most  prolific  of  the  four  species  in  all  seasons  of  the  year 
and  that  none  of  the  others  show  particular  abilities  for  reaching  larvae  of 
the  fruit  fly  that  are  not  as  easily  accessible  to  the  attack  of  O.  humilis. 

The  very  considerable  activity  of  the  parasite  Diachasma  tryoni  and 
the  occasional  heavy  parasitism  by  it  has  made  necessary  the  establish- 
ment of  careful  proof  that  it  is  less  prolific  than  Opius  humilis,  before  it 
can  be  maintained  that  the  introduction  of  the  former  is  a  detrimental 
one.  The  unimportance  of  the  other  two  parasitic  species  eliminates  all 
present  need  for  discussing  them  any  further. 

From  what  has  already  been  shown,  we  know  positively  that  the 
parasite  Diachasma  tryoni  has  a  clear  field  in   Hawaii  for  unchecked 


294  Journal  of  Agricultural  Research  voi.  xii,  No.  s 

reproduction.  Opius  humilis,  though  present  everywhere,  does  not 
hinder  its  activities.  Extensive  records  kept  in  Hawaii  during  a  period 
of  three  years  on  emergences  from  more  than  100,000  fruit-fly  puparia 
from  many  sections  of  the  islands  and  from  all  available  types  of  fruits 
have  clearly  shown  that  the  average  maximum  degree  of  parasitism  by 
D.  tryoni  does  not  exceed  or  equal  that  of  O.  humilis.  This  alone  is  suffi- 
cient proof  of  superior  prolificness  on  the  part  of  the  latter. 

In  no  instance  have  fruits  of  all  varieties  been  found  harboring  larvae 
of  the  fruit  fly  that  are  less  frequented  by  Opius  humilis  than  by  Dia- 
chasma  tryoni.  The  ability  of  the  former  to  find  the  host  is,  then,  equal, 
if  not  superior,  to  that  of  the  other  parasite.  In  no  case  has  evidence 
been  found  to  show  that  the  longer  ovipositor  of  the  D.  tryoni  is  really 
an  advantage  over  that  of  the  O.  humilis,  whose  ovipositor  is  less  than  a 
third  as  long.  The  comparative  difference  is  considerable,  but  the  dif- 
ference when  considered  in  fractions  of  an  inch  is  really  small. 

Preliminary  records  of  individual  females  of  both  species  to  determine 
the  total  number  of  eggs  deposited  by  an  individual  in  a  lifetime  show  no 
superiority  of  the  Diachasma  tryoni  over  Opius  humilis  in  this  respect. 

The  life  cycle  of  Diachasma  tryoni  is  consistently  longer  than  that 
of  Opius  humilis.  From  28,410  records  on  the  length  of  the  life  cycle  of 
the  former,  secured  during  191 6  and  191 7,  it  is  almost  invariably  found 
to  be  from  two  to  four  days  longer  than  that  of  the  latter  during  most  of 
the  year,  and  in  the  winter  months  a  great  number  of  the  Diachasma 
individuals,  hibernating  in  the  larva  stage,  extended  the  cycle  to  from 
one  to  six  months  longer.  From  22,889  cases  under  observation  in  1916 
and  1 91 7  on  the  life  cycle  of  O.  humilis,  in  no  single  case  has  an  individual 
ever  been  known  to  so  hibernate  or  extend  the  length  of  the  life  cycle 
beyond  the  average  for  more  than  a  few  days.  Cool  weather  and  drouth 
seem  most  favorable  for  inducing  this  hibernating  tendency  in  the 
larvae  of  the  D.  tryoni.  Of  3,077  cases  under  observation  by  the  junior 
author  in  January,  February,  and  March,  191 7,  in  which  fruit-fly  larvae 
had  been  parasitized  by  a  Diachasma,  a  total  of  i  ,404  cases  occurred  in 
which  the  parasite  went  into  hibernation  as  a  mature  larva.  This  seems 
to  explain  the  great  reduction  in  abundance  of  D.  tryoni  in  the  field  in 
winter.  Its  capacity  as  a  parasite  in  the  winter  months  is  thus  strik- 
ingly less  than  that  of  O.  humilis.^ 

From  the  standpoint  of  longevity,  all  experiments  so  far  show  no  great 
superiority  of  one  species  over  the  other,  except  as  noted  above  in  regard 
to  hibernation.  Individual  adults  of  both  Opius  humilis  and  Diachasma 
tryoni  have  been  kept  alive  for  about  four  months. 

'  Credit  is  here  due  Mr.  J.  C.  Bridwell  for  valuable  suggestions  offered  in  connection  with  the  study 
of  the  hibernation  of  these  parasites. 


Feb.  4, 191S     Interrelations  of  Fruit-Fly  Parasites  in  Hawaii  295 

The  relative  proportion  of  females  to  males  in  these  two  species  of 
parasites  is  interesting.  Of  26,975  individuals  of  Diachasma  tryoni, 
reared  from  material  collected  in  the  field  in  191 6  and  191 7,  16,845,  or 
62.4  per  cent,  were  males.  Of  10,843  individuals  of  Opius  humilis 
similarly  reared  from  material  collected  in  the  field  in  191 6  and  191 7, 
6,128,  or  56.5  per  cent,  were  males.  Here,  again,  the  advantage,  if  any, 
lies  with  O.  humilis. 

In  view  of  these  several  facts  relating  to  the  comparative  prolificness 
of  the  parasites  Opius  humilis  and  Diachasma  tryoni,  it  appears  that  the 
former  is  superior  to  the  latter  species  or  to  any  of  the  other  introduced 
parasites.  Thus,  when  a  host  larva  is  parasitized  by  both  O.  humilis 
and  D.  tryoni,  the  latter  survives,  and  in  so  doing  produces  an  individual 
less  prolific  than  would  have  been  the  case  had  the  O.  humilis  been  per- 
mitted to  develop.  This  seems  to  point  to  a  certain,  distinct  loss.  If 
one  species  of  larval  parasite  when  working  alone  parasitizes  60  per  cent 
of  the  host,  and  another  species  not  strikingly  different  from  the  first  and 
working  the  same  in  every  known  respect  parasitizes  40  per  cent  of  the 
host  when  working  alone,  there  is  no  reason  to  assume  that  both  com- 
bined can  exceed  a  parasitism  of  60  per  cent.  All  overlapping  by  the 
species  capable  of  only  40  per  cent  parasitism  can  only  serve  to  reduce 
the  total  effect  to  a  point  below  60  per  cent  of  parasitism. 

CONCLUSION 

Sufficient  evidence  has  been  presented  to  prove  the  superiority  of  the 
parasite  Opius  humilis  over  the  other  introduced  fruit-fly  parasites  in 
Hawaii  and  demonstrates  the  decided  restraint  operated  over  it  by  the 
unfailing  cannibalistic  activities  of  the  larvae  of  Diachasma  tryoni  in 
particular  and  of  the  other  parasites  in  part.  Knowing  the  capacity  of 
O.  humilis  for  parasitizing  from  80  to  90  per  cent  of  the  larvae  of  the  fruit 
fly  in  favorable  localities,  such  as  the  large  Kona  coffee  belt  on  the  island 
of  Hawaii,  the  writers  here  maintain  that  detrimental  results  to  a  certain 
extent  have  arisen  from  the  liberation  in  Hawaii  of  parasites  other  than 
O.  humilis  that  attack  the  larva  of  the  fruit  fly.  The  total  parasitism 
has  simply  been  reduced  in  value  to  that  of  a  parasite  of  secondary  value. 

It  is  hoped  that  the  present  analysis  of  the  interrelated  activities  of 
the  imported  fruit-fly  parasites  in  Hawaii  may  serve  to  stimulate  greater 
discrimination  in  the  selection  of  parasites  proposed  for  future  intro- 
duction. 

27808°— 18 5 


PLATE  lo 
Diachasma  tryoni: 

A. — Freshly  hatched  larva  with  its  mandibles  actually  embedded  in  the  body  of  a 
newly  hatched  but  dead  larva  of  Opiiis  humilis.     Xioo. 

B. — ^Newly  hatched  larva  with  its  mandibles  closed,  showing  ventral  serosal  material 
surrounding  the  body  and  the  two  gill-like  appendages  on  the  first  body  segment. 

X  loo. 

(296) 


Interrelations  of  Fruit-Fly  Parasites 


Plate  10 


A 


B 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  5 


Interrelations  of  Fruit-Fly  Parasites 


Plate  1 1 


Journal  of  Agricultural  Researcli 


Vol.  XII,  No.  5 


PLATE  II 
Diachasma  tryoni: 

A. — Lateral  view  of  larva  in  the  second  instar,  showing  particularly  well  the  fat- 
body  of  the  host  recently  taken  in  as  food.     X  loo. 

B. — Lateral  view  of  a  2-day-old  larva  engorged  with  food  and  about  to  molt,  showing 
the  enlarged  and  stiffened  body.     X  loo. 


PLATE  12 
Opius  humilis: 

A,  B. — Dead  larva  in  first  instar;  killed  by  first-stage  larva  of  Diachasma  iryoni, 
showing  cut  on  body  made  by  the  attacking  larva  and  mandibles  extended  in  final 
death  struggle.     X  loo. 

C. — Dead  larva  in  first  instar;  killed  by  first-stage  larva  of  Diachasma  tryoni.  The 
point  of  attack  is  here  clearly  seen.  The  body  contents  have  been  apparently  with- 
drawn from  the  lower  portion  of  the  body  by  the  attacking  larva.     X  loo. 

D. — Dead  larva  in  first  instar:  badly  lacerated  and  distorted  by  attack  of  first-stage 
larva  of  Diachasma  tryoni.  In  this  case  the  dead  larva  was  probably  destroyed  while 
n  the  embryouic  stage  and  a  few  hours  prior  to  the  hatching  of  the  egg.     X  loo. 


Interrelations  of  Fruit-Fly  Parasites 


Plate  12 


Q 


o 


■  '^'^•■^^'■, 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  5 


Interrelations  of  Fruit-Fly  Parasites 


PLATE   13 


^^^H 

H| 

Journal  of  Agricultural  Research 


Vol.  XII,  No.  5 


PLATE  13 

Opius  humilis: 

A. — Dead  larva  in  first  instar,  with  body  shriveled  and  twisted  through  attack  by 
first-instar  larva  of  Diachasma  tryoni.     X  100. 

B. — Dead  larva  in  first  instar;  killed  by  first-instar  larva  of  Diachasma  tryoni.  Here 
the  larva  had  been  feeding  and  developing  for  several  hours  before  being  attacked  by 
the  Diachasma  larva.  The  body  contents  can  be  seen  protruding  from  an  inflicted 
wound  on  the  seventh  and  eighth  body  segments.     X  100. 

C. — Dead  larva  in  first  instar;  killed  by  first-stage  larva  of  Diachasma  tryoni.  This 
is  the  appearance  of  the  dead  Opius  larvae  most  commonly  seen.  The  body  is  pinched 
by  the  mandibles  of  the  Diachasma  larva  in  the  first  or  second  body  segment  back 
of  the  head.     X  100. 

D. — Healthy,  living  larva  in  first  instar.     X  100. 

E. — Healthy,  uninjured,  living  larva  in  first  instar.     X  100. 


ADDITIONAL  COPIES 

Of  THIS  PTJBUCATION  MAY  BE  PBOCtTRED  FBOM 

THE  SUPERINTENDENT  OF  DOCUMENTS 

GOVEBNMENT  PRINTraG  OFFICE 

WASHINGTON,  D,  C. 

AT 

20  CENTS  PER  COPY 
Stjbscbiption  Pbice,  $3.00  Per  Year 


Vol.  XII  KEBRU ARY   11,  1918  No.  6 

JOURNAL  OF 

AGRICULTURAL 
RESEARCH 


CONTE^NXS 


T^mfi 


Water  Eztractions  of  Soils  as  Criteria  of  their  Crop^Pro- 

ducing  Power      --------      297 

JOHN  S.  BURD 

(Contribution  tram  California  Agricultural  Experiment  Station) 

Effect  of  Season  and  Crop  Growth  in  Modifying  the  Soil 

Extract        -       -       -       -       -       -       -       -       -      311 

GUY  R.  STEWART 

(Contribution  from  California  Agricultural  Experiment  Station) 

The  Freezing-Point  Method  as  an  Index  of  Variations  in 

the  Soil  Solution  l)ue  to  Season  and  Crop  Growth       -      369 

D.  R.  HOAGLAND 

(Contribution  from  California  Agricultural  Experiment  Statton) 


PUBUSHED  BY  AUTHORHy  OF  THE  SECRETARY  OF  AGRICULTURE. 

WITH  THE  COOPERATION  OF  THE  ASSOCUTION  OF  AMERICAN 

AGRICULTURAL  COLLEGES  AND  EXPERIMENT  STATIONS 


WASHINGTON,  D.  C- 


'.'A^HINaTO^<:oOVERNUENTPmKTIf<a  OFFICE  ;  lOr; 


.lp'«;MWK.--*««BI»*iVM1Pftl(f^=«»««»*»»«"»'' 


0"  k...^ )^ 


V 


l: 


EDITORIAL  COMMITTEE  OF  THE 
UNITED  pf^Tt'^i  I^lARf^EfW  d^  AND 

THE  IlS^OCiA^IIM  cAiu!^  AOl 

COLLEGES  AND  EXPERIMENT  STATIONS 


fl^URAL 


FOR  THE  DEPARTMENT 

KARL  F.  KELLERMAN,XHAiRMAk 

Physiologist  and  Associate  Chief,  Bureau 
of  Plant  IndusUtv 

EDWIN  W.ALLEN 

Chief,  Office  of  Experiment  Stations 

CHARLES  L.liiARtA'Ift' 

EntomoloQist  and  AssisimU  C^,,  Bureau 


FOR  THE  ASSOCIATION 

RAYMOND  PEARL* 

Biologist,  Maine  Agricultural  Experiment 

H.  p.  ARMSBY 

Dit^ftor,  Institute  of  Animal  Nutrition,  The 
Pennsylvania  State  College 

E.  M.  :^p:Efi^MAN'" 

Botanist,  fla^  jPaShphgist  ftnd^ssfstv!^ 
Dean,  Agricutiiiral  Experiment  Sta^pn  of 

the  University  of  Minnesota  ■  -■  -    ■    ■ 


All  correspondence  regarding  articles  from  the  Department  of  Agriculture  should  be 
addressed  to  K^l  JF,  Kellerjpaii,  Joiup^  pf  Agri<;ultut4  Resmch,  W^ 

*Dr.  Pearl  has  undertaken  special  work  in  cpn^ectioii  with  ti|e  war  epj!?rgency; 
the^ore,  until  further  "notice  all  cbrre^hierice  regarding  articles  from  State  Experi- 
ment Stations  should  be  addressed  to  H.  P.  Armsby,  Institute  of  Animal  Nutrition, 
State  College,  Pa. 


JOMAL  OF  AGRlOimm  ISEARCH 

Vol.  XII  Washington,  D.  C,  February  ii,  1918  No.  6 

WATER  EXTRACTIONS  OF  SOILS  AS  CRITERIA  OF 
THEIR  CROP-PRODUCING  POWER 

By  John  S.  Burd, 
Chemist,  California  Agricultural  Experiment  Station 

THE  PROBLEM  .    V  vuKfe 

Given  sunlight,  suitable  moisture  and  physical  condition,  the  limi-  ^'-^f  ANiCAl 
tation  on  the  power  of  soils  to  produce  crops  is  variously  ascribed  to  the  ^^^Oi^, 
following  general  causes: 

1.  Lack  of  capacity  of  the  soil  to  supply  the  plant  roots  at  all  times 
with  watery  solutions  of  the  essential  elements  in  proper  concentrations, 

2.  Presence  of  toxic  substances. 

3.  Lack  of  physiological  balance  in  dissolved  soil  constituents. 

It  is  clear  that  any  one  or  a  combination  of  these  conditions  may  be 
the  cause  of  low  production,  even  though  the  remaining  conditions  may 
be  entirely  satisfactory.  The  instances  where  the  second  and  third  in- 
hibitive  causes  operate  to  limit  production  are  perhaps  quite  numerous. 
It  may  even  be  that,  because  of  frequent  occurrence,  they  are  of  as  much 
or  more  practical  importance  than  incapacity  of  the  soil  to  supply  the 
plant  with  proper  nutrients.  Nevertheless  we  are  disposed  to  regard 
such  instances  as  special  cases  and  to  lay  greater  stress  on  the  capacity 
of  the  soil  to  deliver  up  solutes  to  the  growing  plant.  Studies  involving 
this  capacity  are  obviously  more  fundamental,  in  that  they  have  to  do 
with  an  important  function  of  all  soils. 

WATER-EXTRACTABLE  SUBSTANCES  IN  SOILS 

When  we  consider  this  function,  we  naturally  turn  to  water  as  the 
logical  agent  for  the  removal  and  determination  of  the  soluble  substances 
in  soils.  The  application  of  water  as  a  solvent  requires  the  use  of  a  suffi- 
cient amount  in  proportion  to  the  soil  to  cause  it  to  come  into  equili- 
brium with  the  true  soil  solution  and  thus  to  insure  the  complete  removal 
of  all  dissolved  matters  when  the  soil  suspension  is  filtered.  Water  in 
excess  unquestionably  removes  not  only  the  substances  already  in  solu- 
tion but  additional  quantities  somewhat  proportional  to  the  amount  of 
water  used.^  Furthermore,  v\e  are  told  that  the  absorption  of  solutes 
by  plants  is  related  to  the  concentration  of  the  soil  solution  and  not  to 

1  HOAGLAND.  D.  R.     THE  FREEZING-POINT  METHOD  AS  AN  INDEX  OF  VARIATIONS  IN  THE  SOn,  SOLUTION 

DUE  TO  SEASON  AND  CROP  GROWTH.     In  Jour.  AgT.  Research,  v.  12,  no.  6,  pp.  369-395,  9  fig.    1918. 

Journal  of  Agricultural  Research,  Vol.  XII,  No.  6 

Washington,  D.  C.  Feb.  n,  1918 

Ik  Key  No.  Cal.— 15 

(297) 


298  Journal  of  Agricultural  Research  voi.  xii,  no.  6 

the  absolute  amounts  present  in  the  soil.  We  heartily  concur  in  this 
opinion;  but  the  impossibility  of  removing  for  investigation  the  true  and 
unmodified  soil  solution,  even  from  soils  whose  optimum  water  content 
is  high,  is  too  well  recognized  to  require  discussion.  It  is  true  that  by 
the  method  of  Bouyoucos  and  McCool  ^  we  can  determine  the  total  con- 
centration of  the  soil  solution,  but  are  left  in  the  dark  as  to  the  concentra- 
tion of  the  individual  solutes  in  that  solution. 

In  the  present  paper,  therefore,  our  data  are  presented  in  terms  of 
the  amounts  of  solutes  extracted  from  soils  and  not  in  terms  of  concen- 
trations. We  shall  endeavor  to  use  this  material  in  such  a  way  as  to 
justify  certain  conclusions  as  to  the  presence  or  absence  of  power  on  the 
part  of  soils  to  supply  the  needs  of  crops  and  to  maintain  optimum 
concentrations  in  the  soil  solution.  Hoagland,^  in  this  laboratory,  has 
shown  that,  while  the  total  amounts  of  material  extracted  from  soils  by 
an  excess  of  water  are  invariably  greater  than  those  contained  in  the  soil 
solution,  they  are  of  the  same  general  order  of  magnitude  (as  2  is  to  4 
or  5),  depending  on  the  type  of  soil.  Obviously,  then,  if  the  amounts  of 
soil  constitutents  extracted  from  cultivated  soils  can  be  shown  to  be 
relatively  high  or  relatively  low,  we  may  legitimately  infer  that  the 
amounts  in  the  soil  solution  are  correspondingly  high  or  low.  We  do 
not  mean  to  imply  that  figures  so  obtained  would  necessarily  indicate 
the  existence  of  an  adequate  supply  of  any  element,  because  even  low 
figures  might  constitute  adequacy.  The  question  of  adequacy  or  in- 
adequacy may  or  may  not  be  answered  by  such  data,  but  it  would  cer- 
tainly reflect  the  relative  magnitudes  of  the  present  supply  although  it 
might  not  indicate  the  soils'  power  of  renewal. 

Data  presented  by  Stewart  ^  in  figures  8  to  20  show  that  normal  seasonal 
fluctuations  in  the  amounts  of  the  essential  elements  extracted  at  different 
times  by  water  (i  to  5)  from  a  cultivated  soil  (cropped  or  uncropped)  are 
likely  to  be  as  great  as  the  variations  between  different  soils.  He  also 
shows  that  soils  under  crop  (barley)  invariably  contain  smaller  amounts 
of  nitrate,  potassium,  calcium,  and  magnesium  than  their  uncropped 
duplicates,  apparently  reflecting  the  inability  of  soils  to  maintain  their 
initial  concentrations  and  at  the  same  time  supply  the  needs  of  growing 
plants.  Since  the  fluctuations  in  the  amounts  of  water-soluble  substances 
in  soils  are  known  to  be  quite  large  and  the  effect  of  withdrawal  important, 
it  follows  that  the  limitations  of  soils  can  best  be  shown  by  data  obtained 
during  the  period  in  which  the  withdrawal  is  actually  taking  place. 

A  comparison  of  the  charts  referred  to  above  shows  that,  even  though 
good  (uncropped)  soils  may  contain  considerably  more  solutes  than  poor 

1  Bouyoucos,  G.  J.,  and  McCool,  M.  M.    The  freezing  point  method  as  a  new  means  of  measuring 

THE  CONCENTRATION  OF  THE  SOn,  SOLUTION   DIRECTLY  IN   THE  SOIL.     Mich.  Agr.    Exp.   Sta.    Tech  Bui. 

24,  p.  592-631,  2  fig.    I916. 

*  HOAGLAND,  D.  R.     Op.  cit. 

3  Stewart,  G.  R.    the  effect  of  season  and  crop  growth  in  modifying  the  sou,  extract.    In  Jour. 
Agr.  Research,  v.  12,  no.  6,  pp.  311-368,  24  fig.,  pi.    1918. 


Feb.  II,  1918     Water  Extractions  of  Soils  and  Crop  Production 


299 


(uncropped)  soils,  both  good  and  poor  soils  are  reduced  to  the  same 
general  level  at  the  time  the  crop  is  growing.  The  condition  of  soils  under 
crop  is  epitomized  in  Table  I. 

Table  I. — Water  extractable  matters  in  soils^^  under  crop.   Average  during  period  ^  of  low 

nitrate  content 

[Results  expressed  as  parts  per  million  of  soil] 


Productivity. 

Good. 

Medium. 

Poor. 

Soil  No  

I 

2 

5 

6 

8 

II 

14 

4 

7 

10 

8 
13 

45 

3 

5 
6 

64 

9 

5 
7 

54 

Nitrate  (NO3) 

Phosphate  (PO4) 

Basic      ions 

(K+Ca+Mg) 

5 

7 

65 

5 

5 

71 

4 
17 

108 

5 
7 

54 

5 
12 

62 

4 
23 

72 

4 
II 

76 

6 
9 

130 

5 
7 

69 

5 
13 

41 

o  For  a  detailed  description  of  soils,  see  Stewart,  G.  R.    Op.  cit. 

b  Period  covered:   Soil  7,  twelfth  to  eighteenth  week  after  planting;  soils  12  and  14,  eighth  to  eighteenth 
week  after  planting;  all  others,  tenth  to  eighteenth  week  after  planting. 

We  observe  in  Table  I  a  general  uniformity  in  the  magnitudes  of  the 
various  solutes  present  at  the  period  of  depletion  in  soils  of  varying  pro- 
ductive capacity.  Apparently  crops  first  reduce  the  supply  of  solutes 
until  these  approach  a  comparatively  low  and  (with  a  few  exceptions) 
fairly  uniform  level.  Subsequent  withdrawals  must  further  reduce  the 
existing  supply  or  be  made  good  by  new  material  coming  into  solution. 

WITHDRAWAL  BY  CROPS 

In  the  second  year's  experiments  reported  by  Stewart  we  installed 
eight  additional  containers  of  soil  i ,  planted  to  barley,  and  harvested  and 
analyzed  the  crop  from  time  to  time.  This  particular  soil  was  chosen 
because  it  was  known  to  be  one  of  our  most  productive  soils  and  the 
amounts  withdrawn  by  the  crop  would  presumably  represent  the  usual 
draft  to  be  expected  from  the  crop  under  consideration.  The  data 
obtained  are  presented  in  Tables  II  and  III.  In  scrutinizing  these  data 
it  must  be  remembered  that  we  are  not  attempting  to  show  the  require- 
ments of  plants  or  their  nutritive  habits,  but  merely  their  effect  on  the 
soil. 

If  we  now  compare  these  data  with  the  water-soluble  content  of  the 
soil,  we  may  develop  the  approximate  relation  between  the  demand  of 
barley  and  the  supply  in  a  productive  soil.  The  relation  is  expressed  in 
the  column  headed  "Ratio."  Literally  interpreted,  the  ratio  for  each 
Ion  expresses  the  number  of  day's  supply  contained  in  the  water  extract 
during  the  several  periods  covered.  Thus,  in  the  case  of  nitrate  we  see 
that  there  was  never  less  than  a  nine  days'  supply  in  the  soil  during  any 
period  of  the  growing  season.     If  the  plant  is  capable  of  absorbing  all  of 


300 


Journal  of  Agricultural  Research 


Vol.  XII,  N'o.  6 


the  nitrate  present  in  the  water-soluble  condition,  we  have  direct  evidence 
that  the  water-soluble  nitrate  need  only  be  renewed  once  in  nine  days  even 
at  this  critical  period,  while  at  all  preceding  and  subsequent  periods  the 
rate  of  renewal  may  be  much  slower.  If  we  apply  this  same  reasoning  to 
all  of  the  ions,  we  develop  the  fact  that  there  are  always  present  in  this 
particular  soil,  in  a  condition  capable  of  solution  in  water,  at  least — 

Nine  days'  supply  of  nitrate. 

Twenty-five  days'  supply  of  phosphate. 

One  hundred  and  forty-four  days'  supply  of  potassium. 

Two  hundred  and  sixty-five  days'  supply  of  calcium. 

Seventy-six  days'  supply  of  magnesium. 

Table  II. — Rate  of  zvithdrawal  by  crop  «  as  shown  by  periodic  harvesting  of  soil  I 
iWithdrawals  computed  to  parts  per  million  of  soil — 50  plants  and  i  ,800  pounds  of  soil  to  the  unit  containerl 


May  I  (planted) 

Jtme  12 

June  26 

July  12 

July  24 

August  7 

August  21 

August  28 


Nitrate  (NO3). 6 


Grams 

per 
plant. 


•405 

(?) 
.  520 

•594 
.  629 
.664 
.694 


Parts  per  million 
of  soil. 


Entire. 


24.85 


31.90 
36.40 
38.40 
40.  80 

42.  50 


Per 

day. 


0-59 


Phosphate  (PO^).^ 


Grams 

per 
plant. 


o.  000 
.038 

•115 
.  120 

•  144 
.  184 

•  194 


Parts  per  million 
of  soil. 


Entire. 


2.32 

7.  02 

7-35 

8.  10 
8.82 

II.  20 
II.  90 


Per 
day. 


0-055 

.  024 
.054 
.051 


Potassium  (K). 


Grams 

per 
plant. 


O.  000 

•  ^33 
(?) 
(?) 
.223 

•233 
.  276 
.294 


Parts  per  million 
of  soil. 


Entir  e. 


Per 

day. 


8.  131  o.  194 


13-70 
14.30 
16.  90 
18.  00 


133 
043 
186 

157 


Date. 


May  I  (planted) 

June  12 

June  26 , 

Jtily  12 

July  24 

August  7 

August  21 

August  28 


Calcium  (Ca). 


Grams 

per 
plant. 


o.  000 

.  012 
•034 
(?) 
.  041 

(?) 
.048 

•043 


Parts  per  million 
of  soil. 


Entire.       Per  day. 


o.  712 
2.  092 


2-530 


2.  940 
2.650 


o.  017 


.  016 
.  014 


Magnesium  (Mg). 


Grams 

per 
plant. 


o.  000 
.013 
.031 
.036 

(?) 
.044 

•049 
•051 


Parts  per  million 
of  soil. 


Entire.       Per  day. 


0.816 

1.  920 

2.  190 


2.  670 

2.  980 

3.  140 


o.  019 

.079 
.  019 


017 
,  022 
023 


o  Analyses  based  on  varying  numbers  of  plants:  June  12,  69  plants;  June  26,  42  plants;  July  12,  42  plants; 
July  24,  29  plants;  August  7.  28  plants;  August  21,  38  plants;  August  28,  87  plants. 

6  Nitrate  (NO3)  and  phosphate  (POi)  computed  from  the  amounts  of  nitrogen  and  phosphorus  in  the 
crop. 


The  entire  time  the  crop  was  in  the  ground  was  119  days.  The  time 
to  elapse  after  the  period  of  greatest  depletion  was  35  days  for  nitrate, 
63  days  for  phosphate,  77  days  for  potassium,  63  days  for  calcium,  and  63 


Feb.  II,  1918     Water  Extractions  of  Soils  and  Crop  Production 


301 


days  for  magnesium.     It  would  appear  that  there  is  actually  present  in 

the  water-soluble  condition  at  the  period  of  greatest  depletion  more  than 
enough  potassium,  calcium,  and  magnesium  to  supply  the  entire  subse- 
quent withdrawals  of  the  crop  at  the  same  rate,  and  that  any  renewal  or 
increase  of  these  amounts  should  be  unnecessary  if  the  plant  is  capable 
of  utilizing  the  entire  supply  present. 


Table  III. 


-Comparison  of  water-extractable  matters  in  soil  I  (cropped)  with  daily  draft 
by  plant 


Nitrate  (NO3). 

Phosphate  (PO<). 

Potassium  (K). 

Weeks. 

Parts  per 

million  of 

soil. 

Ratio. 

Weeks. 

Parts  per 

million  of 

soil. 

Ratio. 

Weeks. 

Parts  per 

million  of 

soil. 

Ratio. 

In 

soil. 

Daily 
draft. 

In 

soil. 

Daily 
draft. 

In 
soil. 

Daily 
draft. 

1-6,  inclusive . 
7-10,  inclusive 
11-12, inclusive 
i3-i4,inclusive 

69 

4 
3 
8 

O.S9 

.14 

117 
16 
9 

57 
18 

1-6,  inclusive .  . 
7-8,  inclusive . . 
9-10,  inclusive. 
11-12,  inclusive 
13-14,  inclusive 
15-16,  inclusive 
17th 

7 
8 
10 
3 

8 

5 

0.055 
.336 
.024 
.054 
.051 
.  170 
.  100 

127 
24 

417 
55 

157 
29 

1-6,  inclusive . . 
7-12,  inclusive. 
13-14,  inclusive 
15-16,  inclusive 
17th 

28 
29 
34 
38 

0.194 
•133 
•043 
.186 
•  157 

144 
217 
790 
204 

1 

i                                   1 

Calcium  (Ca). 

Magnesium  (Mg). 

Weeks. 

Parts  per 

million  of 

soil. 

Ratio. 

Weeks. 

Parts  per 

million  of 

soil. 

Ratio. 

In 

soil. 

Daily 

draft. 

In 
soil. 

Daily 
draft. 

28 
26 
18 
21 

1.6^7 

1-6,  inclusive 

IS 
6 

14 
9 

20 

0.  019 
.079 
.  019 
.0x7 
.  022 
•  023 

789 

.  098]          265 

.016         Iji:5 
.  014         I,  500 

76 

17th     

17th 

1 

It  is  evident  that  nitrate  and  phosphate  are  in  a  class  by  themselves, 
in  that  the  supply  must  be  renewed  at  a  much  more  rapid  rate  than  any 
of  the  other  ions;  but,  inasmuch  as  this  soil  was  highly  productive,  it 
must  have  possessed  the  power  to  replenish  the  soil  solution  as  required. 
But  we  have  shown  heretofore  that  there  is  very  little  difference  in  the 
absolute  amounts  of  solutes  between  good  and  poor  soils  under  a  crop  of 
barley.  It  follows  that  figures  of  the  same  order  of  magnitude  as  those 
just  stated  would  be  obtained  by  similar  computations  from  the  poor  soils. 
We  conclude,  therefore,  that  in  soils  in  any  degree  fit  for  agricultural 
purposes  (with  the  possible  exception  of  very  abnormal  types,  such  as 
peat,  etc.),  the  potentially  soluble  matters  are  sufficient  in  amount  to  meet 
the  requirements  of  crops;  but  it  does  not  follow  that  the  plant  is  capable 
of  drawing  upon  this  supply  at  the  concentrations  corresponding  to 
these  amounts. 


302 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  6 


CONCENTRATION  AND   RATE   OF  FORMATION  OF  SOLUTES 

Since  even  our  poor  soils  appear  to  be  able  to  furnish  adequate  amounts 
of  all  solutes  at  the  most  critical  periods,  we  can  only  ascribe  nutritive 
difficulties  of  soils  to  inability  on  the  part  of  the  plant  to  absorb  this  po- 
tentially soluble  supply.  The  rate  of  absorption  by  any  plant  may  be 
limited  by  the  possible  supply  and  by  the  concentration  of  that  supply, 
but  we  have  seen  that  in  all  cases  which  have  come  under  our  observation 
the  supply  is  adequate  at  all  times.  We  should  therefore  either  measure 
the  concentration  of  the  soil  solution  itself  or  evolve  some  other  means 
of  estimating  the  relative  power  of  soils  to  maintain  an  optimum  concen- 
tration. Our  data  do  not  permit  us,  unfortunately,  to  estimate  the 
concentration  of  individual  ions  in  the  soil  solution,  but  we  may  per- 
haps get  an  idea  of  the  relative  power  of  soils  to  maintain  or  restore  a 
proper  concentration  by  consideration  of  the  data  on  soils  with  and 
without  crop.  We  present  in  Table  IV  the  seasonal  averages  of  the 
water-extractable  substances  in  cropped  and  uncropped  soils. 


Table    IV. — Water-extractable  matters   in  cropped  and  uncropped   soils. 

averages 

[Results  expressed  as  parts  per  million  of  soil] 


Seasonal 


Constituent. 

Yield  in  bushels  per  acre . 
Soil  No 


Nitrate : 

Uncropped. 
Cropped .... 
Difference. . 

Phosphate: 

Uncropped . 
Cropped .  . . . 
Difference. . 

Potassium: 

Uncropped. 
Cropped .  . . . 
Difference. . 

Calcitiin: 

Uncropped. 
Cropped. .. . 
Difference. , 

Magnesium : 
Uncropped. 
Cropped . . . . 
Difference. 


Productivity. 


131 
36 

95 


40 

31 

9 


34 
24 
10 


23 
13 
10 


120 
87 


55 

50 

5 


37 
19 
18 


15 

7 


146 

30 
116 


18 

17 


47 
14 


91 

58 
Z2> 


24 
17 

7 


Good. 


8s- 9 


141 

43 


Z3 

27 

6 


47 
28 

19 


87.0 


180 

28 

152 


17 
12 

5 


49 
31 
18 


66 
34 
32 


15 
10 

5 


154 

27 

127 


30 
26 


67 


19 


51 
29 
22 


13 
7 
6 


96 
23 
73 


13 


Aver- 
age all. 


S4.  7 


137 

30 

107 


14 


52 
40 
12 


51 
30 
21 


17 
10 

7 


Feb.  II,  1918     Water  Extractions  of  Soils  and  Crop  Production  303 


TabIvE  IV. — Water-extr actable    matters    in    cropped   and    uncropped    soils.     Seasonal 

averages — Continued 


Constituent. 


Yield  in  bushels  per  acre 
Soil  No 

Nitrate: 

Uncropped 

Cropped 

Difference 

Phosphate : 

Uncropped 

Cropped 

Difference 

Potassium: 

Uncropped 

Cropped 

Difference 

Calcium: 

Uncropped 

Cropped 

Difference 

Magnesium : 

Uncropped 

Cropped 

Difference 


Productivity. 


Medium. 


141 

33 
108 


45 
40 

5 


66 
18 


26 


69.4 


118 
34 


31 
22 

9 


70.  6 


130 

39 
91 


49 
30 
19 


Aver- 
age all. 


129 

35 
94 


9 

10 

—  I 


42 
31 


66 

44 


16 


56. 2 


88 
30 
58 


43 

37 

6 


32 

25 

7 


14 


4S-8 


54 

18 
36 


44 

35 

9 


83 
23 
60 


30 
20 
10 


41 
24 

17 


Aver- 
age all. 


75 
24 

51 


31 

25 

6 


39 
28 


Examination  of  uncropped  soiIvS. — ^We  regard  the  magnitudes  shown 
as  the  resultant  of  the  combined  effects  of  previous  withdrawals,  the  time 
which  has  elapsed  since  the  soil  was  last  depleted,  and  the  rate  at  which 
the  soil  is  capable  of  replacing  solutes  removed  from  the  soil  solution.  A 
soil  might  show  high  figures  simply  because  previous  withdrawals  by 
crops  had  been  relatively  low  or  remote  in  time.  Nevertheless  it  is 
significant  that  the  average  content  of  the  more  productive  soils  for  each 
constituent  is  relatively  considerably  higher  than  the  average  for  the  poor 
soils.  We  infer  from  this  that  high  figures  ordinarily,  but  not  necessarily, 
indicate  a  relatively  high  rate  of  elaboration  of  solutes  on  the  part  of  the 
soil.  A  conspicuous  example  of  an  exception  to  this  would  be  the  case  of 
a  so-called  alkali  soil,  where  large  accumulations  of  solutes  might  have  no 
relation  to  the  present  elaborate  power  of  the  soil. 

Examinations  of  cropped  soils. — While  all  averages  of  the  good  soils 
under  crop  are  higher  than  the  averages  of  the  poor  soils,  the  magnitudes  of 
the  differences  are  quite  small.  Apparently  the  crop  by  withdrawal  or 
indirect  effects  on  the  soil  tends  to  reduce  the  initial  amounts  of  solutes 


304  Journal  of  Agricultural  Research  voi.  xn,  no.  6 

(and  presumably  their  concentrations)  to  the  same  general  level  in  both 
good  and  poor  soils.  It  seems  probable  that,  when  certain  minima  are 
reached,  the  plant  can  not  absorb  sojutes,  and  subsequent  withdrawal 
then  depends  upon  the  capacity  of  the  soil  to  elaborate  additional  solutes 
as  rapidly  as  the  plant  requires  them.  The  figures  themselves,  however, 
can  not  give  a  direct  measure  of  this  capacity,  but  small  variations  may 
possibly  reflect  important  differences  in  the  rate  at  which  solutes  are 
elaborated  by  different  soils. 

EquivaIvEncs  of  bases. — If  we  compare  the  figures  for  the  individual 
basic  ions  in  good  and  poor  soils  (both  in  the  cropped  and  uncropped 
condition),  we  find  numerous  instances  in  which  some  one  or  more  ions 
are  lower  in  the  good  soils  than  in  the  poor  soils.  These  discrepancies 
need  not  concern  us  if  we  regard  the  figures  merely  as  a  means  of  getting 
at  the  power  of  soils  to  replenish  the  soil  solution.  In  fact,  this  power  is 
more  likely  to  be  reflected  by  the  aggregated  soluble  matters  than  by 
individual  constituents,  and  preferably  by  the  aggregate  of  those  con- 
stituents of  similar  chemical  properties.  We  propose,  therefore,  to  con- 
solidate our  data  for  the  basic  ions  in  one  figure  (by  addition)  for  subse- 
quent consideration.  If  we  further  consider  all  of  our  data  with  reference 
to  the  seasonal  requirements  of  a  good  crop,  we  can  perhaps  indicate  why 
some  soils  are  more  productive  than  others. 

In  Table  II  we  gave  the  amounts  of  the  chemical  elements  found  in  a 
good  crop  of  barley  produced  on  soil  i ,  computed  to  parts  per  million  of 
soil.  These  figures  were  42.50  p.  p.  m.  of  nitrate,  11.90  p.  p.  m.  of  phos- 
phate, 18.00  p.  p.  m.  of  potassium,  2.9  p.  p.  m.  of  calcium,  and  2.98  p.  p.  m. 
of  magnesium;  or  in  round  numbers,  42  p.  p.  m.  of  nitrate,  12  p.  p.  m.  of 
phosphate,  and  24  p.  p.  m.  of  total  bases.  We  assume  these  figures  to 
represent  the  probable  requirements  of  a  good  crop  of  barley  on  all  soils 
and  propose  to  examine  our  data  with  reference  to  the  capacity  of  the 
various  soils  to  meet  this  requirement.  In  Table  V  we  combine  all 
pertinent  data  available. 

Nitrate. — The  figures  for  excess  in  the  good  and  medium  soils  show 
an  extraordinary  discrepancy  between  the  amounts  of  nitrate  in  the 
uncropped  soil  and  the  amounts  found  in  the  cropped  soil  plus  the  with- 
drawal of  a  good  crop.  We  can  not  satisfactorily  account  for  this  dis- 
crepancy, which  appears  to  be  ca,used  in  part  by  an  inhibition  of  nitrifi- 
cation due  to  the  presence  of  the  crop,  but  which  also  represents  an  actual 
loss  of  nitrate  from  the  cropped  soil.  Since  the  soils  were  kept  in  tight 
containers,  there  could  be  no  loss  from  drainage;  and  neither  denitrifica- 
tion,  reduction  to  ammonia,  nor  the  possible  loss  of  ammonia  by  way  of 
the  plant  appeals  to  us  as  an  adequate  explanation.  We  can  only  sug- 
gest that  the  presence  of  a  crop  may  cause  such  a  change  in  the  biological 
environment  of  the  soil  that  the  nitrogen  of  nitrates  is  stored  in  insoluble 
(protein)  forms  in  the  soil.  For  our  present  purposes,  however,  it  is 
only  necessary  to  point  out  that  such  losses  appear  to  be  a  necessary 


Feb.  II,  1918     Water  Extractions  of  Soils  and  Crop  Production 


305 


incident  to  the  production  of  a  good  crop.  If  the  soil  is  not  capable  of 
sustaining  such  losses,  it  is  extremely  probable  that  crops  will  suffer  from 
a  lack  of  nitrates. 

Table  V. — Water-extractable  matters  with  reference  to  seasonal  withdrawals  of  a  good  crop 
[Expressed  as  parts  per  million  of  soil] 


Constituent. 

Productivity. 

Good. 

Medium. 

Poor. 

Soil  No    

I 

2 

5 

6 

8 

II 

14 

4 

7 

10 

30 
42 

9 

18 
42 

12 

Nitrate  (NO3) 

36 
42 

33 
42 

30 
42 

43 
42 

28 
42 

27 
42 

23 
42 

33 
42 

34 
42 

39 

42 

23 

42 

(c)  Suni(a+6) 

78 
131 

75 
120 

72 
146 

85 
141 

70 
180 

69 

154 

65 
96 

75 
141 

76 

118 

81 
130 

72 
88 

60 

54 

f>f 

8^? 

63 

55 

74 

56 

no 

85 

31 

66 

42 

49 

16 

—6 

i8 

Phosphate  (P04): 

6 

12 

5 
12 

17 
12 

8 
12 

12 
12 

26 

12 

II 
12 

10 
12 

7 
12 

13 
12 

6 

12 

6 
12 

12 

12 

(c)Smn(a+6) 

18 
7 

17 
5 

29 
18 

20 

8 

24 

17 

38 
30 

23 
13 

22 
II 

19 
7 

25 
II 

18 
7 

18 
6 

24 

((/)  In  uncropped  soil 

13 

II 

12 

11 

12 

7 

8 

10 

83 
24 

II 

12 

14 

II 

12 

II 

Bases  (KCaMg): 

68 
24 

76 
24 

108 
24 

65 
24 

75 
24 

84 
24 

128 
24 

73 
24 

61 

24 

73 
24 

61 
24 

50 

(c)  Sum  (a+6)  

92 

97 

100 

107 

132 
162 

89 

lOI 

99 

130 

108 
131 

107 
113 

152 
155 

97 
lOS 

85 
112 

97 
89 

85 

72 

74 

80 

5 

7 

30 

12 

31 

23 

6 

3 

8 

27 

-8 

—  13 

6 

When  we  turn  to  the  poor  soils,  we  see  that  they  all  had  in  the  un- 
cropped condition  more  than  enough  nitrate  to  supply  the  actual  amounts 
withdrawn  by  a  good  crop;  but  all  were  incapable  of  sustaining  addi- 
tional losses  of  the  magnitudes,  which  we  must  regard  as  normal. 

Phosphate. — The  figures  for  all  soils  under  crop  plus  the  requirements 
of  a  good  crop  are  invariably  greater  than  the  amounts  in  the  uncropped 
soils.  If  the  figures  for  these  latter  represent  in  each  case  the  maximum 
amount  of  soluble  phosphate  the  soil  is  able  to  hold  in  that  condition, 
it  is  quite  clear  that  either  the  plant  absorbs  insoluble  phosphate,  or  the 
good  soils  replace  the  soluble  phosphate  as  rapidly  as  it  is  required  by 
the  plant.  The  latter  explanation  appears  the  more  probable.  But  the 
computed  deficiencies  of  the  poor  soils  are  no  greater  than  those  of  the 
good  soils,  nor  are  the  soluble  phosphate  contents  of  the  former  less  than 
many  of  the  latter.  The  rate  of  solution  of  phosphate  in  the  good  soils 
must  be  very  high  and  we  can  find  no  reason  to  conclude  that  the  poor 
soils  are  in  any  way  inferior  in  this  respect. 

Basic  ions. — The  amounts  of  basic  ions  in  cropped  soils  plus  the 
requirements  of  a  good  crop  are  usually  less  than  the  amounts  in  the 
uncropped  soils.     We  note  two  exceptions  to  this  in  the  poor  soils  Nos. 


3o6  Journal  of  Agricultural  Research  voi.  xii,  no.  6 

3  and  9.  For  these  soils  to  have  supplied  the  requirements  of  a  good 
crop  and  at  the  same  time  to  have  maintained  their  general  level  of 
concentration  would  have  necessitated  that  more  bases  come  into  solu- 
tion. It  is  not  impossible  that  they  would  have  responded  to  the  demand 
for  bases,  just  as  they  would  probably  have  responded  to  the  demand 
for  phosphate.  The  important  difference  between  these  poor  soils  and 
the  remaining  soils  is  that  the  latter  would  never  have  been  called  upon 
to  furnish  larger  amounts  of  bases  than  the  uncropped  soil  shows  ability 

to  supply. 

CRITERIA   OF  FERTILITY 

The  growth  of  a  crop  reduces  the  average  nitrate  content  of  soils  to  a 
comparatively  uniform  level  in  soils  of  all  degrees  of  productivity  (see 
Tables  IV  and  V).  The  figures  for  nitrate  in  uncropped  soils  are  always 
higher  than  the  known  withdrawals  by  plants,  but  may  not  always  be 
sufficiently  high  to  supply  these  withdrawals  and  certain  other  inherent 
losses.  The  relative  ability  of  soils  to  meet  these  losses  may  be  inferred 
from  the  amounts  of  nitrate  in  the  uncropped  soils. 

Except  in  a  few  soils  containing  large  amounts  of  soluble  phosphate, 
the  growth  of  a  crop  does  not  reduce  this  ion  to  any  considerable  extent. 
(The  small  differences  shown  approach  the  magnitude  of  experimental 
error.)  Furthermore,  the  figures  do  not  show  that  good  soils  possess  the 
power  of  renewing  the  soluble  phosphate  more  rapidly  than  the  poor 
soils.  Figures  for  soluble  phosphate  can  not  be  considered  to  reflect  the 
relative  power  of  soils  to  supply  the  plant. 

The  growth  of  a  crop  reduces  the  basic-ion  content  of  soils,  but  the 
amounts  remaining  are  still  far  in  excess  of  the  crop  requirements.  It  is 
not  improbable,  however,  that  their  concentrations  are  in  some  cases 
falling  below  the  optimum  requirements  of  plants.  If  this  be  the  case, 
comparison  of  the  basic  solutes  of  cropped  soils  may  indicate  deficiency 
in  this  respect.  The  differences  between  the  basic  ion  contents  of  cropped 
and  of  uncropped  soils  are  usually,  but  not  always,  greater  than  the 
demands  of  crops.  These  differences  therefore  express  the  relative 
power  of  soils  to  supply  the  crop  requirements  and  to  maintain  the 
concentration  equivalent  to  the  amounts  shown  by  the  cropped  soil, 
without  drawing  upon  greater  amounts  of  solutes  than  the  uncropped 
soil  indicates  capacity  to  supply. 

Test  op  criteria. — If  we  arrange  our  soils  in  the  order  in  which  they 
possess  the  various  characters  to  which  attention  has  been  called,  we 
may  be  able  to  bring  out  salient  differences  between  good  and  poor  soils. 
The  order  in  which  the  soils  occur  in  each  character  is  frequently  deter- 
mined by  very  small  variations  in  absolute  amount  of  solutes,  but  it  must 
be  remembered  that  these  are  based  on  considerable  numbers  of  analytical 
determinations  and  are  probably  not  vitiated  by  experimental  errors.* 

1  Stewart,  G.  R.    Op.  cit. 


Feb.  II.  1918     Water  Extractions  of  Soils  and  Crop  Production  307 

Very  small  differences  at  critical  points  may  reflect  significant  differences 
in  the  performance  of  soils  (fig.  i). 

All  of  the  poor  soils  appear  among  the  three  lowest  in  two  out  of  three 
characters.  If  we  examine  the  medium  soils,  we  find  that  Nos,  4  and  10 
are  among  the  three  lowest  in  one  character.  If  we  draw  a  line  over  the 
three  lowest  characters  as  a  tentative  indicator  of  the  existence  of  sub- 
optimal  conditions,  we  find  that  five  of  our  six  medium  and  poor  soils 
fall  below  in  one  or  more  characters;  that  the  remaining  medium  soil, 
No.  7,  approaches  it  in  one  character;  that  none  of  the  good  soils  fall 
below  in  any  character.  It  is  true  that  several  of  the  good  soils,  No. 
14,  6,  and  i,  approach  the  line  very  closely,  and  we  do  not  wish  to  be 
understood  as  attaching  too  much  importance  to  slight  differences  in  the 


Fig.  1. — Graphs  showing  soils  arranged  with  reference  to  yield  and  important  characters. 

magnitudes  of  these  characters.  It  would  appear,  however,  that  the 
yields  of  soils  7,  3,  12,  and  9  are  closely  correlated  with  their  nitrate 
content,  but  the  mediocre  yields  of  soils  4  and  10  can  only  be  accounted 
for  by  a  reference  to  other  characters. 

We  conclude  that  the  nitrate  content  of  uncropped  soils  is  the  most 
valuable  single  criterion  for  appraising  the  crop-producing  power  of 
soils;  that  the  amount  of  basic  ions  in  cropped  soils  is  indicative  of  the 
extent  to  which  soils  tend  to  maintain  their  concentrations  when 
subjected  to  depletion  by  crops;  that  the  differences  in  basic-ion  content 
between  cropped  and  uncropped  soils  may  reflect  the  ability  of  the  soil 
to  meet  the  demands  of  the  crop  without  dangerous  diminution  of  con- 
centration and  without  drawing  upon  the  reserve  (iasoluble)  supply 
of  the  soil. 


3o8  Journal  of  Agricultural  Research  voi.  xii,  no.  6 

CONCLUSIONS 

The  evidence  presented  in  the  early  part  of  this  paper  indicates  that 
there  is  always  present  in  soils,  in  a  condition  permitting  ready  solution 
in  water,  enough  of  the  more  important  chemical  elements  to  supply 
the  immediate  needs  of  crops.  It  is  hardly  conceivable  that  substances 
in  this  condition  do  not  represent  a  potentially  available  supply.  Inas- 
much as  this  supply  never  entirely  disappears,  even  in  the  case  of 
nitrates,  it  would  seem  that  there  is  no  such  thing  as  a  lack  of  available 
nutrients  in  soils  which  are  at  all  productive,  but  that  a  plant  may  still 
be  unable  to  satisfy  its  requirements  if  the  concentration  with  reference 
to  the  individual  ions  falls  below  certain  minima.  It  is  furthermore 
highly  probable  that  the  optimum  concentration  varies  with  every  soil 
in  accordance  with  the  physicochemical  system  present  in  the  soil  solution. 
Slight  differences  in  the  character  of  this  system  may  modify  in  a  marked 
degree  the  power  of  a  plant  to  absorb  solutes,  so  that  even  if  we  were 
able  to  obtain  and  analyze  the  true  soil  solution,  we  would  not  necessarily 
be  able  to  say  that  any  figure  for  individual  ions  constituted  inadequacy. 

Attention  has  been  called  to  certain  characters  as  reflecting  the  com- 
position of  the  soil  and  its  power  to  produce  crops.  Inasmuch  as  these 
involve  three  variables  to  which  it  is  impossible  to  assign  definite  rela- 
tive weights,anexactcorrelation  of  productivity  with  the  figures  presented 
is  not  to  be  expected.  Nevertheless,  the  correlation  between  the  general 
magnitudes  of  the  figures  presented  and  the  crop-producing  powers  of 
the  soils  studied  is  sufficiently  close  to  justify  the  belief  that  they  give 
expression  to  the  relative  power  of  soils  to  produce  crops,  although  they 
are  not  an  exact  measure  of  that  power. 

We  believe  that  the  evidence  obtained  is  sufficient  to  justify  the  hope 
that  we  may  be  able  to  predict,  within  reasonable  limits,  thte  relative 
crop-producing  powers  of  soils  by  comparing  their  figures  expressing 
these  characters  with  similar  data  derived  from  soils  whose  productive 
power  is  known.  Before  such  a  method  is  generally  applicable,  how- 
ever, it  will  be  necessary  to  study  the  behavior  of  many  soils  with  numer- 
ous type  crops.  This  is  quite  feasible  if  the  various  characters  can  be 
developed  without  the  enormous  number  of  analytical  determinations 
involved  in  the  present  experiments.  We  believe  that  this  can  be 
accomplished  by  substituting  for  our  figures,  representing  the  sums  of 
the  basic  ions,  figures  obtained  for  total  soluble  salts,  or  preferably 
direct  determinations  of  the  concentration  of  the  soil  solution  by  some 
such  method  as  that  presented  in  the  preceding  paper.^  It  is  quite 
certain  that  we  shall  never  have  a  precise  measure  of  soil  fertility  until 
soils  are  studied  concurrently  in  the  cropped  and  the  uncropped  con- 
dition and  under  strict  control.  The  reasons  are  obvious  in  that  data 
from  the  soil  under  crop  can  not  indicate  its  latent  power,  and  data 

'HOAGLAND,  D.  R.      Op.  cit. 


Feb.  II,  191S     Water  Extractions  of  Soils  and  Crop  Production  309 

from  the  uncropped  soil  taken  alone  do  not  take  into  account  the  fact 
that  the  solutes  in  the  cropped  soils  can  not  be  reduced  below  certain 
minimum  and  probably  variable  limits. 

In  the  present  paper  we  have  dealt  entirely  with  chemical  criteria 
because  they  afiford  the  most  convenient  expression  of  the  results  of  the 
activities  of  the  soil,  chemical,  physical,  and  biological.  We  do  not 
wish  to  be  understood  as  minimizing  the  importance  of  biological  stu4ies 
because  we  regard  living  organisms  as  being  the  most  important  single 
agency,  through  the  formation  of  nitrates  and  carbonic  acid,  in  modify- 
ing the  soil  solution.  While  biological  studies  have  a  most  important 
bearing  on  the  proper  treatment  of  soils,  the  resultant  effects  of  all 
activities  upon  these  heterogeneous  mixtures  can  only  be  developed  by 
the  actual  growth  of  crops  and  observation  of  their  effects. 


EFFECT  OF  SEASON  AND  CROP  GROWTH   IN  MODIFY- 
ING THE  son.  EXTRACT 

By  Guy  R.  Stewart/ 
Assistant  Chemist,  California  Agricultural  Experiment  Station 

HISTORICAL  REVIEW 

The  first  studies  on  the  water-soluble  material  of  soils  were  inspired 
by  the  results  of  the  investigations  of  the  absorption  of  plant  nutrients 
carried  on  by  Thompson  (55)^  and  by  Way  (60).  Their  epoch-making 
discoveries  in  the  year  1850  aroused  a  widespread  discussion  of  the 
manner  in  which  the  essential  compounds  might  be  held  in  the  soil. 

Liebig  (59)  was  greatly  impressed  by  this  work  and  carried  out  a 
series  of  investigations  in  which  he  studied  the  absorption  of  calcium 
phosphate  and  potassium  sulphate.  He  concluded  that,  since  the 
phosphate  and  potassium  radicals  were  so  readily  absorbed  by  soils, 
very  little  of  these  essential  nutrients  could  be  present  in  the  soil  water. 
The  results  of  lysimeter  and  drainage  studies  confirmed  him  in  his  belief, 
and  he  proposed  the  theory  that  the  plant  roots  must  be  able  to  draw 
nourishment  directly  from  the  soil  particles.  At  the  same  time  Grouven 
(22)  had  analyzed  the  extract  from  three  soils  obtained  by  percolating 
6,000  c.  c.  of  distilled  water  through  2,000  gm.  of  soil.  He  attempted 
to  relate  these  figures  to  the  amount  of  material  that  would  be  brought 
into  solution  by  the  season's  rainfall. 

Eichhom  {14)  also  studied  a  soil  near  Bonn,  Germany,  in  an  attempt 
to  obtain  a  solution  which  would  approximate  the  moisture  existing  in 
the  soil.  He  added  36.5  per  cent  by  weight  of  water,  and  allowed  it 
to  stay  in  contact  with  the  soil  for  10  days.  He  concluded  that  the 
soil  contained  all  water-soluble  nutrients  necessary  for  raising  a  crop. 

This  view  was  severely  criticized  by  Wunder  (63),  who  held  with 
Liebig  {39)  that  the  soil  could  not  furnish  sufficient  water-soluble  ma- 
terial. Schumacher  (52)  repUed  to  Wunder  (65),  upholding  Eichhorn's 
views.  Gradually,  through  the  work  of  Peters  (45),  Jarriges  (31), 
Ulbricht  (38),  Hoffman  (28),  Wolff  (62),  and  Cossa  (13),  it  became 
evident  that  the  water  extract  from  soils  contained  the  major  plant 
nutrients. 

1  The  writer  desires  to  make  acknowledgment  of  the  assistance  of  the  Division  of  Soil  Technology,  of 
the  California  Experiment  Station,  in  selecting  the  soils  employed  in  the  investigation  and  in  perform- 
ing the  physical  analyses  reported  in  Table  II.  The  writer  is  also  indebted  to  Mr.  A.  W.  Christie,  of 
this  station,  for  the  performance  of  ammonification  and  nitrification  studies  and  to  Messrs.  H.  E.  BiUings, 
A.  W.  Christie,  and  J.  C.  Martin,  of  this  Station,  for  assistance  in  portions  of  the  analytical  work  per- 
formed in  1916. 

*  Reference  is  made  by  number  (italic)  to  "  Literature  cited,"  p.  364-368. 


Journal  of  Agricultural  Research,  Vol.  XII,  No.  6 

Washington,  D.  C.  ^^^-  "•  ^9i8 

I-  Key  No.  Cal.— 16 

(31O 


312  Journal  of  Agricultural  Research  voi.  xii,  no.  6 

The  existence  of  appreciable  quantities  of  water-soluble  phosphates 
was  the  point  longest  in  doubt,  though  the  results  of  Heyden  {25)  and 
Schulze   {51)   established  this  fact  satisfactorily. 

The  most  notable  of  the  early  investigations  and  those  which  have 
had  the  greatest  influence  on  modern  work  were  performed  by  Schloesing 
(46,  47).  His  method  consisted  of  treating  30  to  35  kgm.  of  soil  with 
an  artificial  rain  and  then  collecting  the  first  portions  of  clear  solution 
which  ran  through.  This  he  believed  to  be  identical  with  the  actual 
soil  solution,  and  his  was  the  first  attempt  to  obtain  it  in  an  unaltered 
condition.  Schloesing  showed  the  presence  of  all  the  principal  elements 
in  this  soil  extract,  and  his  procedure  is  still  occasionally  used  in  Euro- 
pean work.  Schloesing,  jr.  {4.8,  49),  has  continued  this  portion  of  his 
father's  work,  devoting  special  attention  to  water-soluble  phosphates, 
and  has  concluded  that  there  are  differences  in  the  phosphate  content 
of  various  soils,  and  also  that  there  is  almost  enough  soluble  phosphate 
present  in  most  soils  to  supply  an  average  crop. 

The  first  work  performed  in  the  United  States  on  the  water-soluble 
material  of  soils  was  that  undertaken  by  King  (36- j8) ,  and  his  coworkers 
at  Wisconsin.  This  was  largely  devoted  to  studies  of  the  nitrate  con- 
tent of  cultivated  field  soils,  and  was  later  extended  to  studies  of  the 
total  salt  content  by  the  use  of  conductivity  methods.  The  results 
obtained  in  this  work  may  be  considered  to  have  been  merely  prelim- 
inary to  King's  more  extensive  investigations  performed  in  the  Bureau 
of  Soils  of  the  United  States  Department  of  Agriculture  {34,  35). 

Before  this  later  work  of  King's  appeared,  Whitney  and  Cameron  (61) 
issued  a  publication  from  the  United  States  Bureau  of  Soils  which  has 
attracted  more  attention  than  any  other  single  paper  dealing  with  water 
extracts.  In  it  they  gave  the  amounts  of  phosphoric  acid,  nitrates,  cal- 
cium, and  potassium  found  in  the  water  extracts  of  both  good  and  poor 
soils  under  varying  conditions.  The  extracts  employed  for  these  analyses 
were  obtained  by  stirring  100  gm.  of  soil  in  500  c.  c.  of  distilled  water 
for  three  minutes.  After  standing  for  20  minutes  the  liquid  was  decanted 
into  a  cylinder  of  Brigg's  (6)  filtering  apparatus  and  forced  through 
unglazed  Pasteur-Chamberland  filters  under  pressure.  They  concluded 
that  practically  all  soils  gave  essentially  identical  solutions  and  that 
even  where  only  a  small  quantity  of  one  constituent  was  present,  it  was 
sufficient  for  the  growth  of  a  fair  crop  if  the  mechanical  condition  of 
the  soil  was  good. 

It  was  concluded  also  that  the  soil  moisture  was  practically  a  saturated 
solution  of  the  mineral  substances  present  in  the  soil.  Consequently,  as 
fast  as  salts  were  removed  by  the  plant  further  quantities  were  quickly 
dissolved,  thus  keeping  the  solution  at  nearly  the  same  concentration 
throughout  the  growth  of  the  plant.  One  of  the  most  significant  facts 
claimed  to  have  been  shown  by  the  investigation  was  that  the  equilib- 


Feb.  II,  i9i8{  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  313 

rium  of  the  solution  quickly  readjusted  itself,  at  least  as  quickly  as  the 
plant  disturbed  it  by  withdrawing  nutrients. 

It  was  therefore  believed  that  the  controlling  factors  in  fertility  were 
moisture  and  the  physical  condition  of  the  soil,  and  not  fertilizers  or 
plant  nutrients. 

This  was  an  entirely  new  viewpoint  in  soil  investigations  and  has 
proved  extremely  stimulating  to  other  workers,  though  it  should  be 
stated  that  Cameron  {10-12)  has  somewhat  modified  his  conclusions  in 
his  later  writings. 

The  first  portion  of  the  work  of  King  {34),  which  had  been  carried  on 
concurrently  with  that  of  Whitney  and  Cameron  (<5i)  in  the  Bureau  of 
Soils,  appeared  a  year  later.  It  consisted  of  three  papers,  which  were 
published  privately  by  the  author,  while  the  remaining  three  were  issued 
from  the  Bureau  of  Soils  {35). 

This  work  constitutes  the  most  extensive  investigation  so  far  carried 
out  on  the  water  extracts  of  soils.  The  methods  employed  were  the 
same  as  those  used  in  the  previous  investigation  of  Whitney  and  Cam- 
eron {61).  The  analyses  in  the  preliminary  work  were  performed  on  the 
fresh  samples  of  field  soil,  but  later  in  the  investigation,  despite  the  large 
error  involved,  oven-dried  samples  were  employed. 

The  work  of  the  first  season  was  largely  of  a  preliminary  nature  and 
was  principally  carried  out  at  Goldsboro,  N.  C.  Additional  samples 
were  taken  in  Georgia,  Virginia,  Maryland,  New  Jersey,  Pennsylvania, 
and  Wisconsin.  In  the  second  season  the  work  was  more  intensive  and 
consisted  of  a  study  of  eight  soil  types  in  the  four  States  of  North  Caro- 
lina, Maryland,  Pennsylvania,  and  Wisconsin.  The  crops  used  were 
cotton,  peas,  beans,  corn,  and  oats. 

Applications  of  fertilizer  consisting  of  5,  10,  and  15  tons  of  manure 
and  300  pounds  of  guano  were  made  to  two  crops,  com  and  potatoes. 
Analyses  of  the  water  extracts  of  the  soils  from  these  varied  crops  were 
made  from  three  to  six  times  during  the  season,  as  well  as  numerous 
extracts  of  the  plant  saps. 

In  general,  the  results  and  conclusions  drawn  were  diametrically 
opposed  to  those  of  Whitney  and  Cameron.  Relatively  large  amounts 
of  nutrients  were  found  to  be  either  actually  in  solution  or  in  such  form 
that  they  entered  into  solution  when  diluted  with  distilled  water.  It 
was  also  shown  that  the  application  of  fertilizers  materially  increased 
the  amounts  of  salts  recovered  from  the  soils.  The  largest  amounts  of 
salts  were,  as  a  whole,  found  where  the  yields  were  largest,  and  the  same 
results  were  obtained  from  the  examination  of  the  plant  sap. 

The  influence  of  farm  manures  was  found  to  increase  not  only  the  crop 
but  also  the  amounts  of  soluble  salts  which  could  be  recovered  from  the 
soil. 

27809°— 18 2 


314  Journal  of  Agricultural  Research  voi.  xii.  no.  6 

At  this  same  period  Gola  {18-21)  published  the  first  of  his  ecological 
and  chemical  studies  on  the  relation  of  the  soil  solution  to  the  natural 
plant  environment.  He  attempted  to  obtain,  by  a  method  founded  on 
that  of  Schloesing  (46,47),  an  extract  which  would  be  similar  to  the 
actual  soil  solution.  The  soil  was  broken  up  so  that  it  would  pass 
through  a  2-mm.  sieve  and  was  placed  in  a  glass  cylinder,  25  cm.  high 
and  4.5  cm.  in  diameter.  A  gentle  rain  of  distilled  water  at  the  rate  of 
25  to  30  mm.  per  square  centimeter  per  24  hours,  was  then  allowed  to 
fall  upon  it.  After  a  period  of  time,  which  varied  with  the  soil,  drops  of 
solution  issued  from  the  lower  tubulure,  and  the  process  was  allowed  to 
continue  till  25  to  50  c.  c.  of  "pedolytic"  solution  had  been  collected. 
Gola  then  subjected  the  saturated  soil  to  pressure  and  collected  what 
be  called  the  "pedopiezic"  solution. 

The  total  solids,  and  in  some  cases  the  total  colloids,  were  determined 
in  these  solutions.  From  the  figures  so  obtained  Gola  divided  up  the 
possible  habitats  of  various  plants  into  some  32  groups  which  he  believed 
to  be  controlled  by  the  solution  naturally  occurring  in  the  soil.  In  gen- 
eral, he  concluded  that  the  relation  between  the  soil  and  the  organs  of 
absorption  of  the  plant  was  controlled  by  the  osmotic  pressure  of  the  soil 
solution.  High  concentrations,  and,  especially,  rapid  changes  in  the 
solution  were  likely  to  be  harmful  to  plants,  though  many  may  grow 
successfully  in  weaker  solutions  than  were  normal  for  them.  The  prin- 
cipal factor  determining  the  habitat  of  plants  was  the  concentration  of 
the  solution  and,  to  a  lesser  extent,  its  chemical  composition. 

Snyder  (55,  54),  at  Minnesota,  carried  on  a  short  series  of  investiga- 
tions in  which  he  studied  the  absorption  of  nutrients  from  soil  extracts 
which  were  added  to  sand  cultures,  believing  this  condition  comparable 
to  the  absorption  from  the  soil  water.  He  arrived  at  conclusions  op- 
posed to  those  of  Whitney  and  Cameron  (61)  and  also  criticised  their 
results  from  theoretical  considerations. 

Gedroitz  (16),  working  in  Russia,  announced  that  the  concentration  of 
the  soil  solution  fluctuated  so  rapidly  that  it  was  impossible  for  any 
water  extract  to  give  any  indication  of  the  character  of  a  soil. 

Mitscherlich  (42)  issued  an  extremely  valuable  contribution  to  the 
study  of  the  water  extract.  His  method  of  procedure  was  essentially 
different  from  those  previously  discussed.  He  employed  for  the  extrac- 
tion carbon-dioxid-saturated  water  at  30°  C,  the  maximum  temperature 
which  would  probably  be  encountered  in  the  soil.  The  proportions  of 
soil  and  water  varied  from  i  to  5  up  to  i  to  30.  At  least  two  dilutions 
were  used  in  each  study,  generally  i  to  10  and  i  to  25.  The  soil  and 
water  were  placed  in  a  thermostat,  with  a  stirring  apparatus  running 
into  the  center  of  the  flask.  Carbon  dioxid  was  passed  in  constantly 
and  the  extraction  allowed  to  continue  for  ii^  hours. 

From  the  data  obtained  by  these  varied  extractions  Mitscherlich 
graphically  estimated  the  amounts  of  additional  material  which  were 


Feb.  II,  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  315 

dissolved  by  the  water  extract  over  and  above  that  portion  which  was 
actually  in  solution.  This  was  a  very  valuable  differentiation.  His 
conclusion  that  these  graphs  could  be  directly  produced  and  would  fol- 
low the  concentration  of  the  actual  soil  solution  is  discussed  by  Hoag- 
land  (27)  in  connection  with  experimental  data  bearing  on  this  point. 

Mitscherlich  was  able  by  his  procedure  to  distinguish  between  fertil- 
ized and  unfertilized  soils  and  also  between  various  grades  of  fertilizer 
application.  An  excellent  feature  of  his  work  was  the  complete  and 
careful  estimation  of  the  factor  of  error  involved  for  all  determinations 
and  its  influence  on  the  final  result. 

Ishcherekov  (29)  published  the  first  of  numerous  attempts,  which  are 
still  continuing,  to  obtain  the  soil  solution  by  the  use  of  various  reagents. 
Briggs  (6)  and  Briggs  and  McCall  (7)  had  previously  obtained  small 
amounts  of  solution  from  the  soil  by  centrifugal  force  and  capillarity 
when  the  soil  contained  moisture  slightly  in  excess  of  the  optimum, 
Ishcherekov  now  attempted  to  obtain  the  solution  from  a  soil  which  was 
close  to  saturation  by  pouring  a  layer  of  0.5  cm.  of  ethyl  or  methyl 
alcohol  over  the  soil  and  assumed  that  the  first  portion  of  clear  solution 
which  ran  through  was  the  actual  soil  solutio^i. 

The  same  author  later  (50)  reported  a  series  of  studies  in  which  he 
used  the  methods  of  the  U.  S.  Bureau  of  Soils  and  from  which  he  drew 
conclusions  which  were  practically  in  entire  agreement  with  those  of 
Whitney  and  Cameron  {61). 

Engels  {15),  on  the  other  hand,  reported  a  series  of  studies  in  which 
he  used  distilled  water,  carbon-dioxid-saturated  distilled  water,  and  2 
per  cent  citric  acid  and  concluded  that  the  citric  acid  was  the  most 
satisfactory  reagent  to  estimate  the  soluble  material  in  soil. 

Maschhaupt  and  Sinnige  {41)  conducted  an  investigation  in  which 
carbon-dioxid-saturated  water  and  2  per  cent  citric  acid  were  employed 
and  concluded  that  carbon-dioxid-saturated  water  was  preferable. 

Lyon  and  Bizzell  (40)  have  attempted  to  estimate  the  density  of  the 
soil  solution  indirectly  by  determining  the  relation  of  the  dry  matter 
formed  to  the  transpiration  observed.  They  conclude  that  the  addition 
of  f  ertiUzer  caused  an  increase  in  the  density  of  the  solution  and  obtained 
confirmatory  evidence  by  measuring  the  density  of  the  soil  solution  with 
the  Wheatstone  bridge. 

Van  Suchtelen  {17,  59)  has  announced  a  modification  of  Ishcherekov's 
(29)  procedure  in  which  he  uses  paraffin  oil  instead  of  alcohol.  He 
claims  to  obtain  the  soil  solution  in  unaltered  form,  though  the  full 
details  of  his  method  have  not  yet  appeared.  As  the  first  announcement 
of  this  method  appeared  in  1912,  and  the  last  statement  of  its  pros- 
pective full  publication  was  made  by  Morgan  {43)  in  191 6,  it  is  to  be 
hoped  that  it  may  soon  appear  in  its  entirety. 


31 6  Journal  of  Agricultural  Research  voi.  xii.  no.  6 

Ballenegger  (2)  has  used  the  methods  of  the  Bureau  of  Soils,  as  well 
as  determinations  of  electrical  conductivity,  in  the  study  of  75  typical 
Hungarian  soils.  He  concluded  that  the  character  of  the  water  solu- 
tions may  be  used  to  differentiate  the  various  types  of  soil.  The  soils 
investigated  varied  from  the  poor  gray  forest  soils  to  the  very  fertile 
''at fold"  soils. 

Toulaikov  {56,  37)  believed,  like  Gola  (19),  that  the  osmotic  pressure 
of  the  soil  solution  was  the  important  factor  in  plant  growth.  He 
found  the  optimum  to  be  a  pressure  of  three  atmospheres  and  that  the 
growth  of  wheat  was  benefited  by  an  increase  up  to  that  point. 

Pantanelli  {44)  has  attempted  to  study  the  concentrations  of  the 
solutions  of  soils  from  Tripoli  by  determining  the  electrical  conductivity 
of  the  liquids  obtained  by  percolation.  He  was  able  by  this  procedure 
to  differentiate  between  several  classes  of  cultivated  and  virgin  soils. 

Hall,  Brenchley,  and  Underwood  (25)  have  reported  a  noteworthy 
investigation  in  which  the  soils  from  the  Rothamstead  experiment  plots 
were  used.  Solutions  were  prepared  using  20  kgm.  of  soil  and  35  kgm. 
of  water,  and  wheat  and  barley  plants  were  grown  therein.  From  the 
analysis  of  these  solutions  and  the  growth  of  the  plants  in  them  it  was 
concluded  that — 

The  composition  of  the  natural  soil  solution  as  regards  phosphoric  acid  and  potash 
is  not  constant,  but  varies  significantly  in  accord  with  the  composition  of  the  soil 
and  its  past  manurial  history.  Within  wide  limits  the  rate  of  growth  of  a  plant  varies 
with  the  concentration  of  the  nutritive  solution  irrespective  of  the  total  amoimt  of 
plant  food  available.  When  other  conditions  such  as  the  supply  of  nitrogen,  water 
and  air  are  equal,  the  growth  of  crops  will  be  determined  by  the  concentration  of  the 
soil  solution  in  phosphoric  acid  and  potash;  which,  in  its  turn,  is  determined  by  the 
amount  of  these  substances  in  the  soil,  their  state  of  combination  and  the  fertilizer 
applied. 

These  authors'  did  not  find  any  toxic  effect  on  soils  which  had  grown 
wheat  and  barley  for  even  60  years.  Growth  in  the  soil  solutions  agreed 
with  the  growth  in  culture  solutions  containing  equivalent  amounts  of 
phosphoric  acid  and  potash. 

In  a  series  of  nutritive  solutions  of  various  degrees  of  dilution  the 
growth  varied  directly,  but  not  proportionately,  with  the  concentration 
of  the  solution. 

Finally,  the  authors  concluded  that  the  results  of  the  investigation 
restored  the  earlier  theory  of  the  direct  nutrition  of  plants  by  means  of 
fertilizers  and  nullified  the  theories  advanced  by  Whitney  and  Cameron 

(.61). 

Bouyoucos  and  McCool  (5)  have  proposed  an  ingenious  method  for 
determining  the  concentration  of  the  soil  solution  directly  by  means  of 
the  freezing  point.  Use  has  been  made  of  this  procedure  in  the  present 
investigation  by  Hoagland,  and  its  application  and  limitation  are  dis- 
cussed in  a  separate  paper  (27). 


Feb.  II,  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  317 

Bogue  (j)  has  published  a  brief  investigation  in  which  he  leached 
four  soils  with  water  and  also  studied  their  absorptive  capacity  for 
potassium  and  phosphates.  From  his  work  he  agreed  with  Whitney 
and  Cameron  that  the  composition  of  the  soil  moisture  is  not  influenced 
by  the  chemical  composition  of  the  soil,  but  instead  is  dependent  on  the 
mechanical  texture  of  the  individual  soil. 

Jensen  {32)  reported  a  series  of  observations  on  eight  sugar-beet  plots 
at  Rocky  Ford,  Colo.  The  methods  of  investigation  employed  were 
those  of  the  U.  S.  Bureau  of  Soils  {50).  The  plots  receiving  composted 
manure  showed  nearly  twice  as  much  water-soluble  potash  in  the  surface 
foot  as  did  any  of  the  other  plots.  The  seasonal  averages  of  this  element 
were  not  appreciably  influenced  by  any  other  fertilizer  treatment.  There 
was  a  decrease  noted  in  the  quantity  of  water-soluble  potash  from  the 
middle  of  May  till  the  middle  of  July.  After  that  time  the  quantity 
increased  to  approximately  the  amount  that  had  been  found  earlier  in 
the  season.  None  of  the  treatments  resulted  in  a  marked  increase  of 
soluble  phosphates,  and  the  variation  in  this  compound  was  less  than 
in  any  other  element  measured. 

Jordan  {33)  has  recently  published  the  results  of  an  investigation  in 
which  nine  soils  were  analyzed  by  complete  analysis,  acid  extraction 
with  hydrochloric  acid  of  1.115  specific  gravity,  10  days'  leaching  with 
water,  N/200  and  A^/25  hydrochloric  acid,  and  five  hours'  extraction 
with  the  last  three  solutions.  Vegetation  experiments  were  conducted 
with  the  soils  in  the  greenhouse  during  two  years.  At  the  close  of  this 
time  it  was  impossible  to  establish  any  relationship  between  any  of  the 
elements  determined  and  the  crop-producing  capacity.  There  appeared 
to  be  a  slight  relationship  between  the  total  soluble  matter  in  the  soil 
and  crop  production,  but  it  was  not  consistent  in  all  cases.  The  final 
conclusion  was  that  no  method  had  been  developed  by  which  the  fer- 
tility of  a  soil  could  be  measured  through  laboratory  investigation. 

Harris  and  Butt  (24),  working  in  Utah,  have  studied  the  effect  of 
varying  amounts  of  irrigation  water  upon  the  development  of  nitrates 
and  soluble  salts  in  cropped  and  fallow  soils.  They  found  notable 
differences  between  the  cropped  soil  and  the  fallow  duplicate,  but  did 
not  find  that  these  differences  were  related  to  variation  in  the  crop  yield. 

SCOPE  OF  THE  INVESTIGATION 

The  foregoing  discussion  indicates  the  contradictory  nature  of  the 
results  already  obtained.  Two  of  these  moot  points  especially  open  to 
further  study  are : 

1.  The  relationship  between  the  soluble  soil  nutrients  in  cropped  and 
uncropped  soils. 

2.  The  relationship  between  the  soil  extract  and  the  crop  produced 
thereon. 

These  questions  are  the  subject  of  the  present  study. 


31 8  Journal  of  Agricultural  Research  voi.  xii,  No.  6 

METHOD  OF  ATTACK 

It  is  evident  that  in  any  soil  a  large  number  of  complex  factors  deter- 
mine production.  These  factors,  besides  influencing  the  final  crop 
obtained,  can  also  conceivably  modify  the  condition  of  the  plant  food 
in  the  soil  itself.  It  becomes  very  clear  then  that  a  chemical  study  of 
the  soil  solution  can  hope  to  obtain  a  fair  degree  of  success,  only  if  a 
number  of  these  modifying  factors  are  subordinated  or  made  comparable. 

Prominent  among  such  factors  are  the  influence  of  physical  texture, 
climate,  and  moisture.  To  reduce  the  effect  of  physical  texture,  two 
types  of  soils  were  chosen:  silty  clay  loams  and  fine  sandy  loams.  The 
representatives  of  each  class  were  selected  with  as  uniform  a  physical 
texture  as  possible.  Comparisons  can  therefore  be  made  between  the 
soils  within  each  class,  with  the  assurance  that  the  physical  factor  is 
reduced  to  the  minimum. 

The  effect  of  climate  was  made  uniform  by  transporting  exceptionally 
large  samples  of  surface  soil  to  Berkeley.  The  samples  were  then 
sifted  and  placed  in  containers  which  will  be  described  later.  Moisture 
conditions  were  made  as  uniform  as  possible  by  adding  just  sufficient 
distilled  water  to  keep  the  soils  at  their  optimum  content. 

SELECTION  OF  THE  SOILS 

The  soils  used  in  the  investigation  were  chosen  according  to  the  map- 
ping of  the  Bureau  of  Soils  of  the  United  States  Department  of  Agri- 
culture. The  silty  clay  loams  were  all  chosen  from  the  Yolo  series. 
Of  these,  three  were  taken  from  the  Sacramento  Valley  and  three  from 
the  Santa  Clara  Valley.  The  past  history  of  all  these  soils  was  very 
different,  as  shown  in  Table  I.  The  crops  grown  on  them  were  equally 
divided  between  orchard  crops  (prune,  almond,  peach)  and  field  crops. 
None  of  the  orchards  had  received  any  special  treatment.  Among  the 
field  soils,  the  sample  from  the  University  Farm  at  Davis  was  the  only 
one  which  had  been  manured,  though  its  cropping  was  no  more  varied 
than  No.  4  from  the  Santa  Clara  Valley.  Soil  3  from  Yolo  had  a  less 
varied  history  than  any  other  in  either  set.  This  soil  had  been  under 
cultivation  since  about  i860,  and,  except  for  two  years  in  sugar  beets, 
191 1  and  1912,  it  had  been  steadily  cropped  with  wheat  or  barley  each 
year.  Its  past  production  was  stated  to  be  good,  fully  the  average  for 
that  section,  though  no  exact  record  had  been  kept  of  the  yields. 


Feb.  II.  1918  Effect  of  Season  a^id  Crop  Growth  on  Soil  Extract  319 


Table  I. — Classification  and  history  of  the  soils  used  in  this  investigation 


Soil 
No. 

Soil  series  and  type. 

Origin. 

Crop  grown. 

Past  treatment. 

I 

Yolo    silty    clay 

Sacramento  Val- 

Field crops. . . 

Early  planting  of  grain; 

loam. 

ley  (University 
Farm,  Davis). 

1909-1911,  barley; 
1912,  fallow  and  ma- 
nure; 19 13-14,  barley. 

do 

Sacramento  \  al- 

Almond   or- 

Formerly grain;  almond 
orchard  12  years  old. 

ley  (Yolo). 

chard. 

3 

do 

do 

Barley 

Planted  about  i860; 
since  then  barley  and 

wheat,    except    sugar 

beets  in  1911-12. 

4 

do 

Santa  Clara  Val- 

Field crops . . 

Originally  grain;  later 
orchard;  several  years 

ley  (San  Jose). 

alfalfa;   three  years 

field  crops. 

5 

do 

do 

Prunes 

Originally  grain;  prime 

orchard  about  20  years 

old. 

6 

Yolo  clay  loam . . . 

Santa  Clara  Val- 
ley (Lawrence). 

Peaches 

Originally  grain;  peaches 
for  8  years;  heavy 
crop,  about  12  tons  per 
acre. 

7 

H  a  n  f  0  r  d   fine 

Southern       Cali- 

Oats  

Originally  grain;   about 

sandy  loam. 

fornia  (Arling- 
ton). 

1890  put  into  alfalfa 
for  13  years;  potatoes 
2  years,  alfalfa  4  years, 
oats  5  years;  yield  of 
oats,  4  tons  of  hay  per 
acre. 

8 

Fresno     fine 

San  Joaquin  Val- 

Seedless 

Originally    grain;     14 

ley  (Fresno). 

grapes. 

years  in  Sultanina 
(Thompson  Seedless) 
grapes.  Production 
about  2  tons  of  raisins 
per    acre    for    last    6 

Q 

Kimball     fine 

Southern  Caiifor- 

Navel   or- 

25 years  in  oranges;  pre- 

sandy loam. 

n  i  a     (Re  d  - 

lands). 

anges. 

viously  bare  land.  A 
great  variety  of  fertili- 
zers had  been  used. 

10 

T  e  j  u  n  g  a    fine 

Southern  Califor- 

Peaches  

Originally    15    years   in 

sandy  loam. 

nia   (San   Fer- 
nando Valley). 

prunes;  now  10  years 
i  n  peaches;  small 
amount  or  mantue  the 
only  treatment. 

II 

Madera   fine 

San  Joaquin  Val- 

Navel or- 

Orange   trees    about   15 

sandy  loam. 

ley     (Kearney 
Park). 

anges. 

years  old. 

12 

Arnold    fine 

San  Joaquin  Val- 

Oats  

In  cultivation  about  40 

sandy  loam. 

ley  (foothills). 

years;  early  crops 
largely  wheat;  last  foiu- 
or  five  years  biennial 
crops  of  oats;  alternate 
year  summer  fallow. 

13 

Unnamed  fine 

Mendocino  Coim- 

\  irgin 

Very  shallow  soil,  about 

sandy  loam. 

ty  coast. 

I  foot  in  depth  under- 
lain by  clay  subsoil. 

14 

St an dish  fine 

Honey  Lake  re- 

 do 

Desert  soil,  small  shrubs 

sandy  loam. 

gion. 

and  weeds,  natural 
growth. 

320  Journal  of  Agricultural  Research  voi.  xii.  no.  6 

The  fine  sandy  loams  had  an  even  more  varied  past  history.  Three 
of  these  soils  were  from  southern  California.  One  of  these  was  from  a 
Redlands  orange  grove  which  had  been  treated  in  the  past  with  a  great 
variety  of  fertilizers.  A  second  was  from  a  peach  orchard  of  the  San 
Fernando  Valley,  and  the  third  from  the  Riverside  area  had  been  de- 
voted to  field  crops. 

Three  fine  sandy  loams  were  obtained  in  the  San  Joaquin  Valley.  Two 
from  the  vicinity  of  Kearney  Park,  in  oranges  and  Sultanina  (Thomp- 
son Seedless)  grapes,  had  an  excellent  record  for  past  production. 
The  third  from  the  foothills  near  Oakdale  was  purposely  chosen  from  a 
body  of  Arnold  fine  sandy  loam  where  the  growth  of  oats  was  unusually 
small,  though  the  soil  was  normal  in  depth  and  drainage.  The  last  two  fine 
sandy  loams  were  virgin  soils.  One  of  these  (No.  13)  came  from  the  coast 
of  Mendocino  County,  and  was  an  extremely  poor,  shallow  soil.  In 
fact,  it  later  developed  that  this  soil  was  very  acid;  therefore  no  de- 
ductions have  been  drawn  from  the  results  obtained  with  it.  The  other 
virgin  sample  was  obtained  from  the  Honey  Lake  area.  This  soil  was 
considered  to  be  rather  in  the  class  of  desert  soil,  but  was  not  especially 
deficient  in  organic  matter. 

The  soils  of  both  types  were  selected  to  include  as  many  past  treat- 
ments and  crops  as  possible.  They  had,  of  course,  been  exposed  to  very 
different  climatic  influences.  This  is  an  unavoidable  factor  in  Cali- 
fornia, with  its  great  diversity  of  local  conditions.  It  is  believed  that 
these  various  influences  have  been  largely  neutralized  by  the  cropping 
of  the  first  season,  191 5,  under  uniform  conditions.  Greater  emphasis 
is  therefore  attached  to  the  results  of  the  past  year  (1916). 

The  attempt  was  also  made  to  choose  soils  which,  though  of  uniform 
physical  texture,  would  have  a  very  different  crop-producing  power. 
In  this  way  it  was  hoped  that  some  light  might  be  thrown  on  the  relation 
between  yield  and  water-soluble  nutrients. 

In  Table  II  are  given  the  moisture  equivalents,  hygroscopic  coeffi- 
cients, specific  heat,  and  mechanical  analyses  of  the  14  soils.  Moisture 
equivalents  are  obtained  by  the  method  of  Briggs  and  Shantz  (<?); 
hygroscopic  coefficients  by  the  method  of  Hilgard  (26).  Specific  heat 
was  determined  by  heating  20  gm.  of  soil  in  a  tin-foil  container  until  it 
reached  a  temperature  of  approximately  100°  C.  It  was  then  lowered 
into  an  insulated  tank  of  water  of  known  temperature  and  volume.  The 
amount  of  heat  contained  in  the  soil  was  measured  by  the  increase  of 
temperature  of  the  water.  The  mechanical  analyses  were  performed 
by  the  method  of  the  Bureau  of  Soils  (9).  It  should  be  noted  that  the 
analyses  as  reported  were  performed  on  the  same  samples  which  were 
used  for  the  chemical  analyses.  These  samples  had  been  passed  through 
a  i-mm.  sieve.  This  was  done  in  order  that  the  chemical  and  physical 
analyses  should  be  made  on  identical  portions.  The  percentage  of  gravel 
by  this  screening  process  is  also  stated,  so  that  the  mechanical  analyses 


Feb.  II,  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  321 


may,  if  desired,  be  recalculated  to  the  usual  form.  A  study  of  these 
figures  shows  that  the  members  of  each  group  of  soils  in  physical  texture 
are  as  similar  as  was  desired.  From  these  analyses  Mr.  C.  F.  Shaw,  of 
this  Station,  states  that  the  silty  clay  loams  would  be  accurately  classed 
as — 

clay  loams  with  a  very  high  percentage  of  fine  sand  and  silt,  and  that  with  the  excep- 
tion of  No.  6  they  would  function  as  very  tmiform  silty  clay  loams.  No.  6  would  be 
classed  as  a  clay. 

The  fine  sandy  loams  are  considered  to  be  quite  uniform  in  texture,  so 
far  as  porosity,  ease  of  root  penetration,  and  behavior  under  cultivation 
are  concerned.  The  moisture  equivalents  are  very  different,  so  that  at 
first  glance  it  would  appear  that  the  plants  would  be  unequally  sup- 
plied, especially  in  the  fine  sandy  loams.  It  will  be  pointed  out  later, 
however,  that  this  view  does  not  consider  the  fixed  or  "unfree"  water 
of  the  soil.  The  amount  of  moisture,  therefore,  which  is  available  to  the 
plants,  as  shown  by  the  wilting  coefficient,  and  also  the  amount  which 
affects  the  concentration  of  the  soil  solution,  can  not  be  judged  from 
the  moisture  equivalent  alone. 

Table  II. — Moisture  equivalents,  hygroscopic  coefficients,  specific  heat,  and  mechanical 
analysis  of  experimental  soils 

MOISTURE  EQUIVALENTS 


Test  No. 

Soil 

I 

2 

3 

4 

5 

6 

7 

8 

9 

10 

II 

12 

13 

14 

Test  I 

Test  2 

Test3 

Test  4 

27-51 
27-27 
27-08 

27-  21 

25-23 

25-06 
24-93 
24-77 

27.44 

27. 10 
27-  26 

27.69 

23-62 
23-95 
24-53 
24.01 

24. 12 
23.29 
24-38 
23-75 

30.57 
30.09 
30.24 
30.22 

13-61 
13-45 
13.06 
13-10 

7-77 
7-94 
8.24 
8.30 

9-95 
9.27 
9-73 

10-53 

18.41 
17-98 
18.76 

16.  29 
16.00 
16.61 
16.  24 

9.  IS 

10.00 

9-72 
10.01 

18-  57 
19.02 
18.22 
18.43 

17-84 
17-45 
17-99 

Average... 

27.28 

25.00 

27.37 

24-03 

23.88 

30.28 

13-31 

8-06 

9-87 

18.38 

16.28 

9-72 

18.56 

17.99 

HYGROSCOPIC  COEFFiaENTS 

Test  I 

Test  2 

7-9S 
7.92 

6.09 

6.36 

8-33 
8.08 

S-49 
S-77 

8.11 
8.14 

8.82 
8.84 

2.83 
2-94 

I- IS 
I.  20 

1-57 
1-49 

2.93 
2-95 

3-14 
3-18 

1.85 

1-68 

1-31 

1-3° 

4-72 
4-76 

Average. . . 

7-94 

6.23 

8.20 

5-63 

8.12 

8.83 

2.89 

1.17 

1-53 

2-94 

3.16 

1.77 

1.31 

4-74 

SPECIFIC  HEAT 

Test  I 

Test  2 

0-2122 

.2095 

0-2057 

.  2042 

0-193 
.  200 

O-1901 

.2008 

0-  2246 

.2245 

0.243 
.242 

0.1969 

.1876 

0.1758 
.1661 

0.230 
.234 

0- 184 

.188 

0. 2480 
.  2501 

0- 1807 
•  1656 

0- 1933 

.i860 

0-224 

•  230 

Average... 

.211 

.205 

.1965 

•1954 

.2245 

.2425 

.1918 

.1709 

.232 

.186 

.2490 

•1731 

.1897 

.227 

322 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  6 


Table   II. — Moisture  equivalents,  hygroscopic  coefficients,  specific  heat,  and  mechanical 
analysis  of  experimental  soils — Continued 


MECHANICAIv  ANALYSES 

[Averages  of  two  analyses) 


Kind  of  soil. 

Soil 

I 

2 

3 

4 

5 

6 

7 

8 

9 

10 

II 

12 

13 

14 

Fine  gravel 

( 1-2  mm.)- 

o.oo 

0.00 

o-oo 

0.00 

O-OO 

0.00 

0.04 

o-oo 

0.016 

o-oo 

0.00 

0.00 

0.04 

0.00 

Coarse  sand 

(i-i-snun.) 

•  15 

.60 

.60 

2.09 

1.84 

•746 

2.82 

1-95 

9-986 

1-57 

2-77 

4.99 

7-77 

3.09 

Medium 

sand  (0.5- 

0.25mm.) . 
Fine     sand 

•23 

1-50 

1.04 

1.24 

1.70 

-574 

2.87 

3-19 

6.308 

2-94 

2.818 

4-38 

J:4^55 

4^05 

(0.25-0.010 

mm.) 

1.69 

8.00 

3-53 

5-47 

9.61 

1.786 

15.60 

21.  22 

21-730 

16.62 

20.945 

23.27 

13-49 

24.54 

Very  fine 

sand  (o.oi- 

0.05 mm.) . 

32-62 

31-  26 

26.28 

30.32 

32-73 

21-775 

53-67 

50-63 

43-67 

38.76 

39.34 

46.24 

IS- 70 

38.94 

Silt     (0.0s- 

0.00s  mm.) 

42.68 

36-97 

38.12 

37-99 

30.52 

37-415 

16.01 

17-57 

10.87 

26.25 

19-075 

9.61 

34-72 

15.55 

Clay  (0.005- 

omm.) 

24.17 

22.47 

28.47 

25.08 

24.72 

39-380 

9-77 

S-S6 

8.19 

13-37 

14.62 

9.44 

12.9s 

13.87 

Sum  of 

percent- 

ages  

101.54 

100.80 

98.68 

102- 19 

101. 12 

101.676 

100.58 

loo- 12 

100.77 

99-51 

99. 568 

97-93 

99.22 

100.04 

Gravel  r  e  - 

moved  by 

previous 

screening. . 

.  II 

.09 

•39 

6- IS 

3-38 

.11 

3-64 

•99 

6.41 

-59 

2. 12 

1-57 

I- IS 

2.  21 

In  choosing  the  soils  in  the  field  the  following  plan  was  adopted. 
One  soil  of  each  type  was  taken  as  the  typical  sample.  Then  each  suc- 
ceeding member  of  that  group  was  chosen  by  comparison  with  a  small 
sample  of  this  type.  This  same  small  sample  was  carried  while  all  the 
work  was  being  done  on  that  group,  and  constant  comparisons  were 
made  when  there  was  ground  for  doubt.  In  this  way  it  was  believed 
that  each  member  of  a  group  would  approach  very  closely  to  the  other 
members  in  general  physical  texture. 

The  samples  in  each  case  were  taken  from  the  top  foot  of  soil.  Pre- 
liminary borings  with  a  soil  auger  were  made  in  order  to  ascertain  whether 
the  subsoil  was  free  from  hardpan  or  other  abnormal  factors  which  might 
influence  the  surface  soil.  The  amount  of  each  of  these  samples  was 
2>^  tons.  The  soil  was  shipped  in  new,  heavy  burlap  sacks  holding  ap- 
proximately 100  pounds  each.  Every  attempt  was  made  to  prevent 
undue  drying  or  exposure  to  sunshine  after  the  soil  was  sacked. 

Upon  the  arrival  of  the  samples  in  Berkeley  they  were  immediately 
passed  through  X-iiich  concrete  sieves  to  obtain  a  uniform  physical  tex- 
ture. This  work  was  performed  with  all  possible  expedition,  and  it  is 
beUeved  that  they  had  a  normal  bacterial  flora  when  placed  in  the 
containers.  None  of  the  soils,  except  No.  14,  was  actually  air-dry  when 
finally  prepared  for  use.  The  ammonifying  and  nitrifying  powers  of  the 
soils  have  been  determined  as  a  test  of  normal  biological  activity.  This 
work  was  performed  at  the  beginning  and  close  of  the  season  of  191 5 
and  at  the  end  of  the  growing  season  in  191 6.    The  tests  were  carried  out 


Feb.  n.  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  323 

by  the  well-known  tumbler  method.    All  the  soils  showed  a  satisfactory 
capacity  for  ammonification. 

At  the  close  of  the  season  of  191 5,  soils  12  and  14  were  noticeably 
lower  in  nitrifying  power  than  the  others  of  the  group.  At  the  close  of 
1 91 6,  soil  12  alone  showed  a  significant  lower  range  in  nitrification.  With 
these  exceptions  the  soils  gave  results  which  indicated  normally  active 
bacterial  floras.  The  detailed  data  are  not  considered  sufficiently  il- 
luminating to  justify  their  inclusion  in  this  paper. 

SOIL  CONTAINERS  AND  INSTALLATION 

The  soil  containers  selected  were  30  inches  wide,  "60  inches  long,  and 
18  inches  deep.  They  were  made  of  No.  24  galvanized  iron  and  were 
thoroughly  coated  with  asphaltum  varnish.  When  filled,  each  con- 
tained approximately  i  ,700  pounds  of  soil.    The  design  of  the  containers 


I  X  /4-    iron  s>{rao  -y 


Perforated    '/z'  pipe    for    sub  irrigation, 
hofe-i  2"  ap>ort. 

Fig.  1. ^Design  of  soil  containers. 

is  shown  in  figure  i .  It  will  be  observed  than  an  outlet  is  placed  in  the 
bottom  at  one  end.  This  outlet  not  only  prevented  the  accumulation 
of  excess  water,  but  also  gave  additional  aeration  to  the  lower  soil. 
The  escape  pipes  may  possibly  be  seen  in  the  illustration  of  the  wire 
house  inclosing  the  containers  (PI.  14,  A). 

Distilled  water  was  the  only  moisture  used  at  any  time.  This  was 
added  partly  between  the  rows  of  grain  by  a  long-spouted  sprinkling 
can  and  partly  by  subirrigation  through  the  two  perforated  pipes  running 
from  end  to  end  of  the  containers.  So  far  as  possible,  only  enough  water 
was  added  at  one  application  to  keep  the  soil  at  optimum  moisture.  It 
was  found  possible  to  observe  the  moisture  condition  quite  closely  by 
drawing  out  a  core  of  soil  with  an  i8-by-^-inch  cheese  trier,  and  then 
replacing  the  soil. 

The  containers  numbered  28,  2  for  each  soil,  and  were  arranged  on 
level  mud  sills  in  two  duplicate  sets,  as  indicated  in  the  accompanying 
diagram    (fig.    2).     By  the  above  arrangement  the   smallest  possible 


324 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  6 


external  surface  of  each  container  was  exposed  to  any  temperature 
change.  These  changes  were  further  minimized  by  surrounding  the  set 
of  containers  with  a  boxing,  i8  inches  high,  placed  6  inches  from  the 
outside  of  the  entire  group.  This  6-inch  space  was  then  filled  compactly 
with  local  soil.    The  insulation  furnished  by  this  arrangement  was  excel- 


/&-&■ 


/©•-6" 

Fig.  2. — Diagram  of  the  arrangement  of  the  soil  containers. 

lent.    No  difference  in  growth  was  observed  in  any  portion  of  the  con- 
tainers. 

Protection  from  birds  was  assured  by  a  framework,  4^  feet  in  height 
above  the  containers,  covered  with  i -inch-mesh  wire  netting.  The  general 
arrangement  of  the  two  sets  is  shown  in  Plate  14,  A. 


Feb.  II.  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  325 

During  the  growing  season  selected  for  these  experiments,  from  May 
to  September  (the  first  season,  however,  being  from  June  to  October), 
very  little  rain  falls  in  Berkeley.  To  protect  the  soils  from  flooding  by 
any  heavy  showers  that  might  occur,  waterproof  canvas  covers  have 
been  provided  and  have  been  put  over  the  wire  houses  in  a  few 
emergencies.  During  the  season  of  heaviest  rains  these  covers  are  kept 
on,  and  the  containers  are  never  flooded  or  subjected  to  leaching  by 
rain.  In  this  season  between  crops  the  soils  are  watered  with  dis- 
tilled water  at  intervals,  so  that  they  are  close  to  the  optimum  at  all 
times.  It  is,  of  course,  true  that  this  is  not  comparable  to  the  seasonal 
changes  to  which  field  soils  are  subjected.  On  the  other  hand,  such 
seasonal  changes  of  rainfall  and  drouth  are  never  comparable  for  any  two 
seasons  or  any  two  places.  Many  orchard  soils  in  this  State,  by  con- 
stant cultivation  and  irrigation  in  the  summer,  retain  a  very  regular 
moisture  supply  in  the  soil.  The  moisture  conditions  of  this  experiment 
are,  in  fact,  very  similar  to  those  found  in  such  cases.  The  orchard,  in 
addition,  is  subjected  to  occasional  large  excesses  of  water  from  rainfall. 

After  filling  the  containers  with  the  prepared  soils,  the  excess  portions 
of  sifted  soil  were  stored  in  the  set  of  tightly  covered  bins  shown  in 
Plate  14,  B.  This  soil  was  available  for  supplying  the  small  portions 
removed  by  sampling  and  also  furnished  a  stock  for  additional  studies. 

ANALYTICAL  PROCEDURE 

In  the  season  of  191 5  the  analytical  methods  used  were  those  outlined 
in  United  States  Department  of  Agriculture  Bureau  of  Soils  Bulletin  31 
{50).  With  pure  solutions  containing  only  a  single  compound  the  de- 
terminations obtained  by  these  methods  are  very  satisfactory.  Before 
the  close  of  the  season  considerable  doubt  was  felt  in  regard  to  certain 
of  the  results  obtained.  This  applied  especially  to  calcium,  potash,  and 
phosphate.  Before  entering  upon  the  work  of  the  season  of  191 6  a  series 
of  studies  were  outlined  to  test  the  accuracy  of  the  methods  which  had 
been  in  use.  The  preHminary  work  previously  performed  by  the  pro- 
cedures in  Bulletin  31  had  all  been  carried  out  with  solutions  which  con- 
tained only  one  pure  salt  of  the  radical  to  be  determined.  Under  these 
conditions  the  results  were  extremely  accurate. 

In  the  studies  now  undertaken  three  sets  of  solutions  were  prepared 
to  simulate  as  closely  as  possible  the  highest,  average,  and  lowest  con- 
centrations ordinarily  obtained  in  soil  extracts.  The  salts  used  were 
monocalcium  phosphate,  calcium  nitrate,  magnesium  sulphate,  potas- 
sium chlorid,  and  ammonium  chlorid.  The  solutions  were  made  up  with 
distilled  water  which  had  stood  in  ordinary  glass  bottles  and  so  con- 
tained small  amounts  of  sodium  and  silica,  though  not  of  course  com- 
parable to  the  amounts  which  occur  in  water  extracts  from  soils. 


326  Journal  of  Agricultural  Research  voi.  xii,  No.  6 

Quadruplicate  portions  of  these  solutions  were  analyzed  without  further 
treatment.  Large  aliquots  from  the  same  stock  bottles  were  also  passed 
through  new  and  old  Pasteur-Chamberland  filters,  and  some  portions 
were  treated  with  G.  Elf  carbon  black.  These  treatments  were  included 
to  check  the  possible  effect  of  absorption  of  elements  by  the  filter  candles, 
as  well  as  the  effect  of  the  carbon  black  which  was  used  to  decolorize 
soil  extracts  containing  organic  matter. 

Table  III  gives  the  results  of  the  analyses  of  the  untreated  and  of  the 
filtered  solutions.  It  will  be  seen  that  the  determinations  of  ammonia 
and  nitrate  are  satisfactory  throughout.  The  methods  used  are  essentially 
those  almost  universally  applied  in  sanitary  water  analysis  (z)  and  with 
the  proper  precautions  are  not  believed  to  be  open  to  criticism.  The  re- 
sults for  potassium,  calcium,  sulphate,  and  phosphate  are  seen  to  be 
extremely  inaccurate  in  the  lowest  concentration.  Of  these  the  calcium 
is  by  far  the  most  questionable.  The  error  in  the  determination  of  this 
element  is  so  large  that  the  results  for  calcium  obtained  in  191 5  have  con- 
sequently been  discarded.  Even  with  the  potash  and  phosphate  the 
percentage  of  error  is  very  large  in  some  determinations.  In  the  higher 
range  of  concentrations  the  results  are  much  closer  throughout,  and  if  it 
were  possible  to  confine  all  the  work  done  to  such  solutions  there  would 
perhaps  be  no  objection  to  the  methods  employed.  In  many  solutions 
smaller  concentrations  of  one  or  two  elements  are  frequently  encountered. 
In  such  a  case  the  error  may  be  50  per  cent  or  more,  either  plus  or  minus. 
Such  a  variation  can  readily  obscure  any  significant  change  in  concentra- 
tion. It  became  necessary  to  obtain  a  more  reliable  procedure  for  potas- 
sium, calcium,  and  phosphate,  all  of  which  are  extremely  important 
elements  in  any  study  of  soluble  plant  nutrients. 


Feb.  II.  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  327 


Table  III. — Analysis  {in  parts  per  million)  of  dilute,  average,  and  strong  solutions 
by  the  methods  outlined  in  United  States  Department  of  Agriculture  Bureau  of  Soils 
Bulletin  ji 


Ca 

Mg 

SO4 

P04 

NH3 

NO3. 

K. 

(turbidi- 

(colori- 

(turbidi- 

(colori- 

metric). 

metric). 

metric). 

metric). 

Solution  and  treatment. 

.a 

■d 
d 

•d 

.a 

•a 

.a 

•d 
01 

.a 

-s 

•a 
u 
c 

a 

1 

>> 

g 

>. 

a 

H 

& 

a 

v.. 

1 

0 

0 

0 

01 

0 

0) 

0 

0> 

0 

0> 

0 

0> 

0 

(D 

s 

&> 

01 

0> 

01 

-l-l 

0> 

-*-> 

0 

ja 

(U 

ja 

a 

ja 

t< 

J3 

V 

.a 

V 

j: 

01 

j: 

Oi 

H 

0 

H 

0 

H 

Q 

H 

Q 

H 

0 

iri 

Q 

H 

Q 

fo.  50 

0.48 

4.00 

4.08 

I.  2 

S-72 

i-S 

1.76 

0.50 

z. 

1.99 

2.72 

I.O 

2-5 

Dilute,  untreated 

1  .50 
.50 

.48 
.48 

4.00 
4.00 

4.08 
4.00 

I.  2 
I.  2 

1.92 
1.92 

S 

s 

1.68 
3- 80 

•so 
•SO 

.46 

•49 

1.99 
1.99 

2.68 
3^6o 

I.O 

1.0 

2.4 

1-6 

•50 

.48 

4.00 

4. 16 

1.2 

1.80 

s 

2.08 

•50 

•44 

1.99 

2.60 

I.O 

1.9 

f  -50 

•45 

4.00 

4. 16 

I.  2 

3-4° 

5 

.48 

•SO 

•44 

1.99 

2.68 

I.O 

1.6 

Dilute,  filtered  through  Pasteur 

1  .50 

•45 

4.00 

4.08 

1.2 

3^6o 

S 

1.84 

•SO 

•44 

1.99 

2.  60 

I.O 

!•? 

filters 

•  50 

,40 

4. 00 

4.08 
4.08 

1. 2 

3- 20 
3-12 

5 
5 

1.24 
1.60 

•SO 
•SO 

.46 

•44 

1. 99 

2.  76 

I.O 

1.9 

1-7 

•50 

•4S 

4.00 

I.  2 

1.99 

2.72 

I-O 

Dilute,  treated  with  carbon  black 

2.88 

and    filtered    through    Pasteur 

.  -5° 

•43 

4.00 

3-92 

I.  2 

2. 12 

5 

1.44 

.50 

•56 

1.99 

I.O 

!•? 

filters 

•SO 
/.50 
I. SO 
fi-SO 

.40 
.42 

4.00 
4.00 

3-92 
3^68 

I.  2 
I.  2 

2.72 
3^24 

S 
5 

1.28 
1.92 

•SO 
•SO 

•S6 
•54 

1.99 
1.99 

2.80 
2.60 

I.O 
I.O 

1.6 

Dilute,  passed  through  new  Pas- 

^•S 

teur  candles 

.40 
1.66 

4.00 
20.00 

3-84 
19-2 

1. 2 

2.64 
4- 30 

5 
S 

I.  08 

•SO 
2.50 

•49 
2.36 

!•  99 

3-68 

I.  0 

1-7 
6.0 

6.00 

8.00 

9.8 

9.00 

5-0 

Average,     untreated     solu- 

J I- SO 

I.  58I20. 00 

19.6 

6.  GO 

S^io 

5 

7-70 

2.  50  2.41 

9.8 

8.8 

5.0 

5^2 

tion  

|l-SO 

I.  SO 

I.  58120.00 

20. 0 

6.  00 

S-88 
6.  24 

5 
5 

7.502.50 
7.902.50 

2.30 
2.36 

9.8 

10. 0 

S-o 
S-o 

S-2 

5-0 

1.58 

20.00 

20.0 

6.00 

9.8 

8.8 

1.50 

1. 61 

20.00 

19.2 

6.00 

6.  64 

5 

7.902.50 

2.66 

9.8 

9.6 

S-o 

4.8 

Average    solution     filtered 

J  I- so 

1. 61 

20.00 

18.8 

6.00 

8.69 

_ 

S 

8.00^2.50 

2.61 

9.8 

9.2 

5.0 

4.0 

through  Pasteur  filters .... 

1.50 

1. 61 

20.00 

19.2 

6.00 

8.40J  7 

S 

7.60|2.so 

2.61 

9^8 

8.8 

S-o 

4.6 

I.  SO 

t.64 

20.00 

18.8 

6.00 

8.00 

7 

S 

7.  70  2. 50 

2.66 

9^8 

9.2 

S-o 

S-o 

Average    solution     treated 

with    carbon    black    and 

,i-SO 

1.64 

20.00 

18.6 

6.00 

6.00 

7 

5 

8.00  2.50 

2.61 

9.8 

13-4 

S-o 

4-3 

filtered   through   Pasteur 
filters 

1.50 

I.  64 

20.00 

18.6 

6.00 

6.00 

7 

5 

8.  20  2.  50 

2-43 

9^8 

II. 8 

5-0 

4.6 

Average     solution     passed 

through  new  Pasteur  fil- 

.1.50 

1.64 

20.00 

ig.  2 

6.00 

6.96 

7 

S 

8.302.50 

2.30 

9.8 

II. 8 

S-o 

4.8 

ters  

I- SO 
(3.00 

1.64 
3-07 

20.00 
38.00 

18.0 
37-6 

6.00 
12.00 

6.56 

7 

S 
0 

7.802.50 
16. 60,5.00 

2.41 
5-31 

9-8 

19-7 

II. 8 

20.0 

5-0 
10. 0 

4-7 

10.66  15 

10.4 

Strong  solution,  untreated 

I3.00I2.94 

38.00 
38.00 

37^6 
38.4 

12.00 
12.00 

10. 66  15 
13- 12,15 

0 
0 

16.40  5.00 
16. 40  5.00 

4.80 
4-74 

19.7 
19.7 

19.0 

18.2 

10. 0 
10. 0 

10.8 

13.00 

2.94 

10.4 

3-0O 

2.94 

38.00 

38.4 

12.00 

1 1^76  15 

0 

i7.oo'5.oo 

4-93 

19.7 

19.0 

10. 0 

10.2 

[3.00 

2.86 

38.00 

38.4 

12. 00 

12.32  15 

0 

16.  60  5.00 

4-93 

19-7 

19- S 

10. 0 

9.8 

Strong  solution   passed   through 

3-00 

3-07 

38.00 

37-6 

12.00 

12.49  15 

0 

16.  80  5.00 

4.80 

19-7 

18.2 

10. 0 

9-2 

Pasteur  candles 

3- 00 

2.90 

38.00 

37-6 

12.00 

I3^s6i5 

0 

16.  40  5.00 

4-86 

19.7 

18.  s 

10. 0 

9.2 

3- 00 

2.94 

38.00 

38.4 

12.00 

li.6o  IS'O 

16.  oo's- 00 

4^74 

19-7 

18.7 

lO.O 

9-4 

The  determination  of  magnesium,  though  it  is  the  same  procedure  as 
used  for  phosphate,  has,  at  least  in  this  series,  given  more  accurate 
results.  The  same  standard  solution  is  used  in  this  method  as  in  the 
phosphate  determination,  and  has  a  value  approximately  one-fourth  as 
great  when  expressed  in  terms  of  magnesium,  a  fact  which  would  decidedly 
tend  to  reduce  the  percentage  of  error  in  the  final  result.  The  same 
colorimetric  procedure  for  magnesium  is  still  in  use,  as  no  more  satis- 
factory method  for  small  amounts  of  this  element  has  so  far  been 
developed.  The  sulphate  determination  is  seen  to  be  more  accurate 
than  that  of  calcium,  but  the  error  in  low  concentrations  is  from  35  to 
85  per  cent.  This  procedure  has  therefore  been  dropped  and  the  sulphate 
determinations  are  not  included  in  the  191 5  charts. 

In  planning  the  work  for  1 916  it  was  evident  that  a  smaller  number  of 
more  accurate  determinations  were  to  be  preferred  to  a  larger  number 
on  which  little  reliance  could  be  placed.     With  this  idea  in  mind,  work 


328  Journal  of  Agricultural  Research  voi.  xii.  no.  6 

was  undertaken  in  which  larger  aliquots  than  those  recommended  in 
Bulletin  31  were  used.  The  results  obtained  by  the  use  of  larger  portions 
of  solution  yielded,  in  general,  more  reliable  figures.  If  larger  quantities 
of  solution  were  to  be  used,  a  great  deal  of  the  argument  in  favor  of  the 
colorimetric  and  turbidimetric  procedures  disappeared.  The  small  por- 
tions of  solution  required  for  these  methods  have  always  been  considered 
one  of  their  great  merits.  In  the  soil  extracts  which  contained  the 
largest  quantities  of  plant  nutrients,  if  larger  aliquots  were  taken,  it 
was  very  clear  that  the  ordinary  volumetric  and  gravimetric  procedures 
could  be  used.  It  then  became  important  to  learn  whether  the  standard 
methods  of  determination  could  not  be  applied  to  the  weaker  solutions 
if  a  carefully  standardized  technic  were  adopted.  In  the  work  under- 
taken along  this  line  the  dilute  and  average  solutions  previously  men- 
tioned were  used.  As  a  result,  it  is  believed  that  a  more  satisfactory 
procedure  than  the  colorimetric  has  been  developed  for  phosphates, 
calcium,  and  potash.  The  methods  used  are  quite  well  known,  but  a 
great  deal  of  work  was  done  to  establish  the  exact  conditions  for  accurate 
results  with  the  dilute  solutions  employed. 

The  description  of  the  detailed  procedures  used  follows. 

Phosphate. — Evaporate  two  portions  of  soil  extract  of  200  c.  c.  each  in  a  200-c.  c. 
porcelain  casserole.  This  size  of  casserole  stands  ignition  excellently.  Add  a  few 
drops  of  hydrochloric  acid  diluted  i  to  i  before  the  above  solution  reaches  dryness 
to  aid  in  decomposing  soluble  silicates.  Ignite  the  dry  residues  in  the  casserole 
over  a  Meeker  burner  at  a  moderate  temperature  till  a  grayish  white  residue  is 
obtained.  Cool.  Takeup  the  residue  with  loc.c.  of  nitric  acid  diluted  i  tog.  Cover 
with  a  watch  glass  and  digest  on  the  steam  bath  for  approximately  10  minutes  to 
insure  complete  solution.  Filter  into  a  200-c.  c.  Erlenmeyer  flask.  Wash  the 
casserole  with  two  more  portions  of  10  c.  c.  each  of  hot  nitric  acid  diluted  i  to  9. 
Wash  the  dish,  and  filter  with  small  portions  of  hot  water.  Keep  the  total  volume 
of  solution  to  about  50  c.  c.  Cool.  Add  a  few  drops  of  methyl  orange  indicator 
and  neutralize  rapidly  with  concentrated  ammonia.  Bring  just  to  acid  reaction 
with  concentrated  nitric  acid.  Add  2  c.  c.  of  saturated  ammonium  nitrate.  Place 
the  flasks  and  a  quantity  of  properly  acidified  ammonium-molybdate  solution  ^ 
in  a  water  bath  at  50°  C.  When  all  the  solutions  have  reached  this  temperature, 
add  5  c.  c.  of  molybdate  to  each  determination.  Keep  at  50°  C.  for  K  tour.  Remove 
from  the  water  bath,  and  filter  at  once  on  prepared  asbestos  felts.  Wash  with  cold 
distilled  water  till  free  from  acid  by  the  usual  tests. ^  Transfer  filter  felt  to  the  same 
flask,  using  approximately  25  c.  c.  of  distilled  water  free  from  carbon  dioxid.  Add 
15  c.  c.  of  sodium  hydroxid  of  which  i  c.  c.  is  equivalent  to  o.i  mgm.  of  phosphorus- 
pentoxid  and  observe  carefully  whether  the  solution  of  the  yellow  precipitate  is 
complete.  Titrate  the  excess  alkali  with  hydrochloric  acid  of  the  same  strength, 
using  phenolphthalein  as  indicator.  Calculate  results  to  either  elementary  phos- 
phorus or  the  phosphation,  as  desired. 

Calcium. — Evaporate  200  c.  c.  of  water  extract  to  dryness  in  a  200-c.  c.  casserole. 
Ignite  at  a  moderate  temperature  over  a  Meeker  burner  till  a  grayish  white  ash  is 
obtained.     This  step  is  desirable  to  remove  traces  of  organic  matter,  even  though  the 

'Wiley,  H.  W.,  ed.    ofpiciai,  and  provisional  methods  of  analysis,  association  of  official 

AGRICULTURAL  CHEMISTS,    AS  COMPILED  BY    THE  COMMITTEE  ON  REVISION  OF  METHODS.       U.  S.    Dept. 

Agr.  Bur.  Chem.  Bui.  107  (rev.),  p.  4.     1908.     Reprinted  in  1912. 


Feb.  II,  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  329 

extracts  may  be  decolorized  with  carbon-black.  Take  up  with  20  c.  c.  of  hot  distilled 
water  acidified  with  hydrochloric  acid  diluted  i  to  i.  Digest  for  a  short  time  on  the 
steam  bath  to  insure  total  solution.  Filter  and  wash  the  dish  and  the  fimnel  with 
hot  water.  Do  not  permit  the  total  volume  of  solution  to  exceed  50  c.  c.  Make 
faintly  alkaline  with  ammonia,  heat  to  boiling,  add  5  c.  c.  of  10  per  cent  ammonium 
chlorid  and  5  c.  c.  of  saturated  ammonium  oxalate.  Cover  solutions  with  watch 
glasses,  place  on  steam  bath  at  moderate  temperatiue  (about  60°  C),  and  allow  to 
stand  overnight  to  insure  complete  precipitation.  Filter  on  quantitative  filter  paper. 
Wash  till  free  from  chlorids,  using  successive  small  portions  of  hot  water.  Dissolve 
the  precipitate  in  50  c.  c.  of  hot  sulphuric  acid  diluted  i  to  4,  receiving  the  acid  and 
hot  water  subsequently  used  for  washing  the  filter  paper  in  the  beakers  used  for 
precipitation.  Titrate  the  hot  solution  with  N  1I200  potassium  permanganate,  i  c.  c. 
of  which  is  equivalent  to  o.i  mgm.  of  calcium. 

Potash. — Measure  200  c.  c.  of  water  extract  into  a  200-c.  c.  casserole,  acidify  with 
0.5  c.  c.  of  sulphuric  acid  diluted  i  to  i,  and  evaporate  to  dryness  on  the  steam  bath. 
Heat  gently  on  an  asbestos  hot  plate  and  keep  the  casserole  constantly  in  motion  to 
prevent  spattering  imtil  the  excess  of  sulphuric  acid  has  volatilized.  Then  ignite 
over  a  Meeker  burner  until  a  white  or  gra^dsh  white  residue  is  obtained.  Cool  and 
take  up  the  residue  in  successive  small  portions  of  hot  distilled  water.  Filter  through 
5.5  cm.  quantitative  filter  paper,  using  30  to  40  c.  c.  of  water.  Acidify  with  three  or 
four  drops  of  hydrochloric  acid  diluted  i  to  i.  Add  8  to  12  drops  of  platinic-chlorid 
solution  and  evaporate  to  a  pasty  consistency  on  the  water  bath ;  care  shoxild  be  exer- 
cised to  avoid  overheating.  Cool.  Take  up  the  residue  in  5  c.  c.  of  95  per  cent 
alcohol,  observe  that  an  excess  of  platinic  chlorid  is  present.  Tritiu-ate  the  precipitate 
carefully  with  a  rubber  "policeman"  and  filter  at  once  on  a  carefully  prepared  and 
weighed  Gooch  crucible.  Wash  with  95  per  cent  alcohol  till  the  filtrate  is  colorless, 
then  with  30  to  40  c.  c.  of  ammonium  chlorid,  100  gm.  to  500  c.  c.  of  water,  which 
has  been  saturated  with  potassium  platinic  chlorid.  Wash  once  with  80  per  cent 
alcohol  and  then  thoroughly  with  95  per  cent  alcohol.  Dry  in  an  oven  at  100°  C. 
and  weigh.     Calculate  the  weight  of  potassium. 

With  the  use  of  the  known  solutions  previously  prepared  and  the 
above  methods,  the  following  results  (Table  IV)  are  typical  of  the  ac- 
curacy obtained. 

27809°— 18 3 


330 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  6 


Table   IV. — Analyses  of  dilute  and  average  solutions   by   the  methods  used  in   this 

investigation 


K. 

Ca. 

PO4. 

Solution  and  treatment. 

Theory. 

Deter- 
mined. 

Theory. 

Deter- 
mined. 

Theory. 

Deter- 
mined. 

Dilute  solution  (untreated) 

P.p.m. 

I.  2 
I.  2 
j     1.  2 
I.  2 
I.  2 
I.  2 

P.p.m. 
1.05 
I.  OS 
I.  21 

•97 

I.  CO 
I.  22 

P.p.m. 
1.50 
1.50 
1.50 
1.50 
1.50 
1.50 

P.p.m. 
I.  62 

1.47 
1.65 
I.  60 

1-45 
I.  60 

P.p.m. 
I.  0 
I.  0 
I.  0 
I.  0 
I.  0 
I.  0 

P.p.m. 
I.  12 
I.  01 

I.  20 

•94 
I.  10 

•97 

Mean 

I.  2 

1.08 

1.50 

1.56 

I.  0 

I.  06 

Average  solution  (untreated) 

(  6.0 
6.0 
6.0 
6.0 
6.  o 

.  6.0 

6.05 

6.  20 
6.05 
6.60 

5-95 
6.50 

7-5 
7-5 
7-5 
7-5 
7-5 
7-5 

7.91 
7.90 

7.  62 
7.64 
7.66 
7.80 

S-o 
S-o 
S-o 
5-0 
S-o 
S-o 

5.06 
5.02 
4.90 

S-I5 
5.10 

5-19 

Mean  

6.0 

6.  22 

7-5 

7-75 

5-0 

5-07 

Percentage  of  error,  dilute  solution : 
Mean  .   .       

10.  00 
19.  00 

3-6 
10.  0 

4.  00 
10.  00 

3.00 
5-4° 

6.  00 

Maximum 

20.  00 

Percentage  of  error,  average  solution: 
Mean          

I.  40 

Maximum       

3-80 

The  determinations  on  each  element  are  given  on  six  duplicate  portions 
of  solution.  In  addition,  the  mean  percentage  of  error  is  calculated,  as 
well  as  the  maximum  error  for  any  single  determination.  It  will  be  ob- 
served that  the  agreement  with  the  theoretical  result  is  satisfactory  for 
such  dilute  solutions.  The  potassium  determination  shows  the  greatest 
discrepancy,  with  a  mean  error  of  10  per  cent  and  a  maximum  of  19  per 
cent  on  the  dilute  solution.  On  the  average  solution  the  results  are 
much  closer,  ranging  from  3.6  to  10  per  cent.  With  calcium  the  error 
on  both  the  dilute  and  average  solution  is  even  smaller,  4.0  and  3.0  per 
cent  for  a  mean  result,  and  10  and  5.4  per  cent  for  a  maximum.  The 
phosphate  determination  shows  a  slightly  larger  fluctuation;  in  the 
dilute  solution  from  6  to  20  per  cent,  but  the  higher  figure  was  only 
obtained  in  one  analysis.  The  stronger  solution  gave  greater  accuracy 
than  was  obtained  in  any  of  the  other  work. 

Taken  as  a  whole,  it  is  believed  that  the  number  of  determinations  is 
sufficiently  large  to  show  the  small  probability  of  the  occurrence  of  the 
maximum  error.  The  mean  error  will  therefore  be  taken  as  the  usual 
figure  to  be  applied  in  the  interpretation  of  the  subsequent  results. 
All  data  bearing  on  the  seasonal  studies  which  follow  will  be  considered 
in  the  light  of  their  accuracy  as  affected  by  the  mean  and  maximum 
error. 


Feb.  II,  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  331 

EXTRACTION  PROCEDURE 

It  has  been  pointed  out  in  an  earlier  portion  of  this  discussion  that 
several  extraction  procedures  have  been  applied  to  the  study  of  the  water- 
soluble  material  of  soils.  In  Europe  the  methods  of  Schloesing  {46,  47) 
and  Mitscherlich  (42)  are  largely  favored,  though  the  work  of  the  U.  S. 
Bureau  of  Soils  (61)  has  some  followers.  This  latter  system  of  extraction 
is  practically  the  only  one  which  has  been  used  in  this  country.  The 
desire  to  have  the  result  of  this  investigation  somewhat  comparable  to 
previous  work  was  therefore  an  argument  in  favor  of  this  general  pro- 
cedure. This  would  not  have  held  had  there  been  any  closer  method  of 
approximating  the  actual  soil  solution,  for  the  medium  v/hich  nourishes 
the  plant  should  always  be  our  final  object.  As  before  noted,  it  is  not 
known  that  this  actual  solution  has  ever  been  obtained.  It  then  be- 
comes necessary  to  use  some  arbitrary  method  of  extraction,  and  from 
the  data  at  hand  in  beginning  this  work  the  general  procedure  of  the 
U.  S.  Bureau  of  Soils  was  chosen. 

In  all  the  seasonal  studies  which  are  recorded  later  only  fresh  soil 
was  used  for  the  analytical  studies.  The  great  change  which  soil  under- 
goes in  drying  as  seen  from  the  work  of  King  (55),  outweighs  any  con- 
siderations of  expediency.  King  himself  took  the  contrary  view,  but  a 
study  of  his  own  figures  shows  how  large  and  how  variable  were  the 
changes  between  different  soils  caused  by  drying  the  samples  previous 
to  analysis.  His  adoption  of  this  modification  was  influenced  by  a 
desire  to  carry  on  work  at  several  stations  from  one  central  laboratory. 
In  the  present  investigation  the  soils  had  been  brought  together,  so  no 
such  reason  existed.  It  has  been  desired  throughout  to  perform  all 
work  on  the  fresh  soils  and  to  show  what  differences  existed  in  them. 

The  uniform  procedure  of  extraction  is  as  follows : 

Place  the  weighed  portion  of  fresh  soil  in  a  large  mortar  and  mix  with  a  small 
amount  of  distilled  water  till  it  becomes  a  thin  homogenous  paste.  Then  add  the 
rest  of  the  distilled  water,  making  in  all  five  times  the  weight  of  the  soil  and  mix 
the  whole  for  exactly  3  minutes.  Then  transfer  to  wide-mouthed  bottles  all  the 
solution  and  suspended  soil  that  will  pour  off  and  allow  it  to  stand  for  40  minutes. 
At  the  end  of  this  period  pour  the  solution  without  mixing  into  cylinders  of  a  Briggs 
filtering  apparatus,  connected  to  the  pressure  system,  and  apply  compressed  air  at 
approximately  100  pounds'  pressure  till  filtration  is  complete.  Discard  the  first 
portion  of  the  solution,  about  50  c.  c,  to  avoid  any  dilution  by  the  moisture  held  in 
the  Pasteur-Chamberland  candles  and  also  any  small  absorption  by  the  filter  from  the 
solution. 

After  filtration  is  complete  wash  the  filter  candles  and  cylinders  in  tap  water,  rinse  in 
distilled  water,  and  pass  600  to  800  c.  c.  of  distilled  water  through  each  candle  to  wash 
out  any  remaining  portions  of  solution.  After  each  weekly  set  of  filtrations  is  com- 
pleted, thoroughly  dry  the  washed  candles  in  an  air  oven  at  ico°  C.  to  prevent  the 
development  of  bacterial  flora  in  the  outer  pores  of  the  tubes.  The  efficiency  of 
filtration  of  the  candles  is  gradually  impaired  by  continual  use,  owing  to  the  deposition 
of  clay  and  fine  silt  in  the  unglazed  porcelain  of  which  it  is  composed.  This  can  be 
partly  remedied  by  igniting  the  dried  tubes  in  a  mufile  furnace  for  }A  hour  or  more. 


^^2  Journal  of  Agricultural  Research  voi. xii.  No.  6 

Then  cool  the  candles  and  pass  distilled  water  through  them  to  remove  charred 
organic  matter  and  salts  formed  by  the  process  of  ignition. 

In  the  work  carried  out  in  191 5,  with  the  analytical  methods  of  the 
Bureau  of  Soils,  150  gm.  of  fresh  soil  constituted  the  sample  treated 
with  750  c.  c.  of  distilled  water.  In  191 6  the  modified  procedure  which 
has  been  described  required  a  larger  volume  of  solution.  Accordingly 
340  gm.  of  soil  has  been  treated  with  i  ,700  c.  c.  of  distilled  water.  This 
volume  of  solution  is  the  maximum  quantity  that  can  be  filtered  in  two 
cylinders  of  the  filtering  apparatus.  These  two  portions  of  solution 
are  afterwards  united  and  furnish  a  volume  sufficient  for  the  determina- 
tion of  phosphates,  calcium,  potash,  nitrates,  magnesium,  and  total 
solids,  if  desired.  A  moisture  determination  was  in  each  case  performed 
on  the  sample  of  fresh  soil,  and  the  result  used  in  correcting  the  analytical 
data  to  bring  it  to  a  uniform  basis. 

In  this  connection  it  should  be  pointed  out  that  the  form  in  which 
the  final  results  are  to  be  stated  is  a  matter  of  some  importance.  Mitscher- 
lich  (42),  King  {35),  Whitney  and  Cameron  (<5i),  all  expressed  their 
results  on  the  basis  of  percentage  or  parts  per  million  of  the  dry  soil. 
Gola  {18),  on  the  other  hand,  referred  his  to  the  concentration  of  the 
solution  obtained,  which  he  considered  to  be  representative  of  the  soil 
solution.  Inasmuch  as  it  is  the  soil  solution  which  is  our  final  object, 
it  would  be  natural  to  refer  the  analytical  results  to  the  actual  soil 
moisture  determined  in  each  sample.  Desirable  as  it  is,  it  is  not  believed 
that  we  possess  the  necessary  information  in  regard  to  the  state  of  the 
soil  moisture  to  do  this  correctly.  In  discussing  the  results  on  the 
moisture  equivalents  given  in  Table  II  we  have  referred  to  the  unfree 
water  present  in  soils. 

The  work  done  by  Briggs  and  Shantz  (<?)  on  the  wilting  coefficient  of 
soils  has  shown  the  variations  which  occur  among  different  soils,  espec- 
ially where  they  are  of  diverse  types.  Their  work  demonstrates  clearly 
that  a  large  amount  of  moisture  is  unavailable  to  the  plant  in  soils 
which  are  high  in  colloidal  material.  Recently  Bouyoucos  {4)  has 
called  attention  to  the  correlation  which  exists  between  the  moisture 
held  in  unavailable  form  in  the  soil,  as  determined  by  the  above  method, 
and  the  unfree  moisture  determined  by  the  dilatometer  method.  He 
has  also  compared  the  results  obtained  by  these  two  procedures  with 
the  determination  of  the  moisture  which  fails  to  freeze.  A  remarkable 
agreement  is  found  to  exist.  Working  with  some  14  soils  used  in  the 
present  study  Hoagland  (27)  has  found  that  in  all  the  silty  clay  loams 
a  large  percentage,  varying  from  13  per  cent  to  18  per  cent,  does  not 
freeze.  In  the  fine  sandy  loams  the  percentage  is  much  smaller,  from 
4  per  cent  for  soil  8  to  8  per  cent  for  soil  10.  The  extent  of  these  varia- 
tions and  their  significance  is  discussed  at  greater  length  by  Hoagland 
(27).  If  this  moisture  is  held,  either  loosely  chemically  combined  or 
physically  absorbed  or  both,  it  is  at  least  possible  that  it  is  not  available 


Feb.  II.  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  333 

for  the  solution  of  the  soil  nutrients.  In  such  a  case  the  actual  available 
solution  carrying  plant  nutrients  would  not  be  very  different  in  the 
complex  silty  clay  loams  from  that  which  exists  in  the  fine  sandy  loams. 
It  has  therefore  been  decided  to  express  all  results  on  the  basis  of  the 
dry  soil  until  further  evidence  is  obtained  as  to  the  condition  of  the 
moisture  in  the  soil. 

AGREEMENT  OF  DUPLICATE  EXTRACTIONS 

Before  the  results  obtained  by  any  extraction  procedure  can  be  con- 
sidered, it  is  necessary  to  know  the  limits  of  the  accuracy  which  it  is 
possible  to  achieve.  The  first  important  point  in  this  connection  is 
the  possibiUty  of  obtaining  accurate  duplicate  extractions.  A  number 
of  sets  of  duplicate  extracts  have  been  made  from  samples  of  fresh 
field  soils.  The  results  on  four  typical  soils,  two  silty  clay  loams  and  two 
fine  sandy  loams,  are  given  in  Table  V.  It  will  be  observed  that  the 
agreement  is  well  within  the  limits  of  the  mean  factor  of  error  given 
in  Table  IV  for  all  the  elements  except  calcium.  It  will  be  remembered 
that  the  factor  of  error  for  this  element  ranged  from  4  to  10  per  cent. 
With  the  two  fine  sandy  loams  the  variations  between  the  two  duplicates 
is  between  10  and  15  per  cent.  This  result  would  readily  be  accounted 
for  by  the  solubility  of  the  calcium  compounds  present  in  the  soil  and 
indicates  that  it  is  necessary  to  obtain  far  larger  changes  with  this 
element  than  any  others  studied  before  the  result  can  be  considered 
significant.  It  must  be  emphasized,  however,  that  duplicate  results 
such  as  those  given  above  can  only  be  obtained  if  every  step  in  the 
procedure  chosen  is  carefully  followed. 

Table  V. — Typical  duplicate  extractions  of  silty  clay  loams  and  fine  sandy  loams 
[Results  expressed  as  parts  per  million  of  dry  soil.    Ratio  of  soil  to  water  1:5] 


Soil. 

NO3. 

PO4. 

K. 

Ca. 

Mg. 

No.  I  silty  clay  loam 

32.5 

4-3 
3-8 

33-8 
34-8 

16.  7 
16.3 

10.3 
9.9 

Do 

Mean 

32.9 

4.  I 

34.3 

16.5 

10.  I 

No.  5  silty  clay  loam 

27.6 
29.  0 

13- 9 

29-5 
30.  0 

57-1 
63.0 

14.8 
15.0 

Do 

Mean 

28.3 

13- 6 

29.7 

60.  0 

14.9 

No.  7  fine  sandy  loam 

26.3 
26.3 

3-6 

4.  2 

18.0 
18.6 

37-5 
42.  0 

IO-5 

Do 

II.  4 

Mean 

26.3 

3-9 

18.3 

39-7 

No.  8  fine  sandy  loam 

22.  6 
22.  2 

II.  7 
10.  6 

Zi-  7 
30.  6 

35-  I 
30.0 

8.1 

8.S 

Do 

Mean 

22.  4 

II.  I 

32.1 

32.5 

8.3 

334 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  6 


EFFECT  OF  TEMPERATURE  AND  CARBON  DIOXID 

The  next  point  to  be  investigated  was  the  possible  effect  of  compara- 
tively rapid  changes  of  temperature.  This  was  an  important  consider- 
ation in  studying  fluctuations  in  soluble  nutrients  within  a  short  period 
of  time.  Such  fluctuations  were  found  to  occur.  It  was  extremely 
desirable  to  know  whether  the  change  in  solubility  was  due  merely  to  the 
physical  effect  of  the  increased  temperature  or  was  a  secondary  influence 
due  to  increased  activity  of  the  bacterial  flora.  Such  activity  would 
result  in  the  production  of  appreciable  quantities  of  carbon  dioxid  and  a 
possible  increase  in  soluble  soil  compounds. 

A  series  of  studies  were  undertaken  to  investigate  these  various 
effects.  A  silty  day  loam  soil  of  excellent  productivity  was  used.  One 
container  of  this  soil  had  been  kept  fallow  all  season;  the  other  had 
grown  a  crop  of  barley.  Both  samples  contained  practically  identical 
quantities  of  moisture,  slightly  below  the  optimum  for  this  soil. 

One  portion  of  each  soil  was  placed  in  a  bacteriological  incubator  at 
29°  C.  A  second  was  kept  in  a  refrigerator  at  6°  for  two  hours,  a  third 
was  treated  with  carbon  dioxid  for  24  hours,  a  fourth  was  air  dried  at 
20°  for  48  hours,  and  a  control  portion  was  kept  for  24  hours  untreated. 
The  results  are  given  in  Table  VI. 

Table  VI. — Effect  of  loiv  temperature,  increased  temperature,  air-drying,  and  carbon' 
dioxid  on  water-soluble  nutrients 

[Results  expressed  as  parts  per  milUon  of  dry  soil] 


No. 

sA  (Yolo  silty  clay  loam, 

No.  5B  (Yolo  silty  clay  loam,  un- 

cropped). 

cropped). 

mX 

■S 

•S--i 

•t3 

Treatment. 

■-.2 

Si, 

•3 

1% 

1 

•3 

0 

0 

0 

0 

01  r; 

•3  S 

■3 

1 

0 

0 

'1% 

Z 

PL, 

U 

a 

H 

>=. 

"A 

^ 

Pk 

1^ 

0 

H 

291 

254 
404 

"A 

70.7 
68.  4 
62.  1 

17- S 

14- S 
22.  9 

39-9 

4.-i-5 
37-9 

83-3 
83-3 

68.3 

472 
459 

547 

195 
189 
319 

277 
270 
228 

139-5 

143-2 
127.  2 

20.5 
13-0 
22.4 

52.8 
67-3 
49-7 

103.9 
97-7 
70-3 

676 
626 

751 

385 

^72 

48  hours  air-dry 

347 

24  hourscarbon  dioxid  at  i  s  °  C  — 

72-5 

18. 1 

.;2.6 

113- 2 

805 

252 

.S53 

150-7 

21.8 

52-8 

120.  9 

918 

31b 

602 

24  hours  untreated  at  is"  C 

68.4 

20.  2 

47-5 

75- S 

447 

195 

252 

135-8 

17.2 

5«-3 

97.0 

b32 

273 

359 

The  data  in  Table  VI  must  first  be  considered  by  applying  the  mean 
factor  of  error  for  the  several  determinations.  This  correction  shows 
that  the  effect  of  the  temperature  treatments  in  the  incubator  and  re- 
frigerator has  been  negligible.  It  is  therefore  probable  that  sudden  wide 
changes  of  concentration  of  water-soluble  nutrients  are  due  to  other 
factors  than  changes  in  the  solubility  of  plant  food  caused  by  tempera- 
ture alone. 

The  air-drying  to  which  the  two  samples  were  subjected  was  carried 
out  at  20°  C.  without  exposure  to  sunlight.     It  was  consequently  less 


Feb.  II,  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  335 

rigorous  than  the  similar  treatment  which  surface  soil  received  in  the 
field.  Even  so  it  is  seen  to  have  produced  several  striking  changes. 
The  nitrates  and  calcium  are  slightly  decreased,  but  the  greatest  differ- 
ence is  observed  in  the  total  solids.  Here  we  see  a  significant  increase. 
This  is  further  shown  by  the  volatile  solids  to  be  entirely  due  to  a  gain 
in  soluble  organic  matter.  Such  an  increase  in  soluble  organic  materials 
would  suggest  that  the  efifect  of  air-drying  on  the  soil  minerals  would 
largely  be  of  a  secondary  nature.  Any  cultivation  or  stirring  into 
the  soil  mass  of  this  air-dried  portion  of  the  soil  would  result  in  placing 
this  soluble  material  in  circulation  in  the  soil  solution.  It  would  thus 
become  available  as  food  for  the  biological  life  of  the  soil,  and  may  be  one 
of  the  factors  tending  to  produce  the  large  liberation  of  nutrients  after 
cultivation. 

In  the  final  treatment  of  the  two  soil  samples  carbon-dioxid  gas  was 
frequently  passed  through  the  jars  containing  the  soil  during  the  24- 
hour  period.  This  resulted  in  a  striking  liberation  of  soluble  mineral 
compounds.  There  was  also  a  slight  increase  in  the  production  of  ni- 
trates in  comparison  to  the  untreated  soil,  but  it  is  very  noticeable  that 
the  phosphates  remained  practically  constant.  This  is  significant.  It 
will  be  pointed  out  later  that  phosphates  are  the  only  one  of  the  plant 
foods  with  which  the  solution  of  each  individual  soil  appears  to  be  satu- 
rated. The  amount  of  nonvolatile  solids  liberated  by  this  treatment  is 
over  200  p.  p.  m.,  and  shows  that  the  efifect  of  the  carbon  dioxid  e^Jtends  to 
other  compounds  than  those  indicated.  From  this  study  it  seems  prob- 
able that  the  effect  of  the  carbon  dioxid  present  in  the  soil  air  and  dis- 
solved in  the  soil  solution  is  the  most  important  influence  that  renders  the 
soil  minerals  soluble. 

NATURE  OF  THE  WATER  EXTRACT 

The  above  series  of  results  establish  the  accuracy  of  the  extraction  pro- 
cedure which  has  been  employed.  They  do  not  throw  any  light  on  the 
nature  of  the  solution  obtained.  All  of  the  minerals  present  in  the  soil 
possess  a  small  but  appreciable  solubility.  Any  soil  extract  will  there- 
fore contain  not  only  the  material  which  is  actually  in  solution  or  very 
readily  soluble  but  also  an  extra  amount  derived  from  the  soil  minerals 
themselves.  This  latter  portion  of  the  soil  extract  may  in  some  cases 
be  different  for  the  same  soil.  It  is  a  well-known  principle  that  when  a 
compound  is  treated  with  a  solution  containing  a  common  ion  its  solu- 
bility will  be  repressed.  If  we  now  turn  to  the  graphs  showing  the  nutri- 
ents extracted  from  the  same  soils,  cropped  and  uncropped  (fig.  3-6),  it 
will  be  observed  that  there  are  striking  differences,  which  will  be  dis- 
cussed later.  In  the  cropped  soil  there  is  much  less  soluble  material, 
and  the  solution  used  to  extract  it  will  more  nearl}^  approach  distilled 
water  than  in  its  duplicate,  where  nothing  has  been  removed  by  the 


336 


Journal  of  Agricultural  Research  voi.  xii,  no.6 


ERA.  PRY  SOIL 
500 

450 

> 

^  ^'' 

SOIL 

350 

4 

^^^ 

300                                     ^^^'' 

^00/ 

POTAS^IV^  ..CK.) 

150 

100                    _,.---•""' 

- ' ' 

''' 

50    ,.--'" 

/ - 

i!0?i.5  5         10                 i'o 

1 

40 

_ .  P  H. p  5  F  H.AT.&....f-'5^?-^> 

1 
60 

RPiTlO   50!L:VATEII 

Fig.  3. — Graphs  of  the  nutrients  extracted  from  soil  4  by  varying  the  ratios  of  soil  to  water.    Calculated  to 

parts  per  milUon  of  dry  soil. 


F.PA. 
500 

PRY 

50IL 

450 

^ 

400 
350 
300 

50IL5 

£50- 
200 

/<^^" 
^,-^^ 

^^<^ 

_TOT.f.A^)^-^-^- 

ISO 
100  J^ 

:^ 

__,.o.TH/^X^-^^- 



SO   ^,. 

y 

1 
5 

1          1 

10              ^0 

1 

eo 

RATIO  :5oil;vatek 

Fig.  4. — Graphs  of  the  nutrients  extracted  from  soil  5  by  varying  the  ratios  of  soil  to  water.    Calculated  to 

parts  per  million  of  dry  soil. 


Feb.  II.  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  337 


p.p.n. 

50O 

PRY  iOlL. 

45fi 

400 

XI LQ 

350 

300 

250 

2.00 
150 

~ 

n^—  - 

...^^^^^^^^ 

--     ° 

100/ 

1 
5 

.X9T  ?>AVyj^-  SAL 

. , 

7y^ob?n.f^.lX'..^y.9^). 

1 
10 

1 

10 

1 
40 

80 

RATIO 

50IL 

:VATEK 

FlO.  5.— Graphs  of  the  nutrients  extracted  from  soil  8  by  varying  the  ratios  of  soil  to  water.    Calculated  to 

parts  per  million  of  dry  soil. 


T.T.n.  PKY301L 

500 

/ 

// 

450 

O"^ 

f 

400 

50IL  10 

/ 
/ 

350 

V 

500 

4' 

£50 
20Q 

I50; 

^^^ 

'^"^'^^  yo^^-^:>^^--'""'^' 

100 

^-'=^=^^^^^.— --- 

-  -  ""' 

.'r^r-'^'j^"' 

•^ty- 

'-^    . 

.xno^^^-^^-^^-"" 

Q*. 

_^ .-..---•—•"* 

'^hW, 

1 

1 

10                  20 

40 

ea 

RATIO    501L 

IVATER 

Pig.  6. — Graphs  of  the  nutrients  extracted  from  soil  lo  by  varying  the  ratios  of  soil  to  water.    Calculated 

to  parts  per  million  of  dry  soil. 


338 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  6 


plant.  It  may  therefore  be  expected  that  in  the  solution  from  this  soil 
more  material  derived  from  the  soil  minerals  will  be  soluble  than  in  the 
uncropped  extract.  The  effect  of  this  will  be  to  make  the  differences 
which  actually  exist  less  striking  and  evident  than  is  really  the  case. 
The  evidence  given  by  the  freezing-point  method,  which  has  greatly  de-* 
veloped  the  above  theory,  is  discussed  by  Hoagland  (27). 

The  suggestion  was  made  by  Burd  that  it  might  be  possible  to  obtain 
some  conception  of  this  repressive  effect  by  extracting  a  cropped  soil 
with  a  diluted  extract  from  its  uncropped  duplicate.  Two  soils  were 
used  for  this  purpose:  Soil  5,  Yolo  silty  clay  loam,  and  soil  10,  Tejunga 
fine  sandy  loam.  In  each  case  an  extract  was  first  prepared  of  an  un- 
cropped reserve  portion  of  the  soil  and  of  another  portion  upon  which 
a  crop  had  been  grown.  A  determination  of  total  solids  was  immediately 
made.  The  solution  from  the  uncropped  soil  was  then  diluted  with 
distilled  water,  so  that  this  solution  added  to  that  obtained  from  the 
weaker  cropped  soil  would  again  be  equal  to  the  strong  solution.  The 
cropped  soil  was  extracted  with  this  diluted  solution.  The  results  of 
this  extraction  are  compared  in  Table  VII  with  the  amounts  obtained 
by  the  usual  treatment  with  distilled  water  plus  the  quantities  already 
present  in  the  diluted  extract. 

Table  VII. — Extraction  of  two  cropped  soils  with  diluted  extracts  from  duplicate  bin  soils 
[Results  expressed  as  parts  per  million  of  dry  soil] 


Soil. 

Total 
solids. 

Fixed 
solids. 

Volatile 
solids. 

FO*. 

Ca. 

Mg. 

K. 

Cropped  soil  5  extracted  with  di- 
lute extract  from  reserve  bins . . . 

Nutrients  extracted  from  same  soil 
by  distilled  water  (1:5)+  nutri- 
ents contained  in  the  dilute  ex- 
tract   

1,686 

1,575 

1,163 

962 

412 

15-4 
25-7 

187.0 
195.0 

41-5 
46.  4 

66.1 
Rr.  6 

Depression  of  solubility 

439 

10.3 

8.0 

4.9 

15-5 

Cropped  soil  10  extracted  with  di- 
lute extract  from  reserve  bins . . . 

Nutrients  extracted  from  same  soil 
by  distilled  water  (1:5)4-  nutri- 
ents contained  in  the  dilute  ex- 
tract  

426 

589 

157 
298 

269 
291 

8.8 
16.  4 

32.1 

55-8 

6.6 
8.8 

55-4 
76.8 

Depression  of  solubility 

163 

141 

22 

7.6 

23-7 

2.  2 

21.4 

With  the  Yolo  silty  clay  loam  5  there  was  seen  to  be  a  distinct  de- 
pression of  phosphates  and  a  slight  depression  of  potassium,  both  of 
which  were  greater  than  the  maximum  factor  of  error.  The  depression 
of  calcium  was  no  more  than  the  average  factor  of  error.  The  effect 
was  especially  seen  in  the  amount  of  fixed  solids  which  was  held  out 


Feb.  II.  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  339 

of  solution,  but  the  total  solids  were  slightly  increased  as  organic  matter 
replaced  the  nonvolatile  solids. 

With  Tejunga  fine  sandy  loam  10  there  was  a  distinct  depression 
throughout.  Both  solids  and  all  individual  constituents  were  decreased 
in  quantity. 

This  evidence,  together  with  that  obtained  by  Hoagland  (27)  using 
the  freezing-point  method,  clearly  showed  that  the  differences  which 
may  be  found  between  cropped  and  fallow  soils  by  the  extraction  pro- 
cedure was  always  less  than  that  which  actually  exists. 

In  the  same  paper  (27)  Hoagland  has  also  shown,  both  by  special  ex- 
traction studies  and  by  calculation  from  the  amounts  of  total  solids 
contained  in  the  i-to-5  extract,  that  the  relation  of  the  actual  soil  solu- 
tion to  this  water  extract  is  extremely  close.  The  amounts  of  material 
in  the  i-to-5  extract  are  approximately  two  to  five  times  as  great  as  in 
the  soil  solution.  The  range  of  concentration  shown  by  these  extrac- 
tions is  therefore  undoubtedly  too  high.  This  does  not  nullify  the 
differences  between  cropped  and  uncropped  soils  which  are  shown  to 
exist  by  this  procedure.  Hoagland 's  results  show  that,  if  anything, 
these  differences  are  actually  much  greater.  A  decrease  in  the  range  of 
concentrations  would  therefore  have  no  more  effect  than  to  change  the 
scale  upon  which  the  seasonal  curves  have  been  drawn. 

EFFECT  OF  VARYING  THE  PROPORTION  OF  SOIL  AND  WATER 

It  was  important  to  learn  what  relation  the  conventional  i-to-5  extract 
bears  to  other  possible  ratios  of  soil  and  water.  Reference  has  been 
made  to  the  work  of  Mitscherlich  (42),  in  which  he  employed  a  series  of 
varied  extractions  of  soil  with  water  saturated  with  carbon  dioxid. 
This  idea  was  now  applied.  Four  soils,  two  silty  clay  loams  and  two 
fine  sandy  loams,  were  extracted  with  proportions  of  soil  to  water  vary- 
ing from  I  part  of  soil  to  %  part  of  distilled  water  up  to  i  part  of  soil  to 
80  parts  of  distilled  water. 

The  results  in  parts  per  million  of  dry  soil  have  been  plotted  in  figures 
3  to  6.  It  will  be  observed  that  with  each  soil  the  proportions  of  com- 
pounds extracted  in  the  lower  ranges  of  concentration  were  essentially 
the  same.  It  was  not  until  a  range  of  concentrations  of  soil  to  water 
greater  than  i  to  10  were  reached  that  a  distinct  change  in  the  extrac- 
tion of  nutrients  took  place.  These  results  therefore  warranted  the 
employment  of  the  i-to-5  extract  as  a  conventional  procedure. 

SEASONAL  STUDIES  OF  THE  WATER  EXTRACT 
FIRST  SEASON,  1915 

In  the  first  season's  work,  in  1915,  both  containers  of  each  soil  were 
sown  to  barley.  Half  of  the  soils  were  planted  on  June  14,  and  the 
remainder  one  week  later.     At  the  same  time  the  first  soil  samples  were 


340  Journal  of  Agricultural  Research  voi.  xii.  No.  6 

taken  with  the  iS-inch  cheese  trier  previously  mentioned.  From  9  to 
12  triers  of  soil  were  required  for  a  sample  of  approximately  600  gm. 
The  holes  left  by  removing  these  samples  were  filled  with  reserve  soil 
from  the  bins. 

It  was  felt  that  this  procedure  might  be  open  to  some  objections, 
owing  to  the  fact  that  the  soil  in  the  bins  was  partially  air-dried.  In 
any  one  season  10  sampling  periods  is  the  maximum  number,  and  the 
amount  removed  would  not  exceed  6  kgm.  This  amount  was  less  than 
0.8  per  cent  of  the  total  soil  present.  The  error  involved  was  therefore 
smaller  than  it  was  possible  to  detect  by  any  available  methods. 

In  order  to  minimize  this  influence,  the  containers  were  always  sampled 
by  a  definite  system,  so  that  no  two  samples  were  taken  from  the  same 
spot  twice  in  the  same  season.  The  process  of  sampling  was  performed 
with  great  care.  There  was  no  breakage  of  plants  and  no  evidence  of 
injury  from  the  small  roots  cut  by  the  cheese  trier.  The  growth  obtained 
has  been  extremely  satisfactory,  both  in  height  and  vigor.  The  plants 
were  spaced  at  the  usual  field  distance,  6  inches  apart  each  way.  In 
1 91 5  one  seed  was  planted  in  a  place,  and  a  small  number  of  extra  seeds 
were  started  simultaneously  in  the  same  soil.  These  extra  plants  were 
then  transplanted  into  any  spaces  where  the  seed  failed  to  sprout.  An 
even  stand  was  obtained  in  all  except  two  duplicate  containers,  soils 
3  and  5.  The  following  season  the  writers  were  able  to  obtain  a  per- 
fectly uniform  growth  by  planting  three  or  four  seed  in  each  place,  and 
then  thinning  to  one  plant  of  average  size.  The  seed  used  was  furnished 
by  the  Division  of  Agronomy  and  was  a  pure  strain  of  Beldi  barley. 
Before  planting,  it  was  treated  with  formalin  to  prevent  smut;  seed  of 
uniform  size  were  selected. 

A  slight  difficulty  has  been  experienced  toward  the  close  of  each 
season  from  the  tendency  exhibited  by  the  crops  on  several  soils  to  lodge. 
This  has  been  corrected  by  a  system  of  supporting  wires. 

In  1 91 5  the  crop  matured,  and  was  harvested  from  October  21  to 
November  6.  This  was  a  later  date  than  commercial  grain  crops  are 
grown  in  California,  but  it  is  believed  that  the  results  were  strictly  com- 
parable among  themselves.  The  crop  yield  is  given  in  Table  VIII  and 
is  expressed  as  total  air-dry  crop  and  grain  in  grams.  The  grain  is  also 
calculated  as  pounds  and  bushels  per  acre.  It  will  be  observed  that  in 
all  the  soils  except  No.  3  and  5  the  agreement  of  the  duplicates  was 
excellent.  These  two  soils,  it  will  be  remembered,  were  those  in  which 
the  growth  was  least  uniform. 


Feb.  II,  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  341 


Table  VIII. — Crop  yield  of  191$ 


Soil. 


Yolo  siltv  clay  loam : 

No.  lA 

No.  iB 

No,  2A 

No.  2B 

No.  3 A 

No.  3B 

No.  4A 

No.  4B 

No.  5 A 

No.  5B 

Yolo  clay  loam: 

No.  6 A 

No.  6B 

Hanford  fine  sandy  loam: 

No.  7A 

No.  7B 

Fresno  fine  sandy  loam : 

No.  8A 

No.  8B 

Kimball  fine  sandy  loam : 

No.  9A 

No.  9B 

Te junga  fine  sandy  loam : 

No.  loA 

No.  loB 

Madera  fine  sandy  loam : 

No.  iiA 

No.  iiB 

Arnold  fine  sandy  loam: 

No.  12A 

No.  12B 

Standish  fine  sandy  loam : 

No.  14A 

No.  14B 


Average  variations  between 
duplicates 


Total 
yield  of 
air-dry 
grain 
and 
straw. 


Gm. 
2,  092 
1,904 
2,038 
1,649 
1,421 
1,083 

1,331 
1,187 
1,863 
1,742 

2,  287 
2,  182 

1,115 
1,059 

2,  136 
2,  022 

I,  066 

1,  096 

1,401 

•1,337 

2,  020 
2,  194 

796 

775 

1,364 
1,359 


Gm. 

73° 
73° 
605 
548 
474 
333 
534 
459 
529 
420 

739 

705 

432 
363 

827 
780 


405 

495 
500 

742 
774 

242 
214 

390 

403 


Grain 

(pounds 
per  acre). 


5,614 
5,614 
4,653 
4,215 
3,646 
2,561 
4,  106 

3,530 
4,069 
3,230 

5,684 
5,422 

3,323 

2,  792 
6,361 

6,  000 

2,984 
3,115 

3,807 
3,846 

5,707 
5,953 

1,862 
I,  646 

3,000 

3,  100 


Grain 
(bushels 
per  acre). 


93-6 
93-6 
77-5 
70.3 
60.8 
42.7 
68.5 
58.8 
67.8 
53-8 

94-7 
90.4 

55-4 
46.5 

106.  o 
100.  o 

49-7 
SI- 9 

63-5 
64.  I 

95-1 
99.2 

31.0 
27.4 

50.0 
51-7 


Variation 

from 

maximum 

yield  of 

grain. 


Per  cent. 
II.  7 

II.  7 
26.  9 

33-7 
42.  6 

59-7 
35-4 
44-  5 
36.0 
49.2 

10.  7 
14.7 

47-7 
56.1 


5-7 

53-1 
51.0 

40.  I 
39-5 

10.3 
6.4 

70.8 
74.  2 

52.8 

51.2 


Differ- 
ence 
between 
dupli- 
cates. 


O.  O 
6.8 


17.  I 

9.  I 

13-2 


4.0 


8.4 


5-7 
2.  I 


0.6 


3-9 
3-4 
1.6 


5-8 


In  order  to  compare  the  two  seasons'  crops  on  a  uniform  basis  and 
also  to  allow  for  the  mean  and  maximum  difference  in  duplicates,  all 
results  of  the  yield  of  grain  have  been  calculated  to  the  percentage  of 
variation  from  the  highest  yield.  The  figures  given  for  the  duplicate 
containers  have  been  substracted  and  expressed  in  a  separate  column. 
The  difference  between  soils  3A  and  3B  was  the  maximum,  amounting 
to  17. 1  per  cent,  while  that  of  soils  5A  and  5B  was  13.2  per  cent.  The 
mean  figure  given  by  the  above  was  5.8  per  cent.  These  differences 
have  been  expressed  diagrammatically  in  figure  7.  Here  it  is  seen  that, 
by  taking  the  maximum  variation  between  duplicates  as  the  standard, 
soils  8,  11,6,  and  i  have  practically  an  equally  high  production.  Soil  2 
is  significantly  lower,  and  the  others  range  lower,  down  to  soil  12. 


342 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  6 


In  figures  8  to  20  it  is  seen  that  the  graphs  given  on  the  duplicate  con- 
tainers agree  excellently.  In  all  the  soils  except  No.  9,  12,  and  7  there 
was  an  appreciable  increase  of  potassium,  which  occurred  from  four  to 
six  weeks  after  the  time  of  planting.  In  all  except  the  same  soils,  and 
in  addition  No.  14,  there  was  a  liberation  of  nitrates  which,  like  the 
potassium,  began  to  fall  soon  after  the  crop  was  rapidly  developing. 

The  method  employed  for  calcium  had  proved  so  inaccurate  that  the 
results  were  excluded  from  the  final  charts.  It  was  known  that  the 
potassium  method  had  a  high  factor  of  error  in  the  lower  concentrations. 
A  definite  increase  of  nitrates  had  been  shown  in  the  earlier  periods  of 
growth.  It  was  very  clear  that  before  any  interpretation  of  these  figures 
could  be  made  it  was  necessary  to  have  not  only  more  accurate  methods 


100  8A 


ot,h'jKjl^    liiio 


5jElft;^o;i  1916 


Fig.  7. — Graphs  of  the  yield  of  grain  in  1915  and  19:6,  expressed  as  a  percentage  of  the  maximum  yield. 
The  heavily  shaded  portions  in  1915  represent  difference  between  duplicates. 

of  determination  but  also  a  comparison  between  the  conditions  which 
were  found  in  the  planted  soil  and  those  which  would  have  occurred  in 
the  same  soil  had  the  crop  not  been  present. 

The  first  season's  work  therefore  established  the  following  facts: 
(i)  That  a  satisfactory  duplicate  crop  could  be  grown  under  the  con- 
ditions employed;  (2)  that  the  extraction  technic  followed  yielded  prac- 
tically duplicate  graphs  for  the  elements  extracted.  The  period  of  a 
year's  treatment  under  uniform  conditions  had  also  tended  to  bring  the 
soils  into  normal  relation  to  each  other.  It  had  given  an  opportunity 
for  the  soils  to  recover  from  the  sifting  and  handling  which  they  had 
received  and  be  more  comparable  to  a  varied  group  of  field  soils  exposed 
to  the  same  climatic  conditions. 


Feb.  II.  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  343 


j^  ii  ^  ^1  'I  i^  i<.^ 


Fig.  8. — Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  i,  Yolo  silty  clay  loam. 


344 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  6 


SO//L    2/^ 


so 
so 

TO- 


SO/A  2B 


»^     *|     *|     5|    «!^     't'!      ifX  |ii! 


Hi 


5|    <o|    §1    <V^    ^^    !?"(  \ 


Fig.  9. — Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  2,  Yolo  silt  clay  loam. 


Feb.  IX.  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract 


345 


j>o 

eo 

70 

eo 

so 

l\ 

'fO 

1  \ 

JO 

'  V^'      '' 

^o 
/o 

/^■•An 

-^^^^^^^^, 


so 
eo 

TO 


SO/A  33 
C/?OP  B/9/?A£:y 


^ 


^   ^   =^   O   ^   t  S   b   ^ 


SO/L  Oi^ 

/9/e 


%t  tt  t\t  \t\  m 


SO/L  3  s 
C/?OP    /VOA/£ 


nTiiriiniTp 


Fig.  lo.-Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  3.  Yolo  silty  clay  loam. 

27809°— 18 4 


346 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  6 


/9/S 


I  %  %  %  %  % 

^    M^     ?l     ^     «^     (!^ 


Fig.  II.— Graphs  of  the  seasonal  studies  of  the  water  extract  o  fsoil  4.  Yo'o  silty  clay  loam. 


Feb.  II.  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract 


347 


sSO//.  ^y^ 


^e2^ts^^^^_ 


I  ^^  I  i  1^  1^  1^:  i 

^  ^l  >|  ^  ^1  ^1  ^1  ^^ 
Fig.  12.— Graphs  of  the  seasonal 


studies  of  the  water  extract  of  soil  5,  Yolo  silty  clay  loam. 


348 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  6 


.SO/L    s/i 
C/?0/=>   B/^/?L£Y 


...  /y/>gvg3it//Y/%y 


SO/^  SB 

C/?0/=>    B/^/?L£y 

/3/S 


<\3    ;'    ^'    <\i    ^    -^    -M    ")    5;    -vi      &fi      <o 

r^  u  II  u  §  §  ^1  § 


^.  — ? — 

§g  5i^  ^  5S|  ^ 

.—Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  6,  Yolo  clay  loam 


Feb.  II.  1918  EJJect  of  Season  and  Crop  Growth  on  Soil  Extract  349 


ao-. 

70-: 


J^^^M^ 


■      ■-     /   \ 


/  "^•^---. 


Por^ssiU^M- • 


.  ^fA6A/CS/(//T  ^  f) |_. 


I 


9S^cyw7j^^^ 


'— '">r'^^.?'^-"/^-i-dir 


1^-  y  1^  |5 1^  |5 1^  i|5^ 


%^  ?|  5|  ^?  ?^  ^f  i^  ^? 


Fig.  14.— Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  7,  Hanford  fine  sandy  loam. 


350 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  6 


--aj5^5 


I  ^  ^  ^   ^  i^  ;o 

^    ^     ^     ^     ^     '^     <>i     »!     O     O 


so/L  as 


1     I     ^     ^     $     I     n     ^     ^ 


k     k 


30/L  a/i 
C/?Of^  B/^/?^£y 
/9/6 


\  §5    \%    SS   ^   55   %   %    |5i 
^  *»?    »?    *|    5|    5^    5l|   !t^    S| 


FlO.  IS. — Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  8,  Fresno  fine  sandy  loam. 


Feb.  II.  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  351 


so 
so 

70- 

eo 
■50 


/9/e 


5^  ^^ 


5)^  i)^  b^  5^  i!^  b^  l§S 

»^  ^  ^1  ^1 1^  1^  il 

1?^ 


so- 
eo-- 

ZO-- 
60-- 
.30- ■ 

so- 

20 

/o 

o 


.A 


/9/6- 


/y 


s 


1 1 8  i  I  §  ^i  I 

^^    H    *|    ^1    ^^    !^^    ^^    i^^ 


Fig.  16.— Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  9,  Kimball  fine  sandy  loam. 


352 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  6 


W3J 


^  ^  ^  5^  55  k-  k:^  ^• 


i  ^  ^ 


so/z.  /o^ 


% 


>SO/L  /OS 
C/?0/=>    A/OAZ/T 


,"^^^4^ 


<^-"^/0rt(^-" 


^^Ci-J  ^.'^Z-  — 


^  ^^  i^  ^  ^  fi  ^  '^^  iS; 
S  «^  *v  li  1^  «\  1^  *^  % 
'^  *.|  »|  ^  ^1 1|  ^^  ^  ^1 


Fig.  17. — Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  10,  Tejtinga  fine  sandy  loam. 


Feb.  II.  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  353 


SO/L  ///I. 


I  1^  ^vl  1^  1:  ^:  ^'v  S  ^S  b  !^  a 

^    n|  >|  S|  *|  ^  5i|  ^  *|  ^  ^  i^ 

Fig.  i8.-Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  ii,  Madera  fine  sandy  loam. 


354 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  6 


so 

TO 
60 
SO 
fO 
JO 
20 
/O 


SO//L  ZP.^ 


..—-^SOis^ji^,  .por^s^'iT^— 


•SO/L  /2'0 


II  ^  q  ii  1^  1^  ^?  p  ^^ 


SO/L/^^ 


%  ^1  ^1  I?  ^1  f^ 

Fig.  19. — Graphs  of  the  seasonal  studies  of  water  extract  of  soil  la,  Arnold  fine  sandy  loam. 


Feb.  II,  i9i8  Efject  of  Season  and  Crop  Growth  on  Soil  Extract  355 


§5    ^X      *S     *S     ^jj     5Cnj     :to     ■*to    irt&j 

S5  ^^    „^    10^    <b^    5^    5,^    5S    || 


SO/A  MS. 
C/?0/=>    A/OA/E. 

/sue 


Fig.  20.— 


g  12  g  sj  1^  15  i^  r 
^  ^  ?|  1^  ^^  ^  5^  ^^ 

Graphs  of  the  seasonal  studies  of  the  water  extract  of  soil  14.  Standish  fine 


sandy  loam. 


356  Journal  of  Agricultural  Research  voi.  xii,  no.  6 

SECOND   SEASON,  1916 

In  the  season  of  191 6  one  container  of  each  soil  was  again  planted  to 
the  same  strain  of  Beldi  barley,  and  the  duplicate  was  at  all  times  treated 
in  the  same  manner,  except  that  no  crop  was  grown  upon  it.  One  week 
before  planting,  the  top  soil  to  a  depth  of  8  or  10  inches  was  forked  up 
and  put  in  an  excellent  state  of  tilth.  The  soils  were  again  sampled 
every  two  weeks  from  the  time  of  planting  in  May  until  the  crop  was 
harvested  in  August.  During  the  succeeding  fall  and  winter  samples 
were  taken  at  approximately  8-week  intervals. 

The  analyses  of  the  water  extracts  were  performed  by  the  modified 
methods  which  have  been  previously  outlined.  During  the  latter 
portion  of  the  sampling  season  cooperative  work  was  performed  by 
Hoagland  on  the  samples,  using  the  freezing-point  method. 

An  extremely  uniform  stand  of  barley  was  obtained  in  all  containers. 
Shortly  after  sprouting,  the  plants  were  thinned  to  one  vigorous  seedling. 
The  growth  throughout  the  season  was  steady  and  vigorous.  A 
diagrammatic  representation  of  the  height  at  each  soil-sampling  period 
is  given  in  figure  21.  It  will  be  seen  that  the  most  rapid  period  of 
growth  was  from  the  fourth  to  the  tenth  week.  By  that  time  the 
plants  had  almost  attained  their  full  height,  and  the  heads  of  grain 
were  beginning  to  be  formed.  At  the  close  of  the  twelfth  week  the 
maximum  growth  in  height  had  been  attained.  Five  weeks'  additional 
time  was  required  for  complete  development  and  ripening.  It  was 
interesting  to  note  that  in  these  studies,  where  moisture  was  never  a 
limiting  factor,  the  growth  in  height  was  extremely  uniform,  even  though 
the  dry  matter  produced  did  show  great  differences. 

The  results  of  the  crop  yield  in  total  crop  and  grain  are  given  in 
Table  IX.  The  grain  yield  has  again  been  calculated  to  pounds  and 
bushels  per  acre,  and  the  percentage  of  variation  from  the  maximum 
yield  is  shown. 


Feb.  II,  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  357 


5QLL5 


I/1CHE5 
5£1 


I 

VK5.  PLA/{TL]7     ^        4        6        6       10       12.       14  t^EACHtP  A/^X.  HT. 

Fig.  21. — Graphs  of  the  growth  of  crops  in  height,  season  of  1916. 
Table  IX. — Crop  yield  in  igi6 


Soil. 


Yolo  silty  clay  loam : 

No.  lA 

No.  2 A 

No.  3  A 

No.  4A 

No.  5 A 

Yolo  clay  loam ,  6A 

Hanford  fine  sandy  loam,  7A.  .  . 

Fresno  fine  sandy  loam,  8 A 

Kimball  fine  sandy  loam,  qA.  ..  . 
Tejunga  fine  sandy  loam,  loA.  . . 
Madera  fine  sandy  loam,  iiA. . . . 
Arnold  fine  sandy  loam,  12A  .  .  . 
Standish  fine  sandy  loam,  14A .  . 


Total  yield 

of  air-dry 

grain  and 

straw. 


Grams. 
1,474 
1,530 
970 
1.252 
1,652 

1.523 
1,414 

I.  719 

946 

I,  266 

1.547 

970 

1,464 


Grain. 


Grams. 

655 

439 
549 
690 
670 
541 
679 

357 
551 
625 

390 
630 


Grain 
(pounds 
per  acre). 


5.  T92 
5,038 

3,377 
4.223 

5.307 
5,153 

4,  161 

5,  222 

2,  746 
4,238 
4,807 

3,  000 
4,846 


Grain 
(bushel.s 
per  acre). 


85 
69 
87 

45 
70. 
80, 
50.  o 
80.8 


Variation 
from  maxi- 
mum yield 
of  grain. 


Per  cent. 

2.  2 

5-0 

36.4 

20.  4 

.  o 

2.8 
22.  5 

1.6 
48.  2 
20.  I 

9.4 

43-4 
8.6 


358  Journal  of  Agricultural  Research  voi.  xii.no.6 

The  figures  obtained  in  the  preceding  season  for  the  mean  and  maxi- 
mum variation  between  duplicates  will  be  used  in  order  to  find  w^hat  will 
constitute  a  significant  difference  in  yield.  By  taking  the  mean  varia- 
tion as  the  standard,  soils  5 A,  8A,  lA,  6A,  and  2A  are  to  be  considered 
equally  productive,  while  iiA  and  14A  are  slightly  lower.  If,  however, 
the  consen/ative  figure  of  the  maximum  variation  between  duplicates  is 
used,  it  will  be  seen  that  No.  iiA  and  14A  will  also  fall  in  the  group  of 
the  highest  yield.  These  relations  are  presented  diagrammatically  in 
figure  7. 

It  will  be  seen  that  six  soils  were  significantly  lower  than  the  group  of 
highest  productivity.  Of  those  No.  loA,  4A,  and  7 A  were  practically 
equal,  while  3A,  1 2 A,  and  9A  were  distinctly  low  in  yield. 

The  results  of  the  water  extractions  are  again  presented  in  the  form  of 
graphs.  In  figures  8  to  20  the  plot  marked  "A"  is  in  each  case  the  planted 
portion,  while  the  uncropped  duplicate  is  called  "B."  Some  extremely 
striking  differences  are  exhibited  by  these  graphs. 

In  the  planted  soils  the  water-soluble  nutrients  at  the  beginning  of  the 
season,  either  remained  on  practically  the  same  level  or  increased  slightly 
for  the  first  four  weeks.  Then,  without  exception,  the  nitrogen  com- 
menced to  decrease  rapidly  and  was  followed  in  a  smaller  degree  by  the 
calcium  and  potash,  and  very  slightly  by  the  magnesium.  The  con- 
trast shown  by  the  unplanted  soil  was  equally  uniform  in  nature,  though 
variable  in  the  extent  of  the  effect.  In  almost  all  the  soils  the  effect  of 
the  cultivation  was  to  cause  a  considerable  liberation  of  soluble  nutrients 
and,  though  there  was  later  shown  to  be  a  depression  from  this  high 
figure,  yet  the  general  range  of  all  the  nutrients  except  the  phosphates 
continued  to  be  higher  in  the  uncropped  soil. 

The  soils  which  did  not  show  this  stimulation  and  liberation  of  nutri- 
ents were  No.  3B,  9B,  and  12B.  These  were  the  least  productive  soils 
of  the  group  and  also  had  the  lowest  range  of  soluble  nutrients. 

It  is  extremely  significant  that  this  same  period  of  high  soluble  nutri- 
ents in  the  uncropped  soils  corresponds  to  the  period  of  lowered  nutrients 
in  the  cropped  duplicates.  It  was  also  the  period,  as  may  be  observed 
in  figure  21,  in  which  the  plants  were  making  their  most  rapid  growth. 

The  one  compound  which  did  not  exhibit  this  liberation  of  excess 
nutrients  in  the  uncropped  soil  was  the  phosphate.  There  were  large 
differences  between  the  amounts  present  in  various  soils,  but  for  any  one 
soil  the  amount  was  practically  constant  in  the  cropped  and  uncropped 
plot.     It  was  the  only  soil  nutrient  studied  which  behaved  in  this  manner. 

All  the  above  differences  between  the  planted  and  unplanted  soils 
which  have  been  noted  are  greater  than  can  be  accounted  for  by  the 
probable  factor  of  error  which  has  been  determined  in  an  earlier  portion 
of  the  study.  The  maximum  figure  for  any  nutrient  in  the  concentration 
found  is  10  per  cent,  and  that  applies  only  to  potassium  and  calcium. 
On  applying  this  correction  plus  or  minus  to  the  periods  of  greatest  de- 


Feb.  II.  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  359 

pression,  it  is  clear  that  this  will  not  account  for  the  differences  noted. 
With  phosphates  the  probable  maximum  is  4  to  6  per  cent,  while  with 
magnesium  it  would  vary  from  6  to  10  per  cent.  All  these  corrections 
fail  to  change  significantly  the  figures  recorded. 

At  two  periods  during  the  season  of  191 6  daily  studies  of  the  water 
extract  were  made  with  soil  8,  and  at  one  period  with  soil  i.  The  result 
of  these  tend  to  show  that  considerable  fluctuations  may  take  place  from 
day  to  day.  The  results  of  these  two  studies  are  plotted  in  figures  22 
to  24.  It  will  be  seen  that  the  fluctuations  in  general  occur  simulta- 
neously in  both  containers  and  so  will  not  alter  the  relationship  of  the 
graphs  for  the  planted  and  unplanted  soils.  They  show,  however,  that 
small  differences  in  the  range  of  the  graphs  can  not  be  considered  sig- 


f.F.A.-PRY  iOlL 

100 

90 

A          Soil  15 

(SO 

A  /  \    A 

70 
60 

50ILfA 

y                   \ .'hnoi. 

50 
40 
30   \     A 

10  y- 

0  ^r-r 

yY.. 
v^*""-— ^      '\F04. 

AvQ.  7  0  9 

10  II  11     14  15  It  17            l\ 

AvQ.7  6  9  10  11  12     1415  %  IT           U 

Fig.  22. — Graphs  of  the  daily  studies  of  the  water  extract  of  soils  lA  and  iB,  season  of  1916. 

nificant.  The  general  range  of  the  graphs  is  believed  to  represent  an 
average  figure,  but  small  variations  between  soils  should  always  be  dis- 
regarded. This  would  also  indicate  that  an  expression  of  the  differences 
in  water-soluble  material  can  only  be  given  by  a  series  of  related  obser- 
vations. 

DISCUSSION  AND  DEDUCTIONS 

In  the  foregoing  studies  it  is  believed  that  the  limitations  as  well  as 
the  possibilities  of  the  experimental  methods  have  been  pointed  out. 
It  is  earnestly  desired  that  no  deductions  should  be  drawn  which  are  not 
conservatively  justified  by  the  data  presented. 

It  should  be  noted  that  the  results  have  been  obtained  on  13  soils 
which  comprise  only  two  distinct  soil  types.     These  types  are,  however, 


36o 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  6 


P.PA  PRY  50iL 
100 


90 


501L  6A 


m 

60 

50 

\ 

40 
3Q 

/ 

V 

20 

A  A 

1 

iQ 

I 

•J'' 

501L  SB 


CaLCIV/A   (Co.') 


>^^-  -  POTA551Vy<\  (  K) 


N  /     .  / 


\  /V 


\ 


"■--.  PH03PHATL  (VOi).  .,.■•*■■•——■— sj^--'---^-v 


^    I    I    I    I    I    I       I    I    t    I    I    i 
Jsjnr.  \l  15  II 15  16  17     19  m  UliaU    £6 


/)lTROqL/MC/lrt03J 


I    I    I    I   I    I    I  I    I  I    I    I    I 
Jv/tE  12  13  U  15  16  17  10  19  ZO  a  22  i3  IH 


-P04 

—  Alg 

I 
lb 


Fig.  23.— Graphs  of  the  daily  studies  of  the  water  extract  of  soils  8A  and  8B,  July,  1916. 


V.Yt' 

\.  I7KY  501L 

100 

90 

eo 

50IL6A 

70 

60 

,\ 

1  \ 

/    \ 

50 

K     I          \ 

l\   1               ^Ca 

40 

A            /  \/ 

^^'  \        'h^  . 

30 

y     .,   \ -'a 

....-•■.,/          /\^.K..._\^A\9- 

10 

'  ..^x     A  r\  ^po^ 

'••rK. 


""•■•PO4. 


AvQ.  7  8  9  10  II  12      14  15  Ife  \1  i\  AvQ.  7  8  9   10  II  U     14  15  \(>  17  il 


Fig.  34.— Graphs  of  the  daily  studies  of  the  water  extract  of  soils  8A  and  SB,  August,  1916. 


Feb.  II,  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  361 

very  dissimilar  in  their  nature;  the  silty  clay  loams  are  representative 
of  decidedly  complex  soils  with  appreciable  amounts  of  clay  and  other 
colloids,  while  the  fine  sandy  loams  belong  to  the  light  open  soils  which 
are  common  in  the  West.  The  extraction  graphs  yielded  by  both  types 
are  so  similar  that  it  is  believed  that  comparisons  can  safely  be  made 
both  within  the  types  and  between  the  two  groups. 

The  extractions  of  the  season  of  191 6  are  those  to  which  the  greatest 
importance  is  attached  and  are  those  which  will  largely  be  considered 
in  this  discussion.  The  previous  year's  work  brought  the  soils  into 
comparable  condition  and  also  showed  the  probable  limits  of  the  agree- 
ment between  duplicates.  It  is  believed  that  the  results  obtained  in 
this  second  season  throw  some  light  on  several  disputed  points.  The 
method  of  constant  comparison  between  the  cropped  and  uncropped 
soil  is  considered  essential  for  the  success  of  such  a  study.  Only  in  this 
manner  can  any  conception  of  the  inherent  capacity  of  a  soil  be  obtained. 
When  a  large  crop  is  growing  on  a  soil,  it  is  possible  for  it  to  affect  the 
soluble  nutrients  so  that  the  extract  given  by  it  will  be  equal  to  a  moder- 
ately good  or  fairly  poor  soil. 

Such  a  condition  is  seen  with  three  fine  sandy  loams:  No.  7,  Hanford 
fine  sandy  loam,  a  soil  of  moderate  production;  No.  8,  Fresno  fine  sandy 
loam,  a  very  good  soil;  and  No.  9,  Kimball  fine  sandy  loam,  a  soil  which 
has  produced  poor  crops. 

The  nutrients  extracted  from  each  of  these  three  soils  when  cropped 
give  practically  duplicate  graphs.  But  with  the  unplanted  duplicate 
striking  differences  are  seen.  The  better  soils  all  show  greater  differ- 
ences between  the  cropped  and  the  uncropped  soils  than  do  the  three 
poorest  soils,  No.  3,  9,  10,  and  12.  With  these  three  it  is  possible  that 
plant  food  is  the  limiting  factor,  but  with  certain  soils  of  intermediate 
production,  such  as  No.  4  and  10,  it  is  less  evident  why  these  should 
not  fall  in  the  most  productive  group.  It  is  claimed,  however,  that 
this  method  of  study  gives  an  expression  of  the  inherent  capacity  of  the 
soil  to  produce  water-soluble  plant  food. 

Between  these  three  poor  soils  and  the  group  of  highest  production 
there  are  large  differences  in  the  range  of  soluble  nutrients.  These 
three  soils  would  not  be  classed  as  unfertile  soils  in  general  farming. 
The  only  one  which  had  a  small  crop  upon  it  when  the  original  sample 
was  collected  was  soil  No.  12,  Arnold  fine  sandy  loam.  The  oats  grow- 
ing on  this  body  of  soil  were  noticeably  smaller  than  in  other  portions 
of  the  field.  Under  the  controlled  conditions  of  the  experiment  the 
differences  between  these  soils  and  the  more  productive  must  be  con- 
sidered significant,  since  they  have  been  among  the  poorest  soils  each 
year.  It  may,  therefore,  be  stated  that  among  the  10  soils  studied  the 
3  lowest  in  productivity  also  show  the  lowest  inherent  capacity  to  furnish 
soluble  nutrients. 
27809°— 18 5 


362  J our7ial  of  Agricultural  Research  voi.  xii,  no.  6 

This  conclusion  is  notably  at  variance  with  the  earlier  statements  of 
Whitney  and  Cameron  (61).  It  is  believed  that  the  experimental  method 
which  has  been  followed  brings  out  the  relationship  between  different 
soils  with  unusual  clarity  and  effectiveness. 

The  influence  of  the  crop  on  the  soluble  nutrients  is  another  point 
which  has  been  denied  by  the  same  authors.  This  is  shown  by  the 
following  quotation  from  Bureau  of  Soils,  Bulletin  22 : 

At  the  same  time  we  have  detected  no  constant  decrease  in  the  amount  of  soluble 
salts  which  could  be  easily  detected  with  the  methods  used  during  the  advance 
stages  of  growth  of  the  crop,  notwithstanding  the  considerable  withdrawal  there  must 
be  by  the  plants  {61,  p.  60). 

The  contrast  between  the  curves  of  the  planted  and  unplanted  soils  is 
evident  at  the  most  cursory  glance.  It  is  also  clearly  shown  that  the 
soil  did  not  immediately  recover  and  yield  the  same  quantities  of  soluble 
nutrients  as  the  uncropped  plot,  even  after  the  crop  was  recovered.  At 
the  last  observation  recorded,  on  February  12,  there  was  still  an  appre- 
ciable difference  between  the  duplicate  containers.  In  this  connection 
it  is  pertinent  to  refer  to  the  data  obtained  in  a  preliminary  study  in 
which  a  cropped  soil  was  extracted  with  a  diluted  solution  from  an 
uncropped  portion  of  it.  In  that  study  it  was  shown  that  this  diluted 
solution  exerted  a  depressing  effect  on  the  nutrients  extracted  from  the 
cropped  soil.     This  has  an  important  application. 

At  the  period  of  greatest  growth  of  the  crop,  the  water  extract  of  the 
cropped  soil  approached  much  more  closely  to  distilled  water  in  the  case 
of  the  planted  soil  than  it  did  with  the  unplanted  duplicate.  In  conse- 
quence of  this  fact,  a  greater  portion  of  the  extract  obtained  from  it  was 
from  the  solution  of  the  actual  soil  minerals  than  was  the  case  with  the 
uncropped  soil.  The  real  difference  between  the  two  duplicate  con- 
tainers was  therefore  greater  than  the  graphs  would  indicate.  They  can 
only  be  taken  as  the  minimum  difference  which  actually  existed. 

The  data  obtained  by  Hoagland,  using  the  freezing-point  method, 
corroborate  this  difference  between  the  planted  and  unplanted  soils. 
He  has  also  observed  that  soils  9,  Kimball  fine  sandy  loam,  and  12, 
Arnold  fine  sandy  loam,  are  notably  lower  than  any  of  the  other  soils  in 
the  concentration  of  their  solutions.  This  coincided  with  the  observa- 
tions made  from  a  study  of  the  water  extracts.  Considering  soil  3, 
Yolo  silty  clay  loam,  the  next  lowest  soil  in  water  extract,  he  could  not 
draw  any  definite  conclusion. 

Jensen  {32)  in  his  studies  on  sugar  beets  has  described  a  decrease  in 
water-soluble  nutrients  and  has  also  found  this  less  noticeable  in  the  case 
of  phosphates.  King's  {35)  figures  also  showed  some  decrease  in  soluble 
compounds  as  the  crop  advanced  in  growth.  Harris  and  Butt  {24)  have 
observed  differences  in  soluble  salts  and  nitrates  between  cropped  and 
fallow  soils.  It  is  believed,  however,  that  the  use  of  the  unplanted  dupli- 
cate soil  and  the  periodic  observations  made  upon  it  and  the  cropped 


Feb.  II.  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  363 

container  have  given  a  better  expression  of  the  potential  power  of  the 
soil  than  has  been  obtained  in  the  past. 

The  study  of  the  unpTanted  soil  is  especially  valuable  because  of  the 
definite  information. which  it  furnishes  in  regard  to  the  effect  of  fallowing 
and  cultivation.  From  the  results  of  the  present  investigation  it  can 
be  stated  that  the  changes  occurring  in  water-soluble  compounds,  both 
organic  and  inorganic,  as  the  result  of  these  practices,  are  great  and  far- 
reaching  in  effect.  They  entirely  justify  the  importance  attached  to 
such  treatment  in  the  past.  In  biennial  cropping,  alternating  with 
fallowing,  as  practiced  in  California,  it  is  probable  that  fully  as  much  is 
gained  from  the  increase  of  water-soluble  nutrients  as  from  the  moisture 
stored  up  in  the  soil. 

In  the  preceding  studies  the  data  obtained  have  been  considered  solely 
as  a  measure  of  the  water-soluble  nutrients  obtained  by  a  conventional 
procedure  of  extraction.  From  the  work  performed  by  Hoagland  in 
collaboration  the  amounts  extracted  are  definitely  related  to  the  actual 
soil  solution.  Even  though  the  figures  so  obtained  do  indicate  a  range 
of  concentrations  higher  than  the  actual  truth,  it  has  been  previously 
pointed  out  that  this  does  not  alter  the  relationships  which  have  been 
established.  It  is  therefore  believed,  with  this  corroborative  evidence, 
that  the  changes  observed  in  the  water  extract  reflect  actual  changes  in 
the  soil  solution. 

SUMMARY 

(i)  The  water-soluble  nutrients  in  13  soils,  of  two  different  types, 
have  been  periodically  determined  during  two  seasons. 

(2)  Throughout  the  second  season  comparisons  were  made  between 
the  planted  soil  and  its  uncropped  duplicate. 

(3)  Notable  differences  were  observed  between  the  nitrates,  calcium, 
potassium,  and  magnesium  present  in  the  water  extracts  from  the 
cropped  and  uncropped  soils. 

(4)  The  phosphates  did  not  exhibit  corresponding  differences.  Great 
dissimilarities  were  observed  in  the  phosphate  content  of  different  soils, 
but  in  any  one  soil  the  amount  was  practically  constant  in  both  the 
cropped  and  uncropped  plot. 

(5)  Striking  differences  occurred  between  the  soluble  nutrients  present 
in  the  various  uncropped  soils. 

(6)  While  the  crops  were  growing,  the  concentrations  of  nutrients  in 
8  of  the  13  planted  soils  were  practically  the  same.  These  8  included 
both  good  and  poor  soils, 

(7)  The  three  poorest  soils  yielded  the  smallest  amounts  of  water- 
soluble  nutrients  and  the  smallest  differences  between  the  cropped  and 
uncropped  duplicates. 


2^^ A  Journal  of  Agricultural  Research  voi.  xii,  no.  6 

(8)  The  comparisons  between  the  planted  and  unplanted  duplicates 
furnished  valuable  indexes  of  the  inherent  capacities  of  the  soils  to  pro- 
duce nutrients. 

(9)  The  accuracy  of  the  methods  of  analysis  an,d  of  the  extraction 
procedure  employed  was  determined,  and  the  mean  and  maximum  errors 
involved  were  estimated. 

(10)  The  amounts  of  the  water-soluble  nutrients  obtained  by  varying 
the  ratio  of  soil  to  water  were  studied.  The  relationship  of  the  com- 
pounds extracted  did  not  change  essentially  in  the  lower  concentrations. 

(11)  By  comparison  with  freezing-point  determinations  the  concen- 
tration of  the  soil  solution  calculated  from  the  water  extract  was  ^hown 
to  be  from  two  to  four  or  five  times  as  great  as  the  actual  soil  solution. 

(12)  Variations  in  the  water  extract  were  correlated  with  variations 
in  the  freezing  points  of  the  same  samples  of  soil. 

(13)  From  the  results  of  the  freezing-point  determinations  it  is  con- 
cluded that  variations  in  the  water  extract  reflect  actual  changes  in  the 
soil  solution. 

(14)  The  results  of  the  investigation  show  that  large  amounts  of 
water-soluble  nutrients  are  developed  by  cultivation,  fallowing,  and  bi- 
ennial cropping,  and  demonstrate  the  soundness  of  these  practices. 

LITERATURE  CITED 
(i)  American  Public  Health  Association.    Laboratory  Section. 

1912.  STANDARD    methods   FOR   THE    EXAMINATION    OP   WATER     AND    SEWAGE. 

ed.  2,  144  p.     Bibliography,  p.  137-140. 

(2)  BallEnEGGER,  Robert. 

1913.  SOIL  STUDIES  BY  THE  AID  OF  THEIR  WATER  SOLUTIONS.      (Abstract.)      IN 

Exp.   Sta.   Rec,  v.   30,   no.   6,   p.  516.     1914.     (Original  article   in 
Foldtani  Kozlony,  v.  43,  no.  7/9,  p.  359-366.      1913.     Not  seen.) 

(3)  BoGUE,  R.  H. 

1915.  THE  ADSORPTION  OF  POTASSIUM  AND  PHOSPHATE  IONS  BY  T\TICAL  SOILS  OP 

THE  CONNECTICUT  VALLEY.     In  Jour.  Phys.  Chem.,  v.  19,  no.  8,  p.  664- 
695,  13  fig. 

(4)  BouYOucos,  G.  J. 

191 7.    MEASUREMENT  OF  THE  INACTIVE,  OR  UNFREE,   MOISTURE  IN  THE  SOIL  BY 

MEANS  OF  THE  dilatomETER  METHOD.     In  Jour.  Agr.  Research,  v.  8, 
no.  6,  p.  195-217,  I  fig.     Literature  cited,  p.  217. 

(5)  and  McCooL,  M.  M. 

1916.  THE  FREEZING-POINT  METHOD  AS  A  NEW  MEANS  OF  MEASURING  THE  CON- 

CENTRATION OF  THE  SOIL  SOLUTION  DIRECTLY  IN  THE  SOIL.      In  Mich. 

Agr.  Exp.  sta.  Tech.  Bui.  24,  p.  592H531,  2  fig. 

(6)  Briggs,  L.  J. 

1902.   FILTRATION  OF  SUSPENDED  CLAY  FROM  SOIL  SOLUTIONS.      In  U.   S.   Dept. 

Agr.  Bur.  Soils  Bui.  19,  p.  30-40,  fig.  5. 
(7) and  McCall,  A.  G. 

1904.    AN  artificial  ROOT   FOR   INDUCING    CAPILLARY  MOVEMENT  OF  SOIL  MOIS- 
TURE.    In  Science,  n.  s.,  v.  20,  no.  513,  p.  566-569. 


Feb.  II,  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  365 

(8)  Briggs,  L.  J.,  and  Shantz,  H.  L. 

1912.   THE    WILTING    COEFFICIENT    FOR    DIFFERENT   PLANTS    AND    ITS    INDIRECT 
DETERMINATION.     U.  S.  Dcpt.  AgT.  BuT.  Plant  Indus.  Bui.  230,  83  p., 

9fig- 
(p) Martin,  F.  O.,  and  Pearce,  J.  R. 

1904.   THE  centrifugal  METHOD  OP  MECHANICAL  SOIL  ANALYSIS.      U.  S.   Dept. 

Agr.  Bur.  Soils  Bui.  24,  38  p.,  7  fig.,  2  pi. 
(10)  Cameron,  F.  K. 

1910.    AN  INTRODUCTION  TO  THE  STUDY  OF  THE  SOIL  SOLUTION.      In  Jour.  Phys, 

Chem.,  V.  14,  no.  4,  p.  320-372,  3  fig. 

(II) 

(12) 


191 1.    THE  SOIL  SOLUTION,   THE   NUTRIENT   MEDIUM  FOR  PLANT  GROWTH.      136 

p.,  3  fig.     Easton,  Pa.,  London. 


1912.   THE    r6lE    op    THE    LYSIMETER    IN    SOIL    SOLUTION    STUDIES.      In    Orig. 

*  Commun.  8th  Intemat.  Cong.  Appl.  Chem.,  v.  15,  p.  49-50. 

(13)  CossA,  Alphonse. 

1866.    UEBER    DIE    BESTIMMUNG    DER    IM    WASSER    LOSLICHEN    BESTANDTHEILE 

DER  ACKERERDEN.  In  Landw.  Vers.  Stat.,  Bd.  8,  p.  54-59. 

(14)  Eichhorn,  H. 

1863.    UEBER   DIE   AUFNAHME   DER   NAHRUNGSMITTEL   AUS    DEM   BODEN   DURCH 

WASSER.     (Abstract.)    In  Jahresber.  Agr.  Chem.,  Jahrg.  4,  1861/62, 
p.  3-7.     Original  articles  cited  in  footnotes. 

(15)  Engels,  O. 

1912.  die  wirkung  einiger  losungmittel  aup  die  im  boden  enthaltenen 

pflanzennahrstoffe;  phosphorsaure,    kali    und    kalk    im    ur- 

SPRUNGLICHEN  UND  ABSORPTIV  GEBUNDENEN  ZUSTANDE.      In  Landw. 

Vers.  Stat.,  Bd.  77,  Heft  3/4,  p.  269-304. 

(16)  Gedroitz,  K.  K. 

1906.  variability  op  the  concentration  of  the  soil  solution  and  op  the 
content  of  the  soil  op  easily  soluble  compounds  in  dependence 
ON  EXTERNAL  CONDITIONS.  (Abstract.)  In  Exp.  Sta.  Rec,  v.  19, 
no.  4,  p.  315.  1907.  (Original  article  in  Zhur.  Opuitn.  Agron.  [Russ. 
Jour.  Expt.  Landw.],  v.  7,  no.  5,  p.  521-561.     1906.     Not  seen.) 

(17)  GiLTNER,  Ward. 

1913.  REPORT  OF  THE  BACTERIOLOGIST.     In  Mich.  Agr.  Exp.  Sta.,  52d  Ann. 

Rpt.,  1912/13,  p.  149-166. 

Quotes  Van  Suchtelen,  p.  149-155. 

(18)  Cola,  Giuseppe. 

1905.   STUDI  SUI   RAPPORTI   TRA  LA   DISTRIBUZIONE   DELLE  PIANTE   E  LA  COSTI- 

TuziONE  Fisico-CHiNiCA  DEL  suoLO.    In  Ann.  Bot.  [Rome],  v.  3,  fasc. 

3.  P-  455-512,  pi.  13. 
(19) 

(20) 


1910.  SAGGio  Di  UNA  TEORiA  OSMOTICA  dell'Edafismo.    In  Ann.  Bot.  [Rome], 
V.  8,  fasc.  3,  p.  275-548,  I  fig.,  pi.  13-14. 


(21) 


1911.    OSSERVAZIONI  SOPRA  1  LIQUIDI  CIRCOLANTI  NEL  TERRENO  AGRARIO.      In 

Gior.  Geol.  Prat.,  anno  9,  fasc.  3/4,  p.  131-159. 

Also  in  Ann.  R.   Accad.  Agr.  Torino,  v.  54,  1911,  p.  33-67.     1912. 


1912.  LA  vegetazione  dell'appennino  piemontese.    In  Ann.  Bot.  [Rome], 
V.  10,  fasc.  3,  p.  189-338. 


366  Journal  of  Agricultural  Research  voi.  xii.  no.  6 

{22)  Grouven,  H. 

1858.  bEmERkungen  zu  uEbig's  ansichten.  (Abstract.)  In  Jahresber. 
Agr.  Chem.,  Jahrg.  i,  1858/59,  p.  13-14-  i860.  (Original  article  in 
Agron.  Ztg.,  Jahrg.  13,  No.  30,  p.  465-470.     1858.    Not  seen.) 

(23)  Hall,  A.  D.,  BrenchlEy,  Winifred  E.,  and  Underwood,  Lilian  M. 

1913.   THE  SOIL  SOLUTION  AND  MINERAL  CONSTITUENTS  OP  THE  SOIL.      In  Phil. 

Trans.  Roy.  Soc.  London,  ser.  B,  v.  204,  no.  307,  p.  179-200,  2  tig. 

(24)  Harris,  F.  S.,  and  Butt,  N.  J. 

1917.   EFFECT  OF  IRRIGATION  WATER  AND  MANURE  ON  THE  NITRATES  AND  TOTAL 

SOLUBLE  SALTS  OF  THE  SOIL.  In  Joiir.  Agr.  Research,  v.  8,  no.  9,  p. 
333-359.  18  fig.     Literature  cited,  p.  359. 

(25)  HeydEN,  Eduard. 

1865.   UEBER    DEN  PHOSPHORSAUREGEHALT   IN  WASSRIGEN    BODENAUSZUGEN. 

(Abstract.)  In  Jahresber.  Agr.  Chem.,  Jahrg.  8,  1865,  p.  33-34. 
1867.  (Original  article  in  Ann.  Landw.,  Jahrg.  45,  p.  189-194.  1865. 
Not  seen.) 

(26)  HiLGARD,  E.  W. 

1903.    METHODS  OF  PHYSICAL  AND  CHEMICAL  SOIL  ANALYSIS.      Cal.   Agr.    Exp. 

Sta  Cir.  6,  23  p.,  i  fig. 

(27)  HOAGLAND,  D.  R. 

191 7.   THE  FREEZING-POINT  METHOD  AS  AN  INDEX  OF  VARIATIONS  IN  THE  SOIL 

SOLUTION  DUE  TO  SEASON  AND  CROP  GROWTH.  In  Jour.  Agr.  Research, 
V.  12,  no,  6,  p.  369-395.  9  fig- 

(28)  Hoffmann,  Robert. 

1863.    UEBER  QUALITAT  UND  QUANTITAT  DER  AUS  DER  ACKERERDE  DURCH  EINES 

WASSER  AUFNEHMBAREN  BODENBESTANDTHEiLE.  In  Landw.  Vers. 
Stat.,  Bd.  5,  p.  193-197- 

(29)  ISCHEREKOV,  V. 

1907.  THE  OBTAINING  OF  THE  SOIL  SOLUTION  IN  UNALTERED  CONDITION.  (Abs- 
tract.) In  Exp.  Sta.  Rec,  v.  19,  no.  5,  p.  409.  1908.  (Original 
article  in  Zhur.  Opuitn.  Agron.  [Russ.  Jour.  Expt.  Landw.],  v.  8,  no. 
2,  p.  147-166.     1907.     Not  seen.) 

(30)  

1911.  SOIL  SOLUTIONS.  (Abstract.)  /«  Exp.  Sta.  Rec,  v.  28,  no.  4,  p.  320-321. 
19 13.  (Original  article  in  Zhur.  Opuitn.  Agron.  [Russ.  Jour.  Expt. 
Landw.],    v.  12,  no.  4,  p.  558-560,  1911.    Not  seen.) 

(31)  Jarriges,  V. 

1861.  ANALYSE  VON  kleEmuden  bodenarten.  (Abstract.)  In  Jahresber. 
Agr.  Chenu,  Jahrg.  4,  1861/62,  p.  36-37.  1863.  (Original  article  in 
Chem.    Ackersmann,  1861,  p.  83.     1861.    Not  seen.) 

(32)  Jensen,  C.  A. 

1916.  SOLUBILITY  OF  PLANT-FOOD  ELEMENTS  AS  MODIFIED  BY  FERTILIZERS.      In 

Jour.  Amer.  Soc.  Agron.,  v.  8,  no.  2,  p.  100-105. 

(33)  Jordan,  W.  H. 

1916.  MEASUREMENTS  oP  SOIL  FERTILITY.  N.  Y.  State  Agr.  Exp.  Sta.  Bui. 
424,  p.  389-412. 

(34)  King,  F.  H. 

1904.  investigations  in  soil  management,  being  three  of  six  papers  on 
the  influence  op  soil  management  upon  the  water-soluble 

SALTS  IN  SOILS  AND  THE  YIELD  OP  CROPS.      l68p.,2lfig.      MadisoU,  Wis. 
(35) 


1905.  INVESTIGATIONS  IN  SOIL  MANAGEMENT.     U.  S.  Dept.  Agr.  Bur.  Soils  Bul. 
26,  205  p.,  9  fig.,  4  pi- 


Feb.  II,  1918  Effect  of  Season  and  Crop  Growth  on  Soil  Extract  367 

(36)  King,  F.  H.,  and  Jeffery,  J.  A. 

1900.  THE  SOLUBLE  SALTS  OF  CULTIVATED  SOILS.  In  Wis.  Agr.  Exp.  Sta.,  i6th 
Ann.  Rpt.  [iSgSj/gg,  p.  219-243,  3  fig. 

(37)  ^"<i  Whitson,  a.  R. 

1900.  SOLUBLE  SALTS  OP  CULTIVATED  SOILS.    In  Wis.  Agr.  Exp.  Sta.,  17th  Ann. 

Rpt.,  [iSggj/igoo,  p.  204-226,  fig.  27-33. 
(38) 

1901.  DEVELOPMENT  AND  DISTRIBUTION  OK  NITRATES  IN  CULTIVATED  FIELD  SOILS. 

7n  Wis.  Agr.  Exp.  Sta.,  i8th  Ann.  Rpt.,  [igooj/oi,  p.  210-231,  fig.  48-52. 

(39)  LiEBiG,  Justus  von. 

1858.  UEBER  EINIGE  EIGENSCHAFTEN  DER  ACKERKRUME.  In  Ann.  Chem.  U. 
Pharm.,  Bd.  105  (n.  R.,  Bd.  29),  Heft  i,  p.  109-144. 

(40)  Lyon,  T.  L.,  and  Bizzell,  J.  A. 

1913.  THE  PLANT  AS  AN  INDICATOR  OF  THE  RELATIVE  DENSITY  OF  SOIL  SOLU- 

TIONS.    In  Proc.  Amer.  See.  Agron.,  v.  4,  1912,  p.  35-49,  i  fig. 

(41)  Maschhaupt,  J.  G.,  and  Sinnige,  L.  R. 

I912.  DETERMINATIONS  OF  THE  VALUE  OF  PLANT  FOOD  CONSTITUENTS  IN  SOILS 
and  FERTILIZERS  AS  DEPENDENT  UPON  SOLUBILITY.  (Abstract.)  In 
Exp.  Sta.  Rec,  v.  27,  no.  6,  p.  515.  1912.  (Original  article  in 
Verslag  Landbouwk.  Onderzoek.  Rijkslandbouwproefstat.  [Nether- 
lands], 1912,  no.  II,  p.  19-73.     1912.     Not  seen.) 

(42)  Mitscherlich,  E.  a. 

1907.  EINE  chemische  bodenanalyse  Fur   pflanzenphysiologische  for- 
SCHUNGEN.     In  Landw.  Jahrb.,  Bd.  36,  Heft  2,  p.  309-369,  10  fig.,  i  pi, 

(43)  Morgan,  J.  F. 

1917.  SOIL  SOLUTION.     In  Science,  n.  s.,  v.  45,  no.  1149,  p.  19. 

(44)  Pantanelli,  E. 

1914.  ricerche  sulla  concentrazione  del  liquido  circolante  nei  terreni 

LiBici.     In  Bul.  Orto  Bot.  R.  Univ.  Napoli,  t.  4,  p.  371-383. 

(45)  Peters,  E. 

i860.  studien  tjber  den  boden,  aus  dem  laboratorium  zu  tharand.  in 
Landw.  Vers.  Sta.,  Bd.  2,  p.  113-151. 

(46)  ScHLCESiNG,  Th6ophile. 

1866.  suR  l'analyse  des  principes  solubles  de  la  terre  veig^tale.    In 
Compt.  Rend.  Acad.  Sci.  [Paris],  t.  63,  no.  24,  p.  1007-1012. 
(47) . 

1870.  analyse  des  eaux  contenues  dans  les  terres  arables.  In  Compt. 
Rend.  Acad.  Sci.  [Paris],  t.  70,  no.  2,  p.  9S-102. 

(48)  Schlcesing,  Thtophile,  jr. 

1901.  l'acide  phosphorique  dissous  par  les  eaux  du  sol,  son  utilisation 

PAR  LES  PLANTES.     In  Ann.  Sci.  Agron.,  s.  2,  ann.  7,  t.  2,  fasc.  i, 
p.  77-82. 

(49)  ■ 

1902.  SUR  LES  PHOSPHATES  DU  SOL  SOLUBLE  A  l'Eau.     In  Compt.  Rend.  Acad. 

Sci.  [Paris],  t.  134,  no.  23,  p.  1383-1385. 

(50)  ScHREiNER,  Oswald,  and  FailyER,  G.  H. 

1906.  COLORIMETRIC,  TURBIDITY,  AND  TITRATION  METHODS  USED  IN  SOIL  INVESTI- 
GATIONS.    U.  S.  Dept.  Agr.  Bur.  Soils  Bul.  31,  60  p.,  5  fig.,  i  pi. 

(51)  ScHULzE,  Franz. 

1864.   UEBER   DEN  PHOSPHORSAURE-GEHALT   DES   WASSERAUSZUGS   DER   ACKE- 

RERDE.    In  Landw.  Vers.  Stat.,  Bd.  6,  p.  409-412. 

(52)  Schumacher,  Wilhelm. 

1863.  UEBER  DIE  BODENLOSUNGEN  UND  SOLCHE  BETREFFENDE  VEGETATION- 
SVERSUCHE.     In  Landw.  Vers.  Stat.,  Bd.  5,  p.  210-222.  ' 


368  Journal  of  A  gricultural  Research  voi.  xii,  no.  e 

(53)  Snyder,  Harry. 

1904.  THE  WATER-SOLUBLE  PLANT  FOOD  OF  SOILS.     Ill  Science,  n.  s.,  v.  19, 

no.  491,  p.  834-835. 
(54) ^"d  Hummel,  J.  A. 

1905.  SOIL  INVESTIGATIONS.     Minn.  Agr.  Exp.  Sta.  Bui.  89,  p.  189-212,  2  pi. 

(55)  Thompson,  H.  S. 

1850.  ON  THE  ABSORBENT  POWER  OK  SOILS.     In  Joxir.  Roy.  AgT.  Soc.  England, 
V.  II,  p.  68-74. 

(56)  TOULAIKOV,  N.  M. 

1913.   THE  INFLUENCE  OF  THE  OSMOTIC  PRESSURE  OF  THE  SOIL  SOLUTION  ON  THB 

GROWTH   OP  SPRING   WHEAT.     (Abstract.)     In   Intemat.    Inst.    Agr. 
[Rome]  Mo.  Bui.  Agr.  Intel!,  and    Plant    Diseases,  year  5,  no.   11, 
p.   1426-1429.     1914.     (Original  article    in    La    P^dologie,    aim.    15, 
no.  4,  p.  71-103.     1913.     Not  seen.) 
(57) 

1916.   OSMOTIC  PRESSURE  OF  THE   SOIL  SOLUTION  AND  THE  VITREOUS  STATE  OP 

BELOTOURKA  GRAIN.  (Abstract.)  In  Chem.  Abstracts,  v.  10,  no.  23, 
p.  3127-3128.  1916.  (Original  article  in  Zhur.  Opuitn.  Agron.  [Russ. 
Jour.  Expt.  Landw.],  v.  17,  p.  79-92.     1916.     Not  seen.) 

(58)  Ulbricht,  R. 

1863.  EiN  BEiTRAG  zur  METHODS   DER  BODEN-ANALYSE.     In  Landw.   Vcrs. 

Stat.,  Bd.  5,  p.  200-209. 

(59)  Van  Suchtelen,  F.  H.  H. 

I912.   METHODE     zur     GEWINNUNG     DER      NATtJRLICHEN     BODENLOSUNG.      /» 

Jour.  Landw.  Bd.  60,  Heft  4,  p.  369-370. 

(60)  Way,  J.  T. 

1850-52.    ON  THE  POWER  OP  SOILS  TO   ABSORB   MANURE.      In  Jour.   Roy.   AgT. 

Soc.  England,  v.  11,  p.  319-380,  1850;  v.  13,  p.  123-143,  1852. 

(61)  Whitney,  Milton,  and  Cameron,  F.  K. 

1903.   THE  chemistry  of  THE  SOIL  AS  RELATED  TO  CROP  PRODUCTION.      U.   S. 

Dept.  Agr.  Biu".  Soils  Bui.  22,  71  p. 

(62)  Wolff,  Emil. 

1864.  ENTWURF  ZUR  BODENANALYSE.     In  Landw.  Vers.  Stat.,  Bd.  6,  p.  141- 

171. 

(63)  WundER,  Gustav. 

i860.  uEber  DIE  IN  DEM  boden  enthaltenden  losungen.    In  Landw.  Vers. 
Stat.,  Bd.  2,  p.  104-112, 


PLATE  14 
A. — General  views  of  soil  containers. 
B. — Bins  for  storage  of  surplus  soil. 


Effect  of  Season  and  Crop  Growth  on  Soil  Extract 


Plate  14 


Journal  of  Agricultural  Research 


Vol.  Xll,  No.6 


THE    FREEZING-POINT    METHOD    AS    AN    INDEX    OF 

VARIATIONS  IN  THE  SOIL  SOLUTION  DUE  TO 

SEASON  AND  CROP  GROWTH  ^ 

By   D.   R.   HOAGLAND, 

Assistant  Professor  of  Agricultural  Chemistry,  College  of  Agriculture  of  the  University 

of  California 

INTRODUCTION 

In  the  preceding  article  by  Stewart  (8)^  it  has  been  conclusively  shown 
that  water  extracts  of  different  soils  may  have  widely  varying  concen- 
trations of  important  nutrient  elements,  and  that  the  water-soluble 
substances  in  the  13  cropped  soils  were  strikingly  diminished  in  quantity 
when  the  barley  crop  had  reached  its  maximum  power  of  absorption. 
In  the  uncropped  soils  significant  seasonal  variations  were  also  noted. 
These  studies  suggested  the  importance  of  correlating  the  water  extracts 
with  the  actual  soil  solution,  the  immediate  source  of  nutrients  for  the 
plant. 

Various  attempts  have  been  made  to  separate  the  soil  solution,  but  in  no 
case  has  any  appreciable  quantity  of  solution  been  obtained  when  a  soil 
contained  only  the  optimum  percentage  of  moisture.  Recently  Bouyou- 
cos  and  McCool  (2)  have  proposed  a  method  which  seems  capable  of 
giving  direct  experimental  evidence  concerning  the  concentration  of  the 
soil  solution.  The  procedure  consists  in  determining  the  depression  of 
the  freezing  point  in  the  soil  itself  under  varying  moisture  conditions. 
The  soils  used  in  the  investigation  described  by  Stewart  seemed  uniquely 
adapted  for  further  study  by  the  freezing-point  method.  Accordingly 
such  determinations  were  made  on  the  various  soils  under  controlled 
conditions,  and  the  present  paper  deals  with  observations  made  on  the 
depressions  of  the  freezing  point  in  soils  as  affected  by  season,  cropping, 
moisture  content,  and  other  factors  of  significance  in  plant  growth. 

GENERAL  METHOD  OF  PROCEDURE 

The  technic  employed  was  essentially  that  described  by  Bouyoucos 
and  McCool  (2) .  The  experience  of  the  Agricultural  Chemistry  Laboratory 
confirms  their  statements  with  regard  to  the  possibility  of  obtaining 
closely  agreeing  duplicates  and  general  consistency  of  results.  The 
freezing-point  blank  with  distilled  water,  however,  was  not  found  to 
maintain  a  constant  value  from  day  to  day,  and  to  obviate  any  possible 

1  Approved  for  publication  in  the  Journal  of  Agricultural  Research  by  the  Director  of  the  Agricultural 
Experiment  Station  of  the  University  of  California. 

2  Reference  is  made  by  number  (italic)  to  "  Literature  dted,"  p.  394-395. 

Journal  of  Agricultural  Research,  Vol.  XII,  No.  6 

Washington,  D.  C  Feb.  11,  1918 

ly  Key  No.  Cal.— 17 

(369) 


370 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  6 


error  from  this  source,  determinations  on  pure  water  were  checked 
several  times  each  day.  The  general  magnitude  of  error  in  technic  is 
indicated  by  the  following  typical  instances  of  duplicate  experiments: 


Soil  No. 


loA 
loA 
iB. 
iB. 

I5A 


Freezing- 
point  de- 
pression. 


°c. 
0.065 

.  062 

.066 

.  069 

.049 


Soil  No. 


ISA 
14A 
14A 
4B. 
4B. 


Freezing- 
point  de- 
pression. 


°c. 

o.  052 
•  043 
.050 
•157 
•15s 


DESCRIPTION  OF  SOILS  USED 

In  the  article  by  Stewart  {8)  the  reader  will  find  a  detailed  description 
of  the  soils  under  the  same  laboratory  numbers.  For  the  sake  of  con- 
venient reference,  a  list  of  the  soils  used  is  included  here  (Table  I). 
Each  soil  was  investigated  under  three  general  conditions:  (i)  Soil  from 
tanks  in  which  barley  crop  was  grown  for  two  seasons ;  (2)  corresponding 
soils  kept  under  identical  conditions  but  with  no  crop  the  second  season ; 
(3)  soils  air-dried,  sifted,  and  kept  two  years  in  closed  bins.  All  tank 
soils  were  kept  constantly  at  nearly  optimum  moisture  contents  with 
distilled  water. 

Table  I. — Description  of  soils  used  in  this  investigation 


Laboratory  No. 

Soil  series  and  type. 

Origin. 

Yolo  silty  clay  loam 

Sacramento  Valley. 

do 

Santa  Clara  Valley. 

7                    

Hanford  fine  sandy  loam 

Southern  California. 

8                   ... 

Fresno  fine  sandy  loam 

San  Joaquin  Valley. 

Kimball  fine  sandy  loam 

Southern  California. 

10    

Te junga  fine  sandy  loam 

Madera  fine  sandy  loam 

Arnold  fine  sandy  loam 

Do. 

II 

San  Joaquin  Valley. 

Do. 

Standish  fine  sandy  loam 

Virgin  desert  soil. 

UNFREE  WATER  OF  SOILS 

One  of  the  most  important  factors  affecting  the  concentration  of  the 
soil  solution  is  the  moisture  content.  This  has  already  been  pointed 
out  by  Bouyoucos  and  McCool,  and  the  former  has  recently  devised  a 
dilatometer  method  for  more  accurately  measuring  the  unfree  water 
of  a  soil  (7).  In  the  present  investigation  an  attempt  was  made  to 
approximate  the  unfree  water  in  each  soil  by  careful  determinations  of 
the  freezing-point  depressions  at  different  moisture  contents.  These 
data  could  then  be  used  in  reducing  the  observed  depressions  to  definite 
and  comparable  moisture  percentages,  especially  where  the  observed 


Feb.  II,  1918         Freezing-Point  Method  and  Soil  Solution 


371 


moisture  contents  differed  only  slightly  from  the  standard.  It  has 
been  shown  by  Bouyoucos  and  McCool  that  freezing-point  depressions 
vary  with  the  water  content  of  the  soil,  not  usually  in  direct  ratio,  but  in 
such  manner  as  to  necessitate  the  assumption  that  a  certain  fraction 
of  the  total  water  present  is  so  combined  that  it  does  not  form  an  effective 
part  of  the  soil  solution  and  is  not  subject  to  freezing.  The  percentage 
of  combined  water  would  vary  greatly  with  the  type  of  soil,  clay  soils 
having  a  large  proportion  of  their  water  in  the  unfree  state.  The  above 
considerations,  advanced  by  Bouyoucos  and  McCool,  have  been  made 
the  basis  for  the  determinations  of  unfree  water  in  the  soils  used  in  this 
investigation.  Portions  of  each  soil  were  divided  into  two  samples  the 
moisture  contents  of  which  were  so  adjusted  as  to  give  a  difference  of 
about  5  per  cent.  Each  sample  was  thoroughly  mixed  and  kept  over- 
night in  a  tight  jar.  Careful  determinations  of  total  moisture  and  freez- 
ing-point depression  were  then  made.  From  these  data  it  is  possible 
to  calculate  what  proportion  of  the  moisture  must  be  subtracted  from 
the  total  in  order  that  the  percentages  of  free  water  may  give  the  same 
ratios  as  the  freezing-point  depressions.  Such  estimates  may  not  have 
a  high  degree  of  accuracy,  but  it  is  probably  sufficient  for  the  purposes 
in  hand.  Table  II  presents  the  results  for  the  unfree  water  in  each 
soil. 

Table  II. — Estimation  of  unfree  water  in  soils 


Soil  No. 


iC 
iC 
2C 
2C 
3C 
3C 
4C 
4C 

sc 
sc 

6C 
6C 

7C 
7C 


Moisture. 

Freezing- 
point  de- 
pression. 

Unfree 
water. 

Per  cent. 

*C. 

Per  cent. 

20.  0 

0.098 

1      18.0 

22.  4 

•034 

18.0 
23.0 

.  140 
.  069 

}      13-0 

18.4 
24.  0 

•357 
.149 

>      14.0 

19.  6 

23.6 

.362 
.  210 

}      13-0 

18.8 
23.  2 

.089 
.050 

}      13-0 

19.  6 

24.4 

.282 
.079 

}     17- 7 

12.4 
17.6 

.284 
.167 

1       6.0 

Soil  No. 


8C. 
8C. 
9C. 
9C. 
loC 
loC 
iiC 
iiC 
12C 
12C 
14C 
14C 


Freezing- 

Moisture. 

point  de- 

pression. 

Per  cent. 

°C. 

16.6 

0-215 

8.1 

•655 

12.8 

.099 

17.2 

•059 

12.8 

•153 

17.8 

.074 

13.2 

'2,2,^ 

18.8 

•  197 

13.2 

.130 

18.0 

.  089 

14.  0 

•254 

18.4 

.149 

Unfree 
water. 


Per 


cent. 
4.0 

6.0 

8.0 
6.0 

4.0 

7-S 


SEASONAL  VARIATIONS  IN  CROPPED  AND  IN  UNCROPPED  SOILS 

One  of  the  primary  objects  of  this  investigation  was  to  follow  through 
the  season  the  concentrations  of  the  soil  solution  in  each  soil  and  to  com- 
pare at  each  period  of  sampling  the  cropped  soils  with  their  uncropped 
duplicates.  The  first  determinations  of  freezing-point  depressions  were 
made  early  in  July,  1916,  and  at  intervals  from  that  date  until  the  fol- 
lowing May.     The  barley  crop  was  planted  in  May,  1916,  and  harvested 


372 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  6 


in  August.  The  samples  of  soil  used  in  this  work  were  identical  with 
those  employed  by  Stewart  in  making  water  extractions.  In  general, 
the  moisture  contents  of  the  uncropped  and  cropped  soils  were  very 
similar,  but  it  was  not  always  feasible  to  keep  them  in  absolute  agree- 
ment. The  data  are  therefore  presented  in  two  forms:  First,  as  a  table 
showing  the  observed  depressions  and  actual  percentages  of  total  moisture 
present,  together  with  the  corresponding  osmotic  pressures  and  parts 
per  million  of  total  solids  in  solution,  as  calculated  by  the  methods  of 


(&ASIS   or  a2'/«  noiSTURC) 


time:  from 
plantiino. 


(bASISaOF   ZZ.'/'  nOlSTURE) 


I0WIC5  12WKS  I6WKS 
JULIO.  JUL24  AUG  21 


51WKS     time:  FRom 

MAY.  1  PLANTING. 


Fig.  I. — Graphs  of  the  depressions  of  the  freezing  point  in  soils  i  and  2,  with  and  without  crop. 

Bouyoucos  and  McCool.  In  the  second  place,  the  observed  depressions 
have  been  calculated  to  uniform  moisture  contents,  17  per  cent  for  all 
of  the  fine  sandy  loams  and  22  per  cent  for  the  silty  clay  loams.  In  mak- 
ing these  estimates  the  percentages  of  unfree  water  already  presented 
have  been  used.  The  corrected  depressions  have  been  plotted  for  each 
pair  of  soils.  In  this  way  the  cropped  and  the  uncropped  soils,  as  well 
as  the  different  soils  in  each  group,  may  be  compared  readily  (fig.  1-8). 
(Cf.  Table  III.) 


Feb.  II.  1918         Freezing-Point  Method  and  Soil  Solution 


373 


5^s>ls  or  Z2.'A  moisturl) 


5IWKS        TVHL  FBOM 
MAT\.        PLANTING. 


(E>A.51t>    OF    22%  mOlSTUREL) 


lOWKS  IZWK5  I6WKS  Z4W1<:5 

JULIT  JUL3I    AUGZe.         OCT. 30 


5EWK5         40WKS 
DEC.  26         FtM9 


5IWKS       TIME.  fROn 
f-WYT  PLANT  IN  O 


Fig.  2.— Graphs  of  the  depressions  of  the  freezing  point  in  soils  3  and  4,  with  and  without  crop. 


374 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  6 


fDAtj»5    OFZZ'/   MOISTURfL) 


I0WK5 I2WK5  I6WK5  24.WK5 

JULIl  JUL3I  AUGZ6  OCT. 50 


5ZWK3.        40WKS 
DE:C2.6         FE.e>19. 


5IWK5 
MAY  7. 


Fig.  3.— Graphs  of  the  depressions  of  the  freezing  point  in  soil  5,  with  and  without  crop. 


Feb.  II,  1918        Freezing-Point  Method  and  Soil  Solution 


375 


.055 

- 

.050 

^/       \ 

/  \ 

.045 

oy            \ 

.040 

- 

^Diis--''^                      (5A51SOF    2.Z°/.  MOI 

035 

f\ 

/ 

.0^ 

- 

"V^                           N.      ^^^^ 

025 

\ 

f 

.020 

\ 

/ 

\    o7 

.015 

- 

\  / 

.010 

_ 

.005 

- 

C[ 

1 

1          1          t          1            1 

I0WKS12WKSI6WKS 
JULH  JUL31  AU0Z8 


OCT.  50 


^2WK5 
DLC  26 


40WKS 
FEb  19 


.5IWK3     TlhELFROn 
MAY  7.     PLA\NTlN<j 


Fig.  4. — Graphs  of  the  depressioas  of  the  freezing  point  in  soil  6,  with  and  without  crop. 
27809°— 18 6 


376 


Journal  of  A  gricultural  Research 


Vol.  XII,  No.  6 


T/4  MOISTURf:) 


I0WK5I2WK5.I6\NK5  24WK,^  iZNNKS       4-OVVK.S 

JUU7  JUL3I  AU0  2B  OCT  30  DE.C  26        FLb  13 


5IWK5     TIME.  TROM 
HA^T       PLANTING. 


(C)A51S    OF     11'/-     nOISTUPLE-") 


lOWKS.  IZWIO.ieWKSi  24WKS  52WKS         40WKS 

JULI0.JUL24AUGEI  OCT  23         DCC.Ja         FE.B.I2 


■5IWK5     TIML  FROM 
nAY7        PLANT\NG. 


FlG.  5. — Graphs  of  the  depressions  of  the  freezing  point  in  soils  7  and  8,  with  and  without  crop. 


Feb.  M.  i9i8         Freezing-Point  Method  and  Soil  Solution 


377 


(5A5ie>  OF   n'A   noiSTUREi) 


I0WK5 I2WKS I6WKS  Z4WKS        52WK5        40WK.5> 

OULI0JUL24AUCil  OCT  22,        DLCIS        FE.BIZ 


51\N/K5      T\ML  FROM 
MAY  \  PLM^JTING 


^ 


.025 

- 

D2Q 

.0J5 

SOIL    \05_ 

^ 

NONE.. 

(5A515     Of      17"/-    M 

.010 

soilM£^ 

c2^ 

-^^^4^y 

.0^ 

^^^-^r'^"^^'^ 

c 

II               1               1 

1                 1 

I0WK5I2WIC5  16WK5  Z4-VVK5  5ZWK5        4.0WKS 

JULIQ  JUL24  AUG.2I  OCT.23         DEIC.IO         FL5I2 


51WKS       TIME   FCOn 
MAY  I  PLANTING. 


Fig.  6.— Graphs  of  the  depressions  of  the  freezing  point  in  soils  9  and  10,  with  and  \nthout  crop. 


378 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  6 


CgOPjVOlVC 


(bA&IS    OF    \T/.  nOISTURJL) 


IOWKSI2WK5l6WKe>  24-WK.5  5ZWK5         4-OWKS 

JULI0JULa4AUG.21  OCTZi  DE_C  IS        FELb.  17 


5IWK5      time:  FROn 

MA^^Y  I         PLANTING. 


(e>At>(5      OF    \iy-  ri015TUR£.^ 


'  I0WKSIZWKSI6WK5 
JUU7JUL24^UG2I 


Z4WK5  /      52LWK5        40W1C5 
OCT  23         DELCia  TELB  17 


5IWIO    TiMt  rfion 

^^Yl  PLANTING 


Fio.  7. — Graphs  of  the  depressions  of  the  freezing  point  in  soils  ii  and  12, with  and  without  crop. 


OF  IT/.  MOISTURE.^ 


I0WKSI2WK5I6WK5  24.WKt>         5ZWK5.      40WX.5 

JUL.ITJUL5\^UG^6         OCT  30        DE.C.  iS      FEL&O 


51WK.5       TIME.  FR.Or-1 
MAY  7  PLANTING. 


Pig.  S. — Graphs  of  the  depressions  of  the  freezing  point  in  soil  14,  with  and  without  crop. 


Feb.  II.  1918         Freezing-Point  Method  and  Soil  Solution 


379 


Table  III. — Observed  and  calculated  freezing-point  depressions 


Date. 


Moisture. 


Soil 
lA. 


Soil 
iB. 


Observed 

freezing-point 

depression. 


Soil 
lA. 


Soil 
iB. 


Osmotic 
pressure  (atmos- 
pheres). 


Soil 
lA. 


Soil 
iB. 


Calculated  total 

solids  in  soil 

solution . 


Soil  lA. 


Soil  iB. 


Freezing-point 
depression  cal- 
culated to  uni- 
form moisture.a 


Soil 
lA. 


Soil 
iB. 


July  10, 
July  24, 
Aug.  21 
Oct.  23. 
Dec.  18 
Feb.  12 
May  I . 


July  10. 
July  24. 
Aug.  21 
Oct.  23 
Dec.  18 
Feb.  12 
May  I . 


July  10. 
July  24, 
Aug.  21 
Oct.  23 
Dec.  18 
Feb.  12 
May  I . 


July  17, 

July  31- 

Aug.  28 
Oct.  30 . 
Dec.  26 
Feb. 19 

May  7  . 


P.  ct. 
23.2 
22.  5 

23-9 
21.8 

23-4 
22.  7 
21.  9 


Soil 
2A. 


20.  4 
19.7 

23-7 
18.3 
20.  4 
20.  2 
17.4 


Soil 
3A. 


18.6 

19.  o 

22.  O 

20.  5 
20.  9 

20.  6 
19.  I 


Soil 

4A. 


18.8 
18.6 
19.  2 
17-3 
19-3 
17.  6 

16.5 


P.  ct. 
22.  4 
22.  7 
23.0 
21.  I 
22.3 
21.  7 
20.3 


°C. 

0.030 

.  029 

.  019 

•  050 
•037 

•  056 
.082 


"  c. 
o.  064 
.065 
.  062 
.118 

•095 
.  108 

.138 


Soil 
2B. 


Soil 
2A. 


Soil 
2B. 


Soil 

2A. 


21.  4 
20.8 
21.  3 
18.9 
20.  o 
20.  2 
17-3 


o.  026 

034 
008 

055 
047 
056 
131 


o.  044 

.047 
.049 
.  092 

.087 

.  lOI 

.  201 


0.32 
.41 

.  10 
.66 
.56 
.68 

1.47 


Soil 
3B. 


Soil 
3A. 


Soil 
3B. 


Soil 
3A. 


22.  2 
22.  2 
22.  7 
20.  9 
20.  7 
19.9 

17-5 


o.  130 

.078 
.  017 

•075 
.048 
.  062 
.068 


0-055 
.036 

.034 
.  no 
.  060 
.  120 

•  ^2,S 


I.  46 

•94 
.  20 
.90 

•58 
•74 
.82 


Soil 
4B. 


Soil 

4A. 


Soil 
4B. 


Soil 
4A. 


18.5 

19.  6 

20.  o 
16.  7 
18.0 
17.2 
I5^8 


o.  048 
.050 
.  017 
.068 

•055 
.098 
.  106 


0.047 
.066 

•057 
.082 


157 


0.58 
.60 
.  20 
.82 
.66 
I.  19 
I.  29 


0.77 
•78 
•74 
1^43 
I-  IS 
I- 31 
1.56 


P.  p.m. 

900 

900 

600 

I,  600 

1,  200 
1,800 

2,  600 


P.  p.m. 
2,000 
2,000 
I,  900 

3,700 
3,000 
3,400 
4,300 


'  C. 
039 

033 
028 

047 
050 
066 
080 


Soil 
zB. 


Soil  2A. 


0.53 
•56 

•59 
I.  II 
1.05 

1.  22 

2.  42 


800 
1,  100 

300 
1,700 
1,500 
I,  800 
4,  100 


Soil  2B. 


Soil 
2A. 


Soil 
3B. 


Soil  3  A . 


66 

44 
41 

72 

45 
52 


4,  100 
2,  400 
500 
2,300 
1,500 

1,  900 

2,  100 


Soil 
4B. 


S0U4A. 


0.56 
.80 
.68 

■99 
I.  46 
I.  179 
1.8 


1,500 

1,  600 
500 

2,  100 
I,  700 

3,  100 
3,300 


1,  400 
1,500 
1,500 

2,  900 
2,  700 
3,200 
6,300 


o.  021 
025 
010 

032 
039 
045 
064 


SoilsB. 


Soil 
3A. 


I,  700 
I,  100 
I,  100 
3,400 
I,  900 
3,800 
4,  200 


0.075 
049 
017 
061 
041 
051 
043 


Soil4B. 


Soil 

4A. 


1,500 
2,  100 
1,800 
2,  600 
3,800 

4,  600 
4,900 


0.031 
031 
012 

033 
038 
050 
041 


"  C. 
3.  070 
.  076 

•77 
.  ogi 
.  102 
.  100 
•079 


Soil 
2B. 


o.  041 
.  041 

•045 
.  060 
.068 
.081 
.  096 


Soil 
3B. 


D.  056 
•037 
•037 
•095 
.  050 
.089 
•059 


SoU 
4B. 


o.  029 

.048 
.044 
•  034 

.  067 
.  069 

.049 


a  Soils  I,  2,  3,  4,  s,  and  6  calculated  to  22  per  cent  of  moisture.     Soils  7,  8,  9,  10,  11,  la,  and  14  calculated 
to  17  per  cent  of  moisture. 


38o 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  6 


Table  III. — Observed  and  calculated  freezing-point  depressions — Continued 


July  17. 

July  31- 
Aug.  28 
Oct.  30 . 
Dec.  26 
Feb.  19 
May  7. . 


July  17. 

July  31- 
Aug.  28 . 
Oct.  30. 
Dec.  26. 
Feb.  19. 
May  7 .  . 


July  17. 
July  31- 

Aug.  28 
Oct.  30 
Dec,  26 
Feb, 19 
May  7. 


July  10. 
July  2r4. 
Aug.  21 
Oct,  23 
Dec.  18 
Feb. 12 
May  7. 


July  10. 
July  24. 
Aug.  21 
Oct.  23 
Dec.  18 
Feb. 12 
May  I . 


Moisture. 


Soil 

SA. 


P.  ct. 
16.  2 
18.4 
21.  9 
16.8 
18.2 
18.  I 
16,9 


Soil 
6A. 


23-4 
24.  I 

24-  5 
24.8 
23.0 


Soil 
7A. 


12.5 


Soil 

8A. 


13.2 
14.  I 

15-9 
9.9 

12.  7 

10.  6 


Soil 

sB. 


p.  ct. 

18.  7 

20.  6 

21.  I 

17- 5 
18.0 
17.  2 

19.  2 


Soil 
6B. 


23.8 
25.2 
24.9 
24.4 
24.8 
25.  o 
22.  6 


0-399 


Soil 
7B. 


15-9 
16.  9 
16.6 
13-6 
14.7 
14.  o 
12.8 


Soil 
SB. 


Soil 
9A. 


14-3 
14.  o 
16.  9 

10.  7 

11.  o 

9.4 
7-1 


Soii 
9B. 


13-4 
13.2 

13- I 

9-7 

10.  6 

9-5 
6.9 


observed 

freezing-point 

depression. 


Soil 

sA. 


"  C. 
0.093 
029 
017 
079 
066 
106 
079 


Soil 
6A. 


.  040 
.068 
.085 
.066 
.  104 


Soil 
7A. 


0.028 
.  026 
.  007 
.082 
.038 

•053 
.  070 


Soil 

SA. 


038 

015 
,  021 
,  060 
,049 
,062 
,  122 


Soil 
9A. 


O.  OI6 
.  007 
•  014 
.028 
.031 
•057 
.118 


Soil 
sB. 


C. 

o.  069 

•  039 

•  157 
.088 

.  106 

.115 

.  lOI 


Soil 
6B. 


097 
086 
072 

lOI 

103 
126 
153 


Soil 

7B. 


o.  048 
.  042 

•  03s 

•  151 
.  108 

•  093 
.  116 


Soil 
SB. 


o.  064 

•053 
.056 

.138 
.  Ill 

•135 
.  214 


Soil 
9B. 


0.034 
.  026 
.  042 
.  092 
.058 
.087 
.  200 


Osmotic 

pressure  (atmos 
pheres). 


Soil 

sA. 


Soil 
sB. 


0.83 

•4:7 


Soil 

6A. 


4.  til 


.82 
I.  02 

.80 
I.  26 


Soil 
7A. 


0.34 

32 


Soil 
6B. 


I.  17 

I- OS 
.86 
I.  22 
1-25 
1-52 
1.85 


Soil 
7B. 


0.58 

■  50 

.42 

1.82 

1-31 
I-  13 
I.  41 


Soil 

8A. 


o.'46 

.  18 
•25 
•72 
•59 

•74 

1.47 


Soil 
SB. 


0.77 
.64 
.68 
1.56 
1-34 
1-52 
2.58 


Soil 

9A. 


Soil 
9B. 


0.  41 

•  32 

•  50 

1.  II 

.70 

1.  OS 

2.  41 


Calculated  total 

solids  in  soil 

solution. 


Soil  5A. 


P.  p.m. 

2,900 
900 
500 

2,500 

2,  100 

2,500 


Soil 
6A. 


12,  000 


1,300 
2,  100 
2,  700 
2,  100 
3,300 


Soil 

7A. 


900 

800 

200 

2,  600 

I,  200 

1,  700 

2,  200 


Soil  5B. 


P.  p.m. 
2,  200 
I,  200 
4,900 
2,800 
3i30o 
3,600 
3,200 


Soil 
63. 


3,000 
2,  700 
2,  300 
3,200 
3,200 
3>900 
4,  800 


Soil 
7B. 


1,500 
1,300 

1,  100 
4,700 
3,400 

2,  900 
3,600 


Soil 

SA. 


I,  200 
500 
700 
I,  900 
I,  500 
1,900 
3,800 


Soil 
9A. 


600 
200 

400 
900 

I,  000 
I,  800 
3,700 


Soil 
SB. 


2,  000 
I,  700 
I,  800 
4,300 
3,400 
4,  200 
6,  700 


Soil 
9B. 


1,  100 
800 

1,300 

2,  900 
I,  800 
2,700 
6,300 


Freezing-point 
depression  cal- 
culated to  uni- 
form moisture. 


Soil 

sA. 


"C. 

033 
017 
017 

033 
038 
060 
034 


Soil 
5B. 


Soil 
6A. 


Soil 
7A. 


o.  020 

.018 

.  009 

.066 

•  034 

.  041 
.  041 


Soil 
8A. 


o.  027 
.  oia 
.  019 
.  027 

•033 
•  032 

•039 


Soil 
9A. 


014 
005 
014 
012 
014 
018 
012 


'C. 
o.  044 

•033 
.  141 
.044 

•059 
•054 
.  070 


Soil 
6B. 


.  102 

0.138 

.150 

•  053 

.  121 

.  lOI 

•157 

•  135 

.  170 

.  109 

.214 

.  128 

.174 

Soil 
7B. 


0.043 
.  042 
•034 
.  194 
.085 
.068 
,  072 


Soil 
SB. 


o.  051 
.044 
.051 
.  064 

•073 
.  071 
.  071 


Soil 
9B. 


o.  023 
.  017 
.  027 

.031 

.  024 

.028 

.  016 


Feb.  II.  1918         Freezing-Point  Method  and  Soil  Solution 


381 


Table  III. — Observed  and  calculated  freezing-point  depressions — Continued 


Date. 

Moisture. 

Observed 

freezing-point 

depression. 

Osmotic 
pressure  (atmos- 
pheres). 

Calculated  total 

solids  in  soil 

solution. 

Freezing-point 
depression  cal- 
culated to  uni- 
form moisture. 

Soil 
loA. 

Soil 
loB. 

Soil 
loA- 

Soil 
loB. 

Soil 
loA. 

Soil 
loB. 

Soil 
loA. 

Soil 
loB. 

Soil 
loA. 

Soil 
loB. 

July  10 

July  24 

Aug.  21 

Oct.  23 

Dec.  18 

Feb. 12 

May  I 

P.  ct. 

15-7 
16.  0 
19.  6 
14.9 

15-1 
14.  I 

II- 5 

P.  ct. 
16.6 
16.  I 
17.0 
13-6 

14,  6 

13-4 
II.  0 

"  C. 

0.037 
.023 
.  012 
.  041 
.  046 
.066 
.094 

°  C. 

0.056 

.051 

•057 
.099 
.079 
.125 
.228 

0.44 
.28 

•14 
.49 

•55 

-79 

I.  14 

0.  67 
.61 

.68 

1.  19 
•95 

I- 51 
2-75 

P.  p.m. 

I,  200 

700 

400 
1,300 

1,  400 

2,  100 
2,  900 

P.  p.m. 
1,800 
I,  600 
1,800 
3,  100 
2,500 

3,900 
7,100 

'  C. 

0.  028 

.  021 

•015 

•033 
.036 

•  04S 
•033 

'C. 
0.054 
.047 

•057 
.  061 
•058 

•075 
.  076 

Soil 

iiA. 

Soil 
iiB. 

Soil 

1 1  A. 

Soil 
iiB. 

Soil 
iiA. 

Soil 
iiB. 

Soil 

iiA. 

Soil 
iiB. 

Soil 

iiA. 

Soil 
iiB. 

July  10 

July  24 

Aug.  21 

Oct.  23 

Dec.  18 

Feb. 17 

May  I 

14.9 
12.7 
16.6 
13-9 
14-5 

II.  0 

II.  2 
15.8 

16.3 
13.2 
13-6 

13-5 
10.  7 

0.034 

-035 
.027 

•037 
.047 
.065 
.  121 

0.  060 
.  041 
.052 

.  lOI 

.134 
•  133 
.203 

0.  41 
.42 
•32 
.44 
.56 
.78 

1.47 

0.  72 

•49 
.62 

1.  22 

I-  51 
1.50 

2.45 

I,  100 

I,  100 

800 

1,  200 
1,500 

2,  000 
3,800 

I,  900 
1,300 
I,  600 
3,200 
4,200 
4,  2GO 
6,300 

0.  028 
.  021 
.  026 
.027 
•036 
-043 
-055 

0.  028 

•037 
■049 
.066 

•093 
.  091 

.087 

Soil 
12A. 

Soil 
12B. 

Soil 
12A. 

Soil 
12B. 

Soil 
12A. 

Soil 
12B. 

Soil 
1 2  A. 

Soil 

I2B. 

Soil 
12A. 

Soil 
12B. 

July  17 

July  24 

Aug.  21 

Oct.  23 

Dec.  18 

Feb.  17 

May  I 

II.  I 
14.4 
16.3 

14.4 

13-9 
12.7 

10.  0 

12.3 
13-5 
14-5 
II.  4 
II.  7 
II.  4 
9.6 

0.033 
.  001 
.003 
.  061 
.030 

•043 
.048 

0.054 
.036 
.  026 
.138 

•059 
.083 
.077 

0.  40 
.  01 
.04 
•73 
•36 
•52 
.58 

0.65 
•44 
•32 

1.56 

•71 

I.  01 

.92 

I,  000 

50 

100 

I,  900 

900 

1,300 

1,500 

I,  700 
I,  100 

800 
4,300 

1,  800 

2,  600 
2,400 

0.  018 
.  001 
•003 

■049 
.023 
.  029 
.  022 

0.034 
.  026 
.  021 
.079 

•035 
.047 

•033 

Soil 

14A. 

Soil 
14B. 

Soil 

14A. 

Soil 
14B. 

Soil 

14A. 

Soil 
14B. 

Soil 

14A. 

Soil 
14B. 

SoU 
14A. 

Soil 
14B. 

July  17 

July  31 

Aug.  28 

Oct.  30 

Dec.  26 

Feb.  19 

May  7 

IS- 7 
17-9 
20.  4 

17- S 

17.0 
16.  I 
15-7 

18.4 
19.  I 

19.  I 
16.  2 

16.  5 
15.0 

20.  5 

0.  041 

.013 

.  QIC 

•033 
•039 
.063 

-043 

0.045 
.  027 
.031 

•09s 
.068 
.123 
.  060 

0.49 
.16 
.  12 
.40 

•47 
.76 

•52 

c 

I 
I 

54 
32 
37 
IS 

82 

49 

72 

1,300 
400 
300 

1,000 

1,  200 

2,  000 
1,300 

1,400 

800 

I,  000 

3,000 

•2,  100 
3,800 
1,900 

0-035 
.  014 
.  014 
-035 
-039 
•057 
•037 

0.  052 

•033 
.038 
.087 
•  065 

•097 
.082 

382  Journal  of  Agricultural  Research  voi.  xii.  no.  6 

It  is  strikingly  evident  from  these  data  (Table  III)  that  the  freezing- 
point  depressions  are  not  constant  during  the  season,  and  that  the  con- 
centrations of  the  soil  solution  are  uniformly  lower  in  the  cropped  soils 
than  in  the  same  soils  uncropped.  This  latter  observation  is  especially 
noteworthy,  since  comparisons  of  two  samples  of  the  same  soil  are  pecu- 
liarly applicable.  Any  errors  due  to  unfree  water,  type  of  soil,  optimum 
moisture  content,  etc.,  would  be  practically  constant.  The  same  general 
relations  hold,  whether  the  original  data  are  considered,  or  the  corrected 
figures  as  used  in  the  graphs.  The  only  difference  is  that  more  exact 
comparisons  may  be  made  when  the  moisture  contents  are  reduced  to  a 
uniform  percentage. 

The  logical  conclusion  from  the  results  given  in  Table  III  is  that  the 
concentration  of  the  soil  solution  may  vary  widely  at  different  periods  of 
the  year,  and  that  the  growth  of  the  plant  has  a  pronounced  effect  in 
lowering  the  concentration.  Furthermore,  this  effect  is  of  long  duration. 
The  soils  which  had  been  cropped  are  still  decidedly  lower  in  the  concen- 
tration of  the  soil  solution,  as  compared  with  the  fallowed  soils,  eight 
months  after  harvesting  the  crop.  It  does  not  necessarily  follow  that 
the  differences  between  the  cropped  and  the  uncropped  soils  are  to  be 
ascribed  solely  to  the  depletion  of  the  soil  solution  by  the  plant.  The 
evidence  presented  by  Burd  (j)  is  confirmatory  of  the  idea  that  certain 
biological  activities  are  more  intense  in  the  soil  without  crop,  as  shown  by 
the  increased  production  of  nitrates  and  greater  solubility  of  calcium  and 
magnesium.  It  is  nevertheless  true  that  the  lowest  concentrations  in 
the  cropped  soils  occur  at  about  the  time  when  the  crop  has  completed 
its  maximum  absorption  of  nutrients.  In  certain  of  the  fine  sandy  loams 
the  concentrations  have  been  reduced  to  a  very  low  point,  corresponding 
to  only  a  few  hundred  parts  per  million  of  total  solids.  This  condition 
must  be  the  result  of  withdrawals  by  the  plant.  The  magnitudes  of 
absorption,  as  given  by  Burd,  are  quite  comparable  with  quantities 
present  in  the  soil  solution.  After  the  minimum  has  been  reached  at  the 
height  of  the  growing  season,  the  freezing-point  depressions  slowly-  in- 
crease up  to  the  last  date  recorded  in  the  graphs.  May  i.  It  is  highly 
probable  that  several  weeks  after  cultivation  a  more  marked  increase  in 
the  concentration  of  the  soil  solution  would  take  place. 

At  this  point  it  is  desirable  to  compare  the  data  presented  by  Stewart 
{8)  for  the  water  extracts  with  the  determinations  made  by  the  freezing- 
point  method.  It  will  be  noted  that  the  same  general  relations  obtain 
in  both  cases.  All  the  elements  for  which  analysis  was  made  are  subject 
to  marked  fluctuations,  with  the  exception  of  phosphorus.  This  sub- 
ject will  be  given  further  consideration  later  in  the  article. 


Feb.  II.  r9i8         Freezing-Potfit  Method  and  Soil  Solution  383 

VARIOUS  FACTORS  AFFECTING  THE  SOIL  SOLUTIONS 

As  a  result  of  the  observations  on  the  tank  soils  which  have  just  been 
described,  it  was  decided  to  study  various  other  factors  possibly  affecting 
the  depression  of  the  freezing  point.  What  conditions  other  than  crop- 
ping might  increase  or  decrease  concentration?  The  first  step  was  to 
study  the  soils  which  had  been  held  in  a  nearly  air-dry  state  in  bins.  The 
results  of  these  determinations  have  already  been  set  forth  in  Table  II, 
and  comparisons  may  now  be  made  between  tank  soils  and  the  correspond- 
ing bin  soils.  It  will  at  once  be  noted  that  a  number  of  the  latter  show 
very  much  higher  concentrations  of  the  soil  solution  than  either  the 
cropped  or  the  uncropped  soils  during  most  of  the  season.  All  three 
portions  of  each  soil  were  originally  derived  from  the  same  sifted  and 
homogenous  mass  of  soil.  The  existing  dififerences  must  therefore  be 
due  to  subsequent  treatments  which  are  correlated  with  three  levels  of 
concentration:  (i)  A  very  low  concentration  in  the  cropped  soil,  (2)  a 
higher  one  in  the  uncropped  soil,  and  (3)  highest  of  all  in  some  of  the 
stored  soils.  The  latter  in  a  number  of  cases  gave  four  or  five  times  as 
great  a  depression  of  the  freezing  point  as  the  corresponding  cropped  soils. 
The  fact  that  the  uncropped  soils  in  the  tanks  had  in  some  cases  a  less 
concentration  of  the  soil  solution  than  the  stored  soils  may  be  explained 
by  the  treatment  of  the  preceding  year  when  both  tank  soils  were  cropped. 
It  may  also  be  true  that  long  storage  of  the  bin  soils  under  the  special 
conditions  has  brought  about  a  decomposition  of  soil  minerals  as  well  as  an 
increase  in  nitrates,  with  the  result  that  a  more  highly  concentrated  soil 
solution  is  produced  as  soon  as  the  soil  is  mixed  with  its  optimum  quan- 
tity of  water.  The  extractions  of  the  bin  soils  confirmed  the  freezing- 
point  results.  Water  extracts  showed  a  correspondingly  high  solubility 
for  nearly  all  constituents. 

With  the  establishment  of  the  relations  of  freezing-point  lowerings  of 
soils  under  several  conditions  it  became  of  interest  to  make  a  somewhat 
more  detailed  study  in  an  effort  to  determine  what  factors  are  especially 
active  in  changing  the  concentration  of  the  soil  solution. 

EFFECT  OF  INCUBATION  ON  CONCENTRATION  OF  SOIL  SOLUTION 

The  first  experiments  dealt  with  the  effect  of  long  standing  at  laboratory 
temperatures.  Samples  were  obtained  from  the  tank  soils,  and  after  the 
determinations  of  freezing-point  lowerings  portions  were  kept  in  tight 
Mason  jars  for  from  one  to  three  months.  At  the  end  of  the  specified 
periods  the  freezing  points  were  again  determined,  with  the  results  given 
in  Table  IV. 


384. 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  6 


Table  IV. — Effect  of  incubation  at  laboratory  temperature  on  depressions  of  freezing 

point 


Soil 
No 

iMois- 
ture. 

Freezing-point 
depression. 

Soil  No. 

Mois- 
ture. 

Freezing- point 
depression. 

Soil  No. 

Mois- 
ture. 

Freezing-point 
depression. 

Dec.  27 

Mar.  29 

Dec.  19 

Jan. 
9-16 

Oct.  23 

Nov.  3 

4A... 
4B... 
5A... 
5B... 
6A... 

Per  cent 

19-3 
18.0 
18.2 
18.0 

24-5 

24.8 

15-9 
14.7 

13-9 
II.  7 
17.0 
16.5 

'C. 

0.055 
.  121 
,066 
.  106 
,085 
.103 
.038 
.  108 
.030 
•059 
•039 
.068 

'C. 
0.073 

■n3 
.127 

•153 
.  120 

•  149 
•052 
.081 
•013 
.  046 
.  091 
.  096 

lA.  .. 
iB.  .. 
2A.  .. 
2B.  .. 
3A... 
3B... 
8A... 
8B.  .. 
9A... 
9B... 
loA.  . 
loB.  . 

Per  cent 
23-4 
22.3 
20.  4 
20.  0 
20.  9 
20.  7 
12.7 
12.5 
II.  0 
10.  6 

15- I 
14.  6 

"C. 

0.037 

•095 
.047 
.0S7 
.048 
.  060 

•049 
.  Ill 

•031 
•058 
TO46 
•079 

"C. 

0.  050 
.152 
•050 
.  100 
•063 
.  no 
.  062 

.  142 
.038 

.  072 

.058 
•  105 

loB.  . 

loB.  . 
loB.  . 
icB.  . 
loB.  . 
loB.  . 
loB.  . 

Per  cent 
13.6 

12.8 

14.4 

14.  6 
16.  0 
17.0 
17.8 

'C. 
0.  099 

•c. 

0.439 

Dec.  5 

Dec.  13 

6B... 

7A... 

7B 

12A. . 

12B.. 

14A. . 

14B.. 

0-153 
.127 
.  120 
.  092 
.080 
.074 

0.  180 
.  169 
.  148 
.114 
.  104 
.097 

It  will  be  evident  from  Table  IV  that  in  the  majority  of  cases  there 
has  been  a  distinct  increase  in  the  depression  of  the  freezing  point  under 
the  conditions  of  storage.  It  is  significant  that  at  the  end  of  the  storage 
period  the  uncropped  soils  still  maintain  their  superiority  with  regard  to 
the  concentration  of  the  soil  solution.  The  time  at  which  the  sample 
is  taken  may  in  some  cases  influence  the  further  changes  taking  place  at 
laboratory  temperatures.  This  suggestion  is  supported  by  the  data  for 
soil  loB.  The  samples  taken  in  November  gave  a  much  greater  increase 
in  concentration  after  standing  than  was  the  case  with  the  sample  taken 
in  December.  Doubtless  these  differences  may  be  associated  \\-ith  con- 
ditions favorable  for  nitrate  production,  but  the  extraction  experiments 
have  shown  that  other  elements  are  increased  in  solubility  simrltaneously 
with  the  increased  production  of  nitrates. 

EFFECT    OF    CARBON-DIOXID    GAS    ON    CONCENTRATION    OF  SOIL 

SOLUTION 

The  increase  in  concentration  of  the  soil  solution  noted  above,  as  well 
as  the  seasonal  changes,  may  probably  be  related  to  the  activities  of 
microorganisms  and  their  production  of  carbon  dioxid.  Several  experi- 
ments were  therefore  undertaken  to  demonstrate  the  effect  of  carbon/ 
dioxid  gas  on  freezing-point  depressions.  A  sample  of  soil  5A  containing 
17.7  per  cent  of  water  was  placed  in  a  stoppered  bottle,  and  a  current  of 
carbon-dioxid  gas  was  passed  through  the  moist  soil  for  about  five 
minutes.  This  treatment  was  repeated  several  times.  The  bottle  was 
then  closed  tightly  and  the  soil  permitted  to  stand  overnight  in  contact 
with  the  atmosphere  of  carbon  dioxid,  after  which  the  depression  of  the 
freezing  point  was  determined  and  compared  with  another  sample  of 
the  soil  exactly  the  same  but  not  treated  with  carbon  dioxid.  The 
control  sample  gave  a  depression  of  0.083°  C.  and  the  treated  soU  0.138°. 


Feb.  II.  i9i8         Freezing-Point  Method  and  Soil  Solution 


385 


A  similar  experiment  was  performed  on' the  5B  soil,  containing  16.7  per 
cent  of  moisture.  The  results  in  this  case  were  0.156°  depression  for  the 
control  and  0.218°  for  the  sample  treated  with  carbon  dioxid.  It  is 
evident  from  these  data  that  the  gas  has  a  striking  effect  in  increasing  the 
concentration  of  the  soil  solution.  Here,  again,  it  is  interesting  to 
observe  that  the  cropped  and  uncropped  soils  maintain  the  same  general 
relation  to  each  other,  even  after  the  treatment  described.  Water 
extractions  made  on  the  same  soils  showed  that  the  total  water-soluble 
material  was  considerably  increased  by  the  carbon-dioxid  treatment. 
Calcium  was  particularly  affected,  but  several  other  elements  were  also 
made  more  soluble. 


EFFECT  OF  DRYING  ON  CONCENTRATION  OF  SOIL  SOLUTION 

One  of  the  important  influences  which  may  affect  the  chemical  state 
of  the  soil  is  that  of  drying.  King  (4)  has  shown  that  drying  soils  in  the 
oven  causes  a  considerable  increase  in  the  quantity  of  water-soluble 
material.  Bouyoucos  and  McCool  (2)  have  pointed  out  that  the  evidence 
obtained  by  their  freezing-point  method  tends  to  the  same  conclusion. 
The  writers  have  desired  in  the  present  research  to  apply  this  method  to 
the  study  of  the  influence  of  air-drying  on  the  soils  from  the  tanks. 
Freezing-point  depressions  were  first  determined  upon  the  moist  soils, 
and  samples  were  then  spread  out  in  the  laboratory  and  allowed  to 
become  thoroughly  air-dried.  They  were  then  mixed  in  a  mortar  with 
suitable  quantities  of  distilled  water,  allowed  to  remain  in  closed  jars  for 
several  hours,  and  the  freezing  points,  as  well  as  the  percentages  of 
moisture  again  determined.     The  results  are  given  in  Table  V. 

Table  V. — Effect  of  drying  on  freezing-point  depressions 


Soil  No. 


lA. 
iB. 
4A. 
4B. 
5A. 
5B. 
8A. 
8B. 
9A. 

loA 
loB 
iiA 
iiB 


Moisture. 


Per  cent. 
22.  7 
20.  I 

17.  6 

17.  2 

17-7 

16.7 

10.  6 

10.  8 

9.4 

9-5 
14.  I 
13-4 

13- 5 


Freeziag-point  depres- 
sions. 


Orisrinal 
soil. 


°C. 

O.  056 
108 
oq8 
148 
,083 
156 
062 

135 
057 
087 
066 

125 
065 


After  dry- 
ing and 
retnoisten- 
ing.a 


0.050 
.  191 
.089 
.  114 
.068 
.  100 
.058 
.  142 
.083 
.  126 
.094 
•  136 
.  067 
.  110 


o  Calculated  to  same  moisture  content  as  original  sample. 


386 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  6 


It  will  be  noted  that  there  are  no  marked  or  consistent  changes  in  the 
freezing-point  depressions  as  a  result  of  the  air-drying.  In  samples  5A 
and  5B  water  extractions  (i  to  5)  were  made  with  the  result  that  only 
negligible  increases  occurred  in  the  quantity  of  any  of  the  constituents, 
with  the  exception  of  the  organic  matter  which  was  made  somewhat 
more  soluble.  It  may  be  suggested  that  a  further  increase  in  the  con- 
centration of  the  soil  solution  might  subsequently  be  caused  through 
stimulation  of  bacterial  activity  caused  by  the  additional  organic  nutrj- 
ment  made  available.  In  this  case  the  solubility  of  various  elements 
would  be  increased  indirectly  by  drying,  but  there  is  no  evidence  that 
nutrient  elements  become  more  soluble  by  the  simple  process  of  drying 
at  ordinary  temperatures. 

EXPERIMENTS  WITH  LEACHED  SOILS 

All  of  the  foregoing  work  has  consistently  indicated  that  a  given  soil 
may  yield  very  dififerent  soil  solutions  under  different  conditions.  In 
order  to  establish  the  lower  limits,  it  was  decided  to  leach  several  soils 
with  distilled  water  and  then  determine  the  depression  of  the  freezing 
point.  These  were  determined  immediately  after  leaching  and  at  inter- 
vals over  several  months,  during  which  time  the  soils  were  kept  at 
laboratory  temperature  in  closed  jars.  Table  VI  shows  the  results  ob- 
tained. 

Table  VI. — Effect  of  leaching  on  freezing-point  depressions 


Soil  No. 


2B. 


loB. 


8C. 


5C. 


Treatment. 


fSoo  gtn.  leached  with  2  liters  of 
\    water. 


.do. 


1,000  gm.  leached  with  10  liters 
of  water. 


.do. 


Freezing-point  depression. 


Before  leaching. 


Deter- 
mined. 


O.  117 


306 


207 


.  060 


Water. 


Per  cent. 


17.  2 


14-3 


21.5 


After  leaching. 


Date. 


Nov.  3 

Nov.  8 

Dec.  5 

Feb.  26 

May  2 

Nov.  3 

Nov.  8 

<!Dec.  5 

Feb.  26 

May  2 

Jan.  9 

Feb.  6 

iFeb.  26 

[May  2 

fFeb.  6 

{Feb.  26 

[May  2 


Deter- 
mined. 


°C. 

0.034 

.  021 

•033 
•034 
.030 
.  027 
.  020 
.032 
.  040 
.051 
.  004 
.  006 
.  012 
•043 
•  013 
.049 
.  064 


Per  cent. 


20.  4 


17.  2 


17.  6 


22.  4 


Feb.  II,  1918         Freezing-Point  Method  and  Soil  Solution  387 

It  seems  quite  clear  from  an  inspection  of  these  data  that  the  concen- 
tration of  the  soil  solution  may  be  reduced  to  an  extremely  low  point 
through  leaching  and  that  this  condition  is  maintained  over  a  considerable 
period  of  time.  Leaching  has  been  shown  by  Lipman  (5)  to  inhibit 
bacterial  action  markedly,  and  in  the  absence  of  such  activity,  it  seems 
that  the  ability  of  the  soil  to  recover  its  high  level  of  concentration  is 
very  limited. 

RELATION  OF  WATER  EXTRACTS  TO  SOIL  SOLUTION 

Before  entering  upon  the  final  discussion  it  is  desirable  to  consider  at 
this  point  certain  more  detailed  relations  of  water  extracts  to  the  soil 
solutions,  as  indicated  by  the  freezing-point  method.  The  work  of 
Bouyoucos  and  McCool  enables  us  to  obtain  very  useful  and  interesting 
estimates  of  the  concentration  of  the  soil  sola  i  ion,  but  obviously  it  is 
only  the  total  concentration  which  is  measured.  We  are  quite  unable  to 
predict  the  nature  of  the  substances  whose  resultant  effect  is  expressed 
by  the  lowering  of  the  freezing  point.  Only  by  some  method  of  water 
extraction  is  it  possible  to  gain  any  insight  into  the  proportions  of  indi- 
vidual elements.  Various  comparisons  between  osmotic  pressures  in 
the  soil  solution  and  concentrations  in  water  extracts  may,  however,  be 
made,  which,  in  certain  directions,  make  possible  very  interesting 
deductions. 

When  different  soils  are  to  be  compared,  the  first  inquiry  must  be 
concerned  with  the  amount  of  lowering  of  the  freezing  point  which 
occurs  for  each  100  p.  p.  m.  of  total  solids.  This  phase  of  the  question 
has  already  been  investigated  by  Bouyoucos  and  McCool,  who  found 
that  extracts  of  varied  types  of  soil  gave  very  nearly  constant  results, 
approximately  0.0032°  C.  depression  for  each  100  p.  p.  m.  of  total  solids 
in  the  solution.  The  writers  have  obtained  closely  similar  figures  for  the 
soils  used  in  this  investigation  as  is  made  evident  by  the  following  data : 

Freezing-point  depression  per  100  p.  p.  m.  of  total  solids 

Soil  No.  "C. 

2 o.  0032 

4 0031 

9 0032 

10 0030 

8 0034 

5 0030 

Average. 0032 

By  the  use  of  this  constant  the  approximate  strength  of  the  soil  solution 
in  terms  of  parts  per  million  of  total  solids  may  be  calculated,  and  this 
has  been  done  in  Table  VII. 


388 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  6 


Table  VII. 


-Comparison  of  total  solids  in  soil  solution  as  calculated  by  freezing-point 
and  extraction  methods 


Date. 


Nov.  15 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 
Dec.  21 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 
Dec.  27 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 
Feb.  12 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 


Soil 
No. 


4A 

4B 

5A 

sB 

6A 

6B 

7A 

7B 

12A 

12B 

14A 

14B 

lA 

iB 

2A 

2B 

3A 

3B 

8A 

8B 

9A 

9B 

loA 

loB 

iiA 

iiB 

4A 

4B 

SA 

5B 

6A 

6B 

7A 

7B 

12A 

12B 

14A 

14B 

lA 

iB 

2A 

2B 

3A 

3B 

8A 

SB 

9A 

9B 

loA 

loB 

iiA 

iiB 


Total 
mois- 
ture. 


Per  d. 

17-3 
16.  7 
16.8 

17-5 
24 
24 
14 

13 
14 

11.  4 

17-5 

16.  2 

23-4 
22.3 
20.  4 
20.  o 
20.  9 

20.  7 

12.  7 
12.  s 

11.  o 

10.  6 

15- I 
14.  6 

14-5 
13.6 

19-3 
18.0 
18.2 
18.0 

24-  5 
24.8 

15- 9 
14.7 
13-9 

12.  7 

17.  o 

16.  5 
22.  7 

21.  7 
20.  2 
20.  2 
20.  6 
19.9 
10.  6 
10.8 

9.4 

9-5 
14.  I 

13-4 
13-5 


Free 
water. 


Per  d. 
4-3 
3 
3 
4 
6 
6 
8 

7 
10. 

7 

10. 
8 

5 
4 
7 
7 


Freezing- 
point  de- 
pression. 


C. 

068 

082 

079 

088 

068 

lOI 

082 
151 

061 

138 

033 
095 
037 
095 
047 
087 
048 

060 

049 

III 

031 
058 

046 

079 
047 
134 
055 

121 
066 
106 
085 
103 
038 
108 
030 
059 
039 
068 
056 
108 
056 
lOI 
062 
120 
062 
135 
057 
087 
066 

125 
065 


Concen- 
tration of 

total 
dissolved 

solids 
calculated 

from 
freezing- 
point  de- 
pression. 


P.p. 


100 
600 
500 
800 
100 
200 
600 
700 
900 
300 
000 
000 
200 
000 
500 
700 
500 
900 
500 
500 
000 
800 
400 
500 
500 
200 
700 
800 
100 
300 
700 
200 
200 
400 
900 
800 
200 
100 
700 
400 
700 
200 
900 
800 
900 
200 
800 
700 
100 
900 
000 
200 


Concen- 
tration of 

total 
dissolved 

solids 
calculated 

from 
extracts 
to  free 
water. 


P.  p.  m. 
10,  000 


000 

500 
900 
900 
900 
400 
400 
500 
800 
800 
300 
800 
400 
800 
300 
200 
600 
200 
500 
700 
200 
800 
700 
600 
600 
300 
800 
700 
300 
100 
100 
700 
800 
400 
300 
400 
800 
800 
700 
700 
600 
300 
100 
800 
900 
600 
100 
900 
400 
900 
300 


Total 
dissolved 
solids  in 
100  grams 
of  moist 
soil  cal- 
cixlated 

from 
freezing- 
point  de- 
pression. 


Gm. 

O.  009 
.  010 
.  010 

•013 
.013 
.  021 
.023 
.036 
.  020 
.032 
.  010 
.  026 
.  006 
.013 
.  on 
.  019 
.  010 

.013 
.013 

.030 

•  005 

.008 

.  010 
.  017 

.013 
.032 

.  on 
.  019 
.  on 
.  017 
.018 
.023 
.  012 
.030 
.  009 
.  014 
.  on 
.  019 
.  007 
.  on 
.  012 
.023 

•013 
.  026 

■013 
.  029 
.  006 
.  009 
.013 
.  021 
.015 
.032 


Total 
sissolved 
solids  in 
100  grams 
of  moist 
soil  cal- 
culated 

from 
extracts. 


Gm. 
0.043 

•055 
.  029 
.  040 
.031 
.  046 
.030 

•  049 
.  016 
.036 
.028 

•055 
.  026 

•053 
.  021 
.044 
.  029 
.038 
.028 
.  064 
.  019 
.  029 
.  020 
.038 
.031 

•  050 

•  052 
.074 
.  040 
.  067 
.  042 
.  072 
.  027 
.068 
.  024 

■033 
.032 

•043 
•045 
.  069 
.049 
•055 
•03s 
.  042 

•  032 

•054 
.  029 

•039 
.030 
.051 
.036 
.  062 


Feb.  II,  1918         Freezing-Point  Method  and  Soil  Solution 


389 


Table  VII. 


-Comparison  of  total  solids  in  soil  solution  as  calculated  by  freezing-point 
and  extraction  methods — Continued 


Date. 

Soil 
No. 

Total 
mois- 
ture. 

Free 
water. 

Freezing- 
point  de- 
pression. 

Concen- 
trated 
total 
dissolved 

solids 
calculated 

from 
freezing- 
point  de- 
pression. 

Concen- 
trated 
total 

dissolved 
solids 

calculated 
from 

extracts 
to  free 
water. 

Total 
dissolved 
solids  in 
100  grams 
of  moist 
soil  cal- 
culated 

from 
freezing- 
point  de- 
pression. 

Total 
dissolved 
solids  in 
100  grams 
of  moist 
soil  cal- 
culated 

from 
extracts. 

Feb, 19 

Do 

4A 
4B 

5A 

ll 

6B 

7A 

7B 

12A 

12B 

14A 

14B 

Perd. 
17.6 

17.2 
18.  I 
17.2 
24.8 
25.0 
14.  6 
14.  0 
12.7 
II.  4 
16.  I 
15.0 

Perd. 
4.6 
4.2 

4.2 

7-1 
7-3 
8.6 
8.0 
8.7 

7-4 
8.6 

7-5 

"C. 
0.  C598 

.148 
.  106 

•115 
.066 
.  126 
.053 
•093 
•043 
.083 
.063 
.123 

P.  p.  m. 

3,  ICO 

4,  600 
3>300 
3,  600 
2, 100 

3,900 

1,  700 

2,  900 
1,300 
2,  600 
2,  000 
3>900 

P.  p.m. 
8,800 

16,  000 
7,  100 

II,  900 
4,900 
7,500 
2,900 
5,300 
2,300 
3,900 
3,900 
7,  200 

Gm. 

0.  014 
.  019 
.017 
.015 
.015 
.  029 
.015 
.023 
.  Oil 

.  019 

.017 

.  029 

Gm. 
.  041 
.  067 
•036 
.050 
•035 
•055 
.025 
.  042 
.  020 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

.  029 
•034 
•054 

Do 

Do 

Upon  several  occasions  the  total  solids  were  determined  in  the  i-to-5 
extracts  of  all  soils.  These  were  also  calculated  in  terms  of  concentra- 
tion of  total  soil  moisture,  as  well  as  of  free  soil  water.  From  the  depres- 
sions of  the  freezing  point,  calculations  were  made  of  concentrations 
expressed  as  total  solids  in  parts  per  million  of  the  soil  solution  with  the 
constant  just  described  (Table  VII).  The  first  and  most  striking  observ- 
ation based  on  these  results  is  that  the  relation  between  the  total  solids 
in  the  extracts  of  cropped  and  uncropped  soils  is  also  manifested  by  the 
depressions  of  the  freezing  point.  In  fact,  throughout  the  experiment 
this  correlation  was  noted.  The  fluctuations  pf  the  important  nutrient 
elements  in  the  water  extracts  already  described  by  Stewart  were  in 
general  agreement  with  the  changes  in  the  concentration  of  the  soil 
solution.  The  low  and  high  points  came  at  about  the  same  periods  of 
the  growing  season  in  both  cases. 

From  these  considerations  it  might  seem  justifiable  to  assume  that  the 
material  present  in  the  actual  soil  solution  forms  at  least  an  important 
fraction  of  the  total  solids  dissolved  in  the  i-to-5  extract.  It  is,  how- 
ever, highly  improbable  that  an  extract  should  consist  exclusively  of  the 
material  present  in  the  soil  solution.  Necessarily  an  additional  quota 
would  be  derived  from  other  more  or  less  soluble  substances,  and  the 
quantity  dissolved  would  depend  upon  the  conditions  of  extraction.  A 
number  of  years  ago  Mitscherlich  (6),  in  an  important  study  on  extracts 
of  soils  prepared  with  carbonated  water,  emphasized  this  point  of  view. 
His  procedure  consisted  in  making  extracts  with  varying  proportions  of 
carbonated  water,  ranging  from  5  to  i  up  to  30  to  i .     From  the  data  so 


390  Journal  of  Agricultural  Research  voi.  xn.  No.  6 

obtained  he  constructed  graphs  and  proved  that  with  increasing  quan- 
tities of  water  more  material  was  dissolved.  He  therefore  divided  the 
total  dissolved  solids  into  two  fractions :  (a)  That  part  soluble  even  with 
low  moisture  content,  and  (b)  the  additional  material  dissolved  by  the 
excess  of  solvent. 

A  number  of  difficulties  arise  in  connection  with  this  method  of  arriving 
at  the  concentration  of  soil  solution.  In  the  first  place,  it  is  not  safe  to 
assume  that  a  curve  based  on  one  range  of  extractions  can  accurately 
be  extended  to  cover  another  range  of  extractions.  Indeed,  the  experi- 
mental data  indicate  that  with  the  smaller  proportions  of  water  the 
curves  may  change  their  direction  very  appreciably  and  it  is  unfor- 
tunately impracticable  to  obtain  extracts  for  analysis  in  those  concen- 
trations which  correspond  to  optimum  moisture  conditions.  Another 
limiting  factor,  previously  neglected,  has  been  described  by  Stewart. 
This  concerns  the  differential  effect  of  the  solvent.  The  actual  solvent 
in  any  case  is  not  pure  water,  but  pure  water  plus  the  solids  already 
dissolved  in  the  soil  solution,  and  these  vary  enormously  with  changing 
conditions,  even  in  the  same  soil.  It  is  quite  obvious  that  this  factor 
would  modify  any  calculations  of  the  concentration  of  the  soil  solution 
based  on  water  extracts.  Moreover,  a  question  may  arise  with  regard 
to  the  relation  of  the  soil  extract  to  the  residual  solution.  It  is  possible 
that  the  equilibrium  may  not  be  the  same  for  the  low  and  high  propor- 
tions of  water,  especially  in  soils  of  colloidal  character. 

Notwithstanding  these  difficulties  met  with  in  attempting  to  predict 
accurately  the  concentration  of  the  soil  solution  from  soil  extracts,  it  is 
still  possible  to  make  certain  valuable  comparisons  of  the  two  methods. 
If  we  should  contrast  the  concentration  of  the  soil  solution,  calculated 
from  the  extracts  to  the  total  moisture  of  the  soil,  with  the  concentra- 
tions shown  by  the  freezing-point  method,  there  would  be  a  general 
similarity  of  magnitude.  Logically,  however,  a  comparison  is  much 
more  justifiable  when  the  extracts  are  calculated  not  to  the  total  soil 
water,  but  to  the  free  water,  in  the  sense  meant  by  Bouyoucos.  It  is 
then  apparent  that  the  concentration  of  the  soil  solution  calculated  by 
the  extraction  method  is  from  two  to  five  times  that  indicated  by  the 
depressions  of  the  freezing  point  (Table  VII).  In  other  words,  there  is 
a  considerably  greater  quantity  of  total  solids  dissolved  in  the  i-to-5 
extract  than  is  actually  present  in  the  soil  solution  at  any  given  moment. 
The  general  order  of  magnitude  of  the  two  quantities  is,  however,  evi- 
dently not  disproportionate,  the  material  extracted  from  the  sandy 
loams  averaging  only  about  twice  the  quantity  actually  present  in  the 
soil  solution,  as  shown  by  the  freezing-point  method. 

The  total  amounts  of  dissolved  material  in  100  gm.  of  soil  have  been 
calculated  by  multiplying  the  percentages  of  free  water  by  the  concen- 
trations in  terms  of  parts  per  million  of  total  solids.     If  we  use  the 


Feb.  II,  i9i8         Freezing-Point  Method  and  Soil  Solution  391 

freezing-point  depressions  as  a  basis,  it  appears  that  from  0.0 1  to  0.03 
gm.  of  total  solids  is  in  solution  for  each  100  gm.  of  moist  soil,  while 
the  total  solids  obtained  by  i-to-5  extractions  vary  from  0.02  to  0.06 
gm.  per  100  gm.  of  soil. 

In  order  to  obtain  approximate  estimates  of  the  material  dissolved  in 
an  extract  additional  to  that  present  in  the  soil  solution,  two  soils  (No. 
5  and  8)  were  subjected  to  the  procedure  of  successive  extraction.  One 
kgm.  of  each  soil  was  extracted  with  a  total  of  10  liters  of  distilled  water. 
Ten  extractions  were  made,  with  i  liter  of  water  for  each  extraction. 
The  results  of  the  10  extracts  are  plotted  in  figure  9.  It  will  be  noted 
that  the  first  extract  contained  a  very  much  greater  quantity  of  dissolved 
solids  than  any  subsequent  extract.  It  was  suggested  by  Stewart  that 
some  idea  of  the  extra  dissolved  material  might  be  formed  by  using  as  a 
basis  of  calculation  the  quantity  dissolved  in  each  liter  after  the  extracts 
had  become  relatively  constant.  Thus,  for  soil  5  roughly  150  p.  p.  m.  of 
total  solids  were  found  in  each  of  the  later  extracts,  and  for  soil  8,  100 
p.  p.  m.  By  assuming  that  the  rate  of  solubility  is  uniform,  in  a  i-to-5 
extract  there  would  be  5  times  150,  or  750,  p.  p.  m.  of  extra  material  dis- 
solved in  the  case  of  soil  5,  and  500  p.  p.  m.  for  soil  8.  The  total  solids 
dissolved  from  i  kgm.  of  soil  by  5  liters  of  water  aggregate  1,300  and 
1,100  p.  p.  m.  of  total  solids,  respectively.  This  means  that  not  more 
than  50  per  cent  could  have  originally  been  present  in  the  soil  solution. 
If  this  correction  were  applied  to  the  extracts,  the  agreement  with  the 
data  obtained  by  the  freezing-point  method  would  in  some  cases  be 
fairly  close,  but  usually  the  extracts  would  still  give  somewhat  higher 
results.  This  fact  is  reasonably  explained  on  the  assumption  that  the 
first  liter  of  solvent  has  dissolved  out  some  material  relatively  soluble 
although  not  actually  present  in  the  soil  solution,  while  the  later  extracts 
contain  only  difficultly  soluble  substances. 

GENERAL  DISCUSSION 

It  will  now  be  well  to  correlate  certain  broad  relations  which  may  be 
deduced  from  the  data  presented  in  this  paper  through  some  further 
reference  to  the  principal  investigation  of  this  series  by  Stewart.  It 
should  be  emphasized  again  that  neither  from  the  water  extraction  nor 
freezing-point  methods  is  there  any  evidence  that  the  soil  solution  has  a 
constant  composition.  On  the  contrary,  the  soil  solution  even  of  the 
same  soil  may  vary  greatly  under  different  conditions.  The  water- 
extraction  method  indicates  that  this  variation  may  occur  with  all  the 
principal  nutrient  elements  with  the  exception  of  phosphorus.  These 
data  accord  with  the  view  expressed  by  Bouyoucos  and  McCool,  that 
the  soil  solution  should  not  be  considered  as  saturated.  The  opposite 
conception  as  outlined  in  the  earlier  work  of  Whitney  and  Cameron  (9) 
is  not  upheld  by  the  present  investigation.  If  the  great  excess  of  nearly 
insoluble  minerals  in  a  soil  were  the  determining  factor  in  the  soil  solu- 
27809«— 18 7 


>5 
O 


PHOSPHATE(P0,)     SOILS 
L 


£34-567^9 
NUMBLR.    Of     EXTRACT. 

Fig.  g.-Graphs  showiag  the  results  of  successive  extractions  of  soils  s  and  8.    One  liter  of  water  to  i 
kgm.  of  soil  for  each  extraction  was  used. 


Feb.  II,  1918         Freezing-Point  Method  and  Soil  Solution  393 

tion,  then  its  concentration  in  the  cropped  soil  should  not  be  strikingly 
diminished  over  a  long  period  of  time  after  the  crop  is  removed,  which  in 
fact  is  the  case.  Also  the  soil  solution  in  a  leached  soil  has  a  greatly 
diminished  concentration  which  is  not  markedly  increased  by  long 
standing.  In  brief,  the  evidence  is  quite  opposed  to  the  theory  that 
there  is  an  immediate  restoration  of  equilibrium  when  the  soil  solution  is 
diminished  in  concentration  by  the  plant  or  other  agency. 

One  explanation  of  the  fluctuations  in  the  soil  solution  lies  most  prob- 
ably in  the  varying  nature  of  the  solvent,  especially  in  its  content  of 
carbon  dioxid,  which  may  well  be  one  of  the  preponderating  factors. 
The  effect  of  carbon  dioxid  on  the  solution  of  soil  minerals  has  been 
demonstrated  by  Mitscherlich,  and  the  present  investigation  has  indi- 
cated the  effect  of  carbon  dioxid  in  increasing  the  depression  of  the  freez- 
ing point.  It  is  generally  conceded  that  the  partial  pressure  of  carbon 
dioxid  in  the  soil  atmosphere  is  greater  than  that  of  the  outside  air. 
Russell  (7)  has  shown  that  the  content  of  carbon  dioxid  in  the  soil  may 
vary  greatly  at  different  times  of  the  year,  according  to  the  intensity  of 
bacterial  action.  In  addition  to  carbon  dioxid,  other  products  of  bac- 
terial activity  may  have  a  further  influence,  the  extent  of  which  is  still 
problematical. 

It  has  been  a  general  teaching  of  agricultural  art  that  soil  fertility  is 
increased  by  those  operations  which  tend  to  bring  about  optimum 
biological  conditions  in  the  soil.  The  experiments  recorded  in  this 
investigation  seem  to  afford  direct  evidence  that  the  soil  solution  may 
be  greatly  affected  by  the  activity  of  microorganisms  and  indirectly 
therefore  by  cultivation,  temperature,  organic  matter,  etc.  Finally,  the 
results  of  the  experiments  reported  in  this  paper  have  considerable 
significance  to  the  plant  physiologist,  since  comparisons  may  be  made 
between  osmotic  pressures  in  nutrient  solutions  and  in  the  soil  solution 
as  it  actually  exists  in  the  soil  under  conditions  favorable  to  crop 
growth.  It  will  be  observed  that  in  none  of  the  soils  did  the  plants 
obtain  their  nutriment  from  a  highly  concentrated  solution.  The 
general  range  of  concentrations  was  from  a  maximum  of  0.5  or  i.o 
atmosphere  to  a  minimum  of  o.i  or  0.2  atmosphere.  Under  the  condi- 
tions of  approximately  optimum  moisture  contents  there  was  no  sharp 
distinction  between  the  silty  clay  loams  and  the  fine  sandy  loams.  By 
lowering  the  moisture  content  of  the  clay  loams  considerably,  it  is  true 
that  a  very  high  concentration  of  the  soil  solution  results,  but  under 
moisture  conditions  favorable  to  plant  growth  the  solution  is  dilute. 
From  certain  of  the  soils  stored  in  the  bins  a  different  idea  of  the  con- 
centration of  the  soil  solution  might  be  obtained,  since  this  treatment 
has  in  some  cases  greatly  increased  the  soluble  material.  The  true 
conception  of  the  nature  of  the  nutrient  media  in  soils  can  only  be 
attained  when  soils  are  studied  throughout  the  season  under  conditions 
suitable  for  crop  growth. 


394  Journal  of  Agricultural  Research  Voi.  xii.  No.  e 

Preliminary  experiments  have  been  carried  out  with  the  use  of  sand 
and  water-culture  methods  with  the  nutrient  solutions  whose  concentra- 
tions are  comparable  with  those  of  the  soil  solution.  The  results  have 
indicated  that  the  osmotic  pressures  existing  in  the  soil  are  also  most 
favorable  to  the  growth  of  barley  in  culture  solutions.  When  con- 
siderably greater  concentrations  are  maintained,  decreased  yields  result, 
while  very  much  lower  osmotic  pressures  are  suboptimum.  Further 
experiments  are  now  in  progress. 

It  has  not  been  possible  in  this  investigation  to  make  any  general 
correlation  with  crop  yield,  although  it  is  worthy  of  note  that  the  two 
soils  of  lowest  production.  No.  9  and  12,  show  consistently  low  con- 
centrations of  the  soil  solution  and  also  yield  water  extracts  containing 
exceptionally  small  quantities  of  nutrient  elements.  In  the  discussions 
of  Stewart  and  Burd  the  relations  of  chemical  analysis  of  the  medium 
to  crop  production  are  discussed  more  completely. 

SUMMARY 

(i)  Freezing-point  depressions  have  been  determined  on  13  soils 
under  a  variety  of  conditions. 

(2)  The  concentration  of  the  soil  solution  has  been  found  to  vary  with 
the  season  and  also  as  a  result  of  treatment  with  carbon  dioxid,  leaching, 
incubation,  etc. 

(3)  The  growth  of  a  crop  markedly  diminishes  the  concentration  of 
the  soil  solution.  This  effect  is  still  evident  at  the  beginning  of  the 
following  season. 

(4)  The  soil  solutions  under  conditions  favorable  to  crop  growth  were 
found  to  be  very  dilute,  particularly  at  the  height  of  the  growing  season. 

(5)  Certain  general  agreements  between  the  extraction  and  freezing- 
point  methods  are  discussed. 

LITERATURE  CITED 

(1)  BOUYOUCOS,  G.  J. 

I917.    MEASUREMENT    OP    THE    INACTIVE,    OR    UNFREE,    MOISTURE    IN    THE    SOIL 

BY  MEANS  OF  THE  DiLATOMETER  METHOD.     In  Jour.  Agf.  Research,  v.  8, 
no.  6,  p.  195-217,  I  fig.     Literature  cited,  p.  217. 

(2)  and  McCooL,  M.  M. 

1916.  THE  Freezing-point  method  as  a  new  means   of   measuring  the 

CONCENTRATION  OF  THE  SOIL  SOLUTION  DIRECTLY  IN  THE  SOIL.      Mich. 

Agr.  Exp.  Sta.  Tech.  BuL  24,  p.  592-631,  2  figs. 

(3)  BURD,  J.  S. 

1917.  WATER   EXTRACTIONS   OF  SOILS   AS    CRITERIA    OF    THEIR  CROP-PRODUCING 

POWER.     In  Jour.  Agr.  Research,  v.  12,  no.  6,  p.  297-309,  i  fig. 

(4)  King,  F.  H. 

1905.   INVESTIGATIONS  IN  SOIL  MANAGEMENT.      U.  S.   Dcpt.  Agr.  Bur.  Soils  Bul. 

26,  205  p.,  7  figs.,  4  pis. 


Feb.  II,  1918         Freezing-Point  Method  and  Soil  Solution  395 

(5)  LiPMAN,  C.  B. 

1916.  PRELIMINARY    EXPERIMENTS  ON  SOME    EKECTS  OP  LEACHING  ON  THE  SOIL 

FLORA.     In  Soil  Science,  v.  i,  no.  3,  p.  291-297.     Literature  cited, 
p.  297. 

(6)  MiTSCHERLICH,  E.  A. 

1907.   EINE    CHEMISCHE    BODENANALYSE     PUR    PPLANZBNPHYSIOLOGISCHE    POR- 

SCHUNGEN.    In  Landw.  Jahrb.,  Bd.  36,  Heft  2,  p.  309-369,  10  figs., 
ipl. 

(7)  Russell,  E.  J. 

1915.   THE  ATMOSPHERE  OP  THE  soil:  ITS  COMPOSITION  ANDCAUSES OF  VABUATION. 

In  Jour.  Agr.  Sci.,  v.  7,  no.  i,  p.  1-45,  17  figs. 

(8)  Stewart,  G.  R. 

191 7.  effect  of  season  and  crop  growth  in  modifying  the  soil  extract. 

In  Jour.  Agr.  Research,  v.  12,  no.  6,  pp.  311-368,  24  fig.,  i  pi. 

(9)  Whitney,  Milton,  and  Cameron,  F.  K. 

1903.   THE  CHEMISTRY  OF  THE  SOIL  AS  RELATED   TO  CROP   PRODUCTION.      U.   S. 

Dept.  Agr.  Bur.  Soils,  Bui.  22,  71  p. 


ADDITIONAL  COPIES 

OP  THIS  PUBLICATION  MAY  BE  PROCTJBED  FROM 

THE  STTPEKrNTENDENT  OF  DOCUMENTS 

GOVEENMENT  FEINTING  OFHCE 

WASHINGTON,  D.  C. 

AT 

20  CENTS  PER  COPY 
SuBscBiPTioN  Price,  $3.00  Pes  Yeab 


——'"'"'■ ~-~-~— — -— —TnmMi— 


Vol.  XII  KEBRUARY  18,  1918  No.  7 

JOURNAL  OP 

AGRICULTURAL 

RESEARCH 


CONTENXS 

Page 

Efficacy  of  Some  Anthelmintics         _       -       _       -       _      397 
MAURICE  C.  HALL  and  WINTHROP  D.  FOSTER 

(Contribution  from  Bureau  of  Animal  Indastiy) 

Tobacco  Wildfire -449 

FREDERICK  A.  WOLF  and  A.  C.  FOSTER 

(Contribution  from  North  Carolina  Agricultural  Experiment  Station) 

Gipsy-Moth  Larvae  as  Agents  in  the  Dissemination  of  the 
White-Pine  Blister-Rust         -       -        -        -        -        -      459 

G.  FLIPPO  GRAVATT  and  G.  B.  POSEY 

( Contribution  from  Bureau  of  Plant  Industry) 


PDBUSHED  BY  AUTHOMTY  OF  THE  SECRETARY  OF  AGRICULTURE, 

WITH  THE  COOPERATION  OF  THE  ASSOCUTION  OF  AMERICAN 

AGRICULTURAL  COLLEGES  AND  EXPERIMENT  STATIONS 


WASHINOTON,  D.  C. 


WAtHINOTOM  I OOVERNMENT  PNINTINa  OfTIOS  1  1818 


EDITORIAL  COMMITTEE  OF  THE 

UNITED    STATES   DEPARTMENT    OF   AGRICULTURE    AND 

THE  ASSOCIATION  OF  AMERICAN  AGRICULTURAL 

COLLEGES  AND  EXPERIMENT  STATIONS 


FOR  THE  DEPARTMENT 


FOR  THE  ASSOCIATION 


KARL  F.  KELLERMAN,  Chairman    RAYMOND  PEARL* 


Physiologist  and  Associate  Chief,  Burtau 
of  Plant  Industry 

EDWIN  W.  ALLEN 

Chief,  Office  of  Experiment  Stations 

CHARLES  L.  MARLATT 

Entomologist  and  Assistant  Chief,  Burtau 
of  Enfomotogy 


Biologist,  Maine  Agricuitural  Experim$nl 
Station 

H.  P.  ARMSBY 

Director,  Institute  of  Animal  Nutrition,  The 
Pennsylvania  Stale  College 

E.  M.  FREEMAN 

Botanist,  Plant  Pathologist  and  Assistant 
Dean,  Agricultural  EiperimenI  Station  of 
the  University  of  Minnesota 


All  correspondence  regarding  articles  from  the  Department  of  Agricultiu'e  should  be 
addressed  to  Karl  F.  Kellerman,  Journal  of  Agricultiu^  Research,  Washington,  D.  C. 

*Dr.  Pearl  has  undertaken  special  work  in  connection  with  the  war  emergency; 
therefore,  until  further  notice  all  correspondence  regarding  articles  from  State  Experi- 
ment Stations  should  be  addressed  to  H.  P.  Armsby,  Institute  of  Animal  Nutrition; 
State  College,  Pa. 


JOINAL  OF  AGRIdMAL  RESEARCH 


Vol.  XII 


Washington,  D.  C,  February  i8,  1918 


No.  7 


EFFICACY  OF  SOME  ANTHEIyMINTICS 

By  Maurice  C.  Kali,  ^  formerly  Assistant  Zoologist,  and  Winthrop  D.  Foster,  Junior 
Zoologist,  Zoological  Division,  Bureau  of  Animal  Industry,  United  States  Department 
of  Agriculture 

INTRODUCTION 

Although  the  use  of  anthelmintic  treatment  is  an  old  practice  in  human 
and  veterinary  medicine,  the  efficacy  of  the  various  substances  employed 
as  anthelmintics  is  not  well  known.  What  information  is  available  is 
based  largely  on  clinical  observations,  efficacy  being  judged  to  a  large 
extent  on  a  consideration  of  the  improvement  or  lack  of  improvement  in 
the  patient's  health  after  treatment.  In  other  instances  the  efficacy  of 
the  treatment  has  been  checked  by  fecal  examinations  for  worms  passed 
and  for  eggs  persisting  in  the  feces;  but,  while  this  method  gives  real  and 
valuable  information,  it  is  somewhat  inexact.  The  methods  employed 
for  examining  feces  for  worms  passed  are  often  rather  casual;  and  nega- 
tive findings  in  examining  feces  for  eggs,  especially  when  a  small  number 
of  preparations  are  examined  for  only  a  few  days  after  treatment,  are 
not  always  conclusive. 

A  more  satisfactory  method  of  investigating  anthelmintic  efficacy  is  to 
administer  treatment  to  animals,  to  collect  all  feces  passed  for  a  number 
of  days,  to  recover  from  them  all  worms  present,  and  then  to  kill  the 
animals  and  collect  all  worms  remaining.  In  this  way  it  is  possible  to 
arrive  at  a  fair  idea  of  the  anthelmintic  effect  to  be  expected  from  a  drug, 
the  correctness  of  the  conclusions  depending,  of  course,  on  the  number  of 
experimental  animals  used  and  their  degree  of  infestation. 

While  it  is  thus  possible  to  express  the  efficacy  of  a  drug  in  the  form  of 
a  mathematical  ratio,  the  writers  are  fully  aware  that  such  ratios,  except 
when  based  on  extensive  data,  can  not  be  considered  an  accurate  index 
of  the  efficacy  of  the  drug,  since  many  factors,  not  entirely  within  con- 
trol, such  as  the  individual  reaction  of  the  animal,  the  amount  of  material 
in  the  alimentary  tract,  and  the  potency  of  the  drug,  all  enter  into  the 
problem. 

In  carrying  out  this  series  of  experiments  the  plan  of  the  writers  was 
to  test  as  many  drugs  as  possible  having  a  known  or  alleged  anthelmintic 
value,  abandoning  those  which  gave  no  results,  and  making  further 


Journal  of  Agricultural  Research, 

Washington,  D.  C. 

Iz 


'  Resigned  September  19,  1916. 


(397) 


Vol.  xn.  No.  7 
Feb.  18, 1918 
Key  No.  A— 34 


398  Journal  of  Agricultural  Research  voi.  xii.  no.  7 

experiments  with  the  more  promising.  It  is  therefore  possible  that  some 
of  the  drugs  tested  only  once  and  on  a  limited  number  of  animals  may 
have  more  anthelmintic  value  than  the  tests  indicate.  On  account  of 
the  extent  of  the  field  to  be  covered,  the  writers  did  not  feel  justified  in 
devoting  more  effort  to  those  drugs  which  gave  small  promise  of  success. 

Some  such  method  as  the  above  has  been  employed  by  previous  in- 
vestigators. Hutcheon  {iSgiy  made  numerous  tests  of  anthelmintic 
treatments  for  stomach  worms  in  sheep  and  goats  in  South  Africa, 
and  followed  the  treatments  by  post-mortem  examinations  to  determine 
the  immediate  effect  on  the  worms.  Stiles  {1901, 1902)  did  similar  work 
in  this  country,  and  a  number  of  veterinarians  and  stockmen  made  investi- 
gations involving  treatment,  post-mortem  examination,  and  clinical  ob- 
servation. But,  so  far  as  the  writers  are  aware  at  present,  a  detailed 
series  of  experiments  covering  the  treatment  of  animals  and  the  collection 
of  all  worms  from  the  feces  for  a  number  of  days  up  to  the  time  of  making 
a  post-mortem  examination  in  which  all  worms  remaining  were  collected, 
has  not  been  reported. 

In  this  work  the  following  method  was  pursued:  The  animals  were 
given  an  appropriate  dose  of  the  anthelmintic  to  be  tested,  the  method 
of  dosage  varying  with  the  purpose  of  the  experiment  and  the  substance 
to  be  tested,  preliminary  purgation  being  undertaken  or  omitted  as  de- 
sired. Treatment  was  usually  administered  in  the  morning,  and  all  feces 
were  collected  every  morning  thereafter  until  the  animal  was  killed.  The 
feces  were  washed  through  a  set  of  graded  screens  and  the  screens  exam- 
ined for  worms.  The  animal  was  usually  killed  the  morning  of  the  fourth 
day  after  the  administration  of  the  last  dose  of  the  anthelmintic,  and  all 
parasites  remaining  were  collected  and  counted.^  The  percentage  of  effi- 
cacy was  then  estimated  from  the  number  of  worms  found  on  post- 
mortem examination  and  that  number  plus  the  number  passed  after 
treatment.  When  preliminary  purgation  was  resorted  to,  the  feces  were 
collected  on  the  following  day,  the  day  of  administering  the  anthelmintic, 
to  ascertain  what  worms  if  any  were  removed  by  simple  purgation. 

For  convenience  we  have  arranged  our  experimental  data  in  three 
groups:  (i)  Simple  purgatives,  (2)  a  group  including  anthelmintic  shaving 
a  mineral  base  and  coal-tar  products,  and  (3)  a  group  covering  the  vege- 
table anthelmintics.  Tables  I  to  V  of  the  discussion  of  results  summarize 
the  results  of  the  experiments. 

'  Bibliographic  citations  in  parentheses  refer  to  "  Literature  cited."  p.  446-447. 

'  In  conducting  these  experiments  the  writers  occasionally  found  on  post-mortem  examination  dead 
worms  in  the  large  intestine  or  rectum  which  normally  are  found  in  the  small  intestine.  In  such  cases, 
as  the  worms  were  evidently  in  the  process  of  passing  out,  they  are  credited  as  being  removed  by  the  an- 
thelmintic. 


veh.  i8,  i9i8  Efficacy  of  Some  Anthelmintics  399 

EXPERIMENTS  WITH  SIMPLE  PURGATIVES 
CALOMEly 

For  worms  in  dogs. — Calomel  in,  fairly  large  single  doses  followed 
by  castor  oil  proved  inefficacious  as  a  remedy  for  ascarids  in  dogs. 
Four  infested  pups  weighing  3.6  to  4.5  kgm.  were  fasted  from  noon 
of  the  day  before  treatment,  given  4  to  5  gm.  of  calomel,  and  on  the 
day  following  treatment  given  1 5  mils  (milliliters)  of  castor  oil.  Three 
of  the  dogs  passed  no  worms  after  treatment ;  one  passed  a  single  as- 
carid  (Belascaris  marginata).  Two  of  the  dogs  which  passed  no  worms 
when  examined  post-mortem  showed  10  and  8  B.  marginata,  respec- 
tively. The  treatment  was  so  evidently  inefficacious  that  the  other 
two  dogs  were  not  sacrificed  for  post-mortem  examination.  The  results 
of  this  experiment  do  not  accord  with  the  claims  occasionally  made 
as  to  the  anthelmintic  value  of  simple  purgatives  and  cathartics. 

CASTOR  oily 

For  WORMS  in  dogs. — Although  the  efficacy  of  castor  oil  as  an  an- 
thelmintic was  not  tested  as  a  separate  experiment,  in  several  experi- 
ments it  was  given  as  a  preliminary  purge.  The  feces  passed  follow- 
ing its  use  were  examined,  and  the  worms  passed  compared  with  the 
number  found  post-mortem.  In  these  experiments  only  24  hours 
were  allowed  between  the  administration  of  the  oil  and  the  subsequent 
anthelmintic.  It  is  therefore  possible  that  the  oil  alone  might  have 
shown  greater  efficacy  if  the  feces  were  collected  for  a  longer  time. 
However,  in  our  experience  the  greater  number  of  worms  are  passed 
within  24  hours  after  the  administration  of  an  anthelmintic;  and  in  the 
case  of  a  drug  like  castor  oil,  which  usually  acts  within  a  few  hours  after 
its  administration,  it  is  not  likely  that  the  conclusions  of  the  writers  are 
much  in  error. 

In  II  experiments  in  which  castor  oil  was  used  as  a  preliminary  purge 
on  50  dogs  it  removed  27  B.  marginata  out  of  a  total  of  351.  Its  effect 
on  hookworms  (Ancylostoma  caninum)  and  whipworms  (Trichuris  de- 
pressiuscula)  was  practically  nil.  It  was  ineffective  more  frequently  than 
it  was  effective,  and  in  no  case  did  it  remove  all  the  ascarids  from  a 
single  dog.  It  has  therefore  little  to  recommend  it  as  an  anthelmintic. 
It  may,  however,  have  some  diagnostic  value  in  veterinary  practice  to 
confirm  a  clinical  diagnosis  of  ascarid  infestation  in  dogs  when  a  micro- 
scopic examination  of  the  feces  is  impossible,  though  it  is  evident  that 
even  for  this  purpose  it  would  not  always  be  reliable. 

When  worms  are  passed  as  the  result  of  the  increased  peristalsis  due  to 
a  simple  purgative,  there  is  always  the  suspicion  that  the  worms  were  dead 
and  would  have  passed  out  anyway  and  that  the  purge  merely  hastened 
their  removal. 


400  Journal  of  Agricultural  Research  voi.  xii,  no.  ^ 

EPSOM   SALT 

For  worms  in  dogs. — Epsom  salt  was  tested  once  as  a  preliminary 
purge  on  three  dogs  which  were  given  the  salt  in  molasses  at  the  rate  of 
4  gm.  to  each  5  kilos  of  weight.  It  had  no  effect  on  ascarids  {Belascaris 
marginata),  hookworms  (Ancylostoma  caninum),  whipworms  {Trichuris 
depressiuscula) ,  Dipyidium  caninum,  or  Taenia  pisiformis,  all  of  which 
were  found  present  at  post-mortem  examination.  Epsom  salt  is  ex- 
tremely distasteful  to  dogs  and  is  likely  to  promote  vomiting.  The 
salt  is  best  given  in  a  vehicle  like  molasses,  which  serves  to  disguise 
the  taste.  Aside  from  this  objectionable  feature,  the  salt  is  less  effective 
in  dogs  than  other  purgatives,  such  as  castor  oil  or  calomel,  and  is  not 
particularly  useful  in  canine  practice. 

For  worms  in  hogs. — Campbell  {19 17),  in  a  discussion  of  the  treat- 
ment of  swine  for  worms,  mentions  the  use  by  a  western  veterinarian 
of  Epsom  salt  dissolved  in  the  drinking  water.  He  says  that  the  hogs 
are  kept  away  from  water  for  12  to  24  hours,  in  warm  weather  not  so 
long,  in  cold  weather  even  longer.  One  ounce  (28.35  g^i-)  of  the  salt  is 
allowed  for  pigs  and  up  to  eight  ounces  (226.79  g^i.)  for  hogs,  then  twice 
this  quantity  is  added  to  allow  for  waste.  It  is  then  given  in  the  drinking 
trough.  The  pigs  being  thirsty  drink  it.  This  completes  the  treatment. 
A  test  of  this  treatment  was  made  by  the  junior  writer  on  two  pigs  weigh- 
ing about  23  kgm.  (50  pounds)  each.  The  animals  were  kept  in  separate 
pens  on  board  floors  and  deprived  of  both  food  and  water  for  24  hours. 
Each  pig  was  then  given  226.8  gm.  (8  ounces)  of  Epsom  salt  dissolved  in 
about  3.8  liters  (i  gallon)  of  water  and  placed  in  the  drinking  troughs. 
The  pigs  took  a  few  swallows  of  the  water  and  refused  to  taste  more, 
showing  their  resentment  by  overturning  the  troughs  and  spilling  the 
contents.  The  experiment  was  therefore  repeated  a  few  days  later,  the 
troughs  being  nailed  to  the  floor.  After  24  hours  without  food  or  water 
the  pigs  were  given  the  salts  as  before  and  as  in  the  previous  experiment, 
took  only  a  few  swallows,  refusing  to  drink  more.  Although  no  other 
water  was  given  them  for  the  next  24  hours,  the  troughs  remained  appar- 
ently as  full  of  the  salt  solution  as  at  the  beginning  of  the  experiment. 
As  the  pigs  had  now  been  without  food  and  had  drunk  practically  no 
water  for  48  hours,  the  salt  water  was  removed  and  fresh  water  and  food 
given.  The  examination  of  the  feces  during  the  next  three  days  revealed 
no  parasites  excepting  one  nodular  worm  (Oesophagostomum  dentatum). 
The  pigs  were  not  killed,  as  the  experiment  was  evidently  a  failure.  A 
microscopic  examination  of  the  last  fecal  samples  collected  showed 
numerous  ascarid  eggs. 

As  it  was  evident  that  the  Epsom  salt  in  a  solution  of  the  strength 
given  was  extremely  distasteful,  it  was  decided  to  repeat  the  experiment 
with  a  weaker  solution  which  the  pigs  might  be  induced  to  drink  if  suffi- 
ciently thirsty.     Accordingly,  the  pigs  were  again  deprived  of  food  and 


Feb.  i8, 1918  Efficacy  of  Some  Anthelmintics  401 

water  for  28  hours  and  then  given  a  solution  made  by  dissolving  0.45 
kgm.  (i  pound)  of  Epsom  salt  in  22.7  Hters  (6  gallons)  of  water.  As  the 
drinking  troughs  would  hold  only  3.8  liters  (i  gallon)  each,  each  trough 
when  full  contained  75.6  gm.  of  the  salt  in  solution.  It  was  intended 
to  refill  the  troughs  as  fast  as  they  were  emptied  until  each  pig  had  taken 
1 1.35  liters  (3  gallons)  of  the  solution,  or  226.8  gm.  of  Epsom  salt. 

As  in  the  previous  experiment,  the  pigs  merely  tasted  the  solution -and 
refused  to  drink  more.  No  other  water  was  given  them  for  the  next  28 
hours  during  which  time  they  had  two  meals  of  dry  feed.  As  the  pigs 
had  now  been  without  a  drink  for  56  hours,  the  experiment  was  closed. 
Another  microscopic  examination  of  the  feces  passed  at  the  close  of  the 
experiment  revealed  the  presence  of  ascarids  in  both  pigs. 

It  is  evident  from  the  above  that  a  solution  of  Epsom  salt,  even  when 
less  than  2  per  cent  in  strength,  is  so  distasteful  to  pigs  that  they  will 
not  drink  it  even  after  a  relatively  long  period  of  thirst.  It  may  also  be 
noted  that  11.4  liters  (3  gallons)  of  water  is  much  more  than  a  pig  of 
22.67  kgm.  (50  pounds)  weight  could  consume  in  a  day  even  if  very 
thirsty,  so  that  even  if  the  pigs  had  drunk  the  weaker  solution  the  treat- 
ment would  have  had  to  be  prolonged  for  two  or  three  days,  greatly 
,  decreasing  the  purgative  effect  to  be  expected  from  a  single  dose  of  a 
large  quantity  of  Epsom  salt. 

In  this  connection  it  may  be  stated  that  in  an  experiment  with  oil  of 
chenopodium  on  hogs,  conducted  some  years  previously  by  the  junior 
writer,  four  hogs  were  given  113.4  gm.  (4  ounces)  of  Epsom  salt  mixed 
with  a  bran-mash  feed,  all  the  animals  eating  together.  No  difficulty 
was  experienced  in  getting  the  animals  to  eat  the  mixture,  but  the 
amount  of  the  salt  allowed  to  each  hog  was  only  one-eighth  of  the  dose 
allowed  in  the  present  experiment. 

EXPERIMENTS  WITH  ANTHELMINTICS  OF  A  MINERAL  NATURE  AND 

COAL-TAR  PRODUCTS 

TARTAR    EMETIC 

For  worms  in  hogs. — Tartar  emetic  is  commonly  used  as  a  remedy  for 
roundworms  in  horses,  and  this  led  the  writers  to  test  its  effects  upon 
worms  in  hogs.  Two  small  pigs  were  used,  weighing  8.6  and  14.5  kgm. 
They  were  each  given  259  mgm.  of  tartar  emetic  dissolved  in  water, 
followed  after  a  short  interval  with  29.57  mils  of  castor  oil.  Winslow 
(1913)  gives  the  emetic  dose  of  tartar  emetic  for  hogs  as  4  to  10  grains  (259 
to  648  mgm.).  The  drug  in  this  case  exerted  no  emetic  effect.  One  of  the 
pigs  passed  5  ascarids  (Ascaris  suum)  after  treatment  and  showed  the 
presence  of  5  others  on  post-mortem;  the  other  pig  passed  none  and 
showed  19  on  post-mortem.  Owing  to  the  more  or  less  common  habit 
which  pigs  have  of  devouring  ascarids,  it  is  possible  that  the  second  pig 
also  passed  worms  which  were  eaten;  and  the  first  may  have   passed 


402  Journal  of  Agricultural  Research  voi.xu,  No.  7 

more  than  were  recovered  in  the  feces  saved  after  treatment.  A  few 
whipworms  (Trichuris  suis)  and  nodular  worms  (Oesophagostomum  den- 
iatum)  were  passed ;  a  few  of  the  former  were  found  in  one  of  the  pigs 
post-mortem,  not  looked  for  in  the  other,  and  about  250  and  500  nodu- 
lar worms  were  found,  respectively,  in  the  two  pigs  post-mortem. 

This  experiment  demonstrated  that  tartar  emetic  has  an  anthelmintic 
action  on  worms  in  swine,  but  further  trials  will  be  necessary  before  con- 
clusions can  be  drawn  as  to  its  efficacy.  The  method  of  administration 
used  in  the  experiments  is  not  likely  to  prove  suitable  in  practice. 

CHIvOROFORM 

Chloroform  has  been  more  or  less  used  for  some  time,  either  alone  or 
in  combination  with  other  substances,  as  an  anthelmintic.  Recently 
Alessandrini  {191 5)  has  commended  it  very  highly  for  use  against  hook- 
worms and  other  worms.  He  states  that  it  has  these  advantages :  Com- 
bined with  castor  oil  it  requires  no  special  preparation  of  the  patient;  it 
can  be  administered  in  a  single  dose;  it  does  not  cause  local  or  general 
disturbances  either  immediately  or  subsequently;  it  is  perfectly  well  tol- 
erated and  is  not  nauseous.  It  may  be  given  in  3-  to  4-gm.  doses  dis- 
solved in  olive  oil  or  castor  oil  and  is  thoroughly  efficacious  against  hook- 
worms, whipworms,  pinworms,  and  ascarids.  Chloroform  is  a  constitu- 
ent of  Hermann's  mixtures,  the  formula  of  one  of  which  is  given  on  page 
403.  Schultz  {1911)  finds  the  chloroform  to  be  the  active  anthel- 
mintic ingredient  of  Hermann's  white  mixture.  He  found  chloroform 
very  effective  against  hookworms  in  the  dog  and  says  of  it : 

It  has  proven  rapid  in  its  action  and  thus  far  not  followed  by  any  evil  after  effects. 
Should  the  chloroform -castor  oil  mixture  act  as  favorably  in  human  beings  as  it  has 
for  me  in  dogs,  it  will  prove  a  universal  worm  remedy  of  great  importance. 

In  view  of  the  irremediable  damage  resulting  from  overdoses  of  chloro- 
form, Schultz  is  inclined,  however,  to  favor  the  use  of  the  less  effective 
drug  thymol.  Billings  and  Hickey  (1916)  use  a  chloroform-castor-oil 
mixture  following  the  use  of  chenopodium  in  the  treatment  of  hookworm 
and  other  roundworm  parasites  of  aliens  detained  at  the  immigrant  sta- 
tion at  Angel  Island,  California.  For  an  adult  they  administer  20  mils 
of  a  mixture  of  chloroform  and  castor  oil  containing  i  .8  mils  of  chloroform. 
The  dose  is  graduated  according  to  the  apparent  age. 

The  writers'  experiments  with  chloroform  as  a  drug  for  expelling  hook- 
worms from  dogs  inclined  them  to  the  belief  that  it  was  extremely 
satisfactory  compared  with  thymol,  and  on  the  suggestion  of  Dr.  Charles 
Wardell  Stiles,  of  the  United  States  Public  Health  Service,  a  note  was 
sent  to  Dr.  Reid  Hunt,  of  Harvard  University,  asking  his  opinion  as  to 
the  danger  of  the  administration  of  chloroform.     In  his  reply  he  states: 

It  has  been  shown  ....  that  even  one  administration  of  chloroform  causes  distinct 
changes  in  the  liver,  from  which,  however,  animals  almost  always  completely  recover. 
I  presume  that  a  single  dose  is  efficient  and  ordinarily  harmless ;  but  I  think  that  if  the 


Feb.  i8. 1918  Efficacy  of  Some  Anthelmintics  403 

liver  or  heart  had  already  been  injured  or  diseased,  a  single  does  might  have  serious 
consequences.  It  would  seem  advisable  not  to  repeat  a  dose  for  several  days  at 
least. 

The  experiments  of  the  writers  with  chloroform  were  as  follows : 

For  worms  in  dogs. — Five  dogs,  weighing  from  5.5  to  17.7  kgm.,  were 
given,  in  the  morning  after  fasting  from  the  evening  of  the  day  before,  0.2 
mil  of  chloroform  per  kilo  of  body  weight  mixed  with  3  mils  of  castor  oil 
per  kilo  of  body  weight.  All  of  the  dogs  were  infested  with  hookworms 
{Ancylostoma  caninum),  four  very  lightly,  one  heavily,  three  had  a  few 
ascarids  (Belascaris  marginata),iour  a  few  (i  to  27)  whipworms  (Trichuris 
depressiuscula) ,  two  a  few  tapeworms  (Dipylidium  caninum),  and  one 
numerous  (250)  tapeworms  of  the  same  species.  The  dog  which  had  the 
single  hookworm  failed  to  lose  the  parasite  as  a  result  of  the  treatment, 
but  none  were  left  in  the  other  lightly  infested  dogs,  while  more  than 
half  the  hookworms  were  passed  by  the  heavily  infested  dog  (457  passed 
and  355  found  post-mortem).  One  dog  passed  a  single  ascarid  and  showed 
none  post-mortem,  two  others  showed  i  and  2  ascarids,  respectively, 
post-mortem,  but  had  passed  none  after  treatment.  Two  of  the  dogs  out 
of  the  four  infested  with  whipworms  passed  a  few  of  these  worms,  and  all 
four  showed  a  few  on  post-mortem.  After  the  treatment  no  tapeworms 
were  recovered  from  the  feces  of  the  dogs  infested  with  D.  caninum. 

In  this  experiment  chloroform  and  castor  oil  proved  rather  highly 
efficacious  in  removing  hookworms,  removing  all  in  three  cases  of  light 
infestation,  failed  to  remove  any  in  one  case,  and  removed  over  half  of 
the  parasites  in  a  case  of  heavy  infestation.  The  results  in  the  case  of 
ascarids  and  whipworms  are  not  striking.  In  the  case  of  D.  caninum 
there  is  no  evidence  that  chloroform  is  of  value  as  an  anthelmintic. 

The  failure  of  chloroform  to  remove  all  the  hookworms  from  the  heavily 
infested  dog  indicates  that  repeated  treatments  may  be  necessary  in 
cases  of  heavy  infestations,  but  in  view  of  the  dangers  attending  the  use 
of  chloroform  the  advisability  of  repeating  the  dose  is  questionable. 
Several  days  at  least,  perhaps  a  week  or  longer,  should  be  allowed  to 
elapse  to  allow  time  for  the  animal  to  recover  from  the  possible  ill  effects 
of  the  first  treatment.     Further  experiments  are  desirable. 

A  further  test  of  the  efficacy  of  chloroform  in  combination  with  other 
drugs  as  an  anthelmintic  for  hookworms  in  dogs  was  made  with  the 
remedy  in  the  form  of  Hermann's  mixture.  This  preparation  consists, 
according  to  Railliet  {19 15),  of  the  following: 

JOleoresin  of  male-fern 4  g™-*  or) 

|Oil  of  eucalyptus 2  gm.       J 

Chloroform 3  Z^- 

Castor  oil 4°  g™- 

He  suggests  its  use  as  a  substitute  for  thymol.  Although  the  prepa- 
ration is  intended  to  include  either  male-fern  (Dryopteris  filix-mas)  or 
eucalyptus  oil,  through  an  error  both  drugs  were  included.  In  conduct- 
ing this  experiment  two  objects  were  aimed  at :  (i)  To  determine  whether 


404  Journal  of  Agricultural  Research  voi. xii,  no.  7 

or  not  the  chloroform  was  the  sole  or  principal  cause  of  the  efficacy  of 
the  mixture  as  a  remedy  for  hookworms  {Ancylostoma  spp.) ;  (2)  whether 
the  combination  of  castor  oil  and  male-fern  was  likely  to  produce  symp- 
toms of  male-fern  intoxication,  owing  to  its  greater  degree  of  absorption 
when  combined  with  an  oil. 

After  a  fast  of  24  hours,  two  dogs  weighing  9.5  and  16.5  kgm.,  re- 
spectively, were  given  the  full  dose  of  Herman's  mixture  as  detailed 
above.  Two  other  dogs  weighing  9.6  and  12.4  kgm.,  respectively,  were 
given  the  mixture  without  chloroform.  Nearly  three-fourths  of  the 
hookworms  {Ancylostoma  caninum)  in  the  two  dogs  receiving  the  full 
mixture  were  eliminated  by  the  treatment,  one  dog  passing  16  A. 
caninum  after  the  treatment  and  having  only  3  left  at  the  post-mortem 
examination.  On  the  other  hand,  out  of  39  A.  caninum  in  the  two  dogs 
receiving  the  mixture  without  chloroform,  only  3  were  eliminated.  The 
mixture  without  chloroform  showed  some  slight  efficacy  for  whipworms, 
but  the  figures  are  not  striking.  Two  of  the  dogs  were  infested  with 
Dipylidium  caninum,  one  passing  a  number  of  fragments  with  several 
heads  within  30  minutes  after  dosing.  The  other  dog  passed  a  few 
segments.  Both  dogs  were  found  free  from  D.  caninum  at  post-mortem 
examination.  As  these  dogs  had  received  the  mixture  without  chloro- 
form, it  is  evident  that  the  taeniacidal  value  of  the  mixture  was  due  to 
the  male  fern  or  eucalyptol  and  net  to  chloroform.  This  may  reasonably 
be  considered  a  demonstration  of  the  efficacy  cf  male  fern  as  a  taeniacide. 

To  sum  up,  it  appears  from  the  above  that  Hermann's  mixture  (includ- 
ing both  eucalyptol  and  male-fern)  is  an  efficient  vermifuge  for  hook 
worms,  owing  largely,  if  not  entirely,  to  the  chloroform  content.  It  is 
also  efficacious  for  Dipylidium  caninum,  probably  on  account  of  the 
male  fern  it  contains.  It  seemed  to  have  some  slight  effect  on  whip- 
worms, but  the  evidence  of  this  was  by  no  means  convincing.  The 
combination  of  male-fern  and  castor  oil  seemed  to  have  no  deleterious 
effect  on  the  experiment  animals,  in  this  respect  supporting  the  opinion 
of  Lenhartz  (1902)  and  Seifert  {1908).  Further  experimentation,  how- 
ever, is  desirable  to  elucidate  this  point. 

For  worms  in  sheep. — Chloroform  was  tested  on  two  sheep  weigh- 
ing 25  and  28  kgm.,  respectively.  The  dose  for  each  was  mixed  with 
60  mils  of  castor  oil,  the  smaller  sheep  receiving  5  mils  of  chloroform 
(0.2  mil  per  kilo)  and  the  other  10  mils  (nearly  0.4  mil  per  kilo).  Both 
sheep  died  four  days  after  dosing  and  showed,  on  post-mortem  exami- 
nation, lesions  of  gastro  enteritis  and  pneumonia  in  the  congestive  stage. 
The  sheep  were  very  lightly  infested  with  parasites.  None  were  recov- 
ered from  the  feces  of  the  sheep  which  received  the  smaller  dose;  three 
stomach  worms  {Haemonchus  contortus),  five  nodular  worms  (Oesopha- 
gostomum  columbianum) ,  and  five  hookworms  {Bunostomum  trigonoceph- 
alum)  were  found  in  this  sheep  post-mortem.  The  sheep  which  received 
the  larger  dose  passed  nine  stomach  worms  and  one  nodular  worm  after 
treatment,  and  two  stomach  worms  were  found  post-mortem. 


Feb.  i8, 1918  Efficacy  of  Some  Anthelmintics  405 

The  chloroform  had  an  evident  anthelmintic  effect  on  the  sheep  which 
received  the  larger  dose.  The  fatal  results  of  the  treatment  in  both 
cases,  however,  indicate  that  chloroform  is  not  a  promising  anthelmintic 
for  use  on  sheep. 

ETHKR 

For  worms  in  dogs. — Four  dogs  weighing  from  2.38  to  4.89  kgm. 
were  given  ether  in  a  dose  of  0.8  mil  per  kilo  mixed  with  15  mils  of 
castor  oil  for  the  two  smaller  dogs  and  30  mils  of  castor  oil  for  the  two 
larger  dogs.  The  afternoon  of  the  day  before  treatment  the  dogs  were 
given  a  preliminary  dose  of  castor  oil,  and  were  not  fed  until  several  hours 
after  receiving  the  ether.  No  worms  were  recovered  from  the  feces  passed 
between  the  administration  of  the  preliminary  dose  of  castor  oil  and  the 
administration  of  the  ether  and  oil  mixture.  This  mixture  was  much 
resented  by  the  dogs,  and  all  of  them  were  salivated  by  it  and  showed 
more  or  less  evidence  of  collapse.  One  of  the  dogs  passed  a  few  tape- 
worm segments  (Taenia  pisiformis),  but  no  other  tapeworm  material  was 
recovered  from  the  feces ;  nor  were  any  tapeworms  found  post-mortem. 
No  hookworms  were  recovered  from  the  feces,  but  all  showed  infesta- 
tion on  post-mortem  examination,  the  number  of  worms  being  7,  21, 
233,  and  242,  respectively.  One  of  the  dogs  passed  i  ascarid  and  one 
passed  6.  At  the  post-mortem  examination  no  ascarids  were  found  in 
the  former  and  11  in  the  latter.  The  other  two  passed  no  ascarids; 
4  were  found  post-mortem  in  one,  but  i  of  these  was  in  the  rectum, 
evidently  about  to  be  passed;  i  ascarid  was  found  post-mortem  in  the 
other  of  these  two  dogs. 

The  conclusion  from  this  experiment  is  that  the  ether  exhibited  a 
rather  slight  anthelmintic  action  against  ascarids,  no  evident  action 
against  hookworms,  and  probably  was  instrumental  in  the  removal  of  a 
tapeworm  which  was  presumably  present  in  one  of  the  dogs,  in  view  of 
the  discovery  of  segments  in  the  feces. 

iodoform 

For  worms  in  dogs. — Among  the  remedies  prescribed  for  ascarids  in 
human  subjects  iodoform  has  been  occasionally  recommended.  Schid- 
lowski  {quoted  by  Seifert,  1885,  p.  g8)  gives  it  in  the  form  of  a  powder 
mixed  with  sodium  bicarbonate  in  doses  of  o.oi  to  0.06  gm.  three  times  a 
day,  followed  by  a  dose  of  castor  oil  on  the  last  day.  In  the  present  ex- 
periment it  was  intended  to  give  the  maximum  amount,  0.18  gm.,  in  one 
dose  to  each  of  four  dogs  ranging  in  weight  from  11. 8  to  14  kgm.  How- 
ever, through  an  error  in  weighing  the  drug,  the  dogs  were  each  given 
0.018  gm.  instead  of  0.18  gm.  This  error  was  discovered  afterward,  and 
the  next  day  the  full  dose,  0.18  gm.,  was  given.  Thus,  each  dog  received 
a  total  of  about  0.2  gm.  of  iodoform,  given  in  capsule  with  sodium  bicar- 
bonate. The  dogs  were  starved  for  24  hours  before  treatment ,  and  allowed 
one  meal  between  the  first  and  second  doses.  After  each  dose  29.6  mils 
of  castor  oil  were  administered.     One  Taenia  segment  was  passed  by  one 


4o6  Journal  of  Agricultural  Research  voi.xii,  N0.7 

of  the  dogs,  no  worms  being  recovered.  As  the  treatment  was  obviously- 
inefficacious,  only  two  of  the  dogs  were  killed.  The  post-mortem  exami- 
nation showed  these  animals  infested  with  various  numbers  of  intestinal 
nematodes  and  tapeworms.  As  far  as  can  be  judged  from  a  single  experi- 
ment, iodoform  is  valueless  as  an  anthelmintic  for  intestinal  parasites 
in  dogs,  even  when  given  in  doses  in  excess  of  that  prescribed  for  human 
subjects. 

COPPER  SUI.PHATE 

For  worms  in  sheep. — The  use  of  copper  sulphate  as  an  anthelmintic 
against  stomach  worms  in  lambs  was  developed  by  Hutcheon  {1892, 1895), 
who  reported  thousands  of  cases  of  its  successful  use  in  South  Africa. 
His  favorable  reports  were  based  largely  on  clinical  findings,  but  in  a 
number  of  cases  he  treated  animals  with  the  copper-sulphate  solution 
and  killed  them  a  short  time  afterward  to  determine  whether  the  worms 
in  the  stomach  were  dead  or  alive.  The  solution  he  used  is  approxi- 
mately that  which  would  be  obtained  from  0.45  kgm.  of  pure  copper 
sulphate,  powdered  fine  and  dissolved  in  35.96  liters  of  warm  water. 
Only  clear  blue  crystals  are  used,  and  it  is  best  to  powder  these  and 
then  to  dissolve  the  powder  in  a  small  quantity  of  hot  water  and  to  add 
cold  water  to  make  up  the  required  amount.  He  gave  the  solution  in 
the  following  doses. 

Lambs  3  months  old ^  oimce  (22.  i7mils). 

Lambs  6  months  old iK  ounces  (44.  56  mils). 

Sheep  12  months  old 2K  ounces  (73.  9   mils). 

Sheep  18  months  old 3  oimces  (88.  7    mils). 

Sheep  24  months  old 3K  ounces  (103.  5    mils). 

Stiles  and  others  have  tested  and  recommended  copper  sulphate  in 
these  doses. 

It  did  not  appear  to  the  writers  that  such  refinement  in  dosing  was 
called  for  in  treating  sheep;  therefore,  for  the  sake  of  simplicity  a  i  per 
cent  solution  was  made  up  and  administered  in  amounts  of  100  mils  to 
sheep  a  year  old  or  older,  and  in  amounts  of  50  mils  to  lambs  under  a 
year  old. 

Five  sheep  less  than  a  year  old  were  dosed  with  copper  sulphate,  two 
receiving  0.5  gm.  each  of  powdered  copper  sulphate  in  capsule,  and  three 
receiving  50  mils  each  of  the  i  per  cent  solution. 

The  two  sheep  receiving  powdered  copper  sulphate  in  a  capsule  passed 
a  few  stomach  worms,  and  on  post-mortem  showed  over  6,000  in  one 
case  and  over  4,000  in  the  other.  No  nodular  worms  were  recovered 
from  the  feces,  but  the  post-mortem  examination  showed  over  100  in  one 
case  and  nearly  200  in  the  other.  Both  showed  a  few  tapeworms 
(Montezia  sp.)  and  hookworms  post-mortem,  none  having  been  recovered 
from  the  feces  after  treatment.  Several  other  species  of  nematodes  were 
also  found  in  varying  numbers  at  the  post-mortem  examination. 

The  three  sheep  which  received  the  copper-sulphate  solution  passed, 
respectively,  120,  240,  and  314  stomach  worms,  and  showed  on  post- 


Feb.  i8,  1918 


Efficacy  of  Some  Anthelmintics 


407 


mortem  o,  49,  and  3,  respectively.  There  was  no  marked  effect  upon 
the  nodular  worms,  although  one  of  the  sheep  passed  2  worms  of  this 
species,  175  were  found  post-mortem,  and  143  and  21,  respectively,  were 
found  in  the  two  others,  from  whose  feces  none  were  recovered  after 
treatment.  Two  of  the 
sheep  showed  hook- 
worms on  post-mortem 
examination,  and,  al- 
though none  were  found 
in  the  other  one,  it  is 
quite  probable  that  none 
were  present  when  the 
animal  was  treated,  as 
none  were  reco  ve  red 
from  the  feces  after 
treatment. 

Evidently  the  pow- 
dered copper  sulphate 
in  capsule  exhibited  no 
anthelmintic  action.  On 
the  other  hand,  sup- 
porting the  experience 
of  Hutcheon,  Stiles,  and 
others,  a  i  per  cent 
copper  -  sulphate  solu- 
tion in  50 -mil  doses 
proved  very  efficacious 
in  the  removal  of  stom- 
ach worms.  It  had  no 
evident  effect  upon 
other  intestinal  para- 
sites. 

In  view  of  the  efficacy 
of  the  copper-sulphate 
solution,  a  test  was 
made  to  determine  how 
readily  and  easily  it 
could  be  administered, 
with  a  view  to  deter- 
mining whether  large 
numbers  of  animals 
could  readily  be  treated  with  the  solution.  A  dosing  apparatus  devised 
by  the  senior  writer  was  constructed  and  used  as  follows : 

A  I  per  cent  solution  of  copper  sulphate  was  made  up  and  placed  in 
a  small  water-tight  keg  (fig.  i).     At  the  side  of  the  keg,  near  the  bottom, 


Fio. 


I. — Apparatus,  with  control,  for  administering  copper- 
sulphate  solution  to  sheep. 


4o8  Journal  of  Agricultural  Research  voi.  xii,  no.  ? 

a  perforated  cork  with  a  glass  tube  through  the  perforation,  was  inserted 
in  an  auger  hole.  A  rubber  tube  was  connected  with  the  glass  tube. 
The  keg  was  placed  on  a  high  shelf  in  the  barn  and  the  rubber  tube  from 
the  bottom  of  the  keg  connected  with  one  of  two  glass  tubes  that  per- 
forated a  rubber  cork  in  the  bottom  of  a  graduated  glass  cylinder,  the 
top  of  the  cylinder  being  on  a  level  with  the  bottom  of  the  keg.  The 
graduated  glass  cylinder  was  fastened  by  wires  to  two  nails  driven  into 
the  wall  in  such  a  way  as  to  maintain  it  in  a  vertical  position. 

A  second  rubber  tube  was  connected  with  the  second  glass  tube  in 
this  cork,  this  tube  terminating  at  the  other  end  in  a  piece  of  metal 
tubing.  The  glass  cylinder  was  graduated  at  50-mil  intervals  and  had  a 
capacity  of  150  mils.  Close  to  this  cylinder,  pinch  controls  were  fastened 
on  the  rubber  tubing  leading  into  and  out  of  the  cylinder.  By  pinching 
the  control  on  the  rubber  tube  connecting  the  keg  with  the  glass  cylinder, 
the  copper-sulphate  solution  was  allowed  to  flow  from  the  keg  to  any 
desired  graduation  in  the  lower  cyHnder.  By  letting  this  close  and 
pinching  the  other  control,  the  solution  was  allowed  to  flow  from  the 
cylinder  to  the  metal  tubing  at  the  other  end  of  the  dosing  tube,  the  size 
of  the  dose  delivered  being  noted  on  the  cylinder.  The  metal  tubing 
was  held  in  the  sheep's  mouth  by  one  man,  while  another  man  controlled 
the  size  of  the  dose. 

In  actual  practice,  with  two  men  operating  this  apparatus,  as  noted,  and 
with  a  third  holding  the  sheep  and  a  fourth  bringing  them  up,  25  sheep 
were  given  50-mil  doses  in  15  minutes,  and  27  sheep  were  given  loo-mil 
doses  in  25  minutes,  a  total  of  40  minutes  for  52  sheep.  In  the  50-mil  dose, 
which  is  perhaps  enough  for  sheep  of  any  age,  since  it  combines  safety 
with  efficacy,  sheep  can  be  dosed  at  the  rate  of  i  ^  a  minute.  This  is  100 
an  hour,  or  800  sheep  for  an  eight-hour  day. 

It  should  be  noted  that  of  the  sheep  that  received  lOO-mil  doses  of  the 
solution,  the  equivalent  of  i  gm.  of  copper  sulphate,  2  sheep  died  in  the 
course  of  the  next  two  days.  These  sheep  were  10  months  old,  and  these 
d©ses  would  appear  to  be  too  large.  Neither  of  the  dead  sheep  showed 
any  indications  of  traumatic  pneumonia,  but  the  fourth  stomach  of  one 
of  them  was  much  congested. 

On  a  subsequent  occasion  the  50  sheep  surviving  from  this  experiment 
were  dosed  with  50-mil  doses,  using  the  apparatus  described.  No  bad 
results  of  any  sort  were  experienced.  The  sheep  have  been  similarly 
treated  subsequently,  and  occasional  post-mortem  examinations  of  the 
sheep  in  this  lot  show  almost  complete  freedom  from  stomach  worms. 

COPPER   SUIvPHATE 

For  worms  in  dogs. — In  view  of  the  efficacy  of  copper  sulphate  against 
stomach  worms  in  sheep,  an  attempt  was  made  to  determine  whether  the 
well-known  emetic  action  of  this  drug  would  entirely  prevent  its  use  as 
an  anthelmintic  for  dogs.     Four  dogs  were  each  given  0.5  gm.  of  copper 


Feb.  i8. 1918  Efficacy  of  Some  Anthelmintics  409 

sulphate  dissolved  in  10,  20, 30,  and  40  mils,  respectively,  of  water,  and  2K 
hours  later  they  were  given  14.79  to  29.57  mils  of  castor  oil  each.  There 
was  prompt  emesis  in  a  few  minutes  after  the  administration  of  the  cop- 
per sulphate.  Fecal  examinations  for  the  following  four  days  showed  only 
I  worm,  a  whipworm.  Because  of  the  evident  defects  of  the  treatment, 
the  dogs  were  not  killed  and  examined.  It  is  obvious  that  the  emetic 
action  of  copper  sulphate  precludes  the  use  of  this  substance  as  an  an- 
thelmintic for  dogs. 

GASOUNE 

For  worms  in  sheep. — Gasoline  has  been  extensively  used  as  a  treat- 
ment for  stomach  worms,  but  some  authorities  consider  that  there  are 
dangers  attending  its  use.     Stiles  (1901)  says  of  it: 

Gasoline  has  recently  gained  considerable  reputation  as  a  vermifuge.  I  have  used 
it  in  a  number  of  cases  and  have  found  the  claims  made  for  it  to  be  more  or  less  justi- 
fied. Three  objections,  however,  arise  to  its  use,  and  I  can  not,  therefore,  consider 
it  an  ideal  treatment.     These  objections  are: 

(i)  Not  less  than  three  doses,  and  usually  four  to  six,  are  required  to  expel  the 
worms..  Its  use  involves  a  great  expenditure  of  labor,  and  it  is,  therefore,  imprac- 
ticable on  the  large  ranches. 

(2)  While  several  doses  are  not  necessarily  injurious  to  the  stock,  still,  if  the  doses 
are  large,  repeated  drenches  cause  a  more  or  less  severe  congestion  of  the  bowels. 
Not  only  that,  but  repeated  handling  of  range  sheep,  with  the  necessary  preliminary 
treatment  of  withholding  food,  is  injurious  to  the  animals. 

(3)  If  used  on  animals  suffering  from  plevirisy,  it  is  likely  to  be  fatal.  I  have  had 
several  fatal  cases  of  this  kind. 

Luckey  {191 5)  says: 

Gasoline  for  acute  cases  is  a  specific.  One  dose  is  enough.  *  *  *  The  average 
man  can  not  give  gasoline  without  killing  the  animal.  One  can  not  give  a  sheep 
with  a  little  bit  of  pneumonia  gasoline  without  killing  it. 

Arbuckle  {1916)  says  of  himself  and  Joe  Wing: 

We  were  the  first  men  in  the  country  to  employ  the  gasoline  treatment  success- 
fully. Wing  had  heard  of  it  as  a  remedy  used  in  France.  We  were  also  among  the 
first  to  discover  that  this  was  not  a  practicable  treatment. 

Gasoline  is  usually  given  in  milk,  linseed  oil,  or  flaxseed  tea,  which 
makes  the  treatment  considerably  more  expensive,  and,  as  these  feeds 
are  not  always  available  in  large  quantity,  the  treatment  is  not  well 
adapted  to  the  needs  of  large  flocks. 

Stiles  {1901),  Ransom  {1907),  and  others  have  recommended  }{  ounce 
(7.39  mils)  of  gasoline  as  the  dose  for  lambs  and  }4  ounce  (14.79  mils)  as 
the  dose  for  sheep.  Coffey  {191 5)  recommends  larger  doses;  lambs  at 
weaning  to  get  X  ounce,  X  ounce,  and  ^  ounce,  respectively,  on  each  of 
three  successive  days;  sheep  to  receive  >^  ounce,  K  ounce,  and  i  ounce, 
respectively,  on  each  of  three  successive  days,  these  doses  to  be  given 
thoroughly  emulsified  in  5  ounces  (147.87  mils)  of  milk  linseed  oil,  or  flax- 
seed tea. 


4IO  Journal  of  Agricultural  Research  voi.  xii.no.  7 

In  the  writers'  first  experiment  four  lambs  were  treated.  Two  of 
them  weighing  21.09  9-°^  21.77  kgm.,  respectively,  were  given  7.39  mils 
of  gasoline  in  148  mils  of  milk  at  each  dose,  the  equivalent  of  the  dose 
commonly  recommended,  and  the  others,  weighing  26.3  and  22.2  kgm., 
respectively,  were  given  14.8  mils  of  gasoline  in  148  mils  of  milk  at 
each  dose,  the  equivalent  of  Cofifey's  doses  of  X»  Ki  ^^^  H  ounce  for 
lambs.  The  dose  named  was  given  to  the  lambs  on  each  of  three  suc- 
cessive days. 

The  two  lambs  which  received  7.39  mils  of  gasoline  passed  about 
one-fourth  of  the  total  number  of  stomach  worms  (Haemonchus  contortus) 
present,  no  hookworms  {Bunostomum  irigonocephalum) ,  of  which  only  a 
very  few  were  present,  and  several  nodular  worms  (Oesophagostomum 
columbianum) ,  of  which  a  considerable  number  were  found  post-mortem. 
There  was  no  evident  effect  on  various  other  species  of  nematodes  and 
tapeworms  present  in  one  or  both  of  the  lambs.  The  two  lambs  which 
received  14.8-mil  doses  of  gasoline  were  more  lightly  infested;  one  which 
passed  58  H.  contortus  showed  none  on  post-mortem ;  the  other  passed  4 
and  showed  2  post-mortem.  There  was  very  little  effect  on  hookworms 
present  in  small  numbers  in  one  of  the  lambs. 

All  four  of  the  lambs  showed  lesions  of  pneumonia  which  were  sus- 
pected to  be  of  traumatic  origin. 

In  the  second  experiment  three  sheep  were  treated.  The  larger  dose  of 
gasoline,  recommended  by  Coffey  and  found  most  efficacious  by  the 
writers,  was  employed  and  was  given  on  each  of  three  successive  days 
by  a  stomach  tube  passed  down  the  esophagus,  but  not  far  enough  to 
direct  the  fluid  into  the  rumen.  In  this  way  the  writers  expected  to 
avoid  the  possibility  of  causing  traumatic  pneumonia.  None  of  the 
sheep  passed  any  stomach  worms,  and  none  were  found  post-mortem, 
so  the  experiment  throws  no  light  on  this  subject.  The  treatment 
showed  a  slight  efficacy  for  hookworms  {Bunostomum  trigonocephalum) . 
It  failed  to  remove  any  nodular  worms  {Oesophagostomum  columbianum) 
from  one  sheep,  but  removed  all  specimens  of  this  species  from  another 
lightly  infested  sheep.  It  apparently  removed  all  specimens  of  Cooperia 
sp.  from  the  only  animal  infested  with  this  parasite. 

From  this  experiment  it  would  appear  that  gasoline  has  some  sUght 
effect  on  intestinal  worms  in  lightly-infested  sheep,  but  the  evidence  is 
not  sufficient  to  justify  its  use  for  this  purpose  without  further  experi- 
mentation. No  lesions  of  pneumonia  were  observed  in  this  experiment, 
although  the  stomachs  of  all  three  sheep  showed  lesions  suggesting  healing 
ulcers  with  traces  of  hemorrhage. 

As  the  question  of  the  efficacy  of  gasoline  as  a  vermifuge  for  stomach 
worms  in  sheep  and  the  likelihood  of  this  treatment's  causing  lesions  of 
pneumonia  and  gastritis  did  not  seem  to  be  settled,  a  further  trial  of 
gasoline  was  carried  out  by  the  junior  writer.     In  this  experiment  two 


Feb.  i8. 1918  Efficacy  of  Some  Anthelmintics  411 

sheep  were  used  and  treated  by  drenching  with  increasing  doses  of  gaso- 
line in  milk  as  recommended  by  Coffey  {19 13).  The  dosage  given  was 
15,  22,  and  30  mils  in  236.6  mils  of  milk  on  the  first,  second,  and  third 
days,  respectively.     The  same  dose  was  given  to  each  animal. 

The  first  sheep  passed  no  stomach  worms  and  had  none  on  post- 
mortem examination,  so  the  efficacy  of  the  treatment  in  this  case  is 
undetermined.  A  few  hookworms  only  were  passed.  The  second 
sheep,  which  was  lightly  infested  with  stomach  worms,  passed  none,  but 
passed  a  few  specimens  of  hookworms  and  nodular  worms,  forming  a 
small  percentage  of  the  total  number  present. 

In  this  experiment,  although  the  sheep  received  much  larger  doses 
than  in  the  previous  experiments  and  the  medicine  was  given  in  a  drench, 
there  were  no  evidences  of  pneumonia  or  gastritis. 

Considering  the  three  experiments  with  gasoline  as  a  whole,  involving 
the  use  of  nine  experiment  animals,  the  writers  find  that  this  treatment 
removed  over  one-fourth  of  the  stomach  worms  present  and  had  some 
slight  efficacy  for  hookworms  and  nodular  worms.  This  compares  very 
unfavorably  with  the  efficacy  of  copper  sulphate  for  stomach  worms 
when  given  in  solution.  The  gasoline  treatment  has  also  the  further 
disadvantages  that  it  must  be  given  three  times,  and  in  a  vehicle  such 
as  linseed  tea  or  milk,  which  is  an  additional  expense.  There  is  also  the 
possibility  of  causing  traumatic  pneumonia,  although  the  subsequent 
experiments  with  this  drug  indicate  that  gasoline  is  not  necessarily 
more  dangerous  as  a  drench  than  is  copper  sulphate. 

PETROLEUM   BENZIN 

For  worms  in  sheep. — As  already  stated.  Stiles  (1901)  was  favorably  ' 
impressed  with  the  use  of  gasoline  as  a  vermifuge  in  spite  of  certain  objec- 
tions that  he  notes.  It  was  suggested  by  Dr.  B.  H.  Ransom,  Chief  of  the 
Zoological  Division,  Bureau  of  Animal  Industry,  that  our  failure  to  get 
more  satisfactory  results  might  be  due  to  the  difference  in  the  commercial 
gasoline  of  the  present  day  and  that  used  by  Stiles.  It  was  recalled  that 
in  1 90 1,  when  Stiles  conducted  his  experiments,  the  automobile  industry  in 
this  country  was  in  its  infancy,  and  there  was  little  demand  for  gasoline. 
Consequently  the  petroleum  distillers  included  only  the  most  volatile 
hydrocarbons  in  gasoline,  reserving  the  heavier  fluids  for  kerosene,  their 
principal  product.  At  the  present  day,  with  conditions  reversed,  the 
distillation  temperatures  of  gasoline  have  been  greatly  extended,  with  a 
consequent  increase  in  the  specific  gravity  and  lessening  of  volatility. 

In  order  to  determine  whether  the  lessened  volatility  of  the  present- 
day  gasoline  was  related  to  its  inefficacy  as  a  vermifuge,  a  test  was  made 
by  the  junior  writer  of  petroleum  benzin,  U.  S.  P.  This  product,  dis- 
tilled between  45°  and  60°  C,  represents  only  the  most  volatile  hydro- 


412  Journal  of  Agricultural  Research  voi.  xii.  no.  7 

carbons  of  petroleum  and  is  probably  more  like  the  commercial  gasoline 
used  by  Stiles  {1901)  than  the  present-day  commercial  product.^ 

Two  full-grown  sheep  were  used  for  the  experiment.  The  animals  were 
starved  for  24  hours  and  then  given  15  mils  of  petroleum  benzin  in  150 
mils  of  milk,  on  three  consecutive  days.  They  were  drenched  through  a 
tube  held  in  the  mouth.  In  no  case  were  any  symptons  of  intoxication 
from  the  fumes  of  the  benzin  observed  during  or  after  its  administration, 
and  the  post-mortem  showed  no  lesions  which  could  be  attributed  to  the 
action  of  the  drug. 

The  chief  points  of  interest  in  this  experiment  are : 

(i)  The  apparent  superiority  of  refined  gasoline  (petroleum  benzin) 
over  commercial  gasoline  as  an  anthelmintic  for  stomach  worms  {Hae- 
monchus  contortus)  and  hookworms  {Bunostomum  trigonocephalum). 
It  proved  88  per  cent  efficacious  for  stomach  worms  while  the  latter  was 
only  35  per  cent  efficacious,  as  shown  by  the  summary  of  three  experi- 
ments. Its  efficacy  against  hookworms  is  still  more  marked.  It  was 
73  per  cent  efficacious  for  this  parasite,  while  gasoline  proved  only  5  per 
cent  efficacious. 

(2)  Its  entire  inefficacy  against  nodular  worms  (Oesophagostomum 
columhianum) .  No  worms  of  this  species  were  removed  by  petroleum 
benzin,  while  the  three  experiments  with  gasoline  showed  an  efficacy  of 
16  per  cent  for  this  parasite. 

The  post-mortem  examination  of  the  two  sheep  showed  that  they  were 
both  very  lightly  infested  with  stomach  worms  {Haemonchus  contortus), 
most  of  which  had  been  removed  by  the  treatment.  The  remedy  re- 
moved all  hookworms  (Bunostomum  trigonocephalum)  from  one  sheep 
and  more  than  half  the  number  present  from  the  second  sheep.  It  was 
entirely  inefficacious  against  other  intestinal  nematodes  and  tapeworms 
(Moniezia  spp.).  Its  failure  to  destroy  any  of  the  nodular  worms 
(Oesophagostomum  columhianum)  in  the  large  intestine  may  be  explained 
on  the  grounds  that  it  is  perhaps  more  readily  absorbed  than  commercial 
gasoline,  and,  hence,  was  largely  absorbed  before  reaching  the  colon. 
It  should  be  pointed  out  that  in  this  experiment  the  animals  were  starved 
for  24  hours  before  treatment,  while  in  the  experiments  with  gasoline  they 
were  not  starved.  This  may  have  had  something  to  do  with  the  results 
notwithstanding  the  slowness  with  which  the  rumen  becomes  empty  when 
sheep  are  starved. 

In  summary  it  may  be  stated  that  petroleum  benzin  showed  marked 
superiority  over  commercial  gasoline  as  an  anthelmintic  both  for  stomach 
worms  (H.  contortus)  and  hookworms  (B.  trigonocephalum).  While  it 
did  not  quite  equal  the  treatment  with  the  copper-sulphate  drench  as  a 

1  According  to  the  issue  of  the  Journal  of  the  American  Medical  Association  for  October  14,  1916,  the 
present  commercial  gasolineincludeshydrocarbonsdistillingatatemperatureashiKh  as  175°  C.  (Tydeman, 
F.  W.  I,.  NAPHTHALENE  IN  GASOLINE  FOR  AUTOMOBILES.  In  Jour.  Amer.  Med.  Assoc,  v.  67,  no.  16, 
p.  II75-) 


Feb.  i8,  i9i8  Efficacy  of  Some  Anthelmintics  413 

remedy  for  stomach  worms,  it  was  far  superior  to  anything  else  tried  as  a 
remedy  for  hookworms  {B.  irigonocephalum)  and  is  worthy  of  further 
experimentation  to  test  its  efficacy  on  more  heavily  infested  animals. 
For  the  treatment  of  stomach  worms  copper  sulphate  must  still  be  con- 
sidered superior,  not  only  because  it  is  somewhat  more  efficacious  but  on 
account  of  its  lesser  cost  and  the  fact  that  only  one  dose  of  copper  sulphate 
is  required,  whereas  three  doses  of  petroleum  benzin  are  advisable. 

PHENOLS 

The  phenols  a  re  a  group  of  organic  compounds  composed  of  hydroxy 
derivatives  of  the  benzene  series,  the  hydroxyl  radical  being  linked 
directly  to  the  nucleus.  The  refined  phenols  include  phenol  (CgHgOH), 
commonly  called  carbolic  acid;  cresol  (CgH^CHgOH),  commonly  called 
cresylic  acid  or  kresol;  and  the  higher  phenols.  The  term  "crude 
phenols"  is  used  in  general  to  designate  those  unrefined  mixtures  of  the 
phenols  proper  with  certain  hydrocarbon  oils  and  other  impurities  with 
which  they  become  associated  in  the  course  of  their  preparation,  whether 
from  coal-tar,  wood-tar,  or  blast-furnace  gases.  There  are  on  the  market 
numbers  of  trade  preparations  in  the  form  of  soaps,  powders,  ointments, 
or  liquids  which  contain  refined  or  crude  phenols  as  essential  constituents. 
A  number  of  these  liquid  phenol  preparations  have  been  used  and  recom- 
mended as  anthelmintics.  Some  of  these  liquids  are  insoluble  in  water, 
in  which  case  they  may  or  may  not  be  capable  of  being  emulsified,  and 
others  are  soluble  in  water. 

For  worms  in  sheep. — Three  preparations  were  tested,  which  may 
be  referred  to  as  A,  B,  and  C.  A  and  C  are  advertised  as  remedies  for 
stomach  worms,  while  B,  though  not  advertised  on  the  container  for  the 
treatment  of  stomach  worms  {Haemonchus  contorius),  has  since  been 
recommended  by  the  New  Zealand  Department  of  Agriculture  for  that 
purpose.  Two  sheep  were  treated  with  A  and  one  sheep  each  with  B 
and  C.  One  of  the  sheep  treated  with  A  passed  four  nodular  worms 
(Pesophagostomum  columhianum) .  This  sheep  and  the  one  treated  with 
B  died  the  day  after  treatment.     The  other  sheep  passed  no  worms. 

Of  the  two  sheep  receiving  A,  the  one  which  died  had  422  stomach 
worms  (Haemonchntortus)  us  coand  numerous  intestinal  nematodes.  The 
other  sheep  killed  four  days  after  drenching  had  no  stomach  worms  post- 
mortem and  few  intestinal  nematodes.  As  this  sheep  was  evidently  unin- 
fested  with  stomach  w;orms  (H.  contorius),  it  must  be  left  out  of  consider- 
ation. The  sheep  treated  with  C  passed  no  worms  and  had  8  stomach 
worms  (//.  contortus)  post-mortem. 

A  second  trial  was  made  of  A  on  two  sheep,  the  dose  given  being  the 
same  as  in  the  previous  experiment — i  tablespoonful  of  the  product  (14.8 
mils)  to  I  pint  (473  mils)  of  milk,  as  advertised  on  the  label.  The  sheep 
passed  no  worms,  and  at  post-mortem  examination  were  found  lightly 
infested  with  stomach  worms  and  intestinal  nematodes. 
27810°— 18 2 


414  Journal  of  Agricultural  Research  voi.xii.no.  7 

While  the  experiment  is  inconclusive  in  regard  to  the  efficacy  of  coal- 
tar  phenols  for  stomach  worms  in  sheep,  since  the  experiment  animals 
were  so  lightly  infested,  it  failed  completely  as  a  remedy  for  lightly 
infested  animals,  and  there  seems  to  be  no  reason  for  considering  that  it 
would  be  more  successful  in  heavily  infested  sheep. 

There  can  be  little  doubt  that  the  treatment  is  dangerous.  Two  out 
of  four  sheep  in  the  first  experiment  died,  the  post-mortem  showing 
lesions  of  pneumonia,  pleurisy,  and  gastritis.  In  the  second  experiment 
both  sheep  collapsed  after  drenching,  but  seemed  to  recover.  The  pneu- 
monic condition  observed  in  the  dead  sheep  may  be  attributed  to  getting 
some  of  the  fluid  in  the  lungs,  an  error  in  drenching,  to  be  sure,  but  one 
almost  impossible  to  avoid  when  giving  as  much  as  473  mils  of  drench. 
In  the  writers'  experience  sheep  will  usually  take  quietly  118  to  177 
mils  of  fluid  by  drench;  but,  after  more  is  given,  they  begin  to  struggle, 
making  the  drenching  increasingly  difficult.  Furthermore,  the  hemor- 
rhagic lesions  in  the  stomach  of  one  of  the  dead  sheep  would  seem  to  indi- 
cate the  absorption  of  phenols  through  the  gastric  mucosa. 

For  worms  in  dogs. — A  further  test  of  the  anthelmintic  efficacy  of 
phenols  was  made,  using  another  preparation,  which  is  recommended  as 
an  anthelmintic  for  worms  in  dogs,  the  dose  recommended  being  5  to  8 
drops  (0.3  to  0.48  mil)  in  a  tablespoonful  (14.79  mils)  of  castor  oil.  The 
writers  used  0.48-mil  doses,  administering  this  dose  to  each  of  two  dogs, 
weighing,  respectively,  8.6  and  11.3  kgm.  On  the  fourth  day  after 
treatment  one  dog  passed  two  ascarids  and  the  other  passed  a  headless 
chain  of  segments  of  Taenia  hydatigena.  This  latter  dog  showed 
an  infestation  with  ascarids  {Belascaris  marginata),  hookworms 
{Ancylostoma  caninum),  and  tapeworms  {Taenia  sp.),  and,  as  the 
treatment  had  been  unsuccessful  except  for  the  partial  removal  of  a 
tapeworm,  the  animal  was  not  killed.  The  other  dog  was  killed 
and  found  to  have  three  B.  marginata  and  three  A.  caninum. 
The  treatment,  therefore,  was  somewhat  efficacious  against  ascarids 
in  one  case,  but  entirely  inefficacious  in  the  other.  It  was  effective 
against  A.  caninum  in  both  dogs.  It  also  appears  to  be  ineffective 
against  Taenia,  since  it  did  not  bring  away  the  head,  but  this  is  a  point 
that  should  have  been  confirmed  post-mortem. 

We  may  conclude  from  the  experiments  on  dogs  and  sheep,  above 
recorded,  that  the  phenols  in  the  form  of  commercial  disinfectants  and 
dips  are  likely  to  be  of  little  value  and  dangerous  as  anthelmintics. 

EXPERIMENTS  WITH  VEGETABLE  ANTHELMINTICS 
OLEORESIN   ASPIDII 

For  worms  in  dogs. — Oleoresin  aspidii  is  the  classic  remedy  for 
use  against  tapeworm.  In  the  discussion  of  chloroform  as  an 
anthelmintic   for    hookworms    (page    403)    the  writers   have    already 


Feb.  i8, 1918  Efficacy  of  Some  Anthelmintics  415 

shown  that  male-fern  in  a  combination  known  as  Hermann's  mix- 
ture is  efficacious  against  Dipylidium  caninum  and  may  have 
some  slight  value  against  whipworms  (Trichuris  depressiuscula) 
and  hookworms  (Ancylostoma  caninum),  although  the  efficacy  of 
the  mixture  for  hookworms  is  largely  due  to  the  chloroform  contained 
in  it.  The  following  experiments  in  which  male-fern  alone  was  used 
corroborate  these  conclusions. 

In  the  first  experimental  test  of  the  drug  five  dogs  were  used  ranging 
in  weight  from  6.4  to  15.9  kilos.  The  treatment  was  preceded  by 
calomel  (65  to  194  mgm.,  according  to  the  weight  of  the  animals)  the 
afternoon  of  the  preceding  day  and  followed  about  45  minutes  after 
treatment  by  Epsom  salt  in  molasses.  One  to  three  mils  of  male-fern 
were  administered.  Within  an  hour  after  treatment  one  dog  had  passed 
a  mass  of  fragments  of  Dipylidium  caninum,,  including  at  least  four 
heads,  and  another  some  chains  of  segments  of  Taenia  sp.  When  the 
dogs  were  killed,  no  specimens  of  Taenia  sp.  were  found,  and  only  two 
specimens  of  D.  caninum,  one  of  which  was  in  the  colon,  evidently 
about  to  pass  out.  It  is  evident  that  the  remedy  was  entirely  efficacious 
against  Taenia  sp.,  since  the  dog  which  was  observed  to  pass  Taenia  seg- 
ments after  dosing  with  male-fern  was  found  uninfested  on  post-mortem 
examination.  It  was  slightly  less  efficacious  for  D.  caninum,  but  appar- 
ently removed  all  but  one  specimen.  It  is  possible  that  the  male-fern 
removed  all  individuals  of  D.  caninum  not  attached  by  burrowing  into 
the  intestinal  mucosa,  leaving  embedded  heads  to  renew  strobila.  This 
would  account  for  the  failure  to  find  more  than  four  heads  in  the  rela- 
tively large  mass  of  segments  passed. 

Male-fern  had  but  little  effect  on  ascarids  (Belascaris  marginata)  and 
removed  only  one-fourMi  of  the  hookworms  {Ancylostoma  caninum) 
present,  in  spite  of  the  fact  that  at  one  time  it  was  regarded  as  an  appro- 
priate remedy  for  hookworms  {A .  duodenale)  in  man  and  much  used  for 
the  purpose.  It  should  be  stated,  however,  that,  in  view  of  the  small 
number  of  hookworms  involved  in  this  experiment,  only  four  being 
present,  the  conclusion  that  male-fern  is  inefficacious  against  hookworms 
is  hardly  warranted.  None  of  the  three  whipworms  (Trichuris  depressi- 
uscula) present  were  removed  by  this  drug,  although  Miller  (1904) 
reported  the  successful  removal  of  whipworms  from  dogs  with  it. 

A  further  test  of  the  efficacy  of  oleoresin  of  aspidium  against  tapeworms 
in  dogs  was  conducted  by  the  junior  writer.  After  the  usual  24-hour 
fast,  two  dogs  weighing  20.4  and  11.34  kgm.  were  given  2.7  and  1.8 
mils,  respectively,  of  male-fern,  followed  by  162  mgm.  of  calomel. 

Prior  to  treatment  the  smaller  dog  had  been  repeatedly  seen  to  pass 
chains  of  proglottides  identified  as  Taenia  pisiformis,  while  only  a  few 
Taenia  eggs  had  been  seen  in  the  feces  of  the  larger  dog.  The  presence 
of  these  eggs  may  have  resulted  from  contamination  of  the  specimen. 


41 6  Journal  of  Agricultural  Research  voi.xii.  No.  7 

Following  the  anthelmintic  the  smaller  dog  passed  a  few  chains  of 
proglottides  the  first  day  after  treatment,  and  on  the  second  day  17 
fragments  and  4  heads.  No  tapeworms  were  passed  by  the  larger  dog 
during  the  four  days  following  treatment,  and  no  nematodes  were  passed 
by  either  dog.  The  feces  of  the  larger  dog  were  examined  for  Taenia 
eggs  four  days  after  administering  the  vermifuge,  and  were  found  nega- 
tive. As  this  dog  had  passed  no  Taenia,  and  as  the  original  presence 
of  tapeworms  was  somewhat  doubtful,  it  was  dropped  from  the  experi- 
ment. The  post-mortem  examination  of  the  smaller  dog  showed  no 
Taenia,  15  hookworms,  and  13  whipworms.  The  remedy  was  therefore 
entirely  efficacious  for  species  of  Taenia  and  entirely  inefficacious  for 
hookworms  (Ancylostoma  caninum)  and  whipworms  {Trichuris  depres- 
siuscula),  thus  confirming  the  opinion  derived  from  the  previous  experi- 
ment, that  male-fern  is  very  efficacious  for  tapeworms  in  dogs  and 
inefficacious  for  nematodes. 

For  worms  in  cats. — Six  cats  which  had  been  fed  with  Cysticercus 
fasciolaris  fonned  the  subject  of  this  experiment.  Each  cat  was  given 
0.8  mil  of  oleoresin  of  aspidium,  followed  by  130  mgm.  of  calomel  for  the 
four  largest  cats  and  97  mgm.  of  calomel  in  the  case  of  the  two  smaller 
cats.  Within  half  an  hour  one  of  the  cats,  which  was  weak  from  con- 
finement in  a  cage  and  which  was  suffering  from  coryza,  had  died.  The 
post-mortem  examination  revealed  an  intense  congestion  of  the  gastric 
mucosa.  The  characteristic  odor  of  male-fern  was  noticeable  in  the 
stomach,  but  not  in  the  intestines.  A  half -grown  individual  of  Taenia 
taeniaeformis  and  several  ascarids  (Belascaris  cati)  were  found  in  the 
small  intestine,  while  three  dead  B.  cati  were  in  the  colon.  Another  cat, 
which  was  weak  from  a  previous  infestation  with  mange,  passed  several 
chains  of  T.  taeniaeformis  shortly  after  the  administration  of  the  male- 
fern  and  died  three  days  later.  A  third  cat  vomited  one  T.  taeniaeformis 
and  one  B.  cati  and  still  had  one  B.  cati  post-mortem.  Two  of  the  six 
cats  were  uninfested  with  Taenia  spp.,  since  they  passed  no  tapeworms 
and  none  were  found  post-mortem.  Of  the  four  infested  animals  two 
died  from  the  treatment,  but  three  of  the  four  were  entirely  freed  from 
their  Taenia  infestation,  the  fourth  dying  before  the  remedy  had  an 
opportunity  to  reach  the  parasite. 

It  would  appear  from  the  above  that  male-fern  is  efficacious  in  removing 
tapeworms  from  cats,  but  that  it  is  apparently  more  toxic  in  the  case  of 
cats  than  with  dogs,  and  should  only  be  prescribed  for  healthy  animals. 
Probably  some  other  taeniacide  could  be  used  which  might  prove  as 
effective  and  less  dangerous.  It  should  be  noted  that  the  dose  of  male- 
fern  given  was  less  than  the  minimum  dose  15  minims  (0.9  mil),  recom- 
mended by  Winslow  {igis).  With  the  exception  of  the  one  ascarid, 
which  was  vomited,  the  treatment  had  no  effect  on  the  few  ascarids  {Be- 
lascaris cati)  and  hookworms  {Ancylostoma  caninum)  which  were  present. 


Feb.  i8, 1918  Efficacy  of  Some  Anthelmintics  417 

PELLETIERINE  TANNATE 

For  worms  in  cats. — Pelletierine  tannate  has  long  been  recognized 
as  an  effective  taeniacide.     According  to  the  U.  S.  Dispensatory  —  ^ 

The  efficacy  of  pelletierine  as  a  taeniacide  has  been  abundantly  confirmed,  and 
it  appears  to  be  established  that  the  tannate  is  the  most  effective  and  the  least  dan- 
gerous form  of  the  remedy, — probably  because  its  insolubility  prevents  its  rapid  ab- 
sorption and  enables  it  to  come  in  prolonged  contact  with  the  worm. 

A  test  of  this  drug  was  made  by  the  junior  writer  on  two  cats  weighing 
approximately  3  and  4  kilos  each.  Sixty-five  mgm.  of  pelletierine  tannate 
per  kilo  were  administered  in  capsules  to  each  animal  and  were  followed 
one  hour  later  by  25  mils  of  castor  oil.  Prior  to  the  administration  of 
the  anthelmintic  the  cats  were  starved  for  24  hours. 

The  animals  were  fed  shortly  after  the  administration  of  the  castor  oil. 

All  feces  passed  during  the  four  days  following  the  administration  of 
the  drug  were  negative  for  parasites.  Cat  2  vomited  immediately  after 
feeding,  or  a  little  more  than  an  hour  after  taking  the  pelletierine.  On 
post-mortem  examination  cat  i  had  29  Belascaris  cati  and  4  Dipylidiuw, 
caninum;  cat  2  had  i  Taenia  taeniae formis. 

In  this  experiment  pelletierine  tannate  proved  unsuccessful  as  an 
anthelmintic  for  species  of  Taenia,  Dipylidium,  or  ascarids  in  cats.  The 
complete  failure  of  the  drug  is  not  easy  to  understand.  In  the  case  of 
cat  2  the  drug  might  have  been  expelled  in  the  vomitus  before  it  had 
acted  on  the  Taenia,  but  this  is  unlikely,  since  the  vomiting  did  not  occur 
until  more  than  an  hour  after  the  ingestion  of  the  pelletierine. 

As  the  pelletierine  tannate  used  had  been  in  the  laboratory  for  at  least 
six  years  it  was  submitted  for  examination  to  the  Biochemic  Division  of 
the  Bureau  of  Animal  Industry.  The  report  was  returned  that  the  sam- 
ple responded  to  all  the  tests  for  pelletierine  taimate  given  in  the  Pharma- 
copeia. The  Pharmacopeia  specifies  that  the  drug  shall  be  kept  in  small 
well-stoppered  bottles  away  from  the  light,  a  condition  complied  with  in 
this  case.  There  seems  to  be  no  reason  to  suppose  that  the  pelletierine 
tannate  that  was  used  had  undergone  deterioration,  and  its  failure  to 
give  results  remains  unexplained. 

For  worms  in  dogs. — As  pelletierine  tannate  had  shown  a  surprising 
inefficacy  in  regard  to  tapeworms  in  cats,  a  further  test  of  this  drug  was 
made  by  the  junior  writer,  using  dogs  as  experiment  animals. 

The  drug  was  aarainistered  to  three  dogs  weighing  from  11.8  to  14 
kilos,  at  the  rate  of  16  mgm.  per  kilo  and  was  followed  one  hour  later  by 
castor  oil.  Previous  to  the  treatment  35  mils  of  castor  oil  were  given 
to  each  dog  as  a  preliminary  purgative,  and  the  animals  were  then  fasted 
for  24  hours. 

'  Wood,  G.  B.,  and  Bache,  Franklin,    thb  disebnsatory  op  thb  united  states  op  America,    ed.  19. 
p.  6°o.    Philadelphia  and  London,  1907. 


41 8  Journal  of  Agricultural  Research  voi.xii,  no.  7 

Following  the  administration  of  the  preliminary  purgative,  two  dogs 
each  passed  two  Taenia  proglottides.  Following  the  dose  of  pelletierine 
tannate  and  castor  oil,  one  dog  passed  five  Taenia  proglottides,  and  an- 
other dog  passed  two  proglottides.  No  tapeworm  heads  or  nematodes 
were  found  in  the  feces.  The  post-mortem  examination  revealed  a  few 
nematodes  and  Dipylidium  caninum,  but  no  Taenia  sp. 

It  is  evident  that  all  three  dogs  were  infested  with  species  of  Taenia, 
since  proglottides  were  recovered  in  the  feces  of  all  animals.  The  fact 
that  all  dogs  were  free  from  Taenia  on  post-mortem  examination  is  indi- 
cative of  the  efficacy  of  the  drug  as  a  taeniacide.  The  failure  to  recover 
the  worms  in  the  feces  probably  resulted  from  neglect  on  the  part  of  the 
attendant  collecting  the  feces. 

Pelletierine  tannate  apparently  proved  entirely  inefficacious  against 
Dipylidium  caninum  or  intestinal  nematodes  in  dogs,  since  none  of 
these  parasites  was  found  in  the  feces,  but  some  were  present  post- 
mortem. As  there  is  a  strong  probability  that  all  feces  were  not  col- 
lected, the  writers  can  not  be  certain  with  regard  to  the  total  inefficacy 
of  the  drug  for  these  parasites. 

The  sample  of  pelletierine  tannate  used  in  this  experiment  was  pur- 
chased from  the  same  manufacturer  and  at  the  same  time  as  the  sample 
used  in  the  previous  experiment  on  cats.  It  may  therefore  be  assumed 
to  have  the  same  potency  as  the  sample  used  in  the  previous  experi- 
ment, which  was  tested  for  purity  by  the  Biochemic  Division. 

ARECA  NUT 

For  worms  in  dogs. — Areca  nut  (Areca  catechu)  is  not  infrequently 
prescribed  as  an  anthelmintic  for  ascarids  in  dogs.  Railliet  {1915)  gives 
the  dose  of  areca  nut  as  2  to  4  gm.  combined  with  10  to  20  gm.  of  soluble 
cream  of  tartar,^  the  latter  presumably  being  used  merely  as  a  vehicle. 
In  this  experiment  6  gm.  of  areca  nut  and  36  gm.  of  soluble  cream  of  tartar 
were  formed  into  12  pills,  4  of  which  were  given  to  each  of  three  puppies 
weighing  between  1.8  and  2.3  kgm.  Previous  to  the  experiment  the  dogs 
were  each  given  1 4.79  mils  of  castor  oil,  which  failed  to  remove  any  worms, 
and  were  then  starved  for  24  hours.  All  the  dogs  were  infested  with 
ascarids  (Belascaris  marginata)  and  hookworms  {Ancylo stoma  caninum), 
as  shown  by  previous  fecal  examinations.  The  day  following  the  ad- 
ministration of  the  vermifuge  one  of  the  dogs  passed  4  B.  marginata 
and  was  found  dead  the  following  morning.  The  other  dogs  passed  no 
worms.  The  post-mortem  of  the  dog  that  died  showed  hemorrhagic 
areas  in  the  colon  and  feces  stained  with  blood.     The  other  organs  were 

•  Soluble  cream  of  tartar,  a  preparation  seldom  used  by  American  veterinarians,  is  boro-tartrate  of  potas- 
sium, made  by  boiling  together  four  parts  of  cream  of  tartar  with  one  part  of  boric  acid  in  a  large  amoimt  of 
water.  When  most  of  the  water  has  evaporated,  the  process  of  evaporation  is  completed  in  a  drying  oven 
and  the  resultant  salt  is  pulverized  and  stored  in  well-stoppered  bottles.  Soluble  cream  of  tartar  is  very- 
deliquescent,  and  hence  well  adapted  for  making  a  piUmass. 


Feb.  18. 1918  Efficacy  of  Some  Anthelmintics  419 

entirely  normal  in  appearance.  This  puppy  had  23  ascarids  {B.  mar- 
ginata)  and  a  few  hookworms  (^4.  caninum)  and  whipworms  (Trichuris 
depressttiscula).  The  other  dogs  had  varying  numbers  of  ascarids, 
hookworms,  and  whipworms.  The  treatment  removed  only  4  out  of  67 
ascarids  and  had  no  effect  on  the  hookworms  and  whipworms  present. 
It  should  be  stated  in  comment  that  the  areca  nut  used  was  ground  14 
months  previous  to  the  experiment  and  hence  may  have  lost  some  of  its 
potency.  This  leaves  open  the  question  as  to  the  efficacy  of  freshly 
ground  areca  nut,  but  it  appears  from  the  experiment  that  areca  nut 
ground  one  year  previously  is  not  a  very  efficacious  anthelmintic  for 
dogs,  and  may  possibly  cause  serious  digestive  disturbances.  While  it  is 
not  possible  to  determine  whether  the  colitis  noted  in  the  dead  dog 
resulted  from  the  drug,  it  should  be  stated  that  this  puppy  had  been 
weak  and  emaciated  for  some  time  previously  and  was  in  poor  con- 
dition to  undergo  anthelmintic  treatment. 

For  worms  in  poultry. — Areca  nut  is  quite  commonly  prescribed 
as  a  remedy  for  tapeworms  in  poultry.  The  following  experiment  was 
made  to  test  its  efficacy  against  tapeworms  and  incidental  nematode 
parasites  of  the  intestinal  tract  of  chickens : 

One  gm.  of  powdered  areca  nut  was  given  to  each  of  six  chickens, 
weighing  about  453  gm.  each,  the  dose  being  given  in  4  mils  of  olive 
oil.     Prior  to  the  experiment  the  birds  were  fasted  for  24  hours. 

The  total  number  of  worms  passed  following  the  vermifuge  were  20 
Heterakis  papulosa,  which  were  passed  by  three  of  the  six  birds  on  the 
first  day,  and  some  tapeworm  segments,  including  at  least  one  head 
with  these.  On  post-mortem  examination  large  numbers  of  tapeworms 
and  H.  papulosa  were  found,  besides  other  nematodes,  which,  from 
their  location  in  the  esophagus  and  proventriculus,  would  not  be  likely 
to  be  affected  by  anthelmintics.  Probably  more  tapeworms  were 
passed  than  were  counted,  since,  when  a  small  tapeworm  head  is 
passed  and  then  given  an  opportunity  to  dry,  it  is  very  difficult  to 
detect  even  by  a  careful  examination  of  the  feces.  But  at  best  the 
drug  seems  to  have  had  very  little  efficacy  against  H.  papulosa  or  tape- 
worms. 

The  areca  nut  used  in  this  experiment  was  ground  at  least  four  years 
previously.  This  leaves  open  the  question  as  to  the  efficacy  of  the 
freshly  ground  product.  If  it  is  true  that  the  drug  loses  its  potency 
after  grinding,  this  constitutes  a  serious  objection  to  its  use  in  commercial 
products,  since  it  is  unlikely  that  it  will  be  freshly  ground  when  purchased. 

santonin 

Santonin  is  the  classic  remedy  for  ascarids  and  forms  the  basis  of 
most  of  the  worm  remedies  for  children.  It  has  also  been  widely  used 
and  recommended  in  canine  practice.     As  this  drug  is  largely  of  Euro- 


420  Journal  of  Agricultural  Research  voi.xii.no.  7 

pean  or  Asiatic  origin,  the  present  price  of  santonin  is  almost  prohibitive 
in  veterinary  medicine,  and  its  place  is  being  taken  by  the  native  and,  in 
the  writers'  experience,  more  effective  drug  chenopodium.  Santonin  is 
usually  combined  with  or  followed  by  a  purgative  to  promote  elimina- 
tion of  the  parasite.  For  this  purpose  calomel  or  areca  nut  is  much 
used. 

Santonin  and  calomel  for  worms  in  dogs. — The  dose  of  santonin 
for  dogs  recommended  by  Winslow  {19 13)  is  from  i  to  3  grains  (65  to 
195  mgm.).  Taking  130  mgm.  as  the  dose  for  an  average  dog  weighing 
iokgm.,the  writers  gave  seven  dogs,  ranging  in  weight  from  3.8  to  9.5 
kgm.,  doses  graded  from  32  to  130  mgm.,  accompanied  by  the  same 
amount  of  calomel.  The  treatment  was  preceded  by  the  administration 
of  castor  oil,  29.57  mils  to  dogs  weighing  4.5  to  9.5  kgm.,  and  7.39  mils 
to  pups  under  4.5  kgm.  Food  was  withheld  the  previous  day.  Seven 
ascarids  (Belascaris  marginata)  were  passed  following  the  administration 
of  castor  oil  and  13  B.  marginata  and  two  whipworms  (Trichuris  depres- 
siuscula)  after  santonin  and  calomel.  At  the  post-mortem  examination 
it  was  found  that  about  one-fourth  of  the  ascarids  remaining  after  the 
action  of  the  castor  oil  later  had  been  removed  by  santonin  and  calomel. 
The  treatment  was  very  inefficacious  for  whipworms,  removing  2  out  of 
72,  and  entirely  inefficacious  for  Dipylidium  caninum  and  Taenia. 

It  would  appear  from  the  foregoing  that  santonin  and  calomel,  the 
remedy  usually  prescribed  for  ascarids,  is  not  very  efficacious  for  dogs 
in  single  doses. 

As  the  efficacy  of  the  santonin  as  shown  by  the  above  experiment 
was  considerably  less  than  had  been  expected,  considering  the  well- 
established  reputation  that  santonin  has  as  an  ascaricide,  a  second  experi- 
ment was  undertaken  to  determine  the  efficacy  of  santonin  and  calomel 
in  repeated  doses.     In  this  experiment  four  dogs,  weighing  from  1.8  to 

9  kilos,  were  given  graded  doses  of  equal  quantities  of  santonin  and  calo- 
mel, the  dose  varying  from  32  to  130  mgm.,  according  to  weight.  The 
first  dose  was  given  after  a  preliminary  fast,  and  the  second  dose  was  given 
two  days  after  the  first,  food  being  withheld  the  evening  before  the 
second  treatment  also,  the  same  dose  being  given  at  each  treatment. 
The  first  treatment  was  preceded  by  castor  oil,  which  failed  to  eliminate 
any  worms.  One  of  the  dogs  passed  no  ascarids  and  had  none  post- 
mortem, so  that  it  was  evidently  not  infested  and  must  be  left  out  of 
consideration.     From  the  three  remaining  dogs  it  eliminated  7  out  of 

10  ascarids,  a  distinct  gain  in  efficacy  compared  with  the  previous 
experiment,  in  which  a  single  dose  was  given.  In  this  case  the  drug 
also  showed  a  fair  degree  of  efficacy  against  whipworms  {Trichuris 
depressiuscula),  parasites  which  in  the  writers'  experience  are  difficult  to 
remove  with  any  degree  of  certainty  with  any  of  the  anthelmintics  tested. 
The  drug  was  entirely  ineffective  against  hookworms  and  Dipylidium 
caninum. 


Feb.  i8. 1918  Efficacy  of  Some  Anthelmintics  421 

Santonin  and  calomel  for  worms  in  hogs. — Before  the  present 
European  war,  when  the  price  of  santonin  justified  its  use  for  live 
stock,  it  was  much  recommended  for  roundworms  (Ascaris  suum) 
in  swine.  Since  the  separate  dosing  of  a  herd  of  swine  requires 
considerable  time  and  labor  and  the  handling  of  the  animals  is  likely 
to  excite  them  and  may  lead  to  their  being  injured,  it  i§  usually  the 
custom  to  mix  the  medicine  with  the  feed,  allowing  a  few  hogs  to  eat 
from  the  same  trough  at  one  time.  If  the  drug  is  well  mixed  with  the 
feed  and  the  animals  are  of  about  the  same  size,  it  is  assumed  that  they 
will  get  approximately  the  same  dose.  This  method  was  tested  by  the 
junior  writer  for  a  number  of  drugs.  In  this  series  of  experiments  the 
worms  passed  were  not  counted;  nor  were  the  animals  killed.  The 
feces  were  previously  examined  for  ascarid  eggs,  and  the  efficacy  of  the 
drug  was  judged  solely  by  the  presence  or  absence  of  eggs  in  the  feces 
three  days  after  the  administration  of  the  anthelmintic,  care  being  taken 
to  verify  all  negative  findings  by  one  or  more  repeated  fecal  examinations 
at  3-day  intervals. 

In  the  first  experiment,  after  a  preliminary  fast  of  24  hours,  three 
hogs  weighing  15.4  to  18  kgm.  were  given  santonin  and  calomel  allowing 
130  mgm.  of  each  per  hog.  The  drug  was  shaken  up  in  473  mils  of  water, 
poured  over  a  bran-mash  feed  and  thoroughly  stirred  in.  The  feces 
of  all  the  pigs  were  positive  for  ascarids  {Ascaris  suum  or  suilla)  at  the 
beginning  of  the  experiment.  Three  days  after  the  treatment  the  feces 
of  all  the  pigs  were  still  positive.  As  the  treatment  failed  to  free  any 
of  the  pigs  from  ascarids  {A.  suum),  it  was  repeated,  giving  194  mgm. 
of  santonin  and  calomel  instead  of  130  mgm.  The  second  treatment  was 
equally  without  results.  A  third  trial  was  made  in  which  the  pigs  were 
dosed  individually  after  preliminary  starvation,  each  pig  receiving  194 
mgm.  each  of  santonin  and  calomel.  This  treatment  resulted  in  the 
apparent  elimination  of  all  ascarids  from  one  of  the  pigs,  the  other  two 
remaining  infested.  These  hogs  were  then  dosed  individually  with 
259  mgm,  each  of  santonin  and  calomel,  all  subsequent  fecal  examinations 
being  negative. 

It  appears  from  the  above  that  santonin  and  calomel  should  be  given 
in  repeated  and  separate  doses  to  be  effective.  Probably  three  doses  of 
santonin  at  intervals  of  three  days  would  be  effective.  No  symptoms 
of  poisoning  were  observed  either  in  dogs  or  pigs,  although  the  pigs  were 
given  a  dose  more  than  twice  as  large  as  is  usually  prescribed  for  animals 
of  their  size.  According  to  the  Veterinary  College  at  Ames,  lowa,^  not 
over  4  grains  (259  mgm.)  of  santonin  should  be  given  to  the  largest  hog. 
It  would  seem  that  the  treatment  of  swine  with  santonin  requiring 
apparently  three  separate  doses  to  be  effective,  is  far  too  laborious  and 
costly  for  use,  especially  as  we  have  a  more  reliable  and  less  expensive 
drug  in  oil  of  chenopodium. 

•  Treating  pigs  for  worms.    In  Breeder's  Gaz.,  v.  64,  no.  8,  p.  315.    1913. 


422  Journal  of  Agricultural  Research  voi.  xii.no.  7 

Santonin  and  ar^ca  nut  for  worms  in  dogs. — Areca  nut  is 
frequently  prescribed  with  santonin,  since  it  furnishes  the  necessary 
laxative,  and  having  anthelmintic  properties  of  its  own  is  supposed 
to  be  an  adjuvant  to  the  santonin.  In  this  experiment  four  dogs 
weighing  from  2.7  to  6.8  kgm.  were  given  33  to  130  mgm.  of  santonin 
combined  with  0.78  to  1.94  gm.  of  areca  nut.  The  dogs  were  given  a 
preliminary  dose  of  castor  oil  which  removed  7  ascarids  (Belascaris 
marginata).  Following  the  administration  of  the  santonin  and  areca 
nut  3  B.  marginata  were  passed.  At  the  post-mortem  examination  30 
ascarids  {B.  marginata)  were  found  besides  several  hookworms  (Ancy- 
lostoma  caninum),  whipworms  (Trichuris  depressiuscula),  and  i  Dipy- 
lidium  caninum.  One  of  the  dogs  vomited  the  santonin,  which  may 
account  for  its  failure  to  act  in  this  case.  In  this  case  santonin  proved 
less  efficacious  for  ascarids  than  castor  oil,  and  had  no  effect  on  the 
other  intestinal  worms  present. 

Santonin  and  areca  nut  for  worms  in  hogs. — Combinations  of 
santonin  and  areca  nut  are  frequently  prescribed  for  pigs,  to  be  mixed 
with  the  feed.  The  following,  copied  from  the  Breeder's  Gazette,  is 
typical : 

Santonin 2}4  grains. 

Areca  nut i  dram. 

Calomel K  grain. 

Sodium  bicarbonate K  dram. 

This  quantity  is  considered  sufficient  for  a  100- pound  hog.  As  the  pigs 
experimented  with  weighed  only  24.5  kgm.  each,  two-thirds  of  the  amount 
prescribed  was  allowed  for  each  pig  and  mixed  with  the  feed,  after 
starving  the  animals  24  hours.  Three  days  after  the  experiment  the 
feces  were  still  positive  for  Ascaris  suum,  and  a  second  trial  was  made 
using  the  full  amount  prescribed  in  the  formula.  This  experiment  was 
also  unsuccessful  in  ridding  any  of  the  hogs  of  ascarids.  A  third  trial 
of  the  formula  was  made  on  a  pig  weighing  24.5  kgm.  with  the  object  of 
testing  the  efficacy  of  repeated  doses.  Five  times  the  amount  pre- 
scribed was  made  up  and  divided  into  seven  powders,  one  powder  being 
mixed  with  the  feed  every  morning  for  seven  consecutive  days.  At  the 
end  of  the  period  no  diminution  in  the  number  of  ascarid  eggs  in  the 
feces  was  discernible. 

As  far  as  these  experiments  can  be  analyzed,  it  appears  that  santonin 
and  areca  nut  offer  no  advantages  over  santonin  and  calomel,  and  that 
either  combination  must  be  repeated  and  given  in  individual  doses  on 
an  empty  stomach  to  be  efficacious.  The  last  experiment  illustrates 
the  inadvisability  of  giving  drugs  to  hogs,  mixed  with  the  feed,  since  in 
this  case,  a  drug  already  shown  to  be  efficacious  in  repeated  doses  was 
apparently  a  complete  failure  when  given  in  this  manner. 


Feb.  18. 1918  Efficacy  of  Some  Anthelmintics  423 

THYMOL 

For  worms  in  dogs. — Prior  to  the  beginning  of  the  great  European 
war,  thymol  was  the  classic  remedy  for  hookworm  in  man,  and  it  was 
only  after  the  war  had  made  this  drug  practically  unobtainable  that 
serious  efforts  were  made  to  find  a  substitute  for  it.  In  a  general  way 
it  was  regarded  as  a  satisfactory  drug,  and  we  expected  to  find  a  fairly 
high  coefficient  of  efficacy  for  it  in  our  experiments. 

In  our  first  experiment  thymol  was  administered  at  the  rate  of  130 
mgm.  per  kilo  of  body  weight,  giving  the  thymol  in  capsules  to  2  dogs 
and  in  drench  to  2  dogs ;  in  the  latter  case  the  thymol  was  dissolved  in  a 
small  amount  of  alcohol  and  this  added  to  water  to  make  a  fine  suspension. 
After  an  interval  of  a  half  hour  to  an  hour,  calomel  in  doses  of  about  65 
mgm.  per  2.5  kilos  of  weight  of  dog  was  given.  All  dogs  were  fasted  from 
the  previous  day. 

As  regards  hookworms,  thymol  removed  15  out  of  a  total  of  114  worms 
present,  being  entirely  ineffective  in  the  case  of  three  dogs,  two  of  which 
received  the  thymol  in  capsule  and  one  of  which  received  it  in  aqueous 
suspension.     In  one  dog,  however,  it  removed  15  out  of  16  hookworms. 

In  regard  to  ascarids,  thymol  removed  nearly  three-fourths  of  these 
worms  present,  a  very  satisfactory  showing  considering  that  thymol  is 
not  usually  considered  especially  valuable  as  an  ascaricide.  It  was  not 
efficacious  when  administered  in  aqueous  suspension.  Thymol  also 
exhibited  some  efficacy  against  whipworms  and  was  very  slightly  ef- 
fective for  Dipylidium  caninum. 

In  the  above  experiment  thymol  showed  a  low  degree  of  efficacy 
against  hookworms,  the  parasite  for  which  it  is  usually  prescribed,  and 
was  more  efficacious  against  whipworms  and  still  more  efficacious  against 
ascarids.  In  the  experience  of  the  writers  a  number  of  drugs  were 
found  to  be  occasionally  efficacious  against  whipworms,  while  at  other 
times  they  were  decidedly  inefficacious.  This  variation  may  perhaps 
be  explained  as  a  matter  of  accident,  depending  on  whether  or  not  the 
drug  succeeds  in  penetrating  to  the  cecum  in  its  passage  through  the 
alimentary  tract. 

An  objection  to  thymol  at  present  is  its  cost  and  scarcity. 

In  order  to  determine  whether  the  low  degree  of  efficacy  displayed 
against  hookworms  by  thymol  was  due  to  the  mode  of  administration, 
the  experiment  was  repeated.  This  time  the  animals  were  subjected  to 
preliminary  purgation  with  Epsom  salt,  and  the  administration  of 
calomel  after  treatment  was  delayed  for  five  to  six  hours.  The  thymol 
was  given  in  the  same  dosage,  130  mgm.  per  kilo  of  body  weight,  and  was 
given  in  aqueous  suspension. 

Three  dogs  weighing  4  to  9.3  kilos  were  used  for  the  experiment.  They 
were  given  Epsom  salt  at  the  rate  of  3.7  mils  for  each  5  kilos  of  weight, 
the  day  prior  to  the  administration  of  the  anthelmintic,  and  five  to  six 


424  Journal  of  Agricultural  Research  voi.  xii,  no.  7 

hours  after  giving  thymol  they  were  given  calomel  at  the  rate  of  65  mgm. 
for  each  2.5  kilos  of  body  weight. 

One  of  the  dogs  died  immediately  after  receiving  the  thymol,  and  the 
remaining  two  dogs  each  passed  one  ascarid.  The  post-mortem  exam- 
ination showed  that  one  or  both  dogs  were  infested  with  hookworms 
(Ancylostoma  caninum),  whipworms  (Trichuris  depyessiuscula),  asca- 
rids  {Belascaris  marginata),  and  Dipylidium  caninum.  The  treatment, 
therefore,  proved  an  entire  failure  for  all  worms  present  except  ascarids, 
for  which  it  showed  a  rather  low  degree  of  efficacy  in  this  experiment. 

As  the  thymol  used  had  not  proved  as  efficacious  against  hookworms 
as  might  be  expected,  a  sample  was  sent  to  the  Biochemic  Division  of 
the  Bureau  for  analysis.  An  analysis  was  made  by  Dr.  Custis,  who 
reported  that  it  responded  to  all  tests  for  thymol  and  showed  no  impuri- 
ties. Thymol  is  apparently  a  stable  phenol;  the  writers  are  unaware 
of  any  evidence  indicating  that  it  deteriorates. 

Since  thymol  in  a  single  dose  proved  so  ineffective  in  the  foregoing 
experiments,  an  experiment  was  undertaken  to  determine  its  efficacy 
in  repeated  doses.  Four  dogs  weighing  2.3  to  3.9  kilos  were  given  a 
preHminary  purgation  with  20  mils  of  castor  oil  on  the  day  previous 
to  treatment  and  were  then  fasted  for  24  hours.  Thymol  was  given 
at  the  rate  of  130  mgm.  per  kilo  of  body  weight,  and  this  dose  repeated 
three  and  five  days  later  for  a  total  of  three  doses.  After  the  second  and 
third  doses  of  thymol,  97  to  130  mgm.  of  calomel  were  given.  The  first 
dose  of  thymol  was  given  dissolved  in  alcohol  to  one  dog  and  resulted 
in  the  immediate  death  of  the  animal.  The  other  dogs  were  given  the 
thymol  in  aqueous  suspension. 

The  thymol  removed  8  out  of  the  16  hookworms  {Ancylostoma 
caninum)  present  in  the  dogs,  a  rather  poor  showing  in  view  of  the  pre- 
liminary fasting,  purgation,  and  repeated  doses,  but  decidedly  better 
than  in  the  previous  experiments.  It  removed  the  only  ascarid  {Belas- 
caris marginata)  present  and  showed  some  slight  degree  of  efficacy  against 
whipworms  {Trichuris  depressiuscida) . 

It  appears  from  these  three  experiments  that  thymol  to  be  at  all  suc- 
cessful as  an  anthelmintic  for  hookworms  {Ancylostoma  caninum)  must 
be  preceded  by  purgation  and  fasting  and  given  in  repeated  doses.  At 
least  two  such  courses  of  treatment  should  be  given  to  remove  the 
greater  number  of  the  worms  present,  and  further  treatment  combined 
with  prophylaxis  is  desirable.  As  a  matter  of  fact,  this  is  the  usual 
procedure  in  administering  thymol  in  human  practice,  and  it  is  usually 
reaUzed  by  physicians  that  it  is  very  difficult  to  remove  all  the  hook- 
worms present,  as  it  is  observed  that  hookworm  eggs  often  persist  in  the 
feces  after  repeated  thymol  treatments. 

In  the  three  experiments  considered  together,  thymol  removed  over 
half  thje  ascarids  present,  but  showed  very  little  efficacy  for  whipworms  or 


Feb.  i8,  i9i8  Efficacy  of  Some  Anthelmintics  425 

Dipylidium  caninum.  That  the  administration  of  thymol  to  dogs  is 
not  without  danger  is  shown  by  the  death  of  2  out  of  the  1 1  experiment 
animals. 

The  low  efficacy  and  the  danger  in  the  use  of  thymol  do  not  compare  well 
with  the  comparatively  high  efficacy  and  safety  in  the  use  of  chloroform, 
so  far  as  the  experiments  with  dogs  are  concerned.  While  the  writers 
are  impressed  with  the  dangers  in  connection  with  the  administration  of 
most  anthelmintics,  and  these  dangers  are  quite  impressive  for  chloro- 
form as  well  as  thymol,  nevertheless,  it  seems  that,  with  the  exceptions 
already  noted  in  the  discussion  of  chloroform  for  cases  where  there  is  heart 
trouble,  or  lesions  of  the  parenchymatous  viscera,  chloroform  is  not  only 
much  more  effective  than  thymol,  but,  in  therapeutic  doses,  is  safer. 

Turpentine 

Turpentine  is  a  remedy  very  commonly  advocated  for  use  against 
nematode  parasites,  especially  those  in  chickens,  horses,  and  swine.  The 
obvious  objection  to  its  use  is  its  well-known  injurious  effect  on  the 
kidneys. 

For  worms  in  poultry. — Since  the  treatment  of  chickens  for  worms 
is  for  obvious  reasons  so  seldom  undertaken  by  veterinarians  and  so 
commonly  by  owners  of  poultry  who  are  not  especially  trained  in  medi- 
cal lines,  it  is  not  surprising  that  the  dose  advocated  for  use  in  this  con- 
nection is  very  variable.  Some  writers  recommend  a  half  teaspoonful  of 
turpentine  in  an  equal  amount  of  olive  oil ;  others  recommend  i  to  3  tea- 
spoonfuls  of  turpentine  undiluted. 

An  experimental  test  of  the  efficacy  of  the  lighter  dose  of  turpentine 
against  worms  in  poultry  was  made  as  follows:  Six  chickens  weighing 
between  0.45  and  0.9  kgm.  were  given  2  mils  of  turpentine  mixed  with 
2  mils  of  olive  oil,  the  birds  being  fasted  from  the  previous  day  and  the 
dose  being  followed  at  once  with  8  mils  of  castor  oil.  About  five  hours 
after  treatment  all  the  birds  had  passed  some  feces,  the  feces  having  an 
odor  of  turpentine. 

The  treatment  appeared  to  be  fairly  satisfactory  for  the  large  round- 
worm (Ascaridia  perspicillum)  in  the  small  intestine  of  chickens,  since  it 
removed  more  than  three-fourths  of  the  worms  present,  as  shown  by 
post-morten  examination.  It  had  little  effect  in  cases  of  infestation  with 
large  numbers  of  cecum  worms  (Heterakis  papulosa)  with  which  chicks 
are  frequently  infested.  According  to  the  experience  of  the  writers,  this 
worm  is  difficult  to  remove  with  any  anthelmintic,  since  its  location 
protects  it  to  a  greater  or  less  extent  from  contact  with  the  drug. 

Turpentine  was  equally  inefficacious  as  a  remedy  for  tapeworms  in 
fowls,  removing  only  8  out  of  444.  It  should  be  stated,  however,  that  it 
is  very  difficult  to  count  the  tapeworm  heads  which  may  be  present  in  the 


426  Journal  of  Agricultural  Research  voi.  xii,  No.  7 

feces.  On  account  of  their  minute  size  they  dry  quickly,  and  unless  the 
feces  are  examined  very  soon  after  they  are  passed,  many  specimens  will 
be  unrecognized.  The  remedy,  therefore,  may  have  been  more  effective 
for  tapeworms  than  the  figures  indicate.  Its  principal  advantage,  how- 
ever, seems  to  be  as  a  remedy  for  roundworms  (Ascaridia  perspicillum). 

For  worms  in  dogs. — Four  dogs  weighing  2.5  to  y.j  kilos  were  given 
preliminary  purgation  with  castor  oil,  and  dosed  with  turpentine  at  the 
rate  of  i  mil  per  kilo  of  weight.  The  turpentine  was  given  mixed  with 
castor  oil,  15  to  30  mils,  according  to  the  weight  of  the  animals. 

The  preliminary  purgation  with  castor  oil  did  not  bring  away  any 
worms.     One  of  the  dogs  died  on  the  fifth  day  after  treatment. 

Only  one  worm,  an  ascarid,  was  passed;  to  this  should  be  added  one 
ascarid  found  in  the  rectum  of  the  dog  that  died,  making  two  ascarids 
removed  by  the  turpentine. 

In  view  of  the  lack  of  results  from  the  treatment,  the  dogs  in  this 
experiment  were  not  killed.  Just  previous  to  the  experiment,  micro- 
scopic examination  of  the  feces  for  eggs  had  shown  that  all  of  the  dogs 
were  infested  with  either  hookworms  or  ascarids  or  both.  On  the  last 
day  of  the  experiment  the  feces  were  again  examined  for  eggs,  and  eggs 
of  the  same  worms  were  found  to  be  still  present. 

All  the  dogs  showed  symptoms  of  distress  immediately  after  the 
treatment  with  turpentine,  the  principal  feature  being  a  temporary 
paralysis  of  the  hind  quarters. 

It  may  be  concluded  that  turpentine  in  doses  of  i  mil  per  kilo  of 
weight  is  not  very  efficacious  against  ascarids  in  dogs,  is  entirely  ineffi- 
cacious against  hookworms,  gives  rise  to  pronounced  suffering  and  tem- 
porary paralysis  of  the  hindquarters,  and  may  kill  the  dog. 

For  worms  in  hogs. — Turpentine  is  frequently  prescribed  as  a 
remedy  for  roundworms  {Ascaris  suum)  in  swine.  It  is  often  given  in 
repeated  doses  mixed  with  the  feed.  A  better  way  is  to  make  an  emul- 
sion of  the  turpentine  with  equal  parts  of  flaxseed  decoction.  This  is 
more  easily  miscible  with  the  feed,  and  avoids  the  burning  caused  by 
the  ingestion  of  pure  turpentine. 

An  emulsion  of  turpentine  and  flaxseed  decoction,  made  by  boiling 
85  gm.  of  flaxseed  in  296  mils  (10  ounces)  of  water,  straining,  and  adding 
an  equal  amount  of  turpentine,  was  fed  to  three  hogs  by  the  junior  writer. 
The  hogs  weighed  from  45.36  to  68.04  kgm.  The  equivalent  of  7.4  mils 
of  turpentine  per  hog  or  44.36  mils  of  the  mixture,  was  mixed  daily  in 
the  feed  of  the  three  hogs  for  seven  days.  At  the  end  of  seven  days  all 
the  hogs  were  listless,  refused  to  eat,  and  were  constantly  voiding  small 
amounts  of  urine,  the  nephritic  symptoms  continuing  for  a  week  after 
the  treatment  stopped.  All  the  feces  of  the  hogs  contained  ascarid 
(Ascaris  suum)  eggs  after  the  experiment  was  concluded. 


Feb.  i8, 1918  Efficacy  of  Some  Anthelmintics  427 

The  treatment  was  decidedly  inefficaceous  and  highly  dangerous. 
The  animals  never  gained  as  rapidly  as  the  other  hogs  kept  under  the 
same  conditions  but  not  dosed  with  turpentine.  While  separate  dosing 
might  have  shown  some  efficacy  for  the  treatment,  the  injury  to  the 
hogs  precludes  its  recommendation  as  an  anthelmintic. 

FICUS  LAURIFOLIA 

For  worms  in  dogs. — The  latex  of  Ficus  laurifolia  has  been  highly 
recommended  by  Berrio  (1911)  and  by  Mouat-Biggs  (1915)  for  use  in 
expelling  whipworms  from  man,  and  is  also  said  to  have  been  adopted  by 
the  Venezuela  State  Board  of  Health  as  the  official  remedy  for  use  against 
hookworms.  Berrio  gives  doses  of  25  to  30  gm.,  followed  by  castor  oil. 
Mouat-Biggs  gives  the  latex  morning,  noon,  and  night,  mixed  with  water, 
which  may  be  sweetened,  or  with  milk,  in  doses  of  10  to  40  gm.,  accord- 
ing to  the  age  of  the  patient. 

A  sample  of  the  expressed  juice  of  the  latex  kindly  supplied  to  the 
writers  by  Dr.  Gonzalez-Rincones  of  Caracas,  Venezuela,  was  tested  on 
three  dogs  weighing  from  4  to  29  kilos.  Fifteen  to  30  mils  of  the 
juice  were  given,  preceded  by  a  purge  of  30  mils  of  castor  oil,  and 
followed  three  or  four  hours  after  the  administration  of  the  latex  by 
15  to  30  mils  of  castor  oil.  Although  all  the  dogs  were  infested  with 
hookworms,  whipworms,  or  ascarids,  as  shown  by  a  previous  fecal 
examination,  no  parasites  were  passed.  Since  the  remedy  was  entirely 
inefficacious,  the  dogs  were  not  killed. 

The  sample  forwarded  to  the  writers  did  not  conform  to  the  published 
accounts  of  the  latex,  which  describe  it  as  of  thick,  sirupy  consistency 
and  milky  white  in  color,  but  was  instead  a  watery  fluid  with  something 
resembling  curds  floating  in  it. 

A  second  sample  conforming  in  all  respects  to  the  descriptions  of  the 
latex  was  subsequently  received  through  the  courtesy  of  Dr.  Rincones. 

This  drug  was  tested  by  the  junior  writer  on  three  dogs  weighing  4.5 
to  18.2  kilos.  The  latex  was  given  at  the  rate  of  2.6  mils  per  kilo  of 
weight  and  followed  by  castor  oil  at  the  rate  of  3  mils  per  kilo  of  weight. 
The  treatment  was  preceded  by  a  dose  of  castor  oil  the  day  before,  and 
the  animals  were  starved  for  24  hours.  The  drug  removed  9  out  of  a  total 
of  11  ascarids  (Belascaris  marginata  and  Toxascaris  limhata),  i  out  of 
33  hookworms  {Ancylostoma  caninum),  and  i  out  of  50  whipworms 
(Trichuris  depressiuseula) ,  thus  showing  but  little  anthelmintic  value  for 
the  two  worms  (hookworms  and  whipworms)  for  which  it  is  particularly 
recommended,  and  a  very  satisfactory  anthelmintic  efficacy  for  ascarids. 
It  was  ineffective  against  Dipylidium  caninum,  no  specimens  being  passed 
and  several  being  found  at  autopsy.  Of  the  three  nematodes  on  which 
the  latex  had  some  effect,  ascarids  are,  according  to  the  experience  of  the 
writers,  most  easily  removed — probably  because  they  do  not  attach 


428  Journal  of  Agricultural  Research  voI.xii.no.? 

themselves  to  the  mucosa  like  hookworms  and  whipworms.  The  writers 
have  usually  found  that  any  drug  which  is  at  all  efficacious  against  hook- 
worms is  equally  so  against  ascarids. 

Evidently  the  latex  has  defitiite  anthelmintic  properties,  and  although 
its  efficacy  against  hookworms  {Ancylostoma  caninum)  and  whipworms 
(Trichuris  depressiuscula)  was  very  slight  in  this  test,  it  is  worthy  of 
further  experimentation  in  repeated  dosage.  As  the  liquid  expressed 
from  the  latex  appears  to  have  no  value  as  a  vermifuge,  the  active  prin- 
ciple may  be  considered  to  reside  in  the  solid  portion  sustained  in  the 
latex  in  the  form  of  an  emulsion. 

SPIGEUA 

For  worms  in  dogs. — Spigelia,  or  pinkroot,  is  the  dried  rhizome  and 
roots  of  Spigelia  marilandica,  a  plant  native  to  the  Southern  and  South- 
western States.  Its  virtues  as  an  anthelmintic  are  said  to  have  been 
known  to  the  Cherokee  Indians.  According  to  the  U.  S.  Dispensatory, 
19th  edition,*  it  is  generally  considered  among  the  most  powerful 
anthelmintics,  and  is  used  especially  for  roundworms.  The  fluid  extract 
is  official. 

A  test  of  the  fluid  extract  of  Spigelia  was  made  by  the  junior  writer. 
In  this  experiment  four  dogs  weighing  from  4.8  to  18  kgm.,  all  of  which 
were  infested  with  hookworms,  and  one  with  whipworms,  were  treated 
with  Spigelia  in  doses  graduated  from  1.5  to  6  mils,  according  to  the 
weight  of  the  dogs.  The  drug  was  followed  by  194  to  324  mgm.  of  calo- 
mel. The  dogs  were  previously  dosed  with  29.57  mils  of  castor  oil  and 
then  starved  for  24  hours. 

One  dog  was  not  killed,  as  it  passed  no  worms.  As  the  previous  fecal 
examination  for  this  dog  showed  hookworms,  it  is  evident  that  the  drug 
was  ineffective  in  this  case  for  hookworms.  The  three  dogs  remaining 
passed  altogether  6  hookworms,  including  3  found  in  the  rectum  post- 
mortem in  the  process  of  elimination.  There  were  199  hookworms  left 
unaffected  by  the  treatment.  Spigelia  removed  i  out  of  18  ascarids  and 
2  out  of  29  whipworms.  A  few  segments  of  Dipylidium  caninum  were 
passed,  but  no  heads. 

The  remedy  therefore  appears  to  have  but  little  efficacy  as  an  anthel- 
mintic for  any  of  the  common  parasites  of  dogs,  but  further  trial  is  advisa- 
ble to  determine  this  point. 

tobacco 

For  worms  in  chickens. — Herms  and  Beach  {1916)  have  devised  a 
method  of  treating  poultry  for  worms,  consisting  in  the  administration 
of  chopped  tobacco  stems  and  the  liquid  in  which  they  are  steeped. 

Finely  chopped  tobacco  stems,  453  gm.  or  i  pound,  enough  for  100 
birds,  are  soaked  for  two  hours  in  enough  water  to  keep  them  covered. 

1  Wood,  G.  B.,  and  Bache,  Franklin.    Op.  cit.,  p.  1162. 


Feb.  i8. 1918  Efficacy  of  Some  Anthelmintics  429 

Both  the  stems  and  water  are  mixed  with  half  the  usual  ration  of  ground 
feed  and  given  to  the  fowls.  Two  hours  later  one-fourth  the  usual  ration 
is  given  mixed  with  Epsom  salt  at  the  rate  of  312  gm.  for  each  100  fowls. 
The  treatment  is  to  be  repeated  one  week  later.  The  cost  is  said  to  be 
only  10  cents  for  100  fowls. 

In  the  present  experiment  six  chickens  were  deprived  of  food  for  24 
hours  and  then  fed  the  tobacco  mash;  two  hours  later  they  were  fed 
the  Epsom  salt  mash  in  the  proportions  recommended  by  Herms  and 
Beach.  The  next  day  they  were  given  what  was  left  of  the  tobacco 
mash,  since  they  had  refused  to  clean  it  up  the  first  time.  A  mixture  of 
bran  and  tobacco  was  used,  and  the  birds  were  not  very  eager  for  it, 
even  after  the  preliminary  fasting. 

The  remedy  removed  30  out  of  162  Heterakis  papulosa,  3  out  of  39 
tapeworms,  and  had  no  effect  on  Tetrameres  sp.,  a  parasite  which,  from 
its  location  in  the  mucous  glands  of  the  proventriculus,  would  not  be 
likely  to  be  affected.  This  treatment  is  apparently  intended  especially 
for  Ascaridia  perspicillwm,  since  the  "roundworm"  figured  in  the  paper 
by  Herms  and  Beach  {1916)  is  evidently  of  this  species.  Unfortu- 
nately the  efficacy  of  the  treatment  for  this  parasite  could  not  be  deter- 
mined, since  no  species  of  Ascaridia  were  present  in  the  birds  used  in 
the  experiment. 

While  the  tobacco  treatment  failed  to  free  any  bird  from  all  of  the 
worms  of  any  given  species  that  might  be  present,  nevertheless  the 
treatment  seems  to  have  been  more  successful  against  Heterakis  papulosa 
or  against  tapeworms  than  the  other  treatments  tried. 

In  view  of  the  difficulty  of  removing  H.  papulosa  from  the  ceca,  it 
may  be  assumed  on  the  showing  here  that  this  treatment  would  be 
effective  in  removing  Ascaridia  perspicUhim  from  the  small  intestine, 
especially  if  the  dose  is  repeated,  as  recommended  by  Herms  and  Beach. 

OIL   OF  CHENOPODIUM 

Oil  of  chenopodium  is  derived  from  the  distillation  of  the  seeds  or  of  the 
entire  leafy  part  of  Chenopodium  anthelminticiim  L-.,  sometimes  referred 
to  under  the  synonym  "  CJienopodium  a^nbrosioides  anihelminticum 
A.  Gray,"  and  commonly  called  "chenopodium,"  "American  wonn- 
seed,"   or    "Jerusalem  oak."     According  to  Henkel   (1913) — 

Wormseed  has  been  naturalized  in  this  coiintry  from  tropical  America  and  occurs 
in  waste  places  from  New  England  to  Florida  and  westward  to  California. 

Infusions  made  from  chenopodium  were  used  in  the  United  States 
by  the  early  settlers  as  a  treatment  for  infestation  with  ascarids  in  man, 
and  its  anthelmintic  properties  are  said  to  have  been  known  by  the 
Indians.  Oil  of  chenopodium  has  only  recently  come  into  prominence 
as  a  result  of  the  shortage  of  thymol  and  santonin,  for  which  it  has 
proved  an  effective  substitute. 
27810°— 18 3 


430  Journal  of  Agricultural  Research  voi.  xii.  No.  7 

The  chemistry  of  the  oil  of  chenopodium  has  been  studied  by  Nelson 
(jpii,  19 1 3),  who  was  unable  to  ascertain  its  exact  chemical  nature, 
but  concludes  that  it  is  an  unstable  dioxid.  Its  physiological  action 
and  toxicity  have  been  the  object  of  numerous  experiments  by  Salant 
et  al.  (1915)  and  Nelson  (1911,  1913).  They  note  that  it  is  a  respira- 
tory depressant  and  that  it  decreases  vagus  instability  and  diminishes 
frequency  of  heart  action. 

In  regard  to  its  toxicity  the  minimum  lethal  dose  is  twice  as  great 
when  given  by  the  mouth  as  when  given  hypodermatically.  Repeated 
doses  have  a  cumulative  effect,  and  the  toxicity  is  enhanced  when  the 
experiment  animals  are  starved.  It  is  less  toxic  when  combined  with 
nonessential  oils,  such  as  olive,  cottonseed,  or  coconut  oil.  In  cats  0.6 
mil  per  kilo  by  mouth  was  invariably  fatal,  but  in  dogs  0.5  mil  caused 
only  vomiting.  Adrenalin  and  digitalis  were  found  to  be  antagonistic 
to  oil  of  chenopodium. 

Heiser  {19 15)  notes  that  over  100,000  cases  of  hookworm  infestation 
involving  both  Ancylostoma  duodetmle  and  Necator  americanus  have  been 
treated  in  the  Orient  with  chenopodium. 

For  worms  in  dogs. — In  view  of  the  present  great  interest  in  cheno- 
podium, the  writers  have  made  a  considerable  number  of  tests  of  this 
substance  with  a  view  to  determining  its  efficacy  against  worms  in  dogs 
and  other  domestic  animals,  and  in  nearly  all  cases  they  have  found  it 
extremely  efficacious,  especially  for  ascarids.  The  following  experiment 
was  conducted  to  determine  the  efficacy  of  oil  of  chenopodium  when 
administered  in  one  dose  at  the  rate  of  0.3  mil  per  kilo. 

Eight  dogs,  weighing  from  2  to  10.3  kilos,  were  given  a  preliminary 
purge  of  castor  oil,  which  resulted  in  the  removal  of  9  ascarids.  After 
starving  for  24  hours,  they  were  given  chenopodium  in  the  dosage  indi- 
cated, mixed  with  2.5  to  15.5  mils  of  castor  oil,  according  to  weight. 
The  treatment  was  found  extremely  efficacious  for  ascarids,  removing 
160  worms  and  leaving  only  2.  From  all  but  one  of  the  dogs  it  removed 
all  ascarids,  and  removed  10  out  of  12  from  that  one.  Its  effect  on  hook- 
worms and  whipworms  was  less  striking,  about  one-fourth  of  the  total 
number  present  being  removed.  It  had  very  little  efficacy  against 
Dipylidium  caninum.  It  was  evident  that  the  dose  given  was  uimeces- 
sarily  large,  since  the  feces  smelled  strongly  of  chenopodium,  and  six  of 
the  eight  dogs  vomited  from  two  to  four  hours  after  treatment.  An 
objectionable  feature  of  the  chenopodium  treatment,  as  given,  was  the 
excessive  ptyalism  that  it  caused  and  the  fact  that  the  dogs  objected  to 
the  taste  of  chenopodium  in  the  castor  oil  and  resisted  its  administration. 

In  the  second  experiment  with  chenopodium  an  attempt  was  made  to 
overcome  the  ptyalism  by  giving  the  chenopodium  in  a  capsule.  Eight 
dogs,  weighing  from  5.5  to  9  kilos,  were  given  chenopodium  in  capsules 
at  the  rate  of  0.2  mil  per  kilo  on  each  of  three  successive  days,  the  drug 
being  followed  with  five  times  its  volume  of  olive  oil  the  first  two  doses, 


Feb.  18, 1918  Efficacy  of  Some  Anthelmintics  431 

and  five  times  its  volume  of  castor  oil  the  third  dose.  While  this  treat- 
ment removed  15  out  of  17  ascarids  (Belascaris  marginata),  it  was 
ineffective  for  hookworms  (Ancylostoma  caninum),  whipworms  {Tri- 
churis  depressiuscula) ,  and  Dipylidium  caninum.  This  method  was  less 
satisfactory  than  that  employed  in  the  previous  experiment.  The  castor 
oil  given  was  insufficient  to  overcome  the  constipating  effect  of  the 
chenopodium;  the  chenopodium,  being  promptly  released  from  the  cap- 
sules, was  brought  undiluted  against  the  gastric  mucosa,  causing  con- 
siderable irritation,  as  shown  by  the  fact  that  all  dogs  vomited  promptly 
after  the  administration  of  the  drug.  Two  of  the  dogs  died  from  the 
treatment,  one  from  traumatic  pneumonia,  due  to  the  fact  that  the  cap- 
sule opened  in  the  lar3mx,  allowing  oil  to  penetrate  the  lungs.  The  other 
dog  which  died  was  a  bitch,  containing  11  well-developed  fetuses,  and 
it  is  likely  that  the  pregnancy  in  this  case  was  a  condition  contributing 
to  the  animal's  inability  to  withstand  the  large  amount  of  chenopodium 
given. 

A  third  experiment  was  undertaken  to  test  the  efficacy  of  oil  of  cheno- 
poditmi  when  given  in  doses  of  o.i  mil  per  kilo,  and  the  dose  repeated 
daily  for  a  total  of  six  doses.  The  chenopodium  was  given  mixed  with 
10  times  its  volume  of  olive  oil,  and  was  preceded  each  day  by  castor  oil 
in  amount  equal  to  the  olive  oil ;  i  minim  of  chloroform  per  kilo  of  weight 
was  added  to  the  mixture  of  chenopodium  and  olive  oil  the  first  day  and 
to  the  castor  oil  on  the  following  days.  The  animals  were  kept  on  half 
feed  during  the  time  the  treatment  was  being  given. 

In  this  experiment  all  of  the  8  ascarids  present  and  94  out  of  133 
hookworms  were  removed.  This  increased  efficacy  against  hookworms 
is  in  the  writers'  opinion  due  largely  to  the  chloroform  administered 
with  the  chenopodium,  since  in  none  of  their  numerous  experiments  with 
chenopodium  have  they  ever  found  it  so  efficacious  for  hookworms  when 
given  with  castor  oil  only.  The  treatment  was  entirely  ineffective  for 
whipworms  (Trichuris  depressiuscula)  or  for  Dipylidium  caninum. 

This  method  of  administration  proved  generally  satisfactory.  The 
individual  dose  was  sufficiently  diluted  to  prevent  any  undue  ptyalism 
when  administered,  and  there  were  no  symptoms  of  acute  distress,  which 
occurred  when  the  oil  was  given  in  a  capsule. 

The  results  of  this  experiment  led  the  writers  to  make  another  test  of 
chenopodium  and  chloroform  as  a  remedy  for  mixed  infestation,  cheno- 
podium being  in  their  experience  the  most  successful  remedy  for  ascarids 
and  chloroform  the  best  remedy  for  hookworms.  This  combination 
seems  to  be  entirely  compatible  both  physiologically  and  pharmaceu- 
tically.  These  drugs  when  combined  with  castor  oil  form  a  homogeneous 
and  fairly  stable  mixture,  and  the  action  of  each  drug  is  not  inhibited  by 
the  other. 

Four  dogs  having  been  given  a  preliminary  purgative  of  castor  oil, 
followed  by  a  24-hour  fast,  were  given  chenopodium  and  chloroform,  both 


432  Journal  of  Agricultural  Research  voi.  xii.  no.  7 

at  the  rate  of  o.i  mil  per  kilo.  The  drugs  were  given  in  castor  oil,  vary- 
ing in  amount  from  12  to  40  mils,  according  to  the  weight  of  the  animals. 
In  this  case,  as  usual,  the  chenopodium  proved  efficacious  against  ascarids 
(Belascaris  marginata) ,  removing  7  out  of  8.  The  remedy  showed  an 
appreciable  effect  on  the  whipworms  (Trichuris  depressiuscula) ,  removing 
about  one-fifth  of  the  total  number.  In  regard  to  hookworms,  however, 
only  7  out  of  61  were  removed. 

In  order  to  ascertain  whether  the  efficacy  of  the  remedy  for  hookworms 
would  be  enhanced  by  increasing  the  dose,  a  second  trial  was  made  in 
which  0.2  mil  per  kilo,  or  double  the  amount  of  both  chenopodium  and 
chloroform,  were  administered.  As  in  the  previous  case,  the  treatment 
was  preceded  by  the  administration  of  castor  oil  24  hours  before  dosing, 
,and  the  dogs  allowed  no  food  until  after  the  administration  of  the  anthel- 
mintics. The  chenopodium  and  chloroform  were  given  in  castor  oil, 
allowing  29.57  ^^^  for  each  dog.  The  preliminary  purgative  removed 
one  ascarid  (B.  marginata)  and  some  segments  of  Dipylidium  caninum. 

The  combination  of  chenopodium  and  chloroform  in  the  dosage  given 
proved  highly  successful.  All  ascarids  and  over  half  the  hookworms 
present  were  removed.  As  usual,  it  was  entirely  ineffective  for  Dipy- 
lidium caninum,.  Undoubtedly  the  dosage  of  chenopodium  and  chloro- 
form at  the  rate  of  o.  i  mil  per  kilo  is  too  small  for  the  best  results,  and 
the  efficacy  of  both  drugs  is  enhanced  by  doubling  the  amount.  At  the 
same  time  this  increased  dose  seems  to  be  well  within  the  limits  of  safety 
and  has  been  tested  several  times  by  the  junior  writer,  always  with 
satisfactory  results. 

In  view  of  the  fact  that  castor  oil  is  more  or  less  objectionable  to  many 
persons  and  is  not  well  tolerated  by  others,  the  writers  undertook  an 
experiment  in  which  liquid  petrolatum  was  substituted  for  castor  oil  in 
connection  with  oil  of  chenopodium.  Four  dogs  weighing  4.5  to  15  kilos 
were  given  chenopodium  in  doses  of  0.2  mil  per  kilo  mixed  with  10  mils 
of  liquid  petrolatum,  the  drug  being  followed  at  once  by  20  mils  of 
liquid  petrolatum.  The  animals  were  fasted  from  noon  of  the  previous 
day.  One  of  the  dogs  died  on  the  third  day  following  the  administration 
of  chenopodium.  The  treatment  was  entirely  inefficacious  for  ascarids, 
hookworms,  Dipylidium  sp.,  or  Taenia  sp.,  and  removed  only  2  of  28 
whipworms.  While  the  experiment  indicates  that  the  liquid  petrolatum 
diminishes  the  efficacy  of  the  chenopodium,  it  must  be  noted  that  there 
was  only  one  ascarid  present,  and  while  the  treatment  should  have  removed 
the  worm,  this  is  a  rather  small  basis  on  which  to  judge  the  performance 
of  the  anthelmintic.  On  the  other  hand,  the  liquid  petrolatum  seems  to 
be  unequal  to  the  task  of  overcoming  the  constipation  and  toxic  effects 
from  the  chenopodium,  as  the  treatment  resulted  in  the  death  of  one  dog. 

It  is  the  opinion  of  the  writers  that  the  mechanical  lubrication  resulting 
from  the  use  of  liquid  petrolatum  is  not  sufficient  to  overcome  the  con- 
stipation and  that  an  active  purgative,  preferably  castor  oil,  is  indicated. 


Feb.  i8. 1918  Efficacy  of  Some  Anthelmintics  433 

The  foregoing  experiments  with  chenopodium  and  chloroform,  in 
comparison  with  other  standard  anthelmintics  used  for  the  removal  of 
worms  from  dogs,  indicate  that  oil  of  chenopodium  is  the  best  anthel- 
mintic of  those  tried  for  use  against  ascarids  {Belascaris  marginata  and 
Toxascaris  limhata).  In  the  6  experiments,  involving  the  use  of  34 
experiment  animals,  it  removed  194  out  of  200  ascarids,  an  efficacy  of 
97  per  cent.  It  is  at  the  same  time  probably  a  little  more  efficacious 
against  whipworms  (Trichuris  depressiuscula)  than  any  other  anthel- 
mintic tested.  In  the  entire  six  experiments  its  average  percentage  of 
efficacy  was  only  12,  a  figure  which  illustrates  the  difficulty  experienced 
in  dislodging  this  parasite.  Moreover,  it  is  probable  that  the  chloroform 
used  in  three  of  the  experiments  may  be  responsible  for  the  removal 
of  some  of  these  worms,  since  both  chloroform  and  chenopodium  were 
found  to  have  a  limited  efficacy  for  whipworms  when  given  separately. 

As  regards  hookworms  (Ancylostoma  caninum),  the  experiments  of  the 
writers  with  chenopodium  do  not  show  as  high  an  average  efficacy 
for  this  drug  as  they  had  been  led  to  expect,  considering  the  warm 
indorsement  chenopodium  has  frequently  received  as  an  anthelmintic 
for  the  hookworm  of  man.  Undoubtedly  chenopodium  as  well  as  thymol 
would  have  to  be  given  in  repeated  doses  and  the  treatment  renewed 
in  order  to  secure  a  high  percentage  of  efficacy.  The  writers  have  found 
chloroform  much  more  efficacious  than  chenopodium  for  hookworms  in 
dogs,  and,  when  given  at  the  rate  of  0.2  mil  per  kilo,  is  apparently  free 
from  danger. 

For  worms  in  hogs. — A  series  of  experiments  to  test  oil  of  chenopo- 
dium as  an  anthelmintic  for  ascarids  in  swine  were  carried  out  by  the 
junior  writer  and  seemed  to  indicate  the  great  value  of  this  drug  when 
given  to  hogs  individually  after  a  period  of  starvation.  In  these  experi- 
ments the  hogs  were  not  killed,  the  effect  of  the  anthelmintic  being 
judged  by  the  presence  or  absence  of  ascarid  eggs  in  the  feces  passed 
subsequent  to  treatment. 

In  the  first  experiment  with  chenopodium  an  effort  was  made  to 
determine  the  minimum  effective  dose  necessary  to  free  a  hog  from 
ascarids.  In  this  case  the  dose  of  oil  was  combined  with  3.7  mils  of 
areca  nut  and  mixed  with  the  feed,  after  starving  the  pigs.  Each  pig 
was  fed  separately,  thus  assuring  full  dosage  to  each  animal.  Two  pigs, 
weighing  11.32  and  14.06  kgm.,  respectively,  were  given  0.2  mil  of 
chenopodium,  while  a  third  pig  weighing  37.9  kgm.  received  0.4  mil. 
This  latter  pig  was  apparently  free  from  ascarids  after  a  single  dose, 
while  the  two  other  pigs  each  received  three  additional  doses  of  0.27  mil, 
0.36  mil,  and  0.54  mil  before  their  feces  were  negative  for  ascarid  eggs. 

It  is  evident  from  the  above  that,  even  when  given  mixed  with  feed, 
a  method  later  shown  to  be  unsuitable,  a  relatively  small  dose  of  oil  of 
chenopodium  may  prove  efficacious.  It  should  be  stated  in  comment 
that  at  the  time  these  experiments  were  conducted  there  was  little 


434  Journal  of  Agricultural  Research  voi.  xri.  No.  7 

modern  literature  in  regard  to  the  use  of  oil  of  chenopodium,  and,  so  far 
as  the  writer  is  aware,  no  use  had  been  made  of  the  oil  as  an  anthelmintic 
for  swine.  Hence,  the  drug  was  administered  cautiously.  Subsequent 
experiments  have  shown  that  doses  of  chenopodium  relatively  many 
times  larger  are  perfectly  well  tolerated  by  swine  and  are  more  certain 
in  action. 

In  a  second  experiment  with  swine  chenopodium  was  administered 
in  castor  oil,  and  was  not  given  mixed  with  feed  but  was  fed  with  a  spoon 
to  each  hog  individually.  Four  of  the  five  hogs  used  were  each  given 
28.35  gn^-  of  Epsom  salt  mixed  with  their  feed,  while  the  fifth  hog  was 
given  none.  All  animals  were  then  starved  for  24  hours  before  the 
administration  of  the  chenopodium.  The  smallest  pig,  weighing  12.25 
kgm.  was  given  0.8  mil;  the  others,  weighing  12.7  to  28.12  kgm.  were 
given  I  mil.  The  dose  of  chenopodium  in  each  case  was  given  with 
29.57  niils  of  castor  oil.  This  treatment  was  entirely  successful.  All 
five  hogs  had  shown  numbers  of  ascarid  eggs  in  their  feces  prior  to  the 
treatment,  and  all  were  found  free  from  ascarid  eggs  a  week  later.  The 
feces  of  the  hogs  were  again  examined  three  days  later  and  found  negative. 
Two  of  the  five  hogs  received  a  third  fecal  examination,  which  was 
consistently  negative. 

It  appears  from  the  above  that  chenopodium  is  an  excellent  anthel- 
mintic when  given  in  individual  doses  after  preliminary  starvation,  and 
in  most  cases  may  be  relied  on  to  remove  all  the  ascarid  worms  present. 
This  opinion  was  confirmed  by  further  experiments.  The  preliminary 
dosing  with  Epsom  salt  did  not  seem  to  be  of  any  benefit,  since  the  cheno- 
podium was  as  successful  in  the  case  of  the  hog  which  received  no  salt 
as  in  the  case  of  those  that  did.  Subsequent  experience  has  shown  that 
in  the  chenopodium  treatment  for  hogs  preUminary  purgation  is  an  un- 
necessary expense. 

As  there  is  always  a  demand  among  stockmen  for  something  that  can 
be  mixed  with  the  feed  to  rid  their  hogs  of  worms,  an  experiment  was 
undertaken  to  determine  the  possibilities  of  this  form  of  medication. 
As  hogs,  especially  if  large,  are  very  difficult  to  handle  and  as  much  time 
and  eff'ort  would  be  required  to  dose  separately  each  animal  of  a  large 
herd,  the  advantage  of  mixing  the  medicine  in  the  feed  is  obvious.  Unfor- 
tunately this  method  proved  impracticable,  as  shown  by  the  following 
experiment. 

One  mil  of  chenopodium,  29.57  mils  of  castor  oil,  and  237  mils  of  lin- 
seed decoction  made  by  boiUng  113  gm.  of  linseed  in  296  mils  of  water, 
were  thoroughly  shaken  together  into  a  homogeneous  emulsion.  This 
amount  was  allowed  for  each  of  foyr  hogs  weighing  20.4  to  32.2  kgm.  The 
emulsion  was  distributed  evenly  through  a  bran  mash  which  all  four 
hogs  ate  together.  The  animals  had  previously  been  fed  a  mash  con- 
taining Epsom  salt  at  the  rate  of  28.35  gm.  for  each  animal,  then  starved 
for  24  hours  before  treatment.     The  fecal  examination  a  few  days  after 


Feb.  i8. 1918  Efficacy  of  Some  Anthelmintics  435 

the  treatment  showed  that  one  hog  was  freed  from  ascarids,  the  three 
others  still  remaining  infested. 

It  is  evident  that  the  administration  of  drugs  by  mixing  them  with 
the  feed  is  a  method  which  gives  little  promise  of  success,  and  the  far 
greater  efficacy  achieved  when  the  drug  is  given  individually  after  a 
fast  more  than  compensates  for  the  extra  labor  involved. 

In  order  to  be  certain  that  the  comparative  failure  in  this  treatment 
resulted  from  the  method  of  administration  and  not  from  an  insufficient 
dose  of  chenopodium,  the  experiment  was  repeated,  allowing  2.5  mils 
of  oil  to  each  of  the  three  hogs  remaining  infested.  This  experiment 
resulted  in  a  total  failure,  all  the  hogs  remaining  infested  after  the 
experiment. 

The  above  experiments  were  concerned  solely  with  the  action  of  cheno- 
podium against  ascarids,  no  attention  being  paid  to  its  action  on  other 
intestinal  parasites.  In  order  to  test  its  action  on  nodular  worms 
(Oesophagostomum  dentatum)  three  doses  of  oil  of  chenopodium  were  given 
to  one  hog  at  intervals  of  10  days,  two  fecal  examinations  being  made 
after  each  dose.  The  chenopodium  was  given  with  3.7  mils  of  areca 
nut  after  preliminary  starvation,  and,  so  far  as  could  be  judged  from 
the  fecal  examinations,  there  was  no  diminution  in  the  number  of  nodular 
worms.  On  account  of  their  location  in  the  cecum  and  colon,  where 
they  are  protected  by  a  large  mass  of  fecal  material,  it  is  not  surprising 
that  a  vermifuge  given  by  the  mouth  has  little  or  no  effect  upon  nodular 
worms.  The  extensive  colon  of  the  pig  is  rarely,  if  ever,  approximately 
empty,  even  after  prolonged  starvation.  Probably  the  best  way  to 
reach  worms  in  this  location  is  by  the  use  of  diluted  anthelmintics  in  the 
form  of  enemata,  as  suggested  by  Railliet  {1915). 

As  Chenopodium  anthelminticum  grows  abundantly  in  the  vicinity  of 
Washington,  D.  C,  a  test  was  made  to  determine  the  possibilities  of 
feeding  the  entire  plants  to  the  hogs.  The  test  was  made  when  the  plants 
were  in  full  seed  and  therefore  contained  the  maximum  amount  of  oil. 
A  large  armful  of  chenopodium  plants  was  placed  daily  in  a  hogpen  con- 
taining three  hogs.  The  amount  given  represented  about  all  the  hogs 
would  clean  up  in  a  day,  and  the  experiment  continued  for  19  days. 
Although  the  hogs  received  their  usual  daily  ration,  they  ate  the  chenopo- 
dium plants  readily.  At  the  end  of  19  days  one  hog  was  apparently 
free  from  ascarids  (A  scans  suum). 

While  this  treatment  has  the  obvious  disadvantage  of  inaccuracy  in 
dosage  and  is  applicable  only  in  those  regions  where  the  chenopodium 
plant  grows  abundantly,  it  has  the  great  advantage  of  costing  nothing 
but  the  labor  of  gathering  the  weeds.  At  least  it  would  seem  worthy  of 
further  trial. 

A  test  of  the  dried  seeds  of  chenopodium  was  made  in  the  following 
experiment:  Each  of  three  infested  hogs  was  given  daily  3  gm.  of 
chenopodium  seed  and  3  gm.  of  areca  nut  mixed  with  the  feed.     The 


436  Journal  of  Agricultural  Research  voi.  xn,  no.  ? 

treatment  was  continued  for  1 1  days,  at  which  time  each  hog  had  received 
33  gm.  of  chenopodium  seed,  or  the  equivalent  of  about  i  mil  of  the  oil. 
One  of  the  three  hogs  was  apparently  freed  from  ascarids. 

The  treatment,  besides  being  of  uncertain  efficacy,  cost  more  than  a 
single  treatment  with  chenopodium  oil,  since  a  large  amount  of  both 
chenopodium  seed  and  areca  nut  must  be  consumed.  Its  efficacy  would 
doubtless  be  increased  if  given  after  a  period  of  starvation,  but  this  is 
obviously  impossible  when  the  treatment  must  be  continued  for  several 
days. 

It  appears  from  the  above  experiments  that  oil  of  chenopodium  is  an 
excellent  anthelmintic  for  ascarids  in  hogs  when  given  in  suitable  dosage 
after  a  preliminary  period  of  starvation.  The  best  purgative  with  which 
to  combine  it  is  castor  oil.  This  has  the  advantage  of  relative  cheapness, 
certainty  of  its  action,  and  of  being  easily  miscible  with  oil  of  chenopo- 
dium. The  chenopodium  may  be  given  first  and  followed  immediately 
by  castor  oil,  or  the  two  may  be  given  together.  The  efficacy  of  oil  of 
chenopodium  is  greatly  decreased  or  entirely  lost  if  the  drug  is  mixed 
with  the  feed  and  several  animals  allowed  to  dose  themselves  while 
eating  together.  Chenopodium  seed  and  areca  nut  have  but  little 
efficacy  when  given  daily,  mixed  Avith  the  feed,  and  the  expense  of  this 
treatment,  on  account  of  the  number  of  doses  required,  is  actually 
greater  than  a  single  dose  of  the  relatively  high-priced  oil  of  chenopodium. 
Chenopodium  plants  seem  to  have  some  value  as  an  anthelmintic  and, 
when  these  are  available,  the  cost  of  anthelmintic  treatment  is  greatly 
reduced.     However,  further  experimentation  is  desirable  along  this  line. 

The  following  experiments  with  chenopodium,  in  which  the  hogs  were 
subsequently  killed  and  all  worms  counted,  confirmed  the  writers' 
opinion  regarding  the  efficacy  of  oil  of  chenopodium  when  given  under 
suitable  conditions : 

Two  pigs  weighing  8.16  and  14.06  kgm.,  respectively,  were  given  i  mil 
of  oil  of  chenopodium,  while  a  third,  weighing  9.98  kgm.,  was  given  2 
mils.  The  drug  was  followed  by  59  mils  of  castor  oil  in  the  case  of  one 
pig  and  given  in  59  mils  of  castor  oil  in  the  case  of  the  second  pig.  Eleven 
ascarids  and  six  nodular  worms  were  passed  by  the  pigs.  Post-mortem 
examination  showed  that  all  animals  were  freed  from  ascarids  {Ascaris 
suum),  while  numerous  nodular  worms  {Oesophagostomum  dentatum) 
were  found  in  each  animal.  The  experiment  adds  further  proof  of  the 
efficacy  of  oil  of  chenopodium  as  an  ascaricide,  and  demonstrates  that  it 
also  has  some  efficacy  for  nodular  worms,  although  probably  only  a  few 
were  removed.  As  already  stated,  these  worms  are  so  protected  by  the 
large  amount  of  fecal  material  in  which  they  lie  that  it  is  unlikely  that 
any  anthelmintic  given  by  the  mouth  will  have  much  effect  upon  them. 

As  already  noted  in  the  previous  experiments,  oil  of  chenopodium, 
when  mixed  with  the  feed  of  a  number  of  hogs  feeding  together,  had 
shown  little  or  no  efficacy  as  an  anthelmintic.     Since,  however,  in  these 


Feb.  i8, 1918  Efficacy  of  Some  Anthelmintics  437 

experiments  the  pigs  were  not  killed  and  the  results  were  based  solely 
on  the  fecal  examinations,  it  was  decided  to  repeat  the  experiment, 
killing  the  pigs  and  counting  the  worms  and  thus  determining  definitely 
whether  this  method  could  be  recommended. 

A  sow  which  had  been  starved  for  24  hours  was  fed  a  mash  of  corn 
meal  and  bran  into  which  had  been  stirred  an  emulsion  consisting  of 
3.5  mils  of  oil  of  chenopodium,  60  mils  of  castor  oil,  and  473  mils  (i  pint) 
of  milk.  Following  the  treatment  the  sow  passed  19  stomach  worms 
{Arduenna  strongylina)  and  13  nodular  worms  (Oesophagostomwm  den- 
tatum).  A  post-mortem  examination  revealed  the  presence  of  19  ascarids 
(Ascaris  suum),  13  stomach  worms  (A.  strongylina),  and  numerous 
nodular  worms  (O.  dentatum).  The  treatment  proved  a  complete  failure 
in  eliminating  ascarids,  thus  confirming  the  opinion  already  expressed 
regarding  the  inefficacy  of  chenopodium  when  given  mixed  with  the  feed. 
The  treatment  removed  over  half  the  stomach  worms  present  and 
demonstrates  the  possibilities  of  the  drug  for  this  species  of  parasite. 
Although  it  also  had  some  efficacy  against  nodular  worms,  as  usual  the 
percentage  removed  was  very  small. 

In  cooperation  with  Dr.  Ernest,  of  the  Tuberculosis  Eradication 
Division,  of  the  Bureau  of  Animal  Industry,  the  junior  writer  was  recently 
enabled  to  make  a  practical  test  of  the  chenopodium  treatment  of  swine 
on  a  large  scale.  The  test  was  carried  out  at  the  Green  Berry  Point  Farm 
of  the  United  States  Naval  Academy,  where  about  176  swine  ranging 
from  young  pigs  weighing  15.88  kgm.  or  less  to  large  boars  and  brood 
sows,  were  given  the  treatment. 

The  pigs  were  starved  for  24  hours  before  treatment  and  confined  in 
relatively  small  pens  so  that  they  could  be  easily  caught.  The  brood 
sows  and  boars  were  kept  separate.  The  pigs  were  caught  in  the  small 
pen  and  lifted  one  at  a  time  over  the  fence  into  a  larger  inclosure,  where 
they  were  treated.  Four  laborers  were  employed  in  this  work,  two 
capturing  the  pigs  and  two  holding  the  animals  while  they  were  dosed. 
Restraint  was  made  fairly  easy  by  backing  the  pig  into  a  comer  of  the 
inclosure.  While  one  attendant  held  the  pig  in  this  position  the  other 
kept  the  animal's  jaws  apart,  using  two  loops  made  of  harness  straps; 
one  loop,  passing  under  the  upper  jaw,  was  pulled  upward  while  the  other 
loop,  passing  over  the  under  jaw,  was  pulled  downward.  The  head  was 
kept  tilted  upward  and  the  pressure  on  the  straps,  besides  keeping  the 
mouth  wide  open,  served  to  hold  down  the  tongue  and  prevent  the  hog 
from  shaking  its  head. 

A  table  to  hold  the  bottles  of  chenopodium  and  castor  oil  and  the 
measuring  glasses  was  placed  conveniently  near.  Four  mils  of  oil  of  cheno- 
podium and  I  ounce  (29.57  mils)  of  castor  oil  was  allowed  each  hog  under 
100  pounds  (45.36  kgm.),  while  those  over  that  weight  were  given  double 
the  dose.  No  attempt  was  made  to  weigh  the  animals,  but  the  dose  for 
each  hog  was  decided  upon  as  it  was  brought  up  for  treatment.     The 


438  Journal  of  Agricultural  Research  voi.  xii,  No.  7 

two  oils  were  measured  separately,  and  were  then  poured  together  into 
a  large  iron  kitchen  spoon,  which  was  placed  as  far  back  as  possible  in 
the  animal's  mouth.  Subsequently,  however,  the  iron  spoon  was  dis- 
pensed with,  the  oils  being  poured  directly  into  the  hog's  mouth,  care 
being  taken  to  hold  the  graduate  glasses  out  of  reach  of  the  hog's  teeth. 

When  the  medicine  had  been  given  the  pressure  on  the  straps  was 
relaxed  to  enable  the  hog  to  swallow,  the  head  still  being  held  high.  If 
the  hog  refused  to  swallow,  it  could  always  be  induced  to  do  so  by 
plugging  the  nostrils  with  the  finger  tips.  This  forces  the  animal  to  breath 
through  its  mouth,  and  to  do  so  it  must  first  swallow. 

Wefound  this  method  fairly  rapid  for  hogs  weighing  not  over  100  pounds 
(45.36  kgm.),  65  animals  being  dosed  in  one  hour.  It  took  eight  hours 
to  dose  the  entire  herd  of  176  hogs,  the  labor  of  handling  the  heavy 
brood  sows  making  the  operation  much  slower  than  when  pigs  of  medium 
size  were  being  treated.  In  treating  the  brood  sows  it  was  necessary 
to  throw  them  on  their  backs  and  to  hold  them  in  this  position  while 
dosing.  As  this  caused  considerable  excitement  to  the  sows,  those 
which  from  their  appearance  were  soon  to  farrow  were  left  untreated. 

Only  one  of  the  176  hogs  treated  was  injured.  This  animal,  which 
was  accidentally  dropped,  injuring  the  spinal  cord,  was  killed  and  exam- 
ined for  worms.  One  hundred  ascarids  (Ascaris  suum)  were  taken  from 
a  piece  of  intestine  not  over  30  cm.  long,  and  many  more  remained 
uncounted.  Two  days  later  the  manager  of  the  farm  reported  that  a 
great  many  ascarids  were  seen  among  the  feces  on  the  place,  and  about 
two  weeks  later  one  of  the  pigs  which  was  killed  for  food  was  found  to 
be  entirely  free  from  ascarids.  The  treatment  therefore  appears  to  have 
been  very  successful.  The  treatment  required  1.13  kgm.  of  oil  of  che- 
nopodium  and  7.5  liters  of  castor  oil.  With  chenopodium  at  $11  per 
kilo  ($5  a  pound)  and  castor  oil  at  66  cents  per  liter,  the  treatment  in 
this  instance  cost  a  trifle  under  10  cents  per  hog,  exclusive  of  the  labor. 
Even  at  the  present  high  price  of  drugs,  the  cost  is  trifling  compared 
with  the  increased  profit  which  may  be  expected  to  be  derived  from 
healthy  animals. 

For  WORMS  in  sheep. — Four  lambs  weighing  16.6  to  26.1  kilos  were 
dosed  with  oil  of  chenopodium  at  the  rate  of  0.2  mil  per  kilo,  the  dose 
which  was  found  most  effective  for  dogs.  The  medicine  was  given  as  a 
drench  emulsified  with  147.9  mils  of  milk.  Following  the  treatment, 
one  of  the  lambs  contracted  pneumonia,  probably  as  the  result  of 
some  of  the  drench's  entering  the  lungs.  The  treatment  succeeded 
in  removing  all  the  stomach  worms  from  three  lightly  infested  lambs, 
but  failed  completely  in  the  case  of  one  heavily  infested  lamb,  suggest- 
ing that  in  this  case  the  drench  did  not  reach  the  fourth  stomach  di- 
rectly, but  was  modified  or  absorbed  in  the  rumen. 

The  treatment  was  fairly  efficacious  for  hookworms  (Bunostomum  iri- 
gonocephalum) ,  removing  two-thirds  of  those  present,  in  this  respect 


Feb.  i8. 1918  Efficacy  of  Some  Anthelmintics  439 

being  considerably  more  efficacious  than  it  was  found  to  be  against 
hookworms  in  dogs.  It  was  inefficacious  for  other  intestinal  worms. 
On  the  whole,  the  use  of  chenopodium  for  stomach  worms  and  hook- 
worms in  sheep  seems  to  promise  considerable  success  when  properly 
administered,  and  is  at  least  worthy  of  further  trial. 

For  worms  in  poultry. — ^To  test  the  efficacy  of  oil  of  chenopodium 
against  worms  in  poultry,  six  chickens  were  dosed  at  a  rate  of  about 
0.4  mil  per  kilo.  Each  bird  weighed  0.5  kgm.  Each  bird  received  2 
mils  of  castor  oil  followed  at  once  by  0.2  mil  of  oil  of  chenopodium 
mixed  with  2  mils  of  castor  oil,  the  birds  being  kept  without  feed  the 
previous  day.  The  treatment  was  fairly  satisfactory  for  Ascaridia 
perspicillum,  removing  9  out  of  13  worms.  In  this  connection  it  will 
be  recalled  that  the  experiment  in  feeding  tobacco  stems  to  chickens 
was  carried  out  with  birds  not  infested  with  A .  perspicillum,  so  that  the 
writers  have  no  data  of  their  own  to  compare  with  the  showing  made 
by  oil  of  chenopodium.  As  already  stated,  Herms  and  Beach  (1916) 
found  the  treatment  very  efficacious  for  worms  which  were  evidently 
Ascaridia  perspicillum,  to  judge  from  their  illustration  of  the  para- 
sites. 

Chenopodium  removed  only  2  out  of  349  cecum  worms  (Heterakis 
papulosa)  and  was  entirely  inefficacious  against  other  nematodes.  No 
tapeworms  were  passed  in  the  feces,  but  22  were  found  post-mortem, 
2  of  which  were  in  the  large  intestine.  Even  though  these  2  are  con- 
sidered as  having  been  removed  by  the  anthelmintic,  the  showing  is 
not  very  creditable.  It  may  be  stated,  however,  that  none  of  the  sub- 
stances tested  by  the  writers  for  tapeworms  on  poultry  have  proved 
very  satisfactory. 

DISCUSSION  OF  RESULTS 

In  order  to  present  the  foregoing  data  in  a  condensed  comprehensive 
summary,  the  various  experiments  have  been  tabulated  by  hosts 
(Tables  I-V).  These  tables  show,  for  each  host,  the  efficacy  of  the 
different  drugs  tested  against  the  more  important  parasites,  as  indicated 
by  the  percentage  of  worms  removed  compared  with  the  total  number 
present.  Three  columns  of  figures  are  given  for  each  parasite,  in  order 
that  the  reader  may  see  at  a  glance  not  only  the  percentage  of  efficacy 
but  the  data  from  which  this  percentage  is  derived,  and  thus  be  able 
to  judge  how  conclusive  or  otherwise  the  figures  presented  may  be. 
The  tables  also  show  the  number  of  host  animals  used  for  each  drua: 
tested,  and  the  size  of  the  dose,  usually  based  on  the  weight  of  the 
animal.  A  reference  column  gives  the  pagination  of  the  experiments 
described  in  detail  in  the  text.  Where  several  experiments  have  been 
conducted  in  testing  one  drug,  the  results  are  combined  in  the  tables 
into  a  single  set  of  figures. 


440         '  Journal  of  Agricultural  Research  voI.xii.No.  7 

While  it  is  realized  that  in  most  instances  these  data  are  insufficient 
to  express  results  in  percentages,  and  for  this  reason  reference  to  the 
percentage  of  efficacy  of  a  drug  has  been  avoided  in  the  text,  it  was 
thought  that  comprehensive  tables  like  the  following  would  prove 
convenient  in  summarizing  the  results  and  would  at  least  serve  to  indi- 
cate what  drugs  offer  promise  of  success.  Too  much  emphasis,  however, 
should  not  be  placed  on  negative  results  based  in  most  cases  on  insuffi- 
cient data. 

The  tables  do  not  include  some  of  the  earlier  experiments  carried 
out  on  hogs,  in  which  the  efficacy  of  the  drug  tested  was  determined 
solely  by  the  results  of  fecal  examination  without  killing  and  examining 
the  animals  post-mortem. 


Feb.  18,  1918 


Efficacy  of  Some  Anthelmintics 


441 


S  <u  VI 


S  OJ  ui 


SB  c 

030 


^u'2 


h'  c-a 


5  C  osg  o 

(^  8  a  "  3 


e-^ 


5Bc 

o  u  o 

H  5^ 


<Lt    G    tt  f-H    U 


Hfcl 


3Bc 
031; 


s  ci"s  art-- 


1  ■*     "*2^ 


sa 

V  ■* 

to  ro 
O   O 

O  •» 


O         O        -O 


O        O        r^        Ov  ■  O  O\oo 


o  >o       o      o 


O  ■-I  -o  o       o 


^^  fO  >^j  o       o\ 


O         f       o 


^      tn      o 


O^n   •*        O 


•^-•3  a 


C  e  M 

8ga 


aeag 

d  d  o  d  6 


a 

•»  o 


l?^a 

O  -w   O  O 


00         00O0'-« 


■^      O  1^00  o 
t      1 1  ■<»•  -a- 


3  S 

u  g 
•a  3 


I     -     " 

Si    -        ^-  -S 

«       >,     "S  a  tr  " 

^•3    g    las 21-3 

j  a  -  a  a    v^J^-o  3  g  ~  ^  g 

^"•i1=3ia-P^sN  .aSl-o-Slgl 


gs 
a-o 

■u  a 
a -a 


•a    § 

«      o 

V  01 


•rt  c. 

aSi 


.3  a 


V 

•d 

a 

a 

o 

"o 

0 
•0 

•2 

♦J 

L^ 

1 

.a 

a 

o 

■a 
a 

l-n« 


ap.« 


1  I'lp 

i  IB 

a  5^yg 

-  a  ojfc  CI 

.  g  «  a  i'*j 

Sit:  a  a  a  oj 

o  2  a  in       o 

3 1:  a  O' a  V  o  o 
ig  S  o  oj  o  o^ 

'  a;  £  ci<  a,  55  M  ID 


^B-S 
f  til 

"  o  S 
^t:  a. 


442 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  7 


Table  II. — Percentage  of  efficacy  of  various  anthelmintics  for  swine 


Drug  tested. 


Epsom  salt  in  solution  b 

Tartar  emetic  and  cas- 
tor oil. 

Chenopodium  and  cas- 
tor oil. 

Emulsion  of  chenopo- 
dium and  castor  oil 
mixed  in  feed. 


a  . 

It 


Page. 
400 

401 

436 

437 


Dose.a 


75.6  to  227 
gm.  in  water. 

260  mgm.  in 
water. 

I  to  2  mils. . . . 


3.5  mils. 


Ascarids. 


^•0 


3  Q 


I'm-      O 


Whip- 
worms. 


3a 
2 


Is 

t-r    QJ 


Nodular 
worms. 


3  a 


gse 


stomach 
worms. 


sag? 


o  The  dose  indicated  is  for  the  anthelmintic,  not  for  the  laxative  with  which  it  is  combined. 
6  Entirely  ineffective.    Pigs  would  not  drink  either  the  stronger  or  the  weaker  solution. 
c  Present, 
d  Numerous. 

Table  III. — Percentage  of  efficacy  of  various  anthelmintics  for  sheep 


Drug  tested. 


Chloroform  and  castor  oil. 
Copper  sulphate: 

In  capsule 

In  drench 


Gasoline  in  milk.    (Sum 

mary  of  3  experiments.) 
Petrolemn  benzinin  milk 
Coal-tar  phenols  in  milk. 
Chenopodium 


0*3 


Page 
404 


406 
406 


411 

413 
438 


Dose.o 


5  to  10  mils. 


o.s  gm 

(50  mils  of  I  per 

cent  solution.) 
7.5  to  30  mils , 


15  mils 

2.8  to  15  mils 

3.3  to  5.2  mils  in 
147.8  mils  milk. 


Stomach  worms. 


10, S4I 
726 


978 


3  o, 


64.0 


0.3 
93- o 


Hookworms. 


.fl-O 


3  a 
25 


S!fi 


Nodular 
worms. 


•^-g 


■3  y 


0.6 
16.0 


a  The  dose  indicated  is  for  the  anthelmintic,  not  for  the  laxative  with  which  it  is  combined. 
6  Both  sheep  died  following  the  treatment. 
e  Two  sheep  died;  the  others  passed  no  worms. 

Table  IV. — Percentage  of  efficacy  of  various  anthelmintics  for  poultry 


Drug  tested. 


Areca  nut  with  olive  oil 

Turpentine  and  castor  oil 

Tobacco  stems  and  Epsom  salt. 

Chenoiwdiimi  and  castor  oil. . . 


Page. 
419 
42  s 
428 

439 


Dose.o 


1  gm 

2  mils 

454    gni.    per  100 

birds. 
0.2  mil 


Ascaridia 
perspicillum. 


3  a. 


oThe  dose  indicated  is  for  the  anthelmintic,  not  for  the  laxative  with  which  it  is  coi|fbined. 


Feb.  i8,  1918 


Efficacy  of  Some  Anthelmintics 


443 


Table  V. — Percentage  of  efficacy  of  various  anthelmintics  for  cats 


Drug  tested. 


Oleoresin  of  aspidium  and  calomel. 
PeUetierine  tannate  and  castor  oil  < 


^0 


Page, 
416 
417 


66 


Dose.o 


0.8  mil 

64   mgm.  per 
kilo. 


Ascarids. 


S  il 


Taenia. 


|& 


2:       Ph° 


Hookworms. 


3  a 
2: 


a  The  dose  indicated  is  for  the  anthelmintic,  not  for  the  laxative  with  which  it  is  combined. 
b  Two  cats  died  shortly  after  treatment.     The  third  cat  vomited  one  Taenia. 
c  Treatment  was  entirely  ineffective.    No  tapeworms  or  nematodes  removed. 


CONCLUSIONS 

Making  due  allowance  for  the  paucity  of  data  in  regard  to  certain 
drugs,  the  writers  consider  that  the  following  may  be  reasonably  advanced 
as  the  result  of  their  investigations. 

Simple  purgatives,  calomel  and  castor  oil,  may  have  some  slight  value 
as  anthelmintics,  but  it  is  hardly  sufficient  to  justify  their  use  for  this 
purpose.  Ascarids  in  dogs  are  sometimes  removed  by  castor  oil  given  as 
a  preliminary  purge,  and  this  fact  may  prove  of  benefit  in  veterinary 
practice  as  a  diagnostic  measure  when  the  more  accurate  method  of 
microscopic  fecal  examination  can  not  be  carried  out.  However,  castor 
oil  failed  to  remove  ascarids  more  frequently  than  it  succeeded,  and  in  no 
case  were  all  the  ascarids  removed  from  any  one  animal.  As  many  of 
the  experiments  on  dogs  were  preceded  by  a  dose  of  castor  oil,  the  writers 
have  fairly  extensive  data  on  this  subject. 

The  most  reliable  vermifuge  for  ascarids,  whether  in  dogs  or  swine,  is 
oil  of  chenopodium.  This  drug,  which  was  tested  out  on  34  dogs  in  six 
experiments,  showed  an  efficacy  for  the  entire  series  of  97  per  cent. 
It  rarely  fails  to  remove  all  the  ascarids  present  in  a  dog  if  given  at  the 
rate  of  0.2  mil  per  kilo,  preceded  by  a  dose  of  castor  oil  and  the  animal 
starved  for  24  hours  before  treatment. 

The  chenopodium  treatment  is  also  very  efficacious  for  ascarids  in 
swine,  and  when  properly  administered  may  be  expected  to  remove  most, 
if  not  all,  of  the  worms  present.  It  would  seem,  however,  that  neither 
chenopodium  nor  any  other  drug  tested  will  give  satisfactory  results  if 
mixed  with  the  daily  ration  and  the  animals  allowed  to  dose  themselves; 
it  is  best  given  to  each  pig  individually  in  suitable  dosage,  preceded  by 
a  fast.  While  this  method  necessarily  involves  considerable  labor  when 
treating  animals  as  unruly  as  swine,  the  labor  can  be  reduced  by  sort- 
ing the  hogs  roughly  into  classes  according  to  size  and  confining  them 
in  inclosures  which  will  permit  them  to  be  caught  with  a  minimum 
amount  of  struggUng.     The  treatment  has  proved  practical  on  a  large 


444  Journal  of  Agricultural  Research  voi.  xii.  No.  7 

scale  and  the  results,  as  far  as  they  could  be  determined,  were  entirely 
satisfactory. 

Oil  of  chenopodium  appeared  to  be  effective  for  stomach  worms  in 
sheep,  although  the  data  on  this  subject  are  not  sufficient  to  warrant 
its  recommendation.  It  is  also  of  some  efficacy  for  hookworms  in  sheep 
and  in  dogs,  though  in  the  latter  case  chloroform  was  found  more  reliable. 

Other  remedies  which  seem  to  have  more  or  less  merit  as  anthelmintics 
against  ascarids  are  the  latex  of  Ficus  laurifolia,  santonin  in  repeated 
doses,  and  thymol.  Although  thymol  in  repeated  doses  is  fairly  effica- 
cious against  hookworms,  it  was  inferior  to  chloroform  for  this  purpose, 
causing  more  distress.  An  excellent  preparation  for  mixed  infestation 
in  dogs  consists  of  equal  parts  of  oil  of  chenopodium  and  chloroform,  given 
at  the  rate  of  0.2  mil  per  kilo,  combined  with  30  mils  of  castor  oil.  This 
preparation  may  be  expected  to  remove  all  the  ascarids  present,  a  large 
proportion  of  hookworms,  and  possibly  a  certain  percentage  of  whip- 
worms. This  latter  parasite  seems  to  be  very  difficult  to  eliminate,  and 
nothing  tried  by  the  writers  proved  very  efficacious,  almost  any  anthel- 
mintic occasionally  proving  successful.  This  experience  may  perhaps  be 
explained  by  an  intermittent  peristalsis  of  the  cecum,  which  occasionally 
allows  the  anthelmintic  to  enter,  but  which  usually  excludes  it.  Although 
chloroform  was  fairly  successful  in  removing  stomach  worms  from  sheep, 
both  animals  upon  which  it  was  tried  subsequently  died  from  its  effects, 
and  it  would  seem  to  be  too  dangerous  for  use  on  sheep. 

In  the  case  of  stomach  worms  in  sheep,  copper  sulphate  was  found  to  be 
the  most  satisfactory  remedy,  the  experiments  confirming  the  findings  of 
Hutcheon.  A  simple  apparatus  (fig.  i)  devised  by  the  senior  writer 
reduces  the  labor  involved  in  drenching  a  flock  of  sheep  and  insures 
accurate  dosage.  Petroleum  benzin  also  proved  satisfactory  and  was 
more  efficacious  for  hookworms  than  copper  sulphate.  However,  it  is 
much  more  expensive  than  copper-sulphate  solution,  must  be  given  three 
times,  and  in  a  vehicle  like  milk,  which  adds  greatly  to  the  expense.  The 
fact  that  petroleum  benzin  (refined  gasoline)  proved  efficacious,  while 
commercial  gasoline  was  considerably  less  so,  is  perhaps  related  to  the 
differences  in  specific  gravity  and  consequent  volatility  of  the  refined 
product  compared  with  the  commercial  product. 

Among  anthelmintics  intended  for  use  against  tapeworms,  male-fern 
proved  efficacious  when  tested  on  dogs.  In  the  case  of  cats  it  removed 
all  tapeworms  from  75  per  cent  of  the  animals  tested,  though  it  proved 
fatal  to  two  out  of  six  animals  which  were  somewhat  enfeebledf  rom  disease. 
Apparently  it  is  more  toxic  to  cats  than  dogs  and  should  be  prescribed 
with  caution  and  only  given  to  healthy  subjects.  So  far  as  can  be 
judged  from  a  single  experiment  with  dogs,  there  seems  to  be  no  danger 
in  combining  male-fern  with  castor  oil,  as  is  done  in  the  so-called  Her- 
mann's mixture.     In  fact,  the  writers  are  inclined  to  agree  with  Seifert 


Feb.  18, 1918  Efficacy  of  Some  Anthelmintics  445 

{1908)  that  the  administration  of  castor  oil  after  male-fern  will  avoid 
the  toxic  effects  of  the  latter  by  causing  its  rapid  and  thorough  elimina- 
tion, and  that  for  this  purpose  no  other  purgative  is  quite  so  effective. 
This  subject,  however,  should  receive  more  study  before  conclusions  are 
drawn. 

Pelletierine  tannate  was  a  failure  in  the  one  experiment  in  which  it 
was  tested  on  cats,  but  was  efficacious  on  dogs.  No  remedy  was  effica- 
cious against  tapeworms  in  poultry.  Of  the  four  drugs  tested,  chenopo- 
dium  gave  the  best  results  for  this  purpose,  but  its  efficacy  for  tapeworms 
is  very  slight. 

Turpentine  proved  the  most  efficacious  of  the  remedies  tested  on 
poultry  for  the  removal  of  Ascaridia  perspicillum,  while  chenopodium 
was  nearly  as  good.  When  tested  on  dogs  and  pigs,  turpentine  was  not 
very  efficacious;  and,  as  it  caused  grave  symptoms  of  nephritis  in  pigs 
and  caused  the  death  of  some  of  the  experiment  dogs,  its  use  upon  these 
animals  is  inadvisable. 

The  treatment  with  chopped  tobacco  stems  recommended  by  Herms 
and  Beach  for  ascarids  in  poultry  proved  fairly  efficacious  for  Heterakis 
papulosa  and  would  presumably  be  at  least  as  efficacious  for  Ascaridia 
perspicillum,  since  this  latter  worm  is  more  easily  reached  by  anthel- 
mintics than  is  H.  papulosa. 

There  are  a  large  number  of  drugs  showing  a  greater  or  less  degree  of 
efficacy  for  the  various  intestinal  parasites  of  domestic  animals.  Usually 
their  action  is  selective — that  is,  they  show  a  pronounced  efficacy  for 
certain  species  of  intestinal  worms,  while  they  are  decidedly  less  efficacious 
or  entirely  inefficacious  against  other  intestinal  parasites.  If  we  consider 
the  ideal  anthelmintic  one  which  will  remove  all  worms  of  a  given  class 
or  species,  and  do  this  every  time  in  a  single  dose,  we  find  that  very  few 
drugs  approach  this  ideal. 

Among  the  drugs  which  have  given  the  best  results  under  experimental 
conditions  for  the  purposes  intended  and  concerning  which  the  writers 
have  sufficient  data  to  warrant  positive  conclusions  may  be  mentioned 
the  following : 

(i)  Copper  sulphate  in  drench  for  stomach  worms  in  sheep. 

(2)  Oil  of  chenopodium  for  ascarids  in  pigs  and  dogs. 

(3)  Oleoresin  of  male-fern  for  tapeworms  in  dogs. 

(4)  Turpentine  for  Ascaridia  perspicillum  in  fowls. 

(5)  Chopped  tobacco  stems  for  Heterakis  papulosa  in  fowls. 
27810°— 18 i 


446  Journal  of  Agricultural  Research  voi.xii,  No.  7 

LITERATURE  CITED 
Alessandrini,  Giulio. 

191 5.  LE    MALATTIE     DE    PARASSITI     ANIMALI     NEGLI     ESERCITI    COMBATTENTI.      Itt 

Policlinico,  sez.  prat.,  v.  22,  no.  25,  p.  822-827. 
ArbucklE. 

1916.  ROTATE  YOUR  PASTURE  AND  AVOID  DISASTER.    In  Amcf.  Sheep  Breeder, 

V.  36,  no.  7,  p.  402-403. 
Berrio,  Posada. 
1911.  traitement  de  la  trichoc^phalose  par  le  latex  d'higueron.    In  Rev. 
M6d.  et  Hyg.  Trop.,  t.  8,  no.  3,  p.  191-192.     Discussion,  p.  192-193. 
Billings,  W.  C,  and  Hickey,  J.  P. 

1916.  SOME  points  about  hookworm  disease,  its  diagnosis  and  treatment. 

/nJour.Amer.  Med.  Assoc,  V.  67,no.  26,  p.  1908-1912,4 fig. 
Campbell,  D.  M. 

1917.  EPSOM  salt  an  effective  vermifuge.    In  Amer.  Jour.  Vet.  Med.,  v.  12, 

no.  5,  p.  314. 
Coffey,  W.  C. 

1915.  checking  stomach  worms  in  sheep.    In  Breeder's  Gaz.,  v.  68,  no.  6,  p.  192. 
Heiser,  V.  G. 

1915.  RECENT    EXPERIENCES    IN    THE    ORIENT    WITH   CHENOPODIUM    AS    A   REMEDY 

AGAINST  HOOKWORM  AND  OTHER  INTESTINAL  PARASITES.      In  JoUT.  Amer. 

Med.  Assoc,  v.  65,  no.  6,  p.  526-527. 
Henkel.  Alice. 

1913.   AMERICAN     MEDICINAL     FLOWERS,     FRUITS,     AND     SEEDS.     U.  S.  Dept.  AgT. 

Bul.  26,  16  p.,  12  fig. 

Herms,  W.  B.,  and  Beach,  J.  R. 

1916.  ROUND  WORMS  IN  POULTRY— LIFE  HISTORY  AND  CONTROL.      Cal.   AgT.   Exp. 

Sta.  Circ.  150,  7  p.,  3  fig. 

HUTCHEON,  D. 

x89if .  THE  TREATMENT  OF  WIRE- WORMS  IN  SHEEP.    In  Agt.  JouT.  Cape  Good  Hope, 
V.  4,  no.  I,  p.  8-9. 


189IC.   WIRE-WORMS  IN  SHEEP  AND  GOATS,  AND  THEIR  TREATMENT  WITH  SXn<PHATB 

OP  COPPER.    In  AgT.  Jour.  Cape  Good  Hope,  v.  3,  no.  19,  p.  179-181. 


1892.  BLUESTONE  FOR  WIRE-WORMS.    In  Agr.  JouT.  Cape  Good  Hope,  v.  4,  no. 
20,  p.  24c. 


1895.  WIRE-WORMS.    In  Agr.  Jour.  Cape  Good  Hope,  v.  8,  no.  i,  p.  18-19. 
Lenhartz,  Hermann. 

1902.   BEHANDLUNG        DER        DURCH        DARMSCHMAROTZER        HERVORGERUFENEN 

ERKRANKUNGEN.    In  Handbuch  der  speciellen  Therapie  innerer  Krank- 
heiten.     Aufl.  3,  Bd.  4,  p.  604-631.     Jena. 
LUCKEY,  D.  F. 

1915.   HOW   TO   TREAT    A   FLOCK    OP    300    LAMBS    INFESTED   WITH    STOMACH   WORMS 

(STRONGYLUS  CONTORTUS)?    In  Proc  24th  Ann.  Meeting  Mo.  Vet.  Med. 
Assoc,  p.  68-69. 
Miller,  F.  H. 
1904.  haemorrhagic  colitis  of  the  dog  due  to  infection  with  the  tricho- 
CEPHALUS  DEPRESSIUSCULUS.     (True  whipworm).     In  Amer.  Vet.  Rev., 
V.  28,  no.  8,  p.  722-729,  I  pi. 
MOUAT-BIGGS,  C.  E.  F. 

I915.  THE  TREATMENT  OF  ANKYLOSTOMIASIS  IN  VENEZUELA.      In  TranS.  SoC  Trop. 

Med.  and  Hyg.,  v.  8,  no.  7,  p.  216.    Discussion,  p.  217-218, 


Feb.  i8, 1918  Efficacy  of  Some  Anthelmintics  447 

Nelson,  E.  K. 
1911.  a  chemical  investigation  of  the  oil  op  chenopodium.    u.  s.  dept. 
Agr.  Bur.  Chem.  Circ.  73,  10  p. 


19 13.   A  CHEMICAL  INVESTIGATION   OP  THE  COMPOSITION  OF  THE  OH,  OF  CHENOPO- 
DIUM.    U.  S.  Dept.  Agr.  Biir.  Chem.  Circ.  109,  8  p. 
Railliet,  a. 

191 5.  l'EMPLOI   DES  ME;dICAMENTS   dans  LE  TRAITEMENT  DES  MALADIES  causiSes 
PAR    DES    Ni&MATODES.     In  Rcc.  Med.  Vet.,  t.  91,  no.  15,  p.  490-513. 
Ransom,  B.  H. 

1907.  STOMACH  worms  (hAEMONCHUS  CONTORTUS)  IN  SHEEP.     U.   S.   Dept.  Agr. 
Bur.  Anim.  Indus.  Circ.  102,  7  p. 
Salant,  William,  and  Livingston,  A.  E. 

1915.   the  influence  op  the  oh,  op  CHENOPODIUM  ON  THE  CIRCULATION  AND  RES- 
PIRATION.    In  Amer.  Jour.  Physiol.,  v.  38,  no.  i,  p.  67-92,  14  fig. 
and  Mitchell,  C.  W. 

I915.  THE   INFLUENCE    OP    OIL    OF   CHENOPODIUM    ON   INTESTINAL   CONTRACTILITY. 
In  Amer.  Jour.  Physiol.,  v.  39,  no.  i,  p.  37-53,  9  fig. 
and  Nelson,  E.  K. 


1915.  THE  TOXICITY  OF  OIL  OF  CHENOPODIUM.    In  Amer.  Jour.  Physiol.,  v.  36, 
no.  4,  p.  440-463- 

SCHULTZ,  W.  H. 

191 1.  REMEDIES  FOR  ANIMAL  PARASITES.      A  STUDY  OP  THE  RELATIVE  EFFIQENCY 
AND  DANGER  OF  THYMOL  AS  COMPARED  WITH  CERTAIN  OTHER  REMEDIES 

PROPOSED  FOR  HOOKWORM  DISEASE.    In  JouT.  Amer.  Med.   Assoc,  v. 
57,  no.  14,  p.  1102-1106. 
Seifert,  otto. 

1885.  DIE  DARMPARASITEN  DES    MENSCHEN.     In  Dcut.  Med.  Ztg.,    Bd.  6,  No.  8, 
p.  85-87;  No.  9,  p.  97-98;  No.  10,  p.  109-111. 
Cites  Schidlowski,  p.  98. 


1908.  KLINISCH-THERAPEUTISCHER  TEIL.    In  Braun,  Max.     Die  tierischen  Paia- 
siten  des  Menschen.    Aiifl.  4,  p.  477-623.    Wurzburg. 
Stiles,  C.  W. 
1901.  treatment  for  round  worms  in  sheep,  goats,  and  cattle.    u.  s.  dept. 
Agr.  Bur.  Anim.  Indus.  Circ.  35,  8  p. 


1902.  FURTHER  INVESTIGATIONS  ON  VERMINOUS  DISEASES  OF  CATTLE,  SHEEP,  AND 

GOATS  IN  TEXAS.     U.  S.  Dept.  Agr.  Bur.  Anim.  Indus.  i8th  Ann.  Rpt., 
1901,  p.  223-229. 
WiNSLOW,  Kenelm. 

I913.  VETERINARY  MATERIA  MEDICA  AND  THERAPEUTICS,      ed.  7,  781  p.  New  York, 


TOBACCO  WILDFIRE^ 

By  Frederick  A.  Wolf,  Plant  Pathologist,  and  A.  C.  Foster,  Assistant  Plant  Path- 
ologist North  Carolina  Agricultural  Experiment  Station 

INTRODUCTION 

During  the  past  season  (191 7)  a  leaf  disease  of  tobacco  {Nicotiana 
tahacum)  has  been  the  cause  of  much  concern  to  tobacco  growers  because 
of  its  destructiveness.  Attention  was  first  directed  to  it  during  June 
when  tobacco  was  being  transplanted.  Subsequently  during  the  entire 
growing  season  numerous  complaints  of  this  disease  were  received  by 
members  of  the  staff  of  the  North  Carolina  Agricultural  Experiment 
Station  and  of  the  State  Department  of  Agriculture.  Because  of  the 
severity  of  the  epidemic  and  the  insistence  by  growers  that  this  disease 
was  manifestly  different  from  any  they  had  ever  seen,  an  investigation 
was  begun. 

A  preliminary  survey  of  the  literature  on  diseases  of  tobacco  revealed 
the  fact  that  this  disease  was  clearly  unlike  any  which  had  previously 
been  described.  Furthermore,  the  apparently  water-soaked  margin  of 
the  diseased  areas,  the  tissues  of  which,  upon  microscopic  examination, 
were  found  to  be  teeming  with  bacteria,  suggested  that  the  disease  was 
probably  of  bacterial  origin.  Accordingly,  attention  in  the  first  studies 
was  centered  upon  the  etiology  of  the  disorder.  A  brief  statement  (7)  ^ 
concerning  this  work,  in  which  attention  was  directed  to  the  presence  of 
the  disease,  was  duly  prepared. 

It  was  also  pointed  out  that  the  causal  organism  was  undescribed, 
and  the  name  "Bacterium  tahacum  Wolf  and  Foster"  was  suggested. 
Promise  was  made  in  that  report  of  a  description  of  the  morphological  and 
cultural  studies  upon  Bact.  tahacum.  Besides  making  this  description,  it 
is  the  present  purpose  to  adequately  describe  the  disease  and  to  indicate 
our  present  knowledge  of  its  economic  importance,  distribution,  and 
dissemination. 

HISTORY  AND   DISTRIBUTION   OF  THE   DISEASE 

While  the  disease  was  first  definitely  recognized  in  June,  191 7,  near 
Wendell,  N.  C,  it  is  impossible  to  determine  with  certainty  for  how  many 
seasons  prior  to  the  present  one  the  disease  has  existed.  It  is  quite 
probable,  to  judge  from  the  testimony  of  several  reliable  informants,  that 
the  disease  caused  the  loss  of  practically  the  entire  crop  in  one  field,  near 
Wendell,  in  191 6.  Mr.  E.  G.  Moss,  Assistant  Director,  in  Charge  of  the 
Granville  Branch  Tobacco  Station,  Oxford,  N.  C,  is  convinced,  as  is  also 

'  Approved  tor  publication  by  B.  W.  Kilgore,  Director,  North  Carolina  Agricultural  Experiment  Station. 
*  Reference  is  made  by  nimiber  (italic)  to  "Literature  cited,"  p.  458. 

Journal  of  Agricultural  Research,  Vol.  XII,  No.  7 

Washington,  D.  C.  ^  Feb.  18,  1918 

ma  '  KeyNo.  N.C— 9 

(449) 


450  Journal  of  Agricultural  Research  voi.  xii.  No.  ^ 

the  senior  writer,  that  the  malady  was  observed  by  him  during  191 6  at 
Creedmoor,  N.  C.  To  judge  from  the  additional  fact  that  the  disease  has 
been  collected  during  the  past  season  in  19  counties  within  North  Caro- 
lina (Surry,  Stokes,  Forsythe,  Guilford,  Rockingham,  Caswell,  Alamance, 
Orange,  Person,  Durham,  Chatham,  Moore,  Hoke,  Wake,  Johnston, 
Franklin,  Granville,  Vance,  and  Warren)  and  in  3  within  Virginia  (Pitt- 
sylvania, Halifax,  and  Mecklenburg)  it  is  highly  probable  that  the  disease 
existed  prior  to  the  present  year.  Whether  or  not  this  disease  occurs  in 
other  of  the  States  in  which  tobacco  is  grown  is  not  known,  except  in  the 
case  of  Wisconsin.  A  letter  from  Prof.  James  Johnson,  Department  of 
Horticulture,  University  of  Wisconsin,  who  has  had  occasion  to  observe 
the  disease  in  North  Carolina,  states  that  a  similar  bacterial  spot,  although 
not  nearly  so  destructive  as  in  North  Carolina,  has  been  observed  by  him 

in  Wisconsin. 

ECONOMIC  IMPORTANCE 

The  disease  is  universally  conceded  by  growers  of  tobacco  to  be  the 
most  destructive  one  which  attacks  this  crop.  Losses,  ranging  from 
those  which  were  inappreciable  to  those  in  which  almost  the  entire  crop 
was  destroyed,  were  sustained  in  every  locality  where  the  disease  was 
present.  In  some  instances  fields  upon  the  same  farm  were  observed 
to  be  badly  affected,  while  others  had  little  or  none  of  the  disease.  In 
some  sections,  too,  the  disease  occurred  upon  every  farm  within  a  radius 
of  several  miles,  while  in  others  it  was  present  only  in  an  occasional 
field.  It  is  not  possible,  therefore,  because  the  disease  was  not  uniformly 
destructive  over  the  entire  area  in  which  it  is  known  to  occur  and  because 
time  has  not  been  afforded  to  make  a  careful  survey,  to  obtain  a  reasonable 
estimate  of  the  damage  wrought.  Some  idea  of  the  losses,  however,  can 
be  gained  from  statements  taken  from  reports  kindly  furnished  by  a 
number  of  growers,  who  compared  their  leaf-tobacco  sales  with  neighbors, 
whose  crops  were  free  from  wildfire  or  were  at  most  only  slightly  affected. 
Some  of  these  reported  losses  averaging  $100  per  acre  for  their  entire 
crop.  One  correspondent  estimated  his  total  loss  at  $5,000  and  said 
that  hundreds  of  farmers  in  his  section  suffered  an  equal  acreage  loss. 

APPEARANCE  OF  THE  DISEASE 

The  disease  was  first  noted  early  in  June  during  a  period  of  rainy 
weather  accompanied  by  nights  which  were  so  cold  as  to  retard  the 
growth  of  tobacco.  The  affected  plants  in  many  fields  perished,  necessi- 
tating replanting  a  second  or  a  third  time.  A  period  of  relatively  dry, 
warm  weather  of  about  a  month's  duration  followed,  during  which  time, 
the  crop  made  an  extremely  rapid  growth,  as  shown  by  the  fact  that  the 
plants  were  sufficiently  mature  to  be  topped.  At  this  stage  of  their 
development  another  rainy  season  of  about  a  week's  duration  occurred 
and  was  followed  by  another  epidemic  of  the  disease.     The  disease 


Feb.  i8, 1918  Tobacco  Wildfire  451 

appeared  so  quickly,  spread  so  rapidly,  and  affected  the  leaves  so  seri- 
ously that  it  was  commonly  given  the  appropriate  designation  "wildfire." 

The  foliage  alone  seems  to  be  subject  to  attack.  The  first  evidence 
of  disease  is  the  appearance  of  circular,  chlorotic  areas  varying  from  0.5 
to  I  cm.  in  diameter.  Within  24  hours  after  this  chlorosis  is  first  noted 
minute  brown  areas  will  have  formed  at  the  centers  of  the  spots  (PI. 
15,  A).  Within  another  day  these  spots  will  have  enlarged  greatly 
PI.  15,  B),  and  a  border  of  water-soaked  appearance  marks  the  margin 
of  the  necrotic  tissues.  Within  a  few  more  days  the  diseased  areas  are 
2  to  3  cm.  in  diameter  and  are  often  strikingly  concentric  with  shades 
of  tan  to  dark  brown,  the  centers  being  lightest  in  color  (PI.  16,  A). 
Such  spots  have  a  broad  translucent  border,  which  is  in  turn  surrounded 
by  a  chlorotic  halo  that  pales  out  into  adjacent  tissues  (PI,  15,  B;  i6,  A). 
When  the  spots  are  numerous,  they  fuse,  causing  large,  irregular  areas 
of  leaf  tissue  to  become  dry.  These  dead  areas  remain  intact  in  case 
there  is  no  precipitation.  When  dewy  nights  and  intermittent  showers 
occur,  however,  the  dead  areas  rot  out  so  that  the  leaves  present  a  ragged 
appearance  (PI.  16,  C)  which  is  especially  manifest  when  large  numbers 
of  infections  occur  upon  a  single  leaf. 

Not  uncommonly  the  leaves  on  one  side  of  the  plant  are  more  seriously 
diseased  than  those  on  the  opposite  side  and  there  may  even  be  a  uni- 
lateral destruction  of  these  leaves  resulting  in  distortion,  as  shown  in 
Plate  16,  B.  The  vascular  tissues  seem  not  to  be  invaded,  but  the 
organism  confines  its  attack  to  parenchymatous  tissues. 

Two  other  leaf  spot  diseases  of  tobacco,  frogeye  and  speck,  are  present 
within  North  Carolina,  from  both  of  which  wildfire  is  easily  distinguish- 
able. Frogeye  appears  as  circular,  brown  spots,  with  a  darker  border 
and  with  grayish  centers.  Upon  this  gray  center  may  be  seen  the 
fructifications  of  Cercospora  nicotianae  E.  and  E.,  or  other  fungi  associated 
with  the  disease.  No  chlorosis  accompanies  these  spots.  Speck,  which 
results  from  a  deficiency  of  potash,  appears  as  tan-colored,  irregular 
areas  which  are  first  present  at  a  distance  from  the  principal  veins. 
When  this  disease  is  accompanied  by  chlorosis,  there  is  no  definite  halo 
around  the  lesions.  In  the  case  of  neither  of  these  diseases  is  the  margin 
of  the  affected  areas  water-soaked  in  appearance,  and  in  neither  of  them 
do  the  affected  tissues  disintegrate  and  fall  out. 

ISOLATIONS  AND  INOCULATIONS 

On  June  13,  fresh  material  of  tobacco  wildfire  was  collected  and 
isolations  were  made  by  planting  on  poured  plates  of  nutrient  agar  frag- 
ments of  tissue  from  the  margin  of  affected  areas.  Contamination  was 
avoided  by  washing  the  leaves  prior  to  making  the  planting  in  mercuric 
chlorid  and  then  rinsing  them  in  sterile  water.  Several  types  of  colonies 
developed  along  the  margins  of  these  plantings,  the  most  common  of 
which  was  Bad.  tabacum,  which  appeared  as  glistening,  grayish  white 


452  Journal  of  Agricultural  Research  voi.xii.  No.  7 

colonies.  It  was  possible,  in  some  cases,  to  make  transfers  directly  from 
these  colonies  to  tubes  of  agar  and  secure  pure  cultures.  In  others, 
dilution  poured  plates  were  first  made,  and  the  organism  was  transferred 
from  certain  of  the  colonies  which  developed  to  tubes  of  agar.  Diseased 
material  was  collected  several  times  subsequently,  and  numerous  speci- 
mens were  received  by  mail,  so  that  opportunity  was  afforded  during  the 
season  to  isolate  the  organism  from  several  sources. 

Some  preliminary  inoculations  were  made  on  June  13  in  which  the 
inoculum  consisted  of  macerated,  diseased  leaves  upon  which  a  quantity 
of  water  was  poured.  About  100  young  tobacco  plants  growing  in  a 
flat  in  the  greenhouse  were  then  sprinkled  with  this  water.  Four  days 
later  infections  were  evident  by  the  appearance  of  numerous  yellow  spots 
with  pin-point-like  centers. 

On  the  evening  of  June  28  two  potted  tobacco  plants,  about  i8  inches 
in  height,  were  inoculated  with  pure  cultures  of  Bad.  tabacum.  A 
watery  suspension  from  agar  cultures  was  sprinkled  upon  these  plants, 
after  which  they  remained  covered  with  a  bell  jar  for  36  hours.  On  the 
morning  of  July  2  numerous  chlorotic  areas  had  formed,  which  by  July  5 
had  changed  to  large  dry  spots,  typical  of  wildfire.  No  difficulty  was 
experienced  in  reisolating  the  organism  from  these  lesions. 

On  July  10,  16  plants  which  had  been  transplanted  in  the  field  on 
June  14  were  inoculated  by  sprinkling  them  with  suspensions  made  from 
bouillon  cultures.  These  plants  were  not  seen  again  until  July  15,  when 
large  brown  areas  had  formed  abundantly,  whereas  adjacent  uninocu- 
lated  plants  remained  healthy. 

Another  series  of  inoculations,  involving  18  plants  growing  in  pots 
placed  outside  of  the  greenhouse,  was  made  on  July  25.  In  this  case 
the  leaves  were  immersed  in  a  bacterial  suspension.  Seventy-two  hours 
later  the  first  evidence  of  infection  was  observ^ed.  Here  again  the 
organism  was  recovered  from  mature  lesions. 

Another  set  of  inoculations,  involving  12  potted  tobacco  plants,  was 
made  with  what  proved  to  be  Boat,  tabacum  isolated  from  spots  on 
cowpeas  (Vigna  sinensis)  which  had  been  planted  between  the  hills  in 
a  badly  diseased  tobacco  field.  The  spots  on  cowpeas,  from  which  these 
isolations  were  made,  were  very  similar  in  appearance  to  3-  or  4-day-old 
lesions  on  tobacco.  Eight  tobacco  plants  were  inoculated  on  July  27 
and  four  on  August  2.  Inoculation  was  effected  by  sprinkling  the  plants 
with  a  bacterial  suspension.  By  August  i  in  the  first  case  and  August  9 
in  the  second  there  was  no  doubt  that  the  diseased  areas,  which  had 
formed  upon  all  of  the  inoculated  plants,  were  typical  of  wildfire.  The 
organism  was  reisolated  from  these  spots  and,  together  with  transfers 
from  the  original  cultures  from  cowpeas,  was  used  in  inoculating  cowpeas. 
Here  again  the  same  method  as  before  was  used  in  making  inoculations. 
Only  a  few  spots  developed  upon  the  several  plants  employed  in  two  sets 
of  inoculations.     These  spots  were  similar  to  those  on  cowpeas  growing 


Feb.  i8, 1918  Tobacco  Wildfire  453 

in  the  field  of  diseased  tobacco.  Microscopic  examination,  furthermore, 
showed  that  the  dead  tissues  were  filled  with  bacterial  organisms. 

The  lesions,  both  naturally  and  artificially  produced,  are  believed  to 
have  originated  around  punctures  made  by  leaf-hoppers,  which  were 
abundantly  present  on  these  plants  throughout  the  season.  The  wild- 
fire organism  is  capable  of  multiplying  within  the  cells  weakened  as  a 
result  of  the  withdrawal  of  their  contents  by  the  feeding  of  these  insects, 
but  is  not  able  to  parasitize  normal  cells.  Drops  of  moisture  laden  with 
bacteria  certainly  dripped  from  the  diseased  tobacco  plants  to  the  cowpeas 
beneath  them,  and  could  thus  have  supplied  the  inoculum  which  caused 
the  cowpea  foliage  to  become  spotted.  This  explanation  is  supported 
by  the  observation  that  the  lesions  on  cowpeas  did  not  increase  in  size 
beyond  pinpoint-like  dead  areas,  indicating  that  Bact.  tabacum  can  not 
adapt  itself  to  invade  healthy  tissues,  and  by  the  further  fact  that  no 
new  spots  developed  subsequently  on  the  naturally  and  artificially 
inoculated  plants.  Furthermore,  spots  never  developed  on  cowpeas 
growing  at  a  distance  from  diseased  tobacco  plants — that  is,  where  they 
could  not  be  infected  through  the  agency  of  water  dripping  from  diseased 
tobacco  plants.  Bact.  tabacum,  therefore,  is  not  parasitic  upon  cowpeas, 
and  its  chance  occurrence  upon  this  crop  indicates  that  conclusions  as  to 
the  pathogenicity  of  bacteria  when  judged  from  inoculation  experiments 
in  which  the  inoculum  is  introduced  through  wounds  are  not  entirely 
convincing.  In  view,  therefore,  of  the  fact  that  tobacco  is  not  grown  in 
the  vicinity  of  West  Raleigh,  where  the  inoculation  experiments  were 
conducted,  and  that  all  of  the  uninoculated  plants  within  the  greenhouse 
grew  to  maturity  without  any  manifestation  of  wildfire,  there  is  no  doubt 
that  all  of  the  infections  which  were  secured  resulted  from  inoculations 
with  the  organism  in  hand.  When  judged  by  the  readiness  with  which 
infection  occurs,  Bact.  tabacum  is  to  be  regarded  as  a  very  vigorous 
pathogene. 

Aside  from  the  inoculations  upon  cowpeas,  only  two  other  host  species, 
bell  peppers  (Capsicum,  annuum,)  and  Jimson  weed  {Datura  tatula),  were 
employed,  with  negative  results. 

PATHOLOGICAL  ANATOMY 

Affected  tissues  were  fixed  in  95  per  cent  alcohol,  embedded  in  paraffin, 
sectioned,  and  stained  with  carbol-fuchsin.  The  presence  of  a  crystal- 
line substance  whose  nature  is  described  in  a  recent  paper  by  Ridgway  (5) 
interfered  seriously  with  the  cutting  of  suitable  sections.  In  tissues  in 
which  the  cells  had  not  yet  become  dry  and  collapsed,  bacteria  are  abund- 
antly present  within  the  intercellular  spaces  (fig.  i).  In  mature  lesions, 
however,  they  occur  also  within  the  cells.  The  contents  of  such  cells 
appear  to  have  been  completely  destroyed,  whereas  the  walls  have 
undergone  little  disintegration.  The  complete  disintegration  of  diseased 
tissues,  which  occurs  in  the  presence  of  excessive  moisture,  results,  it 


454 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  7 


is  believed,  from  the  activity  of  other  species  of  bacteria  which  enter 
the  cells  following  invasion  by  Baci.  tabacum.  These  species  appear 
always  to  be  present  in  old  lesions  as  judged  by  the  isolation  studies. 

DESCRIPTION  OF  BACTERIUM  TABACUM 

Bacterium  tabacum,  emend. 

The  primary  cause  of  tobacco  wildfire  is  a  grayish  white,  rod-shaped 
organism  with  rounded  ends.     It  is  motile  by  means  of  a  single  polar 


Fig.  1. — Parenchyma  cells  from  the  margin  of  a  lesion  showing  Bacierium  tabacum  in  the  intercellular 

spaces  and  within  the  cells. 

flagellum  which  is  about  twice  as  long  as  the  body  of  the  bacterium 
(fig.  2,  a).     Motility  can  be  observed  when  fresh  material  is  examined  in 

a  drop  of  water  or  when   prepara- 

y       ^  |k  (  •tW  tions    are    made    from    i8-hour-old 

A     V   V     A  f        ^^N     bouillon  or  agar  cultures.     Flagella 
m     { (2. 1      M     ■        ^^^       f     ^^^  easily  demonstrated  when   the 
/    /     /      m  Aw        ^^^^3     organism  from  such  cultures  is  sub- 
's I  ^      /      JF      m    ^H^^  jected  to  the  staining  method  out- 
U  ^ ^^  ^       fl                          lined  by  Morrey  (2).    The  bacterium 
^^^^f    ''^^--^    ^W  usually  occurs  singly  within  the  host 
^^        ^'^"yy^^                  tissues,  but  in  culture  chains  of  as 
^^Btb^ima^^i^mmm^  .^mmmm,u,^^B^      many   as   five  elements  have  been 

noted  (fig.  2,  fc).  The  limits  of  size 
vary  from  2.4  to  5  by  0.9  to  1.5  \x, 
the  most  common  size  being  3.3  by 
1.2  /x.  No  involution  forms  have 
been  observed;  neither  have  endospores  been  demonstrated. 

The  organism  stains  readily  with  aqueous-  and  carbol-fuchsin,  anilin 
gentian-violet,  and  methylene  blue.  It  is  Gram-negative,  however, 
and  is  not  acid-fast.  Neither  is  a  capsule  demonstrable  by  the  methods 
of  Rabiger  or  Welch. 

The  organism  has  been  cultured  mainly  upon  glycerin  agar  and  potato 
agar.     Colonies  appear  in  these  media  on  the  second  day  in  poured  plates 


Fig.  2. — a,  Flagella  of  Bacierium  tabacum  stained 
by  Morrey 's  method;  b.  Bad.  tabacum,  from 
bouillon  stained  with  carbol-fuchsin,  showing 
arrangement  of  the  elements. 


Feb.  i8, 1918  Tobacco  Wildfire  455 

kept  at  a  temperature  of  20°  to  25°  C.  By  the  fourth  day  the  surface 
colonies  have  attained  a  diameter  of  2  to  3  mm.  They  are  grayish  white 
in  color,  circular  in  outline,  are  appreciably  raised,  and  have  a  smooth 
margin  and  a  smooth,  wet -shining  surface.     Buried  colonies  are  biconvex. 

In  stroke  cultures  a  filiform  growth  which  widens  at  the  base  of  the 
slant  is  formed.  Growth  is  moderate  and  does  not  give  rise  to  the 
production  of  odor.  In  stab  cultures  growth  is  best  at  the  surface  of 
the  agar,  and  the  line  of  puncture  is  filiform.  In  stab  cultures  on  nutri- 
ent gelatin  a  filiform  growth  also  appears  along  the  line  of  puncture, 
with  the  greatest  growth  at  the  surface  of  the  medium.  No  evidence  of 
liquefaction  occurs  until  the  tenth  day,  when  it  becomes  crateriform, 
and  is  complete  within  30  days. 

Growth  on  potato  cylinders  is  nontypical  in  appearance,  and  there  is 
no  evidence  of  diastatic  activity. 

With  a  2  per  cent  solution  of  Difco  peptone  as  a  basal  solution,  five 
solutions  were  prepared  by  adding  i  per  cent  of  one  of  the  following 
carbon  compounds :  Dextrose,  saccharose,  lactose,  glycerin,  and  dextrin. 
No  gas  formed  in  fermentation  tubes  containing  any  of  these  media. 
A  vigorous  growth  with  strong  clouding  and  a  surface  pellicle  occurred 
in  all  in  the  open  arm.  In  the  presence  of  dextrose  and  saccharose  a 
distinctly  visible  clouding  gradually  extended  upward  into  the  closed 
arm,  while  in  the  case  of  the  other  carbon  compounds  the  closed  arms 
remained  clear.     The  organism  is  therefore  regarded  as  aerobic  in  general. 

Acid  formation  in  stab  cultures  on  litmus-glycerin,  litmus-dextrose, 
litmus-lactose,  and  litmus-saccharose  agar  begins  within  four  to  six 
days,  but  no  gas  formation  occurs  on  any  of  these  media. 

Growth  on  litmus  milk  presents  a  very  characteristic  appearance. 
During  the  first  three  days  following  inoculation  there  is  a  deepening 
of  the  blue  color.  Two  days  later  coloration  begins  to  become  strati- 
form, and  by  the  seventh  or  eighth  day  four  distinct  layers  are  evident. 
The  upper  one  is  between  plumbeus  and  violaceous  in  color  (5),  the 
next  is  lilacinus,  the  third,  violaceous,  and  the  lowermost  approximates 
caesius.  These  colors  lose  their  intensity  after  a  few  days  and  become 
more  or  less  blended,  and  by  the  tenth  day  there  is  evident  reduction 
of  the  litmus  and  precipitation  of  the  casein.  Reduction  proceeds 
rather  slowly  and  is  complete  by  the  twenty-fourth  day. 

There  is  no  reduction  of  nitrates  in  nitrate-peptone  broth,  although 
a  conspicuous  clouding  occurs.  Furthermore,  no  gas  is  formed;  and 
the  tests  for  ammonia,  indol,  and  skatol  were  negative. 

The  thermal  death  point  of  this  organism,  as  determined  by  exposing 
newly  inoculated  tubes  of  bouillon  in  the  customary  manner,  was  found 
to  be  about  65°  C.  It  is  manifestly  quite  sensitive  to  desiccation, 
since  no  growth  appeared  after  six  days  when  bouillon  cultures  were 
placed  on  sterile  slides  in  sterile  petri  dishes. 

The  group  number  of  Baclerhim  iabactim  according  to  the  numerical 
system  of  the  Society  of  American  Bacteriologists  is  2>2i-2222032. 


456  Journal  of  Agricultural  Research  voi.  xii,  No.  7 

DISSEMINATION  OF  WILDFIRE 

The  fact  that  the  disease  appeared  in  epidemic  form  twice  in  one 
season  and  that  each  epidemic  followed  a  rainy  period,  with  little  or 
no  new  infection  in  the  interim  suggests  that  dissemination  is  primarily 
influenced  by  moisture.  This  is  in  accord  with  observations  on  the 
dissemination  of  certain  other  plant  diseases  of  bacterial  origin,  as 
angular  leafspot  of  cotton  (4)  and  Citrus  canker  (6). 

Very  striking  evidence  was  found  in  two  instances  that  wind  is  also 
a  potent  factor  in  the  spread  of  wildfire.  In  one  instance  a  field  was 
observed  before  and  several  days  following  a  rain  which  was  accom- 
panied by  a  high  wind.  The  disease  had  advanced  in  consequence 
over  a  distance  involving  1 6  to  18  rows  lying  parallel  to  diseased  tobacco. 
The  disease  terminated  rather  abruptly  beyond  this  distance.  In 
another  locality  no  disease  occurred,  except  in  the  case  of  a  field  of 
approximately  an  acre  in  area,  the  plants  for  which  had  been  brought 
for  a  distance  of  several  miles.  Here,  again,  the  disease  spread,  fol- 
lowing the  same  storm  into  an  adjacent  field  to  the  leeward,  was  most 
abundant  near  the  field  of  diseased  plants,  and  gradually  became  less 
in  the  direction  away  from  the  diseased  field.  The  organism  had  very 
evidently  been  spread  by  wind-blown  rain,  a  phenomenon  in  accord 
with  observations  by  Faulwetter  (j)  upon  the  angular  leafspot  of  cotton. 

When  the  first  epidemic  was  prevalent  in  the  vicinity  of  Oxford,  N.  C, 
thrips  were  abundantly  present  upon  tobacco  and  were  popularly  sus- 
pected of  being  responsible  for  the  spread  of  the  disease.  Accordingly, 
diseased  leaves  bearing  numbers  of  these  insects  were  collected,  and 
the  thrips  were  liberated  upon  healthy  plants  in  the  greenhouse  at 
West  Raleigh.  Careful  watch  was  kept,  but  no  evidence  of  wildfire 
developed  upon  any  of  the  plants.  In  the  same  season  only  a  few 
thrips  were  found  in  diseased  fields  near  Wendell,  N.  C,  during  an 
entire  afternoon's  search.  For  these  reasons  it  is  improbable  that 
thrips  are  to  be  regarded  as  agents  of  dissemination. 

Following  the  first  outbreak  of  wildfire,  opportunity  was  afforded  to 
make  numerous  observations  upon  the  origin  of  the  disease.  In  every 
instance  where  the  disease  occurred  in  the  field  it  has  been  possible  to  find 
that  plants  in  the  seed  beds  or  "plant  beds"  were  also  affected.  It  was 
adjudged,  therefore,  that  the  disease  must  have  been  introduced  into  the 
plant  beds  either  through  the  use  of  infected  seed  or  through  the  agency 
of  fertilizers.  One  large  seed  farm  upon  which  diseased  plants  occurred 
was  visited  in  searching  for  wildfire  lesions  upon  seed  pods,  and,  further- 
more, affected  pods  were  carefully  sought  lor  in  many  other  diseased 
fields  with  negative  results. 

Since  the  disease  is  already  so  widely  spread,  which  suggests  that  it 
must  have  had  some  common  agency  of  dissemination,  and  since  tobacco 
stems  were  incorporated  in  certain  fertilizers  as  the  source  of  potash, 
inquiry  was  directed  to  determine  the  possibility  of  the  introduction  and 


Feb.  i8, 1918  Tobacco  Wildfire  457 

dissemination  of  wildfire  by  fertilizer  materials.  Attempts  to  isolate 
Bad.  tahacum  from  diseased  leaves  which  had  passed  through  the  curing 
process  gave  negative  results  in  the  case  of  three  samples  tested.  In 
view  of  the  fact  that  no  growth  occurred  in  bouillon  cultures  exposed  to 
temperatures  above  65°  C.  for  10  minutes,  as  has  previously  been  reported, 
it  is  highly  improbable  that  the  organism  could  survive  for  several  hours 
temperatures  of  180°  F.  and  above,  as  are  maintained  for  several  hours 
in  the  last  part  of  the  curing  process.  Furthermore,  in  the  preparation 
of  tobacco  stems  for  incorporation  with  fertilizer  materials  they  are  sub- 
jected to  a  sufl&cient  degree  of  heat  to  insure  complete  sterilization. 

SUMMARY 

(i)  A  leafspot  disease  of  tobacco  called  "wildfire,"  which  is  more 
destructive  than  any  other  malady  affecting  this  crop,  has  appeared 
within  North  Carolina  and  Virginia. 

(2)  It  has  been  collected  during  the  past  season  in  19  counties  within 
North  Carolina,  3  within  Virginia,  and  occurs  also  in  Wisconsin. 

(3)  Wildfire  first  attracted  attention  at  time  of  transplanting  tobacco 
and  appeared  again  in  epidemic  form  at  time  of  topping  the  crop. 

(4)  The  disease  originated  in  the  seed  bed  or  plant  bed,  but  only  nega- 
tive evidence  had  been  secured  that  infection  comes  from  the  seed. 

(5)  The  leaves  alone  are  attacked,  and  the  symptoms  are  entirely 
unlike  those  of  other  foliage  disease  of  tobacco. 

(6)  The  primary  cause  of  wildfire  has  been  found  to  be  a  wet-shining 
grayish  white,  i -flagellate  organism,  which  is  herein  described  as  Bac- 
terium tahacum.  Its  period  of  incubation  is  about  72  hours,  and  large 
lesions  are  formed  within  a  week. 

(7)  The  disease  is  of  the  necrotic  type,  involving  parenchyma  tissues. 

(8)  Moisture  is  of  prime  importance  in  the  spread  of  wildfire.  When 
rains  are  accompanied  by  wind,  dissemination  is  especially  rapid. 


458  Journal  of  Agricultural  Research  voi.  xii,  No.  ^ 

LITERATURE  CITED 
(i)  Faulwetter,  R.  C. 

1917.   DISSEMINATION    OF   THE    ANGULAR  LEAFSPOT  OP   COTTON.      In  JoUf.    AgT. 

Research,  v.  8,  no.  12,  p.  457-475,  2  fig.     Literature  cited,  p.  473-475. 

(2)  MORREY,  C.  B. 

1917.   THE  FUNDAMENTALS  OP  BACTERIOLOGY.      289    p.,    165   fig.      Philadelphia 

and  New  York. 

(3)  RiDGWAY,  C.  S. 

1916.  GRAIN  OP  THE  TOBACCO  LEAP.     In  Jour.  Agr.  Research,  v.  7,  no.  6,  p.  269- 
288,  2  fig.,  pL  15-17.     Literature  cited,  p.  287. 

(4)  Rolfs,  F.  M. 

1915.  ANGULAR  LEAF  spot  OF  COTTON.      S.  C.  Agr.  Exp.  Sta.  BuL   184,   8  fig., 

30  p.,  9  pL    Literature  cited,  p.  30. 

(5)  Saccardo,  p.  a. 

1894.  chromotaxia.    Ed.  2,  22  p.,  2  pL  Patavii. 

(6)  Wolf,  F.  A. 

1916.  citrus  canker.    In  Jour.  Agr.  Research,  v.  6,  no.  2,  p.  69-100,  8  fig., 

pL  8-11.     Literature  cited,  p.  98-99. 
(7) and  Foster,  A.  C. 

1917.  bacterial  leap  spot  of  tobacco.    In  Science,  n.  s.,  v.  46,  no.  1189, 

p.  361-362. 


PLATE  15: 

Bacterium  tabacum: 

A. — Tobacco  leaf ,  fotir  days  after  artificial  inoculation,  showing  chlorosis  and  lesions. 
B. — Natural  infection  with  brown  lesions  bordered  by  tissues  of  a  water-soaked 
appearance. 


Tobacco  Wildfire 


Plate  15 


Journal  of  Agricultural  Research 


_M.---y 


Vol. XII,  No. 7 


Tobacco  Wildfire 


PLATE  16 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  7 


PLATE  16: 
Bacterium  tabacum: 

A. — Natural  infection.     Lesions  are  large  and  concentrically  zonate. 

B. — Numerous  confluent  lesions  on  one  side  of  the  midrib  have  resulted  in  distortion 
of  the  leaf. 

C. — Almost  the  entire  leaf  is  involved  and  a  portion  of  the  rotted  tissues  have  fallen 
out.     Natiu"al  infection. 
27810°— 18 5 


GIPSY-MOTH  LARV^  AS  AGENTS  IN  THE  DISSEMINA- 
TION OF  THE  WHITE-PINE  BUSTER-RUST^ 

By  G.  Flippo  GravatT,  Assistant  Pathologist,  and  G.  B.  PosEy,  Scientific  Assistant, 
Investigations  in  Forest  Pathology,  Bureau  of  Plant  Industry,  United  States  Depart- 
ment of  Agriculture 

INTRODUCTION 

Very  little  has  been  done  to  correlate  the  widespread  distribution  of 
the  white-pine  blister- rust,  caused  by  Cronartium  rihicola  Fischer,  with 
factors  governing  the  dissemination  of  the  spores  of  the  causal  organism. 
The  early  occurrence  of  telia  on  leaves  of  currant  and  gooseberry  plants 
(Ribes  spp.)  in  localities  distant  from  known  infections  on  pines  (Pinus 
spp.),  together  with  the  absence  of  definite  knowledge  of  instances  of 
overwintering  on  the  former  hosts,  is  suggestive  of  distant  seasonal 
spread  of  the  disease  by  aeciospores  from  pines. 

Larvae  of  the  gipsy  moth  (Porthetria  dispar  L.)  feed  on  the  Peridermium 
stage  of  Cronartium  rihicola  and  carry  thousands  of .  aeciospores  on  their 
bodies.  As  Collins^  found  that  larvae  of  the  gipsy  moth  are  blown  as 
far  as  20  miles,  these  insects  are  a  potential  agent  in  distant  spread  of 
the  blister-rust.  The  gipsy  moth  is  distributed  over  a  large  portion  of 
the  white-pine  region  of  New  England. 

GIPSY-MOTH  INFESTATION  ON  DISEASED  PINE 

In  the  fall  of  191 6  a  stand  of  white  pine  covering  an  area  of  from  5  to 
7  acres  at  Kittery  Point,  Maine,  was  found  to  be  severely  infected  with 
Cronartium  rihicola.  This  growth  ranged  from  young  seedlings  to  mature 
trees  80  feet  tall  and  random  ]4-a.cre  plots  in  this  area  showed  65  to 
100  per  cent  of  the  trees  to  be  diseased.  The  number  of  infections  on 
individual  trees  ranged  from  i  to  more  than  300,  and  it  was  estimated 
that  there  were  75,000  to  100,000  separate  infections  in  trees  on  this 
area. 

In  the  infected  plot  gipsy-moth-egg  clusters  were  found  in  varying 
abundance  on  limbs  and  stems  of  pines  of  all  sizes,  and  were  located 
from  near  the  ground  to  the  tops  of  the  largest  trees.  In  a  number  of 
cases  egg  clusters  were  present  on  the  diseased  bark,  and  in  one  instance 
four  were  located  on  a  single  canker. 

«  The  writers  are  indebted  to  Mr.  A.  F.  Burgess,  of  the  0£Gce  of  Gipsy  Moth  and  Brown-tail  Moth  In- 
vestigations, Bureau  of  Entomology,  for  helpful  suggestions.  Further  work  is  being  carried  on  in  co- 
operation with  the  aboTe  office. 

'  Collins.  C.  W.  dispersion  of  gipsy  moth  larv^  by  the  wind.  U.  S.  Dept.  Agr.  Bui.  273,  23  p., 
7  pi.    1915.    Bibhography,  p.  22-23. 

METHODS  USED  IN  DETERMINING  WIND  DISPERSION  OF     THE     GIPSY     MOTH     AND     SOME    OTHER 

INSECTS.    /«  Jour.  Econ.  Ent..  v.  10.  no.  i.  p.  170-176,  2  pi.     1917- 


Journal  of  Agricultural  Research.  Vol.  XII,  No.  7 

Washington,  D.  C.  Feb.  18,  1918 

mb  Key  No.  G  135 

(459) 


460  Journal  of  Agricultural  Research  voi.  xii,  no.  7 

On  May  25,  1917,  large  numbers  of  gipsy-moth  larvae  were  found  in 
and  around  ruptured  blisters,  and  several  days  later  some  of  the  bUsters 
where  the  presence  of  larvae  had  been  previously  noted  were  empty,  and 
spore  production  was  apparently  arrested.  Subsequent  observation  on 
these  blisters  showed  that  no  further  spore  production  took  place,  while 
on  the  same  cankers  blisters  which  were  artificially  protected  from  larvae 
continued  to  produce  spores  in  abundance  until  June  25.  To  determine 
the  rate  of  blister  destruction  on  June  9  a  number  of  larvae  were  placed 
on  a  twig  infection  which  had  38  sporulating  pustules.  Many  of  the 
larvae  crawled  away  or  dropped  off,  but  a  sufficient  number  remained  to 
destroy  the  fruiting  layer  in  practically  every  blister  by  noon  of  the 
following  day,  with  the  result  that  no  subsequent  spore  production  took 
place. 

Cessation  of  spore  production  in  injured  blisters  was  caused  by  the 
destruction  of  fruiting  hyphae.  Usually  the  spores  and  hyphae  were  eaten 
away  first  and  then  the  larvae  very  often  ate  through  the  base  of  the 
fruiting  layer  to  a  depth  of  several  millimeters.  Apparently,  after  blis- 
ters no  longer  furnished  suitable  food  for  the  larvae,  they  began  feeding 
on  the  areas  of  the  yellowish,  discolored,  infected  bark  outside  the  fruit- 
ing region  and  in  some  cases  a  large  per  cent  of  the  outer  bark  of  next 
year's  sporulating  zone  was  destroyed.  Careful  observations  on  many 
larva-infested  cankers  showed  that  spore  production  was  prematurely 
arrested  in  25  to  100  per  cent  of  the  pustules,  the  percentage  usually  aver- 
aging highest  on  small  twigs. 

SPORES  ON  LARViE 

Larvae  working  in  blisters  collected  so  many  aeciospores  on  their  hairy 
bodies  that  they  appeared  nearly  the  color  of  spores  in  mass.  On  differ- 
ent dates  larvae  were  taken  from  blisters  and  placed  in  separate  capsules, 
precautions  being  taken  against  including  spores  not  on  the  bodies. 
These  were  taken  into  the  temporary  laboratory  and  spore  counts  made 
of  the  bodies  and  the  alimentary  tracts.  Spores  for  counting  were  re- 
moved from  larvae  by  washing  the  bodies  in  series  of  water,  and  alcohol 
mounts  on  slides  followed  by  final  examinations  to  assure  thoroughness 
in  the  method  used.  This  procedure  proved  quite  effective,  and,  where 
carried  through  5  to  10  washings,  practically  all  spores  were  removed 
from  the  outside  of  the  bodies.  Counts  were  made  on  the  spores  adher- 
ing to  the  inside  of  the  capsule  and  added  to  the  total  found  in  the  wash- 
ings. After  bodies  of  the  larvae  had  been  thoroughly  washed,  they  were 
dissected  and  counts  made  of  spores  in  the  alimentary  tracts.  On  May  26, 
June  4,  and  June  1 1  fifteen  small  larvae  were  collected.  Spore  counts  on 
the  bodies  of  these  15  gave  a  minimum  of  1,120,  a  maximum  of  28,320, 
and  an  average  of  18,100.  Counts  of  spores  in  the  alimentary  tract  gave 
a  minimum  of  1,740,  a  maximum  of  48,570,  and  an  average  of  26,022 


Feb.  i8,  i9i8  Gipsy-Moth  Larv(E  and  Blister-Rust  461 

To  determine  the  approximate  amount  of  spore  material  passed  through 
the  alimentary  tract,  20  larvae  were  placed  on  fresh  cankers  in  a  feeding 
tray.  After  they  were  settled  and  had  fed  normally  for  several  hours,  a 
sheet  of  paper  was  placed  under  the  cankers  for  the  collection  of  pellets. 
A  total  of  423  pellets  were  dropped  within  a  period  of  13  hours.  Counts 
of  the  spores  in  these  pellets  gave  from  3,960  to  12,450,  with  an  average 
of  8,160,  which  is  at  the  rate  of  318,616  spores  excreted  per  day  per  larv^a. 

Germination  tests  made  of  the  spores  on  the  bodies  of  larvae  collected 
on  cankers  gave  positive  results,  and  approximately  the  same  percentage 
of  germination  was  observed  as  on  spores  taken  directly  from  these 
cankers.  Germination  tests  of  spores  in  pellets  have  given  very  poor 
results;  in  only  one  case  did  several  spores  germinate.  In  many  cases 
spores  taken  directly  from  these  cankers  also  failed  to  germinate  in 
laboratory  tests. 

WIND  DISPERSION  OF  LARV^  AS  A  FACTOR  IN  BLISTER-RUST  SPREAD 

At  Kittery  Point,  Me.,  aeciospores  were  produced  from  April  29  to 
July  I,  with  maximum  spore  production  from  May  10  to  25.  Collins^ 
gives  the  hatching  period  for  gipsy  moth  larvae  in  this  section  for  191 2, 
1913,  and  1914  as  May  i  to  23,  April  29  to  May  14,  and  May  11  to  28, 
respectively.  The  season  of  191 7  was  approximately  one  week  later  than 
usual.  The  period  of  wind  dispersion  of  larv^se  is  given  as  ranging  from 
18  to  30  days,  starting  one  to  two  weeks  after  the  first  caterpillars  hatched. 
Observations  by  the  writers  showed  varjang  numbers  of  larvae  feeding  on 
blisters  from  May  25  to  June  25. 

Collins^  working  with  wind  dispersion  of  larvae  of  the  gipsy  moth  over 
a  series  of  several  years  showed  that  they  were  carried  in  wind  currents 
to  distances  as  great  as  20  miles.  The  same  author  states  that  approxi- 
mately 50  per  cent  of  the  lar\'ae  caught  at  distances  of  6  miles  or  less  had 
fed  previously. 

The  writers,  using  fly-paper  traps,  placed  10  to  30  feet  from  the  nearest 
pine  infection,  and  so  arranged  as  to  exclude  larvae  that  may  have  reached 
the  trap  by  crawling,  caught  75  small  larvae.  Four  of  these  caterpillars 
had,  respectively,  35,  105,  185,  and  2,180  aeciospores  on  their  bodies, 
which  establishes  the  fact  of  local  wind  dispersion  of  aeciospore-bearing 
gipsy-moth  larvae.  That  spores  carried  on  bodies  of  larvae  may  remain 
viable  for  a  considerable  length  of  time  is  borne  out  in  viability  tests 
under  laboratory  conditions,  wherein  aeciospores  germinated  after 
remaining  in  vials  for  a  period  of  two  months. 

Examination  of  wild  and  cultivated  species  of  Ribes  at  various  points 
throughout  Kittery  Township  showed  an  abundance  of  gipsy-moth 
larvae  feeding  on  the  foliage,  and  in  many  cases  they  were  observ^ed 
crawling  on  the  under  surfaces  of  leaves.     Quite  a  number  of  the  larva- 

'  CoiLiNS.  C.  W.    191S.    Op.  cit.  «  Collins,  C.  W.    1917.    Op.  cit. 


462  Journal  of  Agricultural  Research  voi.  xii,  no.  7 

infested  plants  showed  areas  producing  uredospores,  and  in  four  instances 
the  only  leaves  showing  blister-rust  infections  were  those  which  had  been 
injured  by  insects. 

Sixty  larvae  collected  on  species  of  Ribes  were  examined  for  aeciospores. 
Of  these  one  larva  collected  on  June  14  on  the  und^er  surface  of  a  wild 
gooseberry  plant  showed  280  aeciospores  and  520  uredospores  on  its 
body.  The  gooseberry  plant  was  heavily  infected  with  blister-rust, 
being  located  only  20  feet  from  pine  infections.  Germination  tests  of 
the  spores  from  this  larva  gave  two  germinating  aeciospores  and  many 
germinating  uredospores,  thus  bearing  out  the  fact  that  gipsy-moth 
larvae  do  carry  viable  spores  to  Rihes  spp.  and  also  showing  the  part 
which  insects  may  play  in  local  distribution  of  the  disease  by  uredospores. 

PRACTICAL  importance: 

The  facts  given  in  regard  to  the  gipsy-moth  larvae  show  that  these 
insects  are  certainly  a  factor  in  the  spread  of  the  blister-rust  locally 
from  pines  to  Rihes  spp.  Their  habit  of  feeding  and  crawling  over  the 
lower  leaf  surface,  where  the  stomata  are  located,  gives  the  spores  borne 
on  their  bodies  a  good  opportunity  for  causing  infection.  The  prob- 
ability of  the  spread  of  blister-rust  from  pines  to  distant  Rihes  spp.  is 
undoubted,  since  Collins'  work  shows  that  the  gipsy-moth  larvae  are 
blown  by  winds  of  varying  intensity  for  distances  of  20  miles.  Though 
wind  is  considered  to  be  the  most  important  factor  in  aeciospore  dissemina- 
tion, gipsy-moth  larvae  undoubtedly  play  an  important  part.  Other 
insects  have  been  collected  from  infected  pines  with  thousands  of  aecio- 
spores on  their  bodies,  but  these  insects  were  not  present  in  sufficient 
numbers  to  make  them  of  importance  in  comparison  with  the  number  of 
gipsy-moth  larvae  present. 

SUMMARY 

(i)  The  period  of  hatching  and  of  wind  dissemination  of  gipsy-moth 
larvae  came  within  the  period  of  spore  production  of  the  blister-rust  on 
pines. 

(2)  Larvae  fed  abundantly  on  spores  and  injured  the  fruiting  layer  of 
the  pustules  so  that  further  spore  production  was  arrested. 

(3)  Larvae  from  blister-rust  cankers  had  thousands  of  viable  spores  on 
their  bodies.  A  small  percentage  of  the  larvae  collected  from  fly  paper 
and  from  species  of  Ribes  near  infected  pines  showed  aeciospores  on  their 
bodies. 

(4)  Gipsy-moth  larvae  were  found  feeding  on  leaves  of  Rihes  spp., 
and  in  some  cases  the  only  infected  leaves  on  plants  of  this  genus  were 
those  showing  insect  injury. 

(5)  The  Bureau  of  Entomology  has  shown  that  these  larvae  are 
blown  by  the  wind  up  to  a  distance  of  20  miles.  Within  this  distance 
the  larvae  are  potential  agents  in  the  spread  of  the  white-pine  blister- 
rust  (within  the  area  infested  by  the  gipsy  moth). 


ADDITIONAL  COPIES 

or  THIS  PUBLICATION  MAY  BE  PROCXTRED  FEOM 

THE  SUPERINTENDENT  OF  DOCUMENTS 

GOVERNMENT  PRINTING  OFFICE 

■WASHINGTON,  D.  C. 

AT 

20  CENTS  PER  COPY 
Subscription  Price,  83.00  Per  Year 


Vol.  XII  KBBRUARY  25,  1918  No.  8 

JOURNAL  OF 

AGRICULTURAL 
RESEARCH 


CONTKNTS 


Page 


Influence  of  Carbonates  of  Magnesiimi  and  Calciiun  on 

Bacteria  of  Certain  Wisconsin  Soils    -        -        -        ►      463 

H.  L.  FULMER 

( CffifilribxiUon  trom  Wisconsin  Agricultural  Experiment  Station  ) 

Humus  in  Mulched  Basins,  Relation  of  Humus  Content  to 
Orange  Production,  and  Effect  of  Mulches  on  Orange 
Production    -        --       -        -       -        -        -        -      505 

CHARLES  A.  JENSEN 
(Ccutribtttion  from  Bureau  ol  Plant  Industry ) 

Relation  of  Kinds  and  Varieties  of  Grain  to  Hessian^pFly 

Injury  -       -        -        -       -        -        -        -        -        .      519 

JJOSES  W.  McCOLLOCH  and  S.  C.  SALMON 

( Contrlbntioa  from  Kansas  Agricultural  Experiment  Station) 


PUBUSHED  BY  ADTHOMTY  OF  THE  SECRETARY  OF  AGRICULTURE, 

WITH  THE  COOPERATION  OF  THE  ASS0CL4TI0xN  OF  AMERICAN 

AGRICULTURAL  COLLEGES  AND  EXPERIMENT  STATIONS 


WASMINGXON,  t).  C. 


wASMiNOToit :  aovenNMCNT  pmnTwa  ornce  :  i«ib 


EDITORIAL  COMMITTEE  OF  THE 

UNITED    STATES    DEPARTMENT    OF    AGRICULTURE    AND 

THE  ASSOCIATION  OF  AMERICAN  AGRICULTURAL 

COLLEGES  AND  EXPERIMENT  STATIONS 


FOR  THE  DEPARTMENT 

KARL  F.  KELLERMAN,  Chairman 

Pkyskiogist  and  A  ssociale  Chief,  Buretnt 
of  PI  an  I  Indu-stry 

EDWIN  W.- ALLEN 

Ckie.f,  Office  of  Experiment  Slaitons 

CHARLES  L.  MARLATT 

EntomoloQist  and  Assistant  Chief,  Bureau 
of  Entomology 


FOR  THE  ASSOCIATION 

RAYMOND  PEARL* 

Biologist,  Maine  Agriculfural  Experimtnt 
Station 

H.  P.  ARMSBY 

Director,  Institute  of  Animal  Nutrition,  The 
Pennsylvania  State  College 

E.  M.  FREEMAN 

Botanist,  Plant  Pathologist  and  Assistant 
Dean,  Agricttlturcl  Experiment  Station  of 
the  University  of  Minnesoia 


All  correspondence  regarding  articles  from  the  Department  of  Agriculture  should  be 
addressed  to  Karl  F.  Kellermajti,  Journal  of  Agricultural  Research,  Washington,  D.  C. 

*Dr.  Pearl  has  undertaken  special  work  in  connection  with  the  war  emergency; 
therefore,  until  further  notice  all  correspondence  regarding  articles  from  State  Experi- 
ment Stations  should  be  addressed  to  H.  P.  Armsby,  Institute  of  Animal  Nutrition, 
State  College*  Pa. 


JOMALOFAGMQUIAIISEARCH 

Vol.  XII  Washington,  D.  C,  February  25,  1918  No.    8 


INFLUENCE  OF  CARBONATES  OF  MAGNESIUM  AND 
CALCIUM  ON  BACTERIA  OF  CERTAIN  WISCONSIN 
SOILS  ^ 

By  H.  ly.  FuLMER,^ 
Assistant  Agricultural  Bacteriologist,  Wisconsin  Agricultural  Experiment  Station 

THE  PROBLEM 

In  the  course  of  a  study  of  the  biology  of  certain  acid  soils  it  was 
tound  that  magnesium  carbonate  causes  a  great  increase  in  the  reproduc- 
tion of  soil  bacteria.  Just  what  effect  this  great  increase  in  number  of 
bacteria  has  on  the  fertility  of  the  soil  is  a  problem  beyond  the  scope  of 
this  paper.  However,  it  is  a  well-established  fact  that  the  majority  of 
the  changes  of  soil  constituents  are  brought  about  by  microorganisms. 
These  substances,  which  are  constantly  undergoing  changes,  are  influ- 
enced by  the  number  and  kinds  of  bacteria. 

The  beneficial  effect  of  lime,  calcium  carbonate,  and  magnesium  car- 
bonate on  the  growth  of  higher  plants  is  generally  preceded  by  an  increase 
in  the  number  of  bacteria.  Evidence  is  sufficient  to  warrant  the  con- 
clusion that  an  increase  in  the  number  of  soil  microorganisms  is  usually 
reflected  in  a  more  rapid  decomposition  of  organic  matter  and  a  greater 
liberation  of  the  inorganic  soil  constituents,  chiefly  potassium  and  phos- 
phorous. 

Although  many  important  data  have  been  obtained  in  relation  to  the 
effect  of  liming  on  the  bacteria  in  soil,  there  still  remain  problems  which 
deserve  careful  investigation — for  example,  the  amount  and  kind  of  lime 
and  calcium  and  magnesium  carbonates  that  can  be  applied  most  eco- 
nomically and  yet  give  the  best  results. 

The  investigations  of  this  paper  were  undertaken  with  the  hope  of 
throwing  some  light  jpon  the  problem  of  soil  bacteria  and  their  response 
to  applications  of  magnesium  carbonate,  calcium  carbonate,  and  limestone. 

The  effect  of  these  substances  on  pure  cultures  of  bacteria  and  on  the 
ordinary  soil  flora  was  observed  in — 
(a)  Acid  Colby  silt  loam  soil; 
(6)  Acid  Plainfield  sand; 
(c)  Neutral  Miami  silt  loam. 

'  This  paper  is  submitted  in  partial  fulfillment  of  the  requirements  for  the  degree  of  doctor  of  philosophy 
in  bacteriology  in  th^  Graduate  School  of  the  University  of  Wisconsin,  July,  1917. 

'  The  writer  is  indebted  to  Dr.  E.  B.  Fred,  of  the  Wisconsin  Experiment  Station,  for  supervision  and 
suggestions  in  the  experimental  work  and  preparation  of  the  manuscript. 

Journal  of  Agricultural  Research,  Vol.  XII,   ^fo.  8 

Washington,  D.  C.  Feb.  as,  191S 

me  Key  No.  Wis.— ir 

(463) 


464  Journal  of  Agricultural  Research  volxii.no.s 

REVIEW  OF  LITERATURE 

A  complete  review  of  the  literature,  showing  the  effect  of  lime  (CaO).. 
calcium  carbonate  (CaCOg),  and  magnesium  carbonate  (MgCOg)  on  soil 
microorganisms,  is  not  attempted  in  this  paper.  Only  a  brief  resume  of 
certain  of  the  more  important  papers  is  given. 

Although  this  paper  is  primarily  concerned  with  the  relation  of  soil 
microorganisms  to  calcium  and  magnesium  carbonates,  it  was  deemed 
worth  while  to  include  a  brief  review  of  results  obtained  from  the  use  of 
lime  and  limestone.  Under  ordinary  soil  conditions,  calcium  oxid,  or 
lime,  is  soon  converted  into  calcium  carbonate.  Because  of  the  frequent 
occurrence  of  magnesium  in  limestone,  this  substance  was  included  in 
the  discussion. 

NUMBKR   OF   ORGANISMS 

In  1 90 1  Chester  (7)^  made  plate  counts  of  an  acid  soil  which  had  re- 
ceived lime  at  the  rate  of  i  ,000  to  4,000  pounds  per  acre.  In  every  case 
the  lime  increased  the  total  number  of  bacteria;  the  greatest  gain  was 
noted  where  the  largest  amount  of  lime  was  applied. 

Fischer  {18)  treated  soil  with  lime  and  calcium  carbonate  at  the  rate 
of  0.1  to  3.0  per  cent  by  weight  of  soil.  The  calcium  carbonate  and  lime 
were  added  in  gram-molecular  equivalents.  His  results  showed  that 
0.1  and  0.2  per  cent  of  lime  after  three  days  gave  an  increase  in  the 
total  number  of  bacteria.  Treatments  amounting  to  0.5  per  cent  and 
more  were  harmful  at  first,  but  later  gave  an  increase  beyond  that  of  the 
control.  The  data  showed  that  a  slight  increase  in  the  number  of  bac- 
teria occurred  when  calcium  carbonate  was  added  to  the  soil. 

Several  years  later,  Hutchinson  (25)  made  somewhat  similar  experi- 
ments, using  lime  and  calcium  carbonate.  He  also  found  that  lime  at 
first  exerted  a  depressing  effect  but  later  gave  rise  to  an  increase  in  the 
number  of  bacteria.  He  concluded  that  the  reduction  in  the  number  of 
bacteria  immediately  after  treatment  with  lime  was  due  to  the  caustic 
effect  of  the  oxid,  since  no  injury  was  noted  after  the  oxid  changed  to 
carbonate.  He  believed  the  benefit  from  liming  was  due  in  part  to  the 
gain  in  soluble  organic  matter,  to  the  improvement  in  the  physical  con- 
dition, and  to  the  correction  of  the  acid  reaction  of  a  soil,  all  of  which 
tended  to  bring  about  a  better  environment  for  the  development  of  bac- 
teria. However,  this  investigator  held  that  these  changes  brought 
about  by  the  action  of  liming  did  not  seem  sufficient  to  account  for  the 
enormous  increase  in  plant  growth.  He  assumed,  therefore,  that  the 
action  of  lime  was  much  the  same  as  an  antiseptic — that  is,  it  caused  a 
partial  sterilization. 

In  a  later  publication,  Hutchinson  and  MacLennan  (24)  reported  the 
results  of  experiments  with  lime  and  calcium  carbonate  on  five  different  • 
soils.     The  range  of  reaction  in  these  soils  varied  from  neutral  to  strongly 

1  Reference  is  made  by  number  (italic)  to  "  Literature  cited,"  pp.  500-504. 


Feb.  25. 1918  Influence  of  Carbonates  on  Soil  Bacteria  465 

acid.  Lime  was  added  in  amounts  varying  from  o.i  to  i.o  per  cent, 
while  calcium  carbonate  was  added  in  amounts  of  i.o  per  cent  only. 
Besides  the  total  number  of  bacteria,  these  investigators  determined  the 
nitrate  and  ammonia  nitrogen  of  the  soils  subjected  to  this  treatment. 
The  results  of  the  experiments  showed  that  in  all  of  the  soils,  liming 
increased  the  number  of  bacteria.  At  first  the  heavier  applications  of 
lime  retarded  and  later  stimulated  the  reproduction  of  microorganisms. 
The  acid  soil  required  more  lime  to  retard  the  growth  of  bacteria  than 
the  nonacid  soils.  As  a  result  of  liming,  ammonification  and  nitrifica- 
tion in  all  of  the  soils  was  increased.  The  authors  concluded  that  differ- 
ent soils  varied  in  relation  to  the  amount  of  lime  needed  to  effect  partial 
sterilization.  They  assumed  that  each  soil  absorbed  a  certain  amount 
before  the  antiseptic  action  began. 

Miller  (57)  carried  out  a  rather  intensive  study  of  the  effect  of  lime 
and  calcium  carbonate  in  both  calcium-rich  and  calcium-poor  soils,  and 
also  in  soils  low  and  high  in  organic  matter.  His  results  were  in  accord- 
ance with  those  obtained  by  many  other  investigators — namely,  that 
lime  in  small  amounts  increased  the  number  of  bacteria,  while  larger 
application,  decreased  at  first  and  later  increased  the  number.  To  this 
investigator  it  appeared  that  lime  was  a  direct  stimulant  to  the  soil 
flora.  The  same  general  increase  was  obtained  with  soil  low  and  soil 
high  in  calcium  content,  and  also  in  soils  low  and  high  in  organic  matter. 
Unlike  lime,  calcium  carbonate  brought  about  only  a  slight  increase  in 
the  number  of  bacteria. 

Beckwith,  Vass,  and  Robinson  (5)  applied  lime  at  the  rate  of  2  tons 
per  acre  to  both  acid  and  nonacid  soils.  Only  the  soils  with  an  acid 
reaction  or  where  large  amounts  of  organic  matter  were  added  showed 
an  increase  in  the  number  of  bacteria  from  the  lime  treatment.  Under 
the  same  conditions  ammonification  and  nitrification  responded  in  a 
like  manner. 

Soil  was  treated  with  lime,  magnesium  oxid,  and  magnesium  car- 
bonate by  Lemmermann  and  Fischer  (34).  They  found  that  mag- 
nesium oxid  caused  a  greater  increase  in  the  number  of  bacteria  than 
did  either  lime  or  magnesium  carbonate. 

Engberding  (13)  treated  clay  soil  with  o.i  per  cent  of  lime  and  with 
0.5  per  cent  of  magnesium  oxid.  Both  treatments  increased  the  number 
of  bacteria,  although  the  lime  seemed  to  give  better  results. 

Jenkins  and  Britton  {26)  showed  that  by  using  heavy  applications  of 
lime  with  raw-bone  meal,  the  latter  was  decomposed  more  rapidly. 

Fabricius  and  Van  Feilitzen  (14)  noted  an  increase  in  the  number  of 
bacteria  in  moor  soils  treated  with  lime. 

Lemmermann,  Fischer,  Kappen,  and  Blanck  (55)  reported  a  gain  in 
the  number  of  bacteria  in  cultivated  and  noncultivated  moor  soils  and 
in  clay  soils  when  lime  or  calcium  carbonate  was  added.  However,  an 
increase  in  the  number  of  bacteria  was  not  shown  in  a  sandy  clay  soil 
or  in  a  sandy  soil. 


466  Journal  of  Agricultural  Research  voi.xii.no.  s 

AMMONIA    AND   NITRATES 

Withers  and  Fraps  (62,  63)  added  calcium  carbonate  to  soil  treated 
with  nitrogenous  material  and  observed  a  gain  in  nitrates  which  was 
greater  in  the  calcium-carbonate  soil  than  in  the  control.  Koch  {31) 
reported  similar  results  by  applying  lime.  Fraps  (jp)  treated  acid  and 
nonacid  soils  with  calcium  and  magnesium  carbonates  and  noted  an 
increase  in  nitrates  in  all  cases,  especially  in  the  acid  soil.  However, 
calcium  carbonate  gave  better  results  than  were  obtained  with  mag- 
nesium carbonate. 

By  applying  calcium  carbonate  to  a  soil  treated  with  ammonium 
sulphate,  Lemmermann,  Blanck,  Heinitz,  and  Von  Wlodeck  {36)  no- 
ticed a  loss  in  ammonia.  Lipman,  Brown,  and  their  associates  (4,  5, 
42,  43)  studied  the  effect  of  calcium  carbonate  on  the  actitivies  of  soil 
microorganisms.  Lipman  and  Brown  {44)  showed  an  increase  in  am- 
monification  with  monocalcium  phosphate  [CaH4(P04)2]  and  dicalcium 
phosphate  [Ca2H2(P04)2],  but  a  decrease  with  tricalcium  phosphate 
[Ca3(P04)2]  when  applied  to  soil. 

Wohltmann,  Fischer,  and  Schneider  {64)  inoculated  liquid  media 
with  soil  treated  with  magnesium  oxid  and  lime  and  observed  an  increase 
in  both  ammonification  and  nitrification.  Hutchinson  {22)  observed 
that  in  liquid  cultures  magnesium  carbonate  retarded  nitrification  in 
practically  every  case,  while  calcium  carbonate  had  but  little  influence 
on  this  process. 

Vogel  {57)  inoculated  a  nitrogenous  solution  containing  calcium 
carbonate  with  a  soil  suspension  and  obtained  an  increase  in  ammonia. 
A  similar  test  for  nitrification  was  made,  but  an  increase  in  nitrates  was 
not  obtained.  Somewhat  similar  results  were  obtained  by  Paterson 
and  Scott  (55).  Hutchinson  and  McLennan  (25)  reported  that  calcium 
carbonate  treatments  caused  a  slight  increase  in  nitrification  in  soil. 

Greaves  {21),  working  with  a  Utah  soil  high  in  lime  and  magnesium 
oxid,  found  that  calcium  carbonate  in  all  concentrations  increased  the 
formation  of  ammonia,  while  magnesium  carbonate  retarded  ammonifi- 
cation except  when  applied  in  very  small  amounts.  The  chlorids  of 
magnesium  and  calcium  appeared  to  be  very  toxic.  With  a  Japanese 
soil  Machida  {48),  of  the  Japanese  Experiment  Station  at  Tokio,  noted 
that  calcium  chlorid  retarded  ammonification,  while  magnesium  chlorid 
increased  it.  He  found  that  nitrification  was  favored  more  by  the  use 
of  magnesium  carbonate  than  by  calcium  carbonate. 

The  work  of  McBeth  and  Wright  (49)  showed  that  the  chlorids,  the 
sulphates,  and  especially  the  carbonates  inhibited  nitrification.  Lyon 
and  Bizzell  (47)  reported  that  10  days  after  treatment  with  lime  the 
number  of  bacteria  was  doubled. 

Fred  and  Graul  (20)  concluded  that  the  accumulation  of  nitrates  from 
casein  or  gelatin  in  acid  soils  was  not  materially  benefited  by  calcium 
carbonate.     In  many  cases  in  acid  soil  calcium  carbonate  increased  and 


Feb.  25, 1918  Influence  of  Carbonates  on  Soil  Bacteria  467 

later  decreased  the  nitrate  content,  and  it  was  assumed  that  a  loss  of 
nitrate  nitrogen  occurred  because  of  the  increase  in  the  total  number  of 
microorganisms. 

Allen  and  Bonazzi  (j)  showed  that  the  addition  of  ground  limestone 
to  a  noncalcareous  soil  brought  about  a  more  rapid  nitrification.  Kelley 
{28)  studied  the  effects  of  calcium  and  magnesium  carbonates  on  ammoni- 
fication  and  nitrification  in  a  California  soil  high  in  basic  substances. 
His  results  showed  that  calcium  carbonate  benefited  ammonification 
slightly  and  nitrification  to  a  great  extent,  while  magnesium  carbonate 
was  toxic  to  both  processes.  This  investigator  failed  to  show  any  def- 
inite ratio  of  calcium  and  magnesium  which  favored  the  nitrifying  and 
ammonifying  power  of  this  soil.  With  Hawaiian  soils,  high  in  lime  and 
magnesium  oxid,  Kelley  (29)  obtained  similar  results.  Dolomitic  and 
calcareous  limestones  gave  results  similar  to  those  obtained  with  calcium 
carbonate. 

Kellerman  and  Robinson  (27)  obtained  nitrification  of  ammonium 
sulphate  in  a  soil  with  a  high  magnesium  content.  When  calcium  car- 
bonate was  added,  an  increase  in  nitrification  was  noted;  with  magnesium 
carbonate  a  decrease,  except  when  the  latter  substance  was  added  to  the 
soil  in  very  small  amounts.  Owen  (52)  reported  that  magnesium  car- 
bonate favored  nitrification  more  than  calcium  carbonate. 

White  (6j)  and  Voorhees  and  Lipman  {58)  treated  soil  with  pure  lime 
and  with  a  lime  containing  magnesium.  Better  nitrification  and  in  most 
cases  better  ammonification  was  obtained  from  the  magnesium  than 
from  the  nonmagnesium  limed  soil.  Soil  treated  with  magnesium  car- 
bonate and  inoculated  into  a  liquid  medium  was  shown  by  Lipman  and 
Brown  {41)  to  retard  nitrification. 

It  was  shown  by  Ehrenberg  (12),  Lemmermann,  Aso,  Fischer,  and  Fres- 
enius  {37),  and  Wheeler,  Sargent,  and  Hartwell  {60),  that  when  lime  or 
calcium  carbonate  was  applied  to  soil,  the  decomposition  of  organic 
matter  was  accelerated. 

NITROGEN  FIXATION 

Fischer  {15,  16)  demonstrated  the  fact  that  both  lime  and  magnesium 
oxid  increased  the  reproduction  of  Bacillus  azoiobacter  in  soil.  The  oxid 
of  magnesium  seemed  to  give  better  results  than  lime. 

So  much  is  Bacillus  azotohacter  influenced  by  lime  that  Christensen  and 
Larsen  (9)  suggested  the  use  of  this  organism  to  measure  the  reaction  of 
soil.  In  a  later  publication  Christensen  (8)  showed  that  the  growth  of 
this  organism  in  solution  took  place  only  when  inoculated  with  a  basic 
soil.  In  their  work  with  Danish  soils,  Weis  and  Bornebusch  (^g)  confirmed 
Christensen's  results.  Loew  (45)  showed  that  lime  added  to  a  soil  in- 
creased the  growth  of  the  film  of  B.  azotohacter  formed  in  liquid  cultures. 
Results  similar  to  Loew's  were  obtained  by  Cauda  (6)  when  calcium  car- 
bonate was  used. 


468 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  8 


Fischer  (/d)  obtained  a  better  growth  of  Bacillus  azotobacter  from  limed 
than  from  unlimed  clay  soil.  On  the  other  hand,  Koch,  Litzendorf ,  KruU, 
and  Alves  (32)  reported  that  lime  retarded  free-nitrogen  fixation.  Kriiger 
(5j)  treated  soil  with  lime  and  obtained  an  increase  in  nitrogen  fixation. 

Purer  and  better  film  growth  of  Bacillus  azotobacter  was  obtained  by 
Ashby  (2)  with  magnesium  carbonate  than  with  calcium  carbonate. 

The  effect  of  calcium  carbonate  and  magnesium  carbonate  on  the 
fixation  of  nitrogen  by  Bacillus  azotobacter,  both  in  liquid  cultures  and  in 
soil,  was  studied  by  Lipman  and  Burgess  (40).  In  every  case  magnesium 
carbonate  alone  proved  toxic.  These  authors  observed  that  when  15 
parts  of  calcium  carbonate  were  mixed  with  i  part  of  magnesium  car- 
bonate, the  latter  was  no  longer  toxic  to  this  organism. 

SUMMARY  OF  LITERATURE 

From  the  citations  just  given  it  appears  that  the  addition  of  calcium 
and  magnesium,  either  in  the  form  of  oxid  or  carbonate,  to  soil,  and 
especially  to  acid  soil,  brings  about  conditions  favorable  to  the  growth 
of  certain  groups  of  microorganisms.  There  are  many  factors  which  have 
been  given  little  or  no  consideration — for  instance,  what  relationship 
exists  between  the  total  number  of  bacteria  in  soil  and  the  quantity  of 
soil  acid  neutraHzed?  With  few  exceptions,  little  attention  has  been 
directed  toward  the  relative  effect  of  calcium  and  magnesium  carbonates 
on  the  soil  flora.  There  exists  a  diversity  of  opinion  with  regard  to  the 
relation  of  bacteria  to  these  two  compounds.  This  lack  of  harmony  may 
be  due  to  the  difference  in  the  soil  types  which  have  been  studied. 

Again,  it  seems  that  no  one  has  tried  to  measure  the  effect  of  calcium 
and  magnesium  carbonates  on  pure  cultures  of  bacteria  in  sterihzed 
acid  soil.  To  obtain  information  with  regard  to  these  points,  a  series  of 
experiments  was  planned. 

EXPERIMENTAL  WORK 

For  this  study  three  Wisconsin  soils,  acid  Colby  silt  loam,  acid  Plain- 
field  sand,  and  neutral  Miami  silt  loam,  were  used.  The  Colby  silt  loam 
was  collected  near  Marshfield,  the  Plainiield  sand  from  Hancock,  and 
the  Miami  silt  loam  from  Madison.  At  the  laboratory  each  soil  was 
passed  through  a  4-mm.  sieve  and  thoroughly  mixed. 

The  percentage  composition  of  these  soils  is  given  below : 


Constituent. 


Colby  silt 
loam. 


Miami  silt 
loam. 


Plainfield 
sand. 


Potassium 

Nitrogen 

Phosphorous. .. 
Calcium  oxid .  . 
Organic  matter 


I- 51 
.  198 
.  072 
.  0907 

3-91 


2.  16 
•IS 

•IS 

.185 

2.74 


0-93 
.09 
.032 
.  0023 

I.  41 


Feb.  as;  19x8  Influence  of  Carbonates  on  Soil  Bacteria  469 

The  calcium-carbonate  requirement  of  the  acid  soils  was  determined 
according  to  the  Truog  barium-hydroxid  method.  In  calculating  the 
amount  of  acid  in  each  soil  only  the  active  acidity  was  considered.  For 
every  100  gm.  of  Colby  silt  loam  on  the  dry  basis  1.05  gm.  of  calcium 
carbonate  were  required  to  correct  the  active  acidity,  and  for  100  gm.  of 
Plainfield  sand  0.21  gm.  Three  different  bases  were  used  to  neutralize 
the  acidity  in  these  soils — namely,  pure  precipitated  calcium  carbonate, 
pure  precipitated  magnesium  carbonate,  and  commercial  ground  lime- 
stone. The  limestone,  the  analysis  of  which  showed  53  per  cent  of  calcium 
oxid  and  43  per  cent  of  magnesium  oxid,  was  ground  to  pass  through  a 
loo-mesh  sieve. 

Aside  from  the  compounds  just  named,  monocalcium  phosphate  was 
used  in  certain  experiments.  The  phosphate  was  employed  to  find  out 
whether  or  not  the  calcium  of  this  phosphate  salt  would  serve  in  a  like 
capacity  as  that  of  calcium  carbonate.  Accordingly  the  monocalcium 
phosphate  was  added  to  the  soil  alone  and  in  various  mixtures  with  cal- 
cium carbonate. 

The  calcium  carbonate,  magnesium  carbonate,  and  limestone  were 
added  in  amounts  sufficient  to  satisfy  one-fourth,  one-half,  and  full  cal- 
cium-carbonate requirement — that  is,  to  neutralize  one-fourth,  one-half, 
and  the  total  active  acidity.  The  phosphate  was  added  in  varying 
amounts.  After  the  bases  and  phosphate  were  thoroughly  mixed  with 
the  soil,  the  latter  was  then  poured  into  earthenware  jars  and  the  moisture 
content  raised  to  one-half  saturation  with  distilled  water.  At  definite 
intervals  samples  were  drawn  and  plate  counts  made.  The  soil  of  each 
jar,  after  the  sample  had  been  drawn,  was  poured  on  sterile  paper,  thor- 
oughly mixed,  and  returned  to  the  original  jar.  In  order  to  reduce  evapo- 
ration and  to  prevent  outside  contamination,  the  jars  were  covered  with 
cheesecloth.  The  entire  series  of  jars  was  incubated  in  the  greenhouse 
at  approximately  22°  C. 

At  regular  intervals  the  effect  of  these  compounds  on  the  total  number 
of  bacteria  in  the  soil,  on  ammonification,  and  on  nitrification  was 
studied. 

INFLUENCE     OF    CALCIUM     CARBONATE,     LIMESTONE,     AND    MONOCALCIUM 
PHOSPHATE  ON  THE   NUMBER  OF   BACTERIA   IN   SOIL 

Colby  silt  loam. — Two-kgm.  portions  of  soil  were  treated  as  outlined 
in  Table  I  and  incubated  for  a  period  of  five  months.  During  this  time 
eight  plate  counts  were  made  with  Heyden-Nahrstoflf  agar.  These  plates 
were  incubated  at  27°  C.  for  10  days.  The  influence  of  monocalcium 
phosphate  on  the  number  of  bacteria  in  the  soil  was  tested  simultaneously 
with  that  of  the  carbonates.  The  data  of  this  experiment  are  given  in 
Table  I. 


470 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  8 


Table  I. — Influence  of  calcium  carbonate,  limestone,  and  monocalcium  phosphate  on  the 
number  of  bacteria  in  Colby  silt  loam 


Treatment. 


None 

One-fourth  calcium 
carbonate 

One-half  calcium 
carbonate 

Full  calcium  car- 
bonate   

One  -  fourth  lime- 
stone  

One-half  limestone. . 

Full  limestone. . . . . . 

o.s  sm.  monocalcitmi 
phosphate 

9  gm.  monocalcium 
phosphate 

0.5  gm.  monocalcium 
phosphate  +  one- 
fourth  calcium  car- 
bonate   

o.s  gm.  monocalcium 
phosphate  +  full 
calcium  carbonate . 

s  gm.  monocalcium 
phosphate  +  one- 
f  ointh  calciimi  car- 
bonate   

3  gm.  monocalcium 
phosphate  -f  full 
calcium  carbonate . 


Number  of  bacteria  in  i  gm.  of  dry  soil. 


After  I 
week. 


iS,  600, 000 
21,000,000 

20,  500,  000 

18,  100,  000 

21,000,000 
25,300,000 
22,300,000 

14, 300, 000 
22,300,000 

16,000,000 
21,300,000 

14,  300,  000 

16,  100,000 


Rela- 
tive. 


After  2 
weeks. 


100  xg, 
113 
III 
97 


120  18. 


200,000 

400,000 

000,000 

200,000 

500,000 
500,000 
100,000 

15,500,000 

21,000,000 

15,000,000 
31,700,000 

21,100,000 
20, 900, 000 


Rela- 
tive. 


100 
116 
I3S 
110 
127 

lOI 

94 
80 
109 

78 

109 

108 


After  3 
weeks. 


Rela- 
tive. 


12,800,000 
21,300,000 
14,000,000 
13,100,000 

15,000,000 

15,500,000 
20,700,000, 

18,000,000 

20, 200, 000 

18,300,000 

33,500,000 

26,200,000 
35,600,000 


204 

278 


After  8 
weeks. 


000,000 

000,000 

000,000 

000,000 

,000,000 
,900,000 
,  100,000 

,500,000 

,600,000 


17,300,000 
15,400,000 

15,300,000 

12, 500,000 


Rela- 
tive. 


100 

138 

100 

138 

123 
114 
131 

X03 

104 

132 
119 

117 

97 


After  20 
weeks. 


Rela- 
tive. 


,  700, 000 

,400,000 

,  700, 000 

,600,000 

,500,000 
,500,000] 
, 700,000 

, 400, 000 

,300,000' 


100 
»4S 

230 

203 

20t 
173 
189 

95 

134 


17,400,000       365 
i4f3°o>o<'<'       2*S 


11,600,000 


11,000,000 


173 
164 


It  will  be  seen  from  the  figures  in  Table  I  that  in  practically  every  case 
calcium  carbonate,  either  pure  or  in  the  form  of  limestone  (dolomitic), 
increased  the  growth  of  bacteria  to  a  considerable  extent.  As  compared 
with  the  untreated  soil,  the  favorable  influence  of  the  calcium  compounds 
on  the  number  of  bacteria  was  greatest  3,  8,  and  20  weeks  after  treat- 
ment. Apparently  these  compounds  of  calcium,  especially  the  carbonate, 
have  little  influence  on  the  soil  flora  for  the  first  week.  This  is  to  be 
expected,  since  calcium  carbonate  is  very  slowly  soluble. 

The  most  striking  fact  noted  from  the  results  of  this  experiment  is 
the  marked  stimulation  of  the  microorganisms  following  small  applica- 
tions of  calcium  carbonate.  Figure  i  is  a  diagram  showing  the  effect 
of  calcium  carbonate  and  limestone  on  the  total  number  of  bacteria. 
One-fourth  enough  calcium  carbonate  to  neutralize  the  entire  soil  acidity, 
with  only  one  exception,  showed  the  greatest  increase  in  the  number  of 
bacteria.  If  grouped  according  to  their  effect  on  the  number  of  soil 
organisms,  one-fourth  calcium  carbonate  gave  the  greatest  gain  in  the 
number  of  bacteria,  one-half,  the  next  greatest,  and  full,  the  least.  The 
beneficial  effect  of  calcium  carbonate  extended  over  the  entire  period  of 
five  months — that  is,  the  treated  soil  gave  a  decided  increase  in  the  total 
number  of  bacteria  as  compared  with  the  untreated  soil.  In  general, 
pure  calcium  carbonate  proved  superior  to  limestone  in  its  effect  on  the 
bacteria  of  Colby  silt  loam  soil.    This  superiority  of  calcium  carbonate 


Feb.  2s.  1918  Influence  of  Carbonates  on  Soil  Bacteria 


471 


as  compared  with  limestone  was  due  probably  to  the  difference  in  solu- 
bility of  the  two  compounds;  pure  calcium  carbonate  is  more  soluble 
than  the  dolomitic  limestone. 

The  monocalcium  phosphate  in  small  amounts  apparently  did  not  in- 
crease the  total  number  of  bacteria,  whereas  in  larger  amounts,  applied 
alone,  it  was  slightly  beneficial.  In  two  instances  a  combination  of 
calcium  carbonate  and  phosphate  showed  an  increase  in  the  total  num- 
ber of  microorganisms.  However,  in  most  cases  the  increase  was  no 
greater  than  that  obtained  with  calcium  carbonate  alone. 


I  2  i  4  5  e  7 

I  U/eeA 


/    2  3  4  5  6   7 
ZWeehs 


I    2   i  t   5   b  1 

3  U/ee/(i 


I  z  i4  5  b  r 


t  2  3  ■^  5  b  7 


FiO.  I. — Diagram  showing  the  influence  of  calcium  carbonate  and  limestone  on  the  number  of  bacteria  in 

Colby  silt  loam. 


i=-  no  treatment. 
2"=one-fotirth  cacium  carbonate. 
3=  one-half  calcium  carbonate. 
4=  full  calcium  carbonate. 


5=  one-fourth  limestone. 
6=  one-half  limestone. 
7=  full  limestone. 


A  comparison  of  the  influence  of  calcium  from  calcium  phosphate 
with  that  from  calcium  carbonate  on  the  number  of  bacteria  in  Colby 
silt  loam  is  shown  in  figure  2.  Although  there  are  many  fluctuations, 
the  results  indicate  that  calcium  phosphate  alone  or  with  calcium  car- 
bonate increases  the  number  of  bacteria  much  sooner  than  does  the 
carbonate  alone.  Here  the  maximum  gain  with  the  phosphate  was 
noted  3  weeks  after  treatment  instead  of  20  weeks,  as  in  the  case  of  cal- 
cium carbonate. 

PLAiNFiEiyD  SAND. — ^The  preceding  experiment  was  repeated  with  a 
Plainfield  sand,  a  soil  very  low  in  organic  matter.  The  results  obtained 
in  this  test  are  presented  in  Table  II. 


472 


Journal  of  AgrictUtural  Research 


Vol.  XII,  No.  8 


Tabi,E  II. — Influence  of  calcium  carbonate,  limestone,  and  monocalcium  phosphate  on  the 
number  of  bacteria  in  Plainfield  sand 


Treatment. 


Number  of  bacteria  in  i  gm.  of  dry  soil. 


After  1 
week. 


Rela- 
tive. 


After  2 
weeks. 


Rela- 
tive. 


Afters 
weeks. 


Rela- 
tive. 


Afters 
weeks. 


Rela- 
tive. 


After  2o 
weeks. 


Rela- 
tive. 


None 

One-fourth  calcium 
carbonate 

One-half  calcium  car- 
bonate  

Full  calcium  carbonate. 

One-fourth  limestone. . . 

One-half  limestone 

Full  limestone 

o.s  gm.  monocalcium 
phosphate 

a  gm.  monocalcium 
phosphate 

0.5  gm.  monocalcium 
phosphate  -f  one- 
fourth  calcium  car- 
bonate  

0.5  gm.  monocalcium 
phosphate  -f-  full 
calcium  carbonate . .  . 

a  gm.  monocalcium 
phosphate  +  one- 
fourth  calcium  car- 
bonate  

2  gm.  monocalciimi 
phosphate  +  full 
calciiun  carbonate 


10, 300, 000 

10, 000, 000 

6, 600, 000 
6, 000, 000 
5, 800, 000 
8, 000, 000 
9,  700,000 

8,  soo,  000 

S,  000, 000 

16,000,000 
12,000,000 

11,200,000 


97 


64 


3,500,000 


3,600,000 


4, 660, 000 
60-5,520,000 
56|4, 660,000 
77  3,800, 000 
94'4, 200, 000 


84 


4,200,000 
4,200,000 


100 
102 

130 

158 
130 
109 
120 

120 

120 


120 


80 


5,500,000 

6, 000, 000 

11,200,000 
18,300,000 
11,000,000 
10,000,000 
9,  700,000 

5,700,000 

8,900,000 


14,500,000 
9,400,000 


9,300,000 

150  10,300, 000 


100  4 
10916 

203  7, 
3327. 
200  8, 
180!  7, 
17617, 


700,000 

600,000 

700,000 
300, 000 
500, 000 
700, 000 
200, 000 


263 


4, 100, 000 
3,200,000 

8, 000, 000 
6, 800, 000 

6,600,000 


164 
155 
181 
164 
154 

87 

68 


170 
144 

140 

172 


2,450,000 

4, 850, 000 

4  330,000 
5, 800, 000 
4, 000, 000 
3,000,000 
4,530,000 

2, 650, 000 

2,000,000 


3,120,000 
7,770,000 

3,410,000 
3,660,000 


197 

IJ2 
236 

i6i 
12a 
184 

loS 

83 


86 
317 

139 
148 


12   3  4    5  6    7 


I  z  i  ^  s  b  t 
Z  U/eeHi 


I  Z  i  4  5  6  7 

3  WeeAs 


12  3  4  5  6  7 
6  U/eehi 


ZOU/eehs 


Flo.  a. — Diagram  showing  the  influence  of  calcium  carbonate  and  monocalcium  phosphate  on  the  number 

of  bacteria  in  Colby  silt  loam. 
i=no  treatment. 

2=0.5  gm.  monocalcium  phosphate. 

3=2  gm.  monocalcium  phosphate. 

4=0.5  gm.  monocalcitmi  phosphate  -f-one-fourth  calduza  carbonate. 

5=0.5  gm.  monocalcium  phosphate-1-fulI  calcium  carbonate. 

6="2  gm.  monocalcium  phosphate -f  one-fourth  calcium  carbonate. 

7—  a  gm.  monocalcium  phospbate-l-full  calcium  carbonate. 


Feb.  35, 1918 


Infltience  of  Carbonates  on  Soil  Bacteria 


473 


The  results  reported  in  Table  II  differ  somewhat  from  those  obtained 
with  Colby  silt  loam.  One  week  after  the  treatment  there  was  a  decrease 
in  the  number  of  bacteria  in  the  soil  treated  with  both  the  calcium  car- 
bonate and  limestone.  However,  after  the  first  week  the  treated  soil 
showed  an  increase  in  the  number  of  bacteria. 

In  contrast  with  the  results  of  the  Colby  silt-loam  experiment,  the 
Plainfield  sand,  to  which  one-fourth  calcium  carbonate  or  limestone  was 
added,  did  not  give  any  marked  gain  in  the  total  number  of  bacteria 
after  i  and  2  weeks.  After  3,  8,  and  20  weeks,  one-fourth  calcium  car- 
bonate caused  a  slight  increase  in  the  number  of  soil  organisms.  It  is 
evident  that  one-half  or  full  neutralization  of  the  soil  acids  by  the  calcium 
carbonate  was  required  to  give  the  greatest  increase  in  the  number  of 
bacteria  (fig.  3). 


12  3-^567 

I  Week 


I    Z    i  ^    5  6    7 

2  U/ee/(s 


12^^567 

3  U/eeki 


12   3-9    567 

d  UreeJts 


I    Z  3  i    5  6    7 

20  WeeAi 


Fio.  3.— Diagram  showing  the  influence  of  calcium  carbonate  and  limestone  on  the  number  of  bacteria  in 

Plainfield  sand. 


I— no  treatment. 
3=  one-fourth  calcium  carbonate. 
3— one-half  calcium  carbonate. 
4=  full  calcium  carbonate. 


S=  one-fourth  limestone. 
6=  one-half  limestone. 
7=full  limestone. 


In  general,  light  applications  of  limestone  gave  a  greater  increase  in 
the  number  of  microorganisms  than  did  calcium  carbonate.  A  difference 
in  the  nature  of  the  soil  acid  in  Colby  and  Plainfield  sand  probably  ac- 
counts for  the  difference  in  quantity  of  calcium  carbonate  required  to 
stimulate  the  reproduction  of  bacteria. 

Where  monocalcium  phosphate  was  added  alone  to  the  sandy  soil, 
practically  no  increase  in  the  number  of  bacteria  was  obtained.  A  com- 
bination of  the  phosphate  with  calcium  carbonate  apparently  did  not 
stimulate  the  multiplication  of  bacteria  any  better  than  did  calcium 
carbonate  alone  except  where  a  combination  of  0.5  gm.  of  monocalcium 


474 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  8 


phosphate  and  one-fourth  calcium  carbonate  requirement  was  used.  It 
is  surprising  that  this  combination  should  favor  the  development  of  bac- 
teria more  than  a  heavier  application  of  the  same  combination  (fig.  4). 

INFI^UENCE    OF    MAGNESIUM    CARBONATE    ON    THE    NUMBER    OP    BACTERIA 

IN   SOIIv 

Colby  silt  loam. — Since  pure  calcium  carbonate  or  dolomitic  lime- 
stone failed  to  give  a  large  increase  in  the  total  number  of  bacteria,  an 
attempt  was  made  to  determine  what  effect  magnesium  carbonate  would 
have  on  the  soil  flora.  Accordingly  an  experiment  was  planned  in  which 
pure  magnesium  carbonate  was  added  to  the  soil.     The  magnesium  car- 


12   3  4    5  6   7 

i  Weeh 


12    3   4    5    6    7 

2U/eeki 


I  2  3  4  5  6  r 
3  ureehs 


1^34567 

<3  Uree/o 


I  2  3  4  5  b   7 
20  UreeAa 


Fig.  4. — Diagram  showing  the  influence  of  calcium  carbonate  and  monocalcium  phosphate  on  the  number 

of  bacteria  in  Plainfield  sand. 
x«>no  treatment. 

9=0.5  gm.  monocalcium  phosphate. 
3=  1  gm.  monocalcium  phosphate. 

4=>o.s  gm.  monocalcium  phosphate+one-fourth  calcium  carbonate. 
5—0.5  gm.  monocalcium  phosphate+full  calcium  carbonate. 
6=  a  gm.  monocalcium  phosphate+one-fourth  calcium  carbonate. 
fee 2  gm.  monocalcium  phosphate+full  calcium  carbonate. 


bonate  was  applied  to  the  soil  in  the  gram-molecular  equivalent  of  cal- 
cium carbonate.  The  procedure  in  this  experiment  was  similar  to  that 
just  described.     The  results  of  this  test  are  presented  in  Table  III. 

TablS  III. — Influence  of  magnesium  carbonate  on  the  number  of  bacteria  in  Colby  silt 

loam 


Treatment. 


None , 

One-fourth  mag- 
nesiiun  carbon- 
ate   

One-half  magne- 
sium carbonate. 

Full  magnesium 
carbonate 


Number  of  bacteria  in  i  gm.  of  dry  soil. 


After 
I  week. 


Rela- 
tive. 


25, 200, 00c 

36, 600, 000 
44, 800, 000 
156,000,000 


14s 
178 
61S 


After 
2  weeks. 


21,000,000 

29, 700, 000 
41,000,000 
125,000,000 


Rela- 
tive. 


After 
3  weeks. 


19,300,000 

30, 500, 000 
45, 200,000 
74, 500, 000 


Rela- 
tive. 


157 
234 
386 


After 
8  weeks. 


Rela- 
tive. 


14, 000, 000 

17. 200.000 

19.400.0001  138 
59,100,000       422 


After 
20  weeks. 


8, 600, 000 

II,  700,000 
12,200,000 
26,000,000 


Rela- 
tive. 


136 
141 
30a 


Feb.  25,  191S 


lyifluence  of  Carbonates  on  Soil  Bacteria 


475 


Unlike  calcium  carbonate  and  limestone,  full  applications  of  magne- 
sium carbonate  increased  the  number  of  bacteria  far  beyond  the  increase 
obtained  with  the  one-half  and  one-fourth  treatments.  From  the 
figures  in  this  table  it  will  be  seen  that  the  effect  of  the  application  of 
magnesium  carbonate  to  Colby  soil  invariably  increased  the  reproduction 
of  the  soil  bacteria,  especially  during  the  first,  second,  and  third  week. 


i    Z   d  4 
iUTeek 


/   £  3  4 
ZU/eehs 


I  Z  3  4 

ZWeeh 


6U/eeh 


I  z  3  4 


"PJO,  5. — ^Diagram  showing  the  influence  of  magnesium  carbonate  on  the  number  of  bacteria  in  Colby  silt 

loam. 


i=no  treatment. 

a = one-fourth  magnesium  carbonate. 


3  =  one-half  magnesium  carbonate. 
4=  full  magnesium  carbonate. 


Figure  5  shows  very  conclusively  the  marked  effect  of  magnesium  car- 
bonate on  the  number  of  soil  bacteria.  For  instance,  the  increase  with 
full  magnesium-carbonate  treatment  was  more  than  six  times  as  great 
as  that  of  the  control.  The  absolute  numbers  varied  with  the  time  of 
the  count,  but  the  ratio  of  numbers  between  the  different  quantities  of 
magnesium  carbonate  remained  almost  the  same  throughout  the  entire 


476 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  8 


period  of  20  weeks.  The  increase  was  maintained  for  a  period  of  20 
weeks,  but  the  total  number  of  organisms  decreased  greatly  during  this 
time. 

Plainfield  sand. — It  was  next  arranged  to  determine  the  effect  of 
magnesium  carbonate  on  the  number  of  microorganisms  in  Plainfield 
sand.     The  results  are  summarized  in  Table  IV. 

Here  again,  magnesium  carbonate  caused  a  striking  increase  in  the 
number  of  soil  microorganisms.  The  results  are  very  similar  to  those 
obtained  with  magnesium  carbonate  in  Colby  soil.  An  enormous 
increase  in  the  number  of  soil  organisms  was  noted  after  i  and  3  weeks. 
The  sudden  drop  in  numbers  after  the  3-week  period  was  due  probably 
to  a  mistake  in  weighing  which  resulted  in  a  low  moisture  content. 
After  the  8-  and  20-week  periods,  the  gradual  decrease  in  the  number  of 
bacteria  was  observed,  which  agrees  with  the  results  obtained  in  Colby 
silt  loam  (fig.  6). 

Table  IV. — Influence  of  magnesium  carbonate  on  the  number  of  bacteria  in  Plainfield 

sand 


Treatment. 


Number  of  bacteria  in  i  gm.  of  dry  soil. 


After  I 
week. 


Rela- 
tive. 


A  f ter  2 
weeks. 


Rela- 
tive. 


Afters 
weeks. 


Rela- 
tive. 


After  8 
weeks. 


Rela- 
tive. 


After  20 
weeks. 


Rela- 
tive. 


None 

One-fourth  mag- 
nesium carbonate . . 

One-half  magnesium 
carbonate 

Full  magnesium  car- 
bonate  


10, 300, 000 
21, 100,000 
13, 500,000 
73,300,000 


100 
203 
131 
7" 


3,500,000 
S,  100,000 
5,300,000 
8,100,000 


100  5,500,000 
145  10,400,000 
15119, 300, 000 
23o\sS,  300,000 


100 

190 

350 

1,060 


4, 700,000 
5,800,000 
10,300,006 
16,000,000 


100 
123 
219 
340 


2,450,000 
3,560,000 
5,620,000 
6,530,000 


100 
I4S 
228 
266 


The  results  secured  with  these  two  soil  types  show  clearly  that  mag- 
nesium carbonate  in  certain  soils  is  a  potent  factor  in  the  reproduction 
of  soil  bacteria. 


iNIfLUENCE   OF   A  MIXTURE   OF   CALCIUM    AND  MAGNESIUM  CARBONATE  ON 
THE  NUMBER  OF   BACTERIA   IN    SOIL 

In  every  case  magnesium  carbonate  gave  a  much  greater  increase  in 
the  number  of  bacteria  than  did  calcium  carbonate.  Therefore  the 
question  which  suggests  itself  is,  What  effect  will  a  mixture  of  calcium 
and  magnesium  carbonates  have  on  the  soil  flora? 

It  was  shown  by  many  investigators,  principally  Loew  (46)  and  his 
associates,  that  a  soil  should  contain  calcium  and  magnesium  in  a  certain 
ratio  in  order  to  secure  the  best  plant  growth.  On  the  other  hand, 
Lipman  (38)  made  the  following  statement : 

In  their  behavior  toward  salts,  bacteria  differ  in  some  respects  from  both  plants 
and  animals  and  occupy  a  position  by  themselves. 


Feb.  as.  1918 


Infltcence  of  Carbonates  on  SoU  Bacteria 


477 


From  his  study  of  B.  subtilis,  he  concluded  that  no  antagonism  exists 
between  calcium  and  magnesium.  In  a  later  publication  Lipman  (39) 
gave  an  extensive  review  of  the  literature  bearing  on  the  subject  of  the 


/    2   3  -f 

;   2  3  4 

12    3  4 

I   z  5  A 

i  2  3  4 

\U/eek 

2  Wee  hi 

3  Wee  As 

a  U/eeAs 

^0  Wee/is 

Fig.  6. — Diagram  showing  the  influence  of  magnesium  carbonate  on  the  number  of  bacteria  in  Plainfield 

sand. 


i=no  treatment. 

2=  one-fourth  magnesium  carbonate. 


3-=  one-half  magnesium  carbonate. 
4=  full  magnesium  carbonate. 


proper  lime-magnesia  ratio  in  soils.  He  concluded  that  there  is  little 
evidence  to  support  this  hypothesis  of  the  lime-magnesia  ratio.  In  accord 
with  Lipman's  results,  certain  investigators  {28 ^  29)  showed  that  various 


478 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  8 


mixtures  of  calcium  and  magnesium  carbonates  are  not  favorable  to  the 
development  of  soil  bacteria. 

To  study  the  effect  of  a  mixture  of  calcium  and  magnesium  carbonates 
on  the  total  number  of  bacteria  in  both  Colby  silt  loam  and  Plainfield 
sand,  a  series  of  tests  was  made.  This  mixture  was  employed  in  amounts 
equivalent  to  the  neutralizing  power  of  calcium  carbonate — that  is, 
enough  of  the  mixture  was  added  to  neutralize  one-fourth,  one-half,  and 
all  of  the  soil  acids.  Each  mixture  was  made  by  adding  che  carbonates 
in  gram-molecular  equivalent  amounts.  The  results  are  presented  in 
Table  V. 

Table  V. — Influence  of  a  mixture  of  calcium  and  m,agnesiiim,  carbonates  on  the  number 
of  bacteria  in  Colby  silt  loam  and  Plainfield  sand 

COLBY   SILT   LOAM 


Number  of  bacteria  in  i  gm.  of  dry  soil. 

Treatment. 

After 

Rela- 

After 

Rela- 

After 

Rela- 

After 

Rela- 

After 

Rela- 

I week. 

tive. 

2  weeks. 

tive. 

3  weeks. 

tive. 

8  weeks. 

tive. 

20  weeks. 

tive. 

ig, 200, ooo 

100 

27,600,000 

100 

27,300,000 

100 

8,  700, 000 

100 

One-fourth: 
One-half    caldumj 

carbonate 1 

One-half    magne-j 

slum  carbonate.  J 
One-half: 
One-half    calcium] 

36, 500, 000 

190 

29, 000, 000 

105 

27,500,000 

100 

20, 700,000 

I4S 

15,100,000 

173 

carbonate 1 

One-half    magne-| 

slum  carbonate.  J 
Full: 
One-half   calciumi 

14, 700,000 

76 

32,500,000 

117 

38, 600, 000 

104 

17,000,000 

120 

12, 200,000 

140 

carbonate 1 

One-half    magne-| 
slum  carbonate.  J 

31,800,000 

165 

4S, 300, 000 

163 

44,500,000 

162 

20, 000, 000 

140 

I3j300,ooo 

153 

PLAINFIELD   SAND 


10,300,000 

100 

3,500,000 

100 

S, 000, 000 

100 

4,  700, 000 

100 

2,450,000 

100 

One-fourth: 
One-half   calciumi 

carbonate 1 

One-half    magne-| 

slum  carbonate.  J 
One-half: 
One-half   calciumi 

11,300,000 

109 

3,200,000 

91 

11,700,000 

234 

5,500,000 

117 

4,250,000 

173 

carbonate 1 

One-half    magne-| 
sivun  carbonate.  J 
Full: 
One-half   calciumi 

11,800,000 

"S 

3,800,000 

109 

I3>300,ooo 

266 

8, 400, 000 

180 

3,500,000 

143 

carbonate 1 

One-half     magne-| 
sium  carbonate.  J 

13',  100, 000 

127 

5,000,000 

143  18,400, 000 

370 

8, 600, 000 

184 

2,630,000 

107 

It  will  be  seen  from  the  figures  of  the  table  that  a  mixture  of  calcium 
and  magnesium  carbonates  increased  the  number  of  bacteria  in  CoBby 
silt  loam  and  Plainfield  sand.  The  most  interesting  fact  ascertained 
from  this  test  is  that  magnesium  carbonate  plus  calcium  carbonate  is 
less  efficient  in  its  efifect  on  the  reproduction  of  bacteria  than  the  equiva- 
lent weight  of  magnesium  alone. 


Feb.  35, 1918  Influence  of  Carbonates  on  Soil  Bacteria 


479 


INFLUENCE  OF  CALCIUM  CARBONATE,  CALCIUM  CHLORID,  MAGNESIUM 
CARBONATE,  MAGNESIUM  CHLORID,  DIBASIC  MAGNESIUM  PHOSPHATE, 
AND   MONOCALCIUM    PHOSPHATE   ON  THE  BACTERIA   IN   SOIL 

Colby  silt  loam. — Since  the  increase  in  number  of  bacteria  in  soil 
treated  with  magnesium  carbonate  seemed  far  too  great  to  be  accounted 
for  by  the  correction  of  the  soil  acidity  alone,  an  experiment  was  made 
to  study  the  efifect  of  a  neutral  salt  of  magnesium,  and  a  magnesium 
phosphate,  on  the  number  of  bacteria  in  Colby  silt  loam  soil.  Two 
points  were  considered  in  planning  this  test :  First,  the  action  of  the  mag- 
nesium and  calcium  ions  on  the  bacteria;  and  second,  the  possibility 
of  the  combining  of  the  calcium  or  magnesium,  especially  the  latter,  with 
the  phosphate  of  the  soil,  thus  liberating  the  phosphate  in  a  more  avail- 
able form. 

It  was  reported  by  Truog  (56)  that  magnesium  phosphate  favors  the 
phosphorus  assimilation  by  plants  more  than  any  other  phosphate.  If 
this  be  true,  then  the  bacteria  (lower  plant  life)  should  be  favored  by  a 
phosphate  when  in  this  form.  Therefore  the  question  arises,  Does  the 
magnesium  carbonate  when  added  to  soil  react  with  the  phosphates  to 
form  magnesium  phosphate? 

In  order  to  make  these  tests,  magnesium  carbonate,  magnesium  chlo- 
rid,  dibasic  magnesium  phosphate,  also  calcium  carbonate,  calcium  chlo- 
rid,  and  monocalcium  phosphate  were  added  to  the  soil  alone  and  in 
various  mixtures. 


TabIvE  VI. — Influence  of  calcium  carbonate,  calcium  chlorid,  magnesium  carbonate, 
magnesium  chlorid,  dibasic  tnagnesium  phosphate,  and  m.onocalcium  phosphate  on  the 
bacteria  of  Colby  silt  loam 


Treatment. 


Number  of  bacteria  in  i  gm .  of  dry  soil. 


After  I 
week. 


Rela- 
tive. 


After  2 

Rela- 

Afters 

Rela- 

After 8 

weeks. 

tive. 

weeks. 

tive. 

weeks. 

19, 000, 000 

100 

23,200,000 

100 

12,200,000 

22, 100,000 

116 

14,000,000 

60 

17,700,000 

17,600,000 

92 

27,500,000 

118 

10,500,000 

37,600,000 

197 

43,300,000 

186 

26, 200,000 

121, 100,000 

637 

200,000,000 

862 

58, 700, 000 

13, 100,000 

67 

14,000,000 

60 

15,600,000 

9,300,000 

47 

24,000,000 

103 

11,600,000 

66, 100,  coo 

347 

33,100,000 

141 

26,000,000 

47,300>o°o 

249 

46,200,000 

199 

31,500,000 

190,000,000 

1,000 

106, 000, 000 

456 

85,500,000 

144,000,000 

75S 

197,000,000 

849 

78,000,000 

Rela- 
tive. 


None 

0.1  per  cent  magnesitun  chlo- 
rid  

0.1  per  cent  calcium  chlorid. . . 

Full  calcium  carbonate 

Full  magnesium  carbonate . . . 

0.1  per  cent  dibasic  magne- 
sium phosphate 

0.1  per  cent  monocalcium 
phosphate 

0.1  per  cent  dibasic  magne- 
sium phosphate-f-fuU  cal- 
cium carbonate 

0.1    per    cent    monocalcium 
phosphate-Ffull     calcium 
carbonate 

0.1  per  cent  dibasic  magne- 
sium phosphate-l-full  mag- 
nesium carbonate 

O.I  per  cent  monocalcium 
phosphate  +  full  magne- 
sium carbonate 


28,000,000 

16,000,000 

18, 200,000 
S9, 100,  oco 
165,000,000 

30, 500, 000 

38,000,000 


74, 100, 000 
38,000,000 
191,000,000 
223,000,000 


57 

65 

31S 

590 


264 
I3S 
6S2 
800 


I4S 

86 

314 

481 


a  13 

as8 
700 
639 


The  effect  of  magnesium  phosphate  on  the  number  of  bacteria  was  com- 
pared with  that  of  magnesium  carbonate,  calcium  phosphate,  and  cal- 
cium carbonate.     If  the  action  of  the  magnesium  and  calcium  carbonates 
27811°— 18 2 


48o 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  8 


on  the  bacteria  is  derived  from  the  basic  part  of  these  compounds,  then 
the  salts  of  these  substances — namely,  magnesium  and  calcium  chlorid — 
should  increase  the  number  of  bacteria  in  a  similar  manner.  These  salts 
were  added  in  quantities  equivalent  to  o.i  per  cent  of  the  dry  weight  of 
the  soil.     The  data  for  this  experiment  are  presented  in  Table  VI. 

The  chlorids  of  calcium  and  magnesium  failed  to  increase  the  number 
of  bacteria  in  Colby  silt  loam.  Instead,  the  majority  of  cases  showed  a 
loss,  especially  with  calcium  chlorid.  Calcium  and  magnesium  carbo- 
nates, in  agreement  with  a  previous  study,  increased  the  number  of 
microorganisms.  Magnesium  carbonate  stimulated  the  development  of 
the  bacteria  far  greater  than  did  calcium  carbonate. 

When  the  dibasic  magnesium  phosphate  and  monocalcium  phosphate 
were  applied  alone  there  did  not  seem  to  be  any  gain  in  the  number  of 
bacteria.  The  mixtures  of  monocalcium  phosphate  and  dibasic  mag- 
nesium phosphate  with  magnesium  carbonate  resulted  in  all  except  one 
period  in  an  increase  in  the  number  of  bacteria  beyond  that  obtained  by 
the  use  of  magnesium  carbonate  alone.  The  mixtures  made  by  adding 
these  same  phosphates  with  calcium  carbonate  did  not  result  in  an 
increase  in  the  number  of  bacteria  beyond  that  noted  where  calcium  car- 
bonate was  used  alone. 

From  the  data  of  this  experiment,  which  are  diagramed  in  figures  7 
and  8,  it  seems  safe  to  conclude  that  neither  the  phosphates  nor  the 
chlorids  of  magnesium  and  calcium  alone  caused  a  marked  multiplica- 
tion of  the  bacteria  in  Colby  silt  loam. 

Miami  silt  loam. — A  similar  test  of  these  different  compounds  was 
made  with  Miami  silt  loam,  a  neutral  soil.  The  chlorids  of  magnesium 
and  calcium  were  eliminated  in  this  experiment.  Since  the  effect  of  a 
carbonate  on  the  number  of  bacteria  in  Miami  silt  loam  soil  was  not 
known,  the  carbonates  were  added  in  amounts  equal  to  the  one-half  and 
full  calcium  carbonate  requirement  of  Colby  silt  loam.  The  data  for 
this  experiment  are  presented  in  Table  VII. 

Table  VII. — Influence  of  calcium  carbonate,  magneshim  carbonate,  dibasic  magnesium 
phosphate,  and  monocalcium  phosphate  on  the  number  of  bacteria  in  Miami  silt  loam 


Treatment. 


None 

One-half  calcium  carbonate 

Full  calcium  carbonate 

One-half  magnesium  carbonate 

Full  magnesium  carbonate 

0.1  per  cent  monocalcium  phosphate 

0.1  per  cent  dibasic  magnesium  phosphate.  . 
0.1  per  cent  monocalcium  phosphate -h  full 

calcium  carbonate 

0.1  per  cent  dibasic  magnesium  phosphate 

-t-full  calcium  carbonate 

0.1  per  cent  monocalcium  phosphate  +  full 

magnesium  carbonate 

0.1   per  cent  dibasic  magnesium  phosphate 

+full  magnesium  carbonate 


Number  of  bacteria  in  i  gm.  of  dry  soil. 


After  1  week, 


10, 000, 000 
10,300,000 
13,800,000 
17,000,000 
27,000,000 
10,600,000 
15,  700,000 

12,400,000 

II,  100,000 

20, 000, 000 

21,600,000 


Rela- 


100 
103 
13S 
170 
270 
106 
157 


200 
216 


After  2 
weeks. 


18,500,000 
18,500,000 
23 ,  700, 000 
25,400,000 
52,400,000 
20. 400, 000 
27,000,000 

24, 000, 000 

20, 000, 000 

35,000,000 

43 ,  Soo,  000 


Rela- 
tive. 


100 
100 
128 
137 
283 
no 
146 

130 

108 

1S9 

236 


After  3        Rela- 
weeks.        tive. 


9, 600, 000 
13,800,000 
15, 100,000 
15, 500,000 
21,600,000 
12,300,000 
14,300,000 

13 ,  500, 000 

14, 600, 000 

19, 000, 000 

II,  700,000 


100 
143 
IS7 
161 
22s 
12S 
ISO 

140 

isa 

198 


Feb.  25. 1918  Influence  of  Carbonates  on  Soil  Bacteria 


481 


i    2  5  4  5 

t  U/eeA 


/  i  3  4  5 


12    5  4    5 

3  Wee  As 


i  Z  i  4  5 
6  U/eeA& 


Fig.  7.— Diagram  showing  the  influence  of  the  carbonates  and  chlorids  of  magnesium  and  calcium  on  the 
number  of  bacteria  in  Colby  silt  loam  soil. 


i=no  treatment. 

2=  O.I  per  cent  magnesium  chlorid. 

3=0.1  per  cent  calcium  chlorid. 


4=full  calcium  carbonate. 
S=full  magnesium  carbonate. 


482 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  8 


12  3^667 


12  5^367 


I   2  3  t  5  6   7 


I  2  5  -i   5  6   7 
eOTee/ci 


FiQ.  8.— Diagram  showing  the  influence  of  dibasic  magnesitun  phosphate,  monocalcium  phosphate,  calcium 
carbonate,  and  magnesium  carbonate  on  the  number  of  bacteria  in  Colby  silt  loam. 


1=  no  treatment. 

a=o.i  per  cent  dibasic  magnesium  phosphate. 

3=0.1  per  cent  monocalcium  phosphate. 

4=0.1  per  cent  dibasic  magnesium  phosphate+ 

full  calcium  carbonate. 
S=o.i    per  cent  monocalcium    phosphate+fuU 

calcium  carbonate. 


6=o.i  per  cent  dibasic  magnesium  phosphate+fuU 
magnesium  carbonate . 

7=o.i  per  cent  monocalcium  phosphate+full  mag- 
nesium carbonate. 


Feb.  25, 1918  Influence  of  Carbonates  on  Soil  Bacteria 


483 


From  the  results  which  are  given  in  Table  VII  and  from  figure  9  it 
will  be  seen  that  both  magnesium  and  calcium  carbonate  increased  the 
number  of  bacteria  in  Miami  silt  loam  soil.  Here,  again,  the  increase 
was  much  greater  with  magnesium  carbonate  than  with  calcium  carbo- 
nate. The  greatest  increase  was  obtained  two  weeks  after  treatment. 
In  all  cases,  the  heavier  applications  increased  the  number  of  bacteria 
more  than  did  the  smaller  ones. 


^  3  4  5  6  7 
tu/ceh 


Z  i  ^  5  6    7  Q   9  /O  II 

2  U/eeAi 


2  i  4  5  b  7  &  1  10  ti 
3  U/iseAs 


Fig.  9. — Diagram  showing  the  influence  of  calcium  carbonate,  magnesium  carbonate,  dibasic  magnesium 
phosphate,  and  monocalcium  phosphate  on  the  number  of  bacteria  in  Miami  silt  loam. 

1= no  treatment. 

2=  one-half  calcium  carbonate. 

3= full  calcium  carbonate. 

4=  one-half  magnesium  carbonate. 

S=full  magnesimn  carbonate. 

6=0.1  per  cent  monocalcium  phosphate. 

7=0.1  per  cent  dibasic  magnesium  phosphate. 

8=0.1  per  cent  monocalcimn  phosphate+fuU  calcium  carbonate. 

9=0.1  per  cent  dibasic  magnesium  phosphate-hfuU  calcium  carbonate. 
10=  O.I  per  cent  monocalcium  phosphate  -l-full  magnesium  carbonate. 
11=0.1  per  cent  dibasic  magnesimn  phosphate+full  magnesium  carbonate. 

The  results  differentiate  themselves  from  those  obtained  with  acid 
Colby  silt  loam  soil  since  the  phosphates  increased  the  number  of  bacteria 
in  the  neutral  Miami  silt  loam.  It  was  expected  that  these  phosphates 
would  have  a  more  beneficial  effect  in  neutral  than  in  acid  soil,  since  it 
is  probable  that  the  phosphate  was  hydrolyzed,  thus  adding  more  acid 
to  the  soil.  In  every  test  magnesium  phosphate  gave  a  larger  increase 
in  number  of  bacteria  than  was  obtained  by  the  use  of  calcium  phos- 


484  Journal  of  Agricultural  Research  voi.  xii.no.  s 

phate.  The  dibasic  magnesium  phosphate  increased  the  number  of 
bacteria  to  about  the  same  degree  as  an  application  of  one-half  magne- 
sium carbonate.  A  combination  of  each  phosphate  with  each  carbonate 
did  not  prove  beneficial  in  augmenting  the  number  of  bacteria  in  Miami 
silt  loam  beyond  that  caused  by  each  carbonate  when  used  alone.  It  is 
evident  from  the  foregoing  data,  as  shown  in  figure  9,  that  the  relative 
increase  of  the  number  of  bacteria  from  the  use  of  carbonates  of  calcium 
and  magnesium  in  Miami  silt  loam  was  not  so  great  as  when  Colby  silt 
loam  was  treated  with  the  same  compounds.  However,  the  phosphate 
gave  a  greater  relative  increase  in  the  neutral  Miami  silt  loam  than  in 
the  acid  Colby  silt-loam  soil. 

INFlvUENCS  0]P  MAGNESIUM  CARBONATE,  CAI^CIUM  CARBONATE,  LIMESTONE, 
AND   MONOCALCIUM    PHOSPHATE   ON   THE   BACTERIA   IN    SOIL, 

Another  series  of  jars  was  filled  with  Colby  silt  loam  treated  with 
magnesium  carbonate;  a  second  series  with  the  same  soil  treated  with 
calcium  carbonate;  a  third  series  with  limestone,  and  a  fourth  with 
monocalcium  phosphate.  All  jars  were  placed  in  the  greenhouse  and 
the  soil  moisture  maintained  at  half  saturation.  After  one,  two,  and 
three  months  samples  of  this  soil  were  drawn  and  tested  for  ammonifying 
power,  for  nitrate  content,  and  for  the  number  of  bacteria. 

The  ammonifying  power  was  determined  by  mixing  2  per  cent  of 
dried  blood,  which  contained  13.4  per  cent  of  nitrogen,  with  100  gm.  of 
soil.  After  adding  the  dried  blood  meal,  the  soil  was  placed  in  tumblers, 
the  proper  amount  of  water  added,  the  tumblers  covered  with  petri-dish 
covers,  and  incubated  at  27°  C.  for  six  days.  The  ammonia  was  deter- 
mined by  the  steam-distillation  method. 

In  order  to  measure  the  nitrification  in  the  treated  and  untreated  soil, 
samples  were  taken  from  the  jars  and  the  nitrate  content  determined 
immediately.  This  was  simply  a  study  of  nitrate  accumulation  in 
the  soil;  no  nitrogenous  substance  was  added.  Nitrates  were  deter- 
mined by  the  phenol-disulphonic  acid  method.  The  data  for  the  experi- 
ments with  Colby  silt  loam  and  Plainfield  sand  are  shown  in  Tables 
VIII  to  XI,  inclusive. 

From  the  data  in  Table  VIII  it  will  be  seen  that  after  one  month  the 
increase  in  the  number  of  bacteria  in  Colby  soil,  in  conformity  with 
previous  tests,  was  greatest  with  the  magnesium-carbonate  treatment. 
The  increase  was  very  marked,  about  five  times  greater  than  that  de- 
rived from  the  use  of  calcium  carbonate  or  limestone.  The  favorable 
effect  of  the  magnesium  carbonate  was  noted  after  both  the  2-  and 
3-month  periods.  In  most  cases  the  calcium  carbonate  and  limestone 
increased  the  number  of  bacteria,  notably  after  two  months.  The 
monocalcium  phosphate  apparently  exerted  no  effect  in  increasing  the 


Feb.  as,  1918 


Influence  of  Carbonates  on  Soil  Bacteria 


485 


number  of  bacteria,  since  it  gave  no  increase  except  when  used  in  combi- 
nation with  calcium  carbonate,  and  then  the  increase  was  no  greater 
than  that  obtained  with  calcium  carbonate  alone. 

TablB  VIII. — Influence  of  calcium  carbonate,  magnesium  carbonate,    limestone,   and 
monocakium  phosphate  on  the  number  of  bacteria  in  Colby  silt  loam 


Treatment. 


Number  of  bacteria  in  i  gm.  of  dry  soil. 


After 

Rela- 

I month. 

tive. 

14, 000, 000 

100 

14,370,000 

102 

12,  200,000 

87 

14,370,000 

102 

12,200,000 

«7 

14,370,000 

102 

13.340,000 

95 

15, 180,000 

108 

26,000,000 

185 

101,400,000 

724 

15,000,000 

107 

18,100,000 

130 

11,380,000 

80 

After 
2  months. 


Rela- 
tive. 


After 
3  months. 


Rela- 
tive. 


None 

One-fourth  calcium  carbonate 

One-half  calcium  carbonate 

Full  calcium  carbonate 

One-fourth  limestone 

One-half  limestone 

Full  limestone 

One-fourth  magnesium  carbonate 

One-half  magnesium  carbonate 

Full  magnesium  carbonate 

0.1  per  cent  monocalcium  phosphate 

0.1  per  cent  monocalcium  phosphate  -1-  one- 
fourth  calcium  carbonate 

0.1  per  cent  monocalcium  phosphate  -1-  full 
calcium  carbonate 


10, 700, 000 
13,400,000 
17,000,000 
14, 800, 000 
17,600,000 
II, 200,000 
14, 300, 000 
11,900,000 
25, 700,000 
43  >  700, 000 
10, 700, 000 

10,800,000 

13,600,000 


100 

158 
138 
164 
104 
133 
III 
240 


lOI 

127 


8, 000, 000 

8,800,000 

7, 100,000 

8,  700,000 

10, 300, 000 

8, 700, 000 

10, 700,000 

10, 900, 000 

13, 200,000 

33,500,000 

8,300,000 

7, 600, 000 

10.300,000 


100 
no 
83 
108 
128 
loS 
133 
136 
i6s 
418 
103 

95 

128 


The  data  in  Table  IX  are  in  agreement  with  those  obtained  with  Colby 
silt  loam — that  is,  magnesium  carbonate  increased  the  number  of  bac- 
teria in  Plainfield  sand  to  a  considerable  extent  beyond  that  produced 
by  calcium  carbonate  or  Umestone.  The  effect  of  the  magnesium  car- 
bonate on  the  number  of  bacteria  was  most  noticeable  one  month  after 
the  treatment  was  applied.  In  general,  limestone  proved  inferior  to 
calcium  carbonate  in  stimulating  the  number  of  bacteria  in  Plainfield 
sand,  except  that  in  one  case  the  monocalcium  phosphate  when  applied 
alone  did  not  increase  the  number  of  bacteria. 

Table  IX. — Influence   of  calcium  carbonate,   magnesium  carbonate,    limestone,    and 
monocalcium  phosphate  on  the  number  of  bacteria  in  Plainfield  sand 


Treatment. 


Number  of  bacteria  in  i  gm.  of  dry  soil. 


After 

Rela- 

I month. 

tive. 

6, 800, 000 

100 

6,  220,  000 

91 

6, 100,000 

90 

8,  770,000 

128 

S,  600, 000 

83 

9, 180,000 

I3S 

7,  700,000 

"3 

7,  240, 000 

164 

12,640,000 

185 

44,800,000 

658 

4,380,000 

64 

7,550,000 

III 

6,620,000 

97 

After 

Rela- 

3 months. 

tive. 

6,550,000 

100 

5,440,000 

83 

7,800,000 

119 

8, 440, 000 

128 

5, 900, 000 

90 

4, 900, 000 

74 

7, 660, 000 

116 

S,  400, 000 

82 

7,300,000 

III 

28, 500, 000 

435 

3,000,000 

45 

4, 670, 000 

71 

4,  500, 000 

68 

After 
3  months. 


Rela- 
tive. 


None 

One-fourth  calcium  carbonate 

One-half  calcium  carbonate 

Full  calcium  carbonate 

One-fourth  limestone 

One-half  limestone 

Full  limestone 

One-fourth  magnesium  carbonate 

One-hall  magnesium  carbonate 

Full  magnesium  carbonate 

0.1  per  cent  monocalcium  phosphate 

0.1  per  cent  monocalcium  phosphate-l-one-fourth 
calcium  carbonate 

0.1  per  cent  monocalcium  phosphate-1-full  cal- 
cium carbonate 


2, 100,000 
2,450,000 
3,  200,000 
2,900,000 
I,  780, 000 
2,550.000 
2,550,000 
3,000,000 
5,000,000 
4, 800, 000 
2, 340, 000 

3,230,000 

2, 700,000 


100 
116 
152 
138 
84 
121 
121 
133 
238 
238 
III 

154 
128 


486 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  8 


FORMATION  OP  AMMONIA 

The  ammonia  determinations  were  made  at  the  end  of  two  and  three 
months.  It  will  be  seen  from  the  data  in  Table  X  that  all  treatments 
accelerated  the  formation  of  ammonia  from  dried  blood.  After  two 
months  monocalcium  phosphate  plus  calcium  carbonate  showed  the 
greatest  effect,  although  the  phosphate-alone  treatment  gave  nearly  as 
good  results.  Of  the  carbonates,  calcium  carbonate  appeared  to  have 
the  greatest  effect  in  increasing  the  ammonifying  power,  an  effect  nearly 
as  great  as  that  obtained  from  the  phosphate  applications.  Magnesium 
carbonate  exerted  a  less  effect  than  did  calcium  carbonate,  and  lime- 
stone had  even  a  less  effect  than  was  obtained  with  magnesium  car- 
bonate. After  three  months  the  phosphate  still  retained  a  lead  in 
stimulating  ammonification.  Limestone  caused  a  greater  increase  in 
ammonia-producing  power  than  did  calcium  carbonate,  while  calcium 
carbonate  gave  a  greater  increase  than  was  obtained  with  magnesium 
carbonate.  From  the  data  on  Colby  soil  it  is  shown  that  the  increase 
in  ammonifying  power  is  not  always  parallel  with  the  increase  in  the 
total  number  of  organisms. 

Table  X. — Influence  of  cakium  carbonate,  magnesium  carbonate,  limestone,  and  thouo- 
calcium  phosphate  on  the  ammonification  of  dried  blood  in  Colby  silt  loam  and  Plain- 
field  sand 


Ammonia  nitrogen  in  loo  gm.  of  dry  soil. 

Treatment. 

Colby  silt  loam. 

Plainfield  sand. 

After  2 
months. 

In- 
crease. 

After  3 
months. 

In- 
crease. 

After  2 
months. 

In- 
crease. 

After  3 
months. 

In- 
crease. 

None 

Mgm. 
21- 1 
39- o 

Mgm,. 

Mgm. 

51-7 
S5-5 
61.6 
62.3 
61.6 
64-5 
69.7 

S8.i 

53-7 
52- S 

63.1 

78.0 
80.6 

Mgm. 

Mgm. 
49.0 

47-7 
40. 1 
40. 1 
51.8 
38.7 
37-7 

40-3 

39-9 
35-2 

S6.7 
SO.  5 
48-5 

Mgm.. 

Mgm. 
SS-i 

S2.2 
47.0 
S3- 1 
Sl-o 
51-3 
73-8 

46.7 

47- S 
47.0 

75-3 
73-3 
65.7 

Mgm. 

One-fourth  calcimn  carbonate 

+  1-3 

+  3-8 
-1-  9-9 
-I-I0.6 
+  9-9 

-fl2.8 

-hiS.o 

+  6.4 

+  2.0 
+  1.8 

-i-ll.4 
+26.3 

+  28.9 

-  1-3 

-  8.9 

-  8.9 
-f  2.8 
-10.3 
-1 1. 3 

-  8.7 

-  9.1 

-13-8 

+  7-7 

-t-i-s 

-  o-S 

—  2.9 

Full  calcium  carbonate 

One-fourth  limestone 

S3-0 
45-7 
40.6 
44.2 

49.0 

51.6 
57-1 

6i.  0 

57-5 

72- S 

+  15-3 
-f  8.0 
-1-2.9 
+  6.S 

+  11. 3 

+  13-9 
+  19-4 

+23-3 
+  19-8 

+34-8 

-  2.0 

-  4-1 

-  3-8 

+18.7 

-  8.4 

-  7-6 

-  8.0 

One-half  limestone 

Full  limestone 

One-fourth  magnesium   car- 
bonate   

One-half  magnesium  carbon- 
ate  

Full  magnesium  carbonate . . . 
O.I    per  cent   monocalciiun 
phosphate .... 

O.I    per   cent    monocalciimi 
phosphate-)-one-fourth  cal- 
cium carbonate 

-1-18.  a 

O.I    per   cent    monocalcium 
phosphate-(-full  calcium  car- 
bonate   

-f-io.6 

The  data  in  Table  X  for  Plainfield  sand  showed  very  different  results 
from  those  obtained  with  Colby  silt  loam.  After  two  months  a  decrease 
in  ammonifying  power  was  noted,  except  in  the  case  of  the  one-fourth 
limestone  treatment.  The  phosphate  when  applied  alone  gave  the 
greatest  increase,  but  where  calcium  carbonate  was  added  in  combina- 
tion with  the  phosphate,  the  formation  of  ammonia  was  not  so  great. 


Feb.  as,  1918 


Influence  of  Carbonates  on  Soil  Bacteria 


487 


With  an  increase  of  calcium  carbonate  there  was  a  decrease  in  ammonia 
production.  This  decrease  was  shown  with  all  carbonates,  magnesitun 
carbonate  causing  the  greatest  decrease. 

After  three  months  the  same  order  was  held  by  these  compounds  in 
stimulating  ammonia  formation.  At  this  time  the  full  limestone  had  a 
greater  effect  than  the  smaller  limestone  applications;  the  effect  was 
almost  as  great  as  that  due  to  the  phosphate  treatment.  It  is  evident 
from  the  results  obtained  in  both  Colby  silt  loam  and  Plainfield  sand, 
that  the  greatest  accumulation  of  ammonia  (six  days'  incubation)  does 
not  occur  where  the  largest  increase  in  the  number  of  bacteria  was 
obtained.  Where  the  greatest  number  of  bacteria  developed,  it  seems 
probable  that  the  greatest  amount  of  ammonia  should  be  formed.  Since 
the  substances  which  gave  the  highest  number  of  bacteria  also  neutral- 
ized the  soil  acids,  it  is  probable  that  the  ammonia  partly  escaped  through 
volatilization.  Because  of  the  open  texture  of  the  Plainfield  sand,  more 
ammonia  escaped  from  this  soil  than  from  the  Colby  silt  loam. 

Table  XI. — Influence  of  calcium  carbonate,  magnesium  carbonate,  limestone,  andm,ono- 
calcium  phosphate  on  nitrate  accu7nulation  in  Colby  silt  loam  and  Plainfield  sand 


Treatment. 


None 

One-fourth  calcium  car- 
bonate  

One-half  calcium  carbo- 
nate  , 

Full  calcium  carbonate. . , 

One-fourth  limestone 

One-half  limestone 

Full  limestone 

One-fourth  magnesium 
carbonate 

One-half  magnesium  car- 
bonate  

Full  magnesiinn  carbon- 
ate  

0.1  per  cent  monocal- 
cium  phosphate 

0.1  per  cent  monocal- 
cium  phosphate  -t-  one- 
fourth  calcium  carbon- 
ate  

0.1  per  cent  monocal- 
cium  phosphate  4-  full 
calcium  carbonate 


Nitrate  nitrogen  accumulated  in  100  gm.  of  dry  soil. 


Colby  silt  loam. 


Mgm. 
S-86 

7-23 

6. 10 
6.25 
5- S3 
S-86 
6.51 

6.58 

8.13 

13-28 

S-63 

6. 10 
6.73 


Mgm. 


•24 
•37 

—  -33 
.00 
•6S 

•72 

2.  27 

7.42 

—  -23 


.24 

.87 


•^  o 


Mgm. 
6.05 

6-94 

7-s6 
8.18 
6.0s 
6.  40 
6.  42 

6.75 

8.44 

13.00 

6.05 

6.  42 


Mgm. 


0.89 

I- SI 
2. 13 
.00 
•35 
•37 

.70 

2-39 

6.9s 

.00 


•37 
2-13 


Mgm. 
6.62 

8.36 

9^37 
12.00 

7.2s 
7-25 
8.56 

9-43 
12.00 
IS- 00 

7-2S 


Mgm. 


1.94 

2^7S 
5^38 
•63 
•63 
•94 

2.81 

S-38 

8.38 

.63 


2-75 
4-63 


Plainfield  sand. 


a  5 


Mgm. 
1.74 

2.42 

2.3s 
2.  20 
a.  20 
1.87 
1.97 

a.  10 

2.  20 
2^34 
2.  12 


Mgm.. 


.61 
.46 
.46 
•13 
•23 

•36 

.46 

.60 

•38 

.61 
.06 


^a 


Mgm. 
1.S8 

i.8s 

2.31 
2.  22 
1.80 
1.67 
1.87 

1.24 


2.78 
1.85 


Mgm.. 


•73 
.64 
.  22 
.09 
•29 

■  -34 

.64 

I.  20 

•27 


^a 


Mgm 
^•33 

1.67 

3-55 
2.92 
1-73 

2.  00 
3.80 

1.  SO 

2.  70 
4.62 

3-oS 


Mgm. 


0.34 

1.23 
i^S9 
.40 
.67 
1.47 

•17 

^•37 

3- 29 

I- 75 

•87 
I-7S 


ACCUMULATION   OP   NITRATES 

From  the  data  presented  in  Table  XI  it  will  be  seen  that  nitrates 
accumulated  faster  in  both  the  treated  Colby  silt  loam  and  Plainfield 
sand  than  in  the  untreated.  In  both  soils  the  magnesium  carbonate 
benefited  nitrification  more  than  did  the  other  carbonates.  Next  in 
order  to  magnesium  carbonate  was  calcium  carbonate,  and  lastly,  lime- 


488 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  8 


stone,  in  increasing  the  nitrate  content  of  Colby  silt  loam  and  Plainfield 
sand;  the  heavier  applications  gave  the  highest  nitrate  accumulation. 
In  the  case  of  the  magnesium  carbonate  with  Colby  silt  loam  soil,  the 
increased  accumulation  reached  as  high  as  8  mgm.  The  increase  in 
nitrate  in  the  Plainfield  sand  ran  parallel  with  that  in  the  Colby  silt  loam 
where  similarly  treated,  though  the  accumulation  in  the  sand  was  much 
smaller,  since  this  soil  is  low  in  organic  matter. 


t  :^  3  -^  S  6  7  Q  9  /O  imJ3 
I  month 


I  £  i  'f  5  b    7  6  9 
2mor,ih% 


3  fVoni/is 


Fio.  lo. — Diagram  showing  the  influence  of  calcium  carbonate,  magnesium  carbonate,  limestone,  and 
monocalcium  phosphate  on  nitrate  accumulation  of  Colby  silt  loam. 

i=no  treatment. 

3=  one-fourth  calcium  carbonate. 

3=  one-half  calcium  carbonate. 

4=  full  calcium  carbonate. 

5=  one-fourth  limestone. 

6=  one-half  limestone. 

7=  full  limestone. 

8=  one-fourth  magnesium  carbonate. 

9=  one-half  magnesium  carbonate. 
io=  full  magnesium  carbonate. 
11=  O.I  per  cent  monocalcium  phosphate. 

1 2=  O.I  per  cent  monocalcium  phosphate+one-fourth  calcium  carbonate. 
13=0.1  per  cent  monocaldum  phosphate-1-full  calcium  carbonate. 

The  monocalcium  phosphate  apparently  did  not  favor  nitrification  in 
the  Colby  soil,  while  in  the  sand  this  substance  proved  beneficial.  A 
combination  of  the  phosphate  and  calcium  carbonate  did  not  increase 
nitrification  beyond  that  obtained  from  the  use  of  calcium  carbonate 
alone.  The  effect  of  these  substances  on  the  reproduction  of  soil 
organisms  and  on  the  formation  of  nitrates  was  similar.  A  review  of 
the  entire  data  is  shown  graphically  in  figure  10. 


Feb.  25,  1918 


Influence  of  Carbonates  on  Soil  Bacteria 


489 


INFLUENCE  OF  CALCIUM  CARBONATE,  MAGNESIUM  CARBONATE,  LIME- 
STONE, AND  MONOCALCIUM  PHOSPHATE  ON  THE  NITRIFICATION  OF 
GELATIN  IN   SOIL 

In  order  to  study  the  effect  of  calcium  carbonate,  magnesium  car- 
bonate, limestone,  and  monocalcium  phosphate  on  nitrification  in  Colby 
and  in  Plainfield  sand  to  which  nitrogenous  material  was  added,  a  series 
of  tests  was  made.  After  the  soils  had  been  treated  w^th  the  forenamed 
substances  separately  and  allowed  to  incubate  for  three  months,  loo-gm. 
samples  were  drawn  and  to  each  portion  23.6  mgm.  of  gelatin  were  added. 
The  soil  was  placed  in  tumblers,  which  were  covered  with  glass,  and 
incubated  for  six  weeks  at  27°  C.  At  the  end  of  this  time  the  nitrate 
content  was  determined.  The  results  of  this  experiment  are  given  in 
Table  XII. 


Table  XII. — Influence  of  calcium  carbonate,  magnesium  carbonate,  limestone,  and  mono- 
calcium,  phosphate  on  the  nitrification  of  gelatin  in  Colby  silt  loam  and  in  Plainfield 
sand 


Treatment. 


Nitrate  nitrogen  in  loo  gm.  of  dry  soil. 


Colby  silt  loam. 


After  6 
weeks. 


Increase. 


Plainfield  sand. 


After  6 
weeks. 


Increase. 


None , 

One-fourth  calcium  carbonate 

One-half  calcium  carbonate 

Full  calcium  carbonate 

One-fourth  limestone 

One-half  limestone 

Full  limestone 

One-fourth  magnesium  carbonate 

One-half  magnesiiun  carbonate 

Full  magnesium  carbonate 

0.1  per  cent  monocalcium  phosphate 

0.1  per  cent  monocalcium  phosphateH-one-fourth  calcium  car- 
bonate   

0.1  per  cent  monocalcium  phosphate-t-full  calcium  carbonate. . 


Mgm. 
20.87 
22.31 
25.00 
28.43 
20.87 
23.12 
25.00 

23-12 
26.  12 
25.09 
22.31 

23-12 
28.43 


Mgm. 


1-44 
4-13 
7.56 
.00 
8.  25 
4-13 
3.25 
S-25 
4-13 
1.44 

3.25 

7-56 


Mgm.. 

9.06 
11.70 
10.57 
zi.  12 
II.  70 
II.  70 
II.  12 
II.  12 
13.09 
12.24 

9.06 

13.09 
12.24 


Mgm. 


2.64 
1.51 
2.06 
3.64 
3.64 
3.06 

3.  06 

4-03 

3- 18 
.00 

4-03 
3- 18 


In  reviewing  Table  XII  it  will  be  seen  that  with  the  exception  of  one- 
fourth  limestone  treatment  in  Colby  loam,  and  phosphate  alone  in 
Plainfield  sand,  there  was  an  increase  in  nitrification  in  the  treated 
soils.  The  increase  in  most  cases  was  very  slight,  especially  in  the  sand, 
where  only  about  half  of  the  nitrogen  of  gelatin  apparently  was  nitrified, 
while  in  the  Colby  loam  the  greater  part  of  the  nitrogen  of  gelatin  was 
nitrified.  A  comparison  of  the  amount  of  nitrate  formed  in  the  soil  to 
which  magnesium  or  calcium  carbonate  was  added,  both  with  and  without 
the  addition  of  gelatin  (Table  XI),  showed  a  relatively  greater  nitrate 
formation  in  the  soils  to  which  no  organic  nitrogen  was  added.  These 
results  were  to  be  expected,  since  it  is  likely  that  some  of  the  ammonia 
formed  from  the  gelatin  in  the  neutral  or  partly  neutral  soil  escaped. 
On  the  other  hand,  the  combined  effects  of  the  gelatin  and  the  carbonates 
increased  the  multiplication  of  bacteria  beyond  that  of  the  soil  treated 
with  gelatin  alone.     The  great  gain  in  the  number  of  bacteria  is  no  doubt 


490 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  8 


followed  by  an  increase  in  assimilation  of  nitrates.     These  results  agree 
with  those  obtained  by  Fred  and  Graul  (20). 

In  the  Colby  silt  loam  the  full  calcium  carbonate  gave  the  best  results 
while  magnesium  carbonate  and  limestone  rank  in  the  order  named. 
The  magnesium  carbonate  when  added  to  the  soil  with  gelatin  did  not 
give  as  good  results  as  calcium  carbonate.  This  difference  may  be  due  to 
the  fact  that  the  large  increase  in  the  number  of  bacteria  caused  by  the 
magnesium  carbonate  treatment  favored  a  greater  assimilation  of  nitrate 
by  the  microorganisms.  Plainfield  sand  apparently  gave  better  results 
where  the  one-fourth  or  one-half  neutralization  was  obtained.  Very 
probably  the  sand  releases  more  ammonia  when  neutral  than  does  the 
Colby  silt  loam.  In  both  soils  the  phosphate  benefited  nitrification  but 
slightly. 

INFLUENCE  OF  CALCIUM  CARBONATE,  MAGNESIUM  CARBONATE,  LIMESTONE, 
AND   MONOCALCIUM    PHOSPHATE   ON    NITROGEN    FIXATION    IN    SOIL 

An  effort  was  made  to  study  the  influence  which  carbonate  and  phos- 
phate treatments  would  have  on  the  independent  nitrogen-fixing  organ- 
ism, Bacillus  azotobacter  chroococcum.  Three  months  after  treatment, 
Colby  and  Plainfield  soils  were  sampled  and  the  soil  placed  in  large  soup 
plates,  I  per  cent  of  mannit  was  added  to  each  plate  of  soil,  which  was 
then  inoculated  with  a  culture  of  B.  azotobacter.  After  incubating  for 
four  weeks  in  the  greenhouse,  the  soils  were  dried  and  ground  to  a  very 
fine  powder  in  a  ball  mill.  Total  nitrogen  determinations  (Kjeldahl 
method  modified  to  include  nitrates)  were  then  made.  Duplicate  deter- 
minations were  made  from  each  plate  and  an  average  of  these  taken. 
The  results  for  this  experiment  are  given  in  Table  XIII. 

Table  XIII. — Influence  of  calcium  carbonate,  magnesium  carbonate,  limestone,  and 
monocakium  phosphate  on  nitrogen  fixation  in  Colby  silt  loam  and  Plainfield  sand 
treated  with  mannit  and  inoculated  with  Bacillus  azotobacter 


Treatment. 


None 

One-fourth  calcium  carbonate 

One-half  calcium  carbonate 

Full  calcium  carbonate 

One-fourth  limestone 

One-half  limestone 

FuU  limestone 

One-fourth  magnesium  carbonate 

'  One-half  magnesium  carbonate 

Full  magnesium  carbonate 

0.1  per  cent  monocalcium  phosphate. 

0.1  per  cent  monocalcium  phos- 
phate +  one-fourth  calcium  car- 
bonate  

0.1  per  cent  monocalcium  phos- 
phate +  full  calcium  carbonate 


Nitrogen  in  loo  gm.   of  dry  soil. 


Colby  silt  loam. 


After  4 
weeks. 


Mgm. 

273.0 
277.  o 
273.0 
276.0 
27S.O 
273.0 
278.0 
277.0 
278.0 
275.0 
274.0 


2  79- o 
283.0 


Mgm. 

275.0 
276.  o 
275.0 
276.0 
275-0 
27S-0 
279.  o 
276.0 
276.0 
27S-0 
276.0 


282.  o 
277.  o 


Aver- 
age. 


Mom. 
275.0 
276.5 
274.  o 
276.0 
275- o 
274.0 
278.5 
276.5 
277.  o 
275- o 
275-0 


280.5 
280.  o 


In- 
crease. 


Mgm.. 


+  5-5 
+  S-0 


Plainfield  sand. 


After  4 
weeks. 


Mgm. 
63.0 
59- o 
6i.  o 
64.0 
6i.o 

61.  o 

62.  o 
59- o 
62.  o 
63.0 
61.  o 


61.  o 
65.0 


Mgm. 
61. 


59- o 
65.0 


Aver- 
age. 


Mgm.. 


0 
0 

59 

0 

61 

0 

63 

0 

60 

0 

62 

0 

63 

0 

59 

.0 

62 

.0 

63 

0 

61 

60.  o 
65.  o 


In- 
crease. 


Mgm. 


—2.0 

-I-3-0 


Feb.  as.  1918  Influence  of  Carbonates  on  Soil  Bacteria  491 

From  the  data  presented  in  Table  XIII  it  appears  that  there  was  no 
decided  gain  in  total  nitrogen  in  the  Plainfield  sand,  and  only  a  small  gain 
in  Colby  silt  loam.  It  is  difficult  to  explain  why  no  gain,  and  in  some 
cases  a  loss,  of  nitrogen  was  found.  The  loss  may  have  been  due  to  a 
disturbance  in  the  balance  of  the  flora  of  these  soils.  Since  the  mannit 
which  was  added  to  the  soil,  and  the  neutralization  of  some  of  the  soil 
acids  by  the  basic  substances  caused  a  great  increase  in  the  number  of 
bacteria,  it  is  possible  that  a  part  of  the  nitrate  nitrogen  was  set  free  as 
elemental  nitrogen.  However,  a  gain  in  total  nitrogen  is  shown  in  the 
Colby  silt  loam  soil  due  to  the  carbonate  and  limestone  treatment. 

INFLUENCE   OF  CALCIUM  CARBONATE,   MAGNESIUM  CARBONATE,   AND 
LIMESTONE  ON  ORGANISMS  IN  PURE  CULTURE 

An  effort  was  made  to  determine  the  influence  of  calcium  carbonate, 
magnesium  carbonate,  and  limestone  on  pure  cultures  of  bacteria  in  steril- 
ized soil.     In  order  to  simplify  the  vv^ork,  only  Colby  soil  was  employed. 

It  was  obser\^ed  that  when  this  soil  is  heated  for  a  long  period  at  a 
high  temperature,  there  is  a  reduction  in  soil  acidity.  Other  investi- 
gators have  reported  similar  results  in  acid  soils.  Conner  (10)  showed  a 
decrease  in  acidity  when  soil  was  heated  to  complete  dryness.  He  sug- 
gested that  possibly  the  acid  silicates  were  hydrolyzed  with  the  formation 
of  a  base.  On  heating  soil  to  140°  C.  for  one  hour.  Sharp  and  Hoagland 
(55)  noted  a  decrease  in  acidity.  On  the  other  hand,  Schreiner  and 
Lathrop  (54)  increased  the  acidity  in  an  acid-reacting  soil  by  heating  the 
soil  for  three  hours  at  30  pounds'  pressure,  which  in  all  probability 
changed  the  organic  matter  to  such  an  extent  as  to  increase  the  acidity. 
The  work  of  Kelley  and  McGeorge  (jo)  and  Darbishire  and  Russell  (//) 
showed  a  change  in  the  soil  consituents  on  heating,  especially  in  the  solid 
inorganic  constituents,  which  were  made  more  soluble. 

In  order  to  overcome  as  much  as  possible  this  reduction  in  acidity, 
various  methods  of  sterilization  were  tried.  Both  dry  and  moist  soils 
were  sterilized  in  Erlenmeyer  flasks,  which  were  heated  in  steam  under  a 
pressure  of  15  pounds  for  three  hours.  On  testing  this  heated  soil  for 
its  degree  of  acidity  by  the  Truog  zinc-sulphid  test,  the  dry  soil  showed 
practically  no  change  in  its  acid  content,  while  the  moist  soil,  one-half 
and  full  water-saturated,  showed  a  considerable  decrease.  The  very  wet 
soil  after  sterilization  contained  less  acid  than  that  which  received  one 
half  as  much  water.  From  the  evidence  it  seems  safe  to  conclude  that 
the  silicates  of  this  soil  are  partially  hydrolyzed  with  the  formation  of 
bases. 

Further  tests  were  conducted  with  dry  soil.  It  was  dried  for  24 
hours  at  about  45°  C.  and  placed  in  flasks;  these  were  then  plugged  and 
sterihzed  for  three  hours  at  15  pounds'  pressure.  After  sterilization,  the 
soil  received  water  sufficient  to  bring  it  to  half  saturation.  This  was 
done  to  determine  if  hydrolysis  took  place  by  adding  water  after  the  soil 


492 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  8 


had  been  sterilized.  At  the  end  of  four  days  tests  were  made  on  the 
dry  and  on  the  watered  soil.  The  acidity  of  both  appeared  unchanged. 
From  the  results  of  the  foregoing  experiments  it  appears  that  dry 
Colby  soil  can  be  sterilized  and  sterile  water  added  until  the  soil  is  half 
saturated,  without  interfering  seriously  with  the  original  reaction  of 
the  soil.  Therefore  this  method  of  sterilizing  soil  was  adopted  for  all 
of  the  pure-culture  work. 


INFLUENCE  OF  CALCIUM  CARBONATE,  MAGNESIUM  CARBONATE,  AND  LIME- 
STONE ON  THE  AMMONIFICATION  OF  BLOOD  MEAL  BY  PURE  CULTURES 
OF  BACTERIA   IN   SOIL 

Colby  silt  loam. — One-hundred-gm.  portions  of  dry  soil  were  each 
treated  with  calcium  carbonate,  with  magnesium  carbonate,  and  with 
limestone  separately.  To  the  soil  of  the  entire  series  2  per  cent  of  dried- 
blood  meal  was  applied.  The  treated  soil  was  placed  in  300-c.  c.  Erlen- 
meyer  flasks  and  sterilized  as  previously  stated.  Sterilized  water  was 
then  added  to  bring  the  soil  to  half  saturation.  The  sterilized  soil  was 
then  inoculated  with  water  suspensions  of  Bacillus  tumescens  and  B. 
suhtilis.  The  flasks  were  incubated  at  27°  C.  for  seven  days  and  at  the 
end  of  this  period  the  ammonia  in  each  flask  was  determined.  In  Table 
XIV  are  recorded  the  complete  data  for  this  experiment. 

Table  XIV. — Influence  of  calcium  carbonate,  magnesium  carbonate,  and  limestone  on 
the  am.monification  of  dried  blood  with  pure  cultures  in  Colby  silt  loam 


Treatment. 


Ammonia  nitrogen  in  loo  gm.  of  dry  soil. 


Bacillus  tumescens. 


After 
7  days. 


Increase. 


Bacillus  svbiilis. 


After 
7  days. 


Increase. 


None , 

One-fourth  calcium  carbonate. . . . 

One-half  calcium  carbonate 

Full  calcitmi  carbonate 

One-fourth  limestone 

One-half  limestone 

Full  limestone 

One-fourth  magiiesium  carbonate 
One-half  magnesium  carbonate. . . 
Full  magnesium  carbonate 


Mgm. 


Mgni. 


•3 

15- S 
16.4 
16.4 
II.  2 
14.  6 
14.8 
15-7 
16.  s 
17. 1 


Mgm. 

II. o 


Mgm. 


4.2 
5-1 

S-i 
-o.  I 
3-3 
3-5 
4-4 
S-2 
S-8 


16.0 
16.0 
II.  7 
13- o 
13-9 
12- 5 
17.6 
21.0 


50 
S-O 

0.7 
2.0 

2.9 

1-5 
6.6 
10.  o 


The  data  in  Table  XIV  show  clearly  that  the  ammonification  of  dried 
blood  by  Bacillus  tumescens  or  B.  suhtilis  was  increased  when  the 
sterilized  soil  was  treated  with  calcium  carbonate  or  with  magnesium 
carbonate  or  with  limestone.  Magnesium  carbonate  gave  better  results 
than  did  calcium  carbonate,  while  the  latter  gave  better  results  than  did 
the  limestone.  In  every  case  full  treatment  with  any  of  the  compounds 
gave  the  largest  increase  of  ammonia.  This  is  in  agreement  with  results 
obtained  in  unsterilized  Colby  silt  loam  two  months  after  treatment  with 
the  limestone  or  calcium  carbonate. 


Feb.  2$,  1918 


Influence  of  Carbonates  on  Soil  Bacteria 


493 


INFLUENCE  OF  CALCIUM  CARBONATE,  MAGNESIUM  CARBONATE,   AND  LIME- 
STONE ON  BACILLUS  AZOTOBACTER  IN  STERILIZED  SOIL 

Colby  silt  loam. — Dry  soil  in  200-gm.  portions  was  treated  with  cal- 
cium carbonate,  with  magnesium  carbonate,  and  limestone,  and  placed 
in  500-c.  c.  Erlenmeyer  flasks.  To  each  flask  was  added  i  per  cent  of 
mannit  by  weight  of  dry  soil.  The  flasks  were  plugged,  sterilized,  and 
sterile  water  was  added  to  bring  the  soil  to  the  proper  moisture  content. 
Inoculations  of  Bacillus  azotobacier  in  pure  culture  were  then  made  by 
adding  a  i-c.  c.  suspension  of  the  organisms  to  each  flask.  The  entire 
set  of  flasks  was  incubated  at  27°  C,  and  after  one,  two,  and  three  weeks 
plate  counts,  using  mannit  agar,  were  made.  The  results  for  this  experi- 
ment are  recorded  in  Table  XV. 

Table  XV. — Influence  of  calcium  carbonate,  magnesium  carbonate,  and  limestone  on 
Bacillus  azotobacter  in  Colby  silt  loam  treated  with  mannit 


Treatment. 


Number  of  bacteria  in  i  gm.  of  dry  soil. 


After  I  week. 


After  2  weeks. 


After  3  weeks. 


None 

One-fourth  calcium  carbonate 

One-half  calcium  carbonate 

Full  calcium  carbonate 

One-fourth  limestone 

One-half  limestone 

Full  limestone 

One-fourth  magnesium  carbonate 
One-half  magnesium  carbonate. . 
Full  magnesiiun  carbonate 


602, 
200, 
400, 
<i. 
<i, 
<i, 
200, 
150, 
800, 


<i,ooo 

4, 170,000 
12,500,000 
11,350,000 

<I,000 

<i,ooo 

850,000 

3,  620, 000 

47, 000, 000 

477,000,000 


<I,000 

2,440,000 

32,300,000 

45,500,000 

<i,ooo 

<i,ooo 

26, 000, 000 

10,350,000 

268, 000, 000 

355,000,000 


The  results  of  this  experiment  showed  the  striking  effect  of  magnesium 
carbonate  on  the  multiplication  of  Bacillus  azotobacter  cells  in  Colby  silt 
loam  soil.  Calcium  carbonate  caused  an  increase  in  the  number  of  these 
organisms,  but  the  increase  was  not  as  great  as  that  obtained  with  mag- 
nesium carbonate.  With  limestone  the  increase  was  very  small.  This 
great  gain  with  magnesium  carbonate  was  obtained  where  amounts 
equal  to  full  neutralization  were  applied.  After  the  third  week  the  one- 
half  neutralization  by  magnesium  carbonate  also  gave  a  large  increase, 
while  two  weeks  after  treatment  the  greatest  gain  was  shown. 


INFLUENCE    of    large     APPLICATIONS    OF    MAGNESIUM     CARBONATE     ON 
BACILLUS  AZOTOBACTER   IN   STERILE   SOIL 

Colby  silt  loam.— Because  magnesium  carbonate  when  applied  to  give 
neutralization,  increased  the  number  of  Bacillus  azotobacter  in  sterilized 
Colby  silt  loam  soil  to  a  great  extent,  a  further  test  was  made  to  deter- 
mine if  heavier  applications,  enough  added  to  give  a  distinct  alkaline 
reaction,  would  continue  to  increase  the  reproduction  of  Bacilhis  azoto- 
bacter. These  applications  were  made  by  adding  magnesium  carbonate 
suflBcient  to  make  iXi  iX»  ^^^  double  neutralization. 


494 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  8 


The  data  for  this  test,  presented  in  Table  XVI,  showed  that  i}4  mag- 
nesium— carbonate  treatment  after  one  week  gave  a  greater  increase  in 

the  number  of  Bacil- 

2,500,000 

lus  azotobacter  cells 
than  was  obtained  by 
the  iX  application; 
after  two  and  three 
weeks,  the  iX  treat- 
ment gave  a  greater 
increase  than  was  ob- 
tained by  the  i}4 
treatment.  In  every 
case  the  heaviest  ap- 
plication of  magne- 
sium carbonate  caused 
less  increase  in  the 
number  of  these  or- 
ganisms than  did  the 
lighter  treatments. 
However,  as  compared 
with  the  control,  the 
heavy  application  also 
gave  a  great  increase. 
In  comparing  the 
data  of  Table  XVI 
with  those  of  Table 
XV,  it  will  be  seen 
that  iX  ^^^  ill  most 
cases  i}4  magnesium- 
carbonate  treatment 
caused  a  greater  in- 
crease in  the  number 
of  Bacillus  azotobacter 
organisms  in  sterilized 
Colby  silt  loam  than 
the  full  treatment.  It 
is  difficult  to  explain 
why  such  an  enor- 
mous increase  in  the 
number  of  these  or- 
ganisms was  obtained 
when  more  than 
enough  magnesium 
carbonate  was  added 
to  neutralize  the  soil  acids.  From  the  data  in  Table  XVI,  one  is  led  to 
believe  that  this  great  multiplication  in  Bacillus  azotobacter  cells  was  due 


12  3  4 
lUTeek 


t   Z  3  -t- 
BU/eeh 


Fig.  II.— Diagram  showing  the  influence  of  large  applications  of  mag- 
nesium carbonate  on  Bacillus  azotobacter  in  sterile  Colby  silt  loam. 


I— no  treatment. 

9=»  il4  magnesium  carbonate. 


3=  1^2  magnesium  carbonate. 
4=  double  magnesium  carbonate. 


Feb.  as.  1918 


Influence  of  Carbonates  on  Soil  Bacteria 


495 


to  something  besides  the  correction  of  the  soil  acidity, 
in  Table  XVI  are  presented  graphically  in  figure  1 1 . 


The  entire  data 


Table  XVI. — Influence  of  large  applications  of  magnesium  carbonate  on  Bacillus  azoto- 
bacter  in  Colby  silt  loam  treated  with  fnannit 


Treatment. 

Number  of  bacteria  in  i  gm.  of  dry  soil. 

After  I  week.      After  2  weeks. 

After  3  weeks. 

None 

<i,ooo 

13. 500,000 

147. 500,000 

10, 300, 000 

<I.OOO 

1,030,000,000 
357,000,000 
338,000,000 

One  and  one-fourth  magnesium  carbonate 

481,000,000 

INFLUENCE    OF    MAGNESIUM    CARBONATE    ON    BACILLUS    AZOTOBACTER    IN 
STERILE   NEUTRAL   SOIL 

Miami  silt  loam. — Since  magnesium  carbonate  increased  the  number 
of  Bacillus  azoiobacter  cells  in  an  acid  soil  when  more  than  enough  of  the 
carbonate  was  added  to  neutralize  the  acidity,  an  experiment  was  planned 
to  determine  the  effect  this  compound  would  have  on  Bacillus  azoiobacter 
in  a  neutral  soil.  The  soil  selected  for  this  work  was  Miami  silt  loam, 
previously  described.  Magnesium  carbonate,  along  with  i  per  cent 
mannit,  was  added  in  one-fourth,  one-half,  and  full  neutralization  based 
on  the  acidity  of  Colby  silt  loam.  The  data  for  this  test  are  tabulated  in 
Table  XVII. 


Table  XVII.- 


-Influence  of  magnesium  carbonate  on  Bacillus  azoiobacter  in  sterile 
Miami  silt  loam  treated  with  m,annit 


Treatment. 

Number  of  bacteria  in  i  gm.  of  dry  soil. 

After  I  week. 

Relative. 

After  2  v/ceks 

Relative. 

After  3  weeks. 

Relative. 

None 

3,020,000 

2,870,000 

44, 000, 000 
67, 700,000 

100 
95 

1,456 
2,241 

37,500,000 

52,500,000 

II, 200,000 
26, 200, 000 

J  00 

140 

30 
70 

51,000,000 

127,000,000 

137,000,000 
392,000,000 

One-fourth    magnesium    car- 
bonate  

249 

One-half      magnesium      car- 
bonate  

Full  magnesium  carbonate 

768 

In  studying  the  data  in  this  table,  it  will  be  seen  that  in  Miami  silt 
loam,  a  neutral  soil,  where  magnesium  carbonate  was  applied  in  an 
amount  great  enough  to  neutralize  all  the  active  acidity  in  Colby  soil, 
there  was  a  great  increase  in  the  number  of  Bacillus  azoiobacter  cells. 
At  this  time  no  explanation  can  be  offered  to  account  for  the  drop  in 
number  after  two  weeks  where  one-half  and  full  magnesium  carbonate 
were  applied,  since  the  other  counts,  made  after  one  and  three  weeks 
with  the  one-half  and  full  application  gave  a  decided  increase.  The 
count  made  after  the  3-week  period  showed  the  greatest  increase.  From 
the  results  of  Table  XVII  it  is  clear  that  magnesium  carbonate  plaj's  some 
role  in  stimulating  the  B.  azoiobacter  organism  other  than  that  of 
neutralization. 

27811°— 18 3  ' 


496 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  8 


INFLUENCE     OF     CALCIUM      CARBONATE,      MAGNESIUM      CARBONATE,      AND 
LIMESTONE    ON    BACILLUS   RADICICOLA    (ALFALFa)    IN    STERILE    SOIL 

Colby  silt  loam. — This  experiment  was  planned  to  measure  the  effect 
of  calcium  carbonate,  magnesium  carbonate,  and  limestone  on  the 
reproduction  of  Bacillus  radicicola  in  sterilized  Colby  silt  loam  soil. 
The  alfalfa  strain,  which  is  supposed  to  be  sensitive  to  acidity,  was  selected 
for  this  work.  The  soil  was  treated  with  i  per  cent  of  mannit,  and  the 
magnesium  carbonate  was  added.  Two-hundred-gm.  portions  of  the 
treated  soil  were  placed  in  500-c.c.  Erlenmeyer  flasks.  The  flasks  were 
then  plugged  and  sterilized  according  to  the  method  used  in  the  pre- 
ceding experiments.  Inoculations  were  made  by  introducing  a  i-c.c. 
suspension  of  the  organisms. 

From  the  data  presented  in  Table  XVIII  it  will  be  seen  that  calcium 
and  magnesium  carbonates  increased  greatly  the  number  of  alfalfa 
organisms  in  this  soil.  The  magnesium  carbonate  after  one  week  did 
not  increase  the  number  of  these  organisms  to  any  greater  extent  than 
did  calcium  carbonate.  The  smaller  amounts  were  surprisingly  efifective, 
though  the  increase  was  not  as  great  as  that  obtained  with  larger 
amounts.  Limestone  increased  the  number  of  these  organisms  to  a 
very  slight  extent. 

TablB  XVIII. — Influence  of  calcium  carbonate,  magnesium  carbonate,  and  limestone 
on  Bacillus  radicicola  {alfalfa)  in  sterile  Colby  silt  loam  treated  with  mannit 


Treatment. 

Number  of  bacteria  in  i  gm.  of  dry  soil. 

After  I  week. 

After  2  weeks. 

After  3  weeks. 

None 

12, 500 

340, 000, 000 

1,310,000,000 

J,  750,000,000 

101,000 

227,000 

2,450, 000 

87, 000, 000 

4, 100,000,000 

4, 900, 000, 000 

1,200 

1,350,000,000 

1,650,000,000 

1,620,000,000 

100,000 

5,000 

4,170,000 

1,212,000,000 

1, 112,000,000 

600, 000, 000 

1,170,000,000 

1,870,000,000 

50,000 

240, 000, 000 

1,540,000, 000 

1,770,000,000 

925,000,000 

INFLUENCE     OF     CALCIUM     AND    MAGNESIUM     CARBONATES     ON     BACILLUS 
RADICICOLA    (lupine)    IN    STERILIZED   SOIL 

Colby  silt  loam. — Since  the  alfalfa  strain  of  Bacillus  radicicola  was 
greatly  benefited  by  the  carbonates  of  magnesium  and  calcium,  it  seemed 
desirable  to  see  what  effect  a  similar  treatment  would  have  on  the  lupine 
strain  of  B.  radicicola,  which  is  frequently  termed  an  acid-resistant 
organism.  Limestone  was  eliminated  in  this  experiment;  otherwise  the 
procedure  for  this  test  was  similar  to  that  in  the  preceding  experiment. 
The  results  of  this  test  are  tabulated  in  Table  XIX. 


Feb.  2Si  1918 


Influence  of  Carbonates  on  Soil  Bacteria 


497 


Table  XIX. — Influence  oj  calcium  carbonate  and  magnesium  carbonate  on  Bacillus 
radicicola  {lupine)  in  sterile  Colby  silt  loam  treated  with  mannit 


Treatment. 

Number  of  bacteria  in  i  gfm.  of  dry  soil. 

After  I  week. 

After  3  weeks. 

After  3  weeks. 

None 

112,000 

875,000,000 

1,300,000,000 

1,837,000,000 

886, 000, 000 

9SO,  000, 000 

1, 750, 000, 000 

30,000 
3,  770,000,000 
3, 630, 000, 000 
a,  760, 00c,  000 
4,025,000, 000 
4,837,000,000 
8, 850, 000, 000 

One-fourth  calcium  carbonate 

5,530,000,000 
3,820,000,000 
4, 200, 000, 00c 
125,000,000 
5,000,000,000 
9,170,000,000 

One-half  calcium  carbonate 

FtlJJ  TTifi£:T7'";'1ir"  c^rlffmatr-. , , , 

From  the  data  given  in  Table  XIX  it  is  evident  that  the  lupine  strain 
of  Bacillus  radicicola  was  benefited  to  as  great  an  extent  as  the  alfalfa 
strain  when  Colby  soil  was  treated  with  magnesium  or  calcium  carbo- 
nate. After  the  second  and  third  weeks  the  counts  showed  that  mag- 
nesium carbonate  gave  better  results  than  did  calcium  carbonate.  The 
untreated  soil  gave  a  slightly  higher  number  than  was  obtained  with  the 
alfalfa  strain  under  similar  conditions.  This  difference  may  be  due  to 
the  fact  that  the  lupine  strain  is  slightly  tolerant  to  an  acid  reaction. 

Here,  as  in  the  tests  made  with  unsterilized  soil,  the  smaller  applications 
stimulated  multiplication  to  a  greater  extent,  in  proportion  to  the 
amounts  applied,  than  did  the  larger  treatments.  However,  the  maxi- 
mum gain  in  number  of  bacteria  was  obtained  with  full  magnesium- 
carbonate  treatment.  From  the  data  obtained  in  this  and  the  preceding 
experiment  it  is  clearly  evident  that  either  magnesium  or  calcium  car- 
bonate (the  magnesium  to  a  somewhat  greater  degree  than  the  calcium) 
greatly  increased  the  number  of  legume  bacteria,  alfalfa  and  lupine 
strains,  in  an  acid  soil. 

CONCLUSIONS 

From  a  general  review  of  the  results  of  the  experiments  just  described 
It  is  clearly  shown  that  carbonates  of  calcium  and  magnesium  when 
applied  to  acid  Colby  silt  loam,  acid  Plainfield  sand,  and  neutral  Miami 
silt  loam  increase  the  number  of  bacteria.  A  greater  increase  was 
obtained  in  the  acid  soils  than  in  the  neutral  Miami  silt  loam.  These 
carbonates  increased  the  number  of  bacteria  in  both  sterilized  and  un- 
sterilized soil.  The  sterilized  soil  was  inoculated  with  an  organism  in 
pure  culture.  Not  only  was  the  number  of  bacteria  in  the  soils  increased 
by  applications  of  magnesium  and  calcium  carbonates,  but  an  increase 
in  activity  of  the  various  groups  of  bacteria  was  shown. 

The  form  of  carbonate  which  gave  the  best  results  was  magnesium 
carbonate.  There  was  an  enormous  increase  in  the  number  of  bacteria 
produced  by  the  magnesium  carbonate  over  that  of  calcium  carbonate. 
Ground  dolomitic  limestone  did  not  prove  quite  as  efficient  as  calcium 


498  Journal  of  Agricultural  Research  voi. xir,  no.  s 

carbonate  in  increasing  the  number  of  soil  bacteria.  This  variation  can 
possibly  be  explained  by  the  difference  in  solubility  of  these  compounds. 
It  was  shown  (30)  that  magnesium  carbonate  is  more  soluble  in  carbo- 
nated water  than  is  calcium  carbonate,  and  that  calcium  carbonate  is 
more  soluble  than  limestone.  Dolomitic  limestone,  however,  is  less 
soluble  than  the  nondolomitic.  This  order  of  solubility  is  in  agreement 
with  the  order  in  which  the  bacteria  of  the  soil  responded  to  treatment 
with  these  different  compounds. 

It  does  not  seem  correct  to  say  that  the  entire  influence  which  these 
compounds  had  on  the  soil  bacteria  was  due  to  the  neutralization  of  the 
soil  acids.  When  neutral  soil  was  treated  with  magnesium  carbonate 
or  with  calcium  carbonate,  the  number  of  bacteria  was  increased,  espe- 
cially in  the  case  of  magnesium  carbonate.  Both  in  neutral  and  in  acid 
soil,  the  latter  doubly  neutralized,  magnesium  carbonate  increased  the 
number  of  Bacillus  azotobacter.  All  this  would  indicate  that  carbonates 
either  serve  in  part  as  a  stimulant  or  effect  an  indirect  action  on  other 
compounds  which  are  in  turn  rendered  more  soluble. 

Since  the  magnesium  and  calcium  carbonates  increased  the  number 
of  bacteria  in  acid  soil  when  added  in  small  amounts,  and  since  appli- 
cation in  acid  soil  of  these  compounds  gave  better  results  than  when 
applied  to  neutral  soil,  it  appears  that  the  greater  part  of  this  influence 
on  the  bacteria  was  due  to  neutralization. 

It  may  have  been  that  the  magnesium  carbonate  when  added  to  the 
soil  was  partially  converted  into  magnesium  phosphate.  Truog  (36) 
pointed  out  the  fact  that  for  the  amount  of  phosphate  used,  magnesium 
phosphate  increased  the  phosphorus  content  of  plants  more  than  did 
other  forms  of  phosphates.  If  this  be  the  case,  and  the  formation  of 
magnesium  phosphate  takes  place  in  soil  when  magnesium  carbonate  is 
applied,  then  magnesium  chlorid  and  magnesium  phosphate  should  give 
an  increase  in  the  number  of  bacteria.  The  results  from  the  magnesium- 
chlorid  treatment  did  not  prove  beneficial,  however,  even  in  so  small 
amounts;  the  chlorin  radical  may  have  been  toxic.  Magnesium  phos- 
phate did  not  prove  favorable  to  the  reproduction  of  bacteria  in  acid  soil 
but  did  in  neutral  soil.  In  the  latter  soil,  the  effect  of  magnesium  phos- 
phate was  more  beneficial  than  the  effect  of  calcium  phosphate.  From 
this  evidence  it  appears  that  magnesium  phosphate  in  a  soil  favorable 
for  the  development  of  bacteria  is  a  stimulant  to  the  growth  of  bacteria. 
This  action  of  magnesium  phosphate  on  bacteria  in  soil  may  in  part 
account  for  the  enormous  influence  which  magnesium  carbonate  exerts 
in  increasing  the  number  of  bacteria  in  both  neutral  and  acid  soils. 

From  all  the  data  obtained  in  the  various  experiments  performed, 
magnesium  carbonate  appears  to  play  an  important  part  in  the  devel- 
opment of  soil  bacteria,  much  more  so  than  does  calcium  carbonate. 
Magnesium  phosphate,  when  applied  to  neutral  soil,  caused  an  increase 


Feb.  25. 1918  Influence  of  Carbonates  on  Soil  Bacteria  499 

in  the  number  of  bacteria.  In  the  neutral  soil  magnesium  phosphate 
proved  superior  to  calcium  phosphate  in  stimulating  the  bacteria;  in 
acid  soil  neither  compound  appeared  to  benefit  the  soil  flora. 

SUMMARY 

(i)  The  number  of  bacteria  in  acid  Colby  silt  loam  and  acid  Plain- 
field  sand  is  increased  by  the  applications  of  calcium  carbonate,  mag- 
nesium carbonate,  or  limestone. 

(2)  Magnesium  carbonate  increases  the  number  to  a  much  greater 
extent  than  does  either  calcium  carbonate  or  limestone. 

(3)  Monocalcium  phosphate  and  dibasic  magnesium  phosphate  slightly 
increase  the  total  number  of  bacteria  in  neutral  soil. 

(4)  Nitrification  is  benefited  by  limestone,  calcium-carbonate,  and 
magnesium-carbonate  treatment.  Magnesium  carbonate  in  soil  to 
which  no  nitrogenous  substance  was  added  favors  nitrate  accumulation 
more  than  does  either  calcium  carbonate  or  limestone.  The  phosphates 
increase  the  accumulation  of  nitrate  nitrogen  to  a  very  small  extent. 
When  gelatin  was  applied  to  the  soil,  magnesium  carbonate  did  not 
benefit  nitrification  any  more  than  calcium  carbonate  or  limestone. 

(5)  Ammonification  in  Colby  soil  is  benefited  by  all  three  forms  of 
the  carbonates,  while  in  Plainfield  sand  a  decrease  in  ammonia  is  shown. 
Monocalcium  phosphate  increases  ammonification  in  both  soils. 

(6)  Pure  cultures  of  Bacillus  tuinescens  and  B.  suhtilis  ammonify 
blood  meal  better  when  sterile  Colby  soil  is  treated  with  any  one  of  the 
three  forms  of  carbonates. 

(7)  A  culture  of  B.  azotobacter  in  the  two  soils  treated  with  the  car- 
bonates and  mannit  fails  to  show  an  increase  in  total  nitrogen  in  the 
sand  and  only  a  slight  gain  in  the  Colby  soil. 

(8)  Pure  cultures  of  B.  radicicola,  of  both  alfalfa  and  lupine  strains, 
and  B.  azotobacter  are  greatly  benefited  when  inoculated  into  sterile 
Colby  soil  previously  treated  with  magnesium  or  calcium  carbonate. 
Limestone  barely  increases  the  number  of  B.  azotobacter  in  Colby  soil. 
In  neutral  and  acid  soils  treated  with  magnesium  carbonate  until  the 
soils  were  strongly  alkaline,  B.  azotobacter  is  greatly  increased  in  munber 
over  that  of  the  untreated. 

(9)  From  the  data  considered  as  a  whole,  magnesium  carbonate  is 
superior  to  calcium  carbonate  or  limestone  in  stimulating  the  reproduc- 
tion of  bacteria  in  Colby  gilt  loam  and  Plainfield  soils.  In  general,  the 
smaller  applications  of  either  compound  give  better  results  than  do  the 
heavier  applications. 


500  Journal  of  Agricultural  Research  voi.  xii,  no.  s 

LITERATURE  CITED 

(i)  Allen,  E.  R.,  and  Bonazzi,  A. 

1915.  ON  nitrification:  preliminary  observations.     Ohio  Agr.  Exp.  Sta. 

Tech.  Bui.  7,  42  p.,  6  fig.     Bibliography,  p.  41-42. 

(2)  ASHBY,  S.  F. 

1907.   SOME   OBSERVATIONS   ON   THE   ASSIMILATION    OF   ATMOSPHERIC    NITROGEN 
BY      A     FREE     LIVING     SOIL     ORGANISM — AZOTOBACTER     CHROOCOCCUM 

BEiJERiNCK.     In  Jour.  Agr.  Sci.,  v.  2,  no.  i,  p.  35-51. 

(3)  Beckwith,  T.  D.,  Vass,  a.  F.,  and  Robinson,  R.  H. 

1914.  AMMONIFICATION  AND  NITRIFICATION  STUDIES  OF  CERTAIN  TYPES  OF  OREGON 

SOILS.     Ore.  Agr.  Exp.  Sta.  Bui.  118,  40  p.,  30  fig.     References,  p.  40. 

(4)  Brown,  P.  E. 

1911.  SOME  BACTERIOLOGICAL  EFFECTS  OF  LI^^NG.     lowa  Agr.  Exp.  Sta.  Re- 

search Bui.  2,  p.  49-107,  9  fig. 

(5)  — 

1912.  BACTERIOLOGICAL  STUDIES   OF  FIELD  SOILS.      I.   THE   EFFECTS   OF   LIMING. 

In  Centbl.  Bakt.  [etc.],  Abt.  2,  Bd.  35,  No.  11/13,  p.  234-24S. 

(6)  Cauda,  A. 

1916.  EXPERIMENTS   ON   THE   GROWTH  OF    AZOTOBACTER.      (Abstract.)     In   In- 

ternat.  Inst.  Agr.  [Rome]  Mo.  Bui.  Agr.  Intel,  and  Plant  Diseases, 
year  7,  no.  6,  p.  801-802.  19 16.  (Original  article  in  Staz.  Sper.  Agr. 
Ital.,  V.  49,  pt.  2,  p.  125-131.     1916.     Not  seen.) 

(7)  Chester,  F.  D. 

1901.  STUDIES  IN  SOIL  BACTERIOLOGY.     Del.  Agr.  Exp.  Sta.  13th  Ann.  Rpt., 
[i90o]/oi,  p.  50-73. 

(8)  ChristensEN,  H.  R. 

1915.  STUDIEN  UBER  DEN  EINFLUSS  DER   BODENBESCHAFFENHEIT  AUF  DAS  BAK- 

TERIENLE3EN  UND  DEN  STOFFUMSATZ  IM  erdbodEN.  In  Centbl.  Bakt. 
[etc.],  Abt.  2,  Bd.  43,  No.  1/7,  p.  1-166,20  fig.,  2  pi.  Literaturver. 
zeichnis,  p.  160-166. 

(9)  ^rid  Larsen,  O.  H. 

1911.   UNTERSUCHUNGEN  UBER  METHODEN  ZUR  BESTIMMUNG  DES  KALKBEDURF- 

NissES  DES  BODENS.  In  Centbl.  Bakt.  [etc.],  Abt.  2,  Bd.  29,  No.  9/1 1, 
p.  347-380. 

(10)  Conner,  S.  D. 

1916.  acid  soils  and  the  effect  of  acid  phosphate  and  other  fertilizers 
UPON  THEM.  In  Jour.  Indus,  and  Engin.  Chem.,  v.  8,  no.  i,  p.  35-40, 
2  fig. 

(11)  Darbishire,  F.  v.,  and  Russell,  E.  J. 

1907.  OXIDATION  IN  SOILS  AND  ITS  RELATION  TO  PRODUCTIVENESS.  THE  INFLU- 
ENCE OF  PARTIAL  STERILIZATION.  In  Jour.  Agr.  Sci.,  V.  2,  no.  3,  p. 
305-326,  3  fig. 

(12)  EhrEnbErg,  Paul. 

1904.   DIE  BAKTERIELLE  BODENUNTERSUCHUNG  IN  IHRER  BEDEUTUNG  PUR  DIE 

FESTSTELLUNG  DER  bodEnFruchtbarkeit.     In  Landw.  Jahrb.,  Bd. 
33,  Heft  I,  p.  1-139,  2  pi. 

(13)  EngbErding,  Diedrich. 

1909.  VERGLEICHENDE  UNTERSUCHUNGEN  UBER  DIE  BAKTERIENZAHL  IM 
ACKERBODEN   IN   IHRER    ABHANGIGKEIT   VON   AUSSEREN    EINFLUSSEN. 

In  Centbl.  Bakt.  [etc.],  Abt.  2,  Bd.  23,  No.  21/25,  p.  569-642,  4  fig. 


Feb.  2s.  i9i8  Influence  of  Carbonates  on  Soil  Bacteria  501 

(14)  Fabricius,  Otto,  and  Feilitzen,  Hjalmar  von. 

1905.    UEBER  DEN  GEHALT  AN  BAKTERIEN  IN  JUNGFRAULICHEM  UND  KULTIVIER- 
TEM  HOCHMOORBODEN   AUF  DIE   VERSUCHSFELDE   DEM  SCHWEDISCHEN 

MOORKULTURVEREiNS  BEi  FLAHULT.     In  Centbl.  Bakt.  [etc.],  Abt.  2, 
Bd.  14,  No.  6/7,  p.  161-168,  2  fig. 

(15)  Fischer,  Hugo. 

1905.    EIN     BEITRAG    ZUR    KENNTNIS     DER    LEBENSBEDINGUNGEN     VON    STICK- 

STOFFSAMMELNDEN  BAKTERIEN.     In  Jour.  Landw.,  Bd.  53,  Heft  i,  p. 
61-66. 

(16)  


1905.  EiN  BEITRAG  ZUR  KENNTNIS  DER  LEBENSBEDINGUNGEN  VON  STICK- 
STOFFSAMMELNDEN  BAKTERIEN.  In  Centbl.  Bakt.  [etc.],  Abt.  2,  Bd. 
14,  No.  2,  p.  33-34- 


(17)  

1905.   ZWEITER  BEITRAG  ZUR  KENNTNIS  DER  LEBENSBEDINGUNGEN  VON  STICK- 

STOFFSAMMELNDEN  BAKTERIEN.     In  JouT.  L,andw.,   Bd.   53,   Heft  3, 

p.  289-297. 

(18)   

1909.  UBER   DEN   EINFLUSS   DES   KALKES   AUF  DIE   BAKTERIEN   EINES   BODENS. 

In  Landw.  Vers.  Stat.,  Bd.  70,  Heft  5/6,  p.  335-342. 

(19)  FRAPS,   G.   S. 

1908.  THE  PRODUCTION  OF  ACTIVE  NITROGEN  IN  THE  SOIL.     Texas  Agr.  Exp. 
Sta.  Bui.  106,  31  p.,  5  fig. 

(20)  Fred,  E.  B.,  and  Graul,  E.  J. 

1916.   SOME  FACTORS  THAT  INFLUENCE  NITRATE  FORMATION  IN  ACID  SOILS.      In 

Soil  Sci.,  V.  I,  no.  4,  p.  317-338.     Literature  cited,  p.  337-338. 

(21)  Greaves,  J.  E. 

1916.   THE  INFLUENCE  OP  SALTS   ON  THE  BACTERIAL  ACTIVITIES   OF  THE  SOIL. 

In  Soil  Sci.,  v.  2,  no.  5,  p.  443-480,  4  fig.     Literature  cited,  p.  476-480. 

(22)  Hutchinson,  C.  M. 

1912.  studies  in  bacteriological  analysis  of  indian  soils.    in  mem. 

Dept.  Agr.  India,  Bact.  Ser.,  v.  i,  no.  i,  p.  1-65,  6  pi. 

(23)  Hutchinson,  H.  B. 

1913.  the  partial   STERILIZATION   OP  THE  SOIL   BY  MEANS   OP  CAUSTIC   LIME. 

In  Jour.  Agr.  Sci.,  v.  5,  pt.  3,  p.  320-330,  pi.  14. 

(24)  and  MacLennan,  Kenneth. 

1914.  THE  RELATIVE  EFFECT  OF  LIME  AS  OXIDE  AND  CARBONATE  ON  CERTAIN 

SOILS.     In  Jour.  Agr.  Sci.,  v.  6,  pt.  3,  p.  302-322,  2  fig.,  pi.  9-10.  Bib- 
liography, p.  322. 

(25) 

1915.  STUDIES  ON  THE  LIME  REQUIREMENTS  OF  CERTAIN  SOILS.      In  JoUT.  Agr. 

Sci.,  V.  7,  pt.  I,  p.  75-105,  4  fig.,  I  pi.     References,  p.  105. 

(26)  Jenkins,  E.  H.,  and  Britton,  W.  E. 

1900.    ON  THE  availability  OF  THE  NITROGEN  OF  HARD  RAW  BONE  AS  AFFECTED 

BY  APPLICATIONS  OF  SLAKED  LIME.     In  Conn.  Agr.  Exp.  Sta.,  23d 
Ann.  Rpt.,  1899,  p.  211-216,  i  pi. 

(27)  Kellerman,  K.  F.,  and  Robinson,  T.  R. 

1910.  LIME  AND  LEGUME  INOCULATION.     In  Scicnce,  n.  s.,  V.  32,  no.  813,  p. 

159-160. 

(28)  Kelley,  W.  p. 

1912.  the  effects  of  calcium  and  magnesium  carbonates  on  some  bio- 
LOGICAL TRANSFORMATIONS  OF  NITROGEN  IN  SOILS.     In  Univ.  Cal.  Pub. 

Agr.  Sci.,  v.  I,  p.  39-49. 


502  Journal  of  Agricultural  Research  voi.xii,  No.  8 

(29)  KELtEY,  W.  P. 

I915.   AMMONIFICATION  AND  NITRIFICATION  IN  HAWAHAN  SOILS.      Hawaii    AgT. 

Exp.  Sta.  Bui.  37,  52  p. 

(30)  and  McGeorge,  William. 

1913.  THE  EFFECT  OF  HEAT  ON  HAWAHAN  SOILS.     Hawaii  Agr.  Exp.  Sta.  Bui. 

30.  38  P- 

(31)  KocH,  Alfred. 

19U.    VERSUCHE    t>BER    DIE    SALPETERBILDUNG    IM    ACKERBODEN.        In    Jour. 

Landw.,  Bd.  59,  Heft  3,  p.  293-315. 

(32)  and  others. 

1907.    DIE     STICKSTOFFANREICHERUNG     DES     BODENS      DURCH     FREILEBENDE 
BAKTERIEN   UND  IHRE   BEDEUTUNG  FUR  DIE  PFLANZENNAHRUNG.      In 

Jour.  Landw.,  Bd.  55,  Heft  4,  p.  355-416,  pi.  11-13. 

(33)  KRtJGER,  W. 

1909.    EIN     BEITRAG    ZUR    UNTERSUCHUNG    DER    STICKSTOFPUMSETZUNGEN    IM 

BODEN.     Inaugural  dissertation,  Konigsberg,    1908.     (Abstract.)    In 
Centbl.  Bakt.  [etc.],  Abt.  2,  Bd.  23,  No.  6/9,  p.  236. 

(34)  Lemmermann,  O.,  and  Fischer,  Hugo. 

I911.   DIE    WIRKUNG    VON    MAGNESIA    UND    VON    KALK-MAGNESIAMISCHUNGEN 
AUF  DIE  BAKTERIELLE  TATIGKEIT  IM  BODEN.    In  Landw.  Jahrb.,  Bd. 
40,  Heft  1/2,  p.  244-254. 
(35) et  al. 

1909.   BAKTERIOLOGISCH-CHEMISCHE    UNTERSUCHUNGEN.     In    Landw.    Jahrb., 
Bd.  38,  Heft  2,  p.  319-364. 


(36) 


(37) 


19II.   UNTERSUCHUNGEN  tVBER  DAS  VERHALTEN  DES  AMMONIAKSTICKSTOFFS  IN 

GEKALKTEN  UND  UNGEKALKTEN  BODEN.     In  Landw.  Jahrb.,  Bd.  41, 
Heft  2,  p.  163-216. 


1911.   UNTERSUCHUNGEN    X)BER    DIE   ZERSETZUNG    DER    KOHLENSTOFFVERBIN- 
DUNGEN     VERSCHEIDENER     ORGANISCHER     SUBSTANZEN     IM     BODEN, 

SPEziELL  UNTER  DEM  EINFLUSS  VON  KALK.     In  Landw.  Jahrb.,  Bd. 
41,  Heft  2,  p.  217-256. 

(38)   LiPMAN,  C.  B. 

1910.    ON   THE  LACK   OF   ANTAGONISM   BETWEEN    CALCIUM   VERSUS   MAGNESIUM 
AND  ALSO  BETWEEN  CALCIUM   VERSUS  SODIUM.      In  Bot.   GaZ.,   V.   49, 

no.  I,  p.  41-50,  2  fig. 


(39)  

1916.    A    CRITIQUE    OP    THE     HYPOTHESIS    OF    THE    LIME-MAGNESIA    RATIO.    In 

Plant  World,  v.  19,  no.  4,  p.  83-105;  no.  5,  p.  119-135. 

(40)  and  Burgess,  P.  S. 

I914.   THE  PROTECTIVE  ACTION,  AGAINST  MgCOj,  OF  CaCOs  FOR  A.  CHROOCOCCUM. 

In  Jour.  Agr.  Sci.,  v.  6,  pt.  4,  p.  484-498. 

(41)  LiPMAN,  J.  G.,  and  Brown,  P.  E. 

1907.  LOSSES  OF  AMMONIA  PROM  CULTURE  SOLUTIONS.     In  Jour.  Amer.  Chem. 
Soc,  V.  29,  no.  9,  p.  1358-1362. 


(42) 


1907.   REPORT  OF  THE  SOIL  CHEMIST  AND  BACTERIOLOGIST.      In  N.  J.  Agr.  Exp. 

Sta.  20th  Ann.  Rpt.,  [i9o6]/o7,  p.  139-204,  3  pi. 
(43) arid  Owen,  I.  L. 

19IO.    RELATION  OF  LIME  TO  AMMONIFICATION  IN  SOILS.      N.  J.  Agr.   Exp.   Sta. 

31st  Ann.  Rpt.,  [1909J/10,  p.  114-119. 


(44) 


191 1.    EXPERIMENTS  ON  AMMONIA  AND  NITRATE  FORMATION  IN  SOILS.       II.      1% 

Centbl.  Bakt.  [etc.],  Abt.  2,  Bd.  30,  No.  7/12,  p.  156-181. 


Feb.  25. 1918  Influence  of  Carbonates  on  Soil  Bacteria  503 

(45)  LoEW,  Oscar. 

1913.  STUDIES  ON  ACID  SOILS  OP  PORTO  RICO.      Porto  Rico  AgT.  Exp.  Sta.  Btll. 

13, 23  p.,  I  fig. 

(46)  and  May,  D.  W. 

I901.  THE  RELATION  OP  LIME  AND  MAGNESIA  TO  PLANT  GROWTH.  U.  S.  Dept. 
AgT.  Bur.  Plant  Indus.  Bui.  i,  53  p.  3  pi. 

(47)  IvYON,  T.  L.,  and  BizzELL,  J.  A. 

1909.  SOME  CONDITIONS  FAVORING  NITRIFICATION  IN  SOILS.  In  Science,  n.  s. 
V.  30,  no.  778,  p.  773-774. 

(48)  Machida,  S. 

1905.    ON    THE    influence    OF    CALCIUM    AND    MAGNESIUM    SALTS    ON   CERTAIN 

bacterial  ACTIONS.     In  Japan  Imp.  Cent.  Exp.  Sta.  Bui.,  v.  i,  no.  i, 
p.  1-16,  I  pi. 

(49)  McBeth,  I.  G.,  and  Wright,  R.  C. 

1912.  CERTAIN  FACTORS  LIMITING  NITRIFICATION.  (Abstract.)  In  Science, 
n.  s.,  V.  35,  no.  897,  p.  392. 

(50)  MacIntire,  W.  H.,  Willis,  L.  G.,  and  Hardy,  J.  I. 

1914.  THE     NON-EXISTENCE     OF     MAGNESIUM     CARBONATE     IN     HUMID     SOILS 

Tenn.  Agr.  Exp.  Sta.  Bui.  107,  p.  151-202,  4  pi.     References,  p.  202, 

(51)  Miller,  F. 

1914.  tjber  den  einfluss  des  kalkes  auf  die  bodenbakterien.  in 
Ztschr.  Garungsphysiol.,  Bd.  4,  Heft  2,  p.  194-206. 

(52)  Owen,  W.  L. 

1908.  THE  EFFECT  OF  CARBONATES  UPON  NITRIFICATION.  Ga.  AgT.Exp.Sta. 
Bul.  81,  42  p.,  4  fig.,  4  pi. 

(53)  Paterson,  J.  W.,  and  Scott,  P.  R. 

1912.   INFLUENCE    of    certain    SOIL    CONSTITUENTS    UPON  NITRIFICATION.      In 

Jour.  Dept.  Agr.  Victoria,  v.  10,  pt.  7,  p.  393-400,  2  fig. 

(54)  Schreiner,  Oswald,  and  Lathrop,  E.  C. 

1912.  the  chemistry  of  steam  heated  soils,  /n  Jour.  Amer.  Chem.  Soc, 
V.  34,  no.  9,  p.  1242-1259,  I  fig. 

(55)  Sharp,  L.  T.,  and  Hoagland,  D.  R. 

I916.    acidity    and    adsorption    in    soils    as    MEASURED    BY    THE  HYDROGEN 

ELECTRODE.     In  Jour.  Agr.  Research,  v.  7,  no.  3,  p.  123-145,  i  fig. 
Literature  cited,  p.  143-145. 

(56)  Truog,  E. 

1916.  the  utilization  of  phosphates  by  agricultural  crops,  including 
a  new  theory  regarding  the  feeding  power  of  plants.  wis. 
Agr.  Exp.  Sta.  Research  Bul.  41,  50  p.,  12  fig.,  2  diagr. 

(57)  VOGEL,  J. 

I911.   AMMONIAK-      UND      SALPETERASSIMILATION     DURCH     MIKROORGANISMEN 

DES  BODENS.     In  Centbl.  Bakt.  [etc.],  Abt.  2,  Bd.  32,  No.  6/12,  p. 
169-179. 

(58)  VoorhEES,  E.  B.,  Lipman,  J.  G.,  and  Brown,  P.  E. 

1907.   SOME   chemical  and   bacteriological   EFFECTS  OF  LIMING.      N.  J.  Agr. 

Exp.  Sta.  Bul.  210,  79  p. 

(59)  Weis,  F.,  and  Bornebusch,  C.  H. 

1914.  on  the  PRESENCE  OF  AZOTOBACTER  IN  DANISH  WOODS  AND  ON  THE 
VALUE  OP  AZOTOBACTER  CLOTURES  FOR  THE  DETERMINATION  OP 
THE  LIME  REQUIREMENTS  IN  WOODLAND.  (Abstract.)  In  Exp. 
Sta.  Rec,  V.  34,  no.  9,  p.  814-815.  1916.  (Original  article  in  Forstl. 
Forsogsv.    Danmark,  Bd.  4,  No.  4,  p.  814-815.     1914.    Not  seen.) 


504  Journal  of  Agricultural  Research  voi.  xii,  no.  s 

(60)  Wheeler,  H.  J.,  Sargent,  C.  L.,  and  Hartwell,  B.  L. 

1899.  THE  AMOUNT  OP  HUMUS   IN  SOILS   AND  THE  PERCENTAGE  OP   NITROGEN 

IN     THE     HUMUS,      AS     AFFECTED     BY     APPLICATIONS     OP  AIR-SLAKED 

LIME  AND  CERTAIN  OTHER  SUBSTANCES.     In  Jour.  Amef.  Chem.  Soc, 
V.  21,  no.  II,  p.  1032-1037. 

(61)  White,  J.  W. 

1915.  nitrification  in  relation  to  THE  REACTION  OP  THE  SOIL.    In  Penn. 
Agr.  Exp.  Sta.  Ann.  Rpt.,  1913/14,  p.  70-80,  4  pi.     References,  p.  80. 

(62)  Withers,  W.  A.,  and  Fraps,  G.  S. 

1900.  THE  RELATIVE  VALUE  OF  SOME  NITROGENOUS  FERTILIZERS.     N.  C.  Agr. 

Exp.  Sta.  Bui.  176,  p.  15-22. 

(63) 

1902.  NITRIFICATION  IN  DIFFERENT  SOILS.     In  Jour.  Amer.  Chem.  Soc,  v.  24, 
no.  6,  p.  528-534. 
(64)  WoHLTMANN,  F.,  FiscHER,  Hugo,  and  Schneider,  Philipp. 

1904.    BODENBAKTERIOLOGISCHE     UND     BODENCHEMISCHE    STUDIEN     AUS     DEM 

VERSUCHSFELDE.    In  Jour.  Landw.,  Bd.  52,  Heft  1/2,  p.  97-126,  i  pi. 


HUMUS  IN  MULCHED  BASINS,  RELATION  OF  HUMUS 
CONTENT  TO  ORANGE  PRODUCTION,  AND  EFFECT 
OF  MULCHES  ON  ORANGE  PRODUCTION 

By  Charles  A  Jensen, 

Assistant  in  Plant  Malnutrition,  Office  of  Biophysical  Investigations,  Bureau  of  Plant 

Industry,  United  States  Department  of  Agriculture 

INTRODUCTION 

The  formation  of  humus  in  a  soil  and  its  consen^ation  are  points  that 
are  usually  given  much  weight  in  discussions  of  soil  fertility.  The  free 
formation  of  humus  in  a  soil  from  organic  matter  added  to  it  is  usiially 
held  to  indicate  that  the  soil  its  biologically  active,  and  proper  biological 
activity  in  a  soil  is  conceded  to  be  necessary  under  the  usual  agricul- 
tural practices.  Hence,  whatever  the  absolute  value  of  soil  humus  may 
be  as  a  factor  in  crop  production,  it  is  ordinarily  conceded,  other  fac- 
tors being  equal,  that  a  soil  which  readily  forms  humus  when  organic 
matter  is  added  is  superior  to  one  that  does  not. 

Previous  work  (j)^  by  this  Office  in  southern  California  has  shown 
that  the  mottling  of  Citrus  leaves  varied  inversely  with  the  humus  con- 
tent in  the  soil.  In  other  work  by  the  same  writers  (2)  it  was  shown 
that  on  certain  soil  types  in  the  general  region  of  Riverside,  California, 
the  use  of  the  mulched-basin  system  in  Citrus  groves  improved  the  tree 
conditions  more  than  the  usual  system  of  cultivation  and  furrow  irriga- 
tion, and  that  different  green-manure  substances  when  used  as  a  mulch 
affected  tree  growth  and  fruit  setting  in  varying  degrees. 

It  is  therefore  of  practical  value  to  obtain  information  about  the  rate 
of  humus  formation  from  various  organic  substances  when  employed 
either  as  a  mulch  or  when  worked  into  the  soil,  and  to  ascertain,  if  pos- 
sible, whether  the  increase  in  humus  is  correlated  with  increase  in  tree 
growth  and  fruit  setting. 

This  work  reports  the  study  of  humus  formation  in  mulched  basins  in 
Citrus  groves,  and  the  effect  of  different  mulching  materials  on  fruit 
production  and  tree  growth.     The  work  was  done  at  Riverside. 

The  term  "humus"  as  used  in  this  paper  is  the  brown-colored  organic 
matter  extracted  from  a  soil  by  boiling  it  for  two  minutes  in  a  7.5  per  cent 
sodium-hydrate  solution,  after  the  removal  of  calcium  from  the  soil  with 
I  per  cent  hydrochloric  acid. 

The  percentage  of  humus  was  determined  colorimetrically  by  com- 
paring the  intensity  of  the  color  of  the  soil  extract  thus  obtained  with  the 
intensity  of  a  standard  humus  solution  prepared  from  the  humus  ex- 

'  Reference  is  made  by  number  (italic)  to  "  Literature  cited, "  pp.  sij~si& 

Journal  of  Agricultural  Research,  Vol.  XII,  No.  8 

Washington,  D.  C  Feb.  75,  191S 

md  Key  No.  G136 

(S05) 


5o6  Journal  of  A  gricultural  Research  voi.  xii.  no.  s 

tracted  from  peat.  In  the  great  number  of  humus  determinations  made 
by  the  Office  of  Biophysical  Investigations  in  the  Citrus  areas  studied  in 
California  not  a  single  humus  extract  obtained  has  been  off-color,  when 
compared  with  the  standard  solutions  prepared  for  comparison. 

Gortner  (4)  obtained  a  black  soil  pigment  from  the  soils  he  studied  by 
extracting  them  successively  with  4  per  cent  sodium  hydroxid,  i  per  cent 
hydrochloric  acid,  4  per  cent  sodium  hydroxid  followed  by  water.  The 
soil  residue  obtained  from  the  last  sodium-hydroxid  extraction,  when 
shaken  with  water  in  quantity,  yielded  the  black  pigment. 

Soils  from  various  Citrus  areas  studied  in  this  work  did  not  produce 
black  pigment  in  appreciable  amounts  when  they  had  been  extracted 
with  hydrochloric  acid  to  the  absence  of  calcium,  boiled  for  two  minutes 
in  a  7.5  per  cent  sodium-hydroxid  solution,  and  washed  with  hot  dilute 
sodium-hydroxid  solution  on  the  filter  till  the  "humus  color"  had  been 
removed.  It  might  be  noted  that  a  boiling  5  per  cent  solution  of  sodium 
hydroxid  did  not  remove  all  the  color  from  the  soils,  and  that  a  10  per 
cent  solution  removed  no  more  than  a  7.5  per  cent  solution. 

Following  Gortner's  procedure,  black  pigments  were  obtained  from 
some  of  the  soils  investigated  in  the  Citrus  areas  of  southern  California. 

It  should  be  noted  that  the  colorimetric  reading  should  be  made  as 
soon  as  the  humus  extract  is  obtained,  as  the  color  partially  fades  out 
on  standing. 

As  no  evidence  was  obtained  to  indicate  that  any  coloring  matter  was 
left  in  the  soil  after  the  extraction  with  boiling  sodium  hydroxid  of  the 
strength  noted,  and  as  the  color  tint  was  always  the  same  as  that  of  the 
standard  used,  it  seems  that  this  rapid  method  for  humus  determination 
is  reliable  for  comparative  studies,  especially  when  used  with  soils  in 
the  same  general  area. 

Another  objection  to  the  colorimetric  method  is  that  undecomposed 
organic  matter,  like  dry  alfalfa,  sweet  clover,  etc.,  produces  a  color  when 
treated  in  the  manner  just  described  for  mating  humus  determinations. 
This  objection  is  also  brought  out  by  Gortner  (4).  This  matter  was 
looked  into  when  the  method  was  worked  out.  When  dry  hay  susbtances 
were  extracted  \\dth  hydrochloric  acid  to  the  absence  of  calcium,  the 
chlorin  washed  out,  and  the  residue  boiled  for  several  minutes  with  a 
7.5  per  cent  solution  of  sodium  hydroxid,  the  organic  materials  yielded 
a  yellow-colored  solution  entirely  different  in  color  from  the  brown  to 
black  humus  color. 

A  number  of  readings  of  these  organic  extracts  gave  a  density  of  color 
corresponding  to  about  0.0012  per  cent  humus  in  soil,  if  it  is  assumed 
that  I  per  cent  of  the  substances  was  to  be  added  to  the  soil.  This  is 
about  the  working  error  in  making  humus  determinations  by  the  method 
used,  as  will  be  seen  from  Table  II,  and  such  errors  could  have  no  influ- 
ence on  comparative  results  in  the  kind  of  study  here  reported.  Indeed 
there  could  never  have  been  much  of  this  off-color  in  the  extracts,  as  a 


Feb.  25, 1918  Humus  in  Mulched  Basins  507 

uniform  tint  was  always  obtained  from  the  soils  studied  in  making  humus 
extracts  from  them.  The  straw  color  produced  by  the  extraction  of  soil 
with  I  to  2  per  cent  of  undecomposed  organic  matter  added  was  found 
to  be  so  thoroughly  obscured  by  the  humus  color  of  the  soil  that  its  pres- 
ence did  not  interfere  with  the  colorimetric  reading. 

It  is  well  recognized  that  when  hay  material  of  many  kinds  is  digested 
with  hot  hydrochloric  acid  and  extracted  with  ammonia,  a  solution  is 
obtained  which  in  appearance  and  color  can  not  be  distinguished  from 
a  humus  extract  from  a  soil. 

Much  of  the  literature  relating  to  humus  is  discussed  by  Schreiner  and 
Shorey  (/o). 

HUMUS  FORMATION  IN  MULCHED  BASINS 

Experiment  I. — Mulched  basins,  15  feet  in  diameter,  were  installed 
in  an  orange  grove  near  Riverside,  Cal.,  in  March,  191 5.  The  soil  is  a 
red  clay  loam,  derived  from  granite,  which  underlies  the  soil  at  a  depth 
of  3  to  4  feet.  One  half  of  the  basins  were  mulched  with  about  180 
pounds  of  alfalfa  (Medicago  saliva)  each  and  the  other  half  with  about 
20  cubic  feet  of  good  cow  manure  each.  One  half  of  the  basins  in  each 
mulching  series  received  200  pounds  of  lime  dust  each.  This  lime  was  a 
by-product  from  the  flues  of  a  neighboring  cement  plant.  It  was  com- 
posed of  about  45  per  cent  of  calcium  carbonate,  about  20  per  cent  of 
calcium  hydrate,  about  10  per  cent  of  calcium  sulphate,  about  i  per  cent 
of  potash,  and  a  little  phosphoric  acid.  The  balance  was  mostly  silicious 
material. 

During  Augus^.  of  the  same  year  another  row  of  orange  trees  in  the  same 
grove  was  basined  and  mulched  with  alfalfa.  These  basins  were  smaller 
than  those  just  referred  to,  and  less  mulch  was  used.  Part  of  these 
basins  received  about  100  pounds  each  of  ground  hme  rock  analyzing 
about  90  per  cent  calcium  carbonate.  The  rest  of  the  basins  in  the  row 
were  unlimed. 

All  surface  organic  matter  was  carefully  removed,  and  soil  samples 
were  taken  to  a  depth  of  3  feet  in  these  basins,  usually  from  three  basins 
in  each  experiment,  the  samples  being  separated  into  foot  sections. 
Corresponding  foot  samples  from  the  three  basins  were  composited,  and 
each  composite  sample  was  analyzed  in  duplicate.  The  percentage 
reported  in  the  following  tables  are  the  averages  to  a  depth  of  3  feet. 
Usually  the  samples  were  collected  just  before  irrigation. 

Table  I  shows  the  average  percentage  of  humus  from  time  to  time  to  a 
depth  of  3  feet  of  soil  in  the  basins  under  the  various  treatments  during  a 
period  of  from  12  to  17  months. 

The  low  humus  percentage  is  due  in  part  to  the  removal  of  several 
inches  of  the  surface  soil  in  constructing  the  basins,  which  suggests  inci- 
dentally that  in  making  basins  as  little  surface  soil  as  possible  should 
be  removed. 


5o8 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  8 


Table  I. — Percentage  of  humus  to  a  depth  of  j  feet  in  mulched  basins  in  orange  groves 
on  clay  loam,  soil.     Experiment  I 


Date. 


1915- 

Mar.  31 

Aug.  13 

Sept.  13 

Oct.  12 

Nov.  II 

Nov.  20 

Dec.  II 

Average  for  191  s.* 

1916. 

Mar.  17 

Apr.  17 

May  15 

June  16 

July  13 

Ai»g.  7 

Average  for  1916. . 
Ratio  1916  to  1915 


Basin  treatment  and  percentage  of  htunus. 


Large  basins  installed  in  March. 


Alfalfa 
alone. 


3-075 
•073 
•075 
.084 
.098 
.187 
.  iq8 


"3 


•113 
.  Ill 
.  076 
.  112 
.081 
.185 


113 


Manure 
alone. 


118 
130 
136 

153 
186 
188 
180 


.156 


201 
246 

133 
231 

295 
292 


233 


1.49 


Alfalfa 
and  lime. 


D.  113 
.  106 
.147 
•155 
.185 
.150 
.188 


149 


180 
287 
,181 
288 
272 
215 


237 


1-59 


Manure 
and  lime. 


O.  129 
.  112 

•139 
.  160 

•151 
•151 
.156 


143 


179 
213 
141 
220 
207 
227 


1.38 


Small   basins   in- 
stalled in  August. 


Alfalfa 
alone. 


O.  166 
.  114 

.  118 


130 


171 
257 
^33 
165 
117 
212 


176 


1-35 


Alfalfa 
and  lime. 


0.16s 
.  092 
.  108 


096 


"S 


259 

107 

139 

III 

175 


151 


I- 31 


It  will  be  observed  that  the  determinations  show  a  fluctuation  in  the 
percentage  of  humus  from  time  to  time.  This  is  doubtless  due  partly 
to  the  difficulty  of  getting  uniform  soil  samples.  This  factor,  however, 
does  not  afford  a  complete  explanation,  because  the  same  kind  of  fluc- 
tuation was  noted  when  determinations  were  made  on  soils  kept  in  pots 
in  the  laboratory,  where  better  control  conditions  obtained  and  where 
leaching  was  avoided.  Neither  can  the  fluctuations  be  due  entirely  to 
the  working  error  in  making  humus  determinations,  since  these  errors 
are  of  smaller  magnitude  than  the  variations  in  the  humus  content  of 
the  soils. 

The  following  duplicate  determinations  made  on  March  31,  1915,  the 
averages  of  which  are  given  on  the  first  line  in  Table  I,  illustrate  the 
working  error  in  making  humus  determinations  by  the  method  employed 
(Table  II). 

Heinze  (6)  states  that  after  humus  has  been  formed  other  bacteria, 
such  as  Azotobacter,  commence  to  decompose  it.  McBeth  (9)  lays  stress 
on  the  fact  that  the  cellulose  destroying  organisms  in  the  soil  break  down 
the  organic  matter  in  the  soils  with  the  formation  of  humus,  and  that  the 
nitrifying  organisms  break  the  humus  up  into  still  simpler  compounds 
through  their  nitrifying  activities. 


Feb.  t5,  X918 


Humus  in  Mulched  Basins 


509 


Table  II. — Working  error  in  making  humus  determ,inations 


Soil  No. 


677 
677 
678 
678 
679 
679 
670 
670 
671 
671 
672 
672 


Per  cent. 

0.358 

•358 

.  041 

•039 
•045 
.  046 

•  250 

•  250 
.030 
.028 
.048 
.  046 


vSoil  No. 


676 
676 
677 
677 
678 
678 
682 
682 
683 
683 
684 
684 


Per  cent. 
o.  227 

.  227 

.051 

.  050 

.023 

.  022 
.  192 
.  200 

.056 
.051 

.  022 
.  021 


Soil  No. 


685 
685 
686 
686 
687 
687 
688 
688 
689 
689 
690 
690 


Humus. 


Per  cent, 
o.  284 
.284 
.081 
.081 
.050 
.050 
.284 
.384 
.074 
.  070 
•043 
.  04a 


From  these  considerations  it  might  therefore  be  expected  that  the 
amount  of  humus  in  the  soil  in  the  basins  would  be  subject  to  fluctua- 
tion, the  amount  present  at  any  one  time  depending  upon  the  ascendency 
of  one  or  the  other  of  these  important  groups  of  soil  organisms. 

The  fluctuation  in  the  percentage  of  humus  makes  it  impracticable  to 
attempt  to  determine  the  rate  at  which  humus  was  formed  between 
specific  dates.  It  will  be  seen  in  Table  I  that  there  is  frequently  a  de- 
crease in  the  percentage  of  humus  from  one  period  to  another,  especially 
during  the  second  year,  a  point  that  might  possibly  apply  to  determina- 
tions made  by  Gortner  (5).  He  added  silk,  wool,  flour,  and  alfalfa  meal 
to  soils,  kept  the  materials  in  earthenware  jars,  and  at  the  end  of  a  year 
made  humus  determinations.  In  most  instances  he  found  less  humus 
at  the  end  of  the  year  than  he  found  in  the  treated  soils  when  the  experi- 
ment was  started.  It  would  have  been  of  much  interest  if  these  deter- 
minations had  been  made  frequently  during  the  year.  None  of  the 
losses  of  humus  reported  by  him  are  greater  than  was  sometimes  found 
in  a  month's  time  in  humification  studies  carried  out  under  laboratory 
control  in  this  Office,  to  be  reported  later. 

A  comparison  of  the  amount  of  humus  in  the  basins  from  year  to  year 
is  made  by  averaging  the  percentage  found  during  each  of  the  years 
1915  and  1916,  as  shown  in  Table  I.  At  the  bottom  of  the  table  will  be 
found  the  ratio  of  the  average  percentage  of  humus  in  1916  to  that  in 

1915- 

Experiment  II. — In  April,  191 5,  a  block  of  orange  trees  in  another 
grove  on  sandy  loarri  soil  was  basined.  The  basins  were  about  12  feet 
in  diameter,  and  the  treatments  were  the  same  as  in  Experiment  I, 
except  that  less  lime  was  used. 

The  results  of  the  humus  determinations  are  given  in  Table  III,  which 
shows  the  fluctuation  in  the  amount  of  humus  from  time  to  time.  The 
ratio  of  the  average  percentage  of  humus  in  the  basins  in  191 6  to  that 
in  1 91 5  is  given  in  the  bottom  line  in  the  table. 


5IO 


Journal  of  Agricultural  Research 


Vol.  Xn,  No.  8 


Table  III. — Percentage  of  humus  to  depth  of  ;j  feet  in  mulched  basins  in  orange  groves  on 
sandy  loam  soil.    Experiment  II 


Date. 


1915 

May  3 

Aug.  9 

Sept.  4 

Oct.  4 

Oct.  30 

Nov.  30 

Average  for  191 5. ... 

1916. 

Mar.  17 

Apr.  17 

May  15 

June  16 

July  13 

Aug.  7 

Average  for  1916 

Ratio,  1916  to  1915. . 


Basin  treatment  and  jjercentage  of  humus. 


Alfalfa       Manure       Alfalfa       Manure 
alone.  alone.       and  lime,    and  lime. 


O.  169 
.128 
.  129 
•  230 
.203 
.185 


174 


251 
227 
129 
216 

195 

175 


199 


I.  14 


o.  159 

.  121 
.  146 
.218 
.  209 
•159 


169 


241 
220 
180 
213 
167 
296 


1.30 


D.  214 
.  126 
.187 
.283 

•255 
.181 


208 


.301 
.  296 
.  163 

•309 
.244 

•243 


259 


1.24 


o.  196 

.  164 
.  189 
.  260 

•307 
.  213 


221 


286 
338 
.  180 


248 
309 


272 


23 


Experiment  III. — In  October,  191 5,  a  block  of  60  trees  in  another 
orange  grove  of  heavy  sandy  loam  soil  was  basined  and  mulched.  The 
basins  were  about  6  by  20  feet.  The  mulching  materials  used,  the 
quantity  of  each,  and  the  percentage  of  humus  in  the  soil  are  given  in 
Table  IV.  The  percentages  are  averages  to  a  depth  of  3  feet.  The 
amount  of  bean  straw  used  was  not  enough  to  cover  the  soil,  and  so  did 
not  make  an  effective  mulch.  Some  of  these  basins  received  in  addition 
to  the  mulching  material,  100  pounds  ground  lime  rock,  analyzing  about 
90  per  cent  calcium  carbonate. 

Table  IV. — Average  percentage  of  humus  to  a  depth  of  3  feet  in  mulched  basins  in  an 
orange  grove  on  heavy  sandy  loam,.     Experiment  III 


Percentage  of  humus. 

Humus 

ratio — 

1916  to 

1915- 

Quantity  of  mulch  and  lime 

Oct.  I  J, 
1915. 

June  27, 
1916. 

per  basin. 

Alfalfa  alone 

0.223 
.  160 

.243 
.241 
.  264 
.244 
•319 
•235 
.254 
•237 
.279 
.266 
.265 

0.316 
.185 
.  220 
.  240 
.  226 
.163 
.247 
.194 
.161 
.215 
.203 
.279 
.  206 

I.  41 
I.  16 
.91 
I.  00 
.86 
.67 
.78 
.83 
•63 
.91 

•73 
1.05 

.78 

125  pounds. 

Alfalfa  and  lime 

100  pounds  of  lime. 

Bean  straw  alone 

70  pounds. 

100  pounds  of  lime. 

Bean  straw  and  lime 

Manure  alone 

20  cubic  feet. 

Manure  and  lime 

100  pounds  of  lime. 

Barley  hay  alone 

125  pounds. 

Barley  liay  and  lime 

100  pounds  of  lime. 

Sweet  clover  alone 

125  pounds. 

Sweet  clover  and  lime 

100  potmds  of  lime. 

Bur  clover  alone .        

125  pounds. 
125  pounds. 
100  poimds  of  lime. 

Pine  shavings  alone 

Pine  shavings  and  lime 

Feb.  25,  1918 


Humus  in  Mulched  Basins 


511 


The  basins  with  alfalfa  as  the  mulch  were  the  only  ones  which  increased 
in  percentage  of  humus  during  the  interval  given.  Other  experiments 
have  shown  that  bean  straw  readily  humifies.  It  is  possible  that  if  this 
had  been  applied  in  as  large  a  quantity  as  the  other  substances,  these 
basins  would  have  increased  in  humus  also.  The  lime  was  not  consistent 
in  its  effect  on  the  amount  of  humus  formed. 

It  is  to  be  noted,  however,  that  the  percentage  of  humus  as  found  in 
the  summer,  after  the  application  of  the  mulches  in  the  preceding  fall, 
does  not  necessarily  indicate  that  in  most  of  these  basins  the  average 
amount  of  humus  had  actually  decreased.  As  shown  in  Tables  I  and  III, 
there  are  periodical  fluctuations  in  the  humus  content,  and  the  second  set 
of  determinations  shown  in  Table  IV  may  have  been  made  at  a  time  when 
the  amount  of  humus  was  comparatively  small.  The  figures  in  Table  IV 
probably  give  an  indication  of  the  relative  humifying  activity  in  the 
basins.     Each  percentage  given  is  an  average  of  about  six  determinations. 

Experiment  IV. — A  series  of  basins  was  installed  in  a  lemon  grove 
near  Corona,  Cal.,  in  October,  1915.  The  soil  is  sandy  loam  in  texture 
interspersed  freely  with  gravel  so  that  samples  below  3  feet  can  not  be 
taken  with  the  ordinary  soil  tube. 

The  mulching  materials  used  were  alfalfa  and  manure  and,  in  addition 
to  the  organic  mulch,  some  of  the  basins  received  other  artificial  ma- 
terials— ^viz,  blood,  phosphate,  tankage,  sulphur,  bone  meal,  and  lime. 
In  addition  to  the  basined  rows,  several  control  rows  were  retained,  and 
were  manured,  irrigated,  and  cultivated  in  the  usual  way.  The  amount 
of  manure  applied  to  the  trees  in  the  control  rows  was  the  same  as  appUed 
in  the  basins. 

Soil  samples  were  collected  after  the  basins  were  made,  on  October  20, 
1915,  and  another  set  was  taken  on  June  29,  1916.  The  detailed  humus 
determinations  made  on  the  latter  date  did  not  indicate  that  any  of  the 
artificial  substances  added  to  the  mulches  produced  any  definite  influence 
on  the  humus  content.  The  average  percentage  of  humus  to  a  depth  of 
3  feet  in  the  alfalfa  basins,  manure  basins,  and  in  the  control  rows  is 
given  in  Table  V.  Each  percentage  figure  given  in  the  table  is  made  up 
of  an  average  of  about  21  determinations,  each  determination  repre- 
senting a  composite  of  at  least  three  samples. 

Table  V. — Average  percentage  of  humus  to  a  depth  of  3  feet  in  unmulched  soil  and  in 
mulched  basins  in  a  lemon  grove  on  light  sandy  loatn  soil.     Experiment  IV 


Percentage  of  humus. 

Ratio 

Soil  treatment. 

Oct.  20,       June  29, 
1915.             1916. 

1916  to 
1915- 

Basins  with  alfalfa  mulch 

0.  360 

•345 
•551 

0.306 
.361 

.380 

0.85 

1.05 

.60 

Basins  with  manure  mulch 

Furrow  irrigation  and  siu-face  cultivation ;  manured 

27811°— 18- 


^12  Journal  of  Agricultural  Research  voi.  xii,  No.  8 

The  low  initial  percentage  of  humus  in  the  basins  is  partly  due  to  the 
fact  that  5  or  6  inches  of  the  surface  soil  were  removed  in  making  the 
basins,  owing  to  sloping  ground. 

The  percentage  of  humus  in  the  manured  basins  in  191 6  was  about  the 
same  as  in  191 5,  while  in  the  alfalfa  basins  it  was  less.  The  decrease  in 
the  humus  content  in  the  manured  soil  furrow-irrigated  and  surface- 
cultivated  is  very  appreciable,  indicating  that  this  system  did  not  con- 
serve the  humus  as  well  as  the  basin  system. 

The  grove  in  which  this  experiment  was  conducted  was  deteriorating 
along  with  other  groves  in  the  neighborhood  when  the  basins  were  in- 
stalled. The  whole  grove  was  in  better  condition  in  October,  191 6,  but 
there  was  no  apparent  difference  in  the  appearance  of  or  the  amount  of 
fruit  on  the  trees  basined  as  compared  with  trees  not  basined.  The 
mulch-basin  system  in  experiments  conducted  by  this  office  has  always 
produced  a  quicker  response  on  trees  on  clay  loam  soil  than  on  trees  on 
sandy  loam  or  light  sandy  loam  soil. 

The  deterioration  of  the  grove  above  mentioned  was  evidently  due  to 
bad  soil  conditions,  which  quite  likely  affected  the  bacterial  flora  in  such 
a  way  as  to  make  it  less  efficient  in  converting  the  alfalfa  into  humus. 

In  most  cases  presented  above  the  percentage  increase  in  humus  during 
a  period  of  from  9  to  1 7  months  was  greater  in  the  basins  on  the  clay  loam 
soil  than  in  those  on  the  sandy  loam  soil,  with  both  manure  and  alfalfa 
as  the  mulch.  On  the  whole,  the  manured  basins  gained  slightly  more 
in  humus  content  on  both  types  of  soil  than  the  alfalfa-mulched  basins. 
No  correction  has  been  made  for  the  amount  of  humus  added  with  the 
manure.  This  material  contained  about  5  per  cent  of  humus  when  ap- 
plied, part  of  which  undoubtedly  found  its  way  into  the  surface  soil  when 
the  basins  were  irrigated.  The  percentage  increase  of  humus  in  the 
manured  basins  therefore  does  not  necessarily  indicate  that  the  manure 
had  been  humified  to  a  greater  extent  than  alfalfa.  Data  to  be  presented 
in  another  paper  would  indicate  that  the  increase  is  likely  due  to  the 
humus  added  with  the  manure. 

In  all  cases  where  lime  was  added  to  manure  in  basins  the  increase  in 
humus  was  somewhat  less  than  when  manure  alone  was  used.  In  most 
cases  the  addition  of  lime  to  alfalfa  in  basins  produced  slightly  more 
humus  than  when  alfalfa  was  added  alone.  Frear  and  Hess  (j)  found 
that  field  plots  receiving  manure  and  lime  contained  less  active  humus 
than  unmanured  plots. 

The  amount  of  alfalfa  or  manure  necessary  to  form  a  unit  amount  of 
humus  can  not  be  stated  with  exactness  from  the  data  available.  In 
Experiments  I  and  II  the  average  increase  in  humus  in  the  basins  in 
1 91 6  over  the  average  amount  present  in  191 5  would  indicate  that  i  part 
of  humus  was  formed  from  about  10  to  12  parts  of  alfalfa,  and  i  part  of 
humus  from  about  26  parts  of  manure.     These  calculations  agree  fairly 


Feb.  2s,  1918 


Humus  in  Mulched  Basins 


513 


well  with  results  stated  by  others  (7,  p.  128-129),  but  are  necessarily 
only  estimates. 

The  data  obtained  in  the  humus  studies  in  mulched  basins  did  not 
indicate  that  there  was  any  very  appreciable  accumulation  of  humus  in 
the  lower  soil  depths  due  to  leaching  from  the  surface  foot  of  soil.  The 
detailed  data  showed  that  most  of  the  change  in  humus  content  took 
place  in  the  surface  foot. 

RELATION   OF   HUMUS    CONTENT    OF   THE   SOIL   TO   ORANGE 

PRODUCTION 

Experiment  I. — Picking  records  of  the  earlier  basined  trees  in  Experi- 
ment I  were  obtained  from  the  company  on  whose  grove  the  experiment 
was  conducted,  and  are  given  in  Table  VI. 

Table  VI. — Effect  of  different  mulching  materials  on  orange  production.     Experiment  I. 

Picked  May,  igij 


Basin  treatment. 

Average  number  of  boxes  of  oianges 
per  tree. 

191 7  yield 
corrected  on 
basis  of  pre- 
vious yields 

(boxes  per 
tree). 

Row 
No. 

Tree  nimiben 

1914 

I9I5               !               1917 

in  row. 

Alfalfa  alone 

I.  49 
I.  19 
1.78 
1.32 

0.71 

I.    00 
I.    02 

3-31 
3.61 
2.  16 
I.  81 

4-97 
6.69 
2.  56 

2-55 

3 

3 
3 
3 

I  to    8 

Alfalfa  and  lime 

Manure  alone 

26  to  11 

9  to  15 

16  to  25 

Manure  and  lime 

This  appreciable  difference  in  fruit  production  of  trees  mulched  with 
alfalfa  and  manure,  respectively,  does  not  correlate  with  the  humus  con- 
tent in  the  soil  under  the  respective  mulches,  as  may  be  seen  from  Table  I. 
The  manured  basins  average  a  slightly  higher  percentage  of  humus  than 
the  basins  mulched  with  alfalfa. 

It  will  also  be  noticed  in  Table  VI  that  the  trees  basined  and  mulched 
with  manure  in  March,  191 5,  had  previously  produced  slightly  more 
fruit  than  the  trees  mulched  with  alfalfa. 

When  the  oranges  were  picked  in  the  spring  of  1917,  the  color  of  the 
oranges  on  the  trees  mulched  with  alfalfa  was  discinctly  more  golden 
than  the  color  of  the  oranges  on  the  trees  mulched  with  manure. 

Experiment  II. — Individual  tree  picking  records  had  not  been  kept 
of  the  trees  in  the  grove  in  which  this  experiment  was  conducted.  Indi- 
vidual tree  picking  records  were  obtained  in  the  spring  of  191 7  from  the 
company  on  whose  grove  the  experiment  was  conducted.  In  this  experi- 
ment a  row  of  trees  was  left  unbasined  and  was  manured,  furrow- 
irrigated,  and  cultivated  in  the  usual  manner.  The  amount  of  manure 
per  tree  was  the  same  as  that  put  into  the  basins. 

The  fruit  production  of  the  trees  is  given  in  Table  VII.  The  number 
of  trees  used  in  each  set  of  mulched  trees  was  from  13  to  15. 


514 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  8 


Table  VII. — Influence  on  orange  production  of  different  mulches  in  basins  and  by  furrow 
irrigation  and  cultivation.     Experiment  II.     Picked  JuTie,  1917 


Soil  treatment. 


Average  for 

each  or;ianic 

substance 

with  and 

without 

lime. 


Basins  with  alfalfa  mulch  alone 

Basins  with  alfalfa  mulch  and  lime. . . 

Basins  with  manure  mulch  alone 

Basins  with  manure  mulch  and  lime.  . 
Cultivated  with  manure  alone,  disked 
Cultivated  with  manure  and  lime 


4.8 
4.0 

2-5 


It  is  not  known  that  the  trees  used  in  this  experiment  differed  in 
yields  previous  to  the  installation  of  this  experiment.  However,  as  no 
individual  tree  picking  records  had  been  kept  previous  to  the  experi- 
ment, the  comparative  yields  shown  in  Table  VII  do  not  carry  the  force 
they  would  if  they  could  have  been  compared  with  the  previous  perform- 
ance of  the  same  trees.  In  comparing  the  fruit  production  in  experi- 
ments I  and  III,  it  seems  safe  to  infer  that  the  results  obtained  in  Ex- 
periment II  were  indicative  of  the  effect  of  the  different  organic  mate- 
rials used. 

It  will  be  noticed  in  Table  III  that  the  percentage  of  humus  in  the 
manured  basins  was  slightly  higher  than  that  in  the  basins  mulched 
with  alfalfa,  a  result  agreeing  with  that  obtained  in  Experiment  I. 

No  appreciable  difference  in  the  color  of  the  fruit  from  the  trees  dif- 
ferently treated  was  discernible. 

Experiment  III. — The  block  of  60  orange  trees  used  in  this  experi- 
ment had  been  used  in  a  study  of  individual  tree  performance  by  the 
Office  of  Horticultural  and  Pomological  Investigations,  Bureau  of  Plant 
Industry.  Individual  tree  records  for  the  previous  six  years  had  been 
obtained,  and  these  were  kindly  furnished  by  the  Riverside  officials  of 
that  Office.  The  picking  records  of  the  191 6-1 91 7  crop  are  given  in 
Table  VIII,  together  with  the  basin  treatments.  The  basins  were  in- 
stalled and  mulched  in  October,  191 5. 

No  effect  of  the  lime  on  fruit  production  could  be  definitely  deter- 
mined; hence,  the  yields  of  all  the  trees  mulched  with  the  same  organic 
matter  have  been  averaged,  including  both  limed  and  unlimed  trees. 

The  alfalfa-mulched  trees  produced  more  fruit  than  the  manure- 
mulched  trees,  which  result  agrees  with  the  results  obtained  in  Ex- 
periments I  and  II.  It  does  not  appear  from  Table  VIII  that  any 
legume  whatever  is  superior  to  a  nonlegume  or  to  manure.  The  basined 
trees  mulched  with  sweet  clover  and  bur  clover  did  not  produce  as  much 
fruit  as  those  mulched  with  alfalfa,  bean  straw,  manure,  or  barley  hay. 


Feb.  2$.  1918 


Humus  in  Mulched  Basins 


515 


The  color  of  the  oranges  on  the  trees  mulched  with  alfalfa  and  bean 
straw  was  distinctly  more  golden  than  the  color  of  the  oranges  on  the 
trees  mulched  with  the  other  organic  substances.  The  foliage  on  the 
bean-straw  and  alfalfa-mulched  trees  was  considerably  denser  and 
greener  than  that  on  the  other  trees. 

Table  VIII. — Effect  of  different  mulching  materials  in  basins  on  orange  production. 
Experiment  III.     Picked  May,  igij 


Basin  treatment. 


Yield  of  oranges  corrected  on  ba- 
sis of  previous  performance  of 
trees. 


Pounds 
per  tree. 


Oranges 
per  tree. 


Average 

weight  per 

orange. 


Alfalfa  hay 

Bean  straw 

Manure 

Barley  hay 

Sweet  clover  hay 
Bur  clover  hay. . 
Pine  shavings. . . . 


308 
289 
261 
202 
217 
216 
163 


881 

744 
611 

497 

584 

530 
402 


Pounds. 

350 
389 
427 
407 

372 
408 
406 


On  comparing  the  results  of  the  humus  determinations  shown  in 
Table  IV  with  the  fruit  production  shown  in  Table  VIII  it  is  evident 
that  the  percentage  of  humus  in  the  soil  does  not  correlate  with  the  fruit 
production.  There  is,  however,  a  correlation  between  the  fruit  produc- 
tion and  the  ratio  of  the  humus  content  in  191 6  to  that  in  191 5,  but  such 
correlation  is  not  evident  in  Experiments  I  and  II. 

Closer  examination  of  the  picking  records  in  Experiment  III  showed 
that  the  bean-mulched  trees  produced  more  first-class  fruit  of  the  more 
desirable  sizes  than  the  trees  mulched  with  any  of  the  other  substances; 
the  alfalfa-mulched  trees  were  second  in  this  respect;  manure  third; 
and  barley  fourth;  while  the  trees  mulched  with  sweet  clover  and  pine 
shavings  produced  the  least  number  of  fruits  of  the  first  quality. 

It  may  be  that  some  toxic  substance  is  formed  in  the  decomposition 
of  the  pine  shavings  which  might  account  for  the  poor  condition  of  the 
trees  and  for  the  small  yield  obtained  from  them;  also  the  shavings  prob- 
ably had  a  deleterious  influence  on  nitrification.  Redwood  boxes  for 
use  in  germinating  Citrus  seedlings  proved  unsuitable,  because,  when 
the  root  tips  came  in  contact  with  the  wood,  they  promptly  died. 

Oranges  from  the  trees  in  Experiments  I,  II,  and  III  were  analyzed, 
the  analyses  being  made  by  the  branch  office  of  the  Citrus  B3--products 
Laboratory  in  Los  Angeles.  The  analyses  did  not  bring  out  any  con- 
sistent differences  in  the  fruits  from  trees  receiving  different  fertilizer 
treatments. 


2i6  Journal  of  Agricultural  Research  voi.  xii.no.  s 

There  was  a  consistent  difference  in  the  specific  gravity  of  the  oranges 
from  the  Umed  trees  and  from  the  unUmed,  but  the  difference  was  small. 
The  average  specific  gravity  of  the  oranges  from  the  mulched  and  limed 
trees  was  1.0012,  and  of  those  from  the  mulched  trees,  but  not  limed, 
was  1.0307.  The  specific  gravity  was  calculated  from  the  grading 
results,  and  not  by  individual  determinations  of  separate  fruits.  The 
error  in  sizing  by  the  grading  machine  would  probably  be  compensated 
by  the  number  of  fruits  included,  as  the  entire  crop  was  used  in  the 
calculation. 

On  the  whole,  the  evidence  obtained  from  these  experiments  does  not 
show  that  the  humus  content  of  the  soil  correlates  with  orange  pro- 
duction. Neither  does  the  information  obtained  justify  the  statement 
that  the  humifying  activity  in  the  soil  correlates  with  fruit  production; 
but  the  results  obtained  would  indicate  that  this  point  might  be  worthy 
of  further  study. 

It  appears  from  the  results  here  presented  that  the  most  important 
function  of  organic  matter  as  influencing  orange  production  is  not  that 
of  merely  furnishing  humus;  that  humus  in  itself  is  not  the  most  impor- 
tant product  of  organic  degradation  for  orange  production.  It  appears 
from  other  work  (8)  that  a  more  important  function  of  organic  matter 
is  to  make  the  plant  food  in  the  soil  minerals  more  readily  available. 
It  was  found  that  organic  substances,  especially  in  a  freshly  decom- 
posing condition,  dissolved  plant  food  elements  in  appreciable  amounts, 
even  when  the  organic  solvents  contained  practically  no  electrolytes. 
This  function  and  that  of  promoting  the  biological  activities  of  the  soil 
seem  to  be  more  important  rdles  of  organic  matter  in  the  soil  than  merely 
to  furnish  humus. 

SUMMARY 

This  report  presents  a  study  on  (a)  the  changes  in  humus  content  in 
soils  in  basins  mulched  with  different  organic  substances,  (b)  the  effect 
of  lime  on  humus  content  in  soils  in  mulched  basins,  (c)  the  relation  of 
humus  content  in  the  soil  to  orange  production. 

By  "humus"  is  meant  the  brown- to  black-colored  organic  extract 
obtained  from  soil  leached  with  i  per  cent  hydrochloric  acid  to  the 
absence  of  calcium  and  the  soil  residue  boiled  for  two  minutes  in  a  7.5 
per  cent  sodium-hydrate  solution. 

Humus  determinations  in  mulched  basins  in  citrus  groves  showed  a 
fluctuation  in  the  percentage  of  humus  from  time  to  time. 

The  average  percentage  of  humus  increased  more  in  basins  on  clay 
loam  soil  than  in  basins  on  lighter  soil  tubes,  with  manure  and  alfalfa 
as  mulching  m^aterials. 

Usually  the  percentage  of  humus  in  basins  increased  more  when  manure 
was  used  as  mulch  than  when  alfalfa  was  used  as  mulch.     This  seemed, 


Feb.  J5, 1918  Humus  in  Mulched  Basins  517 

however,  to  be  due  more  to  the  humus  added  with  the  manure,  than 
to  the  greater  "  humification  "  of  the  manure  over  the  alfalfa. 

When  manure  alone  was  used  as  mulch  in  basins  the  increase  in  hu- 
mus was  greater  than  when  lime  was  added  with  the  manure. 

In  most  cases  when  lime  was  added  to  alfalfa  in  basins  greater  increase 
in  the  humus  content  occurred  than  when  alfalfa  alone  was  used. 

Blood,  acid  phosphate,  bone  meal,  tankage,  or  sulphur  did  not  show 
any  appreciable  influence  on  the  changes  of  humus  content  in  mulched 
basins. 

It  was  not  evident  that  there  was  any  appreciable  accumulation  of 
humus  in  the  lower  depths  of  soil  due  to  the  leaching  of  humus  from  the 
surface  foot  of  soil. 

Thsre  was  no  evident  correlation  between  the  amount  of  humus  in  the 
soil  in  mulched  basins  and  the  amount  of  fruit  on  the  trees. 

There  was  no  evident  effect  of  lime  on  orange  production  in  these 
experiments. 

Alfalfa  and  bean-straw  mulch  in  basins  on  the  heavier  soil  types 
produced  from  30  to  100  per  cent  more  oranges  per  tree  than  manure 
mulch.  Manure  mulch  produced  more  oranges  per  tree  than  either 
barley  hay,  sweet  clover,  bur  clover,  or  pine  shavings.  These  differ- 
ences were  obtained  in  the  summer  following  the  application  of  the 
mulches  in  the  preceding  fall. 

Alfalfa  mulch  and  manure  mulch  in  basins  on  the  lighter  types  of 
soil  produced  no  observable  differences  on  fruit  production  of  lemons 
in  the  course  of  one  year.  This  statement  is  based  only  on  observation 
and  not  on  picking  records. 

In  all  experiments  so  far  conducted  by  this  Office  in  the  Riverside  area, 
the  mulched-basin  system  on  the  heavier  soil  types  has  produced  favora- 
ble growth  response  in  a  few  months.  It  usually  takes  longer  to  produce 
appreciable  response  on  the  lighter  soil  types. 

It  would  appear  directly  from  the  work  here  reported,  and  indirectly 
from  work  elsewhere  reported  that  the  degradation  products  from 
freshly  decomposing  organic  substances  are  more  effective  in  orange 
production  than  the  amount  of  "humus"  formed.  And  the  value  of  a 
given  mulch  does  not  necessarily  depend  upon  its  being  a  legume  or 
nonlegume. 

LITERATURE  CITED 

(i)  Briggs,  L.  J.,  Jensen,  C.  A.  and  McLane,  J.  W. 

1916.    mottle-leaf  of  citrus  trees  in  relation  to  soil  conditions.    In 
Jour.  Agr.  Research,  v.  6,  no.  19,  p.  721-739,  4  fig.,  3  pi.     (partly  col.) 
(2) 


1917.   THE    MLTLCHED-BASIN    SYSTEM    OF    IRRIGATED    CITRUS    CLT^TURE     AND    ITS 
BEARING    ON    THE   CONTROL    OF    MOTTLE-LEAF.      U.    S.    Dept.    Agf.    Bul.    499, 

31  p.,  I  pi. 


c  1 8  Journal  of  Agricultural  Research  voi.  xii,  No.  8 

(3)  Frear,  Wiluam,  and  Hess,  E.  H. 

1900.     INFLUENCE   OF  SYSTEMS  OF  FERTILIZING  VFON  THE  AMOUNT  AND  QUALITY 

OF  THE  HUMUS  OF  THE  SOIL.    In  Ftoc.  2ist  Ann.  Meeting  Soc.  Prom.  Agr. 
Sci.,  p.  60-69. 

(4)  GORTNER,  R.  A. 

1916.      THE  ORGANIC  MATTER  OP  THE  SOIL:      I.  SOME  DATA  ON  HUMUS,   HUMUS 

CARBON,  AND  HUMUS  NITROGEN.     In  Soil  Sci.,  V.  2,  no.  5,  p.  395-441,  2  pi. 
Literature  cited,  p.  440-441. 


(5) 

I917.      THE  ORGANIC  MATTER  OF  THE  SOIL:    III.  ON  THE  PRODUCTION  OF  HUMUS 

FROM  MANURES.     In  Soil  Sci.,  v.  3,  no.  i,  p.  1-8.     Literature  cited,  p.  8. 

(6)  Heinze,  B. 

1909.  the  formation  and  decomposition  of  humus  in  cultivated  soils. 
(Abstract.)  hi  Exp.  Sta.  Rec,  v.  23,  no.  7,  p.  621.  1910.  (Original  article 
in  Landw.  Mitt.  Prov.  Sachsen.  u.  Nachbarstaat.  Halle,  1909,  p.  145-146. 
Not  seen.) 

(7)  HacARD,  E.  W. 

1914.  soils:  their  formation,  properties,  composition,  and  RELATIONS 
TO  CLIMATE  AND  PLANT  GROWTH.      593  p.,   89  fig. 

(8)  Jensen,  C.  A. 

191 7.  EFFECT  OF  DECOMPOSING  ORGANIC  MATTER  ON  THE  SOLUBILITY  OF  CER- 
TAIN INORGANIC  CONSTITUENTS  OF  THE  SOIL.  In  Jour.  Agr.  Research,  v.  9, 
no.  8,  p.  253-268. 

(9)  McBeth,  I.  G. 

1916.     STUDIES  ON  THE  DECOMPOSITION  OF  CELLULOSE  IN  SOILS.      In  Soil  Sci., 

V.  I,  no.  5,  p.  437-487.     Literature  cited,  p.  481-487. 
(10)  ScHREiNER,  Oswald,  and  Shorey,  E.  C. 

1909.      THE  ISOLATION  OF  HARMFUL  ORGANIC  SUBSTANCES  FROM  SOILS.      U.   S. 

Dept.  Agr.  Bur.  Soils  Bui.  53,  53  p.,  4  pi. 


RELATION  OF  KINDS  AND   VARIETIES  OF    GRAIN  TO 
HESSIAN-FLY  INJURY^ 

[PRELIMINARY  REPORT] 

By  James  W.  McColloch,  Assistant  Entomologist,  and  S.  C.  Salmon,  Professor  of 
Farm  Crops,  Kansas  Agricultural  Experiment  Station 

It  has  long  been  known  that  certain  varieties  of  wheat  (Triticum  spp.) 
are  injured  less  than  others  by  the  Hessian  fly  {Mayeiiola  destrticior). 
Packard  ^  mentions  the  Underhill,  Mediterranean,  Lancaster,  and 
Clawson  varieties  as  being  noted  for  resistance.  He  states  that  the 
Underhill  variety  has  been  highly  recommended  for  nearly  a  century. 

Woodworth  ^  made  observations  on  125  varieties  of  wheat  grown  at 
the  California  Experiment  Station  in  1886,  1887,  and  1889,  and  noted 
the  damage  by  Hessian  fly  in  each.  The  Volo  and  Washington  Glass 
varieties  were  not  injured.  Forelle;  bearded  wheat  from  Missoyen, 
Palestine;  Polish;  Bluegrass;  Common  March;  Diamond;  and  Egyptian 
Imported  were  practically  free  from  injury. 

Roberts,  Slingerland,  and  Stone,'*  in  summarizing  their  observations  on 
Hessian-fly  injury  in  New  York,  conclude  that 
the  resisting  power  of  varieties  varies  greatly 
and  that 

those  with  large,  coarse,  strong  straw  are  less  liable  to  injury  than  weak-strawed  and 
slow-growing  varieties. 

Six  varieties  are  mentioned  which  were  not  appreciably  affected  by 
the  fly  in  1901,  although  numerous  other  varieties  in  the  same  neighbor- 
hoods were  much  injured.  These  varieties  were  Dawson  Golden  Chaff, 
Prosperity,  No.  8,  Democrat,  Red  Russian,  and  White  Chaff  Mediterranean. 

Gossard  and  Houser^  made  careful  observations  on  75  varieties  of 
wheat  and  other  grains  grown  at  the  Ohio  Experiment  Station  in  1904, 
1905,  and  1906.  They  determined  the  percentage  of  stalks  infested  and 
of  fallen  straws.     Their  observations 

give  but  little  support  to  the  idea  that  there  are  immune  varieties, 
and  they  suggest  that  cases  of  supposed  immunity  may  be  explained  by 
some  other  hypothesis.     They  state,  however,  that  the  most  persistent 

•  Contribution  from  the  Entomological  Laboratory  (Paper  No.  30)  and  the  Department  of  Agronomy 
(Paper  No.  13)  cooperating.  This  paper  embodies  some  of  the  results  obtained  in  the  prosecution  of  projects 
No.  8  and  67  of  the  Kansas  Asricultural  Experiment  Station. 

*  Packard,  A.  S.  the  hessian  fly — its  i^avages,  habits,  and  the  means  of  preventing  its  in- 
crease.   In  3rd  Rpt.  U.  S.  Ent.  Com.,  p.  227-228.    1883. 

»  WOODWORTH,  C.  W.     VARIATION  IN  HESSIAN  FLY  INJURY.     In  Cal.  Agr.  Exp.  Sta.,  Rpt.,  iSgo,  p.  312. 

I89I. 

<  Roberts,  I.  P.,  Si,ingeri.and,  M.  V.,  and  Stone.  J.  L.  the  hessian  fly.  its  ravages  in  new 
YORK  IN  1901.    N.  Y.  Cornell  Agr.  Exp.  Sta.  Bui.  194,  p.  226-260,  fig.  95-98.     1901. 

'  Gossard,  H.  A.,  and  Houser,  J.  S.  the  herslan  fly.  Ohio  Agr.  Exp.  Sta.  Bui.  177,  39  p.,  2  fig., 
I  col.  pi.,  map.    1906. 

Journal  of  Agricultural  Research,  Vol.  XII,  No.  8 

Washington,  D.  C.  Feb.  25.  1918 

mc  Key  No.  Kans.— n 

(SI9) 


520  Journal  of  Agricultural  Research  voi.  xii,  no.  s 

search  has  never  located  a  single  Hessian-fly  egg  on  oats,  although  eggs 
were  found  on  many  grasses. 

Since  1906  the  relation  between  varieties  and  injury  from  Hessian 
fly  has  received  scant  attention  from  investigators.  Presumably  the 
work  of  Gossard  and  Houser  has  been  accepted  as  disproof  of  the  claims 
of  earlier  observers  that  some  varieties  are  resistant  and  others  immune. 

Recently  claims  of  immunity  put  forth  by  growers  of  certain  varieties, 
general  observations  by  farmers  in  eastern  and  central  Kansas  that 
hard  wheats  are  more  susceptible  to  injury  than  soft  varieties,  and  re- 
sults of  experiments  at  the  Kansas  Agricultural  Experiment  Station 
indicate  that  the  subject  is  at  least  worthy  of  further  investigation. 

Experiments  have  been  outlined  to  determine  (i)  the  relative  infes- 
tation and  injury  of  different  kinds,  varieties,  and  strains  of  various 
small  grain,  and  (2)  why  certain  kinds  and  varieties  are  resistant  or 
immune;  or,  if  not,  why  they  escape  injury  in  some  cases  where  others 
are  badly  injured.     This  paper  is  concerned  primarily  with  the  first 

problem. 

EXPERIMENTAL  DATA 

The  data  reported  in  this  paper  were  collected  from  87  kinds  and 
varieties  of  wheat  (Triticum  spp.),  oats  {Avena  saliva),  barley  {Hordeum 
spp.),  rye  (Secale  cereale),  emmer  {Triticxim  dicoccum),  einkorn  {Triti- 
cum, monococcum) ,  and  spelt  (Triticum  spelta),  planted  in  the  Agronomy 
Nursery  of  the  Kansas  Agricultural  Experiment  Station  in  the  fall  of 
1 91 6.  The  different  varieties  were  planted  each  in  a  row  50  feet  long  and 
10  inches  apart.  Two  plantings  were  made  on  each  of  two  dates,  Septem- 
ber 15  and  October  i.  The  soil  was  in  excellent  condition,  moisture 
was  plentiful,  germination  was  prompt,  and  growth  was  normal  in  every 
respect.  Hessian  flies  were  numerous,  and,  as  far  as  known,  there  was 
ample  opportunity  for  all  varieties  to  become  equally  infested.  In  this 
paper  all  varieties  are  tabulated  in  the  order  in  which  they  were  planted. 

Eight  of  the  varieties  tested  were  from  Australia  (rows  i  to  8,  inclu- 
sive, Table  I),  and  had  not  been  grown  previously  at  Manhattan.  The 
spring  varieties  and  about  half  of  the  soft  winter  varieties  had  been  ob- 
tained from  various  Experiment  Stations  in  the  United  States  in  1914 
and  had  been  grown  at  Manhattan  for  two  years  only  before  being  in- 
cluded in  this  experiment.  All  of  the  hard  winter  varieties  and  about 
half  of  the  soft  winter  varieties  had  been  grown  at  Manhattan  for  several 
years  and  were  thoroughly  acclimated. 

The  relative  number  of  eggs  deposited  on  each  variety  was  determined 
by  taking  five  consecutive  plants  from  the  west  end  of  each  row  and 
counting  the  total  number  of  eggs  on  the  leaves.  The  first  count  was 
made  at  the  time  of  maximum  deposition,  September  25.  Subsequent 
determinations  were  made  for  the  early  sown  plots  on  October  2  and  7, 
and  for  the  late  sown  plot  on  October  14.  The  total  number  of  plants 
of  each  variety  examined  was  20. 


Feb.  2s.  1918  Relation  of  Grain  to  Hessian-Fly  Injury  52 1 

The  relative  number  of  flaxseeds  in  each  variety  was  determined  by 
examining  50  consecutive  plants  of  each  row  in  each  plot  sown  on  Septem- 
ber 15,  and  25  consecutive  plants  of  each  row  in  each  plot  sown  on 
October  I ,  making  a  total  of  1 50  plants  of  each  variety.  All  plants  were 
taken  from  the  western  end  of  the  plots,  or,  in  other  words,  adjacent 
to  those  plants  which  were  examined  for  eggs.  The  pertinent  data  for 
each  variety  are  given  in  Table  I. 

That  the  Hessian  fly  is  apparently  able  to  discriminate  between  kinds 
and  varieties  of  grain  is  shown  by  these  data.  For  example,  the  total 
number  of  eggs  per  100  plants  in  the  early  sown  plots  ranges  from  40  for 
Culberson  winter  oats  to  5,600  for  Turkey  winter  wheat  No.  2407.  The 
proportion  of  plants  on  which  eggs  were  laid  varied  from  20  per  cent  for 
Culberson  winter  oats  and  Michigan  winter  barley  to  100  per  cent  for 
Tennessee  winter  barley  and  most  of  the  varieties  of  wheat. 

The  close  agreement  in  determinations  made  at  different  times,  and  the 
striking  differences  in  the  number  of  eggs  laid  on  adjoining  varieties  indi- 
cate that  the  difference  can  scarcely  be  attributed  to  experimental  error. 
Thus,  Turkey  winter  wheat  No.  2407,  which  showed  the  highest  total  num- 
ber of  eggs,  also  had  the  highest  infestation  on  September  25,  the  second 
highest  on  October  10,  when  the  second  determination  was  made,  and  was 
among  the  highest  on  October  7,  when  the  third  determination  was  made. 
On  the  other  hand,  Culberson  Winter  oats,  Michigan  Winter  barley,  and 
einkom  had  the  lowest  total  infestation  and  the  lowest  on  each  date. 

Row  12  (Polish  wheat)  had  a  total  of  2,040  eggs  per  100  plants,  as 
compared  with  280  for  row  13  (einkorn).  Row  15  (spring  emmer)  had  a 
total  of  660  eggs  per  100  plants,  as  compared  with  1,740  for  row  16 
(Black  Winter  emmer). 

On  the  whole,  the  Hessian  fly  appears  to  have  shown  a  preference  for 
common  wheat,  as  compared  with  barley,  oats,  einkorn,  spring  emmer, 
spelt,  and  durum  wheat.  Black  Winter  emmer.  Poulard  wheat,  and 
Polish  wheat  were  as  heavily  infested  with  eggs  as  many  of  the  common 
wheats.  Rye  showed  a  very  heavy  infestation,  the  total  number  of  eggs 
per  100  plants  being  2,500,  which  is  well  above  the  average  for  all  grains 
included  in  the  test. 

Varieties  of  the  hard  winter  wheat  class  were  more  generally  infested 
than  soft  winter  wheat  varieties.  Thus,  27  varieties  of  hard  winter 
wheat  averaged  2,737  ^ogs  per  100  plants,  as  compared  with  an  average 
of  1,835  for  38  varieties  of  soft  winter  wheat.  However,  there  are  wide 
variations  in  each  class.  For  example.  No,  2408  and  Mealy,  which  are 
soft,  or  semihard,  varieties,  were  infested  with  4,720  and  4,320  eggs,  re- 
spectively, per  100  plants,  showing  almost  as  high  an  infestation  as  the 
most  profusely  infested  varieties  of  hard  wheat,  and  more  than  double 
that  of  some  varieties.  On  the  other  hand,  certain  varieties  of  hard 
wheat,  such  as  Defiance  No.  2129,  Red  Winter  No.  839,  Improved  Tur- 
key No.  2382,  and  Pesterboden  No.  205,  had  a  lower  infestation  than  the 
average  of  the  soft,  or  semihard,  varieties. 


522 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  8 


0.0 

•io 

I-* 

o 


«i^ 


;2;.c' 


ij-  'J-  O    M    O    N 


0    •<t  O    M    O    •* 


0000000000000000 

OO00O-<i-OM00iN    'too  00    O  00 


a 


^ 


*> 


10  r-.  M    N    O   t^oO    -^00   i/^oo    'J- O   1^  O   Tl- 


O   O   >0  O   to  N 


0000  «\0  N  >riM  O  Tfoo  00  t>.t^r>. 


O  M   O   O  O  O 


OONMrOM^ON«Ocot>.rO'*         >0 


■§X1  tfl  *J 

a  tj  M  a  >A. 


0000000000000000 

MMOOOf^OO^OOMO-t  'too  00  vO    "t 
row    'troiJ^OfCO    O^'^O    O    O    fl    10\0    J^ 


000000 

't   M  'O    N     'to 

M    M   10  M    ro 


000000000000000 

OCvOOOOvO-tvOvO         OO't'tTtwO 

Tj-rOO  MM'OC<  M  HN 


000000 

't    M   VO     ■^ 

M     M  \0 


0000000000000000  'OOOOOO 

O^OOMVOtNOOMN     ■*00    C<M\000       iTfMPlOOO 

0000M^0M0wvO'tt^C^0^^^'t't^o•         NMrj-CNio 


^ 


M  M  ro  't  100  «^0C  00  «~- 
OOOOOOOOOOv 
'tt't-^'t'tt'tM    cs 


1:13 

«-1 


t4  eq  eq  ffi  u  U  ►S  H  W  S 


•§ 


be  "5 


.5  t« 


B-o  ■"  S  ° 


C    O   r^    VV 


»-i  in  w 

+^  "ti  -^^ 


'^    ^    r-j    <U 

^f-l    5    <U   C 

-  S  "3  S  -y  >^ « •§ 


S  <u  o 


o 


^e 


,:;:i  ^  O  ^  l2  P  P  ;->  Oh  a.'>C^  <^ 


J3 

c   : 


u   O 


•0^  MO-JJ 


Feb.  as.  1918  Relation  of  Grain  to  Hessian-Fly  Injury 


523 


0000000 


0000000000000000000 

>0<N\OoiCivOONO'^NO    "^OO    04  O  vO    o 


N 

'i-'^OcO    i^.cO   w-iO'O- 

Tj-  <N    ro 

ID  1000    ■<J-   ►^    0    t^O    M  00    lOOO    ro  0    LO  0 

00  m  f^  0 

0 

f^ 

N    fO  M    "1-  ■*  ro  CO  lOvO 

10  Tt  W 

rO^Ot~OrO<^»^'!r''5'*N    C^w    N    row    11 

M    0    Ci    0 

1- 

w 

■*  ■*  M  00    'too    "t^O  0 

M    lOOO 

•^M    N    -itlNCOOO^    OmDoOoOoO    MCOO 

0    0   t^cO  00 

0 

00 

M  \C    0    ^00    ro  rooc    't 

MUM 

VOOO    M    lOvO    rj    OP)    M    lOTj-TtrorOroO 

■*  0\vO    CO  CO 

M     04               MM 

00 

I^OmioOmmO    t^oO   10  vooO 


10\0    t>.  tOOO    C4    tOOO    M    On  N    t-~  UOvO    OfO'^'tt^-OO  ■* 

Tj-  tJ-  irj  rf  LOO  toLO-^        •^•^■^        m  r4C4iOM         10 


00     M     04    lOO    t~~MOO    VOOnO-cOO 
li^  t-^  t-~  ir.OO    >J^  C7<  ionO    m  CO    O  O 

N  M  O))-(WMMC404C4 


M    M    ro  M   "^  lOCC  OOoO    O    M    ro'Or-OroPi    CT^cOm    10 
p)    vo  t^  t^OO  0CP4<3l^>^O"*C^OMM  mio  t->.0    m 

MMMMMC4C404M  M  M  M 


0 

0 

0 

0 

0 

0 

n 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

n 

0 

0 

0 

0 

8 

M 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

M 

0 

M 

0 

0 

0 

0 

0 

M 

0 

0 

M 

0 

M 

0  00 

0 

0 

t-H 

0 

M 

u 

U 

0 

0 

0 

0 

0 

0 

M 

0 

0  00 

0 

M 

0 

M 

0 

M 

0  CO 

M 

0 

M 

0 

M 

0 

M 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

Tl-OO 

ct 

0 

-^ 

a) 

0 

PI 

0 

^0 

0 

0  00  vO  CO 

0  0  CO 

1-0 

'to 

^0 

0 

0 

ri 

P4  00 

0 

to 

n 

vO 

10 

0 

I^ 

l^ 

COCO 

*^ 

C) 

'^ 

LO 

LO 

00 

N 

Pt 

»~~ 

^ 

CO 

PI 

CO 

0 

r^ 

0 

CO 

0^  t^ 

t^ 

W 

r^ 

10 

t 

to 

M 

CO  "^ 

C4 

CI 

•^ 

M 

-* 

CO 

" 

" 

cs 

M 

CO 

P4 

M 

C4 

'-' 

P) 

P4 

M 

N 

P4 

-' 

M 

P4 

" 

M 

" 

CO 

" 

'-' 

M 

'-' 

00000000 

O    'toco    Tj-'tP^    PI 


00000000000000000000   o 

piPtPfpfOOPi         O00P40't0O't-t000Pi         00      4* 

t-i  P«MMMCOP4  MM  *i 


0000000000000 

00PiOOP4    0O00000000 

VOP4P4»OM't'tP4COP4C1PIP4 


000000000000000000000 

0'1-P4P40P4     000P40    'tCO     O    P4     O  00     ^tOO  O  O 
H'l-MP4'ttOPinMCOP4't'rl"CO'tCOP4  T|-P1 


^  ^ 


0000000000000  000000000000000000000 

P»    ■t  't  P4  00    TfoO    "to  O    P4    'too  TJ-P4    PI    T^P4000CO    OOOOOOOOOOOO    O    O    -too  00 

00    P4  O    O    too    CO  COOO    t   CO  P4    P4  lOOO    M    100    PI    On  P4    PI    lO  -t  t  COCO    CO  O    "t   0^  lO  CO  CO 

Tti-rpi'TtpTi-rrfi-rcocoHri-rpr  )-rpreri-ri-ri-ri-rprHrp4'"pr     Hri-rhTi-ri-rpr     i-Ti-t 

r~~coMOO  P4  10  o  00-  coo  100  coOPiO'tMiocoOpiO'i-iococoOOcO'tco'o 

O    O    I^  O  CO  00    r^  coo    mOvOm  Ttco-tLOO^OOPici^  coo  O    t^oO    O    O    O    0\  too 

-t     t    O      t    CO    CO    10,00  MMP4P4  -.a-OOOOMMMMMM  MMtMCOCO 

pT  pT  PI    P4  M  H    PI    PI    p)    P4    P4    04   cT  cT  cT  cT  pf  pi' 


Q    Q 


S  a;  «  >^^  '^  3  o  ^  c 
H  C  ^  W  U)  O  H  Pi  S 


3  ^  „ 


rt  ^  _n  "-M 

«Pi<;P 


3 
J-.2T3 


m 


"3  ?=-a 


be  5^ 

M  be  y 

3    !=J3'Z3 


4->    CS  ^.j    c! 

cS  flj  rt  lU 
01  ^    «  ^ 

'^    W  ""^    Ut 

Im    HI  Vh    OJ 

OJ    Ui    Cj    Wi    O 


cBWcBM 


<U     M 

•S'i 

u    o    ^ 


rt  <U  c3 
u  V  u 

.5|.H 

TS  wo 
II  u  4) 
t-^   I-' 


*  I 


S    :t8 


to  CO  PO  t  t  -"t  Tl-  • 


524 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  8 


i2  «J  ff  Si  « 


OnO        OOOnnOOoO 


O   P)   O   O   N   o 


TtOOvOcsOO'* 


oooooooo 

vO  VO    O  00  \0    rt-O  00 

M      M      HI      M  M 


oooooooo 


8 


a 

.IS 

i 


,8 

s 


^ 


O  >oO   iot^t--.M   t^MO   O    t^0"0 


M    O   r^  O   roO   O    0> 


g-oa. 


t/i  rt  Qj 

."■Sen 


rt  Qj  N 
0.W 


i.  ^^  ><^  o  4ij 

>  ^B'U^  Z  9 
J?  "  Sj!  Si  «3~ 
^      5^  ■"  a  Q. 


So' 


«3i2 
a  a 


n  _*j  M  Q,  ^ 
5;  ^  <"«  " 


00  O  00  vO  ■*  ■^OO  O  N  >0  O  M  O  00 
OnO  ^00  r^  "^  r--  r^vO  0^0  ro  fO  O 


N  00  M  >0  00  00  1^0 
00  00"  rOOO'  On  ■^  N  0C5 


■g  O  O  O  o  o  o 
.000000 


00000 
00000 


OOOQQOOOOOOOOO 

o  Tt-00  oooooo'OTl■ooT^o 

M  00    i-i    N    O^vo    000    r^CM    M    r^ioro 


O    O    O    Q    O 

M  VO  \0    O    ■+ 
On  M    LO  C\  w 


00000000000000 

00    'J-oo  00    ■^mS    rf  P)    pi  \0    n    m    ■*\0 
i-i^p<i-iMM         p^H';^rowl-l 


00000000000000 

00  \0  00  vO     ■*   Tj-OO     O     O  \0     O    P<  ^0  00 

Q\\0    Looo    1^  '^^  r-  t^  10  00    ro  ro  O 


OOOOOOOO 

Ptoo    000000    p)vO 
00  00    rooo    Ch  ^00  00 


•    n  S  K  o 


•<4-  lOvO    CN  On  fO  OnnO    O   t^OO   ■*  O    Os 
000         w    po  ^n\0  t^  t^  t^oo    On  O 

•*    ^    Tj-  M 


MVOOO    w    PO'^PO't 
mmOOOOpioO 

l-IMIH(NPJP<C<0 


•o>i  Mo-a 


w 


oj    HI    03 

&  C  & 

?:  &  a  & 

"O  "O  '^J  TJ 
Q»*    C^  'T'    Q^    O    O 

-    -  i-i'O'a 


■CO 


:^c8 


Feb.  25,  191 J 


Relation  of  Grain  to  Hessian-Fly  Injury 


525 


0000000 

T^  N  00  o  00  c<  N 


8000000 
00     -4-   P)  CO     M  00 

roO  00   fO  r^  "^  "^ 


00  \0    t^  M    Ov  "^  O 
00    10  iJOvO  O    r^oo 

O     M      H      M      H      M      H 

cT  cT  N   M   cf  fT  fT 


•g    ;3    ;3    W    OJ 

WfeCJOP 


\C)  t  t  0   0   "^oo 

Q    0    0    0    0    0    0 

0   <N    'too  00   c<   ■* 

c5    N    ro\d    'J-  ro  ro 

000    'J- 
f^O  00 

M     HI 

N  00     N  00 
CO   cL  LO  lO 

M     M     M 

r^  rj-  ro 

H     CI  0     (^ 
\0     M     1-1     ■* 

M    f»   »O00  00   0    ro 
0    ro^    0  ■*  •*  <^ 

888 

M     H     M 

8888 

H     W     H     W 

000 

v6  0  '* 
vo  C^  0 

0000 

W     Tj-CO  00 

t^  0  CO   t^ 

M     M      M 

w    N    1-1 

000 

P)  0    M 

0000 

too  ■too 

N           H    M 

00^0000 
tj  tJ  "^  "d  "13  t3  13 


526  Journal  of  Agricultural  Research  volxii,  no.8 

Perhaps  the  most  significant  result  is  the  high  mortality  of  the  eggs 
and  larvae,  and  as  a  result  the  relatively  low  infestation  with  flaxseed  of  all 
varieties  and  especially  certain  ones.  The  figures  are  open  to  criti- 
cism, so  far  as  the  absolute  mortality  is  concerned,  since  more  eggs  may 
have  been  laid  on  the  ends  of  the  rows,  where  the  egg  counts  were  made, 
than  on  the  adjoining  plants,  which  were  used  for  the  flaxseed  determi- 
nations. However,  this  criticism  would  not  hold  for  the  relative  mortality 
for  the  different  varieties. 

As  will  be  seen,  no  flaxseeds  were  found  in  einkorn,  spring  emmer, 
Culberson  Winter  oats,  rye,  and  Illini  Chief  wheat.  Very  low  infesta- 
tions, 5  per  cent  or  less,  were  recorded  for  Tennessee  Winter  barley  and 
for  Beechwood  Hybrid,  Currell  Selection,  and  Dawson  Golden  Chaff 
wheats.  The  data  for  rye  and  for  Illini  Chief  and  Dawson  Golden  Chaff 
wheats  are  especially  significant,  in  view  of  the  fact  that  2,000  or  more 
eggs  per  100  plants  were  laid  on  each. 

Of  the  hard  wheats.  Red  Winter  No.  2132  is  especially  worthy  of 
mention,  since  it  had  only  30  flaxseeds  per  100  plants,  as  compared 
with  over  200  for  other  varieties  of  the  same  class.  Only  9  per  cent  of 
the  plants  of  this  variety  were  infested  with  flaxseeds,  as  compared  with 
a  range  of  from  28  to  62  per  cent  for  other  varieties  of  this  class. 

In  general,  the  data  show  a  low,  or  no  infestation,  with  flaxseeds 
for  rye,  barley,  oats,  durum  wheat,  Poulard  wheat,  Polish  wheat,  spelt, 
emmer,  and  einkorn.  The  average  number  of  flaxseeds  per  100  plants 
for  38  varieties  of  soft  winter  wheat  was  76.1,  as  compared  with  173.7 
for  27  varieties  of  hard  winter  wheat,  an  increase  for  the  latter  of  more 
than  225  per  cent. 

These  conclusions  are  based  on  the  rows  planted  on  September  15, 
but  essentially  similar  results  were  secured  from  those  rows  planted  on 
October  i.  The  infestation  on  these  plots  was  much  lighter  than  in 
the  former  case,  and  the  data  are  less  conclusive.  It  will  be  noticed, 
however,  that  those  varieties  which  were  not  infested  or  which  show  a 
low  infestation  in  the  first  plantings  exhibit  a  similar  characteristic  in 
the  latter. 

Field  tests. — In  191 5  a  bushel  of  Illini  Chief  wheat  was  secured  by 
the  Entomology  Department  of  the  Kansas  State  Agricultural  College 
and  planted  in  four  localities  in  the  State  where  the  Hessian  fly  was 
abundant. 

The  Illini  Chief  was  practically  free  from  injury  in  all  cases.  At 
Manhattan  it  showed  less  than  a  i  per  cent  infestation,  while  Turkey 
wheat  in  an  adjoining  plot  was  infested  practically  100  per  cent. 

An  examination  of  the  plants  in  the  fall  and  spring  indicated  that 
the  fly  showed  no  preference  for  either  variety,  the  eggs  being  equally 
numerous  on  both  plots.  In  the  Illini  Chief  the  maggots  were  able  to 
work  their  way  down  to  the  crown  of  the  plant,  but  at  this  point  devel- 
opment appeared  to  be  arrested  and  the  larvae  died. 


Feb.  25, 1918  Relation  of  Grain  to  Hessian-Fly  Injury  527 

At  Winfield,  Kans.,  the  Illini  Chief  was  sown  in  the  center  of  a  40- 
acre  field  of  hard  wheat.  Determinations  in  the  spring  showed  that  95 
per  cent  of  the  plants  of  hard  wheat  were  infested,  as  compared  with 
about  10  per  cent  for  the  Illini  Chief. 

In  1 91 6,  a  plot  of  Illini  Chief  wheat  grown  by  the  side  of  Turkey 
wheat  showed  an  infestation  of  from  3  to  5  per  cent,  as  compared  with 
95  to  100  per  cent  for  the  Turkey  variety. 

While  these  tests  appear  to  show  that  Illini  Chief  is  somewhat  resist- 
ant to  the  Hessian  fly,  it  should  not  be  assumed  that  it  is  the  best 
variety  to  grow.  In  Kansas  it  is  one  of  the  least  hardy  of  all  winter- 
wheat  varieties  and  will  survive  none  but  the  mildest  winters. 

CONCLUSIONS 

The  Hessian  fly  is  able  to  discriminate  between  different  kinds  and 
varieties  of  grain.  Eggs  were  laid  on  all  the  kinds  and  varieties  of 
grain  studied,  but  very  sparingly  on  winter  oats,  winter  barley,  einkorn, 
spring  emmer,  spelt,  and  durum  spring  wheat. 

On  the  average,  fewer  eggs  were  laid  on  soft  winter  wheat  than  on 
hard  red  winter  wheat,  but  exceptions  in  both  cases  were  found. 

There  appeared  to  be  a  large  mortality  of  eggs  or  larvae  on  all  kinds 
and  varieties  studied.  This  appeared  to  be  greatest  for  rye,  einkorn, 
spring  emmer,  winter  oats,  and  Illini  Chief  wheat.  Very  few  flaxseeds 
were  found  on  winter  barley,  and  on  Beechwood  Hybrid,  Currell  Selec- 
tion, and  Dawson  Golden  Chaff  wheats. 
27811°— 18 5 


ADDITIONAL  COPIES 

OF  THIS  PUBLICATIOK  MAT  BE  PROCtniED  FROM 

THE  SUPERINTENDENT  OF  DOCUMENTS 

GOVERNMENT  PRINTING  OFFICE 

■WASHINGTON,  D.  C. 

AT 

15  CENTS  PER  COPY 

StmscRiPTioN  Price,  33.90  Pee  Yeab 

A 


Vol.  ::vll  ^^ARCH  ^,  191S  No.  9 

JOURNAL  OP 

AL 
RESEARCH 


CONTKNTS 

Page 
Wilt  Diseases  of  Okra  and  the  Verticillium-V/ilt  Problem      529 

C.  W.  CARPENTER 

(  Contributioa  from  Burcsu  oi  Plant  indtssJry  ) 

Winter  Cycle  of  Egg  Production  in  the  "Sthode  Island  Red 

Breed  of  Domestic  Fowl    ~        -        -        -        -        -      547 
II.  D.  GOODALS 

( Coritdbutioa  from  Massachusetts  Agrfculhirsi  Kiperiirci:.  f.i  Uon") 

Digestion  of  Starch  hj  the  Young  Calf    -        -         ■        -      575 

R.  H.  SHAW,  T.  £.  WOODWARD,  and  E.  P-  ""■ :•  i 

(Contribution  from  Btixeaa  of  Aniraal  ladusL;,  ,, 

Toxicity  of  Volatile  Organic  Compounr  •;  to  I/isect  Eggs  -      579 
WILLIAM  MOORE  and  S; 
( Contribulioa  from  Minnesota  Agr; ;  - . i ) 

Com-Stover  Silage    -        -        -  ,.        „        _      559 

J.  M.  SHERMAN,  and  ^,  ..  ..__..„-'', 
(Ccntribulioa  from  Pennsylvania  AgrJcuItarsl  Excerimt 

Weevils  Which  Affect  the  Irish  Pc  Potato,  and 

Yam    -----  .,        -      601 

W.   DV/IGHT  ±'ii-r:i.-Ji; 
(  Coatrlbutioa  frota  Bureau  of  EnJom'>!'?gy  ) 


PUBLISHED  BY  AUTHORITY  OF  THE  SECRETilRY  OF  AGRIGJLTURB, 

WITH  THE  COOPERATION  OF  THE  ASSOCIATION  OF  AMERICAN 

AGRICULTURAL  COLLEGES  AND  EXPESBIENT  STATIONS 


WASHINOTON,  r>*  p. 


wmhinqton:  QovEnNMENT  pRiKTiNA  orFios )  ran 


EDITORIAL  COMMITTEE  OF  THE 

UNITED    STATES    DEPARTMENT    OF    AGRICULTURE    AND 

THE  ASSOCLATION  OF  AMERICAN  AGRICULTURAL 

COLLEGES  AND  EXPERIMENT  STATIONS 


FOR  THE  DEPASTMENT 
KARL  F.  KELLERMAN,  Chairman 

Physiologhl  and  Associate  Chief,  Bjtreau 
of  Plant  Industry 

EDWIN  W.  ALLEN 

Chief,  Office  of  Experiment  Stations 


CHARLES L.  MARLATT 

Entomoloqisl  and  Assistant  Chief ,  Bureau 
of  EniOTKology 


FOR  THE  ASSOCIATION 

RAYMOND  PEARL* 

Biologist,  Mains  Agricultural  Experiment 
Station 

H.  P.  ARMSBY 

Director,  Institute  cf  Animal  Nutrition,  The 
Pennsylvania  State  College 

E.  M.  FREEMAN 

Botanist,  Plant  Pathologist  and  Assistant 
Dean,  Agricultural  Experiment  Station  of 
the  University  of  Minnesota 


All  correspondence  regarding  articles  from  the  Department  of  Agriculture  should  be 
addressed  to  Karl  F.  Kellerman,  Journal  of  Agricultural  Research,  Washington,  D.  C. 

*  Dr.  Pearl  has  tmdertaken  special  work  in  connection  with  the  war  emergency; 
therefore,  until  further  notice  all  correspondence  regarding  articles  from  State  Experi- 
ment Stations  should  be  addressed  to  H.  P.  Armsby,  Institute  of  Animal  Nutrition, 
State  College,  Pa. 


JOm£  OF  AGMCPLTIAL  ffiSEARCH 

Vol.  XII  Washington,  D.  C,  March  4,  191 8  No.  9 


WILT   DISEASES   OF  OKRA  AND   THE  VERTICILLIUM- 

WILT  PROBLEM 

By  C.  W.  Carpenter,' 

Scientific  Assistant,    Cotton,  Truck,  and  Forage  Crop  Disease  Investigations,  Bureau  of 

Plant  Industry,  United  States  Department  of  Agriculture 

INTRODUCTION 

The  investigation  of  the  okra-wilt  disease  thus  far  reported  has  led 
to  some  confusion  as  to  the  cause  of  this  malady.  Atkinson  (jY  men- 
tions the  fact  that  okra  {Abelmoschtis  esculentus)  is  sometimes  attacked 
by  a  disease  similar  to  the  wilt  of  cotton  (Gossypium  herbaceum).  He 
states  that  a  fungus  for  which  he  proposes  the  name  "Fusarium  vasin- 
fectum"  is  invariably  found  in  the  vascular  system  of  cotton  and  okra 
affected  with  this  disease.  Smith  (25),  in  a  more  extensive  study  of  the 
wilt  diseases  of  cotton,  watermelon  (Citrullus  vulgaris),  and  cowpea 
(Vigna  sinensis),  found  an  ascigerous  fungus  associated  with  the  diseased 
plants.  He  regarded  this  new  genus  as  the  perfect  form  of  Fusarium 
vasinfectum  Atkinson,  and  renamed  the  latter  fungus  "'Neocosmospora 
vasinfecta."  Smith  notes  the  occurrence  of  a  wilt  disease  of  okra,  and 
regarding  A^.  vasinfecta  remarks  (p.  jj) : 

It  probably  occvirs  also  on  okra,  although  the  identification  is  not  complete,  depend- 
ing solely  on  the  character  of  the  symptoms,  on  the  presence  of  similar  macroconidea 
and  microconidia,  and  on  the  occurrence  of  the  disease  in  the  same  localities,  no  cul- 
tures or  cross  inoculations  of  the  okra  ftmgus  having  been  made  and  no  perithecial 
fruits  having  been  discovered. 

Doubts  as  to  the  genetic  relationship  of  Neocosmospora  vasinfecta  and 
Fusarium  vasinfectum  are  expressed  by  Higgins  {12),  Butler  (5),  and 
Wollenweber  {31,  32).  It  appears  from  the  inoculation  and  pure- 
culture  work  of  the  latter  that  A^.  vasinfecta  Smith  is  to  be  regarded  as 
a  saprophyte  entirely  distinct  from  the  vascular  parasite  F.  vasinfectum 
Atk. 

Clinton  {8,  p.  268)  notes  the  occurrence  of  a  wilt  disease  of  okra  in 
Connecticut  which  he  considers  as  the  same  as  that  previously  reported 
by  Atkinson.     The  wilt  of  okra  as  it  occurs  in  North  Carolina  is  described 

'  Since  the  completion  of  these  studies,  the  writer  has  been  transferred  to  the  position  of  Pathologist, 
Hawaiian  Agricultural  Experiment  Station. 
2  Reference  is  made  by  number  (italic)  to  "Literature  cited,"  p.  544-346. 

Journal  of  Agricultural  Research,  Vol.  XII,  No.  9 

Washington,  D.  C.  Mar.  4,  1918 

U                                                                                     ,        ,  KeyNo.  G— IJ7 

(529) 


530  Journal  of  Agricultural  Research  voi.  xii.  No.  9 

by  Stevens  and  Wilson  (25),  with  an  additional  statement  of  the  disease 
by  Wilson  (29).     They  attributed  this  disease  to  Fusarium  vasinfectum. 

No  inoculations  of  okra  plants  with  F.  vasinfectum  are  recorded  by 
Atkinson,  Clinton,  Smith,  or  Stevens  and  Wilson.  A  species  of  Fusarium 
was  found  constantly  associated  with  the  wilt  of  okra,  and  the  disease 
was  regarded  as  the  same  as  the  cotton-wilt.  Orton  {16)  states  that  in 
his  experience  okra  has  contracted  the  disease  when  planted  in  fields 
affected  with  cotton-wilt. 

Wollenweber  (jj,  32)  reports  a  \vilt  disease  of  okra  similar  to  that 
previously  attributed  to  Fusaritim  vasinfectum;  but  Verticillium  albo- 
atrum  Reinke  and  Berthold,  instead  of  the  Fusariumu  fungus,  was  con- 
stantly present  in  the  vascular  ducts.  Successful  inoculations  were  secured 
with  V.  albo-atrum,  and  the  fungus  was  recovered  from  the  wilting  plants. 

It  would  appear  that  there  are  two  similar  wilt  diseases  of  this  crop 
induced  by  two  quite  different  vascular  parasites.  It  was  for  the  purpose 
of  testing  this  latter  hypothesis  that  the  investigation  herein  recorded 
was  undertaken.  While  okra  as  a  crop  is  restricted  in  general  to  the 
home  garden  and  is  relatively  of  little  importance,  a  study  of  the  wilt 
diseases  of  this  plant  is  of  general  significance  to  the  whole  problem  of  wilt 
disease. 

THE  GENERA  VERTICILLIUM  AND  ACROSTALAGMUS  IN  RELATION  TO 

WILT  DISEASES 

Verticillium  albo-atrum  was  first  described  by  Reinke  and  Berthold  (22) 
in  1879  as  the  cause  of  a  wilt  disease  of  potato  {Solanum  tuberosum)  in 
Germany.  This  appears  to  be  the  earliest  record  of  species  of  Verticillium 
associated  with  wilt  disease.  The  work  of  Reinke  and  Berthold  was 
generally  overlooked  until  the  investigation  of  the  leafroll  (Blattroll- 
krankheit)  and  similar  diseases  of  5.  tuberosum  brought  it  again  to  the 
attention  of  pathologists.  Recently  the  presence  of  V.  albo-atrum  in 
wilting  potato  plants  has  been  noted  by  several  investigators :  Appel  (2) , 
Spieckermann  (24),  Pethybridge  (18),  {19),  Stormer  (26),  Wollenweber 
and  Schlumberger  {34),  Wollenweber  (50,  31,  32),  Orton  {17).  Wollen- 
weber {31,  32)  reports  this  organism  as  the  cause  of  a  wilt  disease  of  a 
S.  tuberosum,  S.  melongena,  and  Abelmoschus  (Hibiscus)  esculentus.  A 
similar  disease  of  snapdragon  (Antirrhinum  sp.)  was  described  by 
Brown  (4)  and  the  causal  fungus  determined  by  the  present  writer  as 
V.  albo-atrum.  V.  dahliae,  a  species  closely  related  to  V.  albo-atrum,  is 
described  by  Klebahn  (14)  as  the  cause  of  a  wilting  of  dahlias.  Aderhold 
(i)  also  mentions  a  species  of  Verticillium  in  connection  with  thrombosis 
of  currants  (Ribes  spp. — "Johannisbeere")  and  gooseberry  (Ribes  spp. — 
"  Stachelbeere"). 

The  causal  organisms  of  a  series  of  diseases  of  several  host  plants 
variously  described  as  wilt  disease,  thrombosis,  bluestem,  etc.,  have  been 
relegated  by  the  respective  authors  to  the  genus  Acrostalagmus.     Owing 


Mar.  4. 1918  Wilt  Diseases  of  Okra  531 

to  the  confusion  existing  in  the  genera  Verticillium  and  Acrostalagmus,  a 
discussion  of  the  interrelations  of  these  form  groups  is  necessary  at  this 
time. 

Among  those  writers  who  have  obtained  and  described  Acrostalag- 
mus-like  fungi  may  be  cited  the  following:  Van  Hook  (27)  discusses  a 
wilt  disease  of  ginseng  (Panax  quinque folium)  which  he  found  associated 
with  a  vascular-inhabiting  fungus.  It  is  said  to  closely  resemble  Acro- 
stalagmus alhus  Preuss.  Rankin  (20)  renamed  the  ginseng-wilt  fungus 
"Acrostalagmus  panax"  Rankin,  without  giving  a  technical  description 
of  the  fungus  or  stating  reasons  for  the  change.  Whetzel  and  Rosen- 
baum  (28),  in  mentioning  this  disease  of  ginseng,  attribute  it  to  a  species 
of  Acrostalagmus.  Gueguen  (10)  describes  a  new  species  of  Acro- 
stalagmus causing  a  wilt  disease  of  the  China  aster  Reine-Marguerite 
(Aster  sp.),  and  to  this  fungus  he  gave  the  name  A .  vilmorinii.  A  variety 
of  this  organism  he  later  mentions  (11)  as  associated  with  a  disease  of 
fruits  of  the  cacao  (Theobroma  cacao).  Lawrence  (15)  attributes  a  new 
disease  of  black  raspberry  (Rubus  occidentalis) — that  is,  the  bluestera 
disease — to  a  species  of  Acrostalagmus,  for  which  he  proposes  the  name 
"Acrostalagmus  caulophagus."  Rankin  (21),  in  a  preliminary  report  of 
a  thrombotic  disease  of  maple  trees  (Acer  spp.),  mentions  that  a  species 
of  Acrostalagmus  is  associated  with  the  trouble. 

Corda  (9,  p.  jj)  established  the  genus  Acrostalagmus  to  accommodate 
an  organism,  A.  cinnabarinus ,  which  was  said  to  differ  from  Verticillium 
by  forming  its  spores  in  heads  at  the  tips  of  the  conidiophores.  Hoffman 
(is)  has  rightly  explained  the  complicated  structure  of  the  sterigma  as 
represented  by  Corda,  and  likewise  the  collection  in  heads  of  the  singly 
abscissed  conidia.  The  conidia  are  held  together  by  virtue  of  an  hygro- 
scopic slime,  which  takes  up  water  in  the  presence  of  a  sufficiently  moist 
atmosphere,  forming  a  sphere  of  water  at  the  tip  of  each  sterigma.  In 
these  water  droplets  the  conidia  are  observed  to  float  freely.  With  ex- 
cess moisture  the  drops  rupture  and  the  conidia  slip  down  the  conidio- 
phore.  However,  in  the  absence  of  moisture  the  conidia  cling  tenaciously 
to  the  tips  of  the  sterigma  in  the  form  of  irregular  rounded  masses  of 
varying  sizes.  If  such  material  is  placed  under  a  dry  cover  slip  without 
being  allowed  to  come  in  contact  with  the  latter,  and  examined  micro- 
scopically while  a  small  water  drop  is  brought  into  the  vicinity  of  the 
dry  spore  head,  the  spore  aggregate  will  be  seen  gradually  to  round  up  and 
the  conidia  to  float  more  or  less  actively  within  the  spherule  (PI.  17,  B,D), 

In  the  genera  Acrostalagmus  and  Verticillium  we  now  have  the  follow- 
ing contradictory  and  imperfect  characterization  and  limitation  of  forms. 
In  Verticillium  the  singly  abscissed  and  characteristically  singly  borne 
conidia  may  adhere  in  the  presence  of  moisture,  forming  terminal  heads — 
that  is,  water  droplets  containing  conidia.  In  Acrostalagmus  the 
conidia,  characteristically  united  in  heads,  soon  separate  if  the  humidity 
of  the  environment  is  in  excess  of  the  maximum  for  the  moisture  drops 


532  Journal  of  Agricultural  Research  voi.  xii,  no.  9 

to  retain  their  form,  leaving  but  one  immature  conidium  on  the  sterigma 
tip,  as  in  Verticillium. 

Furthermore,  in  the  genus  VerticilHum,  characterized  by  singly  borne 
or  readily  separating  conidia,  we  have  the  anomalous  condition  of  a 
section  of  the  genus  set  aside  for  those  species  in  which  the  conidia  are 
held  together  by  slime — that  is,  section  Gliocephalum  ^  Saccardo  (1886). 
This  section  of  the  genus  accommodates  forms  in  which  the  conidia  are 
united  in  slimy  heads — that  is,  Acrostalagmus. 

The  fact  that  the  work  of  Reinke  and  Berthold  {22)  was  so  generally 
overlooked  is  responsible  for  the  confusion  of  the  genera  Verticillium  and 
Acrostalagmus.  These  investigators  studied  A.  cinnaharinus,  on  which 
Corda  established  his  genus  Acrostalagmus,  and  concluded  that  this 
form  genus  must  be  united  with  the  older  genus  Verticillium  Nees. 
They  changed  the  name  of  Corda's  fungus  to  Verticillium  cinnaharinum. 
In  view  of  the  above,  there  seems  to  be  no  doubt  that  the  genus  Acro- 
stalagmus Corda  should  be  united  with  the  genus  Verticillium  Nees. 
Klebahn  {14)  supports  this  view. 

The  species  of  Acrostalagmus,  described  as  causing  a  vascular  disease 
of  ginseng,  China  aster,  and  black  raspberry,  may  prove  upon  further 
work  to  be  identical  with  each  other  and  with  V.  alho-atrum,  as  proved 
in  this  paper,  for  the  strains  of  the  latter  fungus  isolated  from  okra, 
cotton,  eggplant  (Solanum  melongena),  and  potato.  Culturally,  at  least, 
the  strains  of  the  species  of  Acrostalagmus  from  ginseng  and  raspberry, 
in  so  far  as  they  have  been  studied  by  the  writer,  are  not  to  be  dis- 
tinguished from  Verticillium  alho-atrum.  The  organisms  A.  vilmorinii 
Gueguen  and  V.  dahliae  Klebahn,  as  described  by  their  authors,  hardly 
differ  sufficiently  from  V.  alho-atrum  to  be  given  specific  rank.  Possibly 
this  is  the  better  treatment  of  such  related  forms  until  their  identity 
is  established  by  careful  cross-inoculation  work  with  pure  cultures. 
The  minor  cultural  differences  found  by  the  writer  in  cultures  of  V. 
alho-atrum  from  different  hosts  are  present  in  parallel  cultures  of  the 
same  strain.  Strains  of  V.  alho-atrum  which  are  morphogically  indis- 
tinguishable, isolated  from  different  hosts,  and  capable  of  producing 
the  same  s)Tiiptoms  of  disease  by  cross-inoculation,  seemingly  should 
be  considered  identical. 

The  available  data  on  the  several  strains  of  Verticillium  and  Acro- 
stalagmus thus  far  described  as  plant  parasites  are  brought  together  in 
tabular  form  (Table  I),  in  order  that  comparison  may  be  readily  made. 
The  hypothetical  identity  of  all  of  these  strains,  suggested  by  the  simi- 
larity of  this  data,  is  strengthened  if  one  consults  and  compares  the 
descriptions  by  the  several  authors.  The  existing  differences  seem  to 
be  those  of  variety  rather  than  of  species. 

1  Saccardo's  section  Gliocephalum  of  the  genus  Verticillium  (Saccardo,  P.  A.  svxlogb  funoorum. 
V.  4,  p.  139.  Patavii,  1886)  is  given  as  Gliocladium  by  Engler  and  Prantl  (1900)  apparently  through  error. 
(Engler,  A.,andPRANTi„K.  naturlichen  PFLANzeNFAMHiEN.  Teil I,  Abt.  I**,  p.  418,432.  Leipzig, 
1900.) 


Mar.  4,  1918 


Wilt  Diseases  of  Okra 


533 


Table  I. — Data  of  several  species  and  strains  of  Verticillium  and  Acrostalagmus  arranged 

for  comparison 


Conidia. 

Sclerotia. 

Sclerotial  cells. 

Length  of 

Organism. 

Limits  of  size.    Average  size. 

Limits  of 

size. 

Average 

size. 

Limits  of 
size. 

Average 
size. 

conidio- 

phore 
branches. 

V.   clbo  -   alTum 
strain  1717  Jrom 
okra. 

V.   albo  -   atrum 

5.1   to    II   by  I  6.8by2. 4... 
2. 1  to  j.C.      ; 

4.  2  to  9. 3  by  !  6. 4  bv  2. 4. . . 

/i. 

21  to  91 

31  to  70 
31  to  87 
28  to  70 

34  by  44 
45  by  54 
SI  by  SI 
39  by  57 

7  to  10 
s  to  10 

8. 4  by  8.  7 
8.  2  by  8.  2 
7-  3  by  9.  4 

A". 

13.  5  to  22.0 

strain  2784  from 
potato. 
V.   albo  -   atrum 

I.  7  to  4.  2. 

5. 1  to  8.  5  by 
2. 1  to  4.  2. 

6.  s  by  2.  6. . . 

strain  1685  from 
eggplant. 

3  to  15 

15. 0  to  38.0 

strain  2985  from 
snapdragon. 
V.  albo  -  atrum  bv 

5.0  to  12.0  by 
2. 0  to  3.  5. 

4. 0  to  6. 0  by 

l.St0  2.5. 

4. 0  to  7.  0  by 
1.5  to  2.0. 

2.  0  to  5.  0  by 

1.  0  to  2.  c. 

3.  0  to  7.  0  by 

2.  0  to  2-  0. 

5.  0  to  7. 0  by 
:.  5  to  3.  0. 

6.  0  to  7.  0  by 
3. 0  to  4. 0. 

Reinke        and 
Berthold. 

54 
50 

12. 0  to  45.0 

16. 0  to  27. 0 
la  ot0  25. 0 

given  by  Kle- 
bahn. 

V.  dakliae  by  Kle- 

Stc6 

bahn. 

Hook. 

7  to  12 

by  Lawrence. 
A.   vilnn)rinii  by 

20  to  70 

50  to  60 
None. 

Guegtien    from 
China  aster. 

cacao. 

For  the  reasons  above  set  forth,  it  is  evident  that  the  genus  Acro- 
stalagmus must  be  united  with  the  older  genus  Verticillium.  In  view 
of  this  condition,  the  following  forms,  if  not  identical  with  V.  albo- 
atrum,  must  at  least  be  regarded  as  belonging  to  the  genus  Verticillium : 
.4.  albus  from  ginseng-wilt  disease,  described  by  Van  Hook  (27)  and 
renamed  A.  panax  by  Rankin  (20);  A.  vilmorinii  Guegen,  from  the  China 
aster  and  the  fruits  of  the  cacao;  and  A.  caulophagus  Lawrence,  the 
cause  of  the  bluestem  disease  of  black  raspberry.  Of  the  remaining 
12  imperfectly  characterized  and  apparently  saprophytic  species  of  the 
genus  Acrostalagmus  some  are  no  doubt  identical  with  species  of 
Verticillium,  and  are  just  as  incompletely  delineated.  The  ubiquitous 
V.  lateritium  Berk.,  commonly  present  on  decaying  potatoes,  is  possibly 
the  same  as  V.    innabariniitn  (Corda)  Reinke  and  Berthold. 

FUSARIUM  VASINFECTUM  AND  VERTICILLIUM  ALBO-ATRUM,  CAUSES 
OF  THE  WILT  DISEASES  OF  OKRA 

FUSARIUM    VASINFECTUM 

The  species  of  Fusarium  from  okra-  and  cotton-wilt  considered  in  this 
work  are  regarded  as  identical  with  each  other  and  with  the  species  of 
Fusarium  isolated  from  "frenching"  cotton  by  Atkinson  (5)  and  named 
by  him  "F.  vasinjccium."  The  cross-inoculation  work  herein  recorded 
proves  the  casual  relation  of  F.  vasinfecium  to  the  wilt  disease  of  okra, 


534  Journal  of  Agricultural  Research  voi.  xn,  Na^ 

and  establishes  the  pathologic  identity  of  the  cotton-  and  okra-wilt 
strains  beyond  reasonable  doubt. 

Certain  strains  of  the  species  of  Fusarium  causing  cotton-wilt  were 
observed  by  WoUenweber  (ji,  32)  to  produce  an  aromatic  odor  (lilac) 
when  cultured  on  starchy  media  such  as  rice;  other  strains  less  com- 
monly isolated  lacked  this  property.  These  latter  strains  were  desig- 
nated "  F.  vasinfectum  var.  inodoratum"  by  WoUenweber.  As  a  further 
indication  of  the  identity  of  the  species  of  Fusarium  on  okra  with  the 
species  of  Fusarium  on  cotton,  it  should  be  noted  that  both  the  odor- 
forming  and  the  non-odor-forming  strains  have  been  isolated  several 
times  from  okra.  While  the  ability  to  generate  this  odor  is  of  doubtful 
specific  value,  since  other  species  of  the  section  Elegans  (WoUenweber, 
32)  possess  this  property,  and  this  ability  has  been  observed  to  be  lost 
in  culture  {Carpenter  7,  p.  206),  the  fact  that  the  species  of  Fusarium 
from  okra  and  cotton  agree  in  this  character  is  significant. 

Normal  ^  cultures  of  this  species  of  Fusarium  develop  in  i  to  3  weeks 
at  room  temperature  (PI.  A).  The  best  results  are  secured  with  plant 
stems,  potato  cylinders,  and  other  vegetable  media.  Morphologically 
F.  vasinfectum  is  scarcely  distinguished  from  the  other  vascular  parasites 
of  the  section  Elegans  of  this  genus. 

Potato-cylinder  cultures  (PI.  A)  develop  an  ocherous-salmon-colored 
pionnotes  with  3-  to  5-septate  conidia  (PI.  17,  L-M).  Blue-gray  sclerotia, 
similar  to  those  of  F.  oxysporum  Schlecht.  (7,  PI.  A,  i).  are  generally 
present  on  this  medium.  The  slight  violet  color  of  the  upper  part  of  the 
pionnotes  as  represented  in  Plate  A,  3,  has  never  been  seen  in  other 
closely  related  species  and  is  possibly  of  differential  value.  Stems  of 
Meliloius  alba  and  Gossypium  sp.  are  useful  in  developing  the  sporodochia, 
which  are  likewise  of  an  ocherous-salmon  color.  The  plate  of  F.  oxy- 
sporum in  an  earlier  paper  (7,  PL  A,  2,  5)  illustrates  stem  cultures  of  F. 
vasinfectum  equally  well.  On  steamed-rice  medium  a  more  or  less 
brilliant  red  color  soon  appears,  later  becoming  tinged  with  various 
shades  of  purple  and  blue,  especially  in  subdued  light.  Normal  conidia 
are  not  usually  present  in  rice  cultures,  the  value  of  this  medium  being 
the  color  reaction  and  the  formation  of  an  aromatic  substance  with  an 
odor  suggesting  lilac  (7,  p.  206).  Chlamydospores  (PI,  17,  I)  are  formed 
in  large  numbers  on  this  medium. 

The  following  measurements  show  the  size  and  percentage  of  the 
variously  septate,  normal,  conidia  (PI,  17,  L,  M)  found  in  strains  of  F. 
vasinfectum  from  okra  and  cotton : 

F.  vasinfectum,  strain  2709,  isolated  from  okra.  Culture,  27-day-old 
stem  of  Melilotus  alba,  without  pionnotes.  Normal  triseptate  conidia, 
60  per  cent.  Limits  of  size:  25.5  to  40.8  by  4.2  to  5.1  ju.  Largest  nor- 
mal triseptate  conidium,  40.8  by  5.1  ix.;  smallest,  25.5  by  4.2  /i.     Average 

'  For  a  discussion  of  the  idea  "nonnal"  and  other  special  terms  as  used  in  relation  to  species  of  Fusarium, 
see  WoUenweber  (jj, />.  255-257). 


Mar.  4,  1918 


Wilt  Diseases  of  Okra 


535 


size  of  10  normal  triseptate  conidia,  32.5  by  4.6  fx.     One-septate  conidia, 
20  per  cent.     Abnormal  conidia  and  microconidia,  20  per  cent. 

Normal  conidia  on  rice  culture  2H  months  old:  Triseptate,  23.8  to 
30.0  by  4.0  to  5.0  IX.     Five-septate  rare,  56.0  by  3.6  fx. 

F.  vasinfectum  strain  3203  from  okra.  Pionnotes  on  a  25-day-old 
stem  culture  of  MeliloUis  alba.  Limits  of  size  of  28  normal  triseptate 
conidia:  23.8  to  34.0  by  3.8  to  4.2  tx.  Average  size,  28.4  by  3.9  /u. 
Largest  triseptate  conidium,  34.0  by  4.2  fx.  Smallest  triseptate  coni- 
dium,  23.8  by  3.8  (x. 

F.  vasinfectum  strain  3242  from  okra.  Normal  triseptate  conidia, 
34.0  to  41.0  by  3.4  to  5.0  fx.     Four-septate  conidium  40.8  by  4.2  [x. 

F.  vasinfectum  var.  inodoratum  strain  3257  from  okra.  Thirty-four- 
day-old  Meliloius  stem  culture  with  pionnotes.  Normal  triseptate  coni- 
dia 60  per  cent.  Limits  of  17  spores,  32.3  to  45.9  by  3.4  to  4.8  /x.  Aver- 
age, 37.4  by  4.2JU.  Four-septate,  35  per  cent.  Limits  of  13  spores,  34.0 
to  44.2  by  4.2  to  4.8  fx.  Average,  39.0  by  4.2  m-  Five-septate,  5  per 
cent.  Limits  of  four  spores,  37.4  to  45.9  by  4.2  to  4.9  /u.,  average  four 
spores  42.0  by  4.4  m- 

F.  vasinfectum  var.  inodoratum,  strain  3258  from  okra.  Thirty-four- 
day-old  pionnotes  culture.  Normal  triseptate  conidia  76  per  cent. 
Limits  of  9  spores,  27.2  to  41.6  by  3.4  to  4.6  /u.  Average,  34.o»by 
4.2  ju.  Four-septate  conidia,  19  per  cent.  Limits  of  13  spores,  35.7  to 
42.5  by  3.4  to  5.0  IX.  Average,  37.8  by  4.1  ix.  Five-septate,  5  per  cent. 
Limits  of  6  spores,  34.0  to  45.9  by  3.8  to  4.7  ix.     Average,  40.0  by  4.2  /u. 

Parallel  cultures  of  F.  vasinfectum,  strain  1855,  from  cotton  and  strain 
3592  from  okra  were  prepared  upon  cotton  stems  as  a  medium.  At  the 
age  of  45  days,  50  normal  triseptate  conidia  from  a  pionnotes  of  each 
culture  were  measured  to  obtain  the  relative  sizes.  The  maximum 
minimum,  and  average  length  and  width  of  conidia  in  each  culture  are 
shown  in  Table  IL  There  is  a  close  correspondence  of  the  measurements 
of  conidia  from  the  two  strains,  considering  that  these  measurements 
were  made  from  but  one  culture  of  each  strain. 

Table  II. — Comparative  size  of  normal  triseptate  conidia  of  Fusarium,  vasinfectum  from 

okra  and  cotton 


Maximum  length. 
Minimum  length . 
Maximum  width. 
Minimum  width . 
Average  length . . . 
Average  width . . . 


F.  vasinfectum 

1855  from 

cotton. 


5-1 

3-4 

37- o 

4.2 


F.  vasinfectum 

3592  from 

okra. 


44. 
27. 

5- 
3- 

35- 
4- 


536  Journal  of  Agricultural  Research  voi.  xii.  no. 9 

Fusarium  vasiofectum  Atkinson. 

Sporodochia  and  perfect  pionnotes  present,  in  mass  ocherous-salmon  colored,  the 
conidia  being  of  the  Elegans  type,  3-  to  5  septate,  sickle-shaped,  constricted  at  the 
apex  and  pedicillate  at  the  base  (PI.  17,  L,M).  Conidiophores  verticillately  branched. 
Normal  triseptate  conidia  present  up  to  100  per  cent,  23.8  to  46.0  by  3.4  to  5.1  m- 
Four-septate  conidia  up  to  35  per  cent,  34.0  to  44.0  by  3.4  to  5.0  m-  Five-septate 
conidia  up  to  5  per  cent,  34.0  to  56.0  by  3.6  to  5.0  n.  Microconidia  may  be  present 
in  subnormal  cultures  up  to  100  per  cent,  4.0  to  14.0  by  2.  o  to  3.5  m-  in  size.  Chlaray- 
dospores  (PI.  17, 1)  ellipsoidal  to  round,  terminal,  intercalary  and  conidial;  when 
unicellular  measvtring  from  8  to  1 5  m-  Blue-gray  sclerotia  on  potato  cylinders.  Strong 
lilac  odor  on  rice  and  other  starchy  media.  Vascular  parasite,  cause  of  a  wilt  disease 
of  Gossypium  Jierbaceum,  G.  barbadense,  and  Abelmoschus  esculentus. 

VERTICILLIUM    ALBO-ATRUM 

Veriicillium  albo-atrum  Reinke  and  Berthold  is  classified  by  Engler 
and  Prantl  in  the  section  Eu-Verticillium  of  the  genus  Verticillium  of 
the  Mucedinaceae-Hyalosporae-Verticillieae.  The  conidophores  (PI.  17, 
A-D)  are  verticillately  branched,  conidia  which  readily  fall  being  formed 
at  the  tips  of  all  the  branches.  The  distinction  between  the  three  sec- 
tions of  the  genus — that  is,  Eu-Verticillium  Sacc,  Oncocladium  Wallr., 
and  Gliocephalum  Sacc.  is  not  sharply  drawn.  In  the  latter  the  conidia 
are  held  together  by  slime,  while  in  Eu-Verticillium  and  Oncocladium 
this  is  not  the  case.  In  the  opinion  of  the  writer,  V.  albo-atrum  should 
more  appropriately  be  placed  in  the  section  Gliocephalum,  for  this  is 
where  it  naturally  belongs,  if  its  characters  are  determined  on  the  sub- 
stratum (PI.  17,  B,  D).  If  examined  in  water  mounts,  rarely  more  than 
one  conidium  would  be  found  on  each  sterigma  tip  (PI.  17,  C),  and  the 
fungus  would  erroneously  be  placed  with  the  section  Eu-Verticillium. 

DESCRIPTION  OF  THE  FUNGUS 

The  conidia  are  ellipsoidal  (PI.  17,  A),  unicellular,  4.0  to  ii.o  by  1.7 
to  4.2  n,  abscissed  singly  from  the  tips  of  verticillate-branched  conidio- 
phores. They  may  or  may  not  cling  to  the  tips  of  the  sterigma  in  rounded 
masses.  In  the  absence  of  sufficient  moisture  in  the  air,  relatively  dry, 
rounded  aggregates  of  spores  accumulate ;  but  with  more  moisture  present 
spherical  drops  appear  on  the  sterigma  tips  by  virtue  of  the  hygroscopic 
slime  in  which  the  conidia  are  embedded  (PI.  17,  B).  With  additional 
moisture  the  drops  rupture,  leaving  one  immature  conidium  clinging  to 
the  sterigma.  These  masses  measure  from  3  /*  to  a  size  where  the  water 
drop  breaks.  The  verticillate  branches  of  the  conidiophores  are  i  to  7  to 
a  whorl  or  virtel,  more  commonly  3  to  5,  and  these  in  turn  may  bear 
secondary  branches  in  virtels.  The  branches  are  from  13  to  38  ^  long, 
disposed  in  virtels  30  to  38  fx  apart  along  the  conidiophore.  Conidiophores 
consisting  of  a  terminal  branch  and  two  primary  virtels,  are  about  100 
to  120  M  in  length,  while  those  bearing  four  primary  virtels  measure  250 
to  300  fi.     Conidiophores  with  7  to  8  primary  virtels  are  occasionally 


Mar.4,  i9i8  WUt  Diseases  of  Okra  537 


seen  in  petri-dish  cultures.  The  terminal  branch  is  usually  i  to  3  times 
longer  than  the  virtel  branches  and  measures  from  1 5  to  60  ju. 

The  mycelium  is  hyalin  at  first,  becoming  brownish  with  age.  It  is 
septate,  2  to  4  jit  in  diameter,  but  is  often  swollen,  all  gradations  from 
slightly  swollen  threads  to  large,  thick-walled,  and  knotted  sclerotia, 
and  chlamydospore-like  cells  being  present  (PI.  17,  G).  The  sclerotium 
is  at  first  but  swollen,  closely  septate  mycelium,  which  enlarges  and 
knots  itself  into  a  variety  of  forms.  The  measurements  of  sclerotia  of 
strains  1685,  1717,  2784,  and  2985,  recorded  in  Table  I,  were  made  only 
from  the  more  or  less  rounded  aggregations.  Such  measurements  of 
irregular  formations  varying  so  greatly  in  size  are  only  of  general  sig- 
nificance. Certain  strains  of  V.  albo-atrum  cultivated  by  the  writer 
produce  sclerotia  abundantly  in  a  few  days  in  petri-dish  cultures.  Macro- 
scopically,  their  presence  is  manifested  by  a  beautiful  black  zonation  of 
the  colonies,  as  illustrated  in  Plate  19.  In  other  strains,  and  sometimes 
in  other  cultures  of  the  same  strain,  these  sclerotial  rings  either  do  not 
develop  at  all  or  only  after  a  long  time ;  yet  these  forms  produce  abun- 
dant sclerotia  if  cultivated  on  potato  cylinders  and  other  vegetable  media. 
The  entire  growth  then  frequently  consists  of  a  black  confluent  layer  of 
sclerotia  and  hyphae.  Parallel  cultures  of  the  same  strain  differ  suffi- 
ciently with  respect  to  the  characters  of  the  conidiophores  and  sclerotia  for- 
mation so  that  specific  determinations  based  on  slight  differences  of  these 
characters  and  unsupported  by  inoculation  tests  are  of  doubtful  value. 
Verticillium  albo-atnun  Reinke  and  Berthold. 

Conidia  ellipsoidal,  iinicellular,  4.0  to  11. o  by  1.7  to  4.2  ft,  abscissed  singly  from 
the  sterigma  tips  of  verticillate  conidiophores.  Primary  whorls  or  virtels  of  branches, 
I  to  8  in  number,  30  to  90  n  apart,  sometimes  bearing  secondar}-  virtels.  Branches 
I  to  7,  usually  3  to  5  in  number,  13  to  38  /x  long,  tapering,  straight  to  slightly  bowed. 
Conidiophores  100  to  300  ^  or  more  in  length.  The  terminal  branch  of  the  conidio- 
phore  is  from  1 5  to  60  m  long.  Conidia  may  or  may  not  collect  in  heads  on  the  sterigma 
tips.  Mycelium  septate,  hyalin  to  brown  with  age;  may  become  swollen  into 
chlamydospore-like  chains  of  closely  septate,  knotted  masses.  These  aggregates 
constitute  the  sclerotia  of  this  fungus.  Vascular  parasite,  cause  of  a  wilt  disease  of 
okra,  potato,  eggplant,  cotton,  snapdragon,  and  probably  of  species  of  Abutilon  and 
Xanthium,  ginseng,  black  raspberry,  China  aster,  and  dahlia.  V.  albo-airum  may 
prove  to  be  the  cause  of  the  Verticillium-wilt  disease  reported  on  currants  and  goose- 
berries by  Aderhold  (/). 

OCCURRENCE  OF  FUSARIUM  VASINFECTUM  .\ND  VERTICILLIUM  ALBO- 
ATRUM  IN  WILT  DISEASES.  OF  OKRA 

VerticUlium  albo-atrum  was  found  constantly  inhabiting  the  vascular 
system  of  wilt-diseased  okra  plants  (PI.  27)  in  New  Jersey,  where  this 
crop  is  of  considerable  importance.  This  organism  was  also  isolated  from 
similar  material  from  Monetta,  S.  C,  Birmingham,  Ala.,  Middle  River, 
Cal.,  and  Medford,  Oreg.,  specimens  from  Middle  River  and  Medford 
having  been  collected  by  Dr.  Wollenweber. 

On  the  other  hand,  a  Fusariura  indistinguishable  from  the  cotton- 
wilt  Fusarium  was  constantly  obtained  from  wilting  okra  collected  at 


538  Journal  of  Agricultural  Research  voI.xii.no.? 

Florence,  Sumter,  and  Charleston,  S.  C,  and  Wrightsboro,  N.  C.  Accord- 
ing to  Clinton  (8),  the  Fusarium-wilt  of  okra  occurs  in  Connecticut. 
Fusarium  vasinfectum  was  not  obtained  by  the  writer  from  okra-wilt  in 
New  Jersey;  and  VerticUlium  alho-atrum  was  isolated  from  this  host  in 
the  South  only  in  the  few  mentioned  localities. 

V.  alho-atrum  was  isolated  from  a  wilt  disease  of  the  weeds  AhutUon 
sp.  and  Xanthium  sp.  in  New  Jersey ;  and  from  spontaneous  wilt  of  cotton 
plants  in  rows  adjoining  the  experimental  plot  at  Arlington,  Va. 

The  wilt  diseases  of  the  several  plants  brought  about  by  F.  vasinfectum 
and  V.  albo-atrum  manifest  the  same  symptoms,  so  that  the  real  cause  of 
the  trouble  is  safely  to  be  determined  only  by  cultural  means.  There  is 
a  lack  of  turgor  in  the  leaves  first  in  evidence  in  those  parts  farthest 
removed  from  the  veins  (PI.  20-23).  The  lower  leaves  are  first  affected, 
wilt,  and  drop  ofif  one  or  two  at  a  time.  Frequently  the  plant  does 
not  die  for  a  long  time,  but  continues  a  dwarfed  existence.  This  is 
especially  true  of  the  Verticillium-wilt.  If  the  plants  are  cut  longitudi- 
nally and  crosswise,  it  will  be  seen  that  the  vascular  tissue  is  brown  or 
black  (PI.  18),  the  discloration  being  traceable  from  the  small  roots  to  the 
top  of  the  stem  and  into  the  petioles.  Microscopic  examination  of  thin 
sections  of  this  material  shows  that  the  vessels  are  plugged  with  the 
mycelium  of  the  parasite,  which  interferes  with  the  conduction  of  moisture 
to  the  aerial  portions  of  the  plant  (PI.  17,  E).  The  host  tissues  do  not 
appear  to  be  invaded  or  broken  down,  the  vascular  inhabitant  merely 
living  as  a  saprophyte  on  the  fluids  of  the  vessels,  and  injuring  the  host 
plant  only  by  mechanical  obstruction  of  the  latter.  Whether  there  are 
injurious  products  secreted  in  the  metabolism  of  the  fungus  detrimental 
to  the  plant  in  other  ways  is  yet  to  be  demonstrated.  The  parasitism  of 
these  fungi  is,  then,  a  mechanical  interference  with  the  nutrition  of  the 
host  and  not  our  usual  conception  of  this  term. 

METHOD  OF  TESTING  PARASITISM 

In  order  to  demonstrate  the  ability  of  V.  albo-atrum  and  F.  vasinfectum 
to  produce  wilt  diseases  of  okra  and  to  gain  a  knowledge  of  the  relation 
of  these  organisms  to  other  host  plants,  approximately  i  ,000  inoculations  * 
and  cross-inoculations  with  pure  cultures  were  made.  Strains  of  V.  albo- 
atrum,  isolated  from  okra,  eggplant,  potato,  and  snapdragon,  were  used 
to  inoculate  okra;  strains  from  okra,  snapdragon,  and  eggplant  were 
used  to  inoculate  eggplant;  and  the  strain  from  okra  was  used  to  inocu- 
late cotton.  Similarly  okra  plants  were  inoculated  with  F.  vasinfectum 
isolated  from  okra  and  cotton ;  and  cotton  was  inoculated  with  this  fungus 
isolated  from  cotton  and  okra. 

The  general  method  used  in  the  inoculations  may  be  summarized  as 
follows:  Selected  okra  seed,  or  seed  of  other  plants  to  be  used,  were  disin- 
fected in  a  solution  of  formalin,  rinsed  in  sterile  water,  and  planted  in 

>  The  writer  is  indebted  to  Mr.  J.  M.  R.  Adams  for  faithful  assistance  in  the  inoculation  tests. 


Mar.  4. 1918  Wilt  Diseases  of  Okra  539 


sterilized  soil  in  the  greenhouse,  or  in  soil  new  to  these  crops  in  field 
plots.  Soil  inoculations  were  made  by  pouring  a  few  cubic  centimeters 
of  a  sterile-water  spore  suspension  of  the  organism  to  be  used  on  the 
steam-sterilized  soil,  either  before  or  after  planting  the  seed,  or  in  the 
vicinity  of  the  seedlings.  The  majority  of  the  inoculations  were  made 
through  wounds  at  the  hypocotyl  below  the  soil  level.  The  wounds 
were  made  with  a  sterile  scalpel,  and  after  the  inoculum  was  intoduced, 
the  soil  was  replaced  to  prevent  drying.  A  few  inoculations  were  made 
through  wounds  made  by  breaking  off  leaves  and  pods  of  the  okra  plants. 

Approximately  as  many  control  plants  were  prepared  as  plants  for 
inoculation.  Half  of  these  were  wounded  or  otherwise  subjected  to  the 
treatment  employed  on  the  plants  inoculated,  and  the  other  half  were 
left  as  imwounded  controls.  For  the  wounded  controls,  sterile  water 
was  used  for  inoculating,  in  lieu  of  the  spore  suspension.  Wilting  plants 
were  examined  microscopically  for  mycelium  in  the  xykm  of  roots, 
stem,  or  petioles;  and  pure  cultures  of  the  organism  used  were  reisolated 
and  identified  as  a  control  on  the  work.  Frequently  such  reisolated 
strains  were  used  for  subsequent  inoculations. 

In  but  one  experiment  was  there  any  wilt  in  the  controls.  In  this 
case  after  six  weeks  6  per  cent  wilted,  and  the  organism  used  in  the 
experiment  was  recovered  from  the  interior  of  the  plants.  However, 
this  was  a  field  test  and  there  was  in  the  duration  of  the  experiment 
an  opportunity  for  infection  from  inoculated  plants  in  the  adjoining 
rows  on  either  side. 

INOCULATION  OF  VARIOUS  ECONOMIC  PLANTS  WITH  VERTICILLIUM 
ALBO-ATRUM  AND  FUSARIUM  VASINFECTUM 

Inasmuch  as  the  inoculation  experiments  were  carried  on  in  a  uniform 
manner  and  controlled  by  a  large  number  of  wounded  and  unwounded 
plants,  as  well  as  by  reisolation  and  identification  of  the  organisms 
causing  the  disease,  the  results  are  comparable.  For  convenience  these 
results  are  brought  together  in  tabular  form  (Tables  III,  IV,  V)  and  the 
discussion  of  the  results  of  the  inoculations  are  arranged  according  to  the 
host  plant  and  the  parasite  used. 

HISTORY  OF  THE  STRAINS  OF  VERTICILUUM   ALBO-ATRUM   AND  FUSARIUM 
VASINFECTUM    USED   FOR   INOCULATION 

The  history  of  the  various  strains  of  V.  albo-atrum  and  F.  vasinfectum 
used  for  inoculating  the  several  host  plants  is  as  follows : 

F.  vasinfectum  strain  1855.  Reisolated  from  the  vascular  system  of  a 
cotton  plant,  which  was  wilting  as  a  result  of  inoculation  with  strain  1733, 
a  reisolation  of  strain  1635,  which  was  in  turn  a  reisolation  of  the  original 
strain  1485,  isolated  from  the  discolored  vascular  system  of  a  wilting 
cotton  plant  at  Florence,  S.  C,  in  191 2. 

F.  vasinfectum  strains  2708,  2709,  3203.  Isolated  from  the  discolored 
vascular  system  of  wilting  okra  plants  collected  at  Florence,  S.  C.  at 


540 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  9 


different  times.  Strain  3592  is  a  reisolation  from  an  okra  plant  inoculated 
with  strain  3203. 

F.  vasinfecium  strain  3210.  Isolated  from  the  vascular  system  of  an 
okra  plant  collected  at  Sumter,  S.  C. 

V.  albo-airum  strain  1717.  Isolated  from  the  interior  of  the  stem  of  a 
wilting  okra  plant  at  Monetta,  S.  C.  Strains  2943,  3075,  3076,  and  3156 
are  reisolations  from  okra  plants  inoculated  with  strain  171 7. 

V.  albo-atrum  strain  2821.  Isolated  from  wilting  okra  plant  collected 
at  Middle  River,  Cal. 

V.  albo-atrum  strain  1685.  Isolated  from  wiltmg  plant  of  Solanum 
melongena. 

V.  albo-atrum  strain  2784.     Isolated  from  wilting  potato  plant. 

V.  albo-atrum  strain  2985.  Isolated  from  wilting  snapdragon  plant 
by  Brown  (4). 

Table  III. — Results  of  inoculating  okra  plants  with  Verticillium  albo-airumfrom  various 

hosts 


Num- 
ber of 
plants 

Age. 

Locality. 

Method  of 
inoculation. 

Species  and  strain  of  organism. 

Incu- 
bation 
pe- 
riod. 

Per- 
cent- 
age 
ulti- 
mate- 
ly 
wilt- 
ing. 

Time. 

40 
60 

Days. 
17 

n 

37 

46 
90 

30 
46 

46 

46 

90 

46 

46 

30 
60 

60 

60 

60 

60 

Greenhouse . 
do 

Soil 

V.      albo-atrum,     okra 

(1717)- 
do 

Days. 

12 

40 
40 

II 

13 
18 
12 

15 

13 
13 
18 

88 

59 

55 

100 

81 

100 

92 

66 

17 
40 

ICO 

0 

0 

0 
75 

71 

36 

73 

45 

Days. 
21 

do 

60 

60 

do 

do 

V.     albo-airum,      Irish 

potato  (2784). 
V.     albo-atrum,      okra 

(1717). 
do 

60 

40 

TlS 

do 

do 

Wounds:    hy- 

pocotj'l. 
....  do 

30 
^O 

do 

do 

do 

^0 

13 

do 

do 

do 

10 

do 

do 

V.      albo-airum,      okra 

(3075)- 
V.     albo-atrum,      okta 

(3076). 
V.      albo-airum,      okra 

(2943)- 

V.      albo-atrum,      okra 
(2821). 

V.     albo-atrum,     snap- 
dragon (2985). 

V.      albo-atrum,      okra 

(1717). 
do 

30 

17 

40 

9 
40 
40 
26 

do 

do 

do 

do 

do 

do 

do 

do 

30 
30 
24 

30 

do 

do 

Wounds:  ped- 
icel. 
....do 

30 

45 
43 

23 

58 
14 
19 
26 

Field 

..    .do..    . 

Wounds:    hy- 

pocotyl. 
do 

do 

10 

14 
10 
10 
10 

V.     albo-atrum,      okra 

(3156). 

V.  albo-airum,  eggplant 
(1685). 

V.    albo-atrum,      Insh 
potato  (2784). 

V.     albo-atrum,     snap- 
dragon (2985.). 

do 

do 

43 

do 

do 

43 

22 

.  do... . 

do 

43 

"  Soil  inoculated  before  planting  the  seed. 


Mar.  4,  igi8 


''ill  Diseases  of  Okra 


541 


Table  IV. — Results  of  inoculating  okra  plants  with  Fusarium  vasinfectum  from  cotton 

and  okra 


Num- 

ber of 
plants 

Age. 

Days. 

40 

17 

40 

17 

40 

37 

99 

21 

40 

37 

40 

37 

30 

24 

50 

49 

100 

10 

20 

60 

17 

21 

Locality. 


Greenhouse 

do 

....  do 


.do. 
.do. 

.do. 

.do. 


.do. 
.do. 
.do. 


.do. 


_  Method  of 
inoculation. 


Soil. 


.do. 


Wounds:    hy- 
pocotyl. 

do 

do 


.do. 


.do. 


.do. 
.do. 
.do. 


.do. 


Species  and  strain  of  organism. 


F.   vasinfectum,      okra 

(2709). 
F.    vasinfectum.,  cotton 

(1855)- 
do 


.do. 


F.     vasinfectum;     okra 

(2709). 
F.    vasinfectum,     okra 

(2708). 
F.     vasinfectum,      okra 

(3203). 

do 

do 

F.     vasinfectum,      okra 

(3210). 
F.     vasinfectum,     okra 

(3592)- 


Incu- 
bation 
pe- 
riod. 


Days. 


17 


17 


Per- 
cent- 
ape 
ulti- 
mate- 
ly 
wilt- 
ing. 


50 


55 


Time. 


Days. 
56 

56 

5f 

36 
56 

56 

15 

30 
18 

23 
36 


Table  V. — Results  of  inoculating  cotton,  eggplant,  and  Brabham  coivpeas  with  tite  okra- 
wilt  organisms  through  wounds  at  the  hypocotyl 


Host. 


Num- 

ber 
of 

Age. 

plants. 

Days. 

14 

60 

20 

60 

18 

21 

43 

21 

7 

45 

5 

45 

7 

45 

16 

24 

Ixxality. 


Species  and  strain  of 
organism. 


Per- 

Incu- 
ba- 
tion 
peri- 
od. 

cent- 
age 
ulti- 
mate- 
ly 
wilt- 

ing. 

Days. 

0 

14 

80 

17 

89 

17 

87 

10 

85 

10 

80 

10 

100 

0 

Time. 


Cotton 

Do 

Do 

Do 

Eggplant 

Do 

Do 

Brabham  cow- 
peas. 


Field 

....do 

Greenhouse  . 

....do 

....do 

do 

....do 

....do 


F.    vasinfectum ,   okra 

(3210). 
V.    albo-atrum,     okra 

(3156). 
F.  vasinfectum,  cotton 

(1855). 
F.     vasinfectum,    okra 

(3592). 
V.    albo-atrum,     okra 

(1717)- 

V.  albo-atrum,  snap- 
dragon (2985). 

V.  albo-atrum,  egg- 
plant (1685). 

F.  vasinfectum,  olcra 
(3203). 


Days. 
23 

23 
36 
36 
15 
15 
15 
60 


^42  Journal  of  Agricultural  Research  voi.  xii,  no.  9 

INOCULATION    OF   OKRA 

No  difficulty  was  experienced  in  securing  typical  wilting  okra  plants 
(PI.  21,  22)  by  the  inoculation  of  the  soil  or  of  the  plants  through  wounds 
at  the  hypocotyl  with  pure  cultures  of  V.  albo-airum  isolated  from  okra, 
eggplant,  potato,  or  snapdragon.  By  the  inoculation  of  the  soil  in  the 
vicinity  of  17-day-old  plants  with  strain  171 7  isolated  from  okra,  88 
per  cent  of  wilting  plants  were  obtained  in  2 1  days.  Inoculations  of  the 
soil  with  this  strain  before  planting  the  seed  resulted  in  59  per  cent  of 
wilt.  Inoculation  of  S7-  to  90-day-old  okra  plants  with  this  strain  through 
the  hypocotyl  gave  from  75  to  100  per  cent  of  wilting  plants.  The 
reisolated  strains  of  V.  alho-atrum  171 7  gave  the  following  results:  Strain 
2943,  40  per  cent;  strain  3075,  66  per  cent;  strain  3076,  16  per  cent; 
strain  3156,  71  per  cent.  V.  alho-atrum,  strain  2821,  isolated  from  a 
wilting  okra  plant  from  Middle  River,  Cal.,  produced  100  per  cent  of 
wilting  plants  through  hypocotyl  inoculations. 

The  introduction  of  V.  alho-atrum  into  wounds  of  the  pedicel  and  the 
stem  of  okra  plants  was  without  apparent  effect,  thus  indicating  that 
there  is  no  danger  of  carrying  the  wilt  disease  from  plant  to  plant  by  the 
cutting  knife. 

The  inoculation  of  okra  plants  with  V  alho-atrum  from  hosts  other 
than  okra  gave  the  following  results:  Strain  2985,  isolated  from  snap- 
dragon, produced  from  none  to  45  per  cent  of  wilt  through  hypocotyl 
inoculation;  strain  2784  from  potato  gave  73  per  cent  through  hypocotyl 
wounds  and  55  per  cent  through  soil  inoculation  previous  to  planting  the 
seed;  strain  1685,  isolated  from  eggplant,  gave  43  per  cent  of  wilt  through 
the  inoculation  of  60-day-old  plants  at  the  hypocotyl. 

The  results  of  inoculating  okra  plants  with  F.  vasinfectum  were  for 
some  time  disappointing.  However,  after  an  insight  was  gained  into 
the  conditions  necessary  to  induce  the  Fusarium-wilt  of  okra,  better 
results  were  obtained.  With  the'  species  of  Fusarium  from  okra  in  two 
trials,  50  and  55  per  cent  of  wilting  plants  were  secured  by  inoculation 
through  the  hypocotyl  (PL  23).  But  with  the  species  of  Fusarium  from 
cotton  only  i  per  cent  was  obtained.  However,  the  cotton  inoculated 
with  the  species  of  Fusarium  (3592)  from  okra  and  the  strain  (1855)  from 
cotton  gave  a  high  percentage  of  wilting  plants  in  a  parallel  test  of  the 
two  strains  (PI.  24-26). 

INOCULATION   OF  COTTON 

Fusarium  vasinfectum  strain  1855  from  cotton,  when  introduced  into 
wounds  at  the  hypocotyl  of  21-day-old  plants,  gave  50  per  cent  of  wilt 
in  17  days  and  a  total  of  89  per  cent  in  5  weeks  (PI'.  25).  Similarly,  in  a 
parallel  test,  F.  vasinfectum  strain  3592  from  okra  gave  42  per  cent  of 
wilt  in  17  days  and  a  total  of  87  per  cent  in  5  weeks  (PI.  26).  In  contrast 
to  these  results  are  those  secured  with  these  strains  on  okra  in  the  parallel 


Mar.  4. 1918  Wilt  Diseases  of  Okra  543 

test  where  but  i  and  6  per  cent  of  wilt  was  secured.  Verticillium  albo- 
airum  stram  3156  inoculated  through  wounds  of  the  hypocotyl  gave  80 
per  cent  of  wilt  in  14  days. 

The  Verticillium- wilt  of  cotton,  as  observed  in  these  tests  can  not  be 
distinguished  from  the  very  destructive  Fusarium-wilt  of  this  crop  imless 
a  critical  examination  be  made.  It  is  possible  that  this  wilt  is  present 
in  the  cotton  fields  of  the  South  and  has  been  overlooked,  owing  to  the 
similarity  of  the  two  diseases. 

INOCULATION   OF   EGGPLANT 

Eggplant  was  found  to  be  susceptable  to  the  strains  of  V.  alho-atrum 
isolated  from  okra,  eggplant  (PI.  20),  and  snapdragon.  Strain  171 7 
from  okra  gave  85  per  cent  of  wilt  in  15  days;  strain  1685  froni  eggplant 
gave  100  per  cent  in  the  same  time,  and  strain  2985  from  snapdragon 
80  per  cent  correspondingly.  The  inoculations  were  made  through 
wounds  of  the  hypocotyl  on  45-day-old  plants. 

INOCULATION   OF   BRABHAM   COWPEAS 

Twenty-four-day-old  seedlings  of  Brabham  cowpeas  were  inoculated 
through  the  hypocotyl  with  F.  vasinjectum  strain  3203  from  okra. 
Although  held  under  observation  for  60  days,  no  cases  of  wilt  were 
detected. 

CONCLUSIONS 

There  are  two  similar  wilt  diseases  of  okra,  caused,  respectively,  by 
Fusarium  vasinfectum  and  Verticillium  alho-atrum.  These  diseases 
can  be  differentiated  only  by  isolating  the  causal  fungi.  The  Fusarium- 
wilt  in  general  is  more  serious  in  the  southern  range  of  okra  growing, 
while  the  Verticillium-wilt  is  more  serious  in  the  northern  range  of  this 
crop.  However,  the  former  has  been  reported  in  Connecticut,  and  the 
latter  occurs  in  South  Carolina  and  Alabama. 

The  wilt  diseases  of  the  Irish  potato  offer  a  parallel  case  of  two  organ- 
isms producing  the  same  disease  symptoms.  Here,  the  Fusarium-wilt 
is  induced  by  F.  oxysporum  and  the  Verticillium-wilt  by  V.  alho-atrum, 
as  in  okra.  Similarly,  it  is  demonstrated  in  this  paper  for  the  first  time 
that  cotton  may  have  both  wilt  diseases.  No  doubt  there  are  several 
other  economic  plants,  which  act  as  host  to  vascular  parasites  of  the 
genera  Fusarium  and  Verticillium — for  example,  China  aster,  ginseng, 
eggplant,  brambles. 

V.  alho-atrum  was  isolated  from  the  discolored  vascular  system  of 
wilting  plants  of  okra,  eggplant,  potato,  and  species  of  Abutilon  and 
Xanthium,  and  was  demonstrated  in  these  studies  to  be  the  cause  of  a 
wilt  disease  at  least  of  okra,  eggplant,  and  cotton.  It  was  identified 
from  a  wilt  disease  of  snapdragon  after  its  causal  relation  to  this  disease 
had  been  established  by  Miss  Nellie  A.  Brown,  of  the  Bureau  of  Plant 


544  Journal  of  Agricultural  Research  voi.  xii,  N0.9 


Industry.  Okra  is  susceptible  to  inoculation  with  V.  albo-arttim  from 
okra,  snapdragon,  eggplant,  and  potato;  and  to  F.  vasinjectum  from 
okra.  Eggplant  is  susceptible  to  V.  alho-atrum  from  eggplant,  okra, 
and  snapdragon.  Cotton  is  susceptible  to  the  strains  of  this  fungus 
from  okra  and  to  F.  vasinjectum  from  okra  and  cotton. 

Thus,  it  is  apparent  that  the  species  of  Fusarium  causing  the  wilt 
disease  of  okra  is  identical  with  F.  vasinjectum.  Likewise  it  is  evident 
that  V.  alho-atrum  is  a  serious  vascular  parasite  of  a  number  of  different 
economic  plants.  In  view  of  the  fact  that  the  genus  Acrostalagmus 
should  be  combined  with  the  earlier  genus  Verticillium,  it  seems  probable 
that  V .  alho-atrum  is  the  fungus  described  by  Van  Hook  under  the  name 
''Acrostalagmus  alhus,"  the  cause  of  ginseng-wilt;  by  Lawrence  as  A. 
caulophagus,  the  cause  of  the  bluestem  disease  of  black  raspberry;  and 
by  Gueguen  as  A.  vilmorinii,  the  cause  of  a  wilt  disease  of  China  aster 
and  associated  with  a  disease  of  cacao  fruits. 

V .  alho-atrum  and  F.  vasinjectum  are  readily  cultivated  artificially, 
indicating  that  both  are  capable  of  persisting  saprophytically  in  the  soil 
for  an  indefinite  period  in  the  absence  of  the  preferred  host.  There 
was  no  indication  that  either  fimgus  is  carried  from  field  to  field  or  from 
year  to  year  by  the  seed  or  from  plant  to  plant  by  the  cutting  knives. . 

As  a  control  measure  it  may  be  suggested  that  seed  be  selected  only 
from  healthy  plants.  If  extreme  precaution  is  to  be  taken,  the  seed 
may  be  disinfected  in  a  formalin  solution  (i  to  240)  for  two  hours. 

Since  okra,  eggplant,  potato,  cotton,  snapdragon,  and  the  weeds 
Xanthium  spp.  and  Abuiilon  spp.  are  all  susceptible  to  the  Verticillium- 
wilt,  as  well  as  ginseng,  China  aster,  and  black  raspberry,  as  seems 
probable,  these  plants  should  be  taken  into  consideration  in  planning 
a  rotation  to  eliminate  wilt  diseases.  Similarly,  okra  and  cotton  are 
hosts  of  F.  vasinjectum  and  should  not  follow  each  other  in  rotation  if 
best  results  are  expected. 

LITERATURE  CITED 
(i)  Aderhold,  Rudolf. 

1907.      UBER  EINE  THROMBOSE  DES  JOHANNIS-  UND   STACHELBEERE.      In  Mitt. 

K.  Biol.  Anst.  Land- u.  Forstw.,  Heft  4,  p.  26-27. 

(2)  Appei<,  Otto. 

1909.  EiNiGES  UBER  DIE  BI.ATTR0LLKRANKHE1T  DER  KARTOFPEL.    /n  Jahresbcr. 

Ver.  Angew.  Bot.,  Jahrg.  6,  1908,  p.  259-265. 

(3)  Atkinson,  G.  F. 

1892.  SOME  DISEASES  OP  COTTON.    Ala.  Agf.  Exp.  Sta.  Bui.  41,  65  p.,  25  fig. 

(4)  Brown,  Nelue  A. 

1914.  A  SNAPDRAGON  WILT  DUE  TO  VERTiciLUUM.    In  Phytopathology,  V.  4, 
no.  3,  p.  217. 

(5)  Butler,  E.  J. 

1910.  THE  WILT  DISEASE  OP  PIGEON-PEA  AND  THE  PARASITISM  OP  NEOCOSMOS- 

PORA  VASINFECTA  SMITH.      Mem.  Dept.  Agr.  India,  Bot.   Ser.,  v.  2, 
no.  9,  64  p.,  6  pi.  (part  col.).    Bibliography,  p.  63-64. 


Mar.  4.  i9ig  IVilt  Diseases  of  Okra  545 

(6)  Carpenter,  C.  W. 

1914.  THE  vERTiciLLiuM  w^LT  PROBLEM.    (Abstract.)    In  Phytopathology,  v.  4, 

no.  6,  p.  393. 

(7)  

1915.  SOME  POTATO  TUBER-ROTS  CAUSED  BY  SPECIES  OP  FUSARIUM.      In  JoUT. 

Agr.  Research,  v.  5,  no.  5,  p.  183-209.    Literature  cited,  p.  208-209. 

(8)  CUNTON,   G.    P. 

1906.  NOTES  OX  FUNGUS  DISEASES,  ETC.,  FOR  1905.  In  Cotui.  AgT.  Exp.  Sta., 
29th  Ann.  Rpt.,  [1904/05],  p.  263-303,  fig.  8-9.    Literatiire,  p.  301-303. 

(9)  CORDA,  A.  C.  I. 

1838.  icoNES  FUNGORUM  HucusQUE  coGNiTORUM.    t.  2.    Prague. 

(10)  Gu6guen,  Femand. 

1906.  acrostalagmus  vilmorinii  n.  sp.,  muceidinee  produisant  une  mala- 
DiE  a  scljSrotes  du  collet  des  reines-marguerites.  In  Bui.  Soc. 
Mycol.  France,  t.  22,  fasc.  4,  p.  254-26^,  5  fig.,  pi.  16. 

(11)  

1910.  SUR   UNE   MALADIE   DU  FRUIT   DE   CACAOYER  PRODUITE  PAR  UNE    MUC^- 

din6e  ET  SUR  LE  mecanisme  DE  l'infection    In  Compt.  Rend.  Soc. 
Biol.  [Paris],  t.  68,  no.  5,  p.  221-222. 

(12)  Higgins,  B.  B. 

1911.  IS  neocosmospora  vasinfecta  (atk.)  smith,  the  perethecial  stage 

OF  THE  FUSARIUM  WHICH  CAUSES  COWPEA    WILT?     In  N.  C.  AgT.  Exp.  Sta., 

32d  Ann.  Rpt.  [i9oS]/o9,  p.  100-116,  16  fig. 

(13)  Hoffmann,  Hermann. 

1854.    SPERMATIEN   BEI  £INEM  FADENPILZE.      In   Bot.    Ztg.,  Jahrg.    12,   No.    l6, 

p.  249-254,  no.  16,  p.  265-269. 

(14)  Klebahn,  H. 

1913.  beitrage  zur  kenntnis  der  fungi  imperpscti.    In  Mycol.  Centbl., 

Bd.  3,  Heft  2,  p.  49-66,  15  fig. 

(15)  Lawrence,  W.  H. 

1912.  BLUESTEM  of  THE  BLACK  R.^SPBERRY.     /w  Wash.  Agr.  Exp.  Sta.  Bui. 

108,  30  p.,  fig.  21-31,  I  pi. 

(16)  Orton,  W.  a. 

1900.   THE  wilt  DISEASE  OF  COTTON  AND  ITS  CONTROL.     U.  S.  Dept.  Agr.  Div. 

Veg.  Physiol,  and  Pathol.  Bui.  27,  16  p.,  4  pi. 

(17)  

1914.  POTATO   WILT,    LEAF-ROLL,    AND  RELATED   DISEASES.         U.  S.    Dept.    AgT. 

t  Bul.  64,  48  p.,  16  pi.    Bibliography,  p.  44-48. 

(18)  Pethybridge,  G.  H. 

1910.  potato  diseases  in  IRELAND.    In  Dept.  Agr.  and  Tech.  Instr.  Ireland, 

Jour.,  V.  10,  no.  2,  p.  241-256,  8  fig. 

(19)  

1911.  investigations  on  potato  DISEASES,     (second  report).     In  Dept. 

Agr.  and  Tech.  Instr.  Ireland,  Jour.,  v.  11,  no.  3,  p.  417-449,  14  fig. 

(20)  Rankin,  W.  H. 

1910.  root  rots  of  ginseng.  In  Special  Crops,  v.  9,  no.  94,  p.  349-360,  14 
fig.   Bibliography,  p.  359-360. 

(21)  

1914.  THROMBOTIC  DISEASE  OP  M.\PLE.  (Abstract.)  In  Phytopathology,  v.  4, 
no.  6,  p.  395-396. 

(22)  Reinke,  J.,  and  Berthold,  G. 

1879.  DIE  zersetzung  der  kartoffel  durch  pilze.     Untersuch.  Bot.  Lab. 

Univ.  Gottingen,  Heft  1,  100  p.,  9  pi. 
38324°— 18 2 


546  Journal  of  Agricultural  Research  voi.  xii,  no.  9 

(23)  Smith,  E.  F. 

1899.    WILT    DISEASE    OF    COTTON,    WATERMELON,     AND    COWPEA    (nEOCOSMOS- 

PORA  NOV.  GEN.)-    U-  S.  Dept.  Agr.  Div.  Veg.  Physiol,  and  Pathol. 
Bui.  17,  72  p.,  10  pi.  (part  col.).    Previous  literature,  p.  50. 

(24)  SpiEckermann,  a. 

191 1.  BEITRAGE   ZUR   KENNTNIS    DER  BAKTEIUENRING  UND  BLATTROLL-KRANK- 

HEiTEN  DER  kartoffelpflanze.     /«  Jahrcsbet.     Angew.  Bot.,  Jahrg 
8,  1910,  p.  1-19,  173-177- 

(25)  Stevens,  F.  L.,  and  Wilson,  G.  W. 

1912.  okra  WILT  (fusariose),  fusarium  vasinfectum,  and  clover  rhizoc- 

TONiosE.     In  N.  C.  Agr.  Exp.  Sta.,  34th  Ann.  Rpt.,  [igioj/n,  p. 
70-73,  fig.  15-18. 

(26)  Stormer,  K. 

1910.  OBSTBAUMSTERBEN  UND  kartoffelblattrollkrankheit.    In  Jahres- 

ber.  Ver.  Angew,  Bot.,  Jahrg.  7,  1909,  p.  1 19-170,  15  fig.,  pi.  5. 

(27)  Van  Hook,  J.  M. 

1904.  SOME   DISEASES   OF   GINSENG.      N.  Y.  Cornell  Agr.  Exp.  Sta.  Bui.  219, 
p.  165-186,  fig.  18-42. 

(28)  WhETZel,  H.  W.,  and  Rosenbaum,  J. 

1912.  THE  DISEASES  OF  GINSENG  AND  THEIR  CONTROL.      U.  S.  Dept.  Agr.  Bur. 

Plant  Indus.  Bui.  250,  44  p.,  5  fig.,  12  pi. 

(29)  Wilson,  G.  W. 

1913.  FUSARIUM  OR  vERTiciLLiuM  ON  OKRA  IN  NORTH  CAROLINA?    In  Phyto- 

pathology, V.  3,  no.  3,  p.  183-185. 

(30)  WOLLENWEBER,  H.  W. 

1911.  UNTERSUCHUNGEN  XJBER  DIE  NATXJRLICHE  VERBRElTUNG  DER  FUSARIEN 

AN  DER  KARTOFFEL.     In  Mitt.  K.  Biol.  Anst.  Land.  u.  Forstw.,  Heft 
II,  p.  20-23. 
(31) 

(32) 

(33) 


1913.   PILZPARASITARE   WELKEKRANKHEITEN   DER   KtrLTURPFLANZEN.      In  Ber. 

Deut.  Bot.  Gesell.,  Bd.  31,  Heft  i,  p.  17-34- 


1913.  STUDIES  ON  THE  FUSARIUM  PROBLEM.     In  Ph3rtopathology,  V.  3,  no.  I, 
p.  24-50,  X  fig.,  pi.  5.     Literatture,  p.  46-48. 


I914.    IDENTIFICATION    OF    SPECIES    OF    FUSARIUM    OCCURRING    ON    THE    SWEET 

POTATO,  IPOMOEA  BATATAS.     In  Jour.  Agr.  Research,  v.  2,  no.  4,  p. 
251-286,  pi.  12-16  (part  col.).     Literature  cited,  p.  284-285. 

(34)   and  SCHLUMBERGER,  Otto. 

191 1.    INFEKTIONSVERSUCHE  MIT  KARTOFrELBEWOHNENDEN  PILZEN.      In  Mitt. 

K.  Biol.  Anst.  Land.  u.  Forstw.,  Keft  11,  p.  15-17. 


PLATE  A 

Fusarium  vasinfecHvm  on  vegetable  media: 

1-3. — Growth  on  steamed  potato.     Both  potato  cultures  show  pionnotes. 

2 ,  4. — Growth  on  rice . 

Cultures  i  and  2  were  grown  in  a  strong  north  light;  3  and  4  in  a  subdued  light. 


Wilt  Diseases  of  Okra 


y 


Journal  of  Agricultural  Research 


Plate  A 


Vol.  XII.  No.  9 


PLATE  17 
A-H. — Verticillium  albo-atrum: 

A,  Simple  conidiophores  and  conidia.     X  1,000. 

B,  Same  showing,  respectively,  the  collection  of  the  conidia  on  the  sterigma  in 
irregular  aggregations  in  dry  air,  and  in  water  drops  in  humid  air.     X  1,000. 

C,  Verticillate  conidiophores  bearing  one  and  three  whorls,  or  virtels,  of  branches, 
respectively.     X500. 

D,  Verticillate  conidiophore  having  conidial  heads,  from  humid  environment — 
that  is,  moistiu-e  drops  in  which  the  conidia  float  as  in  figure  B.  This  is  the  so-called 
Acrostalagmus  type  of  conidial  head.     X500. 

E,  Mycelium  of  V.  albo-atrum  in  the  vascular  ducts  of  an  okra  plant  inoculated 
with  this  fungus.     X250. 

F,  H,  Germinating  conidia:  F,  Xsoo;  H,    Xi,ooo. 

G,  Swollen,  sclerotia-like  mycelium.    X500. 

I-M. — Fusarium  vasinfectwm: 

I,  Terminal,  intercalary  and  conidial  chlamydospores.     i,  Germinating  terminal 
chlamydospore.     2,  Free  mature  chlamydospore.  X  1,000. 
K,  Germinating  macroconidium.    X  1,000. 

1-,,  F.  vasinfectumirova.okxa-WAt.     Twelve  macroconidia.  X  1,000. 
M,  F.  vasinfectum  from  cotton-wilt.     Four  macroconidia.  X  1,000. 


Wilt  Diseases  of  OPcra 


Plate  17 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  9 


Wilt  Diseases  of  Okra 


Plate  18 


Journal  of  Agricultural  Research 


Vol.  XII,  No. 9 


PLATE    i8 

Longitudinal  section  of  an  okra  plant  naturally  infected  with  Verticillium  albo- 
atrum,  showing  the  typical  appearance.  The  vascular  elements  are  discolored  from 
the  roots  to  the  pedicels  and  petioles.     About  natural  size. 


PLATE   19 

Verticillium  albo-atrum: 

Two-weeks-old  colony  on  potato  agar,  showing  the  concentric  rings  of  black  scle- 
rotial  bodies.     X4. 


Wilt  Diseases  of  Okra 


Plate  19 


Journal  of  Agricultural  Rcacarcii 


Vol.  XI  I,  No.  9 


Wilt  Diseases  of  Okra 


Plate  20 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  9 


PLATE   20 
Solarium  melongena,  showing  effect  of  wilt: 

A. — Control  plant  of  the  same  age  as  the  wilted  plant  (B).  Photographed  at  the 
same  time  as  figure  B.     X2/5. 

B. — Wilted  plant  photographed  two  months  after  inoculation  at  the  hypocotyl 
with  Verticillium  albo-airum  isolated  from  wilted  eggplant.  The  organism  was 
recovered  from  the  stem  of  the  small  plant  10  cm.  above  the  point  of  inoculation. 
X2/5. 


PLATE  21 

Abelmoschus  escuknttis,  showing  effect  of  wilt: 

A. — Control  plant.     Photographed  at  the  same  time  as  figure  B.     X  (about)  1/2. 
B. — ^Wilted  plant  photographed  two  weeks  after  inoculation  at  the  hypocotyl  with 
a  pure  culture  of  Verticillium  albo-atrum.     X  (about)  1/2. 


Wilt  Diseases  of  OPcra 


Plate  21 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  9 


Wilt  Diseases  of  Okra 


Plate  22 


Journal  of  Agricultural  Research 


Vol. XII,  No.9 


PLATE   22 

Abelmoschus  esculentus,  showing  effect  of  wilt: 

A. — Wilted  plant  inoculated  with  Verticilium  albo-atrum.     X  (about)  1/3. 
B. — Control  plant  of  the  same  age  as  wilted  plant.     X  (about)  1/3. 
Both  plants  were  photographed  two  months  after  the  wilted  plant  had  been  inocu- 
lated. 


PLATE   23 

Abelmosckus  esculentus,  showing  the  effect  of  wilt  as  a  result  of  inoculation  with 
Fusarium  vasinfecium  isolated  from  okra-wilt.  The  inoculation  was  made  at  the 
hypocotyl  and  the  organism  was  recovered  from  the  vascular  tissues  of  the  petioles 
two  weeks  afterwards — that  is,  immediately  after  the  photograph  was  taken.     X2/3. 


Wilt  Diseases  of  Okra 


Plate  23 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  9 


Wilt  Diseases  of  Ol<ra 


Plate  24 


Journal  of  Agricultural  Researcli 


Vol.  XII,  No. 9 


PLATE   24 

Gossypium  lierbaceum  (Columbia  variety): 

Control  plants  35  days  old.     Photographed  15  days  after  having  been  wounded  at 
the  hypocotyl.     Natural  size. 


PLATE  25 

Gossypium  herhaceum  (Columbia  variety),  showing  effect  of  wilt: 

Wilting  plants  photographed  15  days  after  inoculatiou  at  the  hypocotyl  with  Fusa- 
rium  vasinfecium,  strain  1855,  isolated  from  wilting  cotton  plants.  The  plants  are 
the  same  age  as  the  control  plants  in  Plate  24.    Natural  size. 


Wilt  Diseases  of  Okra 


Plate  25 


Journal  of  Aj/ricultural  Rusuarch 


Vol.  XII,  No.  9 


Wilt  Diseases  of  Ol<ra 


Plate  26 


Journal  of  Agricultural  Research 


Vol. XII,  No. 9 


PLATE   26 

Gossypium  kerbaceum  (Columbia  variety),  showing  effect  of  wilt: 

Wilting  plants  photographed  15  days  after  inoculation  at  the  hypocotyl  with  Fusa- 
Tium  vasinfectum,  strain  3592,  isolated  from  wilting  okra.  The  plants  are  the  same 
age  as  those  in  Plates  24  and  25.     Natiural  size. 


PLATE  27 

Abelmoschus  esculentus,  showing  the  characteristic  symptoms  of  the  wilt  produced 
by  Verticillium  albo-atrum.     Photographed  in  a  field  in  New  Jersey. 


Wilt  Diseases  of  Okra 


Plate  27 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  9 


WINTER  CYCLE  OF  EGG  PRODUCTION  IN  THE  RHODE 
ISLAND  RED  BREED  OF  THE  DOMESTIC  FOWL 

By  H.  D.  GooDALE, 
Massachusetts  Agricultural  Experiment  Station 

INTRODUCTION 

The  winter  cycle  of  egg  production  is  one  of  the  internal  factors 
concerned  in  determining  total  production.  It  was  first  recognized  by 
Pearl  and  Surface  in  Barred  Plymouth  Rocks.  They  state  (7,  p.  gg- 
100) : ^ 

It  will  undeniably  be  advantageous,  in  studying  certain  phases  of  the  problem  of 
egg  production,  to  endeavor  to  use  a  time  unit  which  conforms  to  the  natural 
periodicity  displayed  by  hens  [italics  are  mine — H.  D.  G.].     .     .     . 

The  plan  followed  at  the  present  time  in  the  investigations  in  progress  at  the 
Maine  Station  breaks  the  laying  year  up  into  four  parts.  The  first  of  these  includes 
the  months  of  November,  December,  January,  and  February.  Broadly  speaking, 
this  is  the  period  of  winter  laying  and  is  so  designated.     .     .     . 

The  justification  for  the  conclusion  that  this  division  of  the  year  is  in  general  a 
natiu-al  one  and  corresponds  to  a  real  cyclical  periodicity  in  egg  production  is  in  con- 
siderable measure  to  be  found  in  the  facts  regarding  mean  monthly  egg  production 
and  variation  in  this  character  set  forth  in  this  and  the  preceding  section  of  the  paper. 
The  winter-laying  period  is  a  period  characterized  by  rapid  increase  in  mean  pro- 
duction associated  with  a  relatively  equally  rapid  decrease  in  variability,  both 
absolute  and  relative.  In  this  period  a  large  part  of  the  flock  falls  in  the  A  compo- 
nent of  the  monthly  distributions  (see  p.  142).  This  laying  period  is,  strictly  speak- 
ing, not  a  part  of  the  natural  or  normal  reproductive  cycle  of  the  hen.  Egg  laying 
in  this  part  of  the  year  is  something  which  diu^ing  domestication  has  been  added  on, 
as  it  were,  to  the  natural  reproductive  activity  of  the  wild  Callus.  It  is  a  result  of 
"forcing"  or  special  stimulation.  From  the  evolution  standpoint,  egg  production 
in  these  months  is  a  comparatively  recent  acquisition.  Such  being  the  case,  the 
greater  variability  observed  in  winter  laying  is  only  what  would  be  expected. 

The  limits  of  this  artificial  winter  cycle  of  egg  production  are  fairly  well  defined. 
It  begins  with  the  beginning  of  the  laying  year.  Its  other  limit  is  marked  by  the 
slacking  up  in  egg  production,  which  occurs  in  February  (see  Fig.  i).  This  slacken- 
ing up  in  February,  which  appears  to  be  a  characteristic  of  egg  production,  generally, 
is  to  be  explained,  we  believe,  chiefly  if  not  entirely  as  the  result  of  the  ending  of 
the  winter  cycle  by  the  majority  of  birds  which  have  laid  during  the  early  winter. 
Such  birds  rest  for  a  period  at  about  this  time  before  beginning  the  spring  laying 
cycle.  Of  cotu-se  it  must  be  understood  that  these  statements  are  made  only  with 
reference  to  what  might  be  called  the  general  or  average  course  of  events.  Particular 
birds  may  form  exceptions  in  their  laying.  Many  birds,  of  course,  have  no  proper 
winter  cycle  of  laying  at  all.  They  begin  to  lay  for  the  first  time  in  January  or  Feb- 
ruary, and  keep  on  laying  without  any  large  break  straight  through  the  spring  cycle. 

'  Reference  is  made  by  number  (italic)  to  "  Literature  cited,"  p.  574. 

Journal  of  Agricultural  Research,  Vol.  XII,  No.  9 

Washington,  D.  C.  Mar.  4,  1918 

mf  Key  No.  Mass. — 3 

38324°— 18 3  (547) 


548  Journal  of  Agricultural  Research  voi.  xii.  no.  9 

In  discussing  the  monthly  egg  distribution  Pearl  and  Surface  (7,  p. 
89-90)  state: 

Considering  the  form  of  the  polygon  somewhat  more  in  detail,  we  note  that  the 
line  starts  from  a  low  point  in  November  and  rises  rather  rapidly  and  in  almost  a 
straight  line  to  January.  The  slope  of  the  line  from  January  to  February  is  down- 
ward. In  other  words,  there  is  an  indentation  in  the  ascending  limb  of  the  egg-pro- 
duction polygon  in  the  month  of  February.  This  is  a  very  characteristic  featiire  of 
the  distribution  of  egg  production,  not  only  observed  with  the  birds  here  under 
discussion  but  also  in  published  records  from  other  sources. 

A  study  of  Table  i  shows  that  this  is  generally  true  for  every  year  covered  by  the 
investigation.  While  the  February  mean  production  is  not  necessarily  lower  than  the 
January,  though  this  is  true  in  many  cases,  there  is  a  perceptible  slowing  of  the  rate 
of  increase  in  egg  production  which  has  obtained  up  to  that  time.  The  most  prob- 
able interpretation  of  this  appears  to  us  to  be  that  the  February  indentation  in  the 
egg-production  curve  represents  a  rest  or  reaction  after  the  winter  laying  and  in 
anticipation  of  the  heavy  March  and  April  production.  It  marks  the  completion  of  a 
laying  cycle  on  the  part  of  those  birds  which  have  been  laying  dturing  the  winter 
months. 

The  mean  egg  production  for  February,  however,  seems  to  rest  on  a 
basis  of  28  days.  If  reduced  to  a  basis  of  31  days,  the  mean  production 
is  12.03  eggs,  a  value,  however,  that  produces  a  pronounced  indentation 
in  the  upward  slope  of  the  polygon.  Pearl  believes  that  this  change  in 
the  slope  of  the  polygon,  disregarding  entirely  any  actual  drop  in  pro- 
duction, is  indicative  of  a  winter  cycle.  However,  a  change  in  the 
slope  of  such  a  polygon  would  occur  if  a  flock  of  birds  began  to  lay  in 
some  given  month,  gradually  increased  in  production  for  a  definite 
period  of  some  length  until  they  reached  a  maixmum  and  maintained 
this  maximum  for  a  period  of  several  months.  Pearl,  moreover,  is 
inclined  to  believe  that  the  change  in  mean  temperature  that  occurs  in 
March  bears  no  causal  relation  to  the  increase  in  egg  production  that 
is  observed  at  this  season.  Now,  Avhile  it  is  clear  that  there  are  other 
factors  than  the  change  in  temperature  that  increases  production  at 
this  season,,  it  is  not  at  all  clear  that  temperature  can  be  entirely  elimin- 
ated as  a  factor.  In  the  imaginary  case  just  mentioned  it  seems  pos- 
sible that  the  maximum  production  under  one  set  of  environmental 
conditions  might  be  different  from  that  observed  under  another  set. 
Hence,  it  seems  entirely  probable  that  the  maximum  production 
possible  in  January  and  February  would  not  be  as  high  as  in  March, 
and  that  the  indentation  noted  in  the  curve  of  production  merely  means 
that  the  maximum  production  possible  for  midwinter  conditions  has 
been  reached. 

In  several  other  papers,  notably  those  of  Pearl  (5,  6),  the  winter  cycle 
is  again  discussed.     In  the  former  paper  he  states  (5,  p.  173-174) : 

(2)  The  upper  limit  of  the  winter  period  at  March  i  is  arbitrary,  and  only  approxi- 
mately coincides  with  the  biological  reality.  Actually  with  most  birds  the  spring 
or  reproductive  cycle  of  production  (cf.  37)  begins  in  the  latter  part  of  February. 
In  handling  the  material  it  has  been  found  necessary  (for  reasons  which  will  be 


Mar.  4. 1918      Winter  Egg  Production  of  Rhode  Island  Reds  549 

obvious  upon  consideration  of  the  matter)  to  take  a  fixed  date  for  the  beginning  of 
the  spring  cycle  of  laying  and  the  ending  of  the  winter  cycle.  The  records  of  the 
station  prior  to  1908  are  tabulated  only  for  months  (the  daily  records  unfortunately 
having  been  destroyed  before  I  took  charge  of  the  work),  and  on  this  account  it  is 
necessary  to  take  the  working  limit  of  the  winter  cycle  at  the  end  of  a  calendar  month. 
Since  March  i  comes  the  nearest  to  the  biological  limit  of  any  date  which  is  also  the 
beginning  of  a  calendar  month  it  has  been  chosen.  The  error  introduced  by  taking 
this  arbitrary  date  for  a  point  which  really  shifts  within  rather  narrow  limits  is,  on 
the  average,  small.  However,  it  must  b.e  recognized  as  a  disturbing  element  in  the 
individual  case.  Thus,  some  birds  which  really  lack  any  genetic  factor  for  winter 
production  will  begin  to  lay  in  the  last  days  of  February,  and  consequently  on  the 
arbitrary  "March  i"  basis  will  actually  be  credited  with  a  small  winter  production. 
This  will  tend  to  make  the  number  of  zero  birds  observed  smaller  than  that  expected 
on  theory. 

in  the  latter  paper,  entitled  "Measurement  of  the  Winter  Cycle  in  the 
Egg  Production  of  Domestic  Fowl,"  a  comparison  is  made  betvs^een  the 
egg  production  of  a  pullet  during  the  first  300  days  of  its  life  v^^ith  its  egg 
production  up  to  March  i  of  the  pullet  year.  He  finds  that  there  is  very 
little  difference  in  the  value  of  these  two  measures  of  production 

The  evidence  for  a  winter  cycle  published  by  Pearl  is  all  mass  evi- 
dence. The  possibility  that  a  flock  might  be  heterogeneous  in  respect 
to  the  winter  cycle  is  not  considered  by  him.  It  is  true  that  he  speaks  of 
particular  birds  forming  exceptions  to  the  rule  in  their  laying,  but  the 
sentence  immediatel)-  following  seems  to  indicate  that  Pearl  has  in  mind 
birds  that  begin  to  lay  late  in  the  winter  (7,  p.  100)  rather  than  birds 
that  lay  throughout  the  winter.  However,  a  few  records  that  form  real 
exceptions  to  the  rule  can  be  found  in  the  records  of  individual  Barred 
Plymouth  Rocks  published  by  Gowell  (j,  4). 

Pearl's  mass  evidence  of  the  existence  of  a  winter  cycle  is  supported 
by  a  study  of  the  records  described  belov/.  While  a  cursory  examina- 
tion of  the  data  on  egg  production  of  Rhode  Island  Reds  at  this  Station 
indicates  that  a  considerable  percentage  of  our  records  are  without  vis- 
ible indications  of  a  winter  cycle  (since  many  birds  that  begin  to  lay  in 
November  continue  to  lay  without  noticeable  slackening  of  production 
straight  through  the  winter  and  spring),  at  the  same  time  there  are 
many  instances  in  which  the  existence  of  a  winter  cycle  is  indisputable. 
Records  of  both  sorts  may  also  be  obsers^ed  in  the  reports  from  various 
egg-laying  contests.  Further,  an  examination  of  Gowell's  records  of  the 
monthly  egg  production  of  individuals  (j,  4)  shows  in  most  instances  a 
marked  decrease  in  the  production  for  February  over  January,  indicat- 
ing that  the  birds  either  stop  laying  entirely  for  a  time  during  one  or  both 
months  or  that  they  slow  down  in  their  daily  rate.  The  former  alterna- 
tive, in  view-  of  Pearl's  statements,  appears  to  represent  the  facts.  A 
few  records  published  by  Miss  Curtiss  in  another  connection  also  show 
the  same  thing. 


550  Journal  of  Agricultural  Research  voi.  xii.no. 9 

Since  inspection  of  our  records  shows  the  possibility  of  the  existence  of 
two  types  oif  birds  with  respect  to  a  winter  cycle — namely  (i)  those 
exhibiting  such  a  cycle,  and  (2)  those  that  lay  but  give  no  evidence  of  a 
cycle — it  becomes  necessary  to  examine  the  matter  in  detail,  and  espe- 
cially to  endeavor  to  discover  some  criterion  by  which  any  individual  may 
be  suitably  classified.  This  has  been  attempted  by  a  study  of  individual 
records  associated  with  a  study  of  the  length  and  seasonal  distribution 
of  pauses  in  production  as  well  as  "a  study  of  the  monthly  rate  of 
production, 

MATERIAL  AND  METHODS 

The  material  studied  consisted  of  the  daily  egg  records  of  three  flocks 
of  Rhode  Island  Red  pullets,  hatched,  respectively,  in  1913,  1915, 
and  1 91 6.  While  Pearl  has  never  published  individual  records  bearing 
on  this  point,  Gowell,  who  initiated  the  v/ork  at  the  Maine  Station,  has 
published  the  individual  monthly  egg  records  of  the  flocks  making 
their  pullet  records  in  1899-1900,  1900-1901,  1901-2  (j,  4).  These  have 
been  used  for  comparison  with  our  records.  Further  comparisons  are 
outside  the  scope  of  this  paper. 

It  is  obvious  that  a  cycle  in  production  should  appear  on  the  record  in 
one  of  two  ways.  Either  thiere  is  a  period  of  continuous  egg  production 
followed  by  a  period  in  which  no  eggs  whatsoever  are  laid,  or  else  the 
period  of  production  is  followed  by  a  period  in  which  eggs  are  produced 
at  a  less  rapid  rate  than  previously.  It  is  well  known  to  poultrymen 
that  hens  often  lay  well  for  a  while  and  then  enter  on  a  resting  period 
of  variable  length.  The  egg-producing  period  may  be  designated  as  a 
"litter."  This  is,  of  course,  one  form  of  cycle,  but  is  to  be  distinguished 
in  most  cases  from  the  winter  cycle  through  the  fact  that  the  period  of 
egg  production  is  in  the  latter  relatively  long  and  may  be  composed  of 
more  than  one  litter.  As  a  rule  the  only  difficulty  encountered  is  the 
case  where  a  single  litter  extends  over  the  entire  winter.  In  this  in- 
stance one  can  not  tell  whether  one  has  a  winter  cycle,  or  a  very  long 
litter,  or  whether  the  two  coincide. 

Since  any  cycle  consists  of  a  productive  phase  and  a  nonproductive 
phase,  we  have  put  the  original  records  into  tabular  form  (Table 
I),  to  which  the  reader's  attention  is  directed  in  lieu  of  a  detailed 
account  of  individual  histories  and  variations.  In  studjdng  this  table 
we  have  come  to  pay  particular  attention  to  the  pauses  in  production, 
since  these  serve  as  our  visible  indexes  of  cycles, 

EXPLANATION   OP  TABLE   I 

In  Table  I  the  details  of  the  records  of  several  families  are  shown.  Practically 
all  the  different  sorts  of  pauses  are  illustrated.  Pauses  of  less  than  three  days  in 
length  have  not  been  included.  The  arrangement  of  the  tables  by  families  brings 
out  the  strong  resemblances  betw^een  the  members  of  the  various  families.  The 
record  of  the  mother  is  printed  in  italics  at  the  head  of  the  list  of  her  daughters. 


Mar.  4.  iQis       Winter  Egg  Production  of  Rhode  Island  Reds  551 

Daughters  that  became  broody  before  March  i  are  given  in  bold-faced  type.  Among 
other  things  listed  in  the  column  headed  "Remarks,"  pauses  occiuring  after  March 
I  and  extending  through  April  30  are  listed  and  are  to  be  read  as  follows:  The  date 
of  the  beginning  of  a  pause  is  given,  followed  by  its  length,  and  in  case  more  pauses 
occur  it  is  indicated  by  the  number  of  eggs  (inclosed  in  parentlieses)  followed  by 
the  length  of  the  succeeding  pause.  Thus,  No.  9277  in  the  family  of  male  3617  by 
female  6003,  a  pause  of  7  days  length  began  April  7,  when  the  bird  laid  one  egg 
followed  by  a  3-day  pause. 

Families  sired  by  male  3617. — The  progeny  of  one  father  and  two  mothers.  None 
of  the  members  of  tliese  families  shows  a  well-defined  winter  c)''cle.  The  February 
egg  production,  in  most  cases  where  production  began  January  i  or  before,  is  greater 
than  for  January.  The  families  are  also  noteworthy  because  of  the  number  of  birds 
without  any  pause  exceeding  two  days  in  length. 

Family  sired  by  male  5240. — In  this  family  several  individuals  have  a  well- 
defined  winter  cycle;  others  lack  such  a  cycle. 

Family  sired  by  male  4786. — This  family  is  characterized  by  the  existence  of  a 
.  single  short  pause  in  each  record. 

Family  sired  by  male  6781. — This  family  is  characterized  liy  numerous  pauses  of 
varying  lengths. 

F'amily  sired  by  male  4723. — This  is  a  heterogeneous  family. 


552 


Journal  of  Agricultural  Research 


\-ol.  xn.  No.  9 


iis 

<i 

5r, 

S 

s«. 

>< 

C 

n 

8 

•2 

Q 
B 

fi 

^ 

CO 

PX)g 


o^^-g 


O  7- 


00     OO      M\00<^^00000  ^,  ^o^ 


ss^2?::rS3 


Q  00       TTf^^^wMCTOO 


O    to  O    M    "-MO 


00       O       OOwcir^-Ot"' 


O    &  T  -^30    ^«0 


<  O    "100    O    O    O    O  •-■  t^  < 


o  o  o  o  "-"J" 


2  *.2S 
^  >-^it  '-■5 


•Si? « 


I  +++ I ++++     + 


++++++7 


-^     go     -a 


<  ii  a 


a  t;  "  o  Ji  >"  2 

d  ?;  41'— J3  I-  S 


53  3^ 


a  u  VI  <u  ;h 


H  Si 


d  aj  -r 


d-o-9 


>  >  (J  d  o  e  H^ 


3  i  3 


oooooooo 


o  o  o  o  O  M-< 


.5  53  Si" 


05 


r^  t>.  O  00    O  ' 


O      O      11   Si   O  01   5  1)  5  5 


u  o 
0:2: 


o  9  o  9  fc  oj  « 


t-t       ■  TO 


S.  "^    !i  >•  ^    >>>,>,  o  >.o  >.  o    5,v    Suo 


O    >.  O    O    >^  O  >> 


O*      <*"0    w    ro  f^  O*  O  •-• 
CO       O^O'^QsO^O^O'C* 


o  c 

CO  OQ 


O   -o  n-  -T  -T  -y-^ 
t'T  o  o  o  -r  rrr- 

00    Ot  i7>  9t  0<  QtQQ 


Mar.  4. 1918       Winter  Egg  Production  of  Rhode  Island  Reds 


553 


o  O  r* 


O.  CO  00         HI  O         i-t    '■n 

I    + 1  +    r  +-f 


5   U   0)  tj   O   «J   u   o   « 


)-«t.fO»o<*50»0  cnO 


(J5o   o 


J    J  >  > 

U         o   O   C 


f-00     00 


W         00  »  90 


«   Qi— » 


Ot-.f»5CtOP<       ^01 


1  o  'T  M  *     e>it- 


°^  TTOOMOfO       00k 


^  vO    t^  CI  00  -O    O 


to  M  w     Ti<eij 


I  I    I    II  +  +  +      +  + 


Tf  «o  r-  w      MO 


"•" 

■< 
7 

^ 

.1 

1 

a  a 
sj  id 


I  CO  ^     tata 


t  CO  ^     to«o 


c^  inno  OlO 


»^ 

»^ 

■o 

»^   J 

"  I-'  i^  C  0  >.  >, 

M  P« 

?-a. 

SaaaT)  a  3 

0,al 

X 

■<<<  ;^:::S 

<» 

^ 

2.S  S^"2  r 

S??{ 

r-00  00  00  a> » 

>^ 

30  e» 

554 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  9 


^     W 


>» 
<» 

S 

tJ 

la 

> 

^ 

a 
a 

K 

tj 

Ki 

<j 

r/l 

a 

-w 

> 

o 

,J 

-<A, 

g 

Oi 

< 

C^ 

U< 

<» 

)>. 

S 

?i 

.2^ 

'T^ 


< 


1 

Cl"    "^ 

o 

o 

^ 

( 

"'Mi,                  to  4> 

M 

>^    "■     .           ^-^ 

Jo 

■ts 

U 

^o<a    Om 

d 

o 

1) 
0 

s 

O 

oH  2^—        o 

or 

fA 

1 

fr-    rt 

o 

a 

O 

O                O 

O^ 

O 

is'^-s 

r~               r,                        1- 

PO                             Oi 

00 

»H               r*                        w 

^ill 

o 

3 

u   t>,  >-   -  « 

T*;              ?)                        t^ 

1/1                           OG 

r>. 

00 

t« 

O               O                        t- 

lO                                   Tf 

00 

vO 

n>- 

M   cj 

w 

B  . 

■>» 

cm                    lo 

c<                            00 

t^ 

o 

4)  J3 

0 

Febru- 
ary pro- 
duction 
greater 
or  less 
than 
Janu- 
ary. 

+ 

o                  ■«■ 

+      + 

1 

+ 

Aver- 
age 
numbe 
of  eggs 

be- 
tween 
pauses 

>^ 

^       " 

^0 

V, 

"j- 

Aver- 
age 
length 

of 
pauses. 

v;                                             ^ 

Q 

Num- 
ber 

of  eggs 

before 
the 
first 

pause. 

M                                    H 

t^ 

p« 

•*                                            w 

" 

,.^_— —            — 

. 



1  •«    <U    rl   m 

N  >-■  M 

M 

:  "  "  Si 

r-MH^ 

JHHM 

N  M  >ovo  a 

K     •  vO    H  00 

VO    to 

Nu 
ber 
eggs 
twe 
pau 

tjj: 

a  a 

«s 

N 

■^ 

"S* 

"^ 

^2a 

■    2 

o  oc 

fo 

Xi 

n  3 

c>S 

:  u 

BX 

«  >. 

:§ 

ttj      -    Jo     *    ro  OJ    "    M 

o 

^•^Q  jtiH.:?^  ?^„ 

^«f-^fe 

o  6   . 

^•osi 

O  O  1    A 

.   ??l « r 1 1  i^i< VT 

.^ifM/^S'P^M     -     ^1, 

/-S-^/- 

.vir 

u-tt  3 

>->  H  <\)  IM 

M    M    ^ 

Ua 

8  8-p-c 

ISpHtJjOOQQ^tntH 

•g^lSSti^a'Sifl 

s  °  ^  ^'  s 

si  g  d 

0 

fa^l^SQOO^CQ  " 

.2,:z;Qp£ 

QQ^^ 

^t^O^O   •<t'tt-i   O   -'tT 

1  to  to  M 

S      •"       - 

r^       w  tH       DO  w  M 

« 

a 

i-t"o  a 

Q 

. 

, 

_. 

. 

Total 
length 
of  time 

spent 

in 
pauses. 

M                    ^                               M 

00                                H 

o 

t- 

1        " 

00                  N                             t 

t^                     vt 

S  o^ 

^j- 

N                  -If                            lO 

■*                      -O 

■* 

to 

3S3 

^•^a 

i-«5  " 

M                                     M 

^ 

3fe  S)S 

•n 

to                   -a- 

to 

to 

:5^!fs 

fn 

O                                             PO 

t-)                        0 

V) 

o 

O   M 

•-' 

<^ 

05 

Q 

^               O 

o 

o 

(U 

0 

N 
H 

m             «                      Ov 

^               : 

o 

rt-S 

1-^ 

?«           rt                 S 

C                 o 

o 

Q-si 

a             -a 

•0 

0. 

j: 

■^ 

^3            S 

<              : 

< 

"0    . 

^ 

0              ^                     0 

>o                   o 

o 

s 

P  O 

n^ 

u% 

00 

» 

' 

« 

00 

1 

Mar.4.  i9i8       Winter  Egg  Production  of  Rhode  Island  Reds  555 


^  < 


'O  O  ■*  O         «         Ov  O 


-O  OO^OOvO  10  t^  O^ 


00       f^      11         o         00 


00   O   <^       'O         M  0\  o 


+         I 


1        + 


+11     +     ++ 


'"fc!^ 


i  y  a>  (U  I 


j   >  >  >  ^   rt^  o 

'  o  o  o  9  9ti  o' 


H     f^f^,  \0     f^lt^f'lfi" 


jJ  i  >      o      > 
o  rt  o      a<      o 


_0  C  >; 

•O    Q  ft 


OC  00  00        00 


T3  to 


556  Journal  of  Agricultural  Research  voi.  xii,  no. 9 

As  inspection  of  these  records  shows  that  there  are  pronounced  family 
resemblances  in  the  number  and  length  of  pauses  among  the  members  of 
each  of  several  families,  and  since  environmental  influences  have  been 
excluded,  as  far  as  possible,  it  is  evident  that  our  material  is  not  homoge- 
neous from  the  biometrical  standpoint.  Consequently  most  of  the  usual 
biometrical  constants  have  not  been  calculated,  since  in  this  case  such 
constants  tend  to  conceal  the  differences  we  are  endeavoring  to  find. 

MEANS  BY  WHICH  A  WINTER  CYCLE  MAY  BE  RECOGNIZED 

A. — PAUSES 

LENGTH   AND  SEASONAL   DISTRIBUTION 

Two  questions  arise  in  regard  to  the  length  of  the  pauses:  First,  is 
there  any  significant  difference  in  the  length  of  pauses?  Second,  is  there 
any  difference  in  the  distribution  of  pauses  of  varying  length  through  the 
winter?  Table  II  shows  the  percentage  distribution  of  pauses  beginning 
at  two  days  in  length,  grouped  according  to  the  month  of  origin.  The 
combined  values  for  the  three  years  are  shown  in  the  last  column.  It 
is  clear  that  the  pauses  fall  into  two  classes,  short  and  long,  the  dividing 
line  between  the  two  groups  falling  at  about  10  days.  Further,  it  is 
evident  from  the  table  that  the  long  pauses  originate  mainly  in  December 
and  January,  and  to  a  less  extent  in  February,  but  that  the  short  pauses 
are  distributed  in  a  fairly  uniform  manner  throughout  the  winter.^  An 
examination  of  the  pauses  occurring  in  March  and  April  not  due  to 
broodiness  shows  that  while  all  pauses  are  less  numerous  than  in  the 
winter,  and  while  a  few  long  pauses  may  be  found,  most  are  short  ones 
occurring  in  practically  the  same  proportions  observed  in  the  other 
months. 

'  The  increase  in  long  pauses  originating  in  December  and  January  automatically  decreases  the  percent- 
age of  the  short  pauses  so  that  relatively  they  are  somewhat  less  numerous  in  spite  of  the  percentages 
being  about  equal. 


Mar.  4.I9I8       Winter  Egg  Production  of  Rhode  Island  Reds 


557 


Table  II. — Length  of  pauses  distributed  according  to  month  in  which  they  originate,  for 
IQIJ-14,  IQIS-16,  igi6-i7  « 


1913-14 

1915-16 

Length  of  pause. 

Novem-  Decem- 
ber,        ber. 

1 

Janu- 
ary. 

Febru- 
ary. 

March. 

Novem- 
ber. 

Decem- 
ber. 

Janu- 
ary. 

Febru- 
ary. 

March. 

Days. 
2 

Per  d.    Per  d. 

Per  d. 

SI- 54 
19-23 
6. 15 
3- 8s 

Per  d. 
53-33 
20.00 
9.17 

4.17 

P.r  d. 
48.81 
13.09 
13.09 

4.  76 

Per  d. 

71- IS 
9.62 
3-8s 

Per  d. 

57-01 
14.91 
3- SI 
2-63 

Per  ct. 

53-85 
8.24 
4-95 
3-85 

Per  d. 

53-81 

15-71 

8-57 

1.90 

Per  d. 
59-69 

12.25 

7-14 
8.  i6 

7.84 

17-65 

1.96 

2-5 

80.77 
4.63 
3.08 

I- 54 
3-85 

86.66 
5-83 
2.50 
1.67 

2.50 
.83 

79-76 

10.  71 

1. 19 

3-57 

1.  19 

2.  iH 

84.62 
1.92 

78.07 
S-26 

70.88 
6.59 
3- 30 
i-iO 
2.  20 
2.  20 

1.  10 
1.65 

2.  20 
2.  20 

•55 
2.  20 
•55 
•55 

80.00 
S-71 
2.86 
2.38 
1.90 
3-81 
1-43 
•95 

87-25 
6. 13 
3- 06 

3-92 

11-15 

16-20 

1 

1.92 
1.92 
3-85 

.88 
t-7S 

.88 
2.63 

.88 

.88 
1-75 
I- 75 

.88 
1-75 

.88 

1 

31-35 

6.  IS 

. . . . 
12.50 

1.92 

41-45 

•  SI 

46-50 

1 

'  ■  ■  'i 

51-SS 

• 1. . 

■^\ 

s6-<5o 

1.92 
1.92 

61-65 

66-70 

71-7^ 

.48 

76-80 

1 

•51 

81-85 

..::;;::r. ;:::■:  .:::•■■■ 

1-75 

1 

86-90 

1                  1 

91-95 

1 :.l :: ....:;::  ..:.:.'y.'.vv' 

•55 

96-100 

. . . .     1 . 

1 

101-105 

"1     

1 

■ 

. 

i 

1 

1 

126-150 



I.  19 

Length  of  pause. 


Days. 


1916-17 


Novem- 
ber. 


Per  d. 

57-78 

H.  II 

6.67 

7^40 


Decern-    Janu- 
ber.        ary. 


Per 

48. 
13^ 


Per  d. 
51.  16 
15.  61 
S^65 
3^65 


Febru-  March, 
ary. 


Per  d.  I  Per  d. 

49. 06  j     64. 16 

14-98  I     12.44 

7.  12  i       7-  83 

3-37  I       4-  ij_ 


Three  years  combined. 


Novem-  Decem- 
ber.    I     ber. 


Per  d. 

62.03 
10.  26 
5- 64 


Per  d. 

52.90 
13-  12 
8.39 
2.  80 


Janu- 
ary. 


Per  , 

52- < 


Febru- 
arj'. 


Per  d. 

51-59 
16.  25 
8.04 
3-02 


March. 


Per  d. 
59-76 
12.47 
8.45 

5-84 


2-S 

6-10 

11-15... 
16-20. . . 
21-25. •. 
26-30... 
31-35.-- 
36-40... 
41-45 ■ . - 
46-50... 
51-55- -. 
56-60.. . 
61-65... 
66-70. . . 
71-75- -■ 
76-80... 
81-85... 
86-90... 
91-95- •• 
96-100. . 
101-105. 
121-125. 
126-150. 


82.96 
6.67 
2.  96 


1.48 


76.08 
6.64 
2.99 
I.  00 
2-33 
2.66 
3-32 
1.66 
.66 
1-33 


74-53 
8-99 
4-  12 
2.  25 
3-75 
3-75 
1-50 
•75 
•37 


•33 


1.48 


88.48 
6.  91 
2.30 
2.30 


83.08 
5^13 
2.  56 
1-54 
1-54 
2.05 


1^03 


77-  20 

4-95 

1-  51 

•65 

2- IS 

1.08 

1.72 

.65 

.86 

1.08 

1.72 

I.  29 

I.  72 

1.08 


78.89 
7.  20 
3^35 
2.18 
2.85 
3-18 
I- 17 
■67 
-  17 


7.24 
2.41 
I.  81 
.60 
.80 


a  Broody  pauses  are  not  included. 


Table  III  gives  certain  values  of  interest  in  connection  with  Table  II. 
In  particular,  the  ratios  of  the  number  of  pauses  to  the  number  of  birds 


558 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  9 


laying,  in  spite  of  marked  differences  in  the  value  of  the  ratios  from  year 
to  year,  are  in  agreement  (with  one  partial  exception)  in  having  lower 
ratios  at  each  end  of  the  season  than  in  the  middle. 

Table  III. — Number  of  birds  laying  and  number  of  pauses 


Number  of  birds  laying 

Total  number  of 
pauses 

Ratio  of  pauses  to 
number  of  birds  lay- 
ing (expressed  in 
percentage) 

Number  without  any 
pause 

Percentage  without 
any  pause 

Number  with  one  or 
more  pauses 

Percentage  with  one  or 
more  pauses 


34-7 

65.  22 


34-78 


104 
51 

49.0 

57 
54.81 

47 
45-  19 


1915-16 


1916-17 


161 
130 

81.0 

83 

SI- 55 

78 

48-45 


176      aSj        60 
120        S3 


95 
53-98 


100    86.  6 
35        34 
42.  17  56.  67 
81         48        26 
46.02.57.8345.33 


128 
114 

89.0 

61 

47.66 

67 

52-34 


loi.  7;  100.  c 
88 

41.  9042.  II 

104      121 

58.  10^57.  89 


145 
13s 

89.  I    93-  I 
61 

;o.  91  42.  o 
ic8 


266 
300 


33-46 

177 


49-0957-9366.54 


303 

301 

99-3 
133 

43-89 
170 

56.  II 


352 

267 


75-9 
172 


376 

217 

57-  7 

213 

56-65 

163 


o  A  part  of  this  flock  was  not  trap-nested  after  March  i. 
SIGNIFICANCE   OK   THE  PAUSES 

It  appears  from  a  consideration  of  the  data  presented  in  the  preceding 
paragraphs  that  pauses  differ  in  their  significance.  It  seems  clear  that 
three  classes  at  least  can  be  distinguished :  First,  the  long  pause  that  is 
clearly  indicative  of  the  presence  of  a  winter  cycle.  The  actual  length 
of  this  pause  in  exceptional  instances  need  not  exceed  three  or  four  days, 
provided  it  comes  at  the  proper  season  of  the  year,  and  is  preceded  by  a 
considerable  period  of  egg  production.  Ordinarily,  however,  it  exceeds 
10  days  in  length.  Second,  short  pauses  which  occur  at  frequent  inter- 
vals in  the  records  of  particular  individuals  because  of  their  very 
number  must  be  regarded  as  having  considerable  significance  (see  Table 
I,  sire  6781).  This  type  may  be  called  "multipause"  provisionally 
and  is  to  be  distinguished  from  the  type  in  which  production  is  essen- 
tially continuous,  even  though  no  sharp  dividing  line  can  be  drawn 
between  these  two  groups.  It  is  possible  that  certain  records  of  the 
multipause  type  represent  a  particular  genotype,  since  there  is  a  pro- 
nounced tendency  for  many  multipause  records  to  occur  in  the  same 
families.  This  tendency  has  been  particularly  marked  in  a  small  flock 
of  Brown  Leghorns.  Another  miultipause  type  results  from  intermittent 
egg  production  during  the  winter  pause. 

A  third  class  includes  those  individuals  which  exhibit  only  one  or  two 
short  pauses.  Some  individual  records  probably  represent  extreme 
variants  of  the  multipause  type,  others  pauses  at  the  end  of  the  winter 
cycle,  but  the  majority  are  clearly  without  particular  significance. 


Mar. 4.  i9i8       V/inttr  Egg  Production  of  Rhode  Island  Reds 


559 


NUMBER   OP  PAUSES   PER   INDIVIDUAL 

In  Table  IV  is  shown  the  number  of  pauses  above  two  days  in  length 
per  individual,  for  each  of  the  three  years.  The  data  for  each  of  the 
three  years  are  divided  into  three  groups,  determined  by  the  time  at 
which  egg  production  began.  The  chief  points  of  interest  are:  (i)  the 
large  percentage  of  records  without  a  pause  even  when  production  begins 
before  January,  (2)  the  progressive  diminution  in  number  of  individuals 
as  the  number  of  pauses  rises,  and  (3)  the  agreement  with  expectation — 
viz,  that  those  birds  that  begin  to  lay  early  should  have  more  pauses 
than  those  beginning  to  lay  later  in  the  season. 

Table  IV. — Number  of  pauses  per  individual  {disregarding  I-  and  2-day  pauses),  arranged 
in  three  periods,  according  to  the  time  egg  production  began  (>■ 


Number  of  individual  pauses  in^ 

1913-14 

1915-16 

1916-17 

Number  cf  pauses. 

Where  pro- 
duction 
began — 

"a 
1 

0 

Where  pro- 
duction 
began — 

"a 
1 

G 

Where  pro- 
duction 
began — 

"a 
1 

IS 

a  g 

4 

la 

bet: 

3  9 

be  aj 

.SH 

2i2 

II 

a  nj 
■n  3 

^1 

.ss 

a 
0 

u 

V 

59 
34 
19 
4 
I 
3 

29 

9 

s 
0 

I 

12 

\ 

0 

I 
0 

90 
S4 
29 
9 
2 
4 

47-87 
28.  72 
15-43 
4-  79 
I.  06 
2-13 

33 
61 
40 
20 

5 
3 

■!2 

16 

II 
3 

21 
16 
I 
0 
0 
0 

86 

93 

52 

23 

6 

4 

0 

I 

32-45 
35-09 
19.  62 
8.68 
2.  26 
I.  51 

■■'.'38' 

66 

III 

57 

38 

18 

8 

I 

5 

I 

3 

26 
19 
13 
s 

s 
I 

0 

0 
0 

0 

20 
6 
I 
0 
0 
0 
0 
0 
0 
0 

112 

136 
71 

43 
23 
9 

I 
5 
I 
3 
292 

404 

27.72 

33-66 
17- 57 
10.  64 
S-69 
2.  23 

•25 

1.24 

•2S 

•74 
72.28 

6 

0 

0 

8 

9 

98 

188 

52-  13 

179 

26s 

67-  55 

Total  number  of  in- 

o  A  few  records  in  which  production  began  very  early  in  the  fall  liave  been  omitted. 

In  Table  V  the  number  of  instances  of  different  lengths  of  pauses  is 
shown. 


Table  V. — Number  of  instances 

of  pauses  j 

0  5  days 

trt  length 

(5  ^0  70, 

and 

II  u 

P 

1913-14 

1915-16 

I9I6- 

17 

^ 

^ 

u 

Number  of  pauses. 

>. 

>. 

a 
3 

i 

c 

3 

>. 

a 

3 

nl 

"0 

fl 

0 

a 

•0 

r1 

0 

« 

T3 

a 

u 

b 

•a 

-t- 

0 

•a 

-h 

0 

•O 

+ 

■^ 

0 

1^ 

« 

1 

Jn 

m 

" 

03 

.^ 

pq 

^ 

n 

a 

^^ 

-1- 

— 

+ 

^-' 

+ 

^-^ 

-t- 

-f- 

^ 

-t- 

< 

a 

< 

0 

< 

< 

n 

< 

0 

■< 

<,' 

m 

< 

0 

< 

H 

1 

36 

13 

49 

28 

77 

63 

24 

87 

80 

167 

89 

so 

139 

96 

23  s 

479 

2 

3 

4 

13 
4 

I 
0 
0 

J4 
4 
5 

3 
0 

0 

17 
4 

5 

19 
6 
2 

5 
I 
0 

24 

7 

2 

18 
2 
0 

42 
9 

2 

39 
IS 

10 

II 

I 
0 

so 
16 
10 

36 
9 
0 

86 

2S 

10 

I4S 
38 
17 

5 

6 

0 

I 

0 

I 

4 

0 

4 

0 

4 

S 
0 

7 

8 

I 

0 

I 

0 

I 

I 

Total 

58 

14 

72 

31 

103 

91 

30 

121 

100 

221 

153 

62 

220 

141 

n6o 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  9 


TOTAL  LENGTH   OF   TIME   SPENT  IN  PAUSES 

In  Table  VI  the  percentage  of  the  flock  spending  the  specified  number 
of  days  in  pauses  is  given.  It  should  be  noted  that  pauses  of  one  or  two 
days  in  length  have  been  disregarded  in  compiling  this  table.  Thus,  the 
row  marked  zero  does  not  mean  that  the  birds  laid  continuously  or  that 
they  had  no  2-day  pauses,  but  that  the  total  time  spent  in  pauses,  dis- 
regarding those  of  one  or  two  days,  was  nothing.  The  same  is  true 
for  all  the  other  values  for  the  column  headed  "Total  time  spent  in 
pauses."  An  examination  of  the  individual  records  shows  that  the  2- 
day  pauses,  which  might  be  taken  as  the  lower  limit  instead  of  3 -day 
pauses,  are  distributed  proportionately. 

Table  VI. — Percentage  distribution  of  the  total  time  spent  in  pauses 
[Records  beginning  after  January  i  are  excluded] 


Total  time  spent  in  pauses. 


o. .. 

3-7- 


13-17... 
18-22... 
23-27... 
28-32... 

33-37- •• 
38-42... 

43-47- •  • 
48-52... 

53-57- •• 
58-62 . . . 
63-67... 
68-72... 

73-77- •- 
78-82... 

83-87..- 
88-92... 

93-97- •• 
98-102. . 
103-107. 
108-112. 
113-117. 
118-122. 
123-127. 
128-132. 

133-137- 
138-142. 

143-147- 


Days. 


Percentage  of  flock  pausing. 


March 
hatched. 


60.87 

13.  04 

8.  70 

4-35 

2.  17 


6.52 

2.  17 
2.  17 


April  and 

May- 
hatched. 


53-331 

16.  67 

10.  00 

I.  67 

6.67 

I.  67 

3-33 
1.  67 


I.  67 


3-32 


1915-16 


March 
hatched. 


16.  22 
10.  81 
4-05 
4-05 
5-41 
1-35 


».  II 

^■35 
8.  II 
6.76 
2.  70 
5-41 
5-41 
1-35 
2.  70 

1-35 
5-41 

1-35 

I- 35 
1-35 

1-35 

1-35 
1-35 

1-35 


April  and 

May 
hatched. 


30.88 
II.  76 
5-88 
8.82 
5.88 
5-88 
5.88 
7-35 
2-94 
2.94 
1.47 

4.41 
2.94 


1-47 
1.47 


1916— 17 


March 
hatched. 


'5-38 
7.69 
2.  56 


2.  56 
2.  56 
2.  56 


2.  56 
2.56 
7.69 
2.56 
2.  56 
7.69 

5-13 
2.  56 
10.  26 
2.56 
2.56 
7.69 
2.56 


2.  56 
2.  56 

■2.56 


April  and 

May 
hatched. 


34.08 
17.49 
93 
93 
69 

93 
59 
93 
14 
14 
24 
79 
14 
14 
69 

45 

90 
90 

45 


45 


Mar.  4,  i9i8       WiuieY  Eqq  Production  of  Rhode  Island  Reds  561 

Several  features  of  this  table  require  comment.  First,  the  large  per- 
centage of  the  1 91 3- 1 4  flock  that  falls  in  the  zero  class,  and  the  low 
value  of  the  pause  of  maximum  length.  It  should  be  noted  that  this 
flock  was  not  put  into  the  laying  houses  until  late  October,  and  none  of 
them  began  to  lay  until  after  November  i.  Second,  the  high  percentage 
of  the  flocks  of  the  April  and  May  hatches  of  191 5-1 6  and  191 6-1 7  that 
spent  practically  no  time  at  all  in  pauses.  Thus,  the  sums  of  the  first  two 
rows  are 42.64  and  51 .56  percent,  respectively.  This  high  value  is  the  more 
remarkable  for  the  season  of  1916-17,  since  the  birds  began  to  lay  early 
in  the  fall.  Contrasted  with  this  is  the  third  point — viz,  the  relatively 
low  percentage  of  the  March-hatched  pullets  of  these  two  years  that  fall 
into  these  rows.  This  may  perhaps  be  explained  by  the  tendency  of 
March-hatched  pullets  to  begin  to  lay  early  in  the  fall  and  to  undergo 
a  winter  molt.^  That  the  191 3-1 4  flocks  shows  a  radically  different  dis- 
tribution is  probably  due  to  its  delay  in  beginning  production.  This 
raises  the  question  as  to  whether  or  not  the  time  a  bird  begins  to  lay 
may  not  have  a  considerable  effect  on  the  appearance  or  nonappearance 
of  the  winter  pause.  The  individual  records,  however,  prove  that  no 
necessary  relation  of  the  sort  postulated  exists. 

B. RATE    OF    PRODUCTION 

VALUE  OF  MONTHLY  EGG  PRODUCTION  AS  AN  INDEX  OP  A  WINTER  CYCLE 

A  cycle  in  egg  production  may  be  indicated  by  a  lessened  daily  rate 
of  production  as  well  as  by  a  pause.  Specifically  one  would  expect  of 
the  winter  cycle,  if  it  were  delineated  only  by  a  change  in  rate,  either 
that  production  should  begin  at  a  relatively  high  rate  to  be  followed 
during  the  latter  part  of  the  winter  by  a  period  of  lessened  production 
or  that  the  production  should  start  at  a  slow  rate,  rise  to  a  maximum,  and 
then  decline.  Following  Pearl,  we  should  anticipate  that  February 
would  have  less  eggs  than  January. 

In  handling  the  data  given  in  Tables  VII  to  XI  we  have  proceeded  as 
follows,  unless  otherwise  specified.  Each  lot  of  Rhode  Island  Reds  is  a 
group  of  individuals  selected  for  the  following  reasons:  (i)  A  laying 
period  of  considerable  length  in  order  that  ample  time  should  be  allowed 
for  the  completion  of  the  entire  cycle — that  is,  both  egg  production  and 
pause.  Thus,  birds  beginning  to  lay  after  December  i  are  excluded  from 
the  tables.  (2)  Records  containing  broody  pauses  are  excluded.  (3) 
Each  month  should  be  equal  in  length.  We  utilize,  therefore,  three 
periods  of  31  days  each — viz,  the  months  of  December,  January,  and 
February,  including  in  the  latter  the  first  three  days  of  March.  (4) 
The  data,  furthermore,  are  divided  into  two  groups — viz,  March-hatched 

'  Possibly  the  large  percentage  of  JIarch-hatched  pullets  with  pauses  may  be  interpreted  as  being  due 
to  a  greater  opportunity  for  the  appearance  of  the  winter  cycle,  since  they  begin  to  lay  early  in  the  season. 
Pearl,  however,  does  not  mention  an  association  of  a  molt  with  the  winter  pause. 


562 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  9 


pullets  and  those  hatched  in  April  and  May,  the  hatching  date  of  the 
latter  being  the  same  as  for  the  Maine  birds. 

The  data  for  Barred  Plymouth  Rocks  and  Wyandottes  have  been 
extracted  from  the  Maine  Station  Bulletins  79  and  93  (j,  4)  and  have 
been  handled  in  essentially  the  same  way.  As  these  records  have  been 
given  for  months  only,  only  those  birds  that  laid  at  least  one  egg  in 
November  have  been  included  in  the  tables.  Dr.  Pearl  has  stated  to 
the  writer  that  his  Barred  Plymouth  Rocks  do  not  go  broody  during 
the  winter,  so  that  we  have  assumed  that  broodiness  does  not  enter  as  a 
disturbing  factor.  The  February  records  have  been  corrected  to  a 
basis  of  31  days. 

Table  VII. — Mean  monthly  egg  production  of  three  breeds  of  pullets  beginning  to  lay 

before  December  I 

[Broody  records  are  excluded.  No  record  is  included  where  production  began  later  than  December  i. 
No  Barred  Plymouth  Rocks  or  White  Wyandottes  that  failed  to  lay  in  November  are  included.  Febru- 
ary records  are  reduced  to  a  31-day  basis] 

BARRED  PLYMOUTH   ROCKS 


Flock  of— 

Platched. 

Num- 
ber of 
indi- 
vidu- 
als. 

Decem- 
ber. 

January. 

Feb- 
ruary. 

38 

25 

17-3 
16.  2 
17.2 

19.4 

15-9 
17.2 

13-4 
17.  6 
II.  2 

Total 

96 

16.  9 

X7.6 

14-3 

WHITE   WYANDOTTES 


1899-I9OO . 

I90O-I9OI . 
I9OI-I902. 


Total. 


34 

S8 


103 


12.  7 
15-8 


15.0 


17.4 
16.  9 
13.  2 


16.  7 


14.7 
13-4 
12.7 


I3.8 


RHODE  ISLAND  REDS 


1913-14- 
I915-16, 
I916-17 , 


Total. 


fMarch 

1  April  and  May. 

fMarch 

\ April  and  May. 

|March 

\ April  and  May. 


10 

21.  I 

21.5 

13 

21.  2 

21.  9 

5& 

17.0 

14.  0 

24 

19.9 

14.9 

32 

II.  6 

7-9 

122 

19-3 

15.0 

257 

18.  I 

14-5 

20.  6 

21.  6 
12.  9 
12.8 
12.  9 
16.3 


15-2 


From  Table  VII  it  is  clear  that  the  egg  production  of  February  is  usually 
lower  than  that  for  January.  Some  exceptions  are  to  be  noted,  especially 
in  the  Rhode  Island  Reds,  and  also  for  1 900-1 901  in  the  Barred  Plymouth 


Mar.  4.  i9i8       Winter  Egg  Production  of  Rhode  Island  Reds 


563 


Rocks.  The  Rhode  Island  Reds,  compared  with  the  Barred  Plymouth 
Rocks  and  the  White  Wyandottes,  show  a  less  difference  between  the 
two  months.  In  this  connection  it  may  be  noted  that  the  mean  value 
of  the  February  egg  production  for  eight  years,  as  given  for  the  Barred 
Plymouth  Rocks  by  Pearl  (7)  if  reduced  to  a  basis  of  31  days  is  12.03, 
approximately,  or  slightly  higher  than  for  January,  which  is  11. 71,  The 
mean  egg  production  for  February  (28  days)  for  the  entire  flock  of  our 
Rhode  Island  Reds  has  always  been  higher  than  for  January. 

Table  VIII  is  similar  to  Table  VII,  but  we  have  utilized  the  larger 
numbers  made  available  by  including  in  the  January  and  February 
means  the  records  made  by  birds  beginning  to  lay  as  late  as  January  i. 
The  essential  result  is  the  same  as  shown  in  Table  VII. 


Table  VIII. 


-Mean  monihly  egg  prodticiion  of  pullets  from  the  same  flocks  as  Table  VII , 
but  including  those  beginning  to  lay  as  late  as  January  i 


1899-1900. 
1900-1901. 
1901-1902 . 


Total . 


BARRED  PLYMOUTH   ROCKS 


Hatched. 

Number  of  in- 
dividuals 
laying  in — 

Decem- 
ber. 

January. 

Febru- 
ary. 

Flock  of— 

Decem- 
ber. 

Janu- 
ary and 
Febru- 
ary. 

38 

25 

55 
56 
45 

17-3 
16.  2 
17.2 

18.7 
15-5 
15-4 

II. 8 

16.  5 
12.4 

Total 

96 

156 

16.  9 

16.6 

13.7 

WHITE   WYANDOTTES 


34 

50 

18.8 

18.3 

58 

77 

12.7 

16.  9 

II 

31 

15-8 

16.5 

103 

158 

15.0 

17.2 

14.  o 
14.  6 
12.7 


14.  o 


RHODE   ISLAND   REDS 


I913-14 

I915-16 

I916-17 

Total, 


/March 10 

lApril  and  May j  13 

fMarch '  56 

\ April  and  May 24 

/March 1  32 

\April  and  May 122 


256 


46 
62 
90 
70 

44 
230 


542 


21.  I 
21.  2 
19.9 
17.  I 
II. 6 
19-3 


17.7 


20.  5 
20.6 

14-3 
16.8 

8.3 
16.  I 


14.8 


19.7 
19.9 
13-5 
15-3 
II.  I 
17.  o 


16.  o 


For  the  Barred  Plymouth  Rocks  and  White  Wyandottes  the  December 
production  varies   in   the   individual   years,   being  sometimes   greater, 
38324°— 18 4 


j64 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  9 


sometimes  less  than  for  January.  For  the  Rhode  Island  Reds  the 
December  production  is  considerably  greater  than  for  January  and  also 
for  February.^  - 

Table  IX. — Mean  monthly  egg  production  for  all  Rhode  Island  Reds  that  began  to  lay 
by  December  i,  separated  into  two  groups,  according  to  length  of  pauses 

[Any  record  having  a  single  pause  of  more  than  jo  days  in  length  is  included  in  the  group  of  pauses  of 
II  days  or  more.  A  record  having  several  short  pauses  of  less  than  1 1  days  each,  even  though  aggregating 
more  than  lo  days,  is  included  in  the  groups  of  pauses  of  less  than  ii  days] 


Flock 

cf- 

Grouping  of  records. 

Hatched. 

Num- 
ber of 
indi- 
vidu- 
als. 

De- 
cem- 
ber. 

Janu- 
ary. 

Febru- 
ary (31 
days). 

I9I3-I4 

All  birds 

23 

9 
II 

21.  I 

21.  0 

22.  6 

21.  7 

22.  0 
23.1 

20.  9 

22.  2 
22.3 

Records  with  pauses 
not  exceeding    10 
days  in  length. 

Records  with  pauses 
of  II  days  or  more. 

All  birds 

fMarch 

\  April— May       

Hatches  combined 

("March 

[April-May 

20 

21.  9 

22.  6 

22.3 

I 
2 

22.  0 
13.0 

17.0 
15-5 

I.  I 

16.5 

Hatches  combined 

3 

16.  0 

16.0 

II.  4 

I9I5-I6 

80 

20 
II 

17.9 

19.4 
19-5 

14-3 

19.9 
19.  2 

12.  9 
20.  0 

Records  with  pauses 
not  exceeding   11 
days  in  length. 

Records  with  pauses 
of  1 1  days  or  more . 

All  birds 

/March 

[^  April— May 

18.4 

Hatches  combined 

JMarch 

31 

19-5 

19.  6 

19.4 

36 
13 

15-7 
20.3 

10.8 
II.  2 

9.0 
8.0 

\April— May 

Hatches  combined 

49 

16.9 

10.  9 

8.7 

1916-17 

154 

7 
69 

17.8 

21.7 
22.  4 

13.6 

19.7 
21.  I 

15- 5 

20.  7 
20.  9 

Records  with  pauses 
not   exceeding    11 
days  in  length. 

Records  with  pauses 
of  1 1  days  or  more. 

fMarch 

\  April— May 

Hatches  combined 

/March 

\April-May 

76 

21.7 

19.7 

20.  7 

25 
53 

8.8 
15-6 

4.6 
7.2 

8.8 
10.7 

Hatches  combined 

78 

13-4 

6.1 

10.  I 

If  the  data  for  Rhode  Island  Reds  are  analyzed  further,  as  is  done  in 
Table  IX,  certain  interesting  facts  come  to  light.  In  this  table  the 
records  have  been  divided  into  two  groups,  one  consisting  of  those  records 


'  A  comparison  was  attempted  between  November  and  December  egg  production,  but  no  satisfactory 
results  could  be  obtained,  because  of  the  small  number  of  birds  which  had  complete  records  for  November. 


Mar.  4,  i9i8       IVinier  Egg  Production  of  Rhode  Island  Reds  565 

that  have  no  pauses  exceeding  10  days  in  length,  and  the  second  of  those 
that  have  pauses  of  more  than  10  days.  In  the  former  group  the  egg 
production  remains  at  essentially  the  same  point  in  all  three  months. 
In  the  second  group,  however,  the  January  and  February  production  is 
markedly  lower  than  that  of  December.  The  production  for  January 
may  or  may  not  exceed  that  of  February.^ 

DIFFERENCES  IN  MONTHLY  EGG  PRODUCTION  DURING  DECEMBER,  JANUARY,  AND  FEB- 
RUARY FOR  RHODE  ISLAND  REDS,  BARRED  PLYMOUTH  ROCKS,  AND  WHITE  WYAN- 
DOTTES 

In  Table  X  we  have  examined  the  differences  in  the  monthly  egg 
production  for  the  same  set  of  birds  shown  in  Tables  VII  and  VIII.  For 
the  Barred  Plymouth  Rocks  and  White  Wyandottes  it  is  noticeable  that 
for  each  year  but  one  the  number  of  birds  laying  more  eggs  in  January 
than  February  is  greater  than  the  number  laying  more  eggs  in  February 
than  January  (in  the  exception  the  number  is  equal)  and  that  in  five  out 
of  six  instances  the  average  number  of  eggs  laid  by  the  birds  having  an 
excess  January  production  is  greater  than  for  those  having  an  excess 
production  in  February.  The  percentage  of  birds  having  an  excess 
January  production  is  75.29  for  Barred  Plymouth  Rocks  and  68.23  ^or 
the  White  Wyandottes. 

The  data  for  the  Rhode  Island  Reds  are  quite  different  in  character. 
Only  51 .39  per  cent  of  the  March-hatched  pullets  lay  more  eggs  in  January 
than  February,  while  for  April  and  May  hatched  pullets  the  value  is  47.16 
per  cent,  although  in  two  years  out  of  three  the  number  of  birds  with  a  Janu- 
ary excess  is  greater  than  for  February.  In  only  one  instance,  and  that 
with  a  small  number  of  birds  involved,  is  the  number  much  more  than 
twice  the  number  having  an  excess  February  production.  This  is  to  be 
compared  with  the  Maine  records,  where  in  every  instance  the  number  of 
birds  with  a  January  excess  is  greater  than  those  with  a  February  excess 
and,  except  in  two  instances,  are  two  to  three  times  as  numerous  (in 
one  case  1 6  times).  For  the  Rhode  Island  Reds  hatched  in  April  and  May 
the  average  number  of  excess  eggs  in  two  years  out  of  three  is  greater  for 
those  pullets  with  an  excess  February  production.  For  the  March- 
hatched  pullets  conditions  are  reversed. 

'  It  is  possible  that  the  two  groups,  high  producers  and  mediocre  producers,  should  be  separated  in  an 
analysis  of  this  sort.  So  few  of  the  latter  were  available,  however,  that  it  seemed  unwise  to  attempt  the 
separation. 


566 


Jouryial  of  Agricultural  Research 


Vol.  XII,  No.  9 


^ 


a  S 


•2^ 


*§ 


>^ 


1    - 

«  ■  A 

to 

9. 

w 

0 

K 

10 

M 

sR 

^ 

otc  a  ..• 

a 

vO    •« 

0>  vn 

v  2 

5<i  3  .J, 

0  to 

« 

M     10 

f^ 

t^  10 

c2 

■0     • 

a 

ca 
>-> 

'^ 

^ 

rt  Q.J,   t   ,,; 

to   H 

•* 

0     0 

0 

w  xS 

to        r^  t^ 

t^        0   ^ 

3 

„     ^ 

00     1    M 

§ 

A 

OJ   ^   3I 

to  « 

(« 

t>.\0 

06  -"i- 

vn 

t^ 

M     Tt 

>o       00 

S^- 

,°  S^ 

00               M 

«      0 

rti  *^ 

&I 

^ 

t^cn 

in 

.n  r. 

3 

d 

t 

■<t 

^^ii 

to  >o 

00 

10  0 

>o 

t^  0 

„ 

0 

>—• 

3  ,  • 

c3«  3  ^ 

0  0 
0  -c 

-c 

S>8 

s 

0  ^ 

0 

5 

w  2 
1-2 

05 

to  ■* 

•  to 

VOvO 

lO 

00  tf 

"' 

u> 

2«  s  >. 

O  —  .Si  0 

o. 

Cl 

0    <^ 

00  1^ 

" 

"A 

go.i  -  ^ 

■"  "^ 

CO 

00   ■* 

- 

0  t^ 

■* 

t!.T 

3 

""  X  « 

10 

0   lO 

1       " 

^o  -o 

00 

0 

« 

to  0^ 

■«•  to 

.0 

0  0 

0 

0  in 

-1- 

fe    . 

1    1    7i 

0 

't  to 

t^ 

1- J3 
1'^ 

j2-e 

to  w 

10 

M     •* 

"- 

tH  ^O 

00 

M     0 

- 

M  M 

o\ 

0 

V 

oJ,iS 

to  « 

VO 

C     H 

to 

0  -0 

>o 

00 

« 

^^•ss 

3    ,    < 

f. 

to  to 

0   0 

0        to 

" 

vO    t<- 

0)   ^ 

g-0-3" 

to  to 

to 

tovO 

t 

lo  f 

to      to 

8  a 

2(3  3  ^ 

0   » 

"" 

-0     H 

"   " 

lii^i 

.0    u- 

0        to  t^ 

05 

00 

^  i-  S 

OJ 

2§ 

S  "  ^ 

"  '^ 

" 

*o  t~ 

t^ 

v6  ■■6 

vO        VO 

,0  0 

0  0 

»o^o 

„ 

°"2 

r^ 

0       0 

^»-> 

{-,   X 

M 

0" 

■    1 

0     M 

„ 

%o  td 

0  II   0  ..     1 

1  ::?-^ 

1  - 

O^iS 

n 

C*     H 

to 

1   ° 

II   0 

^^^s 

1 

■a 

J3 

i 

3 

J2 

f^ 

"3 

« 

•3       a  ^ 

g 

c 

fii 

H 

c 

(-1      f-i|l< 

s 

1'^ 

. 

a  c 

-0 

r. 

.^  , 

■^ 

" 

8i 

M 

» 

» 

so 

' 

c 

c 

1 

t^ 

u-  0    ^, 

to  o>  tr 

"•  : 

t^  t^  w 

.0  : 

" 

"   : 

«   tt  m 

w  0 

^ 

" 

0\  0  t^ 

«  moo 

■0 

(-1  0   CT. 

r.. 

N     m   t- 

■«•  0 

to    M     M 

r-. 

M     (M     w 

•r 

t-i 

■^ 

vo   r~  f 

-0 

0    ti  00 

0 

^0    0    .- 

0    to  I' 

00    T 

M      HI 

Cl    ».. 

•r     I 

to     1 

m  t-cc 

■o-  0  0 

00   >n  Q 

'" 

to  «   !/■ 

o\ 

" 

0 

M      •^-     Cl 

t>. 

to  r,  c. 

to  0  0 

tr 

00   t- 

0 

t^    W     M 

0. 

r.. 

^ 

"     -!t« 

m  0 

troo 

m 

00   0  00 

o\ 

m  t-^  «/ 

.        -n 

0    0    I 

tH  m  t. 

1-1   l-l   w 

•^  ^ 

^ 

ti   d   n       0 

> 

^> 

> 

a  ca  0 

3       1 

sss 

5^ 

•c-o-c 

a  S  S 

HOh 

^H  .-^  HH 

"C   C  w- 

a  ac 

<;<< 

0     H     W 

0  0  c 

'oh  I 

"    H 

1 

Mar.  4. 1918       Winter  Egg  Production  of  Rhode  Island  Reds 


567 


rJ-OO    »^ 
1^  '^'O 

CO      ; 

N  00   r^ 

r^    . 

r^O   -"T 

1^0 

000 

0\  rt  ^ 

»^0  OC 

•6    '■ 

- 

&§"' 

-C'      • 

0 

^rt  0          t- 

" 

0 

0  0  0> 

^ 

t^O    f^ 

-0  0 

1.  00    0 
Ci    H    0 

06    0      M 

0 

C>0    vo 
00    fO  >n 

0 

CO 

f1 

ro  <S    l/^ 

0  ^ 

(M     Tf   0 

0 

•3-  0    0 
\0    i^  0 

1- 

H"^ 

T 

"-" 

so     0 

rr  0 

vO    10  0 
0  00   0 

r-  i>^  r^ 

0 

M   CO     T 

r^ 

>0    t<5 

vO     0 

10  T  M 

0 

> 

« 

r 

c 
< 

n 

(a 

-d 

a 
a 

J,  G 

1 

3 

\ 

^ 

568 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  9 


A  consideration  of  the  number  of  individuals  laying  an  excess  number 
of  eggs  for  December  over  January  (and  also  February)  shows  that 
more  of  the  Rhode  Island  Reds  have  a  December  excess  than  a  January 
(or  February)  excess.  For  the  Barred  Plymouth  Rocks  and  White 
Wyandottes  more  individuals  have  a  January  production  in  excess  of 
that  of  December,  but  a  February  deficiency.  The  difference  in  the 
case  of  the  Barred  Plymouth  Rocks,  however,  is  not  great.  Neverthe- 
less it  is  clear  that  the  egg  production  of  these  two  varieties  does  not 
undergo  a  period  of  depression  at  the  same  season  as  that  of  the  Rhode 
Island  Reds. 

TablB  XI. — Distribution  of  individuals  having  a  greater  egg  production  in  January 
over  February,  or  vice  versa,  arranged  according  to  the  maximiim  length  of  any  single 
pause 


January  production  greater  than 
February. 

February  production  greater  than 
January. 

1913-14 

1915-16              1916-17 

1913-14 

1915-16                   I916-I7 

Length  of  pauses,  .days. . 

0 

3-10 

II- 

up. 

0 

3-10 

II- 

up. 

0 

IS 
6 

7 
3 

2 
I 

3-10 

6 
6 

7 
3 

I 
2 

II- 
up. 

5 
3 

I 

2 
2 

3 
3 

0 

7 
S 
S 
S 
.... 

3-10 

3 

I 

2 

3 
3 

II- 

up. 

2 

I 
2 

I 
I 

0 

5 
2 
6 
3 

I 

3-10 
:  2 

3 
3 

2 

II- 
up. 

0 

II 
II 
8 

I 
I 

3-10 

10 
7 
9 
3 
3 

2 

II- 
up. 

EXCESS  NUMBER  OF  EGGS . 

II 

8 
4 

I 

4 
I 
3 

2 
4 
3 

3 
2 

I 
2 

I 

I 
3 

3 
I 
3 
6 
2 
2 
3 
3 
3 
3 
4 
3 
2 

I 
3 
4 

3 

5 

I 

8 

I 

I 

9 

6 

I 

I 

8 

I 

I 

3 

I 

'° 

I 

3 
5 





3 

I 

2 



5 

3 

3 

I 

I 





40 

30 

Total 

24 

II 

9 

9 

17 

44 

34 

26 

22 

12 

10 

17 

12 

32 

36 

6l 

Sum  of  first  two  columns 

35 

S 

26 
0 

60 
6 

34 

29 

68 

Number  with  the  same 
production     in    each 

IS 

0 

6 

4 

18 

9 

SUMMARY 


Production  equal. 

February      produc- 
tion greater  than 
January. 

January    production 
greater  than   Feb- 
ruary. 

Length  of  pause days. . 

0 

3-10 

ii-up. 

0 

3-10 

ii-up. 

0 

3-10 

II-Up. 

39 

II 
so 

13 

71 

60 
131 

103 

67 

S4 
121 

93 

Sum  of  first  two  columns  in  each  sec- 

Sum  of  all  three  colimms  in  each  sec- 

63 

334 

214 

^ 

Mar.  4. 1918       Winter  Egg  Production  of  Rhode  Island  Reds  569 

DrFFERENCE  IN  JANUARY   AND  FEBRUARY   EGG  PRODUCTION,    TAKING   THE   LENGTH  OP 
PAUSE   INTO   CONSIDERATION 

The  results  of  an  examination  of  this  point  are  shown  in  Table  XI, 
which  gives  no  evidence  that  a  difference  in  the  egg  production  of  an 
individual  for  the  two  months  is  more  likely  to  occur  in  one  month  than 
the  other,  since  the  number  of  individuals  falling  in  the  different  cate- 
gories is  approximately  equal,  although  fluctuating  somewhat  from  year 
to  year. 

These  data,  taken  in  connection  with  those  previously  discussed 
regarding  differences  in  monthly  production  and  with  the  observations 
on  rate  of  production,  show  that  a  lower  rate  of  production  for  Feb- 
ruary over  January  in  the  individual  case  is  without  significance,  unless 
associated  with  a  definite  pause.  This  information  is  of  particular  value 
for  those  instances  in  which  the  production  for  February  is  only  a  few 
eggs  less  than  for  January  and  which,  from  Pearl's  data,  might  be  con- 
sidered to  exhibit  a  winter  cycle. 

Table  XII. — Mean  production  of  two  groups  of  pullets  hatched  in  April,  igi6,for  the 
periods  after  the  first  egg  designated  in  the  headings 

BIRDS   BEGINNING   TO   LAY  OCT.    S-DEC.    13,    INCLUSIVE   (80  TO   72    INDIVIDUALS^) 


Days  of  production . . . 


33-63 


63-93 


Mean 
March 
produc- 
tion. 


Mean  number  of  eggs 
per  period 

Mean  date  of  begin- 
ning and  end  of 
each  period 


16. 96 


Nov.  9 

to 
Dec.  10. 


Dec.  II 

to 
Jan.  10. 


14.56 


Jan.  II 

to 
Feb.  10. 


Nov.  9 

to 
Nov.  19. 


17-63 


Nov.  3o 

to 
Dec.  3o. 


15-87 


Dec.  31 

to 
Jan.  30. 


Jan. 31 

to 
Feb.  30. 


BIRDS  BEGINNING  TO  LAY  DEC.    14-JAN.   25,   INCLUSIVE  (69  TO  55  INDIVIDUALS «) 


Mean  number  of  eggs 
per  period 

Mean  date  of  begin- 
ning and  end  of 
each  period 


Jan.  3 

to 
Feb.  I. 


Feb.  3 

to 
Mar.  3. 


Mar.  4 

to 
Apr.  4- 


Jan.  2 

to 
Jan.  12. 


Jan.  13 

to 
Feb.  12. 


19-38 


Feb.  13 
to 

Mar.  16. 


Mar.  17 

to 
Apr.  16. 


o  The  variation  number  of  individuals  is  due  to  broodiness. 


EGG  PRODUCTION   IN  PERIODS   OF   DEFINITE  LENGTH,    BEGINNING   WITH  FIRST   EGG 

In  the  preceding  paragraphs  we  have  used  records  that  begin  before 
a  definite  date — viz,  December  i.  On  being  examined  from  the  stand- 
point of  the  first  egg  it  will  be  observed  that  the  production  for  any  one 
month,  say  December,  is  made  up  of  the  production  of  birds  that  have 
been  laying  for  varying  intervals  of  time,  including  those  well  along  in 
production  and  those  just  beginning.  In  Table  XII  the  rate  of  pro- 
duction has  been  examined  from  another  standpoint.     Three  periods  of 


57©  Journal  of  Agricultural  Research  voi.  xii.no.9 

31  days  each,  beginning  with  the  date  of  the  first  egg,  and  three 
periods  of  31  days  each  beginning  with  the  eleventh  day  of  production 
have  been  employed  in  studying  the  production  of  two  groups  of  April- 
hatched  pullets.  One  group  has  a  mean  date  of  first  egg  on  November  9; 
the  other,  on  January  2.  The  reason  for  employing  two  sets  of  three 
periods  of  31  days,  each  differing  by  10  days,  for  each  lot,  lies  in  the  fact 
that  egg  production  sometimes  is  extremely  slow  and  erratic  at  the  start 
and  that  this  may  reduce  the  egg  production  disproportionately  for  the 
first  3 1 -day  period. 

The  following  points  are  shown  by  Table  XII :  First,  the  egg  produc- 
tion of  the  first  group  is  somewhat  inferior  to  that  of  the  second.  Note 
particularly  the  March  production  of  the  first  group  compared  with  that 
period  of  the  second  that  extends  from  March  4  to  April  4.  Second, 
while  there  is  a  fall  from  the  first  31 -day  period  through  each  of  the  two 
successive  periods  in  the  first  group  until  March,  the  second  group  shows 
a  constant  rise  from  period  to  period,  which  may  mean  that  the  time  of 
year  in  v/hich  the  various  periods  fall  is  concerned  with  the  drop  in  pro- 
duction, for  it  will  be  noted  that  the  first  period  of  the,  second  group 
nearly  coincides  with  the  third  period  of  the  first  group. 

CONCLUSIONS  REGARDING  THE  CRITERIA  FOR  THE  WINTER  CYCLE 

IN  THE  INDIVIDUAL 

A  consideration  of  the  data  presented  in  the  preceding  pages  leads 
to  the  following  conclusions  regarding  criteria  by  which  the  winter  cycle 
can  be  recognized  in  the  individual  record. 

First.  The  rate  of  production,  as  shown  by  the  monthly  egg  records, 
fails  to  furnish  a  satisfactory  index  of  the  existence  of  a  winter  cycle  in 
the  individual  Rhode  Island  Red  pullet. 

Second.  The  best  criterion  of  the  existence  of  a  winter  cycle  in  the 
individual  is  the  existence  of  a  pause  in  production  beginning  in  Decem- 
ber, January,  February,  or,  rarely,  March,  following  a  period  of  con- 
tinuous egg  production,  and  usually  exceeding  10  days  in  length.  A 
single  pause  in  some  instances  may  be  replaced  by  series  of  short  pauses 
separated  by  only  one  or  two  eggs.^ 

Third.  In  some  instances  a  short  pause — that  is,  10  days  or  less  in 
length — occurring  in  February  or  March  and  following  a  period  of  several 
weeks  of  continuous  egg  production  may  delimit  the  winter  cycle. 

It  seems  clear  that  the  period  of  low  flock  production  for  the  Rhode 
Island  Reds,  for  birds  beginning  to  lay  sufficiently  early  in  the  season 
may  come  earlier  in  the  winter  than  at  the  Maine  Station.     In  some 

'  The  second  part  of  the  two  recent  bulletins  from  the  Utah  Station— viz,  Ball,  Alder,  and  Egbert  (i), 
and  Ball  and  Alder  (2)— was  received  after  the  manuscript  of  this  paper  had  been  completed.  Only  a  very 
brief  comment  can  be  made  on  their  discussion  of  the  "'-winter'  egg-laying  period"  in  White  Leghorns. 
They  conclude  "that  there  is  no  apparent  biological  ground  for  either  the  beginning  or  end  of  this  period 
..."  This  conclusion,  which  rests  on  mass  statistics,  needs  reexamination  before  it  can  be  considered  of 
universal  applicability  to  all  White  Leghorns. 


Mar. 4. 1918       Winter  Egg  Production  of  Rhode  Island  Reds  571 

individuals  the  pause  comes  comparatively  early  in  the  winter  so  that  the 
following  [spring  ( ?)  cf .  Pearl  and  Surface  (7)]  cycle  of  production  may 
begin  as  early  as  the  middle  of  January.  This  may  mean,  perhaps,  that 
the  winter  pause  in  Rhode  Island  Reds  is  not  homologous  with  that  of  the 
Maine  Station  birds.  Whether  or  not  this  is  so,  it  would  seem  desirable 
to  look  at  it  from  a  somewhat  dififerent  standpoint.  It  may  be  that  this 
pause  follows  an  initial  cycle  of  production.  It  may  be,  too,  that  we  are 
not  dealing  wholly  with  an  inherent  pause  but  with  a  pause  that  depends 
in  part  on  the  environment  for  its  manifestation,  which  is  due  to  a 
difference  in  resistance  on  the  part  of  individuals  to  the  weather  con- 
ditions at  this  season  of  the  year. 

NUMBER  OF  EGGS  LAID  BEFORE  THE  WINTER  PAUSE  AND  LENGTH 

OF  WINTER  PAUSE 

In  those  instances  where  the  winter  pauses  could  be  determined  with 
some  degree  of  accuracy,  we  have  determined  the  range  and  mean  num- 
ber of  eggs  laid  before  the  pause  and  the  same  constants  for  the  pause 
itself,  for  the  pullets  of  the  April  and  May  hatches  shown  in  Table  XIII. 
It  should  be  borne  in  mind,  however,  that  in  most  instances  the  limit 
specified  earlier  has  been  used  for  the  lowest  number  of  days  indicative 
of  the  winter  pause. 

For  the  number  of  eggs  before  the  pause,  the  range  has  a  value  of 
2  to  96,^  with  a  mean  of  35.98  eggs.  The  length  of  the  pause  has  a 
range  of  8  to  72  days,  with  a  mean  of  34.23  days.  For  the  birds  listed 
under  section  B  of  Table  XIII,  the  values  are  for  eggs;  range  i  to  96,* 
mean,  36.98,  and  for  length  of  pause  range,  3  to  104,  mean  24.87. 

The  possibility  of  a  correlation  between  the  number  of  eggs  laid  before 
the  pause  and  the  length  of  the  pause  has  been  examined  and  found 
to  be  practically  nonexistent. 

MODE  OF  INHERITANCE  OF  THE  WINTER  CYCLE 

From  the  data  that  have  been  presented  in  the  preceding  pages  it  is 
clear  that  some  individual  Rhode  Island  Reds  exhibit  a  definite  winter 
cycle,  while  others  as  definitely  show  no  winter  cycle.  For  purposes  of 
description  we  may  describe  the  former  as  "winter  cycle,"  the  latter 
as  "no  winter  cycle."  It  is  clear,  moreover,  from  an  examination  of 
the  family  records  (Tables  I  and  XIII)  that  the  character  is  inherited 
and  that  segregation  takes  place.  When  the  ratios  are  examined,  how- 
ever, no  evidence  of  an  entirely  satisfactory  character  is  afforded  us  as 
to  the  mode  of  inheritance  of  the  winter  cycle,  although  there  is  some 
evidence  that  the  winter  cycle  is  inherited  according  to  the  simple 

1  The  rather  absurd  values  for  the  lower  end  of  the  range  result  from  the  inclusion  of  a  few  records  that 
are  obviously  out  of  the  ordinary  but  which  can  not  be  excluded.     The  next  lower  value  is  9. 


572 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  9 


Mendelian  scheme,  "no  winter  cycle"  being  the  dominant  allelomorph, 
"winter  cycle"  the  recessive. 

Before  considering  the  evidence  for  this  scheme  there  are  certain 
pecuUarities  of  the  character  that  should  be  listed. 

First.  The  character  does  not  come  to  visible  expression  in  the  male; 
hence,  his  gametic  composition  can  be  made  out  only  through  a  con- 
sideration of  his  maternal  ancestors,  his  daughters,  and  his  sisters. 

Second.  The  character  is  a  physiological  one,  subject  to  possible  in- 
fluences by  the  environment,  and  perhaps  to  other  internal  factors  con- 
cerned with  &gg  production. 

Third.  Pauses  distinct  from  those  indicative  of  the  winter  cycle  but 
likely  to  be  mistaken  for  them  may  occur. 

Fourth.  Difficulties  are  encountered  in  classifying  certain  individual 
records,  such  as  records  with  several  very  short  pauses,  a  single  very 
short  pause,  a  simple  slackening  in  rate  of  production  without  any  defi- 
nite pause,  and  pauses  occurring  at  the  limits  of  the  season.  In  such 
instances  we  have  proceeded  somewhat  arbitrarily.  There  have  been  ex- 
cluded, first,  all  birds  that  have  laid  less  than  20  eggs  after  the  first  of 
January;  second,  all  birds  showing  broody  pauses;  third,  March-hatched 
birds  that  begin  to  lay  early  in  the  season,  and  which  molt. 

TablB  XI 1 1 . — Progeny  of  individual  pairs  of  birds  of  flock  of  igi6-iy,  showing  the  number 
with  a  winter  cycle  (P),  those  without  (N),  and  those  on  which  a  definite  determination 
could  not  be  m,ade  {X) 


Presence 
or  absence 

Method  A. 

Method  B. 

(X4 

u 
0 

Presence 
or  absence 
of  winter 
cycle  in 
mother's 
record  as 

deter- 
mined by 

two 
methods. 

Method  A. 

Method  B. 

u 

lU 

0 

of  winter 
cycle  in 
mother's 
record  as 

deter- 
mined by 

two 
methods. 

P 

N 

X 

P 

N 

X 

"3 

0 

P 

N 

X 

"a 

P 

N 

X 

"3 
1 

4723 

SS8i 

507  7 
6265 
4354 

43" 

4473 

5476 
5463 
4178 
4845 
4138 
4553 
4533 
5S95 
2564 
5771 
4529 
2565 
5703 
6840 
6767 
5719 
5159 
5982 
4518 
5412 
4S28 
5090 
4030 
4254 
49 

NX 

NP 

NP 

NN 

PP 

PP 

XP 

XP 

XP 

XP 

XP 

XX 

XX 

XP 

NN 

NP 

NP 

XP 

NN 

PP 

XP 

XP 

XP 

NN 

XP 

NN 

6 

I 
2 
3 
3 
4 
0 
0 
2 
3 
3 
0 
I 
1 
0 
2 
I 
I 
0 
I 
I 
2 
0 
0 
0 
I 

5 
9 

3 
3 
8 
6 
I 
5 
3 
3 
4 
0 
0 

2 
0 

3 
4 
4 
4 
3 
7 
6 
8 
0 
0 
0 

12 
0 
2 
4 

I 
S 
6 
0 

2 
5 
2 
6 
2 
I 
4 
0 
7 
I 
3 
4 
I 
2 
2 

» 

I 

7 

23 
10 
7 
10 
12 
IS 
7 
S 

8 
6 

3 
4 
4 
5 
12 
6 
7 
8 
9 
10 
10 
2 
I 
8 

12 
2 
3 
7 
5 
5 
0 
0 

2 

6 

4 
3 
I 
I 
0 
3 
4 
I 
3 
3 
3 
5 
3 
0 
0 
2 

4 
8 
3 
1 
6 
6 
I 
5 
3 
2 
3 
0 
0 

2 

I 
2 

4 
4 
2 
3 
6 
4 
6 
0 
0 
0 

7 
0 

I 
2 

I 

t 
0 
2 

3 

I 

i 

I 
3 
0 
4 
I 
2 
2 

° 
I 

I 
2 
I 

6 

23 
10 

7 

10 

12 

IS 

7 

S 

7 

II 

8 

6 

\ 

12 
6 
7 
8 
9 

10 

10 

2 
I 
8 

6781 
6373 
4378 

4786 
3617 

S477 

5240 
5776 

4882 

4509 
3719 

3482 
3600 
174 
284 
4043 
4754 
4012 
4592 
6684 
5881 
4409 
4716 
5230 
6751 
6003 
4844 
6366 
6404 
6232 
5098 
2332 

PP 

NX 

NN 

XX 

XP 

XX 

NN 

PP 

NN 

PP 

XX 

XX 

PP 

NP 

NP 

NN 

NN 

XX 

NN 

XX 

PP 

NN 

XP 

Total... 

I 

I 

3 

2 

2 
I 
2 
0 

4 
I 
0 
0 
0 
0 
0 
I 
0 

2 

I 

3 
0 
0 
0 

61 

8 
2 
7 
2 
0 
I 
1 
0 
2 
0 
0 
6 
4 
8 
2 
5 
9 

2 

3 
4 
5 
0 
13 

175 

2 
3 
4 

7 
2 
3 
4 
8 
5 
3 
2 
2 
I 
2 
I 
I 
I 
7 
0 
0 
3 
2 
13 

158 

II 
6 

14 

II 

4 

5 

7 

8 

II 

4 

2 

8 

5 

10 

3 

7 

10 

II 

4 

7 

8 

.1 
394 

s 

2 

6 

5 
2 
I 
3 
0 
4 
I 
0 
I 
3 
I 
I 
I 
3 

2 

I 

4 

2 

2 

4 
132 

4 

I 
5 
2 
0 
I 
0 
0 
2 
0 
0 
5 
I 
7 
2 
5 
6 
2 
3 
3 
3 
0 
12 

140 

2 

3 
3 
4 

2 

3 
4 
8 
5 
3 
2 
2 
I 
2 
0 
I 
I 

i 
3 
0 
10 

122 

II 
6 

14 

II 
4 
5 
7 
8 

II 
4 
2 
8 
S 

10 
3 
7 

10 

II 
4 
7 
8 
2 

26 

394 

Mar.  4. 191S       Winter  Egg  Production  of  Rhode  Island  Reds 


573 


In  Table  XIII  is  shown  the  classification  of  the  flocks  of  191 6-17, 
arranged  by  families.  Each  individual  is  classified  twice.  According  to 
one  classification  (method  A)  our  endeavor  has  been  to  ascertain  as 
closely  as  our  best  judgment  would  permit,  the  true  status  of  each  in- 
dividual. In  method  B,  however,  we  have  given  the  benefit  of  any  di  .ubt, 
to  the  winter  cycle  and  have  listed  every  bird  in  the  positive  column 
that  could  possibly  be  considered  as  having  a  winter  cycle. 

Section  A  gives  the  ratio  of  61  individuals  with  a  winter  cycle  to  174 
without.  This  is  close  to  the  i  to  3  ratio — viz,  expected  58%"  to  176X — 
for  a  simple  Mendelian  case  of  inheritance.  According  to  method  B, 
however,  the  ratio  approaches  closely  to  equality. 

Table  XIV  gives  the  ratios  for  the  flocks  as  a  whole  for  1913-14  and 
191 5-1 6,  in  addition  to  1 916-17,  and  the  grand  total.  The  years  vary 
somewhat,  but  the  total,  147  to  388,  is  perhaps  merely  a  deviation  from 
the  expected  1 33^  to  401  X>  the  deviation  being  in  the  direction  expected, 
on  the  assumption  that  birds  without  the  genes  for  a  winter  cycle  may 
exhibit  a  false  cycle. 

Table  XIV. — Number  of  individuals  classified  according  to  the  presence  or  absence  of  the 
winter  cycle  for  the  years  igij-14,  igi^-i6,  and  igid-iy 


.  Method  A. 

Method  B. 

Flock  of— 

Birds 

with  a 
winter 
cycle. 

Birds 

with- 
out a 
winter 
cycle. 

Unde- 
termin- 
able. 

Total. 

Birds 
with  a 
winter 
cycle. 

Birds 
with- 
out a 
winter 
cycle. 

Unde- 
termin- 
able. 

Total. 

IQI'?— 14. 

25 

61 
61 

114 
100 

85 
139 
158 

224 
300 
394 

53 
104 
132 

99 

79 
140 

72 
117 
122 

224 
300 
394 

ioic;-i6 

1016—17 

Total 

147  ^ 

388 
401X 

289 

318 
(?) 

Expected. 

For  method  B  the  observed  ratio  for  the  three  seasons  is  289  to  318. 

Mass  figures  of  the  sort  just  given  are  merely  suggestive,  since  the 
ratio  I  to  3  holds  only  under  certain  conditions.  However,  an  examina- 
tion of  the  proportions  in  which  the  two  types  occur  among  the  progeny 
of  a  single  female,  and  in  some  instances  of  the  progeny  of  one  male  by 
several  females,  favors  the  suggestion  given  above.  At  the  same  time, 
the  gametic  constitution  of  the  parents  can  not  be  made  out  with  a 
satisfactory  degree  of  accuracy.  To  be  sure,  one  can  assign  a  gametic 
constitution  to  many  individuals,  but  it  is  impossible  to  check  these  by 
reference  to  preceding  years,  mainly  because  the  number  of  progeny 
from  a  single  pair  in  the  earlier  years  was  too  small  to  afford  critical 
evidence.  Since  prospective  matings  are  likely  to  furnish  critical 
evidence  on  the  point  in  question,  it  seems  advisable  to  defer  any  attempt 
at  a  solution  of  this  phase  of  the  problem  for  the  present. 


574  Journal  of  Agricultural  Research  voi.  xii, N0.9 

SUMMARY 

(i)  An  examination  of  the  data  published  by  Gowell  confirms  the 
statements  of  Pearl  and  Surface  (7)  and  Pearl  (3,6)  regarding  the  presence 
of  a  winter  cycle  in  Barred  Plymouth  Rocks. 

(2)  The  winter  cycle  is  much  more  characteristic  of  the  Maine  flocks 
as  a  whole  than  it  is  of  our  Rhode  Island  Reds,  where  it  can  be  demon- 
strated in  only  a  portion  of  the  flock. 

(3)  The  period  of  decreased  flock  egg  production  for  Barred  Plymouth 
Rocks  and  White  Wyandottes  comes  in  February.  For  Rhode  Island 
Reds  it  may  come  in  January  as  well  as  in  February. 

(4)  A  pause,  or  series  of  pauses,  usually  exceeding  10  days  in  length 
and  following  a  considerable  period  of  regular  egg  production,  is  the  best 
index  of  the  existence  of  a  winter  cycle  in  the  individual  Rhode  Island 
Red. 

(5)  The  rate  of  production  does  not  furnish  a  satisfactory  index  of  the 
presence  or  absence  of  a  winter  cycle. 

(6)  Evidence  is  presented  which  indicates  that  the  winter  cycle  may 
be  inherited  in  some  definite  but  undetermined  manner. 

LITERATURE  CITED 
(i)  Ball,  E.  D.,  Alder,  Byron,  and  Egbert,  A.  D. 

1916.  BREEDING    FOR   EGG    PRODUCTION.      PART   I.      A   STUDY   OP   ANNUAL   AND 

TOTAL  PRODUCTION.     Utah   Agr.  Exp.   Sta.  Bui.   148,   60  p.,    10   fig. 
Bibliography,  p.  59-60. 
(2) 

1917.  BREEDING    FOR  EGG    PRODUCTION.      PART  II.      SEASONAL  DISTRIBUTION  OF 

EGG  PRODUCTION  WITH  SPECIAL  REFERENCE  TO  "  WINTER"  EGG  PRO- 
DUCTION.    Utah  Agr.  Exp.  Sta.  Bui.  149,  71  p.,  28  fig.     Bibliography, 

P-  55-56- 

(3)  Gowell.  G.  M. 

1902.  BREEDING  FOR  EGG  PRODUCTION.     In  Maine  Agr.  Exp.  Sta.  Bui.  79,  p. 

26-40. 

(4)  

1903.  BREEDING  FOR  EGG  PRODUCTION.     In  Alainc  Agr.  Exp.  Sta.  Bui.  93,  p. 

69-76. 

(5)  Pearl,  Raymond. 

I912.   THE  mode  of  INHERITANCE   OF    FECUNDITY  IN   THE   DOMESTIC  FOWL.      In 

Jour.  Exp.  Zool.,  V.  13,  no.  2,  p.  153-268,  2  fig.  Literature  cited,  p. 
266-268.  Also  published  as  Maine  Agr.  Exp.  Sta.  Bui.  205,  p.  283-394. 
1912. 

(6)  

I915.    MEASUREMENT    OF    THE    WINTER     CYCLE    IN     THE     EGG    PRODUCTION     OF 

DOMESTIC  FOWL.  In  Jour.  Agr.  Research,  v.  5,  no.  10,  p.  429-437. 
Literature  cited,  p.  436-437. 

(7)  and  Surface,  F.  M. 

I911.    A    BIOMETRICAL    STUDY    OF    EGG    PRODUCTION    IN    THE    DOMESTIC    FOWL. 
II.    SEASON.'^L  DISTRIBUTION    OF  EGG    PRODUCTION.        U.   S.    Dept.    Agt. 

Bur.  Anim.  Indus.  Bui.  no,  p.  81-120,  30  fig. 


DIGESTION    OF   STARCH   BY  THE   YOUNG  CALF 

By  R.  H.  Shaw,  T.  E.  Woodward,  and  R.  P.  Norton,  of  the  Dairy  Division,  Bureau 
of  Animal  Industry,  United  States  Department  of  Agriculture 

PREVIOUS  INVESTIGATIONS 

There  is  considerable  evidence  that  young  animals  thrive  on  a  ration 
containing  starch,  but  a  rather  extensive  search  in  the  literature  failed 
to  discover  any  data  concerning  the  question  as  to  how  scon  after  birth 
the  calf  can  begin  to  digest  starch.  The  investigation  here  described  was 
undertaken  for  the  purpose  of  ascertaining  how  early  in  its  life  the  calf 
can  utilize  starch  or  starch -containing  feeds.  The  practical  application, 
of  course,  is  in  supplementing  or  supplanting  the  milk  ration  of  the 
young  calf  with  other  feed. 

The  literature  contains  many  accounts  of  feeding  experiments  with 
young  animals  where  starch  alone  or  as  the  principal  component  of 
some  feed  has  been  used.  The  purposes  of  these  experiments,  however, 
have  been  largely  to  determine  the  effect  of  starch  upon  the  health,  the 
rate  of  gain  in  weight,  the  cost  of  raising,  or  the  digestibility  of  some 
other  component  of  the  ration  rather  than  the  actual  digestibility  of 
the  starch  itself.  The  record  of  but  one  experiment  was  found  in  which 
the  feces  of  young  starch-fed  calves  were  tested  for  the  presence  of 
starch. 

Ewing  and  Wells  ^  report  the  use  of  starch  in  combination  with  corn 
silage  and  cottonseed  meal  in  the  ration  of  12 -month-old  steers  on  di- 
gestion trial.  In  their  summary  they  state  that  when  as  much  as 
47.3  per  cent  of  the  net  energy  of  the  ration  was  supplied  in  the  form 
of  starch  the  iodin  test  did  not  indicate  the  presence  of  starch  in  the 
feces. 

There  are  recorded  in  the  medical  literature  on  the  diet  and  hygiene 
of  children  several  investigations  in  which  the  actual  digestibility  of 
starch  by  children  was  studied.  Kerley,  Campbell,  and  Mason  ^  report 
the  examination  for  starch  of  324  stools,  collected  under  controlled  con- 
ditions at  the  New  York  Infant  Asylum  from  60  children,  all  under  i 
year  of  age,  who  had  been  fed  either  wholly  or  in  part  on  barley  water. 
The  barley  water  was  prepared  by  boiling  raw  barley  flour  for  1^4  hours. 
The  stools  were  examined  for  starch  by  the  Von  Jaksch  iodin  test,  with 


1  Ewing,  Perry  van,  and  Wells,  C.  A.     the  associative  digestibility  of  corn  silage,  cotton- 
seed meal,  and  STARCH  IN  steer  RATIONS.     Ga.  Agt.  Exp.  Sta.  Bui.  115,  p.  269-296,  7diagr.,  1915. 

2  Kerley,  C.  G.,  Campbell,  W^  C,  and  Mason,  H.  N.     a  further  contribution  to  the  study  of 
STOOLS  of  starch-fed  INFANTS.     In  Jour.  Amer.  Med.  Assoc,  v.  47,  no.  10,  p.  763-765-    1906. 


Journal  of  Agricultural  Research.  Vol.  XII.  No.  9 

Washington.  D.  C.  Mar.  4,  1918 


mg 


(575) 


Key  No.  A — 35 


t^y6  Journal  of  Agricultural  Research  voi.xii, no. 9 

Lugol's  solution.  Of  the  60  children,  33  always  gave  negative  iodin 
tests,  indicating  complete  utilization  of  the  starch.  Among  the  re- 
mainder, 8  usually  gave  a  negative  test,  12  usually  gave  a  positive  test, 
and  7  always  a  positive  test  for  starch.  Of  the  41  children  showing  a 
good  capacity  for  starch  utilization  one  19-day -old  child  received  9.2 
gm.  of  barley  flour  daily  for  2  days;  one  21 -day-old  child  received  14.6 
gm.  every  24  hours;  one  child  i  month  and  22  days  old  received  25.9  gm. 
the  first  day  and  25.3  gm.  a  day  for  the  following  four  days;  and  another 
child  I  month  and  19  days  old  received  12  gm.  daily  for  three  days. 

Heubner,*  in  a  paper  presented  before  the  Berlin  Medical  Society, 
describes  an  investigation  conducted  at  Leipzig  for  the  purpose  of 
determining  the  digestibility  of  starch  in  the  food  of  artificially  fed 
children.  The  children  received  during  i-day  and  2-day  periods  a  care- 
fully prepared  starchy  gruel  which  was  fed  in  place  of  milk,  at  the  same 
intervals  and  in  the  same  quantity  as  the  milk  feeding.  Carbon  was 
used  to  identify  the  experimental  stools,  and  the  feces  from  each  child 
were  assembled,  dried,  and  analyzed  for  starch.  A  7-weeks-old  child 
received  24.6  gm.  of  rice  flour  during  a  25-hour  period,  and  no  starch 
was  found  in  the  feces.  Another  child  14  weeks  old  received  53  gm. 
of  rice  flour  during  a  39-hour  period,  and  0.1689  g"^-  o^  starch  was  found 
in  the  feces.  A  third  child  i  year  old  received,  in  addition  to  72  gm. 
of  butter,  133  gm.  of  rice  flour  during  a  48-hour  period,  and  0.2804  g^' 
of  starch  was  found  in  the  feces.  A  fourth  child,  14  weeks  old,  received 
57  gm.  of  a  specially  prepared  oatmeal  during  a  34-hour  period,  and 
0.261 1  gm.  of  starch  was  found  in  the  feces. 

In  an  elaboration  of  the  work  done  by  Heubner  at  Leipzig,  Carstens ' 
gives  the  results  of  digestion  experiments  on  eight  children  from  5  to  14 
weeks  old.  Some  of  these  children  received  starch  from  rice  flour,  some 
from  a  prepared  oatmeal  flour,  and  some  from  two  different  proprietary 
infant  foods.  The  same  methods  were  followed  as  in  the  Heubner 
investigation.  The  quantity  of  undigested  starch  varied  from  a  trace 
in  the  feces  of  two  children,  one  9  weeks  and  one  15  weeks  old,  respec- 
tively, to  5.08  gm.,  or  6.23  per  cent  of  the  amount  ingested  by  a  child 
6)4  weeks  old. 

Kriiger  *  who  worked  with  fetal  and  newborn  calves  found  that  the 
ptyalin  is  secreted  in  the  salivary  glands  as  early  as  the  seventh  month 
of  fetal  life,  but  that  while  the  quantity  increases  up  to  birth,  even  at 
that  time  it  is  too  small  to  be  of  any  importance  in  the  digestion  of  food. 

'  Heubner,  O.  ueberdieadsnutzung  des  mehls  im  darm  jungersaugunge.  /ra  Berlin.  Klin. 
Wchnschr.,  Bd.  32,  No.  10,  p.  201-204,  1S95.    Literatur,  p.  204. 

'Carstens,  J.  H.  weitere  erfahrungen  iber  die  ausnutzung  des  mehls  im  darme  junger 
Saugunge.    In  Verhandl.  Gesell.  Kinderheilk.,  Bd.  12,  p.  169-176.    1895. 

'  Kruger,  Friedrich.  die  verdauungsfermente  bhim  embryo  und  neugeborenen.  80  p. 
Wiesbaden,  1891.     Literatur,  p.  79-80. 


Mar.  4, 1918  Digestion  of  Starch  by  the  Young  Calf  577 

EXPERIMENTAL  WORK 

Two  male  calves,  each  4  days  old,  were  selected.  Each  was  fed 
5.44  kgm.  of  whole  milk  a  day  in  two  feedings.  Beginning  at  4  days  of 
age,  each  calf  received  40  gm.  of  ordinary  cornstarch  per  feeding,  in 
addition  to  the  milk,  for  a  period  of  three  days.  The  starch  ration  was 
prepared  as  follows:  The  weighed  quantity  of  cornstarch  was  placed 
in  a  pail  and  mixed  with  a  little  milk,  then  the  bulk  of  the  milk  was 
added  and  the  mixture  well  stirred.  The  calf  consumed  the  mixture 
with  eagerness  and  without  any  apparent  digestive  disturbance.  To 
make  sure  that  all  the  starch  was  consumfd,  the  pail  was  rinsed  once 
or  twice  with  milk  and  the  calf  permitted  to  drink  the  rinsings.  The 
starch-feeding  period  was  followed  by  a  rest  period  of  about  five  days, 
during  which  only  whole  milk  was  fed;  then  the  calves  again  received 
starch  in  addition  to  their  whole-milk  ration  exactly  as  in  their  first 
3-day  period.     The   length   of  the   periods   and   the   duration  of  the 


Fig.  I.— Bag  for  receiving  feces  and  harness  for  supporting  it. 

experiment  are  shown  in  Table  I.  The  cornstarch  fed  to  calf  i  con- 
tained, according  to  analysis,  77.02  per  cent  of  pure  starch,  and  that 
fed  calf  2,  76.32  per  cent.  The  feces  were  received  in  a  closely  fitting 
rubber  bag  supported  by  a  harness  as  shown  in  figure  i.  Collections 
were  made  during  the  three  starch-feeding  days  and  the  three  days 
following.  The  feces  were  removed  from  the  bag  each  day  and  imme- 
diately dried  on  the  premises  in  an  electric  oven  at  about  the  temperature 
of  boiling  water.  The  dried  feces,  representing  a  starch-feeding  period 
and  the  three  subsequent  days,  were  united  and  ground  in  a  mill.  The 
analyses  were  made  according  to  the  methods  adopted  by  the  Association 
of  Official  Agricultural  Chemists,  the  malt-diastase  method  being  selected 
for  the  starch  determinations. 

The  feeding  was  conducted  at  the  Bureau  of  Animal  Industry's  Ex- 
periment Farm,  Beltsville,  Md.,  and  the  analytical  work  was  done  at 
the  Dairy  Division  laboratories  in  Washington.  The  results  are  given 
in  Tables  I  and  II. 


578 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  9 


Table  I. — Composition  of  feces  of  calves 


Animal  and  period. 


Calf  i 
I. .. . 
2  .  . . 

3-  •• 

4-  ■  ■ 

Calf  2 
I .  . . 

2«  . . 

3-  •• 

4.  . . 


Age  of 
calf 
when 
feces 
were 
col- 
lected. 


Days. 

4 

12 

20 

30 

39 

4 
14 


Weight 
of  dry 
feces. 


Gin. 
288.  98 
270.  28 
244.  87 
179.  40 
202.  00 

283.  49 


190.  21 
152-95 


Nitrogen  in  feces. 


Gm. 
14.07 

13-74 
14.  20 
11.66 
17-25 

14.27 


13-54 
10.  28 


Per  cent. 
4.87 
5.08 
5.80 
6.  50 
8.54 

5-03 


7.  12 
6.  72 


Ether  extract  infeces. 


Starch  in  feces 


Gin. 
6.36 

Per  cent. 
2.  20 

Gm. 
144.2 

"•95 

13-75 

7-36 

14.71 

4.42 
5.62 
4.  10 
7.28 

101.  5 

67.8 

IS-  2 

2.  2 

12.08 

4.  26 

146.  0 

24.  25 
17-58 

12.75 
11.49 

6-7 
1.6 

Per  ci. 
49.91 
37-56 
27.  70 

8.49 
I.  07 

51.48 


3-53 
1.03 


a  Sample  was  lost  at  the  farm. 
Table  II. — Proportion  of  starch  digested  by  calves 


Animal  and  period. 

Starch  fed. 

Starch  in 
feces. 

Starch  digested. 

Calf  i: 

I 

Gm. 
184.9 
184.9 
184.9 
184.9 
184.  9 

183.2 
183.2 
183.2 
183.2 

Gm. 

144.2 

loi.  5 

67.8 

15-2 
2.  2 

146.  0 

Gm. 

40.7 

83-4 

117.  I 

169.7 

182.7 

37-2 

Per  cent. 
22.  02 

2 

45-  II 

2 

63.  34 

A     .                        .                                         

9*1.  79 
98.81 

20.30 

e                                                            

Calf  2: 

I 

2  a 

•1 

6-7 
1.6 

176.5 
181.  6 

96.  32 

A...             .      .          .                                            

99.10 

<»  Sample  was  lost  at  the  farm. 

CONCLUSIONS 

The  figures  in  Table  II  for  digested  starch  show  that  the  calves  when 
from  4  to  7  days  old  were  able  to  digest  about  one-fifth  of  the  quantity 
consumed;  in  one  case  22.02  per  cent  and  in  the  other  20.30  per  cent. 
When  calf  i  was  12  to  15  days  old,  the  percentage  of  starch  digested 
had  more  than  doubled  and  when  3  weeks  old  it  had  nearly  tripled, 
while  at  4  weeks  in  the  case  of  calf  i  and  at  3  weeks  in  calf  2,  the  per- 
centage of  starch  digested  was  well  over  90. 

While  it  is  quite  probable  that  a  calf  but  a  few  hours  old  can  not 
digest  an  appreciable  amount  of  starch,  it  can  readily  be  seen  that  the 
quantity  of  starch-splitting  enzyms  must  increase  very  rapidly  in  the 
first  few  days  of  life,  for  the  calves  under  experiment,  when  only  3  to  4 
weeks  old,  were  able  to  digest  a  ration  nearly  10  per  cent  of  the  dry 
matter  of  which  was  starch. 

These  results  indicate  that  the  milk  ration  of  a  calf  but  a  few  days 
old  may  be  supplemented  with  a  starchy  food  and  that  the  starchy 
material  may  be  rapidh'  increased  as  the  calf  grows  older. 


TOXICITY  OF  VOLATILE  ORGANIC  COMPOUNDS  tO 

INSECT  EGGS^ 

By  William  Moore,  Head  of  Section  of  Research  in  Economic  Zoology,  and  Samuel  A. 
Graham,  Assistant  in  Entomology,  Minnesota  Agricultural  Experiment  Station 

INTRODUCTION 

A  general  sur\"ey  of  the  literature  has  failed  to  reveal  any  extensive 
study  of  the  toxicity  of  different  materials  to  insect  eggs.  Certain  spray 
solutions  have  been  studied,  but  they  have  been  considered  individually 
and  not  in  comparison  with  other  related  compounds. 

Cooley  (j)^,  working  with  the  oyster-shell  scale  {Lepidosaphes  tdmi  L.) 
has  shown  that  linseed  oil,  cottonseed  oil,  and  lime-sulphur  were  effective. 
In  the  cases  of  the  oils  some  of  the  eggs  were  killed,  while  other  eggs 
hatched,  but  the  young  insects  died  during  or  shortly  after  emergence. 
Lime-sulphur  failed  to  kill  the  eggs,  but  the  young  were  killed  very  soon 
after  hatching.     Pure  kerosene  apparently  had  no  effect  on  the  eggs. 

Gillette  (2),  working  with  aphid  eggs  of  different  species,  has  shown 
that  eggs  treated  with  kerosene  emulsion  containing  less  than  25  per 
cent  of  keresene  were  unaffected  by  the  spray.  Scalecide,  Thompson's 
Soluble  Oil,  lime-sulphur,  and  different  soaps  had  little  effect  unless  used 
in  very  large  doses.  Tobacco  extracts  containing  nicotine  or  nicotine 
sulphate  were  found  to  be  very  effective. 

Safro  (9),  in  the  study  of  lime-sulphur  as  an  ovicide  for  the  codling 
moth  (Carpocaspa  pomenella  L.),  shows  that  this  material  is  only  effective 
to  eggs  in  which  the  embryo  is  almost  fully  developed. 

Woodworth  (lo)  has  studied  the  toxicity  of  hydrocyanic-acid  gas 
to  the  eggs  of  scale  insects,  but  does  not  consider  the  factor  of  age. 

Recently  particular  attention  has  been  given  to  the  study  of  the 
toxicity  of  volatile  organic  compounds  to  the  eggs  of  lice  (Pedictdus 
capitis  and  P.  corporis,  and  Phthiris  pubis).  Kerosene  has  been  used 
against  head  lice  and  their  eggs  for  many  years,  but  in  the  recent  work 
many  other  materials  have  been  recommended. 

Postnikov  (7)  recommends  amyl  alcohol,  ethyl  alcohol,  benzene, 
chloroform,  carbon  tetrachlorid,  methane,  and  birch  tar  for  the  destruc- 
tion of  head  lice  and  their  eggs.  Gasoline  has  been  used  for  the  destruc- 
tion of  the  eggs  of  the  clothes  louse,  and  Klinloch  (4)  claims  that  im- 
mersion in  this  material  for  one  minute  will  kill,  while  exposure  to  its 
vapor  is  fatal  in  one-half  hour.     He  considers  that  benzene,   toluene, 

1  Published,  with  the  approval  of  the  Director,  as  Paper  88  of  the  Journal  Series  of  the  Minnesota  Agri- 
cultural Experiment  Station. 
*  Reference  is  made  by  number  (italic)  to  "Literature  cited,"  p.  586-587. 

Journal  of  Agrictiltural  Research,  Vol.  XII,  No.  9 

Washington,  D.  C.  Mar.  4,  1918 

™^  (579)  ^^^  ^°'  ^^""i-  — 24 

38321°— 18 5 


580  Journal  of  A  gricultural  Research  voi.  xii,  no.  9 

and  acetone  are  as  toxic  as  gasoline.  Von  Prowazek  (8)  recommends 
xylol  and  ether  for  the  destruction  of  lice  and  their  eggs. 

In  view  of  the  work  of  the  senior  author  (5,  6)  showing  that  the  toxicity 
of  organic  compounds  to  insects  is  related  to  their  volatility,  of  which 
the  boiling  point  is  a  general  index,  it  was  thought  advisable  to  make 
a  similar  study  of  the  toxicity  of  a  series  of  volatile  organic  compounds 

to  insect  eggs. 

METHOD  OF  EXPERIMENTATION 

For  these  experiments  it  was  considered  desirable  to  use  eggs  which 
were  not  protected  from  the  action  of  the  chemical  by  any  covering. 
Eggs  of  the  bedbug  and  the  clothes  louse  were  considered;  but,  owing 
to  the  fact  that  many  of  these  eggs  were  found  to  be  infertile  under 
the  artificial  conditions  of  breeding,  and,  further,  that  they  were  hard 
to  obtain  in  large  enough  quantities  for  the  purpose  of  the  experiments, 
they  were  discarded.  Potato-beetle  eggs  (Leptinotarsa  decemlineata 
Say)  were  finally  decided  upon  as  fulfilling  all  requirements.  As  many 
as  50,000  eggs  were  used  in  these  experiments,  and  it  was  found  that 
in  every  case  untreated  eggs  hatched  100  per  cent.  They  were  also 
convenient  to  use,  as  a  cluster  of  20  to  30  or  more  eggs  could  be  easily 
treated  as  a  unit.  The  effects  of  the  chemicals  were  studied  in  three 
different  ways: 

1.  By  dipping  the  clusters  in  the  chemical  to  be  tested. 

2.  By  spraying  the  clusters  with  the  chemical  by  means  of  an  atomizer. 

3.  By  exposing  the  eggs  to  the  action  of  the  vapor  of  the  chemical. 
In  the  exposure  to  the  vapor  the  eggs  were  fumigated  in  a  similar 

manner  to  that  employed  in  the  study  of  the  toxicity  of  the  vapor  to 
houseflies  recorded  in  a  previous  paper  (6).  It  was  found  necessary, 
however,  to  use  a  longer  time  limit  than  400  minutes,  15  hours  being 
finally  selected  for  this  purpose. 

Eggs  were  fumigated  for  15  hours  with  varying  quantities  of  the 
chemical,  after  which  they  were  removed  from  the  flask  and  placed  in 
open  pasteboard  pill  boxes  until  they  hatched  or  were  undoubtedly 
dead.  The  smallest  dose  necessary  to  kill  the  eggs  in  this  length  of 
time  was  thus  determined  and  reduced  to  millionths  of  a  gram-molecule, 
making  possible  an  accurate  comparison  of  the  different  chemicals  used. 
In  experiments  where  the  eggs  were  dipped  or  sprayed  they  were  placed 
in  open  pill  boxes  after  treatment  and  handled  in  a  manner  similar  to 
those  fumigated. 

RESULTS  OF  THE  EXPERIMENTS 

The  results  of  dipping  and  spraying  the  eggs  are  given  in  Table  I. 
The  compounds  used  are  arranged  in  the  order  of  their  boiling  points, 
from  the  lowest  to  the  highest.  It  will  be  noted  that  in  general  the 
eggs  treated  with  compounds  having  the  lowest  boiling  point — that  is, 
the  most  volatile  compounds,  permitted  most,  if  not  all,  of  the  eggs  to 
hatch.     Exceptions    may    be    noted    of    compounds    extremely    active 


Mar.  4. 1918      Toxicity  of  Organic  Compounds  to  Insect  Eggs 


581 


chemically,  such  as  allyl  alcohol,  which  contained  ammonia  as  an  im- 
purity, and  chlorpicrin.  In  general  those  sprayed  showed  a  higher 
percentage  of  hatching  than  those  dipped.  Some  compounds  of  this 
series,  the  vapor  of  which  had  been  previously  shown  (6)  to  be  non- 
toxic to  houseflies,  owing  to  the  fact  that  they  formed  gummy  masses 
on  exposure  to  the  air,  were  found  to  be  toxic  to  the  insect  eggs.  Pinene, 
terpineol,  and  geranyl  acetate  are  examples  of  such  chemicals.  Kurther, 
it  is  noted  that  compounds  which  are  so  slightly  volatile,  owing  to 
their  high  boiling  point,  that  they  were  ineffective  against  flies,  were 
found  to  be  toxic  to  the  insect  eggs.  Such  chemicals  were  eugenol, 
alpha  naphthol,  ethyl  ether,  and  trimethylene  cyanid.  These  com- 
pounds are  only  effective,  however,  when  brought  into  actual  contact 
with  the  eggs,  as  in  spraying  or  dipping,  and  are  no  more  effective  as  a 
fumigant  against  the  eggs  than  they  are  against  adult  insects. 

Table  I. — Relation  of  the  boiling  point  to  the  toxicity  of  organic  compounds  used  in 
dips  and  sprays  for  potato-beetle  eggs 


Organic  compound. 


Ethyl  ether 

Ethyl  mercaptan .  . . 
Carbon  bisulphid .  . . 
Petroleum  ether .  . . . 

Acetone 

Chloroform 

Methyl  alcohol 

Carbon  tetrachlorid . 

Ethyl  alcohol 

Gasoline 

Benzene 

Thiophene 

Allyl  alcohol 

Amyl  nitrite 

Nitromethane 

Propyl  acetate 

Toluene 

Chlorpicrin 

Pyridin 

Acetic  acid 

Chlorbenzene 

Amyl  alcohol 

Xylene 

Amyl  acetate 

Bromoform 

Pinene 

Ethyl  malonate 

Allyl  isosulphocyan 

ate 

Furfural 

Butyric  acid 


"C. 

35 

36.2 
46 
40-70 

56-3 
61 

66.  s 
78.1 
78.4 
70-90 
80.3 
84 

97 
98 

lOI 

IC2 

III 

112 

116.  7 

119 

132 

137 

140 

148 

151.  2 

160 

160 

161 
162 
163 


O  v. 


2-oS 


g  bio 


100 
97 


95 
85 


100 

44 
100 
100 


.S-d 
•S  >. 

CO   u 


100 
ICO 

100 

100 
100 

ICO 

100 
100 


100 

100 

100 

0 

100 

0 

0 

16 

100 

0 

0 

0 

20 

74 

100 

ICO 

20 

Organic  compound. 


Trimethylene  bromid 

Terpineol 

Benzonitrile 

Thiophenol 

Benzaldeliyde 

Anilin 

Ortho-bromtoluene . . . 

Valeric  acid 

Ortho-creosol 

lodobenzene 

Salicylic  aldehyde. .  . 

Para-cresol 

Meta-cresol 

Nitrobenzene 

Benzyl  alcohol 

Kerosene 

Ortho-nitrotuluene.  .. 

Bromxylene 

Citral 

Quinolin 

Eugenol 

Nitroxylene 

Nicotine 

a-Napthol  ethyl 

ether 

Trimethylene  cyanid 

Geranyl  acetate 

Brbmmethylphenyl- 

ketone 

Ethyl  aceto-acetate.  . 


°C. 
165 
168 
170 

172.5 
179.  I 
182 
182 

184.5 
190 

193 
196 
201.8 
202.  8 
205 
206.  5 
150-300 
223 
225 
225 
239 
247-5 
250 


274 


Si 

0  »-     . 

i     >. 

bt'2'3 

.a  c3  ce 

C4   O    H 

01  o.  n 
M--  — 

Ssfa 


17 


582 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  9 


Kerosene,  although  having  a  high  boihng  point  of  150''  to  300°  C, 
allowed  83  percent  of  the  dipped  eggs  and  100  percent  of  the  sprayed 
eggs  to  hatch. 

Table  II. — Toxicity  of  various  organic  compounds  to  eggs  of  different  ages  when  dipped 

for  -various  periods 


Compound. 


Ether. 


Carbon  bisulphid. 


Methyl  alcohol. 


Chloroform . 


Carbon  tetrachlorid . 


Benzene . 


Toluene . 


Acetic  Acid. 


Xylene. 


Ortho-bromtoluene . 


Age. 


Fully       developed 

embryos 

Partially  developed 
Freshly  laid — . ... 


'Fully  developed.  . .  . 
Partially  developed 
.Freshly  laid 


Fully  developed . . .  . 
Partially  developed 
Freshly  laid 


'Fully  developed ... 
Partially  developed 
.Freshly  laid 


I  Fully  developed ... 
Partially  developed 
Freshly  laid 


Fully  developed.  . . 
Partially  developed 
Freshly  laid 


Fully  developed. . 
Partially  developed . 
Freshly  laid 


Fully  developed. . . 
Partially  developed 
Freshly  laid 


Fully  developed 

Partially  developed 
Freshly  laid 


Fully  developed. . . 
Partially  developed 
Freshly  laid 


Percentage  hatching  after  dippmg  for- 


I  sec- 
ond. 


100 
100 


IOC 

100 
100 

100 
100 
100 


100 
100 
100 

100 
100 
100 


100 
100 
100 


S  sec- 
ond. 


10  sec- 
onds. 


100 
IOC 
ICO 

100 
100 
100 

3 
100 


100 
100 
100 


o 
100 
100 


100 
100 


o 
100 

100 
100 
100 

100 
100 


15  sec- 
onds. 


100 
100 


100 
100 


30  sec- 
onds. 


o 

100 


100 
100 

100 

o 

100 


100 
100 


60  sec- 
onds 


120 
sec- 
onds. 


100 
100 


5 
100 
100 


100 
100 

100 


100 
O 


In  the  experiments  it  v^as  noted  that  compounds  with  low  boiling 
points  evaporated  from  the  surface  of  the  eggs  very  quickly,  this  giving 
the  material  very  little  time  to  penetrate.  On  the  other  hand,  com- 
pounds with  high  boiling  points  remained  upon  the  eggs  for  hours,  or 
even  days.  In  view  of  this  fact,  a  series  of  experiments  was  conducted 
with  the  lower  boiling-point  compounds,  in  which  the  eggs  were  dipped 


Mar.  4, 1918      Toxicity  of  Organic  Compounds  to  Insect  Eggs 


583 


for  periods  from  i  to  120  seconds,  eggs  of  different  ages  being  used. 
The  results  are  given  in  Table  II.  This  table  shows  that  in  general  with 
compounds  of  very  low  boiling  points  the  freshly  laid  eggs  or  those  con- 
taining embryos  only  partially  developed  were  more  easily  killed  than 
those  in  which  the  embryo  was  fully  developed.  Gortner  and  Banta  (5), 
in  working  with  the  toxicity  of  certain  phenolic  compounds  to  amphibian 
eggs,  found  that  the  youngest  eggs  were  more  susceptible  than  the  older 
eggs.  This  may  be  due  to  disturbances  in  the  permeability  of  the  egg, 
and  compounds  with  low  boiling  points  such  as  ether  would  have  more 
influence  on  the  permeability  than  compounds  with  higher  boiling  points. 
With  an  increase  in  the  boiling  point  it  was  found  that  the  eggs  in  which 
the  embryo  was  fully  developed  were  most  easily  killed.  Compounds 
with  very  high  boiling  points,  such  as  nicotine  and  kerosene,  often 
remained  on  the  eggs  and  killed  the  larvae  in  the  act  of  hatching.  It 
seems  that  compounds  with  high  boiling  points  are  not  able  to  penetrate 
the  egg  as  readily  as  compounds  having  low  boiling  points.  In  general  a 
slightly  longer  exposure  to  the  chemical  will  result  in  the  death  of  a  larger 
number  of  eggs. 

In  these  experiments  the  writers  they  have  been  unable  to  remove 
compounds  of  high  boiling  points  from  the  surface  of  the  egg  without 
injury  to  the  egg,  and,  hence,  have  no  data  as  to  the  length  of  time 
necessary  to  kill  with  these  compounds. 

Table  III. — Toxicity  of  kerosene  to  eggs  of  various  ages 


Age  of  eggs. 


1  day 

2  days 

3  days 

4  days 

Freshly  laid 

Slightly  developed 

Well  developed 

Spiracles  visible  through  shell . 

Within  I  day  of  hatching 

Within  2  days  of  hatching 

Within  3  days  of  hatching 

Within  4  days  of  hatching 


100 


100 
100 

33 
o 
100 
100 
100 
100 


Percentage  hatching  after  dipping  for- 


I  min.     sniin.     loxnin.   is  jnin.i  aojnin.  Brand 


13 
100 
100 


100 
100 
100 
100 
100 
o 
100 


o 

100 
100 
100 
100 
o 


100 
100 


Unknown. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 

Do. 
Pearl  oil. 

Do. 

Do. 

Do. 


A  special  study  was  made  of  the  action  of  kerosene  on  the  eggs,  inas- 
much as  it  was  a  high-boiling-point  compound  which  did  not  always 
kill.  The  results  of  dipping  eggs  of  different  ages  in  kerosene  are  given 
in  Table  III.  Eggs  i  day  old  in  general  did  not  hatch  after  being  dipped. 
Eggs  classed  as  freshly  laid,  which  may  have  been  i  or  2  days  old,  hatched 
in  every  case,  while  eggs  with  well  developed  or  fully  developed  embr3'os 
failed  to  hatch.     This  is  interesting  in  that  by  killing  the  young  eggs  it 


584 


Journal  of  A  gricultural  Research 


Vol.  XII,  No.  9 


acts  in  a  similar  manner  to  compounds  with  a  low  boiling  point,  while  by 
killing  the  eggs  with  fully  developed  embryos  it  acts  as  a  compound  hav- 
ing a  high  boiling  point.  The  Pearl  oil  used  in  the  experiments  is  a  prod- 
uct from  the  California  oil  fields  and  therefore  differs  chemically  from  the 
other  kerosenes  used. 

In  order  to  obtain  further  data  on  the  toxicity  of  kerosene  to  eggs,  two 
brands  of  kerosene  were  broken  up  into  fractions  of  different  boiling 
points.  The  two  brands  used  were  Pennsylvania  Best  sold  by  the  Pure 
Oil  Co.  and  Perfection  oil  put  out  by  the  Standard  Oil  Co.  of  Indiana. 
The  data  obtained  by  this  experiment  are  shown  in  Table  IV.  The 
fraction  of  Pennsylvania  Best  boiling  between  140°  and  187°  C.  killed 
only  freshly  laid  eggs.  The  fraction  between  187°  and  234°  C.  killed 
both  the  fully  developed  and  freshly  laid  eggs,  while  the  fraction  between 
234°  and  280°  C.  killed  all  the  eggs.  Apparently  the  lower  fraction  did 
not  remain  long  enough  upon  the  fully  developed  eggs  to  produce  any 
influence.  It  was  noted  in  the  kerosene  experiments  that  the  higher 
fractions  killed  at  the  time  of  hatching.  This  may  explain  the  fact  that 
the  partially  developed  eggs  treated  with  the  second  fraction  allowed 
100  per  cent  to  hatch.  These  eggs  being  too  far  developed  were  not 
effected  by  changes  in  permeability  such  as  would  kill  a  freshly  laid  egg, 
while  the  compound  had  sufficient  time  to  evaporate  from  the  surface  of 
the  egg  before  the  emergence  of  the  larva. 

Table  IV. — Toxicity  of  different  fractions  of  kerosene  on  eggs  of  different  ages 


Brand. 


Pennsylvania  Best. . 


Age. 


Fully  developed ... 
Partially  developed 

Freshly  laid 

General 

do 


Percentage  hatching  after  dippinur  for  5  seconds  ia 
fractions  with  boiUng  points  of — 


140  -187 
C. 


100 
100 
o 
100 
100 


187  -234 
c. 


100 

o 

100 

100 


234  -2 So 
C. 


O 

o 

o 

100 


Residue 
above 

280°  C. 


Undis- 
tiUed. 


O 
O 

100 

50 


Brand. 


Age. 


Percentage  hatching  after  dipping  for  s 
seconds  in  fractions  with  boiling 
points  of — 


i35°-i83°  C. 


i83°-23i''  C. 


23i°-28o''  C. 


Perfection 


Fully  developed . . . . 
Partially  developed 
Freshly  laid 


Mar.  4.  i9i8      Toxicity  of  Organic- Compouuds  to  hisect  Eggs 


585 


The  Perfection  oil  killed  both  the  fully  developed  and  freshly  laid 
eggs,  leaving  the  partially  developed  uninjured  except  for  the  fraction 
between  231°  and  280°  C.  Other  experiments  with  these  two  oils,  which 
will  be  published  later,  have  shown  that  the  Perfection  oil  is  more  toxic 
than  the  Pennsylvania  Best.  Kerosene  therefore  has  oils  of  low  boiling 
points  which  kill  the  freshly  developed  eggs  and  oils  of  high  boiling  points 
which  destroy  the  fully  developed  eggs,  while  the  partially  developed 
eggs  are  most  likely  to  hatch. 

Nineteen  compounds  were  tested  to  determine  the  toxicity  of  their 
vapor  to  insect  eggs.  The  results  of  these  experiments  are  given  in  Table 
V,  in  which  the  compounds  are  arranged  according  to  their  boiling 
points,  from  the  lowest  to  the  highest.  The  increase  in  toxicity  with  an 
increase  in  boiling  point  is  indicated  by  the  smaller  portion  of  a  gram- 
molecule  necessary  to  kill  in  15  hours.  That  the  boiling  point  is  not  so 
good  an  index  of  the  toxicity  of  a  chemical  as  its  volatility  is  shown  in 
Table  VI,  where  the  compounds  are  arranged  according  to  their  volatility, 
from  the  most  volatile  to  the  least  volatile.  A  comparison  of  the  two 
tables  shows  a  better  correlation  between  volatility  and  toxicity  than 
between  boiling  point  and  toxicity.  As  was  found  in  the  work  with  the 
housefly  (5),  chemicals  having  low  boiling  points  are  more  valuable 
for  fumigation  purposes  due  to  the  fact  that  a  larger  amount  of  the 
vapor  may  be  contained  in  the  air. 

Table  V. — Toxicity  of  the  vapor  of  certain  organic  compounds  (in  millionths  of  a  gram- 
molecule  required  to  kill  in  75  hours'  exposure).  Compoundf  arranged  according  to 
their  boiling  points 


Compound. 


Ethyl  ether 

Carbon  bisulphid .  . 
Petroleum  ether. .  . 

Acetone 

Chloroform 

Methyl  alcohol .... 
Carbon  tetrachlorid 

Ethyl  alcohol 

Gasoline 

Thiophene 


Boiling 
point. 


°C. 

35 
46 
40-70 

56.3 

61 

66.5 

78.1 
78.4 
70-90 
84 


Toxicity 

(mil- 
lionths of 
a  gram- 
molecule). 


3,  468.  2 

555-5 
243.  2 

558-0 

I,  232.  o 

404.  o 

284.  o 

41. 1 

344- o 


Compound. 


Toluene 

Chlorpicrin . . 
Chlorbenzene 

Xylene 

Bromoform .  . 

Furfural 

Nitrobenzene 
Nitroxylene . . 
Nicotine 


Boiling 
point. 


°C. 
Ill 
112 
132 
140 

151- 
162 
205 
250 
250 


Toxicity 

(mil- 
lionths of 
a  gram- 
molecule). 


104.  O 
2.8 

160.8 

29-5 

3-9 

1.8 

2.7 

3-3 
1.8 


586 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  9 


Table  VI. — Toxicity  of  the  vapor  of  certain  organic  compounds  {in  millionihs  df  a  gram- 
molecule  required  to  kill  in  15  hours'  exposure).  Compounds  arranged  from  the  m-ost 
volatile  to  the  least 


Compound. 

Toxicity 
(millionth  s 
of  a  gram- 
molecule). 

ComiX)und. 

Toxicity 
(millionths 
of  a  gram- 
molecule). 

Kthyl  ether 

3,  468.  2 

243.  2 

1,232.  0 

I;  363-0 

555-0 

558.0 

404.  0 

284.  0 

2.8 

344- 0 

Chlorbenzeiie 

160  8 

Petroleum  etlier 

Toluene 

104.  0 

29-5 
41.  I 

3-9 
I  8 

Methyl  alcohol 

Acetone 

Xylene 

Gasoline 

Carbon  bisulphid 

Chloroform 

Bromoform 

Furfural 

Carbon  tetrachlorid 

Nitrobenzene  .   .   . 

2.7 
1.8 

Ethyl  alcohol 

Nicotine 

Nitroxylene 

Chlorpicrin 

Thiophene 

3-3 

Fumigation  in  a  saturated  atmosphere  with  ether,  ethyl  mercaptan, 
carbon  bisulphid,  benzene,  carbon  tetrachlorid,  and  chloroform  will  kill 
all  the  eggs  in  one  hour. 

SUMMARY 

(i)  In  general,  compounds  with  high  boiling  point  and  slight  volatility 
are  more  eiffective  in  dipping  and  spraying  insect  eggs  than  compounds 
with  low  boiling  point  and  high  volatility. 

(2)  Compounds  with  low  boiling  points  kill  freshly  laid  eggs  more 
readily  than  eggs  in  which  the  embryo  is  partially  or  fully  developed. 

(3)  Compounds  of  higher  boiling  points  are  more  toxic  to  eggs  with 
fully  developed  embryos  than  they  are  to  eggs  in  which  the  embryo  is 
only  slightly  formed. 

(4)  Kerosene  containing  both  high  and  low  boiling  points  is  destruc- 
tive to  both  young  and  old,  but  is  only  slightly  toxic  to  partially  devel- 
oped eggs. 

(5)  The  toxicity  of  the  vapor  of  organic  compounds  to  insect  eggs  is 
related  to  the  boiling  point  and  the  volatility.  As  the  boiling  point 
increases  and  the  volatility  decreases,  the  toxicity  increases. 

LITERATURE  CITED 

(1)  COOLEY,  R.  A. 

1910.    NOTES    ON    SPRAYING    EXPERIMENTS    FOR    THE    OYSTER    SHELL    SCALE    IN 

MONTANA.     In  Jour.  Econ.  Ent.,  v.  3,  no.  i,  p.  57-64. 

(2)  Gillette,  C.  P. 

I910.   SOME     insecticide     tests     for     the     destruction     op     APHIDIDAE     AND 

THEIR  EGGS.     In  Jour.  Econ.  Ent.,  v.  3,  no.  2,  p.  207-210. 

(3)  GoRTNER,  R.  A.,  and  Banta,  A.  M. 

1914.  notes  ON  the  toxicity  of  dilute  solutions  of  CERTAIN  PHENOLIC 
COMPOUNDS  AS  INDICATED  BY  THEIR  EFFECT  ON  AMPHIBIAN  EGGS  AND 
EMBRYOS,  TOGETHER  WITH  REFERENCES  ON  MODIFICATIONS  OP  PIG- 
MENT DEVELOPMENT.  In  Biochem.  Bui.,  v.  3,  no.  11/12,  p.  357-368. 
Literature  cited,  p.  368. 


Mar.  4,  i9i8      Toxicity  of  Organic  CompouTids  to  Insect  Eggs  587 

(4)  KiNLOCH,  J.  p. 

191 5.    AN    INVESTIGATION    O?    THE    BEST    METHODS    OF    DESTROYING    LICE    AND 

OTHER  BODY  VERMIN.     In  Brit.  Med.  Jour.,  1915,  no.  2842,  p.  1038- 
1041. 

(5)  Moore,  William. 

I917.   toxicity  OV  various  BENZINE  DERIVATIVES  TO  INSECTS.      In  JotlT.  Agr. 

Research,  v.  9,  no.  11,  p.  371-381,  4  fig.     Literature  cited,  p.  380-381. 

(6)  


1917.    VOLATILITY   OP  ORGANIC  COMPOUNDS   AS    AN   INDEX   OP   THE  TOXICITY   OP 

THEIR  VAPORS  TO  INSECTS.  In  JoxiT.  Agr.  Research,  v.  10,  no.  7,«p. 
365-371.  7  fig- 

(7)  POSTNIKOV,  A.  I. 

1915.   ON    THE    QUESTION    OF    THE    dbNTROL    OP    LICE    IN    THE    ACTIVE    ARMY. 

(Abstract.)  In  Rev.  Appl.  Ent.,  s.  B,  v.  3,  pt.  8,  p.  122-123.  1913. 
(Original  article  in  Proc.  Conf .  Bacteriologists  and  Representatives  of 
Medical-Sanitary  Authorities  on  the  Campaign  against  Infectious  Dis- 
eases in  Connection  with  the  War.     p.  70-71.     Moscow.     Not  seen.) 

(8)  Prowazek,  S.  von. 

19x5.   BEMERKUNGEN    UBER    DIE    BIOLOGIE    UND    BEKAMPFUNG    DER    KLEIDER- 

LAUS.     In  Mtinchen.  Med.  Wchnschr.,  Jahrg.  62,  No.  2,  p.  67-68. 

(9)  Safro,  V.  I. 

1912.   LIME-SULPHUR   WASH   AN   INEFFIOENT   OVICIDE   FOR   CODLING    MOTH.      In 

Jour.  Econ.  Ent.,  v.  5.  no.  5,  p.  385-395. 

(10)  WOODWORTH,  C.  W. 

1915.  TOXICITY  OF  INSECTICIDES.     In  Science,  v.  41,  no.  1053,  p.  367-369. 


CORN-STOVER  SILAGE 

By  J.  M.  Sherman,  Bacteriologist,  and  S.  I.  Bechdel,  Assistant  Dairy  Husbandman, 
Pennsylvania  Agricultural  Experiment  Station 

INTRODUCTION 

The  ensiling  of  com  stover  and  cured  com  fodder  is  not  a  new  idea. 
In  a  few  localities  farmers  have  followed  this  practice  to  a  limited  extent 
for  some  years,  and  there  have  appeared  several  Experiment  Station 
publications  which  deal  briefly  with  the  process.  Experimental  data 
on  the  subject  are,  however,  very  meager;  little  has  been  established  with 
reference  to  the  practicability  of  stover  silage,  while  nothing  is  known 
concerning  the  nature  of  the  fermentation  which  takes  place  and  the 
factors  operative  other  than  what  may  be  deduced  from  knowledge 
of  ordinary  silage.  The  present  necessity  for  more  economic  produc- 
tion, with  the  conservation  of  concentrates  and  the  utilization  of  more 
roughage  in  the  live-stock  and  dairy  industries,  makes  a  reconsidera- 
tion of  corn-stover  silage  of  especial  pertinence. 

The  object  of  the  present  study  was  partly  to  test  the  practicability 
of  ensiling  stover,  and  partly  to  determine  the  nature  of  the  fermenta- 
tion which  takes  place  therein. 

PRACTICABILITY   OF   ENSILING   CORN   STOVER 

The  stover  used  in  this  experiment  was  ensiled  early  in  April,  191 6; 
the  material  had  been  kept  in  a  shed  since  fall  and  was  quite  dry.  The 
condition  of  the  material  was  not  good;  it  was  moldy  in  spots  and  on 
the  whole  represented  an  inferior  grade  of  stover.  This  stover  was 
run  through  a  silage  cutter  and  packed  in  the  silo  by  means  of  tramp- 
ing. Water  was  added  in  a  continuous  stream  through  a  hose  which 
was  cairied  and  the  water  distributed  by  the  man  who  did  the  packing. 
A  water  meter  was  attached  to  the  hose  so  as  to  enable  the  regulation  of 
the  amount  used.  A  wooden-stave  silo  16  feet  in  diameter  was  filled 
with  32,000  pounds  of  stover  to  which  were  added  66,000  pounds  of 
water. 

vSamples  of  the  silage,  which  were  taken  at  frequent  inter\'als,  were 
examined  for  general  appearance,  texture,  and  aroma.  The  stover 
was  soon  observed  to  undergo  a  fermentation  with  the  formation  of  a 
product  quite  similar  to  normal  silage  made  with  green  com.  The 
material  softened,  regained  a  slightly  greenisli  color,  and  developed  an 
aroma  simulating  that  of  normal  silage,  though  inferior  in  all  these 
respects  to  silage  made  in  the  usual  way  from  green  com. 

Feeding  tests  made  at  the  end  of  the  experiment  showed  that  cattle 
ate  this  silage  with  little  waste  and  apparently  with  a  relish.     While 

Journal  of  Agricultural  Research,  Vol.  XII,  Ko.  9 

Washington,  D.  C.  Mar.  4,  1918 

mi  Key  No.  Pa. — 5 

(589) 


590  Journal  of  A  gricultural  Research  voi.  xn,  No.  9 

it  is  not  believed  that  the  stover  silage  is  as  palatable  as  is  that  made 
from  fresh  com,  it  did  prove  to  be  a  very  acceptable  feed  which  was 
preferred  by  cattle  to  any  of  the  dry  roughages  furnished.  Some  ex- 
periments in  which  the  actual  feeding  value  of  stover  silage  is  deter- 
mined in  comparison  with  ordinary  silage  and  other  roughages  are 
desired,  and  it  is  hoped  that  such  tests  may  be  carried  out  at  this  Station 
in  the  near  future. 

The  keeping  quality  of  stover  silage  appears  to  be  excellent,  pro- 
vided sufficient  water  is  added.  Although,  as  is  the  case  with  other 
types  of  silage,  the  surface  material  undergoes  a  moldy  spoilage 
accompanied  with  heat  formation,  this  condition  does  not  extend  to 
more  than  the  ordinary  depth.  The  silage  made  in  this  experiment 
was  not  all  used  during  the  following  winter,  and  the  remaining  material, 
at  the  time  of  this  writing  nearly  1^2  years  old,  is  still  in  excellent  con- 
dition. In  view  of  the  very  satisfactory  results  obtained  wdth  such 
an  inferior  raw  product,  we  do  not  hesitate  to  predict  success  in  the 
ensiling  of  any  stover  which  is  in  reasonably  good  condition. 

Probably  the  most  important  consideration  for  the  successful  pro- 
duction of  stover  silage  is  the  amount  of  Vv^ater  to  add.  This  obviously 
will  vary  according  to  the  quantity  of  water  contained  in  the  stover,  and 
this  factor  should  be  taken  into  consideration.  While  it  would  be 
more  scientific  and  exact  to  determine  the  most  desirable  amount  of 
water  to  add  by  means  of  moisture  tests  on  the  stover,  such  a  recom- 
mendation would  find  no  place  in  farm  practice.  In  our  experimental 
silo  the  proportion  was  about  2  parts  of  water  to  i  of  stover,  but  the 
stover  was  probably  somewhat  drier  than  would  usually  be  the  case. 
As  may  be  seen  from  the  moisture  determinations,  which  are  reported  in 
another  section  (Table  I),  the  quantity  of  water  added  was  none  too 
much;  in  general  appearance  and  to  the  touch  some  of  the  samples 
seemed  to  be  considerably  below  the  most  desirable  point. 

A  laboratory  test  was  carried  out  upon  this  point  by  making  stover 
silage  in  small  jars  with  varying  quantities  of  water  and  examining 
after  about  one  month  for  general  appearance  and  condition  of  moisture. 
The  stover  used  in  these  tests  was  very  dry.  Samples  which  were 
made  with  equal  parts  by  weight  of  water  and  stover,  as  well  as  those 
made  with  i  }4  parts  of  water  to  i  of  stover  were  too  dry  to  undergo  a 
typical  fermentation  and  form  good  silage.  Those  which  had  water  in 
the  proportion  of  2  parts  to  i  of  stover  made  good  silage,  but  did  not 
appear  to  have  as  much  moisture  as  would  be  best.  Samples  put  up 
with  2)4  and  with  2^  parts  by  weight  of  water  to  i  part  of  stover  were  in 
good  condition  when  opened  and  apparently  did  not  contain  an  excess 
of  water. 

It  seems  that  in  general,  when  reasonably  fresh  stover  is  used,  about 
2  parts  of  water  by  weight  to  i  of  stover  would  be  advisable,  while  for 
older  and  drier  stover  a  slightly  larger  proportion  of  water  may  be  more 


Mar.  4,  i9i8  CoYn-StoveY  SUagc  591 

desirable.  It  should  be  kept  in  mind  that  these  tests  were  all  made 
with  rather  dry  stover.  In  the  case  of  ensiling  soon  after  the  corn  is 
husked,  2  parts  of  water  might  be  too  much.  However,  it  appears 
from  our  observ^ations  that  there  is  less  danger  of  adding  too  much 
water  than  of  getting  too  small  an  amount,  and  that  considerable  water 
may  be  added  above  the  required  amount  without  injury  to  the  product. 
The  water  should  be  added  uniformly  as  the  silo  is  being  filled  so  that 
all  the  dry  cut  stover  becomes  thoroughly  wet  down.  If  this  precau- 
tion is  not  taken,  the  water  may  follow  channels  down  through  the  silage 
and  waste  away  at  the  bottom  of  the  silo.  In  such  an  instance  spoiled 
silage  might  result  in  some  parts  of  the  silo  because  of  an  insufficient 
amount  of  water. 

FERMENTATION  OF  STOVER  SILAGE 

To  obtain  information  on  the  nature  of  the  fermentation  which  takes 
place  in  silage  made  from  stover,  determinations  were  made  of  the  vola- 
tile and  nonvolatile  acids,  temperatures,  and  numbers  and  types  of 
bacteria  at  various  stages  of  the  ripening  process. 

Samples  for  examination  were  obtained  by  means  of  a  2-inch  auger 
provided  with  an  extension  shaft  of  8  feet,  thus  making  it  possible  to 
penetrate  to  the  center  of  the  silo.  By  repeatedly  boring  in  a  short  dis- 
tance and  withdrawing  the  auger  until  the  center  of  the  silo  was  reached, 
no  difficulty  was  experienced  in  securing  sufficient  material  for  the 
different  tests.  The  sample  obtained  in  this  way  represented  the  silage 
mass  from  the  wall  to  the  center  of  the  silo.  Not  more  than  one  boring 
was  made  in  one  place,  the  different  samples  being  removed  at  points 
all  the  way  around  the  silo  and  from  3  to  8  feet  from  the  ground.  The 
material  so  obtained  was  subjected  to  pressure  in  an  ordinary  lard 
press  and  sufficient  juice  collected  for  the  various  examinations  made. 

Acid  formation. — There  can  be  no  doubt  that  the  amount  and  char- 
acter of  acids  in  silage  influence  its  quality  profoundly.  Since  it  has 
been  the  general  experience  that  extremely  green  com  produces  a  very 
sour  silage,  while  com  more  nearly  mature  produces  a  silage  with  less 
acid  and  of  a  much  better  quality,  the  acid  fermentation  in  silage  made 
from  dry  material  is  of  interest. 

The  volatile  acidity  was  determined  by  subjecting  a  loo-gm.  sample 
of  juice  to  steam  distillation  under  reduced  pressure  until  4  liters  of 
distillate  were  secured.  These  were  titrated  directly  after  collection  with 
Njio  barium  hydrate,  with  phenolphthalein  as  the  indicator.  The  non- 
volatile acidity  was  obtained  by  the  difference  between  the  volatile 
acidity  and  a  total  acid  determination  made  by  the  titration  of  20  gm. 
of  juice,  diluted  to  500  c.  c.  with  carbon-dioxid-free  water,  against 
Njio  barium  hydroxid.  In  Table  I  the  nonvolatile  and  volatile  acids 
(calculated  as  lactic  and  acetic,  respectively)  are  reported  in  terms  of 
percentage  of  air-dry  material. 


592 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  9 


Table  I. — Acid  formation  in  stover  silage 


Age. 


Week 

3/7 

1 

2 

3 

4 

5 

6 

8 

10 

12 


Total 
solids." 


Per  cent. 
28.9 

35-  5 

26.  7 
26.8 

27.  I 
29.7 
31.2 

27-5 
32.  I 
25.8 


Acid  in  total  solids. 


Nonvola- 
tile.^ 


Per  cent. 

Trace 
o.  16 

•95 
1.36 

1-54 
1-51 
2.  00 

2-53 
2.  64 

3-15 


Volatile,  c 


Per  cent. 
0.51 
87 
36 
41 
49 
55 
69 
92 
82 
24 


Ratio    of 
nonvolatile 
to  volatile. 


5- 40 

1-43 
I.  04 

•97 
I.  02 

•85 

•7S 
•75 
•71 


u  Air-dried. 


b  Calculated  as  lactic  acid. 


c  Calculated  as  acetic  acid. 


It  is  seen  from  the  figures  in  Table  I  that  for  the  first  week  the  vola- 
tile acidity  was  greatly  in  excess  if  the  nonvolatile  portion ;  that  from  the 
second  to  the  fifth  week  the  two  were  apparently  present  in  about  equal 
amounts;  and  that  from  the  sixth  week  on,  the  nonvolatile  acids  were  in 
excess,  the  proportion  of  nonvolatile  acidity  increasing  to  the  end  of  the 
experiment.  The  total  acidity  obtained  was  somewhat  lower  than  is 
usually  found  in  ordinary  silage.  This  is  probably  to  be  expected,  in 
view  of  the  chemical  differences  in  the  raw  materials.  In  this  con- 
nection, however,  it  should  be  noted  that  in  this  experiment  samples 
were  not  taken  after  the  twelfth  week.  It  is  not  unlikely  that  the  acidity 
increased  some  after  the  last  sample  was  secured.  In  regard  to  the  pro- 
portion of  nonvolatile  to  volatile  acids,  if  we  accept  the  ratio  i  to  0.75 
reported  by  Dox  and  Neidig  (5)^  as  representing  a  general  average  for 
ordinary  silage,  it  will  be  seen  that  our  results  on  com  stover  silage  in- 
dicate a  remarkable  agreement  in  this  respect  between  these  two  types  of 
ensilage. 

Fermentation  temperatures. — Temperature  records  were  obtained 
by  means  of  four  resistance  bulbs  with  about  60  feet  of  insulated  cable 
attached  to  each  which  w^ere  buried  in  the  silage  as  the  silo  was  filled. 
The  bulbs  were  located  as  follows : 

No.  I.  About  2  feet  from  the  bottom  and  in  the  center  of  the  silo. 
No.  2.  About  6  feet  from  the  bottom  and  about  3  feet  from  the  center  of  the  silo. 
No.  3.  At  the  same  height  as  No.  2  but  about  3  feet  from  the  center  in  the  opposite 
direction. 

No.  4.  About  12  feet  from  the  bottom  and  in  the  center  of  the  silo. 

The  ends  of  the  cables  were  located  at  a  convenient  place  on  the  out- 
side of  the  silo  so  as  to  allow  easy  attachment  for  temperature  readings. 
Table  II  gives  the  temperatures  obtained  from  April  4,  the  day  the  silo 
was  filled,  until  June  16. 


•  Reference  is  made  by  number  (italic)  to  "Literature  cited,"  p.  6co. 


Mar.  4,  1918 


Corn-Stover  Silage 


593 


Table  II. — FeTtnentaiion  temperatures  of  stover  silage 


Date. 


Apr,    4 

5 
6 

7 
8 

9 
10 
II 
12 
13 
14 
15 
16 

17 
18 

19 
20 
21 

22 
23 
24 
25 
26 
27 
28 
May  I, 
2, 

4 

6 

8 

16 

22 

31 
June  16. 


Temperature  CF.). 


Bulb  I.  Bulb  2 


42.  o 

42.5 
46.5 

46.  5 
47.0 

47-5 
49.0 
49.0 

49-5 
50.0 

50-5 
50.0 
50.0 
50.0 
50.0 
50.0 
50.0 
50.  o 
49-  5 
49-5 
50.0 
50.  o 
49-5 
49-5 
50.  o 

49-5 
49.0 
50.0 
50.0 
50.0 
50-5 
50-5 
51.0 
52.0 


42 
49. 

SC- 
52- 
54- 
54- 
54- 
54 
56- 
56. 
56. 
56. 
56- 
56. 
57- 
57- 
57- 
57- 
57- 
57- 
57- 
57- 
57- 
57- 
57- 
57- 
58- 
58- 
58. 
58. 
58. 
58. 
58. 


Bulb  3. 


.  0 

42. 

•5 

47- 

-5 

48. 

.  0 

48. 

-5 

49. 

.  0 

51- 

.0 

51- 

•5 

51- 

5 

52- 

.  0 

52- 

.  0 

53- 

.  0 

53- 

.  0 

53- 

.  0 

53- 

•5 

54- 

.0 

54- 

.0 

54- 

.  0 

54- 

.  0 

54- 

•5 

55- 

-5 

55- 

•5 

55- 

•5 

55- 

-5 

55- 

•5 

55- 

.  0 

55- 

-5 

55- 

•5 

50. 

•5 

5^ 

.  0 

5f 

.  0 

sf 

.  0 

50 

.  0 

57 

•5 

57 

Bulb  4. 


46. 

47. 
48. 
49. 
SO' 

SI' 

52' 
52' 

55' 

55' 

55' 

55' 

56. 

56.0 

56.0 

55- o 
56.  o 

57- o 
57- o 
57-0 
57-5 
57-5 
57 


58.  5 

59- 

59- 

59- 

59- 

59- 

59- 

61. 

63- 


Atmos- 
phere. 


33' 
42. 

43' 
46, 

33 

33 

37 

42 

48.0 

58.0 

51.0 

49.0 

50.  o 

52.0 

49-5 
46.0 

55-0 
49 
49 
41, 

47 
47 
48, 

48.  5 
47.0 
66.5 
63-5 
53-5 
60.5 

65-5 
65.0 

55- o 
64.  o 

63-5 


Recent  investigations  have  furnished  abundant  proof  that  high  tem- 
peratures are  not  essential  in  silage  preservation,  and,  in  fact,  do  not 
occur  except  at  the  surface,  which  undergoes  an  aerobic  decomposition. 
Bechdel  (3)  has  recorded  an  instance  in  which  the  maximum  temperature 
attained  in  the  center  of  a  concrete  silo  during  the  curing  period  was  only 
60°  F.  As  has  been  shown  by  Eckles,  Oshel,  and  Magruder  (6),  the 
atmospheric  temperature  at  time  of  filling  influences  greatly  the  tem- 
perature attained  during  the  fermentation  of  the  silage. 

Table  II  shows  that  the  temperature  at  the  start  was  42°  F.  and  gradu- 
ally increased  until  the  readings  were  discontinued.  The  maximum 
temperature  attained  was  63°  F.  in  the  case  of  bulb  4;  but  bulb  i,  which 
was  buried  to  a  depth  slightly  below  the  surface  of  the  surrounding 
soil,  showed  a  maximum  temperature  of  only  52°  F.  An  examination 
of  the  column  giving  the  atmospheric  temperature  during  this  period 
suggests  that  the  continued  increase  in  the  silage  temperature  during 
the  latter  part  of  the  time  may  be  accounted  for  by  a  similar  increase  in 


594 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  9 


the  outside  air.  However,  the  more  rapid  increase  during  the  early 
part  of  the  period  was  entirely  independent  of  this  factor.  A  compari- 
son of  the  temperature  records  secured  in  this  experiment  with  the  data 
which  have  been  obtained  on  ordinary  silage  at  this  and  other  Experi- 
ment Stations  indicates  that  there  is  no  wide  difference,  if  any,  between 
the  rate  and  amount  of  increase  in  temperature  in  silos  in  which  ordi- 
nary silage  and  corn-stover  silage  have  been  stored. 

BactERIOIvOGICAIv  observations. — A  quantitative  bacterial  exami- 
nation was  made  on  each  of  the  samples  taken.  The  juice  obtained  was 
plated  in  proper  dilutions  and  the  counts  obtained  were  reported  as 
numbers  per  cubic  centimeter  of  juice.  Lactose  agar  was  used,  and  the 
plates  were  counted  after  six  days'  incubation  at  33°  C.  As  may  be 
seen  from  Table  III,  it  appears  that  the  bacterial  comit  increases  during 
the  first  week  and  is  followed  bv  a  continued  decrease  thereafter. 

Table  III. — N timber  of  bacteria  in  stover  silage  at  different  stages  of  curing 


Age. 

Number  of  bacte- 
ria    per     cubic 
centimeter. 

Age. 

Number   of  bacte- 
ria per  cubic  cen- 
timeter. 

Weeks, 
oh 

528, 000,  000 

3,  630,  000,  000 

I,  850,  000,  000 

975,000,000 

400,  000,  000 

Weeks. 
c 

510,  000,  000 

235, 000,  000 

186, 000,  000 

98,  000,  000 

71,000,000 

I 

6 

2 

8 

•J 

10 

4 

12 

Direct  microscopic  examinations  of  the  silage  juice  were  made  in 
order  to  follow  in  a  general  way  any  marked  changes  which  take  place 
in  the  bacterial  flora  during  the  curing  process.  This  at  best  could  only 
give  suggestive  data,  but  such  examinations  are  sometimes  important 
in  connection  with  cultural  studies.  At  first  a  great  variety  of  cells 
were  observed.  During  the  first  two  weeks  rods  and  cocci  were  appar- 
ently present  in  about  equal  numbers,  after  which  the  rods  became 
increasingly  predominant.  Toward  the  end  of  the  experiment  prac- 
tically nothing  but  rods  were  found  in  the  microscopic  preparations. 
Because  of  the  high  acidity  it  is  not  likely  that  the  cocci  were  active 
nearly  as  long  as  they  appeared  under  the  microscope.  But  it  is  prob- 
able that  the  acid  medium  would  tend  to  peserve  the  cells  so  that  they 
would  appear  for  some  time  after  they  were  inactive  or  even  dead. 

A  qualitative  bacterial  study  was  also  carried  out.  All  of  the  colonies 
from  a  representative  lactose-agar  plate  from  each  sample  were  isolated 
and  subjected  to  a  cultural  study.  For  the  present  purpose  they  may  be 
divided  roughly,  according  to  their  action  on  litmus  milk,  into  acid- 
forming,  casein-digesting,  alkali-forming,  and  inert  groups.  The  acid 
formers  may  be  further  divided  according  to  whether  they  produced 
sufficient  acid  to  cause  coagulation  of  the  milk.  The  distribution  of 
these  groups  in  silage  at  various  stages  of  its  fermentation  is  shown  in 
Table  IV. 


Mar.  4,  1918 


Corn-Stover  Silage 


595 


Table  IV. — Groups  of  bacteria  present  at  different  stages  of  curing 


Age. 


Percentage  of  total  number. 


Acid-non- 
coagulat- 
ing group. 


Weeks. 

3/7 

I 

2 

3 

5 

6 

8 

10 

12 


10 

53 

14 

57 

20 

71 

II 

63 

21 

67 

60 

30 

68 

32 

60 

40 

--,t> 

39 

Casein- 

Alkali- 

digesting 

fonniug 

group. 

group. 

10 

4 

9 

6 

0 

0 

5 

0 

0 

0 

0 

0 

0 

0 

0 

0 

5 

0 

Inert 
group. 


23 

14 

9 


10 
O 

o 
o 


Table  IV  shows  that  the  rather  complex  bacterial  flora  which  is  present 
at  the  beginning  of  the  process  gives  way  to  one  which  is  almost  entirely 
acid-producing  as  the  fermentation  progresses.  The  proportion  of 
acid-forming  and  coagulating  organisms  to  the  noncoagulating  ones 
also  increases  as  the  curing  period  advances.  A  comparison  of  these 
figures  with  those  given  in  Table  III  indicates  that  the  change  in  flora 
is  not  to  be  accounted  for  by  an  actual  increase  in  the  high  acid-forming 
organisms  during  the  latter  part  of  the  fermentation  period,  but  rather 
to  the  fact  that  they  do  not  decrease  as  rapidly  because  of  their  greater 
resistance  to  the  unfavorable  hydrogen-ion  concentration. 

The  division  of  the  acid-forming  organisms  into  coagulating  and  non- 
coagulating  types,  though  convenient  and  significant  for  the  present 
purpose,  probably  does  not  separate  them  into  natural  groups.  From 
early  in  the  fermentation  the  predominating  organisms  were  acid  formers, 
most  of  which  probably  belonged  to  the  same  general  group'.  We  have 
found  cultures  which  were  apparently  identical,  as  indicated  by  the 
cultural  and  fermentative  reactions  studied,  but  which  varied  in  the 
amount  of  lactic  acid  produced  in  milk  from  only  0.3  to  more  than  2.0 
per  cent.  All  of  these  probably  belonged  to  the  same  general  group  as 
the  aciduric  bacteria  which  have  previously  been  noted  as  occurring 
abundantly  in  silage.  In  the  first  two  samples  examined  organisms  of 
the  colon-aerogenes  group  were  found  and  ^Iso  a  few  cultures  which 
were  probably  Streptococcus  lacticus,  but  tests  were  not  applied  which 
would  definitely  identify  the  latter. 

NATURE    OF    SILAGE    FERMENTATION 

Until  recently  the  cell-respiration  theory  of  silage  fermentation  estab- 
lished by  Babcock  and  Russell  (/,  2)  has  not  been  seriously  challenged. 
During  the  past  year,  however,  several  publications  ha\e  appeared  in 
support  of  the  bacterial  explanation  of  this  phenomenon.  In  view  of 
the  recent  contributions  to  the  subject,  it  is  not  out  of  place  to  examine 
38324"— 18 6 


596  Journal  of  Agricultural  Research  voi.  xii,  no.  9 

critically  the  present  status  of  the  question.  As  the  older  literature 
has  been  reviewed  so  many  times  that  further  elaboration  is  not  neces- 
sary, we  shall  pass  in  review  only  those  papers  of  very  recent  date. 

Hunter  and  Bushnell  (8)  have  demonstrated  the  presence  in  silage 
of  large  numbers  of  high  acid-producing  bacteria,  and  have  furnished 
strong  evidence  that  these  organisms  are  mainly  responsible  for  the 
acid  fermentation.  Although  their  work  is  a  most  valuable  one,  it  should 
be  borne  in  mind  that  the  evidence  is  circumstantial  and  perhaps  not 
conclusive.  It  is  rather  doubtful  if  the  data  submitted  justify  the  positive 
conclusion  that — 

The  present  investigation  warrants  the  statement  that  acid  production,  common 
to  ail  normal  silage,  is  largely  the  result  of  fermentation  by  the  Bulgarian  group  of 
bacteria. 

The  fact  that  these  bacteria  formed  considerable  acetic  acid  when 
grown  in  alfalfa  extract  to  which  was  added  i  per  cent  of  glucose  hardly 
warrants  the  assumption  that — 

Although  these  organisms  evidently  do  not  produce  all  of  the  acetic  acid  found 
in  normal  silage,  they  must  be  responsible  for  a  large  per  cent  of  it. 

Sherman  (75)  also  noted  the  presence  of  large  num.bers  of  the  aciduric 
bacilli  in  silage  and  reported  some  observations  which  indicated  that 
they  are  of  significance  in  the  fermentation  process.  The  evidence, 
however,  was  not  direct  and  by  no  means  conclusive. 

Hunter  (7)  in  his  work  on  heat  production  in  silage  has  added  further 
weight  to  the  bacterial  theory  of  silage  fermentation.  Unfortunately 
it  is  not  possible  to  evaluate  properly  some  of  the  interesting  points 
contained  in  his  paper,  as  they  are  obscured  by  insufficient  description. 
For  example,  some  data  are  given  which  show  the  difference  in  heat 
formation  between  green  kafir  heated  and  green  kafir  inoculated  with 
Bacillus  bulgaricus.  The  exact  treatment  in  this  case  is  not  clear:  If 
the  inoculation  v\^as  made  into  heated  kafir,  the  results  are  of  utmost 
significance;  if,  on  the  other  hand,  as  the  caption  of  the  graph  would 
indicate  (see  7,  fig.  9),  the  inoculation  was  made  into  unheated  kafir 
and  that  compared  with  heated  kafir,  the  test  contributes  nothing  to 
the  solution  of  the  moot  question.  In  his  assumption  that  cell  respira- 
tion can  play  no  part  in  the  fermentation  of  silage  made  from  dry  for- 
age. Hunter  has  arrived,  we  think,  at  conclusions  which  are,  in  part  at 
least,  erroneous. 

In  a  very  interesting  paper  Lamb  (9)  concludes  that  both  factors 
are  of  importance,  btit  that  microorganisms  play  the  larger  part,  espec- 
ially in  the  production  of  acid.  Following  the  suggestion  of  Rahn  {10), 
he  has  attempted  to  determine  the  cause  of  the  process  by  the  rates 
of  change  in  the  fermentations  studied.  The  course  of  the  curve  obtained 
when  such  data  were  plotted  was  interpreted  as  indicating  whether 
the  action  was  of  bacterial  or  enzymic  origin. 


Mar.  4,  igis  Corn-Stover  Silage  597 

Although  it  is  true  that  purified  enzym  preparations  acting  under 
favorable  conditions  give  time  curves  which  follow,  with  certain  modi- 
fications, the  law  of  mass  action,  while  cur\res  representing  bacterial 
action  take  an  entirely  different  course  because  of  an  increase  in  the 
active  mass  with  the  multiplication  of  the  organisms,  we  are  inclined 
to  believe  that  Lamb  has  placed  undue  confidence  in  this  method, 
especially  when  we  consider  the  complexity  of  the  material  studied 
and  the  factors  concerned.  A  few  considerations  will  suffice  to  illustrate 
some  of  the  possibilities  of  error  in  such  a  method.  In  the  first  place 
conditions  in  silage  are  not  constant,  but  are  undergoing  continual 
change.  For  example,  the  temperature  and  acidity,  which  are  of  utmost 
significance  in  enzym  action,  increase  as  the  fermentation  progresses. 
In  view  of  the  great  increase  in  activity  of  some  enzyms  as  the  temper- 
ature is  increased,  and  the  stimulating  effect  on  some  enzyms  of  an 
increased  hydrogen-ion  concentration  (within  certain  limits),  it  is  not  at 
all  impossible  that  these  factors  might  so  modify  the  course  of  action 
of  an  enzym  as  to  produce  a  curve  resembling  that  typical  of  bacterial 
action. 

Again,  the  phenomenon  of  adsorption  and  the  action  of  the  so-called 
antienzyms  in  many  cases  may  so  suppress  the  activities  of  an  enzym 
during  the  early  stages  of  the  reaction  as  to  cause  it  to  follow  a  course 
not  at  all  characteristic  of  enzym  action.  This  has  been  beautifully 
illustrated  by  Rosenthal  (11)  in  his  work  on  the  antitryptic  action  of 
egg  albumen.  It  was  shown  that  the  trypsin  was  at  first  suppressed, 
but  gradually  regained  its  power  and  increased  in  activity  so  as  to  give 
the  appearance  of  an  increase  in  the  "active  mass"  as  indicated  by  a 
curve  convex  to  the  axis  of  abscissae  (the  typical  bacterial  curve).  On 
the  other  hand,  the  same  trypsin  preparation  when  acting  on  egg  albu- 
men which  had  been  previously  heated  to  destroy  the  antitrypsin  gave 
a  time  curve  characteristic  of  enzym  action. 

These  illustrations  will  suffice  to  show  the  fallacy  of  such  a  method, 
but  the  number  of  possibilities  of  error  in  its  application  to  such  a  com- 
plex mixture  as  silage  might  be  increased  almost  indefinitely.  That  the 
limitations  of  this  method  of  studying  biochemical  phenomena  were 
appreciated  by  Rahn  (10)  is  shown  in  the  following  paragraph  from 
his  valuable  paper: 

It  is  hardly  necessary  to  mention  that  the  curve  of  a  process  will  be  an  absolute 
means  of  discussion  only  in  case  of  pure  cultures.  In  natiu-al  fermentations,  there 
is  always  the  possibility  that  different  processes  taking  place  at  the  same  time  de- 
stroy the  regular  form  of  the  curve.  A  simple  example  would  be  the  growth  of  an 
acid-producing  and  an  alkali-producing  organism  in  the  same  liquid.  It  is  also 
possible  that  an  enz;jTnic  curve  imder  certain  conditions  shows  the  form  of  a  fermen- 
tation curve :  We  can  imagine  that  an  enzym  is  acting  slowly  at  first,  because  of  an 
imsatisfactory  acidity  of  the  medium.  By  a  chemical  or  microbial  process,  inde- 
pendent of  the  enzymic  action,  the  acidity  may  be  made  more  suitable  for  the  en- 
zyme, and  this  will  cause  an  increased  rate  of  action  of  the  enzyme  and  give  the  type 


598 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  9 


of  a  fermentation  curve  without  the  presence  of  organisms.  The  value  of  the  curve 
is,  therefore,  not  an  absolute  one  and  no  conclusions  ought  to  be  drawn  without  con- 
sideration of  the  possibilities  of  error. 

It  would  indeed  seem  that  the  application  of  such  methods  to  the 
study  of  silage  fermentation  is  entirely  without  foundation. 

To  the  evidence  in  support  of  the  bacterial  theory  of  silage  fermenta- 
tion should  be  added  the  very  suggestive  observation  of  Clark  (4)  that 
the  hydrogen-ion  concentration  of  mature  silage  is  coir;cident  with  the 
limiting  hydrogen-ion  concentration  obtained  in  cultures  of  BacUlus  bui- 
garicus,  which  organisms  are  considered  by  some  workers  as  of  para- 
mount importance  in  the  ripening  process.  In  support  of  the  respiration 
theory,  on  the  other  hand,  it  is  pertinent  to  call  attention  to  the  recent 
work  of  Round  (12),  which  indicates  that  cell  respiration  is  of  greater 
importance  in  the  fermentation  of  sauerkraut  than  has  been  generally 
recognized. 

It  was  thought  at  the  beginning  of  this  experiment  that  a  study  of 
the  fermentation  in  silage  made  from  dry  stover  would  throw  much 
light  on  the  nature  of  the  process  in  ordinary  silage.  The  belief  was 
held  that  the  activity  of  the  plant  cells  (which  have  been  demon- 
strated to  play  an  important  part  in  the  fermentation  of  silage  made 
from  green  com)  would  be  eliminated,  to  a  large  extent  at  least,  in  the 
stover  silage.  But,  as  may  be  seen  from  an  examination  of  the  fore- 
goiag  results,  the  fermentation  of  ensiled  stover  appears  to  resemble, 
in  its  main  characteristics,  that  which  takes  place  in  green-corn  silage. 
In  an  effort  to  determine  the  factors  responsible  for  the  fermentation, 
laboratory  tests  were  made  by  ensiling  stoVer  under  different  condi- 
tions. Glass  jars  containing  175  gm.  of  cut  stover  and  400  gm.  of  water 
were  used.  Some  were  untreated,  some  put  up  with  antiseptics,  while 
some  were  sterilized  in  the  aiitoclave  and  reinoculated  with  i  per  cent 
of  raw-silage  juice.     The  results  of  this  study  are  given  in  Table  V. 

Table  V. — Fermentation  of  stover  ensiled  under  different  conditions 


Sample; 
No. 


Treatment. 


Untreated 

....do 

Sterilized  and  inoculated 

...do 

2  per  cent  of  toluene 

....do 

2  per  cent  of  ether 

do 


Aj;e. 

Acidity. 

Weeks. 

Per  cent. 

4 

15-5 

4 

10.4 

7 

2.4 

5 
6 
6 

3-3 
4-3 
4.6 

5 

10.8 

8 

12.4 

Bacterial  count. 


210,  000,  000 

240, 000,  000 

320,  000,  000 

36,  000,  000 

180,  000 

460, 000 

7,  500,  000 

Not  made. 


The  main  points  brought  out  by  this  test  are  that  stover  silage  is 
capable  of  undergoing  a  fermentation  when  preser\^ed  with  ether,  while 
bacteria  alone  are  apparently  unable  to  produce  the  typical  fermenta- 
tion, even  though  conditions  are  favorable  for  their  active  development. 
The  predominating  organisms  in  the  sterilized  and  inoculated  samples 


Mar.  4, 1918  Corn-Stover  Silage  599 


were  of  the  same  type  as  those  characteristic  of  normal  silage.  The 
organisms  found  in  the  samples  preserved  with  antiseptics,  on  the  other 
hand,  were  a  more  miscellaneous  group;  and,  although  many  probably 
belonged  to  the  same  group  as  the  aciduric  bacilli  of  normal  silage,  the 
cultures  isolated  were  mostly  very  weak  acid  producers.  The  fact  that 
fermentation  took  place  under  ether  indicates  that  the  activity  of  the 
plant  cells,  whether  it  be  called  "respiration"  or  "autolysis,"  is  present 
in  silage  made  from  dry  stover.  As  silage  preserv^ed  with  ether  fer- 
mented, whereas  in  that  kept  with  toluene  the  process  was  checked 
suggests  that  conclusions  drawn  from  experiments  conducted  with  only 
one  antiseptic  are  of  doubtful  value.  When  opened,  the  ether-preserved 
samples,  after  the  evaporation  of  the  ether,  appeared  to  resemble  the 
untreated  material,  while  the  sterilized  and  inoculated  silage  were 
"flat"  and  lacking  in  the  characteristic  aroma.  The  results  with  ether 
were  checked  by  another  test  in  which  triplicate  samples  were  pre- 
served, and  again  an  active  fermentation  took  place  as  was  indicated 
by  the  development  of  acidity  in  each  case. 

Although  our  results  would  tend  to  support  the  respiration  theory 
of  silage-curing  rather  than  the  bacterial,  we  do  not  feel  that  the  data 
thus  far  collected  warrant  definite  conclusions  on  this  point.  It  is 
difficult  to  believe  that  such  active  acid-forming  organisms  should 
occur  in  silage  in  large  numbers  without  taking  some  part  in  the  acid 
fermentation;  perhaps  they  supplement  in  some  important  way  the 
action  of  the  plant  cells.  It  is  not  inconceivable  that  a  preliminary 
cleavage  due  to  cell  respiration  is  an  essential  prerequisite  for  the 
vigorous  action  of  the  aciduric  bacteria.  In  fact,  the  continued  increase 
in  the  ratio  of  nonvolatile  to  volatile  acidit}'  as  the  fermentation  pro- 
gressed (see  Table  I)  might  lead  one  to  suspect  that  such  was  the  case. 
On  the  other  hand,  the  great  increase  in  the  nonvoladle  acidity  from 
the  fifth  to  the  twelfth  week,  during  which  time  the  bacterial  count 
was  rapidly  decreasing,  might  be  interpreted  as  strong  evidence  against 
that  view.  It  is  clear  that  microorganisms  are  not  solely  responsible 
for  the  fermentation  of  silage,  and  the  weight  of  evidence  at  the  present 
time,  in  our  opinion,  indicates  that  their  role  is  not  as  important  as  that 
of  the  plant  cells. 

Although  not  committing  ourselves  definitely  qn  the  nature  of  silage 
fermentation  in  general,  in  regard  to  the  present  problem  we  do  maintain 
that  the  fermentation  which  takes  place  in  stover  silage  is  similar  in  its 
essential  points  to  that  of  ordinary  silage  and  is  caused  by  similar 
factors.  SUMMARY 

Com  stover  when  ensiled  with  a  suitable  quantity  of  water  undergoes 
fermentation  with  the  production  of  a  palatable  silage  of  good  -keeping 
quality,  which  resembles  ordinary  com  silage  in  aroma  and  appearance. 

The  fermentation  which  takes  place  in  corn-stover  silage  appears  to 
be  essentially  the  same  as  that  of  silage  made  from  green  com,  as  is 


5oo  Journal  of  Agricultural  Research  voi.  xii.  N0.9 

indicated  by  the  total  acidity  developed,  the  ratio  of  nonvolatile  to 
volatile  acids,  temperature  observations,  and  bacterial  studies. 

From  a  review  of  the  present  status  of  the  question  as  to  whether 
bacteria  or  plant  cells  are  mainly  responsible  for  silage  fermentation, 
it  is  concluded  that  the  data  thus  far  published  are  inconclusive. 
Although  the  results  of  the  present  study  tend  to  support  the  cell- 
respiration  theory,  conclusions  on  this  point  are  withheld. 

LITERATURE  CITED 

(i)  Babcock,  S.  M.,  and  Russell,  H.  L. 

1900.  CAUSES  OPERATIVE  IN  THE  PRODUCTION  OF  SILAGE.     In  Wis.  Agr.   Exp. 

Sta.  i7tli  Ann.  Rpt.,  [1899V1900,  p.  123-141,  fig.  17. 
(2) 

1901.  CAUSES    OPERATIVE    IN    THE    FORMATION    OF    SILAGE-      (SECOND    PAPER.) 

In  Wis.  Agr.  Exp.  Sta.  i8th  Ann.  Rpt.,  [i90o]/i90i.  p.  177-1S4,  fig.  44. 

(3)  BechdEL,  S.  I. 

1917.  STUDIES  IN  THE  PRESERVATION  OF  CORN  SILAGE.  3°  P-  Reprint  (separate 
No.  y)from  Pa.  Agr.  Exp.  Sta.  Ann.  Rpt.     1915/16. 

(4)  Clark,  W.  M. 

191 7.  THE  ACID  PRODUCTION  OF  BACILLUS  BULGARicus.  (Abstract.)  In  Abs. 
Bact.,  V.  I,  no.  i,  p.  59-60. 

(5)  Dox,  A.  W.,  and  Neidig,  R.  E. 

1913.  lactic  ACID  in  CORN  SILAGE.     lowa  Agr.  Exp.  Sta.  Research    Bui.   10, 

P-  363-37S.  4  fig- 

(6)  EcKLES,  C.  H.,  OsHEL,  O.  I.,  and  Magruder,  D.  M. 

1916.  SILAGE  investigations:  normal  temperatures  and  some  factors 
INFLUENCING  THE  QUALITY  OF  SILAGE.  Mo.  Agr.  Exp.  Sta.  Research 
Bui.  22,  32,  p.,  7  fig. 

(7)  Hunter,  O.  W. 

i917.  microorganisms  and  heat  production  in  silage  fermentation. 
In  Jour.  Agr.  Research,  v.  10,  no.  2,  p.  75-83,  10  fig.  Literature 
cited,  p.  82-83. 

(8)  and  BushnELL,  L.  D. 

1916.  SOME   IMPORTANT   FERMENTATIONS    IN   SILAGE.     Kan.    Agr.    Exp.    Sta. 

Tech.  Bui.  2,  32  p.     Bibliography,  p.  32. 

(9)  Lamb,  A.  R. 

1917.  THE   RELATIVE   INFLUENCE   OF    MICROORGANISMS    AND  PLANT    ENZYMS    ON. 

THE  FERMENTATION  OF  CORN  SILAGE.     In  JouT.  Agr.  Research,  v.   8, 
no.  10,  p.  361-380,  13  fig.     Literature  cited,  p.  37S-380. 

(10)  Rahn,  Otto. 

I910.    THE     USEFULNESS     OF    CURVES    IN    THE    INTERPRETATION   OF   MICROBIAL 

AND    BIOCHEMICAL    PROCESSES.     Mich.    Agr.    Exp.   Sta.   Tech.    Bui. 
5,  29  p.,  18  fig. 

(11)  Rosenthal,  Eugen. 

I9IO.    UNTERSUCHUNGEN       UBER       die      ANTrPROTEOLYTlSCHE      WIRKUNG      DES 

BLUTSERUMS.     In  Folia  Serol.,  Bd.  6,  p.  285-300. 

(12)  Round,  L.  A. 

I917.    THE    BACTERIOLOGY    OF    SAUERKRAUT,    A    FURTHER  STUDY.      (Abstract.) 

In  Abs.  Bact.,  v.  i,  no.  i,  p.  50. 

(13)  Sherman,  J.  M. 

I916.    A  CONTRIBUTION  TO  THE  BACTERIOLOGY  OF  SILAGE.      In  JoUT.  Bact.,  V.  I, 

no.  4,  p.  445-452.     Bibliography,  p.  452. 


WEEVILS  WHICH  AFFECT  IRISH  POTATO,  SWEET 
POTATO,  AND  YAM 

By  W.  DwiGHT  Pierce, 

Entomologist,  Southern  Field  Crop  Insect  Investigations,  Bureau  of  Entomology,  United 
States  Department  of  Agriculture 

INTRODUCTION 

In  a  previous  article  ^  the  writer  discussed  three  important  Andean 
weevil  pests  of  the  potato  tuber  {Solanum  tuberosum).  In  the  present 
paper  a  fourth  potato  tuber  weevil  is  described  and  notes  are  presented 
on  three  weevils  which  attack  the  tubers  of  sweet  potato  (Jpomoea 
batatas)  and  one  which  attacks  the  tubers  of  the  yam  {Dioscorea  batatas). 

WEEVILS  WHICH  AFFECT  IRISH  POTATO  TUBERS 

The  native  home  of  the  Irish  potato  is  the  west  coast  of  South  America, 
and  here  we  find  that  the  crop  has  a  series  of  characteristic  pests  which 
may  be  easily  disseminated  in  shipments  of  potatoes.  As  stated  above, 
the  writer  has  described  three  of  these  species  in  a  previous  paper.  The 
description  of  the  larva  of  one  of  them  is  now  added  and  a  new  species 
described.  In  order  that  these  weevils  may  be  easily  distinguished  one 
from  another  by  the  man  in  the  field  the  following  table  has  been  con- 
structed : 

TABLE   OF   IRISH   POTATO   TUBER   WEEVILS 

1.  Prostemum  grooved  for  reception  of  beak;  mandibles  without  deciduous  piece; 

pronotum  with,  a  deep  median   furrow  widened  angularly  at  middle  and  also 

behind Rhigopsidius  tucumanus  Heller. 

Prostemum  not  grooved  for  reception  of  beak;  mandibles  with  deciduous  piece .  .  2 

2.  Mandibles  with  tooth  beneath;  scrobes  abruptly  and  broadly  terminated,   not 

extending  beneath ? 

Mandibles  without  tooth  beneath;  scrobes  narrowing  and  extending  beneath;  pro- 
thorax  not  as  wide  as  elytra,  angulate  and  broadest  in  front  of  base 

Premnotrypes  solani  Pierce. 

3.  Prothorax  broader  than  elytra,  subquadrate,  with  sides  parallel  to  apical  third, 

thence  strongly  narrowed ;  sides  of  elytra  smooth .  .  Trypopremnon  laiithorax  Pierce. 
Prothorax  acutely  angulate  at  sides,  widest  before  base;  sides  of  elyixa  tuberculate, 

Trypopremnon  sanfordi,  n.  sp. 

'  Pierce,  W.  D.    new  potato  weevus  from  andean  south  America.    In  Jour.  Agr.  Research,  v.  i, 
no.  4,  p.  347-352,  3  fig.,  pi.  39-41-    1914. 

Journal  of  Agricultural  Research,  Vol.  XII,  No.  9 

Washington,  D.  C.  (601)  Mar.  4,  191S 

ms  Key  No.  K— 61 


6o2  Journal  of  Agricultural  Research  Voi.  xii.No.  9 

FAMILY   PSALIDURIDAE   PIERCE  (1914) 

Rhigopsidius  tucumanus  Heller  (1906) 

This  weevil  has  been  recorded  by  the  writer  ^  from  Tucuman,  Argen- 
tina; Cuzco,  Temuco,  and  Arequipa,  Peru;  Oruro,  Bolivia;  and  Ancud  or 
San  Carlos  and  Castro  Islands,  Chile.  The  weevil  belongs  to  the  sub- 
family Rhytirhininae  and  the  tribe  Rhytirhinini. 

FAMILY    PSALLIDIIDAE   PIERCE  (1916) 

Premnotrypes  solani  Pierce  (1914)  ^ 

This  weevil  was  described  from  the  mountain  districts  of  Peru.  It 
belongs  to  the  subfamily  Entiminae  and  the  tribe  Ophryastini. 

Trjrpopremnon  latithorax  Pierce  (1914)  ^  (PI.  29,  30) 

This  weevil  was  described  from  Cuzco,  Peru.  It  belongs  to  the  same 
tribe  as  Premnotrypes.  The  ventral  tooth  of  the  mandible  does  not 
belong  to  the  deciduous  piece  as  stated  in  the  original  description. 

On  June  ii,  1914,  Mr.  H.  L.  Sanford  found  several  larvae  of  this  species 
in  potatoes  from  La  Paz,  Bolivia,  collected  by  Mr.  H.  T.  Knowles,  under 
Federal  Horticultural  Board  No.  2475.  On  June  20-26  pupae  were 
noticed,  and  an  adult  emerged  on  June  26. 

This  enables  the  writer  to  describe  these  stages. 

Larva  (PI.  29). — Length  12.5  mrn.  when  crawling,  10  mm.  when  slightly  curved. 
It  is  typically  rhynchophorid  in  form,  white,  with  light  reddish  brown  head  and  dark 
mandibles.  The  essential  diagnostic  characters  are  illustrated  by  the  author  in 
Plate  29. 

From  the  base  of  the  head  a  median  pale  line  passes  forward.  This  is  the  epicranial 
suture.  It  divides  behind  the  frons  and  forms  the  two  frontal  sutures.  The  frons  is 
subtriangular,  rounded  behind  and  margined  in  front  by  the  epistoma.  The  epicranial 
areas  are  the  two  large  areas  at  each  side  of  the  epicranial  suture,  further  bounded  by 
the  frontal  sutiu-e,  the  pleurostoma,  and  the  hypostoma.  In  front  of  the  frons  is  the 
clypeus,  and  in  front  of  this  is  the  labrum.  The  clypeus  and  labrum  partly  overlap 
the  mandibles  which  arise  at  the  side  of  the  clypeus  based  on  the  pleurostoma.  Below 
the  hypostoma  at  the  sides  of  the  mandibles  arise  the  maxillae,  of  which  the  cardo  is 
a  Very  large  basal  area.  Located  on  the  median  line  below  the  mouth  opening,  which 
is  covered  by  labrum  and  mandibles,  is  a  shield-shaped  area  known  as  stipes  labii. 
Below  and  around  this  is  the  large  basal  area  consisting  of  mentum  and  submentum. 

There  is  a  small  abortive  branch  of  the  frontal  suture  extending  back  on  the  epicra- 
nium,  on  each  side  of  and  not  far  from  the  epicranial  suture.  It  is  terminated  by  a 
setigerous  puncture.  On  the  epicranium  there  are  setae  arranged  as  follows  on  each 
lobe:  One  at  terminus  of  branch  of  frontal  suture,  one  on  the  frontal  suture,  two  oppo- 
site middle  of  frons,  one  basal,  two  discal ,  one  opposite  base  of  mandible,  two  on  hypo- 
stoma. On  the  frons  there  are  three  pairs  of  setas,  the  two  posterior  pairs  being  about 
equidistant,  the  anterior  close  to  the  antennae.  At  base  of  clypeus  there  are  four 
tiny  hairs.  On  labrum  are  four  subbasal,  six  subapical,  and  six  marginal  hairs.  The 
mandibles  have  one  hair  each.  The  maxillae  are  provided  with  two-jointed  palpi 
(PI.  29,  F),  and  a  very  broad  setose  lacinia,  two  setae  near  base  of  palpi  and  one 
near  base.  Some  of  the  hairs  are  clavate  as  shown  in  the  illustration,  but  this  is  not 
always  true,  some  specimens  having  normal  hairs.  The  stipes  labii  has  one  pair  of 
hairs.     Each  lobe  of  the  mentum  has  one  pair. 

>  Pierce,  V»'.  d.    Op.  cit. 


Mar.4.  i9i8  WeevUs  Affecting  Potatocs  603 

The  pronotum  is  simple,  undivided;  the  mesonotum  and  metanotum  are  composed 
of  praescutum  and  scutoscutellum.  The  first  six  abdominal  sclerites  are  composed 
of  a  spindle-shaped  praescutum,  a  transverse  scutum  terminated  by  the  spiracles,  a 
spindle-shaped  scutellum,  and  a  transverse  postscutellum  very  greatly  narrowed  on 
the  dorsum.  The  praescutum  has  a  few  hairs.  The  scutellum  has  a  row  of  hairs. 
Just  above  each  spiracle  is  a  tiny  hair.  On  each  epipleural  lobe  beneath  the  spiracles 
there  is  one  hair.  There  are  eight  abdominal  spiracles  and  one  on  the  mesothorax. 
The  seventh  and  eighth  segments  are  more  crowded  than  the  preceding.  The  ninth 
and  tenth  are  small  and  reduced. 

Pupa  (PI.  30). — Length  10  mm.,  white.  The  most  interesting  features  of  this  pupa 
are  the  rudimentary  wing  pads  seen  only  when  the  elytra  are  spread.  The  elytral  pads 
are  not  as  large  as  often  found  in  Aveevil  pupae.  The  antennae  are  not  geniculate. 
The  beak  is  short.  There  are  five  pairs  of  hairs  located  on  the  head  and  beak  as 
illustrated.  On  the  thorax,  which  is  subquadrate  with  truncate  angles,  there  are 
setigerous  tubercles  as  follows:  Four  on  anterior  margin,  two  antemedian  and  two 
postmedian  on  the  disc,  two  pairs  of  antemedian  and  two  pairs  of  postmedian  on  lateral 
margin.  Mesonotum  and  mettootum  with  one  pair  of  setae  each.  The- first  abdominal 
segment  has  two  pairs  of  setse,  and  the  remaining  segments  have  a  long  line  of  setigerous 
tubercles.  Each  femur  has  two  apical  hairs,  and  a  few  ventral  hairs  are  found  as 
illustrated.  It  is  interesting  to  note  that  the  processes  of  the  ninth  segment  are  acute 
but  reduced  almost  to  the  size  of  the  tubercles.  The  tenth  segment  is  ventral  to  the 
ninth. 

Trypopremnon  sanfordi,  n.  sp.  (PI.  28) 

Described  from  a  single  specimen  collected  in  quarantine  by  Mr.  H.  1,. 
Sanford  September  24, 191 5,  from  a  potato  tuber  sent  by  Mr.  O.  F.  Cook 
from  Cuzco,  Peru.^  The  excellent  illustrations  of  the  type  (PI.  28)  were 
made  under  the  writer's  supervision  by  Mr.  H.  B.  Bradford. 

Length  8  mm.,  greatest  breadth  4.5  mm.  Beak  longer  than  head  and  narrower 
than  eyes;  the  dorsal  squamose  portion  being  gradually  narrowed  from  eyes  to  nasal 
plate.  Alae  strongly  flared,  making  the  scrobes  open  above.  Head  tumid  above 
the  eyes.  Median  line  slightly  depressed  on  head,  strongly  in  frontal  fovea,  and 
very  faintly  on  beak  except  just  behind  nasal  plate.  Lateral  depressions  on  beak 
strong.  Apex  of  beak  brownish  black,  -with  nasal  plate  polished,  convexly  raised 
around  margin,  emarginate  at  apex.  Mandibles  shining  brownish  black;  deciduous 
piece  reddish  brown,  lightest  at  tip,  moderately  long,  arcuate,  with  sharp  edges; 
the  ventral  tooth  is  not  as  acute  as  in  T.  latithorax;  there  is  a  slight  denticle  on  the 
right  deciduous  piece,  and  the  left  mandible  is  denticulate  as  shown  in  the  figure. 
The  antennal  scrobes  are  strongly  flexed  downward,  very  much  broadened  and 
evanescent  behind;  scape  clavate;  fimicle  with  first  two  joints  elongate,  the  others 
progressively  shorter,  the  last  moniliform;  club  as  long  as  the  foiu-  preceding  joints. 
Head,  beak,  and  scape  densely  clothed  with  fine  silky-bronze  scales,  and  with  scat- 
tered white  setae;  funicle  sparsely  setose;  club  minutely  pubescent,  sparsely  setose. 

Prothorax  basally  truncate,  slightly  broadly  emarginate  at  middle;  apically  sinuate; 
with  very  strong  supraocular  lobes,  which  have  vibrissae  on  the  inner  surface;  surface 
coarsely  irregularly  punctured,  finely  densely  squamose  with  golden  metallic  scales," 
sparsely  setose  with  white  curved  setae;  surface  very  uneven,  with  median  depression 
bordered  by  antemedian  ridges  and  two  postmedian  tubercles;  sides  prominently 
produced  by  two  angulate  tubercle's;  widest  at  posterior  tubercles. 

Elytra  at  base  narrower  than  thorax;  humeri  tuberculate;  sides  subparallel  but 
very  roughly  tuberculate,  abruptly  narrowed  at  posterior  declivity  which  is  nearly 
perpendicular.     Scutellum  triangular.    Surface  densely  minutely  squamose,  sparsely 

'  Recorded  under  Federal  Horticultural  Board  No.  4348. 


6o4  Journal  of  Agricultural  Research  voi.  xii,  N0.9 

setose;  striae  irregular,  with  strong  isolated  punctures;  entire  surface  covered 
with  tubercles  which  are  largest  on  the  third,  fifth,  and  seventh  intervals. 

Front  coxae  contiguous.  Prostemum  strongly  arcuately  emarginate.  Meso- 
stemal  coxae  narrowly  separated.  Intercoxal  piece  of  first  abdominal  segment 
broad  and  deeply  punctate.     Second  segment  as  long  as  third  and  fourth  together. 

Type. — Cat.  No.  21613,  United  States  National  Museum. 

WEEVILS  WHICH  AFFECT  SWEET-POTATO  TUBERS 

At  least  four  species  of  weevils  attack  the  tubers  of  the  sweet  potato — 
namely,  Euscepes  batatae  Waterhouse,  Cylas  formicarius  Fabricius,  C. 
turcipcnnis  Boheman,  and  C.  jemoralis  Faust.  The  adults  of  Euscepes 
and  Cylas  can  not  be  confused.  Those  of  Cylas  are  differentiated 
as  shown  in  the  "Table  of  sweet-potato  weevils  of  the  genus  Cylas,"  p. 
605.  The  pupffi  of  Cylas  can  be  recognized  from  the  fact  that  the  direction 
of  the  appendages  is  anteriad,  while  in  Euscepes  it  is  posteriad.  Refer- 
ence to  the  illustrations  will  be  of  great  assistance  in  separating  them. 
The  larvse  can  not  be  so  easily  distinguished,  as  both  are  of  the  same 
general  shape.  It  will  be  noticed,  however,  that  they  have  quite  a 
different  system  of  abdominal  folds.  The  larva  of  Euscepes  is  more  com- 
pact. That  of  Cylas,  when  alive,  is  often  attenuate  and  tightly  drawn 
so  that  no  folds  can  be  distinguished.  When  killed  in  a  liquid  which 
shrinks  it  slightly,  however,  it  will  be  noticed  that  the  praescutal  areas 
are  proportionately  larger  and  often  subdivided  transversely.  The 
praescutum  of  Euscepes  is  not  subdivided.  This  sclerite  is  the  anterior 
sclerite  of  a  segment  and  almost  always  has  a  few  tiny  hairs.  The  only 
other  dorsal  sclerite  with  hairs  is  the  scutellum.  In  Cylas  the  scutellum 
adjoins  the  praescutum,  and  the  scutum  is  only  lateral.  In  Euscepes 
the  scutellum  is  separated  from  the  praescutum  by  the  scutum. 

FAMILY   APIONIDAE  LE  CONTE  (1876) 

Subfamily  Cyladinae  Pierce  (1916) 

genus  cylas  latreille  (1802) 

Cylas  Latreille,  1802,  Hist.  Nat.  Gen.  et  Part.  Crust,  et  Insects,  t.  3,  p.  196. 

Type. — Cylas  brunneus  Fabricius,  monotypic. 

This  genus  contains  twenty  named  species,  of  which  two  are  widely 
known  under  the  names  formicarius  and  turcipcnnis.  There  is  consid- 
erable confusion  about  these  two  species,  due  in  part  to  the  claims  of 
lyC  Conte  and  Faust  that  they  are  synonymous.  Fabricius  described  a 
piceous-brown  Indian  species  with  reddish  thorax  as  formicarius;  Olivier 
illustrated  the  species  as  almost  pink  but  described  it  as  brownish; 
Schonherr  cited  it  as  piceous;  Gyllenhal  described  a  species  from  Java 
with  greenish-blue  elytra,  red  thorax,  and  black  head  as  turcipcnnis; 
Labram  and  Imhoff  illustrated  a  blue  species  under  this  name.  Finally 
Wagner  presented  an  illustration  of  a  species  with  green  elytron  as 
turcipcnnis,  and  he  probably  is  right.  The  bluish  species  was  named 
elegantulus  by  Summers.      It  is  a  common  sweet-potato  weevil.     If  it 


Mar  4,  iyi8  WeevUs  Affectiug  Potatoes  605 

should  prove  to  be  a  variety  of  formicarius  as  here  treated,  after  exam- 
ination of  the  type,  it  must  still  be  considered  very  distinct  specifically 
from  iurcipennis.  The  National  Museum  collection  contains  six  species. 
Sketches  have  been  made  of  the  side  view  of  the  head  and  thorax  ^of 
the  three  species  which  presumably  attack  sweet  potato,  formicarius 
variety  elegantulus  (Pi.  31,  A),  femoralis  (Pi.  31,  F),  and  iurcipennis 
(PI.  31,  B),  and  also  of  brunneus  (PI.  31,  C-E),  which  was  erroneously 
recorded  by  the  writer  from  sweet  potato  in  the  Manual  of  Dangerous 
Insects.^ 

TABLE  OP  SWEET  POTATO  WEEVILS  OP  THE  GENUS  CYLAS 

1.  Male  club  twice  as  long  as  funicle  or  longer;  antennae  as  long  as  head  and  thorax; 

head  not  more  than  one-fifth  shorter  than  beak;  elytra  greenish,  thorax  red, 

head  black,  legs  red  with  dark  band  (PI.  31,  B) iurcipennis  Boheman. 

Male  club  not  twice  as  long  as  funicle 2 

2.  Male  club  half  to  three-fourths  longer  than  funicle,  female  club  almost  one-third 

shorter  than  funicle;  male  antennas  almost  as  long  as  head  and  thorax;  head 
one-fourth  to  one-third  shorter  than  beak;  el^-tra  bluish,  thorax  red,  head  black, 

legs  red  (PI.  31,  A) .formicaritis  Fabricius,  var.  elegantulus  Summers. 

Male  club  half  longer  than  funicle ;  head  as  long  as  beak ;  antennae  as  long  as  thorax 
plus  head  behind  eye;  el;>i;ra  black  with  blue  or  green  luster,  suture  piceous; 
thorax  black,  margins  piceous;  head  black;  legs  dark  red  with  black  ring  on 

femora  (PI.  31,  F) .femoralis  Faust. 

Cylas  formicarius  Fabricius  (1798) 

Brentus  formicarius  Fabricius,  1798.  Sup.  Ent.  Syst.,  p.  174,  no.  5. 

Fabricius  -  gave  the  following  description : 

Habitat  Tranquebariae. 

Parvus  in  hoc  genere.  Rostrum  cylindricum,  atrum,  antennis  rufis,  monili- 
formibus:  articulo  ultimo  longiori,  cylindrico,  clavato.  Thorax  rufus,  antice 
globosus.  Elytra  laevia,  atra,  nitida.  Pedes  rafi,  femoribus  clavatis,  at  inermibus: 
annulo  nigro. 

In  altero  sexu  antennarum  clava  brevior,  ovata. 

Olivier  and  Schonherr  described  the  species  as  piceous  with  ferru- 
ginous thorax,  antennae,  and  legs.  It  hardly  seems  possible  that  this 
can  be  the  same  species  as  the  common  sweet  potato  wee\'il  with  shiny 
blue-black  elytra,  red  thorax  and  appendages,  and  black  head  and  beak. 

For  this  reason  it  is  considered  best  to  apply  to  the  sweet-potato 
weevil  a  name  which  certainly  applies  to  it — elegantulus  Summers. 
In  order  that  economic  entomologists  may  not  be  inconvenienced  greatly, 
and  in  deference  to  the  many  writers  who  have  assigned  Fabricius's 
name  to  the  sweet-potato  weevil,  elegantulus  may  be  considered  as  a 
variety  of  formicarius  until  there  can  be  an  examination  of  the  type. 
Cylas  formicarius  elegantulus  Summers  (1875),  the  Sweet  Potato  Weevil  (PI.  31, 
A;  PI.  32,  A,  B;  PI.  2>Z<  E-H;  PI.  34,  A-D) 

Otidocephalus  elegantulus  Summers,  1875,  in  New  Orleans  Home  Jour.,  Jan.  and  Dec. 

,Cylas  formicarius  Le  Conte,  1876,  in  Proc.  Amer.  Phil.  Soc.,  v.  15,  p.  327.     O .  elegantulus  is  QVLOted  in 
synonymy. 

>  Pierce,  W.  D.    a  manual  of  dangerous  insects  ...  p.  209.     1917-    Published  by  the  United  States 
Department  of  Agriculture,  Office  of  Secretary. 
2  Fabricius,  J.  C.    systema  eleutheratorum  ...  v.  2  p.  549.    Kiliae,  1801. 


6o6  Journal  of  Agricultural  Research  voi.  xii, N0.9 

This  is  the  common  sweet-potato  weevil  (PI.  31,  A;  32,  A,  B)  with 
bluish  elytra,  red  thorax  and  appendages,  and  black  head.  The  illus- 
tration of  the  adult  is  drawn  from  a  New  Orleans  specimen.  The  side 
view  of  head  and  thorax  is  from  a  Hawaiian  specimen.  There  is  quite 
a  r^nge  of  difference  in  measurements  of  the  species  but  the  analysis  of 
these  differences  is  reserved  for  a  more  technical  paper  now  in  prepara- 
tion. The  immature  stages  are  described  from  specimens  collected  at 
Victoria,  Texas. 

Larva.  (PI.  34,  A-D). — The  larva  of  this  species  measures  from  5  to  8  mm.  in 
length  and  is  white,  with  light  brownish  head  and  darker  brown  mandibles.  The 
head  shield  is  slightly  angulately  emarginate  behind.  From  the  center  of  the  emar- 
gination  on  the  median  line  the  epicranial  suture  passes  forward,  separating  the  epi- 
cranium  into  two  parts;  this  suture  divides  behind  the  frons  and  forms  the  two  frontal 
sutures.  The  frons  is  subtriangular,  emarginate  at  anterior  angles  for  antennae, 
and  emarginate  along  epistoma  for  attachment  of  clypeus.  The  median  line  is  im- 
pressed and  darkened.  The  frons  has  three  pairs  of  large  setae,  the  posterior  pair 
being  closest  and  the  median  pair  but  little  more  separated.  The  anterior  pair  are 
located  very  close  to  the  antennal  fossae.  A  tiny  pair  of  setae  are  located  in  such  a 
way  as  almost  to  form  an  equilateral  triangle  with  the  posterior  and  median  setae. 

The  epicranial  areas  are  located  on  each  side  of  the  epicranial  suture.  Each  lobe 
bears  the  following  setae:  One  very  close  to  the  apex  of  frons,  one  slightly  posterior 
to  this  and  farther  from  the  median  line,  one  opposite  the  middle  of  the  frons,  one 
a  little  farther  from  the  median  line  on  the  same  line  as  the  preceding,  one  toward 
the  base  of  the  frons,  one  opposite  the  mid(Jle  of  the  mandible,  one  opposite  the  hy{X)S- 
tomal  angle  of  mandible,  one  on  hypostoma  near  base  of  mandible,  one  opposite 
but  distant  from  mandible,  and  three  forming  a  triangle  on  disc  of  epicranium. 

The  antenna  is  a  fleshy,  two-jointed  appendage  located  at  the  lateral  angle  of  the 
frons.  The  mandibles  are  very  bluntly  bidentate.  Each  mandible  has  a  tiny  hair 
about  the  middle.  The  clypeus  is  attached  in  front  of  the  frons  and  is  broadly  trans- 
verse. It  bears  on  the  epistomal  margin  four  tiny  hairs.  The  labrum  is  not  as  broad, 
is  rounded  in  front,  and  has  a  row  of  four  setae  behind  the  middle,  a  seta  on  each  side 
in  front  and  closer  than  the  outermost  setae,  and  four  marginal  setae.  The  maxillae 
are  elongate,  terminated  by  a  two-jointed  palpus  and  a  setose  lacinia.  The  maxillae 
are  provided  with  four  setae,  two  near  palpus,  one  at  middle,  and  one  at  base.  The 
stipes  labii  is  cordate,  bearing  two-jointed  palpi  and  a  single  pair  of  setae.  Each 
lobe  of  the  mentum  is  provided  with  two  pairs  of  setae. 

The  entire  surface  of  the  body  is  covered  with  tiny  pubescence. 

The  prothorax  is  dorsally  not  divided  but  has  the  praescutal  and  scutal  areas  in- 
dicated by  rows  of  setae.  The  mesothoracic  spiracle  is  located  on  a  lobe  very  close 
to  the  prothorax.  The  praescutum  of  the  prothorax  and  that  of  the  mesothorax  are 
provided  with  a  few  small  hairs.     The  scutellum  is  marked  with  a  row  of  hairs. 

The  first  eight  abdominal  segments  are  normal  and  each  bears  a  spiracle.  The 
praescutal  area  is  more  or  less  transversely  divided,  and  its  posterior  lobe  is  marked 
with  a  few  tiny  setae.  The  scutellum  is  large  and  prominent  and  provided  with  a 
row  of  setae.  The  scutum  is  only  lateral  and  has  just  above  the  spiracle  a  tiny  seta. 
Each  epipleural  lobe  is  provided  with  a  single  seta,  and  each  hypopleural  lobe  with 
two  setse.  The  coxal  lobes  of  the  thorax  bear  several  setae,  and  those  of  the  abdomen 
a  single  seta  each.  The  last  two  abdominal  segments  are  simple  and  provided  with 
a  number  of  setae. 

Pupa  (PI.  35,  E-H).  Elongate,  about  6  ram.  long,  white.  This  pupa  is  especially 
characterized  by  the  nongeniculate  antennae  which  lie  parallel  to  the  legs.  The 
antennae  and  two  anterior  pairs  of  legs  are  directed  cephalad  instead  of  posteriad. 


Mar.4.  i9i8  WeevUs  Affecttfig  Potatoes  607 

as  in  the  Curculionidae.  Another  characteristic  is  that  the  femora  and  tibiae  of  the 
posterior  pair,  being  directed  in  the  same  manner,  are  completely  covered  by  the 
wings.  The  head  and  beak  are  elongate  and  provided  with  setigerous  tubercles 
as  follows:  One  pair  between  the  eyes  at  base,  one  pair  immediately  behind  eyes, 
two  tiny  pairs  between  eyes,  and  two  pairs  on  beak;  the  posterior  pair  being  close 
to  the  eyes,  and  the  anterior  behind  the  middle.  The  antennae  are  roughly  tuber- 
culate. 

The  prothorax  is  margined  anteriorly  by  four  pairs  of  setigerous  tubercles  and  has 
one  pair  of  discal  setae.  The  mesothorax  has  two  pairs  of  lar-ge  setigerous  tubercles 
and  a  lateral  pair  of  tiny  setae.  The  femora  bear  two  or  three  setae.  The  knees  of 
the  posterior  femora  are  visible  dorsal ly  only. 

The  mesothorax  bears  three  pairs  of  small  setae  between  the  bases  of  the  elytra. 
The  metathorax  is  provided  with  two  rows  of  setae  on  tiny  tubercles,  the  anterior  row 
having  two  pairs  and  the  posterior  row  six  pairs.  The  abdominal  segments  have 
dorsally  five  pairs  of  setigerous  tubercles  near  the  posterior  margin,  a  pair  of  tiny 
setae  near  the  middle  of  the  segment,  and  lateral  setigerous  tubercles. 

The  ninth  segment  is  provided  with  two  large  curved  processes.  The  tenth  seg- 
ment is  ventral  to  the  ninth. 

Cylas  turcipennis  Boheman  (1833)  (PI.  31,  B) 

Cylcs  turcipenfiis  'Bohew.a.B,  1833,  f«  Schonherr,  Gen.  etSpec.  Cure.  v.  i.  p.  369-370. 

The  brief  preliminary  diagnosis  of  the  species  presented  by  Boheman 
is  as  follows : 

Elongatus,  viridi-coerulescens,  nitidus,  antennis  thorace  pedibusque  rufis,  capite 
cruciatim  impresso,  rostro  punctulato,  elytris  modice  convexis,  subtiliter  striato- 
punctatis.     Habitat  in  Java,  in  India  orientali. 

The  following  dimensions  are  included  in  the  detailed  description:  Length  3  lines 
(6  mm.);  antennae  as  long  as  thorax  and  head;  club  of  male  antenna  longer  than  pre- 
ceding joints;  beak  not  longer  than  head;  elytra  twice  as  wide  as  thorax  at  base,  and 
twice  longer  than  wide.  The  color  description  is  as  follows:  Head  obscurely  viridi- 
cceruleous;  beak  almost  black;  antennae  rufo-ferruginous;  thorax  shining  rufous; 
elytra  coerulescent-virescent;  thorax  beneath  rufous,  remainder  of  body  beneath 
coeruleo-virescent;  legs  rufous;  tarsi  beneath  fulvous,  spongy;  female  with  femora 
in  middle  annulate  virescent. 

Two  specimens  from  Palembang,  Sumatra,  collected  by  Mr.  M.  Knappert, 
are  here  considered  as  this  species.  They  differ  only  in  having  the  beak 
slightly  longer  than  the  head,  and  a  statement  to  th\s  effect  might  have 
been  made  if  the  description  had  been  based  on  examination  with  a  low- 
power  lens.  Two  other  specimens  are  at  hand  from  Bay  Laguna  Prov- 
ince, Philippine  Islands,  collected  by  Mr.  P.  L.  Stangl.  A  part  of  a 
body  of  a  weevil  from  Guatemala,  collected  by  Mr.  D.  G.  Eisen,  is  also 
undoubtedly  this  species.  Pascoe  records  the  species  from  Sarawak, 
Java,  and  India. 
Cylas  femoralis  Faust  (1898).  (PI.  31,  F) 

Cylasfc-moralis  Faust,  1899,  in  Deut.  Ent.  Ztschr.,  p.  24. 

This  species  was  collected  by  Mr.  Rolla  P.  Currie  at  Mount  Coffee,  Liberia, 
in  February  to  April,  1897,  and  he  has  informed  the  writer  that  it  was  a 
serious  sweet-potato  pest  in  that  country.  It  is  described  from  Kamerun. 
In  the  Manual  of  Dangerous  Insects  ^  this  species  was  referred  to  as 
C.  brunneus  by  mistake. 

'Pierce,  W.D.  a  m.\nuai,  of  dangerous  insects  ...  p.  209.  1917.  Published  by  the  United  States 
Department  of  Agriculture,  OfKce  of  Secretary. 


6o8  Journal  of  Agricultural  Research  voi.  xii.Ng.g 

FAMILY    OROBITIDAE    PIERCE    (1916) 

Subfamily  Orobitinae  Pierce  (19 16) 

genus  euscepes  schonherk  (1844) 

Euscepes  Schonherr,  1844,  Gen.  et  Spec.  Cure,  v.  8,  pt.  i,  p.  4:9-  Type— porcellus  Boh.  by  original  desig- 
nation. 

Euscepes  Lacordaire,  1866,  Gen.  Coleop.,  v.  7,  p.  loo-ioi.     Type. — porcellus  Boh. 

Hyperonwrpha  Blackburn,  1885,  in  Sci.  Trans.  Roy.  Dublin  Soc,  s.  2,  v.  3,  p.  182-183.  "Typt— {squamosa. 
Blackbum)=fcato<ae  Waterhouse. 

Euscepes  Champion,  1905,  in  Biol.  Centr.-Amer.,  Coleopt.,  v.  4,  pt.  4,  p.  496-49S.    Type.— porcellus  Boh. 

Lacordaire  caused  a  confusion  of  genera  by  wrongly  interpreting  the 
number  of  funicular  joints,  of  which  there  are  seven.  This  error  was 
corrected  by  Champion.  The  two  genera  Euscepes  and  Hyperomorpha 
are  strictly  congeneric ;  in  fact,  the  two  type  species  differ  principally  in 
size.  A  large  series  of  porcellus  from  various  parts  of  Central  America  is 
at  hand.  These  have  been  carefully  compared  w^th  Blackburn's  descrip- 
tion of  Hyperomorpha,  but  no  generic  difference  can  be  found. 

The  rostral  canal  extends  along  the  prosteraum  and  ends  in  a  meso- 
stemal  pocket.  The  beak  when  at  rest  fits  tightly  into  this  canal.  The 
prothorax  is  lobed  to  cover  the  eyes  when  at  rest.     The  body  is  elongate. 

Euscepes  batatae  C.  O.  Waterhouse  (1849),  ^Iie  Scarabee  of  the  Sweet  Potato  (PI.  32, 

C,  D;  PI.  Z2,>  A-D;  PI.  34,  E-H) 
CryptorhynchusbataiaeWnterhouse,  1849,  in  Trans.  Ent.  Soc,  London,  v.  5,  p.  LXix. 
Hypercjiwrpha  sguamosa  Blackburn,  1885,  in  Sci.  Trans.  Roy.  Dublin  Soc,  s.  2,  v.  3,  p.  182-183. 
Euscepes  bataiae  Champion,  1905,  in  Biol.  Centr.-Amer.,  Coleopt.,  v.  4,  pt.  4,  p.  497. 

This  weevil  (Pi.  32,  C,  D)  is  one  of  the  most  serious  cosmopolitan  pests 
of  the  sweet  potato,  although  hitherto  it  has  been  recorded  only  from 
Barbados,  St.  Vincent,  and  Antigua,  St.  Kitts,  Nevis,  and  Hawaii.  In 
all  of  these  places,  however,  it  is  reported  as  damaging  sweet  potatoes. 
The  receipt  of  tvv^o  specimens  from  Dr.  Da  Costa  Lima,  of  Brazil,  with 
the  statement  that  they  were  injuring  sweet  potatoes  at  Rio  de  Janeiro, 
caused  the  writer  to  make  a  search  through  the  undetermined  collections 
of  the  National  Museum  with  the  result  that  the  known  distribution  of 
the  species  is  hereby  greatly  extended.  Specimens  are  at  hand  from 
Barbados,  injuring  sweet  potatoes  May  22,  1900,  and  more  recent 
material;  Hope,  Kingston,  Jamaica,  on  sweet  potatoes,  Mr.  S.  F.  Ashby; 
Campinas,  Brazil,  injuring  sweet  potatoes,  August,  191 3,  Mr.  A.  Hempel 
(No.  100);  Rio  de  Janeiro,  Brazil,  injuring  sweet  potatoes,  July,  1917, 
Carlos  Moreira;  Honolulu,  Oahu,  Hawaii,  bred  from  sweet  potato ;  Kaimuki, 
Oahu,  Hawaii,  bred  from  sweet  potato;  Guam,  on  sweet  potato,  Mr. 
®.  T.  Fullaway;  Norfolk  Island,  New  Zealand,  March,  1883,  Mr.  P.  H. 
Metcalfe;  Mayaguez,  Porto  Rico,  injuring  sweet  potatoes,  1912,  1914, 
1 91 7,  Mr.  C.  W.  Hooker,  Mr.  R.  H.  Van  Zwaluwenburg. 

This  extensive  distribution  indicates  that  there  are  probably  many 
other  countries  where  sweet  potatoes  are  grown  that  may  have  the 
weevil.  If,  fortunately,  it  should  prove  to  be  absent  in  other  countries, 
rigid  quarantines  should  be  put  into  effect,  such  as  that  recently  estab- 


Mar. 4.  I9I&  Weevils  Affecting  Potatoes  609 

lished  by  the  United  States.  In  fact  it  has  been  only  because  of  tke 
excellent  system  of  quarantine  inspection  in  California  that  the  species 
has  not  already  come  into  the  United  States  with  Hawaiian  potatoes. 
Many  shipments  of  infested  sweet  potatoes  have  already  been  inter- 
cepted at  the  California  ports.^ 

The  following  is  a  description  of  the  species  redrawn  to  include  all 
the  material  at  hand.  The  variations  of  color  will  be  mentioned  in 
subsequent  paragraphs. 

Length  about  4  mm.  Brown,  mottled  with  lighter  areas,  especially  by  a  transverse, 
irregular,  postmedian  band  on  the  elytra.  Squamose,  bristling  with  upright  setse. 
Beak  curved,  carinate  and  laterally  bifurcate.  Front  foveate.  Front  and  beak 
bristling  with  erect  scales.  Vertex  provided  with  more  decumbent  scales.  Prothorax 
constricted  in  front,  laterally  impressed  on  disk  behind,  mottled  with  erect  scales, 
except  on  posterior  margin  which  is  provided  with  smaller,  more  decumbent  scales. 
Elytral  strise  composed  of  rather  distant  punctures,  each  bearing  a  small  scale ;  stu^ace 
closely  set  with  overlapping  scales  and  each  interspace  with  a  single  series  of  elongate 
squamiform  seta.  Undersides  more  sparsely  clad  with  semierect  scales.  Legs 
provided  with  scales  and  seta.  Rostral  canal  deep,  terminating  in  a  prominent 
pocket  of  the  mesostemum.  Intercoxal  process  broad,  angulate  on  anterior  margin. 
First  segment  behind  coxae  subequal  to  the  second,  which  is  but  slightly  longer  than 
the  subequal  third  and  fourth  segments.     The  femora  are  minutely  toothed. 

• 

Mr.  Bradford's  excellent  illustrations  will  be  very  helpful  in  identifying 

this  weevil.     His  illustration  of  the  adult  is  from  Hawaiian  material. 

The  species  varies  from  very  light  brown  to  almost  black  and  on  the  darkest  speci- 
mens the  mottling  and  the  postmedian  vitta  have  practically  disappeared. 

On  light  specimens  the  scales  of  the  thorax  are  mostly  dark  brown,  with  flecks  of 
pale  scales  and  with  the  basal  scales  orange  colored.  The  scales  of  the  eljixa  are 
mottled  in  many  shades  of  brown.  The  postmedian  fascia  extends  to  the  fifth  inter- 
space and  is  bordered  by  very  dark  scales  and  divided  by  a  wa\^  dark  line.  The 
erect  setae  are  mixed  dark  and  white.  The  ventral  scales  are  pale,  but  on  the  legs 
they  are  mottled  dark  brown  and  pale.  This  description  fits  best  some  of  the 
Jamaican,  Barbados,  and  Brazilian  specimens.  Almost  black  material  comes  from 
Brazil,  Jamaica,  and  Guam.  The  Hawaiian  specimens  are  a  duller  brown,  and  the 
New  Zealand  material  is  the  lightest  of  all.  There  is,  however,  no  doubt  of  the  specific 
identity  of  the  entire  series. 

In  order  that  the  immature  stages  may  be  readily  distinguished  from 
those  of  Cylas  jormicarius  a  series  of  very  careful  drawings  of  the  essen- 
tial characters  of  the  larva  and  pupa  have  been  made  by  Mr.  Harry 
Bradford  under  the  writer's  direction.  The  drawings  of  the  vertex  and 
face  are  by  the  writer.     Barbados  material  was  used  for  these  drawings. 

Larva  (PI.  34,  E-H).  The  larva  of  this  species  measures  about  5  mm.  in  length 
and  is  white,  with  a  yellowish  head  and  reddish  brown  mandibles  tipped  with  black. 
The  maxillae  and  labium  are  slightly  tinged  with  brown. 

The  head  shield  is  broadly,  angulately  emarginate  behind;  from  the  center  of  the 
emargination  on  the  median  line  the  epicranial  suture  passes  forward,  separating  the 
epicranium  into  two  parts.     This  suture  divides  behind  the  frons  and  forms  the  two 

'  Whitney,  L.  A.  the  small  sweet  potato  weevil  (crvptorhynchus  batatae  waterh.).  In  Mo 
Bui.  State  Com.  Hort.  [Cal.],  v.  4,  no.  3,  p.  162-164,  fig.  24-28.    1915. 


6jo  Journal  of  Agricultural  Research  voi.  xii,  no.  9 

fj-ontal  sutures.  The  frons  is  sub  triangular,  rounded  at  anterior  angles,  and  slightly 
emarginate  for  antennae;  its  front  margin  is  the  epistoma.  There  are  three  pairs  of 
large  setae  on  the  frons,  the  posterior  pair  being  located  rather  close  together  and  near 
the  apex  of  the  triangle.  The  second  pair  are  farther  apart  and  halfway  to  the  front. 
The  anterior  pair  are  located  near  the  lateral  angles  just  behind  the  antennae.  A  tiny 
pair  of  setae  are  located  just  in  front  of  the  posterior  pair  of  large  setae.  A  tiny  pair 
are  located  slightly  behind  and  outside  of  the  median  pair  of  large  setae.  The  median 
line  of  the  frons  is  impressed  from  the  posterior  angle  almost  to  the  middle. 

The  epicranial  areas  are  the  two  large  areas  on  each  side  of  the  epicranial  suture 
further  bounded  by  the  frontal  suture,  the  pleurostoma,  and  the  hypostoma.  The 
following  setae  occur  on  each  lobe  of  the  epicranium:  One  opposite  the  apex  of  the 
frons,  one  on  tlie  disk  of  the  epicranium  in  the  line  with  the  preceding;  one  opposite 
the  middle  of  the  frons  and  near  the  frontal  suture;  one  near  the  hypostoma  toward 
the  base  of  the  mandible;  one  as  close  to  the  hypostoma  and  opposite  the  base  of  the 
maxilla;  one  near  the  pleurostoma;  the  last  three  forming  a  triangle.  Forming  a 
semicircular  line  with  the  setse  opposite  the  frons  are  a  tiny  seta,  a  longer  one  opposite 
the  antenna,  and  a  long  one  opposite  the  pleurostomal  seta. 

The  antenna  is  a  small,  fleshy,  two-jointed  appendage  at  the  angle  of  the  frontal 
suture  and  pleurostoma.  The  mandibles  are  bluntly  bidentate  and  have  two  small 
setae.  The  clypeus  is  attached  in  front  of  the  frons  and  is  broadly  transverse.  It 
bears  on  the  epistomal  margin  four  tiny  hairs.  The  labrum  is  not  as  broad;  it  has  a 
pair  of  median  setae  and  three  pairs  of  marginal  setae.  The  maxillae  attached  at  the 
side  of  the  mandibles  are  terminated  by  a  two-jointed  palpus  and  a  setose  lacinia. 
The  maxillae  are  provided  with  three  setae,  two  near  the  palpi  and  one  toward  the  base. 
The  stipes  labii  is  appendiculate,  bilobed,  bearing  two-jointed  palpi  and  a  single  pair 
of  posterior  setae.  Each  lobe  of  the  mentum  is  provided  with  one  seta.  The  thoracic 
segments  are  simple,  being  composed  of  only  praescutum  and  scutoscutellum.  In 
the  prothorax  there  is  no  separation  of  these  parts,  but  they  are  indicated  by  the 
arrangement  of  the  setae.  The  abdominal  segments  are  dorsally  composed  of  four 
single  sclerites,  namely,  praescutum,  scutum,  scutellum,  and  postscutellum. 

The  thoracic  spiracle  is  located  on  a  lobe  of  the  meso thorax,  very  close  to  the  head. 
The  abdominal  spiracles  are  located  on  a  small  lobe  at  the  side  of  the  scutum  on  the 
first  eight  segments.  The  ninth  and  tenth  abdominal  segments  are  smaller  and 
considerably  modified. 

Setae  are  arranged  as  follows :  Four  pairs  of  tiny  hairs  on  anterior  margin  of  pro- 
thoracic  praescutal  area;  a  row  of  longer  hairs  on  the  scutal  area  of  the  prothorax; 
each  segment  from  the  mesothorax  back  with  two  pairs  of  long  lateral  setae,  near 
which  are  located  smaller  and  inconspicuous  setae;  each  praescutum  with  a  single 
pair  of  setae;  on  each  segment  of  the  abdomen  the  lateral  lobe  of  the  scutellum  is 
provided  with  a  tiny  seta,  behind  and  close  to  the  spiracle;  each  epipleural  lobe 
below  the  spiracle  with  two  setse;  each  coxal  lobe  with  several  setas  which  are  more 
conspicuous  on  the  prothorax. 

Pupa  (PI.  33,  A-D). — Length  4  mm.,  white.  This  is  a  normal,  characteristic 
ciu-culionid  pupa  with  geniculate  antennae  and  legs  turned  posteriad.  Head  oval, 
beak  short.  The  head  bears  four  pairs  of  basal  setigerous  tubercles,  two  pairs  of 
interocular  tubercles,  and  one  pair  of  tiny  setae  at  base  of  beak.  Prothorax  with 
two  pairs  of  antero-marginal  setigerous  tubercles;  one  pair  of  antero-lateral  and  one 
pair  of  postero-lateral  setigerous  tubercles;  four  pairs  of  dorsal  tubercles  and  one 
pair  of  ventral.  The  femora  are  apically  armed  with  two  setae.  Mesonotum  and 
metanotum  each  provided  with  two  pairs  of  setae.  Each  abdominal  segment  bears 
four  dorsal  and  one  or  more  lateral  setae.  The  ninth  segment  is  armed  with  two  very 
large  processes.  The  tenth  segment  is  very  small  and  located  on  the  venter  of  the 
ninth. 


Mar. 4,  i9is  Weevils  Affecting  Potatoes  6ii 

A  WEEVIL  WHICH  ATTACKS  THE  TUBERS  OF  YAMS 

Palaeopus  dioscoreae,  n.  sp.  (PI.  32,  E,  F) 

Described  from  two  specimens  reared  from  tubers  of  Dioscorea  batatas, 
Hope,  Kingston,  Jamaica,  by  Mr.  S.  F.  Ashby  in  April,  191 4.  Belongs 
in  the  same  subfamily  as  Euscepes. 

Length  4.5  mm.,  breadth  1.75  mm.  Piceous  black,  with  reddish  brown  append- 
ages. It  is  sparsely  clad  with  dark  brown  or  whitish  oblong  decumbent  scales 
and  erect,  longer  truncate  scales  of  variable  color.     The  punctuation  is  very  coarse. 

Head  smooth,  beak  separated  from  head  by  strong  transverse  constriction.  Beak 
longitudinally  five-striate  and  bristling  in  basal  half  with  erect  brown  scales;  apical 
half  smoother,  with  confluent  pimctures  and  flat  scales.  Scrobes  beginning  beyond 
middle,  diagonal,  reaching  eyes  beneath.  Eyes  lateral,  separated  by  width  of  beak, 
covered  when  at  rest  by  pronotal  lobes.  Antennal  funicle  seven-jointed,  first  joint 
a  little  longer  than  second.  Pro  thorax  broad,  depressed,  convexly  rounded  on 
sides,  bisinuate  at  base,  lobed  over  eyes,  trtmcate  at  apex  which  is  about  half  as 
wide  as  base;  median  line  broadly  elevated;  punctuation  very  coarse.  Elytra  10- 
striate,  the  tenth  striae  abbreviated;  strial  ptmctures. large,  rounded,  well  separated, 
and  setigerous;  interspaces  not  wider  than  striae,  clad  with  a  single  row  of  erect  squa- 
mose  setae  of  variable  color.  Scutellum  indistinct.  Elytra  broader  at  base  than 
thorax;  with  distinct  humeri. 

Sternal  canal  deep,  sharply  margined,  limited  by  a  cuplike  depression  of  the 
mesostemum.  Posterior  coxae  very  broadly  separated.  Metastemum  at  middle  as 
long  as  first  abdominal  segment  behind  coxae.  Intercoxal  piece  of  first  abdominal 
segment  angulate  at  middle.  Second  segment  not  as  long  as  third  and  fourth 
together.  * 

Femora  dentate  and  canaliculate  beneath.  Tibiae  curved  at  base,  strongly  hooked 
at  the  apex.     Tarsal  claws  simple. 

Type. — Cat.  No.  21612,  United  States  National  Museum. 
38324°— 18— 7 


PLATE  28 

Trypopremnon  sanfordi:  Adult  from  Cuzco,  Peru 

A. — Dorsal  view.    Actual  length  8.025  mm. 

B. — Face  of  same.    Actual  length  of  head  and  beak  3.5  mm. 

C. — Side  view  of  thorax  and  head. 

D. — Ventral  view  of  adult. 

Drawn  by  Mr.  H.  B.  Bradford. 

(612) 


Weevils  Affecting  Potato  and  Yam 


PLATE  28 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  9 


Weevils  Affecting  Potato  and  Yam 


Plate  29 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  9 


PLATE  29 

Trypopremnon  latithorax:  Larva  from  La  Paz,  Bplivia 

A. — Prothoracic  spiracle. 

B. — Larva,  lateral  view. 

C. — Lateral  view  of  head. 

D. — Right  side  view  of  apex  of  labium. 

E. — Corresponding  hair  on  left  side. 

F. — Maxillary  palpiger  and  palpus,  lateral  view. 

G. — Face. 

Drawn  by  the  author. 


PLATE  30 

Trypbpremnon  laiiihorax:  Pupa  from  La  Paz,  Bolivia 

A. — Dorsal  view. 

B. — Ventral  view. 

C. — Enlarged  sketch  of  eighth,  ninth,  and  tenth  abdominal  segments. 

Drawn  by  Mr.  H.  B.  Bradford. 


Weevils  Affecting  Potato  and  Yam 


Plate  30 


Journal  of  Agricultural  Research 


Vol. XII,  No.  9 


Weevils  Affecting  Potato  and  Yam 


Plate  3 1 


Journal  of  Agricultural  Research 


Vol. XII,  No. 9 


PLATE  31 

Species  of  the  genus  Cylas: 

'  A. — Cylas  formicarius  elegantulus  from  Honolulu,  Hawaii,  side  view  of  head  and 
thorax. 

B. — Cylas  iurcipennis  from  Sumatra,  side  view  of  head  and  thorax. 

C. — Cylas  brunnetis  from  East  Africa,  dorsal  view  of  thorax. 

D. — Cylas  brunneus,  side  view  of  head  and  thorax. 

E. — Cylas  brunneus,  ventral  view  of  thorax. 

F. — Cylas  femoralis ,  side  view  of  head  and  thorax. 

The  abbreviations  used  on  this  plate  are  as  follows:  pr,  Presegmental  ring;  sc, 
scutum;  si,  scutellum;  psl,  postsegmental  ring;  pi,  pleurum;  in,  trochantin;  c,  coxa; 
bs,  basistemite;  stl,  stemellum;  cstl,  centrostemellum. 

Drawn  by  the  author. 


PLATE  32 

Sweet-potato  and  yam  weevils: 

A. — Cylas  formicarius  eleganfulus,  female,  from  sweet  potatoes,  New  Orleans,  La. 

B. — Same,  head  of  male. 

C. — Euscepes  batatae,  from  sweet  potatoes,  Hawaii. 

D. — Same,  side  view  of  head. 

E. — Palaeopus  dioscoreae,  from  yams  {Dioscorea  batatas),  Jamaica. 

F. — Same,  side  view  of  head. 

Drawn  by  Mr.  H.  B.  Bradford. 


Weevils  Affecting  Potato  and  Yam 


Plate  32 


Journal  of  Agricultural  Research 


Vo'.  XII,  No.  9 


Weevils  Affecting  Potato  and  Yam 


Plate  33 


Journal  of  Agricultural  Researcn 


Vol.  XII,  No.y 


PLATE  33 

Pupae  of  sweet-potato  weevils : 

A. — Euscepes  batatae,  Barbados,  venter  (length  4  mm.). 
B. — Same,  latere- ventral  view  of  fifth  to  tenth  segments. 
C. — Same,  dorsal  view. 

D. — Same,  venter  of  seventh  to  tenth  segments  (length  of  this  portion  i  mm.). 
E. — Cylas  formicarius  elegantulus ,  Victoria,  Texas,  ventral  view  of  sixth  to  tenth 
segments  (length  of  this  portion  i  mm.). 
F. — Same,  ventral  view  (length  6  mm.). 
G. — Same,  latero- ventral  view. 
H. — Same,  dorsal  view. 
Drawn  by  Mr.  H.  B.  Bradford. 


PLATE  34 

Larvae  of  sweet-potato  weevils : 

A. — Cylas  formicaritis  elegantulus,  Victoria,  Texas,  lateral  view. 

B. — Same,  dorsum  of  head. 

C. — Same,  face. 

D. — Same,  side  of  head. 

E. — Euscepes  batatae,  Barbados,  dorsum  of  head. 

F. — Same,  face. 

G. — Same,  side  of  head. 

H. — Same,  lateral  view  of  larva. 

Drawn  by  Mr.  H.  B.  Bradford. 


Weevils  Affecting  Potato  and  Yam 


PLATE  34 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  9 


^ 


vtmBKammnammum 


Vol.  XII  IVIARCM  11,  191S  No.  lO 

JOURNAL  OF 

AGRICULTURAL 
RESEARCH 


CONXENXS 

Page 

Sterility  in  the  Strawberry        -        -        -        -        -        -613 

W.  D.  VALLEAU 

( Contrlbutigo  bom  Minnesota  Agrictilttiral  Exj>etimeat  Station ) 

Effect  of  Nitrifying  Bacteria  on  the  Solubility  of  Trical- 

cium  Phosphate  -       -       -        -        -       -        -671 

W.  P.  KELLEY 
(Contiibation  from  California  Agricultura)  Bzpezbneol:  Station) 


PUBLISHED  BY  AUTHORITY  OF  THE  SECRETARY  OF  AGRICULTURE, 

WITH  THE  COOPERATION  OF  THE  ASSOCIATION  OF  AMERICAN 

AGRICLXTURAL  COLLEGES  AND  EXPERIMENT  STATIONS 


WASHINOXON,  D.  C. 


WAtHIKOTON  t  OOVERMMENT  PRINTINO  Omce  :1916 


EDITORUL  COMMITTEE  OF  THE 

UNITED    STATES    DEPARTMENT    OF    AGRICULTURE   AND 

THE  ASSOCIATION  OF  AMERICAN  AGRICULTURAL 

COLLEGES  AND  EXPERIMENT  STATIONS 


FOR  THE  DEPARTMENT 

KARL  F.  KEI/LERMAN,  Chairman 

Physiologist  and  Associate  Chief ,  Bureau 
of  Plant  Industry 

EDWIN  W.AXtEN 

Chief,  Ojfflce  of  Experiment  Stations 

CHARLES  L.  MARLATT 

Entomologist  and  Assistant  Chief,  Bureau 
of  Entomology 


FOR  THE  ASSOCIATION 
RAYMOND  PEARL* 

Biologist,  Maine  Agrictdtural  ExperinunI 
Station 

H.  P.  ARMSBY 

Director,  Institute  of  Animal  Nutrition,  The 
Pennsylvania  State  College 

E.M.  FREEMAN 

Botanist,  Plant  Pathologist  and  Assistant 
Dean,  Agricultural  Experiment  Station  of 
the  University  of  Minnesota 


All  correspondence  regarding  articles  from  the  Department  of  Agriculture  should  be 
addressed  to  Karl  F.  Kellerman,  JoiU"nal  of  Agricultural  Research,  Washington,  D.  C. 

*Dr.  Pearl  has  undertaken  special  work  in  connection  with  the  war  emei^ency; 
therefore,  until  further  notice  all  correspondence  regarding  articles  from  State  Experi- 
ment Stations  should  be  addressed  to  H.  P.  Armsby,  Institute  of  Animal  Nutrition, 
State  College,  Pa. 


JOINAL  OF  AGRiaiLTIIAL  RESEARCH 

Vol.  XII  Washington,  D.  C,  March  ii,  1918  No.  10 


STERILITY  IN  THE  STRAWBERPvY  ^ 

By  W.  D.  Valleau 

Research  Assistant  in  Fruit-Breeding  Investigations,  Agricultural  Experiment  Station 

of  the  University  of  Minnesota 

INTRODUCTION 

This  paper  is  a  report  of  studies  on  the  sex  condition  in  strawberries 
(Fragaria  spp.)  which  have  been  carried  on  during  the  past  four  years. 
The  study  of  pistil  steriHty  and  anther  abortion  in  the  cultivated  varieties 
and  wild  species,  which  are  the  result  of  the  strong  tendency  of  this 
genus  toward  dieciousness,  has  received  considerable  attention;  but  the 
primary  object  of  the  investigation  has  been  to  find,  if  possible,  some 
satisfactory  explanation  for  the  phenomenon  of  pollen  abortion,  which 
is  so  prevalent  among  heterozygous  plants  or  plants  of  hybrid  origin.^ 

MORPHOLOGY  OF  THE   FLOWER  PARTS 

INFLORESCENCE 

The  inflorescence  of  our  cultivated  strawberry  and  of  Fragaria  vir- 
giniana,  which  it  closjely  resembles,  is  a  dichasial  cyme  or  sometimes, 
especially  in  certain  cultivated  varieties,  a  pleiochasium.  The  two 
lateral  branches  of  the  relatively  main  axis  are  not  always  equal  either 
in  size — that  is,  number  of  flowers  borne — or  in  time  of  flowering.  The 
pedicel  of  the  primary  flower  is  generally  inserted  a  short  distance  from 
the  joint  of  the  two  secondary  branches  and  on  the  smaller  of  the  two. 
The  primary  flower  of  the  largest  lateral  branch  usually  opens  directly 
after  the  primary  flower  and  before  that  of  the  smaller  lateral  branch. 
The  arrangement  of  the  flowers  and  order  of  blooming  are  shown  in 
figure  I. 

Variations  from  these  types  are  not  uncommon.  In  some  cases  the 
primary  flower  is  lacking ;   in  others,  the  primary  stalk  seems  to  be  made 

1  Published,  with  the  approval  of  the  Director,  as  Paper  94  of  the  Journal  Series  of  the  Minnesota  Agri- 
cultural Experiment  Station. 

2  The  work  was  begun  at  the  suggestion  of  Dr.  M.  J.  Dorsey,  of  the  Minnesota  Agricultural  Experiment 
Station ,  and  I  wish  to  express  my  thanks  for  the  help  and  encouragement  which  he  gave  diuing  the  progress 
of  the  work.  I  also  wish  to  express  my  appreciation  of  the  assistance  given  by  Dr.  C.  E.  Allen,  of  the 
Department  of  Botany,  University  of  Wisconsin,  in  a  portion  of  the  cytological  studies  and  for  space 
kindly  furnished  me  in  his  laboratory  during  a  month  in  1915. 

Journal  of  Agricultural  Research.  '  Vol.  XII,  No.  10 

Washington,  D.  C.  Mar.  11,  1918 

mj  Key  No.  Minn,  aj 

(613) 


6i4 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  lo 


up  of  two  which  separate  at  varying  distances  from  the  ground,  thus 
producing  two  typical  cymes  from  one  main  stalk.  The  peduncle  and 
pedicel  lengths  vary  greatly  within  a  variety,  but  there  is  apparently  a 
rather  constant  negative  correlation  between  the  two — that  is,  as  the 
peduncle  or  primary  stalk  decreases  in  length  the  pedicels  or  secondary 
branches  increase,  resulting  in  a  rather  constant  ratio  between  fruit 
stalk  length  and  leaf  petiole  length  or  height  of  plant. 

Opposite  the  point  of  insertion    of  the  small  secondary  branch  is 
usually  a  large   bract.     This  may  be  and  usually  is  in  the  wild  forms 

a  monophyllous  leaf,  while  in  manv 
m  01  the  cultivated  vaneties  it  may 
be  a  well-developed  di-  or  tri-phyl- 
lous  leaf.  The  bracts  or  bracteoles 
subtending  the  branches  of  lesser 
degree  are  usually  only  rudimentary 
structures,  being  made  up  of  the 
stipules  often  much  reduced,  but 
with  an  occasional  slight  broaden- 
ing of  the  midvein  to  form  a  small 
leaf  blade. 

The  flowers  are  hypogynous,  and 
typically  pentamerous  with  regard 
to  all  parts  except  the  carpels. 
The  perianth  consists  of  three 
whorls  of  members,  the  outer  five 
epicalyx  lobes  alternating  with  the 
five  sepals  and  opposite  to  the  five 
white  petals. 

STAMENS 

Stamen  arrangement. — The 
stamens  in  typical  flowers  are  ar- 
ranged in  multiples  of  five.  The 
number  is  not  constant  in  pure 
species  or  within  a  variety.  The 
stamens  are  arranged  in  three 
whorls.  The  outer  consists  of  lo 
parapetalous  stamens  located  at 
either  side  of  the  base  of  the  petals  (fig.  2,  a,  PP).  These  have  long 
filaments.  Their  number  is  the  most  constant  of  any  of  the  whorls.  The 
second  whorl  consists  of  five  antipetalous  stamens  located  opposite  the 
petals  and  inside  of  the  parapetalous  whorl  (fig.  2,a,AP).  The  filaments 
are  shorter  than  those  of  the  outer  and  inner  whorls.  The  third  whorl 
consists  of  five  antisepalous  stamens  inserted  opposite  the  point  of  inser- 
tion of  the  sepals,  and  inside  of  the  two  other  whorls  (fig.  2,  a,  AS). 


0  PR/MARY  AXIS 
P/^IMAfiy  FL. 


'SECONDARY 


«7- 


'TE.RTIARY 


43 

Fjg.  1. — Diagram  showing  the  arrangement  of 
flowers  of  the  strawberry  and  the  order  of  blos- 
soming. The  approximate  order  of  opening  is 
indicated  by  the  figures. 


Mar.  II,  i«i8  Sterility  in  the  Strawberry  615 

Variations  in  stamen  number  from  the  above  arrangement,  if  slight, 
are  usually  due  to  the  addition  or  loss  of  one  or  more  stamens  from  the 
antipetalous  series.  If  a  definite  increase  of  five  takes  place,  it  may  be 
the  result  of  an  increase  in  one  of  two  places:  either  the  5  single  anti- 
petalous stamens  may  be  replaced  by  5  pairs  to  form  a  whorl  of  10  (fig.  2, 
b,  AP),  or  the  5  antisepalous  stamens  may  have  been  replaced  by  5  pairs 
of  parasepalous  stamens  located  at  the  same  points  as  the  whorl  of  5 
(fig.  2,  c,  PS).  A  further  increase  in  the  number  of  antipetalous  stamens 
may  consist  in  the  development  of  a  pair  located  on  either  side  and 
slightly  inside  of  the  single  antisepalous  stamens  (fig.  2,  d,  AP).  These 
are  characterized  by  the  same  short  filament  that  is  found  in  the  anti- 


Fig.  2. — Flower  diagrams  of  Fragaria  spp.,  showing  stamen  arrangement:  A.S  represents  antisepalous; 
AP,  antipetalous;  PP,  parapetalous;  and  PS,  parasepalous  stamens,  a  represents  the  20-stamen 
arrangement  found  commonly  in  F.  vhginiana  and  F.  americana  and  many  cultivated  varieties;  6  and  c, 
a  25-stamen  arrangement  found  in  F.  virginiana  and  some  cultivated  varieties;  d,  a  35-stamen  arrange- 
ment sometimes  found  in  cultivated  varieties;  while  e  and/represent  a  10  and  15  stamen  arrangement 
found  in  some  clones  of  F.  ainericana, 

petalous  whorl  of  5.  This  increase,  plus  the  20-stamen  arrangement, 
gives  a  30-stamen  arrangement,  or  with  either  of  the  two  25-stamen 
arrangements,  gives  35. 

Rydberg  {34,  p.  loY  has  pointed  out  that  the  antipetalous  stamens  which 
appear  to  be  the  middle  whorl  in  Fragaria  spp.  are  truly  an  itmer  whorl 
which  has  been  pushed  out  to  form  apparently  a  middle  one  (fig.  2,d),  and 
that  the  outer  parapetalous  stamens  are  in  reality  younger  with  regard  to 
development  than  the  antisepalous  or  inner  series.  A  study  of  the 
position  of  the  accessory  stamens  of  the  antipetalous  series  (which  can 
readily  be  distinguished  by  their  short  filament)  and  of  the  order  of 
stamen  development  indicates  that  this  view  is  correct. 

1  Reference  is  made  by  number  (italic)  to  "Literature  cited,"  p.  666-669. 


6 16  Journal  of  Agricultural  Research  voi.  xii.  No.  lo 

In  F.  americana,  lo,  15,  and  20-stamen  arrangements  are  common. 
The  20-stamen  arrangement  is  the  same  as  that  described  for  F.  vir- 
giniana  and  the  cultivated  varieties.  When,  however,  a  decrease 
below  20  to  15  takes  place,  it  is  due  to  the  loss  of  the  short  filamented 
middle  whorl  of  antipetalous  stamens  (fig.  2,  /) — further  proof  that 
this  is  truly  an  inner  and  not  a  middle  series.  The  next  series  to  be 
lost  is  that  consisting  of  the  inner  long  filamented  antisepalous  stamens, 
thus  leaving  the  parapetalous  stamen  arrangement  (fig.  2,e).  This  seems 
peculiar,  in  view  of  the  fact  that  they  are  older  than  the  parapetalous 
stamens  and  therefore  should  remain  longest.  This  might  be  consid- 
ered as  proof  that — 
the  parapetalous  stamens  must  be  regarded  as  abnormal  supernumary  parts, 

as  Rydberg  {34,  p.  11)  considers  them.  The  genus  Fragaria  differs 
from  other  species  of  the  Potentilleae  in  this  respect,  as  the  more  usual 
order  of  loss  is  first,  the  parapetalous,  followed  by  the  antipetalous 
stamens  {34,  p.  11),  the  long  filamented  antisepalous  stamens  being 
the  most  permanent.  When  stamens  are  dropped  in  Fragaria  spp., 
they  are  lost  completely  and  do  not  form  the  staminoids  or  partially 
developed  stamens  which  are  found  in  the  pistillate  flowers,  so  that  a 
decrease  in  stamen  number  can  not  be  considered  as  a  step  toward 
dieciousness. 

Staminodia. — Typically  F.  virginiana  and,  as  will  be  shown  later, 
some  other  species  of  strawberry  are  diecious,  although  the  separation 
into  strictly  staminate  and  pistillate  forms  is  not  complete.  The  flowers 
of  pistillate  plants  bear  staminodia  showing  varying  degrees  of  develop- 
ment (fig.  3,  11-16),  which  never,  as  far  as  I  have  observed,  produce 
pollen.^  The  staminate  plants  develop  pistils  which,  as  far  as  outward 
appearances  are  concerned,  are  normal,  but  which  do  not  set  fruit.  As 
a  result  of  this  inconiplete  separation  of  the  sex-bearing  organs,  there 
are  variations  in  the  stamen  and  pistil  condition  on  individual  clones 
and  also  on  the  flowers  of  an  infloresence  within  a  clone. 

The  most  common  condition  in  the  staminate  plants  is  that  in  which 
all  of  the  flowers  produce  normal  stamens  bearing  good  pollen.  Occa- 
sionally clones  are  found  in  which  the  first  flower  bears  only  staminodia 
in  place  of  the  normal  stamens.  In  flowers  of  this  type  the  pistils  are 
practically  always  fertile  and  produce  normal  fruits.  On  the  other 
hand,  the  primary  flower  may  produce  normal  stamens  and  no  fertile 
pistils  while  one  or  both  of  the  secondary  flowers  may  be  of  the  pistillate 

1  A  possible  case  of  pollen  production  in  a  pistillate  variety  is  that  of  the  Crescent.  Plants  of  it  which 
I  have  observed 'are  strictly  pistillate,  although  producing  very  large  prominent  staminoids  (fig.  3, 16)- 
which  are  entirely  devoid  of  reproductive  tissue  (PI.  36,  A).  Castle (7  ,p.  150)  states  that  in  England  "it 
produces  perfect  flowers  and  sets  its  fruit  most  readily,  cropping  heavily  in  favorable  seasons."  As  most 
other  English  varieties  are  hermaphrodites,  it  is  very  possible  that  large  crops  might  set  as  a  result  of  cross- 
pollination,  and  that  the  presence  of  the  extremely  large  staminoids  has  been  misleading  with  regard  to 
the  exact  sexual  condition  of  these  plants. 

Fletcher  (75,  p.  132)  also  states  that  Crescent  may  vary  in  stamen  condition  becoming  "a  true  stami, 
nate  on  rich  soils,"  but  gives  no  ftuther  evidence  on  the  point. 


Mar.  II,  1918 


Sterility  in  the  Strawberry 


617 


type,  in  which  case  they  set  fruit.  A  few  cases  have  been  found  in  which 
one  side  of  a  flower  produced  normal  stamens  and  sterile  pistils  while 
the  other  half  produced  staminodia  and  set  fruit.  A  similar  range  of 
conditions  with  regard  to  stamens  has  been  noticed  in  seedlings  of  cer- 
tain cultivated  varieties.  Figures  G  and  H,  Plate  35,  are  from  photo- 
graphs of  primary  and  secondary  flowers  of  the  seedling  Minnesota  1017  X 
Progressive  32-1,  both  of  which  bear  only  staminoids,  while  I  and  J  repre- 
sent secondary  and  tertiary  flowers  of  the  same  variety,  I  producing 
both  staminodia  and  normal  anthers  and  J  producing  only  normal 
anthers.     The  production  of  pistillate  flowers  on  the  primary  and  on 


Fig.  3. — Outline  camera-lucida  drawings  of  perfect  and  intermediate  anthers  and  staminodia  of  straw- 
berry: I,  2,  and  J,  Normal  anthers  from  tertiary  flowers  of  a  seedling  of  Minnesota  1017  X  Progressive, 
Progressive,  and  another  seedlingof  Minnesota  1017  X  Progressive (40-1),  respectively.  4,  Staminodiuin 
from  a  primary  flower  of  Minnesota  loi  7  X  Progressive  (40-1 ).  5  and  6,  Normal  anthers  from  Miimesota, 
3  and  a  late  primary  flower  of  Minnesota  loi  7,  respectively.  7,  Staminodia  from  a  primary  flower  of  Minne- 
sota 1017  produced  early  in  the  spring.  8  and  p,  Intermediate  anthers  from  primary  flowers  of  Minnesota 
3.  JO,  Intermediate  anther  from  a  primary  flower  of  Progressive.  11.  Staminodia  from  a  pistillate  flower 
cf  F.virginiana.  J 2.  13,  and  1$,  Staminodia  from  pistillate  flowers  of  seedlings  of  Minnesota  1017  X  Pro- 
gressive, 2-25,  13-40,  and  i:-59,  respectively.  14,  A  staminodimn  from  a  flower  of  Haverland.a  pistil- 
late variety.  16,  Staminodia  from  a  flower  of  Crescent,  a  pistillate  variety  which  produces  extremely  large 
and  prominent  abortive  anthers. 

some  of  the  secondary  flowers  throughout  the  season  seems  to  be  the  nor- 
mal condition  in  a  number  of  seedlings  of  the  cross  1017  Minnesota  X  Pro- 
gressive (fig.  3,  3,  4),  while  among  the  commercial  varieties  there  are  a 
number  which  produce  pistillate  primary  flowers  early  in  the  flowering  sea- 
son, while  those  produced  later  are  all  perfect.  A  few  varieties  which 
show  this  peculiar  condition  early  in  the  spring,  are  Brandywine,  Minne- 
sota 3,  Bederwood,  Tennessee  Prolific,  and  Staples.  In  figure  3,  6 
represents  an  outline  drawing  of  a  normal  anther  from  a  primary  flower 
of  Minnesota  1017,  produced  late  in  the  season,  while  7  represents 
staminoids  of  the  same  variety  taken  from  flowers  produced  early  in  the 
flowering  season. 


5i8  Journal  of  Agricultural  Research  voi.  xii.no.  lo 


Although  typically  there  are  two  rather  distinct  types  with  regard  to 
stamen  development  in  both  cultivated  and  wild  clones  of  the  straw- 
berry— namely,  the  perfect  stamens  bearing  normal  pollen  and  the 
staminodia  of  the  pistillate  varieties — there  are  apparently  a  series  con- 
necting these  two  conditions. 

Figures  ii,  12,  13,  14,  15,  and  16  of  text  figure  3  show  the  range  of  de- 
velopment in  size  of  the  staminodia  on  some  pistillate  plants,  and  i,  2,  j, 
5,  and  6  show  the  relative  size  of  normal  anthers.  In  actual  size  the  two 
types  closely  approach  one  another.  With  regard  to  development  of  re- 
productive tissue,  there  is  considerable  difference.  The  normal  stamens 
naturally  carry  pollen  development  through  to  completion.  The  stami- 
nodia, on  the  other  hand,  never  produce  normal  pollen,  but  show  some 
variations  in  the  extent  to  which  development  is  carried.  Figures  A, 
B,  C,  and  D,  Plate  36,  are  photographs  of  cross  sections  of  staminodia, 
figures  A  and  B  being  cross  sections  of  two  loculi  of  staminodia  of  Crescent 
and  Columbia,  respectively,  in  both  of  which  varieties  the  staminodia  are 
rather  prominent.  Plate  36,  C,  is  from  the  seedling  Minnesota  1017  X 
Progressive,  1 1-59,  which  produces  very  large  and  prominent  staminodia. 
Here  there  are  distinct  evidences  of  early  degeneration  of  the  repro- 
ductive tissue,  probably  pollen  mother  cells,  although  the  early  stages  of 
this  variety  have  not  been  studied.  Figure  D  (PI.  36)  is  from  a  stami- 
nodium  of  a  pistillate  variety  which  produces  extremely  small  stami- 
nodia.    There  is  no  evidence  of  any  reproductive  tissue  whatever  having 

been  produced. 

Janczewski  {24)   has  studied  the  stamen  condition  in  some  of  the 

diecious  species  of  Ribes  and  finds  that  in  the  pistillate  flowers  small  sta- 
mens develop.  Their  internal  development  soon  ceases  and  abortion  of  the 
reproductive  tissue  takes  place.  He  considers  that  the  small  dark  stain- 
ing mass  which  he  found  in  the  staminodia  was  made  up  of  the  decom- 
posed pollen  mother  cells.  Often  the  cavity  left  by  the  breaking  down 
of  the  pollen  mother  cells  was  filled  with  parenchyma  which  had  grown 
in  from  the  wails.  Gates  (17)  found  the  same  condition  in  some  anthers 
of  Oenothera  lata.  Apparently  the  parenchymatous  tissue  filling  the 
staminodia  of  Crescent  and  Columbia  is  not  of  this  origin,  as  early  stages 
show  no  signs  of  the  formation  of  pollen  mother  cells. 

Intermediate  anthers. — In  studying  the  anther  types  of  F.  virginiana 
certain  clones  were  found  which  on  first  examination  appeared  to  be 
producing  normal  stamens,  but  on  closer  examination  were  found  to 
contain  either  a  dark  staining  disintegrated  mass  or  completely  aborted 
microspores,  the  walls  of  which,  in  some  cases,  were  disintegrating. 
These  are  apparently  intermediate  types  between  the  staminodia  and 
normal  anthers.  Similar  types  of  anthers  are"  not  infrequently  found 
in  the  primary  flowers  of  many  wild  staminate  clones. 

A  study  of  the  intermediate  stamens  of  F.  -virginiana  indicates  that 
pollen    development  is  generally  carried  to  the  homceotypic    division 


Mar.  II,  1918  Sterility  in  the  Strawberry  619 

or  to  the  formation  of  the  tetrads  when  degeneration  occurs.  This  be- 
comes apparent  first  through  degeneration  of  the  mother-cell  wall  and 
the  cytoplasm,  if  the  homoeotypic  division  is  taking  place,  leaving  the 
spindles  and  chromosomes  standing  out  sharply  in  this  degenerate  mass; 
or  if  the  tetrads  have  already  been  formed,  the  material  in  which  the 
microspores  are  embedded  disintegrates  and  is  follov/ed  directly  by  the 
disintegration  of  the  microspores.  Plate  36,  E,  shows  degenerating 
tetrads;  F  shows  a  later  stage  of  the  same  thing  in  which  the  micro- 
spores have  completely  degenerated;  and  G  shows  the  condition  found 
in  mature  anthers  of  this  type. 

Occasionally  development  may  proceed  to  the  formation  and  libera- 
tion of  the  microspore  when,  following  a  slight  thickening  of  the  walls, 
degeneration  of  the  contents  and  disintegration  of  the  microspore  walls 
takes  place.  The  same  type  of  degeneration  is  found  here  as  where 
earlier  abortion  takes  place.  The  walls  become  thickened  and,  as  degen- 
eration proceeds,  show  a  beaded  appearance  and  finally  break  up  into 
drops  of  a  yellow  oily  appearing  substance  which  makes  up  the  mass 
shown  in  figure  G  (PI.  36) .  In  other  clones  of  F.  virginiana  development 
proceeds  to  the  liberation  of  the  microspores  from  the  tetrad  when,  fol- 
lowing a  slight  development  of  the  microspore  wall,  degeneration  of  the 
cell  contents  takes  place,  leaving  aborted  pollen  of  the  type  so  charac- 
teristic of  hybrids. 

In  the  cultivated  hermaphroditic  varieties  which  produce  staminodia 
on  the  early  primary  flowers  (Pi.  36,  D,  G),  and  on  some  other  varieties, 
such  as  Lovett,  Glen  Mary,  and  Minnesota  1017  X  Progressive  9-24  (PI.  36, 
B,  E),  these  same  types  of  anthers  characterized  by  being  small,  shrunken, 
and  bleached  yellow  or  deep  ocher  are  common.  They  show  both  types 
of  degeneration — i.  e.,  complete  disintegration  of  the  anther  contents 
and  abortion  of  the  microspore  contents  following  their  liberation  from 
the  tetrad.  Figure  H,  Plate  36,  shows  a  section  of  a  whitish  yellow 
anther  of  the  type  shown  in  figure  3,  8  and  9,  from  a  primary  flower  of 
Minnesota  3,  a  variety  which  for  the  most  part  produces  normal  stamens. 

Jeffrey  and  others  have  recently  given  emphasis  to  the  relationship 
between  aborted  pollen  and  hybridity  and  have  attempted  to  correlate 
any  considerable  amount  of  pollen  abortion  with  a  hybrid  condition  of 
the  plant.  Apparently  in  the  strawberry  the  above  type  of  pollen 
sterility  and  the  tendency  toward  dieciousness  are  very  closely  related. 
As  all  degrees  of  stamen  development  may  be  found  on  a  single  culti- 
vated variety  of  the  strawberry,  and  on  some  wild  plants  also,  from  the 
small  staminodia  to  well-developed  stamens  bearing  normal  pollen,  it 
seems  safe  to  conclude  that  these  intermediate  stamen  types  bearing  100 
per  cent  aborted  pollen  and  found  in  apparently  pure  F.  virginiana  are 
not  the  result  of  hybridization  but  are'  really  the  expression  of  various 
degrees  of  dieciousness. 

Whether  the  clones  of  F.  virginiana  bearing  these  intermediate  anther 
types  are  able  to  develop  fruit,  thus  indicating  whether  they  have  been 


620 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  lo 


derived  from  the  pistillate  forms  continuing  pollen  development  in  the 
staminoids  beyond  the  usual  time  or  whether  they  are  staminate  forms 
in  which  pollen  development  is  inhibited  has  not  been  determined  ex- 
perimentally. However,  they  appear  to  be  of  the  latter  type  as  they 
have  been  seen  in  flower  a  number  of  times  and  have  shown  no  signs  of 
setting  fruit. 

RELATION  OF  FLOWER 
PART  NUMBER  TO 
SIZE   OF   FRUIT 

It  is  generally  rec- 
ognized by  growers 
that  toward  the  end 
of  and  in  fact  during 
the  whole  progression 
of  the  picking  season 
of  strawberries,  there 
is  a  progressive  de- 
crease in  the  size  of 
berries  produced,  but 
the  relationship  be- 
tween this  decrease 
and  the  position  of 
the  flowers  on  the  in- 
florescence producing 
these  smaller  berries 
has  not  been  so  gen- 
erally recognized.  As 
has  already  been  men- 
tioned, strawberry 
flowers  are  typically 
pentamerous,  but  un- 
der cultivation  there 
has  been  an  increase 
in  the  number  of 
parts  in  a  portion  of 
the  flowers.     This  in- 

Fig.  4.— Graphs  showing  the  relation  between  sepal  nmnber  and  flower      cj-g^sg    jg    mOSt    Strfk- 
position  in  the  seedling  varieties  No.  373. 968,  and  1006.    The  sepal     ...  _. 

number  is  indicated  on  the  abscissas  and  the  frequencies  on  the     ing  lU  the  primary  and 

ordinates.  sccoudary  flowcrs  and 

is  only  apparent  to  a  very  slight  degree  in  the  later  ones.  An  increase  in 
calyx-lobe  number  is  practically  always  accompanied  by  an  increase  in  both 
petal  and  epicalyx  lobe  number  and  as  the  stamens  are  arranged  with  regard 
to  petal  position  there  is  necessarily  an  increase  in  stamen  number  also. 
Figure  4  shows  the  relationship  between  flower  position  and  sepal  num- 
ber in  three  seedling  strawberry  varieties.     These  show  a  condition  typical 


Mar.  II,  1918 


Sterility  in  the  Strawberry 


621 


of  our  cultivated  varieties.  It  will  be  seen  that  there  is  a  very  direct 
relationship  between  flower  position  and  number  of  flower  parts.  In 
these  same  varieties  the  relationship  between  calyx-lobe  number  and 
size  of  berries  was  studied.  The  results  are  shown  in  Table  I.  These 
results  indicate  that  there  is  a  high  correlation  between  flower-part  number 
and  fruit  size,  and  as  a  definite  relationship  has  been  pointed  out  between 
flower  position  and  flower-part  number,  it  follows  that  the  larger  fruits 
will  be  developed  on  the  early  blooming  primary  flowers  and  that,  as  the 
season  progresses,  there  will  be  a  decrease  in  fruit  size  due  to  their  being 
borne  on  later-blooming  flowers  of  a  higher  order. 


Table  I. — Relation  of  fruit  size  to  calyx-lobe  number  in  strawberries 


Popu- 
lation. 

Calyx-Iobe  number. 

Diameter  of  berry. 

Variety. 

Range. 

Mean. 

Stand- 
ard 
devia- 
tion. 

Kange. 

Mean. 

stand- 
ard 
devia- 
tion. 

Coefficient  of 
correlation. 

Seedling  1006. . . . 
Seedling    373---- 
Seedling    968 

274 
288 
275 

4-8 
5-9 
5-9 

6-350 
6.827 

6-313 

0.925 

I.  040 

•956 

Cm. 

I.  4-3-  8 
I-  4-3-  5 
I.  7-3.  8 

2.326 
2.403 
2.427 

0-343 
•532 
.4361 

0.  482  ±0.  031 
•  667 ±    .  022 

•5i9±  -03 

Even  a  casual  observation  of  normal  large  and  small  berries  of  any 
variety  will  indicate  that  there  must  be  a  relationship  between  berry  size 
and  carpel  number.  Table  II  shows  the  direct  relationship  which  exists 
between  fruit  position,  size  of  fruit  and  achene  or  carpel  number  in  all  of 
the  fruits  produced  on  the  inflorescences  studied.  The  fact  that  there  is 
such  a  definite  relationship  between  size  of  berry  and  flower  position 
should  be  constantly  kept  in  mind  in  the  selection  of  breeding  stock. 
This  is  apparently  the  point  that  Mr.  Hubach,  a  southern  strawberry 
breeder,  has  in  mind  in  selecting  for  stalks  which  bear  only  one  fruit 
per  inflorescence  {see  Darrow,  9). 

Table  II. — Relationship  between  fruit  position,  number  of  achenes,  and  size  of  fruit 

in  the  strawberry 


Pr 

imary. 

Secondary. 

Tertiary. 

Quaternary. 

A-n- 

<n 

u 

fc 

u 

in 

u 

<Kv 

J3 

J3 

J3 

X 

Variety. 

^  3 

0 

Eg 

ci 

0 

is 

ng 

V 

0 

ci 

0 

0 

I/) 

J3 

w 

^ 

Vi 

t/i 

<UU 

vlj 

h 

DO 

u  0 

ti  ™ 

M 

s 

•^  C 

""S 

a 

«-, 

a 

•O 

«_ 

a 

XI 

ts_ 

ts 

C  0 

s 

S° 

S 

a 

fco 

to 

H 

ijo 

i 

S 

fco 

Ui 

> 

> 

3 

> 

> 

3 

> 

> 

> 

> 

iz  t: 

z 

< 

<; 

•z 

< 

< 

2 

< 

< 

Z 

< 

< 

A/w. 

Mm. 

Mm. 

Mm. 

Minnesota  3 

6 

7 

3^2.  28 

28.0 

14 

224.  27 

17-7 

21 

150. 9 

9.8 

7 

92.42 

7 

Wiidwood 

7 

4 

2  29-  70 

23-8 

17 

142.7 

19.0 

20 

88.15 

II.  7 

3 

72-7 

7 

F .  virginiana  9  

I 

I 

112.  00 

15.0 

2 

116.  5 

13- 5 

4 

77.0 

8-3 

I 

70.  0 

6 

622  Journal  of  Agricultural  Research  voi.  xii,  no.  io 

PISTILS 

MoRPHOivOGY. — The  carpels  bearing  a  single  orthotropous  ovule  are 
arranged  in  a  spiral  on  the  fleshy  cone-shaped  receptacle.  At  maturity 
they  form  dry  achenes  either  set  on  the  surface  of  the  receptacle,  as  in 
F.  aynericana,  or  in  shallow  or  deep  pits,  as  in  F.  virginiana.  The  style 
is  inserted  laterally  on  the  inner  side  of  the  carpel  and  extends  well  above 
the  upper  portion  of  the  ovary.  The  pistil  number  is  not  constant  on 
the  flowers  of  an  inflorescence  but  is  directly  dependent  upon  the  posi- 
tion of  the  flower. 

Pistil  sterility  and  dieciousness. — As  has  been  previously  men- 
tioned, F.  virginiana  is,  in  the  wild,  typically  diecious,  the  stamens  having 
been  reduced  to  staminodia  in  the  pistillate  plants  and  the  pistils,  although 
present  and  apparently  normal,  as  far  as  can  be  seen  superficially,  in  the 
staminate  flowers  are  nearly  always  functionless.  This  condition  of 
dieciousness  has  apparently  remained  unrecognized  by  systematists  of 
Fragaria.  Apparently  dieciousness  is  not  confined  to  F.  virginiana  alone, 
but  is  typical  of  most  of  the  American  species,  except  F.  americana, 
which  is  hermaphroditic,  as  is  also  the  European  species  F.  vesca. 

As  early  as  1760  dieciousness  was  recognized  in  F.  elatior  by  Duchesne 
{see Fletcher,  j(5),  who  showed  that  the  apparent  sterility  of  the  Hautboy 
was  in  reality  due  to  the  weeding  out  of  the  unproductive  male  plants. 
He  also  recognized  partial  separation  of  F.  chiloensis  into  male  and 
female  plants.  A  study  of  herbarium  material  of  some  of  the  American 
species  of  Fragaria  indicates  that  F.  chiloensis  from  Alaska  to  Bolivia, 
F.  cuneijolia  on  Vancouver  Island  and  in  Washington  and  F.  platypetala 
from  the  north  moraine  of  Sir  Sandford  Glacier  are  all  diecious,  at  least 
some  plants  produced  pistillate  flowers  bearing  staminodia,  while  others 
bore  flowers  with  well-developed  stamens  and  apparently  normal  pistils 
which  showed  no  signs  of  setting  fruit.  On  two  plants  of  F.  chiloensis 
from  Lake  Merced,  California,  the  primary  flowers  of  staminate  inflores- 
cences were  found  to  have  set  fruit,  while  the  remaining  flowers,  although 
well  beyond  the  fruiting  stage,  showed  no  signs  of  setting.  Supposedly 
hermaphroditic  plants  of  F.  chiloensis  from  Alaska,  grown  at  University 
Farm  for  a  number  of  years,  produced  few,  if  any,  fruits,  although  they 
blossomed  profusely. 

Georgeson  (19,  p.  ij),  in  speaking  of  F.  chiloensis,  which  he  used  in  his 

hybridization  experiments,  says : 

There  is  a  decided  variation  among  the  plants;  some  are  much  more  productive 
than  others,  and  some  appear  to  bear  only  staminate  flowers,  though,  as  a  rule,  the 
flowers  are  perfect. 

and  again  (20,  p.  11) : 

The  flowers  are  large  and  white  and  many  of  them  staminate  and  sterile. 

The  first  plants  of  F.  chiloensis  brought  to  Europe  by  Frezier  were  all 
pistillate  and  had  probably  been  selected  by  him  because  of  their  fruiting 
propensities. 


Mar.  II,  1918  Sterility  in  the  Strawberry  623 

Richardson  (jj,  p.  1-6)  mentioned  receiving  male  plants  of  F.  virgin- 
iana  from  America.  I  have  grown  plants  of  F.  virginiana  illinoensis 
from  near  Dresden,  Ohio,  which  proved  to  be  strictly  diecious.  The 
literature  on  strawberry  growing  in  this  country  during  the  early  part  of 
the  last  century  makes  many  references  to  the  necessity  of  growing 
staminate  varieties  for  the  purpose  of  fertilizing  the  pistillate  forms  and 
to  the  fact  that  these  plants  were  unproductive  of  fruit,  but  very  pro- 
ductive of  plants  and  would  soon  dominate  the  garden  if  attention  was 
not  paid  to  them.  Apparently  they  were  using  staminate  F.  virginiana 
plants  as  pollen  producers. 

A  study  of  F.  virginiana  in  various  sections  of  Minnesota  shows  that  for 
the  most  part  this  species  is  diecious,  although  some  few  hermaphroditic 
plants  may  be  found.  Of  a  total  of  1,615  pistillate  flowers  of  this  species 
borne  on  304  plants  located  in  four  distinct  regions  of  the  State,  1,180 
set  fruit,  while  393  were  still  in  the  bud  or  blossom  stage,  thus  indicating 
that  practically  all  of  the  flov/ers  of  the  pistillate  forms  are  fertile  if 
pollinated.  One  pistillate  clone,  composed  of  11  plants  was  found, 
however,  which  produced  a  total  of  57  flowers,  18  of  which  were  still  in 
the  bud  or  flower  stage;  of  the  remaining  39  only  one  set  seed.  Another 
clone  of  this  same  type  was  found  in  a  different  region.  Material  of  both 
these  has  been  saved  to  test  further  their  fertility.  Plants  of  both  of 
these  clones,  when  grown  under  conditions  more  favorable  to  pollination, 
proved  fertile. 

In  contrast  to  the  striking  fertility  of  most  of  the  pistillate  forms  is 
the  condition  in  the  apparently  hermaphroditic  plants.  Of  1,640 
flowers  of  this  type  borne  on  381  fruiting  stalks  of  separate  plants,  403 
were  in  the  flower  or  bud  stage,  while  only  152  of  the  remaining  1,237 
set  fruit,  leaving  a  total  of  1,085  flowers  which  were  definitely  sterile; 
286  plants  of  the  381  studied  bore  no  fruit. 

The  position  on  the  inflorescence  of  the  flowers  which  set  is  interesting 
in  connection  with  the  problem  of  nubbins  and  pistil  sterility  in  our 
cultivated  varieties.  The  fact  has  previously  been  mentioned  that  a 
few  flowers  borne  on  male  plants  may  bear  only  staminodia  in  place  of 
stamens  and  that  these  are  generally  fertile.  Of  the  152  fruits  which 
set,  31  were  developed  from  this  type  of  flower.  The  other  flowers  on 
these  stalks  w^ere  of  the  usual  staminate  form  and  were  generally  sterile, 
although  a  few  instances  were  noted  in  which  one  of  the  flowers  bearing 
normal  stamens  set  a  few  achenes.  Of  these  31  fruits,  17  were  borne  on 
primary,  10  on  secondary,  3  on  tertiary,  and  one  on  a  quinary  flower. 
Five  of  the  152  which  set  were  borne  on  flowers  bearing  the  intermediate 
type  of  anther  and  of  these  2  were  primary  and  3  secondary.  The  other 
flowers  of  these  inflorescences  bore  normal  anthers  and  were  sterile. 
Fifteen  resulted  from  flowers  which  produced  some  staminodia  and 
anthers  either  distinctly  segregated  in  definite  portions  of  the  flower  or 
mixed  indiscriminately;  of  these,   12  were  from  primary  and  3  from 


624  Journal  of  Agricultural  Research  voi.  xii,  no.  i© 

secondary  flowers.  The  remaining  loi  fruits  were  borne  on  flowers 
bearing  a  full  quota  of  normal  anthers.  Of  these  flowers  54  were  primary, 
40  secondary,  and  7  tertiary.  While  most  of  the  fruits  which  set  on  the 
pistillate  plants  were  of  a  regular  shape,  indicating  a  perfect  or  nearly 
perfect  set  of  achenes,  those  borne  on  the  staminates  were  for  the  most 
part  very  irregular  in  shape,  as  the  achenes  which  set  were  often  few  in 
number  and  irregularly  scattered.  Often  not  more  than  one  or  two 
achenes  per  flower  developed.  Where  only  a  few  achenes  developed, 
the  typical  nubbins  which  are  so  common  in  the  latter  part  of  the 
picking  season  in  commercial  plantings  were  produced.  These  results 
prove  that  F.  virginiana  is  a  species  well  on  its  way  toward  diecious- 
ness,  and,  reasoning  from  analogy  with  F.  virginiana  and  F.  elatior, 
it  may  be  concluded  that  those  other  American  species  which  produce 
two  types  of  plants — that  is,  pistillate  and  somatic  hermaphrodites — are 
also  diecious. 

Recent  investigations  by  Bunyard  (6)  and  Fletcher  (74)  into  the  origin 
of  our  cultivated  strawberries  tend  to  show  that  they  have  originated 
from  hybrids  of  F.  virginiana  and  F.  chiloensis,  both  of  which  are  ap- 
parently diecious.  If  this  is  the  case,  it  raises  the  question  of  the  origin 
of  our  cultivated  hermaphroditic  forms.  A  study  of  the  pistil  ste- 
rility in  these  forms  seems  to  indicate  that  they  may  have  been  derived 
from  males  which  have  varied  in  regard  to  pistil  fertility. 

Table  III  shows  the  relationship  between  flower  position  and  pistil 
sterility  in  10  hermaphroditic  and  4  pistillate  varieties.  This  table  was 
prepared  regardless  of  the  degree  of  setting,  whether  perfect  or  whether 
the  resulting  fruit  was  a  nubbin,  all  flowers  which  set  any  achenes  being 
put  under  the  heading  "Set."  The  lower  horizontal  row  under  each  of 
the  two  groups  indicates  the  percentage  of  the  flowers  of  each  position 
which  set  fruit.  It  shows  the  very  great  increase  in  sterility  from  the 
first  flowers  to  the  last  in  both  the  pistillate  and  hermaphroditic  forms, 
being  greater  in  the  latter  than  the  former.  This  is  the  condition  which 
would  be  expected  if  the  cultivated  hermaphrodites  have  been  derived 
from  males  of  the  wild  type,  as  the  males  which  do  set  fruit  in  the  wild 
exhibit  a  high  percentage  of  their  low  fertility  in  the  primary  flowers. 

The  conclusion  that  the  hermaphrodites  have  been  derived  from  stam- 
inate  forms  rather  than  from  pistillate  is  in  keeping  with  the  results 
found  in  other  species,  as  Lychnis  spp.  (55)  and  the  grape  {43).  A  further 
study  of  pistil  sterility  was  made  in  1 5  other  varieties  of  hermaphrodites 
and  3  pistillates  to  determine  the  relationship  between  nubbins  or  irregu- 
larly set  fruit  and  flower  position. 

Nubbins  and  lack  of  setting  of  flowers  have  been  attributed  by  horti- 
culturists for  the  most  part  to  lack  of  proper  pollination  or  to  frost  injury. 
The  first  of  these  factors  may  be  eliminated,  however,  as  pollen  is  usually 
very  plentiful  and  in  a  mixed  planting,  such  as  the  data  given  in  Tables 
in  and  IV  were  taken  from,  was  always  abundant  and  especially  so  when 


Mar.  II,  1918 


Sterility  in  the  Strawberry 


625 


the  later  more  sterile  flowers  were  in  blossom.  Further,  the  fact  that 
pistillate  varieties,  grown  in  proximity  to  hermaphrodites,  set  fruit  even 
in  the  early  part  of  the  season,  when  pollen  is  admittedly  scarce,  would 
argue  for  pollination  having  little  to  do  with  nubbin  formation.  To 
those  who  have  noticed  the  effect  of  frost  on  strawberry  flowers  it  will 
be  clear  that  this  factor  may  also  be  eliminated  as  a  cause  of  irregularly 
set  fruit,  as  frost,  if  it  injures  the  flower  at  all,  will  blacken  the  entire 
receptacle.  The  possibilities  of  the  primary  flowers  being  "frostbitten" 
are  much  greater  than  the  later  ones,  but  it  is  the  latter  which  generally 
form  nubbins  or  are  entirely  sterile. 

Table  III. — Relationship  between  flower  position  and  number  of  fruits  set  in  hermaphro- 
ditic and  pistillate  varieties  of  strawberries 


Variety. 


Ses. 


Num- 
ber of 
stalks 


Primary. 


Set. 


Not 

set. 


Secondary. 


Set. 


Not 

set. 


Tertiary. 


Set. 


Not 
set. 


Quater- 
nary. 


Set. 


Not 
set. 


Quinary. 


Set. 


Not 
set. 


Reasoners  324. 
Seedling  947... 

Orem 

Lovett 

Seedling  893.. . 
Seedling  1023. . 

Abington 

Everbearing.  . 
Glen  Mary .  .  .  . 
Seedling  924. . . 


Total . .  . 
Per  cent . 


Paul  Jones . 

Marie 

July 

Wildwood  . 


Total . .  . 
Per  cent . 


66 
III 
79 
72 
83 
77 
90 
6j 
74 


87 
55 
68 
77 
S8 
90 
82 
107 


6 
1-3 


68  s 
65.8 


356 

34-2 


IS5 
65-7 


5 
83.3 


48 


286 
90.8 


177 
66.5 


80 
47   I 


6 
100 


Table  IV  shows  the  relation  between  flower  position,  imperfectly 
developed  fruit  or  nubbins,  and  complete  pistil  sterility.  It  corroborates 
what  has  already  been  pointed  out,  namely,  that  the  first  flowers  of  an 
inflorescence  are  much  more  fertile  than  the  later  ones.  With  regard 
to  nubbins  the  same  relationship  is  shown — that  is,  there  is  a  gradual 
increase  in  the  percentage  of  nubbins  formed  from  the  primary  to  the 
last  flowers  which  open.  This  condition  can  hardly  be  construed  as 
indicating  anything  but  a  morphological  sterility  of  a  portion  of  the 
pistils  in  those  flowers  which  result  in  nubbins,  if  viewed  with  the  facts 
in  mind  of  the  condition  shown  in  the  hermaphrodites  of  the  wild  parent 
species,  the  unquestionable  sterility  of  many  of  the  later  flowers,  and  the 
fact  that  the  greater  percentage  of  these  partially  sterile  flowers  are  in 
bloom  when  pollen  is  most  abundant.  If  it  were  a  question  of  pollina- 
tion, we  would  expect  the  pistillate  forms  to  exhibit  much  more  sterility 
than  the  hermaphrodites,  whereas  they  exhibit  decidedly  less,  both  with 
regard  to  the  actual  number  of  sterile  flowers  as  well  as  nubbins. 


626 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  lo 


Table  IV. — Relationship  between  flower  position  and  the  degree  of  setting  in  hermaphro- 
ditic and  pistillate  varieties  of  strawberries 


i 

to 

0 

Primary. 

Secondary. 

Tertiary. 

Quaternary. 

Quinary. 

Variety. 

1 
>-• 

2; 

01 

0 

a 
IS 

3 

2; 

1 

0 

a 

3 

t 
0 

1 

■§ 

-4-1 

•4-t 
pi 

a 

3 

Dorman 

20 
20 
20 
20 
24 
20 
20 
20 
20 
20 
20 
20 
20 
20 
20 

17 

17 

21 
15 

36 
19 
21 
18 
22 

37 
30 

26 

27 
25 
24 

2 

2 
0 
5 

18 
I 
I 
5 

1 
I 
0 
0 

5 

53 
42 
84 
36 
27 
60 
45 
47 
64 
73 
68 
39 
73 
64 
67 

13 
II 
7 

16 
40 
12 
16 
23 
2 
4 
4 
9 

2 

3 

"46 

3 

4 

43 

74 
62 
74 
3 
74 
35 
33 
70 
63 
95 
59 
85 
72 
88 

23 
23 
24 

28 

0 

51 
19 

47 
25 
2 
6 
4 
I 
17 
5 

45 

I 
35 

30 

12 

48 

33 

6 

7 

9 

7 

2 

21 

8 
32 
13 

28 

II 
II 
IS 

14 

55 

6 

39 

5 

Haverland  o 

Parson's  Beauty 

Bederwood 

Steven's  Late  Champion 

Helen  Davies 

31 
9 

16 
26 
I 
24 
II 
7 
6 
29 

10 
II 
16 
16 

27 
44 
36 
10 

Senator  Dunlap 

Pride  of  Delaware 

Minnesota  3 

Seedling  924 

Seedling  937 

8 

I 
I 
4 

I 

Seedling  947 

I 

14 
8 
7 

Seedling  876 

Seedling  (number  lost) 

s 
2 

1 

Seedling  778 

Total 

355 
89-4 

19 
15 
20 

5-1 

35 

8.8 

0 
4 

1.8 

842 
79.1 

62 
46 
56 

164 

82.  8 

164 

IS- 4 

16 

8 
10 

34 
17-2 

59 

5-5 

930 

('3-3 

95 

52 
90 

237 
69-5 

284 
19-3 

44 
27 
21 

92 

27 

256 

17.4 

4 

2 

6 

12 
3-5 

241 
39-4 

46 
24 

57 

127 
62 

119 

19-5 

19 
10 
10 

39 
19 

251 
41.  I 

23 
10 
6 

39 
19 

Per  cent 

Enormous 

9 
9 
9 

20 
20 
20 

==;; 

Warfield 

Crescent 

2 

2 
40 

1 

I 
20 

Total 

4 
6-9 

Per  cent 

"  Not  certainly  true  to  name. 

The  most  characteristic  type  of  nubbin  is  that  in  which  all  of  the 
achenes  set  except  those  situated  at  the  tip,  thus,  producing  a  berry 
with  a  dead,  dry  tip.  This  raises  the  question  as  to  why  the  tip  pistil 
should  be  more  sterile  than  those  at  the  base. 

We  have  already  seen  that  when  a  reduction  in  stamen  number  takes 
place  it  is  the  youngest  which  are  lost  first,  and  apparently  this  is  true 
also  of  the  pistils.  The  pistils  of  the  oldest  flowers  are  decidedly  more 
fertile  than  those  of  the  later  flowers,  and  it  seems  logical  that  the  older 
pistils  Avithin  a  flower  should  be  the  more  fertile.  Observations  on  a 
seedling  everbearing  variety  which  in  the  spring  produced  only  stamens 
and  a  small,  white  dome  destitute  of  pistils  in  place  of  the  ordinary 
receptacle,  but  which  later  in  the  season  gradually  produced  normal 
flowers,  illustrates  this  point.  It  was  noticed  that  the  first  flowers  which 
produced  pistils  developed  only  a  few  normal  ones  around  the  base  of 
the  receptacle,  while  those  above  gradually  decreased  in  size  until  at 
the  tip  there  were  none.  The  later  fruiting  stalks  increased  the  number 
of  normal  pistils  until  in  the  last  fruiting  stalks  the  entire  receptacles  of 
the  first  flowers  were  covered  with  normal  pistils.  This  series  of  flowers, 
although  an  extreme  with  regard  to  sterility,  still  indicates  the  portion 
of  a  flower  in  which  there  is  the  greatest  likelihood  of  its  appear^ance. 
The  question  of  pistil  sterility  should  be  kept  in  mind  in  the  selection 
of  breeding  stock,  as  it  is  most  certainly  inherited  in  the  pure  species 
and  apparently  is  in  the  cultivated  varieties,  as  selfed  seed  of  Glenville, 
a  variety  which  rarely  sets  more  than  an  occasional  primary  flower. 


Mar.  II,  1918  Sterility  in  the  Strawberry  627 

produced  a  number  of  seedlings  which  were  as  sterile  as  the  parent. 
(All  the  progeny  have  not  yet  flowered.) 

Sufficient  pistillate  varieties  have  not  been  studied  to  indicate  defi- 
nitely whether  there  is  always  a  distinct  difference  between  the  ability 
of  these  and  of  the  hermaphrodites  to  set  fruit  on  the  later  flowers; 
however,  a  comparison  of  the  percentage  of  fruit  set  on  the  tertiary 
and  quarternary  flowers  of  the  pistillate  and  staminate  varieties  reported 
in  Tables  III  and  IV  would  seem  to  indicate  that  the  former  are  de- 
cidedly more  fertile  than  the  latter,  as  is  also  the  case  in  the  wild  forms. 
Darrow  (9)  reports  that  Mr.  Hubach  will  use  only  pistillate  varieties 
as  the  fem.ale  parents  because  of  the  decrease  in  fruit  production  when 
hermaphrodites  are  used  as  both  parents. 

A  cytological  comparison  of  the  sterile  pistils  of  wild  males  and  the 
sterile  pistils  of  hermaphrodites  may  give  further  evidence  as  to  the 
origin  of  sterility  in  the  varieties  which  produce  many  nubbins  and 
entirely  sterile  flowers.  Strasburger  (jp)  has  shown  that  in  a  male 
resulting  from  a  cross  between  F.  mrginiana  9  and  F.  elatior  3- ,  the 
pistils  which  are  apparently  ready  for  fertilization  already,  in  longi- 
section,  show  a  mass  of  degenerating  material  which  contains  the  embryo 
sac  mother  cell.  This  may  or  may  not  be  the  condition  in  the  pure 
forms  of  these  species  and  in  the  sterile  pistils  of  cultivated  varieties. 

To  the  commercial  grower  of  strawberries  as  well  as  to  the  breeder 
the  question  of  variation  of  fertility  under  varying  conditions  of  environ- 
ment or  culture  is  of  importance.  Evidence  which  indicates  that  fer- 
tility is  affected  somewhat  by  environmental  or  seasonal  conditions  is 
given  by  the  seedling  which  in  the  spring  produced  strictly  male  flowers 
but  which  in  the  summer  and  early  autumn  produced  fertile  pistils  as 
well  as  stamens.  Further,  it  is  a  matter  of  common  observation  that  a 
bed  of  berries,  if  allowed  to  fruit  more  than  one  year,  will  produce  an 
increasing  number  of  small  berries  and  nubbins.  Of  actual  observational 
evidence  the  following  indicates  that  growth  conditions  have  something 
to  do  with  sterility:  A  variety,  named  "Glenville"  for  convenience,  was 
sent  to  the  Station  with  inquiries  as  to  why  it  did  not  set  fruit.  Plants 
of  it  were  grown  in  the  greenhouse  during  the  late  winter,  and,  although 
they  produced  numerous  fruiting  stalks  ancl  an  average  of  13  flowers 
per  stalk,  only  an  occasional  primary  or  secondary  flower  set  fruit. 
Some  of  these  plants,  after  having  been  grown  in  benches,  were  put 
into  pots  and  given  little  attention.  On  June  5  they  were  fruiting, 
and  a  count  was  made  of  the  flowers  which  had  set.  On  6  fruiting 
stalks  there  was  a  total  of  43  flowers,  an  average  of  slightly  over  7  per 
stalk.  Of  these,  22  set  fruit.  Of  6  primaries  there  was  i  which  set,  of 
12  secondaries  10,  of  19  tertiaries  11,  and  of  6  quaternaries  none  set. 
The  previous  3'car  some  plants  of  this  variety  had  been  planted  in  the 
field ;  and  20  days  after  taking  the  above  notes,  observations  were  made 
on  the  field  plants.  Of  105  fruiting  stalks  examined,  bearing  a  total  of 
1,292  flowers,  an  average  of  12.3  flowers  per  stalk,  there  was  a  total  of 


628  Journal  of  Agricultural  Research  voi.  xii.  no.  io 

20  fruits,  of  which  15  were  borne  on  primary  flowers,  4  on  secondaries, 
and  I  on  a  tertiary.  Although  this  variety  shows  an  extreme  case  of 
sterility,  the  condition  found  as  regards  variability  of  sterility  may  be 
an  indication  of  what  will  be  found  when  a  thorough  study  is  made  of 
this  point  in  our  cultivated  varieties. 

Thus  far  the  study  of  sterility  has  dealt  mainly  with  those  types  of 
sterility  induced  by  a  decided  tendency  toward  dieciousness  in  species 
of  Fragaria.  Another  type  of  sterility  very  prevalent  in  cultivated 
varieties  and  undoubtedly  a  factor  in  pollination  is  expressed  in  the 
appearance  in  ripe  pollen  of  varying  amounts  of  defective  grains.  It  is 
with  this  type  of  sterility  that  the  remainder  of  this  paper  deals. 

POLLEN  DEVELOPMENT  AND  STERILITY 

A  careful  cytological  examination  of  the  pollen  condition  in  the  straw- 
berries, both  wild  and  cultivated,  was  made  with  the  objects  of  deter- 
mining (i)  the  amount  of  viable  pollen  in  cultivated  varieties  and  its 
relation  to  the  setting  of  fruit  and  (2)  the  cause  of  pollen  abortion  in 
plants  of  hybrid  origin. 

The  material  used  as  a  basis  in  determining  the  general  pollen  condi- 
tion in  Fragaria  spp.,  consisted  of  (a)  F.  virginiana  from  various  parts  of 
Minnesota,  (b)  F.  americana,  (c)  a  considerable  number  of  cultivated  varie- 
ties, and  (d)  seedlings  under  test  in  the  course  of  the  fruit-breeding  work. 
The  cytological  study  was  carried  on  principally  on  the  self -fertile  variety 
Minnesota  No.  3,  a  cross  of  Senator  Dunlap  X  Pocomoke,  recently  intro- 
duced by  the  Minnesota  Agricultural  Experiment  Station.  It  produces, 
on  an  average,  about  50  per  cent  of  aborted  grains  and  so  furnishes 
desirable  material  for  the  study  of  normal  and  abnormal  pollen  develop- 
ment .    The  stages  in  normal  development  we  re  also  studied  in  F.  virginiana. 

POLLEN    CONDITION    IN   WILD   FORMS 

The  recent  work  of  Jeffry  and  his  students  on  the  pollen  condition  in 
wild  forms  puts  under  suspicion  the  genetic  purity  of  the  Rosaceae  in 
general.  The  forms  which  have  been  studied  most  intensively, 
Onagraceae  (25),  Crataegus  spp.  {37),  Rubus  spp.  (23),  and  Rosa  spp.  (8), 
show,  in  some  species,  a  relatively  large  proportion  of  aborted  pollen 
and  the  appearance  of  many  subspecies,  some  of  which  appear  to  be 
hybrids.  Because  of  this  fact  and  with  a  view  to  comparing  the  pollen 
condition  of  the  wild  with  the  cultivated  forms,  pollen  of  F.  virginiana 
and  F.  americana  was  examined. 

The  methods  used  in  determining  the  amount  of  abortive  pollen  were 
as  follows:  Fresh  flowers  were  collected  and  either  were  allowed  to  dry 
or  were  kept  with  their  pedicels  immersed  in  water  until  the  anthers 
had  dehisced.  The  pollen  was  then  transferred  to  slides  by  holding  the 
flower  over  a  slide  and  giving  it  a  few  sharp  taps.  In  this  way  the 
anthers  dehisced  completely  onto  the  slide.  A  drop  of  lactic  acid  was 
then  added,  and  a  small  cover  slide  placed  over  the  drop,  forming  a 


Mar.  II,  1918  Sterility  in  the  Strawberry  629 

fairly  permanent  mount  if  handled  carefully.  The  lactic  acid  has  the 
advantage  over  water  or  alcoliol  for  this  purpose,  as  it  is  not  volatile 
and  seldom,  if  ever,  breaks  the  pollen  grains  through  osmotic  pressure. 
It  readily  enters  and  expands  the  normal  grains,  while  it  leaves  the 
aborted  ones  collapsed.  It  has  another  very  distinct  advantage  over 
the  more  mobile  liquids,  in  that  its  viscosity  holds  both  the  normal  and 
aborted  grains  in  place  until  the  cover  has  settled  firmly,  while,  if  water 
is  used,  the  empty  grains  have  a  decided  tendency  to  wander  toward 
the  edge  of  the  cover  slide,  thus  invalidating  the  count.  Where  no 
aborted  pollen  was  present,  of  course  no  actual  counts  were  made;  but 
where  present,  counts  were  made  varying  from  200  to  2,000,  the  object 
in  each  case  being  to  include  enough  grains  to  indicate  within  a  close 
range  the  percentage  of  abortion.  A  record  was  kept  of  the  position 
on  the  inflorescence  of  the  flower  from  which  the  pollen  was  collected 
in  order  to  determine  whether  a  similar  relationship  existed  between 
flower  position  and  abortion  of  pollen  as  was  found  to  exist  between 
flower  position  and  stamen  type  in  certain  clones. 

The  results  of  the  pollen  counts  on  223  flowers  of  F.  virginiana  are 
shown  in  Table  V.  They  indicate  that  F.  virginiana  produces  for  the 
most  part  morphologically  perfect  pollen,  although  a  few  plants  were 
found  in  which  the  percentage  of  aborted  grains  was  high.  If  pollen 
condition  may  be  taken  as  a  criterion  of  species'  purity  it  may  be  said 
that  F.  virginiana  in  this  region  is  nearly  a  pure  form.  The  appearance 
of  considerable  amounts  of  aborted  pollen  in  a  few  plants  might  be 
considered  as  the  results  of  the  "conditions"  under  which  these  plants 
have  been  grown,  but  the  fact  that  several  flower  types  have  been  found 
in  the  wild  may  indicate  that  abortion  is  due  to  a  condition  arising 
from  a  slight  degree  of  hybridity  consequent  on  the  intercrossing  of 
these  forms.  The  one  primary  flower  which  produces  100  per  cent 
aborted  pollen  bore  intermediate  anthers  and  abortion  was  more  likely 
due  to  the  partial  suppression  of  stamens,  in  some  w^ay  connected  with 
the  tendency  toward  dieciousness,  than  to  other  causes  which  may  result 
in  sterility.  This  was  not  included  in  the  average  percentage  of  aborted 
pollen  in  the  primary  flowers  for  this  reason. 

A  comparison  of  the  pollen  condition  in  the  flowers  borne  on  various 
positions  shows  that  pollen  i.,bortion  is  in  no  way  related  to  flower  position 
and  thus  to  dieciousness  as  are  the  various  anther  types  and  sterile  pistils 
before  mentioned. 

In  the  hermaphroditic  species  F.  americana  nearly  the  same  condition 
exists  with  regard  to  the  degree  of  perfection  of  pollen  as  in  F.  virginiana. 
The  result  of  pollen  counts  on  49  flowers  taken  from  an  equal  number  of 
plants  are  shown  in  Table  VI  and  they  indicate  that  the  pollen  condition 
of  F.  americana  is  normal.  As  it  is  very  difficult  to  cross  F.  virginiana 
and  F.  americana,  it  is  probable  that  where  these  two  species  are  growing 
in  close  proximity  they  remain  pure. 
38325°— 18 2 


630 


Journal  of  Agricultural  Research 


.  Vol.  XII,  No.  10 


Table  V. — Pollen  condition  in  wild  plants  of  Fragaria  virginiana  with  reference  to  the 
position  of  ike  flower  on  the  inflorescence 


Aborted  percentage. 

Number  of 
primary 
flowers. 

Number  of 

secondary 

flowers. 

Number  of 
tertiary 
flowers. 

Nmnber  of 

quaternary 

flowers. 

Number  of 
flowers 
position 

not 
recorded. 

0 

9 

15 

16 

6 

4 

5 

I 
0 

I 

3 
4 

2 

I 
I 

II 

18 
5 
5 
3 

2 

3 
4 
2 

2 
I 

I 
I 

5 
6 

4 

I 

I     

14 
II 

2       

2 

6 

4 

I 
I 

5 
3 
7 
2 

1;          

6 

7 

8 

2 

I 

0 

10      

3 
6 

11-15 
16—20 

3 

I 

4 
2 

21-25 
26-30 

31-35 
36-40 
100 .  . 

I 

2 

i 

I 

I 

a  T 

i 

Total 

72 
4.9 

58 

3-2 

23 
4-3 

2 
?>  28.0 

68 

Average       percentage       of 
abortion 

6.9 

a  Not  included  in  average  because  borne  in  an  intermediate  type  of  anther. 
b  Of  no  significance  because  of  few  flowers. 

Table  VI. — Pollen  condition  in  wild  plants  of  F.  americana  with  reference  to  the  position 
of  the  flower  on  the  inflorescence 


Aborted  percentage. 

Number  of 
primary 
flowers. 

Number  of 

secondary 

flowers. 

Number  of 
flowers 
position 

not 
recorded. 

0 

I        

I 

7 
2 

4 
4 

6 

2 

3 
2 

•5 

4. 

3 

c 

6 

I 

I 
I 
I 

I 

7        

8           

Q 

I 

ic                            

II 

I 

2 

12 

12 

I 

2 

14      

JC            

I 

Total 

14 

2.8 

13 
4-4 

22 

Average  percentage  of  abortion 

5' 2 

Mar.  II,  1918 


Sterility  in  the  Strawberry 


631 


Observations  on  the  pollen  condition  in  herbarium  material  of  other 
species  than  those  above  mentioned  are  not  conclusive  with  regard  to 
the  species  examined  because  of  the  scarcity  of  material.  Plants  of 
F.  chiloensis  from  Sequin,  Washington,  from  two  localities  in  San 
Francisco  County,  California,  and  from  Bolivia,  South  America,  pro- 
duced perfect  pollen.  A  single  plant  from  Vark  Hill,  Cal.,  produced 
small  amounts  of  defective  grains  while  two  plants  of  a  clone  from  Lake 
Merced,  Cal.,  which  set  fruit  on  the  primary  flowers  of  an  otherwise 
staminate  cluster,  produced  in  the  neighborhood  of  50  per  cent  of  aborted 
grains.  A  plant  of  F.  chiloensis  var.  Scouleri,  from  Klantaak  Island, 
Yakutat  Bay,  Alaska,  produced  perfect  pollen.  F.  platypetala  from  Mt. 
Carleton,  Wash.,  also  produced  perfect  pollen.  Although  flowers  of 
comparatively  few  plants  of  these  two  species  have  been  examined,  the 
facts  seem  to  indicate  that  the  pollen  condition  is  much  the  same  as  that 
found  in  F.  virginiana  and  F.  americana. 

POLLEN    CONDITION    IN    CULTIVATED   FORMS 

In  contrast  to  the  nearly  normal  pollen  condition  of  the  wild  species 
is  the  variable  condition  found  in  the  cultivated  forms  in  which  practi- 
cally all  have  a  larger  or  smaller  percentage  of  aborted  grains.  The 
pollen  conditions  in  varieties,  controlled  seedlings  of  varieties,  and  some 
species-variety  hybrids  are  given  in  Table  VII.  The  percentages  given 
are  based  on  an  average  of  over  600  grains  per  count,  and  indicate  fairly 
accurately  the  pollen  condition  of  the  flowers  studied. 

Table  VII. — Percentage  of  aborted  pollen  in  flowers,  of  various  positions,  from  120 
cultivated  varieties,  l8  controlled  seedlings  of  cultivated  varieties,  J3  selfed  seedlings  of 
one  of  these,  3  F,  plants  of  F.  chiloensis  X  Wilson,  7  Fj  plants  of  F.  cuneifolia  X 
Magoon,  and  10  Fg  plants  of  F.  vesca  X  F.  cuneifolia.  The  percentages  are  based 
on  an  average  of  over  600  grains  per  count 


Variety. 


Position  of  flower. 


Primary. 


Second- 
ary. 


Tertiary. 


Position 

not 
recorded. 


Cultivated: 
Abington . .  . 

Abundance . 


Amanda . 


Arizona . . . . 

Aroma 

Barrymore . 
Beacon 


Bederwood. 


Bradley 

Brandywine . .  . 

Brown  Beauty. 

Charles 

Chesapeake . .  . . 


26.  5 


63.0 

1-5 

13-6 


90.8 


37- o 

20.  O 
10.5 

33- o 


4.8 
I.  o 

97.0 


10.  s 
39- o 
47.0 


7.0 

19.  8 

2.  8 

10.  0 

1             16.3 

1       l\ 

22.4 
9-7 

99-3 

88.7 
71.4 
14.  I 

9-7 
36.7 

{ 

31-7 
94-3 
100.  0 

{ 

12- S 
S-9 

632 


Jouryial  of  Agricultural  Research 


Vol.  XII,  No.  10 


Table  VII. — Percentage  of  aborted  pollen  in  flowers,  of  various  positions,  from  120 
cultivated  varieties,  18  controlled  seedlings  of  cultivated  varieties,  jj  selfed  seedlings  of 
o-ne  of  these,  5  Fj  plants  of  F.  chiloensis  X  Wilson,  7  Fj  plants  of  F.  cuneifolia  X 
Magoon,  and  10  F2  plants  of  F.  vesca  X  F.  cuneifolia — Continued 


Variety. 


Position  of  flower. 


Primary. 


Second- 
ary. 


Tertiary. 


Position 

not 
recorded. 


Cultivated — Continued: 

Clara 

Climax 

Clyde 

Commonwealth 

Cooper 

Corsican 

Darlinsiton 

Deacon 

Dewdrop 


Dorman. 


Duncan 

Early  Giant 

Early  Jersey  Giant. 
Early  Ozark 

Ekey 

Enhance 

Evcrbearin? 

Ewell's  Early 

Excelsior 

Fendell 

First  Quality 

Frances  E.  Willard. 
Fitting  Eclipse 

Candy 

Gill 


Glen  Mary. 


Glenville 

Gold  Dollar 

Gold 

Good  Luck 

Goree 

Grand  Marie 

Hanbeck  Beauty. 
Haverland  o 

Ideal 


Indiana 

Jas.  Todd 

Jerome 

Jessie 

Jewell  Improved. 

Jewell 

Jocunda 

Joe 

Kevitt  Wonder . . 
King  Edward 

Klondike 


Late  Jersey  Giant. 

Lea 

Longfellow 


Lovett. 


45-5 
13- O 


42-5 
2.0 


27.  o 
56.0 

1.8 


38.5 


23.  2 
38-0 
50.0 

i-S 
o.  o 
7-9 


79.6 
61.  S 


S4-0 
4.9 


f         87.4 
\  S9-I 


63-7 


63.1 


56.0 
S3-0 

I.  o 
69.9 
27.  I 
63.0 
39-7 
20.5 
83.3 

8.6 

6.S 
95 


39- S 
60.0 

80.8 
33-8 
73-7 
S4-0 
6.8 

2-7 

86.6 
44.9 

97-7 
46.0 
99.0 
84-3 

76.9 


84-3 


5-6 
70.  2 
46.5 


83-9 


24.8 
61.4 


32-3 
9.4 


16.6 
35- 6 
6.0 
42-3 
10.6 


10.8 
1-3 


68.4 
34-2 

5-2 

61.0 


48.0 


46.7 


SO.O 


44.0 
42.9 
42.9 

36.  S 
56.0 

43-.'; 

IS.  8 

69.6 


38.  S 

72.4 

33-1 


12.6 

50.1 


13- o 
6.8 


54- 7 

43-5 


38.2 


10.3 
39-6 

38.5 


Si-S 

44.9 

100,0 
97-3 

roo.  o 
98.0 
17.6 


3-0 
67.8 


Si-o 
14-3 


53-7 
II.  8 
iS-o 


100.  o 
100.  o 
95-0 
92- S 
85.8 
84.0 


3  Not  certainly  true  to  name. 


Mar.  II,  1918 


Sterility  in  the  Strawberry 


633 


Table  VII. — Percentage  of  aborted  pollen  in  flowers,  of  various  positions,  from  120 
cultivated  varieties,  18  controlled  seedlings  of  cultivated  varieties,  JJ  selfed  seedlings  of 
one  of  these,  j  Fj  plants  of  F.  chiloensis  X  Wilson,  7  F^  plants  of  F.  cuneifolia  X 
Magoon,  and  10  Fg  plants  of  F.  vesca  X  F-  cuneifolia — Continued 


Variety. 


Position  of  flower. 


Primary. 


Second- 
ary. 


Tertiary. 


Position 

not 
recorded. 


Cultivated — Continued : 

Magoon 

Maiinda 


Manhattan . 


Marstiall . 


Mascot 

Michel!  Early. 
Miller 


Minnesota  3 . 

Missionary . . 
Model 


New  Home. . 

New  York . . . 
Nick  Ohmer . 


II-3 
32-6 
44-5 
44.1 


82.0 
65-6 
97-5 
28.7 
41.9 
47-5 


45-3 
i-S 

47-4 

3-5 


Ohio  Boy. 


Orem 

Oswego 

Palmer 

Panama 

Pan  American .  . . 
Park  Beauty  . . . . 
Parson's  Beauty. 

Pearl 


94.2 
10.  2 
96-3 


Pennell . 


Pineapple. 


Pitcher  Eclipse. 
Progressive 


Prolific 

Pride  of  Delaware 

Providence 

Purcell  Early 

Purcell 

Reasoner  324 

Rewatisco 


Ridgeway. 


Sample. 


Saratoga 

Saunders 

Seedling  373. 

Seedling  585. 
Seedling  702. 
Seedling  753. 

Seedling  776. 


75-3 
26.  7 
24.0 


29.6 


62.0 

7.0 
85.1 
54.0 

12.5 


60.  4 
14.2 


38.0 


26.4 


8.6 
SO.  8 


43-5 
21.  7 
18.2 
47.1 
41.8 
41.9 
39- S 
57-5 
81.4 

4.0 
67.  2 
50-S 

i.o 
15- o 

45- S 

25.0 
19-8 
72-5 
11.5 
10.3 
29.6 
22.0 
53-8 
43- S 
39-9 


42.9 

7.8 


H 


SS-7 
S3- 5 

91- S 
23-1 

19.  2 

61.  7 
46.7 


3-0 
3-9 
6.0 


24.8 

15-5 
33-3 


38-4 


50.0 
56.1 
9.0 
18.0 
16.9 
15.4 
49.8 
63-8 
28.1 
14- S 


62.0 
100.  o 
100.  o 


35-6 
37-4 
40.0 
45-4 
82.6 
44-3 

60.5 
29.6 


84.4 
85-9 


634 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  lo 


Table  VII. — Percentage  of  aborted  pollen  in  flowers,  of  various  positions,  from  120 
cultivated  varieties,  i8  controlled  seedlings  of  cultivated  varieties,  33  selfed  seedlings  of 
one  of  these,  3  F^  plants  of  F.  chiloenis  X  Wilson,  /  F^  plants  of  F.  cuneifolia  X 
Magoon,  arid  10  F^,  plants  of  F.  vesca  X  F.  cuneifolia — Continued 


•Variety. 


Position  of  flower. 


Primary. 


Second- 
ary. 


Tertiary. 


Position 

not 
recorded. 


Cultivated — Continued: 
Seedling  778 


24.8 


Seedling  813. 
Seedling  823. 
Seedling  893. 

Seedling  908. 


79.8 


Seedling  923 

Seedling  924 

Seedling  927 

Seedling  947 

Seedling  loio 

Seedling  1017 

Seedling  1043 

Seedling   104s — 

Senator  Dunlap. 


Son's  Prolific 

South  Dakota 

Splendid 

Splendid  X  Bunlap. 


9.0 
13- o 


18.5 


Staples. 


Steven's  Late  ChaxajMon. 


St.  Louis. 
Success .  . 


Sweetheart , 

Tennessee  Prolific. 

Texas 

Three  Ws 

Twilley 

Uncle  Jim 

Warren 

Warfield  a 

Wm.  Belt 

Wilson 

Winner 

Wolverton 

Wonder 


Average. 


Selfed  seedlings  of  Seedling  778: 


24-4. 
25-1. 
25-2. 

25^- 

25-6. 
26-1. 
26-4. 

27-5- 
28-3. 

S3-4- 
54-2. 


77.0 
17.0 
49.2 


52-3 

24.5 
23.  S 


o  Not  certainly  true  to  name. 


25-4 
15-9 


55-4 
25.0 


77- S 
S4-I 
61.8 

29.6 


IS- 2 
31.0 
10.0 

S9-3 

14-3 


13-8 
4S-9 


20.  o 
34-9 


46-3 
28.6 


S7-9 

1.  o 
52-1 
69.  o 
75-8 
40.  o 
45-2 

4.6 

2.  2 
10.  2 
80.6 

9S 
1.6 


I.  2 

5S-4 
7.8 


38.3 


47.1 
52-4 
27.9 


Mar.  II,  1918 


Sterility  in  the  Strawberry 


635 


Table  VII. — Percentage  of  aborted  pollen  in  flowers,  of  various  positions,  from  120 
cultivated  varieties,  18  controlled  seedlings  of  cultivated  varieties,  jj  selfed  seedlings  of 
one  of  tliese,  j  F^  plants  of  F.  chiloensis  X  Wilson,  7  Fj  plants  of  F.  cuneifolia  X 
Magoon,  and  10  F^  plants  of  F.  vesca  X  F.  cuneifolia— -Continued 


Position  of  flower. 

Variety. 

Primary. 

Second- 
ary. 

Tertiary. 

Position 

not 
recorded. 

Selfed  seedlings  of  Seedling  778— Continued: 

2S-3 

9-9 

f       22.5 

I          19- 4 
/         52-4 
I         42.5 

66-1 

32-0 

66-2 

94.2 

99.0 
43-9 
38.1 

66-3 

67-1 

25.8 

41-3 

99.0 
9.1 

f         10.6 

\         22. 3 

21-3 

71-5 

46.0 
100. 0 

33-1 

25.9 

0.4 
49.1 

SO.  8 
52- 7 

34-9 

72-6 

1.8 
62.8 

.     34-7 

61.  0 

Average 

47.1 

36.1 

32.7 

3  Fi  plants  of  F.  chiloensis  X  Wilson: 

77- X 

58.0 

76.7 

7  Fi  plants  of  F.  cuneifolia  X  Magoon: 

S-2 

IS-O 

2.0 

SI.  4 

18.2 

16.4 

29.1 



..1 

II.  2 

7-7 
8-3 
8.0 

19-3 

7-3 

7.0 

8                                                 

II.  9 

S-2 

3-  I 

8.9 

The  most  striking  fact  exhibited  by  the  counts  in  Table  VII,  other 
than  the  general  presence  of  some  aborted  pollen,  is  the  variability  of 
the  pollen  condition  within  a  variety.  Abington,  for  instance,  shows  a 
range  of  from  7  to  37  per  cent  of  abortive  pollen.  Abundance  from  2 
to  33  per  cent,  and  Bederwood,  which  usually  produces  a  high  per- 


636 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  10 


centage,  in  one  instance  produced  as  low  as  31.7  per  cent  of  abortive 
grains.  Numerous  other  equally  striking  variations  in  pollen  abortion 
will  be  evident  by  referring  to  the  table. 

In  view  of  the  variability  shown  above,  the  pollen  condition  was  deter- 
mined in  all  of  the  flowers  from  two  inflorescences  of  Minnesota  3 
to  determine  what  variations  occur  in  flowers  grown  under  so  nearly 
similar  conditions.  The  results  of  these  counts  are  presented  in  Table 
VIII  and  exhibit  as  great  variability  within  the  flowers  of  a  cluster  as 
is  found  between  flowers  from  separate  plants.  In  one  stalk  the  range 
is  from  31.2  to  91. i  per  cent,  in  the  other  it  is  from  21.4  to  40.3  per  cent, 
while  the  range  in  other  counts  of  the  same  variety  (Table  VII)  from 
different  plants  is  from  14.2  to  60.4  per  cent. 

Table  VIII. — Percentage  of  aborted  pollen  in  all  of  the  flowers  from  two  stalks  of 

Minnesota  j 


Stalk. 

Primary. 

Secondary. 

Tertiary. 

Quaternary. 

Stalk  I 

40.3 
(a) 

/          29.9 
I          30.6 
f         91.  I 
63.5 
1           70-7 
I        7^-9 

21.  4 

Stalk  II 

42.3 
68.2 
31.  2 
49-  7 

42.  6 
42.7 

o  Anthers  intermediate. 


In  view  of  the  variability  shown  between  different  plants  of  the  same 
variety  and  between  different  flowers  of  a  single  inflorescence,  a  study 
was  made  of  the  variation  in  pollen  conditions  of  individual  anthers 
within  a  single  flower.  The  pollen  of  nine  anthers,  from  a  flower  of 
Seedling  778  was  studied,  four  of  which  were  from  the  outer  para- 
petalous  series,  three  from  the  middle  antipetalous  series,  and  two  from 
the  inner  long  filamented  antisepalous  series.  The  results,  given  in 
Table  IX,  show  a  greater  range  of  variability  between  the  anthers  of  a 
single  flower  than  was  exhibited  by  the  five  separate  flowers  of  the  same 
variety  reported  in  Table  VII.  Because  of  the  extremely  great  varia- 
bility in  pollen  condition  shown  by  some  varieties,  as  Abington,  Abun- 
dance, Aroma,  Darlington  and  others,  which  often  produce  nearly  perfect 
pollen  (as  far  as  can  be  determined  by  this  method  of  study),  while  at 
other  times,  apparently  under  the  same  conditions,  high  percentages  of 
abortive  pollen  are  produced,  too  much  stress  should  not  be  laid  on  a 
few  scattered  observations  in  determining  whether  a  species  is  pure  or 
of  hybrid  origin  until  more  is  known  of  the  factors  which  produce  such 
great  variability.  It  may,  of  course,  be  argued,  and  logically,  that  a 
large  number  of  the  grains  in  the  anthers,  which  produce  nearly  perfect 
pollen,  are  truly  abortive,  but  have  developed  beyond  the  stage  where 


Mar.  II,  191S 


Sterility  in  the  Strawberry 


637 


degeneration  can  be  discovered  by  a  superficial  examination.  The 
possibility  of  this  being  the  case  will  be  shown  later.  Nevertheless  the 
factors  causing  a  relatively  high  degree  of  pollen  abortion  in  apparently 
pure  species  have  been  so  little  studied  that  to  assign  hybridity  as  the 
only  cause  is,  to  say  the  least,  presumptive. 

Table  IX. — Variation  in  percentage  of  pollen  abortion  in  anthers  from  one  flower  of 

Seedling  yjS 


Stamen  position. 


Parapetalous. 


Antipetal- 
ous. 


29.18 
50.00 
41.70 
34.61 


33-65 
38.65 
17.09 


Antisepal- 
ous. 


27.  27 
41.87 


POLLEN   GERMINATION  TESTS 

In  view  of  the  possibility  of  many  of  the  apparently  normal  grains 
being  in  reality  abortive,  attempts  were  made  to  determine  the  exact 
amount  of  fertile  pollen  without  regard  to  its  apparent  morphological 
condition.  The  usual  method  of  pollen  germination  in  Van  Tieghem 
cells  was  employed.  The  results  were  disappointing,  so  far  as  a  deter- 
mination of  actual  condition  of  individual  grains  was  concerned;  but 
nevertheless  some  suggestive  conclusions  may  be  drawn  from  them.  In 
all,  450  tests  were  made,  comprising  28  of  F.  americana,  45  of  F.  virginiana, 
and  377  of  cultivated  varieties.  The  results  obtained  under  carefully 
controlled  conditions  were  very  erratic.  Solutions  of  cane  sugar  in 
distilled  water  were  used  in  concentrations  var}dng  from  7  to  60  per  cent. 
Pollen  germinated  to  a  very  slight  degree  at  both  of  these  extremes, 
but  the  optimum  concentration  ranged  between  35  and  45  per  cent. 

Temperature  is  an  important  factor  in  pollen  germination,  since  at 
ordinary  room  temperature  there  was  practically  no  germination,  while 
if  the  cultures  were  placed  in  a  warmer  portion  of  the  room,  at  a  tempera- 
ture of  about  90°  F.  sometimes  a  high  germination  resulted.  Other 
tests,  carried  on  under  as  nearly  identical  conditions  as  possible,  often 
gave  entirely  negative  results.  Tests  made  in  an  electric  oven,  at  95°  F., 
showed  at  times  a  high  percentage  of  germination,  while  at  others  the 
germination  was  very  low.  Although  the  proper  conditions  for  germi- 
nation could  not  always  be  produced,  yet  enough  evidence  was  obtained 
to  conclude  that  wherever  any  normal  pollen  is  present,  a  portion  of 
it  is  likely  to  have  the  power  of  germination,  for  in  several  cases  where 
over  95  per  cent  of  the  pollen  was  abortive  some  of  the  morphologically 
perfect  grains  produced  normal  tubes. 


638 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  10 


BAGGING   TESTS 

The  final  test  for  the  fertility  of  pollen  is  its  behavior  when  used  in 
pollination.  The  simplest  method  of  testing  pollen  fertility  is  to  bag 
the  flowers  before  they  open.  This  method  has  been  used  very  exten- 
sively by  several  investigators  in  self-sterility  studies  of  the  pear,  apple, 
cherry,  peach,  and  grape  and  has  been  found  to  be  efficient  under  favor- 
able conditions.  Observations  by  growers  upon  large  blocks  of  any  of 
the  hermaphroditic  varieties  of  strawberries  agree  that  strawberries  are 
all  self-fertile,  physiologically,  wherever  normal  pollen  is  produced,  so 
that  the  question  of  self -sterility  does  not  enter  into  the  problem. 

Bagging  tests  were  made  on  106  varieties  and  40  unnamed  seedlings 
produced  at  the  Minnesota  Fruit  Breeding  Station.  The  detailed  results 
and  summary  of  these  tests  are  given  in  Table  X,  and  show  (i)  that  in 
the  hermaphroditic  varieties  studied  no  physiological  self -sterility  exists; 
(2)  that  wherever  morphologically  normal  pollen  is  present  fertilization 
takes  place;  and  (3)  that  the  extent  of  fertilization  is  dependent  upon 
the  percentage  of  normal  pollen  produced. 

Table  X. — Degree  of  setting  of  the  fruits  of  various  positions  on  the  inflorescence  on  g8 
hermaphroditic  varieties  and  jg  seedlings  of  strawberries  when  bagged 


Variety. 


Primary. 


Secondary. 


Tertiary. 


Quaternary. 


Total. 


Abington 

Abundance 

Amanda 

Arizona 

Arojna 

Barryittore 

Bederwood 

Bradley 

Brandywine 

Brown  Beauty. . . . 

Charles  I 

Chesapeake 

Clara 

Climax 

Clyde 

Cooper 

Commonwealth 

Corsican 

Darlington 

Dorman 

Duncan 

Early  Jersey  Giant 

Early  Ozark 

Ekey 

Enhance 

Ewell  Early 

Excelsior 

First  Quality 

Candy 

Gill 

Glen  Mary 

Gold  Dollar 

Good  Luck 

Goree 

Grand  Marie 

Hanbeck  Beauty . . 
Haverland  "■ 


a  Not  certainly  true  to  name. 


Mar.  II,  1918 


Sterility  in  the  Strawberry 


639 


Table  X. — Degree  of  setting  of  the  fruits  of  various  positions  on  the  inflorescence  on 
hermaphroditic  varieties  and  jg  seedlings  of  strawberries  when  bagged— Continued 


Variety. 


Primary. 


Secondary. 


Tertiary. 


Quaternary. 


Total. 


Ideal 

Indiana 

Jas.  Todd 

Jerome 

Jessie 

Jewell  Imp 

Joe 

Kevitt  Wonder 

King  Edward 

Klondike 

Late  Jersey  King 

Lea 

Longfellow 

Lovett 

Manhattan 

Marshall 

Mascott 

Michell  Early 

Miller 

Missionary 

Model 

Molinda 

New  Home 

New  York 

Nich  Ohmer 

Ohio  Boy 

Orem 

Oswego 

Palmer 

Pan  American 

Parson  Beauty 

Pearl 

PenneU 

Pineapple 

Pitcher  Eclipse 

Pride  of  Delaware 

Pride  of  Minnesota 

Prolific 

Providence 

Reasoner  324 

Rewatisco 

Ridgway 

Sample 

Saratoga 

Saunders 

Senator  Dunlap 

Son's  Prolific 

South  Dakota 

Staples 

Steven's  Late  Champion 

St.  Louis 

Success 

Tennessee  Prolific 

Three  Ws 

Twilley 

Uncle  Jim 

Warfield 

Warren 

Wm.  Belt 

Winner 

Wolverton 

Minnesota  3 

Seedling  14 

Seedling  114 

Seedling  130 

Seedling  97 

Seedling  15 

Seedling  123 

Seedling  123 

Seedling  89 


38 


2S 

17 
3 

19 
8 

25 

IS 
37 
27 

12 
9 
8 

I 
16 
17 
17 


1  One  bag  split  or  tip  open,  allowing  the  possibility  of  cross-pollination. 

6  Of  a  total  of  11  flowers,  none  set.    The  pollen  of  this  variety  has  not  been  ejcamined. 


640 


Joitrnal  of  Agricultural  Research 


Vol.  XII,  No.  10 


Table  X. — Degree  of  setting  of  the  fruits  of  various  positions  on  the  inflorescence  on 
hermaphroditic  varieties  and  jp  seedlings  of  strawberries  when  bagged — Continued 


Variety. 


Seedling  40 -  -  -  - '     2 

Seedling  168 |     2 

SeedlingsSs '     2 

Seedling  753 [ "  2 

Seedling  876 

Seedling  703 

Seedling  778 

Seedling  937 

Seedling  1065 

Seedling  1009 

Seedling  845 

Seedling  901 

Seedling  1018 

Seedling  1017 

Seedling  585 

Seedling  702 

Seedling  753 

Seedling  778 

Seedling  776 

Seedling  908 

Seedling  923 

Seedling  924 

Seedling  937 

Seedling  947 - 

Seedling  loio 

Seedling  101 7 

Seedling  1023 

Seedling  1026 

Seedling  1045 

Splendid  XDunlap 


Total.... 
Per  cent . 


Primary. 


113 
26.8 


48 
II.  4 


Secondary. 


28s 
24.9 


624 
54.6 


234 
20.5 


Tertiary. 


337 

21-5 


587   646 
37. 441. 1 


Quaternary. 


70 
II.  8 


114 
19.  2 


Total. 


a  Two  bags  split  or  tip  open,  allowing  the  possibility  of  cross-pollination. 

Pollination  in  unbagged  flowers  is  for  the  most  part  dependent  upon 
bees  and  small  insects  and  upon  the  anthers  becoming  erect  and  partially- 
folding  about  the  receptacle  while  dehiscing,  thus  dropping  the  pollen 
on  the  stigmatic  surfaces.  In  the  bagged  flowers  the  insects  are  elimi- 
nated, and  thus  the  most  efficient  natural  means  of  pollination  is  lost. 
As  a  consequence  many  of  the  bagged  varieties  produced  many  nubbins, 
but  no  case  of  complete  self-sterility  was  found  where  morphologically 
perfect  pollen  was  present. 

If  the  percentages  of  total  perfect  fruits,  nubbins,  and  sterile  flowers 
of  each  flower  position  in  Table  IV  are  compared  with  those  in  Table  X,  it 
will  be  seen  that  (i )  the  setting  under  bags  is  very  much  poorer  than  in  the 
open,  both  with  regard  to  the  number  of  perfect  fruits  set  and  also  with 
regard  to  the  actual  number  of  flowers  which  set  any  achenes,  and  (2)  the 
primary  flowers  are  decidedly  more  fertile  than  the  later  ones,  as,  even 
under  the  adverse  conditions  of  pollination  within  the  bags  only  ii.o 
per  cent  of  the  primary  flowers  were  sterile,  while  20  per  cent  of  the 
secondary.  40.7  per  cent  of  the  tertiary,  and  68.5  per  cent  of  the  quater- 
nary flowers  were  sterile,  in  spite  of  the  very  great  increase  in  amount  of 


Mar.  II,  1918  Sterility  in  the  Strawberry  641 

pollen  present  in  the  bags  while  the  later  flowers  were  in  condition  for 
pollination. 

In  order  to  determine  to  what  extent  parthenogenesis  or  parthenocarpy 
might  possibly  enter  into  the  above  results,  bags  were  put  over  67  clusters 
of  22  pistillate  varieties.  Of  a  total  of  661  flowers  covered,  55  set  some 
achenes.  Of  these,  52  were  found  in  6  bags  which  had  been  accidentally 
split,  thus  accounting  for  the  probable  pollination  by  insects.  Of  the 
remaining  three  fruits,  which  developed  in  apparently  tight  bags,  two  bore 
I  achene  each  and  the  other  7.  Significance  can  hardly  be  attributed 
to  the  setting  of  these  few  achenes,  since  the  chances  for  accidental 
pollination,  to  this  extent,  are  relatively  great.  It  may  therefore  be 
concluded  that  parthenogenesis  does  not  exist  in  the  cultivated  strawberry. 

A  condition  which  might  possibly  be  attributed  to  parthenocarpy 
occurred  in  the  Buster  variety,  in  which  9  flowers  of  the  22  bagged 
showed  a  very  decided  development  of  the  achenes  with  no  accompany- 
ing development  of  the  receptacle.  These  achenes  contained  no  embryos. 
In  the  strict  sense  of  the  word  parthenocarpy  in  the  strawberry  could 
only  be  applied  to  a  development  of  this  kind.  A  more  comprehensive 
use  of  the  term  might  include  the  development  of  the  fleshy  receptacle. 
Ordinarily,  flowers  which  set  only  one  or  a  few  achenes  develop  the 
fleshy  receptacle  only  at  corresponding  points,  due  probably  to  the 
stimulus  of  fertilization.  In  perfect  varieties  receptacles  are  often 
found  in  which  development  has  taken  place  not  only  at  the  base  of 
the  pistils  but  also  about  the  base  of  the  stamens.  In  3  out  of  10 
bagged  flowers  of  the  pistillate  Red  Bird  variety,  the  fleshy  receptacle 
developed  about  the  base  of  the  staminodia  forming  a  red  fleshy  circle 
about  the  dried  pistils.  Two  flowers  of  Crescent,  also  an  imperfect 
variety,  developed  normal  fleshy  berries,  one  bearing  one  achene  and 
the  other  none.  Apparently  these  receptacles  developed  without  the 
stimulus  of  fertilization  in  the  same  way  as  that  at  the  base  of  the  stami- 
nodia in  Red  Bird. 

POLLEN   DEVELOPMENT 

As  Mendelian  and,  in  fact,  most  genetic  results  are  dependent  upon 
the  segregation  of  determiners  during  the  formation  of  gametes  and  to 
their  recombination  again  at  the  time  of  fertilization,  any  processes 
which  interfere  with  the  normal  procedure  should  be  carefully  studied 
and,  if  possible,  their  nature  determined. 

There  are  a  number  of  ways  in  which  the  normal  order  may  be  dis- 
turbed, at  least  there  are  various  outward  expressions  of  them.  The 
condition  in  the  Phylloxera  spp.,  as  pointed  out  by  Morgan,  in  which 
half  of  the  spermatids  degenerate  regularly,  while  the  other  half  con- 
tinue and  form  normal  spermatozoa,  and  the  relationship  between 
degeneration  and  the  absence  of  the  accessory  chromosome  is  so  well 
known  that  it  needs  no  comment.     Gates    (iS)   has  shown   that  in 


642  Journal  of  Agricultural  Research  voi.  xii,  no.  10 

Oenothera  lata  the  early  abortion  of  the  male  generative  organs  and 
partial  abortion  of  the  female  is  in  some  way  associated  with  the  presence 
of  an  extra  chromosome,  while  in  Oenothera  semilata  an  extra  chromo- 
some is  present,  but  only  a  portion  of  the  pollen  grains  abort.  Morgan 
has  shown  that  slight  mutations  are  continually  occurring  in  Drosophila 
spp.,  which  inhibit  the  development  of  the  2X  individual,  while  Bridges 
(5)  has  shown  that  certain  chromosome  combinations  can  not  bring 
about  normal  development  in  the  zygote.  Is  it  too  much  to  expect 
that  like  conditions  may  affect  the  i  X  generation  also  ? 

Dorsey  {12)  pointed  out  a  different  type  of  pollen  degeneration  from 
that  which  expresses  itself  in  the  production  of  empty  grains.  He 
showed  that  in  functionally  pistillate  grapevines  pollen  developnemt 
proceeds  normally  through  the  microspore  division  and  the  formation 
of  the  normal  content  of  cytoplasm  found  in  mature  fertile  grains. 
During  the  period  of  development  following  the  microspore  division, 
one  or  both  of  the  nuclei  of  a  portion  of  the  pollen  grains  aborted,  leaving 
the  grain  normal  as  far  as  cytoplasm  was  concerned.  Associated  \^^th 
the  complete  sterility  of  the  pollen  of  the  functionally  pistillate  varieties 
were  the  reflexed  type  of  stamens,  an  entire  lack  of  sutures  and  germ 
pores  in  the  mature  pollen,  and  dieciousness.  It  is  probable  that  the 
lack  of  germ  pores  is  the  direct  cause  of  stei-ility  in  the  numerous  grains 
which  otherwise  appear  entirely  normal.  Tischler,  Rosenberg,  and  others 
have  shown  that  pollen  abortion  in  hybrids  may  follow  either  normal  or 
irregular  reduction  divisions,  when  the  parents  have  both  an  equal  and 
unequal  number  of  chromosomes.  Shull  (35,  36)  has  recorded  some 
consistent  irregularities  in  sex  ratios  in  Lychnis  dioica  which  will  later 
be  shov/n  possibly  to  have  been  due  to  slight  mutations  causing  pollen 
abortion.  Goodspeed  {21,  22)  gives  further  evidence  on  the  sterility  of 
hybrids  oi  Nicotiana  spp,,  when  A^  sylvestris  is  used  as  one  parent  andshovv^s 
that  not  only  the  pollen  is  sterile  but  that  the  F^  plants  are  incapable  of 
forming  any  very  appreciable  amounts  of  seed.  Rimpau  {31)  and 
Jesenko  (26)  have  shown  that  in  hybrids  between  wheat  and  r5'e  there  is 
complete  male  sterility,  while  some  of  the  egg  cells  are  able  to  produce 
viable  seed  if  pollinated  with  either  wheat  or  rye  pollen. 

The  cytological  investigation  of  pollen  development  in  the  strawberry, 
reported  in  this  paper,  has  two  main  objects  in  view:  (i)  the  determina- 
tion of  the  mechanism  and  nature  of  pollen  abortion  and  \vith  these 
facts  at  hand,  (2)  the  determination  to  which  of  the  many  categories  of 
sterility  the  very  prevalent  pollen  abortion  in  the  strawberry  varieties 
belongs. 

Material  and  methods. — The  variety  used  as  the  basis  of  this  study 
is  Minnesota  3.  As  was  previously  stated,  it  furnishes  desirable  ma- 
terial for  this  type  of  study,  as  about  50  per  cent  of  its  pollen  aborts, 
while  the  remainder  develops  normally.  The  egg  cells  in  this  variety 
and  in  the  cultivated  varieties  in  general  do  not  show  a  corresponding 


Mar.  II,  1918  Sterility  in  the  Strawberry  643 

degree  of  abortion,  as  it  is  very  common  for  practically  all  of  the  achenes 
to  develop  on  perfectly  formed  strawberries. 

The  material  was  prepared  for  microscopic  examination  according 
to  the  ordinary  cytological  methods.  Carnoy's,  Flemming's  strong, 
medium,  and  weak,  and  chromacetic-acid  fixing  solutions  were  used. 
All  g'ave  very  good  results,  except  Carnoy's  fluid.  Sections  were  cut 
from  4  to  20  ^t  thick,  the  best  results  being  obtained  from  those  4  to  6  /i 
thick.  The  triple  stain  and  Haidenhain's  iron-alum-hymatoxylin  stains 
were  used,  both  giving  good  satisfaction. 

The  drawings  in  Plates  B  to  E  were  outlined  w^th  the  help  of  an  Abbe 
camera  lucida.  All,  with  a  few  exceptions  noted,  are  drawn  to  the  same 
scale,  in  order  that  comparisons  of  cell  size  and  cytoplasm  content  may  be 
readily  made. 

Anther  tissues. — The  walls  of  the  young  anthers  are  made  up  of 
four  oblong  layers  of  cells  of  about  equal  size ;  the  outer  epidermal  layer 
and  three  inner  layers.  Inside  of  these  layers  there  is  usually  one 
layer  of  tapetal  cells  and  about  five  layers  of  pollen  mother  cells,  both 
of  which  at  this  time  are  easily  distinguished  from  the  wall  cells  by 
their  large  size  and  different  staining  reaction. 

The  growth  and  development  of  the  wall  layers  should  be  followed 
because  of  its  relation  to  the  increase  in  size  of  the  pollen  mother-cell 
cavity  during  the  formation  of  the  tetrads.  As  the  pollen  mother  cells 
prepare  for  reduction,  and  during  the  division,  rapid  cell  division  is  tak- 
ing place  in  the  parietal  cells,  so  that  by  the  time  the  heterotypic  division 
is  complete  there  is  an  appreciable  increase  in  the  size  of  the  anther 
cavity.  Gradually  the  two  inner  layers  of  cells  flatten  out,  owing  prob- 
ably to  the  growth  of  the  outer  layer,  the  cells  of  which  rapidly  increase 
in  size,  although  showing  no  further  cell  divisions. 

At  the  tetrad  stage  the  inner  layer  is  very  much  flattened,  while  the 
middle  layer  is  still  plainly  visible  and  the  cells  are  still  full  of  cytoplasm. 
The  cells  of  the  outer  layer  are  now  very  appreciably  larger  than  the 
epidermal  cells,  which  have  also  grown  slightly.  After  the  liberation 
of  the  microspores  and  while  they  are  increasing  in  size,  the  epidermal 
cells  sometimes  collapse,  as  their  contents  have  become  scant.  By  the 
tirhe  the  microspores  have  nearly  completed  their  growth  and  have 
begun  to  divide,  the  cells  of  the  outer  layer  have  become  deeper  than 
long  and  are  nearly  as  large  as  the  tapetal  cells.  At  this  time  they  show 
distinctly  the  spiral  thickenings  which  have  to  do  with  dehiscence.  Both 
of  the  inner  layers  have  now  collapsed  or  show  very  scant  cytoplasm. 
Before  dehiscence  the  walls  separating  the  members  of  the  two  pairs  of 
loculi  break  down,  leaving  two  large  loculi  in  each  anther.  The  relation 
between  the  increase  in  size  of  the  anther  cavities  both  during  prepara- 
tion for  the  first  meiotic  division  and  subsequent  to  it,  and  a  diff'erence 
which  Fragaria  spp.  shows  from  some  other  forms  in  the  history  of  the 
mother-cell  wall  will  be  pointed  out  later. 


644  Journal  of  Agricultural  Research  voi.  xii,  no.  10 

The;  tapetum. — The  tapetum,  usually  one  cell  layer  thick,  is  com- 
posed of  large  angular  cells  similar  in  size,  and  staining  reactions  to  the 
pollen  mother  cells.  The  tapetum  in  Fragaria  spp.  differs  in  its  greater 
persistence  from  most  other  forms  reported.  Division  in  the  tapetal 
cells  begins  at  about  the  time  of  synapsis  of  the  pollen  mother  cells  and 
has  been  observed  as  late  as  the  metaphase  of  reduction  division.  The 
divisions  are  all  mitotic,  no  evidences  of  amitotic  divisions  having  been 
observed.  Following  reduction  division  the  tapetal  cells  are  binucleate 
and  remain  so  until  the  liberation  of  the  microspores,  when  they  degen- 
erate and  completely  disappear.  The  disappearance  of  the  tapetal 
layers  is  gradual.  They  first  separate  from  the  wall  layers  and  then 
proceed  to  dissolve,  the  wall  which  was  in  contact  with  the  anther  wall 
first  disappearing,  followed  by  gradual  dissolution  of  the  entire  cell 
layer.  During  this  process  the  walls  become  thick  and  laminate  and  in 
places  the  middle  lamella  dissolves,  partially  freeing  the  individual  cells. 
When  the  microspores  are  three-quarters  grown  the  tapetal  cells  have 
entirely  disappeared.  This  persistence  of  the  tapetal  cells  will  be  shown 
to  be  correlated  with  a  like  persistence  of  the  pollen  mother-cell  walls. 
The  history  of  the  tapetum  in  F.  virginiana  is  identical  with  that  just 
described  which  is  of  Minnesota  3. 

Pollen  mother  cells. — The  study  of  pollen  development  was  begun 
with  the  so-called  resting  stage  of  the  pollen  mother  cells  between  the 
last  archesporial  division  and  the  first  meiotic  division.  They  do  not, 
however,  show  the  characteristic  chromatin  and  linin  condition  found 
in  true  resting  cells  of  Fragaria  spp.  (compare  fig.  i,  PI.  B  and  fig.  2, 
PI.  D).  The  cells  are  angular  and  contain  a  large  nucleus  (PI.  B,  i). 
The  contents  of  the  nucleus  are  irregular  dark-staining,  very  small  masses 
of  chromatin  held  in  a  network  of  linin  fibers.  The  number  of  chro- 
matin bodies  is  very  much  larger  than  the  number  of  chromosomes. 
Ordinarily  one  large  nucleolus  is  present,  although  it  is  not  uncommon 
to  find  two. 

Synapsis. — ^The  first  indication  of  the  onset  of  the  prophase  is  to  be 
seen  in  the  gradually  increasing  size  of  these  chromatin  bodies  which 
still,  however,  appear  very  flaky  and  irregular.  The  linin  and  chromatin 
during  this  period  are  so  indistinct  that  it  is  impossible  to  determine 
whether  there  is  any  definite  pairing  of  the  individual  particles  and 
threads  as  has  been  shown  to  exist  in  Lilium  spp.  by  Allen  (2)  and  in 
several  forms  by  Overton  (jo).  These  larger  masses  gradually  move  to 
one  side  of  the  nuclear  cavity  and  congregate  about  or  near  the  nucleolus 
in  a  loose  indefinite  mass  (Pi.  B,  2).  At  this  time  a  few  rather  defi- 
nite threads  appear  in  the  mass,  some  extending  out  from  it  as  loops. 
Where  the  loops  are  long  enough,  they  are  seen  to  be  distinctly  double 
(PI.  B,  2,  3).  This  is  the  only  evidence  of  any  pairing  during  the 
presynaptic  stages.  Gradations  between  the  conditions  shown  in 
figures  I  and  2  occur  within  a  single  loculus  of  an  anther  and  are  proof 


Mar.  II.  1918  Sterility  in  the  Strawberry  645 

that  these  stages  bearing  bivalent  loops  are  presynaptic.  Gradually  the 
synaptic  mass  tightens  until  it  is  close  and  compact,  occupying  a  very 
small  portion  of  the  nuclear  cavity  (PL  B,  4).  During  this  contrac- 
tion there  are  refractive  particles  present  both  in  the  mass,  some  of  which 
are  in  contact  with  the  nucleolus,  and  outside  of  the  nuclear  membrane, 
which  give  the  same  staining  reactions  as  the  nucleolus.  Similar  masses 
to  these  Digby  (jj)  has  considered  to  be  synaptic  extrusions. 

The  synaptic  stage  is  of  long  duration.  Gradually  the  chromatin  in 
synapsis  takes  on  the  appearance  of  being  made  up  of  a  closely  tangled 
mass  of  threads.  Soon  loops  are  pushed  out  from  it,  which  are  bivalent 
often  for  their  entire  length  (PI.  B,  5).  The  fact  that  these  loops  are 
often  double  from  the  point  at  which  they  leave  the  mass  to  the  point  of 
entrance,  and  can  sometimes  be  traced  through  a  portion  of  the  synaptic 
mass  gives  the  impression  that  they  are  made  up  of  two  continuous 
threads  which  closely  approximate  each  other  over  their  entire  length. 
This  view  is  supported  by  the  later  stages,  especially  those  at  and  follow- 
ing segmentation,  which  in  Fragaria  spp.  are  very  clear. 

Loops  continue  to  push  out  from  all  sides  of  the  synaptic  mass,  often 
shifting  it  to  the  center  of  the  nucleus.  Gradually  the  bivalent  thread 
becomes  more  or  less  regularly  distributed  about  the  nuclear  cavity, 
usually  having,  however,  a  somewhat  tangled  center  near  the  nucleolus. 
The  spireme  thread  at  this  stage  often  appears  to  be  a  single  strand 
due  to  the  dose  approximation  of  its  univalent  portions  (PI.  B,  6). 
However,  no  anthers  at  this  stage  of  development  have  been  found  which 
do  not  contain  many  portions  of  the  spireme  which  are  double  for  eon- 
siderable  distances.  It  is  probable  that,  during  the  post  synaptic  stages 
up  to  segmentation  of  the  spireme  into  chromosomes,  the  univalent 
portions  of  the  thread  are  never  fused  throughout  their  whole  length  to 
form  a  single  spireme.  It  is  even  possible  that  no  fusion  takes  place, 
but  that  the  imivalent  threads  only  approach  each  other  so  closely  that 
in  such  delicate  threads  the  line  of  demarcation  can  not  be  distinguished. 

There  is  no  distinct  second  contraction,  but  there  is  a  semblance  of 
one  following  the  loosely-distributed  spireme  stage.  The  thread  con- 
tracts gradually  but  unevenly  throughout  its  whole  length,  its  univalent 
portions  as  a  consequence  becoming  separated  from  one  another  and 
appearing  thicker  (PI.  B,  7).  The  portions  of  loops  which  are  in 
contact  with  the  nuclear  membrane  remain  so  and  often  extend  long 
distances  on  the  periphery  and  then  turn  at  relatively  sharp  angles  and 
again  extend  in  fairly  straight  lines  toward  the  central  mass,  still  situ- 
ated usually  near  the  nucleolus.  Many  of  the  loops  in  this  way  form 
equilateral  triangles.  There  is  no  doubt  at  this  time  of  the  double 
nature  of  these  loops.  The  paired  threads  are  evidently  identical  with 
those  which  appeared  while  passing  out  of  synapsis  and  may  be  identical 
with  the  bivalent  loops  seen  extending  out  from  the  loose  presynaptic 
38325°— 18 3 


646  Journal  of  Agricultural  Research  voi.  xii,  no.  10 

mass.  Continued  contraction  of  the  bivalent  spireme  results  in  seg- 
mentation. 

Segmentation. — In  order  to  determine  definitely  whether  a  telosyn- 
aptic  or  a  parasynaptic  arrangement  of  the  univalent  chromosomes 
prevails  during  the  synaptic  and  postsynaptic  stages,  it  would  seem  nec- 
essary to  determine  the  exact  procedure  from  the  bivalent  condition  just 
previous  to  segmentation,  through  segmentation  to  the  paired  condition 
in  diakenesis.  If  it  can  be  shown  that  the  bivalent  threads  appearing 
during  the  pre  and  post  segmentation  stages  are  identical  in  Fragaria 
spp.,  it  will  be  a  strong  argument  in  favor  of  the  parasynaptic  arrange- 
ment of  the  chromosomes.  As  Fragaria  spp.  is  not  complicated  by  a 
second  contraction  and,  as  the  segmentation  stages  are  rapid,  all  stages 
from  that  shown  in  Plate  B,  figure  7,  to  diakenesis  being  found  within 
a  single  loculus;  and  as  the  stages  during  this  period  are  unusually 
distinct,  such  a  determination  is  not  difficult. 

Digby  {11)  has  recently  presented  the  results  of  a  very  detailed  study 
of  the  cytology  of  Crepis  virens  in  which  the  conclusion  is  reached  that 
there  is  an  end-to-end  arrangement  of  the  chromosomes  during  the 
synaptic  stages.  The  details  from  the  loosening  of  the  synaptic  mass  to 
segmentation  are  very  similar  to  those  in  Fragaria  spp.,  although  the 
chromatin  in  C.  virens  is  apparently  much  more  viscous  and  gives  less 
clear-cut  images  than  does  Fragaria  spp.  Her  figures  76,  78,  and  79 
may  be  considered  as  in  the  same  stages  as  my  figure  7  of  Plate  B  and  to 
present  the  same  condition — that  is  to  say,  a  split  or  a  double  spireme  the 
pairs  of  which  are  somewhat  twisted  about  one  another.  She  considers 
that  these  figures  do  not  show  two  univalent  threads  lying  side  by  side, 
but  that  the  bivalent  loops  are  due  to  the  folding  back  upon  each  other 
of  univalent  segments  during  second  contraction.  At  this  stage  the 
chromatin  mass  became  so  viscous  that — 

it  is  generally  impossible  to  individualize  the  future  three  bivalent  chromosomes. 
The  chromosomes  are  in  fact  evolved  out  of  what  appears  to  be  a  tangle  of  viscous 
nuclear  contents. 

It  appears  hardly  logical  to  conclude  from  this  evidence  that  there  is 
an  end  to  end  rather  than  a  side  to  side  pairing  of  the  univalent 
chromosomes. 

The  nimiber  of  loops  present  it  segmentation  in  species  of  Fragaria 
is  always  less  than  the  number  of  chromosome  pairs  which  appear  at 
diakenesis.  It  has  been  mentioned  that  these  loops  extend  for  long 
distances  on  the  periphery  of  the  nucleus,  forming  more  or  less  regular 
equilateral  triangles.  When  segmentation  takes  place,  it  is  usually  at 
the  outer  angles  of  these  loops  and  at  or  near  the  nucleolus,  which  gen- 
erally forms  more  or  less  of  a  center  from  which  the  loops  radiate.  Thus 
the  bivalent  loops  are  often  divided  into  three  pairs  of  bivalent  chromo- 
somes (PI.  B,  9).     The  pairs  continue  to  contract,  those  attached  to 


Mar.  II.  1918  Sterility  in  the  Strawberry  647 

the  periphery  remaining  in  that  position  and  those  which  have  one  end 
attached  to  the  nucleolus  assuming  a  position  alongside  of  it  (PI.  B, 
II,  14).  Occasionally  one  pair  may  be  attached  both  to  the  nucleolus 
and  to  the  periphery ;  when  there  is  evidence  of  considerable  force  exerted 
by  the  pair  in  contraction  (PI.  B,  11),  The  pairs  continue  to  contract 
(PI.  B,  12,  13),  forming  various  figures  which  have  often  been  described 
in  other  forms,  but  very  few  circles  have  ever  been  seen  at  this  time. 
The  contraction  continues  until  the  typical  diakenesis  stage  is  reached 
when  it  is  often  difficult  to  distinguish  between  the  two  univalent 
chromosomes  of  a  pair  (PI.  B,  14,  15).  Apparently  they  often  fuse, 
as  in  the  multipolar  spindle  stage  they  sometimes  appear  as  single  entities. 
At  diakenesis  26  chromosome  pairs  can  readily  be  counted.  At  this 
stage  in  Minnesota  No.  3,  of  22  counts  made,  19  showed  definitely  26 
chromosome  pairs  and  three  others  showed,  respectively,  24,  25,  and  27. 

A  very  similar  condition  to  the  diakenesis  of  the  pollen  mother  cells  is 
shown  in  the  prophase  of  the  tapetal  cell  divisions.  Here,  however,  in 
place  of  the  26  pairs  of  chromosomes  52  pairs  appear  arranged  about  the 
periphery  of  the  nucleus.  Five  counts  made  at  this  time  showed  in 
three  cases  52,  and  in  two  50  and  54  pairs,  respectively. 

Heterotypic  division. — Diakenesis  in  the  pollen  mother  cells  is  of 
somewhat  long  duration,  but  the  period  between  it  and  the  metaphase 
of  the  heterotypic  division  is  extremely  short,  usually  not  more  than  two 
or  three  multipola'r  spindles  appearing  in  a  loculus  simultaneously.  The 
small  oval  chromosomes  now  arrange  themselves  on  the  equatorial  plate. 
Whether  there  is  any  definite  order  or  arrangement  could  not  be  deter- 
mined, as  the  chromosomes  appear  identical.  They  are  arranged  close 
together  and,  while  their  number  can  not  be  readily  determined,  26  have 
been  counted  on  one  plate  and  24  on  another. 

The  chromosomes  are  then  gradually  pulled  apart  and  drawn  to  the 
poles.  No  irregularities  in  cell  division  or  extrusion  of  chromatin  matter 
have  been  seen  during  this  process.  The  daughter  chromosomes  show 
only  slight  evidence  of  fission  for  the  following  division.  The  disk- 
shaped  daughter  nuclei  are  soon  formed  (PI.  C,  i),  and  directly  after 
prepare  for  the  second  meiotic  division. 

HoMEOTYPic  DIVISION. — The  two  spindles  of  the  homeotypic  division 
may  be  parallel  to  one  another  or  their  axes  may  be  at  right  angles. 
The  metaphase  of  the  division  is  also  characterized  by  great  uniformity, 
the  daughter  chromosomes  separating  and  advancing  toward  the  poles 
simultaneously.  After  separation  they  could  be  readily  counted  and 
showed  in  7  counts  26  chromosomes  (PI.  C,  2). 

The  daughter  nuclei  are  soon  formed,  and  walls  are  laid  down  between 
them,  dividing  the  cytoplasm  evenly.  The  cells  gradually  split  apart, 
separating  the  four  microspores  and  allowing  the  entrance  between 
them  of  the  viscous  material  which  has  up  to  this  time  surrounded  either 


648  Journal  of  Agricultural  Research  voi.  xn,  no.  » 

partially  or  entirely  the  original  mother  cytoplasm   (PL  B,  9;  PL  C, 

I,  3,  4)- 

Pollen  mother  cell  wall. — The  history  of  the  pollen  mother  cell 
wall  is  of  interest  as  it  differs  somewhat  from  that  generally  reported  for 
the  higher  plants. 

In  the  lily  (Allen,  i),  grape  (Dorsey,  12),  and  in  many  other  plants  in 
which  pollen  development  has  been  studied,  it  is  usual,  during  prepara- 
tion for  the  first  meiotic  division,  for  the  pollen  mother  cells  to  separate 
from  one  another,  due  apparently  to  dissolution  of  the  middle  lamella 
and  to  growth  of  the  anther  walls,  forming  a  greater  space  into  which 
the  cells  can  round  up  and  float  free  from  one  another.  Allen  (/,  p.  200) 
considers  that  the  separation  is  due  to  a  dissolution  of  the  cell  walls 
from  between  the  mother  cells,  and  that  each  is  "surrounded  only  by  a 
plasma  membrane."  Following  separation,  a  very  decided  thickening 
of  the  material  surrounding  the  cytoplasm  takes  place  (Pi.  B,  9; 
PI.  C,  i).  This,  Allen  (j),  Tischler  {41),  Stevens  {38),  and  others  speak 
of  as  a  thickening  of  the  mother  cell  wall.  Following  the  formation  of 
the  tetrads,  this  material  increases  and,  as  the  cells  of  the  tetrad  separate 
from  one  another,  flows  between  them.  This  material  is  usually  of  a 
rather  firm  nature  and  in  buckwheat  {38)  often  persists  for  some  time 
after  the  liberation  of  the  microspores  from  it. 

There  is  evidence  in  Fragaria  spp.  which  indicates  that  this  material 
is  entirely  distinct  from  the  mother  cell  wall  and  is  in  no  way  dependent 
on  it  for  its  increase  in  volume,  thus  appearing  to  be  more  of  the  nature 
of  the  gelatinous  sheath  which  surrounds  groups  of  cells  in  many  of  the 
algae. 

In  Fragaria  spp.,  in  place  of  the  pollen  mother  cells  rounding  up  just 
before  or  during  reduction  division,  while  there  is  taking  place  a  rapid 
growth  of  the  anther  walls  and  a  consequent  increase  in  size  of  the 
anther  cavities,  the  cytoplasm  separates  at  the  angles  from  the  walls 
and  rounds  up  independently  (Pi.  B,  9,  14,  15;  Pi.  C,  t).  The  walls 
remain  in  contact  with  one  another  and  adjust  themselves  to  the 
increasing  space  by  stretching.  As  soon  as  evidence  of  rounding  up  of 
the  cytoplasm  appears,  a  gelatin-like  material  is  secreted  unevenly 
about  the  cytoplasmic  mass  (PI.  B,  9;  Pi.  C,  i).  This  material  is 
apparently  identical  with  that  laid  down  about  the  plasmic  mass  in 
the  grape,  lily,  and  forms  like  them  in  which  the  mother  cell  wall  rounds 
up,  supposedly  following  the  dissolution  of  the  mother  cell  wall.  In 
these  cases  it  is  generally  spoken  of  as  the  thickening  of  a  new  mother 
cell  wall.  In  the  strawberry  this  material  increases  in  amount  until  at 
the  completion  of  tetrad  formation  and  before  liberation  of  the  micro- 
spores, the  spores  are  completely  embedded  in  it  (PL  C,  3,  4).  The 
mother  cell  walls  are  still  present,  but  simply  divide  the  anther  cavity 
into  large  spaces,  which  are  only  partially  filled  by  the  tetrad  (PL  C, 
3).     An  examination  of  analogous  stages   to    these   in   the   lily   and 


Mar.  II,  191S  Sterility  in  the  Strawberry  649 

grape  shows  the  mother  cell  wall  lying  closely  about  the  gelatinous 
material  in  which  the  spores  are  embedded,  it  having  not  disappeared 
at  the  period  of  rounding  up  of  the  pollen  mother  cells,  as  is  generally 
assumed.  At  the  time  of  rounding  up  of  the  pollen  mother  cells  in  these 
two  plants,  there  is  no  evidence  of  old  walls  being  left  behind,  but  there 
seems  rather  to  be  a  separation  of  the  contiguous  walls,  due  to  the  dis- 
solution of  the  middle  lamella  and  rounding  up  of  them  with  the  plasma 
masses.  The  walls  give  a  slightly  different  staining  reaction  from  the 
thick  secreted  mass  and  so  can  be  readily  distinguished  from  it.  During 
the  liberation  of  the  spores  in  the  grape,  it  is  not  uncommon  to  see  this 
gelatinous  sheath  completely  disappear,  leaving  portions  of  the  original 
mother  cell  wall  about  the  spores.  This  gradually  dissolves,  liberating 
the  spores.  In  the  strawberry,  the  procedure  is  much  the  same.  The 
gelatinous  sheath,  which  shows  no  wall  closely  about  it  but  only  a  limit- 
ing membrane,  dissolves,  liberating  the  spores  into  the  large  cell  bounded 
by  the  mother  cell  wall.     This  soon  disappears. 

At  the  time  of  liberation  the  spores  have  a  distinct  wall  about  them- 
selves, which  is  independent  of  the  surrounding  sheath.  Apparently 
the  condition  in  the  lily,  grape,  and  strawberry  is  identical,  as  far  as 
structures  are  concerned,  but  differs  primarily  in  the  separation  of  the 
walls  of  contiguous  cells  in  the  former  plants,  while  in  the  latter  only 
the  cytoplasm  of  the  pollen  mother  cells  rounds  up,  leaving  the  walls 
and  middle  lamella  intact. 

Following  the  liberation  of  the  tetrads  and  the  disappearance  of  the 
mother  cell  walls,  degeneration  of  the  tapetum  takes  place.  Tapetal 
degeneration  seems  to  be  in  some  way  correlated  with  the  disappearance 
of  the  middle  lamella  from  between  the  mother  cell  walls,  for  in  those 
forms  which  show  an  early  rounding  up  of  the  mother  cell  walls  there 
is  a  correspondingly  early  degeneration  of  the  tapetal  cells. 

Growth  of  the;  microspores. — Following  the  liberation  of  the 
microspores,  there  is  a  period  of  very  rapid  growth  in  their  walls.  At 
first  this  causes  the  microspores  to  become  very  irregular  in  outline 
(PI.  C,  7),  but  as  growth  continues  and  the  wall  becomes  thicker  the 
cells  become  more  spherical.  The  growth  of  the  microspore  wall  is 
so  rapid  that  there  is  not  a  corresponding  growth  of  the  cytoplasm, 
a  condition  which  results  in  a  large  vacuole  occupying  the  greater  portion 
of  the  cell  cavity.  When  gr«wth  of  the  microspores  is  nearly  complete, 
the  original  cytoplasmic  content  of  the  spore  is  spread  out  over  the 
periphery  of  the  cell  and  about  the  nucleus  (PI.  C,  8).  The  compara- 
tive size  and  consequently  the  very  great  decrease  in  relative  cytoplasmic 
content,  of  the  newly  liberated  microspores  and  those  ready  for  the 
microspore  division  may  be  realized  by  comparing  figures  5  and  8  of 
Plate  C,  which  are  drawn  to  the  same  scale. 

Early  in  the  development  of  the  liberated  microspore  the  wall  is 
differentiated  into  the  extine  and  intine  (PI.  C,  7).     The  extine  gradually 


650  Journal  of  Agricultural  Research  voi.  xii,  no.  10 

thickens  and  a  series  of  scales  are  formed  over  its  surface.  The  ex- 
ternal appearance  of  the  extine  is  shown  in  Plate  C,  figures  15  and  16; 
Plate  D,  figures  6  and  15.  Soon  after  the  microspore  division  the  extine 
development  is  practically  complete. 

Division  of  the  microspore  nucleus. — ^With  the  growth  of  the 
microspore  there  is  an  apparent  decrease  in  chromatin  content;  for, 
in  the  nuclei  which  are  just  about  to  divide,  the  chromatin  is  distributed 
in  small  particles  about  the  periphery  of  the  nucleus  and  appears  very 
scant.  Transition  stages  between  that  seen  in  Plate  D,  figure  8,  and 
the  completion  of  the  spireme  have  not  been  observed.  The  spireme 
(PI.  D,  9,  10)  is  a  continuous,  heavy,  dark-staining  thread.  It  follows, 
more  or  less,  the  periphery  of  the  nucleus  and  surrounds  the  large 
nucleolus  which  has  at  this  time  begun  to  break  down.  The  nucleolus 
takes  on  a  very  irregular  outline,  which  is  in  some  way  related  to  the  nu- 
merous threadlike  processes  which  extend  from  it  to  the  spireme  thread 
(PI.  D,  10).  The  nucleolus  is  now  very  light-staining  and  seems  for  the 
most  part  homogeneous,  but  it  may  contain  one  or  more  small  vacuoles. 
Soon  after  the  segementation  of  the  spireme  into  its  26  constituent  parts 
the  nucleolus  completely  disappears. 

Following  segmentation  of  the  spireme  the  nuclear  membrane  dis- 
appears, the  chromosomes  are  drawn  to  one  plane,  and  the  spindle 
is  formed  (PI.  D,  11).  The  most  usual  position  of  the  spindle  in  the 
microspores  of  many  plants  is  near  the  wall  and  perpendicular  to  it; 
the  pole  which  is  to  form  the  generative  nucleus  being  nearly  in  con- 
tact with  it.  As  a  result  of  this  position  the  generative  cell  usually  lies 
against  the  wall  and  the  vegetative  nucleus  in  the  cell  cavity  (PI.  D,  15). 
This  arrangement  of  the  spindle  is  also  found  commonly  in  Fragaria 
spp.,  but  is  not  universal.  Often  spindles  are  found  which  lie  parallel 
to  the  wall  (PI.  D,  12)  and  result  in  the  arrangement  of  the  nuclei  shown 
in  Plate  D,  figure  14,  both  of  which  lie  against  the  wall. 

The  spindle  is  always  broad  at  the  poles  and  short.'  The  chromo- 
somes on  the  equatorial  plate  are  small  and  oval  in  shape  and  may 
readily  be  counted  if  a  section  can  be  obtained  in  which  the  heavy  wall 
of  the  microspore  is  removed  both  above  and  below  the  chromosomes. 
In  these  sections  26  chromosomes  are  plainly  visible.  The  daughter 
chromosomes  are  drawn  to  the  poles  simultaneously  (Pi.  D,  12),  no 
instances  having  been  seen  in  which  chromosomes  lagged  behind  on 
the  spindle  or  in  which  there  was  an  extrusion  of  chromosomes. 

Directly  after  the  rounding  up  of  the  daughter  nuclei  and  the  dis- 
appearance of  the  spindle  fibers  there  is  visible  no  sign  of  a  cell  wall  or 
limiting  membrane  (PI.  D,  13)  between  the  two  nuclei,  which  is  even- 
tually to  set  off  the  generative  nucleus  in  a  separate  cell  (Pi.  D,  14,  15). 
Soon,  however,  the  wall  appears  and  the  generative  cell  begins  to  round 
up  (PI.  D,  14),  eventually  to  lie  free  in  the  cytoplasm  of  the  pollen  grain 
(PI.  C,  I). 


Mar.  II,  1918  Sterility  in  the  Strawberry  651 

Pollen  maturity. — After  the  microspore  division  there  is  again  a 
period  of  growth  of  the  pollen  grain  and  a  very  marked  increase  in  the 
amount  of  cytoplasm.  Eventually  the  pollen  grain  is  completely  filled 
mth  cytoplasm  of  a  distinctly  alveolar  structure.  During  this  time  the 
cytoplasm  nearly  disappears  from  about  the  microspore  nucleus,  leaving 
the  microspore  cell  wall  loosely  surrounding  the  nucleus,  the  chromatin 
of  which  is  graduallv  taking  on  the  condition  typical  of  resting  nuclei 
(PI.  D,  2). 

When  pollen  formation  is  complete,  changes  take  place  in  the  anthers 
preparatory  to  dehiscence.  These  consist  primarily  in  breaking  down 
the  wall  between  each  pair  of  loculi,  thus  throwing  all  of  the  grains  of 
one-half  of  the  anther  together.  There  is  also  a  general  drying-out 
process  which  results  in  the  disappearance  of  the  liquid  material  which 
previously  surrounded  the  developing  grains,  and  of  a  considerable 
portion  of  moisture  from  the  pollen  grains,  thus  causing  them  to  collapse 
along  three  meridial  sutures  which  fold  in,  thus  giving  in  cross  section 
somewhat  of  a  clover-leaf  shape,  while  the  general  shape  of  the  grain  is 
long-oval.  The  three  germ  pores  are  located  at  the  middle  points  of  the 
sutures.  On  placing  the  dry  pollen  in  water  it  soon  swells  to  form  a 
sphere. 

The  position  of  the  generative  cell  and  vegetative  nucleus  within  the 
collapsed  grains  could  not  be  determined,  for  as  soon  as  the  killing  fluid 
penetrated  the  anthers  the  grains  immediately  swelled.  In  the  swelled 
grains  the  nuclei  lie  in  various  positions  with  regard  to  one  another,  but 
usually  in  close  proximity.  The  vegetative  nucleus  is  generally  spherical 
while  the  generative  cell  contents  are' fusiform,  owing  to  the  folding  of 
the  dried  pollen  grains  and  are  not  closely  surrounded  by  the  cell  wall. 
The  chromatin  of  both  nuclei  now  shows  the  typical  resting  condition 
(PI.  D,  2). 

Up  to  and  including  the  liberation  of  the  microspores  from  the  tetrad, 
the  cells  have  shown  marked  regularity  in  division,  the  stages  of  short 
duration  proceeding  regularly  from  one  end  of  a  loculus  to  the  other, 
while  during  the  stages  of  longer  duration  the  cells  of  a  loculus  all  show 
the  same  degree  of  development.  Up  to  the  time  of  liberation  of  the 
microspores,  the  development  which  has  taken  place  has  depended 
entirely  on  materials  furnished  by  the  sporophyte,  the  one  group  of 
chromosomes  merely  being  the  tools  by  which  the  materials  of  the 
pollen  mother  cell  were  divided  into  four  parts,  the  microspores.  So 
far  as  can  be  seen  up  to  this  point,  no  growth  process  resulting  directly 
in  an  increase  of  cell  material  can  be  directly  attributed  to  the  one  aggre- 
gation of  chromosomes.  At  the  time  of  liberation  from  the  tetrad 
the  spores  are  strikingly  alike  in  size  and  all  other  visible  characters. 
There  has  been,  up  to  this  time,  no  differentiation  in  rate  of  development 
of  individual  microspores;  and  the  spores,  as  liberated,  are  all  normal. 


652 


Journal  of  Agricultural  R,esearch 


Vol.  XII.  No.  10 


DEGENERATION    OF    THE    MICROSPORE 

Liberation  of  the  microspore  from  the  tetrad  places  these  individuals 
upon  their  own  resources  for  future  development.  It  is  true  that  the 
spores  are  surrounded  by  a  nourishing  medium  furnished  by  the  sporo- 
ph3rte,  but  the  abihty  to  use  this  material  depends  upon  the  individual 
cell  metabolism  of  the  microspores. 

A  study  of  the  actual  change  in  cytoplasmic  content  from  the  pollen 
mother  cell  stage  to  the  stage  of  complete  formation  of  the  pollen  grains 
indicates  the  degree  to  which  the  young  spores  are  dependent  upon 
their  own  metabolism  from  the  time  of  liberation  from  the  tetrad  to 
maturity  of  the  male  gametophyte  or  mature  pollen  grain.  Measure- 
ments were  made  of  the  diameter  of  the  spherical  cytoplasmic  mass  at 
the  following  stages  during  pollen  formation:  Rounded  pollen  mother 
cell  at  diakenesis,  young  microspores  which  have  just  been  liberated 
from  the  tetrad,  microspores  which  have  completed  growth  and  in 
which  the  nuclei  are  either  dividing  or  are  about  to  divide,  and  the 
mature  pollen  grains.  A  summary  of  the  measurements  is  given  in 
Table  XI.  It  will  be  seen  by  comparing  the  volume  of  the  pollen  mother 
cells  and  of  the  microspores  at  liberation  that  the  latter  show  slightly 
less  than  one-fourth  the  volume  of  the  former,  indicating  that  during 
reduction  division  and  the  period  subsequent  to  the  liberation  of  the 
microspores  no  increase  in  cytoplasm  has  occurred.  Following  libera- 
tion from  the  tetrad,  however,  when  the  microspores  are  floating  in 
the  nourishing  medium  of  the  loculus  as  independent  units,  very  rapid 
wall  growth  takes  place.  From  <the  time  of  liberation  to  division  of 
the  microspores  nucleus  the  volume  of  the  cells  increases  about  6.4 
times,  but  shows  no  corresponding  increase  in  cytoplasm.  At  the 
mature  pollen  stage  the  cells  have  increased  to  nearly  7.5  times  that 
of  the  liberated  microspore.  The  changes  in  volume  and  cytoplasmic 
content  from  the  pollen  mother-cell  stage  to  mature  pollen  are  also 
well  illustrated  by  Plate  B,  figure  14;  Plate  C,  figures '4  and  8;  and 
Plate  D,  figure  2,  all  of  which  are  drawn  to  one  scale. 

Table  XI. — Volume   of    pollen   mother   cells   and   microspores   at   various   stages   of 

development  . 


Stage. 


Number 

Average 

ured. 

diameter. 

M. 

263 

15-25 

300 

9.41 

200 

17.44 

500 

18.39 

Volume. 


Pollen  mother  cell  at  diakenesis 

Liberated  microspores 

Microspores,  division  stage 

Mature  pollen 


Cm.  /I. 
1,859.99 
436.  28 
2,  777.  40 
3.255-39 


Mar.  II,  1918  Sterility  in  the  Strawberry  653 

Although  the  gametophytic  generation  must  properly  be  considered 
as  beginning  with  the  first  appearance  of  the  haploid  chromosome 
number,  yet  the  liberation  of  the  spores  from  the  tetrad  may  be  con- 
sidered as  marking  the  beginning  of  the  independent  growth  period  of 
this  generation  and  the  rapid  growth  of  the  spore  wall,  division  of  the 
microspore  nucleus,  the  increase  in  cytoplasm,  and  finally  the  germi- 
nation of  the  pollen  grain  and  the  production  of  sperms  as  develop- 
mental stages  in  this  very  much  foreshortened  plant  generation.  The  very 
rapid  wall  growth,  division  of  the  nucleus,  and  finally  the  increase  in 
cytoplasm  to  7.5  times  its  original  volume,  all  within  a  relatively  short 
period,  go  to  make  this  period  of  a  plant's  life  history  probably  the 
most  finely  adjusted  and  critical  one  which  it  is  called  upon  to  survive. 

It  is  during  the  periods  of  rapid  enlargement  and  of  increasing  cyto- 
plasm that  degeneration  of  the  microspores  takes  place. 

Although  there  is  some  evidence  in  the  strawberry  that  degeneration 
may  begin  before  liberation  of  the  microspores  from  the  tetrad,  so  many 
anthers  have  been  examined  which  contain  only  normal  microspores, 
both  in  the  tetrad  stage  and  directly  following  their  liberation,  that 
degeneration  before  this  time  can  be  considered  as  negligible.  It  is 
probable  that  poor  fixation  may  account  for  the  few  apparent  instances 
which  have  been  found.  Indeed,  poor  fixation  constantly  enters  as  a 
disturbing  factor  in  the  interpretation  of  the  condition  of  the  supposedly 
aborting  grains,  and  it  is  only  by  finding  the  same  types  at  later  and 
more  advanced  stages  of  degeneration  when  there  can  be  no  question  as 
to  the  condition  of  the  microspore  that  the  factor  of  poor  fixation  can 
be  eliminated.  No  aborted  microspores  have  so  far  been  found  which 
could  be  referred  to  as  degeneration  within  the  tetrad. 

Directly  following  microspore  liberation  evidences  of  pollen  abortion 
may  be  noticed  and  from  this  period  on  through  the  various  stages  of  de- 
velopment, microspores  and  pollen  grains  are  continually  aborting. 

Plate  D,  figures  3  and  4,  shows  microspores  shortly  after  liberation,  in 
which  the  contents  have  completely  broken  down  into  a  yellow  oily  mass 
which  turns  black  on  exposure  to  osmic  acid.  The  lighter  areas  are 
globules  evidently  of  a  different  substance. 

Ordinarily  during  the  growth  period  of  the  wall  following  liberation, 
the  cytoplasm  becomes  spread  over  the  periphery  in  a  thin  layer.  Plate 
D,  figures  5,  6,  and  7,  represent  various  conditions  in  which  this  normal 
process  has  not  occurred.  Figures  7  and  1 1  are  later  stages  of  this  type 
of  abortion  in  which  degeneration  of  the  nucleus  and  cytoplasm  is  grad- 
ually taking  place.  Figure  15  apparently  represents  a  further  stage  of 
this  series,  in  which  the  nucleus,  although  visible  in  outline,  is,  with  the 
cytoplasm,  entirely  functionless.  Plate  E,  figure  11,  shows  the  com- 
pletely degenerate  microspore  of  this  type  found  among  normal  mature 
pollen.     The  dark-staining  mass  is  yellow  and  oily  before  killing. 


654  Journal  of  Agricultural  Research  voi.  xii,  no.  io 

Although  there  has  been  a  considerable  amount  of  degeneration  up  to 
the  time  of  the  completion  of  growth  of  the  microspore  wall,  probably 
more  takes  place  between  the  time  of  the  formation  of  the  large  vacuole 
(Pi.  C,  8)  and  the  completion  of  the  microspore  division  than  at  any 
other  period. 

Plate  D,  figure  12,  shows  a  full-grown  microspore  in  which  degeneration 
is  taking  place,  both  in  the  cytoplasm  and  in  the  nucleus.  This  was  found 
in  an  anther  containing  full-grown  i -nucleate  microspores,  and  probably 
is  a  case  in  which  degeneration  has  begun  during  the  period  of  enlarge- 
ment. 

Plate  D,  figure  13,  and  Plate  E,  figure  2,  represent  early  stages  in  de- 
generation of  I -nucleate  microspores  which  are  at  the  stage  at  which 
division  should  take  place  as  they  were  found  in  anthers  containing  both 
I-  and  2-nucleate  grains,  as  well  as  division  stages.  The  microspore  rep- 
resented in  Plate  E,  figure  7,  was  found  among  microspores  of  the  stage 
of  development  shown  in  Plate  C,  figure  15,  while  that  of  Plate  E,  figure 
8,  was  found  among  mature  pollen  grains.  Both  are  evidently  later 
stages  of  the  condition  represented  in  Plate  D,  figures  12  and  13,  and  Plate 
E,  figure  2. 

During  the  period  of  division  a  few  microspores  contain  a  very  scant 
amount  of  cytoplasm.  Such  spores  are  seen  in  Plate  D,  figures  14  and  16, 
and  Plate  E,  figures  i  and  4.  Eventually  these  completely  degenerate. 
Those  of  the  type  shown  in  Plate  D,  figure  16,  and  Plate  E,  figure  i ,  prob- 
ably are  the  forerunners  of  the  completely  degenerate  microspore  shown 
in  Plate  E,  figure  10,  which  is  a  common  type  in  mature  anthers. 

Although  degeneration  takes  place  to  a  greater  or  less  degree  at  all 
stages,  from  microspore  liberation  to  microspore  nucleus  division,  no 
instances  of  degeneration  occurring  during  the  process  of  division  have 
been  noted.  Directly  following  division,  however,  evidences  of  degener- 
ation again  become  apparent,  although  much  less  numerous  than  in  the 
period  just  prior  to  division.  Figure  3  of  Plate  E  represents  an  early 
stage  of  degeneration  directly  following  microspore  division.  This  is 
probably  an  early  stage  of  the  more  advanced  degeneration  stage  shown 
in  Plate  E,  figure  I.  This  young  pollen  grain  was  found  among  grains  in 
which  the  cytoplasm  was  increasing  rapidly  in  amount.  It  will  be  noticed 
that  the  generative  cell  is  aborting,  while  the  remainder  of  the  grain 
is  normal.  Apparentiy  grains  of  this  type  continue  to  increase  normally 
with  regard  to  the  vegetative  portion,  while  degeneration  of  the  genera- 
tive cell  occurs.  Figure  1 3  of  Plate  E  represents  such  a  grain  found  in  a 
mature  anther.  Figure  12  represents  another  type  of  degeneration 
which  takes  place  subsequent  to  division  of  the  microspore  nucleus.  The 
generative  cell  has  completely  degenerated.  The  vegetative  nucleus, 
although  still  present  in  outline,  is  f unctionless,  while  the  cytoplasm  bears 
no  resemblance  to  the  normal.    It  still  retains  the  property,  however,  of 


Mar.  ir.  1918  Sterility  in  the  Strawberry  655 

absorbing  liquids  when  placed  in  them,  and  so  the  grains  of  the  type 
shown  in  Plate  E,  figure  13,  are  likely  to  cause  inaccuracies  when  the  ordi- 
nary methods  of  determining  the  percentage  of  good  and  abortive  pollen 
are  employed. 

It  seems  clear  thus  far  that  degeneration  of  the  microspores  and  pollen 
grains  is  a  phenomenon  closely  related  to  the  very  active  metabolic  proc- 
esses which  are  taking  place  during  this  period  of  the  plant's  life  history. 

DISCUSSION  OF  RESULTS 

T"hus  far  there  has  been  shown  to  exist  in  the  strawberry  types  of 
sterility  due  to  at  least  two  distinct  causes. 

In  some  of  the  wild  species,  including  F.  elatior,  F.  platypeiala,  F. 
cuneifolia,  and  F.  chiloensis,  it  seems  highly  probable  that  the  species 
are  diecious,  while  F.  virginiana  is  unquestionably  so  for  the  most  part. 
Dieciousness  is  expressed  in  the  production  of  pistillate  plants  bearing 
staminodia,  which,  so  far  as  I  have  observed,  never  produce  pollen. 
The  staminate  plants  bear  normal  stamens  and  pistils,  which  appear 
normal  but  which  seldom  are  fertile.  Certain  types  are  also  found  which 
are  intermediate  between  these  two  types.  A  few  staminate  plants  may 
bear  fruit  on  one  or  more  of  the  early  flowers.  These  flowers  may  or 
may  not  bear  stamens,  but  both  staminodia  and  intermediate  anthers 
are  found  on  them.  Other  plants  which  are  apparently  staminate 
develop  only  intermediate  anthers  in  which  abortion  takes  place  at  the 
tetrad  division  or  shortly  after,  resulting  in  a  degenerate  mass  in  the 
anther.  Another  staminate  type  has  recently  been  studied  which  bears 
fruit  on  the  primary  and  occasionally  on  some  of  the  other  flowers. 
The  anthers  of  the  primary  flowers  are  reduced  to  staminodia.  The 
secondary  flowers  carry  pollen  development  through  to  the  liberation 
of  the  microspores  when  40  to  80  per  cent  of  them  abort  and  degenerate 
completely,  forming  a  yellow  oily  mass.  The  remaining  grains  develop 
normally  and  are  fertile.  These  types  are  all  found  in  plants  of  pure  F. 
■virginiana  and  represent  varying  degrees  of  expression  of  dieciousness. 

The  diecious  condition  of  the  wild  species  from  which  the  cultivated 
forms  have  been  derived  probably  explains  the  greater  sterility  of  the 
later  flowers  of  the  cultivated  hermaphroditic  varieties,  than  is  found 
in  those  of  the  pistillates ;  if  we  can  accept  the  origin  of  the  hermaphro- 
dites as  being  from  males  which  have  developed  partial  fertility  of  the 
female  organs.  The  appearance  of  staminodia  showing  varying  degrees 
of  development  in  the  cultivated  varieties  is  also  the  direct  result  of 
dieciousness,  while  the  intermediate  types  of  anthers  in  the  cultivated 
forms  are  of  the  same  nature  as  those  found  in  wild  staminate  clones. 
It  is  an  interesting  fact,  in  connection  with  the  problem  of  sex  deter- 
mination and  dieciousness,  that  where  intermediate  anthers  or  stami- 
noids  occur,  either  on  wild  clones  or  cultivated  varieties  which  are  able 


656  Journal  of  Agricultural  Research  voi.  xii,  No.  10 

to  bear  normal  anthers,  they  are  practically  always  borne  on  the  primary 
flowers,  while  the  anthers  produced  later  have  a  greater  tendency  toward 
normal  development.  The  tendency  toward  the  production  of  stami- 
nodia  is  much  greater  in  the  early  spring  than  later.  On  the  other  hand, 
pistil  sterility  is  much  more  frequent  on  the  later  flowers  of  an  inflores- 
cence and  when  fruits  set  on  wild  staminate  clones  it  is  practically  always 
on  the  first  flowers  of  a  cluster  which  open. 

A  second  type  of  sterility  often  associated  with  the  above  type  but  due 
to  a  different  cause  is  that  which  results  in  aborted  microspores  and 
pollen  grains  in  otherwise  normal  anthers.  Aborted  pollen  has  been 
shown  to  be  present  in  relatively  small  amounts  in  pure  species  of  Fragaria, 
but  appears  often  in  large  quantities  in  many  of  the  cultivated  varieties. 
This  type  of  abortion  has  long  been  recognized  in  hybrids,  and  recently 
Jeffrey  and  his  students  have  gone  so  far  as  to  consider  any  plant  bearing 
ever  15  or  20  per  cent  of  this  type  of  pollen  a  hybrid. 

Selfed  seedlings  and  F^  plants  of  crosses  between  varieties  of  cultivated 
strawberries  are  so  extremely  variable  for  many  factors  that  it  seems 
self-evident  that  they  are  of  hybrid  origin,  and  this  is  to  a  great  extent 
confirmed  by  what  is  known  of  the  origin  of  the  numerous  cultivated 
varieties,  many  of  which  are  the  result  of  variety  crosses,  while  by  far 
the  larger  number  are  chance  seedlings.  It  thus  seems  evident  that 
pollen  abortion  in  the  cultivated  varieties  is  due  to  the  same  causes 
which  produce  sterility  in  other  hybrids. 

As  would  be  expected,  there  are  varying  degrees  of  sterility  resulting 
from  hybridization  and  varying  degrees  of  irregularities  in  the  stages 
which  lead  up  to  the  final  abortion  of  pollen.  There  appear  in  the 
literature  numerous  instances  of  abortion  in  both  male  and  female  repro- 
ductive organs  following  irregular  reduction  divisions.  The  irregular 
divisions,  especially  in  the  pollen  mother  cells,  result  in  the  production 
of  more  than  four  cells  of  unequal  size  in  the  tetrad.  These  produce 
microspores  of  varying  sizes,  few  of  which  ever  come  to  maturity.  Gates 
(77,  p.  98)  pointed  out  that  most  of  the  forms  studied  by  Wille  (44) 
showing  supernumerary  cells  in  the  tetrad  are  either  hybrids  or  have 
been  under  cultivation  for  some  time  and  are  open  to  the  suspicion  of 
being  hybrids.  Other  plants,  some  of  which  are  knowTi  to  be  hybrids 
while  others  which  have  been  cultivated  as  horticultural  varieties  and 
are  under  suspicion  as  hybrids,  have  been  studied  in  more  detail  by 
various  workers.  Tischler  (40,  41,  42),  who  has  done  much  work  with 
plants  of  this  type,  finds  that  in  hybrids  of  Ribes  spp.  and  in  the  sterile 
hybrid  Mirahilis  jalapa  X  M.  iubiflora  pollen  degeneration  usually  takes 
place  following  normal  divisions.  In  the  hybrids  Poientilla  tabernae- 
montani  X  P.  rubens,  Syringa  chinensis,  and  Bryonia  alba  X  B.  dioica, 
and  in  three  varieties  of  banana  {Musa  paradisiaca)  having  different 
chromosome  numbers,  irregular  divisions  are  common  and  are  always 
followed   by   much   pollen   abortion.     In   these   banana   varieties,    the 


Mar.  II.  1918  Sterility  in  the  Strawberry  657 

origins  of  which  are  unknown  but  which  dififer  cytologically  in  having 
8,  16,  and  24  chromosomes  as  the  reduced  number,  it  is  significant  that 
the  most  frequent  irregularities  in  cell  division  during  reduction,  and 
most  complete  pollen  sterility  occurs  in  the  two  varieties  having  the 
greater  chromosome  numbers.  Thus,  pollen  abortion  may  or  may  not 
be  the  result  of  irregularities  at  reduction  division,  but  is  apparently 
related  to  hybridity  and  is  associated  with  heterozygosity. 

After  working  with  several  hybrid  plants  showing  both  normal  divi- 
sions and  irregularities  during  reduction  and  formation  of  tetrads, 
Tischler  (41,  p.  144)  concluded  that — 

Die  Sterilitat  bei  Hybriden  hangt  nicht  von  irgendwelcher  Chromatin  repulsion  ab. 

He  concluded  further  that  irregularities  during  tetrad  division  can  not 
be  considered  as  characteristic  only  of  hybrids.  He  thought  that  ste- 
rility of  hybrids  was  due  to  the  coming  together  of  two  sex  cells  which 
did  not  contain  identical  developmental  tendencies  and  that  these  were 
expressed  at  the  critical  time  of  the  formation  of  the  reproductive  organs. 
Actual  abortion  of  the  grains  he  thought  was  due  to  insufficiency  of  cyto- 
plasm in  the  enlarged  microspores. 

One  of  the  most  striking  cases  of  sterility  following  hybridization,  the 
cytological  details  of  which  have  been  worked  out,  is  that  of  the  hybrid 
Drosera  longifolia  X  D.  rotundifolia,  reported  by  Rosenberg  (jj).  The 
striking  feature  of  this  hybrid  is  that  it  is  between  parents  having  differ- 
ent chromosome  numbers,  the  diploid  number  of  D.  longifolia  being  40, 
while  that  of  D.  rotundifolia  is  only  20.  As  a  consequence  the  hybrid 
contains  30.  At  reduction  division  in  both  the  megaspore  mother  cells 
and  pollen  mother  cells  there  appeared  10  pairs  of  chromosomes  and  10 
single  ones,  the  pairs  supposedly  being  made  of  the  10  D.  rotundifolia 
chromosomes  paired  with  lo  from  the  D.  longifolia  parent,  while  the  10 
single  chromosomes  were  the  remaining  10  D.  longifolia  chromosomes. 
Reduction  division  resulted  in  the  separation  of  the  paired  ones,  these 
being  drawn  regularly  to  opposite  poles.  The  unpaired  chromosomes, 
on  the  other  hand,  were  either  drawn  to  one  or  the  other  pole  or  were 
left  in  the  cytoplasm  to  form  another  small  nucleus.  The  homoeotypic 
divisions  took  place  normally.  Following  the  organization  of  the  micro- 
spores within  the  tetrad,  many  proceeded  to  increase  in  size;  in  some, 
division  of  the  microspore  nucleus  proceeded  normally,  and  then  in 
practically  all  cases  abortion  of  the  pollen  took  place.  Following  tetrad 
formation  in  the  feinale  reproductive  organs,  three  of  the  tetrads  usually 
aborted,  as  is  common,  while  the  other  proceeded  to  form  the  egg  sac. 
Egg-sac  formation  was  carried  to  various  stages,  but  it  was  only  very 
rarely  that  a  perfect  egg  sac,  capable  of  further  development,  was  formed. 
Rosenberg  concluded  {p.  39)  that  because  of  the  fact  that  the  micro- 
spore division  was  able  to  proceed  normally,  the  degeneration  of  the 
pollen  grains  was  not  the  result  of  the  irregular  distribution  of  chromo- 


658  Journal  of  Agricultural  Research  voi.  xii.  No.  i© 

somes  during  reduction  division,  but  was  due  to  a  lack  of  cytoplasm. 
The  abortion  of  the  egg  sacs,  he  again  concluded,  was  not  due  to  irregular 
divisions,  as  all  of  the  divisions  following  reduction  were  normal.,  but  was 
due  to  poor  nutrition. 

Nakao  (29) ,  working  on  the  cytology  of  certain  grain  hybrids  in  which 
very  striking  irregularities  in  reduction  division  took  place,  followed  by 
complete  abortion  of  the  microspore  after  liberation,  concluded  that 
abortion  in  this  case  was  due  to  an  insufficiency  of  cytoplasm  previous 
to  reduction  division  which  resulted  in  abnormally  early  division  and  con- 
sequent irregularities.  He  did  not  consider  why  these  irregularities 
should  cause  abortion  of  the  microspores. 

In  view  of  the  conditions  existing  within  an  anther  at  the  time  of 
degeneration  of  microspores  and  because  of  certain  genetic  results  which 
can  only  be  explained  on  the  basis  of  selective  elimination  of  certain 
gametic  combinations,  it  is  difficult  to  agree  with  the  conclusions  of 
Tischler  and  Rosenberg  that  degeneration  has  nothing  to  do  either  with 
the  irregularities  or  normal  repulsion  which  occur  at  reduction  division 
in  hybrids. 

Although  there  are  striking  differences  in  the  regularity  with  which 
reduction  takes  place  in  sterile  or  partially  sterile  hybrids,  there  are 
certain  conditions,  in  both  those  which  proceed  normally  and  those 
which  show  irregularities,  which  are  alike  and  must  be  taken  into  account 
in  the  consideration  of  the  causes  of  pollen  abortion.  These  conditions 
are  as  follows : 

(i)  At  reduction  division  there  is  a  sorting  out  of  the  parent  chromo- 
somes, resulting  in  new  combinations  in  the  daughter  cells,  the  mnuber 
of  which  depends  upon  the  degree  of  difference  between  the  two  parents. 
If  division  proceeds  normally,  there  is  an  equal  number  of  chromosomes 
in  each  daughter  nucleus.  If  it  takes  place  irregularly,  imequal  numbers 
are  found  in  the  resulting  daughter  cells.  In  either. case  the  combina- 
tions are  new  and  may  or  may  not  contain  all  of  the  properties  necessary 
for  perfect  metabolism  of  the  cell. 

(2)  If  divisions  take  place  regularly,  there  is  an  equal  quantitative 
and,  as  far  as  can  be  determined,  also  an  equal  qualitative  division  of 
cytoplasm  between  the  quadrants  of  a  tetrad.  If  divisions  have  pro- 
ceeded irregularly,  the  cytoplasm  is  divided  between  the  members  of  the 
tetrad  in  proportion  to  the  amoimt  of  chromatin  which  they  contain. 
In  either  case  at  the  time  of  liberation  from  the  tetrad,  or  if  liberation 
does  not  take  place,  as  in  Drosera  spp.,  at  the  period  previous  to  enlarge- 
ment, all  of  the  microspores  appear  normal — that  is,  they  contain  an 
organized  nucleus  and  are  filled  with  cytoplasm.  In  Minnesota  3 
there  is  at  this  time  as  great  uniformity  in  size  and  cytoplasmic  content 
of  the  individual  microspores  as  is  fotmd  in  entirely  fertile  plants  of 
F.  virginiana. 


Mar.  II,  1918  Sterility  in  the  Strawberry  659 

(3)  It  is  not  until  rapid  growth  of  the  microspores  takes  place  and  the 
necessity  of  active  cell  metabolism  appears  that  evidences  of  degenera- 
tion appear.  The  necessity  of  active  metabolism  becomes  apparent 
when  it  is  remembered  that  the  microspores  increase  7.5  times  their 
original  volume  during  this  growth  period. 

(4)  There  is  no  specific  time  at  which  degeneration  of  the  grains 
within  a  single  anther  takes  place.  In  most  of  the  sterile  forms  thus  far 
studied  a  series  of  degenerating  stages  appear  from  the  first  period  of 
growth  of  the  microspores  to  the  formation  of  nearly  mature  pollen. 

It  is  becoming  more  and  more  evident  that  the  growth  and  develop- 
ment of  plants  and  animals  are  directly  dependent  upon  the  chromosome 
combination  which  they  contain.  Boveri  {see  Morgan,  28,  p.  55),  in 
working  with  dispermic  sea-urchin  eggs,  found  that  they  very  rarely 
develop  normal  individuals,  while  if  separated  at  the  4-celled  stage 
normal  individuals  often  developed.  This  seems  dependent  upon  the 
fact  that  in  the  first  four  cells,  which  are  the  result  of  a  single  division, 
the  chances  of  one  of  the  cells  receiving  at  least  one  of  each  kind  of 
chromosome  are  relatively  high  and  thus,  when  separated,  some  may 
develop  normally.  On  the  other  hand,  the  chances  of  each  of  the  four 
cells  receiving  one  of  each  kind  of  chromosome  necessary  for  perfect 
development  are  small;  and  as  a  result  the  individual  develops  abnor- 
mally. Bridges  (5)  has  shown  that  in  the  fruit  fly  Drosophila  ampelo- 
phila  certain  variations  from  the  normal  chromosome  combinations 
have  a  definite  effect  upon  the  development  of  the  zygote.  Zygotes 
containing  3X  chromosomes  die,  while  those  containing  2X  and  a  Y 
chromosome  develop  normally.  Male  individuals  may  develop  which 
contain  an  X  but  no  Y  chromosome,  but  were  found  to  be  entirely  sterile; 
while  those  zygotes  which  received  only  a  Y  chromosome  died  as  did 
also  those  which  received  2Y  chromosomes,  but  no  X.  Zygotes  con- 
taining 2Y  chromosomes  plus  an  X,  however,  were  able  to  develop  into 
normal  males.  Apparently  the  presence  of  an  X  chromosome  is  neces- 
sary for  the  development  of  an  individual,  while  the  presence  of  a  Y  in 
males  is  necessary  if  the  male  is  to  be  fertile. 

In  plants  the  evidence  for  the  dependence  of  development  upon 
chromosome  combinations  is  becoming  indisputable  if  the  Mendelian 
interpretation  of  the  inheritance  of  factors  is  admitted.  In  F^  progeny 
of  hybrids,  if  the  parents  are  homozygous,  there  is  generally  as  much 
uniformity  as  is  shown  by  either  parent,  while  the  Fj  progeny  shows  a 
wide  range  of  types,  often  overstepping  the  limits  of  the  parents.  If 
such  a  variety  of  types  with  regard  to  hardiness,  rust  resistance,  adapta- 
bility to  various  regional  and  soil  conditions,  and  vigor  of  the  individual 
plant  are  produced  in  the  2X  generation  as  the  result  of  new  chromo- 
some combinations,  why  is  it  not  possible  for  a  similar  series  to  exist  in 
the  iX  generation  with  regard  to  the  ability  of  the  individuals  to  develop 
in  a  given  environment? 


66o  Journal  of  Agricultural  Research  voi.  xii,  no.  lo 

I  have  already  pointed  out  that  up  to  the  time  of  liberation  of  the 
microspores  from  the  tetrad,  in  the  strawberry,  cell  divisions  have 
resulted  merely  in  an  equal  division  of  the  cytoplasm  of  the  original 
mother  cell  between  its  four  granddaughter  cells  with  no  evidence  of 
any  metabolic  changes  resulting  in  an  increase  of  cytoplasm.  A  similar 
condition  exists  in  those  forms  which  show  irregular  divisions  during 
reduction  with  uneven  distribution  of  cytoplasm  between  the  resulting 
cells.  Liberation  of  the  microspores  from  the  tetrad  marks  the  end  of 
the  period  of  dependence  of  these  cells  upon  the  2X  generation,  as  far 
as  future  growth  and  development  is  concerned.  At  this  time  there  is 
no  difiference  between  the  conditions  surrounding  the  microspores  of  a 
hybrid  and  those  of  a  genetically  pure  individual.  Both  groups  of 
microspores  are  set  free  in  a  homogeneous  anther  sap  to  complete  their 
own  further  development. 

The  progress  of  the  developmental  stages  in  plants  of  pure  F. 
virginiana  is  characterized  by  great  regularity  of  development  of  the 
individual  microspores  within  a  loculus  with  regard  to  rate  of  enlarge- 
ment, time  of  division  of  the  microspore  nucleus,  and  the  subsequent 
development  of  cytoplasm.  In  fact,  there  is  as  great  uniformity  shown 
in  these  stages  as  was  shown  in  the  stages  leading  up  to  the  mature 
tetrad  stage  in  either  F.  virginiana  or  Miimesota  3.  Minnesota  3,  on 
the  other  hand,  shows  great  irregularities  during  this  developmental 
period  in  rate  of  growth  of  the  individual  microspores,  in  the  time 
of  division  of  the  microspore  nucleus,  and  in  the  rapidity  of  formation 
of  cytoplasm.  This  lack  of  uniformity  is  in  striking  contrast  to  the 
uniformity  shown  in  earlier  stages  of  the  same  plant,  while  the  cells 
were  dependent  on  the  2X  individual.  The  liberated  microspores  of 
Minnesota  3  are  strikingly  uniform  in  size  and  cell  contents. 

During  any  period,  following  liberation  of  the  microspore  to  the 
completion  of  development,  microspores  or  pollen  grains  may  be  found 
degenerating.  As  all  of  the  grains  within  a  loculus  are  free  in  a 
homogeneous  nutrient  liquid,  it  seems  difficult  to  believe  that  the  varia- 
tions in  development  can  be  due  to  anything  but  the  individual 
constitution  of  the  microspores. 

Indeed,  there  is  constantly  accumulating  an  increasing  amount  of 
evidence  which  points  to  the  continual  elimination  of  gametes  bearing 
certain  chromosome  combinations.  In  1894  Millardet  (27)  reported  on 
a  series  of  hybrids  between  species  of  strawberry  the  progeny  of  which, 
he  said,  formed  an  exception  to  the  general  rule  of  hybrids,  as  the  specii&c 
type  of  one  or  the  other  parent  was  always  produced  in  the  first  and  later 
generations.  The  specific  type  shown  in  the  second  generation  was, 
with  one  exception,  the  same  as  that  shown  by  its  parent  in  the  F^ 
generation.  Millardet  mentions  complete  sterility  in  one  species  of  cross 
and  high  percentages  of  sterility  in  the  F^  generation  of  some  of  the  other 
combinations.     Bellair  (j)  reports  that  in  the  tobacco  cross  Nicotiana 


Mar.  II.  1918  Sterility  in  the  Strawberry  661 

sylvesiris  X  N.  tahacum  the  F^  generation  resembled  the  A'',  tabacum 
parent  and  was  partially  fertile.  From  these  he  was  able  to  obtain  F3 
plants  apparently  identical  with  the  two  parents  and  fully  fertile.  The 
reappearance  of  types  similar  to  the  parents  in  large  numbers  in  the 
F3  generation  suggests  the  elimination  of  gametes  containing  combinations 
which  would  result  in  intermediate  types. 

Detlefsen  {10),  working  with  animals,  reports  results  obtained  from 
cavy  crosses  which  may  readily  be  explained  on  the  basis  of  the  elimina- 
tion of  certain  combinations  in  the  gametes  of  the  males.  He  crossed 
.tame  females  to  wild  males.  The  F^  males  were  all  sterile.  The  cross  F^ 
female  with  wild  male  was  not  very  successful  and  produced  one  sterile 
male  and  a  sterile  female.  The  F^  females  crossed  to  tame  males  gave 
sterile  males  of  which  a  few  produced  some  nonfunctional  sperms.  The 
females  of  this  back-cross  again  crossed  to  tame  males  produced  males 
showing  a  low  degree  of  fertility.  As  this  process  was  continued,  always 
crossing  back  to  tame  males,  the  fertility  of  the  male  progeny  increased 
as  they  became  more  nearly  homozygous  for  the  tame  condition.  In  the 
sixth  generation  all  of  the  males  produced  sperms,  and  66.7  per  cent  of 
these  males  were  readily  fertile.  Apparently  the  more  chromosomes  of 
one  parent  type  which  were  present  in  the  sperm,  the  greater  its  chance 
of  complete  development. 

East  {13)  in  a  short  abstract  gives  the  conditions  which  he  found  in 
the  progency  of  the  partially  sterile  hybrid  Nicotiana  rustica  humilis  X  iV. 
paniculata.  The  Fj  progeny  of  this  hybrid  were  very  uniform,  but  only 
I  to  6  per  cent  of  the  female  gametes  were  functional,  and  2  to  6  per 
cent  of  the  pollen  grains  were  morphologically  perfect.  In  the  V^  gener- 
ation some  perfectly  fertile  plants  were  found,  many  possible  Fg  combina- 
tions were  omitted,  many  more  homozygous  combinations  occurring 
than  should  be  expected,  and  the  parent  types  appeared  once  in  everv 
100  to  200  plants,  whereas  if  all  of  the  possible  Fj  combinations  appeared, 
the  parent  types  would  be  much  more  rare.  East  considered  that  the 
results  might  be  explained  on  the  basis  either  of  selective  elimination  of 
F2  zygotes  or  selective  elimination  of  F^  nonfunctional  gametes.  He  con- 
sidered further  that  the  elimination  of  the  nonfunctional  gametes  might 
be  due  to  irregularities  of  chromosome  distribution,  which  scheme  seemed 
improbable;  or  the  facts  might  be  interpreted  without  this  assumption  if 
certain  conditions  were  met  which  are  as  follows : 

If  (i)  there  is  a  group  of  chromosomes  in  each  parent  that  can  not  be  replaced  by- 
chromosomes  from  the  other  parent;  if  (2)  there  is  a  group  of  chromosomes  from  each 
parent,  a  percentage  of  which  may  be  replaced  by  chromosomes  from  the  other  parent, 
but  where  functional  perfection  of  the  gametes  varies  as  their  constitution  approaches 
that  of  the  parental  forms;  if  (3)  there  are  other  chromosomes  that  have  no  effect  on 
fertility  and  therefore  can  promote  recombinations  of  characters  in  the  progeny  of 
fertile  Fg  plants;  if  (4)  a  naked  male  nucleus  entering  the  normal  cytoplasm  of  the 
egg  in  the  immediate  cross  can  cause  changes  in  the  cytoplasm  that  will  affect  future 
reduction  divisions;  if  (5)  this  abnormally  formed  cytoplasm  is  not  equitably  dis- 
38325°— 18 4 


562  Journal  of  Agricultural  Research  voi.  xii,  no.  w 

tributed  in  the  dichotomies  of  gametogenesis  in  the  Fj  generation;  if  (6)  it  follows 
from  (4)  and  (5)  that  Fj  zygotes  may  be  formed  which  are  less  perfect  in  their  gamete 
forming  mechanism  than  those  of  the  Fj  generation;  and  if  (7)  the  heterotypic  division 
of  gametogenesis  does  not  necessarily  form  two  cells  alike  in  their  viability. 

In  the  strawberry,  in  which  no  irregular  distribution  of  chromatin 
occurs,  certain  of  these  assumptions — namely  i,  2,  3,  and  7 — would  apply 
in  explaining  partial  male  sterility  in  many  partially  sterile  varieties, 
but  assumptions  4  and  5  and  assumption  6,  which  is  dependent  upon 
them,  can  hardly  be  considered  applicable;  assumption  4  because  there 
is  no  cytological  evidence  that  there  has  been  any  disturbance  of  reduc- 
tion division;  5,  because  there  is  apparently  equal  distribution  of  the 
mother  cell  cytoplasm  to  each  member  of  the  tetrad,  and  6,  because 
there  is  no  cytological  evidence  which  would  lead  one  to  believe  that  the 
cytoplasm  of  the  zygote  had  anything  to  do  with  the  perfection  of  its 
gamete  forming  mechanism.  In  the  strawberry  sterility  seems  to  be 
due  to  the  inability  of  certain  chromosome  combinations  to  use  the  food 
material  in  which  they  are  embedded  in  the  growth  and  development  of 
the  liberated  spore  to  a  ripe  pollen  grain. 

If  we  can  accept  the  hypothesis  of  pollen  abortion  being  due  in  hybrids 
to  certain  chromosome  combinations  affecting  the  normal  metabolism 
of  the  microspore  in  its  development,  the  question  at  once  presents 
itself:  Is  pollen  abortion  the  result  of  the  presence  of  one  particular 
chromosome  or  of  certain  combinations  of  two  or  more,  or  do  all  of  the 
chromosomes  play  some  part  in  it  ?  This  question  can  not  be  answered 
from  the  facts  so  far  obtained  in  the  strawberry  because  of  the  heterozy- 
gous condition  of  the  material  which  has  been  used;  but  there  is  other 
published  evidence  which  throws  some  light  on  this  question. 

Belling  (4)  has  made  a  very  careful  study  of  partial  sterility  of  hybrids 
between  four  species  of  the  "bean"  Stizolobium.  He  found  that  the  F^ 
plants,  of  those  crosses  in  which  the  velvet  bean  {S.  deeringianum)  was 
used  as  one  parent  continually,  aborted  one-half  of  the  pollen  grains  and 
one-half  of  the  egg  sacs.  Of  the  second  generation  plants  one-half  were 
completely  fertile  and  one-half  partially  sterile,  as  in  Fj.  The  progeny 
of  fertile  F2  plants  continued  to  be  fertile,  while  the  progeny  of  the 
partially  sterile  plants  were  one-half  fertile  and  one-half  partially  sterile. 
Belling  explained  these  results  on  the  basis  of  the  presence  in  the  velvet 
bean  of  the  factor  K,  which  was  not  present  in  the  other  three  varieties. 
These  three,  however,  contained  the  factor  L  not  present  in  the  velvet 
bean.  The  presence  of  either  K  or  L  he  assumed  to  be  necessary  for  the 
normal  development  of  either  egg  sacs  or  pollen  grains,  the  presence  of 
both  K  and  L  causing  abortion.  We  may  extend  this  working  hypoth- 
esis slightly  and  put  it  on  a  chromosome  basis,  in  which  case  we  must 
consider  that  Belling's  factors  K  and  k  form  one  allelomorphic  pair, 
being  situated  in  a  certain  loculus  of  a  specific  chromosome  in  the  velvet 
and  the  other  three  varieties  of  beans,  respectively,  and  that  the  factors 


Mar.  II,  1918  Sterility  in  the  Strawberry  663 

L  and  1  form  an  allelomorphic  pair  located  in  a  definite  but  different  chro- 
mosome of  the  three  varieties  and  the  velvet  bean,  respectively.  Then, 
to  follow  out  Selling's  scheme,  the  presence  of  the  chromosomes  bearing 
the  factors  K  and  L  in  the  same  member  of  a  tetrad  causes  abortion,  and 
likewise  the  presence  of  both  chromosomes  lacking  the  factors  K  and  L 
causes  abortion.  Belling  states  that  abortion  of  the  microspores  takes 
place  in  the  vacuolate  stage  and  that  there  are  no  intermediates  between 
the  completely  aborted  and  the  most  perfect  grains,  thus  strengthening 
the  idea  that  in  this  case  no  more  than  two  chromosomes  have  to  do 
with  abortion. 

Another  set  of  studies  which  point  to  one  instead  of  two  chromosomes 
being  the  cause  of  pollen  abortion  are  those  of  Shull  {35,  36)  on  the 
inheritance  of  sex  and  of  a  sex-linked  factor  in  Lychinis  dioica.  In 
the  first  of  these  studies  Shull  showed  that  very  probably  L.  dioica 
9  is  homozygous  for  the  sex  determining  factors,  while  L.  dioica  $  is 
heterozygous.  In  crossing  these  forms  an  approximate  ratio  of  i  pis- 
tillate to  I  staminate  usually  resulted,  but  with  nearly  always  a  slight 
excess  of  pistillate  plants,  suggesting,  if  the  females  are  homozygous 
for  sex,  an  elimination  of  a  portion  of  the  male  gametes  bearing  the 
determiner  for  maleness.  In  a  later  study  he  was  able  to  show  that  the 
determiner  for  maleness  was  linked  with  a  factor  for  narrow  leaves  while 
in  normal  plants  the  determiners  for  femaleness  were  linked  with  broad 
leaves.  In  a  narrow-leaved  mutant  male  found  by  Baur  he  showed 
that  the  determiner  for  femaleness  as  well  as  maleness  was  linked  with 
the  narrow-leaved  determiner.  It  was  as  a  result  of  the  discovery  of 
this  homozygous  (for  leaf  width  only) ,  narrow-leaf  male  that  the  factor 
for  narrow  leaves  linked  to  maleness  was  able  to  be  discovered  as,  being 
a  recessive  character  and  always  in  a  heterozygous  condition,  it  was 
hidden  in  normal  males. 

When  these  homozygous  narrow-leaved  males  were  used  in  crosses 
with  either  normal  broad-leaved  females  or  heterozygous  females, 
there  was  always  produced  a  great  excess  of  males,  the  females  appear- 
ing only  in  very  small  numbers.  These  results  were  apparently  in  con- 
tradiction to  those  previously  obtained  in  which  females  were  more 
abundant.  Shull  gave  no  explanation  of  these  irregularities.  They 
suggest,  however,  that  there  is  a  fairly  constant  elimination  of  certain 
gametes.  A  study  of  all  of  Shull's  results,  with  this  idea  in  mind,  in- 
dicates that  an  explanation  based  on  the  elimination  of  certain  male 
gametes  will  cover  all  cases  of  irregularity  so  far  reported  by  him  except 
the  nonappearance  of  homozygous  hermaphrodites  and  of  heterozygous 
hermaphrodites  containing  male  determiners.  These  two  instances,  if 
we  may  draw  analogies  between  plants  and  animals,  are  of  the  same 
nature  as  the  YY  zygotes  in  species  of  Drosophila,  and  die  (Bridges,  5).* 

'  Shull  (55-56)  has  shown  that  the  hermaphrodites  have  undoubtedly  been  derived  from  males;  and 
therefore  the  presence  of  two  hermaphrodites  or  a  male  and  an  hermaphrodite  determiner  would  be  anal- 
ogous  to  the  presence  of  two  male  determiners. 


664 


Journal  of  Agricultural  Research 


Vol.  Xri.  No.  10 


In  every  other  instance  in  which  irregularities  in  sex  ratios  occurred,  a 
male  or  hermaphroditic  parent  was  used  in  which  the  condition  for  nar- 
row leaves  was  linked  with  either  a  determiner  for  maleness,  femaleness, 
or  hermaphroditeness.  If  the  irregularities  were  relatively  slight,  as 
was  the  case  when  normal  broad-leaved  males  and  females  were  crossed, 
maleness  and  narrow  leaf  were  linked.  A  partial  elimination  of  these 
male  gametes  would  produce  the  actual  results  obtained.  Hermaphro- 
dites of  Lychnis  dioica  acted  in  the  same  manner  as  the  males.  Hermaph- 
rodites of  Melandrium  album,  which  we  may  assume  to  bear  the  narrow- 
leaf  and  hermaphrodite  determiners  linked,  as  they  have  undoubtedly 
been  derived  from  males,  acted  in  the  same  manner  as  the  narrow- 
leaved  males — that  is,  they  produced  only  females  when  crossed  to 
normal  broad-leaved  females,  in  place  of  a  i  to  i  ratio.  These  results 
can  be  explained  on  the  assumtion  of  complete  elimination  of  the  male 
gametes  of  the  M.  album  hermaphrodite,  which  carry  the  hermaphrodite 
mutant  and  its  linked  factor,  narrow  leaf,  and  in  the  case  of  the  narrow- 
leaved  L.  dioica  males  of  the  nearly  complete  elimination  of  the  male 
gametes  bearing  the  mutant  factor  for  femaleness  and  narrow  leaf. 

In  all  cases  it  seems  that  the  factor  for  narrow  leaves  has  an  inhibiting 
action  on  the  formation  of  the  male  gametes  and  results  in  the  partial 
or  complete  elimination  of  them.  Elimination  in  the  normal  males  is 
not  complete;  otherwise  this  line  would  long  ago  have  disappeared.  In 
the  mutants  in  which  narrow  leaf  is  linked  with  femaleness,  elimination 
of  male  gametes  bearing  this  mutant  factor  is  nearly  complete.  Shull 
has  also  given  much  evidence  which  shows  that  there  is  also  some  elim- 
ination of  female  gametes  bearing  this  mutant  factor  but  to  a  less  extent 
than  in  the  males.  Such  partial  elimination  of  the  female  gametes  was 
shown  in  the  cross  heterozygous  broad-leaved  female  (of  the  formula 
FBFb)  by  a  narrow-leaved  male  Fbfb,  which  produced  two  heterozygous 
broad-leaved  females,  no  homozygous  narrow-leave'd  females  (the  nearly 
complete  absence  of  both  of  these  classes  evidently  being  due  to  the 
elimination  of  male  gametes  bearing  the  factor  for  narrow  leaves  and 
femaleness),  630  broad-leaved  males,  and  463  narrow-leaved  males.  In 
this  cross  all  four  types  should  have  appeared  in  equal  numbers.*     The 

1  The  effect  of  the  linked  factors  "narrow  leaves"  and  "femaleness"  on  the  productioa  of  male  and 
female  gametes  can  be  most  readily  seen  by  the  use  of'this  simple  diagram 


(fFb 

fb 

? 

Fb 

fb 

FB 

FB 

FB 

3 

6jo 

Fb 

fb 

Fh 

Fb 

Fb 

0 

463 

in  which  the  male  gametes  are  placed  on  the  upper  side  of  the  square  and  the  female  gametes  on  the  left 
side,  while  the  number  of  each  of  the  types  of  progeny  are  placed  within  the  small  squares,  with  their 
respective  gametic  combiaations. 


Mar.  II,  1918  Sterility  in  the  Strawberry  665 

inequality  of  the  last  two  classes  must  be  due  to  inequality  in  production 
of  the  two  kinds  of  female  gametes  FB  and  Fb ;  the  latter,  which  carries 
femaleness  and  narrow  leaf  linked,  appearing  less  frequently  than  FB. 
Apparently  with  the  suppression  or  loss  of  the  determiner  for  broad 
leaves  in  the  sex  chromosome,  there  has  also  been  a  partial  suppression 
of  a  factor  necessary  for  the  normal  development  of  male  and  to  a  less 
extent  of  female  gametes.  In  personal  correspondence  with  Dr.  ShuU 
he  informs  me  that  there  is  actually  a  great  deal  of  pollen  sterility  in 
the  narrow-leaved  Lychnis  male. 

CONCLUSIONS 

(i)  The  flowers  of  Fragaria  are  pentamerous  with  regard  to  all  parts 
except  pistils.  The  stamens  are  arranged  in  three  whorls;  the  outer 
parapetalous  series  of  10  stamens,  the  middle  antipetalous,  short  fila- 
mented  series  of  5,  and  the  inner  antisepalous  series  of  5.  Increases  in 
stamen  number  are  due  to  the  addition  of  5,  or  a  multiple  of  it,  to  either 
the  antipetalous  or  the  antisepalous  series.  Decreases  in  stamen  number 
are  due  to  the  loss  of  first  the  antipetalous  and  next  the  antisepalous 
series.  Apparently  the  parapetalous  series  are  permanent.  Decrease 
in  stamen  number  is  in  no  way  related  to  dieciousness. 

(2)  There  is  a  positive  correlation  between  flower  position,  flower  part 
number,  and  size  of  fruit  in  the  strawberry. 

(3)  The  wild  American  species  of  strawberry,  from  which  the  cultivated 
varieties  have  been  derived,  are  for  the  most  part  diecious.  The  pistillate 
plants  bear  staminodia,  which  rarely  develop  as  far  as  the  pollen  mother 
cell  stage,  and  the  staminate  plants  bear  pistils  which  superficially  ap- 
pear to  be  perfect  but  which  are  only  occasionally  functional.  In  a  few 
wild  clones  of  F.  virginiana,  which  appear  to  be  sterile,  pollen  develop- 
ment is  carried  as  far  as  the  tetrad  division  or  slightly  beyond  this  to 
the  liberation  of  the  microspores,  when  complete  disintegration  of  the 
anther  contents  to  an  oily  mass  takes  place.  In  other  instances  a  por- 
tion of  the  microspores  develop  normally  while  the  remainder  within  the 
same  anther  disintegrate,  while  in  other  clones  shortly  after  liberation, 
and  following  a  slight  growth  of  the  microspores,  complete  abortion  of 
the  same  type  as  that  found  in  hybrids  takes  place.  These  anther  types, 
in  wild  clones,  all  appear  to  be  various  expressions  of  a  tendency  toward 
dieciousness  and  are  not  the  result  of  hybridization.  Similar  anther 
types  are  common  in  certain  cultivated  varieties,  on  the  early  flowers  of 
an  inflorescence,  and  especially  on  those  appearing  early  in  the  season. 

(4)  There  is  a  correlation  between  flower  position  and  fertility  of 
pistils,  fertility  decreasing  in  the  later  flowers  of  an  inflorescence.  Pistil 
sterility  is  expressed  in  the  production  of  irregularly  shaped  berries  or 
entirely  sterile  flowers.  Sterility  of  the  later  flowers  of  an  inflorescence 
is  more  general  in  hermaphrodites  than  in  pistillates,  suggesting  that 
the  hermaphrodites  have  been  derived  from  staminates  of  the  diecious 
wild  forms. 


666  Journal  of  Agricultural  Research  voi.  xii,  no.  io 

(5)  The  appearance  of  considerable  amounts  of  aborted  pollen  in 
wild  plants  of  F.  virginiana  and  F.  americana  is  rare  except  in  anthers 
of  the  intermediate  type.  Most  cultivated  varieties  produce  considerable 
amounts  of  aborted  pollen  of  the  type  common  in  hybrids.  The  percentage 
of  aborted  grains  is  not  constant  in  the  individual  flowers  of  a  variety, 
and  neither  is  it  constant  in  the  individual  anthers  of  a  single  flower,  as 
just  as  great  variations  appear  within  the  anthers  of  a  flower  as  are 
shown  by  composite  pollen  samples  of  individual  flowers, 

(6)  In  those  varieties  producing  high  percentages  of  aborted  grains 
a  portion,  at  least,  of  the  morphologically  normal  pollen  grains  are 
functional,  as  shown  by  germination  and  bagging  tests.  There  is  no 
evidence  of  a  physiological  self-sterility  in  the  strawberry. 

(7)  In  the  partially  sterile  variety  Minnesota  3  pollen  development 
is  carried  on  normally  up  to  the  liberation  of  the  microspores  from  the 
tetrad.  At  this  time  all  of  the  microspores  appear  normal  and  alike. 
Following  liberation,  variations  in  rate  of  growth,  time  of  division  of  the 
microspore  nucleus,  and  ability  of  the  individual  microspores  to  develop 
normally  are  shown.  At  all  stages,  during  this  growth  period  micro- 
spores were  found  in  various  stages  of  abortion.  F.  virginiana  exhibits 
as  great  regularity  during  this,  growth  period  as  is  shown  in  the  stages 
leading  up  to  liberation  of  the  microspores. 

(8)  Liberation  of  the  microspores  from  the  tetrad  marks  the  beginning 
of  an  independent  gametophytic  generation,  so  far  as  the  metabolic 
processes  of  growth  are  concerned.  The  individual  microspores  float  in 
a  homogeneous  nourishing  medium  provided  by  the  sporophyte,  but  the 
use  of  this  food  material  in  cell  metabolism  depends  entirely  upon  the 
individual  organization  of  the  microspores. 

(9)  Specific  chromosome  combinations  have  been  shown  by  various 
investigators  to  be  a  potent  factor  in  the  developm^t  or  lack  of  develop- 
ment of  individual  plants  or  animals.  In  plants  heterozygous  for  a 
number  of  factors,  as  are  the  varieties  of  strawberries,  numerous  new 
chromosome  combinations  occur  for  the  first  time  in  the  microspores. 
The  varying  rates  of  growth,  time  of  microspore  division,  ability  to 
increase  the  cytoplasm,  and  inability  in  many  cases  to  develop  normally 
seem  to  be  the  outward  expression  of  the  differential  ability  of  these 
new  chromosome  combinations  to  carry  on  cell  metabolism. 

LITERATURE  CITED 
(i)  Allen,  C.  E. 

1905.    NUCLEAR  DIVISION  IN  THE  POLLEN  MOTHER-CELLS  Olf  LILIUM  CANADENSE. 

/n  Ann.  Bot.,  v.  19,  no;  74,  p.  189-258,  pi.   6-9.     Literature  cited, 
p.  252-256. 


(2) 


1905.   DAS    VERHALTEN    DER    KERNSUBSTANZEN    WAHREND    DER    SYNAPSIS    IN 
DEN     POLLENMUTTERZELLEN       VON     LILIUM     CANADENSE.      In     Jahtb. 

Wiss.  Bot.  [Pringsheim],  Bd.  42,  Heft  i,  p.  72-82,  pi.  2. 


Mar.  II,  1918  Sterility  in  the  Strawberry  667 

(3)  Bellair,  Georges. 

1913.  RECROIS^ES  ENTRE  ELLES  DEUX  ESP^CES  QUI  SE  SONT  D^GAG^ES  d'uN 

HYBRIDE  n'OB^ISSENT  PLUS  A  LA  LOI  MENd6lIENNE  DE  LA  DOMINANCE. 

In  Compt.  Rend.  4th  Conf.  Intern.  Genetique,  1911,  p.  201-203. 

(4)  Bellinc,  John. 

1914.  THE    MODE    OP    INHERITANCE    OP  SEMI-STERILITY  IN  THE    OFFSPRING   OP 

CERTAIN  HYBRID  PLANTS.     In  Ztschr.  Induk.  Abstam.  u.  Vererbungsl., 
Ed.  12,  Heft  5,  p.  303-342,  17  fig.     Literature  cited,  p.  341-342. 
(s)  Bridges,  C.  B. 

1916.  NON-DISJUNCTION  AS  PROOF  OP  THE  CHROMOSOME  THEORY  OP  HEREDITY. 

In  Genetics,  v.  i,  no.  i,  p.  1-52;  no.  2,  p.  107-163,  9  fig.,  i  pi.     Bib- 
liography, p.  162-163. 

(6)  BUNYARD,  E.  A. 

1914.   THE  HISTORY  AND  DEVELOPMENT  OP  THE  STRAWBERRY.      In  Jour  Roy. 

Hort.  Soc.  [London],  v.  39,  pt.  3,  p.  541-552,  fig.  164-168.     Bibfiog- 
raphy,  p.  551-552- 

(7)  Castle,  R.  L. 

1904.  POMOLOGY  AS  A  STUDY.    In  Jouf.  Roy.  Hort.  Soc,  V.  39,  pt.  1/3,  p. 
146-160. 

(8)  Cole,  Ruth  D. 

191 7.  IMPERFECTION    OP   POLLEN    AND    MUTABILITY   IN    THE   GENUS    ROSA.      In 

Bot.  Gaz.,  V.  63,  no.  2,  p.  110-123,  pL  4-6. 

(9)  Darrow,  G.  M. 

1916.  SOUTHERN  STRAWBERRIES.     In  Jour.  Heredity,  v.  7,  no.  12,  p.  531- 

540,  6  fig. 

(10)  Detlepsen,  J.  A. 

1914.  GENETIC  STUDIES  ON  ACAVY  SPECIES  CROSS.    Carnegie  Inst.  Washington 
Pub.  205,  134  p.,  2  fig.,  10  pi.     Bibliography,  p.  129-132. 

(11)  Digby,  L. 

I914.    A  CRITICAL  STUDY  OF  THE  CYTOLOGY  OF  CREPIS  V^RENS.      In  Arch.  Zell- 

forsch.,  Bd.  12,  Heft  i,  p.  97-146,  pi.  8-10.     Bibliography,  p.  138-141. 

(12)  DORSEY,  M.  J. 

1914.  POLLEN    DEVELOPMENT    IN    THE    GRAPE    WITH    SPECIAL    REFERENCE    TO 

STERILITY.     Minn.  Agr.   Exp.  Sta.   Bui.   144,  60  p.,  4  pi.     Bibliog- 
raphy, p.  50-60. 

(13)  East,  E.  M. 

1915.  AN  INTERPRETATION  OP  STERILITY  IN  CERTAIN  PLANTS.     In  Proc.  Amet. 

Phil.  Soc,  V.  54,  no.  216,  p.  70-72. 

(14)  Fletcher,  S.  W. 

191 5.  fragraria  virginian  a  in  the  evolution  op  the  garden  straw- 
BERRY IN  NORTH  AMERICA.     In  Proc.  Soc  Hort.  Sci.,  1915,  p.  125-137. 

(15)  

1917.  STRAWBERRY-GROWING.     325  p.,  23  fig.,  24  pi.     New  York. 

(16) 

I917.   THE  STRAWBERRY   IN    NORTH   AMERICA.      234  p.,  26  fig.      NeW  York. 

(17)  Gates,  R.  R. 

1907.  pollen  development  in  hybrids  op  oenothera  lata  x  o.  lamarck- 

lANA,  AND  ITS  RELATION  TO  MUTATION.      In  Bot.  GaZ.,  V.  43,  nO.  2,  p. 

81-115,  pi.  2-4.     Literature  cited,  p.  110-112. 

(18) 

1915.  THE  MUTATION  FACTOR  IN  EVOLUTION.    353  p.,  114  fig.     Bibliography, 

p.  323-342. 


668  Journal  of  Agricultural  Research  vd.xii,  no.  lo 

(19)  Georgeson,  C.  C. 

1907.  BRIEF  SUMMARY  OF  WORK.     SITKA  STATION.     In  Alaska  AgT.  Exp.  Sta. 

Ann.  Rpt.,  1906,  p.  9-15,  2  pi. 

(20) 

1911.  WORK  AT  SITKA  STATION.  In  Alaska  AgT.  Exp.  Sta.  Ann.  Rpt.,  1910, 
p.  10-29,  4  pi. 

(21)  GooDSPEED,  T.  H. 

1913.  ON  THE  PARTIAL  STERILITY  OF  NICOTIANA  HYBRIDS  MADE  WITH  N.  SYL- 

VESTRIS  AS  A  PARENT.     I.     In  TJniv.  Cal.  Pub.  Bot.,  v.  5,  no.  4,  p. 

189-198. 

(22)  and  Kendall,  J.  N. 

I916.  ON  THE  PARTIAL  STERILITY  OF  NICOTIANA  HYBRIDS  MADE  WITH  N.  SYL- 
VESTRIS  AS  A  PARENT.  III.  AN  ACCOUNT  OF  THE  MODE  OP  FLORAL 
ABSCISSION  IN  THE  Fj  SPECIES  HYBRIDS.  In  Univ.  Cal.  Pub.  Bot., 
V.  5,  no.  10,  p.  293-299. 

(23)  Hoar,  C.  S. 

1916.  sterility  as  the  result  of  hybridization  and  the  condition  of 
POLLEN  IN  RUBUS.    In  Bot.  Gaz.,  v.  62,  no.  5,  p.  370-388,  pi.  10-12. 
^       Literature  cited,  p.  386-387. 

(24)  JanczEwski,  Ed. 

1908.  SUR  LES  ANTH^RES  ST^RILES  DES  GROSEILLIERS.      In  Bul.   Acad.  Sci« 

Cracovie,  1908,  p.  587-596,  i  pi. 

(25)  Jeffrey,  E.  C. 

1914.  spore  conditions  in  hybrids  and  the  mutation  hypothesis  op  de 

VRiES.     In  Bot.  Gaz.,  v.  58,  no.  4,  p.  322-336,  pi.  22-25. 

(26)  JESENKO,  F. 

1913.  SUR    UN    HYBRIDE   FERTILE    ENTRE    TRITICUM    SATIVUM  9  9     (blE    MOLD- 

squarehead)  ETSECALECEREALE^   (SEIGLE  DE  PETKUS).      /w  Compt. 
Rend.   4tli   Conf.    Intern.     Genetique,    1911,   p.   301-311. 

(27)  Millardet,  a. 

1894.   NOTE  SUR  l'HYBRIDATION  SANS  CROISEMENT  OU  FAUSSE  HYBRIDATION. 

In  Soc.  Sci.  Phys.  et.  Nat.  Bordeaux,  s.  4,  t.  4,  cahier  2,  p.  347-372* 
I  fig. 

(28)  Morgan,  T.  H. 

1914.  HEREDITY  AND  SEX.    ed.  2,  248  p.,  121  fig.    N^sv  York. 

(29)  Nakao,  M. 

191 1.  cytological  studies  on  the  nuclear  di\^sion  of  the  pollen  mother- 
CELLS   OF   SOME    CEREALS    AND    THEIR    HYBRIDS.      In  JotUT.    Col.    Agr. 

Tohoku  Imp.  Univ.,  v.  4,  no.  3,  p.  173-190,  4  pi. 

(30)  Overton,  J.  B. 

1905.   tJBER    REDUKTIONSTEILLUNG     IN    DEN     POLLENMUTTERZELLEN    EINIGER 

DiKOTYLEN.     In  Jahrb.  Wiss.  Bot.  [Pringsheim],  Bd.  42,  Heft  i,  p. 
121-153,  pi.  6-7. 

(31)  Richardson,  C.  W. 

1914.  A  preliminary  NOTE  ON  THE  GENETICS  OF  FRAGARiA.  In  Jout.  Gene- 
tics, V.  3,  no.  3,  p.  171-177,  4  fig-,  pl-  7- 

(32)  Rimpau,  W. 

1891.   KREUZUNGSPRODUKTE     LANDWIRTSCHAFTLICHER     KULTURPFLANZEN.     In 

Landw.  Jahrb.,  Bd.  20,  p.  335-339- 

(33)  Rosenberg,  O. 

1909.  cytologische  und  morphologische  studien  an  drosera  longlfolia 
X  ROTUNDIFOLIA.  K.  Svcnsk.  Vetensk.  Akad.  Handl.,  Bd.  43,  No. 
II,  64  p.,  33  fig.,  4  pi.     Literaturverzeichnis,  p.  61-62. 


Mar.  II,  1918  Sterility  in  the  Strawberry  669 

(34)  Rydberg,  p.  a. 

1898.   ▲   MONOGRAPH   OF  THE   NORTH   AMERICAN  POTENTILLEAE.      Mem.    Dept. 

Bot.  Columbia  Univ.,  v.  2,  223  p.,  112  pi. 

(35)  Shull,  G.  H. 

I910.   INHERITANCE  OP  SEX  IN  LYCHNIS.      In   Bot.   Gaz.,  V.  49,  HO.  2,  p.   IIO- 

125,  2  fig.     Literature  cited,  p.  125. 

(36)  

1915.  SEX-LIMITED    INHElilTANCE    IN    LYCHNIS    DIOICA   L.      In   Ztschr.    Induk. 

Abstam.  u.  Vererbimgsl.,  Bd.  12,  Heft  5,  p.  265-302,  5  fig.     Literature 
cited, p.  299-302. 

(37)  Standish,  L.  M. 

1916.  what  is  happening  to  the  hawthorns?    In  Jour.  Heredity,  v.  7,  no. 

6,  p.  266-279,  fig.  7-17.     Literature  cited,  p.  279. 

(38)  Stevens,  N.  E. 

1912.  observations  on  heterostylous  plants.  In  Bot.  Gaz.,  v.  53,  no.  4, 
p.  277-308,  pi.  21-23.     Literature  cited,  p.  305-307. 

(39)  Strauburger,  Eduard. 

1909.  ZEITPUNKT  DER  BESTIMMUNG  DES  GESCHLECHTS,   APOGAMIE,  PARTHENO- 

GENESIS UND    REDUKTiONSTEiLUNG.     In  his  Histologische  Beitrage, 
Heft  7,  124  p.,  3  pi.     Jena. 

(40)  TiSCHLER,  G. 

1906.   tJBER    DIE    ENTWICKLUNG    DES   POLLENS    UND    DER    TAPETENZELLEN    BEl 

RiBESHYBRiDEN.     In  Jahrb.  Wiss.  Bot.  [Pringsheim],  Bd.  42,  Heft  4, 
P-  545-578,  pi.  15-     Literatur-Verzeichnis,  p.  575-577- 
(41) 

1908.   ZELLSTUDIEN    AN    STERILEN    BASTARDPFLANZEN.      In    Aich.    Zellforsch., 

Bd.  I,  Heft  I,  p.  33-151,  120  fig.     Literatur,  p.  104-106,  147-151. 
42)  

1910.  UNTERSUCHUNGEN  tJBER   DIE  ENTWICKLUNG    DES   BANANEN-POLLENS.   I. 

In  Arch.  Zellforsch.,  Bd.  5,  Heft  4,  p.  622-670,  4  fig.,  pi.  30-31.     Ver* 
zeichnis  der  zitierten  Literatur,  p.  67. 

(43)  VallEau,  W.  D. 

1916.  INHERITANCE  OF  SEX  IN  THE  GRAPE.  In  AmcT.  Nat.,  V.  50,  no.  597, 
P-  554-564-     Literature  cited,  p.  563-564. 

(44)  WiLLE,  N. 

1886.  UEBER  DIE  ENTWICKLUNGSGESCHICHTE  DER  POLLENKORNER  DER  ANGIO- 
SPERMEN  UND  DAS  WACHSTHUM  DER  MAMBRANEN  DURCH  INTUSSUS- 
CEPTION. Forhandl.  Vidensk.  Selsk.  Christiania,  no.  5,  71  p.,  3  pi. 
Original  not  seen. 


PLATE  B 

Minnesota  3 : 

1. — Pollen  mother  cell  previous  to  synapsis.     X  2,000. 

2. — Presynapsis  in  the  pollen  mother  cell  showing  loops  extending  out  from 
synaptic  mass.     X  2,000. 

3. — ^Two  loops  and  portion  of  a  loop  extending  from  the  presynaptic  mass.     Same 
stage  as  figure  2.     X  3,280. 

4. — Synapsis  in  a  pollen  mother  cell.     X  2,000. 

5. — A  postsynaptic  stage.     The  synaptic  mass  is  unraveling  into  bivalent  loops. 
X  2,000. 

6. — Open  spireme  stage.     X  2,000. 
7. — A  presegmentation  stage  of  the  spireme.     X  2,000. 

8. — A  portion  of  a  bivalent  spireme  thread  of  the  same  stage  as  figure  7.     X  3)28o. 
9. — Segmentation  of  the  bivalent  spireme  into  chromosome  pairs.     The  cytoplasm 
is  roxmding  up  and  is  partially  surrounded  by  a  gelatin-like  sheath.     X  2,000. 
10. — A  portion  of  the  bivalent  spireme  during  segmentation.     X  2,340. 
II,  12. — Chromosome  pairs  during  the  contraction  period  following  segmentation. 
X  2,000. 

13. — Individual  chromosome  pairs  showing  various  figures  commonly  formed  during 
contraction.     X  3.280. 

14. — Diakenesis  in  the  pollen  mother  cell.     X  2,000. 
15. — Multipolar  spindle  stage  of  pollen  mother  cell.     X  2,000. 
16. — Early  anaphase  of  the  heterotypic  division.     X  3,280. 

(670) 


Strawberry 


Plate  B 


\ 


A'^' 


\.i 


'    !^*.»^C 


4 


13 


12 


I? 


^ 


15 


^ 


8 


16 


Journal  of  Agricultural  Research 


Vol.  XII,  No.   10 


I  B-einHun  CO  wash'  d c 


PLATE  C 

I. — Late  anaphase  of  the  heterotypic  division.  The  cytoplasm  is  rounding  up 
from  the  mother  cell  wall  and  is  partially  surroimded  by  a  thick  gelatin-like  sheath. 
X  2,000. 

2. — Chromosomes  on  the  equatorial  plate  of  the  homeotypic  division. 

3. — A  portion  of  an  anther  in  the  tetrad  stage,  showing  the  microspores  embedded 
in  the  gelatin-like  sheath.     The  original  mother  cell  walls  are  still  present. 

4. — A  tetrad  at  the  same  stage  as  those  shown  in  figure  3.  The  mother  cell  wall 
is  not  shown.     X  2,000. 

5. — A  microspore  shortly  after  liberation  from  the  tetrad.     X  2,000. 

6. — A  liberated  microspore  in  which  growth  has  commenced.     X  2,000. 

7. — A  later  stage  than  figure  6,  showing  the  slight  thickening  of  the  wall  and 
the  irregularities  due  to  growth  of  the  wall.     X  2,00c. 

8. — Microspore  growth  completed  previous  to  division  of  the  microspore  nucleus. 
X  2,000. 

9. — A  section  through  a  microspore  nucleus  in  prophase  showing  the  continuous 
univalent  spireme.     X  3,280. 

.  10. — Another  section  of  the  same  nucleus,  showing  the  first  stages  of  the  disappear- 
ance of  the  nucleolus.     The  nucleolar  strands  are  attached  to  the  spireme.     X  3,280. 

II. — Metaphase  of  the  division  of  the  microspore  nucleus.  A  spindle  in  this  posi- 
tion results  in  the  nuclear  arrangement  shown  in  figure  15.  The  thickened  extine  is 
shown.     X  2,000. 

12. — F.  virginiana.  Anaphase  in  the  division  of  the  microspore  nucleus.  The 
spindle  lying  parallel  to  the  wall  results  in  the  nuclear  arrangement  shown  in  figiu-e  14. 
X  2,000. 

13. — Telophase  of  the  division  of  the  microspore  nucleus.  The  wall  which  eventu- 
ally surrounds  the  generative  nucleus  is  not  always  apparent  at  this  time. 

14. — A  later  stage  than  figure  13  in  which  the  generative  cell  has  been  definitely 
cut  off.     X  2,000. 

15. — A  yotmg  pollen  grain  shortly  after  division,  showing  an  increase  in  cytoplasm 
content.     The  thickened  extine  is  shown.     X  2,000. 

16. — End  view  of  a  pollen  grain  showing  the  pattern  of  the  laminate  layers  shown 
in  figure  15  and  Plate  D,  figures  i,  6,  and  15.  The  arrows  mark  the  ends  of  the  three 
sutures  which  bear  the  germ  pores. 


Sterility  in  tlie  Strawberry 


Plate  C 


:'?^ 


c 


'^«^ 


10 


m 


II 


12 


14 


15 


16 


# 


Journal  of  Agricultural  Research 


Vol.  XII.  No.   10 


A  B  eKAHAM  CO.  WASHf  DC. 


PLATE  D 

I. — Nearly  mature  pollen  grain.  The  central  body  is  the  vegetative  nucleus 
while  the  other  is  the  generative  cell.     X  2,000. 

2. — Mature  pollen  grain.  The  extine  is  not  shown.  The  killing  fluid  causes  the 
dry  folded  grains  to  become  spherical.     X  2,000. 

3.  4>  5.  7- — Various  types  of  degenerate  microspores  from  anthers  bearing 
microspores  of  the  stage  shown  in  Plate  C,  figure  6.     X  2,000. 

6. — An  aborting  microspore  from  an  anther  containing  half-grown  microspores. 
X  2,000. 

8. — An  aborting  microspore  of  the  same  type  as  that  shown  in  figure  6  from  an 
anther  containing  nearly  full-grown  microspores  as  in  Plate  C,  figure  8.  As  ia  Plate 
D,  figiu-e  6,  the  cytoplasm  and  nucleus  still  appear  normal.     X  2,000. 

9,  II. — Microspores  of  the  same  types  and  same  age  as  figures  6  and  8,  in  which 
degeneration  has  proceeded  farther      X  2,000. 

10. — An  aborted  microspore  from  an  anther  containing  microspores  of  the  stage 
shown  in  Plate  C,  figure  8.  Apparently  abortion  took  place  shortly  following  libera- 
tion from  the  tetrad. 

12. — An  early  stage  of  degeneration  in  a  full-grown  i-nucleate  microspore.     X  2,000. 

13. — An  early  stage  of  degeneration  in  a  full-grown  i-nucleate  microspore.  A 
ntmiber  of  normal  microspores  in  this  anther  are  already  dividing. 

14. — An  aborting  microspore  containing  an.  abnormally  small  amoimt  of  light 
staining  cytoplasm;  from  an  anther  containing  i-  and  2-nucleate  microspores.     X  2 ,000. 

15. — An  aborted  microspore  from  an  anther  containing  i-  and  2-nucleate  micro- 
spores. Apparently  this  is  a  late  stage  of  the  type  of  degeneration  shown  in  figures 
6  and  8.     X  2,000. 

16. — An  aborted  microspore  containing  very  scant  cytoplasm.  The  nucleus  has 
completely  degenerated  and  degeneration  of  the  cytoplasm  has  begxm. 


Sterility  in  the  Strawberry 


PLATE    D 


t;--,--:i.u6.  j--  _ 


14 


15     ""'^V^Wv./ 


16 


Jniirnal   of   Atrriculturai   Research 


Vol.  XII,  No.   10 


A  a  euHAH  CO.  wiLSHf  o  c. 


PLATE  E 

I, — A  slightly  more  advanced  stage  of  the  condition  shown  in  Plate  D,  figure  i6. 

2. — An  early  stage  in  the  abortion  of  a  full-grown  i-nucleate  microspore.     X  2,000. 

3. — An  early  stage  of  abortion  directly  following  microspore  division.     X  2,000. 

4. — A  full-grown  i-nucleate  microspore  containing  very  scant  light-staining  cyto- 
plasm; from  an  anther  containing  i-  and  2-nucleate  microspores.     X  2,000. 

5. — Another  type  of  degeneration  of  a  full-grown  I-nucleate  microspore.     X  2,000. 

6. — An  aborted  microspore  fotmd  among  i-  and  2-nucleate  microspores.     X  2,000. 

7. — A  later  stage  of  the  type  of  degeneration  shown  in  Plate  D,  figure  13;  from 
an  anther  containing  microspores  of  the  stage  of  development  shown  by  Plate  C, 
figure  15.     X  2,000. 

8. — Degeneration  of  the  generative  cell  shortly  after  division.  The  vegetative 
nucleus  and  cytoplasm  are  still  normal.     X  2,000. 

9,  10. — Common  types  of  aborted  microspores  found  with  mature  pollen.  Evidently 
abortion  took  place  before  the  division  of  the  microspore  nucleus.     X2,ooo. 

II. — An  aborted  microspore,  of  the  same  type  as  that  shown  in  figure  7.  Found 
with  mattu"e  pollen.     X  2,000. 

12. — A  pollen  grain  showing  abortion  of  the  generative  cell  and  an  abnormal  vacuo- 
late condition  of  the  cytoplasm.     X  2,000. 

13. — A  later  stage  of  the  type  of  degeneration  shown  in  figure  8.  The  vegetative 
nucleus  and  cytoplasm  are  normal.     X  2,000. 


Sterility   in  the  Strawberry 


Plate  E 


A 


(a 


-^ 


13 


10 


Journal  of  Agricultural  Research 


Vol.  XII.   No.    10 


1  a  SHkHt  1  CO .  WASH  "  0  c 


38325"— 18 5 


PLATE  35 

A. — ^Tertiary  flower  of  the  pistillate  variety,  Minnesota  roiyX  Progressive  — 13-40, 
showing  prominent  staminodia. 

B,  C. — Primary  and  secondary  flowers  of  the  perfect  variety,  Minnesota  loiyXPro- 
gressive~9-24;  B  showing  intermediate  and  C  perfect  anthers. 

D,  E,  F. — Two  primary  and  a  secondary  flower  of  the  perfect  variety,  Minnesota 
loiyXProgressive— 2-55,  showing  pistillate,  intermediate,  and  perfect  types  of 
flowers. 

G,  H,  I,  J. — Flowers  from  the  perfect  variety,  Minnesota  1017X  Progressive— 32-1. 
G  and  H  are  primary  and  secondary  flowers,  respectively,  and  are  pistillate;  I  a 
secondary  imperfect  flower  with  a  few  normally  developed  stamens  and  J  a  tertiary 
perfect  flower. 


sterility  in  the  Strawberry 


Plate  35 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  10 


sterility  in  tlie  Strawberry 


Plate  36 


Journal  of  Agricultural  Research 


Vol. XII,  No.  10 


PLATE  36 

A,  B,  C,  D. — Cross  sections  of  two  loculi  of  staminodia  of  the  pistillate  varie- 
ties, Crescent,  Columbia,  Minnesota  loi 7  X  Progressive  — 11-59,  and  Seedling  140, 
respectively. 

E. — Degeneration  of  the  tetrads  in  an  intermediate  anther  of  Fragaria  virginiana. 

F,  G. — Later  stages  of  the  condition  shown  in  figure  E. 

H. — A  portion  of  an  intermediate  anther  from  the  first  flower  of  Minnesota  3. 


EFFECT    OF    NITRIFYING    BACTERIA  ON    THE  SOLU- 
BILITY OF   TRICALCIUM   PHOSPHATE^ 
By  W.  P.  Kblley^ 

Professor  of  Agricultural   Chemistry,   Graduate   School  of  Tropical  Agriculture  and 
Citrus  Experiment  Station,  University  of  California 

INTRODUCTION 

The  solution  of  tricalciurn  phosphate  and  the  chemical  changes  through 
which  it  passes  in  soils  are  subjects  of  special  interest.  The  phosphorus 
compounds  of  soils  have  been  mainly  derived  from  tricalcium  phosphate, 
and  are  relatively  insoluble  in  water.  A  considerable  part  of  the  phos- 
phorus of  soils  probably  actually  occurs  as  tricalcium  phosphate.  The 
soluble  phosphates  of  processed  fertilizer  also  become  relatively  insoluble 
soon  after  being  mixed  with  soil,  some  of  which  may  be  converted  into 
tricalcium  phosphate.  In  addition,  tricalcium  phosphate,  in  the  form 
of  untreated  rock  phosphate,  has  been  recommended  as  a  fertilizer  at 
various  times  in  the  past  and  at  present  is  being  applied  in  considerable 
amounts  in  different  parts  of  America. 

While  tricalcium  phosphate  is  relatively  insoluble  in  water,  it  is  well 
known  that  this  substance  is  notably  soluble  in  water  saturated  with 
carbon  dioxid.  For  this  reason  it  has  long  been  suspected  that  the 
carbonic  acid  of  soils  promotes  the  solution  of  the  phosphates  present. 
Likewise  the  increased  fertilizing  effect  resulting  from  the  application 
of  rock  phosphate  in  conjunction  with  decaying  organic  matter  has 
commonly  been  assumed  to  be  due  to  the  solvent  effects  of  the  carbonic 
acid,  and  possibly  other  organic  acids,  that  are  formed  in  the  decom- 
position of  the  organic  matter. 

As  a  rule,  however,  investigators  have  not  been  able  to  detect  any 
increase  in  the  solubility  of  rock  phosphate  when  left  in  contact  with 
decaying  organic  matter.  For  example,  Lupton  ((5)^  McDowell  (7), 
Truog  (9),Tottinghara  and  Hoffmann  (8),  and  various  European  workers* 
found,  as  a  result  of  composting  rock  phosphate  with  various  fermenting 
mixtures,  that  in  no  case  more  than  slight  increases  in  the  solubility  of 
the  phosphate  took  place  and  in  certain  instances  decreases  in  solubiUty 
were  noted. 


>  Paper  No.  45,  University  of  California  Graduate  School  of  Tropical  Agriculture  and  Citrus  Experiment 
Station,  Riverside,  Cal. 
»  Credit  is  due  Mr.  A.  B.  Cummins  for  assistance  in  this  investigation. 
'  Reference  is  made  by  number  (italic)  to  "  Literature  cited,"  p.  685. 
*  A  very  complete  bibliography  of  this  subject  is  given  by  Lipman,  McLean,  and  Lint  (5). 


Journal  of  Agricultural  Research,  Vol.  Xil,  No.  10 

Washington.  D.  C.  March  n,'  1918 

n*  ,,      ,  KeyNo.  Cal.  18 

(671) 


672  Journal  of  Agricultural  Research  voi.  xii. No.  10 

A  number  of  investigators  have  shown  that  the  solution  of  tricalcium 
phosphate  may  be  efifected  by  the  biochemical  oxidation  of  sulphur. 
Recently  Lipman,  McLean,  and  Lint  (5)  found  that  large  amounts  of 
rock  phosphate  may  be  made  soluble  by  the  sulphur  bacteria  in  fer- 
menting mixtures  containing  elemental  sulphur. 

The  effects  of  the  nitrifying  bacteria  on  the  solubility  of  tricalcium 
phosphate  have  recently  been  investigated  by  Hopkins  and  Whiting  (j). 
They  found  that  the  nitrite  bacteria  (Niirosomonas  spp.)  have  the  power 
of  oxidizing  ammonium  sulphate  in  solution  cultures  containing  trical- 
cium phosphate  but  no  carbonate  or  free  base,  and  that  under  these  con- 
ditions the  nitrous  acid  and  sulphuric  acid,  formed  from  the  ammonium 
sulphate,  attacked  the  tricalcium  phosphate  and  rendered  notable 
amounts  of  phosphorus  and  calcium  soluble  in  water.  Similar  effects 
were  found  when  ammonium  nitrate  was  substituted  for  ammonium  sul- 
phate. By  calculation  they  found,  as  an  average  of  13  tests,  that  for 
every  56  pounds  of  nitrogen  oxidized,  115  pounds  of  phosphorus  and  211 
pounds  of  calcium  were  made  soluble  in  water.  They  also  found  that  no 
change  in  the  solubility  of  tricalcium  phosphate  takes  place  as  a  result 
of  the  action  of  the  nitrate  bacteria  (Nitrobacter  spp.).  As  pointed  out 
by  them,  the  oxidation  of  nitrite  to  nitrate  does  not  necessitate  an  in- 
crease in  acidity,  but  is  merely  a  matter  of  adding  an  atom  of  oxygen  to 
the  nitrite. 

Hopkins  and  Whiting  have  discussed  their  views  regarding  the  prac- 
tical bearing  of  these  experiments  at  considerable  length,  and  have  as- 
signed special  importance  to  the  nitrite  bacteria  as  agents  in  promoting 
the  solution  of  rock  phosphate  in  the  field.  In  commenting  on  these 
investigations  Davenport  ^  even  suggested  that  the  importance  of  the 
nitrite  bacteria  as  agents  in  the  solution  of  rock  phosphate,  is  on  a  par 
with  that  of  the  legume  bacteria  in  nitrogen  fixation. 

Since  the  formation  of  nitrous  acid  is  commonly  considered  to  be  an 
essential  step  in  the  nitrification  process  and  as  nitrification  is  generally 
active  in  productive  soils,  it  is  at  once  apparent  that  relatively  large 
amounts  of  phosphate  will  be  made  soluble  by  this  group  of  bacteria, 
provided  the  reactions  that  take  place  in  soils  be  similar  to  those  found  in 
solution  cultures. 

The  practical  importance  of  phosphates  in  agriculture  and  the  general 
interest  in  the  several  phases  of  the  nitrification  process  justify  further 
investigation  of  this  problem.  Accordingly,  the  writer  has  made  some 
studies  on  it  at  the  University  of  California  Citrus  Experiment  Station. 
In  these  studies  both  soil  and  sand  cultures  have  been  employed.  The 
formation  of  nitrate  and  the  solubility  of  calcium  and  phosphoric  acid  in 
water  were  used  in  this  investigation  as  measures  of  the  biochemical 
action  of  the  nitrifying  organisms. 

•  Foreword  to  the  publication  by  Hopkins  and  Whiting  (3). 


Mar.  II,  1918     Nitrifying  Bacteria  and  Tricalcium  Phosphate  673 

EXPERIMENTAL  RESULTS 

The  soil  used  in  this  investigation  was  drawn  from  one  of  the  plots 
(F)  now  being  used  in  a  fertilizer  experiment  with  Citrus  trees.  This 
plot  has  been  treated  annually  for  10  years  with  light  applications  of 
stable  manure,  but  no  commercial  fertilizer  or  lime  has  been  applied 
to  it.  The  soil  is  a  light  sandy  loam  of  granitic  origin,  the  coarser  par- 
ticles of  which  are  composed  largely  of  granite.  It  contained  8.5  p.  p.  m. 
of  nitric  nitrogen,  and  0.188  per  cent  of  total  phosphorus  pentoxid 
(P2O5),  of  which  17.4  p.  p.  m.  were  soluble  in  water  when  the  experiment 
was  begun.  The  total  carbonate  (CO3),  as  determined  by  the  Gaither 
(2)  method,  amounted  to  0.03  per  cent,  but  the  sample  was  free  from 
water-soluble  carbonate  (CO3). 

Portions  of  2,000  gm.  each  of  fresh  soil  were  weighed  into  half -gall  on 
(1.89  Hters)  fruit  jars.  A  solution  of  ammonium  sulphate  (c.  p.)  was 
added  to  certain  portions  at  rates  supplying  o.oi  gm.  of  nitrogen  per 
100  gm.  of  dry  soil.  To  other  portions  an  equal  quantity  of  nitrogen  was 
added  in  the  fonr)  of  dried  blood.  Still  other  portions  were  employed  with- 
out the  addition  of  any  nitrogenous  substance.  Baker's  analyzed  trical- 
cium phosphate  was  added  in  certain  cases  at  the  rate  of  o.io  gm.  per  100 
gm.  of  soil  and  calcium  carbonate  (c.  p.)  at  the  rate  of  0.25  gm.  per  100 
gm.  of  soil.  The  experiments  were  made  in  duplicate.  After  a  thorough 
mixing,  adding  suitable  amounts  of  water  and  mixing  again,  the  jars 
were  loosely  covered  and  incubated  at  room  temperature. 

It  is,  of  course,  well  known  that  the  purest  tricalcium  phosphate  is 
somewhat  soluble  in  water  and  that  ammonium  sulphate  affects  the 
solubility  of  certain  soil  constituents,  notably  calcium,  without  the  inter- 
vention of  bacteria.  Consequently  it  was  deemed  necessary  to  deter- 
mine the  solubility  of  calcium  and  phosphoric  acid  after  the  above- 
named  substances  had  been  mixed  with  the  soil,  but  before  sufficient  time 
had  elapsed  to  permit  measurable  bacterial  action.  It  is  obvious  that 
the  amounts  of  soluble  calcium  and  phosphoric  acid  present  in  the  soil 
at  the  beginning  of  the  experiment  should  not  be  considered  as  having 
been  dissolved  by  subsequent  bacterial  action.  Accordingly  portions  of 
soil  each  containing  200  gm.  were  placed  in  flasks,  the  same  relative 
proportions  of  tricalcium  phosphate,  calcium  carbonate,  and  ammonium 
sulphate  added  as  in  the  incubated  series,  and  the  contents  thoroughly 
mixed.  Distilled  water  was  added  at  the  rate  of  250  parts  per  100 
parts  of  dry  soil,  and  the  contents  were  vigorously  shaken  once  every 
10  minutes  during  an  hour,  and  were  then  filtered  through  Chamberland- 
Pasteur  filters.  Calcium  was  determined  in  the  filtrates  by  the  volu- 
metric permanganate  method  and  phosphoric  acid  by  the  Pemberton 
volumetric  method.  The  average  results  obtained  from  closely  agreeing 
duplicate  solutions,  expressed  in  parts  per  million  of  dry  soil,  are  sub- 
mitted in  Table  I. 


674 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  lo 


Table  I. — Solubility  of  calcium  and  phosphate  in  soil  immediately  after  adding  calcium 
carbonate,  tricalcium  phosphate,  and  ammonium  sulphate 


Treatment. 


Soil  only 

Soil  and  calcium  carbonate 

Soil  and  tricalcium  phosphate 

Soil,  tricalcium  phosphate,  and  calcium  carbonate.  . 

Soil  and  ammonium  sulphate 

Soil,  ammonium  sulphate,  and  tricalcium  phosphate 


Soluble 

calcium 

(Ca). 

Soluble 
phosphoric 
acid  (P2O5). 

P.  p.  TO. 

P.  p.  m. 

27-5 
26.8 

17.4 
13.8 

31- I 

28.6 

33-5 
90.  I 

27.  I 
16.6 

92.  6 

25.  I 

These  data  show  that  the  addition  of  calcium  carbonate  produced  no 
effect  on  the  immediate  solubility  of  the  calcium  already  in  the  soil  or 
that  added  as  tricalcium  phosphate,  but  the  addition  of  tricalcium  phos- 
phate produced  an  increase  of  about  5  p.  p.  m.  of  soluble  calcium  and 
II. 2  p.  p.  m.  of  soluble  phosphoric  acid.  The  most  notable  effect  was 
produced  by  ammonium  sulphate,  which  caused  an  increase  in  water- 
soluble  calcium  from  27.5  to  90.1  p.  p.  m. 

The  data  submitted  in  Table  I  should  not  be  considered  as  represent- 
ing true  solubility  determinations,  for  it  is  not  certain  that  equilibrium 
was  completely  established,  either  between  the  various  solids  present 
and  the  solvent  (water),  or  between  the  constituents  of  the  soil  and  the 
chemical  substances  that  were  added  to  it.  A  longer  period  of  contact 
might  have  yielded  extracts  either  more  or  less  concentrated,  depending 
on  whether  or  not  the  rate  of  solution  was  greater  or  less  than  the  rate 
of  fixation.  The  same  procedure  was  followed  in  making  these  deter- 
minations,  however,  as  was  used  at  the  close  of  the  incubation  periods, 
and,  although  the  results  are  not  strict  solubility  determinations,  they 
are  believed  to  be  comparable,  and  that  any  difference  between  the 
amounts  of  calcium  and  phosphoric  acid  found  at  the  beginning  and  the 
close  of  the  periods  of  incubation  may  be  assumed  to  have  arisen  mainly 
through  the  action  of  biochemical  agents.^ 

After  incubation  periods  of  28,  57,  and  157  days,  quantities  containing 
200  gm.  of  dry  soil  were  transferred  from  the  incubation  jars  to  flasks, 
distilled  water  was  added  at  the  rate  of  250  parts  per  100  parts  of  dry 
soil,  was  shaken  vigorously  once  every  10  minutes  during  an  hour,  and 
was  then  filtered  through  Chamberland-Pasteur  filters  as  in  the  pre- 
ceding series.  Soluble  calcium  and  phosphoric  acid  were  determined 
in  the  filtrates  by  the  methods  already  referred  to,  and  nitric  nitrogen 
by  the  phenoldisulphonic-acid  method.  The  filtrates  were  also  tested 
for  nitrite,  but  not  more  than  0.5  p.  p.  m.  was  found  in  any  case.  The 
average  results  of  closely  agreeing  duplicates  are  recorded  in  Table  II. 

*It  was  not  deemed  advisable  to  maiataia  separate  portions  in  a  sterile  condition,  owing  to  the  fact 
that  the  various  methods  now  in  use  for  bringing  about  complete  sterilization  in  soils  probably  affect  the 
solubility  of  the  various  constituents. 


Mar.  II.  1918     Nitrifying  Bacteria  and  Tricalcium  Phosphate  675 


Table  II. — Effects  of  nitrification  on  the  solubility  of  tricalcium  phosphate  in  soil 


Materials  added. 


Control 

Calcium  carbonate 

Tricalcium  phosphate. . 

Calcium  carbonate  and 
tricalcium  phosphate. 

Ammonium  sulphate .  . 

Ammonium  sulphate 
and  calcium  car- 
bonate   

Ammonium  sulphate 
and  tricalcium  phos- 
phate  

Ammonium  sulphate, 
calcium  carbonate, 
and  tricalcium  phos- 
phate   

Dried  blood 

Dried  blood  and  cal- 
cium carbonate 

Dried  blood  and  trical- 
cium phosphate 

Dried,  blood  calcium 
carbonate ,  and  trical- 
cium phosphate 


After  2S  days. 


Nitric 
nitro- 
gen. 


P.  p.  m. 

20.  O 
22.  o 

21.  o 


22.  o 
98.0 


Cal- 
cium. 


P.  p.  m. 
45- o 
56.5 
53-5 

59-1 
219.4 


97.0   254.4 


99.0   217.  7 


100.  o 
91.  o 

89.0 

82.0 


253-4 
107.7 

107.  2 


Phos- 
phoric 
acid. 


P.  p.  m. 

13-  I 
II.  9 

24.  2 

17-3 
18.5 


18.5 


52.1 


26.6 
9-7 

9.8 


After  s7  days. 


Nitric 
nifiro- 
gen. 


111.71      24.3 
118.  2I      19.  5 


P.  p.m. 

25-5 
29.  o 

28.0 

28.0 
99.0 


99.0 

lOI.  o 

90.  o 
90.  o 
88.0 

87.5 


Cal- 
cium. 


P.  p.  m. 
50.  6 
70.8 
58.8 

70.  I 
225.4 


Phos- 
phoric 
acid. 


P.  p.  m. 


13.2 
25.0 


22.  4 
19.4 


270.  5!       7-  4 


229.  6 

230.4 
113-9 

140.  2 
117.  7 


38-0 

13-9 
10.  o 

II- 5 
22.  2 

18.3 


After  IS7  days. 


Nitric 
nitro- 
gen. 


P.  p.m. 


114.  O 


94.0 


Cal- 
cium. 


Phos- 
phoric 
acid. 


P.  p.  fn.  \P,  p.  m. 


232.  I 


218.4      30.0 


116.  4 


5-7 


A  series  of  experiments  with  the  use  of  silica  sand  corresponding 
closely  with  the  preceding  soil  series  was  conducted  at  the  same  time. 
The  silica  sand  was  obtained  from  Monterey,  Cal.,  and  was  free  from 
carbonate,  but  contained  small  amounts  of  feldspar,  hornblende,  and 
mica  particles,  and  possibly  traces  of  other  minerals.  Portions  of  1,000 
gm.  each  were  placed  in  fruit  jars,  and  quantities  of  ammonium  sulphate, 
dried  blood,  calcium  carbonate,  and  tricalcium  phosphate  were  added 
in  duplicate  at  the  same  rates  and  arranged  after  the  same  plan  as  in  the 
preceding  soil  series. 

Mixed  cultures  of  bacteria  were  supplied  by  adding  150  c.  c.  of  an 
ordinary  soil  infusion  obtained  from  the  soil  used  in  the  preceding  series. 
The  infusions  were  quite  clouded  with  suspended  matter,  which  probably 
included  small  amounts  of  various  soil  constituents.  In  addition,  50  c.  c. 
of  a  nutrient  solution,  composed  of  2  gm.  of  sodium  chlorid,  0.2  gm.  of 
magnesium  sulphate,  0.5  gm.  of  potassium  sulphate,  and  6  drops  of  a  10 
per  cent  solution  of  ferric  chlorid  per  liter,  were  thoroughly  mixed  with 
the  sand  in  each  jar.  The  jars  were  loosely  covered  and  incubated  at 
room  temperature. 

After  periods  of  28,  56,  98,  and  157  days,  quantities  containing  200  gm. 
of  dry  sand  were  withdrawn,  500  c.  c.  of  distilled  water  added,  and  after 
shaking  vigorously  as  in   the  preceding  series,   were  filtered   through 


676 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  10 


Chaniberland-Pasteur  filters.  Nitrate,  calcium,  and  phosphoric  acid 
were  determined  in  the  filtrates  by  the  methods  already  referred  to. 
Nitrite,  when  present,  was  also  determined  by  the  Greiss-Ilosovay  method. 
The  calcium  determinations  for  the  98-day  period  have  been  omitted 
from  the  table,  owing  to  an  error  in  the  analytical  procedure.  The 
average  results  from  duplicate  incubations  are  recorded  in  Table  III. 

Table  III. — Effects  of  nitrification  on  the  solubility  of  fricalcium  phosphate  in  sand 

cultures 

[Results  expressed  in  parts  per  million) 


Aiter  28  days. 

After  56  days. 

After  98 
days. 

After  IS 7  days. 

Materials  added. 

6 

a 

d 

3  M 
0  0 

i 

0 

0    . 

1 

a  a 

0  0) 

•a 

•n 

0 
0    . 

0  (U 

'3  . 

ss 

0   . 

0  « 

6 

■3 
1° 

a 

3 
•0 

0    . 

0  " 

S 

iS 

6 

iz; 

6 

p4 

s 

g 

g 

S 

2 

c3 

S 

Calcium  carbonate 

7-S 

0.0 

72-5 

0.0 

g.o 

0.0 

77-6 

0.0 

12.  2 

0.0 

0.0 

n.o 

0.0 

77.6 

0.0 

Tricalcium  phosphate 

7.2 

.0 

52-1 

67.  S 

8.1 

.0 

SI.  8 

S2.3 

ICO 

.0 

ss-  s 

10. 2 

.0 

44.6 

4S-7 

Calcium  carbonate  and  tricalcium 

phosphate 

7-3 

.0 

75-1 

23-7 

8.3 

.0 

61.3 

18.0 

"•5 

.0 

21.6 

10.5 

.0 

64-3 

19.8 

Ammonium  sulphate 

24-5 

I- 5 

6.9 

28. 7 

Ammonium  sulphate  and  calcium 

carbonate 

92.0 

.0 

308.0 

.0 

79- 0 

.0 

298.0 

.0 

82.  s 

.0 

.0 

92.0 

.0 

345-1 

■  0 

Ammonixmi  sulphate  and  trical- 

cium phosphate 

2. 1 

.0 

62.6 

77-9 

7-S 

.0 

87.1 

79-9 

14.0 

.0 

83.9 

33- 0 

.0 

120.4 

8s.  I 

Ammonium  sulphate,  calcium  car- 

bonate,   and    tricalcium    phos- 

phate  

33-5 

8.2 

207.8 

17.7 

80.0 

.0 

269.2 

12.5 

80.0 

.0 

X5-4 

92.0 

.0 

304- 0 

15-8 

Dried  blood 

1. 1 

41.2 

23- 0 

.0 

3.0 

^0.0 

20.5 

.0 

7-2 

23.7 

.0 

29.0 

S.O 

35-5 

.0 

Dried  blood  and  calcium  carbonate. 

4S-0 

IS.  6 

150-3 

.0 

7S-0 

.0 

191-5 

•  0 

81.0 

.0 

.0 

91.0 

.0 

210.6 

.0 

Dried  blood  and  tricalcium  phos- 

phate     

•S 

2S.0 

^S.'! 

■I?-! 

I.O 

Si.o 

40.3 

"^S-? 

2.0 

2S.O 

S2.  2 

12.2 

22.  S 

4S-S 

46.  s 

Dried  blood,  calcium    carbonate. 

and  tricalcitun  phosphate 

1-3 

62.5 

100.2 

15-1 

2.9 

S2.0 

115- 7 

12.0 

2.6 

50.0 

18.3 

II.  0 

75- 0 

162.0 

12.7 

DISCUSSION  OF  EXPERIMENTAL  RESULTS 

Soil  series. — The  data  submitted  in  Table  11  show  that  active  nitri- 
fication took  place  in  the  soil  series.  For  example,  the  nitric  nitrogen 
increased  in  the  control  portions  from  8.5  p.  p.  m.,  originally  present,  to 
20  p.  p.  m.  in  28  days.  Where  ammonium  sulphate  was  added,  the  con- 
centration increased  to  98.5  p.  p.  m.,  while  dried  blood  yielded  91.0 
p.  p.  m.  of  nitric  nitrogen.  The  addition  of  calcium  carbonate  and  tri- 
calcium phosphate  either  singly  or  together  produced  very  slight,  if  any, 
effects  on  nitrification  in  this  series.^  After  subtracting  the  amounts  of 
nitric  nitrogen  in  the  controls,  it  is  found  that  78  per  cent  of  the  am- 
monium sulphate  and  71  per  cent  of  the  dried-blood  nitrogen  were 
oxidized  in  28  days. 

When  allowance  is  made  for  the  soluble  calcium  found  at  the  beginning 
of  the  experiment  (Table  I),  the  data  show  that  in  every  case  nitrification 


1  In  other  experiments  with  this  soil  the  addition  of  calcium  carbonate  has  slightly  stimulated  the  nitri- 
fication of  ammonium  sulphate,  but  not  of  dried  blood. 


Mar.  II,  1918     Nitrifying  Bacteria  and  Tricalcium  Phosphate  677 

was  accompanied  by  increases  in  the  solubility  of  the  calcium  present.* 
This  was  noted  to  some  extent  in  the  portions  to  which  no  nitrogenous 
additions  were  made ;  was  considerably  greater  when  dried  blood  was  sup- 
plied; and  was  greatest  with  the  addition  of  ammonium  sulphate.  The 
data  show,  however,  that  the  concentration  of  soluble  calcium  was  not 
increased  as  a  result  of  adding  tricalcium  phosphate.  On  the  other  hand, 
soluble  calcium  was  considerably  increased  in  a  number  of  cases  by  the 
addition  of  calcium  carbonate.  It  would  seem,  therefore,  that  calcium, 
in  the  form  either  of  the  carbonate  or  of  such  silicates  as  occur  in  this 
soil,  will  be  dissolved  by  the  biochemical  oxidation  products  in  preference 
to  tricalcium  phosphate. 

In  contrast  to  the  effects  on  the  solubility  of  calcium,  the  data  show 
that  nitrification  of  the  soil  nitrogen  and  that  added  as  dried  blood  was 
accompanied  in  each  case  by  a  well-defined  decrease  in  soluble  phos- 
phoric acid.  Where  ammonium  sulphate  was  added  alone,  the  amounts 
of  soluble  phosphoric  acid  found  at  the  28-  and  57-day  periods  were 
approximately  the  same  as  found  at  the  beginning  of  the  experiment. 
It  would  appear,  therefore,  that  the  solvent  action  of  the  bacteria  in 
this  case  was  almost  exactly  equal  to  the  precipitating  action  that  evi- 
dently took  place  in  the  control  and  dried-blood  portions.  After  157 
days,  however,  more  than  half  of  the  soluble  phosphoric  acid  originally 
present  in  this  portion  had  disappeared.  It  is  also  shown  that,  while 
larger  amounts  of  soluble  phosphoric  acid  were  found  where  tricalcium 
phosphate  had  been  added  than  in  the  control  portions,  the  increases 
can  not  be  definitely  ascribed  to  the  action  of  bacteria  in  any  case  except 
where  ammonium  sulphate  was  also  added,  and  then  only  without  the 
addition  of  calcium  carbonate.  For  in  all  other  cases  the  solubility 
immediately  after  adding  tricalcium  phosphate  was  equal  to  or  greater 
than  that  at  the  end  of  the  incubation  periods. 

In  the  absence  of  calcium  carbonate,  however,  the  oxidation  of  am- 
monium sulphate  dissolved  tricalcium  phosphate,  as  shown  by  the  fact 
that  the  concentration  of  soluble  phosphoric  acid  was  increased  in  28 
days  from  24.2  to  52.1  p.  p.  m.  Later  the  solubility  steadily  declined 
until  at  the  end  of  157  days  the  concentration  had  been  reduced  to 
30  p„  p.  m. 

Assuming  in  this  case  that  the  increase  in  nitric  nitrogen  over  the 
amounts  found  in  the  controls  was  due  to  the  oxidation  of  ammonium 
sulphate,  we  find  that  the  oxidation  of  78.0  p.  p.  m.  of  nitrogen  resulted 

'  It  should  be  clearly  understood  that  the  results  obtained  in  this  investigation  represent  the  algebraic 
sumanddifferenceoi  the  results  of  a  number  offerees.  In  the  first  placeit  is  highly  probable  that  biochem- 
ical agent"!  other  than  the  nitrif  jTng  organisms  are  capable  of  affecting  the  solubility  of  calcium  and  phos- 
phoric add  in  soils.  A  part  of  this  effect  may  be  referred  to  as  positive  and  a  part  as  negative,  since,  on  the 
one  hand,  carbonic  acid,  formed  in  the  life  process  of  bacteria,  is  a  solvent  for  calcium  and  phosolioric  acid, 
andon  the  other  hand,  the  organisms  themselves  absorb  phosphoric  acid  {8).  in  the  second  place  diffusion 
tends  to  bring  about  more  or  les^  fixation  in  soils.  The  concentration  at  a  given  moment,  therefore,  is 
really  dependent  on  the  Interaction  of  a  number  of  forces.  Consequently  a  full  explanation  of  the  results 
obtained  is  not  possible  at  present. 


678  Journal  of  Agricultural  Research  voi.  xii.  no.  10 

in  the  solution  of  27.9  p.  p.  m.  of  phosphoric  acid,  or  12.2  p.  p.  m.  phos- 
phorus. By  comparing  these  data  with  the  rate  of  solution  reported  by 
Hopkins  and  Whiting  from  solution  cultures,  it  will  be  seen  that,  while 
the  oxidation  of  i  pound  (454  gm.)  of  nitrogen  was  accompanied  by  the 
solution  of  2.033  pounds  (922  gm.)  of  phosphorus  in  their  experiments;^ 
in  these  experiments  with  soil  cultures  the  maximum  amount  of  phos- 
phorus made  soluble  was  only  0.156  pound  (70.8  gm.)  per  pound  of 
nitrogen  oxidized.  Therefore,  the  oxidation  products  of  ammonium 
sulphate  were  approximately  13  times  as  effective  in  dissolving  trical- 
cium  phosphate  in  solution  cultures  as  in  this  soil. 

It  is  interesting  to  note  that  the  addition  of  calcium  carbonate  tended 
to  lower  the  solubility  of  tricalcium  phosphate  wherever  applied. 

Sand  series. — Ammonium  sulphate,  when  added  alone,  underwent 
almost  no  nitrification  in  the  sand  series  (Table  III)  until  the  last  period 
of  the  experiment,  during  which  small  amounts  of  nitrate  were  formed.^ 
The  presence  of  calcium  carbonate,  however,  promoted  very  active 
nitrification  of  ammonium  sulphate.  In  this  case  the  concentration  of 
nitric  nitrogen  reached  its  maximum  (92  p.  p.  m.)  in  28  days.  On  the 
other  hand,  the  effects  resulting  from  the  addition  of  tricalcium  phos- 
phate only  began  to  be  manifested  by  the  fifty-sixth  day.  Later  the 
nitrate  content  slowly  increased  until  the  close  of  the  experiment,  when 
33  p.  p.  m.  were  found.  Nitrification  of  ammonium  sulphate  in  the 
proportions  containing  both  calcium  carbonate  and  tricalcium  phosphate 
was  not  so  pronounced  during  the  first  28  days  as  with  calcium  carbon- 
ate only,  but  later  the  effects  were  almost  identical. 

The  nitrification  of  dried  blood  in  sand  proved  to  be  especially  inter- 
esting in  that  the  intermediate  formation  of  nitrite  proceeded  much 
more  rapidly  than  the  oxidation  to  nitrate.^  When  dried  blood  alone 
was  added,  42.1  p.  p.  m.  of  nitrite  nitrogen  were  found  after  28  days  and 
only  I.I  p.  p.  m.  of  nitric  nitrogen.  Later  no  further  accumulation  of 
nitrite  took  place,  but  the  formation  of  nitrate  set  in  slowly  with  the 
result  that  29  p.  p.  m.  had  been  formed  by  the  close  of  the  experiment, 
but  even  then  5  p.  p.  m.  of  nitrite  still  remained. 

>  It  should  not  be  inferred  that  Hopkins  and  Whiting  claim  that  the  products  of  nitrification  will  dis- 
solve rock  phosphate  at  the  same  rate  in  soil  as  in  solution  cultures.  They  pointed  out  (j.  />.  405), for 
example,  that  nitrous  acid  may  combine  with  calcium  sihcate,  calcium  carbonate,  and  other  compounds 
in  soils  as  well  as  with  tricalcium  phosphate.  Nevertheless  they  hold  that  the  nitrite  bacteria  are  important 
agents  in  bringing  about  the  solution  of  rock  phosphate  in  field  soils. 

2  Appreciable  amounts  of  nitrate  were  formed  in  the  portions  which  were  intended  to  be  free  from  com- 
Kined  nitrogen.  The  nitrate  in  these  instances  probably  originated  from  organic  matter  held  in  suspen- 
sion in  the  soil  infusions  that  were  added.  The  amounts  formed,  however,  were  small  and  consistent  ia 
every  case,  increasing  steadily  from  an  average  of  7.3  p.  p.  m.  at  the  28-day  period  to  a  maximum  of  ii.i 
p.  p.  m.  at  98  days. 

5  Data  showing  that  nitrites  may  accumulate  in  nitrification  experiments  have  previously  been  reported 
(4),  but  this  is  a  condition  not  commonly  met  in  the  field.  The  accumulation  of  nitrites  indicates,  of 
course,  that  some  factor  in  the  medium  was  abnormal,  but  more  favorable  for  the  nitrite  bacteria  than 
for  the  nitrate  bacteria.  It  is  known  that  the  nitrate  bacteria  are  more  sensitive  to  adverse  conditions 
than  the  nitrite  bacteria. 


Mar.  u,  1918      Nitrifying  Bacteria  and  Tricalcium  Phosphate  679 

The  application  of  calcium  carbonate  notably  stimulated  the  nitrifi- 
cation of  dried  blood,  and  after  56  days  the  yield  was  approximately  the 
same  as  from  ammonium  sulphate.  On  the  other  hand,  tricalcium 
phosphate  produced  no  stimulation  in  the  nitrification  of  dried  blood  at 
any  period,  but  the  application  of  both  calcium  carbonate  and  tricalcium 
phosphate  promoted  active  nitrite  formation,  which  during  the  last 
period  of  the  experiment  resulted  in  the  oxidation  of  approximately  as 
much  nitrogen  as  in  any  other  case  in  the  experiment.  The  final  oxida- 
tion to  nitrate  in  this  case,  however,  was  very  feeble  throughout  the 
entire  experimental  period. 

It  is  evident  from  the  above  results,  therefore,  that,  while  tricalcium 
phosphate  may  promote  nitrification  in  the  absence  of  carbonate,  more 
favorable  conditions  for  nitrification  were  produced  by  calcium  carbonate 
than  by  tricalcium  phosphate. 

Large  amounts  of  calcium  carbonate  were  made  soluble  by  the  nitrifi- 
cation of  ammonium  sulphate  in  the  sand  series,  but  since  enfeebled 
nitrification  of  ammonium  sulphate  took  place  when  tricalcium  phos- 
phate only  was  added,  relatively  small  increases  in  soluble  calcium  were 
produced.  At  the  final  period  of  the  experiment,  however,  the  increase 
in  soluble  calcium  resulting  from  tricalcium  phosphate  and  the  small 
amounts  of  nitrogen  that  had  been  oxidized  agree  closely  with  theo- 
retical calculation. 

Again,  considerably  less  calcium  was  dissolved  in  the  nitrification  of 
dried  blood  than  in  that  of  ammonium  sulphate,  a  result  which  is  in 
harmony  with  generally  accepted  vievv's  regarding  the  nature  of  the 
oxidation  products  formed  in  the  two  cases.  With  the  latter  sulphuric 
acid  is  formed  in  addition  to  nitrous  acid,  while  with  the  former  carbonic 
acid  is  probably  one  of  the  end  products. 

The  limited  nitrification  of  dried  blood,  found  where  tricalcium  phos- 
phate had  been  added,  was  associated  with  a  lower  soluble-calcium 
content  than  occurred  where  tricalcium  was  added  alone.  But  in  view 
of  the  fact  that  not  more  than  30  per  cent  of  the  nitrogen  was  oxidized, 
it  is  probable  that  the  medium  remained  alkaline  as  a  result  of  the 
ammonification  of  the  dried  blood,  and,  consequently,  the  lower  yields 
of  soluble  calcium  may  have  been  due  to  slight  precipitarion  of  calcium 
as  calcium  carbonate. 

It  is  especially  interesting  that  smaller  amounts  of  calcium  were 
dissolved  by  the  nitrification  of  ammonium  sulphate  and  dried  blood 
in  the  presence  of  both  calcium  carbonate  and  tricalcium  phosphate 
than  with  calcium  carbonate  alone. 

Considering  the  phosphoric-acid  determinations,  it  is  at  once  apparent 
that  nitrification  did  not  produce  an  increase  in  the  solubility  of  tri- 
calcium phosphate  when  carbonate  was  present.  In  the  absence  of 
carbonate,  however,  an  increase  in  solubility  resulted  from  the  nitrifica- 
tion of  ammonium  sulphate.     In  this  case,  although  nitrification  was 


68o  Journal  of  Agricultural  Research  vot.  xii.  no.  w 

less  active  than  elsewhere,  the  data  show  that  the  small  amounts  of 
oxidation  products  formed  dissolved  tricalcium  phosphate.  On  the 
other  hand,  wherever  calcium  carbonate  was  also  present,  the  oxidation 
products  not  only  combined  with  it,  but  the  initial  solubility  of  the 
tricalcium  phosphate  was  very  materially  lowered  as  well.  This  was 
true  in  the  nitrification  of  both  ammonium  sulphate  and  dried  blood.^ 

INTERPRETATION  OF  RESULTS 

In  making  a  practical  interpretation  of  these  investigations  it  should  be 
borne  in  mind  that  fertile  soils  commonly  contain  at  least  small  amounts 
of  carbonate,  that  even  the  so-called  acid  soils  frequently  contain  con- 
siderable amounts  of  bicarbonate,  and  that  the  presence  of  calcium 
carbonate  in  soils  is  generally  considered  to  promote  conditions  that  are 
favorable  for  the  growth  of  most  crops.  Furthermore,  large  amounts 
of  calcium  carbonate  are  being  applied  to  soils  in  many  localities, 
especially  in  the  humid  sections.  In  the  Central  West,  for  example, 
ground  limestone  is  being  applied  on  a  large  scale,  and  generally  it  is 
recommended  that  the  application  be  repeated  every  few  years. 

Under  the  conditions  that  result,  chemical  reasoning  (j)  and  the  experi- 
mental results  reported  above  agree  in  suggesting  that  the  action  of  the 
biochemical  oxidation  products,  formed  in  the  nitrification  of  organic 
nitrogen,  would  be  spent  on  the  carbonate  and  not  on  tricalcium  phos- 
phate. Furthermore,  it  seems  doubtful  whether  this  could  be  avoided 
by  the  application  of  limestone  and  rock  phosphate  at  different  times 
in  a  rotation,  as  was  suggested  by  Hopkins  and  Whiting.  Although 
it  is  possible  that  under  this  condition  the  particles  of  rock  phosphate 
may  chance  to  occur  in  local  centers  that  are  somewhat  removed  from 
solid  particles  of  calcium  carbonate,  and  nitrification  may  happen  to 
take  place  in  these  centers,  it  does  not  even  then  necessarily  follow 
that  the  phosphate  would  be  dissolved.  For  such  centers  would  probably 
always  be  in  contact  with  soil  silicates,  and  the  above  data  indicate 
that  at  least  some  soil  silicates  may  be  attacked  by  the  products  of 
nitrification  in  preference  to  tricalcium  phosphate.  When  the  con- 
ditions permit  the  accumulation  of  considerable  acidity,  however,  such 
as  probably  obtain  when  ammonium  sulphate  is  applied  to  a  soU  low 
in  carbonate,  it  was  found  that  small  amounts  of  the  phosphate  were 
dissolved.  But  then  a  degree  of  acidity  that  is  distinctly  injurious  to 
crops  may  soon  develop,  as  has  been  found  by  field  trials  in  a  number 
of  localities.  Furthermore,  bicarbonate,  arising  from  the  action  of 
carbonic  acid  on  the  solid  particles  of  calcium  carbonate,  would  certainly 
tend  to  diffuse  toward  the  supposed  centers  of  acidity,  thus  precipitating 

•  Soluble  phosphoric  acid  may  have  been  utilized  to  some  extent  in  the  life  processes  of  the  bacteria 
present,  as  was  found  by  Tottingham  and  Hofimann  (,8).  But  the  simultaneous  losses  m  soluble  calcium 
also  suggest  the  precipitation  of  phosphoric  acid,  a  view  that  is  in  harmony  with  the  well-known  fact  that 
calcium  carbonate  may  precipitate  phosphoric  acid  irom  solution. 


Mar.  II,  i9i8     Nitrifying  Bactevla  and  Triccdcium  Phosphate  68 1 

the  phosphoric  acid.  The  results  of  many  studies  on  the  concentration 
of  nitric  nitrogen  in  soils  supporting  growing  crops  show,  for  example 
that  diffusion  is  an  important  force  in  maintaining  equilibrium  in  soils. 

Tricalcium  phosphate  can  not  be  converted  into  monocalcium  phos- 
phate (the  water-soluble  form)  without  active  acidity  being  developed.^ 
Nor  are  acids  neutralized  by  converting  tricalcium  phosphate  into  mono- 
calcium  phosphate,  since  the  latter  is  an  acid  compound.  But  when  the 
acidity  necessary  to  the  solution  of  tricalcium  phosphate  is  neutralized, 
the  phosphoric  acid  will  be  precipitated.  If  it  be  desirable  that  a  con- 
dition either  of  neutrality  or  slight  alkalinity  obtain  in  soils,  as  has  been 
widely  taught,  it  is  difficult  to  see  how  more  than  traces  of  monocalcium 
phosphate  can  exist  therein  at  the  same  time. 

It  should  also  be  recalled  that  untreated  rock  phosphate  generally 
contains  considerable  amounts  of  calcium  carbonate  intimately  com- 
mingled with  the  phosphate.  Before  dilute  acids,  formed  by  the  action 
of  bacteria  or  otherwise,  can  dissolve  the  phosphate,  the  carbonate  must 
first  be  neutralized,  as  is  commonly  recognized  in  the  commercial  pro- 
cesses employed  in  the  manufacture  of  acid  phosphate. 

It  seems  possible,  however,  that  the  nitrifying  bacteria  may  dissolve 
limited  amounts  of  rock  phosphate  in  acid  soils.  The  soil  used  in  this 
investigation,  although  low  in  carbonate,  was  not  acidic,  and  more  active 
solution  of  tricalcium  phosphate  would  probably  take  place  in  an  acid 
soil.  But,  in  any  case,  with  the  possible  exception  of  very  sandy  types 
of  soil,  it  is  probable  that  phosphoric  acid,  made  soluble  by  the  nitrifying 
bacteria,  would  tend  to  become  fixed  through  being  brought  into  con- 
tact with  other  soil  constituents  by  diffusion.  Many  investigations  have 
shown,  for  example,  that  acid  phosphate  soon  becomes  fixed,  even  in 
acid  soils. 

From  these  investigations  it  seems,  therefore,  that  while  the  nitrite 
bacteria  are  capable  of  effecting  the  solution  of  tricalcium  phosphate 
under  restricted  conditions,  they  are  not  the  potent  agents  in  the  solu- 
tion of  rock  phosphate  in  the  field  that  Hopkins  and  Whiting  were  led  to 
infer  from  their  experiments  with  solution  cultures. 

In  the  opinion  of  the  author  the  results  obtained  in  this  investigation 
should  not  be  interpreted  as  being  definitely  opposed  to  the  use  of  un- 
treated rock  phosphate  as  a  fertilizer.  It  is  true  that  the  means  by  which 
tricalcium  phosphate  is  made  soluble  in  soil  have  not  been  definitely 
determined,  but  the  important  fact  remains  that  in  various  localities 
beneficial  effects  on  the  growth  of  crops  have  frequently  been  produced 
by  rock  phosphate.  It  therefore  remains  for  further  investigation  to 
lay  bare  the  reasons.  The  suggestions  offered  by  Truog  (9)  in  this  con- 
nection seem  to  be  deserving  of  special  consideration. 

•  It  is,  of  course,  understood  that  hydrolysis  is  excepted  in  this  case.  The  absolute  increases  in  soluble 
phosphoric  acid  in  soib  resulting  from  the  hydrolysis  of  tricaldum  phosphate  are  probably  quite  small. 


682  Journal  of  Agricultural  Research  voi.  xii,  no.  lo 

SUMMARY 

The  investigations  reported  in  this  paper  include  a  study  of  (i)  the 
effects  of  adding  calcium  carbonate,  tricalcium  phosphate,  and  ammo- 
nium sulphate  on  the  immediate  solubility  of  calcium  and  phosphoric 
acid  in  a  light  sandy  loam  soil;  (2)  the  effects  of  nitrification  of  the  soil 
nitrogen,  ammonium  sulphate,  and  dried  blood  on  the  solubility  of  the 
naturally  occurring  calcium  and  phosphoric  acid;  (3)  the  effects  of  nitri- 
fication in  soil  and  sand  cultures  on  the  solubility  of  tricalcium  phosphate 
both  with  and  without  the  application  of  calcium  carbonate. 

The  following  results  were  obtained: 

(i)  The  addition  of  calcium  carbonate  produced  no  effect  on  the  imme- 
diate solubility  of  the  soil  calcium  or  that  added  as  tricalcium  phosphate. 
The  addition  of  tricalcium  phosphate  produced  an  increase  of  about  5 
p.  p.  m.  of  soluble  calcium  and  11.2  p.  p.  m.  of  soluble  phosphoric  acid, 
while  the  addition  of  ammonium  sulphate  brought  about  an  increase  in 
water-soluble  calcium  from  27.5  to  90.1  p.  p.  m. 

(a)  Active  nitrification  of  ammonium  sulphate  and  dried  blood  took 
place  in  the  soil  series,  and  at  the  same  time  notable  increases  in  soluble 
calcium  were  produced. 

(3)  No  increase  in  the  solubility  in  water  of  the  soil  phosphates  or  of 
tricalcium  phosphate  was  produced  by  bacterial  action  except  in  the 
nitrification  of  ammonium  sulphate  when  added  wdthout  calcium  car- 
bonate. In  this  case  0.156  pounds  (70.8  gm.)  of  phosphorous  were  dis- 
solved for  every  pound  of  nitrogen  oxidized,  whereas  Hopkins  and  Whit- 
ing found  from  solution  cultures  that  2.033  pounds  (922  gm.)  were  dis- 
solved for  every  pound  of  nitrogen  oxidized. 

(4)  The  addition  of  calcium  carbonate  brought  about  an  increase  in 
soluble  calcium  but  tended  to  lower  the  solubility  of  tricalcium  phosphate. 

(5)  In  the  absence  of  calcium  carbonate  the  nitrification  of  ammonium 
sulphate  in  sand  cultures  was  accompanied  by  the  solution  of  theoretical 
amounts  of  tricalcium  phosphate.  When  calcium  carbonate  was  present, 
however,  the  solubility  of  tricalcium  phosphate  was  not  increased  by 
nitrification. 

(6)  The  formation  of  nitrite  from  dried  blood  took  place  more  rapidly 
in  the  sand  cultures  than  the  formation  of  nitrate. 

(7)  Tricalcium  phosphate  was  not  dissolved  by  the  nitrification  of  dried 
blood  in  the  sand  series. 

(8)  It  was  found  that  calcium  carbonate  promoted  more  active  nitri- 
fication than  tricalcium  phosphate. 

(9)  The  experimental  results  indicate  that  the  nitrification  of  organic 
forms  of  nitrogen  does  not  increase  the  solubility  of  rock  phosphate  under 
field  conditions  that  are  favorable  to  crop  growth.  It  is  possible,  how- 
ever, that  the  nitrification  of  ammonium  sulphate  may  result  in  the  solu- 
tion of  small  amounts  of  tricalcium  phosphate  in  soil  low  in  carbonate. 


Mar.  ij,  is,i8     Nitrifying  Bacteria  and  Tricalcium  Phosphate  683 

LITERATURE  CITED 
(i)  Cameron,  F.  K.,  and  Bell,  J.  M. 

1907.  THE  ACTION  OF  WATER  AND  AQUEOUS  SOLUTIONS   UPON   SOIL  PHOSPHATES." 

U.  S.  Dept.  Agr.  Bur.  Soils  Bui.  41,  58  p.,  5  fig. 

(2)  Gaither,  E.  W. 

191 2.  a  new  apparatus  for  the  determination   op  carbon  dioxide.    in 

Jotir.  Indus,  and  Engin.  Chem.,  v.  4,  no.  8,  p.  611-613,  2  fig. 

(3)  Hopkins,  C.  G.,  and  Whiting,  A.  L. 

1916.  soil  BACTERIA  AND  PHOSPHATES.     111.  Agr.  Exp.  Sta.  Bui.  190,  p.  395-406. 
Foreword  by  E.  Davenport. 

(4)  Kellev,  W.  p. 

1916.  NITRIFICATION  IN  SEMIARID  SOILS,  I.     In  JouT.  Agr.  Research,  v.  7,  no. 
10,  p.  417-437.     Literature  cited,  p.  436-437. 

(5)  LiFMAN,  J.  G.,  McLean,  H.  C,  and  Lint,  H.  C. 

I916.    SULFUR  OXIDATION   IN  SOILS  AND     ITS    EFFECT    ON    THE    AVAILABILITY  OK 

MINERAL  PHOSPHATES.     In  Soil  Sci.,  V.  2,  no.  6,  p.  499-538,  5  fig. 
Literature  cited,  p.  535-538. 

(6)  LuPTON,  N.  T. 

1893.   THE    EFFECT     OF    DECOMPOSING     ORGANIC     MATTER    ON     NATURAL    PHOS- 
PHATES.    In  Ala.  Agr.  Exp.  Sta.  Bui.  48,  p.  i-io. 

(7)  McDowell,  M.  S. 

1908.  is  the  phosphoric  acid  op  floats  made  soluble  by  rotting  manure? 

In  Pa.  Agr.  Exp.  Sta.,  Ann.  Rpt.,  1907-08,  p.  175-178. 

(8)  ToTTiNGHAM,  W.  E.,  and  Hoffmann,  C. 

1913.  NATURE    of    the    CHANGES    IN    THE     SOLUBILITY    AND    AVAILABILITY    OF 

PHOSPHORUS  IN  FERMENTING  MIXTURES.    Wis.  AgT.  Exp.  Stfi.  Research 
Bui.  29,  p.  275-321,  3  fig. 

(9)  Truog,  E. 

1912.  factors  influencing  the  availability  of  rock   phosphate.    wjs. 

Agr.  Exp.  Sta.  Research  Bui.  2c,  p.  17-51,  4  fig. 
38325°— 18 6 


ADDITIONAL  COPIES 

OP  THLS  PtJBLICATION  MAY  BE  PKOCUBED  FROM 

THE  SUPERINTENDENT  OF  DOCUMENTS 

GOVERNMENT  PRINTING  OFFICE 

WASfflNOTON,  D.  C. 

AT 

30  CENTS  PER  COPY 
SoBSCEiPTiON  Price,  J3.00  Peb  Year 

V 


Vol.  XII  NIARCH    IS,  1918  No.   11 

JOURNAL  OF 

AGRICULTURAL 
RESEARCH 


CONTKNTS 

Page 

Respiration  of  Stored  Wheat    -        -        -        -        -        -      685 

C.  H.  BAILEY  and  A.  M.  GURJAR 

( Contribution  Irom  Minnesota  Agricultural  Experiment  Station ) 

E£fects  of  Mistletoe  on  Young  Conifers  -        -        -        -      715 

JAMES  R.  WEIR 

(Contribution  from  Bureau  of  Plant  Industry ) 

Determination  of  Fatty  Acids  in  Butter  Fat    -        -        -      719 
E.  B.  HOLLAND  and  J.  P.  BUCKLEY,  Jr. 

(Contribution  from  Massachusetts  Agricultural  Experiment  Station) 


PUBLISHED  BY  AUTHORITY  OF  THE  SECRETARY  OF  AGRICULTURE, 

WITH  THE  COOPERATION  OF  THE  ASSOCIATION  OF  AMERICAN 

AGRICULTURAL  COLLEGES  AND  EXPERIMENT  STATIONS 


\VASMINGTON,  D.  C. 


WAtHINOTON:  OOVERNMENT  PRINTINO  OFFICE  :  Itll 


EDITORIAL  COMMITTEE  OF  THE 

UNITED    STATES    DEPARTMENT    OF    AGRICULTURE    AND 

THE  ASSOCIATION  OF  AMERICAN  AGRICULTURAL 

COLLEGES    AND    EXPERIMENT    STATIONS 


FOR  THE  DEPARTMENT 


FOR  THE  ASSOCIATION 


KARLF.KELLERMAN.  Chairman    RAYMOND  PEARL 


Physiologist  and  Associate  Chief,  Bureau 
of  Plant  Industry 

EDWIN  W.  ALLEN 

Chief,  Office  of  Experiment  Stations 

CHARI,EvS  L.  MARLATT 

Eniomologisl  and  Assislajit  Chief,  Bureau 
of  Enionwloay 


Biolpqisl.  Maine  Agriculliiral  Experiment 
Station 

H.  P.  ARMSBV 

Director .  Institute  of  Animal  A'utriliim,  The 
Pennsyhania  State  College 

E.  M.  FREEMAN 

Botanist,  Plant  Pailwlot/ist  and  Assistant 
Dean,  Agricultural  Experiment  Station  of 
the  University  of  Minnesota 


All  correspondence  regarding  articles  from  the  Department  of  Agriculture  should  be 
addressed  to  Karl  F.  Kellerman,  Journal  of  Agricultural  Research,  Washington,  D.  C. 

*  Dr.  Pearl  has  undertaken  special  work  in  connection  with  the  war  emergency ; 
therefore,  imtil  further  notice  all  correspondence  regarding  articles  from  State  Experi- 
ment Stations  should  be  addressed  to  H.  P.  Armsby,  Institute  of  Animal  Nutrition, 
State  College,  Pa. 


mm 


Vol..  XII 


Washington,  D.  C,  March  i8,  1918 


No.  II 


RESPIRATION  OF  STORED  WHEAT  ^ 

By  C.  H.  Bailey,  Cereal  Technologist,  and  A.  M.  Gurjar,  Assistant  in  Agrkultural 
Biochemistry,  Division  of  Agricultural  Biochemistry,  Minnesota  Agricultural  Experi- 
ment Station 

INTRODUCTION 

The  preservation  in  storage  of  large  quantities  of  thrashed  grain  in- 
volves certain  difficulties.  In  addition  to  protection  from  vermin,  it  has 
long  been  known  that  the  grain  must  be  dry  when  stored,  and  the  ingress 
of  moisture  prevented.  The  ancients  were  aware  that  damp  grain,  or 
similar  vegetable  material,  will  heat  and  become  decomposed  when 
stored  in  quantities.  It  is  evident  that  in  prehistoric  times  carefully 
constructed  receptacles  were  employed  for  the  protection  of  the  reserve 
supplies  of  cereals.  Those  ancient  people  of  India,  the  Hindus,  some- 
times resort  to  the  use  of  receptacles  which  are  submerged  in  cool  water, 
reducing  the  temperature  of  the  stored  grain  and  also  the  supply  of  air. 
In  modern  practice  carefully  constructed  tanks  or  silos,  frequently  of 
large  capacity,  are  chiefly  used  for  the  storage  of  grain. 

CAUSE  OF  THE  HEATING  OF  GRAIN 

The  cause  of  heating  of  damp  vegetable  matter  was  not  known,  how- 
ever, until  comparatively  recent  times,  and  it  is  only  mthin  the  last 
decade  or  two  that  any  data  have  been  accumulated  which  indicate  the 
exact  effect  of  various  factors  on  the  rate  of  heating  of  grain  and  similar 
material.  That  the  phenomenon  known  as  respiration  is  responsible  for 
the  heat  energy  released  in  a  mass  of  damp  grain  is  shown  by  modern 
research.  Loew  (i8pc>y,  in  discussing  the  fermentation  and  heating  of  leaf 
tobacco,  maintained  that  the  release  of  energy  and  rise  in  temperature 
was  occasioned  by  the  activity  of  the  oxidizing  enzyms  of  the  leaf  cells. 
The  microbial  flora  were  not  believed  to  play  any  considerable  part  in 
these  changes.  In  advancing  this  view  he  opposed  vigorously  the  bac- 
terial-fermentarion  theory  of  Suchsland  {i8gi). 

Rahn  {1910)  states  that  the  curve  of  the  process  of  spontaneous  heat- 
ing of  organic  matter,  including  grain,  would  not  in  itself  indicate  whether 

00  '  Published  with  the  approval  of  the  Director,  as  Paper  97,  Journal  Series,  Minnesota  Agricultural  Experi- 

^ir  jnent  Station. 

_  2  Bibliographic  citations  in  parentheses  (italic)  refer  to  "  Literature  cited,"  pp.  710-713. 


.  -J         Journal  of  Agricultural  Research, 
Washington,  D.  C. 


(685) 


Vol.  XII,  No.  n 
Mar.  18,  1918 
Key  No.  Minn.  26 


686  Journal  of  Agricultural  Research  voi.  xii.  No.  n 

the  heat  was  produced  by  chemical  or  microbial  causes.  Gore  {191 1, 
P-  33)  concluded  that  in  self-heating,  physiological  processes  are  probably 
the  first  to  operate.  Oxidizing  enzyms  are  the  active  agents  in  many 
cases.  Chemical  oxidation  would  intervene  only  when  the  temperature 
had  been  raised  to  the  combustion  point  of  the  substances  present.  Gore 
presents  formulas  and  graphs  illustrating  the  theoretical  progress  of  the 
accumulation  of  heat. 

The  studies  of  E.  M.  Bailey  {1912)  on  the  ripening  of  bananas  indicate 
that  bacterial  activity  is  not  responsible  for  the  heat  produced  or  other 
changes  resulting  during  the  ripening  of  the  fruit. 

Nabokich  {1903)  found  that  seeds  which  had  been  sterilized  by  chemi- 
cals such  as  mercuric  chlorid  respired  during  the  first  few  days  as  much 
carbon  dioxid  as  did  the  controls.  In  some  instances  the  respiration 
of  the  sterilized  seeds  exceeds  that  of  the  controls. 

Respiration  may  be  briefly  defined  as  the  release  of  energy  through 
the  biochemical  oxidation  of  organic  compounds  as  accelerated  by  certain 
enzyms.  Carbon  dioxid  and  water  are  the  characteristic  chemical  end 
products.  It  is  shown  by  De  Saussure  {1804)  that  respiration  was  ac- 
companied by  the  disappearance  of  oxygen.  Pfliiger  {1875)  maintained 
that  the  inspired  oxygen  combines  in  some  manner  with  cyanogen  radi- 
cals of  the  living  protoplasm.  This  effects  a  readjustment,  as  the  result 
of  which  carbon  dioxid  and  water  are  eliminated.  The  decomposition 
is  of  an  explosive  character,  and  the  reaction  liberates  heat  energy. 

Verworn  {1899,  p.  483)  evolved  the  "  biogen"  theory,  according  to  which 
the  oxygen  enters  the  "biogen  molecules,"  which  are  thus  rendered  less 
stable.  Slight  impulses  are  then  required  to  bring  about  a  chemical 
union  of  this  oxygen  with  the  carbon  in  the  cyanogen  group. 

Since  the  grain  itself  is  a  poor  conductor  of  heat,  it  follows  that  the  heat 
energy  released  through  respiration  accumulates  in  the  mass  in  propor- 
tion to  its  bulk  so  that  the  increase  in  temperature  may  in  time  become 
very  marked. 

MATERIAL  OXIDIZED  IN  RESPIRATION 

In  the  case  of  resting  tissues  and  storage  structures,  such  as  the  grain 
kernel  or  caryopsis,  the  exact  character  of  the  substrate  or  material 
oxidized  in  respiration  is  of  significance.  Wehmer  {1892)  fed  Aspergillus 
on  peptone,  and  found  it  capable  of  satisfying  its  requirements  from 
this  source.  Gore  (19 14)  found  that  in  ripening  bananas  the  rate  of 
starch  hydrolysis  paralleled  the  rate  of  respiration.  There  was  slight 
change  in  the  quantity  of  protein  and  fats  in  the  fruit.  Maige  and 
Nicolas  {1910)  found  that  the  immersion  of  etiolated  leaves,  shoots,  or 
seedlings  in  sugar  solutions  resulted  in  an  increased  rate  of  respiration. 

Langworthy  and  Milner  (1913)  observed  that  in  ripening  bananas  the 
"thermal  quotient,"  or  quantity  of  heat  produced  per  unit  of  carbon 
dioxid  respired,  indicated  the  combustion  of  carbohydrate.     The  thermal 


Mar.  is,  lyib  Respuation  of  Stored  Wheat  687 


quotient  found  for  one  period  of  their  studies,  when  conditions  for 
accurate  measurement  were  at  an  optimum,  was  2.6.  This  is  so  nearly 
the  theoretical  for  the  combustion  of  carbohydrates  as  to  indicate  that 
little  else  was  involved.  As  pointed  out  by  Dr.  Milner  in  a  private 
communication,  the  thermal  quotient  when  fat  is  burned  is  3.4,  while 
that  when  protein  is  burned  is  2.9. 

Hasselbring  and  Hawkins  (1913)  found  no  general  correlation  between 
the  total  sugar  content  of  the  sweet  potato  and  its  respiratory  activity. 
Cane  sugar  is  relatively  stable  and  does  not  appear  to  be  used  in  the 
process  of  respiration.  A  simultaneous  decrease  in  the  reducing  sugars 
and  the  respiratory  activity  was  observed.  The  reducing  sugars  are, 
in  their  opinion,  the  immediate  source  of  respiration  material. 

That  the  fat  or  ether  extract  of  the  wheat  embryo  is  not  the  principal 
substrate  upon  which  the  respiratory  enzyms  act,  and  that  it  is  not 
burned  or  destroyed  during  germination  is  indicated  by  the  experiments 
of  Le  Clerc  and  Breazeale  (1911,  p.  12).  Their  data  show  that  there  w^as 
more  fat  in  the  total  plant  (seed  residues  plus  axes)  at  practically  all 
stages  of  germination  than  there  was  in  the  original  seeds.  Accordingly 
little  energy  could  have  been  derived  from  the  oxidation  of  the  fats;  on 
the  contrary,  energy  derived  from  some  other  source  must  have  been 
utilized  in  the  synthesis  of  the  additional  fat  produced  in  the  seedling. 
All  available  evidence  therefore  seems  to  indicate  that  the  heat  of  respira- 
tion is  produced  by  the  oxidation  of  reducing  sugars. 

SEAT  OF  RESPIRATION  IN  THE  WHEAT  KERNEL 

There  are  a  number  of  reasons  for  believing  that  the  germ  or  embryo 
of  the  wheat  kernel  is  the  location  of  the  larger  part  of  the  biological 
oxidation  that  occurs  incidental  to  respiration.  The  embryo  is,  in  a 
general  w^ay,  decidedly  richer  in  enzyms  than  is  the  endosperm  or  any 
other  kernel  structure.  Brown  and  Morris  {i8go)  consider  that  the 
endosperm  of  the  ripened  kernel  is  no  longer  a  vital  tissue.  The  secretion 
of  diastase  and  other  enzyms  is  assigned  to  the  scutellum,  an  organ  of  the 
embryo.  Maim  and  Harlan  (1915)  concluded  that  in  germinating 
barley  the  conversion  of  the  endosperm  is  effected  by  enzyms  secreted 
by  the  epithelial  layer  of  the  scutellum. 

Karchevski  (1903)  found  the  energy  of  carbon-dioxid  respiration 
to  be  12  times  as  great  in  the  wheat  embryos  as  in  the  seeds  themselves. 
Burlakow  {1898)  states  that  the  respiratory  activity  of  the  germ  is  20 
times  greater  than  is  that  of  the  endosperm.  Wender  {^1905)  called 
attention  to  the  pronounced  catalase  activity  of  the  germ  structures, 
as  contrasted  with  mill  products  derived  almost  exclusively  from  the 
endosperm.  This  has  also  been  observed  in  an  unpublished  study  of 
catalases  which  was  made  in  this  laboratory.  This  fact  is  more  signifi- 
cant, in  view  of  Appleman's  {1915)  discovery  that  catalase  activity  in 
potato  juice  shows  a  striking  correlation  with  respiratory  activity  in  the 


688  Journal  of  Agricultural  Research  voi.  xu.no.  n 

tubers.  If  catalase  activity  parallels  respiratory  activity,  it  may  be 
reasoned  by  analogy  that  those  structures  of  the  wheat  kernel  which 
exhibit  the  greatest  catalase  activity  are  the  seat  of  the  larger  part  of 
respiration. 

Barnes  and  Grove  {1916)  lend  further  support  to  the  hypothesis  that 
the  seat  of  oxidation  activity  is  in  the  embryo  by  their  obsenl^ation  that 
in  air-dry  wheat  the  embryo  becomes  shrunken  after  a  time,  while  the 
food  materials  of  the  endosperm  are  unimpaired.  This  is  interpreted 
by  them  to  indicate  a  destruction  of  the  material  of  the  embryo  itself 
as  the  result  of  respiration  when  the  dryness  of  the  kernel  suppresses 
diffusion.  Were  respiration  equally  pronounced  in  the  endosperm  it 
too  should  exhibit  a  similar  loss  of  material. 

Osterhout's  (1917)  observation  that  oxidation  is  more  rapid  in  the 
nucleus  than  in  the  cytoplasm  might  lead  to  the  deduction  that,  since 
the  embryo  cells  of  the  wheat  caryopsis  have  a  much  larger  proportion 
of  nuclei  than  the  endosperm  cells,  oxidation  should  proceed  more  rapidly 
in  the  embryo  tissues. 

All  of  the  above  facts  are  in  harmony  with  what  might  logically  be 
expected.  The  principal  release  of  energy  as  the  result  of  biological 
combustion  should  occur  in  the  structure  where  such  energy  is  required 
for  the  synthesis  of  new  organic  compounds.  Since  the  embryo  is 
endowed  particularly  with  that  function,  respiration  must  of  necessity 
be  most  pronounced  in  it,  if  not  confined  to  it. 

MEASUREMENT  OF  THE  RATE  OF  RESPIRATION 

There  are  two  general  methods  which  may  be  used  in  the  quantita- 
tive estimation  of  the  rate  of  respiration  of  vegetable  material.  One 
that  is  employed  where  faciHties  are  available  is  to  measure  in  terms 
of  Calories  the  heat  energy  released  per  unit  of  time  and  material.  The 
elaborate  device  for  this  purpose  is  known  as  the  respiration  calorimeter. 
The  second  general  method  includes  the  determination  of  one  of  the  end 
products  of  the  reaction,  carbon  dioxid.  This  method  may  easily  be 
made  decidedly  accurate  without  entailing  the  assembling  of  a  calori- 
meter. In  using  the  calorimeter  both  carbon  dioxid  respired  and  heat 
evolved  may,  if  desired,  be  determined  simultaneously. 

Inasmuch  as  the  writers  were  not  provided  with  a  respiration  calori- 
meter suited  to  this  purpose,  the  carbon  dioxid  evolved  b}^  stored  wheat 
was  measured,  and  from  this  data  the  rate  of  respiration  was  calculated. 
Truog's  {1915)  method  and  absorption  tower  was  used  for  this  purpose, 
the  tower  being  slightly  modified,  or  rather,  added  to,  in  order  to  adapt 
it  to  the  present  work.  The  procedure  followed  and  a  description  of 
the  apparatus  has  been  published  by  the  junior  author  (Gurjar,  19 17). 
To  compute  the  Calories  of  heat  evolved,  the  factor  found  by  Langworthy 
and  Milner  may  be  employed;  i  gm.  of  respired  carbon  dioxid  equals 
2.6  Calories  of  heat. 


Mar.  18, 1918  Respiration  of  Stored  Wheat  689 

Calcium-chlorid  towers  were  used  as  respiration  chambers.  Paraffined 
wire  netting  was  fitted  into  the  constriction  near  the  base,  and  on  this 
the  grain  rested.  Rubber  connections  were  made  as  short  as  possible, 
and  all  stoppers  and  tubing  were  thoroughly  paraffined  to  prevent 
selective  absorption  of  carbon  dioxid.  In  all  instances  where  the  same 
lot  of  grain  was  worked  with  at  different  moisture  contents,  the  several 
portions  were  brought  to  approximately  the  desired  percentage  of 
moisture  by  adding  water  from  a  burette,  at  the  same  time  stirring 
thoroughly.  The  samples  were  allowed  to  stand  in  sealed  jars  for  three 
days,  in  order  to  insure  uniform  distribution  of  the  water  through  the 
kernels.  It  had  been  found  by  determining  the  rate  of  swelling  of  the 
kernels  that  they  reached  their  maximum  size  in  considerably  less  than 
three  days,  and  from  this  it  was  concluded  that  the  moisture  distribution 
would  be  complete  in  the  3-day  interval.  When  the  grain  was  ready 
to  work  with,  a  weighed  quantity  v^^as  sealed  into  the  tower.  Samples 
were  taken  at  this  time  for  the  determination  of  moisture.  The  amount 
of  grain  placed  in  the  tower  varied  with  the  moisture  content,  about  500 
gm.  being  employed  in  the  case  of  the  lower  moisture  limits,  while 
about  300  gm.  were  used  when  the  percentage  of  moisture  exceeded 
15.5  per  cent.  In  this  manner  convenient  and  accurately  determinable 
quantities  of  carbon  dioxid  were  obtained.  After  sealing  the  grain  into 
the  glass  cylinders,  the  air  was  removed  and  replaced  by  carbon-dioxid- 
free  air.  The  towers  were  then  placed  in  the  thermostat,  which,  except 
when  temperature  was  the  variable,  was  maintained  at  37.8°  C.  (100°  F.). 

The  period  of  incubation  was  fixed  at  four  days,  the  exact  number 
of  hours  being  noted  at  the  time  of  removing  the  respiration  chambers 
from  the  thermostat.  The  accumulated  carbon  dioxid  v/as  then  re- 
moved through  the  tubulure  at  the  bottom  of  the  tower,  carbon-dioxid- 
free  air  being  simultaneously  admitted  through  the  top.  The  carbon 
dioxid  was  absorbed  in  N/4  barium  hydroxid  [Ba(0H)2]  solution  in 
the  special  absorption  tower,  as  described  by  Gurjar  in  the  above- 
mentioned  paper.  The  respiration  data  given  in  the  tables  are  stated 
in  terms  of  milligrams  of  carbon  dioxid  respired  per  24  hours  by  each 
100  gm.  of  dry  matter. 

RELATION  OF  THE  MOISTURE  CONTENT  OF  WHEAT  TO  THE  RATE  OF 

RESPIRATION 

The  observation  of  Bonnier  and  Mangin  {1885)  that  respiration  of 
living  plants  varies  directly  with  the  humidity  of  the  air  might  be 
interpreted  to  mean  that  the  moisture  content  of  the  tissues  increased 
in  a  humid  atmosphere.  This  increase  in  turn  may  have  occasioned 
the  rise  in  the  rate  of  respiration.  Maquenne  {1900)  concluded  that 
a  reduction  in  the  moisture  content  of  seeds  is  accompanied  by  a  reduc- 
tion in  the  rate  of  respiration,  and  Lund  (1894)  discovered  that  the 
desiccation  of  roots  and  tubers  reduced  their  rate  of  respiration. 


690  Journal  of  Agricultural  Research  \'o;.  xn,  no.  h 

Kolkwitz  (1901)  found  that  barley  grains  containing  19  to  20  per 
cent  of  moisture  respired  3.69  mgm.  of  carbon  dioxid  per  kilo  in  24 
hours  at  summer  temperature,  while  at  the  same  temperature  barley 
containing  14  to  15  per  cent  of  moisture  respired  1.4  mgm.  per  kilo, 
and  0.35  mgm.  per  kilo  when  containing  10  to  12  per  cent  of  moisture. 
White  (1909)  found  that  all  cereals  gave  off  appreciable  quantities 
of  carbon  dioxid  when  stored  in  an  air-dried  condition,  the  respiration 
of  wheat  containing  11.9  per  cent  of  moisture  being  especially  pro- 
nounced. Wheat  dried  for  eight  days  at  45°  C.  did  not  respire  a  determin- 
able quantity  of  carbon  dioxid. 

Babcock  (1912)  states  that  respiration  is  practically  suspended  in 
dry  seeds  and  spores,  and  is  most  pronounced  when  vital  processes 
are  most  active,  as  during  the  germination  of  seeds.  The  metabolic 
water  produced  as  respiration  proceeds  is  believed  to  play  an  important 
part  in  the  vital  phenomena  of  the  cells. 

Duvel  (1904),  in  studying  the  vitality  of  stored  seeds,  observed  that 
the  rate  of  respiration,  as  indicated  by  the  carbon-dioxid  content  of 
the  air  in  the  closed  container,  was  increased  on  increasing  the  moisture 
content  of  the  seed.  At  the  same  time  there  was  a  marked  diminution 
in  the  percentage  of  viable  seed. 

Qvam  {1906)  observed  an  increased  rate  of  respiration  in  barley  as 
the  percentage  of  moisture  was  increased. 

Duvel  (1909)  held  a  lot  of  corn  in  storage  in  an  elevator  bin.  The 
moisture  content  averaged  17.8  per  cent,  and  the  initial  temperature 
on  February  17,  1909,  was  36°  to  40°  F.,  which  increased  near  the  surface 
of  the  grain  to  133°  F.  on  April  27,  1909.  The  temperature  from  the 
middle  to  the  bottom  of  the  bin  was  only  about  40°  F.  A  portion  of 
the  hot  corn  from  the  top  of  the  bin  was  artificially  dried  to  an  average 
of  14.57  per  cent  of  moisture.  This  was  put  in  a  car,  and  as  a  control 
a  lot  of  the  cool,  undried  corn  from  the  same  bin  was  put  in  another 
car.  The  latter  had  an  average  moisture  content  of  17.5  per  cent. 
The  dried  corn  remained  for  37  days  in  as  good  condition  as  when  put 
in  the  car,  its  temperature  rising  from  57°  to  67°  F.,  or  a  total  increase  o£ 
10  degrees.  The  cool,  undried  corn  began  to  show  signs  of  deterioration 
in  23  days,  and  five  days  later  a  point  near  the  surface  reached  a  tem- 
perature of  122°  F.  This  indicates  the  increased  tendency  of  the  damp 
grain  to  heat  in  storage. 

Shanahan,  Leighty,  and  Boerner  {1910)  examined  cargoes  of  American 
corn  on  arrival  at  European  ports  and  observed  an  increased  tendency 
to  heat  and  "go  out  of  condition"  as  the  moisture  content  increased. 

Duvel  and  Duval  (1913)  Studied  the  temperature  changes  in  carloads 
of  corn  containing  different  percentages  of  moisture.     In  one  experiment 


Mar.  i8,  1918 


Respiration  of  Stored  Wheat 


691 


running  from  March  2  to  March  29,    191 1,  the  following  temperature 
changes  were  observed  in  cars  held  on  track  at  Baltimore,  Maryland: 


Moisture. 

Average  temperature  of  corn  when — 

Loaded. 

Unloaded. 

Per  cent. 
21.  6 
19.9 

17.4 
14.  I 

"F. 
40.  0 
40.  0 

40.  0 
40.  0 

•F. 
109.7 

41-5 
40.5 
41-3 

In  a  similar  experiment,  running  from  May  11,  191 1,  to  June  i  and  3, 
191 1,  corn  containing  16.9  per  cent  of  moisture  or  more  was  heating 
when  unloaded,  while  that  which  contained  13.9  per  cent  of  moisture 
was  still  cool.  The  extent  of  heating  bore  a  fairly  definite  relation  to 
the  shrinkage  or  loss  in  weight  of  the  grain. 

Bailey  (1917a)  reported  to  the  Second  Interstate  Cereal  Conference  in 
1916  the  results  of  storage  experiments  with  wheat  at  Duluth,  Minn. 
In  the  cool  climate  of  that  city  it  was  found  that  wheat  containing  15.5 
per  cent  of  moisture  when  put  in  a  bin  in  the  fall  kept  333  days  before 
it  developed  a  sufficiently  high  temperature  to  necessitate  turning  it, 
while  wheat  containing  16.5  per  cent  of  moisture  was  actively  heating 
in  49  days. 

The  exact  reason  for  an  accelerated  respiration  with  an  increased 
moisture  content  had  not  been  adequately  explained.  In  the  discussion 
referred  to  in  the  paper  mentioned  in  the  preceding  paragraph  an  hy- 
pothesis was  advanced  to  account  for  this  relation.  Moisture  in  grain 
may,  in  the  light  of  recent  discoveries  in  the  field  of  physical  chemistry, 
be  assumed  to  exist  as  imbibed  water  in  loose  combination  with  the 
organic  colloids.  The  organic  colloids  which  form  the  principal  con- 
stituents of  the  wheat  kernel  have  the  property  of  imbibing  considerable 
quantities  of  water  and  forming  elastic  gels.  The  gel  swells  as  the  water  ^ 
is  increased,  although  the  total  volume  of  the  dry  colloid  plus  the  added 
water  is  diminished.  The  water-imbibing  capacity  of  the  several  colloids 
varies  widely,  starch  having  an  imbibing  capacity  materially  lower  than 
that  of  wheat  gluten.  There  is  no  fixed  amount  which  a  given  dry 
colloid  will  imbibe;  thus,  gels  of  varying  viscosity  can  be  produced, 
depending  upon  the  proportion  of  water  present,  and  other  variables, 
such  as  temperature,  mineral  salts,  and  other  substances.  The  rate  of 
diffusion  in  a  gel  varies  with  the  viscosity,  as  pointed  out  by  Plimmer  ■ 
(1975,  p.  386).  In  dilute  gels  diffusion  takes  place  as  in  water,  while  in 
strong  gels  the  rate  is  slower.  It  is  probable  that  in  very  dry  grain  the 
imbibed  water  is  not  sufficient  to  produce  a  gel  in  the  endosperm  struc- 
tures.    The  colloidal  material  there  located  accordingly  does  not  have  a 


692 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  n 


continuous  structure,  and  the  possibilities  of  diffusion  are  decidedly 
reduced  under  such  conditions.  The  exact  percentage  of  moisture  below 
which  this  discontinuous  structure  exists  in  a  normal  wheat  kernel  is 
not  known;  it  probably  varies  with  the  percentage  of  gluten  in  the  grain 
since  gluten  possesses  a  greater  water-imbibing  capacity  than  starch. 
Increasing  the  moisture  content  above  the  maximum  at  which  discon- 
tinuity exists  results  in  the  formation  of  a  gel  through  which  diffusion 


AS 

OF  /'TO/SrU/PS 


Fig.  I. — Graph  showing  the  relation  of  the  moisture  content  of  wheat  to  the  rate  of  respiration. 

can  occur.  Further  increases  in  moisture  content  up  to  maximum  imbi- 
bition produce  progressively  less  viscous  gels,  and  correspondingly  in- 
crease the  possible  rate  of  diffusion.  Since  the  rate  of  respiration  in 
grain  doubtless  depends  in  part  upon  the  rate  of  diffusion  between  the 
various  kernel  structures,  it  follows  that  the  less  viscous  the  gelatinous 
material  of  which  the  cell  contents  are  composed,  the  more  rapid  the 
production  of  heat  through  respiration.  To  restate,  the  production  of 
heat  is  dependent  upon  the  activity  of  the  oxidizing  enzyms  of  the 


Mar.  i8,  1918 


Respiration  of  Stored  Wheat 


693 


kernel,  the  complex  phenomenon  being  known  as  respiration.  The  latter 
is  accelerated  by  an  increase  in  the  rate  of  diffusion,  which  in  turn  is 
dependent  upon  the  existence  of  a  gel,  and  the  relative  viscosity  of  that 
gel.  For  these  reasons  the  moisture  content  of  sound  grain  determines 
to  a  considerable  extent  the  rate  of  respiration  and  consequent  liability 
of  heating  when  bulk  grain  is  stored. 

To  determine  the  relation  of  moisture  content  to  respiration  in  stored 
wheat,  a  large  sample  of  spring-sown  Haynes  Bluestem  wheat  known  as 
Minnesota  169  was  obtained.  The  weight  per  bushel  of  the  sample  was 
57K  pounds  (26.08  kgm.) ;  the  w^eight  per  i  ,000  average  kernels,  24.62  gm. ; 
and  it  contained  2.21  per  cent  of  nitrogen  on  the  dry  basis.  It  was  then 
divided  into  several  portions,  and  each  portion  was  brought  to  a  different 
moisture  content,  the  percentages  of  moisture  ranging  from  12.50  to 
17.07  per  cent.  The  quantity  of  carbon  dioxid  respired  per  24  hours  by 
each  100  gm.  of  dry  matter  is  given  in  Table  I  and  is  shown  graphically 
in  figure  i.  The  rate  of  increase  in  respiration  is  fairly  gradual  from 
12.50  to  14.78  per  cent,  but  after  the  latter  percentage  is  exceeded  the 
rate  is  markedly  accelerated.  The  break  in  the  curve  occurs  when  the 
moisture  content  slightly  exceeds  14.5  per  cent,  and  it  is  probable  that 
this  represents  about  the  maximum  percentage  of  moisture  that  this  class 
of  wheat  may  safely  contain  without  danger  of  heating  when  stored  in 
bulk. 


Table  I. — Respiration  of  Haynes  Bluestem  (Minnesota  i6g)  -wheat,  <^  incubated  at  j^ -8°  C 

for  four  days 


Moisture. 

Carbon  dioxid 

respired  per  24  hours 

for  each  100  gm.  of 

dry  matter. 

Moisture. 

Carbon  dioxid 

respired  per  24  hours 

for  each  100  gm.  of 

dry  matter. 

Per  cent. 
12.  50 

13-93 
14.78 
15.42 

\IC7n. 

0-54 

.65 

.86 

I.  62 

Per  cent. 
16.08 
16.65 
17.07 

Mgm. 
2.88 
6.86 

II.  72 

"Weight  per  bushel  of  sample,  57^4  pounds, 
basis,  2.21  percent. 


Weight  per  1,000  kernels,  24.62  gm.  Nitrogen  on  dry 


The  acceleration  of  respiration  as  the  moisture  content  increases  is 
shown  in  Table  II.  These  data  are  based  upon  the  estimated  respiration 
values  at  even  percentages  of  moisture,  and  the  computed  increase  in 
respired  carbon  dioxid  for  each  increase  of  i  per  cent  of  moisture.     In 

computing  these  data  the  following  formula  was  employed :  —^ — ^^^ 

in  which  Km  represents  the  respiration  value  at  a  particular  percentage 
of  moisture,  and  Km  _  i  represents  the  respiration  value  for  the  same  wheat 
containing  i  per  cent  of  moisture  less  than  Km.  It  is  evident  that  the 
acceleration  between  12  and  14  per  cent  of  moisture  is  very  gradual,  while 
it  increases  markedly  after  14  per  cent  is  exceeded. 


694 

Journal  of  Ag 

ricultural  Research 

Vol.  XII,  No.  II 

Table  II. 

— Acceleration  of  the  rate  of  respiration  of  hard  spring  wheat  with  increasing 
moisture  content 

Formula. 

Acceleration  at  following  percentages  of  moisture. 

12  to  13 

13  to  14 

14  to  IS 

IS  to  16 

16  to  17 

Km  — Km  — 

t                                

o.  i6 

0.17 

0.66 

I.  41 

Km— 1 

3.  02 

/^(?/^. 


24 


RELATION  OF  THE  CON- 
SISTENCY OF  THE 
WHEAT  KERNEL  TO 
THE  RATE  OF  RESPI- 
RATION 

Reference  has  already 
been  made  to  the  differ- 
ence in  the  relative 
water-imbibing  capacity 
of  the  various  organic  col- 
loids of  the  wheat  kernel. 
Since  starch  and  gluten 
constitute  a  large  propor- 
tion of  the  endosperm, 
their  differences  in  this 
regard  are  of  principal 
interest.  Simple  tests  in- 
dicate that  the  water-im- 
bibing capacity  of  gluten 
is  materially  greater  than 
that  of  starch.  In  con- 
sequence it  follov\rs  that, 
as  a  general  rule,  those 
kernels  which  contain  a 
high  percentage  of  gluten 
will,  at  any  particular 
moisture  content,  be  more 
viscous  than  will  kernels 
containing  a  lower  per- 
centage of  gluten. 

The  gluten  content  is  also  related  to  the  relative  consistency  or  hard- 
ness of  the  wheat  berry.  In  general,  the  hard,  vitreous  grains  contain 
a  higher  percentage  of  gluten  than  do  the  soft,  starchy  grains.  Accord- 
ingly we  may  expect  exactly  what  we  find — viz,  that  the  soft  wheats 
are  "tougher"  and  of  a  lower  viscosity  at  any  given  moisture  content 
(within  the  limits  found  in  ordinary  commercial  grain)  than  the  hard, 
vitreous  wheats. 


/2 


/3        Af         /s         /e        /7 


Fig.  2.— Graphs  showing  the  comparative  rate  of  respiration 
hard  spring,  soft  red  winter,  and  soft  white  winter  wheat. 


Mar.  i8,  19 1  £ 


Respiration  of  Stored  Wheat 


695 


It  is  commonly  recognized  in  the  grain  trade  that  the  keeping  quali- 
ties of  soft  wheats  are  inferior  to  those  of  hard  wheat.  Because  of  the 
relation  of  respiratory  activity  to  rate  of  diffusion,  it  should  follow  that 
with  the  same  moisture  content  respiration  would  proceed  more  rapidly 
in  a  soft  than  in  a  hard  or  vitreous  kernel.  A  sample  of  soft  red  winter 
wheat  of  the  Fultz  variety  was  obtained  from  the  Experiment  Station 
at  Columbia,  Mo.,  and  another  of  white  winter  wheat  from  Grand  Blanc, 
Mich.  The  rate  of  respiration  in  these  soft  wheats  at  different  moisture 
contents  was  studied,  and  it  was  found  that,  except  at  the  lower  per- 
centages of  moisture,  the  rate  of  respiration  was  higher  in  the  soft  red 
winter  wheat  than  it  was  in  the  hard  spring  wheat,  and  still  higher  in 
the  white  winter  wheat.  These  data  are  given  in  Tables,  III,  IV,  and 
V,  and  graphically  in  figure  2.  As  shown  by  the  graph,  the  curves  tend 
to  converge  at  about  1 2  per  cent  of  moisture,  indicating  that  at  less  than 
this  moisture  content  the  discontinuity  of  endosperm  structure  referred 
to  above  may  exist  in  sound  wheats  and  respiration  proceed  at  the  ex- 
pense of  substances  in  the  germ  rather  than  by  oxidation  of  materials 
which  diffuse  to  it  from  the  endosperm. 

It  may  further  be  seen  that  the  quantity  of  heat  evolved  by  hard  spring 
wheat  containing  14.5  per  cent  of  moisture,  as  evidenced  by  the  rate  of 
respiration  at  that  moisture  content,  was  evolved  by  these  soft  wheat 
samples  when  they  contained  about  13.6  to  13.8  per  cent  of  moisture. 
This  is  of  interest,  in  view  of  the  moisture  limits  prescribed  in  the  United 
States  Grain  Standards  for  wheat,  which  are  I4>^  per  cent  for  No.  2  hard 
spring,  and  1 3  per  cent  for  No.  2  soft  red  winter,  and  the  same  for  com- 
mon white  and  white  club  wheat. 

Table  III. — Respiration  of  soft  red  winter  wheat  (^  from  Missouri,  incubated  at  j/.8'°C. 

for  four  days 


Moisture. 

Carbon  dioxid 

respired  per  24  hours 

for  each  100  gm.  of 

dry  matter. 

Moisture. 

Carbon  dioxid 

respired  per  24  hours 

for  each  100  gm.  01 

dry  Eoatter. 

Per  cent. 
13-07 

14.70 

Mgm. 

0.  65 

.80 

•95 

Per  cent. 
15-45 
16.37 
17.  40 

Mgm. 
2.  00 
5.06 

22.03 

o  Weight  per  bushel  of  sample,  61  pouadi.    Weight  per  1,000  Iceruels,  29.97  gm-    Nitrogen  on  dry  basis. 
1.54  per  cent. 


696 


Journal  of  Agricultural  Research 


Vol.  XII,  Mo.  II 


Table  IV. — Respiration  of  white  winter  wheat  "■  from  Michigan,  incubated  at  27-8°C. 

for  four  days 


Moisture. 

Carbon  dioxid 

respired  per  24  hours 

for  each  100  gm.  of 

dry  matter. 

Moisture. 

Carbon  dioxid 

respired  per  24  hours 

for  each  100  jrm.  of 

dry  matter. 

Per  cent. 
11.94 
13.04 
14.32 

Mgm. 

0.48 

.60 

.89 

Per  cent. 

15-57 
16.83 

Mgm. 
3.  20 

22.  77 

<J  Weight  per  bushel  of  sample,  59  pounds.     Weight  per  1,000  kernels,  37.73  gm.    Nitrogen  on  dry  basis, 
1.53  per  cent. 

Table  V. — Interpolated  quantity  of  carbon  dioxid  respired  per  unit  of  time  and  material, 
at  even  percentages  of  m,oisture 


Class  of  wheat. 


Hard  spring . . . . 
Soft  red  winter. 
White  winter.  . . 


Carbon  dioxid  respired  per  24  hours  for  each  100  gm.  of  dry  matter. 


per  cent 
moisture. 


Mgm. 
O.  CO 


49 


13 
per  cent 
mojsture. 


Mgm. 
C.  K& 

■  h 

.  60 


14 
per  cent 
moisture. 


Mgm. 
0.68 
.81 
■83 


IS 
per  cent 
moisture. 


Mgm. 
I-  13 
1-37 
4-  15 


16 
per  cent 
moisture. 


Algm.. 
2.  72 
3-84 
9-85 


17 
per  cent 
moisture. 


Mgm,. 

IO-73 

15-51 
25.18 


RELATION  OF  THE  RELATIVE  PLUMPNESS  OF  THE  WHEAT  KERNEL 
TO  THE  RATE  OF  RESPIRATION 

It  is  generally  recognized  that  the  velocity  of  enzym  action  conforms 
quite  closely  to  the  law  of  mass  action.  Thus  any  condition  which 
affects  the  quantity  of  either  the  substrate  or  the  enzym  will  cause 
variations  in  the  rate  of  the  reaction.  Since  respiration  is  occasioned 
by  enzyms,  the  rate  of  respiration  of  the  wheat  kernel  should  vary  with 
these  conditions. 

A  shriveled  condition  of  the  wheat  kernel  is  due  generally  to  factors 
operating  during  the  later  stages  of  kernel  development.  The  trans- 
location of  reserves  to  the  kernel  is  interfered  with  by  rust,  drouth, 
desiccation  by  hot  winds,  or  some  other  agency;  and  an  incomplete 
filling  of  the  endosperm  results.  According  to  Brenchley  {igog),  the 
germ  portion  of  the  kernel  is  developed  earlier  than  the  endosperm, 
and  tends  to  escape  injury  from  the  agency  causing  shrunkenness  of 
the  endosperm  more  than  does  the  latter.  The  diminished  size  and 
weight  of  the  shriveled  kernel  is  therefore  due  principally  to  the  decreased 
quantity  of  endosperm. 

The  enzymic  activities  of  the  kernel  seem  to  be  mainly  invested  in  the 
embryo.  This  was  discussed  in  one  of  the  foregoing  paragraphs.  In 
the  embryo  of  the  shriveled  wheat  berry  the  enzyms  are  probably  repre- 
sented practically  as  they  are  in  the  normal  kernel.  In  instances  where 
the  average  weight  of  the  individual  kernel  is  only  about  half  the  normal, 


Mar.  i8,  1918 


Respiration  of  Stored  Wheat 


697 


it  follows  that  there  is  approximately  twice  the  enzymic  activity  per 
unit  of  mass  than  that  shown  by  normal  wheat.  The  normal  spring 
wheat  used  by  the  writers  weighed  24.62  gm,  per  1,000  average  kernels. 
A  shriveled  sample  of  the  same  type  of  wheat  was  obtained  which  weighed 
11.73  g™-  per  1,000  kernels,  or  less  than  half  the  weight  of  the  normal. 
The  rate  of  respiration  of  these  two  lots  was  compared,  and,  as  shown 
in  figure  3,  the  respiratory  activity  of  the  shriveled  sample  decidedly 
exceeded  that  of  the  normal,  or  plump  wheat.  Thus  the  quantity  of 
carbon  dioxid  respired  by  the  latter  when  it  contained  14.5  per  cent  of 
moisture  was  respired  by  the  shriveled  wheat  used  when  it  contained 
only  12.8  per  cent  of  moisture.  The  curves  tend  to  converge  at  moisture 
contents  slightly  below  12  per  cent. 

The  respiration  data  for  the  shriveled  sample  are  given  in  Table  VI, 
while  in  Table  VII  is  shown  the  interpolated  quantity  of  carbon  dioxid 
respired  by  the  normal  and  shriveled  wheats  at  even  percentages  of 
moisture. 


Table  VI. — Respiration  of  shriveled  spring  wheat, '^  incubated  at  27-8°  C .  for  four  days 


Moisture. 

Carbon  dioxid 

respired  per  24  hours 

for  each  100  gm.  of 

dry  matter. 

Moisture. 

Carbon  dioxid 

respired  per  24  hours 

for  each  100  gm.  of 

dry  matter. 

Per  cent. 
12.68 

13-  19 

14.  29 

15-30 

Mgm. 

0.75 

.94 

1.38 

3.02 

Per  cent. 
15.68 
16.  09 

16.44 
16.80 

Mgm. 
4-  50 
10.  51 
16.  92 
21.  65 

a  Weight  per  bushel  of  sample,  47M  pounds.     Weight  per  1,000  kernels,  11.73  gm.    Nitrogen   on  dry 
basis,  2.03  per  cent. 

Table  VII. — Interpolated  quantity  of  carbon  dioxid  respired  per  unit  of  time  and  material 
at  even  percentages  of  moisture 


Carbon  dioxid  respired  per  24  hours  for  each  100  gm.  of  dry  matter. 

Class  of  wheat. 

12 
per  cent 
moisture. 

13 
per  cent 
moisture. 

14 
per  cent 
moisture. 

IS 
per  cent 
moisture. 

16 
per  cent 
moisture. 

17 
per  cent 
moisture. 

Plump  spring  wheat          

Mgm. 
0.  50 
•65 

Mgm. 
0.58 
.88 

Mgm. 

0.  68 

1.  26 

Mgm.. 
I-  13 
2.54 

Mgm. 
2.  72 
9.41 

Mgm. 
10.93 
22.  65 

Shriveled  spring  wheat 

SOUNDNESS  OF  THE  WHEAT  KERNEL  IN  ITS  RELATION  TO  THE  RATE 

OF  RESPIRATION. 

A  form  of  unsoundness  recognized  as  such  by  the  grain  trade  and 
frequently  occurring  in  spring  wheat  is  the  frosted  condition.  This 
results  from  the  freezing  of  the  plants  before  the  grain  is  matured  and 
desiccated.    The  plants  usually  thaw  later,  and,  vv^hile  the  protoplasm 


698 


Journal  of  Agricultural  Research 


Vol.  XII.  No.  II 


is  killed  or  disorganized  and  its  synthetic  activities  reduced  or  destroyed, 
certain  hydrolytic  enzyms  are  activated,  and  hydrolysis  or  splitting  of 
certain  of  the  kernel  constituents  ensues.  As  a  result,  there  is  an  accumu- 
lation of  the  split  products  of  starch  and  proteins,  particularly  dextrose 
and  amino  acids.  It  is  probable  that  the  extent  of  starch  hydrolysis 
by  amylases  depends  in  large  part  upon  the  percentage  of  moisture  in  the 
grain  after  it  thaws  out.  If  the  kernels  are  nearly  dry,  less  change  \vill 
occur  than  if  the  kernels  contain  considerable  moisture. 

The  dextrose  which  thus  ac- 
cumulates in  the  kernel  is  pre- 
sumably available  as  substrate 
for  the  respiratory  enzyms. 
In  accordance  with  the  law 
of  mass  action,  a  greater  con- 
centration of  substrate  should 
accelerate  the  rate  of  respira- 
tion. There  is  an  additional 
factor  in  the  case  of  frosted 
wheat  that  would  also  tend  to 
result  in  an  increased  rate  of 
respiration  at  any  particular 
percentage  of  moisture.  The 
hydrolysis  of  the  gluten  sub- 
sequent to  thawing  results  in 
products  having  a  materially 
lower  water-imbibing  capacity 
than  the  normal  gluten.  In 
fact,  the  amino  acids  formed 
are  not  colloids,  and  form  true 
solutions.  Consequently  the 
relative  viscosity  of  frosted 
grain  at  any  moisture  content 
will  be  less  than  in  normal 
grain,  the  difiference  depending 
upon  the  extent  of  hydrolysis. 
To  ascertain  the  efifect  of 
frosting  upon  the  rate  of  respiration,  two  samples  of  commercial 
wheat  containing  frosted  kernels  were  secured.  These  were  marked 
"moderately  frosted"  and  "badly  frosted,"  respectively.  The 
respiratory  activity  of  these  frosted  samples  was  determined  with 
five  different  percentages  of  moisture  present.  Tables  VIII  and  IX 
show  the  respiration  data  of  these  two  lots,  while  in  Table  X  are 
given  the  interpolated  values  at  even  percentages  of  moisture  in  com- 
parison with  sound  spring  wheat.  The  same  data  are  shown  graph- 
ically in  figure  4.     There  is  a  slight  overlapping  of  the  curves  for  the 


I 


I 

\ 


to 

Q      /i?         /S         /^         /S         /&         /y' 

Fig.  3. — Graphs  showing  the  rate  of  respiration  of  shriveled 
■wheat  and  of  plump  wheat  of  the  same  class. 


[■ 

1 

J 

1 

^1 

r.l 

if 

fi 

1 

'  3 

/ 

1 

/ 

y 

"^y 

A 

y 

Mar.  i8,  1918 


Respiration  of  Stored  Wheat 


699 


moderately  and  badly  frosted  samples  between  14.5  and  15.5  per  cent  of 
moisture,  the  reason  for  which  is  not  clear.  The  discrepancy  is  not 
great,  and  the  similarity  of  the  curves  indicates  that  possibly  our  judg- 
ment was  at  fault  in  describing  the  two  lots  of  frosted  wheat  as  "moder- 
ately" and  "badly"  damaged.     The  decided  differences  in  the  rate  of 


/P 


/O 


\ 

w 

•5^ 


li 

/ 

^1 

/ 

if. 

1 

ii 

1  i 

»/ 

/ 

J 

1 

,(C 

f/ 

/ 

/ 

1 

V 

/ 

.y^j 

/ 

/ 

^ 

'^ 

y 

r 

**' 

— 

y4? 


/C? 


/-S*  /^  /6- 


/T" 


Fig.  4. — Graphs  showing  the  rate  of  respiration  of  frosted  wheat  and  sound  wheat  of  the  same  class. 

respiration  of  frosted  and  of  normal  wheat  containing  up  to  16.5  per 
cent  of  moisture  substantiates  what  had  previously  been  empirically 
observed  by  the  senior  author  (Bailey,  1917b) — viz,  that  frosted  wheat 
tends  to  heat  more  readily  when  stored  than  does  sound  wheat  contain- 
ing the  same  percentage  of  moisture. 


700 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  ir 


Table   VIII. — Respiration   of  moderately  frosted  wlieatfl  incubated  at  J/.8°   C.  for 

four  days 


Moisture. 

Carbon  dioxid 

respired  per  24  hours 

for  each  100  gm.  of 

dry  matter. 

Moisture. 

Carbon  dioxid 

respired  per  24  hours 

for  each  100  gm.  of 

dry  matter. 

Per  cent. 
12.32 

14.44 

Mgm. 
0.74 
I.  04 
1.89 

Per  cent. 

14-95 

15-42 

Mgm. 
3-75 

5-21 

oWeight  per  bushel  of  sample,  58  pounds.    Weight  per  i,ooo  kernels,  36.94  gm.     Nitrogen  on  dry  basis, 
2.02  per  cent. 

Table  IX. — Respiration  of  badly  frosted  wheatfl  incubated  at  37.8°  C.  for  four  days 


Moisture. 

Carbon  dioxid 

respired  per  24  hours 

for  each  100  gm.  of 

dry  matter. 

Moisture. 

Carbon  dioxid 

respired  per  24  hours 

for  each  100  gm.  of 

dry  matter. 

Per  cent. 

13-79 
14.67 

15-74 

Mgm. 
1.63 
2.  64 
5-24 

Per  cent. 

16.81 

16.95 

Mgm. 

9.82 

II.  40 

o  Weight  per  bushel  of  sample,  53  pounds.     Weight  per  1,000  kernels,  22.52  gm.    Nitrogen  on  dry  basis, 
2.10  per  cent. 

Table  X. — Interpolated  quantity  of  carbon  dioxid  respired  per  unit  of  time  and  m,ate- 
rial,  at  even  percentages  of  moisture 


Class  of  wheat. 


Carbon  dioxid  respired  per  24  hours  for  each  100  gm.  of  dry  matter. 


per  cent 
moisture. 


13 
per  cent 
moisture. 


14 
per  cent 
moisture. 


per  cent 
moisture. 


16 
per  cent 
moisture. 


17 
per  cent 
moisture. 


Sound  spring  wheat 
Moderately  frosted. , 
Badly  frosted 


Mgm. 

0.  50 

-65 

1.  12 


Mgm,. 
0.58 

•94 
I.  20 


Mgm. 
0.68 
1.52 
1.87 


Mgm.. 
1-13 
3-90 
3-46 


Mgm. 
2.  72 


Mgm. 
10.93 


6-35 


11.97 


Table  XI. — Respiration  of  several  samples  of  frosted  wheat  from,  car  lots,  and  containing 
varying  percentages  of  moisture.     Incubated  at  37.8°  C.  for  four  days 


Laboratory  No. 

Moisture. 

Weight  per 
bushel. 

Weight  per 

1,000 

kernels. 

Nitrogen  on 
dry  basis. 

Carbon 
dioxid  re- 
spired per 
24  hours  for 
each  100  gm. 
of  dry 
matter. 

Interpolated 

respiration 

of  sound 

wheat  of 

same  water 

content. 

G122 

Per  cent. 
14.30 
14.82 
16.16 
16.  19 

Pounds. 
60K 
61 

56K 

57 

Gm. 
28.36 
28.  64 
23-56 
28.  88 

Per  cent. 
2.  23 
2.  20 
2.  27 
2.  00 

Mgm. 
I.  71 

2-34 
8.71 
8.24 

Mgm. 

1.30 

1-73 
6  86 

Gl2^ 

GII8 

GI02 

7-15 

Several  wheat  samples  taken  from  car  lots  by  samplers  of  the  State 
Grain  Inspection  Department  were  frosted,  and  the  respiratory  activity 


Mar.  i8,  1918 


Respiration  of  Stored  Wheat 


701 


of  four  of  these  was  measured.  In  every  case  the  rate  of  respiration 
was  greater  than  in  sound  wheat  of  the  same  moisture  content  which 
was  taken  from  car  lots  at  the  same  time.  The  tests  of  these  frosted 
samples  are  shown  in  Table  XI,  the  last  column  of  which  shows  the 
interpolated  respiration  data  of  sound  wheat. 

THE  PERIOD  OF  DAMPNESS  AS  INFLUENCING  THE  RATE  OF  RESPI- 
RATION OF  WHEAT 

When  damp  wheats  are  stored,  the  excess  moisture  which  is  responsible 
for  their  dampness  has  been  present  for  varying  lengths  of  time.  This 
is  called  for  the  purposes  of  this  discussion  the  "period  of  dampness." 

In  connection  with  these  studies  a  series  of  wheat  samples  was  taken 
from  car  lots  by  the  State  Grain  Inspection  Department  between  the  15th 
and  27th  of  March,  which  contained  varying  percentages  of  moisture. 
The  respiratory  activity  of  each  of  these  was  determined,  and  on  plot- 
ting a  curve  it  was  observed  that  these  samples  gave  materially  higher 
values  than  the  samples  which  had  been  dampened  and  allowed  to 
stand  for  three  days  before  measuring  the  rate  of  respiration.  The 
curves  tend  toward  convergence  at  a  moisture  content  of  12  per  cent. 
These  data  are  given  in  Tables  XII  and  XIII  and  are  shown  graphically 
in  figure  5. 

Table  XII. — Respiration  of  natural  hard  spring  wheat  collected  from  car  lots,  and 
incubated  at  ^7.8°  C.  for  four  days 


Laboratory  No. 


G128 
G127 
G124 
G120 
G104 
G103 
G107 
G109 


Moisture. 


Per  cent. 
12.47 
13.  II 
14.70 
15-51 
15-73 
16.  00 

16.53 
16.  90 


Weight  per 
bushel. 


Pounds. 
62 

63 

58i 

52 

59 

59 

56 

51 


Weight  per 
1,000  kernels. 


Gm. 
30-56 
32-94 
24.  60 
26.  92 
28.64 
25.72 
22.88 


Nitrogen  in 
dry  matter. 


Per  cent. 
2.  14 
2.  16 
2. 
2. 
2. 
2. 


05 
31 

47 
37 


2.25 


Carbon 
dioxid  re- 
spired per 
24  hours  for 
each  100  gm. 
of  dry  matter. 


Mgm. 
O.  61 

•75 
1.49 
3.26 
3-94 
5- 69 
10.  65 
13.86 


Table  XIII. — Interpolated  quantity  of  carbon  dioxid  respired  per  unit  of  time  and  mate- 
rial, at  even  percentages  of  moisture 


Carbon  dioxid  respired  per  24  hours  for  each  100  gm.  of  dry  matter. 

Class  of  wheat. 

13                         13 

per  cent      per  cent 
moistiu-e.   moisture. 

14 
per  cent 
moisture. 

IS 
per  cent 
moisture. 

16 
per  cent 
moisture. 

17 
per  cent 
moisture. 

Freshly  dampened   wheat .... 
Natural  wheat       

Mgm. 
0.  50 
•51 

Mgm. 
0.58 

•73 

Mgm. 
0.68 

I-  15 

Mgm. 

I-  13 
2.  14 

Mgm. 
2.  72 
5-69 

Mgm. 
IO-73 
15-03 

3832G°— 18 2 


702 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  II 


Since  the  principal  difference  between  the  samples  dampened  by  the 
writers  and  the  damp  grain  obtained  from  the  freight  cars  was  the  length 
of  the  period  of  dampness,  an  experiment  was  conducted  to  ascertain 
the  extent  to  which  that  variable  affected  the  rate  of  respiration.  Two 
samples  of  Bluestem  wheat  that  had  been  dampened  some  time  previous 


Af&/*f. 


/v?  M  /s  /e 


/7 


Fig.  s.— Graphs  showing  the  comparative  respiratory  activity  of  naturally  damp  wheats  and  of  wheats 
dampened  in  the  laboratory  three  days  before  they  were  incubated. 

and  stored  at  a  temperature  of  about  25°  C.  in  the  laboratory  vault 
were  incubated  in  the  usual  manner  at  37.8°  C.  with  the  results  shown 
in  Table  XIV.  The  quantity  of  carbon  dioxid  respired  by  the  sample 
containing  15.21  per  cent  of  moisture,  which  was  stored  for  55  days  after 
it  was  dampened,  was  about  four  times  as  great  as  that  respired  by 
freshly  dampened  wheat  of  the  same  moisture  content,  while  that  from 


Mar.  18,  1918 


Respiration  of  Stored  Wheat 


703 


the  sample  containing  15.71  per  cent  of  moisture,  which  was  stored  for 
108  days,  was  about  eight  times  as  great  as  for  freshly  dampened  wheat 
containing  that  percentage  of  moisture. 

Table  XIV. — Respiration  of  dampened  wheat  after  storage  at  about  25°  C.     Incubated 

at  37.8°  C.  for  four  days 


Moisture per  cent. . 

Number  of  days  stored 

Carbon  dioxid  respired  per  100  gra.  of  dry  matter  in  each  24 
hours mgm .  . 

Carbon  dioxid  respired  per  100  gm.  of  dry  matter  in  each  24  hours, 
of  same  lot  of  wheat  4  days  after  dampening mgra . . 

Carbon  dioxid  respired  by  naturally  damp  wheat  from  car  lots, 
of  same  moisture  content mgm .  . 


Lot  B. 


15-72 
108 


17.  00 

2.  17 

3-85 


Table  XIV  also  shows  a  decided  difference  in  the  comparative  rate  of 
respiration  of  wheat  stored  for  a  time  in  a  warm  room  and  that  obtained 
from  freight  cars.  The  latter  had  no  doubt  been  damp  for  a  longer  time 
than  that  which  was  dampened  and  held  in  the  room.  These  data 
suggest  that  not  only  does  the  period  of  dampness  affect  respiration, 
but  the  conditions  of  storage  may  have  an  equally  important  effect. 
The  grain  taken  from  cars  was  cold  at  the  time,  and  had  probably  been 
exposed  to  the  cold  atmospheric  conditions  of  the  preceding  winter 
months.  If,  as  postulated  by  the  writers,  dextrose  tends  to  accumulate 
in  the  stored  damp  grain  the  rate  of  accumulation  would  depend  upon  the 
temperature  of  the  grain  as  well  as  upon  the  time  elapsed.  Accord- 
ingly there  may  have  been  more  substrate  (dextrose)  for  the  respiratory 
enzyms  in  the  grain  which  had  been  stored  in  the  warm  room  than 
there  was  in  that  stored  out  of  doors  during  the  winter  months.  A 
series  of  experiments  with  both  the  period  of  dampness  and  the  tempera- 
ture of  storage  as  variables  have  accordingly  been  begun,  and  the  results 
will  be  reported  later  in  another  publication. 

INFLUENCE   OF  TEMPERATURE   ON  THE   RESPIRATION   OF   STORED 

WHEAT 

Pfeffer  (1878)  observed  that  the  intensity  of  respiration  increases 
with  the  temperature  until  the  latter  begins  to  injure  all  the  vital  proc- 
esses. Hoff  {1896,  p,  12^)  stated  that  the  rate  of  respiration  increases 
two  or  three  times  for  each  lo-degree  rise  in  temperature  in  accordance 
•vvith  the  usual  rule  of  chemical  reactions.  Ziegenbein  {i8gj)  found 
that  temperatures  above  45°  C.  were  injurious 

Clausen  {i8po)  studied  the  respiration  of  germinating  wheat  at  differ- 
ent temperatures  and  found  the  optimum  to  be  about  40°  C.  The  rate  of 
respiration  was  2.86  times  as  great  at  10°  as  at  0°  and  1.09  times  as  great 


704  Journal  of  Agricultural  Research  \o\.  xii,  no.  h 

at  40°  as  at  30°.  The  average  increase  between  0°  and  40°  was  2.71 
times  for  each  lo-degree  increase  in  temperature. 

Qvam  (1906)  found  the  rate  of  respiration  increased  up  to  at  least  45°  C, 
which  was  apparently  the  highest  temperature  at  which  observations 
were  made.  The  grain  was  very  moist,  100  gm.  of  water  having  been 
added  to  200  gm.  of  grain. 

Matthaei  (1905)  investigated  the  respiration  of  cherry-laurel  leaves  at 
different  temperatures  and  reported  an  increasing  rate  of  respiration 
between  5.8°  and  33.1°  C.  At  5.8°  2  gm.  of  green  leaves  respired  o.i 
mgm.  of  carbon  dioxid  per  hour,  and  at  33.1°  the  rate  was  1.35  mgm.  per 
hour. 

The  experiments  of  Duvel  and  Duval  (1913)  with  shelled  corn  indicate 
the  relation  of  air  temperatures  to  the  heating  of  such  material.  Shelled 
corn  in  transit  and  on  track  containing  16.9  per  cent  of  moisture  began 
to  heat  and  go  out  of  condition  between  May  1 1  and  June  3,  191 1,  while 
in  the  period  from  December  24,  1910,  to  January  20,  191 1,  corn  con- 
taining 22  per  cent  of  moisture  gained  only  a  few  degrees  in  temperature. 
The  purely  physical  factor  of  heat  loss  into  the  cold  winter  atmosphere, 
of  course,  served  to  reduce  the  rate  of  rise  in  temperature,  but  there  was 
probably  a  diminished  rate  of  evolution  of  heat  as  well. 

Attention  has  been  called  by  Bailey  {1917a),  to  the  influence  of  at- 
mospheric temperatures  upon  the  rate  of  heating  of  damp  wheat.  A 
lot  of  wheat  containing  16.5  per  cent  of  moisture  required  but  11  days 
to  increase  in  temperature  from  70°  to  80°  F.,  when  the  mean  air  temper- 
ature was  62.1°  F.,  while  later  in  the  year,  when  the  mean  air  tempera- 
ture was  44.3°  F.,  another  lot  of  wheat  containing  the  same  percentage 
of  moisture  was  stored  49  days  before  its  temperature  increased  to  the 
same  extent. 

Another  comparison  of  the  influence  of  temperature  on  the  rate  of 
heating  in  storage  was  afforded  by  two  lots  of  wheat  put  into  bins  at  about 
the  same  time  and  containing  nearly  the  same  percentage  of  moisture. 
The  initial  temperature  of  one  lot  was  74°  F.,  and  of  the  other  70°.  The 
latter  required  over  five  times  as  long  to  reach  a  temperature  of  80°  as 
did  the  former,  owing  to  the  slow  increase  in  temperature  at  the  outset 
as  contrasted  with  the  rapid  rate  of  increase  as  the  temperature  mounted 
higher. 

To  ascertain  the  relation  of  temperature  to  the  rate  of  respiration  in 
stored  grain,  a  large  sample  of  Minnesota  i69Bluestem  wheat  was  damp- 
ened until  it  contained  14.96  per  cent  of  moisture.  Aliquots  of  this 
sample  were  sealed  in  glass  jars  and  kept  in  a  refrigerator  until  they  were 
used.  This  was  done  to  minimize  enzymic  changes  in  the  grain.  The 
necessary  quantities  were  drawn  from  the  refrigerator  for  incubation  at 
the  several  temperatures. 

The  lowest  temperature  at  which  observations  were  made  was  4°C., 
and  since  the  increase  in  rate  of  respiration  between  4°  and  25°  was  rela- 


Mar.  i8,  1918 


Respiration  of  Stored  Wheat 


705 


lively  small,  no  intermediate  temperatures  were  employed.  Regular 
increases  in  the  temperature  of  the  thermostat  by  lo-degree  intervals 
were  then  made  until  the  respired  carbon  dioxid  showed  a  marked 
diminution. 

The  data  in  Table  XV  and  the  graph  in  figure  6  show  that  the  rate  of 
respiration  increased  to  a  maximum  at  55°  C.  This  is  therefore  the  tem- 
perature at  which  the  most  rapid  evolution  of  heat  would  occur.  A  dis- 
coloration of  the  seed  coat  of  the  wheat  kernel  begins  to  show  on  some 
kernels  at  about  35°  C,  while  at  55°  the  whole  mass  is  of  a  mahogany 
color.     At   65°   the   respiratory   enzyms   have   been  '  partially   but   not 


J 

\ 

/ 

\ 

/ 

\ 

/ 

\ 

/ 

\ 

\ 

/ 

f 

0 

..    .     « 

1 

/s 


2'S       35        '^S       .5i5 


es 


7^ 


Fig.  6. — Graph  showing  the  relation  of  temperature  to  the  rate  of  respiration. 

wholly  inactivated,  while  at  75°  this  inactivation  has    proceeded    still 
further,  and  some  roasting  of  the  grain  has  occurred. 

Table  XV. — Respiration  of  hard  spring  wheat  at  different  temperatures 


■    Carbon  dioxid 

Carbon  dioxid 

Temperature. 

respired  per  24  hours 

Temperature. 

respired  per  24  hours 

for  each  100  gm.  of 

for  each  icxj  gm.  of 

dry  matter. 

dry  matter. 

•c. 

Mgm. 

"C. 

Mgni. 

4 

0.  24 

55 

31-73 

25 

•45 

65 

15-71 

35 

1.30 

75 

a  10.  28 

45 

6.61 

"A  part  of  this  carbon  dioxid  may  have  resulted  from  roasting  the  grain. 


7o6 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  II 


The  proportional  change  in  respiration  for  each  lo-degree  rise  in  tem- 
perature is  shown  in  Table  XVI.     The  data  in  this  table  were  obtained 

by  employing  the  conventional  formula  -^  in  which  Vt  represents  the 

rate  of  respiration  at  the  specified  temperature,  and  F^+^o  represents 
the  rate  at  a  temperature  lo  degrees  higher.  The  values  at  5°  and 
15°  were  computed  by  integrating  the  actual  data  obtained  at  4°  to 
25°  and  35°  C. 

Table  XVI. — Acceleration  of  respiration  for  each  lo-degree  increase  in  temperature 


t. 

Vt 

■ 
t. 

Vt 

°  c. 

5 
25 

I.  16 

1-55 
2.89 

°  C. 

35 
45 
55 

S.oS 
4.  So 

•49 

INFLUENCE  OF  ACCUMULATED  CARBON  DIOXID  UPON  RESPIRATION 

Muntz  {1881)  observed  that  ten  times  as  much  carbon  dioxid  was 
respired  when  grain  had  access  to  free  air  as  when  sealed  air-tight  and 
that  the  yield  of  carbon  dioxid  was  much  greater  after  the  moisture 
exceeded  13  to  14  per  cent. 

Mangin  (i8g6)  found  the  evolution  of  carbon  dioxid  and  absorption 

of  O2  to  be  reduced  when  germinating  seeds  were  put  in  air  containing 

CO 
up  to  5  per  cent  of  carbon  dioxid.     The  -^r  ratio  was  increased,  indi- 

eating  that  absorption  of  oxygen  was  diminished  more  than  the  evolu- 
tion of  carbon  dioxid. 

Duvel  {1904),  Babcock  (191 2),  Barnes  and  Grove  (19 16),  and  others 
have  called  attention  to  the  reduced  vitality  and  germination  of  seeds 
stored  in  carbon  dioxid,  or  in  tight  containers  in  which  the  respired  car- 
bon dioxid  accumulated. 

In  the  case  of  grain  stored  under  ordinary  commercial  conditions  it 
follows  that  the  oxygen  in  the  space  surrounding  the  kernels  must  be 
replaced  with  respired  carbon  dioxid.  The  rate  of  such  replacement 
will,  of  course,  hinge  upon  the  factors  influencing  the  rate  of  respira- 
tion. To  determine  the  relative  change  in  the  respiration  of  grain  stored 
in  a  tight  container  in  which  the  respired  carbon  dioxid  must  accumu- 
late, the  following  experiment  was  conducted:  A  sample  of  wheat 
containing  15.05  per  cent  of  moisture  was  divided  into  four  portions, 
which  were  incubated  in  the  usual  manner  at  37.8°  C.  At  the  end  of  i, 
4,  8,  and  12  days  a  cylinder  was  removed  and  the  respired  carbon  dioxid 
determined.  The  rate  of  respiration  for  the  first  day  and  the  average 
rate  for  each  of  the  4-day  periods  are  shown  in  Table  XVII  and  graph- 


Mar.  iS,  1918 


Respiration  of  Stored  Wheat 


707 


ically  in  figure  7.  Both  table  and  graph  show  plainly  that  the  rate  of 
respiration  is  reduced  by  the  accumulated  carbon  dioxid,  and  it  is  prob- 
able that  a  further  reduction  in  respiratory  activity  would  result  if  the 
proportion  of  carbon  dioxid  were  increased.  This  fact  has  an  important 
bearing  on  commercial  practices.  It  indicates  that  stored  grain  should 
not  be  disturbed  so  long  as  its  temperature  does  not  exceed  that  of  the 
atmosphere,  since  exposing  it  would  serve  to  ventilate  the  grain  and  thus 


P2 

Af(P/V. 

N 

^ 

.^ 

^\ 

y 

A 

y 

/ 

> 

y 

/ 

y 

t 

r 

/ 

/ 

J 

/ 

/ 

/ 

/ 

3  ^  S  £^  7-         <9  ^ 


/iP      // 


/^ 


Fig.  7. — Graph  showing  the  rate  of  respiration  during  successive  intervals  when  the  respired  carbon 
dioxid  was  permitted  to  accumulate  in  the  mass  of  grain. 

remove  the  carbon  dioxid.  On  returning  such  grain  to  the  bin  after 
aerating  it  the  rate  of  respiration  would  be  increased  over  that  of  the 
grain  which  was  not  handled  and  exposed,  since  the  rate  of  respira- 
tion of  the  unaerated  grain  has  been  depressed,  owing  to  the  accumulated 
carbon  dioxid.  The  aerated  grain  will  accordingly  heat  more  rapidly 
than  before  aerating,  unless  its  temperature  has  been  materially  reduced 
by  exposing  it  to  the  air. 


7o8 


Journal  of  Agricultural  Research  voi.  xii,  no.  n 


Table  XVII. — Rate  of  respiration  per  day  for  several  successive  periods 


Period. 


Carbon  dioxid 

respired  per 

loo  gm.  of  dry 

matter. 


First  day 

Average  rate  per  day  for  first  4-day  period.  . . 
Average  rate  per  day  for  second  4-day  period 
Average  rate  per  day  for  third  4-day  period. . 


Mgm. 

4.  II 
2.68 
1.49 
I.  II 


RESPIRATION  IN  OXYGEN-FREE  ATMOSPHERE 

That  respiration  may  occur  in  the  absence  of  oxygen  was  first  dis- 
covered by  Rollo  in  1798  {Hill,  191 3)  in  vi^orking  with  barley  grains. 
Other  investigators  confirmed  this  observation,  and  Pfefifer  (1S78) 
suggested  the  term  "intramolecular  respiration"  for  this  class  of  phe- 
nomena. Takahashi  (1903)  reported  that  rice  can  germinate  in  water 
without  the  presence  of  sugar  and  in  the  entire  absence  of  any  air. 
Hill  (1913)  determined  the  rate  of  respiration  of  water-soaked  and 
sterilized  wheat  in  air,  nitrogen,  and  hydrogen.  The  decrease  in  respira- 
tion in  a  continuous  current  of  hydrogen  and  nitrogen  below  that  in  a 
continuous  current  of  air  was  about  50  per  cent  in  seeds  sterilized  in 
alcohol  and  about  80  per  cent  in  seeds  sterilized  in  formalin. 

It  appeared  desirable  to  determine  the  effect  of  the  elimination  of 
oxygen  upon  the  respiration  of  stored  wheat,  and  to  this  end  the  follow- 
ing experiment  was  conducted:  Two  lots  of  wheat  containing  15.6  per 
cent  and  17.6  per  cent  of  moisture  respectively  were  secured.  A  por- 
tion of  each  lot  was  sealed  in  cylinders,  the  air  was  removed  and  replaced 
by  nitrogen.  These  lots  and  controls  in  ordinary  atmospheric  air  freed 
from  carbon  dioxid  were  then  incubated  for  four  days  at  a  temperature 
of  23.9°  C.  The  data  in  Table  XVIII  show  the  rate  of  respiration  in 
the  oxygen-free  atmosphere  to  have  been  reduced  to  about  two-fifths 
of  that  in  a  normal  atmosphere. 

Table  XVIII. — Comparative  rate  of  respiration  in  oxygen-free  and  normal  atmosphere 


Moisture. 

Carbon  dioxid  respired  per  100  gm. 
of  dry  matter  in  eacli  24  hours. 

Oxygen-free 
atmosphere. 

Normal 
atmosphere. 

Per  cent. 
15.6 
17.6 

Mgm. 

0.43 
2.80 

Mgm. 
I.  10 
6.80 

Mar.  i8. 1918  Respiratiou  of  Stored  Wheat  709 

CONCLUSIONS 

(i)  Deductions  from  these  investigations  support  the  findings  of 
earlier  investigators  that  spontaneous  heating  in  damp  grain  is  occa- 
sioned by  the  biological  oxidation  of  dextrose  and  similar  sugars,  chiefly 
in  the  germ  or  embryo  of  the  kernel. 

(2)  Moisture  is  one  of  the  determining  factors  in  respiration.  It 
establishes  the  comparative  rate  of  diffusion  between  the  several  kernel 
structures.  Any  gain  in  the  moisture  content  of  the  kernel  accord- 
ingly increases  the  rate  of  diffusion  and,  simultaneously,  the  rate  of 
respiration.  The  increase  is  gradual  and  fairly  uniform  until  the  mois- 
ture exceeds  14.5  per  cent,  in  the  case  of  plump  spring  wheat,  when  it 
is  markedly  accelerated. 

(3)  Density  of  the  wheat  Kernel  generally  parallels  the  gluten  con- 
tent. Gluten  possesses  the  property  of  imbibing  more  water  than  starch, 
and  thus  varying  percentages  of  gluten  result  in  varying  degrees  of 
viscosity  at  the  same  moisture  content.  The  relative  viscosity  affects 
the  rate  of  diffusion  and  this  in  turn  directly  affects  the  rate  of  respira- 
tion. The  soft,  starchy  wheats  thus  respire  more  rapidly  than  hard, 
vitreous  wheats  containing  the  same  percentage  of  moisture. 

(4)  Plumpness  of  the  wheat  kernel  affects  the  rate  of  respiration,  as 
shown  by  contrasting  plump  and  shriveled  grain.  The  shriveled  wheat 
respired  two  to  three  times  as  much  as  did  the  plump  wheat  at  moisture 
contents  above  14  per  cent.  At  percentages  of  moisture  below  14  per 
cent  the  difference  is  not  very  marked.  The  high  acceleration  of  respira- 
tion in  shriveled  wheat  containing  more  than  14  per  cent  of  moisture 
is  attributed  to  the  higher  ratio  of  germ  to  endosperm  and  hence  the 
larger  percentage  of  enzym  to  substrate  as  compared  with  plump  wheat. 

(5)  The  period  of  dampness — that  is,  the  length  of  time  the  excess 
moisture  has  been  present  in  the  wheat — bears  a  relation  to  the  rate  of 
respiration.  This  is  shown  by  comparing  the  respiration  of  freshly 
dampened  wheat  with  that  of  naturally  damp  grain  and  with  grain 
that  had  been  dampened  and  stored  for  varying  lengths  of  time.  The 
curve  of  respiration  diverges  from  that  of  freshly  dampened  wheat  when 
the  moisture  content  exceeds  12  per  cent,  and  this  divergence  is  more 
marked  after  13  per  cent  of  moisture  is  reached.  In  the  case  of  wheat 
dampened  and  stored,  the  quantity  of  carbon  dioxid  respired  varies 
directly  with  the  number  of  days  the  wheat  remained  in  storage.  The 
temperature  at  which  the  grain  is  stored  affects  the  rate  of  diastatic 
action,  thus  increasing  the  quantity  of  substrate  available  to  the  respira- 
tory enzyms.  This  is  indicated  by  the  greater  rate  of  respiration  of 
wheat  stored  at  room  temperature  than  that  stored  at  the  outdoor 
temperature    during    the    winter    months. 

(6)  Unsoundness  of  wheat  caused  by  the  freezing  of  the  unripe  plant 
results  in  higher  respiratory  activity  in  the  threshed  grain.     This  was 


7IO  Journal  of  Agricultural  Research.  voi.  xii.  no.  n 

shown  by  comparing  moderately  and  badly  frosted  wheats  with  sound 
wheat.  The  frosted  wheat  respired  more  vigorously  than  the  sound 
wheat.  This  was  attributed  to  the  arresting  of  the  synthetic  processes 
on  freezing,  and  subsequent  activities  of  the  hydrolytic  enzyms  on  thaw- 
ing of  the  frozen  wheat.  The  accumulation  of  glucose  as  the  result 
of  starch  hydrolysis  furnishes  larger  quantities  of  substrate  to  the 
respiratory  enzyms. 

(7)  Increasing  temperatures  accelerate  the  rate  of  respiration  until 
55°  C.  is  reached.  As  the  temperature  rises  the  diastatic  action  upon 
starch  increases.  A  point  is  reached,  however,  at  which  the  enzym 
activity  diminishes. 

(8)  Accumulation  of  carbon  dioxid  in  the  respiration  chamber 
decreases  the  rate  of  respiration.  The  mean  rate  by  four-day  intervals 
is  highest  for  the  first  four  days  and  diminishes  materially  in  successive 
periods. 

(9)  Respiration  is  reduced  in  an  oxygen-free  atmosphere,  the  ratio 
to  that  occurring  in  a  normal  atmosphere  being  about  i  to  2.5. 

LITERATURE  CITED 
ApplEman,  CO. 

1915.    RELATION  OP  CATALASE  AND  OXIDASES  TO  RESPIRATION  IN  PLANTS.      Md.  AgT. 

Exp.  Sta.  Bul.  191,  16  p.,  2  fig.     Literature  cited,  p.  16. 
Babcock,  S.  M. 

I912.    METABOLIC  water:    ITS  PRODUCTION   AND  r6lE  IN    VITAL  PHENOMENA.      Wis. 

Agr.  Exp.  Sta.  Research  Bul.  22,  p.  87-181. 
Bailey,  C.  H. 

1917a.   THE    HANDLING    AND    STORAGE    OF    SPRING    WHEAT.      hi    Jour.-  Amer.    SoC. 

Agron.,  V.  9,  no.  6,  p.  275-281,  fig.  16-20. 


1917b.  THE  MOISTURE  CONTENT  OF  HEATING  WHEAT.     In  Jour.  Amer.  Soc.  Agron., 
V.  9,  no.  5,  p.  248-251. 

Bailey,  E.  M. 

1912.   biochemical    and    bacteriological    STUDIES    ON    THE    BANANA.      In    JoUr. 

Amer.  Chem.  Soc,  v.  34,  no.  12,  p.  1706-1730. 
Barnes,  J.  H.,  and  Grove,  A.  J. 

I916.  THE  INSECTS  ATTACKING  STORED  WHEAT  IN  THE  PUNJAB,  AND  THE  METHODS 
OF  COMBATING  THEM,  INCLUDING  A  CHAPTER  ON  THE  CHEMISTRY  OF  RESPI- 
RATION. In  Mem.  Dept.  Agr.  India,  Chem.  s.,  v.  4,  no.  6,  p.  i65-28od, 
12  fig.,  4  col.  pi. 

BoNNiER,  Gaston,  and  Mangin,  Louis. 

1885.  LA  FONCTiON  RESPiRATOiRE  CHEZ  LES  v^G^TAUX.  In  Ann.  Sci.  Nat.  Bot., 
s.  7,  t.  2,  p.  365-380. 

Brenchley,  W.  E. 

1909.   ON   THE  STRENGTH   AND   DEVELOPMENT   OP   THE   GRAIN   OP   WHEAT   (TRITICUM 

vulgare).     In  Ann.  Bot.,  v.  23,  no.  89,  p.  117-139,  5  fig.,  pi.  8-9.     Litera- 
ture cited,  p.  138. 


Mar.  i8, 1918  Respiratiofi  of  Stored  Wheat  711 

Brown,  H.  T.,  and  Morris,  G.  H. 

1890.    RESEARCHES    ON    THE    GERMINATION    OP   SOME    OP    THE    GRAMINEAE.      I.      In 

Jour.  Chem.  Soc.  [London,]  v.  57,  p.  458-528,  2  pi. 

BURLAKOW,  G. 

1897.    UEBER  ATHMUNG  DES  KEIMES  DBS  WEIZENS,  TRITICUM  VULGARE.      (Abstract.) 

In  Bot.  Centbl.,  Bd.  74,  No.  11,  p.  323-324.  1898.  (Original  article  in 
Arb.  Naturf.  Gesell.  K.  Univ.  Charkow,  Bd.  31,  Beilage,  p.  i-xv.  1897. 
Not  seen.) 

Clausen,  H. 

1890.    BEITRAGE  ZUR  KENNTNISS  DER  ATHMUNG  DER  GEWACHSE.      In  Landw.  Jahrb., 

Bd.  19,  p.  893-930,  pi.  19-20. 
DuvEL,  J.  W.  T. 

1904.    THE    VITALITY    AND   GERMINATION    OP   SEEDS.      U.  S.    Dcpt.    Agr.    BUT.    Plant 

Indus.  Bui.  58,  96  p.,  I  fig.     Literature  cited,  p.  90. 


1909.  THE  DETERIORATION  OP  CORN  IN  STORAGE.     U.  S.  Dept.  Agr.  Bur.  Plant 
Indus.  Circ.  43,  12  p.,  3  fig. 

,  and  Duval,  Laurel. 


I913.   THE  SHRINKAGE  OF  SHELLED  CORN  WHILE  IN  CARS  IN  TRANSIT.      U.  S.  Dcpt. 

Agr.  Bui.  48,  21  p.,  8  fig. 

Gore,  H.  C 

1911.  STUDIES  ON  FRUIT  RESPIRATION.     U.  S.  Dept.  Agr.  Biu".  Chem.  Bui.  142,  40  p., 
17  fig- 


1914.  CHANGES  IN  COMPOSITION  OP  PEEL  AND  PULP  OP  RIPENING  BANANAS.      In  JoUT. 

Agr.  Research,  v.  3,  no.  3,  p.  187-203,  i  fig.     Literature  cited,  p.  202-203. 

GURJAR,  A.  M. 

I917.    THE   ADAPTATION   OP  TRUOG'S   METHOD  FOR  THE   DETERMINATION   OP  CARBON 

DioxiD  TO  PLANT  RESPIRATION  STUDIES.     In  Plant  World,  V.  20,  no.  9, 

p.  288-293, 1  fig- 

Hasselbring,  Heinrich,  and  Hawkins,  L.  A. 

1915.  RESPIRATION  EXPERIMENTS  WITH  SWEET  POTATOES.     In  Jour.  Agr.  Research, 

V.  5,  no.  12,  p.  509-517.     Literature  cited,  p.  517. 

HiLL,  G.  R.,  Jr. 

I913.    RESPIRATION  OP  FRUITS  AND  GROWING  PLANT    TISSUES  IN  CERTAIN  GASES, 
WITH  REFERENCE  TO  VENTILATION  AND  FRUIT  STORAGE.      N.  Y.  Cornell  Agr. 

Exp.  Sta.  Bui.  330,  p.  373-408.     Bibliography,  p.  407-408. 
Cites  Rollo,  p.  379- 

HoPP,  J.  H.  van't. 

1896.  STUDIES  IN  CHEMICAL  DYNAMICS.     Translated  by  Thomas  Ewan.     286  p.,  49 
fig.     Amsterdam,  London. 

KarchEvski. 
1901.  influence  of  the  fluctuation  op  temperature  on  THE  respiration  OP 
SEEDS  AND  EMBRYOS  OP  WHEAT.     (Abstract.)     In  Exp.  Sta.  Rec,  v.  14, 
no.  9,  p.  839.     1903.     (Original  article  in  Mem.  Warsaw  Univ.,  v.  9,  114  p. 
1901.     Not  seen.) 

KOLKWITZ,  R. 

1901.  UEBER  DIE  ATHMUNG  RUHENDER  SAMEN.     In  Ber.  Dcut.  Bot.  Gesell.,  Bd.  19, 
Heft  4,  p.  285-287. 


712  Journal  of  Agricultural  Research  voi.  xii.  No.  n 

Langworthy,  C.  F.,  and  Milner,  R.  D. 

I913.   SOME  RESULTS  OBTAINED  IN  STUDYING  RIPENING  BANANAS  WITH  THE  RESPIRA- 
TION CALORIMETER.     In  U.  S.  Dept.  Agr.  Yearbook,  1912,  p.  293-308. 

LE  Clerc,  J.  A.,  and  Breazeale,  J.  F. 

191 1.   TRANSLOCATION  OF  PLANT  FOOD  AND  ELABORATION  OP  ORGANIC  PLANT  MATE- 
RIAL IN  WHEAT  SEEDLINGS.     U.  S.  Dept.  Agr.  Bur.  Chem.  Bui.  138,  32  p. 

LoEw,  Oscar. 

1899.  CURING  AND  FERMENTATION  OF  CIGAR  LEAF  TOB.'VCCO.      U.  S.  Dept.  Agr.  Dept. 

Rpt.  59,  34  p. 

Lund,  J.  F. 

1894.   NOTE  SUR  L'iNFLUENCE  DE  LA  DESSICCATION  SUR  LA  RESPIRATION  DBS  TUBER- 

CULES.    In  Rev.  Gen.  Bot.,  t.  6,  p.  353-355. 
Maige,  a.,  and  Nicolas,  G. 

1910.   RECHERCHES  SUR  L 'INFLUENCE  DES  SOLUTIONS  SUCR^ES  DE  DIVERS  DEGR^S 
DE  CONCENTRATION  SUR  LA  RESPIRATION ,  LA  TURGESCENCE  ET  LA  CROISSANCE 

DE  LA  CELLULE.     In  Ann.  Sci.  Nat.  Bot.,  s.  9,  t.  12,  no.  2/6,  p.  315-368. 
Mangin,  Louis. 

1896.   SUR   LA    V]6g6TATION    DANS    UNE   ATMOSPHl^RE    VIClfiS   PAR   LA    RESPIRATION. 

In  Compt.  Rend.  Acad.  Sci.  [Paris],  t.  122,  no.  12,  p.  747-749. 
Mann,  Albert,  and  Harlan,  H.  V. 

191 5.   MORPHOLOGY  OF  THE  BARLEY  GRAIN  WITH  REFERENCE  TO  ITS  ENZYM-SECRETING 
AREAS.     U.  S.  Dept.  Agr.  BuL  183,  32  p.,  7  fig.,  8  pi. 

Maquenne,  L. 

1900.  RECHERCHES  SUR  LA  GERMINATION.    In  Ann.  AgTon.,  t.  26,  no.  7,  p.  321-332. 

Matthaei,  G.  L.  C. 

1905.   EXPERIMENTAL  researches  ON  VEGETABLE  ASSIMILATION  AND  RESPIRATION. 
III.   ON   THE   EFFECT  OF  TEMPERATURE  ON  CARBON-DIOXIDE   ASSIMILATION. 

In  Phil.  Trans.  Roy.  Soc.  [London],  s.  B,  v.  197,  no.  228,  p.  47-86,  6  fig. 
Bibliography,  p.  85-86. 

MtJNTZ,  A. 

1881.  SUR  LA  CONSERVATION  DES  GRAINS  PAR  l'ensilage.    In  Compt.  Rend. 
Acad.  Sci.  [Paris],  t.  92,  no.  3,  p.  137-139. 

Nabokich,  a.  J. 

1903.  UBER  DEN  EINFLUSS  DER  STERILISATION  DER  SAMEN  AUP  DIE  ATMUNG.     In 

Ber.  Deut.  Bot.  Gesell.,  Bd.  21,  Heft  5,  p.  279-291. 

OSTERHOUT,  W.  J.  V. 

1917.  THE  r6le  of  the  nucleus  IN  OXIDATION.    In  Science,  n.  s.,  v.  46,  no.  1189, 
P-  367-369- 
Pfeffer,  W. 

1878.   DAS  WESEN  UND  DIE  BEDEUTUNG  DER  ATHMUNG  IN  DER  PFLANZE.     In  Landw. 
Jahrb.,  Bd.  7,  p.  805-834. 

PFLtJGER,  E. 

1875.    UEBER  DAS  PHYSIOLOGISCHE  VERBRENNUNG  in  DEN  LEBENDIGEN  ORGANISMEN. 

In  Arch.  Physiol.  [Pfluger],  Bd.  10,  p.  251-367. 

Plimmer,  R.  H.  a. 

1915.  PRACTICAL  ORGANIC  AND  bio-chemistry.    635  p.,  86  fig.,  I  col.  pi.  London, 
New  York. 


Mar.  i8,  igis  RespiratioYi  of  Stored  Wheat  713 

QvAM,  Olaf. 

1906.    KORNETS   AANDING.    EN    RELATION    MELLEM  SPIREEVNE  OG  AANDINGSINTEN- 

siTET.     Tidsskr.  Norske  Landbr.,  Aarg.  13,  p.  263-284,  14  fig. 

Rahn,  Otto. 
1910.  THE  usefulness  of  curves  in  the  interpretation  op  microbial  and 
BIOCHEMICAL  PROCESSES.    Mich.  AgT.  Exp.  Sta.  Tech.  Bui.  5,  p.  347-372, 

18  fig. 

SaussurE,  Theodorde. 
1804.  recherches  chimiques  sur  la  v6g6tation.    327  p.,  i  pi.    Paris. 

Shanahan,  J.  D.,  Leighty,  C.  E.,  and  Boerner,  E.  G. 

1910.  AMERICAN  EXPORT  CORN  (maize^  IN  EUROPE.  U.  S.  Dcpt.  AgT.  BuF.  Plant 
Indus.  Circ.  55,  42  p.,  7  fig. 

Suchsland,  Emil. 

1891.  UEBER  TABAKSFERMENTATiON.    In  Bet.  Deut.  Bot.  Gesell.,  Bd.  9,  Heft  i, 

p.  79-81. 

Takahashi,  T. 

1905.  is  GERMINATION  POSSIBLE  IN  THE  ABSENCE  OF  AIR?  (Abstract.)  In  Exp. 
Sta.  Rec,  v.  16,  no.  11,  p.  1090.  1905.  (Original  article  in  Bui.  Col.  Agr. 
Tokyo  Imp.  Univ.,  v.  6,  no.  4,  p.  439-442.     1905.    Not  seen.) 

Troug,  E. 

I915.  METHODS  for  the  DETERMINATION  OF  CARBON  DIOXIDE  AND  A  NEW  FORM  OF 
ABSORPTION    TOWER    ADAPTED    TO    THE    TITRIMETRIC    METHOD.      In    Jour. 

Indus,  and  Engin.  Chem.,  v.  7,  no.  12,  p.  1045-1049,  i  fig. 

Verworn,  Max. 
1899.  GENERAL  PHYSIOLOGY.    Translated  by  F.  S.  Lee.    632  p.,  285  fig.    London. 

Wehmer,  C. 

1892.  oxalsaures  ammon  als  pilzliches  stoffwechselproduct  die  ernahrung 

durch  eiweiss.  (Abstract.)  In  Bot.  Centbl.,  Bd.  51,  No.  11,  p.  337. 
1892.  (Original  article  in  Jahresber.  Natiu-h.  Gesell.  Hannover,  1892, 
p.  99-106.    Not  seen.) 

Wender,  Neumann. 

1905.  peinheitsbestimmung  der  mehle.  In  Ztschr.  Untersuch.  Nahr.  u.  Gen- 
ussmtl.,  Bd.  10,  Heft  12,  p.  747-756. 

White,  Jean. 

1909.  the  ferments  and  latent  life  of  resting  seeds.  In  Proc.  Roy.  Soc. 
[London],  s.  B,  v.  81,  no.  550,  p.  417-442.     Special  bibliography,  p.  441-442. 

Ziegenbein,  Ernst. 

1893.  UNTERSUCHUNGEN  USER  DEN  STOFPWECHSEL  UNO  DIE  ATHMUNG  KEIMENDER 

kartoffelknollen  sowie  anderer  pflanzen.  In  Jahrb.  Wiss.  Bot. 
[Pringsheim],  Bd.  25,  Heft  4,  p.  563-606. 


EFFECTS  OF  MISTLETOE  ON  YOUNG  CONIFERS 

By  James  R.  Weir 

Forest  Pathologist,  Investigations  in  Forest  Pathology,  Bureau  of  Plant  Industry, 

United  States  Department  of  Agriculture 

During  the  years  191 1  to  1917  the  writer  has  been  engaged  in  the 
investigation  of  the  injurious  effects  of  mistletoes  of  the  genus  Razou- 
mofskya  (Arceuthobium)  on  conifers.  A  part  of  these  investigations 
have  already  been  published.*  During  the  course  of  the  work,  a  number 
of  studies  were  made  which  have  not  yet  been  reported.  It  seems  that 
these  studies  are  of  sufficient  importance  to  be  presented  at  this  time. 

The  fact  that  these  parasites  are  a  cause  of  suppression  ^  in  forest 
trees  is  readily  appreciated  after  the  injury  has  become  acute.  In 
middle-aged  and  older  trees  where  accumulated  injury  has  resulted  in 
small  diameters  and  broomed  scraggly  crowns  with  reduced  leaf  surface, 
the  effect  of  the  parasites  is  quite  evident  (PI.  37,  A).  The  effects  of 
mistletoe  on  its  host  is  by  no  means  so  apparent  on  trees  ranging  in  age 
from  4  to  10  years.  Trees  which  have  become  infected  early  in  life 
may  begin  to  react  in  marked  degree,  but  the  extent  of  the  suppression 
is  not  always  apparent  to  the  eye.  It  may  be  noticed  that  the  branches 
of  young  trees  are  broomed  along  with  other  types  of  infection  of  a  more 
general  nature,  but  the  retardation  of  the  excurrent  or  elongated  growth 
of  the  main  stem  may  not  be  readily  recognized.  This  is  where  detailed 
measurements  are  of  value  and  prove  or  disprove  the  early  suppression 
of  trees  of  the  younger  age  classes  by  mistletoe. 

In'  order  to  demonstrate  the  suppression  in  young  trees,  a  series  of 
measurements  of  the  height  growth  of  w^estern  yellow  pine  (Pinus  ponde- 
rosa)  were  made  in  Spokane  County,  Washington.  Two  representative 
plots  of  I  acre  each,  consisting  of  infected  and  uninfected  reproduction 
and  representing  all  age  classes,  were  selected.  The  site  was  level  bench 
land  with  a  sparse  stand  of  merchantable-sized  trees.  Some  of  these  trees 
were    severely    infected    with    mistletoe    {Razoumofskya    campy lopoda 

'  Weir,  J.  R.  larch  mistletoe:  some  economic  considerations  of  its  injurious  effects.  U.  S. 
Dept.  Agr.  Bui.  317,  25  p.,  13  fig.  1916. 

mistletoe  injury  to  conifers  in  the  northwest.    U.  S.  Dept.  Agr.  Bui.  360,  39  p.,  27  fig. 

X916.     literature  cited,  p.  39.' 

some  suggestions  on  the  control  of  mistletoe  in  the  N.^TION.^L  forests  op  the  northwest. 

In  Forestry  Quart.,  v.  14,  no.  4,  p.  567-577.  1916. 

'  The  term  "suppression"  is  not  used  here  in  the  ordinary  forestry  sense,  but  refers  to  a  retardation  of 
growth  induced  by  parasitic  organisms. 

Journal  of  Agricultural  Research,  Vol.  XII,  No.  n 

Washington,  D.  C.  (/^S)  VltiT.  18,  1918 

ml  Key  No.  G— 138 


7i6 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  ir 


(Engelm.)  Piper]  and  were  the  source  of  infection  of  the  reproduction. 
The  young  trees  were  well  distributed  over  the  area  and  were  dominant. 
The  results  of  this  study  are  summarized  in  Table  I  and  demonstrate 
very  clearly  the  damage  which  may  result  from  attack  by  this  parasite 
on  yellow  pine. 

Table  I. — Effect  of  mistletoe  on  yellow  pine 


Age  of  tree. 


Basis  No. 


Infected. 


Years. 

4 '  8 

5 lo 

6 14 

7 !  9 

8 j  19 

9 1  10 

10 1  7 

I 

Total 77 


Unin- 
fected. 


50 
27 

42 

35 
20 

23 
16 


213 


Average  height  of  trees. 


Infected. 


Cm. 

15.  22 
22.  84 

35-53 
48.  22 

53-29 
63-45 
73.60 


44,72 


Uninfected. 


Cm 

25 

35' 
50' 
71 
78. 

93 
109 


■38 
•53 
.76 
.06 
.68 
.90 
-  13 


57-87 


DiEference  in 

height  be- 
tween infect- 
ed and  un- 
infected 
trees. 


Cm.. 
10.  15 
12.  69 
15.22 
22.  84 
25-38 
30-45 
35-  53 


13-15 


During  the  above  study  it  was  plainly  observed  that  the  length  of  the 
intemodes  of  the  infected  trees,  also  the  length  of  the  terminal  and  lateral 
buds  of  the  main  shoot,  were  much  shorter  than  those  of  the  uninfected, 
and  led  to  the  following  study  on  Douglas  fir  {Pseudotsuga  taxijolia)  in 
the  Missoula  River  region  near  Missoula,  Mont.  The  trees  were  infected 
with  Razoumofskya  douglasii  Engelm.  For  this  study  trees  were  selected 
over  an  area  without  recourse  to  sample  plots.  The  only  point  adhered 
to  was  the  selection  of  trees  on  same  type  of  site,  condition  of  growth, 
and  an  average  age  of  18  years.     The  results  are  given  in  Table  II 

Table  II. — Effect  of  mistletoe  on  Douglas  fir 


Trees  seriously  infected. 

Trees  vigorous,  not  infected. 

Average 

Average  di- 

Average 

Average  di- 

V, 

Average  height 

dimen- 

mension and 

u 

Average  height 

dimen- 

mension and 

■w 

growth  of  last  four 

sion  ter- 

number of 

4-» 

growth  of  last  four 

sion  ter- 

number of 

■3 

intemodes. 

minal 

lateral  termi- 

a 

intemodes. 

minal 

lateral  termi- 

1-> 

St 

bud. 

nal  buds. 

a 

bud. 

nal  buds. 

a 

J3 

.a 

A 

5 

ja 

.a 

A 

5 

•a 

n 

I 

3 

3 

4 

1 

n 

No. 

bo 

% 

P3 

.52 

I 

" 

3 

4 

ti 

S 

No. 

ti 

1 

Cm. 

Cm. 

Cm. 

Cm. 

Mm. 

Mm. 

Mm. 

Mm. 

Cm. 

Cm. 

Cm,. 

Cm. 

Mm. 

Mm. 

Mm. 

Mm. 

so 

28 

23 

33 

18 

8 

3 

loS 

S 

3 

so 

36 

33 

36 

35 

13 

4 

157 

7 

3  + 

The  foregoing  results  clearly  demonstrate  the  eflFect  of  the  formation 
of  brooms  and  burls  on  the  storage  of  food  materials  in  the  terminal  buds 
and  shoots.     It  is  a  well-known  fact  that  in  the  terminal  bud  are  stored 


Mar.  i8.  i9is  Effects  of  Misiletoe  Oil  Young  Couifers  717 

the  elaborated  food  materials  for  its  early  development  the  following 
season.  If  this  food  material  is  reduced  in  amount  b}'  its  becoming 
localized  in  other  parts  of  the  tree,  the  growth  of  the  main  shoot  must  be 
necessarily  retarded  (PI.  37,  B),  and  the  bud  itself  will  form  earher  in 
the  season  and  be  much  reduced  in  size.  Two  yellow-pine  trees,  each 
8  years  of  age,  one  with  a  conspicuous  infection  with  broom  formation, 
the  other  entirely  free  from  infection,  were  carefully  observed  to  deter- 
mine this  point.  The  former  not  only  started  the  elongation  of  the  main 
shoot  nine  days  later  than  the  other  tree,  but  ceased  to  develop  altogether 
at  the  end  of  the  first  month.  The  shoot  of  the  uninfected  tree  continued 
to  elongate  for  two  months  and  showed  a  gain  of  1 1  inches  over  that  of  the 
infected  tree.  The  trees  grew  under  exactl}'  similar  conditions,  and  their 
root  systems  were  practically  equal  in  extent.     Kirkwood  ^  states  that — 

It  is  improbable  that  the  whole  growth  of  the  new  leader  is  at  the  expense  of  the  food 
stored  in  the  bud  alone.  That  from  other  parts  also  doubtless  contributes,  but  the 
tendency  is  to  crowd  the  formative  materials  toward  the  extrmities  of  the  main  shoot 
and  the  branch.  In  the  sharing  of  these  materials  the  main  shoot  leads  and  the 
branches  follow  in  order  of  their  importance.  The  principal  growth,  however,  is 
undoubtedly  at  the  expense  of  the  locally  stored  materials,  the  substances  stored  else- 
where having  their  part  to  play  in  the  development  of  the  tissues  in  their  immediate 
proximity. 

The  storage  of  food  in  the  shoot  and  branches  is  exactly  what  the 
formation  of  brooms  and  burls  prevents  in  a  large  measure  in  all  parts  of 
the  tree  above  the  seat  of  infection,  and  eveaatually  results  in  retardation 
and  the  appearance  of  spike  top  or  staghead.  The  writer  has  repeatedly 
called  attention  to  this  in  previous  publications.  In  order  to  demonstrate 
that  there  is  an  actual  storage  of  food  materials  in  mistletoe  brooms  greater 
than  that  of  normal  branches  the  results  of  an  experiment  may  be  given. 
Late  in  the  m.onth  of  October,  191 4,  after  the  leaves  had  fallen,  10  mistle- 
toe brooms  and  10  uninfected  branches  from  points  ranging  from  5  to  15 
feet  above  the  fomier  were  cut  from  a  western  larch  and  thrown  in  a  damp, 
shady  ravine.  In  June  of  the  following  year  the  brooms  and  branches 
were  examined  with  the  following  result:  Practically  all  foliar  spurs  of 
the  brooms  developed  needles  in  proportion  of  about  one-third  of  the 
normal  length.  A  fev/  of  the  foliar  spurs  of  the  uninfected  branches  pro- 
duced needles  barely  protruding  from  the  bud  scales,  and  in  most  cases 
there  was  no  leaf  production  whatever.  Observation  of  brooms  in  the 
crowns  of  larch  cut  late  in  the  fall  during  logging  operations  showed  in 
the  production  of  needles  in  the  spring  that  there  must  be  a  great  localiza- 
tion of  elaborated  food  materials  in  the  branches  of  the  brooms  over  that 
of  the  normal  branches.'*     The  latter  showed  no  signs  of  foliation. 


'  Kirkwood,  J.  E-  the  influence  op  preceeding  seasons  on  the  growth  of  yellow  pine.  In 
Torreya,  v.  14,  no.  7,  p.  118.     1914. 

*  Certain  parasitic  fungi  also  cause  a  flow  of  building  materials  to  the  place  of  infection.  See  GoebEL, 
K  .  E.     EINLEITUNG  IN  DIE  EXPEKIMKNTELLE  MOKPHOLO'^IE  DER  PFLANZEN.     p.  75.      Leipzig  uud  Berlin. 

1908. 

3832G°— 18 3 


71 8  Journal  of  Agricultural  Research  voi.  xn.  xo. 


-^v- 


From  the  foregoing  studies  it  may  be  concluded  that  the  false  mistle- 
toes are  serious  agents  in  the  suppression  of  young  forest  growth. 
Moreover,  young  growth  once  infected  on  the  main  stem  (and  it  is 
usually  so  infected)  can  not  recover  and  produce  merchantable  material. 
Suppression  in  young  yellow  pine  even  up  to  the  sapling  stage  is  of 
serious  consequence.  Very  seldom  do  such  trees  ever  overcome  the 
early  influence.  This  is  also  noticeable  in  trees  of  dense  stands  or 
when  overtopped  by  older  classes.  The  ultimate  effect  on  the  growth 
of  the  tree  is  exactly  equivalent  to  lack  of  light,  only  worse.  In  the 
former  case  the  excurrent  growth  is  arrested,  and  the  tree  either 
develops  into  one  continuous  broom  or  dies;  in  the  latter,  a  sapling 
of  some  merchantable  value  may  result.  It  is  clear  that  every  effort 
should  be  made  in  regions  of  heavy  mistletoe  infection  ro  reduce  the 
infection  of  reproduction  by  cutting  all  infected  overtopping  trees. 
Care  should  be  taken  to  prevent  the  introduction  of  mistletoe-in- 
fected transplants  in  regions  where  the  parasites  do  not  occur/'  Such 
resfions,  for  examole,  are  the  Black  Hills  of  South  Dakota  and  several 
of  the  southeastern  Montana  forests. 

SUMMARY 

The  height  growth  of  young  trees  is  greatly  retarded  by  mistletoe. 
The  effects  to  be  observed  are  reduction  in  the  length  of  the  internodes, 
small  dimensions  of  the  terminal  buds,  and  reduction  in  number. 
This  result  is  caused  by  the  localization  of  food  materials  at  the  seat 
of  infection. 

To  reduce  the  chances  of  infecting  young  growth,  all  overtopping 
infected  trees  should  be  killed.  Infected  trees  of  any  age  should  be 
killed  if  possible.  Care  should  be  taken  that  infected  trees  are  not 
planted  in  regions  w^here  mistletoe  does  not  occur. 

PLATE  37 
A. — Pseudotsuga  taxifolia  infected  with  Razoumofskya  douglasii. 
B. — Effect  of  an  inoculation  with  Razoumofskya  campylopoda  on  the  height  growth 
of  6-year-old  Pimts  jeffreyi.    Culture  made  in  a  greenhouse  at  Missoula,  Mont. 


5  Weir,  J.  R.    mistletoe;  injury  to  conifers  in  the  northwest.    U.  S.  Dept.  Agr.  Bui.  360,  39  p. 
27  fig.    1916.    Literature  cited,  p.  39. 


Effects  of  Mistletoe  on  Young  Conifers 


Plate  37 


Journal  of  Agricultural  Research 


Vol.  XII,  No.  II 


DETERMINATION  OF  FATTY  ACIDS  IN  BUTTER  FAT:  V 

By  E.  B.  Holland,  Associate  Chemist,  and  J.  P.  Buckley,  Jr.,  Assistant  Chemist, 
Massachusetts  Agricultural  Experiment  Station  . 

INTRODUCTION 

Since  reporting  a  method  (14)  ^  for  the  determination  of  stearic  acid 
in  butter  fat,  work  has  been  continued  with  a  \'iew  of  evolving  a  process 
for  determining  some  of  the  remaining  fatty  acids.  The  object  of  the 
investigation  was  to  deduce  practical  analytical  methods  which  might 
serve  to  measure  the  effect  of  feed  upon  the  composition  of  the  resulting 
butter  fat.  Such  methods,  however,  if  quantitative  and  reasonably 
workable,  would  be  applicable  to  other  fixed  oils  and  fats  and  play  a  part 
in  the  so-called  technical  examination  of  such  products. 

There  are  several  distinct  lines  of  procedure  upon  which  methods 
for  determining  different  fatty  acids  in  a  mixture  might  be  based : 
(i)  Crystallization  of  the  acids ; 

(2)  Solubility  of  various  salts; 

(3)  Fractionation  of  the  acids  in  vacuo; 

(4)  Fractionation  of  methyl  or  ethyl  esters. 

The  iodin  absorption  and  the  acetyl  substitution  are  measures  of 
unsaturated  and  of  hydroxy  acids,  respectively.  They  are  valuable 
adjuncts,  but  of  limited  rather  than  of  general  application. 

Attention  has  been  called  in  several  instances  to  some  of  the  inherent 
faults  (jj,  14)  of  the  different  schemes  of  separation,  and  a  discussion 
of  their  relative  merits  appears  unnecessary  at  this  time.  It  will  suffice 
to  summarize  the  present  application  and  limitation  of  the  schemes. 

Crystallization  methods  have  been  employed  for  the  quantitative 
separation  of  stearic  and  of  arachic  (27)  acids  but  seldom  for  other  acids. 

The  lead-salt-ether  method,  or  Gusserow  (7)-Varrentrapp  (29)  process, 
for  the  separation  of  liquid  from  solid  acids  is  the  most  prominent  illus- 
tration of  the  salt-soluble  methods  which,  as  a  rule,  have  not  proved 
sufficiently  discriminative  for  quantitative  use. 

Fractionation  of  the  acids  in  vactw  has  failed  as  an  analytical  process. 

Fractionation  of  ethyl  esters  appeared  applicable  to  many  fatty  acids, 
although  chemists  generally  consider  the  process  as  having  little  quanti- 
tative significance.  The  experience  gained  in  purifying  fatty  acids  indi- 
cated that  the  method  was  practicable  and  at  least  worthy  of  additional 
study. 


>  From  the  Department  of  Chemistry,  Massachusetts  Agricultural  Experiment  Station.     Printed  with 
the  permission  of  the  Director  of  the  Station. 
2  Reference  is  made  by  number  (italic)  to  "  Literature  cited,"  pp.  73i-7,i2- 


Journal  of  Agricultural  Research,  Vol.  XII,  No.  ii 

Washington,  D.  C.  Mar.  i8.  igiS 

mn  (719)  Key  No.  Mass.  4 


-20  Journal  of  Agricultural  Research  voi.  xii, No.  n 

EARLIER   INVESTIGATIONS 

At  the  present  writing  esterification  methods  have  so  wide  an  appli- 
cation and  contributors  to  the  literature  on  the  subject  are  so  numerous 
that  only  references  bearing  directly  on  the  analysis  of  oils  and  fats 
will  be  cited. 

Rochleder  (28)  studied  the  action  of  absolute  alcohol  and  dry  hydro- 
chloric acid  gas  upon  castor  oil  in  the  separation  of  glycerol. 

Berthelot  (2,  p.  311-312)  applied  the  reaction  to  a  number  of  glycerides 
and  showed  that  esters  of  the  fatty  acids  were  formed  in  addition  to 
glycerol. 

Juillard  {15,  p.  239)  prepared  methyl  and  ethyl  esters  of  dihydroxy- 
stearic  acid  by  boiling  the  acid  with  10  times  its  weight  of  alcohol  and 
2  or  3  drops  of  sulphuric  acid. 

Fischer  and  Speier  (5)  esterified  various  organic  acids  with  different 
amounts  of  absolute  ethyl  alcohol  and  dry  hydrochloric  or  concentrated 
sulphuric  acid,  also  of  methyl  alcohol  and  hydrochloric  acid  and  noted 
the  yield.     They  did  not  deduce  a  general  method. 

Haller  {8-1 1)  heated  a  variety  of  oils  and  fats  with  twice  their  weight 
of  absolute  methyl  alcohol  containing  i  to  2  per  cent  of  hydrochloric 
acid,  removed  the  glycerol  and  excess  alcohol  with  water,  or  preferably 
brine,  and  fractionated  the  esters  up  to  194°  C.  (methyl  caprylate) 
at  atmospheric  pressure  and  the  residual  esters  in  vacuo.  The  methyl 
oleate  present  in  the  myristic  and  higher  fractions  was  removed  by 
chilling  and  absorption  on  porous  plates.  When  complete  alcoholysis 
was  not  obtained,  he  recommended  a  second  treatment  with  a  new  por- 
tion of  acidulated  alcohol,  the  employment  of  a  larger  amount  of  alcohol 
at  the  outset,  or  the  addition  of  an  inert  solvent  such  as  ether  to  facilitate 
the  reaction,  particularly  in  the  case  of  butter  fat  and  of  drying  oils  which 
readily  oxidize  and  polymerize.  The  production  of  a  small  amount  of 
aldehyde  was  noted  in  some  instances.  Haller  neutralized  the  esters 
wAth.  barium  carbonate  or  a  solution  of  sodium  carbonate  and  dried  over 
calcium  chloride  or  anhydrous  sodium  sulphate. 

Phelps  and  Hubbard  {18)  esterified  succinic  acid  with  ethyl  alcohol 
and  hydrochloric  acid,  and  secured  the  greatest  yield  with  maximum 
dehydration.  In  other  experiments  {ig)  the  addition  of  anhydrous  zinc 
chlorid  enhanced  the  reaction. 

Complete  esterification  of  50  gm.  of  benzoic  acid  {26)  was  obtained 
by  treating  for  4  hours  with  400  c.  c.  of  absolute  alcohol  containing  1.25 
per  cent  of  dry  hydrochloric  acid  and  10  gm.  of  anhydrous  zinc  chlorid, 
or  for  3  hours  with  200  c.  c.  of  absolute  alcohol  and  2  gm.  of  sulphuric 
acid. 

Several  other  chlorids  {24,  p.  296-297;  20-25)  proved  nearly  as  effi- 
cient as  zinc  chlorid  under  like  conditions  of  operation. 


Mar.  i8, 1918  Fatty  Acids  ifi  B litter  Fat  721 

Meyer  (17)  employed  substantially  the  Haller  method  with  cotton- 
seed oil,  but  increased  the  amount  of  methyl  alcohol  to  four  times  the 
weight  of  the  oil,  and  obtained  a  yield  of  about  90  per  cent. 

Elsdon  {3,  4)  employed  the  Haller  method  with  coconut  oil  and  palm- 
kernel  oil,  fractionated  the  resulting  methyl  esters  in  vacuo,  and  refrac- 
tionated  to  constant  boiling  point.  He  stated  that  the  process  had 
qualitative  and  a  considerable  amount  of  quantitati\"e  value,  but  was 
too  lengthy  for  ordinary  use. 

Kailan  {16)  found  that  ether,  benzene,  and  carbon  tetrachlorid  did 
not  accelerate  the  esterification  of  benzoic  acid  with  absolute  alcohol 
and  hydrochloric  acid  or  with  dilute  alcohol  and  acid. 

Wolff  and  Scholze  (50)  used  a  dilute  sodium-bicarbonate  solution  to 
purify  the  esters  when  shaking  out  with  ether. 

Abderhalden  and  Kautzsch  (j)  esterified  the  silver  salt  of  an  amino 
acid  by  boiling  with  an  excess  of  ethyl  iodide. 

Grandmougin,  Havas,  and  Guyot  (6)  showed  how  an  organic  acid  after 
treatment  with  sodium  methylate  might  be  converted  by  means  of  an 
excess  of  dimethyl  sulphate  into  the  methyl  ester  of  the  organic  acid  and 
sodium  methyl  sulphate. 

Possibly  methyl  sulfonic  acid  might  be  substituted  in  some  instances 
for  dimethyl  sulphate  or  methyl  halide  as  indicated  by  an  English  patent, 
No.  9359-^ 

Hauser  and  Klotz  {12)  esterified  organic  acids  with  alcohols  by  pass- 
ing the  vapors  over  glucinum  oxide  heated  to  310°  C. 

Permissible  space  does  not  allow  one  to  do  justice  to  the  articles  cited. 

PRELIMINARY  WORK 

Further  study  of  esterification  was  undertaken  with  a  view  of  secur- 
ing a  method  for  determining  the  percentage  of  another  insoluble  acid 
besides  stearic  and  oleic  in  butter  fat.  Laurie  acid  was  naturally  the 
most  promising,  on  account  of  the  lower  boiling  point  of  its  esters, 
although  myristic  acid  was  also  considered  a  possibility.  The  per- 
centage of  lauric  acid,  together  with  ordinary  analytical  data  and  the 
amount  of  stearic  acid  determined  by  crystallization,  would  permit  a 
satisfactory  calculation  of  the  remaining  insoluble  acids. 

Material. — At  the  outset  the  insoluble  acids  were  employed  for 
esterification  with  an  idea  that  the  previous  elimination  of  soluble  acids 
and  of  glycerol  would  be  an  advantage.  In  reality  such  did  not  prove 
the  case,  as  water,  a  limiting  factor  in  esterification,  was  produced  in 
the  reaction  between  fatty  acids  and  alcohols,  while  glycerol  was  pro- 
duced in  the  case  of  fats  (glyceryl  esters)  and  alcohols  as  shown  bv  the 
following  equations : 

RCOOH  -f-  RiOH  -^catalyzer=RCOOR^-^  H.,0 
fatty  acid    alcohol  ester        water 

(RCOO),C3H54-3RiOH-f-catalyzer=3RCOOR,-|-C3H5(OH), 
fat  alcohol  ester  glycerol 

'  Cliem.  iVhs..  v.  6.  no.  4,  p.  5.55.     lorj. 


722 


Journal  of  Agricultural  Research 


Vol.  XII,  Ne.  II 


F'urthennore,  the  employment  of  the  original  product  was  preferable 
from  an  analytical  standpoint  to  say  nothing  of  the  time  and  labor 


Fig.  1. — Apparatus  employed  in  esterification. 


involved  in  the  preparation  of  a  large  stock  of  insoluble  acids.     Butter 
fat  was  used  in  all  subsequent  work. 


Mar.  38.  igts  Fatty  Acids  in  Butter  Fat  723 

EsTERiFiCATiON  (fig.  i). — Earlier  investigators  have  indicated  various 
methods  of  esterification.  Absolute  alcohol  with  a  mineral  acid  catalyzer 
appeared  the  most  practicable  for  analytical  purposes  and  was  adopted 
tentatively.  Between  methyl  and  ethyl  alcohols  there  was  little  choice, 
except  as  to  cost  and  convenience,  although  the  methyl  esters  have  a 
somewhat  lower  boiling  point.     Ethyl  alcohol  was  used  exclusively. 

The  preparation  of  dry  ethyl  alcohol  substantially  free  from  aldehydes 
was  found  considerable  of  a  problem.  For  dehydration  neither  metallic 
calcium  nor  any  quicklime  from  the  usual  sources  proved  efficient.  A 
granulated  caustic  lime  containing  about  95  per  cent  of  calcium  oxid  was 
eventually  obtained  from  the  manufacturers  which  would  produce  a  dry 
alcohol  on  the  third  distillation. 

Different  schemes  have  been  suggested  for  the  removal  of  aldehydes, 
such  as  oxidation  with  silver  nitrate,  potassium  permanganate,  or 
potassium  bichromate,  polymerization  with  caustic  alkali  or  fractional 
distillation,  and  all  were  tried  in  some  form  or  other.  The  following 
process  finally  proved  satisfactory  and  w^as  adopted : 

Approximately  2  liters  of  alcohol  w^ere  fractionated  in  a  water  bath 
over  600  to  700  gm.  of  caustic  lime  and  2  to  3  gm.  of  caustic  soda.  The 
main  portion  of  the  distillate,  the  first  and  last  being  rejected,  w^as  refrac- 
tionated  twice  in  a  similar  manner  over  fresh  lime  and  soda.  The 
rejected  portions  were  united  and  retreated. 

Dry  hydrochloric  acid  or  concentrated  sulphuric  acid  has  been  the 
catalyzer  almost  invariably  employed  by  different  workers  for  esterifica- 
tion. The  former  with  a  greater  hydrogen-ion  concentration  is  apparently 
less  efficient,  gram  for  gram,  than  the  latter,  and  is  generally  used  in 
larger  amounts.  This  may  be  due  in  part  to  loss  of  hydrochloric  acid 
as  ethyl  chlorid  by  volatilization,  but  more  likely  to  the  dehydrating 
effect  of  sulphuric  acid.  By  using  a  larger  quantity  of  alcohol,  together 
with  certain  chlorids,  the  yield  of  esters  with  hydrochloric  acid  may  be 
increased,  as  shown  by  Phelps. 

A  considerable  excess  of  alcohol  is  also  required  to  insure  the  necessary 
mass  action  irrespective  of  its  dehydrating  action.  For  esterif\ing  150 
gm.  of  butter  fat,  400  c.  c.  of  absolute  ethyl  alcohol  containing  8  gm.  of 
dry  hydrochloric  acid  or  4  c.  c.  of  concentrated  sulphuric  acid  were 
employed.  This  amount  of  alcohol  furnished  about  1 1  times  that  needed 
for  combination  with  the  fatty  acids. 

The  use  of  a  neutral  solvent,  such  as  ether,  did  not  appear  to  accelerate 
esterification  and  w^as  omitted  after  a  few  trials.  Butter  fat  diffused 
rapidly  through  the  acid  alcohol  on  boiling,  and  the  solution  generally 
cleared  in  a  few  minutes.  Short  boiling  periods  were  tried,  but  24  hours 
were  considered  more  reliable  and  adopted. 

Purification  of  esters. — After  completing  the  esterification  and 
cooling  the  solution,  the  esters  must  be  precipitated  and  freed  from 
mineral  acid,  glycerol,  and  excess  alcohol. 


724 


Journal  of  Agricultural  Research 


Vol.  Xlf,  No.  II 


Attention  has  already  been  called  to  various  methods  of  neutralizing, 
purifying,  and  drying  the  esters,  as  described  by  other  workers.  Most 
of  these  processes  and  innumerable  modifications  were  given  careful  trial, 
but  the  readiness  with  which  at  least  a  portion  of  the  butter-fat  esters 
hydrolyzed  precluded  the  use  of  water  in  their  purification.  This  was  an 
extremely  exacting  condition  and  called  for  a  salt  soluble  in  alcohol, 
neutral  in  reaction,  and  with  dehydrating  properties.     Dry  magnesium 


Fig.  3. — Apparatus  employed  in  fractionation. 

chlorid  satisfactorily  met  these  requirements,  and  furthermore  was  cheap 
and  easily  procured.  On  the  addition  of  the  dry  salt  in  the  presence  of 
ether,  a  rapid  separation  of  esters  was  obtained.  The  underlying  solu- 
tion was  drawn  off  by  means  of  a  separate ry  funnel,  and  the  esters  were 
purified  by  "shaking  out"  several  times  with  ether  and  a  saturated 
alcoholic  solution  of  magnesium  chlorid.  The  ether  facilitates  the 
separation  and  shouid  always  be  added  before  the  magnesium  chlorid. 


Mar.  1 8,  1918 


Faity  Acids  in  Butter  Fat 


725 


All  the  agitation  necessary  can  be  obtained  by  reversing  the  separatory 
funnel  several  times,  which  gives  a  clear  separation  much  quicker  than  a 
more  violent  shaking.  The  resulting  esters  contain  ether  and  probably 
some  alcohol,  but  do  not  require  neutralizing  or  drying. 

Fractionation  (fig.  2). — Fractionation  of  the  esters  in  vacuo  by  means 
of  a  Bruehl  or  other  type  apparatus  was  found  impracticable  as  a  quanti- 
tative process  for  the  reason  that  a  constant  level  of  the  liquid  in  the  dis- 
tilling flask  and  a  constant  pressure  were  prime  requisites  for  a  definite 
fraction,  neither  of  which  co^ld  be  successfully  maintained  with  the 
facilities  at  hand.  Furthermore,  the  use  of  such  apparatus  necessitates 
a  certain  aptitude  or  technic  not  possessed  by  all  analysts. 

Fractionation  at  atmospheric  pressure  required  a  high  temperature, 
but  proved  feasible.  Gas  could  not  be  used  as  a  direct  source  of  heat 
owing  to  fluctuations  in  pressure,  influence  of  air  currents,  tendency  to 
decompose  the  esters,  etc.,  but  when  applied  to  a  bath  of  superheated 
valve  oil  proved  entirely  satisfactory.  This  oil  is  a  cheap  commercial 
product  that  will  safely  withstand  a  temperature  of  OA^er  400°  C.  when 
covered  and  may  be  used  three  times,  possibly  more,  without  appreci- 
able loss  of  efficiency.  All  the  exposed  surface  of  the  side-neck  flask 
should  be  wound  with  asbestos  paper  to  prevent  chilling  the  vapors  and 
breakage  of  the  flask. 

For  some  reason  fractionation  of  the  purified  esters  proved  imprac- 
ticable, possibly  due  to  the  influence  of  a  relatively  large  amount  of  the 
higher  esters  particularly  oleic,  but  was  readily  accomplished  after  a  pre- 
liminary distillation.  Accurate  results  are  dependent  in  large  measure 
on  a  slow,  steady  rise  in  temperature  during  the  first  distillation  and 
subsequent  fractionation.     Glass  beads  were  found  helpful  in  boiling. 

T.'^BLE  I. — Boilinq  point  and  range  of  fractions  {uncorrected)  of  esters  of  butter  fat 


Ester. 

Boiling  point. 

No. 

I 
2 

3 
4 
5 
6 

Esters. 

Range  of 
I'ractioiij. 

Ethyl  butyrate         .... 

"C. 
119.  9-1 2 1 
165      -167 
205      -208 

243      -245 
269 

Ethyl  butyrate,   caproate,  and 
oleate 

'C. 

Ethyl  caproate    

I25-IS0 
180—221; 

Ethyl  caprylate 

Ethyl  caprate 

Ethyl  caproate,  caprylate,  and 
oleate 

Ethyl  laurate 

Ethyl   caprylate,   caprate,   and 
oleate 

Ethyl  myristate 

Ethyl  palmitate 

Ethyl  stearate 

22C-270 

Ethyl     caprate,     laurate,     and 
oleate 

270-300 

Ethyl  oleate .... 

Ethyl    laurate,    myristate,    and 
oleate 

•^oo— •;2'; 

Ethyl  myristate,  palmitate,  and 
oleate 

325-365 

The  range  of  the  several  fractions  is  more  or  less  arbitrary,  being  de- 
pendent on  the  speed  of  distillation  and  the  distance  the  vapors  have  to 
rise.  The  object  was  to  secure  fractions  that  did  not  contain  more  than 
two  esters  in  addition  to  oleic  ester,  and,  furthermore,  adjacent  fractions 
should  contain  approximately  one-half  of  the  saturated  ester  appearing 


726 


J  Of.  r  rial  of  AgricuJiiiYal  Research 


Vol.  Xn,  No.  II 


in  each.  The  reported  boiling  point  of  the  ethyl  ester  of  a  number  of 
fatty  acids  that  occur  in  butter  fat,  together  with  the  range  of  the  sev- 
eral fractions  as  determined  by  analysis  for  a  "high"  side-tube  500-c.  c. 
distillation  flask,  may  be  noted  in  Table  I. 

The  apparent  lack  of  agreement  is  probably  due  to  the  greater  distance 
the  vapors  have  to  rise  in  practical  fractionation  and  to  the  influence  of  a 
gradually  increasing  amount  of  ethyl  oleate  in  the  several  fractions. 
Analysis  seemed  to  be  the  only  method  for  accurately  establishing  the 
required  range.  Hempel  tubes  or  similar  apparatus  could  not  be  em- 
ployed to  break  up  the  distillate. 

As  the  preliminary  work  advanced,  the  results  became  more  concord- 
ant and  indicated  that  it  was  possible  to  determine  not  only  lauric  and 
myristic  acids  but  also  caproic,  caprylic,  and  capric  acids.  A  part  of 
the  butyric  acid  was  recovered,  but  the  main  portion  was  evidently  lost 
during  the  purification  of  the  esters  or  distilled  over  with  the  ether,  owing 
to  the  greater  solubility  and  volatility  of  this  ester. 

Analysis  of  fractions. — Only  the  determinations  of  saponification 
and  of  iodin  numbers  by  the  usual  methods  employed  wth  oils  and  fats 
were  required. 

Calculation  of  results. — Having  determined  the  weight,  saponifi- 
cation and  iodin  numbers  of  the  several  fractions,  the  analyst  must 
ascertain  whether  the  range  of  the  fractions  had  been  accurately  estab- 
lished. If  correctly  fractionated,  the  percentage  of  the  different  ethyl 
esters  in  the  several  fractions  can  be  calculated  algebraically  and  their 
weight  computed,  from  which  the  percentage  of  the  corresponding  acids 
in  the  butter  fat  can  be  determined. 

As  a  matter  of  convenience,  the  data  necessary  for  these  calculations 
have  been  compiled  in  Table  II. 

Table  II. — Fatty  acids  and  their  ethyl  esters 
[C,  12.005;  H,  1.008;  0, 16.000;  K,  39.10;  1, 1^6.92] 


Add. 


Molecular 

weight  of 

acid. 


Molecular 

•weight  of 

ester. 


Saponifica- 
tion No. 
of  ester. 


Iodin  No. 
of  ester. 


Recipro- 
cal. 


Conver- 
sion fac- 
tor, ester 
to  acid. 


Acetic '  60.  042 

Butyric [  8S.  084 

Valeric ;  102.  105 

Caproic 116.  126 

Caprylic 1  144-  168 


Capric . 

Lauric 

^lyristic 

Palmitic 

Stearic 

Arachic 

Oleic 

Erucic 

Linolic 

Linolenic 

Clupanodonic .  .  . 

Ricinoleic 

Dihydroxystearic . 


172.  210 
200.  252 
228.  294 
256-  2,3^ 
284.  378 
312.  420 
282.  362 
338.  446 
280.  346 
278.  330 
276.314 
298.  362 
316.378 


88.084 
116.  126 
130.  147 
144.  168 
172.  210 
200.  252 
228.  294 

256-  z?,(> 

284.  378 
312.  420 
340.  462 
310.  404 
366.  488 
308.  388 
306.  372 
304-  356 
326.  404 
344.  420 


Mgm. 
636.  983 
483.  165 

431-  "3 
589.  185 
325.812 
280.  187 
245.  771 
218.  885 
197.  301 

179.  592 
164.  800 

180.  758 
153.096 

181.  940 

183.  137 

184.  350 
171.897 
162.  906 


0.81777 
. 69263 
I.  64624 
2. 48561 
3- 33609 
•  77769 


I.  22284 

I-  44377 
.60744 
.  40232 

•  29975 
I. 


68164 
75852 
78454 
80540 
83716 

85997 

87717 

, 89060 

90139 
, 91024 
, 91764 
,  90966 
,  92348 
,  90907 
,90847 
. 90786 
,91409 
,  91858 


Mar.  is,  191S  Fatty  A  cids  in  Butter  I- at  727 

As  an  illustration  of  the  method  of  calculation,  take  the  fifth  fraction 
of  a  sample  of  butter  fat  (Table  III)  vv^hich  weighed  50.69  gm.,  had  a 
saponification  number  of  219.280  and  an  iodin  number  of  14.055,  and 
was  intended  to  contain  ethyl  laurate,  myristate  and  oleate. 

The  iodin  number,  convertible  to  ethyl  oleate  by  the  factor  1.22284,  is 
equivalent  to  17.187  per  cent. 

Then  i.oo— 0.17187,  or  0.82813,  is  the  percentage  of  the  residual  esters 
in  the  fraction  and  2i'9.28o—  (0.17187X  180.758),  or  188.213,  their  alkali- 
consuming  power.  The  latter  divided  by  the  former  gives  a  saponifica- 
tion number  of  227,  a  figure  between  that  of  ethyl  laurate  (245.771)  and 
that  of  ethyl  myristate  (218.885),  ^-s  anticipated. 

Let  X  indicate  ethyl  laurate  and  y  ethyl  myristate,  then 

rc+r  =  0.82813 
245. 77ix+2i8.885>'=  188.213 

3;=  25.842  per  cent  of  ethyl  laurate 
)/=  56.971  per  cent  of  ethyl  myristate. 

The  percentages  0.25842,  0.56971,  and  0.17187  multiplied  by  the 
weight  of  the  fraction  (50.69  gm.),  gives  the  weight  of  the  several  esters 
in  the  fraction  13.099,  28.879,  ^^^  8.712,  which  by  their  respective  fac- 
tors 0.87717,  0.89060,  and  0.90966  may  be  converted  into  the  weight  of 
the  several  fatty  acids  in  the  fraction.  The  total  weight  of  a  fatty  acid 
divided  by  the  weight  of  fat  taken  (300)  gives  the  percentage  of  that 
acid  in  the  butter  fat.  The  same  method  of  calculation  is  followed  in 
each  of  the  other  fractions. 

METHOD   IN   DETAIL 

Reagents. — Alcohol :  absolute. 

Hydrochloric  acid :  dry,  generated  by  dropping  concentrated  sulphuric 
acid  into  a  mixture  of  concentrated  hydrochloric  acid  and  sodium  chlorid 
and  dried  by  passing  through  concentrated  sulphuric  acid. 

Sulphuric  acid:  concentrated,  heated  to  225°  C. 

Ethyl  ether:  anhydrous,  freshly  distilled  over  metalic  sodium. 

Magnesium  chlorid:  dry  powder,  neutral.     Dried  at  75°  to  80° C. 

Magnesium  chlorid  solution:  250  gm.  of  dry  powder  to  500  c.  c.  of 
absolute  alcohol. 

ESTERIFICATION  AND  PURIFICATION  OF  ESTERS. — Into  a  I,000-C.  C.  flat- 

bottom  globe  flask  are  brought  150  gm.  of  filtered  fat,  together  with 
400  c.  c.  of  alcohol  previously  charged  with  8  gm.  of  dry  hydrochloric 
acid,  or  4  c.  c.  of  concentrated  sulphuric  acid  and  a  number  (25)  of  glass 
beads.  The  flask  is  connected  with  a  long  spiral  or  other  form  of  reflux 
condenser,  and  the  mixture  carefully  boiled  on  50-nicsh  Nichrome  wire 
gauze  lor  24  hours. 


728  Journal  of  AgrimiHural  Research  voi.  xii,  No.  n 

After  esterification  the  contents  of  the  flask  are  cooled,  50  c.  c.  of 
ether  and  150  gm.  of  magnesium  chloride  added,  rotated  to  hasten 
saturation,  transferred  to  a  i,ooo-c.  c.  pear-shaped  separatory  funnel 
and  allowed  to  stand  until  a  clear  separation  is  secured.  The  underlying 
layer  is  drawn  off  into  the  original  flask  and  the  esters  carefully  shaken 
out  two  or  three  times  with  25  to  50  c.  c.  of  ether  and  50  c.  c.  of  an  alco- 
holic solution  of  magnesium  chlorid.  Violent  shaking  causes  a  slow 
separation.  After  the  removal  of  the  final  washing,  the  clear,  purified 
esters  are  filtered  through  a  firm,  close-textured  paper  into  a  500-c.  c. 
round-bottom  "low"  side-tube  distillation  flask.  The  filter  is  extracted 
with  ether  which  is  run  into  the  original  globe  flask  containing  the 
alcoholic  layer  and  washings  from  the  esters.  More  ether  is  added  until 
a  ready  separation  is  obtained  and  the  solution  again  transferred  to 
the  separatory  funnel  and  allowed  to  stand  several  hours  to  recover  any 
occluded  esters  which  are  then  washed  several  times  vath  ether  and 
magnesium  chlorid  solution  as  previously  described,  and  filtered  into 
the  distillation  flask  containing  the  first  portion. 

Fractionation. — A  number  (50)  of  glass  beads  are  placed  in  the 
side-neck  distillation  flask  which  is  connected  with  a  12-inch  Liebig 
condenser  and  heated  in  a  bath  of  superheated  valve  oil  for  distillation 
of  the  esters.  The  exposed  portion  of  the  flask  should  be  covered  with 
asbestos  paper  and  the  condenser  filled  with  cold  water  at  the  outset 
but  no  circulation  should  be  permitted  during  the  distillation.  The 
temperature  should  be  raised  slowly.  After  the  ether  and  alcohol  are 
expelled,  the  entire  distillate  between  85°  and  ^65°  C.  is  collected  and 
constitutes  from  no  to  120  gm.  with  butter  fat. 

The  distillate  from  two  portions  representing  300  gm.  of  butter  fat 
and  a  number  (50)  of  glass  beads  are  brought  into  a  500-c.  c.  round- 
bottom  "high"  side-tube  distillation  flask  which  is  connected  with  a 
Liebig  condenser  and  heated  as  previously  described.  Particular  care 
should  be  exercised  in  heating  the  oil  bath  so  as  to  insure  a  slow,  steady 
rise  in  temperature  taking  at  least  80  minutes  from  the  beginning  of 
the  first  fraction  to  the  completion  of  the  last.  The  required  range  of 
every  fraction  must  be  accurately  established  with  the  apparatus  em- 
ployed by  analysis  (See  Table  I). 

The  volatility  of  the  esters  and  the  readiness  with  which  they  hydrolyze 
necessitates  careful  treatment  of  the  fractions,  which  should  be  collected 
in  tared  flasks,  weighed,  and  the  saponification  and  iodin  numbers 
determined  as  soon  as  possible. 


Mar.  18,  1918 


Fatty  Acids  in  Butter  Fat 


729 


APPLICATION   OF  THE  METHOD 

A  sample  of  dry  filtered  butter  fat  churned  from  sweet  cream  from 
mixed  milk  of  the  Experiment  Station  herd  was  taken  for  examination. 
An  analysis  of  the  fat  and  of  its  insoluble  acids  gave  the  following  results, 
which  indicate  a  normal  product. 

FAT 

Saponification  number mgm .  .  231.  453 

Acid  number mgm. .  2.  183 

Ether  number  (e) mgm.  .  229.  270 

lodin  number 27.  999 

Equivalent  in  oleic  acid per  cent. .  31.  145 

Total  fatty  acids  (1.00—0.00022594  e) per  cent. .  94.  819 

Insoluble  fatty  acids  (by  alcoholic  saponification) per  cent. .  87.  500 

Soluble  fatty  acids  (by  difference) per  cent. .  7-  319 

Glycerol  (0.00054703  e) per  cent.  .  12.  542 

INSOLUBLE  ACIDS 

Neutralization  number mgm.  .  221.  890 

lodin  number 28.  125 

Equivalent  in  oleic  acid per  cent.  .  31.  285 

Stearic  acid  (by  crj'stallization) percent.  13.  010 

The  fat  was  esterified,  with  dry  hydrochloric  acid  as  a  catalyzer,  the 
esters  purified  and  fractionated  as  described  in  the  method. 

In  Table  III  are  given  the  range,  weight,  saponification,  and  iodin 
numbers  of  the  several  fractions.  From  these  data  the  percentage  and 
weight  of  the  different  esters  in  the  fractions  were  calculated  (Table  lY) 
and  the  weight  of  the  corresponding  acids  and  their  percentage  computed 
on  the  basis  of  the  original  fat  (Table  Y). 

Table  III. — Weight  and  analysis  of  fractions 


No. 

Range  of 
fraction. 

Weight  of 
fraction. 

SaponiSca- 
tjon  No. 

lodin 
No. 

Ethyl 
oleate. 

I                                                  .          

"C. 

125-180 
180-225 
225-270 
270-300 
300-325 
325-365 

Gm. 

5-  0835 

4.  4400 

6.  8855 

15.  2990 

50.  6900 

84.  8800 

Mgm. 
417.  209 
346.  396 
279.  Ill 
240.  983 
219.  280 

205-  955 

4-753 
8.  091 

11.  471 

12.  846 
14-055 
16.  691 

Per  cent. 
5.812 
9.894 
14.027 

15-  709 
17.  187 

2                                                            

■3             .                                              

A                                                                           

6 

20.  410 

Table  IV. — Percentage  and  weight  of  esters  recovered 


No. 

Ethyl 
butyrate. 

Ethyl 
caproate. 

Ethyl 
caprylate. 

Ethyl 
ca  prate. 

Ethyl 
laurate. 

Ethyl              Ethyl 
myristate.      palmitate. 

1 

Ethyl 
oleate. 

P.ct. 

42.708 

Gm. 

2.  171 

P.ct.  Gm. 

51.480  2.  617 
SS.  128  2.  448 

P.ct. 

Gm. 

P.ct. 

Gm.. 

P.ct. 

Gm. 

P.ct. 

Gm.     P.  ct. 

Gm. 

P.cU 

5.812 
9.894 
14.  027 
15-709 
17.187 
20. 410 

Gm. 
0.29S 

34-978 
28.  210 

I- 553 
1.942 

57-  763 

IS-  763 

3-977 
2.412 

.966 

68.528 
25.  842 

10. 484 
13.099 

3. 403 

56.  971 

55.  736 

2d.  879 

8.  712 

6 

::::.. 1 

47.309 

23-  854 

20.  247 

17.324 

•••   1 

Total 

2.  171 



5-065 

3-495 

16.389 

'23-583 

76.188 

20.247 

3°-  139 

730 


Journal  of  Agricultural  Research  voi.xii.No.it 


Table  V 

. — Weigh 

and  percentage  of  fatty  acids  recovered 

Acid. 

Gm. 

Per  cent. 

Acid.                               Gm. 

Per  cent. 

Butyric  acid  (partial  re- 
covery)  

1.647 
4,  080 
2.  926 

5-494 
20.  686 

0-  549 
1.360 

•975 
I.  831 
6.89s 

Myristic  acid 67.  853 

Palmitic   acid   (partial 

recovery) iS.  250 

Oleic  acid  (partial  re- 

22. 6r8 

Caproic  acid 

6.  083 

Caprylic  acid 

covery) 

27.416 

9-  139 

Laurie  acid 

1 
1 

The  recovery  of  ethyl  butyrate  was  incomplete,  owing  to  its  high 
solubility  and  volatility;  that  of  ethyl  palmitate  was  due  to  inability  to 
continue  the  distillation,  owing  to  insufficient  volume  of  higher  esters  and 
excessive  heat  requirements.  Ethyl  oleate  appeared  in  all  fractions, 
gradually  increasing  with  the  temperature. 

The  percentage  of  the  different  fatty  acids  in  the  butter  fat  are 
presented  in  Table  VI.  Butyric  acid  of  the  soluble  acids  and  palmitic 
acid  of  the  insoluble  were  determined  by  difference,  stearic  acid  by 
crystallization,  as  previously  stated,  and  oleic  acid  from  the  iodin  number 
of  the  insoluble  acids.  No  allowance  was  made  for  unsaponifiable  matter 
in  the  calculation. 

The  alkali-consuming  and  glycerol -combining  powers  of  the  several 
acids  are  also  recorded  and  confirm  the  results  secured  in  large  measure. 
The  glycerol  requirements  of  the  constituent  acids  slightly  exceed  that  of 
the  fat,  for  the  reason  that  no  allowance  was  made  for  the  free  fatty  acids 
present. 

Table  VI. — Percentage  of  fatty  acids  in  butter  fat 


Fatty  acids. 


Amount 
present. 


Sapoaifi- 

cation 
niuuber. 


Glycerol. 


Soluble  acids: 
Butyric  acid.  , 
Caproic  acid. . 
Caprylic  acid . 
Capric  acid. . . 


Total 

Calculated . 


Insoluble  acids: 
Laurie  acid .  . 
Myristic  acid. 
Palmitic  acid. 
Stearic  acid.  .. 
Oleic  acid .  .  . 


Per  cent. 

"■o-  153 
I.  360 

•975 
I.  831 


Mgm. 
20.  084 

6-  571 
3-  795 
5.966 


Per  cent. 
1.099 

■359 
.208 
.326 


7-319 


36.  416 
37-  299 


1.992 


Total 

Determined , 


Soluble  and  insoluble  acids: 

Total 

Determined  or  calculated  from  fat. 


6.89s 

22.  618 

°'  19.  229 

11.384 

27-374 


87.  500 


94.  819 


19-  319 
55-  588 
42.  089 
22.  461 
54-  395 


1-057 
3.041 
2.302 

1.  229 

2.  976 


193.  852 

194.  154 


10.  605 


230.  268 
231-453 


12.  597 
12.  542 


a  By  difference. 


Mar.  18, 1918  Fatty  Acids  in  Butter  Fat  731 

In  a  later  article  attention  will  be  called  to  additional  analyses  of 

butter  fat  and  to  the  results  reported  by  other  workers  employing  different 

methods. 

SUMMARY 

The  direct  esterification  of  butter  fat  v/ith  subsequent  fractionation 
of  the  resulting  esters  has  proved  an  accurate  and  practical  method  for 
the  determination  of  five  of  the  fatty  acids  in  butter. 

LITERATURE  CITED 

1)  Abderhalden,  Emil,  and  Kautzsch,  Karl. 

1912.  VERSUCHE      iJBER      VERESTERUNG      VON      MONOAMINOSAUREN      MITTELS 

JODATHYL.     In  Ztschr.  Physiol.  Chem.,  Bd.  78,  Heft  2,  p.  115-127. 

2)  Bertkelot,  Marcellin. 

1S54.    SUR  LES  COMBINAISONS   DE   LA   GLYCERINE   AVEC   LES   ACIDES    ET   SUR  LA 
SYNTHASE    DES    PRINCTPES    IMM^DTATS     DES    GRAISSES     DES     ANIMAUX. 

In  Ann.  Chim.  et  Phys.,  s.  3,  t.  41,  p.  216-319,  pi.  2. 

3)  Elsdon,  G.  D. 

1913.  ALCOHOLYSis  AND  THE  COMPOSITION  OP  cocoANUT  OIL.     In  Analyst,  V.  38, 

no.  442,  p.  8-II. 

4) 

1914.  THE  COMPOSITION"  OP  PALM-KERNEL  OIL.     In  Analyst,  v.  39,  no.  455, 

p.  78-80. 

5)  Fischer,  Emil,  and  SpEiEr,  Arthur. 

1895.  DARSTELLUNG  DER  ESTER.     In  Ber.  Deut.  Chem.  Gesell.,  Jahrg.  28,  No.  19, 
P-  3252-3253. 

6)  Grandmougin,  E.,  Havas,  E.,  and  Guyot,  G. 

I913.    METHYLIERUNG      ALIPH.\TISCHER      VERBINDUNGEN      MITTELS      DIMETHYL 

SULFATS.     In  Chem.  Ztg.,  Jahrg.  37,  No.  81,  p.  812-813. 

7)  GussEROW,  C.  A. 

1828.  ueber  die  einwirkung  des  bleioxyds  auf  die  organischen  korper, 

WELCHE    IM    ALLGEMEINEN     UNTER     DIE     KLASSE     DER     FETTE     GESTELLT 
WERDEN   UND   DIE  DADURCH   ENTSTEHENDEN   VERBINDUNGEN.      Ill  Afch. 

Apoth.  Ver.  Nordl.  Teutschland,  Bd.  27,  Heft  2,  p.  153-244. 

8)  HallER,  a. 

1906.  SUR  l'alcoolyse  DES  CORPS  GR.^s.     In  Compt.  Rend.  Acad.  Sci.  [Paris], 
t.  143,  no.  19.  p.  657-661. 

0)  - 

ic) 


1907.  alcoolyse  DE  l'huile  dE  ricin.     In  Compt.  Rend.  Acad.  Sci.  [Paris], 
t.  144,  no.  9,  p.  462-466. 


1908.  alcoolyse  DE  l'huile  DE  LIN.     In  Compt.  Rend.  Acad.  Sci.  [Paris], 
t.  146,  no.  6,  p.  259-262. 

i:)  and  Youssoufian. 

1906.  ALCOOLYSE  DU  BEURRE  DE  coco.     In  Compt.  Rend.  Acad.  Sci.  [Paris] 
t.  143,  no.  22,  p.  803-806. 
12')  Hauser,  O.,  and  Klotz,  A. 

I913.    K.\T.\LYTISCHE     REAKTIONSBESCHLEUNIGUNG     MITTELS     BERYLLIU.MVER- 
BINDUNGEN    BEI    DARSTELLUNG    VON    ESTERN    ORGANISCHER    SAUREN. 

In  Chem.  Ztg.,  Bd.  37,  No.  15,  p.  146-148. 
13)  Holland,  E.  B. 

191 1.  PURIFICATION  OF  INSOLUBLE  FATTY  ACIDS.     In  Jouf.  Indus.  and  Engin. 
Chem.,  V.  3,  no.  3,  p.  171-173. 

14')  Reed,  J.  C,  and  Buckley,  J.  P.,  Jr. 

1916.  DETERMINATION  OP  STEARIC  ACID  IN  butter  FAT.     /n  Jour.  Agr.  Research, 
V.  6,  no.  3,  p.  ior-113,  fig.  1-2.     Literature  cited,  p.  113. 


y-^2  Journal  of  Agricultural  Research  voi. xii, No.  n 

(15)  JVILLARD,  Paul. 

1895.  surl'acidE  dioxyst^arique  naturel.     In  Bui.  Soc.  Chim.  Paris,  s.  3 , 
t.  13,  p.  238-240. 

(16)  Kailan,  Anton. 

1914.  UBER  DIE  durch  chlorwasserstofp  katalysierte  esterbeldung  in 
LosungsmittelgemischEn.  In  Ztschr.  Phys.  Chem.,  Bd.  88,  Heft  i, 
p.  65-102. 

(17)  Meyer,  V.  J. 

1907.  UBER  DAS  BAUMWOLLSAMENOL.     In  Chem.  Ztg.,  Bd.  31,  No.  64,  p. 

793-794. 

(18)  Phelps,  I.  K.,  and  Hubbard,  J.  L. 

1907.  ON  THE  ESTERiFiCATiON  OF  SUCCINIC  ACID.  In  Amer.  Jour.  Sci.,  s.  4, 
V.  23,  no.  137,  p.  368-374- 

(19)  and  Phelps,  M.  A. 

1907.  THE   use  op  ZINC  CHLORIDE  IN   THE   ESTERIFICATION   OF  SUCCINIC   ACID. 

In  Amer.  Jour.  Sci.,  s.  4,  v.  24,  no.  141,  p.  104-196. 

(20) 

1908.  ON  THE  ESTERIFICATION  OF  MALONic  ACID.     In  Amer.  Jour.  Sci.,  s.  4. 

V.  26,  no.  153,  p.  243-252,  I  fig. 
(21) 

1908.  RESEARCHES  ON  THE  INFLUENCE  OF  CATALYTIC  .\GENTS  IN  ESTER  FORMA- 
TION. ON  THE  ESTERIFICATION  OF  CYANACETIC  ACID.  In  Amer.  Jour. 
Sci.,  s.  4,  V.  26,  no.  153,  p.  264-266. 

(22)  .    Palmer,  H.  E.,  and  SmilliE,  R. 

1908.  RESEARCHES  ON  THE  INFLUENCE  OP  CATALYTIC  AGENTS  IN  ESTER  FORMA- 
TION. THE  EFFECT  OF  CERTAIN  SULPHATES  ON  BENZOIC  AND  SUC- 
CINIC ACIDS.    In  Amer.  Jour.  Sci.,  s.  4,  v.  26,  no.  153,  p.  290-295. 

(23)  .     Phelps,  M.  A.,  and  Eddy,  E.  A. 

1908.  CONCERNING  THE  PURIFICATION  OF  ESTERS.    In  Amer.  Jour.  Sci.,  s.  4, 
V.  26,  no.  153,  p.  253-256. 
(24) 

1908.   RESEARCHES  ON  THE  INFLUENCE  OF  CATALYTIC  AGENTS  IN  ESTER  FORMA- 
TION.     THE   ESTERIFICATION  OF  BENZOIC    .A^CID    V/ITH    CERTAIN    CHLO- 
RIDES.    Li  Amer.  Jour.  Sci.,  s.  4,  v.  26,  no.  153,  p.  296-300. 

(25) 

1908.  RESEARCHES  ON  THE  INFLUENCE  OF  CATALYTIC  AGENTS  IN  ESTERFOR- 
MATION.  HYDROBROMIC  ACID  AND  ZINC  BROMIDE  IN  THE  FORMA- 
TION OF  ETHYL  BENZOATE.  In  Amer.  Jour.  Sci.,  s.  4,  v.  26,  no. 
153,  p.  281-289. 

(26) and  Osborne.  R.  W. 

1908.  ON  THE  ESTERIFICATION  OF  BENZOIC  ACID.  In  Amer.  Jour.  Sci.,  s.  4, 
V.  25,  no.  141,  p.  39-48. 

(27)  Renard,  A. 

1871.   RECHERCHE       ET      DOSAGE      DE      l'hUILE      D'aRACHIDE      DANS      l'hUILE 

d'olivE.    In  Compt.  Rend.  Acad.  Sci.  [Paris],  t.  63,  no.  25,  p.  1330- 
1332  • 

(28)  ROCHLEDER,   F. 

1846.  UEBER  DAS  GLYCERIN.  In  Ann.  Chem.  u.  Pharm.,  Bd.  59,  Heft  2, 
p.  260-261. 

(29)  VarrEntrapp,  Franz. 

1840.  UEBER  DIE  OELSAURE.  In  Ann.  Chem.  u.  Pharm.,  Bd.  35,  Heft 
2,  p.  196-215. 

(30)  Wolff,  H.,  and  Scholze,  E. 

I914.   t)BER     KOLOPHONIUMBESTJMMUNG     in     FIRNISSEN,     OLEN     UND     SEIFEN. 

In  Chem.  Ztg.,  Jahrg.  38,  No.  34,  p.  369-370;  No.  35,  p.  382-383. 


Vol.  XII  MARCH  25,  1918  No.  12 

JOURNAL  OF 

AGRICULTURAL 

RESEARCH 


CONTENTS  AND  INDEX 
OF  VOLUME  XII 


PUBUSHED  BY  AUTHORITY  OF  THE  SECRETARY  OF  AGRICULTCRE. 

WITH  THE    COOPERATION   OF  THE   ASSOCIATION  OF  AMERICAN 

AGRICULTURAL  COUEGES  AND  EXPERIMENT  STATIONS 


WASHINGTON,   D.   C. 


WASMINOTONt  OOVCRNMCNT  PmNTUM  OFFIOE  I IMI 


i 

EDITORIAL  COMMITTEE  OF  THE 

UNITED  STATES  DEPARTMENT  OF  AGRICULTURE  AND 

THE  ASSOCIATION  OF  AMERICAN  AGRICULTURAL 

COLLEGES  AND  EXPERIMENT  STATIONS 


FOR  THE  DEPARTMENT 
KARL  F.  KELLERMAN,  Chairman 

Physiologist  and  Associate  Chief,  Bureau 
of  Plant  Industry 

EDWIN  W.  ALLEN 

Chief,  Office  of  Experiment  Stations 


CHARLES  L.  MARLATT 

Enfomoiogist  and  Assistant  Chief,  Bttreau 
of  Entomology 


FOR  THE  ASSOCIATION 

RAYMOND  PEARL* 

Biologist,  Maine  Agricultural  Experiment 
Station 

H.  P.  ARMSBY 

Director,  Institute  of  A  nim  al  Nutrition,  The 
Pennsylvania  State  College 

E.  M.  FREEMAN 

Botanist,  Plant  Pathologist,  and  Assistant 
Dean,  Agricultural  Experiment  Station  of 
the  University  of  Minnesota. 


All  correspondence  regarding  articles  from  the  Department  of  Agriculture  should  be 
addressed  to  Karl  F.  Kellerman,  Journal  of  Agricultural  Research,  Washington,  D.  C. 

*  Dr.  Pearl  has  tmdertaken  special  work  in  connection  with  the  war  emergency; 
therefore,  until  further  notice  all  correspondence  regarding  articles  from  State  Experi- 
ment Stations  should  be  addressed  to  H.  P,  Armsby,  Institute  of  Animal  Nutrition, 
State  College,  Pa. 


INDEX 


Abelmoschus  csculentus —  Page 

host  plant  of — 

,        Fusarium  vasiiifectutn 529-546 

Veriicillium  albo-airum 530-546 

wilt  diseases  of 529-546 

Acetic  acid.    See  Acid,  acetic. 

Acetone,  toxicity  to  insect  eggs 5S1-5S7 

Acid — 

acetic,  toxicity  to  insect  eggs 581-582 

butyric — 

percentage  in  butter  fat 726-731 

toxicity  to  insect  eggs 581 

capric,  percentage  in  butter  fat 726-731 

caproic,  percentage  in  butter  fat 726-731 

caprylic,  percentage  in  butter  fat 726-731 

fatty,  determination  in  butter  fat 719-732 

formation  in  corn-stover  silage 591-592 

lauric,  percentage  in  butter  fat 726-731 

myristic,  percentage  in  butter  fat 726-731 

oleic,  percentage  in  butter  fat 726-731 

palmitic,  percentage  in  butter  fat 726-731 

phosphate,  effect  on  soil  reaction 27-28 

phosphoric,  in  soil,  solubility 674-676 

stearic,  percentage  in  butter  fat 729-731 

valeric,  toxicity  to  insect  eggs 581 

Acidity,  soil,  definition  of 139 

Alcohol — 

allyl,  toxicity  to  insect  eggs 581 

amyl,  toxicity  to  insect  eggs 581 

benzyl,  toxicity  to  insect  eggs 581 

ethyl,  toxicity  to  insect  eggs 581-586 

methyl,  toxicity  to  insect  eggs 581-586 

Alfalfa.    See  Medicago  sativa. 
Allyl— 

alcohol,  toxicity  to  insect  eggs 581 

isosulphocyanate,  toxicity  to  insect  eggs. . .       581 
Ammonia — 

effect  of  nitrates  on  production  of 200-201 

formation  in  Wisconsin  soil 486-487 

Ammonium — 
nitrate — 
influence  on — 

Bacillus  radicicola 210-211 

growth  of  Azotobacter 190 

nodule  formation 219-220 

sulphate — 
effect  on — 
hydrogen-ion  concentration  of  soil-film 

water 28-30 

solubility  of  calciixm  in  soil 674 

solubility  of  phosphoric  acid  in  soil.  .  . .       674 
Amygdalus  pcrsica,  food  plant  of  Cer otitis 

capitate 105 

Amyl — 

acetate,  toxicity  to  insect  eggs 581 

alcohol,  toxicity  to  insect  eggs 581 

nitrite,  toxicity  to  insect  eggs 581 

Ancylostoma — 

caninum,  in  dogs,  effect  of  anlhclmintics 

on 399ff 

duodenale.  in  man,  effect  of  anthelmintics  on       415 


Page 

Anilin,  toxicity  to  insect  eggs 581 

Anthelmintic,  efficacy  of 397~477 

Apple.    See  Malus  sylvestris. 

Apple-blister 132 

Areca  catechu,  use  as  anthelmintic 419 

Areca  nut.    See  Areca  catechu. 

Ascarid.    See  Ascaris  suum;  Belascaris  cati; 

Belascaris  marginata;  Toxascaris  limbata. 
Ascaridia  perspicillum,  in  chickens,  effect  of 

anthelmintics  on 42sff 

Ascaris  suum,  in  hogs,  effect  of  anthelmintics 

on 4oiff 

A  vena  sativa,  relation  to  injury  by  Mayetiola 

destructor 519-527 

Averrhoa  carambola.iood  plant  of  Ceratitis  capi- 

tala 105 

Azotobacter — 
influence  of  nitrates  on — 

fLxation  of  nitrogen  by 193-201 

formation  of  volutin  bodies  in 205-208 

growth 187-208 

production  of  pigment  by 203-205 

influence  on  nitrates  in  solution  by 200-203 

Bacillus — 
arnylovorus,  causal  organism  of  fireblight. . .       130 
azotobacter — 
influence  of — 

calcium  carbonate  on 490-495 

limestone  on 490-495 

magneshim  carbonate  on 490-495 

radicicola — 
influence  of — 

calcium  carbonate  on 496-497 

limestone  on 496-457 

magnesium  carbonate  on 496-497 

nitrates  on — 
fixation  of  atmospheric  nitrogen  by.  214-217 

growth 20S-226 

infection  by 220-222 

production  of  gum  by 217-219 

influence  on  nitrates  in  solution  by. .   21 1-2 14 
Bacteria — 

cause  of  tobacco  wildfire 449-458 

influence  of — 

calcium  carbonate  on 469-499 

calcium  chlorid  on 479-480 

dibasic  magnesium  phosphate  on 479-499 

limestone  on 469-499 

magnesium  carbonate  on 477-499 

magnesium  chlorid  on 479-480 

monocalcium  phosphate  on 469-499 

nitrification  of,  effect  on  solubility  of  trical- 

cium  phosphate 671-683 

nitrogen-assimilating,  influence  of  nitrates 

on 183-230 

Bacterium  tabacum — 

causal  organism  of  tobacco  wildfire 454-457 

description 4S4-4SS 

Bailey,  C.   H.   and   Gurjar,   A.  M.   (jxiper): 

Respiration  of  StorcKl  Wheat 685-713 

Baldwin-spot.     See  Bitter-pit. 
Barley.    See  Hordeum  spp. 


(733) 


734 


Journal  of  Agricultural  Research 


Vol.  XII 


Batchelor,  L.  D.,  and  Reed,  H.  S.  (paper):  Page 
Relation   of  the  Variability  of  Yields  of 
Fruit    Trees    to    the    Accuracy    of    Field 
Trials 245-283 

Bechdel,  S.  I.     Corn-Stover  Silage 589-600 

Beetle,  potato.    See  Lepiinotarsa  decemlincala. 

Behavior  of  Sweet  Potatoes  in  the  Ground 
(paper) 9-^7 

Belascaris — 
cati,  in  cats,  effect  of  anthelmintics  on.  .  .   416,417 
viarginata,  in  dogs,  effect  of  anthelmintics 
on 399ff 

Benzaldehyde,  toxicity  to  insect  eggs 581 

Benzene,  toxicity  to  insect  eggs 581-582 

Benzonitrile,  toxicity  to  insect  eggs 581 

Benzyl  alcohol,  toxicity  to  insect  eggs 581 

Bestill.    See  Theveiia  neriifoUa. 

Black  myrobalan.     See  Tcrminalia  chebula. 

Bitter-pit— 

deFcription iio-iii 

effect  of  heavy  irrigation  on 126 

resemblance  of  other  spot  diseases  of  Malus 

sylveslris  to 130-i.^S 

resemblance  to  rosy-aphis  stigmonose. . . .  iio-iii 
susceptibility,  relation  of  size  of  Malus  syl- 
veslris to 126 

"  Blister,"  resemblance  to  cork  disease 134 

Blister-rust,  white-pine— 

caused  by  Cronarlium  ribicola 459 

dissemination 461-462 

Boiling  point,  index  to  toxicity  of  organic 
compounds 580-586 

Brazilian  plum.    See  Eugenia  braziliensis. 

Brommethylphenylketone,  toxicity  to  insect 
eggs 581 

Bromoform,  toxicity  to  insect  eggs 581-586 

Bromtoluene,  ortho,  toxicity  to  insect  eggs . .       581 

Bromxylene,  toxicity  to  insect  eggs 581 

Brooks,  Charles,  and  Fisher,  D.  F.  (paper): 
Irrigation  Experiments  on  Apple-Spot  Dis- 
eases   109-138 

Buckley,  jr.,  J.  P.,  and  Holland,  E.  B.  (pa- 
per): Determination  of  Fatty  Acids  in  But- 
ter Fat 719-732 

Bunostomum  Irigonocephaluvi,  in  sheep,  effect 
of  anthelmintics  on 404, 410, 412 

Burd,  J.  S.  (paper):  Water  Extractions  of 
Soils  as  Criteria  of  their  Crop- Producing 
Power 297-309 

Butter  fat,  determination  of  fatty  acids  in. .  719-731 

Butyric  acid.    See  Acid,  butyric. 

Calcium. — 
carbonate — 
effect  on — • 

solubility  of  calcium  in  soil 674 

solubility  of  phosphoric  acid  in  soil ....       674 

bacteria  in  Wisconsin  soils 469-499 

chlorid,  influence  on  bacteria  in  Wisconsin 

soils 479-480 

in  soil — 

effect  of  crop  growth  on 311-364 

effect  of  season  on 311-364 

solubility 674-676 

nitrate — 
influence  on — 

Bacillus  radicicola 209 

fixation  of  nitrogen  by  Azotobacter. .   198-199 

growth  of  Azotobacter 1S9, 192 

nodule  formation. . ; 219-220 


Calf—  Page 

digestion  of  starch  by S74-S78 

use  of  starch  food  in  addition  to  milk  for . .  577-578 

Calomel,  use  as  anthelmintic 399 

Calophyllutn  inophyllum,  food  plant  of   Cera- 

iitis  capHaia loj 

Capric  acid.     See  Acid,  capric. 
Caproicacid.     See  Acid,  caproic. 
Caprylic  acid.    See  Acid,  caprylic. 
Carbohydrate,  in  Ipomocc  batatas,  changes  in 

content 10-15 

See  also  Starch. 
Carbon — 

bisulphid,  toxicity  to  insect  eggs 581-586 

dioxid — 
accumulated,  influence  upon  respiration 

of  stored  wheat 706-708 

effect  of  gas  of  on  soil  solution 384-385 

effect  on  soil  reaction 140-145 

effect  on  water-soluble  nutrients  in  soil .  334-339 

tetrachlorid,  toxicity  to  insect  eggs 581-586 

Carpenter,  C.  W.  (paper);  Wilt  Diseases  of 

Okra  and  the  Vcrticillium-Wilt  Problem .  529-546 
Carruth,  F.  E.,  and  Withers,  W.  A.  (paper): 

Gossypol,  the  Toxic  Substance  in  Cotton- 
seed    83-102 

Casein,  hydrolysis  of  in  presence  of  starch, 

effect  of  time  of  digestion  on 1-7 

Castor  oil,  use  as  anthelmintic 399 

Cectun  worms.     See  Heterakis  papulosa. 
Ceratitis  capitata — 

control 104 

infestation  of  fruits  by  larvae  of 103-108 

parasites  of 103-108,  285-296 

Chenopodiuni — 

ambrosioides   anthclminticuni,    syn.    Cheno- 
podiuin  anthelminticujn . 

anthelminticum,  use  as  anthelmintic 429-439 

Chenopodiiun,  oil  of,  use  as  anthelmintic. .  429-439 
Cherry,  French.     See  Eugenia  unifiora. 
Chinese  orange.     See  Citrus  japonica. 

Chlorbenzene,  toxicity  to  insect  eggs 5S1-586 

Chloroform — 

toxicity  to  insect  eggs 5S1-586 

use  as  anthelmintic 402-403 

Chlorpicrin,  toxicty  to  insect  eggs 581-586 

Chrysophylluin  monopyrenum,  food  plant  of 

Ceratitis  capitata 105 

Citral,  toxicity  to  insect  eggs 581 

Citrus — 

japonica,  food  plant  of  Ceratitis  capitata 105 

limonia — 

individual  tree  yield 255, 258 

variability  of  yield 245-283 

sinensis — 

individual  tree  yield 252-255, 258 

variability  of  yield 245-283 

spp.,  effect  of  mulches  on  production  of. .  513-516. 
Cojj'ea  arabica,  food  plant  of  Ceratitis  capitata . .       105 
Coffee.    See  Coffca  arabica. 
Collins,  G.  N.  (paper):  New-Place  Effect  in 

Maize 231-243 

Conifer,  young,  effect  of  mistletoe  on 71S-718 

Cooperia  sp.,  in  sheep,  effect  of  anthelmin- 
tics on 410 

Copper  sulphate,  use  as  anthelmintic 406-409 

Corerot,  form  of  cork  disease 132 

Cork  disease — 

description 131-134 

resemblance  to  drouthspot 131-134 


Jan.  7-Mar.  25,  1918 


Index 


735 


Com.    See  Zea  mays.  Page 

Com-stover  silage — 

fermentation 591-599 

importance  of  water  in  production  of 590-591 

practicability 589-591 

Corn-Stover  Silage  (paper) 589-600 

Cottonseed — 

kernels,  toxicity 83-102 

meal,  toxicity 83-102 

toxic  substance,  method  of  removing  from 

ether  extract 87 

toxicity 83-102 

toxicity.    See  also  Gossypol. 

Cresol — 

meta,  toxicity  to  insect  eggs 581 

ortho,  toxicity  to  insect  eggs 581 

para,  toxicity  to  insect  eggs 581 

Cronartiiim  ribicola,  causal  organism  of  white- 
pine  blister-rust 459 

Crop  growi;h — 

effect  on  soil  extract 1 1-368 

freezing-point  method  as  index  of  variations 
in  soil  solution  due  to 369-395 

Crop  production,  relation  of  water  extraction 
of  soils  to 297-309 

Cyladinae,  subfamiily  of  Apionidae 604-607 

Cylas — 

femoralis 607 

formicarius 605 

formicariiis  elcgantulus 605-607 

turcipennis 607 

Determination  of  Fatty  Acids  in  Butter  Fat 

(paper) 719-732 

Diachasma — 
fullawayi — 

cannibalism 288-290 

parasite  of  Ceratilis  capital'a 106-10S,  280-296 

tryoni — 

cannibalism 286-290 

life  cycle 294 

parasite  of  Ceratilis  capitata,  104,  106, 108, 286-296 
prolificness,  comparison  with  Opius  hu- 

milis -95 

seasonal  abundance 291 

Dibasic  magnesium  phosphate,  influence  on 

bacteria  in  Wisconsin  soils 469-499 

Digestion,  effect  of  time  of,  on  hydrolysis  of 

casein  in  presence  of  starch 1-7 

Digestion    of    Starch    by    the  Young  Calf 

(paper) 575-578 

Dwicoreafcatetaj,  weevils  affecting 611 

Dipylidium  caninum,  effect  of  anthelmintics 

on 400  ff 

Douglas  fir.    See  Pseudotsuga  laxifolia. 
Drouthspot — 

description 130-1J1 

resemblance  to  cork 131-134 

Dryrot  of  Malus  sylvestris,  form  of  cork 132 

Effect  of  Nitrifying  Bacteria  on  the  Solubility 
of  Tricalcium  Phosphate  (paper) 671-683 

Effect  of  Season  and  Crop  Growth  in  Modify- 
ing the  Soil  Extract  (paper) 3 '  '-368 

Effect  of  Time  of  Digestion  on  the  Hydrolysis 
of  Casein  in  the  Presence  of  Starch  (paper) . .       1-7 

Effects  of  Mistletoe  on  Young  Conifers 
(paper) 715-718 

Efficacy  of  Some  Anthehnintics  (paper) 397-447 

Egg,  insect,  toxicity  of  organic  compounds 
to 579-587 


Page 

Egg,  winter  cycle  of  production  in  domestic 
fowl 547-574 

Einkom.   See  Triticum  monococcum. 

Emetic,  tartar,  use  as  anthelmintic 401-402 

Emmer.    See  Triticum  dicoccum. 

Epsom  salt,  use  as  anthelmintic 400-401 

Errata  and  authors'  emendations IV 

Ether- 
ethyl,  toxicity  to  insect  eggs 581-586 

petroleiun,  toxicity  to  insect  eggs 581 

toxicity  to  insect  eggs 582-5S6 

use  as  anthelmintic 405 

Ethyl— 

aceto-acetate,  toxicity  to  insect  eggs 581 

alcohol,  toxicity  to  insect  eggs 581-586 

ether,  toxicity  to  insect  eggs 581-586 

malonate,  toxicity  to  insect  eggs 581 

mercaptan,  toxicity  to  insect  eggs 581 

Eugenia  spp.,  food  plants  of  Ceralitis  capitala.       105 

Eugenol,  toxicity  to  insect  eggs 581 

Euscepes  batalae — 

description 608-610 

distribution 608-609 

Fat,  butter.    See  Butter  fat. 

Feed,  calf,  use  of  starch  ration  in  addition  to 
milk 577-578 

Fern,  male,  use  as  anthelmintic 415 

Fertilizer — 

dissemination  of  tobacco  wildfire  by 456 

effect  on  soil  reaction 25-26 

Ficuslaurifolia,  use  as  anthelmintic 427-428 

Fir,  Douglas.    See  Pseudotsuga  taxifolia. 

Fireblight,  resemblance  to  drouthspot 130 

Fisher,  D.  F.,and  Brooks,  Charles  (paper):  Ir- 
rigation Experiments  on  Apple-Spot  Dis- 
eases    109-138 

Fly,  fruit,  Mediterranean.  See  Ceratiiis  capi- 
tala. 

Fly,  Hessian.     See  Mayetiola  destructor. 

Fames  spp. ,  cultural  characters  of 4off 

Foster,  A.  C.and  Wolf,  F.  A. (paper): Tobacco 
Wildfire 449-458 

Foster,  W.  D.  and  Hall,  M.  C.  (paper): 
Efficacy  of  Some  Anthelmintics. . 397-447 

Fowl,  domestic.    See  Poultry-. 

Fragaria  spp. — 

degeneration  of  microspore 652-655 

dieciousness 622-628 

inflorescence 613-614 

pistils,  morphology 623 

pistils,  steriUty 622-628 

pollen,  development 628, 641-651 

pollen,  germination 637-641 

pollen,  sterility 628 

stamens,  arrangement 614-621 

Freezing-point  depression,  determination  of 
in  soils 369-395 

Freezing-Point  Method  as  an  Index  of  Vari- 
ations in  the  Soil  Solution  Due  to  Season 
and  Crop  Growth,  The  (paper) 369-39S 

French  chern'.    See  Eugenia  uniflora. 

Frogeye,  comparison  with  tobacco  wildfire .  .  .       451 

Fruit  fly,  Mediterranean.  See  Ceratilis  capi- 
tala. 

Fruit-Fly  Parasitism  in  Hawaii  during  1916 
(paper) 103-108 


736 


JouYJial  of  Agricultural  Research 


Vol.  XII 


Page 

Fruitpit 109 

Fulmer,  H.  L.  (paper):  Influence  of  Carbo- 
nates of  Magnesium  and  Calcium  on  Bac- 
teria of  Certain  Wisconsin  Soils 463-504 

Fumigation,  organic  compounds,  toxicity  to 

insect  eggs 585-586 

Fungus,  wood-rotting — 

cultural  characters,  identification  by 63-64 

hymenium,    influence    of    substratum    on 

character  of "6 

mycelium,  density  of 79 

pileus  formation,  influence  of  light  on 77-78 

pore  formation,  influence  of  gravity  on 78 

pore  formation,  influence  of  substratum  on.   7S-79 

sporophore  production 65-80 

Furfural,  toxicity  to  insect  eggs 581 

Fusarium  vasmfectum — 
causal  organism   of    wilt   disease  of   Abel- 

moschus  esculentus 533-537 

description 536-537 

occurrence  in  okra- wilt 537-538 

parasitism 538-539 

Gas,  carbon-dioxid,  effect  on  soil  solution . . .  384-385 
Gasoline— 

toxicity  to  insect  eggs 581-586 

use  as  anthelmintic 409-411 

Geranyl  acetate,  toxicity  to  insect  eggs 581 

Germination,  pollen,  of  Fragaria  spp 637-641 

Gibbs,  Joshua,  cylindrical  plow  bottom,  de- 
scription        iSi 

Gipsy  moth.    See  Portketria  dispar. 
Gipsy-Moth  Larvae  as  Agents  in  the  Dissemi- 
nation   of    the    White- Pine    Blister-Rust 

(paper) 459-462 

Goodale,  H.  D.  (paper):  Winter  Cycle  of  Egg 
Production  in  the  Rhode  Island  Red  Breed 

of  the  Domestic  Fowl 547-574 

Gooseberry.     See  Rihes  spp. 
Gossypol — 

"acetate,"  toxicity 87 

toxicity  to  pigs 89-100 

toxicity  to  rabbits 8S-89 

Gossypol,  the  Toxic  Substance  in  Cottonseed 

(paper) 83-102 

Graham,  S.  A.,  and  Moore,  W.  (paper):  Tox- 
icity  of   Volatile   Organic  Compounds   to 

Insect  Eggs 579-5S7 

Gravatt.  G.  F.,  and  Posey,  G.  B.  (paper): 
Gipsy-Moth  Larvae  as  Agents  in  the  Dis- 
semination of  the  White-Pine  Blister- 
Rust  459-462 

Grain,  cause  of  heating 6S5-6S6 

Guava.     See  Psidium  guajava. 
Guava,  strawberry.    See  Psidium  caitleianutn. 
Gurjar,  A.  M.,  and  Bailey,  C.  H.   (paper): 
Respiration  of  Stored  Wheat ;  .  685-713 

Haemonchus  conlorius,  in  sheep,  effect  of 
anthelmintics  on 404 

Hall.M.C.and  Foster,  W.D.,  (paper):  Effi- 
cacy of  Some  Anthelmintics 397-447 

Harsch,  R.  M.,  and  Long,  W.  H.  (paper): 
Pure  Cultures  of  Wood-Rotting  Fungi  on 
Artificial  Jledia 33-82 

Hasselbring,  Heinrich  (paper):  Behavior  of 
Sweet  Potatoes  in  the  Groimd 9-17 


Page 
Hawaii,  pa.xaisit<isol  Ceraliiis  capitata'm  ....    103-108 

285-296 
Hessian  fly.     See  Mayetiola  destructor. 
Hclerakis    papulosa,    in    poultry,    effect    of 

anthelmintic  on 425, 439 

Hills,  T.  L.  (paper):  Influence  of  Nitrates  on 

Nitrogen-Assimilating  Bacteria 183-230 

Hoagland,  D.  R.  (paper):  The  Freezing- 
Point  Method  as  an  Index  of  Variations  in 
the  Soil  Solution  Due  to  Season  and  Crop 

Growth 369^395 

Hoagland,  D.  R.,  and  Sharp,  L.  T.  (paper): 
Relation  of  Carbon  Dioxid  to  Soil  Reac- 
tion as  Measured  by  the  Hydrogen  Elec- 
trode    139-148 

Hog,  toxicity  of  cottonseed  to 89-100 

Holbrook,  F.  F.,  plow  bottom,  description. .       1S2 

Hollow-apple 132 

Holland,  E.  B.,  and  Buckley,  Jr.,  J.  P. 
(paper):  Determination  of  Fatty  Acids  in 

Butter  Fat 719-732 

Hookworm.    See  Ancylostoma;  Bunosiomum 

trigonocephalu'in. 
Hordeuni  spp.,  relation  to  injury  by  Maye- 
tiola destructor 519-527 

Humus — 

formation  in  mulched  basins 507-513 

in  soil,  relation  to  orange  production 513-516 

Humus  in  >Iulched  Basins,  Relation  of 
Humus  Content  to  Orange  Production, 
and  Effect  of  Mulches  on  Orange  Produc- 
tion (paper) 505-518 

Hybrid— 

Fragaria  spp.,  aborted  pollen 631-635 

Zea  7nays,  new- place  effect 231-243 

Hydrogen  electrode — 

use  in  indicating   soil  reaction 19-31 

use    in    determining    relation    of    carbon 

dioxid  to  soil  reaction 139-148 

Hydrogen-ion  concentration  of  soils . .   19-30, 139-146 

Incubation,  effect  on  freezing-point  depres- 
sion   383-384 

Influence  of  Carbonates  of  JMagnesimn  and 
Calcium  on  Bacteria  of  Certain  Wisconsin 
Soils  (paper) 463-504 

Influence  of  Nitrates  on  Nitrogen-Assimi- 
lating Bacteria  (paper) 183-230 

Insect  eggs,  toxicity  of  organic  compounds 
to 579-587 

Interrelations  of  Fruit-Fly  Parasites  in  Ha- 
waii (paper) 285-296 

lodobenzene,  toxicity  to  insect  eggs 581 

Iodoform,  use  as  anthelmintic 405-406 

Ipoinoca  batatis — 

composition,  changes  in  ground 9-1? 

weevils  aflecting 604-610 

Irish  potato.    See  Solanum  tuberosum. 

Irrigation  Experiments  on  Apple-Spot  Dis- 
eases (paper) 109-138 

Jefferson,  Thomas,  plow  bottom,  descrip- 
tion    173-174 

Jensen,  C.  A  (paper):  Humus  in  Mulched 
Basins,  Relation  of  Humus  Content  to 
Orange  Production,  and  Effect  of  Mulches 
on  Orange  Production 505-518 


Jan.  7-Mar.  25,  1918 


Index 


737 


Jonathan-spot —  Page 

description 127 

resemblance  of  other  spot  diseases  of  Malus 

sylvesiris  to 130-135 

Juglans  regia — 

individual  tree  yield 256-258 

variability  of  yield 245-283 

Kamani.  See  Calophyllum  inophyllutn:  Ter- 
minalia  catappa. 

Kelley,  \V.  P.  (paper):  Effect  of  Nitrifying 
Bacteria  on  the  Solubility  of  Tricalcium 
Phosphate 671-683 

Kerosene,  toxicity  to  insect  eggs 581-584 

Knox,  Samuel  A.,  plow  bottom,  descrip- 
tion    179-181 

Lambruschini,  R.,  plov,-  bottom,  descrip- 
tion     174-175 

Laurie  acid.    See  Acid,  lauric. 

Leaching,  soil,  effect  on  freezing-point  de- 
pressions    386-687 

Lemon.    See  Citrus  linionia. 

Leplinotarsa  dccunlineala,  toxicity  of  organic 
compounds  to  eggs  of 580-586 

Light,  relation  to  sporophore  production  in 
wood-rotting  fungi 65-76 

Lime- 
effect  on  soil  reaction 27-28 

requirement,  confusion  with  soil  acidity. .   139-140 

Limestone,  influence  on  bacteria  of  Wiscon- 
sin soils 469-499 

Long,  W.  H.,  and  Harsch,  R.  M.  (paper): 
Pure  Cultures  of  Wood-Rotting  Fungi  on 
Artificial  Media si-ft^ 

Magnesium — 
carbonate,   influence  on   bacteria  of  Wis- 
consin soils 474-499 

chlorid,  influence  on  bacteria  of  Wisconsin 

soils 479-480 

extraction  from  soil,  effect  of  crop  growth 

on 311-368 

extraction  from  soil,  effect  of  season  on. .  311-368 
phosphate,  dibasic,  influence  on  bacteria  of 

Wisconsin  soils 479-499 

Maize.    See  Zea  viays. 
Male-fem,  effect  on — 

A  ncylostoma  duodenale 415 

Belascaris  mcrginala 415 

Trickuris  depressiuscula 415 

use  as  anthelmintic 415 

"Malformation"  of  Malus  syhestris 132 

Malus  syhestris — 
bitter-pit  of — 

description iio-iii 

effect  of  heavy  irrigation  on 126 

resemblance  of  other  spot  diseases  to. . .   130-135 

corerot  of,  form  of  cork  disease 132 

cork  disease  of,  description 131-134 

drouthspot  of,  description 130-131 

dryrot  of,  form  of  cork  disease 132 

individual  tree  yield 257-238 

Jonathan-spot  of — 

description 127 

resemblance  of  other  spot  diseases  to. . .   130-133 
relation  of  time  of  picking  to  development 

of  spots  in  storage 119, 127, 129 

size,  relation  to  bitter-pit  susceptibility 126 

spot  diseases  of,  effect  of  irrigation  upon .    109, 137 
variability  of  yield 245-283 


Page 

Mangifera  indica,  food  plant  of  Ceratitis  capi- 
tata 105 

Mango.     See  Mangifera  indica. 

Manure,  mulch  of,  effect  on  orange  produc- 
tion     513-S16 

Manure-and-lime,  mulch  of,  effect  on  orange 
production 513-516 

Mayetiola  destructor,  relation  of  grain  to  in- 
jury by 519-527 

McColloch,  J.  W.,  ana  Sahnon,  S.  C.  (paper): 
Relation  of  Kinds  and  Varieties  of  Grain  to 
Hessian-Fly  Injury 519-527 

McHargue,  J.  S.  (paper):  Effect  of  Time  of 
Digestion  on  the  Hydrolysis  of  Casein  in 
the  Presence  of  Starch 1-7 

Mead,  plow  bottom,  description 181 

Medicago  saliva — 
mulch  of — 

effect  on  orange  production 513-516 

formation  of  humus  in 507-513 

roots  of,  influence  of  nitrates  on 222-226 

Mediterranean  fruit  fly.  See  Ceratitis  capi- 
tata. 

Mercaptan,  ethyl,  toxicity  to  insect  eggs 581 

Meta-cresol,  toxicity  to  insect  eggs 581 

Methyl  alcohol,  toxicity  to  insect  eggs 581-586 

Milk,  use  with  starch  food  for  young  calf. . .   577-578 

Mimusops  eleiigi,  food  plant  of  Ceratitis  cap- 
itata los 

Mistletoe.     See  Razoumofskya  spp. 

Moisture- 
content  in  Ipomoea  batatas 10-15 

importance   in    dissemination    of    tobacco 

wildfire 450-451,453,456 

in  stored  wheat,  relation  to  rate  of  respira- 
tion    689-696 

soil,  effect  on  soil  solutions 370-371 

See  also  Water. 

Moniezia  spp.,  in  sheep,  effect  of  anthelmin- 
tics on 406,412 

Monocalcium  phosphate,  effect  on — 

bacteria  of  Wisconsin  soils 469-499 

hydrogen-ion     concentration     of     soil-fihn 
water 28-30 

Moore,  W.,  and  Graham,  S.  A.  (paper):  Page 

Toxicity  of  Volatile  Organic  Compounds  to 
Insect  Eggs 579-587 

Moth,  gipsy.    See  Porthelria  dispar. 

Mulch — 

alfalfa,  effect  on  orange  production 513-5x6 

alfalfa-and-lime,  effect  on  orange  produc- 
tion    513-316 

alfalfa-hay,  effect  on  orange  production. . . .       515 
bean-straw,  effect  on  orange  production. .   514-515 
bur-clover-hay,  effect    on    orange  produc- 
tion    514-516 

manure,  effect  on  orange  production S13-51S 

manure-and-lime,  effect  on  orange  produc- 
tion   513-516 

pine-shavings, effect  on  orange  production.      515 
swcet-clover-hay,  effect  on  orange  produc- 
tion     314-315 

Mulched  basins,  formation  of  humus  in 507-513 

Myristic  acid.     See  Acid,  myristic. 

Myrobalan,  black.     See  Terminalia  chcbula. 

Neotopism.    Sec  New-place  effect. 
I   New-Place  Effect  in  Maize  (paper) .'31-243 


738 


Journal  of  Agricultural  Research 


Vol.  xn 


Nicotiana  tabacum —  Page 

host  plant  of  Bacterium  lahacuin 449-458 

infected  seed  of,  factor  in  dissemination  of 

tobacco  wildfire 456 

use  as  anthelmintic 428-429 

wildfire  of 449-462 

Nicotine,  toxicity  to  insect  eggs 581 

Nitrates — 

in  Wisconsin  soils,  accumulation  of 486-487 

in  solution — 

influence  of  Azotobacter  on 200-203 

influence  of  Bacillus  radicicola  011 21 1-2 14 

influence  on — 

alfaKa  roots 222-226 

Azotobacter 187-208 

Bacillus  radicicola 208-226 

fixation    of   atmospheric    nitrogen    by 

Bacillus  radicicola 2 14-2 1 7 

fixation  of  nitrogen  by  Azotobacter 193-201 

formation  of  volutin  bodies  in  Azoto- 
bacter    205-208 

infection  by  Bacillus  radicicola 220-222 

nitrogen-assimilating  bacteria 1S3-230 

nodule  formation 219-220, 222-226 

production  of  gum  by   Bacillus  radi- 
cicola    217-219 

production  of  pigment  by  Azotobacter.   203-205 
Nitrification — 
effect  on  solubility  of  tricalcium  phosphate 

in  soil 675-676 

gelatin,  influence  of — 
calcium  carbonate  in  Wisconsin  soils  on  489-449 

limestone  in  Wisconsin  soils  on 489-490 

magnesium  carbonate  in  Wisconsin  soils 

on 589-490 

raonocalcium    phosphate    in    Wisconsin 

soils  on 489-490 

Nitrobenzene,  toxicity  to  insect  eggs 5S1 

Nitrogen — 
accumulation  in  Wisconsin  soils,  influence 

of  calcium  carbonate  on 487-488 

fixation — 

in  Wisconsin  soils 490-491 

influence  of  nitrates  by  Azotobacter  on .   193-201 
influence  of  nitrates  by  Bacillus  radicicola 

on 214-217 

in  soil,  effect  of  crop  growth  on 311-364 

in  soil,  effect  of  season  on 311-364 

relation  to  plant  life 183-230 

Nitromethane,  toxicity  to  insect  eggs 581 

Nitrotoluene,  ortho,  toxicity  to  insect  eggs  . .       581 

Nitroxylene,  toxicity  to  insect  eggs 581 

Nodular  worm.     See  Oesophagostomum. 

Nodule  formation,  influence  of  nitrates  on. .  222-226 

Noronhia  emarginata,  food  plant  of  Ceratitis 

capitate 105 

Norton,  R.   P.,  et  al.   (paper):  Digestion  of 

Starch  by  the  Young  Calf 575-578 

Nut,  areca.     See  Areca  catechu. 


Oats.    See  Avena  sativa. 

Ochrosia  elliplica,  food  plant  of  Ceratitis  capi- 

tata 105 

Oesophagostomum — 
columbiantim,  in  sheep,  effect  of  anthelmin- 
tics on 404, 410, 412 

dentatum,  in  hogs,  effect  of  anthelmintics 
on 4ooff 


Oil—  Page 

castor,  use  as  anthelmintic 399 

chenopodium,  use  as  anthelmintic 429-439 

Okra.     S^e  Abeimoschus  esculentus. 

Okra-wilt 529-546 

Oleic  acid.    See  Acid,  oleic. 

Oleoresin  aspidii,  use  as  anthelmintic 414-416 

Opius  humilis — 

cannibalism 286-290 

life  cycle 294 

parasite  of  Ceratitis  capitata 104, 286-296 

prolificness 295 

seasonal  abundance 291 

Orange.    See  Citrus. 

Orange  production — 

effect  of  mulch  on 513-516 

relation  of  humus  content  of  soil  to 513-516 

Organic  compounds — 

boiling  point,  index  to  toxicity 580-586 

toxicity  to  insect  eggs 579-587 

volatility,  index  to  toxicity 580-586 

Orobitidae 608-610 

Orobitinae,  subfamily  of  Orobitidae 60S-610 

Ortho-brom toluene,  toxicity  to  insect  eggs.  581-582 

Ortho-cresol,  toxicity  to  insect  eggs 581 

Ortho-nitrotoluene,  toxicity  to  insect  eggs. . .       581 

Palaeopus  dioscoreae,  n.  sp 611 

Palmitic  acid.    See  Acid,  palmitic. 

Para-cresol,  toxicity  to  insect  eggs 581 

Parasites,  fruit-fly 103-108, 285-296 

Peach.     See  A  mygdalus  persica. 

Pelletierine  tannate,  use  as  anthelmintic. .  .  417-418 

Pemberton,  C.  E.,  and  WiUard,  H.  F.   (pa- 
per)— 
Fruit-Fly    Parasitism    in    Hawaii    during 

1916 103-108 

Interrelations    of    Fruit-Fly    Parasites    in 
Hawaii 285-296 

Petrolemn — 

benzin,  use  as  anthelmintic 411-413 

ether,  toxicity  to  insect  eggs 581, 585-586 

Phenol,  use  as  anthelmintic 413-414 

Phosphate.    See  Phosphorus. 

Phosphoric  acid.     See  Acid,  phosphoric. 

Phosphorus — 

in  soil,  effect  of  crop  growth  on 311-364 

in  soil,  effect  of  season  on 311-364 

Pierce,  W.  Dwight  (paper):  Weevils  Which 
Affect  Irish  Potato,  Sweet  Potato,  and 
Yam 601-612 

Pig,  toxicity  of  cottonseed  to 89-100 

Pine.    See  Pinus. 

Pinene,  toxicity  to  insect  eggs 581 

Pinus — 

ponderosa,  effect  of  mistletoe  on 715-718 

spp.,  food  plant  of  Porthclria  dispar 459-460 

Plow  bottom — 

forms  of 150-173 

history  of  development  of 173-183 

Plowing,  motion  of  soil  particles  in 162-173 

Plmn,  Brazilian.     See  Eugenia  braziliensis. 

Plummer,  J.  K.  (paper):  Studies  in  Soil 
Reaction  as  Indicated  by  the  Hydrogen 
Electrode 19-31 

Plymouth  Rock,  Barred,  egg  production 
in  winter 562-570 

Poisoning, cottonseed 83-102 

Polyporus  spp.,  cultural  characters  of. .  45-47, 60-61 


Jan.  7-Mar.  25,  1918 


Index 


739 


Page 
Polyslictus  kirsutus,  hymenium,  influence  of 

substratum  on  character  of 76 

Porlhetria — 
dispar — 

distribution 459 

factor    in    dissemination    of    white-pine 

blister-rust 459-462 

infestation  of  Pinus  spp.  by  larvae  of . . .  459-460 
infestation  of  Ribes  spp.  by  larvae  of . . .  459-462 

spp.,  injury  to  Ribes  spp 461-462 

Posey,  G.  B.,  and  Gravatt,  G.  F.  (paper): 
Gipsy-Moth  Larvae  as  Agents  in  Dissemi- 
nation of  the  White- Pine  Blister-Rust. . .  459-462 
Potash.    See  Potassium. 
Potassium — 
chlorid,    effect    on    hydrogen-ion    concen- 
tration of  soils 146 

in  soil — 

effect  of  crop  growth  on 311-364 

effect  of  season  on 311-364 

nitrate — 

influence  of  Bacillus  radicicola  on 21 1-2 13 

influence  on — 

alfalfa  nodules 224-226 

alfalfa  roots 222-224 

fixation  of  nitrogen  by  Azotobacter. .   193-196 
fixation  of  nitrogen  by  Bacillus  radici- 
cola   214-217 

growth  of  Azotobacter 188, 191 

growth  of  Bacillus  radicicola 208 

nodule  formation 223-226 

sulphate,  effect  on  soil  reaction 1 .  26-27 

Potato  beetle.    See  Lepiinotarsa  decemlincala. 
Potato- 
Irish.    See  Solanum  tuberosum. 
sweet.    See  I pomoea  batatas. 

Poultry,  egg  production  in  winter 547-574 

Premnotrypes  solani 602 

Propyl  acetate,  toxicity  to  insect  eggs 581 

Prunus  persica,  syn.  Amygdalus  persica. 

Psaliduridae 602 

Pseudotsuga  iaxifolia — 

effect  of  mistletoe  on 716-718 

host  plant  of  Razoumofskya  spp 716-718 

Psidium — 

cattleianum,  food  plant  of  Ccratitis  capitala  .  .       105 
guajaxa,  food  plant  of  Ccratitis  capitala  ....       105 

Pimky  disease 132 

Pure  Cultures  of  Wood-Rotting   Fungi   on 

Artificial  Media  (paper) 33-82 

Purgative,  use  as  anthelmintic 399-401 

Pyridin,  toxicity  to  insect  eggs 581 

Rabbit— 

toxicity  of  cottonseed  to 88-89 

toxicity  of  gossypol  to ^ 88-89 

Rahm,  W.  L-,  plow  bottom,  description. . .  178-179 
Rain,  wind-blown,  factor  in  dissemination  of 

tobacco  wildfire 456 

Rat,  toxicity  of  cottonseed  to «3-io4 

Razoumofskya  spp.,  parasite  of  Psevdotsuga 

taxifolia 716-718 

Reed.  H.  S.,  and  Batchelor,  L.  D.  (paper): 
Relation  of  the  Variability  of  Yields  of 
Fruit  Trees  to  the  Accuracy  of  Field 
Trials 245-283 


Page 
Relation  of  Carbon  Dioxid  to  Soil  Reaction 
as  Measured  by  the  Hydrogen   Electrode 

(paper) 139-148 

Relation  of  Kinds  and  Varieties  of  Grain  to 

Hessian-Fly  Injury  (paper) 519-527 

Relation  of  the  Variability  of  Yields  of  Fruit 
Trees    to    the    Accuracy    of    Field    Trials 

(paper) 245-283 

Respiration  of  Stored  Wheat  (paper) 685-713 

Rhigopsidius  iucumanus 602 

Rhode  Island  Red,  winter  cycle  of  egg  pro- 
duction in 547-574 

Ribes  spp. — 

food  plants  of  Porlhetria  dispar 459-462 

Rose-apple.    See  Eugenia  jainbos. 
Rosy-aphis     stigmonose,      resemblance     to 

bitter-pit no-iii 

Rot,  identification  by  cultural  characters. . .     63-64 
Roimd-worm.    See  Ascaridia  perspicillum. 
Rye.    SeeSecale  cereale. 

Salicylic  aldehyde,  toxicity  to  insect  eggs 581 

Salmon,  S.  C,  and  McColloch,  J.  W.  (paper): 
Relation  of  Kinds  and  Varieties  of  Grain  to 

Hessian-Fly  Injury 519-527 

Salt,  Epsom,  use  as  anthelmintic 400-401 

Sand.    See  Soil. 

Santonin  and  calomel,  use  as  anthelmintic.  420-422 

Santonin,  use  as  anthelmintic 419-420 

Season — 

effect  on  soil  extract 31 1-368 

freezing-point  method  as  index  of  variations 

in  soil  solution  due  to 369-395 

relation  to  abundance  of  Diachasma  tryoni. .      291 

relation  to  abundance  of  Opiiis  humilis 291 

variation  of  concentration  of  soil  solution 

with. 371-382 

Secale  cereale,  relation  to  injury  by  Mayetiota 

destructor 519-527 

Seed,  dissemination  of  tobacco  wildfire  by. . .       456 

Sharp,  L.  T.,  and  Hoagland,  D.  R.  (paper): 

Relation  of  Carbon  Dioxid  to  Soil  Reaction 

as  Measured  by  the  Hydrogen  Electrode.  139-148 

Shaw,  R.H.,  Woodward,  T.  E.,  and  Norton, 

R.  P.  (paper):  Digestion  of  Starch  by  the 

Young  Calf 575-578 

Sherman,  J.  M.,  and  Bechdel,  S.  I.  (paper): 

Corn-Stover  Silage 589-600 

Silage,  corn-stover 5S9-600 

fermentation 591-599 

imix)rtance  of  water  in  production  of .  .  .  .   590-591 

Small,  James,  plow  bottom,  description 175-177 

Sodium  nitrate — 
effect  on  hydrogen-ion  concentration  of  soil 

suspensions 26 

influence  on — 

fixation  of  nitrogen  by  Azotobacter 197 

fixation    of    nitrogen    by    Bacillus    radi- 
cicola    214-217 

growth  of  Azotobacter 188, 191, 207 

growth  of  Bacillus  radicicola 209 

nodule  formation 219-220 

Soil- 
acidity,  definition 139 

solubility  of  calcium  in,  effect  of — 

ammonium  sulphate  on 674 

calcium  carbonate  on 674 

tricalcium  phosphate  on 674 


740 


Journal  of  Agricultural  Research 


Vol.  XII 


Soil— Continued.  Page 

cropped  and  uncropped — 
relation   between  soluble  soil   nutrients 

in 317-363 

seasonal  variations 371-382 

effect  of  leaching  on  freezing-point  depres- 
sions    386-3S7 

extract — 

effect  of  crop  growth  on 311-368 

effect  of  season  on 311-368 

relation  to  crop  produced 317-363 

filni  water,  reaction 23-24 

humus,  relation  to  orange  production. . . .  513-516 
hydrogen-ion     concentration.      See     Soil 
reaction. 

potassiiun  chlorid  on 146 

influence  of  nitrates  on — 

Azotobacter  in 1S7-208 

Bacillus  radicicola  in 208-226 

moisture,  effect  on  soil  solutions 370-371 

nutrients  in — 

effect  of  carbon  dioxid  on 334-339 

effect  of  temperature  on 334-339 

particles,  motion  in  plowing 162-173 

solubility  of  phosphoric  acid  in — 

effect  of  ammonium  sulphate  on 674 

effect  of  calcium  carbonate  on 674 

effect  of  tricalcium  phosphate  on 674 

reaction — 

effect  of  acid  phosphate  on 27-28 

effect  of  ammonium  sulphate  on 2S-30 

effect  of  carbon  dioxid  on 140-145 

effect  of  lime  on 27-28 

effect  of  monocalciinn  phosphate  on 28-30 

effect  of  potassium  chlorid  on 146 

effect  of  potassium  sulphate  on 26-27 

solution — 
effect  of — 

carbon-dioxid  gas  on 384-385 

drying  on 385-3S6 

incubation  on 383-384 

moisture  on 370-371 

relation  of  water  extracts  to 387-391 

water  extracts,  index  of  crop  production . .   297-309 
Wisconsin,  bacteria  of — 

influence  of  calcium  carbonate  on. . . .   469-499 

influence  of  calcium  chlorid  on 479-480 

influence  of  dibasic  magnesium  phos- 
phate on  479-499 

influence  of  limestone  on 469-499 

influence  of  magnesivun  carbonate  on .  474-499 
influence  of  magnesium  chlorid  on. . .  479-480 
influence  of  monocalcium  phosphate 

on 469-499 

Solanum  tuberosum — 

origin 601 

weevils  affecting 601-604 

Species — 

emended 454-457 

new 603-604, 611 

Speck,  comparison  with  tobacco  wildfire. .  451 

Spelt.     See  TriticuJn.  Page 

SpigeUa,  use  as  anthelmintic 428 

Sporophore,  wood-rotting  fungus — 

influence  of  inoculum  on  development 80 

position  on  media 79 

production  of 65-80 

relation  of  hght  to  production  of 65-76 

relation  of  temperature  to  production  of . .  .  67-68 


Starch—  Page 

content  of  Ipomoea  batatas 10-15 

digestion  by  calf S74-S78 

hydrolysis  of  casein  in  presence  of,  effect  of 

time  of  digestion  on 1-7 

use  as  feed  in  addition  to  milk  for  young 
calf 577-578 

Stearic  acid.    Sec  Acid,  stearic. 

Stephen,  plow  bottom,  description 177-178 

Sterility  in  the  Strawberry  (paper) 613-670 

Stewart,  G.  R.  (paper):  Effect  of  Season  and 
Crop  Growth  in  Modifying  the  Soil  Ex- 
tract    311-360 

Stippen.     See  Bitter-pit. 

Stomach  worms.     See  Haemonchus  contortus. 

Storage — 

apple,  relation  of  time  of  picking  to  develop- 
ment of  spots  in 119, 128, 129 

wheat 685-714 

Strawberry.     See  Fragaria  spp. 

Strawberry  guava.     See  Psidium  cattleianum. 

Studies  in  Soil  Reaction  as  Indicated  by  the 
Hydrogen  Electrode  (paper) 19-31 

Study  of  the  Plow  Bottom  and  Its  Action 
upon  the  Furrow  Slice,  A  (paper) 149-182 

Sugar  content  in  Ipomoea  batatas 10-15 

Sweet-polato.     See  Ipomoea  batatas. 

Taenia  spp.,  in  dogs,  effect  of  anthelmintics 
on 4oofi 

Tapeworm.  See  Dipylidium  caninum;  Monie- 
zia  spp..  Taenia  spp. 

Tartar  enjetic,  use  as  anthelmintic 401-402 

Temperature — 
effect  on  water-soluble  nutrients  in  soil. .  334-339 
influence  on  respiration  of  stored  wheat. .   703-706 
relation    to    corn-stover    silage    fermenta- 
tion    S92~594 

relation  to  sporophore  production  in  wood- 
rotting  fungi 67-68 

Terminalia  spp.,  food  plants  of  Ceraiitis  capi- 
tata los 

Terpineol.  toxicity  to  insect  eggs 581 

Tetraslichus  giffardianus — 

cannibalism 288-296 

parasite  of  Ceratitis  capitata 106-108,  288-296 

Thievelia  neriifolia,  food  plant  of  Ceratitis  capi- 
tata        10s 

Thiophene,  toxicity  to  insect  eggs 581-586 

Thiophenol,  toxicity  to  insect  eggs 581 

Thymol,  use  as  anthelmintic 423-425 

Tobacco.     See  Nicotiana  tabacuin. 

Tobacco  wildfire — 
anatomy  of  plant  tissues  affected  with. . .  453-454 

appearance  of 450-451 

causal  organism  of 4S4-4S7 

comparison  with  frogeye 451 

comparison  with  speck 451 

dissemination 456-4S7 

distribution 449-450 

economic  importance 450 

history 449-450 

inoculation 4SI-4S3 

isolation 45I-4S3 

origin 456 

symptoms 45' 

Tobacco  Wildfire  (paper) 449-4s8 

Toluene,  toxicity  to  insect  eggs 581-586 


Jan.  7-Mar.23,  191S 


Index 


741 


Page 
Toxascaris  limbata,  in  dogs,  effect  of  anthel- 
mintics on 427-433 

Toxicity,  cottonseed 83-102 

Toxicity  of  Volatile  Organic  Compounds  to 

Insect  Eggs  (paper) 579-587 

Tree- 
coniferous,  effect  of  mistletoe  on 715-718 

fruit,  relation  of  variability  of  yield  to  ac- 
curacy of  field  trials 245-283 

walnut,  relation  of  variability  of  yield  to 

accuracy  of  field  trials 245-2S3 

Tricalcium  phosphate — 
effect  on  solubility  of — 

calcium  in  soil 674 

phosphoric  acid  in  soil 674 

solubility  of,  effect  of  iiitrifying  bacteria 

on 671-683 

Trichuris  depressiuscula,  in  dog,  effect  of  an- 
thelmintics on 399ff 

Trimethylene — 

bromid,  toxicity  to  insect  eggs 581 

cyanid,  toxicity  to  insect  eggs 581 

Triticum  spp. — 
relation  to  injury  by  Mayetiola  destructor .  519-527 

stored,  cause  of  heating 686-686 

stored,  respiration  of — 

consistency  of  kernel  to 694-696 

influence  of  accumulated  carbon  dioxid 

on 706-708 

influence  of  temperature  on 703-706 

location  in  kernel 627-688 

measurement 688-6S9 

oxidizable  material  in 686-687 

period  of  dampness  in  relation  to. . . .   701-703 

rate  in  oxygen-free  atmosphere 708 

relation  of  moisture  to 689-694 

relation  of  plumpness  of  kernel  to.  . . .  696-697 
soundness  of  kernel  in  relation  to. .  . .  697-701 
Trypopremnon — 
lalithorax — 

description  of  lar\'a 602-603 

description  of  pupa 603 

sanfordi,  n.  sp 603-604 

Turpentine,  use  as  anthelmintic 425-427 

Valeric  acid.     See  Acid,  valeric. 

Valleau,  W.  D.  (paper):  Sterility  in  the  Straw- 
berry   613-670 

Van  Slyke  method  for  protein  analysis,  use. .       1-7 

Vermifuge.    See  Anthelmintic. 

Verticilliutn  albo-atruwi — 

causal  organism  of  okra-wilt 533-537 

description 537 

inoculations 539-544 

occurrence  in  okra-wilt S37-538 

parasitism 538-539 

Verticillium-wilt 529-546 

Walnut.    See  Juglans  regia.  Page 

Water- 
content  of  Iponwea  batatas 10-15 

important  factor  in    production   of   com- 
stover  silage 590-591 


Water — Continued.  Page 

soil,   effect     of   crop   growth    on    extracts 

from 311-368 

soil,  effect  of  season  on  extracts  from 311-368 

soU-film,  reaction  of 23-24 

See  aho  Moisture. 

Water  Extractions  of  Soils  as  Criteria  of 
Their  Crop-Producing  Power  (paper) . . .  297-309 

Weather,  factor  in  dissemination  of  tobacco 
wildfire 456 

Weevil,  occurrence  on — 

Ipomoea  batatas 604-610 

Dioscorea  batatas 611 

Solarium  tuberosum 601-604 

Weevils  Which  Affect  the  Irish  Potato,  Sweet 
Potato,  and  Yam  (paper) 601-612 

Weir,  J.  R.  (paper):  Effects  of  Mistletoe  on 
Young  Conifers 715-718 

Wheat.    See  Triticum  spp. 

Whipworm.    See  Trichuris  depressiuscula. 

White,  E.  A.  (paper):  A  Study  of  the  Plow 
Bottom  and  Its  Action  upon  the  Furrow 
Slice 149-182 

White- pine  blister- rust.  See  BUster-rust, 
white-pine. 

Wildfire,  tobacco.    See  Tobacco  wildfire. 

Willard.    H.    F.,    and     Pemberton,    C.    E. 
(paper)— 
Fruit-Fly   Parasitism   in   Hawaii   During 

1916 103-108 

Interrelations    of    Fruit-Fly    Parasites   in 
Hawaii 285-296 

Wilt  Diseases  of  Okra  and  the  Verticillium- 
Wilt  Problem  (paper) 529-546 

Wind,  factor  in  dissemination  of — 

Portheiria  dispar 461-462 

Tobacco  wildfire 456 

WinterCycIeof  Egg  Production  in  the  Rhode 
Island  Red  Breed  of  Domestic  Fowl 
(paper) 547-574 

Wisconsin  soil.    See  Soil,  Wisconsin. 

Withers,  W.  A.,  and  Carruth,  F.  E.  (paper): 
Gossypol,  the  Toxic  Substance  in  Cotton- 
seed    83-102 

Wolf,  F.  A.,  and  Foster.  A.  C.  (paper):  To- 
bacco Wildfire 449-458 

Wood-rotting  fungus.  See  Fungus,  wood- 
rotting. 

Woodward,  T.  E.,  ct  al.  (paper):  Digestion 
of  Starch  by  the  Young  Calf 575-578 

Worm — 
cecum.    See  Helerakis  papulosa. 
nodular.    See  Oesophagostomuin. 
round.    See  Ascaridia  pcrspicillum. 
stomach.    See  Haemonchus  conlortus. 
whip.    See  Trichuris. 

Wyandotte,  White,  egg  production  in  win- 
ter    562-570 

Xylene,  toxicity  to  insect  eggs 581-586 

Yam.     See  Dioscorea  batatas. 

York-spot,  resemblance  to  cork 132 

Zea  mays,  new-place  effect 231-243 


o 


New  York  Botanical  Garden   Library 


3   5185  00263  3889 


/