Full text of "Journal of agricultural research"

See other formats

JOURNAL OF
AGRICULTURAL
RESEARCH

Volume VII

OCTOBER 2— DECEMBER 26, 1916

•••

DEPARTMENT OF AGRICULTURE

WASHINGTON, D. C.

• * •
• • •
• •• #

'"1

V-'

CONTENTS

Page

Aspergillus niger Group. Charles Thom and James N. Currie. i
Some Effects of the Blackrot Fungus, Sphaeropsis malorum, upon
the Chemical Composition of the Apple. Charles W. Cul¬
pepper, Arthur C. Foster, and Joseph S. Caldwell . 17

Formation of Hematoporphyrin in Ox Muscle during Autolysis.

Ralph Hoagland . * . 41

Comparison of the Nitrifying Powers of Some Humid and Some
Arid Soils. C. B. Lipman, P. S. Burgess, and M. A. Klein. . 47

Immobility of Iron in the Plant. P. L. GilE and J. O. CarreRO . . 83

Effects of Nicotine as an Insecticide. N. E. McIndoo . 89

Acidity and Adsorption in Soils as Measured by the Hydrogen

Electrode. L. T. Sharp and D. R. Hoagland . 123

Life History of Habrocytus medicaginis, a Recently Described
Parasite of the Chalcis Fly in Alfalfa Seed. Theodore D.

Urbahns . . . . 147

Daily Transpiration During the Normal Growth Period and Its
Correlation with the Weather. Lyman J. Briggs and H. L.

Shantz . 155

Studies of Spongospora subterranea and Phoma tuberosa of the
Irish Potato. I. E. Melhus, J. Rosenbaum, and E. S.

Schultz . 213

Growth of Parasitic Fungi in Concentrated Solutions. Lon A.

Hawkins . 255

Freezing-Point Lowering of the Leaf Sap of the Horticultural
Types of Persea americana. J. Arthur Harris and Wilson
Popenoe . 261

Grain of the Tobacco Leaf. Charles S. Ridgway . 269

Host Plants of Thielavia basicola. James Johnson. . . 289

Chemical Composition, Digestibility, and Feeding Value of Vege¬
table-Ivory Meal. C. L. Beals and J. B. Lindsey . 301

Rosy Apple Aphis. A. C. Baker and W. F. Turner . 321

Use of Two Indirect Methods for the Determination of the Hygro¬
scopic Coefficients of Soils. F. J. Alway and V. L. Clark. . 345

Correlation Between the Size of Cannon Bone in the Offspring and

the Age of the Parents. Christian Wreidt . 361

Laspeyresia molesta, an Important New Insect Enemy of the

Peach. A. L. Quaintance and W. B. Wood . 373

Energy Values of Red-Clover Hay and Maize Meal. Henry Pren¬
tiss Armsby, J. August Fries, and Winfred Waite Bra-
man . 379

(m)

Journal of Agricultural Research

Vol. VII

Page

Relationship between the Wetting Power and Efficiency of Nico¬
tine-Sulphate and Fish-Oil-Soap Sprays. Loren B. Smith. . . 389

Life History and Poisonous Properties of Claviceps paspali. H.

B. Brown . 401

Effect of Sodium Salts in Water Cultures on the Absorption of

Plant Food by Wheat Seedlings. J. F. BreazealE . 407

Nitrification in Semiarid Soils. W. P. Kelley . 417

Factors Affecting the Evaporation of Moisture from the Soil. F.

S. Harris and J. S. Robinson . . . 439

Macrosiphum granarium, the English Grain Aphis. W. J. Phil¬
lips . 463

A Specific Mosaic Disease in Nicotiana viscosum Distinct from the

Mosaic Disease of Tobacco. H. A. Allard. . 481

Syntomaspis druparum, the Apple-Seed Chalcid. R. A. Cush¬
man . 487

Assimilation of Iron by Rice from Certain Nutrient Solutions.

P. L. GilE and J. O. Carrero . 503

Influence of Bordeaux Mixture on the Rates of Transpiration from
Abscised Leaves and from Potted Plants. William H. Mar¬
tin

529

ERRATA

Page is, line 10, “A. Schiemanni (Schiemann) Thom, n. comb. = A . fuccus Schiemann” should read
“A. Schiemanni Thom, new name= A . fuscus Schiemann."

Page is, line 14, "combination" should read "name."

Page 103, lines 4 and 5 from bottom, “ PI. I, fig. 7” should read “ PI. I, fig. G." Line 6 from bottom
"PI. I, fig. 6" should read "PI. I, fig. F.”

Page 128, line 3, “OH- of 0.5X10'7” should read " H+of o-sXio-7.”

Page 254, Plate 13, last line, “Plate 12" should read “Plate 11."

Page 256, line 2 from bottom, “the salt (sodium chlorid) or sugar solution" should read “the salt or sugar
solution.”

Page 294, Table II, column 3, “ Gratz” should read “ Grah” and “ Nicotiana atropurpurea ” should read
“Nicotiana atropurpureaa"

Page 382, Table IV, column 9, line 5, ”3733” should read “377 3." Column 10, line 7, “25.26" should
read "25.36.”

Page 386, line 13 from bottom, “863" should read “862.”

Page 387, bottom line “ 1,91” should read “1,913.”

Page 417, line 12, “ Lipman (16)" should read “ Lipman (17).”

Page 418, line 9, “ Lipman (16)” should read “Lipman (17)." Lines 27 and 28, “Lipman and Burgess
(17, 18, 19, 22, 23)” should read “Lipman and Burgess (16, 18, 19, 22, 23).”

Page 424, Table III, column 1, "0.75 per cent” should read “0.075 per cent.”

Page 446, Table IV, column 4, “ Loss in 34 days” should read “ Loss in 81 days.”

Pages 481 to 486, “ Nicotiana viscosum” should read “ Nicotiana glutinosa.”

(V) .

ILLUSTRATIONS

PLATES

Effects of Nicotine as an Insecticide

Plate i. A.— Portion of the large longitudinal trachea of the house fly cut
crosswise obliquely, showing the carmine acid “ precipitate. ” B. — Combi¬
nation drawing from two consecutive sections of a green peach aphis, show¬
ing the indigo-carmine “ precipitate. ’ * C. — Cross section of a large longitu¬
dinal trachea of larva of lesser wax moth, showing the indigo-carmine
“ precipitate * * adhering to the tracheal wall. D. — Longitudinal section of
one of the smallest tracheae of the same larva as in figure C. E. — Longitu¬
dinal section of a large trachea and one of its branches of a coccid, showing
the “ precipitate ” resulting from the union of pure nicotine and phospho-
molybdic acid. E. — Portion of a cross section of an aphid, showing the
indigo-carmine “precipitate ’ ’ in a spiracle. G. —Portion of a cross section
of the same aphid as in figure F, showing no precipitate in the trachea but
much on the outside of the integument. H-O. — Longitudinal sections of
spiracles with connecting tracheae, showing how it is practically impos¬
sible for aqueous spray solutions to enter spiracles, owing to hairs, a closing
plate, and a peculiar arrangement of rims at mouths of spiracles. H. — Spi¬
racle of a coccid . I . — Spiracle of a caterpillar of A tteva aurea. J . — Spiracle
of a larva of lesser wax moth. K. — Spiracle of a caterpillar of Datana sp.
L. — Spiracle of a caterpillar of a catalpa sphinx. M. — Spiracle of a larva of a
Colorado potato beetle. N. — Spiracle of fall webworm. O. — Spiracle of
the tomato worm, showing the closing plate .

Plate 2. A to J. — Cross sections of portions of the alimentary canals and Mal¬
pighian tubules of worker honeybees, showing “ precipitated’ ' indigo-
* carmine that had been fed with pure nicotine and honey to bees three days
before they were fixed in absolute alcohol. A. — Portion of the wall of the
ventriculus, showing the “ precipitate ’ ' in inner ends of the epithelial cells.
B. — Portion of the wall of the ventriculus, showing the “precipitate” in
the middle of the epithelial cells. C. — Portion of the wall of the ventricu¬
lus, showing the “ precipitate ” in the outer ends of the epithelial cells and in
the transverse muscle layer. D. — Portion of the wall of the honey stomach
joining the pro ventriculus, showing the “ precipitate ” in the chitinous and
muscular layers. E.— Portion of the wall of the anterior part of the valve
of the pro ventriculus, showing the “precipitate” in muscles, tracheae, and
epithelial cells. F. — Section through the small intestine, showing the
“precipitate” in the center of the lumen and lining epithelium, but none
in the walls of this organ nor in the Malpighian tubules by it. G. — Section
through two Malpighian tubules against the ventriculus, showing the
“precipitate” in their cells and lumens. H. — Section through two Mal¬
pighian tubules near the ventriculus, tracheal branch and blood, showing
the “precipitate” in these tissues. I. — Section of one-third of the

rectum in a compressed condition, showing the “precipitate” in the
lumen, but none in the chitinous layer, rectal glands, or muscular layer.
J. — Section through the middle of the ventriculus, showing the dis¬
tribution of the “ precipitate ” in the lumen, between the peritrophic mem¬
branes, in the epithelial and muscular layers of the ventriculus and in the

Malpighian tubules .

(vn)

Page

122

VIII

Journal of Agricultural Research

Vol. VII

Page

Plate 3. A to I, L, M. — Drawings and diagrams representing the distribution
of precipitate resulting from the fumes of 40 per cent nicotine sulphate and
phosphomolybdic acid. A. — Transverse-longitudinal section of a trachea of
an aphid, showing the precipitate inside the trachea and in fat cells near by.

B. — Cross section of a portion of an aphid just molting, showing the precipi¬
tate on the outer surfaces of the old and the new integuments and between
them, but none in the fat cells. C.: — Combination drawing from six con¬
secutive sections through thoracic ganglion of an aphid, showing the pre¬
cipitate in three tracheal branches in the cortical layer and in the inner
layer of a ganglion . D . — Diagram of the dorsal tracheal system of an aphid ,
showing the dorsal trunk and dorsal arch. E. — Diagram of the ventral
tracheal system of an aphid, showing the anterior ventral arch, posterior
ventral arch, and the ventral trunk. F. — Combination drawing from five
consecutive sections through the thorax of an aphid, showing the precipi¬
tate on the outer surface of the integument, in the tracheae, and in the sub-
esophageal ganglion. G. — Portion of cross section of an optic lobe of an
aphid, showing the precipitate inside and outside a tracheal branch.

H. — Cross section of the brain and optic lobes of an aphid, showing the pre¬
cipitate in the tracheal branch and in the cortical layer of the brain.

I. — Cross section of two ovaries of an aphid, showing the tracheal branch
containing precipitate passing between them. L. — Cross section of two
tracheae and a fat cell of a house fly, showing the precipitate in the tracheae
and in the fat cell outside its nucleus. M. — Longitudinal section through
a spiracle and its connecting trachea of a house fly! showing the precipitate
in the neck of the spiracle and along the tracheal wall. J and K. — Cross
sections of the small tracheae, showing the precipitate in newly formed
tracheal walls resulting from the union of pure nicotine and phosphomolyb¬
dic acid. N and Q. — Cross sections, showing how well Camoy ’s fluid passes
through hard chitin, as indicated by remaining crystals of mercuric chlorid.

O. — Cross section of a medium-sized trachea of a lesser wax-moth larva,
showing that pure nicotine did not pass into an older tracheal wall under *
the same conditions as stated for figure K. P. — Cross section of portion of

the integument of an aphid, showing that pure nicotine did not pass into
chitin under same conditions as stated for figures J, K, and 0 . 122

Life History of Habrocytus medicaginis, a Recently Described Para¬
site of the Chalcis Fly in Alfalfa Seed

Plate 4. Habrocytus medicaginis: A. — Adult. B. — Cages for rearing parasite

larvae. C. — Larva. D.— Larva destroying its host larva. E. — Pupa . 154

Daily Transpiration during the Normal Growth Period and its Corre¬
lation with the Weather

Plate 5. A. — Six pots of alfalfa used in transpiration measurements. B. — Six

pots of com used in transpiration studies . . 213

Plate 6. The type of spring balance and lifting device used in the transpira¬
tion measurements . 212

Spongospora subterranea and Phoma tuberosa on the Irish Potato

PLATE A. Spongospora subterranea and Phoma tuberosa on Solanum tuberosum:

1-5. — Spongospora subterranea as found on different varieties of the Irish
potato. 6, 7. — Stages in the development of dryrot caused by Phoma
tuberosa . 254

Oct. a-Dec. 26, 1916

Illustrations

Page

PLATE 7. Spongospora subterranea on Solanum tuberosum: A. — Stem of a potato
showing formation of a gall caused by Spongospora subterranea. B. — Part
of a stolon showing galls caused by Spongospora subterranea. C. — Discol¬
oration so often found on the root near the point where the galls form. D . —

Spongospora subterranea as found on the root system of the potato . 254

Plate 8. Spongospora subterranea in the roots of various hosts: A. — Section
through a potato root affected with Spongospora subterranea. B. — Several
cells from Solanum warscewicziiy showing the formation of “giant cells”
and their division into daughter cells. C. — Section through a tomato root,

showing effects of infection by Spongospora subterranea . 254

PLATE 9. Spongospora subterranea on the roots of various hosts: A. — Galls
caused by Spongospora subterranea on the roots of Solanum warscewiczii.

B , C. — Galls caused by Spongospora subterranea formed on the roots of the

tomato . . . . 254

PLATE IO. Injuries caused by Spongospora subterranea and other agencies: A. —

Tuber showing the effect of flea-beetle injury. B . — Tuber showing a very
early stage of infection by Spongospora subterranea. C, D. — Tubers grown
in infected soil in the greenhouse under exceptionally moist conditions
and allowed a long growing season. E. — A potato from Ireland showing
the cankerous stage. P. — A tuber showing enlargement of the lenticels. . . 254

Plate ii. Dryrot associated with Spongospora subterranea: A. — A potato tuber
showing natural infection with Phoma sp. B, C. — Sections through tubers
showing more advanced stages of a rot caused by a species of Phoma. D. —

A potato tuber showing injury immediately around the sori, due partially
to the work of the plasmodium. The lower side of the tuber also shows
the beginning of the rot caused by Phoma sp. E. — Infection due to Phoma
sp. on a potato tuber infected with Spongospora subterranea , followed by
another, due probably to Fusarium coeruleum. F, H. — Potato tubers in¬
fected with Spongospora subterranea about three weeks after harvesting,
showing the effects of desiccation injury. G. — Section through a tuber,

showing the depth to which rot caused by Phoma sp. extends . 254

Plate 12. Spongospora subterranea and Phoma tuberosa: A. — Section of potato
tuber through a sorus around which no dryrot has as yet set in. B. — Sec¬
tion of a potato tuber made through a sorus of Spongospora subterranea after
the tuber had been held in storage and some dryrot due to desiccation had
developed. C, D. — Two views of the pycnidia of Phoma tuberosa as grown
in pure culture. E. — Pycnospores. F. — Mature “bulbils” of Papulo-
spora coprop hila Hotson, which in the tissues of potato tubers may be
mistaken for spore balls of Spongospora subterranea. G. — Spores of fungi
associated with Spongospora subterranea and referred to Verticillium
sp. and Stysanus sp. by Home, of whose drawing this figure is a repro¬

duction . . . 254

Plate 13. Phoma tuberosa on Solanum tuberosum: A, B. — Stages of the rot
caused by Phoma tuberosa on the Irish potato. C, D, E- — Results of arti¬
ficial inoculation with pure cultures of Phoma tuberosa . 254

Plate 14. Scab caused by Phoma tuberosa and Oospora scabies on Solanum
tuberosum: A. — Section through a tuber affected with common scab. B. —
Section through a tuber affected with the rot caused by Phoma tuberosa. . . 254

Journal of Agricultural Research

Vol. VII

Grain of the Tobacco Leaf

Page

Plate 15. A, B. — Well-cased tobacco leaves stretched over the closed end of
a test tube; showing very pronounced grain development. C. — A portion
of a cigar wrapped with a leaf containing very coarse grain. D. — The
same as figure C, but after a portion of the cigar had been smoked, showing
the white pimples in the ash produced by the burning and swelling of the

grain bodies . % . . 288

Plate 16. A. — Grain bodies of Connecticut Broadleaf tobacco as seen in ordi¬
nary transmitted light, a, Idioblasts containing sand crystals of calcium
oxalate. B. — Representative grain bodies of class 1. C. — Representa¬
tive grain bodies of class 2. D. — Representative grain bodies of class 3.

E. — Representative grain bodies of class 4. F. — Grain substance in the

form of minute spherites . 2 88

Plate 17. A. — Green tobacco leaf killed in absolute alcohol and showing idio¬
blasts of calcium oxalate and minute, scattered, single crystals of an unde¬
termined substance, but no grain. B. — Representative sample of the
poor burning 1909 Pennsylvania tobacco. C. — Flue-cured tobacco. Poor
burning. D. — Connecticut Broadleaf tobacco. Good burning. E. —
Fermented tobacco. Poor burning. F. — Fermented tobacco. Good
burning. G. — Tobacco. Cured only . 288

Host Plants of Thielavia basicola

Plate 18. Fairly typical diseased spots and lesions caused by Thielavia basicola
on various host plants. A. — Citrullus vulgaris . B. — Onobrychis viciae-
folia . C. — Lupinus luteus. D. — Arachis hypogaea . E* — Robinia pseudo¬
acacia. F. — Sclotis chinensis . 300

Plate 19. A. — Part of a field infected with Thielavia basicola in foreground,
with newer soil planted to tobacco in the background, illustrating the
marked pathogenic powers of this organism. B, — A tobacco plant showing
diseased roots from infected soil. C. — Healthy roots from uninfected soil
of a semiresistant type of tobacco. Figures B and C show the relative
growth of plants and amount of root system after equal care in removing
roots from the soil . 300

Rosy Apple Aphis

Plate 20. A. — Aphis sorbi : Spring migrant. B. — Aphis kochii: Spring mi¬
grant. C. — Aphis malifoliae: Spring migrant . 344

Plate 21. Aphis tmalifoliae: A. — Fall migrant. B. — Male. C. — Spring wing¬
less female. D. — Intermediate form . . . 344

Plate 22. A. — Aphis malifoliae: Summer wingless form, A. — Aphis malifoliae:
Oviparous female. C. — Structural details of Aphis malifoliae , A. sorbi ,
and A. kochii. a, A. sorbi: Segment VI of antenna of winged form, b, A.
malifoliae: Cornicle of spring wingless form, c, A. malifoliae: Cornicle of
summer wingless form. dt A. malifoliae: Cauda of summer wingless form.
e, A. malifoliae: Cauda of spring wingless form. /, A. malifoliae: Segment
VI of antenna of winged form, g, A . malifoliae: Segment VI of antenna of
stem mother. hf A. kochii: Segment VI of antenna of winged form, i, A.
sorbi: Cauda of winged form, j, A. kochii: Cornicle of spring migrant.
k, A. sorbi: Cornicle of spring migrant. lf A. malifoliae: Cornicle of spring
migrant, m , A. sorbi: Segment III of antenna of spring migrant, n, A.

malifoliae: Segment III of antenna of spring migrant . 344

Plate 23. Aphis malifoliae on its alternate host, Plantago lanceolata . 344

Oct. 2-Dec. 26, 1916

Illustrations

Page

Pi, ATS 24. A. — Broad-leaved plantain showing the effect of an attack by Myzus

plantaginis. B . — Apple leaves curled by colonies of Aphis malifoliae . 344

Plate 25. A. — Rhode Island Greening apples deformed by Aphis malifoliae .

B. — Apple twigs twisted by colonies of Aphis malifoliae: Beginning of
twisting. C. — Apple twigs twisted by colonies of Aphis malifoliae: Twisted
twig. D. — Winesap apples deformed by Aphis malifoliae . 344

Laspeyresia molesta, an Important New Insect Enemy op the Peach

Plate 26. Laspeyresia molesta: A. — Injury to shoot of a Domesticaplum. B. —

Injury by larva to cherry . 378

Plate 27. Laspeyresia molesta: One-year budded peach nursery tree, showing

injury of caterpillars . 378

Plate 28. Laspeyresia molesta: A. — Typical appearance of peach twigs in fall
injured by larva. B.— -Peach twig, showing large mass of dried gum and

leaf fragments due to attack by the caterpillar . 378

Plate 29. Laspeyresia molesta: A. — Typical exterior appearance of larval in¬
jury to peach shoot. B. — The same shoot cut open, showing the larva in

its burrow . 378

Plate 30. Laspeyresia molesta: A. — Cavity excavated in peach by larva enter¬
ing at the side. B. — Larval injury at stem end of peach; also the summer

cocoon of the insect . 378

Plate 31. Laspeyresia molesta: Peach cut open to show larval injury at the pit . 378

Life History and Poisonous Properties op Claviceps paspali

Plate 32. A. — Section through a mature stromatic head of Claviceps paspali ,
showing perithecia containing asci. B. — Spike of Paspalum dilatatum
with mature sclerotia attached. C; — Tufts of hyphae producing sphacelial
spores. D. — Section of mass of tissue within grass spikelet during spha-
celia stage of Claviceps paspali; spores are produced by tufts of hyphae
along edge of section. E. — Schlerotium of Claviceps paspali with stromata.

F. — Spikes of Paspalum dilatatum , showing a number of sclerotia attached . 406

Macrosiphum granarium, the English Grain Aphis

Plate B. Forms of Macrosiphum granarium: 1. — Mother of males and grand¬
mother of oviparous females. 2. — Typical green viviparous female. 3. —

Pupa of male. 4. — Pupa of the mother of oviparous females. 5. — Ovipa¬
rous female . 480

Plate 33. Macrosiphum granarium: A. — Winged viviparous female: a, Cor¬
nicle. B. — Winged male . 480

Plate 34. Macrosiphum granarium: A. — Antenna of male. B. — Antenna of
winged viviparous female. C. — Hind tibia of oviparous female. D. — An¬
tenna of wingless viviparous female. E. — Antenna of wingless oviparous
female. Antenna of stem mother . 480

A Specific Mosaic Disease in Nicotian a viscosum Distinct from the
Mosaic Disease of Tobacco

Plate 35. Leaves of Nicotiana viscosum affected with the mosaic disease . 486

PLATE 36. A. — Normal blossoms from healthy plants of Nicotiana viscosum. B.

— Depauperate blossoms from mosaic plants affected with the mosaic dis¬
ease peculiar to N. viscosum. C, D. — Blossoms showing catacorolla, etc. , as
a result of the mosaic disease affecting Nicotiana viscosum . 486

Journal of Agricultural Research

Vol. VII

Syntomaspis druparum, the Apple-seed Chalcid

Page

PLATE 37. Syntomaspis druparum: A. — Adult female. B. — Adult male ; outline

of abdomen, lateral view, at right . 502

Plate 38. Syntomaspis druparum: Apple injury and hibernating larvse. A. —

Usual type of injury resulting from oviposition. B, C. — Extreme type of
injury resulting from oviposition. D. — Hibernating larvae within seeds of

an apple . . 502

Plate 39. Syntomaspis druparum: Infested and sound seeds of apples. A. —

Infested seeds. B. — Sound seeds . 502

Plate 40. Syntomaspis druparum: Oviposition. A. — Female ovipositing in
fruit of Crataegus sp. B, C. — Oviposition in apples. D. — Mica cage used
in the life-history studies of Syntomaspis druparum . 502

TEXT FIGURES

Acidity and Absorption in Soils as Measured by the Hydrogen Elec¬
trode

Fig. 1. Diagram of the hydrogen-electrode apparatus and of the apparatus for

generating pure hydrogen . 140

Life History of Habrocytus medicaginis, a Recently Described Para¬
site of the Chalcis Fly in Alfalfa Seed

Fig. 1. Map of the United States, showing the known distribution of Habro¬
cytus medicaginis . 148

Daily Transpiration During the Normal Growth Period and its Cor¬
relation with the Weather

Fig. 1. Graphs showing the daily transpiration from the individual pots of
plants which constituted the first five sets in the transpiration meas¬
urements in 1914 . . . . . 157

2. Graphs showing the daily intensity of environmental factors and the

daily transpiration of 22 crops for the year 1914 . . . 164

3. Graphs showing the daily intensity of environmental factors and the

daily transpiration of 23 crops for the year 1915 . 174

4. Determination of the area, 01a a plane normal to the sun’s rays, of the

shadow of a cylinder of diameter and height in terms of the angular
departure of the sun from the vertical . „ 186 •

5. Ratios of the daily transpiration of Kubanka wheat to the daily inten¬

sity of various weather factors, plotted with approximately the same
amplitude . 188

6. Ratios of the daily transpiration of Minnesota Amber sorghum to the

daily intensity of various weather factors, plotted with approximately
the same amplitude . 188

7. Ratios of the daily transpiration of alfalfa to the daily intensity of

various weather factors, plotted with approximately the same
amplitude . 190

8. The ratio of daily transpiration of different crops in 1914 to daily evap¬

oration (shallow tank) plotted in percentage of the maximum . 191

9. The ratio of daily transpiration of different crops grown in 1915 to daily

• evaporation (shallow tank) plotted in percentage of the maximum. . 192

10. Graph showing a linear relation between the logarithm of the transpira¬
tion-evaporation ratio of Sudan grass and the time . . . 196

Oct. 2-Dec. 26, 1916

Illustrations

XIII

Page

Fig. 11. Graph showing a linear relation between the logarithm of the trans¬
piration-evaporation ratio of Sudan grass (grown in the open) and the
time . . . , 197

12. Graph showing a linear relation between the logarithm of the trans¬

piration-evaporation ratio of Algeria corn and the time . 198

13. Graph showing a linear relation between the logarithm of the trans¬

piration-evaporation ratio of Northwestern Dent com and the time. . 199

14. Graph showing a linear relation between the logarithm of the trans¬

piration-evaporation ratio of Minnesota Amber sorghum and the time . 1 99

15. Graph showing a linear relation between the logarithm of the trans¬

piration-evaporation ratio of alfalfa (in the open) and the time . 200

16. Observed daily ratios of transpiration to evaporation during early

stages of growth of Sudan grass compared with exponential graph
computed from the relationship shown in figure 11 . 200

17. Graphs of the daily ratios of the transpiration of the different crops

grown in 1914 plotted logarithmically . 201

18. Graphs of the daily ratios of the transpiration of the different crops

grown in 1915 plotted logarithmically . . . 203

Spongospora subterranea and Phoma tuberosa on the Irish

Potato

Fig. i. Map of the experimental plots at Caribou, Me., showing their arrange¬
ment, distribution of Spongospora subterranea , percentage by weight
of the progeny infected with the disease, and the yield per acre. . . . 239

Grain op the Tobacco Leaf

Fig. i. Curves plotting data relative to the burning quality of tobacco from

fertilizer treatment at Red Lion, Pa., for crops of 1913 and 1914 . ... 282

2. Curves plotting data relative to the burning quality of the tobacco

grown on fertilizer plots at Red Lion, Pa., in 1914 . . 283

Use of Two Indirect Methods for the Determination of the Hygro¬
scopic Coefficients of Soils

Fig. i. Diagram showing the amounts of free water at different levels in eight
fields, illustrating the concordance of the values obtained for the
hygroscopic coefficient by calculation from the hygroscopic moisture
with those directly determined . . 356

Correlation Between the Size of Cannon Bone in the Offspring and

the Age of the Parents

Fig. i. Curve showing the percentages of mares with various-sized cannon

bones, sired by stallions under 11 years old . . . 363

2. The percentages of mares with various-sized cannon bones, sired by

10 selected stallions when these were under 11 years old . 365

3. The percentages of mares of various-sized cannon bones bred from

dams under 1 1 years old . 365

4. The percentages of mares of various-sized cannon bones bred from both

parents under 1 1 years old . . 366

5. The percentages of mares in various classes deviating from their dams

when both parents were under 11 years old . 366

XIV

Journal of Agricultural Research

Vol. VII

Relationship Between the Wetting Power and Efficiency of Nicotine-
Sulphate and Fish-Oil-Soap Sprays

Page

Fig. i. Efficiency and wetting-power graphs for sprays in group i, containing
io ounces of nicotine sulphate and varying quantities of soap, and

group 4, containing various amounts of soap with no nicotine . 394

2. Efficiency and wetting-power graphs for group 2, containing 5 pounds
of soap, and group 3, containing 1 pound of soap plus varying amounts
of nicotine sulphate . 397

Life History and Poisonous Properties of Clavtceps paspali

Fig. i. Clavtceps paspali: a , Mature ascus; b , ascus breaking up to liberate

spores; c, ascospore . 402

2. Claviceps paspali: Tip of tuft of hyphae, showing the production of

sphacelia spores . 403

Effect of Sodium Salts in Water Cultures on the Absorption of Plant
Food by Wheat Seedlings

Fig. 1. Graphs showing the effect of sodium chlorid in nutrient solutions on

the nitrogen, potash , and phosphoric-acid content of wheat seedlings . 408

2. Graphs showing the effect of sodium sulphate in nutrient solutions on

the nitrogen , potash , and phosphoric-acid content of wheat seedlings . 409

3. Graphs showing the effect of sodium carbonate on the nitrogen, potash*,

and phosphoric-acid content of wheat seedlings. First series . 410

4. Graphs showing the effect of sodium carbonate on the nitrogen, potash,

and phosphoric-acid content of wheat seedlings. Second series . 41 1

5. Graphs of the mean values of the first and second series showing the

effect of sodium carbonate on the nitrogen, potash, and phosphoric-
acid content expressed in percentage of the dry weight of wheat
seedlings . 412

6. Graphs showing the effect of sodium chlorid on the absorption of nutri¬

ents by wheat seedlings . 414

7. Graphs showing the effect of sodium sulphate on the absorption of

nutrients by wheat seedlings . 414

8. Graphs showing the effect of sodium carbonate on the absorption of

nutrients by wheat seedlings . 415

Factors Affecting the Evaporation of Moisture from the Soil

Fig. j. Evaporation from' Greenville loam containing different initial per¬
centages of moisture . 445

2. Evaporation from sand containing different initial percentages of

moisture . . 447

3. Evaporation from clay containing different initial percentages of

moisture . 448

4. Evaporation from muck containing different initial percentages of

moisture . 448

5. Loss of moisture from Petri dishes containing different percentages of

soil moisture and kept in a saturated and unsaturated atmosphere . . 449

6. Evaporation of water from wet soils with different wind velocities. . . . 450

7. Loss of water from soil and temperatures in the sun and under cheese¬

cloth and board shade . . . 451

8. Time required at different temperatures to drive off half and all the

water from Greenville loam containing 12 per cent moisture . 452

Oct. 2-Dec. 26, 1916

Illustrations

Page

Fig. 9. Time required at different temperatures to drive off half and all the

water from sand containing 20 per cent moisture . . 453

10. Evaporation of water in 66 days from sand of different sizes with a

water table maintained 1 cm. below the surface . 454

xi. Evaporation of water in 36 days from loam and sand of different sizes

with a water table maintained 3 cm. below the surface . 454

12 . Evaporation of water in 1 1 5 days from quartz and river sand of different

sizes with a water table maintained 3 cm. below the surface . 455

13. Toss of water from glasses having dry mulches of sand of various sizes

suspended above free water . 455

14. Loss of water in 180 days from glasses having dry mulches of various

kinds suspended above free water . 456

15. Evaporation from distilled water and from sodium-chlorid solutions

of different concentrations . 457

16. Evaporation from sand wet with distilled water and with sodium-

nitrate solutions of different concentrations . 458

17. Evaporation of water from Greenville loam containing different

quantities of sodium chlorid . 459

MACROSIPHUM GRANARIUM, THE) ENGLISH GRAIN APHIS

FlG. i. Map showing the distribution of Macrosiphum granarium in the United
States as indicated by records on file in the Bureau of Entomology,

1916 . 465

Syntomaspis druparum, the) Apple-Seed Chalcid

Fig. i. Syntomaspis druparum: Apple and seed showing oviposition puncture . 493

2. Syntomaspis druparum: Eggs . 494

3. Syntomaspis druparum: Newly hatched larva . 494

4. Syntomaspis druparum: Mouth parts of larva of first instar . 494

5. Syntomaspis druparum: Mandibles of larvae of various instars _ 495

6. Syntomaspis druparum: Full-grown larva . 495

7. Syntomaspis druparum: Mouth parts of full-grown larva . 496

8. Syntomaspis druparum: Pupa of female . ' . 500

DEPARTMENT OF AGRICULTURE

Vol. VII Washington, D. C., October 2, 1916 No. 1

ASPERGILLUS NIGER GROUP

By Charles Thom, Mycologist , Bureau of Chemistry , and James N. Currie, Dairy
Chemist , Bureau of Animal Industry

OXALIC-ACID PRODUCTION OF SPECIES OF ASPERGILLUS

The recent discovery by the writers of an oxalic-acid-forming species of
Penicillium 1 led from a review of the subject of oxalic-acid production
by molds to a study of the black forms of Aspergillus spp. Wehmer, in
1 891, 2 showed that A.niger is a very active oxalic-acid-producing fungus.
The question whether this ability to produce oxalic acid is possessed in
equal degree by all strains of A. niger has not been heretofore discussed.
Culture hi, received from Amsterdam, Netherlands,3 as A . niger , was
selected for comparison with Penicillium oxalicum , because this strain
was supposedly obtained originally from Wehmer.

Aspergillus niger has been the subject of a large number of biochemical
researches. Back of this selection lies its apparent ease of specific
identification, together with, as a corollary, the assumption that the study
may be repeated elsewhere by isolating a black species of Aspergillus.
Antithetic to this point of view there occurs in the literature a series of
specific names and descriptions of black or dark-brown species, most of
which rest upon minor . morphological characters plus the assumption
that occurrence upon hosts or substrata of widely different nature is evi¬
dence of specific difference.

Wehmer, in 1901, 4 cited 18 such names; at least 25 may now be found
in the literature. Comparative culture may ultimately show how many of
these may be separated by characters definite enough to be used in
descriptive work.

The possible bearing of the comparative study of oxalic-acid pro*
duction upon the problem of relationship among this lot of strains or

1 Currie, J. N., and Thom, Charles. An oxalic add producing Penidllium, In Jour. Biol. Chem., v. 22,
no. 2, p. 287-293, 1 fig. 1915.

2 Wehmer, Carl. Entstehung und physiologische Bedeutung der Oxalsaure in Stoffwechsel einiger
Pilze. In Bot. Ztg., Jahrg. 49, p. 233ft. 1891.

8 By courtesy of Dr. Johanna Westerdijk.

* Wehmer, Carl. Die Pilzgattung Aspergillus in morphologischer, physiologischer und s ystematischer
Beziehung unter besonderer Berucksichtigung der mitteleuropaeischen spedes. V. Sy sterna tik. C.
Schwarzbraune Arten. In Mem. Soc. Phys. et Hist. Nat. Gen&ve, t, 33, pt. 2, no. 4, p. 103-m. 1901.

Vol. VII, No. r
Oct. 2, 1916

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

Journal of Agricultural Research

Vol. VII, No. i

species led to the selection of 20 cultures for a comparative test. Two
forms which could be readily distinguished from the typical black group
were included: 4030.4, a strain of A . ochraceus , and 3522.30, possibly A .
violaceo-fuscus of Gasperini. Two other forms, 2580 and 4030.1, have
differences discoverable readily with the microscope; but most of the
series shade into each other morphologically. The history of these forms,
so far as known, follows: 111, A . niger , received in September, 1909, by
courtesy of Dr. Johanna Westerdijk, Amsterdam, Netherlands; 3534-a,
A. niger , var. altipes , 3534-b, A . cinnamomeus , 3534-c, A. fuscus , the
three forms described by Schiemann 1 as mutants from a strain of A. niger
obtained from Amsterdam and probably identical with No. 111 ; 142,
marked A. ficuum , P. Hennings, received from Amsterdam; 2396, from
Missouri, by Prof. G. M. Reed; 2580, isolated from interior of red pepper
(capsicum) from Barcelona, Spain; 2469.4, from Delaware, marked Sterig-
matocystis violacea , by Prof. M. T. Cook; 2657, from soil, England, by
Miss E. Dale; 2774, from ulcerated human ear, by Dr. A. B. Stout; 2766,
from fermenting mash, consisting of oak galls from China; 3522.30, pos¬
sibly A. violaceo-fuscus of Gasperini from soil, Porto Rico, by Dr. J. R.
Johnston; 3528.7, from Pittsburgh, the Mellon Institute, by Mr. F. A.
McDermott; 4049, from sardine paste, Bureau of Chemistry; 3547.254-b;
from Kansas soil; 4050, from Chinese galls, sent by Eastman Kodak Co.,
4020.33, from soil, Texas; 4030.1, A. carbonarius(?)t 4030.4, A. ochraceus ,
4030.5, from Dr. A. F. Blakeslee, Storrs, Conn.

Table I gives the chemical results with the strains arranged approxi¬
mately in the order of their relative activity in oxalic-acid production.
Two determinations are given: The direct titration of free acid expressed
in cubic centimeters of N/10 sodium hydroxid per 50 c. c. of medium,
and the determination of the oxalic acid as oxalates as found in the
same amount of medium.2

Oxalic acid was first precipitated as calcium oxalate, then dissolved in
dilute hydrochloric acid and again precipitated.3 It was redissolved in
dilute sulphuric acid and titrated with standard potassium permanganate
to a slight permanent pink.

The cultures were held two days at 30° C. and for the remaining period
at 200.

1 Schiemann, Elisabeth. Mutationen bei aspergillus niger Van Tieghem. In Ztschr. Induk. Abstam.
u. Vererbungslehre, Bd. 8, Heft 1/2, p. 1-35, 16 figs., 2 pi. (1 col.). 1912.

s The cultures were grown upon a modified Czapek’s solution with the following composition:

Water . 1,000 c. c.

Sodium nitrate . 3 gm.

Potassium phosphate (KH2PO4 ) . 1.0 gm.

Magnesium sulphate . o. 25 gm.

Potassium chlorid . . 0.25 gm.

Ferrous sulphate . . . o. 01 gm.

Cane sugar . 50.0 gm.

* For a description of the method, see Currie, J. N., and Thom, Charles. An oxalic acid producing Peni-
cillium. In Jour. Biol. Chem,, v. 22, no. 2, p. 290. 1915.

Oct. 2, 1916

Aspergillus niger Group

Table) I. — Comparative oxalic-acid production in strains of Aspergillus niger and related
organisms grown in flasks of 50 c. c. each of Czapek’s solution containing 5 per cent
of sugar

[Acidity and oxalate radical expressed in cubic centimeters of Njio sodium hydroxid required to neutral¬
ize 50 c. c. of medium]

Culture

No.

Name of source.

7 days.

10 days.

14 days.

18 days.

Acidi¬

ty.

Oxa¬

late.

Acidi¬

ty.

Oxa¬

late.

Acidi¬

ty.

Oxa¬

late.

Acidi¬

ty.

Oxa¬

late.

142 .

2469-4 .

hi .

3528.7 .

4020. 33 .

4049 .

2657 .

4050 .

2766 .

2774 .

3534b.......

3547- 354b...

2396 .

2580 .

4030.5 .

3534C . . .

3522.30 .

3534a .

4030- 1 .

4030.4......

A. ficuum (?) .

Soil, Delaware .

A . niger (Amsterdam) . .

Pittsburgh .

Soil, Texas .

Sardine paste .

Soil, England .

Chinese galls .

Human ear .

A. cinnamomeus .

Soil, Kansas .

Missouri .

Pepper, Spain .

Unknown .

A . fuscus .

A. violaceo-fuscus (?).,..

A. niger altipes .

A . carbonarius .

A . ochraceus .

116. 17
97-57
62.30

103. 74
62. 30
61.03
29. 78
50. 16
39*77
52. 70

72. 10
47.80
46. 24

81. 11
34- 97
85-23
17-73
31-74
31-54

1-57

xi6. 21
48. 76
67. 82

48.57
65-95
36-43
30.64
28. 58
29. 88
34- 94

63-71

45-96
31-01
84.07
11.96
81.08
9-15
25-41
14-57
7. 10

142. 85
103. 15
80. 52
97.96

69.94
69. 30
58.78

54-36

53-10

52.41
47. 80
So- 74
45-55
49-57
28.60
26. 64
12-73
II. 17
5-58
I.47

155-82
94-35
94. 72
79- 9S
79-03
59- 41
74- 73
51-00
62.77
56. 05

61.47

65. 20
56.05
63.90
17-Si
40- 73
18. 87
23-54
18.31
13.08

167. 90
in. 88
79- 54
89.92
67.80
75* 24
29- 58

51. 12
26.06
37- 62
28. 22
58. 78

33-90

32.90

13.90
19- 39

9. 01
9. 21
6. 66
.78

182. 90
96. 59
94-35
85-56
82. 87
66. 5i

45- 59

47. 64
41.48

46- 71
43-34
74-92
48.95

48. 20
15-32
33-07

I7- 56
23-35
21-33
14. 70

153* 21
99*53
71. 12
61. 14
80. 72

32- 33
22. 92
37-32
14. 69
38. 20
28. 80
34- 68
15*28

33- 50

2- 35
12.34

3*13

8.03

3- 53
•39

170. 01
99.02
87.05
60. 34
94* 53
40-54
39* 42
44.09

30- 83
49- 70
44*09
54- 93

31- 39
49-66

16.44

28.02

16.44

23-54

13*85
xi. 58

A study of Table I disposes effectually of the idea that because a
species of Aspergillus is black or fuscous it must possess in specific measure
the power to produce oxalic acid. It is noteworthy that all of the series
possess this power in some degree and that some of the series show it in
excessive degree. Others, however, produce this acid in no greater
amounts than do members of other groups, as has been shown in the
authors' previous paper.1 These wide variations therefore indicate
either a group of heterogeneous ancestry or a series of races of a single
ancestry, which show great variation in the ability to produce a par¬
ticular reaction. If such variation can be correlated with morphological
characters, it is a valuable accessory in the identification of species.
If cultures exhibit only quantitative differences in the reactions selected
for study, such differences may be exceedingly important economically
without justifying the description of separate species.

Nine forms representative of the range of variation found in Table I
were selected for further experiment. No. 4047, growing upon strong
lemon juice, was added. Tubes of Czapek’s solution agar, Raulin’s fluid,
wort agar, and beef-peptone agar were prepared. Duplicate tubes of
each of the four media were inoculated from a single tube of each of these
10 strains. The cultures were incubated at 3 70 C., until ripe spores were
abundant, which was usually in about three days. Then transfers were
made to fresh tubes of the same medium. In this way in a period of
about five weeks each strain was transferred seven times upon each of

1 Currie, J. N.p and Thom, Charles. Op. cit.

Journal of Agricultural Research

Vol. VII, No. i

these media and grown at 370. At the end of that time they were
reexamined for oxalic-acid production. The tabulated results are given
in Table II, with the forms arranged in the same order as in Table I. The
quantity of N/10 sodium hydroxid required to neutralize the free acid and
the oxalate radical is shown in separate columns. The column marked
‘average ’ ’ represents the average of the four previous figures.

Table; II. — Comparative oxalic-acid production of 10 strains of a black species of Asper¬
gillus grown at S7° C. , through seven transfers in parallel culture upon the hnedia
indicated

[Acidity and oxalate radical expressed in cubic centimeters oiNjio sodium hydroxid required to neutralize
50 c. c. of Czapek's solution containing 5 per cent of cane sugar, after growth for 10 days]

Culture No.

Czapek’s

solution.

Raulin’s

solution.

Wort agar.

Beef-peptone

agar.

Average.

Comparison
of 10 days’
growth from
Table I.

Acid¬

ity.

Oxa¬

late.

Acid¬

ity.

Oxa¬

late.

Acid¬

ity.

Oxa¬

late.

Acid¬

ity.

Oxa¬

late.

Acid¬

ity.

Oxa¬

late.

Acid¬

ity.

Oxa¬

late.

142 .

165. 18

190. 62

150- 89

192.58

120. 57

135- 30

125-57

142. 22

143-05

165. 23

142. 85

155- 82

2469.4 .

69-35

67.90

72-85

86. 18

7i- 13

68. 40

112. 29

74-32

80. 95

74. 20

103. 15

94-35

4047 .

in .

69.77
82. 49

46. 66
98. 76

71.04

84.52

56. 78
100. 98

53-70
73- 76

39- 76
89. 88

65. 80

83-51

56- 54
100. 98

65.08
81. 07

49- 93
97- 65

80. 52

94* 72

4049 _ / .

46. 26

48.36

85. 12

58.52

31. 81

48.88

62. 99

56. 54

56- 54

53-07

69. 30

59- 41

2766 .

33- 15

40.74

80. 50

107.66

49- 85

67.40

43-27

56.30

51.69

68. 02

53- 10

62. 77

40. 28
53-39

45.18

65. 92

54* 53

50. 62

38. 70

40. 00

65. 40

62.96

49- 73

52. 41

56. 05

2580 .

73- 45

77. 28

50. 86

67. 16

38. 23

54-32

53-98

66. 14

49-57

63.90

28. 14
7- 33

45-i8

17. 28

23-95

39- 76

63. 02

55-30

22.04

35- 56

34- 29

43- 95

26. 64

40. 73

4030.1 .

11. 16

21. 48

11. 16

4-43

15-80

11. 08

13- 22

5.58

18. 31

A study of Table II shows a somewhat higher acid production for
cultures propagated upon the Czapek’s and Raulin’s solutions than for
those given upon wort and beef-peptone agar. In spite of occasional
contrasting figures, the entire table fails to show any marked increase or
decrease in acid-producing power by the treatment. This experiment
tends to the conclusion that there are many strains or varieties of black
Aspergillus spp. which differ .markedly in the production of this reaction.
These differences have persisted through many repetitions of the work
with certain forms. Some of these forms, notably some of those pro¬
ducing the largest quantities of oxalic acid, have been in continuous
culture by one of the writers for six years. Whether these strains would
continue to produce oxalic acid iri the same quantity if cultivated for a
longer period is not known. The mutants described and distributed by
Schiemann 1 (3534-a, 3534-b, and 3534-c) differ from each other in the
quantity of acid produced. This difference is accompanied by a differ¬
ence in color on the part of two of them, A. fuscus and A. cinnamomeus.
In the case of A. niger, var. altipes, the strain as studied in this laboratory
has lost the single 'morphological difference — long stalks — originally de¬
scribed; hence, at present it is not distinguishable from our Amsterdam

1 Schiemann, Elisabeth. Op. cit.

Oct. 2, I916

Aspergillus niger Group

stock culture (No. m). This strain (No. 3534-a) is, however, the lowest
oxalic-acid producer of all the strains closely related morphologically to
the typical A . niger . These three forms, therefore, known to be closely
related to each other and probably also to the Amsterdam culture
(No. in), differ markedly enough in this one reaction to suggest that the
mutation which occurred was probably a quantitative readjustment
among the enzyms.

In both tables some cultures show much more free acid, as indicated
by titration, than is shown by the determination of oxalate. This
difference, as redetermined for certain of these forms in the second experi¬
ment, is maintained in approximately the same relative proportion in
Table II. It has held true also in experiments not included in this paper.
No. 142, hi, 4020.33, 2657, 3547, 3546, 2580, 3534-a, 3534--C, and
3522.30 show slight differences between these determinations at 7 and
10 days. The other forms show much greater differences. As all the
cultures become older, the oxalate determination equals, and even be¬
comes greater than, the free-acid figure. Unfinished work to be reported
later shows that citric acid forms part at least of the excess transiently
found in the determination of free acid. Comparative study of the
colonies themselves does not correlate these differences in acid production
with morphology. The difference between the strains used appears,
therefore, to be one of rate and quantity of reaction rather than a differ¬
ence in kind of activity.

Although most of these forms had previously been studed carefully, a
microscopic examination of each strain was made, in order to seek in
morphology a possible basis for separation. The range of morphological
characters found points to the existence of a series of closely related
strains in which the differences are in measurement of parts, intensities
of color, and quantitative differences in the production of particular
reactions.

The members of this group grow under a wide range of cultural condi¬
tions. When colonies of a particular strain are grown simultaneously
upon substrata of markedly different composition, distinct differences
appear. All these strains, however, when grown under the same condi¬
tions, have so many common characters and so many intergradations
that a group characterization upon the lines recently used by one of the
writers for Penicillium spp.1 will be more useful than any attempt to
describe strains separately. This work harmonizes with the conclusion
of Schiemann 2 that A. niger as commonly understood is an unstable or
mutating group comparable to Oenothera spp.

1 Thom, Charles. The Penicillium luteum-purpurogenum group. In Mycologia, v. 7, no. 3, p. 134-142,
1 fig. 1915.

2 Schiemann, Elisabeth. Op. cit.

Journal of Agricultural Research

Vol. VII, No. i

GROUP CHARACTERIZATION
COLONY CHARACTERS

Colonies spread rapidly upon Czapek’s solution agar. They are at
first white with abundant submerged mycelium and with more or less
prostrate or trailing hyphse radiating toward the periphery. Floccose
aerial mycelium is occasionally developed later. Conidiophores arise as
colorless branches from submerged or, more rarely, from aerial hyphse
and constitute the whole surface growth in most strains. With conidial
production the colony color changes from white to fuscous black, but
never shows any shade of green.1 In certain forms Saito 2 records the
color as passing through yellow to brownish black. The color found is
rarely evenly distributed in the fruiting parts. Yellow or yellow-brown
color may be present or absent in the upper ioo fx of the stalk; in old
cultures it is usually found in the vesicle and sterigmata, which frequently
are deeply colored, even becoming brownish black or carbonaceous in
certain races. Most of the color, as indicated by the researches of
Linossier,3 is deposited in the ridges or warts of the conidial wall. The
submerged mycelium is uncolored in some forms, more or less deeply
yellowed in others. The agar remains uncolored. All forms grow at
370 C. and, with the exception of A. carbonarius , are favored by that
temperature.

MORPHOLOGY

Stalk. — The stalks or conidiophores vary in thickness from 6 to 25^,
and in lengths from 0.5 to 10 mm. They are unseptate or indistinctly
septate and have thick walls, smooth on the outside, and on the inside
either smooth or, in some cases, with irregular thickenings. When
broken these walls split lengthwise into strips. There is a great differ¬
ence in the length of stalks in the same colony. In the denser center of
the colony, developed upon fresh media with abundant food, the stalks
are crowded together and shortest. At the margin the scattered fruiting
heads are borne upon stalks which may be twice to several times the
length of those in the center. If the center of the young colony be taken
as typical of the race, the strains in culture fall roughly into three groups :
(1) Short-stalked, with stalks 500 to i,oooju; (2) intermediate, with
stalks 1,000 to 3,ooom; (3) long-stalked, with stalks 3 to several milli¬
meters in length. The first two groups, however, certainly shade into
each other.

1 Ridgway, Robert. Color Standards and Color Nomenclature. PI. 46, 13"". Washington, D. C., 1912.

2 Saito, Kendo. Mikrobiologtsche Studien iiber die Zubereitung des Batatenbranntweines auf der Insel
Hachijo (Japan). In Centbl. Bakt. [etc.], Abt. 2, Bd. 18, No. 1-3, p. 31. 1907.

* Linossier, Georges. Sur une h&natine veg^tale, l'aspergilline. In Compt. Rend. Acad. Sci. [Paris],
t. 112, no. 15, p. 807-808. 1891.

- Sur une h&natine vegetale: l’aspergilline, pigment des spores de l’Aspergillus niger. In Compt.

Rend. Acad. Sci. [Paris], t. 112, no. 9, p. 489-492. 1891.

Oct. 2, I916

Aspergillus niger Group

Heads. — Conidial heads vary exceedingly in size; the common forms
show simultaneously mature heads varying from 100 to 400/A in diameter.
This maximum becomes 1,000/1 or more in the gigantic forms. The
chains of conidia at first radiate uniformly, but as they lengthen they
adhere into black masses or columns which separate more and more as size
increases.

VesiceE. — The vesicle or enlarged apex of the stalk varies commonly
from 20 to 50 /a, but occasionally reaches 80 to ioo/a; it is continuous with
the lumen of the stalk, thick -walled, and with walls and often contents
yellow to brown in age.

Sterigmata. — The vesicle is fertile over its whole surface and bears
sterigmata usually in two series. Examination of young or growing
colonies commonly shows individual heads producing conidia upon
sterigmata of the first series, other heads with both simple and branched
sterigmata, and heads with well-differentiated primary sterigmata, each
bearing its quota of three to several secondary sterigmata. Here, as
elsewhere, the variations of measurement in the same colony destroy all
faith in such figures as an exact means of separating forms. The classic
descriptions of A . niger as summarized by Wehmer 1 give the primary
sterigmata as 26 by 4 to 5 /a, the secondary as 8 by 3/A. The secondary
sterigmata vary perhaps within limits of 6 to io/a by 2 to 4 /a in the whole
series, with 8 by 3/4 as a fair average figure. The variation in primary
sterigmata makes a length of 26/A an occasional average for sleeted heads
only. Examination of many heads in some strains gives lengths of primary
sterigmata averaging between 12 and 20/A with an occasional longer cluster
and frequently much of this variation occurs within the individual head.
In another group the maximum length lies between 20 and 30/A. Again,
in a few strains, lengths of 40 to 6o/a are seen, while two of this series in
their largest heads show primary sterigmata up to 120/t in length, as
given by Bainier 2 for S. carbonaria.

Conidia. — The formation of conidia in the A. niger series follows the
process described by Thom 3 for Penicillium spp. This probably agrees
with the process designated as endogenous by Bainier and Sartory 4 in
that each conidium is first cut off from the conidial tube of the sterigma,
then rounds itself up after secreting for itself a new wall, while the original
cell wall is frequently distinguishable between the ripe conidia as a con¬
nective. Ripe conidia are subglobose, with, in most strains, a variation
in the same culture of about i/a in diameter; some run from 2.5 to 3.6/A,

1 Wehmer, Carl. Die Pilzgattung Aspergillus in morphologischer, physiologischer imd systematischer
Beziehung unter besonderer Beriicksichtigung der mitteleuropaeischen Species. V. Systematik. C.
Schwarzbraune Arten. In Mem. Soc. Phys. et Hist. Nat. Geneve, t. 33, pt. 2, no. 4, p. 103-107. 1901.

2 Bainier, Georges. Sterigmatocystis et nematogonum. In Bui. Soc. Bot. France, t. 27 (s. 2, t. 2), p.

29^30. 1880.

8 Thom, Charles. Conidium production in Penicillium. In Mycologia, v. 6, no. 4, p. 211-215, 1 fig.
1914-

4 Bainier, Georges, and Sartory, Auguste. Etude d’un Aspergillus pathogene (Aspergillus fumigatoides
n. sp.). In Bui. Soc. Mycol. France, t. 25, fasc. 2, p. 112. 1909.

Journal of Agricultural Research

Vol. VII, No. i

others from 3 to 4 /x, a few from *3 to 4.5/i, and in A. carbonarius from 6 to
10/x. Conidial walls are typically rough, with irregular, but usually
oblong or barlike masses of coloring matter [the aspergillin of Linos-
sier 1 ] which run lengthwise of the conidial chain, and are absent at the
ends of the spore or are replaced there by the connective when such is
present. In some forms these bars are more or less completely broken
up to form irregularly disposed rough tubercles. In experiments this
coloring substance was dissolved by soaking in hot water, after which
the original outer cell wall became visible. This experiment indicates
that the difference in color between strains is due to the varying amount
of aspergillin deposited between the outer, or primary, and inner, or
true, spore walls. Upon careful study, A. fuscus (one of SchiemamTs2
mutants from A. niger) showed delicate traces of the typical bars of
color; A. cinnamomeus with its much lighter color and smooth spores
showed only diffused, not localized color.

Details in colonies of the same strain differ in successive cultures.
These differences are nearly all quantitative, but they indicate great
power of response to the stimulation of environment. The mutants,
A. cinnamomeus and A. fuscus , separated by Schiemann,2 differ from
the usual form only in intensity of color, yet maintain these characters
consistently in culture. A. carbonarius (4030.1) is a gigantic form in
which the proportions are approximately quadrupled, while No. 2580
(A. strychni ?) shows the same measurements except that the conidia re¬
main with a maximum diameter of 4 to 4.5^. Perhaps A. niger is a form
comparable to Oenothera spp. in its tendency to produce mutants. There
arise, therefore, a few forms with characters sufficiently tangible to
separate by description. In a majority of the strains met with in cul¬
ture, morphological differences are not sharp enough for diagnostic pur¬
poses. Nevertheless great and fairly stable differences in physiological
activity are found among them. Two forms morphologically alike may
thus differ greatly in economic importance.

Nomenclature. — Current literature has accepted the name “ As -
pergillus niger ” Van Tieghem,3 for the black species of Aspergillus.
Sterigmatocystis autacustica Cramer, obtained from a human ear, was
undoubtedly one of the series of organisms, but was insufficiently
described. A. ficuum (Reich.) P. Hennings, first named uUstilago
ficuum” by Reichardt in 1867, differs slightly in measurements, but the
presumption of identity is based upon the constant occurrence of the

1 Linossier, Georges. Sur une h^matine v£g4tale l’aspergilline. In Compt. Rend. Acad. Sci. [Paris];

t. 112, no. 15, p. 807-808. 1891.

- Sur une hdmatine v^getale: raspergilline, pigment des spores de l’Aspergillus niger. In Compt.

Rend. Acad. Sci. [Paris], t. 112, no. 9, p. 489-492. 1891.

2 Schiemann, Elizabeth. Op. cit.

8 Van Tieghem, P. E. L. Rec&erches pour servir k l’histoire physiologique des Mucedin^es. Fermenta¬
tion gallique. In Ann. Sci. Nat. Bot., s. 5, t. 8, no. 4, p. 240. 1867.

Oct. 2, 1916

A spergillus niger Group

usual type of a black species of Aspergillus in practically pure culture
in figs. A. phoenicis (Corda) Patouillard and Delacroix, recorded as
Ustilago phoenicis by Corda1 2 in 1840, differs also in the shape and smooth¬
ness of its conidia, although other data point to close relationship. Van
Tieghem in his article discussed A. nigrescens of Robin (1848) and A .
nigricans of Wreden (1867) and offered reasons for separating A. niger
from these organisms, neither of which is adequately described.

The conspicuous character of these black colonies and their frequency
in all sorts of decaying food make it difficult to believe that the species
remained undescribed until 1867. A review of the literature of As¬
pergillus spp., and of those generic names used interchangeably with it
by some of the earlier botanists, has included Micheli, Linnaeus, Sowerby,
Persoon, Link, Ehrenberg, Fries, Greville, Corda, the Tulasnes, and Bon-
orden. Several names are found which might refer to this group, but
are unaccompanied by either figures or descriptions which can be
definitely shown to represent this species. The usage of Raulin 3 probably
suggests the true explanation. The reference of A. niger to the genus
Ascophora (syn. Rhizopus) by the French workers preceding Van
Tieghem points to the conclusion that A. niger had been constantly
confused with the mucors. The recognition by Robin, Cramer, and
Wreden of black forms of Aspergillus spp. as the cause of mycotic
diseases in the ear seems to have led directly to the recognition of the
separateness of the black species of Aspergillus from the black mucors
as a cosmopolitan organism.

The generic name “Sterigmatocystis,” proposed by Cramer,3 is based
upon the assumption that in the forms of Aspergillus to which it was
applied, there were always two sets of cells (basidia and sterigmata of
some authors; primary and secondary sterigmata of Wilhelm and
Wehmer) between the vesicle, or enlarged end of the conidiophore, and
the actual conidial chains. This distinction was disregarded by Wilhelm,4
reaffirmed by Eidam,5 and again discarded by Wehmer.6 Examination
of thousands of cultures does not, in the opinion of the writers, justify
the use of the separate generic name “ Sterigmatocystis.” There appears,
therefore, no good reason to displace the name “ Aspergillus niger ” for
at least a section of the group.

In a classification on color alone A. cinnamomeus and A. fuscus of
Schiemann would be excluded from the series. If structure and meas-

1 Corda, A. C. I. leones Fungorum . . . t. 4, p. 9, pi. 3, fig. 26. Pragae, 1840.

2 Raulin. Etudes chimiques sur la vegetation des Muc&iinees, particuli£rement de lf Ascophora nigrans.

In Compt. Rend. Acad. Sci. [Paris], t. 57, no. 4, p. 228-230. 1863.

3 Cramer, Carl. Ueber eine neue Fadenpilzgattung: Sterigmatocystis. Cramer. In Vrtljschr. Naturf .

Gesell. Zurich, Jahrg. 4, Heft 4, p. 325-337, pi. 2. 1859.

i Wilhelm, K. A. Beitrage zur Kentniss der Pilzgattung Aspergillus. 70 p. Strassburg-Berlin, 1877.
Inaug. Diss.

6 Eidam, Eduard. Zur Kenntniss der Entwicklung bei den Ascomyceten. In Beitr. Biol. Pfianz.,
Bd. 3, p. 377-433. pl. 19-23. 1883.

9 Wehmer, Carl. Op. cit., p. 28, 34-35*

io Journal of Agricultural Research voi. vii, No. i

urement of conidial apparatus be used to define the group, certain strains
of A . ochraceus Wilhelm and A . wentii Wehmer certainly fall within it.
All of these forms are clearly related and may properly constitute a
section of the genus Aspergillus without venturing a guess as to their
genetic connection.

The group, therefore, may be held to consist of a series of forms, some
of which seem to be connected so closely by intergrading forms as to
make separation difficult if not impossible. Other members of the group
notably represented by the mutations of Schiemann1 show permanent
and striking differences. A. carbonarius of Bainier may be a similar
case. Upon morphology alone we may therefore be justified in retain¬
ing certain specific names as well-defined representatives of the sections
of the group. The arrangement proposed begins with A. nanus Mont, for
the diminutive form, A. niger Van Tieghem for the most numerously
occurring section with primary sterigmata 20-30/* in length, A. phoenicis
(Corda) Pat. and Delacr. with primary sterigmata about 50/*, A. pul-
verulentus McAlpine, or A . strychni Lindau, with very long sterigmata,
and end our series with A . carbonarius , which has the long sterigmata
and very large conidia.

In suggesting that the following names be retained as designating
representative cultures falling within fairly well-defined sections of the
group, it remains uncertain how many names may ultimately be re¬
quired to designate forms permanently considered as species. The
sections and citations follow:

I. — FORMS WITH SIMPLE STERIGMATA UP TO 20/* IN LENGTH

A. nanus Monthgne, 1856, Syll. Gen. Spec. Crypt., p. 300, no. 11 12. Saccardo,
1886, Syll. Fung. v. 4, p. 71.

11a (black or brown). — FORMS with both primary and secondary
sterigmata; primary 20-30/* in length

A. niger Van Tieghem, 1867, in Ann. Sci. Nat. Bot., s. 5, t. 8, no. 4, p. 240. As
probable synonyms the following may be listed: 5. antacustica Cramer, 1859, in
Vrtljschr. Naturf. Gesell. Zurich, Jahrg. 4, Heft 4, p. 325; A. echinosporus Sorok.,
Paras.2 p. 40, pi. 7, fig. 82-87. Ref. in Saccardo, 1895, Syll. Fung., v. 11, p. 592;
A . ficuum (Reich.) Hennings, 1895, in Hedwigia, Bd. 34, Heft 2, p. 86; A.fuliginosus
Peck, 1873, *n Bui. Buffalo Soc. Nat. Sci., v. 1, p. 69; 1874, in 26th Ann. Rpt. N. Y.
State Mus. Nat. Hist. [1872], p. 79; A. nigrescens Robin, 1853, Hist. Nat. V6g. Paras.,
p. 518, atlas, pi. 5, fig. 2; A. nigricans Wreden, 1867, in Compt. Rend. Acad. Sci.
[Paris], t. 65, no. 9, p. 368. A. phaeocephalus (Durieu and Montagne), 1881, in
Saccardo, Fungi Ital., fig. 903; 1886, in Saccardo Syll. Fung., v. 4, p. 76; S. pseudo -
nigra Costantin and Lucet, 1903, in Bui. Soc. Mycol. France, t. 19, fasc. 1, p. 33-44;
A . ustikigo Beck, 1888, in Wawra. Itin. Princ. S. Coburgi, T. 2, p. 148; 1892, in
Saccardo, Syll. Fung., v. 10, p. 526; A. welwitschiae (Bresadola) Henn.

1 Schiemann, Elisabeth. Op. cit.

2 Not seen by the author.

Oct. 2, 1916

Aspergillus niger Group

Jib. — FORMS DIFFERING FROM A. NIGER ONLY IN COLOR

A. cinnamomeus Schiemann, 1912, in Ztschr. Induk. Abstain. 11. Vererbungslehre,
Bd. 8, Heft p. 1-35, 16 fig., 2 pi. (1 col.).

A.Schiemanni (Schiem.) Thom, n. comb. Syn. A. fuscus Schiem. (Schiemann,
Elizabeth. Op. cit., 1912).

III. — FORMS WITH BOTH PRIMARY AND SECONDARY STERIGMATA; PRIMARY

ABOUT 50 jU IN LENGTH

A. phoenicis (Corda) Patouillard and Delacroix, 1891, in Bui. Soc. Mycol. France,
t. 7, p. 118-120, pi. 9; syn. Ustilago phoenicis Corda, 1840, leones Fung., t. 4, p. 9,
pi. 3, fig. 26.

IV. — FORMS WITH BOTH PRIMARY AND SECONDARY STERIGMATA; PRIMARY

UP TO I20JU IN LENGTH

A. pulverulentus McAlpine, 1896, in Agr. Gaz. N. S. Wales, v. 7, pt. 5, p. 302;
probable synonym A. strychni Lindau, 1904, in Hedwigia, Bd. 43, Heft 5, p. 306-307.

V. — FORMS WITH BOTH PRIMARY AND SECONDARY STERIGMATA; PRIMARY

UP TO 120 H IN LENGTH, SPORES DOUBLE SIZE

A. carbonarius (Bain.), Thom (Bainier, 1880, in Bui. Soc. Bot. France, t. 27 (s. 2,
t. 2), p. 27-28).

The following list contains the names, original and often secondary
citations, for the forms described as black or brown. Among these are
some forms certainly not related to A . niger, but described in terms which
might suggest such relationship unless critically examined. Wherever
possible, the proper placing of the form is suggested. One original
reference, that of Sorokin, has not been seen; all others have been crit¬
ically examined. The forms are arranged alphabetically to species
without reference to the describees placing in Aspergillus or Sterigmato-
cystis.

S. antacustica Cramer (Cramer, Carl. Ueber eine neue Fadenpilzgattung: Sterig-
matocystis. Cramer. In Vrtljschr. Naturf. Gesell. Zurich, Jahrg. 4, Heft 4,
p. 325. 1859).

In the external ear of man; considered A. niger by Wilhelm (Wilhelm, K. A.
Beitrage zur Kenntniss der Pilzgattung Aspergillus. 70 p. Strassburg-Berlin,
1877. Inaug. Diss.).

A. airopurpureus A. Zimm. (Zimmermann, A. Ueber einige an tropischen Kultur-
pflanzen beobachtete Pilze. II. In Centbl. Bakt. [etc.], Abt. 2, Bd. 8, No. 5, p.
218. 1902).

Not in the A. niger series; sterigmata and spores suggest the A. glaucus series.
A. batatae Saito (Saito, Kendo. Mikrobiologische Studien fiber die Zubereitung des
Batatenbranntweines auf der Insel Hachijo (Japan). In Centbl. Bakt. [etc.],
Abt. 2, Bd. 18, No. 1/3, p. 34. 1907).

Morphology given suggests A . niger in details, but in colony characters indi¬
cates a close relationship to A. wentii , Wehmer. This is confirmed by culture
of a strain of A . wentii which showed this morphology.

Journal of Agricultural Research

Vol. VII, No. i

A. brunneus Bain. (Bainier, Georges. Sterigmatocystis et nematogonum. In Bui.
Soc. Bot. France, t. 27 (s. 2, t. 2), p. 29. 1880).

Colonies at first green, then black-brown; conidia 15 in diameter; probably
A . glaucus series.

A, brunneus Delacr. (Delacroix, Georges. Esp&ces nouvelles de Champignons in-
ferieurs. In Bui. Soc. Mycol. France, t. 7, p. 109, pi. 7. 1891).

Related to the preceding by color,

5. carbonaria Bain. (Bainier, Georges. Sterigmatocystis et mematogonum. In Bui.
Soc. Bot. France, t. 27 (s. 2, t. 2), p. 27-28. 1880).

A culture from Dr. Blakeslee reproduces the morphology recorded by Bainier.
A redescription of the form is therefore givep.

A. carbonarius (Bainier), Thom., n. comb.

Colonies grown in Czapek’s solution agar show vegetative mycelium white or
with some yellow in submerged areas, broadly spreading, more or less zonate;
sclerotia produced upon the surface of the substratum in old cultures; fruiting
areas carbon-black; stalks colorless below, yellow to yellow-brown toward the
apex, 4 to 6 mm. or longer and up to 254 in diameter, with walls smooth, up to
4ju in thickness; heads globose varying in diameter up to 500/*, vesicles up to
QOfi in diameter, fertile over entire surface, commonly with contents yellow-
brown to black and in old heads forming with the primary sterigmata a hard brit¬
tle, carbonaceous mass; sterigmata in two series, primary sometimes 1 -septate,
from 20 to 40M long in young or small heads and up to 120 m long in large heads,
5 to 13M in diameter at the apex, secondary 8 to 14M by 3 to 64 ; conidia at first
smooth becoming rough When ripe, 5 . 5 to 10. 5m in diameter. Colonies grow well
upon all culture media used, with temperature optimum below 37 0 C. Growth
at 370 C. slow and more or less dwarfed.

A . cimmerius Berk, and Curtis (Berkeley, M. J. Notices of North American fungi.
In Grevillea, v. 3, no. 27, p. 108, no. 656. 1875. Saccardo, P. A. Sylloge

Fungorum . . . v. 4, p. 71. Patavii, 1886).

The color reference “aterrimus” and the spore size, elliptical 7 m, suggest A.
carbonarius , but the data are inadequate.

A. cinnamomeus Schiem. (Schiemann, Elisabeth. Mutationen bei Aspergillus niger
Van Tieghem. In Ztschr. Induk. Abstam. u. Vererbungslehre, Bd. 8, Heft 1/2,
p. 1-35, 16 fig., 2 pi. (i col.). 1912):

Differentiated by color and smoothness of spores, but maintains these distinc¬
tions uniformly. Obtained as a mutant from A. niger by Schiemann.)

A . cookeii (Cooke) Sacc. (Saccardo, P. A. Sylloge Fungorum ... v. 4, p. 71. Pa¬
tavii, 1886. )=A. mucoroideus Cooke (Cooke, M. C. Australian fungi. In Gre¬
villea, v. 12, no. 61, p. 9. 1883).

Not recognizable from the description, but suggests a form of Syncephalas-
trum which we have had in culture.

A. echinosporus Sorok. (Sorok. Paras.1 p. 40, pi. 7, fig. 82-87. Ref* in Saccardo, P. A.
Sylloge Fungorum , . . v. n, p. 592. Patavii, 1895).

The description suggests a Haplographium.

A. ficuum (Reich.) Henn. (Hennings, P. C. Ustilago Ficuum Reich. = Sterigmato¬
cystis Ficuum (Reich.) P. Henn. In Hedwigia, Bd. 34, Heft 2, p. 86. 1895.)=

Ustilago ficuum Reich. (Reichardt, H. W. Ein neuer Brandpilze. In Ver-
handl. K. K. Zool. Bot. Gesell. Wien, Bd. 17, p. 335. 1867).

Regarded as A . niger by Wehmer (Wehmer, Carl. Zur Kenntnis einiger Asper-
gillus-Arten. In Centbl. Bakt. [etc.], Abt. 2, Bd. 18, No. 13/15, p. 394-395*
1907).

Slight differences in morphology are reported by Hennings, but disregarded by
Wehmer.

1 Not seen by the author.

Oct. 2, 1916

Aspergillus niger Group

A . fuliginosus Peck (Peck, C. H. Descriptions of new species of fungi. In Bui.
Buffalo Soc. Nat. Sci., v. 1, p. 69. 1873. Peck, C. H. Reportof the botanist.

In 26th Ann. Rpt. N. Y. State Mus. Nat. Hist. [1872], p. 79. 1874).

No valid data are offered for separating this strain from A. niger.

A.fuscus Bonorden (Bonorden, H. F. Beitrage zur Mykologie. In Bot. Ztg., Jahrg.
19, no. 29, p. 202. 1861).

This is probably a very common form of Haplographium.

S.fusca Bain. (Bainier, Georges. Sterigmatocystis et nematogonum. In Bui. Soc.
Bot. France, t. 27 (s. 2, t. 2), p. 29. 1880).

The morphology recorded differs from A . niger in the size of the conidia, “ less
than 9.4M. ” The form has not yet been found by us.

A.fuscus Schiem. (Schiemann, Elisabeth. Mutationen bei Aspergillus niger van
Tieghem. In Ztschr. Induk. Abstam. u. Vererbungslehre, Bd. 8, Heft 1/2,
p. 1-35, r6 fig., 2 pi. (r col.). 1912).

This mutant from A . niger is distinguished by the color and smoothness of its
spores from the parent strain. The name of this and the preceding form are
invalidated by the occurrence of the name A.fuscus Bonorden.

A . japonicus Saito (Saito, Kendo. Nachtrag zu der Abhandlung “ Untersuchungen
iiber die atmosphaerischen Pilzkeime, I.” In Bot. Mag. [Tokyo], v. 20, no. 233,
p. 61, 5 fig. 1906).

This species as described differs from A . niger in having but one set of sterig-
mata and showing granules at times upon the walls of the conidiophore. It is
evidently very closely related to the regular forms of A. niger.

A. luchuensis (Inui, T. Untersuchungen iiber die niederen Organismen welche sich
bei der Zubereitung des alkoholischen Getrankes. "Awamori” betheiligen.
In Jour. Col. Sci. Imp. Univ. Tokyo, v. 15, pt. 3, p. 469. pi- 22> fig- 1-8 • I901)-

This form differs from A. niger by showing green colors when young and by
morphological details. It does not, therefore, belong in this series.

A. mucoroideus Cooke (Cooke, M. C. Australian fungi. In Grevillea, v. 12, no. 61,
p. 9. 1883).

A . cookei Sacc.

A. mucoroideus Corda (Corda, A. C. I. leones Fungorum . . . t. 2, p. 18, pi. n, fig.
76. Pragae, 1838).

This is a different organism and not in this group. The description suggests
A . castaneus Patterson.

A nanus Montague (Montague, J. F. C. Sylloge Generum Specierumque Cryptogama-
rum . . . p.300, no. 1112. Parisiis, 1856. Saccardo, P. A.- Sylloge Fungorum

... v. 4, p. 71. Patavii, 1886).

The description clearly distinguishes this organism as a member of the group
and as having a single series of sterigmata about 1 5/z in length and spores 3 At m
diameter.

A. niger Van Tieg. (Van Tieghem, P. E. E. Recherches pour servir a l’histoire
physiologique des Muc6din6es. Fermentation Gallique. In Ann. Sci. Nat.
Bot., s. 5, t. 8, no. 4, p. 240. 1867. )=S. nigra Van Tieg. (Van Tieghem, P. E. E.

Sur le developpement de quelques Ascomycetes. In Bui. Soc. Bot. France, t.
24, p. 102-103. 1877).

Described by Van Tieghem as the active agent in fermenting tannin solutions.
A. nigrescens Robin (Robin, C. P. Histoire Naturelle des V^getaux Parasites . . .
p. 518; atlas, pi. 5, fig- 2. Paris, 1853).

The organism of Robin has been called A. niger by Wilhelm (Wilhelm, K. A.
Beitrage zur Kenntnis der Pilzgattung aspergillus.' 70 p. Strassburg-Berlm,
1877 Inaug. Diss.), A. fumigatus by Siebenmann (Siebenmann, Friedrich.

Journal of Agricultural Research

Vol. VII, No. i

Die Fadenpilze Aspergillus flavus, niger u. fumigatus; Eurotium repens (u.
Aspergillus glaucus) und ihre Beziehungen zu Otomycosis Aspergillina. p. 82,
Wiesbaden, 1883 . Inaug. Diss.), and is judged undeterminable from the informa-
given, by Wehmer (Wehmer, Carl. Zur Kenntnis einiger Aspergillus- Arten.
In Centbl. Bakt. [etc.], Abt. 2, Bd. 18, No. 13/15, p. 394-395. 3 fig. 1907).
Robin's figures represent A . fumigatus much more closely than A . niger.

A . nigricans Wreden (Wreden, Robert. Recherches sur deux nouvelles esp&ces de
v£g6taux parasites (Aspergillus flavescens et Aspergillus nigricans) de l’homme.
In Compt. Rend. Acad. Sci. [Paris], t. 65, no. 9, p. 368. 1867).

This is regarded as A. niger by Wilhelm (Wilhelm, K. A. Beitrage zur Kennt-
niss der Pilzgattung Aspergillus. 70 p. Strassburg-Berlin, 1877. Inaug. Diss.)
by Siebenmann (Siebenmann, Friedrich. Die Fadenpilze Aspergillus flavus,
niger u. fumigatus; Eurotium repens (u. Aspergillus glaucus) und ihre Bezie¬
hungen zu Otomycosis Aspergillina. p. 82. Wiesbaden, 1883. Inaug. Diss.);
and by Wehmer (Wehmer, Carl. Zur Kenntnis einiger Aspergillus- Arten, In
Centbl. Bakt. [etc.], Abt. 2, Bd. 18, No. 13/15, p. 394-395. 1907).

A. nigricans Cooke (Cooke, M. C. Some remarkable moulds. In Jour. Quekett
Micros. Club, s. 2, v. 2, no. 12, p. 140, pi. 9, fig. 3. 1885. Saccardo, P. A. Sylloge

Fungorum . . . t. 4, p. 76. Patavii, 1886).

The occurrence in the human ear of forms known to be identical with A . niger
makes a separate name unnecessary. The description of this organism corre-

rnds with A. nanus Montagne, except that the spores are given as 5m in
meter.

A. phaeocephalus Durieu and Montagne (Saccardo, P. A. Fungi Italic! ... fig. 903.
Patavii, 1881. Saccardo, P. A. Sylloge Fungorum , . . v. 4, p. 76. Patavii,
1886).

No differences are given to warrant a separation of this form from A. niger.

S. phoenicis (Corda) Patouill. and Delacr. (Patouillard, Narcisse, and Delacroix,
Georges. Sur une maladie des dattes produite par le Sterigmatocystis Phoenicis
(Corda) Patouill. et Delacr. In Bui. Soc. Mycol. France, t. 7, p. 119, pi. 9.
1891.)= Ustilago phoenicis Corda (Corda, A. C. I. leones Fungorum . . . t. 4,
p. 9, pi. 3, fig. 26. 1840).

In describing the black Aspergillus as found upon dates, the authors com¬
pared this material to specimens in the museum labeled (t Ustilago phoenicis ”
and attributed to Corda, thus establishing the identity of the organism of Corda,
which was not recognizable from any previous references. A. ustilago Beck is
described with the same measurements. The measurements of sterigmata place
this form in the section of the group with primary sterigmata 50-60M in length.

5. pseudo-nigra Costantin and Eucet (Costantin, and Eucet. Sur le Sterigmatocystis
pseudonigra. In Bui. Soc. Mycol. France, t. 19, fasc. 1, p. 33-44. 1903).

This is regarded as A . niger by Wehmer (Wehmer, Carl. Zur Kenntnis einiger
Aspergillus-Arten. In Centbl. Bakt. [etc.], Abt. 2, Bd. 18, No. 13/15, p. 394-
395. 1907).

The distinctions proposed by the authors are based upon pathogenicity and
cultural reactions, not upon definite morphological characters.

S. pulverulenta McAlp. (McAlpine, Daniel. Australian fungi. In Agr. Gaz. N. S.
Wales, v. 7, pt. 5, p. 302. 1896).

This is regarded as A . niger by Wehmer (Wehmer, Carl. Zur Kenntnis eini¬
ger Aspergillus-Arten. In Centbl. Bakt. [etc.], Abt. 2, Bd. 18, No. 13/15,
p. 394-395 • 1907) after examining material received from McAlpine.

5. purpurea Van Tieg. (Van Tieghem, P. E. E. Sur le d6veloppement de quelques
Ascomyc£tes. . In Bui. Soc. Bot. France, t. 24, p. 101-103. 1877).

The information given is regarded as insufficient by Wehmer to separate this
form (Wehmer, Carl. Die Pilzgattung Aspergillus in morphologischer und
systematischer Beziehung unter besonderer Beriicksichtigung der mitteleuro-
paeischen Species. V. Systematik C. Schwarzbraune Arten. In M6m. Soc.
Phys. et Hist. Nat. Geneve, t. 33, pt. 2, no. 4, p. 103-111. 1901).

Oct. 2, 1916

Aspergillus niger Group

A. purpureofuscus Fries (Fries, E. M. Systemata Mycologicum. v. 3, p. 388. Gry-
phiswaldae, 1829).

The hyphae are described as purpureofuscus. This suggests certain of the
A. glaucus series.

A. purpureofuscus Schw. (Schweinitz, L. D. von. Synopsis fungorum in America
boreali media degentium. Secundum observationes. In Trans. Amer. Phil.
Soc., n. s. v. 4, p. 282, no. 2680. 1834. Saccardo, P. A. Sylloge Fungorum

... v. 4, p. 68, Patavii, 1886).

Not to be regarded as a member of this series from the description given.

A. Schiemanni (Schiemann) Thom, n. comb.=A. fuscus Schiemann (Schiemann,
Elisabeth. Mutationen bei Aspergillus niger Van Tieghem. In Ztschr. Induk.
Abstam. u. Vererbungslehre, Bd. 8, Heft 1/2, p. 1-35, 16 fig., 2 pi. (1 col.).
1912).

A new combination was necessary because the specific name fuscus had been
already used by Bonorden for a species of Aspergillus and fusca by Bainier for
a species of Sterigmatocystis. This form is distinguished only by its fuscous
conidia which are smooth or very delicately roughened.

A. sirychni Lindau (Lindau, Gustave. Aspergillus (Sterigmatocystis) strychni nov.
spec. In Hedwigia, Bd. 43, Heft 5, p. 306-307. 1904).

This is regarded as A. niger by Wehmer (Wehmer, Carl. Zur Kenntnis einiger
Aspergillus- Arten. In Centbl. Bakt. [etc.], Abt. 2, Bd. 18, No. 13/15, p. 394-
395- W)-

The details of structure reported by Lindau for this form have been found in a
culture (No. 2580) obtained in a study of red peppers (capsicum species) from
Barcelona, Spain. The interior of these fruits had been destroyed and there
remained a shell containing a mass of Aspergillus spores. The morphology
obtained in cultures is identical with that of A. carbonarius, as described above,
except that the conidia remain 4/* in diameter, as in the ordinary form of A .
niger.

A. ustilago Beck (Wawra, Heinrich. Itinera Principum S. Coburgi. T. 2, p. 148.
Wien, 1888. Saccardo, P. A. Sylloge Fungorum . . . v. 10, p. 526. Patavii,
1892).

The description relates this form to the section of the A . niger group showing
sterigmata 50JU or more in length.

A . violaceo-fuscus Gasper. (Gasperini, G. Sopra un nuovo morbo che attacca i limoni
e sopra alcuni ifomiceti. In Atti Soc. Toscana Sci. Nat. Pisa, Mem. , v. 8, fasc. 2 ,
p. 326. 1887).

The conidia are reported as ovoid, 3 to 6 by 3 to 3.5^. This form' probably
belongs elsewhere.

A. welwitschiae (Bresadola) Hennings.

This is regarded by Wehmer as A. niger (Wehmer, Carl. Zur Kenntnis
einiger Aspergillus- Arten. In Centbl. Bakt. [etc.], Abt. 2, Bd. 18, No. 13/15,

p- 394-393- 1907)-

No data are reported which warrant the establishment of a separate species.

SOME EFFECTS OF THE BLACKROT FUNGUS, SPHAE-
ROPSIS MALORUM, UPON THE CHEMICAL COMPOSI¬
TION OF THE APPLE1

By Charles W. Culpepper, Biologist , West Tennessee State Normal School; Arthur
C. Foster, Assistant in Botany , North Carolina Agricultural and Mechanical College;
and Joseph S. Caldwell, Specialist in Fruit By-Products Investigations, Washington
Agricultural Experiment Station

INTRODUCTION

Various investigators have concerned themselves with the taxonomy,
morphology, and life history of Sphaeropsis mcUorum Peck. This fungus
occurs upon the twigs and branches of the apple (Malus sylvestris) , produc¬
ing cankers (15) ; 2 upon the foliage, where it produces a characteristic leaf-
spot (22, 24); and upon the fruits, where its growth results in a decay
that is unique because the diseased tissues undergo a rapid and charac¬
teristic blackening, because there is neither softening nor breaking down
of the affected areas, and because loss of water and consequent shriveling
do not begin until some time after the entire fruit has become involved.
The occurrence of the disease upon the fruits of the apple appears to be
very much less common in the North and East than in the South, where
it frequently destroys 25 to 50 per cent of the crop (27).

The fact that the organism is able to attack tissues which vary so
widely in chemical composition indicates that a study of its physiology
should yield results of considerable interest. Very little attention has
thus far been paid to the nature and extent of the chemical changes
produced in the fruit by this organism, and it was to obtain information
in regard to these changes that the present study was undertaken.

The present paper presents the results of a quantitative analytical
study of the chemical composition of normal mature apples in comparison
with others of the same variety attacked by Sphaeropsis malorum. The
work was done in the laboratories of plant physiology of the Alabama
Polytechnic Institute and Agricultural Experiment Station in the summer
and autumn of 1915. The problem was suggested to Messrs. Foster and
Culpepper, who were at that time graduate assistants in the laboratory,
by Dr. Caldwell, who outlined the general methods of study to be em¬
ployed and supervised the work. The major portion of the analytical
work was done by Messrs. Foster and Culpepper. Upon their removal to
other laboratories at the end of the summer, certain portions of the
analyses were necessarily left incomplete; these have been completed,

1 Published with the permission of the Director of the Alabama Agricultural Experiment Station.

2 Reference is made by number to “ Literature cited,” p. 39-40.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.
fo

(!7)

Vol. VII, No. 1
Oct. 2, 1916
Ala. — 3

55855°— 16 - 2

Journal of Agricultural Research

Vol. VII, No. i

some portions of the work have been repeated, and the results have been
prepared for publication by Dr. Caldwell, who is responsible for the
conclusions drawn from the analytical data.

ANALYTICAL METHODS

The analytical methods employed are based upon those devised by
Waldemar Koch and his pupils (9, 10, n, 12) for use in the quantitative
analysis of animal tissues. With the subsequent application of these
methods to the study of plant tissues, a considerable number of modifica¬
tions of the original scheme of analysis have been found necessary, and
these have been incorporated into the analytical procedure. For some
of these we are indebted to Dr. Fred C. Koch or to workers in his labora¬
tory. Others of them were worked out by the writers in the course of
this and similar investigations. A number of determinations of indi¬
vidual constituents have also been made by the employment of special
methods, so that the results reported are believed to present a rather
complete statement of the chemical differences between sound and
diseased fruits.

The apples used in the experiments were of the Red Astrachan variety
and were grown in the orchards of the Alabama Agricultural Experi¬
ment Station. The various lots employed in the work were gathered
between June 17 and June 23, all those employed in any one series of
analyses having been collected at one time and from the same tree. At
the earliest date mentioned, the sound fruits were fully mature but had
not begun to soften. As all the trees of this variety showed very severe
blackrot infection, it was possible to secure at any desired time and
from a single tree specimens showing any desired stage of the disease,
together with sound fruits of the same age.

The procedure employed in collecting samples was the same in every
case; a considerable number of wholly sound fruits, an equal number
of fruits, each of which was approximately half-decayed, and a third
lot, all of which were entirely decayed, were picked from the trees and
carried at once to the laboratory. A careful examination of each fruit
was then made in order to be sure that the diseased apples were free from
organisms other than Sphaeropsis malorum, and three lots of material
were prepared. The first lot, designated throughout this paper as
“normal,” was made up of selected portions of perfectly sound, normal
fruits; the second lot consisted of decayed portions of fruits, one-half or
three-fifths of which were invaded by blackrot, and is here termed <£ half-
decayed’ J; the third lot, made up of fruits in which decay was complete,
is designated “wholly decayed.” Each lot of material was ground to a
pulp by passing it through a food chopper and completing the grinding
in a mortar. A portion was then removed, accurately weighed in a pre¬
viously dried and tared crucible, and employed in a moisture determina-

Oct. 2, I916

Effect of Blackrot Fungus on the Apple

tion. From the remainder, 100 gm. were weighed out, transferred to a
closely stoppered flask, and preserved in such a quantity of redistilled 95
per cent alcohol as sufficed to give an alcohol concentration of 85 to 87
per cent. As the moisture content ranged from 87.39 per cent in the
sound fruits to 86.14 per cent in completely decayed apples, each 100-
gm. sample received 830 to 860 c. c. of alcohol. Preliminary moisture
determinations had been made in order to ascertain the approximate
water content. The whole operation of pulping, weighing samples, and
transferring to the alcohol could be carried out in 8 to 10 minutes, so
that little opportunity was offered for chemical changes resulting from
exposure of the ground tissues to the air. The material was allowed to
stand in the alcohol for seven days, with frequent shaking, and was occa¬
sionally warmed on a water bath to hasten the coagulation of the proteins
and the extraction of constituents readily soluble in alcohol. The analy¬
ses of any given sample of material were begun on the eighth day after
it had been collected.

The initial procedure in the method of analysis employed consists
of successive extractions of the material with alcohol, ether, water, and
alcohol, which divides the material into two portions, a portion consisting
of the constituents extracted by these solvents and designated, for a
reason which will presently appear, as “Fractions 1 and 2,” and an in¬
soluble residue designated “Fraction 3." The details of the method of
extraction follow.

The extractions were carried out in an apparatus originally designed
for the analysis of rubber insulating materials (6). This apparatus has
very great advantages over the ordinary Soxhlet apparatus in that it has
only two glass parts, a small Erlenmeyer extraction flask and an ex¬
traction cup bearing a side-arm siphon. The condenser, which is made of
block tin, fits within the neck of the flask and bears a flange which forms
a perfect seal, so that the dropping of cold water from the condenser upon
the flank is prevented. The extraction cup is suspended from the con¬
denser at such a height that the bottom is just above the level of the boil¬
ing solvent ; hence, the material is kept constantly at a temperature equal
to that of the liquid, which usually boils vigorously in the cup. The
absence of ground-glass connections obviates the danger of loss from
breakage, while the compact character of the apparatus permits the
carrying out of a large number of extractions simultaneously upon an
ordinary hot plate or water bath.

The extraction was begun by transferring the preserved material to
previously dried and weighed Schleicher and Schull extraction thimbles,
in which all the subsequent extractions were made. Three thimbles of
the size designated by the makers as “Hiilsen fur Extractions -apparate
Nr. 603, 30X80 mm./’, with the upper 20 mm. cut off, usually sufficed
to contain a sample of 100 gm. In filling, the thimbles were placed
in the glass siphon cups, which were then set into funnels dripping into

Journal of Agricultural Research

Vol. VII, No. i

flasks, and the material was transferred by filtering the alcohol used for
preservation repeatedly through the thimbles until it had been freed
from solid particles. As the alcohol used in preserving the samples
contained such constituents as were readily soluble in cold alcohol, it
constituted a cold extract and was preserved for addition to the products
of the subsequent extractions.

The extraction flasks were prepared by placing about 75 c. c. of redis¬
tilled 95 per cent alcohol in each, fitting the condensers, and warming
gently on an asbestos plate over a gas flame. As soon as all dripping
from the filled extraction cups had ceased, the material in each was pressed
down with a glass rod, a weighed circle of ashless filter paper was fitted
carefully over the top to prevent loss of finer particles, the cups were sus¬
pended from the condensers, and the flame so regulated that the filling
and emptying of the siphon cups occurred about three times per minute.
The extraction was continued for 12 hours; but at intervals of 3 to 4 hours
the flame was turned out, the solutions collected from the extraction
flasks, and equal quantities of 95 per cent alcohol introduced. This pro¬
cedure has the twofold advantage that it removes the possibility of altera¬
tion in the dissolved materials by prolonged boiling in alcohol, while it
at the same time prevents the slowing down of the extraction through
an elevation of the boiling point of the solvent.

At the close of the alcohol extraction, the flasks were emptied, washed
out with boiling 95 per cent alcohol, the material in the cups was pressed
with a rod to remove the alcohol as completely as possible, and the
extracts and washings combined and preserved. The extraction flasks
now received 75 c. c. of ether each, and a 12-hour extraction with ether
was made. While ether removes only very small quantities of material,
which, except in the case of very heavily cutinized tissues, is quanti¬
tatively removed in the first two hours of the extraction, long-continued
treatment with boiling ether greatly facilitates subsequent comminution
of the tissues.

At the end of the ether extraction the material was warmed to drive
off the ether, removed from the cups to a mortar, and ground into a
fine powder. Bits of cuticle and fragments of vascular tissue were cut
into fine bits with scissors and separately ground. The powdered mate¬
rial was then transferred to a stoppered flask with 100 c. c. of distilled
water, placed on a boiling water bath, and heated with frequent shaking
for 12 hours. Enough 95 per cent alcohol was then added to make the
concentration of the solution 85 per cent, and the heating on the water
bath was continued for a second 12-hour period. The solids were then
collected into the original extraction thimbles by filtering the solution
repeatedly through them, and a final alcohol extraction of 12 hours*
duration was made. The residue remaining afjter this extraction con¬
stituted the alcohol-ether-water-insoluble fraction 3. It was dried to
constant weight in the extraction thimbles, an aliquot part taken for

Oct. 2, 1916

Effect of Blackr ot Fungus on the Apple

duplicate or triplicate ash determinations, and the remainder preserved
in a desiccator for subsequent analysis in the manner to be presently
described.

The soluble portions of the material were now contained in the alcohol
originally employed for the preservation of the material, in that used for
the first extraction and for washing the flasks, in the water-alcohol mix¬
ture in which the powdered material had been heated after extraction
with ether, and in the ether extract. This last was now heated on a
water bath until nearly all the ether had been driven off, when it was
taken up with hot 70 per cent alcohol. All the solutions were then com¬
bined, whereupon some precipitation occurred. The precipitate was
brought again into solution by warming the flask and adding sufficient
boiling water to reduce the alcohol concentration to 70 per cent. When
a perfect solution had been secured, the solution was transferred to a
volumetric flask, made up to 2,000 c. c. with 70 per cent alcohol, and
200 c. c. were removed for the determination of the total solids. The
remainder of the solution was transferred to beakers, placed upon a
water bath kept at 750 C., and evaporated down to a sirupy consistency,
a little absolute alcohol being added from time to time. The material
was finally allowed to become alcohol free and was then taken up with
sufficient distilled water, in most cases about 700 c. c., to form a perfect
emulsion, and transferred to a 1,000 c. c. volumetric flask. Twenty
c. c. of chloroform were added, the flask was vigorously shaken for sev¬
eral minutes, and the shaking was continued while 10 c. c. of concen¬
trated hydrochloric acid were very gradually added. The flask was
then filled to the mark with distilled water, stoppered, shaken at short
intervals for 2 hours, and finally submerged to the neck in cold running
water for 48 hours. At the end of this time the solution was clear, the
lipoids having been partially carried down by the chloroform layer,
while a part adhered to the walls and neck of the flask.

The solution was now filtered through a dry filter, great care being
taken to prevent the transfer of the chloroform layer or the precipitated
material upon the flask walls to the filter paper. After the filter had
been allowed to drain well, the volume of the filtrate was noted. It
constitutes the water-soluble portion of the alcohol-ether-water extract,
and is designated as fraction 2.

The material precipitated by the chloroform, together with that held
in the chloroform layer, constituted the lipoid precipitate, fraction 1.
It was next collected by thoroughly washing the filter paper and the
neck and walls of the precipitation flask with a large volume of boiling
95 per cent alcohol from a wash bottle, 500 to 600 c. c. being necessary
with most samples. When a uniform solution had been secured, the
flask was transferred to the water bath and kept at 70° C. until the
chloroform had been completely driven off. The flask was then filled
to the mark with warm 95 per cent alcohol, and aliquot parts of the

Journal of Agricultural Research

Vol. VII, No. i

solution were at once taken for determinations of dry weight, phosphorus,
nitrogen, carbohydrates as reducing sugars before and after hydrolysis,
and ash.

The following determinations were then made upon fractions i and 2 :
Dry weight, ash, total phosphorus, total nitrogen, ammonia (in fraction

2) , carbohydrates (as reducing sugar before and as invert sugar after
acid hydrolysis in fraction 2 ; as hydrolyzable, polysaccharids in fraction

3) , and total acids (in fraction 2). Details of the determinations follow.

Dry weight. — The dry weight was ascertained in the case of fractions

1 and 2 by taking an aliquot part of the solution, evaporating to con¬
stant weight in a previously weighed platinum or porcelain dish at 102°
to 105° C., weighing, and calculating the total weight from the results.
Fraction 3 was weighed entire after drying to constant weight in the
tared diffusion thimbles in which the extraction had been carried out.

Ash. — Ash was determined upon the portions of fractions 1 and 2
which had been taken for dry-weight determinations, and upon an
aliquot' of fraction 3. The procedure was the same in all cases. A
preliminary test determined the presence or absence of free acids; if
such acids were present, the sample was neutralized with sodium hydroxid
and a correction for the sodium chlorid thus formed was made in the
final weighings. The samples for ashing were transferred to previously
dried, weighed porcelain crucibles, dried to constant weight in the oven,
weighed, then placed within larger nickel crucibles. The contents were
very gradually charred over a low flame, and finally burned to constant
weight, averages of duplicate or triplicate samples being taken as true
readings.

Total, phosphorus. — Determinations were made by the Neumann-
Pemberton method as described by Plimmer (16, p. 543) and by
Mathews (13, p. 893-895). The method consists essentially in the
conversion of organic to inorganic phosphates by the addition of con¬
centrated sulphuric acid and nitric acid, conversion of the phosphates
into ammonium phosphomolybdate, dissolving the phosphomolybdate
in a known excess of sodium hydroxid, and titrating the reduction in
alkalinity of the sodium hydroxid with sulphuric acid. Determinations
of inorganic phosphorus were attempted, employing the usual methods,
but the amounts were so small — in no case amounting to 4 mgm. — that
separate determinations were abandoned.

Total, nitrogen. — Nitrogen was determined by the employment of
the Gunning- Arnold modification of the Kjeldahl method. By reason of
the very small amount of nitrogen present in fraction 2, the original
intention to separate the nitrogenous constituents of the fraction into
proteoses, peptones, polypeptids, amino acids, and nitrogen bases by
separate determinations was abandoned. Accuracy in such determina¬
tions would have necessitated the extraction of very large amounts of
material, which was impossible in the limited time available for the work.

Oct. 2, 1916 Effect of Blackrot Fungus on the Apple 23

Ammonia determinations were made in all cases upon fraction 2 by
aerating the alkaline solution, passing the escaping air through an ab¬
sorption tube containing Nfio sulphuric acid, titrating the acid, and
estimating the ammonia from the decrease in acid strength. In the case
of both normal and diseased fruits, fraction 1 gave by the Kjeldahl
method only slight traces of nitrogen, and the figures are consequently
omitted from the tables.

Carbohydrates. — In the case of fraction 2, the portion taken for
sugar estimation was freed from organic acids and tannins by adding
normal lead acetate in excess, diluting, filtering, removing the excess
of lead by the addition of saturated sodium-sulphate solution, and again
filtering (26, p. 43). An aliquot portion of the neutral solution was then
taken for determination of reducing sugar by the Bertrand volumetric
method (16, p. 228-229) and the amount of sugar calculated as dextrose
by the use of the Munson and Walker tables (26). Total sugars were
determined by the same methods upon a second portion of the solution,
after inversion by heating for 10 minutes in a water bath kept at 70°
with 3.5 per cent of hydrochloric acid and subsequent neutralization of the
cooled solution. Total sugars were estimated as invert sugar, and the
disaccharids were determined by a difference of readings before and after
inversion.

The polysaccharids in fraction 3 were estimated as dextrose, by five
hours hydrolysis with 2.5 per cent of hydrochloric acid under a reflux
condenser, neutralizing the cooled solution, clearing of nonsugars with
lead acetate, filtering, and employing the Bertrand method. No attempt
was made to secure a quantitative separation of the mixture of sugars
resulting from the hydrolysis into its various constituents. It is probable
that the figures are low and that longer hydrolysis would have slightly
increased the yield. The figures given, however, afford a safe basis for
such general comparisons of the chemical composition of normal and
diseased fruits as it is the purpose of the writers to make.

Very considerable difficulties were encountered in attempting to esti¬
mate the sugars in the lipoid precipitate, fraction 3. In the analytical
scheme originally developed by Koch and his co workers (10), sugars in
this fraction are determined by evaporating a portion of the solution to
an alcohol-free paste, taking up with water, hydrolyzing for 24 hours with
4 per cent of hydrochloric acid under a reflux condenser, concentrating
the solution, drawing off the aqueous layer through a separatory funnel,
clearing the solution with anhydrous sodium sulphate, and making a
determination of sugar by the Bertrand method. The lipoid precipitate
obtained from plant tissues contains true fats, waxes, lecithins, chloro¬
phyll and derived or associated pigments, tannins and their derivatives,
and resins. In the course of a 24-hour hydrolysis very considerable
changes occur in some of these constituents, with the production of com¬
pounds which reduce Fehling’s solution. Glycerol, which is absent

Journal of Agricultural Research

Vol. VII, No. i

before the hydrolysis, is present, apparently as a result of the hydrolysis
of some lipoid substance in the presence of hydrochloric acid as a catalyst.
The glycerol can not be quantitatively removed without the removal of
some sugar. Precipitation with normal lead acetate, with ferric salts, or
with other ordinarily used precipitants does not completely free the
hydrolysate from nonsugar constituents which combine with the copper
of Fehling’s solution. In consequence, the determinations for sugar in
this fraction are uniformly high. While devoid of absolute quantitative
value, they were made under identical conditions and, hence, have a
certain comparative value.

RESULTS OF THE ANALYSES
COMPOSITION OF THE FRUITS USED

Prior to the discussion of the analytical results, it is pertinent to
present data upon the composition of the apples used, which differ
somewhat widely from those given by other investigators. The only
analyses of Red Astrachan apples found in the literature are those made
by Browne (i, 2) and Jones and Colver (7). Browne analyzed fruit
grown at State College, Pa., while Jones and Colver analyzed two sam¬
ples, one grown under irrigation and the other without irrigation, from
unknown localities in Idaho. The data presented in Table I show that
the apples employed in this work were somewhat lower in solids and
reducing sugar, while higher in acid, in disaccharids expressed as cane
sugar, and in insoluble residue than were those analyzed by the authors
named. These differences are due in part to differences in the degree
of maturity of the fruits used, in part to the very different climatic con¬
ditions under which the samples were grown, and in part to the use of
varying methods of analysis. The apples used were taken prior to the
beginning of the rapid digestion of carbohydrates attendant upon ripen¬
ing, as attested by the presence of small quantities of starch, and the
differences between normal and decayed fruits which appear in the
results are those consequent upon the activities of the fungus.

MOISTURE LOSSES CONSEQUENT UPON ATTACK BY THE FUNGUS

The moisture content of the normal and the diseased fruits taken for
analysis was determined by drying weighed portions of the finely ground
and uniformly mixed masses of pulp, from which samples for analysis
were subsequently taken, in an oven at 105° C. until constant weight
was obtained. The writers are aware that the use of this method in¬
volves some error as a result of the decomposition of the sugars of the
fruit, particularly of the levulose, as has been shown by Browne (2), but
the fact that no vacuum drying apparatus was available rendered its
use imperative.

In order to gain some idea of the magnitude of the constant and in¬
creasingly rapid loss of weight which accompanies decay, a number of

Oct. 2, 1916

Effect of Blackrot Fungus on the Apple

fruits showing just perceptible initial areas of tissue attacked by Sphaerop -
sis malorum were removed from the tree, weighed, the stems carefully
sealed with paraffin, the increase in weight due to this treatment noted,
and the fruits kept under observation. They were again weighed when
decay had involved approximately one-half the tissues of each fruit,
and again as soon as decay had become complete. The weight of the
normal fruits was 566 gm.; when half-decayed, this had been reduced to
559 gm., or 98.76 per cent the original weight; when wholly decayed
the weight was 539.6 gm., or 95.39 per cent of the original weight. In¬
asmuch as the moisture contents of the normal, half-decayed, and wholly
decayed tissues taken for analysis were 87.39 Per cent, 86.62 per cent,
and 86.14 per cent, respectively, of the total weights, it is obvious that
the loss in weight is due principally to other causes than the escape of
water, partial proof of which is given by the fact that in the apples from
which these weights were obtained the stems had been sealed, while
the peel of every fruit remained intact.

Table I. — Composition of normal mature fruit of Red Astrachan apple

Analyst.

Source of fruit.

Total

solids.

Ash.

Acidity
as malic.

Crude

fiber.

Reduc¬

ing

sugar.

Cane

sugar.

Protein.

Browne (i) .

State College, Pa.

Percent .

Per cent.

15*30

0-37

1*038

6.67

6.98

3*53

O. TC

Jones and Colver (7)

Nonirrigated or¬

18. 10

•9457

0.288

chard, Idaho.

Do .

Irrigated orchard,

14*73

.890

6.08

.560

Idaho.

2 • 9r

Culpepper, Foster,
and Caldwell.

Auburn, Ala .

12. 94

.2548

.9288

2 . 10

*574

4*960

■245

A comparison of the transpiration of normal apples given similar
treatment with that of decaying apples would lead one into error,
since the release of energy resulting from decomposition would mani¬
festly increase transpirational water loss from the decaying fruits.
While losses of weight due to transpiration and to respiration conse¬
quently can not be separately measured, it is clear from a comparison
of the figures given that the primary and chief cause of loss in weight
must be accelerated respiration. For these reasons, the analytical data
now to be presented are in all cases based upon 100 gm. of fresh normal
tissue, and have been corrected to this basis by the division of all figures
for half-decayed fruit by 1. 01255, of all figures for completely decayed
material by 1.0492. These ratios are based upon the figures for losses
in weight during decay obtained in the experiment just described,
and the corrected figures are regarded as very close approximations to
equal amounts of normal tissue, especially since the fruits employed in
any one group of analyses were carefully selected for equal size and
degree of maturity. Table II presents in tabular form a summary of
the results of the analyses.

Journal of Agricultural Research

Vol. VII, No. i

Table II. — Composition of ioo gm. of Red Astrachan apples , normal and mature , half-
decayed as a result of attack by Sphaeropsis malorum , and wholly decayed as a result of
attack by Sphaeropsis malorum

NORMAL AND MATURE

Fraction.

Dry weight.

Phosphorus.

Nitrogen.

Ammo¬

nia.

Actual.

Percent¬
age of
total
weight.

Actual

weight.

Percent¬
age of
total
phos¬
phorus.

Actual

weight.

Percent¬
age of
total ni¬
trogen.

Gw.

o. 624
10. 216

2. IOO

4. 82
78. 94
16. 24

Gm.

0. 0101
. 0191
. 00418

30.25
57. 22
12. 53

Gm.
Trace.
0. 0210

• 0393

34- 82

6s- 17

Gm.

0.00375

Total

12. 940

• 03338

• 0603

half-decayed

0. 519
9.396
2.458

4. 26
75- 94
19. 80

0.0097
.0237
, 00217

27.40
66. 48
6. 12

Trace.

0.01659

• 0443

26. 79

73.21

To+iO

12.373

• 03557

.06089

WHOLLY DECAYED

0.758
7-857
3- 445

6. 29
65. 14

28. 57

0. 009055
.0200
.00714

25.OO
57- 89
17. 11

Trace,

0.00933

.0476

16. 39

83. 16

0. 003485

'TVvI-al

12. 060

. 03619

• 05693

NORMAL AND MATURE

Fraction.

Ash.

Carbohydrates.

Hydrolyzable

polysaccharids.

Actual

weight.

Percent¬
age of
total ash.

Lipoid

sugars.

Reduc¬

ing

sugars.

Disac-

charids.

Starch.

Other

than

starch.

Total.

Gm.

0. 0017
• i75
.0798

0. 66
68. 22

31. 11

Gm.

0. 4797

Gm.

0. 574

4. 960

0. 321

0. 315

0. 636

• 2565

1 otai .

half-decayed

0. 0014
. 2050
.0542

0. 533

78. 66
21. 81

0. 02488

* . * .

0. 1619

3-651

3 .

0. 344

0. 771

I.II5

T^tal

. 2606

WHOLLY DECAYED

Trace.
0. 277
. 0482

85. 11
14. 89

0. 1269

* .

0. 0609

2. 234

0.318

1. 086

1.404

TV»fa1

.3252

Oct, 2, I916

Effect of Blackrot Fungus on the Apple

The most immediately obvious fact observed upon inspection of Table
II is the very considerable decrease in the total dry weight of solids
recovered in the analyses of partially and totally decayed fruits. In
comparison with a total dry weight of 12.940 gm. in normal fruits, the
half-decayed material had 12.373 a loss of 0.567 gm., or 4.38 per
cent, while completely decayed material had a dry weight of 12.060 gm.,
with a loss of 0.880 gm., or 6.80 per cent. When the distribution of the
dry weight between the three fractions is considered, it is apparent that
the progress of the disease is accompanied by a very marked decrease in
the constituents soluble in alcohol, ether, or water and consequently
recovered in fraction 2. These make up 78.94 per cent of the dry weight
of the normal fruit, 75.94 per cent of the weight of the half-decayed
fruit, and 65.14 per cent of the weight of the wholly decayed material, a
total decrease of 13.85 per cent. This reduction in fraction 2 is accom¬
panied by a decrease, followed by an increase, in the lipoid material
constituting fraction 1 , and by a steady increase in the quantity of insol¬
uble residue (fraction 3). In half-decayed fruits, lipoids make up 4.26
per cent and in wholly decayed fruits 6.29 per cent of the total dry
weight, as compared with 4.82 per cent in normal tissues. There is con¬
sequently an absolute increase of 30.7 per cent in the lipoids as a result
of the diseased condition. Concurrently, there is in the insoluble con¬
stituents making up fraction 3 an increase from 16.24 Per cent m sound
apples to 19.80 per cent in half-decayed fruit and 28.57 per cent in
wholly decayed fruits. With a total dry weight equaling 12.060 gm.
for wholly decayed fruits, the insoluble residue totals 3.445 gm., while
a total dry weight of 12.940 gm. of sound fruit has an insoluble residue
of only 2.100 gm. Stated in summary fashion, the result of the progress
of disease to complete decay of the tissues is an increase of the insoluble
residue to 176.23 per cent and of the lipoid fraction to 130.7 per cent
of the normal, with a concurrent decrease of the water-soluble portion
of the alcohol-ether- water extract to 82.29 Per cent of normal. The
significance of these differences will appear as the details of the analysis
are discussed.

The nitrogenous constituents are practically wholly contained in frac¬
tions 2 and 3, since fraction 1, both in normal and in decayed fruits, uni¬
formly gave amounts too small to be included in the tables. The total
amounts of nitrogen found are very small, and there is a slight decrease,
amounting to only 3.4 mgm. in an original total of 60.3 mgm., in the com¬
pletely decayed fruit. The significant feature of the results lies in the
fact that there is a steady decrease in the nitrogen of fraction 2, which rep¬
resents proteoses, peptones, polypeptids, amino acids, and nitrogen bases,
as the disease progresses. The amounts in this fraction are 21, 16.59,
and 9.33 mgm. for sound, half-decayed, and completely decayed fruits,
respectively. There is a corresponding, though not absolutely compen¬
satory, increase in the nitrogen of fraction 3. The figures for that fraction

Journal of Agricultural Research

Vol. VII, No. i

are 39.3, 44.3, and 47.6 mgm., and since these amounts may fairly be con¬
sidered as the nitrogen of proteins, the employment of the factor 6.25 gives
245.6, 276.8, and 297.5 mgm. of protein for the three lots of tissue, or a
gain of 51.9 mgm. of protein in the course of the process of decay. This
unquestionably represents the synthesis of protein by the attacking
fungus. It may be further pointed out that the gain in nitrogen in frac¬
tion 3 is not sufficient to account for the decrease in fraction 2 ; only about
two-thirds of the 11.67 mgm. lost from fraction 2 by completely decayed
tissues is represented by the gain of 8.3 mgm. occurring in fraction 3, and
the total nitrogen figures are 60.3 mgm. for sound and 56.93 mgm. for
completely decayed fruits, indicating a slight loss of nitrogen. Determi¬
nations of ammonia made upon aliquot parts of fraction 2 immediately
upon completion of the separation showed a slight decrease in the amount
present as the disease progressed; for sound tissue the figures were 3.757
mgm. of ammonia; for partially decayed tissue, 3.620 mgm.; for entirely
decayed tissue, 3.485 mgm. Although these differences are small and
subject to a relatively large experimental error, as must always be the
case when extremely low concentrations of ammonia are determined, they
nevertheless are concordant and compel the conclusion that the peptone
and amino nitrogen of the fruit are steadily decreased with the progress
of the disease by two processes, one a degradation process which results
in the formation and escape of ammonia, the other an anabolic process
which converts the soluble nitrogenous cons titu tents into insoluble forms,
presumably within the cells of the invading organism. Reed and Stahl
(2 1 ) reported the presence of erepsin in cultures of Sphaeropsis malorum ,
as evidenced by the formation of tryptophane from the peptone of Dun¬
ham's solution, but no further studies of the proteases of the genus
Sphaeropsis appear to have been made, although Reed (18) has recorded
the presence of an amidase able to form ammonia from alanin and aspar-
agin in Glomerella rufomaculans .

Many of the statements made in the preceding discussion are rendered
conclusive by inspection of the data for phosphorus. In the normal
fruit the distribution of phosphorus between the three fractions is as
follows: Lipoid fraction 30.25 per cent, water-soluble portion of alcohol-
ether-water extract 57.22 per cent, and the insoluble portion 12.53 Per
cent. In the half-decayed fruits there is a very marked decrease in the
insoluble fraction from 12.53 to 6.12 per cent; a less marked decrease from
30.25 per cent to 27.40 per cent in the lipoid fraction, and a concurrent
gain in water-soluble phosphorus from 57.22 to 66.48 per cent of the
total. These figures show that in the earlier stages of the attack the
changes which involve phosphorus are predominantly katabolic in
nature, and that they affect both lipoid and protein phosphorus, reducing
both to simpler water-soluble forms. With the further progress of the
disease, there is a further reduction of lipoid phosphorus to 25 per cent,

Oct. 2, 1916

Effect of Blackrot Fungus on the Apple

a decrease of water-soluble phosphorus to 57.89 per cent, which is
practically the proportion found in normal tissues, and an increase in
fraction 3 to 17.11 per cent, as compared with 12.53 per cent in sound
fruits. Here constructive processes are predominant; a portion of the
lipoid phosphorus, together with that derived from destruction of the
host proteins, has been utilized in the construction of new and complex
material, presumably protein in nature, by the blackrot organism. We
therefore possess, in the analytical data fox nitrogen and phosphorus, an
index to the character and amount of the changes in nitrogenous con¬
stituents of the fruit brought about during decay. While no attempt
to do so has been made in this case, it would appear that a quantitative
estimation of the several phosphorus-containing groups and a determina¬
tion of the relative amounts of proteoses, peptones, and amino acids
present at various stages of the progress of the disease would contribute
much to our knowledge not only of the changes wrought in the host
proteins by the attacking fungus, but also of the extent to which these
materials are wholly destroyed or utilized in constructive processes
by the parasite. Such determinations will be attempted by the senior
author, who contemplates continuing the work here reported.

The most obvious changes in carbohydrate contefit are those occurring
in fraction 2, which contains reducing sugars, disaccharids, trisaccharids,
and glucosids, and in fraction 3, which contains the insoluble carbohy¬
drates, as starch, cellulose, and cellulose derivatives. The reducing
sugars in fraction 2 are reduced from 0.574 gm. in normal fruit to 0.1619
gm. in half-decayed fruit, and to 0.0609 gm. in wholly decayed tissue; in
other words, there is a loss of 89.4 per cent of reducing sugars with the
progress of the disease. Disaccharids, which total 4.960 gm. in normal
apples, are reduced to 3.651 gm. in half-decayed fruit and to 2.234 gm.
in wholly decayed fruits, a loss of 56.94 per cent. The data obtained
from the determinations of acidity will be discussed later. It may be
said here, however, that the acid content of the sound fruit, which equaled
0.9288 per cent, calculated as malic, was reduced to 0.3563 per cent in
the wholly decayed fruit. The steady decrease in sugar content is
accompanied by an even larger proportional decrease in acid content,
since 71.64 per cent of the acid of sound fruits had disappeared con¬
currently with a loss of 89.4 per cent of the reducing sugars and 56.94
per cent of disaccharids.

There is also a reduction in the amount of lipoid sugar in fraction
1 as decay proceeds. The first determinations of sugar in this fraction
were vitiated by the fact that sufficient precautions were not taken to
remove the last traces of chloroform. When the senior author attempted
to repeat the determinations upon portions of the lipoid fraction which
had been reserved for the purpose, he encountered further difficulties.
When a 24-hour hydrolysis with 4 per cent of hydrochloric acid is carried

Journal of Agricultural Research

Vol. VII, No. i

out upon these fractions, a great variety of products are present in the
aqueous solution. Glycerol derived from the hydrolysis of lipoid mate¬
rials was present in small amounts ; phloroglucinol and catechol or some
very closely related substance were present in sufficient quantities to give
all the usual tests (16), and the presence of other unknown com¬
pounds capable of reducing Fehling’s solution was not precluded. Pre¬
cipitation with normal lead acetate and the subsequent removal of the
excess of lead with saturated sodium-sulphate solution removed tannin
derivatives, but it is impossible to remove glycerol by any method of
treatment which does not also cause some loss of sugars. The figures
obtained therefore have a comparative rather than an absolute value;
they indicate a reduction from 49.97 mgm., the quantity found in normal
fruit, to 12.69 mgm. in totally decayed apples, a reduction of 74.61 per
cent.

The results of the determination of polysaccharids were distinctly sur¬
prising and effectually destroyed any preconceived ideas as to the utiliza¬
tion of these compounds in the metabolism of the fungus. The starch
content of the normal apples was 0.321 gm., that of half-decayed fruits
0.344 gm** alld that of wholly decayed fruits 0.318 gm. It was conse¬
quently evident that conversion of starch into simpler forms had not
occurred under the conditions of the experiments. Since Reed (18)
demonstrated the presence of a diastase capable of acting vigorously
upon both com and arrowroot starch in enzym powders prepared by
alcoholic extraction of apples completely rotted by Sphaeropsis malo -
rum , while Thatcher (25) found no diastase in ripe sound apples, this
point was subjected to further investigation. Microscopic examination
of starch grains from completely decayed apples showed no discoverable
erosion of the starch grains. Since this result might be due to inhibition
of the secretion of diastase by the fungus in the presence of sugars, as
Katz (8) found to be the case with several fungi, the senior author
tested the diastatic activity of fresh extracts of pure cultures of Sphae¬
ropsis malorum grown upon a variety of media poor or lacking in sugar.
In no case was there the slightest activity either upon com or apple
starch. It would therefore appear that the amylase found by Reed was
derived from the apple and not from the attacking fungus, and that the
disagreement between the results of Reed and of Thatcher can be ex¬
plained only by the assumption that they were dealing with fruits of
differing degrees of maturity. Hawkins (5) found that Fusarium oxy-
sporum9 F. radicicola , and F. coeruleum were unable to attack the
starch of potato tubers, showing no action upon it even after one week,
although extracts of these fungi rapidly digested soluble starch. The
same investigator has shown (4) that in the case of the peach, rotting
induced by Sclerotinia cinerea (Bon) Schroter produces only very slight
decrease in the amount of alcohol-insoluble material capable of reducing

Oct. 2, 1916

Effect of Blackrot Fungus on the Apple

Fehling's solution after hydrolysis with dilute hydrochloric acid. About
one-third of this material appears to be starch, as determined by diges¬
tion with malt diastase.

There was a large and consistent increase in the yield of reducing sugar
given by the insoluble residue after five hours' hydrolysis with 2.5 per
cent of hydrochloric acid under a reflux condenser. In the normal tissue,
the amount of hydrolyzable carbohydrate material, after deducting the
amount of starch found by separate determinations, was 0.315 gm. In
half-decayed tissue, the quantity had risen to 0.771 gm., and in wholly
decayed tissue there was a further increase to 1.086 gm., or a total gain
of 0.771 gm. That this hydrolyzable material is not derived from partial
decomposition of cellulose or other structural materials is conclusively
shown by the fact that after deducting from the total weight of the
insoluble residue of normal fruits, 2.100 gm., the weight of the hydro¬
lyzable portion, 0.636 gm. plus the weight of the protein therein con¬
tained as indicated by the nitrogen content, 0.2456 gm., we have remain¬
ing 1.2184 gm. as the weight of the nonnitrogenous, nonhydrolyzable
residue. In the half-decayed fruit, after making similar deductions,
there remain 1.0662 gm. as the weight of this residue, while in the wholly
decayed fruit this residue amounts to 1.7435 £m- These results were
seriously questioned when first obtained; but repeated determinations
yielded results which were entirely concordant, leaving no question that
there occurs in the course of the disease a very marked increase in both
the hydrolyzable and the nonhydrolyzable residues of fraction 3. In
the early stages of the disease the nonhydrolyzable portion decreases
very considerably, but subsequently increases to 143.08 per cent of its
original amount. The increase in hydrolyzable constituents aside from
starch is 244.76 per cent. These very large increases can be explained
only as being the results of constructive processes carried on by the
fungus, in which connection the increase of nitrogen and phosphorus in
this fraction is significant.1 In the face of our almost total absence of
knowledge as to the amount of the carbohydrate synthesis occurring in
any parasitic fungus and particularly as to the nature of the materials
loosely designated as “glycogen,” speculation as to the meaning of these
results would be fruitless, but they indicate an interesting field of inves¬
tigation. It is hoped that the study may be continued to the end that
some information in regard to the nature of the compounds formed in
these synthetic processes may be gained.

The variations in total ash content shown by the analyses are ac¬
counted for by the fact that determinations are made upon small aliquot

1 Hawkins (5) found that there was rather vigorous, construction of both hydrolyzable and nonhydrolyzable
material by Fusarium oxysporum growing on potatoes; the crude fiber content of rotted quarters of tubers
was uniformly considerably greater than that of sound quarters used as checks, as was the content of
material reducing Fehling’s solution after acid hydrolysis. The relative amounts of such materials pro¬
duced by F . oxysporum are very much smaller than those found by us for Spkaeropsis malorum.

Journal of Agricultural Research

Vol. VII, No. i

parts of each fraction and not upon large single samples. The presence
of waxy and resinous bodies in the lipoid fraction also contributes to the
difficulty met in securing accordant results. In normal fruits 68.22 pel
cent of the total ash is present in fractions 2, with 31.11 per cent in frac¬
tion 3. In half-decayed material these percentages become 78.66 and
21.81, respectively, while when decay becomes complete 85.11 per cent is
present in fraction 2 and only 14.89 per cent in fraction 3. There is
clearly a steady transfer of mineral elements from insoluble combina¬
tions with constituents of fraction 3 into less complex, readily soluble
forms as the disease proceeds. At the same time there is a reduction
in the originally very small quantity of ash in the lipoid fraction practi¬
cally to zero, the amounts found in this fraction for wholly decayed
fruits being uniformly too small to weigh.

As originally planned, the present study also contemplated the deter¬
mination of the amounts of tannins and tannin derivatives present in
normal and diseased fruits, to the end that some information as to the
effect of the growth of the fungus upon these compounds might be
secured. A review of the literature resulted in the bringing together of
a number of methods which were tried out in a comparative way, in
part by the authors upon apples and pears, in part by Dr. F. A. Wolf in
the course of his work in the same laboratory upon other plant
material (28). The results were of such discordant character as to be
entirely valueless. It is clear that we have as yet no methods of esti¬
mating tannins which are sufficiently quantitative to be dependable
when a fruit very low in tannin content is employed and when the pur¬
pose in view is the recognition of small alterations in this content.

CHANGES PRODUCED BY SPHAEROPSIS MALORUM IN APPLES IN

ARTIFICIAL CULTURE

In order that the analytical results reported in the preceding pages
might be compared with results obtained when pure cultures of Sphae-
ropsis malorum were allowed to act upon mature sterile apple tissue
under the most favorable conditions for growth obtainable in the labora¬
tory, a number of experiments with such cultures were made.

On June 21 a carefully selected lot of sound, mature apples were
ground in the manner previously described, 100-gm. samples were weighed
into Erlenmeyer flasks, plugged with cotton, and sterilized in an auto¬
clave for three successive days. Half the flasks were then inoculated
from a pure culture of Sphaeropsis malorum; the remaining flasks were
kept as sterile checks. All were then incubated until August 15 at
320 C. Dry-weight determinations upon a 100-gm. portion of the original
material gave a weight of 12.600 gm. At the end of 54 days of incuba¬
tion, sterile and inoculated flasks were opened and the contents of each
separated into the three fractions. The dry weight of each fraction was
then determined. The results are given in Table III.

Oct. 2, I916

Effect of Blackrot Fungus on the Apple

Table III. — Dry weights (in grams) of fractions 1, 2, and 3 in sterile checks and in
artificial cultures of Sphaeropsis malorum after 34 days’ incubation

Item.

Fraction 1.

Fraction 2.

Fraction 3. j

Total.

Sterile check .

0. 850
. 470
- . 38o

9. 270
4. 170
- 5. 100

2. 384
3- 065
+ . 681

12. 404
7- 705
- 4.699

Inoculated flask .

Changes in fractions in the inoculated
flasks . . .

Percentage of gain or loss in inoculated
flasks .

-44. 7

“55- 02

+28- 55

-37-97

It is unnecessary to point out that the conditions in an artificial inocula¬
tion upon finely ground pulp which has been sterilized under pressure differ
fundamentally from those surrounding an infection of the whole fruit
under natural conditions. Sterilization brings about a very considerable
hydrolysis of disaccharids and of more complex carbohydrate materials, ‘
so that a much larger supply of reducing sugars is available to the fungus.
The finely ground character of the material permit% the rapid penetration
of the material by the fungus, and there is consequently an exceedingly
rapid growth. The enzyms of the fruit are destroyed, but there is
opportunity for rapid spontaneous oxidations as a result of the free
access of air through the cotton plugs, as is evidenced by a loss of 0.196
gm., or 1.55 per cent, of the dry weight of the check during the period of
incubation. The inoculated flask meanwhile lost 4.699 gm., or 37.97 per
cent, a loss almost 25 times as great as that occurring in the sterile
control.

In artificial inoculations there is a considerable increase in the amount
of alcohol-ether-water-insoluble material constituting fraction 3. In the
sample for which data are given in Table III, the dry weight of this
fraction was 3.065 gm., as compared with 2,384 gm. in the normal tissue,
a gain of 0.681 gm., or 28.55 per cent. These results afford further
evidence of the fact brought out by the detailed analyses of naturally
infected fruits, namely, that there is a very considerable conversion of
soluble to insoluble material as a result of the activities of the fungus.
This conversion takes place with greater rapidity in comparison with the
loss of soluble constituents from fraction 2 in unsterilized fruits, since in
the fruit entirely decayed as a result of natural infection a loss of 2.659
gm. from fraction 2 was accompanied by a gain of 1.345 gm. in fraction 3,
while in the artificial inoculations, a loss of 5.100 gm. from fraction 2 was
accompanied by a gain of only 0.681 gm. in fraction 3. These very con¬
siderable differences indicate that there are large differences between the
metabolic activities of the fungus when grown upon sterilized, aerated
pulp and those occurring in natural infections. They emphasize the
obvious fact that very great possibilities of error exist wherever one
attempts to reason back from the results obtained with artificial cultures
and to construct therefrom a picture of the life processes of a pathogenic
organism pursuing its life cycle under natural conditions.

55855°— 16 - 3

Journal of Agricultural Research

VoJ. VII, No. i

Comparisons of the carbohydrate content of artificial cultures with
sterile checks were made upon material collected June 27, when the fruits
had become fully ripe and somewhat soft. Four identical 100-gm.
samples were prepared from one lot of pulp, sterilized for three suc¬
cessive days, and two were then inoculated while two were kept as
sterile checks. All were incubated for 42 days. One each of the inocu¬
lated and the sterile flasks were then opened and the contents extracted
for 12 hours with 95 per cent alcohol, the other pair being extracted
for the same period with water. After the removal of the noncarbo¬
hydrate material from the extracts, determinations of the reducing and
nonreducing sugars were made by the methods previously described.
The results are summarized in Table IV.

Table IV. — Carbohydrate content {in grams) of water and alcohol extracts of sterile
checks and cultures of Sphaeropsis mahrum upon apple pulp after 42 days’ incubationt
100 gm . of finely ground pulp in each flask

Extract.

Reducing

sugars.

Nonredudng

sugars.

Total sugar
after

hydrolysis.

Water extract:

Sterile check .

1. 736

0. 328

2. 064

Inoculated .

.960

• *336

I. 0936

Alcohol extract:

Sterile check . .

7.496

. 1427

7. 638

Inoculated .

2. 128

.068

2. I96

Alcohol extraction recovered from the sterile check a total of 60.62
per cent of the original dry weight as sugars, and from the inoculated
material only 17.42 per cent, a difference of 43.20 per cent, or 5.442 gm.
in an original dry weight of 12.600 gm. That monosaccharids are rapidly
attacked by the fungus while disaccharids are much more slowly re¬
duced is indicated by the data of Table VI as well as by that presented
in Table II.

Since Sphaeropsis malorum obtains the energy necessary for its growth
mainly from the oxidation of carbohydrates, an attempt was made to
secure information as to the nature of the products resulting from carbo¬
hydrate decomposition. Determinations of the acid and alcohol content
of fruits in various stages of decay from blackrot, with check determina¬
tions upon sound fruits, were carried out.

ACIDITY OF NORMAL AND DISEASED APPLES

Sound, partially decayed, and wholly decayed apples were selected
and material was prepared and ground exactly as in the preparation of
other samples. Duplicate 40-gm. portions of each lot were weighed
off, each portion was placed in a 500 c. c. flask, and 200 c. c. of distilled
water added thereto. One set of flasks was then placed on a water bath
and kept at ioo° C. for four hours, the duplicate set meanwhile being
kept at room temperature. All were thoroughly shaken at short intervals.

Oct* 2, 1916

Effect of Blackrot Fungus on the Apple

At the end of four hours the contents of the flasks were poured upon
large filters, allowed to drain, and washed with distilled water until
the total volume of the filtrate and washings equalled 500 c. c. A
convenient portion of each filtrate was then titrated with iV/20 sodium
hydroxid. As the tissues of the apple contain a natural indicator which
develops a marked color upon neutralization, the method of titration
employed was that described by Schley (23). The results given in
Table V state in each case the number of cubic centimeters of N/20
sodium hydroxid required to neutralize the acids in the clear filtrate
from 40 gm. of fresh pulp.

Table V. — Acidity in normal and diseased tissue of apple expressed in cubic centimeters
of N 1 20 sodium hydroxid required to neutralize the clear filtrate from 40 gm, of fresh
pulp after four hours’ extraction with water

Extraction and date.

Normal.

Half-

decayed.

Wholly

decayed.

Average for samples taken June 17:

Cold water. .

8-5

4-3

3- 4

Hot water . ; .

ix. 7

4. 6

4- 5

Average for samples taken June 22 :

Cold water .

9- 7

6. 0

3-3

Hot water .

9- 7

7- 5

4. 0

It is, of course, recognized that this method gives only comparative
results, that a considerable portion of the acid content is not extracted
by water, and that much higher figures would have been obtained had the
titration been made directly upon the pulp suspended in water. It may
be pointed out, however, that when titrations are made by the last-
mentioned method, the diffusion of acids out of the tissues continues for
many hours and at slower rates in diseased than in normal fruits, as
experiments in this laboratory have shown. Hence, the employment of
a method which combines rapidity of performance with the attainment*
of consistent approximate results. As a check upon determinations made
upon samples from the field, artificial inoculations and sterile checks
containing 40 gm. each of pulp were prepared, sterilized, and inoculated
for 33 days in all respects as previously described, and 150 c. c. of distilled
water were then added to each flask. The flasks were heated on the
water bath for 30 minutes, and duplicate portions were taken for titration.
The neutralization of the acids of the sterile check required 66 c. c. of
N/20 sodium hydroxid, for the inoculated pulp 3 c. c. were required. The
acidity of the sterile check was therefore 22 times that of the culture. In
natural infections, there is a reduction of acidity to about one- third that
found in sound fruits. The growth of Sphaeropsis malovum upon the apple,
whether under natural or artificial conditions, is therefore accompanied
by a considerable reduction in the acidity of the tissues. Reed (18), in
the course of his studies of the enzymic activities of Glomerella rufomacu-
lansy observed that this fungus produced an alkaline reaction in originally

Journal of Agricultural Research

Vol. VII, No. i

acid synthetic media, and also that the acidity of juice from infected
apples was materially decreased. Further studies of this point by Reed
and Grissom (20) brought out the fact that this production of alkalinity
in culture media is due, at least in part, to the production of carbon dioxid
and the resulting formation of carbonates, in part to the formation of
ammonia from the peptone of the media, and in part to the formation of
ammonia and purin and hexone bases as a consequence of the autolysis
of the proteins of the fungus (19). That none of these causes is active
in the reduction of acidity produced by Sphaeropsis malorum is evident
from the fact that the nitrogen of the alcohol-ether-water-soluble frac¬
tion, as well as the ammonia therein present, steadily decreases as the
disease proceeds, a fact quite precluding the possibility of the formation
of hexone or purine bases. The results point rather conclusively to the
destruction of organic acids by oxidation as a cause of the observed facts.
That this is the true explanation is borne out by the fact that Wolf, in the
course of studies conducted in this laboratory (28), has observed that a
species of Phoma is capable of decreasing the acidity of several species
of Citrus without a concurrent increase in the soluble nitrogenous con¬
stituents. Cooley (3) and Hawkins (4), in their studies of the chemical
changes produced in the peach by Sclerotinia cinerea , found that the devel¬
opment of the fungus was accompanied by a marked rise in the acidity of
the tissues, and Cooley showed that this was due to the production of
oxalic acid, which was absent from the plum and peach juices used as
culture media. That we are here dealing with activities of an entirely
different character is evident; there is absence of acid formation with
progressive decomposition of the acids naturally present in the fruit.

ALCOHOL DETERMINATIONS IN SOUND AND DISEASED APPLES

Samples for alcohol determination were prepared from sound, half-
decayed, and wholly decayed fruits by grinding, weighing 200-gm. sam¬
ples into distillation flasks, and adding 500 c. c. of distilled water to each
sample. The flasks were then attached to condensers and distillation
continued until alcohol had ceased to come over and a total of 100 c. c.
of distillate had been obtained from each flask. Determinations of
alcohol in these distillates were made according to the method originally
devised by Nicloux (14), as described by Pringsheim (17), which is based
upon the oxidation of the alcohol by potassium bichromate. Five c. c.
of each distillate was placed in a beaker, a small measured quantity of
N/20 potassium bichromate added, 35 c. c. of concentrated sulphuric
acid poured in, and the solution heated. The solution was then con¬
tinuously stirred while N/20 potassium bichromate was cautiously added
from a burette until the completion of the oxidation was indicated by
a change in the color of the solution from greenish blue to yellowish
green. The comparative results are as follows : For normal apples, 5 c. c.
distillate required 19.4 c. c. of potassium bichromate; for half-decayed
fruits, 5 c. c. required 22.3 c. c. of potassium bichromate; for completely
decayed fruits, 5 c. c. required 30.6 c. c. of potassium bichromate.

oct. 2, 1916 Effect of Blacky ot Fungus on the Apple 37

A further comparison of the alcohol content of sterile tissues and the
artificially inoculated material was made. Forty-gm. samples of pulp
prepared and sterilized as usual and incubated for 30 days after inocula¬
tion of half the number of flasks were employed. Two hundred c. c. of
water were added to each flask and distillation continued until alcohol
ceased to come over and 50 c. c. of distillate had been collected, when 5
c. c. samples were titrated. These distillates are two-fifths as strong as
those obtained from the fresh material, when amounts of pulp and dis¬
tillate are compared. When corrections are made for this difference, the
results for the check and the inoculated material are for 5 c. c. of distil¬
late, 7.5 and 10.5 c. c., respectively.

The experiments agree in showing that there is a considerable increase
in the alcohol content of tissues invaded by the blackrot fungus, whether
under natural or artificial conditions; moreover, they show that this
increase proceeds at an equal pace with the disappearance of soluble
carbohydrates from the affected fruits.

SUMMARY

A quantitative analytical comparison of the chemical composition of
normal mature Red Astrachan apples with that of similar fruits in two
stages of decay as a result of attack by Sphaeropsis malorum reveals cer¬
tain well-marked alterations in composition which proceed at an equal
pace with the progress of the disease.

(1) The loss of water from the affected tissues is small, amounting to
4.61 per cent of the original weight in the case of fruits which had just
reached the stage of complete decay.

(2) There is a very considerable reduction in the total solids present,
amounting to 6.80 per cent in totally decayed fruits.

(3) There is a very marked reduction, concurrently with the progress
of the disease, in the amount of the constituents removed by successive
extractions of the pulp with alcohol, ether, water, and alcohol. In sound
fruit these make up 78.94 per cent of the total solids, in wholly decayed
fruits 65.14 per cent. This reduction goes on very slowly in the earlier
stages of the disease but quite rapidly in the later stages.

(4) There is a decrease, followed by an increase, in absolute as well
as in relative amount of lipoid constituents extracted by alcohol or ether,
and precipitated from watery emulsion by chloroform, with the progress
of the disease. In half-decayed material these constituents are reduced
to 83.17 per cent of the absolute amount found in sound fruits; in wholly
decayed material they are increased to 121.48 per cent of the total content
of the sound apples. In the onset of the disease there is active attack and
transformation of the lipoid constituents of the host; later there is rapid
construction of lipoid materials by the parasite.

(5) The nitrogen extracted by alcohol, water, and ether, which rep¬
resents the nitrogen of proteoses, peptones, polypeptids, amino acids,
and nitrogen bases, steadily decreases with the progress of the disease, as

Journal of Agricultural Research

Vol. VII, No. i

does the ammonia extracted by these solvents. There is consequently
no reduction of acidity through the formation of purin and hexone bases,
as was found to be the case in Glomerella rufomaculans .

(6) The nitrogen of the alcohol-ether- water-insoluble fraction, repre¬
senting protein nitrogen, increases steadily with the progress of the
disease. This increase runs nearly parallel with the decrease in non¬
protein nitrogen in the soluble fraction, indicating that the parasite
utilizes these forms of nitrogen in the synthesis of proteins.

(7) There is a steady but small decrease in the total nitrogen present,
due to complete decomposition of some of the nitrogenous constituents
with the escape of ammonia.

(8) The phosphorus of both lipoid and insoluble fractions is materially
decreased in the half-decayed fruits, with a concurrent increase in the
amount of phosphorus extracted by alcohol, water, and ether. In com*
pletely decayed fruit there is a further reduction in lipoid phosphorus, a
marked increase in soluble phosphorus, and a very large increase in
insoluble or protein phosphorus. The organism is able to reduce both
the lipoid and the protein phosphorus of the host tissues to simpler
water-soluble forms and to utilize their phosphorus in the construction
of new and complex phosphorus-containing materials.

(9) There is a steady transfer of mineral elements from the insoluble
to the soluble fraction, the percentage of the total ash present in fraction
2 increasing from 68.22 to 85.11 per cent in the course of decay.

(10) There is a rapid decrease in the content of reducing sugars,
disaccharids, and lipoid sugars as the disease proceeds. Of these classes
of carbohydrates, disaccharids are least completely utilized, totally
decayed fruits having 45.04 per cent of the disaccharid content, but only
10.6 per cent of the monosaccharid content, of normal fruits.

(11) Starch is not attacked by the fungus, its amount remaining
unchanged throughout the progress of decay.

(12) There is a rapid increase in the hydrolyzable carbohydrate material
other than starch, present in fraction 3. In the early stages of decay
there is a slight decrease in the nonhydrolyzable portion of this fraction,
which is followed by a large increase. Invasion by species of Sphaeropsis
is characterized by a large increase in materials convertible into reducing
sugars by hydrolysis for five hours with 2.5 per cent hydrochloric acid and
by a less rapid increase in substances not affected by such hydrolysis.
The natures of the compounds thus synthesized has not been determined.

(13) There is a progressive decrease in the acid content of the fruits
from 0.9288 per cent for sound apples to 0.3086 per cent in those which
become completely decayed. In artificial inoculations upon sterile apple
pulp the reduction of the acid content is more rapid and more complete.

(14) The progress of the disease, both in natural infections and in
artificial inoculations upon sterile apple pulp, is accompanied by a large
increase in the alcohol content of the tissues.

Oct. 9, Z916

Effect of Blackrot Fungus on the Apple

LITERATURE CITED

(1) Browne, C. A., Jr.

1900. A chemical study of the apple and its products. Penn. Dept. Agr. Bui.
58, 46 p. Literature cited, p. 46.

(2) -

1901. The chemical analysis of the apple and some of its products. In Penn.
Agr. Exp. Sta. Ann. Rpt. 1899/1900, p. 262-276.

(3) Cooley, J. S.

1914. A study of the physiological relationsof Sclerotinia cinerea (Bon.) Schrdter.
In Ann. Mo. Bot. Gard., v. 1, no. 3, p. 291-326. Literature cited, p. 324-326.

(4) Hawkins, L. A.

1915. Some effect of the brown-rot fungus upon the composition of the peach.
In Amer. Jour. Bot., v. 2, no. 2, p. 71-81. Literature cited, p. 80-81.

(5) - —

1916. Effect of certain species of Fusarium on the composition of the potato
tuber. In Jour. Agr. Research, v. 6, no. 5, p. 183-196. Literature cited,
p. 196.

(6) Joint Rubber Insulation Committee.

1914. Tentative specifications and analytical procedure for 30% Hevea rubber
insulating compound. Preliminary report of . . . committee. In Jour.
Indus, and Engin. Chem., v. 6, no. 1, p. 75-82, fig.

(7) Jones, J. S., and Colver, C. W.

1912. The composition of irrigated and non- irrigated fruits. Idaho Agr. Exp.
Sta. Bui. 75, 53 p.

(8) Katz, J.

1898. Die regulatorische Bildung von Diastase durch Pilze. In Jahrb. Wiss.
Bot. [Prmgsheim], Bd. 31, Heft 4, p. 599-618.

(9) Koch Waldemar.

1909. Methods for the quantitative chemical analysis of animal tissues. I.

General principles. In Jour. Amer. Chem. Soc., v. 31, no. 12, p. 1329-
. J335*

(10) - and Carr, Emma P.

1909. Methods for the quantitative chemical analysis of animal tissues. III.
Estimation of the proximate constituents. In Jour. Amer. Chem. Soc.,
v. 31, no. 12, p. 1341-1355, 1 fig.

(11) - and Mann, S. A.

1909. Methods for the quantitative chemical analysis of animal tissues. II.
Collection and preservation of material. In Jour. Amer. Chem. Soc., v. 31,
no. 12, 1335-1341.

(12) - and Upson, F. W.

1909. Methods for the quantitative chemical analysis of animal tissues. IV.
Estimation of the elements, with special reference to sulphur. In Jour.
Amer. Chem. Soc., v. 31, no. 12, p. 1355-1364, 1 fig.

(13) Mathews, A. P.

1915. Physiological chemistry . . . 1040 p., 77 fig., 1 chart. New York.

(14) Nicloux, Maurice.

1906. Dosage de l'acool dans le chloroforme. In Bui Soc. Chim. Paris, s. 3,

t- 35> P- 330-335-

(15) Paddock, Wendell.

1899. The New York apple-tree canker. N. Y. Agr. Exp. Sta. Bui. 163, p.
179-206, 6 pi.

(16) Plimmer, R. H. A.

1915. Practical Organic and Bio-Chemistry. 635 p. 86 fig., 1 col. pi. London,
New York.

Journal of Agricultural Research

Vol. VII, No. i

(17) Pringsheim, Hans.

1910. Nachweis und Bestimmung der biologisch wichtigen niederen Alkohole.
In Abderhalden, Emil. Handbuch der biochemischen Arbeitsmethoden,
Bd. 2, p. 1-13, fig. 1. Berlin, Wien.

(18) Reed, h. s.

1913. The enzyme activities involved in certain fruit diseases. In Va. Agr.
Exp. Sta. Ann. Rpt. 1911/12, p. 51-77.

(19) -

1914. The formation of hexone and purine bases in the antolysis of Glomerella.
In Jour. Biol. Chem., v. 19, no. 2, p. 257-262.

(20) - and Grissom, J. T.

1915. Development of alkalinity in Glomerella cultures. In Jour. Biol. Chem. ,
v. 21, no. 1, p. 159-163.

(21) - and Stahl, H. S.

1911. The erepsins of Glomerella rufomaculans and Sphaeropsis malorum.
In Jour. Biol. Chem., v. 10, p. 109-112.

(22) Roberts, J. W.

1914. Experiments with apple leaf -spot fungi. In Jour. Agr. Research, v. 2,
no. 1, p. 57-66, pi. 7. Literature cited, p. 65.

(23) Schley, Eva O.

1913. Chemical and physical changes in geotropic stimulation and response.
In Bot. Gaz., v. 56, no. 6, p. 480-489, 6 fig. Literature cited, p. 489.

(24) Scott, W. M., and Rorer, J. B.

1908. Apple leaf-spot caused by Sphaeropsis malorum. In U. S. Dept. Agr.
Bur.Plant Indus. Bui. 12 1, p. 47-54, pi. 3-4. Bibliography, p. 54.

(25) Thatcher, R. W.

1915. Enzyms of apples and their relation to the ripening process. In Jour.
Agr. Research, v. 5, no. 3, p. 103-116. Literature cited, p. 116.

(26) Wiley, H. W., ed.

1908. Official and provisional methods of analysis, Association of Official Agri¬
cultural Chemists. As compiled by the committee on the revision of
methods. U. S. Dept. Agr. Bur. Chem. Bui. 107 (rev.), 272 p., 13 fig.

(27) Wolf, F. A.

1913. Control of apple black-rot. In Phytopathology, v. 3, no. 6, p. 288-289.

(28) -

1916. Citrus canker. In Jour. Agr. Research, v. 6, no. 2, p. 69-100, 8 fig., pi.
9-1 1. Literature cited, p. 98-99.

FORMATION OF HEMATOPORPHYRIN IN OX MUSCLE
DURING AUTOLYSIS

By Ralph Hoagland 1

Senior Biochemist , Biochemic Division , Bureau of Animal Industry
INTRODUCTION

In the course of a series of autolysis experiments with ox muscle certain
changes in the color of the tissue were noted, which, on closer study,
were found to be due to the formation of hematoporphyrin. While this
compound is a well-known derivative of hemoglobin, yet this is the first
instance, so far as the author has been able to determine by a careful
search in the literature, where the formation of hematoporphyrin in this
manner has been reported. Since the method and place of the produc¬
tion of hematoporphyrin in the human body are not at all clearly under¬
stood, the results of these observations are offered with the hope that
they may throw some light on that subject.

EXPERIMENTAL WORK

A series of experiments was conducted for the purpose of studying the
chemical changes which take place in ox muscle during autolysis. A
detailed description of these experiments has been reported in a previous
paper by Hoagland and McBryde,2 and only such details as seem pertinent
to this paper will be described here.

Both aseptic and antiseptic experiments were conducted, but the
observations now to be reported pertain to the aseptic experiments.
These were carried on under strict bacteriological control. Thirty-six
samples of muscular tissue from the hind quarter of an ox were secured
by aseptic methods and were transferred to sterile covered crystallizing
dishes. The dishes were sealed by means of adhesive tape and were
incubated at 3 70 C. for periods ranging from 7 to 220 days. The samples
weighed from 274 to 51 S gm., the average being 377 gm. Out of the 33
samples 21 showed visible evidences of bacterial contamination and were
discarded. Of the remaining 12 samples 9 were submitted to careful
bacteriological examination and proved to be sterile. The remaining 3
samples, which were incubated for 103/123, and 220 days, respectively,
showed no apparent evidences of bacterial growth, and while they were
not submitted to careful bacteriological examination, yet their appearance

1 The author desires to extend his thanks to Dr. C. N. McBryde, of the Biochemic Division, who prepared
the samples of muscle for incubation and who exercised bacteriological control over the experiments.

2 Hoagland, Ralph, and McBryde, C. N. Effect of autolysis upon muscle creatin. In Jour. Agr.
Research, v. 6, no. 14, P- 535_547- 1916. Literature cited, p. 546-547-

(41)

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.
fn

Vol. VII, No. 1
Oct. 2, 1915
A-2S

Journal of Agricultural Research

Vol. VII, No. i

as compared with that of the 9 other samples was reasonably good
evidence that they also were sterile.

Early in the course of the experiments it was noted that the exposed
surfaces of the samples had turned light brown in color, in contrast to
the red color of fresh lean beef, while the surface that rested on the bottom
of the dish had become bright pink in color. When a cross section was
cut through a sample, it was found that the brown color extended to a
depth of about one-fourth of an inch, the width of the zone increasing
somewhat with the period of incubation, while the interior of the sample
was of a uniform bright pink color. In a few cases the meat samples
rested in the dishes in such a way as to pocket some of the exuded juice
and protect it from the air. In these instances the juice had a peculiar
dark purplish red color, as compared with the muddy brown color of the
juice that had been exposed to the air. After the meat was ground for
analysis it soon turned brown in color. At first it was assumed that the
production of the pink to purplish red color was due to a simple reduction
of oxyhemoglobin to hemochromogen. A spectroscopic examination of
the 0.9 per cent sodium-chlorid extract of the sample that had been
incubated for 7 days showed that the coloring matter was not hemo¬
chromogen, and that oxyhemoglobin and hemoglobin also were absent.
These facts led to a careful study of the color of a number of samples
incubated for various periods of time, with the following results:

Sample incubated eor 7 days. — The exposed surface of the sample
was light brown in color, while the surface that rested on the bottom of
the dish was bright pink. When a cross section was cut through the
sample, the brown color was found to extend to a depth of about one-
fourth of an inch from the surface, while the interior of the sample was
uniformly bright pink in color. The 0.9 per cent sodiumfchlorid extract
of the meat had a light-pink color. On spectroscopic examination the
extract showed a fairly distinct narrow band at the left of, and extending
just over, the D line, and a wider and less distinct band midway between
the D and the E lines. On the addition of hydrazin hydrate the bands
at first became more distinct, but after a time disappeared. The fact that
the absorption bands were not readily affected by hydrazin hydrate and
that no absorption bands of either hemoglobin or hemochromogen
appeared after the addition of this reagent indicates that the color of the
solution was due neither to oxyhemoglobin nor to hematin. The spectrum
of the solution was, of course, neither that of hemoglobin nor that of hemo¬
chromogen.

Sample incubated por 103 days. — This sample of meat showed the
brown outer zone and the pink interior common to all of the incubated
samples. The meat rested in the dish in such a way as to pocket a con¬
siderable quantity of meat juice, perhaps 20 c. c., out of contact with the
air. This juice had a peculiar dark purplish red color and showed the

Oct. 2, 1916 Formation of Hematoporphyrin in Ox Muscle 43

following spectrum: A heavy narrow band, with sharply defined edges,
immediately at the left of the D line; and a second band, two to three
times as wide as the first band, but not quite so heavy, and with less
sharply defined edges, midway between the D and the E lines.. The addi¬
tion of hydrazin hydrate did not affect the color or spectrum of the
solution.

Sample incubated for 123 days. — This sample had the usual brown
outer zone and pink interior. When ground and extracted with water,
the pink inner portion yielded a light straw-colored extract tinged with
pink. The following spectrum was observed: A fairly heavy band im¬
mediately at the left of the D line; and a lighter and wider band between
the D and the E lines.

The extract was tested for bile pigments by means of Hammarsten's
test, with negative results.

Sample incubated for 220 days. — The brown outer zone had ex¬
tended to a depth of nearly an inch, leaving only a small inner portion
that was pink. The pink-colored portion when ground and extracted
with water yielded a light straw-colored extract that showed the following
spectrum: A heavy narrow band just at the left of the D line; and a wider,
heavy band midway between the D and the E lines. The addition of
hydrazin hydrate did not affect the color or the spectrum of the solution.

A summary of the observations which have been made concerning the
effect of autolysis upon the natural red color of ox muscle leads to the
conclusion that the pink or purplish red color which was developed in
the interior of the samples of muscular tissue and in the exuded juice that
had been protected from the air was due to hematoporphyrin that had
been formed by the reduction of oxyhemoglobin. This conclusion is
supported by the following evidence: (1) The spectrum of the color and
its behavior toward reducing agents correspond with those of hemato¬
porphyrin; (2) the color was formed only in the absence of oxygen, a
condition necessary for the formation of hematoporphyrin, and it was
destroyed on exposure to the air; (3) conditions under which hemato¬
porphyrin is formed in the body indicate that this compound probably
results from the action of certain intracellular enzyms upon free hemo¬
globin, and it is reasonable to expect that a similar change might occur in
the coloring matter of muscular tissue during autolysis, provided that
suitable conditions are maintained; (4) the pink to purplish red color
developed in the tissue and exuded juice during autolysis is characteristic
of hematoporphyrin. In all essential properties, the substance which im¬
parted the pink to ’purplish red color to the interior of the incubated meat
samples, and to the exuded meat juice that was protected from the air,
correspond to hematoporphyrin.

Journal of Agricultural Research

Vol. VII, No. t

SIGNIFICANCE OF THE FORMATION OF HEMATOPORPHYRIN DURING
THE AUTOLYSIS OF MUSCULAR TISSUE

It seems proper to call attention to the possible significance of the
formation of hematoporphyrin under the conditions which have been
described not only as regards the excretion of this compound by the
body under pathological conditions but also as related to the normal
transformation of hemoglobin into bile pigments.

Occurrence. — Hematoporphyrin occurs in traces as a constituent of
the normal urine of man and of the higher animals, and it has been
found in the feces. It occurs most frequently, and in largest quantities,
as a constituent of the urine under pathological conditions, particularly
in cases of poisoning with sulphonal, trional, tetronal, lead, and phos¬
phorus; in case of fevers and of gastric and intestinal hemorrhages; in
diseases of the liver; and in various cases of acute infectious diseases —
for example, tuberculosis, nephritis, pleuritis, rheumatism, and Addi¬
son’s disease — and under certain other conditions.

Formation in The body. — Comparatively little appears to be known as
to how or where hematoporphyrin is formed in the body.

Oswald 1 states that very little is known concerning the place or
method of production of hematoporphyrin in the body. He discusses
the hypotheses proposed by various investigators concerning the sub¬
ject and concludes that the evidence seems to indicate that hemato¬
porphyrin originates in the blood stream, particularly since this pigment
is often found in the urine in cases of hemoglobinuria. In concluding the
discussion on the subject the author states:

Jedenfalls ist die Frage nach dem Orte der Hamatoporphyrinbildung noch nicht
entschieden. Sie bedarf noch weiterer Bearbeitung.

Although a diligent search has been made in the literature relating to
hematoporphyrin, no satisfactory explanation, based upon experimental
evidence, has been found as to the method of formation or the seat of
production of that compound in the body.

The experiments which are reported in this paper show that the
striated muscular tissue of the ox contains enzyms which, under ana¬
erobic conditions, readily reduce oxyhemoglobin to hematoporphyrin.
These findings appear to offer a satisfactory explanation as to the method
and source of production of hematoporphyrin in the body.

As has been previously noted, hematoporphyrin may occur in the
urine in very small quantities under physiological conditions; but it
occurs most often and in largest quantities under certain pathological
conditions, notably (i) those obtaining in cases of poisoning or disease
where the liver cells are destroyed or inactivated and (2) those obtaining in
case of certain diseases or other conditions that cause an abnormal libera¬
tion of free hemoglobin into the blood stream.

1 Oswald, Adolph. Dehrbuch der chemischen Pathologie. S. 17J-176. I^eipzig, 1907.

Oct. 2, Igi6

Formation of Hematoporphyrin in Ox Muscle

The generally accepted view concerning the disposal of the free hemo¬
globin in the blood stream under normal conditions is that the hemo¬
globin is converted directly into bile pigments by the liver. Such being
the case, it would seem that when an abnormal quantity of hemoglobin is
liberated into the blood stream, or when the activity of the liver is
greatly decreased by disease or other causes, that hemoglobin rather
than hematoporphyrin would be excreted in the urine.

The results of the experiments reported in this paper, which show
that the striated muscular tissue of the ox has the property of readily
and completely reducing oxyhemoglobin to hematoporphyrin, not only
indicate a source and method of production of this compound in the
body but likewise suggest that hematoporphyrin may be an intermediate
product in the transformation of hemoglobin into bile pigments. The
presence of hematoporphyrin-producing enzyms in muscular tissue can
hardly be regarded as without physiological significance.

In this connection a statement by Matthews 1 is significant. He
asserts that it is one of the functions of the liver to pick out the hema¬
toporphyrin from the blood, where it occurs normally in small quantities,
and convert it into a harmless bile pigment.

SUMMARY

The results of the experiments reported in this paper may be sum¬
marized as follows :

(1) The striated muscular tissue of the ox contains enzyms which,
under anaerobic conditions, readily and completely reduce oxyhemoglobin
to hematoporphyrin.

(2) It appears very probable that hematoporphyrin may be a regular
intermediate product in the transformation of hemoglobin into bile
pigments.

1 Matthews, A. P. Physiological Chemistry, p. 422. New York, 1915.

ADDITIONAL COPIES

OF THIS PUBLICATION MAY BE PROCURED FROM
THE SUPERINTENDENT OF DOCUMENTS
GOVERNMENT PRINTING OFFICE
WASHINGTON, D. C
AT

10 CENTS PER COPY
Subscription Price, $3.00 Per Year
V

JOURNAL OF MCOLTOAL RESEARCH

DEPARTMENT OF AGRICULTURE

Vol. VII Washington, D. C., October 9, 1916 No. 2

COMPARISON OF THE NITRIFYING POWERS OF SOME
HUMID AND SOME ARID SOILS 1

By C. B. Ljpman, Soil Chemist , California Agricultural Experiment Station , P. S.

Burgess, Chemist , Hawaiian Sugar Planters* Experiment Stationt and M. A. Klein,

Assistant Soil Chemist , California Agricultural Experiment Station

INTRODUCTION

It will be remembered by those who are interested in the subject
under consideration that Stewart (ii)2 took occasion in 1912 to question
the correctness of the view then prevalent, due principally to the high au¬
thority of Hilgard (3, p. 68) , that nitrification proceeds with great intensity
in arid soils. It will also be remembered that in the statement above
cited and in subsequent papers (12, 13) Stewart controverted the theory
of Headden (1, 2) with respect to the cause of unusually large nitrate ac¬
cumulations in certain Colorado soils on similar grounds. The evidence
presented for the latter controversion consisted of numerous analyses
of irrigated and unirrigated soils at the Utah Agricultural Experiment
Station and of soil and soil-forming material of the Book Cliff areas in
Utah and Colorado. The analyses represented the quantity of nitrogen
in the form of nitrates or the quantities of other “alkali” salts, or both,
in the various soils and soil-forming materials at the time the sampling
was done. As a result of his studies of these data, Stewart con¬
cluded that the intense power of nitrification attributed to arid soils,
because of the large amounts of nitrates present in the field, is
nonexistent, that it may actually be feeble as compared with that
of humid soils, and that the large quantities of nitrates found as noted
can be more readily and plausibly explained on the basis of accumulation
by water movement from adjacent soil or soil-forming materials rich
in salts (including nitrates). The latter contention is supported by
numerous analyses of the large variety of soils, sandstones, and shales

1 The samples of soil from the different States were sent us by the chemists or agronomists of the several
Stations, and acknowledgment in addition to that already made is here publicly expressed. We owe and
express our sincere thanks also to Prof. C. F. Shaw, in charge of the soil-survey work of the California
Agricultural Experiment Station, for allowing us a portion of every soil sample collected iu the soil-survey
areas which have already been mapped in this State.

2 Reference is made by number to “ Literature cited,” p. 82,

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.
fs

(47)

Vol. VII, No. a
Oct. 9, 1916
Cal.— 6

Journal of Agricultural Research

Vol. VII, No. 2

involved, which indicate that an increase in the nitrate content of soils,
such as those described by Headden (i, 2) in Colorado, is always accom¬
panied by an increase in the quantity of the other alkali salts. Not only
on the basis of his own data obtained as above but also on those of
Headden himself, Stewart was able to make out a strong case in sup¬
port of his explanation anent the origin of the heretofore-mysterious
“niter-spots.”

It will be noted that both Hilgard and Stewart employed actually or
by implication an indirect method for arriving at their conclusions.
While Stewart's method was less indirect in that seasonal variations in
nitrate accumulations were studied, thus allowing something like a meas¬
urement of the soil's nitrifying power in the field, it can at best be pro¬
ductive only of results which call for further investigation. In other
words, the determination of nitrates in field soils, no matter how fre¬
quently made, as a basis for determining the power of the soil to nitrify
any nitrifiable material, is subject to more objections because of a lack
of control of conditions than is a direct, if arbitrary, method of deter¬
mining the nitrifying power of a soil in the laboratory. In an attempt,
therefore, to obtain data by a more direct method, which might serve to
reveal the truth about the nitrifying powers of humid and arid soils, the
writers have carried out a series of studies the results of which form the
principal topic of this paper.

PLAN, MATERIAL, AND METHODS OF EXPERIMENTS

In the experiments the plan was to compare under controlled and
uniform conditions in the laboratory the nitrifying powers of a large
number of representative soils from both humid and arid regions. In
seeking for a scheme to use as the basis for the selection of such soils, it
occurred to the senior author that it could be arranged by employing
one soil type at least from every State in the Union to compare with a
large number of California soils collected in connection with the soil-
survey work of the State. Fortunately, there had just been completed
a collection of soils from all the States and Territories in the Union
(at least one soil and one subsoil from each State or Territory) and the
writers were therefore supplied with what might be regarded as repre¬
sentative soil material. For the illustration of arid soils the samples
collected to represent the different types of four soil-survey areas in
California were employed. Approximately 45 humid were compared
with 150 arid soils.

The nitrifying power of each soil was determined, using the nitrogen in
the soil, in sulphate of ammonia, in dried blood, and in cottonseed meal
for every soil. The last-named was employed in 100-gm. portions in
tumblers, as in the methods now common among soil bacteriologists. The
soils containing as nearly as possible optimum amounts of water were

Oct. 9, 1916

Nitrifying Powers of Humid and Arid Soils

incubated for one month at 28° to 30° C. The usual devices were em¬
ployed for preventing an undue evaporation of moisture, the mainte¬
nance of a uniform water supply, and for mixing soil with fertilizer
materials. The soil nitrogen was, of course, employed as naturally occur¬
ring in the 100-gm. soil portions, sulphate of ammonia (in solution) was
employed at the rate of 0.2 per cent, dried blood at the rate of 1 per
cent, and cottonseed meal at the rate of 1 per cent, all based on the air-
dry weight of the soil. The phenoldisulphonic-acid method for deter¬
mining nitrates, as described by Lipman and Sharp (8), was employed
throughout the experiments, except as otherwise stated. The nitrate
content of the original soil was subtracted in all cases, and calculations
were made of the absolute amounts of nitrates produced, of the total
nitrogen present (whether soil nitrogen alone or soil nitrogen plus fer¬
tilizer nitrogen), and of the percentage of the latter which was trans¬
formed in a month’s incubation period into nitrates. For the purposes
of the last-named determination only the complete whole numbers for
the percentage concerned were computed, a plus sign being used after
every one to indicate that the exact percentage was less than 1 per cent
in excess of the number given.

Throughout these experiments the writers have been cognizant of
the weaknesses which obtain in any method yet devised to obtain results
in the laboratory with soil-bacterial activities which are directly trans¬
latable into terms of field conditions and magnitudes. For example,
the fact that uniform moisture and such exceptional air conditions as
are present in a constant-temperature incubator are not to be found in
the field has not been overlooked. Nor yet have the writers assumed
that the large amounts of fertilizer employed by them in the experiments
exercise the same effect as the much smaller quantities employed in farm
practice. There has been employed as a basis what seems to be the
reasonable hypothesis that soils and fertilizers, particularly the former,
bearing a certain relationship to one another as regards nitrifying powet
in the laboratory, should bear approximately the same relationship to one
another in the field. This should be particularly so as regards results
obtained with the soil’s own nitrogen.

Further, the writers also realize two other serious difficulties which
beset the investigator engaged on problems such as the one in hand.
Seasonal variations in the nitrifying powers of soils are of great magni¬
tude. This has been demonstrated by numerous investigators. Not the
least disquieting data on that subject are in the hands of the writers,
representing the most extensive study yet carried out on seasonal varia¬
tions in the ammonifying and nitrifying powers of soils. The other
difficulty is that the soils which are compared with arid soils were
collected in most of the States on Experiment Station lands, which
are not truly representative of average conditions obtaining among

Journal of Agricultural Research

Vol. VTI, No. a

humid soils. A minor objection which has been urged may also be
mentioned here — namely, that of a given lot of soil collected at the same
time two ioo-gm. samples may give widely different nitrifying powers.
It may be observed here that the writers have never been able to sub¬
stantiate such a claim. They have merely called attention to some of the
most important and perhaps the only important objections in the path
of validating such comparative studies and other similar ones. It will
be noted that scarcely any comment has been made regarding the objec¬
tions and difficulties in question. Occasion will be taken to examine
them critically and evaluate their importance in a more proper place.
It may be said that in these experiments the attempt has been made to
obtain relative and not absolute values, and despite the accompanying
weaknesses of the methods, it is believed that current ideas with respect
to the intensity of nitrification in the soils of humid and of arid regions
have been improved and rendered more definite.

EXPERIMENTS WITH “FOREIGN”1 SOILS

The word “foreign” is applied in this paper to soils coming from other
States than California, unless specifically qualified in some other manner.
While subsoils were available in every case and were studied in the
foreign soils and while several soil types were available from some
States, only one such type of surface soil and the results obtained there¬
with will be discussed. The results obtained with the foreign soils,
arranged as above described, follow in Table I, which gives the results
of nitrification of the soil, sulphate of ammonia, dried blood, and cotton¬
seed-meal nitrogen.

Table; I. — Comparison of the nitrification in soils from various Statesa

State.

Soil type.

Soil nitrogen
(Group I).

Soil nitrogen
and sulphate
of ammonia
(Group II).

Soil nitrogen and
dried blood
(Group III).

Soil nitrogen
and cotton¬
seed meal.
(Group IV).

Nitrate nitro¬
gen produced.

Total soil ni¬
trogen present.

Nitrogen in
soil nitrified.

Nitrate pro¬
duced.

Total nitrogen
present.

Nitrogen nitri¬
fied.

Nitrate pro¬
duced.

Total nitrogen
present.

Nitrogen nitri¬
fied.

Nitrate pro¬
duced.

Total nitrogen
present.

Nitrogen nitri¬
fied.

Per

Mg.

Mgm.

ct.

Mg.

Mgm.

ct.

Mg.

Mgm .

cent.

Mg.

Mgm.

ct.

Alabama .

Norfolk sandy

9. 20

23. 60

38+

1.50

63. 60

0. 10

155-60

Tr.

4. 10

73- 60

loam.

Alaska .

Rich peat .

225. 20

1. 10

265. 20

Tr.

81. 10

357- 20

22+

2<. IO

272. 20

Arizona .

Fine sandy loam

46. 00

86. 00

178.00

93-oo

(Alkali).

Arkansas .

Huntington clay

r3-8o

68.40

20+

5-40

108.40

7.80

200.40

3 +

13-30

115-40

11 +

loam.

Colorado .

Silt loam .

7. 5°

108. 30

I7- 5°

148. 30

11+

29* 50

12 +

IQ. CO

Tee. 40

1 2 “f"

Connecticut ....

Gloucester fine

5- 60

209. 80

7.00

249. 80

9.00

341-80

2 +

18.00

256. 80

sandy loam.

Delaware .

Sassafras loam . . .

54- 75

109.00

50+

34- 75

149. 00

23 +

66.75

241.00

27+

54- 75

156. OO

35+

Florida .

Calcareous peat

54-oo

1894. 00

3 +

32. 00

193400

1 +

52.00

2026. 00

42.00

194I.OO

(Everglades).

1 This term was borrowed from Sackett (io).
a Does not include Hawaii, Porto Rico, or California.

Oct. 9.

Nitrifying Powers of Humid and Arid Soils

■Table I. — Comparison of the nitrification in soils from various States — Continued.

Soil nitrogen
(Group I).

Soil nitrogen
and sulphate of
ammonia
(Group II).

Soil nitrogen and
dried blood
(Group III).

Soil nitrogen
and cotton¬
seed meal
(Group IV).

State.

Soil type.

Nitrate nitro¬
gen produced.

Nitrogen in

soil nitrified.

Nitrate pro¬

duced.

Total nitrogen

present.

Nitrogen nitri¬

fied.

Nitrate pro¬

duced.

Total nitrogen

tvresent.

Nitrate pro¬

duced.

Total nitrogen

present.

Nitrogen nitri¬

fied. j

Georgia .

Cedi loam .

Mg.
9- 8s
18.80

Mgm.

27.80

177.60

Per

ct.

35+

10-i-

Mg.

6.85

28.80

Mgm.

67.80

Per

ct.

10+

Mg.

2.85

Mgm.
139. 80

Per

cent.

Mg.

11.85

Mgm.

74.80

Per

ct.

15+

Guam .

217.60

13+

35* 80

309.

11+

30.80

224. 60

13+

Idaho .

Siit ioam. . .

12.05

152.40

5*25

192.40

25.25

284. 40

18. 25

199.40

Illinois .

Marshall silt loam

32.00

219.60

14+

4.00

259. 60

27.00

351*

15*00

266.60

Indiana .

Miami silt loam. .

8.42

in. 80

is* 42

151. 80

10+

33*42

243*

13+

20.42

158.80

12 +

Iowa .

Wisconsin drift. . .

17. 40

146. 80

11+

3*50

186.80

61. 50

278. 80

22-f

14. 50

193*80

Kansas .

Oswego silt loam .
Lexington lime-

24JO

177. 60

t!3+

6. 10

217.60

30.50

309.60

17- so

224. 60

Kentucky .

149.60

*95

189.60

Tr.

23-95

281. 60

n-95

196. 60

T^nisiana .

stone soil.
Orangeburg sandy

4-35

64. 20

•75

104. 20

Tr.

•55

186. 20

Tr.

21.95

hi. 20

19+

Maine .

loam.

Aroostook loam . .

*6. 05

148. 20

10+

11.25

188. 20

31-25

280. 20

11+

20. 25

195* 20

10+

Maryland .

Sassafras clay

13*98

72.60

22+

7.38

112.60

18. 78

204.60

io- 78

119.60

Massachusetts . .

loam.

Poduhk fine

7.88

151.00

12.88

191.00

17.88

283.OO

19.88

198.00

10+

Michigan .

sandy loam.
Miami sandy

17.40

112. 5

iS+

6.40

152.5

31.40

244. 50

12+

14.40

159-50

Minnesota .

loam.

Par go day loam, .

37-00

316. 20

11+

20.00

356. 20

53*oo

448. 20

11+

18.00

363. 20

"Loam .

7. 70

34- 80
89.40

4. 70

74.80
129. 40

— .30

166.80

4. 70

81. 80

Missouri . . .

Upland silt loam.

XO. 15

11+

5*75

26. 75

221.

12 +

14* 75

136. 40

10+

Montana... .

Yakima silt loam .

18. 00

218. 20

11. 50

258. 20

35* 50

350.

IO+

18. 50

265. 20

Nebraska .

Fine sandy loam .

39-30

76.80

51+

14.30

im. 80

12+

•30

20. 80

Trace.

20. 30

123.80

16+

Clay loam .

15-00

I3-S7

67.00
2x1. 20

22+

4.00

5*57

107.00
251. 20

22.00

199.

11+

17. 00

114.00

14+

New Hampshire

Boulder day .

18.57

343*

9- 97

258.20

New Jersey .

New Mexico....

Sassafras loam. . . ,

5-68

83.80

.68

123. 80

Tr.

13-68

215.80

11. 68

130. 80

Anthony fine

15.00

39*00

38+

12.00

79.00

15+

-1. 00

171.

9.00

86. 00

10+

New York .

sandy loam.
Dunkirk clay

24.85

160.80

iS+

16.85

200. 80

39* 85

392-

14+

22.85

207.80

10+

North Carolina .

loam.

Cedi sandy loam.

12. 52

40.40

30+

2. 87

80.40

19.87

172.

11+

14. 87

87.40

17+

North Dakota. .

Clay loam .

6. 20

279. 80

3 +

12.00

319. 80

3 +

42.00

411.

10+

17.00

326.80

Ohio .

Wooster silt loam

23- 70

120. 20

19+

3*70

160. 20

26* 70

252.

10+

14. 70

167. 20

OVIatinma

Loam .

6- 75
13. 20

82.40

935*oo

2. 75
47- 20

122. 40
975.00

12. 75

214. 40

11. 75

129.40

Oregon .

Willamette Val-

57.20

1067.00

50. 20

982.00

Pennsylvania. . .

ley sandy loam.
Hagerstown silt

19. 00

127. 20

14+

7.00

167. 20

26.00

259*

10+

17.00

174. 20

Rhode Island. . .

loam.

Miami silt loam..

8-55

163.60

3*25

203. 60

1 +

19.25

295*

6++

12. 25

210. 60

South Carolina .

Cedi sandy loam,.

13. 20

43.20

30+

11. 00

83. 20

12+

— . 20

175*

19.00

90. 20

21+

South Dakota . .

Clay loam .

19.60

193*00

10+

13.60

233.OO

29. 60

325*00

22.60

240.00

Tennessee .

Upland loam .

11. 06

102.00

10+

4. 26

142.00
143.4O
l68. 60

21. 86

234-

235-

14. 86

149-00

Texas .

Black adobe .

15.40

22.90

103. 40
128. 60

14+

17+

24.40

28. 90

17+

49.40

20+

19.40

150. 40

112 +

Utah .

Greenville loam . .

17+

48.90

260.60

18+

24. 90

175* 6oji4+

Vermont .

Sandy loam .

8-8o

85. 20
95*oo

10+

2. 80

125. 2C
135*00

25.80

217.

11+

14. 80

132. 20 11+

Virginia .

Hagerstown loam

15.20

16+

6. oc

27. 00

227.OO

11+

18.00

142.00

12 +

Washington. . . .

Shot clay .

6.60

41. 80

15+

2.6c

87. 8c

2 +

35*8o

173*

20+

9. 00

94.80

West Virginia. .

Meigs sandy loam

7. 84

184. 60

i* 44

224. 6c

9.84

316.6c

3 +

6. 04

231.6c

Wisconsin .

Carrington sandy

16. 50

1 74. 80

11.5c

214. 8c

28. 5c

306.80

22. 50

221. 80

10+

Wyoming .

loam.

Loam .

15*15

99. 20

r5+

24-iS

139* 20

17+

57*15

231

24+

27*15

146. 20

18+

Before discussing nitrification proper it is of importance to note the
conditions which obtain in the foreign soils with respect to total nitrogen
content. This consideration is important because it emphasizes what
has been only vaguely appreciated in the past — namely, the great dis¬
crepancy between the nitrogen content of humid and of arid soils. More¬
over, it is important because the soil’s total nitrogen content may have
a bearing on the absolute quantities of nitrates produced through the

Journal of Agricultural Research

Vol. VII, No. a

agency of nitrification. Out of 49 foreign soils, about 9 of which should
be reckoned as arid rather than humid, 30, or about 61 per cent of the
whole number, contain more than 0.1 per cent of total nitrogen (only
in very small part, including nitrates). If allowances are made for the arid
or semiarid soils among the foreign soils, about three-quarters of the soils
from the humid region are found to contain more than 0.1 per cent of
nitrogen. Only 8 soils, or about 14 per cent of the total of 49, contain
less than 0.05 per cent of nitrogen each, and 4 of these belong properly
to the arid or semiarid class. Only 2 soils contain less than 0.03 per
cent total nitrogen, but neither of these is an arid soil. These figures
are very interesting and worth remembering for comparison later with
similar statistics anent the California soils. The second column of
figures in Group I of Table I can be made to show percentages of total
nitrogen by moving the decimal point three places to the left.

With respect to the absolute quantities of nitrate produced, only 16
out of 47 soils tested produced from their own nitrogen less than 10
mgm. of nitrate nitrogen, and only 2 of them produced less than 5 mgm.
of nitrate nitrogen in the same period. No relationship whatever is
discernible between the total amount of nitrogen present in the soil
and the amount which was rendered into nitrate. About two-thirds of
the soils tested were therefore able to produce in every case more than
10 mgm. of nitrate nitrogen in a month's incubation period. Likewise,
nearly two-thirds of the soils tested rendered more than 10 per cent of
the nitrogen present in the soil into nitrate, and several more approached
the 10 per cent mark very closely. Moreover, nearly 40 per cent of the
soils tested transformed in every case more than 15 per cent of the
total nitrogen present into nitrate. If the few characteristically arid
or semiarid soils among the foreign soils are disregarded, less than 35
per cent of the soils would fall in the class last named. It is also of
great interest to note that over 12 per cent of the soils tested trans¬
formed more than 30 per cent of the nitrogen in them into nitrate
under the circumstances noted; and two of the soils, the Delaware and
Nebraska samples, transformed into nitrate more than 50 per cent
of the nitrogen which they contain. It may be purely a matter
of coincidental interest but possibly worthy of note that of the 9 soils
which transformed between 20+ per cent and 38+ per cent of their
nitrogen into nitrate, thus placing them in a class next to the two very
exceptional soils just referred to, 7 belong to the Southern or South Atlan¬
tic group of States, and the other 2 are from New Mexico and Nevada,
which more properly belong with the arid or semiarid group.

The results obtained when sulphate of ammonia is added to the soil
nitrogen and the whole incubated are found to be opposite to that taken
by the data for soil nitrogen. The addition of sulphate of ammonia to the
foreign soil has not induced, as might be expected, an increase in the
production of nitrate over that produced from the soil's nitrogen alone,

Oct. 9, 1916

Nitrifying Powers of Humid and Arid Soils

but has, on the contrary, caused a reduction in the soil's power to render
nitrogen into nitrate. This holds, of course, only in general, and several
exceptions may be found to the rule. If the absolute amounts of nitrate
produced as given in Group II of Table I are compared specifically with
those given in Group I, it will be found that all the soils tested except six —
namely, those from Indiana, Oregon, Texas, Utah, Wyoming, and Guam —
produced far less nitrate when sulphate of ammonia was added to them
than they did from their own nitrogen alone. It will be noted that of the six
soils which did produce more nitrate under the conditions noted than from
their own nitrogen only one is a soil belonging strictly to the humid
region. The effects of sulphate of ammonia on the nitrification of soil
nitrogen in humid soils is very striking and difficult to explain. The
acidity of soils appears to be inadequate to explain the situation.

As is to be expected, smaller absolute transformations of nitrogen into
nitrate in soil and ammonium sulphate than in soil nitrogen mean a
smaller percentage transformation of the total nitrogen present. Hence,
whereas in the case of the soil nitrogen alone, 11 soils out of 47,
or 23 per cent, transformed more than 20 per cent of nitrogen in
every case into nitrate, only one soil belongs in that class when the
series containing soil nitrogen plus sulphate of ammonia nitrogen is
considered. Also, whereas nearly 66 per cent of the soils in the soil-
nitrogen group transform in every case more than 10 per cent of the total
nitrogen into nitrate, only about 23 per cent of the same soils in the
ammonium-sulphate group fall in that category. These limited statis¬
tical illustrations on the differences obtaining between the experimental
series resulting in the soil-nitrogen group and the ammonium-sulphate
group are sufficiently emphatic to need no further comment at this point.

Group I of Table I is compared with Group III, which sets forth the
results of the dried-blood series with the foreign soils, only 10 soils out of
47 or 48, or about 20 per cent, will be found which produced less nitrate
when blood (1 per cent) was added to them than when only their own
nitrogen supply was allowed to nitrify. Again, attention is called to
what is probably a purely coincidental but interesting circumstance like
the one above mentioned. Just as 9 of the soils which transformed more
than 20 per cent and less than 39 per cent of their own nitrogen
into nitrate included 7 which came from States of the Southern
or South Atlantic group, so in this case of the 10 soils which produced less
nitrate from dried-blood nitrogen plus the soil nitrogen than from the
latter alone, 7 belong to the group of States which are in nearly all cases
the same. Of the three other soils, two are from the semiarid region —
namely, Nebraska and New Mexico — and only one is from the northern
portion of the humid region — namely, Illinois. It is worthy of note that
in all but the Florida and Illinois soil, of the group of 10 just considered,
the nitrogen content is below 0.08 per cent, and in most of them is below
0.05 per cent.

Journal of Agricultural Research

Vol. VII, No. a

The amounts of nitrate produced are smaller in Group III of Table I
than in Group I for reasons which are obvious and which have been
referred to previously. Nevertheless, in spite of the disadvantage which
the added dried-blood nitrogen creates in this series with respect to the
relative considerations, about 23 soils, or nearly 50 per cent of the whole
number, transform more than 10 per cent of the total soil plus dried-blood
nitrogen into nitrate, and several other soils approach that record closely.
In five cases more than 20 per cent, of the total nitrogen present is trans¬
formed into nitrate. Taken as a whole, therefore, and in spite of the
large quantities of blood used, the foreign soils must be adjudged efficient
nitrifiers of dried-blood nitrogen. This is particularly to be kept in mind
for comparison with data from the California soils.

Group IV of Table I, which gives the results obtained in the cotton¬
seed-meal series, is not strikingly different from Group III, which rep¬
resents the dried-blood series. Nevertheless some distinct points of
dissimilarity between the two require some comment. Thus, it must
be noted that in the dried-blood series, 34 out of 48 soils produced more
nitrate nitrogen than they did with the soil nitrogen alone. In the cotton-
seed-meal series only 26 out of the same total of soils accomplished that
task. On the other hand, certain soils which induced only losses of
nitrate nitrogen with dried blood, like the Georgia, Louisiana, and
South Carolina soils, gave with cottonseed meal increases of nitrate over
those produced with the soil nitrogen alone.

The percentage of total nitrogen which is transformed into nitrate in
the cottonseed-meal series with the foreign soils is not strikingly unlike
that of the dried-blood series when the soils are regarded as a whole.
Nevertheless the individual soils show marked differences in the direc¬
tion noted. Thus, for example, eight of the soils in the dried-blood
series transform nothing or less than 2 per cent of the total nitrogen
present into nitrate, whereas in the cottonseed-meal series no soil is pro¬
ductive of no nitrate, and only two fall in the 2 per cent class, or there¬
abouts. In other words, it would appear that while the dried blood is
better suited to the foreign soils if a few soils are eliminated from con¬
sideration, cottonseed meal is better suited to the average soil, provided
the influence of the amount of fertilizer used is disregarded. In general,
it would appear that dried blood is a more readily and more efficiently
nitrifiable material for the soils of the humid region than cottonseed meal.

In a general survey of the results obtained with the foreign soils, it
seems to be true beyond question that with respect to relative quantities
of nitrates produced from the different forms of nitrogen, the soil nitro¬
gen is the most efficiently nitrified of the four forms tried. Sulphate of
ammonia is the least efficiently nitrified, while dried blood and cotton¬
seed meal differ very little. Table II summarizes the situation with
respect to one degree of nitrate formation only.

Oct. 9, 1916

Nitrifying Powers of Humid and Arid Soils

Table; II. — Total nitrogen present transformed into nitrate in foreign soils

Nitrifiable material.

Soils transform¬
ing zo per cent
or more of total
nitrogen into
nitrate.

Soil nitrogen alone . .

Per cent.

Soil nitrogen plus ammonium sulphate . . . .

Soil nitrogen plus dried blood .

Soil nitrogen plus cottonseed meal . . . . .

Based on the absolute criterion of the production of 20 mgm. of nitrate
nitrogen, or more, under the circumstances noted, sulphate of ammonia
still remains the lowest in the scale with only 8 soils possessing such a
record, the soil nitrogen is next with 9 soils in that class, cottonseed-meal
nitrogen is next with 13 such soils, and dried-blood nitrogen stands best,
with 31 such soils. The corresponding figures obtained when the pro¬
duction of 15 mgm. of nitrate is taken as a criterion are as follows: Soil
nitrogen, 23 soils; ammonium-sulphate nitrogen, 11 soils; dried-blood
nitrogen, 36 soils; and cottonseed-meal nitrogen, 28 soils.

In brief, therefore, dried-blood nitrogen takes first place for the abso¬
lute amount of nitrate produced therefrom by the foreign soils, cotton¬
seed-meal nitrogen being second, soil nitrogen third, and ammonium-
sulphate nitrogen last. On the relative basis, however, the dried-blood
nitrogen goes from first to second place, and the soil nitrogen from
third to first place, with the sulphate of ammonia remaining last, and
the cottonseed meal third in order.

EXPERIMENTS WITH CALIFORNIA SOILS

In order to make the choice of California soils representative not only
of the arid region but also of parts of the State with widely varying
climatic and other conditions, soil types were chosen from two central to
northwest soil-survey areas and from two southern California areas.
These areas were the Ukiah and the Bay areas and the Riverside and
the Pasadena areas. The rainfall for the first two areas varies from 20
to 40 inches or more a year, depending oil the location; and nearly all
of it falls during the winter. The precipitation for the Riverside and
Pasadena areas varies from 7 to 12 inches or more, also limited almost
entirely to the winter months. The nitrification tests were arranged as
described above and in the same way as those of the foreign soils. The
first area to be considered here is the Bay area, and Table III sets forth
the results obtained with the soil types of that area and with the different
forms of nitrogen.

Journal of Agricultural Research

Vol. VII, No. 2

Table III. — Nitrification in types from the Bay area series

Soil type.

Soil nitrogen
(Group I).

Soil nitrogen and
sulphate of am¬
monia (Group II).

Soil nitrogen and
dried blood
(Group III).

Soil nitrogen and
cottonseed meal
(Group IV).

Ni¬

trate

pro¬

duced.

Total

nitro¬

gen

soil.

Ni¬

tro¬

gen

nitri¬

fied.

Ni¬

trate

pro¬

duced.

Total

nitro¬

gen

soil.

Ni¬

tro¬

gen

nitri¬

fied.

Ni¬

trate

pro¬

duced.

Total

nitro¬

gen

soil.

Ni¬

tro¬

gen

nitri¬

fied.

Ni¬

trate

pro¬

duced.

Total

nitro¬

gen

soil.

Ni¬

tro¬

gen

nitri¬

fied.

Mgm.

Mgm,

P.ct.

Mgm.

P.ct.

Mgm.

P.ct .

Mgm.

P.ct.

Tidal marsh clay .

Tr.

5-oo
9- 75

Altamont heavy loam .

2-35

112.00

7- 75

152.00

14- 75

244.00

64-

159-00

64-

Yolo silty clay loam. . .

1. 20

93.80

.80

133-80

Tr.

. 60

225.80

Tr.

*45

140. 80

Tr.

Dublin clay adobe ....

1*50

65.80

4- 00

105.80

4. 00

197.80

2-h

6.00

112. 80

Dune sand .

.80

35*oo

.72

75-oo

Tr.

167.00

1. 20

82.00

Oakley light sandy

loam .

3-oo

51-80

4.00

91.80

—1. 00

183.80

3-50

98. 80

Altamont clay adobe. .

•30

50.40

Tr.

.40

90.40

Tr.

182. 40

97-40

Diablo clay adobe .

2.00

6s- 80

12.00

105. 80

7.80

197.80

8.00

112. 80

Residual day adobe. . .

3*oo

1 16. 20

3*00

156. 20

3.00

248. 20

2. 50

163. 20

Santa Rosa loam .

• 9c

86.80

4* 50

126. 80

9.50

218.80

44-

7- 50

133- 80

Altamont light types. .

3.00

72.80

3-oo

112. 80

204.80

7.00

1 19. 8c

No. 150 Brown type

heavy loam .

5- 50

183.40

2-f

20.90

223-40

x8. <0

3*5- 40

II- 50

230. 40

Tuscan stony loam. . . .

•43

32. 20

•43

72. 20

Tr.

2. 13

164. 20

14-

- 63I

79. 20

Tr.

Do .

I.40

138,60

5-30

178.60

4.80

270. 60

1 4-

4-30

185.60

24*

Santa Rosa loam .

i- 57

61. 60

1. 28

roi. 60

2 +

193. 60

.48

to8. 60

Tr.

No. 150 Gray phase

loams .

1. 80

82.60

1.80

122. 60

1 +

.60

214.60

Tr.

5- 10

129.60

Newark loam .

4.10

75.60

4*30

115.60

3 +

25.90

207.60

12+

6. 10

122.60

Antioch loams and

clay loams .

4-36

72.80

6. 56

112. 80

15.96

204. 80

16. 96

1 19. 80

14+

No. 12 black phase _

1.60

99.40

6.40

139- 40

• 4+

12.00

231.40

7.60

146. 40

Auburn clay loam .

.80

32. 20

£+

•50

72. 20

Tr.

i. 00

164. 20

Tr.

. 10

79. 20

Tr.

Coming loam .

11+

Pleasanton .

4.00

64.40

2. 80

104. 40

3 J

-30

196. 40

Tr.

20.00

11 1. 40

17+

Montezuma clay loam.

. 20

32. 20

Tr.

•30

72. 20

Tr.

166. 20

79. 20

Montezuma clay adobe

2. 20

86.80

8. 20

126.80

25. 20

218. 80

11+

16. 20

133- 80

124-

Yolo clay .

3-70

113-40

153-40

245. 40

160.40

Tr.

Yolo light type silt

loam .

4.00

131.60

3 +

18. 20

171.60

10+

19. 20

263. 60

23. 20

178. 60

124-

Coming loam .

58.80

98. 80

— .06

5 +

Yolo gravelly loam. . . .

2.18

82.60

2 +

4-68

122. 60

33-68

214. 60

iS+

23.68

129. 60

i84-

Antioch loam .

•50

64.40

Tr.

•50

104. 40

Tr.

196. 40

.40

ill. 40

Tr.

Antioch light sandy

loam .

2.84

105.00

14.04

145-00

12.04

237.00

21.04

152.00

13+

BAY AREA SOILS

The figures for total nitrogen in the second column of Table III show
that only 8 soils out of 30, or approximately 26 per cent, contain more
than 0.1 per cent of nitrogen, and that only one of them contains more
than 0.14 per cent of nitrogen. In other words, the foreign soils contain
relatively 2% times as many soils which contain nitrogen in excess of 0.1
per cent as do the soils of the Bay area; and, moreover, many of the
first-named group contain very much more nitrogen than 0.15 per cent.
The effects of the arid climate are therefore quite evident on soils of the
Bay area and are only emphasized by comparison with the foreign soils
existing under a humid climate. It must be further remarked that the
comparison gains in significance from the reflection that while the Bay
area soils are subjected to a long season of drought, they receive annually
between 20 and 30 inches of rainfall, depending on the part of the area
concerned, and are in addition protected from excessive oxidation influ¬
ences by much fog and cool weather. In respect to the number of soils

Oct. 9, 1916

Nitrifying Powers of Humid and Arid Soils

in the Bay area containing less than 0.05 per cent of total nitrogen, the
last-named group of soils is not unlike the foreign group. The soils of
the latter contain 14 per cent of such soils as against 16 per cent for the
Bay area. No soil in the Bay area contains less than 0.023 per cent of
total nitrogen.

After a study of one form of nitrogen at a time and in the same
order as before, a striking difference is found between the absolute quanti¬
ties of nitrates produced by the soils of the Bay area out of their own
nitrogen and those produced by the foreign soils under similar circum¬
stances. Thus, for example, only 1 soil out of 30 in the Bay area pro¬
duces a little more than 5 mgm. of nitrate nitrogen, and all other soils
produce less. If this observation is compared with the corresponding one
for the foreign soils of the humid region, the feeble nature of the nitri¬
fying power of the soils of the Bay area for their own nitrogen is very
noticeable, as is the very vigorous power in that direction possessed by
the foreign soils. In the latter only 2 soils out of a total of 47 produced
less than 5 mgm. of nitrate nitrogen, while in the former only 1 soil pro¬
duced more than 5 mgm. of nitrate nitrogen. The absolute magnitude
of nitrate production in soils of the Bay area varies from 0.1 mgm. to
5.50 mgm., thus making a very small range. Again, there is evidence
in Table III that the magnitude of nitrifying power for soil nitrogen is
independent of the total amount of nitrogen present. Thus, for exam¬
ple, the largest production of nitrate occurs in the soil with the highest
total nitrogen content in the whole series. On the other hand, the
lowest nitrate production occurs in the soil with the third highest quan¬
tity of total nitrogen.

It follows from what has been said in the preceding paragraph that
only small percentages of the nitrogen present in the soils of the Bay
area could have been transformed into nitrate. The best record made
consists in a conversion by the Pleasanton soil of over 6 per cent of the
total nitrogen present into nitrates. In three other soils 5 per cent of
the nitrogen was thus converted, and in the other 26 soils the records are
much poorer. Comparing the relative data of Table III with those set
forth in Table I, section 1, one can not help being struck by the remarka¬
bly high nitrifying efficiency of the foreign soils as compared with that
of the Bay area soils.

The use of sulphate of ammonia as a nitrifiable material with the Bay
area soils shows also the relatively low nitrifying power of the latter.
Nevertheless Table III, Group II, brings out a very important fact —
that is, that in the case of the absolute amounts of nitrate produced
the Bay area soils gave increases over the amount of nitrate produced
from the soil alone in all but 5 out of a total of 30 soils. This is a dia¬
metrically opposite effect to that induced by sulphate of ammonia in the
foreign soils. On a relative basis the results in the sulphate-of -ammonia

Journal of Agricultural Research

Vol. VII, No. 2

series were superior to those in the soil-nitrogen series as is shown clearly
in Table III, Group II, whereas in Table III, Group I, the highest per¬
centage of nitrogen transformed into nitrate was slightly in excess of 6.
The corresponding figure for Group II is n, and several other soils, be¬
sides, approach that record. In general, it would seem that sulphate of
ammonia stimulates nitrification in soils of the Bay area, while it depresses
nitrification in the foreign soils.

With dried blood the Bay area soils show a loss in nitrate production
over that of which the soil is capable on its own nitrogen supply in over
43 per cent of the soil types, so far as absolute quantities of nitrates are
concerned. In a number of these cases, moreover, there is not only less
nitrate produced than from soil nitrogen alone, but the soil’s original
nitrate supply is lost besides. On the other hand, there are about 9 soils
in the Bay area series which, on the basis of production of nitrate in the
absolute sense, surpass any in the series with soil nitrogen alone, and,
similarly, all but one in the sulphate-of-ammonia series. It may be
noted, however, that none of these soils consists of coarse, sandy material
with a small absorbent surface, and none contains less than 0.073 percent
of total nitrogen, nearly all of them containing considerably more than
0.08 per cent. It is the last group of 9 soils showing naturally a number
of good results on the relative basis in Table III, Group III, which reveals
the highest percentage transformation of total nitrogen into nitrate yet
noted with the Bay area soils in these experiments. The record is a
15 per cent transformation in the case of the Yolo gravelly loam. One
transformation of 12 per cent and one of 11 per cent are also noted, but
all the rest are considerably below those figures. In general, it would
seem that the heavier soils and such as are better supplied with nitrogen
than the average of the Bay area series will give better results with
dried blood than with sulphate of ammonia or soil nitrogen, but the
rest (more than two-thirds of the total number) will not do as well with
dried blood as with the other forms of nitrogen.

Considering the cottonseed-meal results in the case of the Bay area
soils, as set forth in Table III, Group IV, from the absolute amounts of
nitrate produced cottonseed meal is to be regarded as of less value than
dried blood from some points of view and of more value from others.
To be more specific, no soil produces as much nitrate from cottonseed
meal as does the Yolo gravelly loam from dried blood. On the other
hand, there are less soils in the cottonseed-meal series than in the dried-
blood series which are induced to lose in nitrifying power by the incor¬
poration of the fertilizer and none at all which lose part or all of the
nitrate nitrogen originally contained in them. Relatively, the cot¬
tonseed-meal series is ahead of all others with the Bay area soils in
that the largest amount of nitrogen transformed into nitrate is there
noted. This record is attained by the same soil as that having the
record in the dried-blood series and amounts to an 18 per cent trans¬
formation of the total nitrogen present into nitrate. The Pleasanton

Oct. 9, 1916

Nitrifying Powers of Humid and Arid Soils

soil accomplishes a 17 per cent transformation in this series, and three
other soils transform more than 12 per cent of the nitrogen present into
nitrate. The other soils are considerably inferior in the direction noted.

PASADENA AREA SOIES

The soil types of the Pasadena area number 33, and data are here
given either for the whole number or for one or two less, in accordance
with the circumstances attending the experimental work. Table IV
sets forth the results obtained with the Pasadena area soils with the
different forms of nitrogen as above described.

Table IV. — Nitrification in Pasadena area soil type

Name of type.

Soil nitrogen
(Group I).

Soil nitrogen and
sulphate of am¬
monia
(Group II).

Soil nitrogen and
dried blood
(Group III).

Soil nitrogen and
cotton-seed meal
(Group IV).

Ni¬

trate

pro¬

duced.

Total
ni¬
tro¬
gen
pres¬
ent in
soil.

Ni¬

tro¬

gen

nitri¬

fied.

Ni¬

trate

pro¬

duced.

Total
ni¬
tro¬
gen
pres¬
ent in
soil.

Ni¬

tro¬

gen

nitri¬

fied.

Ni¬

trate

pro¬

duced.

Total
ni¬
tro¬
gen
pres¬
ent in
soil.

Ni¬

tro¬

gen

nitri¬

fied.

Ni¬

trate

pro¬

duced.

Total
ni¬
tro¬
gen
pres¬
ent in
soil.

Ni¬

tro¬

gen

nitri¬

fied.

Mgm.

Mgm

P.ct.

Mgm.

P.ct .

Mgm .

Mgm.

P. cf.

Mgm

Mgm.

P.ct.

Dublin clay adobe ......

7. 60

109. 20

39.60

149. 20

26+

2* 60

241. 20

50.60

156. 20

32+

TYiahln clay .

5.68

10+

15.88

94. 60

16+

— • 32

186. 60

7.68

101. 60

Zeinab li ght loam .

20. 30

ft. 80

IT'S. OO

— . 20

209. 00

5. 80

124. 00

Altamon t clay .

9. 80

154-00

47. 80 194. 00

24+

39- 80

286.00

13+

27. 80

201. 00

13+

Zelzah gravelly loam. . .

9.44

61. 60

15+

15.44 101.60

15+

— • 56

193- 60

20. 44

108. 60

18+

TTnllatiH Inam .

67. 20

— .08

1 14. 20

24-b

Hanford coarse sandy

loam . .

4. 28

54- 60

9- 38

94. 60

— .13

186. 60

.68

101. 60

Trace

Hanford gravelly sandy

loam .

5. 59

68.60

17. 10

108. 60

15+

- .81

200. 60

- .81

115. 60

Hanford sandy loam. . .

3- 70

40. 60

5.90

80. 60

— .30

172. 60

3.90

87-60

Hanford fine sandy

]nam .

15. 57

— . 4^

13. 58

Hanford loam .

100. 80

14- 79

140. 80

10-f

9- 79

232. 80

39- 79

147. 80

26+

18. 20

58. 20

4-t-

150. 20

65. 20

Zelzah stony loam .

6. 00

64.40

7. 00

104. 40

196. 40

24. OO

in. 40

21 +

Hanford stony loam. . . .

3-88

28. 00

13 +

6. 28

68. 00

9 +

- - 13

160. 00

- -13

75-oo

Placentia loam .

5. 50

68.60

3- 00

108. 60

2 +

— I- 50

170. 60

28. 50

115. 60

24 4-

Holland sandy loam . . .

5* 39

46. 20

11 +

3-39

86. 20

3 +

— . 61

178. 20

5- 39

93- 20

5 +

Hanford stony sandy

loam .

3- 80

67. 20

0. 30

107. 20

— 2. 20

199. 20

11. 80

1 14. 20

10+

Antioch clay loam .

8. 00

182. 00

4 +

38. 00

222. 00

17 +

122. 00

314.00

38+

36. 00

229. 00

15+

Chino silty clay loam. . .

14-40

151. 20

•9 +

42.40

191. 20

22 +

79-40

283. 20

27+

30.40

198. 20

15 +

Chino clay adobe .

II. 00

382. 20

2 -j-

61. 00

422. 20

14+

119. 00

514- 20

23 +

33-oo

429. 20

Chinn loam .

— 9. 00

187. 60

10. 00

227. 60

36. 00 3 iq. 60

11 +

68. 00

234. 60

28+

Hanford fine sandy

loam .

13- 20

88. 20

14+

45- 20

128. 20

35 +

8. 70 220. 20

3 +

77. 20

135- 20

57 +

Chino silty clay loam. . ,

7. 00

219. 80

3 +

55- 00

259. 80

21 +

81. 00351. 80

23 +

31. 00:266. 80

n +

Hanford fine sand .

1. 20

35-oo

3 +

4. 20

75.00

5 +

— 1. 80 167. 00

IO. 20

82. 00

12 +

24+

55. 40

5 +

— . 20

I A'7. AO

— . 20

62. 40

Yolo loam . . .

— 2. 00

123* 20

13. 60

163. 20

42. 00.255. 20

16+

8. 30

170. 20

Dublin clay loam .

7.80

105. 00

11. 80

145-00

19. 8o]237- 00

75- So

152. 00

49+

Dtihlin clay .

Q. <K

13 1. 60

7 +

64. q«:

171. 60

37+

47. Q5 263. 60

18+

31. 95

178. 60

17+

Hanford sand .

4. IO

25. 20

16+

3- 80

65. 20

5 +

157. 20

. 20

72. 20

Trace

Zelzah clay loam .

9. OO

109. 20

8 +

21. 00

149. 20

14+

3. 00 241. 20

1 +

53-oo

156. 20

33 +

Altamont loam .

7- 67

74- 20

10+

14. 67

1 14. 20

12 +

17. 70 206. 20

23. 67:121. 20

19+

Diablo clay adobe .

7- 95

100. 80

7 +

16. 15

140. 80

11 +

35.95 232. 80

15 +

19- 95 I47-8o

13+

Altamont clay loam. . . .

■ 50

93- 80

Trace

13- 00

133- 80

9 +

17. OOj225* 80

13. 80^140. 80

The Pasadena area contains a larger percentage of soils having more
than 0.1 per cent of nitrogen than does the Bay area. In the latter,
for example, 26 per cent of the soils contained more than 0.1 per cent
of nitrogen, while in the former more than 40 per cent of the soils belong
in that class. Nevertheless it must still be observed that even the

Journal of Agricultural Research

Vol. VII, No. a

Pasadena area falls short, approximately by 21 per cent, of having as
large a number of soils with more than 0.1 per cent of nitrogen as the
foreign-soil group. While, therefore, the soils of the Pasadena area
approach more closely in nitrogen content those of the foreign group
than do the soils of the Bay area, they are still distinctly inferior to the
foreign soils. The effect of aridity of climate therefore makes itself
plainly manifest in the Pasadena as it does in the Bay area series. It
must be further remarked, however, that again owing to climatic con¬
ditions, a larger number of soils with less than 0.05 per cent of nitrogen
are found in the Pasadena area than in the Bay area, and more of that
class than among the foreign soils. The percentage of such is greater
than 21 in the Pasadena area, as against 16 for the Bay area, and 14 for
the foreign soils. Two soils out of the thirty-three in the Pasadena area
contain less than 0.02 per cent of nitrogen, and two others contain less
than 0.03 per cent of nitrogen.

On the basis of the absolute values for the nitrate from the soil nitro¬
gen (Table IV, Group I) it would seem that the Pasadena area soils,
though manifestly superior in nitrifying power as a class to the Bay area
soils, are still far from being equal in that direction to the foreign soils.
Thus, for example, 85 per cent of the soils in the Pasadena area produced
less than 10 mgm. of nitrate nitrogen in 100 gm. of soil under the cir¬
cumstances described, while the foreign soils numbered only 27 per cent
of such soils among them, it is therefore very clear that so far as abso¬
lute quantities of nitrate produced are concerned even the fertile Pasa¬
dena area soils are inferior transformers when compared with foreign
soils. On the other hand, when compared with the Bay area soils on the
basis of a 5-mgm. production of nitrate, the Pasadena soils are clearly
superior. Thus, among the latter there are 20 soils of the class last
mentioned, whereas among the Bay area there is but 1 such soil.

On the relative basis, or that the criterion of which is percentage of
soil nitrogen transformed into nitrate, the Pasadena area soils make
even a better showing than on the absolute basis when compared with
the Bay area soils. Likewise, they are less inferior when so judged in
comparison with the foreign soils for obvious reasons concerned with
the total nitrogen content. Thus 4 soils out of 33 (12 per cent) in the
Pasadena area transforms more than 20 per cent of their total nitro¬
gen content into nitrate as against 11 out of 44 soils of that class, or

25 per cent, in the case of the foreign soils. The highest individual
record for percentage nitrogen transformation is attained equally by
the Zelzah light loam and the Holland loam, which transform more than

26 per cent of the total nitrogen present into nitrate. This, while very
high, is below the record attained by several of the foreign soils. All
of these considerations, moreover, must be viewed in conjunction with
the fact that even the group of foreign soils contains a few arid or semi-
arid soils similar to the California soils which are being studied here.

Oct. 9, 1916

Nitrifying Powers of Humid and Arid Soils

Some very interesting data are available in Table IV, Group II, which
gives the results with sulphate of ammonia. Thus, only about 2 1 per cent
of the soils in the Pasadena area produced less nitrate from the sulphate
of ammonia nitrogen plus the soil nitrogen than from the latter alone.
It is interesting to note that the percentage of such soils is so nearly
the same in the two arid-soil series thus far considered, even though
the latter are in other respects very different. To emphasize again the
wide difference existing in respect to the sulphate-of -ammonia nitrogen
between the humid (as illustrated by the foreign) and the arid soils,
one need but recall that 88 per cent of the humid soils failed to respond
to sulphate of ammonia, whereas only 20 per cent of the arid soils
behaved in that manner.

As other points of interest in Table IV, Group II, may be mentioned
the following: (1) Only four soils in the whole series transform less than
5 per cent of the total nitrogen present into nitrates. (2) Six soils of
the series transform more than 20 per cent of the total nitrogen present
into nitrates. (3) In one soil, the Dublin clay, over 37 per cent of
the total nitrogen present is nitrified, and another soil, the Hanford fine
sandy loam, approaches closely to that record. (4) Both soils, the
Zelzah light loam and the Holland loam, which have the highest record
on the relative basis in Table IV, Group I (soil nitrogen alone), lose in
nitrifying power very markedly when sulphate of ammonia is added to
them, while the Dublin clay and the Hanford fine sandy loam, which do
only moderately well with soil nitrogen alone, make the highest records,
as above indicated, with soil nitrogen plus sulphate-of -ammonia nitrogen.
(5) It should be noted that in the sulphate-of-ammonia series no soil
loses its original nitrate content without replacing and adding to it by
nitrification. A loss of the soil's original nitrate content does, however,
occur in the case of two soils, the Chino loam and the Yolo loam, in the
series with soil nitrogen alone. All of these points, moreover, are of great
interest in comparison with the results for the Bay area soil series as
obtained by the use of sulphate of ammonia. In general, of course, the
superiority at nitrification of the soils in the Pasadena series to that of
the Bay area series is more emphasized in Table IV, Group II, than here¬
tofore.

In the experiments with dried blood in the case of the Pasadena area
soils (see Table IV, Group III), results totally different in nature from
those obtained with sulphate of ammonia are noted. Thus, 63 per cent
of all the soils tested produce less nitrate under the circumstances above
described when dried blood plus the soil nitrogen are available for nitri¬
fication than when only the soil nitrogen. is present. This is 20 per cqnt
in excess of the number of such soils in the Bay area, despite the fact
that the latter area contains less soils than the Pasadena area with a
percentage of nitrogen higher than 0.1. Again, however, as in the case
of the Bay area, several of the soils produce in absolute quantities much
55856°— 16 — —2

Journal of Agricultural Research

Vol. VII, No. 2

more nitrate than is produced by any of the same series when either soil
nitrogen alone is present or when it is present with sulphate of ammonia.
The number of such soils is about the same in the Pasadena as in the
Bay area series, and in no one of them does the nitrogen content go quite
as low as o.i per cent. They are, besides, soils of large internal surface
throughout. Of the two other soils, which, in addition to the ones just
mentioned, produce more nitrate in the blood series than in either of the
foregoing, both are of large internal surface, and one contains very nearly
o.i per cent of nitrogen, while the other contains nearly 0.075 per cent
of nitrogen.

While on the absolute basis in the dried-blood series the Pasadena and
Bay area soils are much alike, with the former in some respects superior
and in other respects inferior to the latter, the difference is more marked
on the relative basis. Thus, the records made for a percentage trans¬
formation of nitrogen into nitrate attain higher values in the Pasadena
area than in the Bay area soils, and four soils of those above noted trans¬
form more than 20 per cent of the nitrogen present into nitrate, while
three others pass the 15 per cent mark. It will be seen that there is
only one soil in the Bay area even in the latter class in the dried-blood
series. The foreign soils behave as a class of humid soils in a diametrically
opposite manner from the Pasadena area soils with respect to dried blood.
For the most emphatic proof of this, the reader can compare this para¬
graph with that discussing Table I, Group III.

Cottonseed meal gives in many respects results similar in the Pasadena
area soils to those obtained with it in the Bay area soils, though in one
or two respects the two are very different. Thus, for example, 26 per
cent of the Bay area soils produce less nitrate from soil nitrogen plus
cottonseed-meal nitrogen than from the former alone. The correspond¬
ing figure for the Pasadena area soils is 21 per cent. It is also interesting
to observe that the last-named value is exactly or very nearly that of
the analogous figure for the sulphate-of-ammonia series in the two soil
areas above compared. Most striking of all are the very high absolute
and almost necessarily high relative amounts of nitrates produced by
many of the Pasadena area soils in the cottonseed-meal series. Thus,
while there are among the foreign soils but two which transform more
than 20 per cent of the total nitrogen in the soil and cottonseed meal
into nitrate and none such in the Bay area soils, there are nine, or 28
per cent, of such soils in the Pasadena area group. Moreover, four of
these nine transform, as indicated, more than 30 per cent of the nitro¬
gen into nitrate, and one of these reaches the very high figure of a 57
pejr cent transformation. That cottonseed meal can be more readily
and efficiently nitrified in the Pasadena soils as a class than it can in
the foreign and the Bay area groups of soils is patent. Since, however,
the Pasadena soils are under the most arid conditions of the three and

Oct. 9, 1916

Nitrifying Powers of Humid and Arid Soils

the Bay area soils are the intermediate group in that respect, it appears
that cottonseed meal is better suited to arid than to humid soils. It
may be well to note here that in the different groups of soils thus far
studied the percentages of soils producing less nitrate from cottonseed
meal plus soil nitrogen than from the latter alone are as follows: For¬
eign soils, 37 per cent; Bay area, 2 6 per cent; and Pasadena area, 21 per
cent. The first figure is doubtless too low, because the foreign soils
include several arid and semiarid soils which were not separated for
purposes of this calculation.

RIVERSIDE) AREA SOILS

Only 3 out of 52 soils of the Riverside area contain as much as 0.1
per cent of nitrogen, or more. That is equivalent to something over
5 per cent of the total number of soils and is strikingly low when
contrasted with 66 per cent of such soils for the foreign soils, 40 per
cent for the Pasadena area, and 26 per cent for the Bay area. What is
even more striking is that including the three soils just mentioned there
are but 8 soils (15 per cent) in the Riverside area which contain as
much or more than 0.05 per cent of nitrogen. Over 36 per cent, or
more than one-third of all the soils in this area, contain less than 0.03
per cent nitrogen. Two soils contain less than 0.0 1 per cent nitrogen
and four others less than 0.015 Per cent of nitrogen. Of the four groups
of soils thus far studied, including the Riverside area, this last is clearly
one in which the total nitrogen content is distinctly below that of all
other groups. From the discussion already given this subject, such a
circumstance should not be unexpected, since the general tendency is for
more arid climates to produce soils with a lower nitrogen content than
that of soils in a humid climate.

When the situation is reviewed with respect to nitrate formation from
the soil-nitrogen supply in the Riverside areas (see Table V) , some further
interesting data become evident. In the first place it is plain that the
absolute amounts of nitrate formed are very small and, while of a
slightly greater magnitude than those of the Bay area, are still of about
the same order. No soil in the whole area produces, under the conditions
noted, as much as 10 mgm. of nitrate nitrogen, the largest amount pro¬
duced being 7.40 mgm., produced by the Montezuma silty clay loam.
Moreover, there are but 13 out of 53 soils, or about 24 per cent, which
produce as much or more than 5 mgm. of nitrate nitrogen under the same
conditions. There are, thus, more than three-fourths of the total number
of soils tested in the Riverside area which produce less than 5 mgm. of
nitrate nitrogen, and most of them form from 1 to 3 mgm. only.
Here again, as in the case of the Bay area soils and the others, it appears
impossible to find evidence for establishing a definite relation between

Journal of Agricultural Research

Vol. VII, No. 2

total nitrogen present and nitrate produced. Thus, for example, in the
Riverside area, as in the Bay area, the largest amount of nitrate formed
from the soil nitrogen is found in the soil with the highest total nitrogen
content, but no nitrates are produced by the soil with the third highest
total nitrogen content.

Tabi,B V. — Nitrification in Riverside area soil types

Soil nitrogen
Soil nitrogen and sulphate of
(Group I). ammonia

(Group II).

Soil nitrogen
and dried blood
(Group III).

Soil nitrogen
and cottonseed
meal (Group IV),

Name of type.

Total
nitro¬
gen
pres¬
ent in
soil.

Nitro¬

gen

nitri¬

fied.

Ni¬

trate

pro¬

duced

Total
nitro¬
gen
pres¬
ent in
soil.

Nitro- Ni-
gen trate

nitri- pro-
fied. duced

Total
nitro¬
gen
pres¬
ent in
soil.

Nitro¬

gen

nitri¬

fied.

Ni¬

trate

pro¬

duced

Total!
nitro¬
gen
pres¬
ent in
soil.

Nitro¬

gen

nitri¬

fied.

Mgm.

P.ct.

San Joaquin sandy loam. . .

1.80

32.20

Placentia loam .

5*80

42.00

13+

Zelzah sandy loam .

4. 00

26. 60

15+

Tejunga fine sandy loam . . .

5.20

47-60

10+

Sierra loam .

.60

15.40

3 +

Placentia sandy loam .

5.18

22.40

23 +

Tejunga fine sand . .

3-92

28.00

Zelzah loam, .

3*74

32. 20

11+

Zelzah clay loam .

3*40

53*20

Holland sandy loam .

2.80

29.40

Hanford loam .

2.80

28.00

10+

Hanford clay loam .

6. 80

74-20

Sierra loam .

1.90

32. 20

Hanford sandy loam .

5*96

28.00

21+

Hanford fine sandy loam . . .

5-35

33* 60

15+

Hanford gravelly sandy

loam .

1.40

36.40

Hanford stony gravelly

sandy loam .

3- 80

44.80

Hanford loam .

5- 60

49*00

11+

Hanford stony sand .

2*35

5- 60

41+

Hanford sandy loam .

4. 05

43-40

Tejunga sand .

3*00

14.00

21 +

Tejunga gravelly sand .

1-52

11. 12

13+

Hanford sand .

S-6o

21.00

26+

Hanford gravelly sand .

4. 10

37.80

10+

San Joaquin loain .

2.00

35-oo

5 +

Tejunga sandy loam .

3.20

33- 60

Hanford fine sandy loam. . .

•4.00

28.00

14+

Hanford gravelly loam .

4.92

47-60

10+

Placentia gravelly loam. . . .

5- 00

36. 40

13+

Hanford stony loam .

no. 60

Hanford coarse sandy loam.

3-io

37.80

Montezuma loam .

6. 20

161.00

Montezuma clay adobe .

4- 45

33* 60

13 +

Hanford coarse sand .

*•35

15.40

Pedley fine sandy loam ....

2.40

32. 20

Montezuma silty clay loam.

7.40

182.00

Hanford coarse sandy loam .

1.85

21.00

Olympic loam .

3*60

40. 60

Oakley sand .

3*oo

26. 60

11+

Tejunga stony sand .

Kimball fine sandy loam. . .

2.40

9. 80

24+

Aiken loam . . .

6.00

50.40

11+

Mendocino loam .

3-oo

21.00

14+

Holland fine sandy loam , . .

4.00

39- 20

10+

Corona gravelly sandy loam

5- 10

54- 60

Corona clay loam .

3.80

50.40

Yolo gravelly loam .

4.00

47- 60

Rincon loam .

2. 80

39- 20

Antioch silty clay loam. , . .

3*80

35-oo

10+

Hanford silty clay loam. . . .

I.SO

28.00

Placentia clay .

2.00

42.00

Sierra sandy loam .

4.80

25. 20

19+

Holland loam .

2.40

28.00

Mgm.

P.ct.

Mgm.

P.ct.

Mgm.

P.ct.

1.50

72.20

—1. 00

164. 20

10. 00

79. 20

12+

15-30

82.00

18+

47- 90

174.00

27+

21.90

89.00

24+

40. 00

66.60

60+

1 58. 60

32. 00

73. 60

A?* f*

47. 20

87.60

53+

95-20

179. 60

53+

50. 40

94-60

53 +

1. 50

55.40

2 +

147. 40

3. 00

62. 40

6.07

62.40

— , 12

154.40

11.88

69.40

17+

63.92

68.00

-.08

160.00

14. 92

75-00

19+

9*74

72. 20

13 +

8. 54

164. 20

15-74

79. 20

19+

13.90

93-20

14+

32. 10

185. 20

17+

15.90

i 00. 20

15 +

5. 00

69. 40

161. 40

12. 80

76. 40

x64-

6.40

68.00

6.00

160.00

20. op

75-oo

26+

23.40

1 14. 20

20+

63.60

206. 20

30+

38.40

121. 20

31+

4.80

72. 20

23.80

164. 20

14+

11.80

79* 20

14+

9-36

68.00

13+

—.04

160.00

6.96

7S-oo

15-75

73- 60

21 +

5- 75

165.60

31-75

80. 60

39+

17.40

76.40

22+

—.60

168.40

27.40

83.40

32+

12. 60

84. 80

14+

176.80

17. 50

91. 80

19+

27.60

89.00

31 +

33*6o

181.00

18+

63.60

96.00

66+

1*15

45- 60

-*05

137-60

•95

52.60

8.85

83, 40

10+

38.85

175-40

22+

27- 85

90. 40

30+

1. 90

54- 00

146.00

3* 00

61. 00

19-93

5X- 12

38+

—.08;

143-12

• ••*««

4. 22

58.12

.80

61.00

.60

i53-oo

Tr.

2. 80

68.00

4. 10

77-8o

-1.50

169. 80

15.20

84. 00

18+

14.00

75.00

18+

1.60

167.00

Tr.

25.40

82.00

30+

7. 20

73-60

8.40

165. 60

15.20

80.60

18+

12.00

68.00

17+

.90

160. 00

Tr.

16. 00

7S-oo

21+

7.92

87. 60

1.42

179. 60

Tr.

13.92

94- 60

14+

2.00

76.40

.90

168. 40

Tr.

.40

83.40

Tr.

15. 70

150. 60

10+

47.40

242 60

19+

17.40

157-60

11+

4.00

77-8o

2.40

169. 80

24.00

84.80

28+

44-oo

201. 00

21+

49-60

293.00

16+

32.50

208.00

15+

5-95

73- 60

2,15

168. 60

3-35

80.60

•45

55-40

Tr.

— -05

286.00

•45

62.40

Tr.

3*30

72. 20

.80

164. 20

Tr.

6. 50

79. 20

46.40

222.00

20+

70.40

314-00

22+

38. 40

229. 00

16+

I- 95

61. 00

3 +

-.15

153.00

3-85

68.00

6.00

80. 60

4.40

172. 60

4.00

87.60

6. 40

66.60

158.60

1. 30

73- 60

3- 10

1.40

35- 00

8. 00

49. So

16+

5-oo

141. 80

3 +

18.00

56.80

3i +

7.00

90.40

20.00

182. 40

10+

14. 20

97-4°

14+

5-20

61. 00

16.00

i53-oo

10+

7-50

68. 00

11+

10.00

79. 20

12+

•50

171. 20

Tr.

15.60

86. 20

18+

13.80

94. 60

14+

58.40

186. 60

31+

28. 40

101.60

27+

5-oo

90.40

5 +

20.00

182.40

10+

14.00

97-40

14+

7.00

87. 60

32.00

179. 60

17+

15*80

94.60

16+

3.20

79. 20

4.00

171. 20

3-oo

86. 20

3 +

7.80

75.00

10+

26. 30

167.00

15+

14. 80

82. 00

18+

5*oo

68.00

2. 10

160.00

2. 80

75-00

- 3 +

8.00

82.00

12.00

174.00

14.60

89. 00

16+

5. 20

65. 20

.70

157- 20

13- 20

72. 20

18+

3.00

68.00

160. 00.

9. 20

75-00

12 +

Oct. 9, 1916

Nitrifying Powers of Humid and Arid Soils

Owing to the very low total nitrogen content of the Riverside area
soils, the relative figures for nitrate transformation given in Table V,
Group I, are, if wholly expected, exceedingly large in many instances.
Thus, there are six soils, or n per cent of the whole number studied,
which nitrify more than 20 per cent of the total nitrogen present and 24,
or 46 per cent of all the soils, which transform more than 10 per cent of
the total amount of nitrogen present into nitrate. On the relative basis,
therefore, the Riverside area soils are far superior in nitrifying power to
the Bay area soils, but not so in relation to the Pasadena soils; and
when compared with the foreign, consisting very largely of humid soils,
the Riverside soils are much inferior even on the relative basis.

The results obtained with the soils of the Riverside area employing
sulphate of ammonia are perhaps the most striking of all thus far studied.
Thus, on the absolute basis there are but 6 soils, or a little over
1 1 per cent of the whole number tested, which yielded less nitrate (see
Table V, Group II) with the combined nitrogen of the soil and of the
sulphate of ammonia present than from the former alone. It will be
recalled that the corresponding figures for the foregoing series were as
follows: Foreign soils, 87 per cent; Bay area soils, 16 per cent; and Pasa¬
dena area soils, 21 per cent. In other words, it would appear that the
more distinctly arid a soil's character is the more likelihood there is of its
being favorably affected by sulphate of ammonia, or to put it otherwise,
to have its nitrifying flora stimulated to greater activity. It will be
noted that the 6 soils of the Riverside area which were unfavorably
affected as to nitrifying power by the sulphate of ammonia are all
either sandy, coarse sandy, or gravelly soils. The actual amounts
of nitrate produced in many of the soils of this series under the
influence of sulphate of ammonia are very large — for example, there
are 5 soils which produce more than 40 mgm. of nitrate nitrogen under
the conditions of the experiment and 2 other soils which produce more
than 20 mgm. of nitrate nitrogen. There are 18 soils in the area, or
more than 33 per cent of the whole number studied, which produce more
than 10 mgm. of nitrate nitrogen, and 1 soil which produced as much as
63.92 mgm. of nitrate nitrogen. It will be remembered, however, that
even this record is surpassed by one of the Pasadena area soils with
sulphate of ammonia and that much larger absolute productions of nitrate
nitrogen are accomplished with other forms of nitrogen in a number of
instances by the foreign soils.

Partly as a result of the high absolute quantities of nitrate produced by
soils in the Riverside area and partly owing to the relatively small
quantity of total nitrogen present, the percentages given in the last
column of Table V, Group II, make a very good showing. Thus, there
are 10 soils, or about 18 per cent of the number tested, which transform
in every case more than 20 per cent of the total nitrogen present into

Journal of Agricultural Research

Vol. VII, No. a

nitrate, and 23 soils, or about 43 per cent of the whole number in this
area, which nitrify 10 per cent or more of the nitrogen present. When
compared with 66 per cent of the soils with such a record among the
foreign group with the soil nitrogen alone, it still seems clear that the arid
soils are inferior as nitrifying media. Nevertheless one remarkable figure
must not be lost sight of among the relative data of Table V, Group II—
that is, the 94 per cent transformation of the totai nitrogen present
into nitrate by the Tejunga fine sand. This means an almost complete
nitrification of both soil nitrogen and sulphate of ammonia nitrogen
added in one month under the conditions noted.

In the case of dried-blood nitrogen, the Riverside area soils again
behave typically. Fifty per cent of all the soils produce in the abso¬
lute, as reference to Table V, Groups III and IV, will show, less nitrate
from dried blood plus soil nitrogen than from the latter alone. In the
Riverside area, however, even more markedly than in the two preceding
California areas, there is a considerable number of soils producing large
quantities of nitrates from soil plus dried-blood nitrogen. Two points
of difference are noted between the Riverside area soils on the one
hand and the two groups just mentioned on the other. There are
more soils of the class just described in the Riverside area soils than in
the Bay area or Pasadena area soils, and very few of them contain
nearly as much as 0.1 per cent of total nitrogen. Besides, the absolute
quantities of nitrate produced are in four or five cases exceptionally
high. It must be noted, however, that in only one case does the nitrifi¬
cation of dried blood exceed that of the soil nitrogen when the soil is
stony, sandy, or gravelly and contains little loam or clay. Even that
one exception shows but feeble powers of nitrification. In the other
cases the soils vary from fine sandy loams with large internal surface
to clay loams and clays with larger internal surface. Likewise, it will
be noted that in the cases of the 3 soils with more than 0.1 per cent
total nitrogen every one made a very good showing in the nitrification
of dried-blood nitrogen. It will further be observed that of the 12 soils
producing more than 25 mgm. of nitrate nitrogen in this series only one
contained less than 0.04 per cent of total nitrogen.

On the relative basis it follows that the Riverside area soils must
surpass the Bay area soils in efficiency in the dried-blood series and that
they must equal the Pasadena area soils, but they do not do either with
respect to the foreign soils. For example, five soils in the group now
under consideration transform more than 20 per cent of the blood plus
soil nitrogen into nitrate, and six more transform more than 15 per
cent of the total nitrogen present in that manner. This is a record only
slightly behind that of the Pasadena area soils, but as far behind that of
the foreign soils as it is ahead of the Bay area soils. So far as maximum
transformation is concerned, however, the Tejunga fine sandy loam in
the Riverside area surpasses any soil in the Pasadena area by nitrifying

Oct. 9, 1916

Nitrifying Powers of Humid and Arid Soils

53 per cent of the total nitrogen of soil and dried blood under the con¬
ditions of the experiment. In comparison with the foreign soils as
regards dried blood, the Riverside soils must be regarded in the same
light as the Pasadena soils, which will be shown more fully later.

Much the same situation as exists in the cottonseed-meal series of the
Pasadena area soils is to be found in the Riverside area soils (see Table V,
Group IV). There are two principal differences between them. One is
that there are only 11 per cent of the Riverside soils, as against 21 per
cent of the Pasadena soils, which produce less nitrate in the absolute
from cottonseed meal plus soil nitrogen than from the latter alone. The
other is that, on the whole, more vigorous nitrification of cottonseed-
meal nitrogen occurs in the Riverside than in the Pasadena soils. The
latter superiority is based mainly on the fact that more than 17 per cent
of all the soils in the Riverside area transform more than 30 per cent of
the total nitrogen present into nitrate, whereas the corresponding figure
for the Pasadena area is 13 per cent. It may be added also in connec¬
tion with the preceding that the highest percentage transformation of
cottonseed meal plus soil nitrogen found anywhere among the California
soil series studied occurs in the Riverside area soil known as the
Hanford loam, the record being 66 per cent. On the whole, therefore,
the Riverside area soils are the most efficient of any in the nitrification
of cottonseed meal plus soil nitrogen. This result is further strengthened
by the fact that nearly 70 per cent of all the soils present produce 10
mgm. or more of nitrate nitrogen in this series, as shown in Table V,
Group IV. It is clear also that in this group there is further evidence
of the definite relationship of degree of aridity in climate to its effect on
the nitrifying power of soils for a given form of nitrogen.

THE URIAH AREA SOILS

Table VI gives the results obtained in our experiments with the
Ukiah area soils which were carried out in a manner similar to those
described for the other series. Of the Ukiah series, 10 soils, or over 35
per cent of the whole number, contain more than 0.1 per cent of nitrogen.
This is 9 per cent in excess of the corresponding figure for the Bay area,
which ranks highest in the areas studied. Of the 10 soils just mentioned,
6 contain more than 0.14 per cent of nitrogen, a marked contrast to the
corresponding number for the Bay area, which is 1. Nevertheless,
while among all the California soil series here studied the Ukiah area
stands easily first with respect to the number of soils containing more
than 0.1 per cent of total nitrogen, it still falls short more than 26 per
cent of equaling the record in the same regard attained by the foreign
soils. That, as compared with the other California series, the Ukiah
area soils receive the place to which theoretically they should be entitled
on the basis of total nitrogen content is also indicated in Table VI. Thus,

Journal of Agricultural Research

Vol. VII, No. a

it is clear that the Ukiah area should stand first among the California
series here studied. This is so because it receives more rain in a longer
rainy season and is not subjected to the extremely high tenperatures
and other conditions favorable to loss of nitrogen which exist in the
more arid areas. There are only three soils in the Ukiah area, or about
io per cent of the whole number, which contain less than 0.05 per cent
of total nitrogen. Two of these approach that point rather closely and
the other contains nearly 0.03 per cent of nitrogen and is the only soil
which is seriously deficient in nitrogen in the whole area. The 10 per
cent value just given, when compared with corresponding values for
other California areas, again supports the relationship drawn above
between climate and soil nitrogen content and which for the first time is
being properly emphasized.

Table) VI. — Nitrification in Ukiah area soil types

Name of type.

Soil nitrogen
(Group I).

Soil nitrogen and
sulphate of ammo¬
nia (Group II).

Soil nitrogen
and dried blood
(Group III).

Soil nitrogen
and cottonseed
meal (Group IV).

Ni¬

trate

pro¬

duced

Total
ni¬
trate
pres¬
ent in
soil.

Ni¬

tro¬

gen

nitri¬

fied.

Ni¬

trate

pro¬

duced

Total
nitro¬
gen
pres¬
ent in
soil.

Ni¬

tro¬

gen

nitri¬

fied.

Ni¬

trate

pro¬

duced

Total
nitro¬
gen
pres¬
ent in
soil.

Ni¬

tro¬

gen

nitri¬

fied.

Ni¬

trate

pro¬

duced

Total
ni¬
trate
pres¬
ent in
soil.

Ni¬

tro¬

gen

nitri¬

fied.

Ukiah gravelly loam .

Vichy fine sandy loam .

Russian silt loam . . .

Pinole gravelly loam. . .

Finn gravelly loam .

Qrr loam .

Large gravelly loam .

Russian silty clay loam .

Russian silt loam .

Largo silty clay. . . .

Largo loam .

Sanelloam .

Guidiville sandy loam .

Tehama loam .

Russian fine sandy loam . . .

Diablo day .

Finn loam.... .

Dublin clay .

Russian loam...,, .

Largo silt loam .

Mendocino loam .

Tehama gravelly loam .

Finn gravelly day loam. . . .

Pinole loam .

Diablo loam .

Orr clay loam .

Pinole sandy loam .

Mendocino gravelly loam. . .

Mgm.
0. 10

.02

1.42

2.94

•30

4-79
2*47
S*5°
2.08
3*63
6. 10
.48
3*59
•03
•8s
2*39
1.42
3*52
6*45
7. 00
2.69
1. 00
1.24
3*oo
3*3°
5*io
2. 72
5*46

Mgm.

156.80

29.40
91.00
77.00
70.00
75* 60

203.00

147*00

140.00

152.60

163. 80
58.80
42.00
49.00

53*20

63.00
61.60
147.00
120.40
135* 80

61.60
67. 20

75.60
54* 60

85.40
112.00

85.40
95* 20

P.ct.

Tr.

_3+

Tr.

Mgm.
— 0.40

— .28
2.42
3*64

.44
15*79
4*37
7* 50
5* 58
6*93
12.00
.28

8.39

— . 12
8*35

13.89
1.32
5*82
10. 95
10.80
3*99
1. 20
.28

2.40
2.90

13* 70
5. 62
15*66

\Mgm.

196. 80
69.40

131*00
117.00
no. 60

115.60
243*00
187.00
180. 00

192.60

203.80
98. 80
82.00
89.00

93*20

103.00

101.60
187.00

160. 40
I75*8o
ioz.60
107. 20

115.60
94* 60

125.40
152.00
125.40
135. 20

P.ct.

'1+

Tr.

^3+
1+
4+
3+
3+
^5 +
Tr.
10+

8+
13+
1 +
3 +
6+
6+
3+
1+
Tr.
2+
2+
9+
4+
1 +

Mgm.
0.60
. 12
1.82

— .06
1.90
6. 29

18. 97

28. 50
29. 58

20.93

23.60

— .02

— ,41
.98

8-35

11.89

— .08
23.92

— .05
4*30

— .01
6.60
1.64

. 02

39* 70

— .08
26.66

Mgm.

288. 80

161.40
223.00
209. 00
202. 00

207. 60
335-oo
279.00
272.00

284.60

295. 80

190.80
174. 00
181.00
185. 20
i95*oo
193* 60
279.00

252. 40

267. 80
193*60
199. 20

207. 60

186. 60

217.40
244.00
217.40
227. 20

P.ct.

Tr.

10+

Tr.

^3+

Tr.

16+

11+

Mgm.
0. 50
•52
1. 82
3*54
. 20
2. 19
8.77
12.00
8. 58
9*53
13*60
■07

14. 79

*36
8-3S
21.49
■ 72
10.32
5- 15
4.00
.01

*32

1.04
.68
2.58
19. 70
9. 02
11.86

Mgm.

203.80
76.40

138. 00
124. 00
117.00

122. 60
250.00
194.00
187. 00

199.60

210. 80

105. 80
89-00
96.00

100. 20

no. 00

108.60
194. 00

167. 40
182. 8b

108.60
114. 20

122. 60

101. 60

132.40
159-00
132. 40

142. 20

P. ct.
Tr.
Tr.
1+
2+
Tr.
1+
3+
6+
4+
4+
6+

16+

Tr.

x4+

Tr.

5+
3 +
2+

Tr.

12+

With reference to the absolute amounts of nitrate produced from soil
nitrogen, it will be seen that the Ukiah soils are distinctly superior to
the Bay area soils. Thus, 6 soils in the Ukiah area produce more than
5 mgm. of nitrate nitrogen, as against only 1 in even a larger total number
of soils in the Bay series. Moreover, 1 of the 6 soils under discussion
reaches a nitrate production equivalent to 7 mgm. Two soils in the

Oct. 9, 1916

Nitrifying Powers of Humid and Arid Soils

Ukiah series produce almost no nitrates, but no soil loses nitrates during
the period of incubation. The average nitrate production is 2.84 mgm.
in the Ukiah series and 2.09 mgm, in the Bay area series. The data for
the percentage transformation of nitrogen into nitrates in the Ukiah
series are also correspondingly larger than those of the Bay area, as is
the case with the absolute data.. Thus, an 8 per cent maximum trans¬
formation is attained in the Ukiah series, as against one of 6 per cent in
the Bay area series, and 5 soils besides transform more than 5 per cent
of the nitrogen present into nitrate, as against 4 such in the Bay area.

Table VI, Group II, helps to emphasize again the several points made
in the foregoing discussion with respect to the behavior of sulphate of
ammonia in the Bay area soils. Of the total number of soils in the
Ukiah area, 8, or about 28 per cent, produce less nitrate from sulphate
of ammonia plus soil nitrogen than from the latter alone. The corre¬
sponding figure for the Bay area is 20 per cent. From these facts, it
appears that parallelism between degree of aridity of climate and the
nitrifying power of soils for different forms of nitrogen is more firmly
supported than ever. Thus, the Ukiah area soils as a class having more
nitrogen and greater internal surface act more nearly like the humid soils
than any of the other California series here studied, even though they
are still far removed from the humid soils in that direction. Just as
sulphate oi ammonia stimulates nitrification in the Bay area soils, so it
it does in commensurate degree in the Ukiah area soils. As the average
absolute nitrate production is higher from soil nitrogen alone in the
Ukiah as against the Bay area group of soils so it is with respect to the
sulphate-of -ammonia series of the two soil groups. The maximum trans¬
formation of nitrogen into nitrate is also a little higher in the sulphate-
of -ammonia series with the Ukiah soils, it being 13 percent in the case of
two soils, as against 11 per cent, which was the corresponding figure in
the Bay area soils. On the other hand, there are about equal numbers
of soils in the two areas (4 or 5) which transform more than 9 per cent
of the total nitrogen present in this series into nitrate. The other soils
are all considerably more feeble in nitrification. The average nitro¬
gen transformation on the relative basis is only slightly greater in the
Ukiah than in the Bay area soils with sulphate of ammonia as the nitri-
fiable material.

Only 35 per cent of the soils in the Ukiah area produce less nitrate
from combined soil and dried-blood nitrogen than from the former alone
(see Table VI, Group III). This is in sharp contrast to the Bay area
soils, for which the corresponding figure is 50 per cept, and with the
Pasadena area, for which the corresponding value is 63 per cent. All of
these, moreover, are in sharp contrast to the corresponding value for the
foreign soils, which is 20 per cent. It will be noted next that very large
nitrate productions are accomplished in the dried-blood series. Such
quantities are in all cases much in excess of those produced by the same

Journal of Agricultural Research

Vol. VII, No. a

soils in the sulphate-of -ammonia or the soil-nitrogen series. Thus, 8
soils in a total of 28 produce more than 20 mgm. of nitrate nitrogen
under the conditions of the experiment. Allowing for one or two excep¬
tional soils, the break between the high nitrate-producing soils and the
rest in the blood series is very abrupt. As against the 8 soils just referred
to in the Ukiah series, there are but 3 in the Bay area series of a corre¬
sponding class. In general, nitrification of dried-blood nitrogen proceeds
very much better in the Ukiah series than in the Bay area soils. In
fact, the discrepancies in nitrifying powers between the two soil groups
are better exemplified with dried blood than with any other form of
nitrogen. With one exception, all the soils in the dried-blood series
which produce more than 20 mgm. of nitrates contain considerably more
than 0.1 per cent of nitrogen, and are all possessed of large internal sur¬
face. The one exception mentioned is Mendocino gravelly loam, which
contains very nearly 0.1 per cent of nitrogen and has a large interna^
surface besides. Owing to the higher nitrogen content in many of the
soils of the Ukiah area than in corresponding soils of the Bay area, the
relative figures for the two areas using dried-blood nitrogen as the nitri-
fiable material do not differ as much as the absolute figures. Never¬
theless the relative values are again distinctly in favor of the Ukiah
area soils.

A study of Table VI, Group IV, brings us to a consideration of cotton¬
seed-meal nitrogen in its relations to the soils of the Ukiah area. As a
result of such consideration we find that 40 per cent of the soils concerned
produce less nitrate from cottonseed meal plus soil nitrogen than from
the latter alone in every case. This is a behavior with respect to cot¬
tonseed-meal nitrogen very similar to that evinced by the foreign soils,
the corresponding percentage for which was 37. The latter is obtained
even with a few arid and semiarid or otherwise peculiar soils included, as
previously pointed out. On the other hand, the corresponding figure for
the Bay area soils is 26 per cent, and it is much lower for the other areas.
In other words, it would seem that the Ukiah soils not only approach very
closely in their behavior to cottonseed meal that of humid soils, but also
that the curve in that respect shows a gradual decline with an increase in
the aridity of the climate concerned. The maximum absolute produc¬
tion of nitrate is 29.49 nigm. as seen in Table VI, Group IV, accomplished
by the Diablo clay. This record is surpassed in two cases in the Bay area
series, and nearly equaled in two others.

It may be well to summarize briefly from one or two points of view the
results which were obtained in the experiments with the California soils.
In a similar manner, therefore, to that employed in Table II for the
foreign soils, Table VII gives for California the percentage of soils in
each area which transformed more than 10 per cent of the total nitrogen
in a given culture into nitrate, giving every form of nitrogen separately.

Oct. 9, 1916

Nitrifying Powers of Humid and Arid Soils

TablB VII. — Percentage of California soils transforming more than 10 per cent of
nitrogen present in culture into nitrate

Form of nitrogen.

Soil area.

Riverside.

Pasadena.

Bay.

Ukiah.

Soil nitrogen alone . ,. .

Soil nitrogen plus ammonium sulphate .

Soil nitrogen plus dried blood .

Soil nitrogen plus cottonseed meal .

It appears, therefore, that in the Riverside, Pasadena, and Bay areas
the cottonseed-meal nitrogen gives the best results in the largest number
of soils. In the Ukiah area cottonseed meal takes second place and
divides honors with sulphate of ammonia. Sulphate of ammonia takes
second place in the Riverside, Pasadena, and Ukiah area soils, but third
in the Bay area soils. The soil nitrogen does, however, contend with it
for second place in the Riverside soils. Dried blood is last in both the
Pasadena and Riverside soils, but first in the Ukiah soils and second in the
Bay area soils. Soil nitrogen is either second or third in the Riverside
soils, is third in the Pasadena soils, and last in the other two areas.

Table VIII gives relative values on another basis than that employed
in Table VII. This latter criterion consists in computing the percentages
of soils which produce with every form of nitrogen more than 1 5 mgm.
of nitrate in the foreign soils and is based on the amounts of total nitro¬
gen present in corresponding quantities in the other soil areas. Thus,
the figure is 11 instead of 15 for the Ukiah area, 10 for the Bay and
Pasadena areas, and 5 for the Riverside area.

TablU VIII. — Percentages of soils producing more than 15 mgm. of nitrate in foreign

soils with every form of nitrogen

Form of nitrogen.

Soil area.

Riverside.

Pasadena.

Bay.

Ukiah.

Foreign.

Soil nitrogen alone .

Soil nitrogen plus ammo¬

nium sulphate .

Soil nitrogen plus dried •

blood . .

Soil nitrogen plus cottonseed

meal .

Similar relations appear to hold in this manner of computation as in
that previously employed. Some minor differences, however, are appar¬
ent. Ammonium sulphate very definitely takes second place in all the
California soils except the Bay area. Cottonseed meal is still first in
the Riverside and the Pasadena soils, but blood is first in the Ukiah and
Bay area soils. The soil nitrogen is fourth in all the California soil areas.

Journal of Agricultural Research

Vol. VII, No. a

In addition to these observations, it may be remarked that in the last
column, in the foreign soils, dried blood stands first, as it does in the Bay
and Ukiah areas, and soil nitrogen stands third. Sulphate of ammonia
in the foreign soils again takes last place instead of a close second, as in
the Riverside and Pasadena areas, afid cottonseed meal stands second
instead of first.

COMPARISON OF FOREIGN AND CALIFORNIA SOILS

It has been shown that 52 per cent of the foreign soils, which include
several arid or semiarid soils, produce more than 15 mgm. of nitrate nitro¬
gen out of the total soil nitrogen present. Neither the Bay nor the Ukiah
areas includes any soil of such nitrifying activity. However, it was not
expected that they would produce the same number of milligrams
of nitrate nitrogen, but merely an amount having an approximately
similar ratio to the total nitrogen as that in the foreign soils. While
the Riverside and the Pasadena areas, with 24 per cent and 18
per cent, respectively, of soils with an equivalent nitrifying power to
that of the foreign soils mentioned, are much more active than those of
the other two arid-soil areas just referred to, their records are still far
behind those of the foreign soils. These comparative data were arranged
as noted on the basis of equivalent quantities of total nitrogen, a basis
employed because of the claim which has been made that the quantity
of nitrogen rendered available in a soil is always a certain constant pro¬
portion of the total nitrogen present in soils. If the latter theory is
tenable, arid soils are certainly very much more feeble in the nitrification
of soil nitrogen than humid soils. But if the theory above mentioned
is incorrect, the data are all the more emphatic as to the considerable
disparity (in favor of the humid soils) between the nitrifying power for
soil nitrogen of soils of the arid and humid regions.

On the assumption that the stimulating or depressing effect of a certain
nitrogenous material on the soil's nitrifying power under laboratory con¬
ditions is a justifidble criterion, the percentage of soils in every group
which produced less or more nitrate from soil plus fertilizer nitrogen than
from soil nitrogen alone have been noted. Table IX has been arranged
on the basis of all those calculations.

Table IX. — Percentage of soils in every area which produced less nitrate with fertilizer
plus soil nitrogen than from soil nitrogen alone

Source of nitrogen.

Soil.

Sulphate of
ammonia
and soil.

Dried blood
and soil.

Cottonseed
meal and
soil.

Foreign .

Ukiah . . .

Bay . .

Pasadena . .

Riverside .

Oct. 9, 1916

Nitrifying Powers of Humid and Arid Soils

It seems clear from Table IX that sulphate of ammonia induces under
laboratory conditions a larger yield of nitrate in the large majority of the
soils of arid regions than could be produced by those soils with their own
nitrogen. The range in the percentage of such soils is not large and
varies between 89 per cent in the Riverside area and 72 per cent in the
Ukiah area. On the other hand, it is clear also that in the large majority
of soils in the foreign group ammonium sulphate has the opposite effect —
that is, to depress the nitrifying power of a soil for its own nitrogen.
The effects of sulphate of ammonia are thus almost exactly reversed in a
comparison of the Riverside and the foreign soils, for example. In the
comparison made by Sackett (10) of the nitrifying power of certain
Colorado soils (using ammonium sulphate) with certain others from
several other States, it will be noted that the foreign soils, with the excep¬
tion of the California soils, acted as a class like those studied by the
writers. Likewise, the Colorado soils are more like the arid soils which
were described previously. Inotherwords, the Colorado soils were superior
in nitrifying power for sulphate-of-ammonia nitrogen to soils from other
States, even not excepting the California soils. It must be remembered,
however, that anything more than a general analogy or comparison
between the writers1 results and Sackett’s is not permissible for the fol¬
lowing reasons: Sackett used 100 mgm. of ammonium-sulphate nitrogen,
while the writers used 40 mgm. in the soil cultures. He inoculated the
cultures with a fresh soil suspension in the case of every soil studied,
which was equivalent to 5 gm. of soil, while the writers merely employed
the soil flora existing in the air-dried soil. Sackett used a six weeks’
period of incubation, whereas the writers employed only a four weeks’
period. In several other minor respects the conditions of Sackett’s ex¬
periments were different, among which may be particularly mentioned
the method of analysis employed for nitrate determinations.

In the case of dried blood, conditions seem to be almost the reverse of
those with ammonium sulphate. The soils of the foreign group which
produce less nitrate from dried blood than from their own nitrogen alone
are decidedly in the minority, but the reverse is true of the arid soils of
California as they become more and more humid in character. Dried
blood therefore induces a loss in nitrate-producing power in from 35 to
63 per cent of the soils of the arid region. On the other hand, the same
substance affects only 20 per cent of the total number of foreign soils in
that manner. Hence, there is a stimulation to nitrification in some soils
of the arid region induced by the presence of dried blood in them. The
opposite is true with humid soils.

With respect to the stimulating or retarding action on the nitrifying
powers of the soils studied, cottonseed meal acts almost exactly like
ammonium sulphate in the Pasadena and Riverside areas. The differ¬
ence between the two materials is, however, much greater in the other two
arid -soil groups; but still in both cases the percentage of soils in which
apparent losses in nitrifying power are induced is below that of the foreign

Journal of Agricultural Research

Vol. VII, No. a

group. The latter is none the less almost half the magnitude of the
corresponding figure for sulphate of ammonia with the foreign soils.

From all the foregoing considerations it appears evident that different
forms of nitrogen exercise widely different effects on different groups of
soils. Thus, for example, ammonium sulphate seems to enhance the
nitrifying powers of arid soils and to depress those of humid soils. Dried
blood seems, in general, to operate in a reverse manner, while cottonseed
meal seems to act more like the ammonium sulphate. It will be remem¬
bered that conclusions similar to those reached for arid soils were drawn
by Lipman and Burgess (7) with respect to another group of arid soils with
which they worked. Moreover, Sackett, in the experiments cited, noted
that the relationship of the Colorado soils to sulphate of ammonia and
to dried blood was the reverse of the relation of the other soils which he
studied to those substances. The position of the Colorado soil was
similar to that occupied by the California soils with respect to the foreign
soils. The apparently mystifying feature of the comparison lies, how¬
ever, in the fact that the California soils tested were in Sackett's foreign
group (10) and, therefore, in his hands gave different results from those
obtained by us. This may perhaps be accounted for by the fact that
the foreign soils are considered as a whole and are not separated from
the California soils, which in reality give much better results than, for
example, the eastern soils with sulphate of ammonia.

Table X has been arranged to bring together some of the figures above
discussed with respect to the relative powers of soils to nitrify different
forms of nitrogen. The first group shows the percentage of soils in each
of the areas studied which transformed more than 10 per cent of the
total nitrogen present into nitrate in the case of every form of nitrogen
employed. The second group shows the percentage of soils in every
area which produced a quantity of nitrate equivalent to 1 5 mgm. in the
case of the foreign soils. The equivalent amount for soils other than
those of the foreign group was determined by using the same ratio of*
nitrate to total nitrogen which is employed in the last-named groups of
soils, the average total nitrogen content of all soils in a given area being
used as a basis.

Table X. — Transformation of nitrogen in various soil areas

Form of nitrogen.

Percentage of soils which
, transformed 10 per cent
or more of nitrogen into
nitrate.

Percentage of soils produc¬
ing 15 mgm. of nitrate ni¬
trogen in the foreign area
and an equivalent quan¬
tity based on nitrogen
present in other areas.

Soil nitrogen only .

Soil nitrogen and ammonium sulphate .

Soil nitrogen and dried blood .

Soil nitrogen and cottonseed meal. . . .

oct. 9, 1916 Nitrifying Powers of Humid and Arid Soils

By the different arrangement of the data in Tables VIII to X a reversal
of indications in minor ways has perhaps been brought about. But,
in general, certain differences of a marked character in the nitrifying
powers of humid and arid soils are obvious. Thus, by whatever method
compared, the soil nitrogen of humid soils seems to become nitrified more
readily than that of arid soils. Likewise, the opposite is true of sulphate-
of-ammonia and cottonseed-meal nitrogen and their effects on the nitri¬
fication of soil nitrogen, and, as pointed out, the difference may amount
to veritable inverse relationships. On the other hand, the opposite of
the effects noted for the forms of nitrogen just discussed is true in gen¬
eral for dried-blood nitrogen. The most marked differences are, of course,
evident between the foreign soils on the one hand and the Pasadena and
Riverside soils on the other, because the latter are more distinctly arid
in character than the Ukiah and Bay soils.

REVIEW OE RESULTS

Since the results obtained by the writers are very striking and offer the
first direct evidence, so far as they are aware, of the differences between
the nitrifying powers of humid and arid soils, it is essential that their find- .
ings be viewed critically. The first question which arises is that of the
representative nature of the samples of soil employed with respect to the
different climatic regions. The four soil areas chosen to represent arid
soils in California may be taken as representative because they exemplify
as nearly as possible interior valley and coast conditions in both southern
and northern California. Since more than twice as much rainfall is
normally received by the Bay and Ukiah areas as that received by the
Riverside and Pasadena areas, these soils, all of which are none the less
arid, should be considered as representative of average conditions in the
State. If, however, it is desired to apply the term “arid” only to regions
receiving less than 20 inches of rain a year, it is necessary to consider only
the Riverside and Pasadena soils. So far as soil nitrogen and nitrifica¬
tion are concerned, the Riverside area will fairly represent a large, if not
the largest, part of the San Joaquin Valley, a small part of the Sacra¬
mento Valley, the Coachella and Imperial Valleys, and nearly all of the
Mojave and the Colorado Desert regions besides. The Pasadena and
the Bay areas may be taken as nearly representative of the coast valley
conditions from San Francisco to San Diego, the first being only in
behavior more typical of the southern and the second of the northern
valleys. The Ukiah area will partly represent a large portion of the
northern half of Sacramento Valley and much of the coast region above
San Francisco. In general, therefore, the four soil areas are fairly
representative of California conditions, and in particular may serve, as
above indicated, to represent the typically arid conditions of the State
if only a certain area or areas be chosen.

Journal of Agricultural Research

Vol. VII, No. a

In regard to the humid soils, the situation is by no means as satisfac¬
tory. As already indicated, the soil samples from the other States were
sent by Experiment Station officers and may therefore, in many instances
at least, have been chosen from exceptionally well-managed fields. On
the other hand, it would not be fair so assume that the land of every
State Experiment Station is the best to be had in the State. It is
unlikely that the samples here studied would represent anything more
than average conditions of the Eastern States. It must be added, of
course, that when the soils of individual States of the humid group are
considered, most of them may be found either very deficient or very excel¬
lent in respect to nitrifying power. Such of the Eastern States, however,
as contain soils throughout of a low nitrifying power are decidedly in the
minority. It is therefore gratifying to be able to point to the relatively
low nitrifying power of the Sassafras loam from New Jersey, as given in
Table I, and to show in Table XI the nitrifying powers, similarly deter¬
mined, of a number of other New Jersey soil types for comparison.
It was fortunate that these samples of soil were made available from
an exhibit sent to the Panama Pacific Exposition by the New Jersey
Experiment Station.

Table XI. — Comparative nitrifying power of soil types from New Jersey

Name of type.

Nitrate pro¬
duced. ‘

Total soil ni¬
trogen present.

Soil nitrogen
nitrified.

Mgm.

Per cent .

Hoosic gravelly loam .

4. 30

149. 80

2 +

Norfolk sand .

. IO

1 1?. 40

Colbington sandy loam .

3- 75

103. 6O

Dover loam. . . .

5- 30

12 1. 80

Dutchess loam . . .

• 30

8l. 20

Sassafras loam .

Ov'

2. 40

13 1. 60

Penn loam .

4. IO

99. 40

Light sandy loam .

2. 50

96. 60

2 +

Do . .

3- 40

74.20

Since the samples described in Table XI were kept in a dry condition
in sacks for considerably over a year, it is possible that the soils may
have lost to a relatively slight degree their powers to nitrify their own
supply of nitrogen, and the figures obtained might be regarded as a little
low on that account. This effect on the New Jersey soils of drying
could not have been very great, as the evidence of the other soils would
show. It is therefore seen that of nine types of New Jersey soils studied,
not one has the power to change much more than 4 per cent of the total
nitrogen present into nitrate, and the samples described in Table I
transformed only a little more than 6 per cent of the soil nitrogen into
nitrates. It is interesting to note that the last-named sample, while
classified as a Sassafras loam, is very different in total nitrogen content

Oct. 9, 1916

Nitrifying Powers of Humid and Arid Soils

and in nitrifying power from the sample of the same name described
in Table XI. This question of variation in any one type is now being
investigated.

In general, therefore, it appears possible that some States may con¬
tain very few good nitrifying soils, others very few poor nitrifying soils,
and still others contain a fair proportion of both. If this possibility is
allowed, then the samples described in this paper as the foreign soils
must approach closely the average conditions of humid soils, despite
the criticisms which are above suggested.

The next question is that of the influence of the seasonal variation
in the nitrifying activity of soils on the validity of nitrification data, and
particularly of those above presented. As pointed out in the intro¬
ductory part of this paper, there are now in the writers' possession many
data on the monthly variation in the nitrifying powers of several dif¬
ferent soil types. These data, which cover a period of 1^ years of
monthly tests, indicate even more strikingly than former results in this
regard the great variability to which such determinations are subject.
Hence, the low nitrifying powers of the Bay and Ukiah area soils which
were gathered from November to December, 1914, may be regarded as
due to the depressing effect of the conditions of the late fall, particularly
as regards a lack of moisture. It is also realized that the much greater
relative activity of the soils of the Pasadena and Riverside areas may
in part be accounted for by the collection of the samples in June and
July, when conditions were more favorable to nitrification. The humid
soils were collected in various parts of the year, but the bulk of them
arrived at Berkeley between September 15, 1914, and January 1, 1915,
and soon thereafter were tested. If anything, therefore, the seasonal
effects should have caused the samples from the humid soils to be some¬
what depressed in nitrifying efficiency, but they show themselves su¬
perior as a class to the arid soils in the nitrification of soil nitrogen.
In view of this fact and of the opportunity offered for the comparison
of a variety of seasonal effects in both arid and humid soils, great signifi¬
cance can not be attached to the influence of the seasonal variations,
which are characteristic of nitrification determinations, on the validity
of the results.

The quantity of dried blood employed in the cultures in which that
material was tested will next be considered. It has recently been
pointed out (5) for dried blood that the nitrification of small quantities
may proceed normally in certain soils, as has also previously been shown
to be true for calcium cyanamid and for goat manure (7), whereas large
quantities would not only permit no nitrification but would actually
induce losses of nitrate from the soil. In comparative studies like those
described previously the absolute values obtained are not significant
except as side issues, provided all classes of soils tested are treated alike.

55856°— 16 - 3

Journal of Agricultural Research

Vol. VII, No. 2

Such like treatment has, of course, been accorded all the soils. It
would therefore seem that the figures are illuminating if not exactly for
the use of dried blood in general, at least for i per cent of dried blood as
the nitrifiable material. Of course, the possibility still remains that
with 0.05 per cent or 0.1 per cent of blood, instead of the 1 per cent
employed in the cultures, the results in the blood series may have favored
the arid and not the humid soils, as they did in these experiments. That,
however, is quite unlikely, since it is unreasonable to suppose that smaller
quantities of dried blood would have been nitrified with less vigor in the
humid soils than the large quantity employed. Except, therefore, as
it might have changed absolute values, the procedure in the use of dried
blood appears to be no factor in the comparison here made between the
powers of humid and of arid soils to nitrify dried-blood nitrogen.

The depression of nitrification in most arid soils by 1 per cent of blood
in nitrification experiments was briefly explained by the senior author
(6, 7) on the following hypothesis: Ammonification of dried blood
proceeds very rapidly. Ammonia is poisonous to the nitrifying organism.
If a soil has a large internal surface for adsorption, as, for example, in
the presence of large quantities of decaying organic matter, in organic
colloids, or a similar material, the ammonia produced by the process of
ammonification is quickly adsorbed and removed from harmful action
on the nitrifying bacteria. If the contrary is true of the internal surface
of soils, such as would obtain in sands poor in humus or in closely packed
soils of finer texture, the ammonia given off in the ammonification of
dried blood would be toxic to the nitrifying bacteria. This would take
place by a direct toxic effect and also by the depressing effect on nitrifi¬
cation of large quantities of soluble organic matter, in the soil solution
introduced through the solvent effects of ammonia on the dried blood.
Very few arid soils have been found incapable of nitrifying 1 per cent of
blood which did not also yield a very dark soil solution, indicating the
active dissolution of the organic matter present. Very few such soils,
moreover, have been noted which do not give off a strong odor of ammo¬
nia during the period of incubation. That both dissolved organic matter
in the medium and free ammonia deter or inhibit the process of nitrifi¬
cation by pure cultures has been proved for solution cultures by Wino¬
gradsky’s experiments (14) . That the same is true for soils in the presence
of mixed cultures has not been known, however, until now. Since the
hypothesis was first formulated, the writers have carried out experiments,
not possible of description here, which clearly point to the highly toxic
nature of relatively small quantities of ammonia to the nitrification of
dried blood in soils which otherwise transform the nitrogen of that
material into nitrates without difficulty. It is further gratifying to
note that Hutchinson (4) has arrived at a similar result to that which
the writers have obtained on the toxicity of ammonia to the process of
nitrification.

Oct. 9, 1916

Nitrifying Powers of Humid mid Arid Soils

If the hypothesis briefly explained is correct, the difference in the
behavior of the humid and arid soils toward the nitrification of dried-
blood nitrogen is at once explicable, since the amounts of organic and
inorganic colloids are usually much larger in the humid than in the arid
soils and would act toward the ammonia produced from dried blood as
above explained. Even if it should prove desirable to use smaller
quantities of dried blood in cultures to determine its availability in
arid soils, the mejthod used heretofore may serve as an excellent means
for the comparison of groups of soils and as an index to the soil's internal
surface and its status with reference to colloid content.

In connection with this discussion it is cogent to refer to the results
obtained by Sackett (10), showing that Colorado soils are superior in
nitrifying power to 22 soils from localities outside of that State. These
results would seem to imply that for some reason Colorado soils are
in general superior in nitrifying efficiency to other soils. A comparison
of Sackett's data with the writers', however, does not bear out such an
implication when the nitrification of the soil's own nitrogen and not
that of fertilizer nitrogen is considered. Thus, for example, in the
Pasadena area out of 33 soils about 45 per cent produced 7 mgm. or more
of nitrate in 100 gm. of soil in a month's incubation period. Of the 23
Colorado soils tested by Sackett only 21 per cent of such soils were found;
yet the average total nitrogen content of the Pasadena soils is prob¬
ably below that of the Colorado soils and Sackett's incubation period
was six weeks and the writers' only four weeks. When a relative instead
of an absolute basis of comparison is used, similar results are obtained
in other California areas. But the humid soils outstrip the Colorado
soils even farther than the Pasadena soils and show clearly a very supe¬
rior nitrifying power as a class for the soil's own nitrogen to that
possessed by the Colorado soils.

So far as the nitrification of sulphate-of -ammonia nitrogen is con¬
cerned, the Colorado soils do seem to be superior to other soils if the soils
chosen by Sackett are fairly representative of Colorado soils. They
are, however, to be considered as a class only slightly superior to the
Riverside and Pasadena soils when it is considered that we employed a
much shorter period of incubation, much less sulphate of ammonia, and
did not inoculate our soils with fresh infusions. In fact, it appears now
that an equal comparison of the Riverside or Pasadena soils with the
representative Colorado soils would probably show them to be very
similar in respect to nitrifying powers for sulphate of ammonia. While
it would be difficult to establish any fixed criterion, it would seem, how¬
ever, that the soil's own nitrogen would for all ordinary purposes best
subserve the purposes of soil fertility. If such a criterion is adopted,
then the Colorado soils as well as the arid soils of California can not only
be said not to be superior in nitrifying power to the humid soils but it
is barely possible that they are appreciably inferior in that respect.

Journal of Agricultural Research

Vo!. VII, No. a

Attention must be called to the fact that the classification for purposes
of discussion of the Pasadena area as a coast valley is not done with any
idea of so classifying it permanently. It is merely done to emphasize
its closeness to the sea and its greater rainfall and attendant factors
which affect many of the soils in that area so as to make them more
characteristic of coast than of interior valley conditions under which
the Pasadena area would normally be classed..

Several other minor points of interest with respect to the data which
have been discussed deserve consideration here. Most of the soils above
studied from the South Atlantic States are deficient in nitrogen, and in
that respect resemble, for example, the truly arid soils of the Riverside
area. Likewise, with respect to the nitrification of nitrogen in dried
blood and cottonseed meal, but particularly the former, the South
Atlantic soils behave more like the truly arid soils. Their behavior
could therefore be accounted for on the hypothesis explaining the
behavior of the truly arid soils toward dried-blood nitrogen, taking the
low nitrogen content of the soils as an index of the organic matter present
and, hence, indirectly of the total internal surface. That, however,
aside from the total nitrogen considerations and their indications that
climate exerts additional effects on the soil's nitrifying power, is exem¬
plified by the fact . that the South Atlantic soils, while behaving
toward dried blood and partly toward cottonseed meal like arid soils,
stand, with respect to soil nitrogen and sulphate of ammonia, in the
position of the truly humid soils. The latter is more easily explicable
on the basis of soil acidity with respect to sulphate of ammonia and that
of superior moisture conditions with respect to the soil nitrogen. It
will be noted that two soils which did not nitrify their own nitrogen well
are the Kentucky and Louisiana soils. The first was collected in De¬
cember and the second in March, which would probably account for the
relative inactivity of the nitrifying organisms. With the exception of
the last two soils, it will be noted in Table I, Group I, that the South
Atlantic soils are far superior in their nitrifying powers for soil nitrogen
to the soils of the Riverside area.

It will be noted that the total nitrogen content of soils in connection
especially with the nitrification of dried-blood nitrogen has several times
been emphasized. This is done to indicate, in a general way, the likeli¬
hood of the presence of certain quantities of organic colloids in the soil
on the assumption that the soil's nitrogen content bears a more or less
intimate relation to the quantity of organic matter present. The latter,
moreover, is considered as an indication in its turn of the amount of
internal surface contributed to the soil from that source in addition to
that present there by virtue of the surfaces of inorganic constituents,
including sand, silt, and clay.

Oct, 9, 1916

Nitrifying Powers of Humid and Arid Soils

SUMMARY

A study was made of the nitrifying powers, under incubator conditions,
of about 40 humid and about 150 arid soils. The soil was used as a
medium and the forms of nitrogen employed were soil nitrogen, sulphate
of ammonia plus soil nitrogen, dried blood plus soil nitrogen, and cot¬
tonseed meal plus soil nitrogen. The humid soils were obtained from
the different State and Territorial Experiment Stations, one type of
surface soil from each one being used in these experiments. The arid
soils employed represented the soil types of four typical soil-survey
areas in California. The results obtained appear to justify the follow-
lowing statements:

(1) The conclusion appears ineluctable that the nitrifying powers of
soils of the arid region are no more intense than those of the humid
region. This denies Hilgard's (3) teaching to the contrary and confirms
the statement of Stewart (11), which was based on more indirect and
less extensive evidence.

(2) While indications are not by any means positive now, it is possible
that the data of the writers justify the further conclusion that the
nitrifying powers of humid soils are greater than those of arid soils.
If such a conclusion could be drawn, it would have to be based merely
on the nitrification of soil nitrogen and dried-blood nitrogen. The
former being the natural source of supply of most of the available
nitrogen obtained by crops, it should really be regarded as the most
valuable basis for forming a judgment.

(3) Arid soils, on the other hand, nitrify the nitrogen of sulphate of
ammonia and cottonseed meal with much greater vigor than do the
humid soils. A reversal of efficiency is manifest between the two groups
of soils as regards sulphate of ammonia and cottonseed meal, on the one
hand, and dried blood and soil nitrogen, on the other.

(4) These results not only throw new light on the question of nitrifica¬
tion in soils of different climatic regions but also tend to confirm the
earlier findings of the writers and others on the important relation of
climate to soil properties (9).

(5) The foregoing considerations apply to the two groups in general
and not to parts of such groups in particular. For example, the Bay
and Ukiah soils do not nitrify soil nitrogen as efficiently as the more arid
Pasadena and Riverside soils. This may, however, be accounted for by
the seasons at which the collection of the soils was made and by the
physical condition in which the Bay and Ukiah soils were received.

Journal of Agricultural Research

Vol. VII, No. 2

LITERATURE CITED

(1) Headden, W. P.

1910. Fixation of nitrogen in some Colorado soils. Colo. Agr. Exp. Sta. Bui.

155. 48 p., 8 pi.

(2) -

1911. Fixation of nitrogen in some Colorado soils. Colo. Agr. Exp. Sta. Bui.

178. 96 p., 6 pi.

(3) Hiegard, E. W.

* 1906. Soils; Their Formation, Properties, Composition, and Relations to Climate
and Plant Growth in the Humid and Arid Regions. 593 p., illus. New
York, London.

(4) Hutchinson, C. M.

1912. Studies in bacteriological analysis of Indian soils. No. 1, 1910-1911. In

India Dept. Agr. Mem., Bact. Ser., v. 1, no. 1, p. 1-65, 6 pi., 2 fold, charts.

(5) KeelEy, W. P.

1916. Some suggestions on methods for the study of nitrification. In Science, n. s.,
v. 43, no. 1097, p. 30“33*

(6) Lipman, C. B.

1915. The nitrogen problem in arid soils. In Proc. Nat. Acad. Sci., v. 1, no. 9,
p. 477-480.

(7) - and Burgess, P. S.

1915. The determination of availability of nitrogenous fertilizers in various Cali¬

fornia soil types by their nitrifiability. Cal. Agr. Exp. Sta. Bui. 260, p.
107-12 7.

(8) - and Sharp, L. T.

1912 . Studies on the phenodisulphonic acid method for determining nitrates in soils.

In Univ. Cal. Pub. Agr. Sci., v. 1, no. 2, p. 21-37.

(9) - and Waynick, D. D.

1916. A detailed study of effects of climate on important properties of soils. In

Soil Science, v. 1, no. 1, p. 5-48, pi. 1-5. Literature cited, p. 48.

(10) Sackett, W. G.

1914. The nitrifying efficiency of certain Colorado soils. Colo. Agr. Exp. Sta. Bui.
193- 43 P-i3pl-
(n) Stewart, Robert.

1913. The intensity of nitrification in arid soils. In Proc. Amer. Soc. Agron., v. 4,

1912, p. 132-149. Also in Centbl. Bakt. [etc.], Abt. 2, Bd. 36, No. 19/25,
p. 477-490.

(12) - and Peterson,* William.

1914. The origin of the “nitre spots” in certain western soils. In Jour. Amer. Soc.

Agron., v. 6, no. 6, p. 241-248.

(13) -

1916. The origin of the “nitre spots ' * in certain western soils. In Science, n. s. , v.
43, no. 1097, p. 20-24.

(14) Winogradsky, M. S.

1890. Recherches sur les organismes de la nitrification. In Ann. Inst. Pasteur,
ann. 4, no. 4, p. 213-231.

immobility of iron in the plant

By P. L. GilE, Chemist , and J. O. CarrERO, Assistant Chemist , Porto Rico Agricultural

Experiment Station

INTRODUCTION

Work at the Porto Rico Experiment Station on the assimilation of
iron by certain plants, including rice ( Oryza sativa) , has afforded results
which seem to show that iron is relatively immobile in the plant after it
has once been transported to the leaves. In respect to mobility in the
plant iron would thus be similar to silicon and calcium and different
from nitrogen, phosphorus, potassium, and magnesium, which are
generally considered mobile.

These observations on the immobility of iron are chiefly concerned with
the nontransference of iron from leaf to leaf under conditions where the
plant was insufficiently supplied with iron. That the mobile mineral
elements and nitrogen are translocated from leaf to leaf under such con¬
ditions seems proved by Schimper (8)1 as well as by observations on the
growth of plants in media lacking one of these elements. The translo¬
cation of nitrogen, potassium, and phosphorus from old leaves and
stems to the fruiting parts has, of course, been well established by ash
analyses of plants during maturation and by direct experiments (6, p.
585-586). This, however, is not different in principle from the translo¬
cation of these elements from old to new leaves, as the constituents
must in this case also pass from the old leaf into the stem.

The various facts which seem to point to iron being relatively immo¬
bile in the plant are given below.

OBSERVATIONS ON RICE GROWN IN NUTRIENT SOLUTIONS LACKING

IRON

Rice grown in nutrient solutions without iron is quite different in
appearance from rice grown without nitrogen or phosphorus. In iron-
free solutions the leaves of the plants commence to die at the top rather
than at the base of the plant. The newer leaves that are formed are
almost pure white, very thin, and generally wither before the old leaves.
The phenomena in rice grown in absolutely iron-free solutions are not so
marked as in plants grown for a time with iron and then transferred to
an iron-free solution; since either the amount of iron in the seed does
not suffice for the needs of the first leaves or it is incompletely translo¬
cated, so that even the first two or three leaves formed are yellowish
green in color.

1 Reference is made by number to “literature cited,” p. 87 .

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.
fp

(83)

Vol. VII, No. 3
Oct. 9, 1916
B— 11

Journal of Agricultural Research

Vol. VII, No. 2

A lot of rice was grown for 13 days in a nutrient solution well supplied
with iron and then transferred for 13 days to a solution identical except
for the absence of iron. The leaves formed during growth in the com¬
plete nutrient solution were dark green, while the leaves that formed in
the 13 days after change to the iron-free solution were yellowish green
to creamy-white, the old leaves retaining their dark -green color during
this change. The chlorosis or lack of green in the newer leaves was
obviously associated with a lack of iron and due to a nontransference of
iron from the green lower leaves.

Similar phenomena in regard to the appearance of chlorosis were
observed by Molisch (4, p. 92) on growing Cucurbita pepo , Helianthus
annuus , Zea mays , etc., in iron-free solutions.

The appearance of rice grown in nitrogen or phosphorus-free solutions
was quite distinct from that of plants grown in iron-free solutions; the
leaves commenced to die from the base upwards, and new leaves were
continually formed at the top of the plant. Here there evidently was a
translocation of nitrogen or phosphorus from the old leaves to the new,
as growth continued after all material in the seed had been exhausted,
and new tissue could not have been formed without these elements.

OBSERVATIONS ON PLANTS AFFECTED WITH LIME-INDUCED

CHLOROSIS

Experiments at this station showed that the chlorosis of some plants
on strongly calcareous soils was due in part at least to lack of iron; at
all events appropriate treatment with iron salts cured the chlorosis.
Phenomena similar to that observed in rice grown in iron-free solutions
were observed with rice and pineapples (Ananas sativus) , grown in calca¬
reous soils. Rice when not immediately affected with chlorosis showed
the chlorosis in the new leaves, even though the old leaves were green.
The new leaves formed came out yellowish green to cream y-white and
withered completely, the plants dying from the top down, while the lower
leaves remained sound and green. Pineapples behaved similarly, in that
the chlorosis appeared first in the new leaves, although frequently the
lower leaves followed the new ones in becoming chlorotic, until eventually
the whole plant became chlorotic.

If iron were mobile in the plant after once being transported to the
leaves, evidently the phenomena would have been different. Iron would
have been translocated from the old to the new leaves, the scene of most
active growth, and the old leaves would have withered or become chlo¬
rotic first.

BRUSHING WITH IRON SALTS THE LEAVES OF PLANTS LACKING IRON

Rice plants grown in certain nutrient solul ions with a lack of available
iron developed chlorosis and were brushed with a 0.2 to 0.4 per cent
solution of ferrous sulphate. In the course of this treatment the tips of
leaves just emerging from the stalk were brushed. As these leaves grew

Oct. 9, 1916

Immobility of Iron in the Plant

out, the part that had been brushed was a normal green, while the lower,
unbrushed part of the leaf was strongly chlorotic and remained so until
treated with iron. This would hardly have been the case if the iron
were mobile in the leaf tissue. Of course, if a great excess of iron had
penetrated the epidermis, it might have been translocated to other parts
of the leaf.

The inefficiency of spraying the leaves with iron salts, as a means of
curing the chlorosis of pineapples or grapevines, is probably partially due
to the immobility of iron in the leaves.1 With pineapples this treatment
completely restored the green color to the leaves treated, but new leaves
formed after the treatment were chlorotic.

EVIDENCE FROM ASH ANALYSES

Ash analyses of old, young, and withered leaves generally support the
view that iron once conducted to the leaf is immobile, although they by
no means afford proof. It should be borne in mind, however, in judging
many of the old ash analyses that the determination of the relatively
small amounts of iron by precipitation as ferric phosphate is not particu¬
larly accurate. Also the accuracy of many results where iron was not the
chief element sought may well have been- affected by contamination.
The writers have found the colorimetric potassium-sulphocyanate
method, either in usual form or as modified by Stokes and Cain (9), more
accurate than the usual method.

Czapek in his compilation states that old leaves as a rule contain more
iron than young ones (1, p. 800). The work of Fliche and Grandeau on
the composition of leaves of different trees at various stages of growth
supports this (3, p. 487). Analyses made at the Porto Rico Experiment
Station of leaves from 1 -year-old rough-lemon trees {Citrus limonum)
show the lower or older leaves to be higher in iron than the young leaves.
The lower and upper leaves of plants from four different soils were
sampled. The percentages of iron in the dry substance are given in
Table I.

Tabi.£ I. — Percentage of iron in young and in old rough-lemon leaves

Soil No.

Age of leaves.

Iron (ReaOa) in dry
substance.

I . . .

Young . .

Per cent.

0. 017
. O44
. 021

*039
. 012
. 025
. OI4
. 023

I. . .

Old .

II .

Young .

II .

Old .

HI .

Young .

Ill .

Old. . . .

IV .

Young .

IV . .

Old .

1 It is significant that with certain kinds of chlorosis of grapevines, treatment of cut stems with the iron
salts (method of Rassiguier) has proved more efficacious than spraying the foliage.

Journal of Agricultural Research

Vol. VIIt No. 2

With respect to the accumulation in old leaves, iron is similar to
silicon and calcium, which also seem immobile in the plant when once
transported to the leaves. Young leaves are generally relatively higher
in the mobile elements than old leaves. This parallelism, however,
really affords little proof, as the relative amount of the different ash
constituents present in young and old leaves is probably governed more
by the function than the mobility of the elements. Also the accumula-
tion of iron in the old leaves of plants well supplied with iron may merely
show that a functioning leaf has a continual need of iron. As iron is in
some way associated with the formation of chlorophyll and as chloro¬
phyll is apparently undergoing a continual formation and destruction
under the. influence of light (2, p. 193; 5) and enzyms (10, p. 746), one
can conceive of a physiological necessity for an accumulation of iron in
old leaves.

The most that can be said concerning the evidence of ash analyses is
that the results are such as one would expect if iron were immobile
after being located in the leaf.

DISCUSSION OF RESULTS

The only references to the mobility of iron in the plant which the
authors have found in the literature are a discussion by Sachs (7) and
a statement by Pfeffer that “in a starved green plant, as well as in a
fungus, the iron and potassium may be removed from the older dying
organs and transferred to the younger growing parts so that the growth
may not immediately cease” (6, p. 417). No data or reference are given
in support of this statement. While the movement of potassium under
such conditions is generally recognized, the same can hardly be said
of iron.

Lack of information on the movement of iron is partially due to the
fact that iron has ordinarily been considered of so little interest in plant
nutrition as to be disregarded in ash analyses. Also, as already pointed
out, the usual method of determining iron in plant analyses is probably
not sufficiently accurate to show significant changes.

Sachs, in his interesting discussion of the chlorosis of various plants
grown under garden conditions (on a calcareous soil), points out the
apparent slowness with which iron moves in plants (7).

From the data presented it is not intended to assert that the non¬
translocation of iron from leaves is an absolutely general rule for all
plants, since the foregoing observations were based chiefly on rice and
pineapples.

Various observations on rice and pineapples grown with insufficient
iron seem to show that iron after once being transported to the leaves
is immobile.

Oct. 9, 1916

Immobility of Iron in the Plant

LITERATURE CITED

(1) Czapek, Friedrich.

1905. Biochemie der Pflanzen. Bd. 2. Jena.

(2) EudER-Chelpin, H. K. A. S. von.

1908. Grundlagen und Ergebnisse der Pflatizenchemie. ... T. 1. Braun¬
schweig.

(3) Fdiche, P., and Grandeau, L.

1876. Recherches chimiques sur la composition des feuilles. In Ann. Chim.
et Phys., s. 5, t. 8, p. 486-511.

(4) Mousch, Hans.

1892. Die Pflanze in ihren Beziehungen zum Eisen. . . . 119p.jC0l.pl. Jena.

(5) - . ' '

1902. Ueber voriibergehende Rothfarbung der Chlorophyllkomer in Laub-
blattem. In Ber. Deut. Bot. Gesell., Bd. 20, Heft 8, p. 442-448.

(6) Pfeffer, W. F. P.

1900. The Physiology of Plants. . . . Trans, by A. J. Ewart, ed. 2, v. 1.
Oxford.

(7) Sachs, Julius.

1888. Erfahrungen fiber die Behandlung chlorotischer Gartenpflanzen. In
Arb. Bot. Inst. Wfirzburg, Bd. 3, p. 433-458.

(8) SchimpER, A. F. W.

1890. Zur Frage der Assimilation der Mineralsalze durch die griine Pflanze.
In Flora, Jahrg. 73 (n. R. Jahrg. 48), p. 207-261.

(9) Stokes, H. N., and Cain, J. R.

1907. On the colorimetric determination of iron with special reference to chem¬
ical reagents. In U. S. Dept. Com. and Labor, Bur. Standards Bui.,
v. 3, no. i, p. 115-156, 5 fig.

(10) Woods, A. F.

1899. The destruction of chlorophyll by oxidizing enzymes. In Centbl. Bakt.
[etc.], Abt. 2, Bd. 5, No. 22, p. 745-754.

ADDITIONAL COPIES

OP THIS PUBLICATION MAY BE PROCURED FROM
THE SUPERINTENDENT OP DOCUMENTS
GOVERNMENT PRINTING OFFICE
WASHINGTON, D. C.

10 CENTS PER COPY
Subscription Price, 13.00 Per Year

JOURNAL OF AdlllLTlAI, BEAM

DEPARTMENT OF AGRICULTURE

Vol. VII Washington, D. C., October 16, 1916 No. 3

EFFECTS OF NICOTINE AS AN INSECTICIDE

. By N. E. McIndoo 1

Insect Physiologist , Deciduous Fruit Insect Investigations , Bureau of Entomology
INTRODUCTION AND EXPERIMENTAL METHODS

The pharmacological effect of nicotine (C10H14N2) on the higher animals
is well understood, but there is practically nothing known about the
pharmacological effects of nicotine as an insecticide. Owing to the
high cost of nicotine, it is desirable to have a substitute for this insecticide.
Before being able to discover, if possible, such a substitute, it is first
necessary to ascertain how nicotine affects insects.

In the investigation herein recorded two chief objects have been
kept in view: (1) To determine the physiological effects of nicotine as
an insecticide, and (2) to trace the nicotine into the insects after it
has been applied to them. A brief account of the pharmacological
effects of nicotine on other animals and the views pertaining to the
physical and chemical effects of nicotine on the cells are also given.

Owing to the small size of the insects utilized in the experiments the
usual method of procedure employed by pharmacologists could not be
used, because it was impossible to operate on living insects in order to
ascertain what tissue is vitally affected by nicotine. Consequently the
behavior of the insects treated with nicotine was compared with the
behavior of normal and untreated ones; and immediately after the
treated ones had died, they were fixed in a fluid containing a nicotine
precipitant. By this means the nicotine was precipitated wherever it
had gone into the insects; and after making microscopical sections from
these insects, it was not a difficult task to trace the precipitated nicotine.

Shafer (20, 21), 2 from the standpoint of a physiological chemist, carried
on investigations to determine how contact insecticides kill. He did not

1 The writer is grateful to the following persons: To Mr. A. F. Sievers, Chemical Biologist in Drug-Plant
and Poisonous-Plant Investigations, for extracting pure nicotine from a commercial nicotine material
and for verifying the percentage of nicotine in a sample of 40 per cent nicotine sulphate; to Dr. D. E.
Jackson, of the Department of Pharmacology, Washington University Medical School; and to Mr. O. D.
Swett, Assistant Professor of Chemistry in George Washington University, for reading and criticizing
the manuscript of this paper.

2 Reference is made to Literature cited, pp. 1 20121 .

* Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.,
fr

Vol. VII, No. 3
Oct. 16, 1916
K— 43

(89)

Journal of Agricultural Research

Vol. VII, No. 3

experiment much with nicotine and did not endeavor to ascertain what
tissue is vitally affected by any particular insecticide, but he seems to
infer that nicotine affects the cells chemically in the same way as do the
other contact insecticides. Even if nicotine has no therapeutic use, it
is classified as a poisonous drug; and for this reason the investigations in
which it is used give the best results when it is considered from the
standpoint of toxicology.

PHYSIOLOGICAL EFFECTS OF NICOTINE

At the outset it was decided to select one of the most specialized insects
and to feed it nicotine so that the results might be compared to those
previously obtained after administering nicotine to certain higher animals.
Although nicotine as an insecticide is rarely used as a stomach poison,
nevertheless the experimentation was begun with this phase of the work.
Since pharmacologists have determined that, as a rule, nicotine, regard¬
less of how it is administered, has practically the same general effects, it
seems logical that nicotine as an insecticide will also have practically the
same effect, regardless of how it is applied.

I. — NICOTINE AS A STOMACH POISON

Since the writer, during the past four years, has made a special study
of the behavior of the honeybee, and as the honeybee is one of the highest
forms of insects, it was first selected for making a special study of the
physiological effects of nicotine on this class of animals.

(a) BEES FED PURE NICOTINE

To avoid the complications which often arise when a drug composed
of more than one constituent is administered, pure nicotine was fed to
bees in the following manner: Honey and pure nicotine were thoroughly
mixed in the proportion of i part of nicotine to 100 parts of honey.
Ten c. c. of this mixture were poured into a small tin feeder covered with
parallel pieces of wire; then the feeder and contents were placed inside a
triangular observation case, previously described by the writer (17).
Fifty bees (guards) were next introduced into this case. On account of
the faint nicotine odor emitted from the mixture of honey and nicotine,
the bees did not eat the food readily. To be certain that the bees had
eaten some of this poisoned food before they died, the honey stomachs
of several dead bees were examined. It was found that each stomach
contained more or less honey, and this was certainly not eaten before the
bees were put into the case, because the honey stomachs of guards never
contain honey. Testing this supposedly poisoned honey for the presence
of nicotine by using alkaloidal reagents, such as silico tungstic acid and
phosphomolybdic acid without first attempting to isolate the alkaloid*
means nothing, for these reagents also precipitate honey and many other

Oct. 16, 1916

Effects of Nicotine as an Insecticide

organic substances. This test did not appear sufficiently significant to
warrant the expenditure of more time.

The 50 bees lived from 10 hours to 72 hours, with 33 hours as an
average, whereas 50 bees fed honey containing no nicotine lived 8 days,
on an average. To facilitate description, the behavior of the bees dying
of nicotine poisoning may be divided into three stages, and since nico¬
tine kills the higher animals by paralysis and since, as will be shown, it
kills insects similarly, the words “paralyze” and “paresize” may be
used from the outset. The word “paresis ” means partial motor paralysis,
while the word “paralysis” includes both motor and sensory paralysis.

First stage;. — Shortly after being poisoned, bees become more or
less inactive and are seldom seen eating. They “pay little or no atten¬
tion” to hive mates or to strange bees and never attack the latter. They
soon become stupid, and from then on their behavior is quite abnormal.
All their senses are perhaps benumbed, for they do not offer to attack
bees carrying foreign hive odors, and they are not very sensitive to
mechanical stimuli of any kind. A little later one or both, but usually
one, of the hind legs becomes partially paralyzed (paresized), and there¬
after they are of little use. Or the front legs may be stricken partially
or totally with motor paralysis at the same time, but occasionally the
middle legs may be similarly affected before either of the other two
pairs is stricken. The wings seem to be paresized before the legs are
affected, because a stupid bee removed from the case is able to walk
normally and can vibrate its wings, but can not lift itself from the table.
Whenever the motor paralysis has not extended further than to paralyze
partially the wings and to paralyze totally only one pair of legs, the bees
in almost every instance recover when removed from the case to fresh air
and when given pure honey. They eat the honey readily and soon
throw off their stupor, and the paresized wings and legs soon recover so
that after half an hour the bees are again able to fly.

Second stage. — Soon after one pair of legs is stricken, all three pairs
and the wings become paresized. During this stage bees act somewhat
like a man intoxicated with alcohol. They walk in a staggering manner,
drag the paresized legs, and frequently fall down, but never walk upright
in the normal manner. Sometimes all three legs on one side may be
affected totally by motor paralysis, while on the other side one or more
legs may not be stricken. In such a case as this, the bee lies flat on its
thorax and abdomen and turns in a circle. Often a bee falls over on its
side or on its back and can not get up. Sometimes the middle and hind
legs are totally stricken with motor paralysis, while the front legs are
apparently not affected. In this case the bee crawls along by dragging
its abdomen. A little later when all the legs and wings are affected
totally by motor paralysis, the bees are entirely helpless. If removed
from the case at this instant, a bee thus paralyzed is still able to extend
its tongue and to eat honey, but a few moments later the men turn

Journal of Agricultural Research

Vol. VII, No. 3

becomes paresized, and the tongue can no longer be extended. Subse¬
quently the mandibles and antennae are stricken.

Third stage. — At the beginning of this stage the bees are apparently
dead, except that an occasional twitching of a tarsus or a slight movement
of the end of an antenna or of the abdomen may be seen. Sometimes an
abdomen passes through a series of convulsions, and occasionally a small
amount of feces is voided; in one instance a small amount of liquid was
seen issuing from the mouth.

If the behavior of bees dying of nicotine poisoning is interpreted in
the same way as is interpreted the behavior of higher animals likewise
poisoned, it seems that nicotine as a stomach poison really kills bees
by motor paralysis, and that the paralysis travels along the ventral
nerve cord from the abdomen to the head, first affecting the abdominal
and thoracic ganglia, then the subesophageal ganglion, and last the
.brain. Considerable light is thrown on this point in the portion of this
paper dealing with the tracing of the nicotine from the time it is applied
to the time it reaches the nervous system. Having decided that nicotine
kills insects by paralysis, we shall now consider the effects of nicotine as
an insecticide when applied in practical work.

2. — NICOTINE SPRAY SOLUTIONS

The following results are not meant to test the efficiency of any of the
commercial nicotine spray materials, or of even various dilutions of them,
but merely to determine how nicotine affects insects when it is applied
as in practice.

(a) APHIDS DIPPED INTO SOLUTION OP PURE NICOTINE

Carolina poplar leaves bearing many aphids {Aphis populifoliae Davis)
were dipped into a solution of pure nicotine (i : ioo) . At once the aphids
began to exhibit an abnormal behavior and soon showed signs of dying.
Half an hour later all of them were dead.

( b ) APHIDS SPRAYED WITH SOLUTION OP PURE NICOTINE

Many more aphids on Carolina poplar leaves were sprayed heavily
with the above solution. Half an hour later nearly all the aphids were
dead, every one being dead 15 minutes still later. Before being sprayed,
these aphids were quiet and seldom moved from place to place on the
leaves. They usually stood on the first two pairs of legs, with the hind
pair of legs and abdomen high in the air and with the beaks stuck into
the leaves. Occasionally an aphid elevated its abdomen higher into the
air and simultaneously moved its body sidewise in a jerky manner.
The legs and antennae were moved little, and no liquid was seen issuing
from the cornicles or from the anal openings. Immediately after being
sprayedj these same aphids lay flat on the leaves, apparently dead.

Oct. 16, 1916

Effects of Nicotine as an Insecticide

They were covered more or less with white “wool,” which was less con¬
spicuous after being wetted by the spray. Five minutes later the aphids
stood up, began to move their legs, and most of them were comparatively
active for a few moments. They removed their beaks from the leaves,
moved about considerably by lifting the legs nervously, and their peculiar
jerky movements became more conspicuous. Later they were more
quiet and the legs became paresized, the hind legs being affected first,
the middle legs next, and the front legs and antennae last. At this stage
the hind legs generally are totally stricken with motor paralysis. When
paresized, many of the aphids fell from the leaves, and for a few moments
they seemed to be recovering from their stupor; but they finally died.
However, all of those that fell from the leaves lived several moments
longer than those that remained on the leaves. When almost inactive,
the aphids fell either over on their sides or on their backs and were com¬
pletely helpless. The last signs of life were twitchings of the tarsi and
slight movements of the antennae. Before death, the bodies of the aphids
appeared perfectly dry. When dead, the legs are usually folded and are
stiff. During the various stages of paralysis, it was common to see
small drops of clear and dark fluids issuing from the cornicles and anal
openings.

( c ) insects sprayed with solution op nicotine sulphate

Many aphids on leaves of the Carolina poplar (Populus deltoides ) were
heavily sprayed with a solution of nicotine sulphate, made in the propor¬
tion of 1 ounce of the nicotine sulphate to % gallon of water, this being
1 part of the insecticide to 64 parts of water, which is 12^ times as strong
as recommended for the more resistant sucking insects, such as the black
aphis and woolly aphis. This nicotine sulphate is guaranteed to be at
least 40 per cent nicotine, and the analysis of this sample showed that it
contained a fraction more than 40 per cent. Four hours after being
sprayed, all these aphids were apparently dead.

In practical work nicotine as an insecticide is rarely used for caterpillars
and probably never in the form of spray for the imagoes of coleopterous
and hymenopterous insects, but it was desirable to ascertain how nicotine
affects various kinds of insects and to obtain material for the study of the
tissues after the insects had died of nicotine poisoning. For this reason
various kinds of insects were heavily sprayed with the above solution.
An hour elapsed before the large caterpillars of the catalpa sphinx
(i Ceratomia catalpae Bdv.) died; however, a much weaker solution
(1 : 1,200) apparently killed the small caterpillars (6 to 10 mm. in length)
of the same moth in five minutes, and the same was true of an extract
made of powdered tobacco and water (50 gm. of tobacco boiled in 1,000
c. c. of water). The stronger solution of the nicotine sulphate quickly
killed the small caterpillars of Atteva aurea Fitch, and of Datana sp.,
but it was not so effective on the larger larvae of the lesser wax moth

Journal of Agricultural Research

Vol. VII, No. 3

(Achroia grisella Fab.) and on bagworms, larvae of Thyridopteryx epheme¬
rae j or mis Haw. Of the four adult blister beetles (Epicauta penns ylvanica
DeG.) sprayed, only three died, and worker bees could not be killed by
spraying them.

From the foregoing it is thus seen that there is considerable difference
in the responses of various insects to nicotine spray solutions. The
youngest and smallest individuals of any given species always succumb
first and some imagoes, such as bees, can not be killed at all.

3. - NICOTINE AS A FUMIGANT

The following apparatus was devised : To the neck of a 50 c.c. retort
supported on a ring stand was connected a piece of rubber tubing 12
inches in length, with its free end projecting into a battery jar 9 inches in
diameter and 12 inches in height. The jar was covered with a piece of
glass.

(a) APHIDS FUMIGATED WITH PURE NICOTINE

Carolina poplar leaves bearing many aphids {Aphis populifoliae) were
supported in a bottle, and the bottle with its contents was placed inside the
battery jar so that the aphids did not touch the sides of the jar. Twenty-
five c. c. of pure nicotine were poured into the retort, which was then
heated gently. The free end of the tubing was removed from the battery
jar, and the heat was still applied. Brownish fumes soon arose from the
nicotine; they immediately condensed upon striking the upper, colder
portions of the retort, which soon became too warm to condense them.
Other fumes then passed into the neck of the retort, where they were
likewise immediately changed into liquid, which ran in little streams back
into the retort. The rubber tubing was next warmed by the fumes.
As soon as drops of the liquid ceased to fall from the free end of the tubing
the fumes were passing freely from this end. The tubing was then inserted
into the battery jar. At once the aphids began to squirm, and the jar
was soon filled with dense fumes. At this instant the burner was removed
from under the retort, whereupon the fumes began to condense. A
little later small streams of the liquid ran down the sides of the jar, and
small drops collected on the underside of the glass cover. The leaves and
the aphids seemed to be covered with a fine spray.

So far as could be observed through the dense fumes, the behavior of
these aphids was similar to that of sprayed aphids. Before dying
many of them dropped from the leaves. Most of them appeared dead
within three minutes after the introduction of the fumes; two minutes
later still all of them were dead.

The preceding mode of procedure has been described in detail in order
to make evident the ease with which the liquid can be applied by fumi¬
gating to cool surfaces with which the vapors may come in contact.
Since the temperature of insects is practically the same as that of the

Oct. 16, 1916

Effects of Nicotine as an Insecticide

air surrounding them, it seems evident that the nicotine fumes would be
condensed upon striking the integuments and tracheal walls of the
insects fumigated.

The preceding experiment was repeated by using aphids (Aphis
rumicis L.) on nasturtiums. Small pots containing these plants were
placed inside the battery jar. Five minutes after introducing the fumes
all the aphids were dead.

(b) INSECTS FUMIGATED WITH 40 PER CENT NICOTINE-SULPHATE SOLUTION

The preceding experiments were repeated by fumigating the follow¬
ing insects with a 40 per cent solution of nicotine sulphate: Aphids on
nasturtiums (Tropaeolum majus) and those (Myzus persica Sulz.) on
potato plants (Solanum tuberosum ), coccids (Orthezia insignis Dougl.), fall
web worms (caterpillars of Hyphantria cunea Dru.), larvae of potato
beetles (Lepiinoiarsa decemlineata Say), imago house flies (Musca domes -
tica T.) and worker honeybees (Apis mellifica T.)*

The aphids and coccids died a few minutes after the introduction of
the fumes, and the plants which bore them were also affected consider¬
ably by the fumes. The leaves on the potato plants soon wilted, and
some of them finally turned brown. They emitted a comparatively
strong nicotine odor for several days, and even a very faint nicotine odor
was perceptible 15 days after the plants were fumigated.

The fall webworms and potato-beetle larvae (two-thirds grown) were
not so easily killed, although after being confined for a period of 15 or 20
minutes in dense fumes, they die. While dying, the caterpillars wriggle
about considerably and exude a yellowish fluid from the mouths. Bees
die in the same length of time, but house flies do not succumb so readily.
Bees, when apparently dead, often revive if they are removed from the
jar to fresh air.

The preceding experiments indicate that nicotine as a fumigant kills
insects by paralysis and that part, if not all, of the fumes, which strike
the integuments and which pass into the tracheae of the insects, are
condensed before they enter the various tissues. On page no it is
shown that the nicotine never passed far from the tracheae into the
tissues. This supports the view that nearly all of the fumes in the
tracheae were changed into liquid which did not pass readily through
the tracheal walls. It is also seen that the most delicate insects yield
first to nicotine fumes.

4. - NICOTINE ODOR AND VAPOR

To determine the effects of nicotine odor on insects, leaves were either
sprayed with or dipped into nicotine spray solutions. Their stems were
then inserted into bottles of water which were placed in the sun or in
front of an electric fan. The leaves were alwavs left in the current of

Journal of Agricultural Research

Vol. VII, No. 3

air from the electric fan for an hour, at the end of which time they were
perfectly dry, but still emitted a very faint odor of nicotine. A slightly
longer time in the sun was required before they became perfectly dry.
Each bottle with its contents was placed inside a battery jar 5 inches in
diameter by 1 1 inches in height. Normal and untreated insects were
then removed with a camel's-hair brush from other leaves to the leaves in
these bottles. A glass cover was placed over each bottle, and the insects
were observed at regular intervals.

To ascertain the effects of nicotine vapor on aphids and bees, the
insects were either inclosed in a battery jar with nicotine spray material
below them or with the spray solutions placed in watch glasses or on the
leaves near the insects in the open.

(a) ODOR FROM SOLUTION OR PURE NICOTINE

Carolina poplar leaves were dipped into a solution of pure nicotine
(1 : 100), placed in the current of an electric fan for an hour, and were
then arranged as already described. At 11 o'clock aphids (Aphis popu-
lifoliae) from other leaves of the same tree were transferred to the leaves
treated with the nicotine solution. At 4.30 o'clock the aphids were
slightly stupid. The next morning all of them were dead. Not one of
the aphids used as controls died.

One day aphids were killed by being placed in vials which a week
before had contained some of the nicotine solution. These vials after
having been used had not been washed, and a week later two of them
were unintentionally used for collecting aphids in the greenhouse. By
the time a dozen aphids had been put into each vial and closed with
stoppers which had also been used a week before, most of the insects
were dead, and the remainder of them were in the last stage of paralysis.
An examination showed that the vials still gave off a very slight odor
of nicotine.

(6) VAPOR PROM PURE NICOTINE

At 10.30 o'clock a large Carolina poplar leaf bearing many aphids
was put into one of the battery jars. A small beaker containing 5 c. c.
of pure nicotine was also placed inside the jar about 5 inches below the
leaf. At 12 o'clock a few aphids were stupid; at 1 o'clock several were
dead; at 4.30 most of them were dead; the next morning all of them
were dead.

A few cubic centimeters of a pure nicotine solution (1 : 100) were
poured into each of seven watch glasses. A small wire screen was laid on top
of each watch glass so that it did not touch the nicotine solution. Sev¬
eral cabbage aphids ( Aphis hrassicae L.) were then placed on each wire
screen. The smallest aphids died within 10 minutes, the medium-sized
ones within 16 minutes, and the largest ones within 22 minutes.

Oct. 16, 1916

Effects of Nicotine as an Insecticide

At 11 o'clock the upper surface of a large dock leaf (Rumex sp.) was
sprayed heavily with the nicotine solution (1 : 100) in the greenhouse.
A comparatively large nasturtium leaf was placed directly over and
one-half inch above the dock leaf. The under surface of each leaf
bore many aphids. At 2 o'clock all the aphids on the underside of the
nasturtium leaf were dead, while none on the dock leaf apparently had
been affected.

Catalpa leaves were sprayed heavily with the extract of powdered
tobacco described on page 93. After these leaves had become per¬
fectly dry in the sun, they still emitted a faint nicotine odor. At 1
o'clock many small caterpillars (6 to 10 mm. in length) of the catalpa
sphinx were then tested by being placed inside battery jars in the man¬
ner already described. At 1.10 o'clock a few of them were dying; at
2.25 four-fifths of them were dead. These had not eaten of the leaves.
The remaining ones did not die until three days later; they had eaten
the leaves to a limited degree.

The preceding experiment was repeated twice by using large fall
webworms on mulberry leaves (Morns sp.) that had been submerged
for two minutes in the extract and had been dried in the current of air
from the electric fan. The first lot of caterpillars ate the leaves readily
and apparently were not affected, but a third of the second lot was
dead the following morning, after having slightly eaten the leaves.

(d) ODOR FROM SOLUTION OF NICOTINE SULPHATE

The experiment just preceding was repeated twice by using a solution
of nicotine sulphate (1 : 1 ,200). The results obtained with the first lot of
fall webworms showed that only, one caterpillar was killed, but the
leaves were not much eaten. Relative to the second lot, the leaves were
not eaten at all. Two hours after being placed on the leaves, many
of the caterpillars became stupid and a few showed signs of dying, but
were not found dead until the following morning. Two days later still all
of them were dead.

The preceding was repeated by heavily spraying catalpa leaves with
a much stronger solution of nicotine sulphate (1 : 64) and by using
small caterpillars (6 to 10 mm. in length) of the catalpa sphinx. At
12.52 o'clock these leaves had become perfectly dry in the sun; at 12.56
a few of the caterpillars acted as if dying; at 2.25 all of them were dead.
The leaves gave off a slight nicotine odor and had not been eaten.

The preceding was repeated by dipping Carolina poplar leaves into
the same solution and by using the aphids removed from these leaves
after the latter had been dried in the current of air from the electric
fan. The following morning most of these aphids were dead.

Journal of Agricultural Research

Vol. VII, No. 3

(e) VAPOR FROM A 40 PER CENT NICOTINE-SUIPHATE SOLUTION

Fifty worker bees in an observation case were introduced into a large
battery jar. A small quantity of a 40 per cent nicotine sulphate solu¬
tion in a Petri dish 10 cm. in diameter was placed inside the jar 8 inches
beneath the case of bees. During all the following day the bees remained
more or less inactive and appeared slightly stupid. The next day follow¬
ing they were still slightly abnormal in behavior, but none died.

The upper surfaces of large dock leaves were heavily sprayed with
two solutions of nicotine sulphate (1 : 100 and 1 : 500) in the greenhouse.
The aphids on the under surfaces of these leaves apparently were not
affected by the vapor from either solution. On the other hand, when
a small amount of the stronger solution was placed on the under surfaces
of nasturtium leaves near the aphids but not against them, most of
the insects were found dead three hours afterwards. When a piece
of cheesecloth wet with the weaker solution was placed an inch beneath
the branches and leaves of a nasturtium, a few of the many insects on
this plant were found dead.

In view of the results of all the preceding experiments in which the
spray solutions had been evaporated, it may be argued that many of the
insects died of nicotine poisoning by eating the leaves which had previ¬
ously been treated with nicotine solutions. That these leaves still emitted
a faint odor of nicotine indicates that their surfaces still bore many traces
of the alkaloid. It is also probable that some of the nicotine passed into
the tissues of the leaves. Since some of those insects that did eat the
leaves died so quickly after being placed inside the jars, it does not seem
logical that they died primarily from the effects of nicotine as a stomach
poison, because a small amount of nicotine as a stomach poison acts
slowly. In view of the preceding reasoning and since some of the insects
did not eat the leaves at all, it seems safe to say that most of them were
killed by the odoriferous particles of the nicotine passing into the tracheae.
In all of those experiments in which the insects were subjected to nico¬
tine vapor, although they did not actually touch the nicotine solutions,
there can be no doubt that the vapors killed the insects; and it is also
probable that the vapors passed into the tracheae and killed by paralyzing
the nervous system. These experiments demonstrate that nicotine spray
solutions are not necessarily contact insecticides, although they are more
effective when actually used as such, for by this means the insects are
constantly brought near the vapor under the most favorable conditions.

TRACING NICOTINE TO TISSUES

Owing to the small sizes of the insects used, it was not considered possi¬
ble to operate successfully on live individuals in order to determine what
particular tissue is vitally affected when nicotine is used as an insecticide.
Drawing conclusions solely from the behavior of the insects dying of nico¬
tine poisoning, the author states in the preceding pages that they die of

Oct. i6, 1916

Effects of Nicotine as an Insecticide

paralysis. Since paralysis is an affection of the nervous system, it still
remains to be shown that nicotine applied as an insecticide reaches the
nervous system and how it affects the nerve cells. The following pages
deal with this portion of the work. Many difficulties were encountered,
and the experiments performed to determine how nicotine affects the
nerve cells gave no definite answer to this question. The latter phase of
this subject is presented mostly by giving a brief discussion of the various
views pertaining to the physical and chemical effects of drugs on cells.

Tracing the nicotine into the various insect tissues was accomplished
by precipitating this alkaloid immediately after it had killed the insects
and then by carefully studying the microscopical sections made from the
insects thus tested. A study of this nature involves considerable tech¬
nique and many precautions in making sections, because two objects
instead of one must be successfully accomplished at the same time. It
is an easy matter to obtain good sections of most larvae and soft-bodied
insects under ordinary conditions, but it was found quite difficult to
obtain good sections and at the same time not to lose the precipitates
held in the tissues while the slides were being run through the various
reagents. This is appreciated when we consider the solubility of various
substances in the clearing oils, in the alcohols, and in water.

In addition to the difficulties enumerated above, there are still three
more to be considered: (1) After a certain period has elapsed following
death as a result of having been treated with nicotine, the tissues of the
insects were unusually abnormal upon fixation. As soon as life is ex¬
tinct, and probably a short time before, the cells gradually change from
normal to abnormal ones. This was particularly noticeable when small
caterpillars were sprayed with solutions of nicotine. A short time
after death they turned brown and the tissues were found to be more or
less disintegrated. For this reason it was always necessary to fix the
insects just before the last signs of life had disappeared in order to avoid
mistaking post-mortem changes in the cells for physical ones caused by
the nicotine. (2) On the other hand, if nicotine really causes physical
changes in the cells of insects, these changes are always masked by the
large physical ones caused by the fixing reagents. (3) It is often difficult
and sometimes impossible to distinguish the precipitated insecticide in¬
side the tissues from the coagulated constituents of the cells caused by
the fixative. Fischer (8) regards the coagulation of these constituents,
which really constitutes fixation, as a true precipitation, but of course
it is a milder form. The coagulated particles are, nevertheless, frequently
as large and sometimes larger than the precipitated ones.

I . — TRACING COLORED LIQUIDS INTO INSECTS

Before determining whether or not nicotine spray solutions as applied
under practical conditions reach the tissues by passing through the
spiracles, many preliminary experiments were performed to ascertain

IOO

Journal of Agricultural Research

Vol. VII, No. 3

whether water and nicotine solutions, applied under the most favorable
conditions, are able to enter the spiracles, mouths, and anal openings of
various insects. In order to follow these liquids, they were colored with
various stains.

(a) ABILITY OP COLORED LIQUIDS TO ENTER THE SPIRACLES AND ALIMENTARY CANAL

A small quantity each of water and of pure nicotine solution (i : ioo)
was colored with each of the following aqueous stains: Carmine acid
(Griibler's Carminsaur), eosin, gentian violet, Delafield’s, Ehrlich's, and
pure hematoxylins, methylene blue and safranin. Cabbage aphids
were submerged in each of these colored liquids for an hour. After
removal from the liquids they were well washed in water, then crushed,
and finally mounted on slides in glycerin. The glycerin did not mix with
nor scatter the stains. Various parts of the integuments were colored
more or less with each stain, but methylene blue seemed to penetrate
the integuments the most readily of any of these stains. Each stain
seemed to pass into the tracheae more or less, but the four following ones
entered most readily: Carmine acid, gentian violet, methylene blue, and
safranin. Of these four, carmine acid proved to be the best. The
tracheae in most of the aphids showed scarcely any of the stain, while
those in the remaining ones showed it conspicuously. In these few
insects all the larger tracheae in the abdomen, thorax, and head were
stained ; and occasionally a stained trachea was traced into a leg. There
seemed to be no difference in permeability between the stains dissolved
in water and those dissolved in the nicotine solution.

The preceding experiments were repeated by submerging roaches
{Periplaneta americana L.), croton bugs (. Blattella germanica E.), house
flies, and larvae of blow flies ( Calliphora vomitoria L.) for an hour in
water and in a pure nicotine solution (i : 500), each being colored with
carmine acid. The stain was observed in the esophagus and anus of the
roaches and croton bugs; in a few of the larger tracheae and in many
of the smaller ones and in the hind gut of the fly larvae. The house flies
were fixed in absolute alcohol, which readily throws down carmine acid.1
One of the flies that had been submerged in the colored nicotine solution
was sectioned, and the sections were placed in xylol alone without being
stained, in order that none of the “precipitated” carmine acid might be
lost. A study of the sections showed the “precipitated” carmine in
several of the larger tracheae (PI. 1, fig. A, pr). An examination of the
other flies showed that the colored liquid had passed into some of the
larger tracheae and into the rectums.

The preceding experiments were repeated by submerging green peach-
aphids ( Myzus persicae) in pure nicotine solution (1:500), colored with

1 Absolute alcohol does not precipitate carmine acid nor indigo-carmine, but merely throws them out
of solution, because they are not soluble in it. For lack of an appropriate term to describe these stains
when thrown out of solution the word precipitate in quotation marks may be used.

Oct. 16, 1916

Effects of Nicotine as an Insecticide

101

indigo-carmine (sodium sulphindigotate) for 45 minutes. These insects
were fixed in absolute alcohol, which readily throws down indigo-carmine.
The resulting “ precipitate” is totally insoluble in xylol and absolute
alcohol, but its solubility in the other alcohols increases as the water in
them increases. Tor this reason the sections of these aphids were stained
in absolute alcohol containing safranin. The blue “precipitate" was
common on the outside of the integument, but it was not seen inside the
integument anywhere, except occasionally in the larger tracheae and then
usually not far from the spiracles. Plate 1, figure B, represents the
“ precipitate’ ' ( pr ) observed in two places in the same trachea. This
drawing was made from two consecutive sections and shows the most
“precipitate" that could be found.

The experiment just preceding was repeated by submerging larvae of
wax moths (Achroia grisella) and small nymphs of croton bugs in the
above liquid for 30 minutes. A small amount of blue “precipitate"
(PL 1, fig. C, pr) was observed in most of the larger tracheae of the wax
moths and occasionally some (PI. 1, fig. D, pr) in the smaller ones.
Each of the sections of .the croton bugs, containing parts of the ali¬
mentary canal, shows more or less blue “precipitate" inside this tube;
but none was observed elsewhere inside the integument.

Thirty worker bees were submerged for 30 minutes each in water and
in pure nicotine solution (1 : 500), each liquid being colored with carmine
acid. When removed from the liquids, the bees were thoroughly washed
in water. The apparently dead bees were then laid on blotting paper in
the sun to become dry and to revive from the effects of the liquids. All
30 bees submerged in the colored water revived and were walking about
in from 12 to 18 minutes, with an average of 15 minutes, after being placed
in the sun. Only 60 per cent of those submerged in the colored nicotine
solution revived, and these never became able to fly as did those sub¬
merged in the colored water. The time required for them to recover
sufficiently so that they could crawl about varied from 45 fninutes to
3^ hours, with about 2 hours as an average. •

An examination of the live bees just described showed that the thin
chitin between the segments was often colored red and that when a
thorax was crushed, the red liquid usually issued from the mouth.
When the bees that had been submerged in the colored water were dis¬
sected, the stain was seen in 90 per cent of the honey stomachs, in 64
per cent of the ventriculi, in 50 per cent of the rectums, and in 50 per
cent of the anal openings and around the stings. When the bees that
had been submerged in the colored nicotine solution were dissected,
the stain was observed in 45 per cent of the honey stomachs, in only
6 per cent of the ventriculi, never in the rectums, and in 45 per cent
of the anal openings and around the stings. The behavior of the bees
when placed into the nicotine solution may be used to explain why
such a small amount of the colored solution passed into the alimentary

102

Journal of Agricultural Research

. Vol. VII, No. 3

canal. While the bees placed into the colored water struggle and cling
to one another for several moments after being submerged, those placed
into the colored nicotine solution struggle little and never cling to one
another. They seem to be slightly paresized as soon as put into the
solution, and perhaps for this reason alone they swallow little of the
liquid. Paresis may also be used to explain why the colored liquid is not
forced from the honey stomachs into the ventriculi.

That the colored, liquids were never seen in any of the tracheae of
the 60 bees submerged demonstrates that the valves guarding the
spiracles closed water-tight at the instant of placing the bees into the
liquids. That one-half of the bees submerged in the nicotine solution
did not swallow any of it indicates that these valves can not be closed
air-tight, because there seems to be no way of explaining why the bees
were paresized other than by supposing that vapor from the nicotine
solution passed the valves and entered the tracheae. Of course, the vapor
might have entered the insects through the mouths and anal openings,
but this view is highly improbable, and the liquid had not penetrated
the integuments even at the thinnest places. .

To determine whether the red liquid passed through the thin chitinous
layer of the honey stomachs, several of these organs which were almost
full of the red liquid were removed after both ends of a honey stomach
had been securely ligatured with thread. Immediately after a honey
stomach had been dipped into water to moisten its walls, it was gently
rolled on white paper. No red liquid was seen issuing through its walls;
nor was any observed on the paper. The same experiment was repeated
by using the ventriculi. In this case red liquid was plainly seen to issue
from the walls of each ventriculus, and it made the paper red. Thus,
it seems that the red liquid usually seen surrounding the viscera of these
bees when cut open had not passed through the walls of the honey stom¬
achs, but through those of the ventriculi and probably to a limited degree
through the walls of the small intestines and those of the rectums,
although it is shown on page 108 that a nicotine solution containing
indigo-carmine does not pass through the walls of the small intestine
and rectum.

In all the preceding experiments liquids colored with stains have
been used. It is probably true that many stains increase the permea¬
bility of their solvents and consequently may also increase the ability
of the solvents to pass into small openings, such as the spiracles. For
this reason coccids (Orthezia insignis) were submerged for 30 minutes
in a pure nicotine solution (1 moo) which had not been colored with
a stain. They were then fixed for 15 minutes in a mixture consisting of
two parts of absolute alcohol and one part of phosphomolybdic acid.
To insure the removal of all the phosphomolybdic acid not united with
the nicotine and to insure better fixation, the insects were put into a
mixture consisting of two parts of absolute alcohol and one part of

Oct. 16, 1916

Effects of Nicotine as an Insecticide

103

glacial acetic acid. After remaining in this mixture for about three
hours, they were placed into absolute alcohol for another hour.

Phosphotnolybdic acid is one of the alkaloidal reagents, and it pre¬
cipitates nicotine even in a dilution of 1 to 40,000. It was prepared
according to the directions of Autenrieth (1) with modifications as follows:
A sodium-carbonate solution was saturated with pure molybdic acid;,
one part of crystallized disodium phosphate (Na2HP04 — i2H30) to five
parts of the acid was added and the mixture evaporated to dryness. The
residue was fused and the*cold melt was dissolved with absolute alcohol.
This mixture was filtered and enough nitric adid added to produce a
golden-yellow color. The resulting mixture, called “ phosphomolybdic
acid,” was used full strength when mixed with absolute alcohol to serve
both as a fixative and as a precipitant. The precipitate resulting from
the union of this mixture and nicotine is neither soluble in water nor in
any of the alcohols; but for fear of losing some of the precipitate the
sections were stained in safranin dissolved in 95 per cent alcohol.

A study of the sections of the coccids treated as described above
showed a brownish yellow precipitate inside many of the tracheae
(PI. 1, fig. E, pr)j but it was not seen elsewhere inside the integument.
In the sections of coccids used as controls no precipitate was seen any¬
where. It is thus seen that nicotine solutions containing no stains are
able to pass into the tracheae of coccids that have been submerged in
the solution for 30 minutes.

(b) Jt BILITY OP NICOTINE} SPRAY SOLUTIONS TO ENTEJR SPIRACLES

Aphids {Aphis hrassicae , A . rumicis L., and Macrosiphum sanborni Gill.) ,
and coccids (Orthezia insignis) were sprayed with a pure nicotine solution
(1:500), colored with carmine acid, until they were wet with spray.
An hour later they were mounted on slides, as described on page 100,
and were examined. The spray had evaporated, leaving the red stain
adhering to various parts of the integuments. Nearly all of the tracheae
showed no signs of the stain; but a few seemed to be slightly pink,
although this kind of an examination is not entirely reliable.

The preceding experiments were repeated by heavily spraying aphids
{Aphis rumicis) with a pure nicotine solution (1:500), colored with
indigo-carmine. The sections were stained as described on page 101.
A thorough study of these slides showed that the colored nicotine solu¬
tion had not passed through the integuments nor into the tracheae. At
only one place was it found that the “precipitate” had lodged in a
spiracle (PI. 1, fig. 6, sp); but it was- never observed in the tracheae
(PI. 1, fig. 7, tr), nor elsewhere inside the integuments, although it was
commonly seen adhering to the outer surfaces of the integuments (PI. 1 ,
fig. 7, inf).

Aphids of the same species as just described were heavily sprayed
with a pure nicotine solution (1 : 100), not colored with any stain. These

104

Journal of Agricultural Research

Vol. VII, No. 3

aphids were fixed with the mixture of absolute alcohol and phospho-
molybdic add and were further treated as described on page 102. No
precipitate was found in the tracheae nor dsewhere inside the integu¬
ments.

Some of the caterpillars of the catalpa sphinx (Ceratomia catalpae ) , of
Atteva aurea, of Datana sp., and larvae of the lesser wax moth, Achroia
grisella , that had been sprayed with a solution of nicotine sulphate
(1:64) were fixed in Carnoy's fluid. This fixative is a mixture com¬
posed of equal parts of absolute alcohol, chloroform, and glacial acetic
acid, saturated with mercuric chlorid (HgCl3). Mercuric compounds are
among the general alkaloidal reagents and the mercury in mercuric
chlorid readily predpitates nicotine. The sections of the above larvae
were stained with Ehrlich's hematoxylin and the crystals of the mercuric
chlorid remaining in the tissues after the fixative had been washed out
were removed by 95 per cent alcohol containing tincture of iodin. The
iodin unites with the mercuric chlorid, forming a compound which readily
dissolves in alcohol, but iodin apparently has no effect on the gummy
precipitate resulting from the union of nicotine and Carnoy's fluid. For
this reason, if the spray solution passed into the tracheae, the nicotine in
it should have been precipitated and should not have been affected by
the iodin, and all of it certainly could not have been washed out while
the slides were being run through the reagents, At any rate, after the
sections had been treated with tincture of iodin, no precipitate of any
kind was observed inside the integuments of these larvae. This indicates
that a spray solution of nicotine sulphate does not enter the spiracles
nor pass through the integument; but this fiiethod is not fully reliable,
on account of having to remove the crystals of mercuric chlorid.

Since it has been shown that spray solutions, as applied under prac¬
tical conditions, do not pass into the tracheae, a study of the spiracles
of the aphids, coccids, and larvae that have been used in the experi¬
ments shows that it is practically impossible for aqueous solutions to
enter the spiracles. The mouths of the spiracles of all these insects,
except the coccids, are guarded by pseudohairs, which are outgrowths
from the chitinous linings of the spiracles and by the rims (PI. 1, fig. J, r)
turning inward and sometimes downward. The spiracles (PI. 1, fig. H,
sp) of the coccids are practically unprotected, while those of aphids
(PL 1, fig. B, sp) bear a few hairs. The small size of these spiracles
seems to be the best reason why aqueous solutions can not readily pass
into the tracheae. The hairs guarding the spiracles of some of these
insects vary from short, stout ones, as in the larvae of Atteva aurea (Pl.i,
fig. I, sp), in the lesser wax moth (PI. 1, fig. J, sp) and in Datana sp.
(PI. 1, fig. K, sp) to long, stout ones, as in the caterpillar of the catalpa
sphinx (Pl. 1, fig. E, sp) and in the larva of the Colorado potato beetle
(PL 1, fig. M, sp). The hairs (Pl. 1, fig. N, hr) in a spiracle of the fall

oct. 1 6, 1916 Effects of Nicotine as an Insecticide 105

webworm are branched, and they nearly close the entrance, while the
entrance of a spiracle (PL 1, fig. O, sp) in the tomato worm (larva of
Phlegethontius sexta Joh.) is closed by a hairy and porous plate (Pl. 1,
fig. O, p), which has an oblong opening through its center.

Shafer (20) colored kerosene with Sudan III; and after thoroughly
spraying or dipping grasshoppers ( Melanoplus femoratus Burjn.), tomato
worms, and aphids into this oil and after dissecting these insects, he
found more or less of the red oil in the larger tracheae. He colored kero¬
sene emulsion and the emulsions of the miscible oils with indigo-carmine
and with saf ranin and found that they also enter the spiracles. Shafer
repeated these experiments by treating aphids with creolin emulsion
containing indigo-carmine. After fixing the insects in absolute alcohol
and after studying the sections he observed plugs of “precipitated”
indigo-carmine in the larger tracheae, which were sufficiently large to
close them.

Dewitz (7), in discussing contact insecticides, does not believe that
either liquids or powders can enter the spiracles in sufficiently large
quantities to cause the death of the insects by suffocation.

Without attempting to apply the physical law governing the surface
tension of liquids, the following experiment was performed to determine
roughly the surface tensions of water, different solutions of 40 per cent
nicotine sulphate, pure nicotine, kerosene, gasoline, and kerosene emul¬
sion. Fresh nasturtium leaves were spread out flat on a table, with the
under surfaces upward. With pipettes drops of water, solutions of
nicotine sulphate (1:500 and 1:100), undiluted 40 per cent nicotine
sulphate, undiluted pure nicotine, pure nicotine solutions (1:500 con¬
taining indigo-carmine and 1:100), kerosene, gasoline, and kerosene
emulsion were dropped upon the nasturtium leaves. Of all these liquids
the surface tension of gasoline was weakest and that of water the strongest;
that of kerosene was second weakest, while those of pure nicotine and
kerosene emulsion were about equal, but still much stronger than that
of kerosene. So far as practical work is concerned, the ability of 40
per cent nicotine sulphate and its solutions and of the two enumerated
solutions of pure nicotine to spread over the surfaces of these leaves is
about equal to that of water. The drops of each one of these liquids
upon striking the leaves form small spheres and are not retained when
the leaves are somewhat inclined.

In regard to the evaporation of the solutions of nicotine sulphate and
of pure nicotine, the drops of the solutions of pure nicotine evaporated
rather quickly while those of the nicotine sulphate did not disappear
for some time. The more nicotine contained in the drops of the pure
nicotine solutions, the more quickly they evaporated. The evapora¬
tion of the drops of the solution of nicotine sulphate (1:500) was about
equal to that of the water drops.

55857°— 16 - 2

106 Journal of Agricultural Research voi. vn, no. 3

From what we know about the relative surface tensions of nicotine
solutions, of kerosene, and of various emulsions, it is easily understood
why kerosene and the emulsions are able to pass into the tracheae while
the nicotine solutions can not.

2. — tracing nicotine: as a stomach poison to tissues

To obtain material for tracing nicotine as a stomach poison to the
tissues and to determine the effects of an extremely small amount of
nicotine and of indigo-carmine on bees, the following experiments were
performed: 200 drops of pure honey were put into a feeder; 200 drops
of honey mixed thoroughly with 40 drops of water colored blue with
indigo-carminh were poured into a second feeder; 200 drops of honey
mixed thoroughly with 40 drops of pure nicotine solution (1 : 500)
colored blue with indigo-carmine were poured into a third feeder. Each
of these feeders, with its contents, was placed inside an observation
case, and 50 worker bees were introduced into each case. Before all
the bees died nearly ail the food had been eaten. Since bees confined
in observation cases can not void their feces, the abdomens of these
bees became much distended with the blue-green food. The bees that
ate the pure honey lived eight days, on an average, while those in the
two other cases lived about seven days, on an average, showing that the
extremely small amount of nicotine did not affect their longevity,
whereas the indigo-carmine seemed to shorten their lives by one day.

The preceding experiments in feeding bees nicotine and indigo-carmine
were repeated ; and three days later , when several bees showed signs of
dying, they were placed into absolute alcohol for two days. The anterior
portions of their abdomens, and occasionally the base of a leg, appeared
blue-green from the outside. When cut open under absolute alcohol,
all the tissues in the abdomen appeared blue-green. A closer examina¬
tion, however, showed that the alimentary canal was blue, while the
other tissues in the abdomen as a rule were pale blue-green, with now
and then darker colored streaks running through them. A few muscle
fibers and some parts of the chitin were pale blue, and other parts of
the chitin were pale blue-green. Under alcohol the tissues in the thorax
and head did not appear colored at all; but after being removed from
the alcohol and dried, they assumed a pale blue-green color, and occa¬
sionally darker colored blue streaks were seen in the muscles and brain.
It seems that the indigo-carmine had colored the blood or body fluid
pale blue-green and that this fluid in turn had colored all the tissues, but
the stain was diluted too much to be “precipitated,” except in a few
organs.

Parts of the alimentary canal and various tissues were dissected out,
and sections were made of them. Sections through the anterior and
middle portions of the honey stomach failed to show any blue “precipi-

Oct. 16, 1916

Effects of Nicotine as an Insecticide

107

tate” either in the lumen or in the walls of this organ, but sections
through the posterior portion of the honey stomach and other parts of
the alimentary canal distal to the honey stomach usually showed more
or less blue “precipitate.” There was no difference in distribution of
the stain, whether or not it contained nicotine, but since the distribu¬
tion of nicotine only is of interest, the discussion will be confined to the
distribution of the stain which formerly contained this insecticide.

Plate 2, figure D, represents the blue “precipitate” as seen in the
wall of the lower portion of the honey stomach. The stain seemed to
have united with the alcohol as the former was passing through the
chitinous layer {chi) of this organ. A little “precipitate” was also seen
in the muscular layer {m) of this organ.

Sections through the anterior portion of the valve of the proven-
triculus show small particles of “precipitate” in the muscles (Pi. 2, fig.
E, w), epithelial cells (ep), and tracheae ( tr ).

Sections through the ventriculus show “precipitate” in various places
of the epithelium, indicating that the stain was in the act of passing
through the wall when it was overtaken by the alcohol. From the loca¬
tion of the blue “precipitate” some of the stain was just ready to pass
into the inner ends of the epithelial cells, while some had just entered
these cells (Pi. 2, fig. A, pr). Other portions of the stain were “pre¬
cipitated” midway between the inner and outer walls of the epithelium
(Pi. 2, fig. B, pr), while still other portions were overtaken by the alco¬
hol when they were passing through the outer wall of the epithelium
(Pi. 2, fig. C, pr ). At this location a small amount of “precipitate”
was also seen in the transverse muscular fibers (PI. 2, fig. C, tm), indi¬
cating that, while most of the stain passed between the muscular fibers,
some of it passed into and probably through the fibers.

At only one place was blue “precipitate” (PI. 2, fig. H, pr) observed
in the blood (bl). This was seen a short distance from the ventriculus
near a small trachea (tr) and two Malpighian tubules {mat), which also
contain a little “precipitate.” Two particles of this “precipitate” are
lying in the outer walls of these tubules, indicating that the stain was
passing into these organs when it was thrown down. It was also
observed that the trachea contained several small particles of “precipi¬
tate.”

Many of the Malpighian tubules, particularly those near the honey
stomach, small intestine, and rectum, showed no traces of the stain,
while those near the ventriculus contained a small amount of it, as repre¬
sented in Plate 2, figure H, and whereas those against the ventriculus
contained large amounts of the “precipitate,” as represented in Plate 2,
figure G.

The “precipitated” particles in sections through the middle of the
ventriculus are more numerous and more compact than in sections
through either end of this organ. These sections are never perfect,

io8

Journal of Agricultural Research

Vol. VII, No. 3

and since the tissues were fixed with absolute alcohol, which must have
passed into the bees chiefly through the mouths and anal openings, the
cells in the epithelial lining of the alimentary canal were not well pre¬
served, so they have been drawn diagrammatically in outline from
Snodgrass (22). In sections through the middle of the ventriculus, the
“ precipitate ” is arranged more or less in concentric circles (Pi. 2, fig. J,
pr ). This arrangement is probably caused by the peritrophic mem¬
branes being likewise arranged. In the lumen (/) and between the con¬
centric circles of * ‘ precipitate ” the “precipitated” particles are scattered
irregularly. In Plate 2, figure J, the epithelial ( ep ) and muscular (m)
walls have been drawn diagrammatically, showing how the stain proba¬
bly passes through these walls into the blood where most of it is taken up
by the Malpighian tubules. The other parts of this figure were drawn
with the aid of a camera lucida.

Since no blue “precipitate” was observed in the epithelium (PI. 2, fig.
F, ep) of the small intestine, it seems that the stain did not pass through
the walls of this organ, although “precipitate” was easily seen in the
lumen (/) of the small intestine. Most of the “precipitate” usually
occurred in large particles near the center of the lumen, while the inner
wall of the epithelium was often lined with a layer of “precipitate”
composed of innumerable small particles.

Despite the fragmentary sections of the rectum, a careful study was
made of this organ, but no blue “precipitate” was seen in its walls;
nevertheless it was quite conspicuous in its lumen (PI. 2, fig. I, /). In
the rectum, as well as in the other parts of the alimentary canal, there
was considerable “precipitated” material which was not stained. This
and the pollen grains were easily distinguished from the “precipitated”
indigo-carmine by the blue color of the latter.

The preceding results obtained in tracing nicotine as a stomach poison
by means of “precipitating” indigo-carmine is not meant to be conclu¬
sive, but merely to point out the possibilities for future investigations
along this line. For these results to be conclusive, the nicotine should
have been traced without the aid of a stain like indigo-carmine; buf
owing to the odor from this insecticide bees can not be forced to eat
food containing a large quantity of nicotine. For this reason it did
not seem possible in preliminary work of this nature to be able to trace
an extremely small amount of nicotine without using some compara¬
tively harmless stain with it.

In passing through the walls of the ventriculus, it is scarcely possible
that the nicotine and indigo-carmine were separated from one another,
and the experiments on page 91 show that nicotine as a stomach
poison kills by paralysis. It must therefore be concluded that nicotine
in passing through the walls of the ventriculus is not so materially
changed as to destroy its effectiveness. That all the tissues, even in¬
cluding the brain, were stained more or less with the indigo-carmine

Oct. 16, 1916

Effects of Nicotine as an Insecticide

109

shows that this substance was widely distributed, and it is also logical
to think of the nicotine accompanying the stain wherever it went, except
when the stain penetrated hard tissues, such as chitin. In the higher
animals nicotine is chiefly excreted through the kidneys, because it is
found in the urine soon after it has been administered. Since the
Malpighian tubules take up indigo-carmine so readily, it seems that
these organs would also readily excrete poisons contained in the blood.
As indigo-carmine does not seem to pass through the walls of the small
intestine and those of the rectum, they are either impermeable to this
substance or the stain has been so changed that it has lost its original
permeability. The same reasoning might also be used for nicotine or
any other stomach poison which acts similarly, although according to
Cushny (6) iron behaves quite differently when administered to the
higher animals. He says:

Iron injected into the veins of animals is stored up in the li\er, spleen and bone
marrow, but is taken up from these organs again, and is excreted by the epithelium of
the caecum and colon. When iron is given by the mouth, therefore, it may either pass
along the canal and be thrown out in the faeces, or it may be absorbed, make a stay in
the liver, be excreted in the large intestine, and again appear in the stools.

He also states that iron has been followed in its course through the
tissues by histological methods, but nothing is known about the changes
which iron preparations undergo in the stomach and intestine, or the form
in which iron is absorbed.

3. — TRACING NICOTINE AS A FUMIGANT TO NERVOUS SYSTEM

While experimenting to determine the physiological effects of nicotine
as a fumigant, various insects were fumigated with pure nicotine and a
40 per cent nicotine-sulphate solution. The results indicated that the
nicotine fumes were condensed wherever they went. Several of the green
peach aphids (Myzus persicae) that had been killed by the fumes from
the solution of 40 per cent nicotine sulphate were fixed with the mixture
consisting of absolute alcohol and phosphomolybdic acid. A careful
study of the sections made from these aphids gives the following results:

As already stated, these aphids appeared to be coveted with fine spray
before they died, indicating that the fumes had changed into tiny drops
of liquid. In the sections it is easily seen that the entire integument is
covered with minute particles of precipitate. Plate 3, figure B, taken
from a molting aphid, well represents the precipitate ( pr ) on the integu¬
ments which have been cut obliquely. It is to be noted that the minute
precipitated particles lie on the outside of both the old (intf) and new
integuments (inf), and even between them, but never on the inside of the
new integument. Aphids, not fumigated, put into the same fixative
occasionally show a little precipitated material on the outside of the
integuments, but usually it is easily distinguished from the precipitate

Journal of Agricultural Research

Vol. VII, No. 3

resulting from the union of the fumes of nicotine sulphate and phospho-
molybdic acid. It is practically impossible to find an aphid or perhaps
any other insect that does not carry at least a little organic matter on the
integument. Bees, for instance, when placed into silicotungstic acid or
into any other alkaloidal reagent soon become more or less covered with
a white precipitate, showing that the hairs are full of organic matter.

It is supposed that some of the nicotine fumes which had passed into
the tiacheae had not changed into liquid by the time the insects were
fixed, and in order to prove that the fixative precipitates nicotine, whether
in a liquid or in a gaseous state, a test tube was filled with the fumes from
the nicotine sulphate. Immediately after a small quantity of the fixative
was poured into the test tube a yellowish precipitate wras thrown down.

Upon examining the sections that had not been stained, a few of them
were observed to have a light-tan color and the chitin in places assumed
a darker tan color. All the aphids after being fumigated assumed a light-
tan color; this color was particularly noticeable when the insects were
embedded in white paraffin. The same species of aphids, not fumigated,
had a whitish appearance. The light-tan color in most of the sections is
caused by minute particles of tan-colored precipitate on the integument,
in the tracheae, and to a limited degree in the tissues, but in a few cases
the tan-colored tissues contain no perceptible precipitate. In such
instances the fumes must have penetrated the cells and mixed with the
protoplasm before the cell constituents were coagulated.

Most of the tracheae contain more or less tan-colored precipitate, but
very little of it lies outside the tracheal walls. Plate 3, figure A, represents
a large trachea ( tr ) cut both crosswise and longitudinally near a spiracle,
showing the precipitate (pr) inside the tracheal walls and some scattered
in the fat cells (jc) which surround the trachea. Tracheae may be traced
for short distances between the cells of any tissue, but the precipitate is
never found further from the tracheae than that shown in Plate 3, figure
A. It is quite probable that some of it which lies in the cells has been
dragged there from the tracheae by the microtome knife.

Baker (2) has well described the respiratory system of the woolly apple
aphis (Eriosoma lanigerum Hausm.), and he gives two drawings, one repre¬
senting the dorsal tracheal system and the other the ventral tracheal system.
To illustrate how well a gaseous or vaporous form of an insecticide may be
distributed to all tissues, these two figures are again given (PI. 3, fig. D, E).
It is to be noted that there are seven pairs of abdominal spiracles and two
pairs of thoracic ones and that in the abdomen a short distance from
each spiracle the trachea divides into two smaller branches, one of which
passes dorsally to help form the dorsal tracheal system (PL 3, fig. D)
and the other passes ventrally to help form the ventral tracheal sys¬
tem (PL 3, fig. E). In the ventral system of the thorax there are two
ventral arches, while in the dorsal system of the abdomen there is only
one, the dorsal arch (PL 3, fig. D, da). The anterior ventral arch (PL 3,

Oct. i6t 1916

Effects of Nicotine as an Insecticide

fig. E, ava ) in the thorax unites the pair of spiracles in the prothorax
and aerates the subesophageal ganglion, whereas the posterior ventral
arch ( pva ) connects the pair of spiracles in the metathorax and aerates
the large thoracic ganglion. The two ventral arches are of the greatest
interest, because they and a few other smaller branches carry nicotine
fumes directly to the nervous system, and for this reason it is under¬
standable why the fumes so quickly paralyze aphids.

Plate 3, figure E, is a reproduction of a combination drawing from
five consecutive sections through the thorax of an aphid that had been
fumigated with a solution of 40 per cent nicotine sulphate, showing the
precipitate ( pr ) on the integument {ini), in the tracheae (tr), and in the
subesophageal ganglion (sg). The large trachea was cut crosswise near
the spiracle, and the branches are drawn in only their approximate posi¬
tions. It is to be noted that the anterior ventral arch (ava) passes over
the subesophageal ganglion, but sends one of its branches under and into
this ganglion. Another large branch from the main trachea also sends
one of its branches to the same ganglion, penetrating its dorsal surface.
These sections did not actually show the small tracheae penetrating this
ganglion, but sections from several other aphids did.

Plate 3, figure C, reproduces a combination drawing from six consecu¬
tive sections of the same aphid as above described, showing three tracheal
branches entering the thoracic ganglion. Attention is to be called to the
precipitate (pr) in these tracheae and in the ganglion. Often large gran¬
ules resembling precipitated particles lie in and near the ganglia. Three
groups of them are represented in this figure, two being near the largest
trachea and one by the smaller trachea. These fine particles may be
either the precipitate resulting from fumes that had passed through the
tracheal walls, or that from some other source. There can be no doubt
about the large particles of the precipitate, because they are never found
in aphids used as controls.

While it is easy to find tracheae and precipitate in the ganglia, it is
quite difficult to find them in the brain. This seems to be due chiefly to
the absence of the larger tracheal branches in the brain. Plate 3, figure
H, shows a small tracheal branch in the brain cut crosswise, containing
three particles of the precipitate, and there are a few more scattered in
the adjacent brain. Plate 3, figure G, shows a small tracheal branch
running into an optical lobe, containing a few particles of the precipitate.

A critical study of any given tissue would certainly show that it con¬
tains as much precipitate as found in the nerve tissue; but no other tissue
was thus studied, because all the evidence indicated from the outset that
nicotine kills insects by- paralysis. One more illustration from the aphid
may be used to show that the precipitate may also be found in tracheae
aerating other tissues. Thus, Plate 3, figure I, represents a tracheal
branch containing the precipitate running between two ovaries.

112

Journal of Agricultural Research

Vol. VII, No. 3

Two of the recently emerged house flies that had been fumigated were
fixed with the mixture of absolute alcohol and phosphomolybdic acid,
and were sectioned. Most of the sections were not cleared well and con¬
sequently are not reliable for a study of this kind, but a feW of them are
fairly reliable. Plate 3, figure M, represents a spiracle and a portion of a
trachea taken from the abdomen of a fly. The neck of the spiracle was
almost closed with the precipitate ( pr ), while scattered particles of it
were seen along the walls of the trachea. Plate 3, figure L, represents
two medium-sized tracheae ( tr ) and a large fat cell (/c), taken from another
section through the abdomen, showing fine particles of precipitate (pr)
inside the tracheae- and in the fat cell outside the nucleus.

In conclusion, under this head a few more remarks may be made. A
ganglion is composed usually of two more or less round or oblong halves
which are securely united to one another. The outer or cortical layer is
cellular, while the center of each half never shows definite cell walls or
nuclei like those in the cortical layer. There is also usually a difference
in coloration between these two portions after being stained, although
this difference in the aphids stained with safranin was scarcely noticeable,
and the cortical layer was not cellular. This was true not only for those
that had been fumigated, but also for the controls that had been fixed
and stained the same way. In the illustrations the two portions are dis¬
tinguished by a difference in stippling. At no time was any anatomical
change observed in any insect that could actually be attributed to the
effect of nicotine. The failure to see such changes, if they existed, is not
significant, because the physical changes effected by the fixation probably
mask the smaller physical changes brought about by the nicotine. In
the higher animals, however, it has been observed that nicotine causes
slight anatomical changes in the cortical layer of the brain.

In cross sections of caterpillars the tracheae are easily traced into the
ganglia. Most of them penetrate the neurilemma and pass between the
two halves, where they ramify considerably by sending minute branches
through both portions of a ganglion. Occasionally a small tracheal
branch may enter a ganglion near or at the base of a nerve. The ramifi¬
cations of tracheae inside the ganglia of aphids are not so easily observed,
but they seem to exist, although perhaps not so abundantly.

All the preceding histological work has shown that nicotine spray solu¬
tions, and even nicotine used as a fumigant, do not penetrate chi tin. To
determine whether pure nicotine, undiluted, is able to penetrate chitin,
larvae of house flies, of the lesser wax moths, and aphids (Aphis rumicis)
were submerged in this fluid for 35 minutes. A study of the sections
made from these insects showed that the nicotine had passed into the
newly-formed chitinous wails of the tracheae in the larvae of the house
flies and wax moths, but had not passed all the way through them. Plate'
3, figures J and K, representing cross sections of small tracheae, well illus-

Oct. 16, 1916

Effects of Nicotine as an Insecticide

trate this point. These tracheal walls certainly were not much harder
than the other tissues, because they stained more deeply than did the
older chi tin. Plate 3, figure O, represents another trachea of the wax
moth, showing an older chitinous fracheal wall; the nicotine did not pass
into this wall. Plate 3, figure P, represents a small portion of the integu¬
ment (ini) of an aphid, showing that the pure nicotine did not pass into
the chitin. Plate 3, figures N and Q, illustrates how well Camoy's fluid
penetrates hard chitin. The black dots in the illustrations (pr) are crys¬
tals of mercuric chlorid that have remained after the fixative was removed.
Plate 3, figure N, represents a trachea from a wax-moth larva, and Plate
3, figure Q, a portion of the integument (inf) and fat cells (fc ) from an
aphid. Attention is called to the mechanical or physical changes brought
about in the fat cells by the fixation. Live fat cells never have a netlike
appearance, but appear more or less granular, and usually# contain many
globules. Sections from other insects that had been fixed in Carnoy ’s fluid
showed better than does figure Q the ability of this fluid to penetrate the
integument. In a few cases the crystals lie in rows penetrating the integu¬
ment, indicating that the fluid had passed through the chitin in streams.

If the most important results recorded under this large heading are
briefly summarized, the following conclusions may be drawn: (1) Nico¬
tine spray solutions neither enter the spiracles nor pass through the
integuments of insects; (2) nicotine as a stomach poison seems to be
distributed to all the tissues, including the nervous system; (3) nearly
all the nicotine fumes that strike the integuments and pass into the
tracheae are immediately condensed, so that in regard to nicotine as a
fumigant the integuments and tracheal walls are more or less covered
with fine spray; (4) this fine spray is well distributed through the
many small tracheal branches to all the tissues, where some of it passes
into the cells; (5) the nervous system receives its quota of the fine spray
and vapors from the spray, which immediately paralyzes the nerve cells;
(6) the statement just preceding explains how odoriferous particles and
vapors from nicotine spray solutions kill insects by paralysis.

HISTORICAL REVIEW

After making a few remarks concerning the chemistry and properties
of nicotine, a brief review pertaining to the pharmacological effects of
nicotine on various classes of animals, and a few other observations by
the writer will be given in order that these results may be compared ^ith
those obtained on the insects discussed in the preceding pages.

(l) GENERAL REMARKS ABOUT NICOTINE

Nicotine (C10H14N2) was conclusively prepared synthetically by Pictet
and Rotschy (19) in 1904. This investigation concluded a long series
of works pertaining to the structure of this deadly poisonous alkaloid.
These authors showed that it is a pyridin-methyl-pyrrolidin.

Journal of Agricultural Research

Vol. VII, No. 3

114

Blyth (3, p. 271-272) says that nicotine —

When pure , is an oily, colorless fluid , of 1 . 01 1 1 specific gravity at 1 5 0 . It evaporates
under ioo° in white clouds, and boils at 240°, at which temperature it partly distils
over unchanged, and is partly decomposed — a strong resinous product remaining.
... It has a strong alkaline reaction . . . and a sharp caustic taste. It absorbs
water exposed to the air, and dissolves in water in all proportions, partly separating
from such solution on the addition of a caustic alkali.

The aqueous solution acts in many respects like ammonia, saturating
acids fully; and by the action of light pure nicotine soon becomes yellow,
then brown and thick, in which state it leaves, on evaporation, a brown
resinous substance.

(2) PHARMACOLOGICAL EFFECTS OF NICOTINE ON VARIOUS CLASSES OF

ANIMALS

Greenwood *(9) experimented on certain protozoa, coelenterates, the
earthworm, certain echinoderms, crustaceans, and on certain mollusks
by using nicotine. He found that the toxic effect of this alkaloid on any
organism is determined mainly by the degree of development of the
nervous system. Thus, for the protozoa that he used it can not be re¬
garded as exciting or paralyzing, but is rather inimical to continued
healthy life. He states (p. 604) that —

As soon as any structural complexity is reached the action of nicotin is discriminat¬
ing, and discriminating in such a fashion that the nervous actions which are the
expression of automatism — which imply coordination of impulse — are stopped first.
This is seen dimly in Hydra, and it is more pronounced among the medusae, where
spontaneity, irradiation of impulse and direct motor activity are affected successively.

He asserts that relative to the higher invertebrates the paralyzing
action of nicotine is preceded by a phase of stimulation; and as the
positively exciting action becomes noticeable, nicotine becomes more and
more a medium in which life is impossible. He found that animals
closely allied structurally may also often behave quite differently toward
nicotine.

The present writer carried on one preliminary experiment to ascer¬
tain the action of nicotine on the lower invertebrates. A piece of scum
containing many paramecia and nematodes was placed on a slide under
a cover glass. A drop of pure nicotine was then placed at the edge of
the cover glass and the following results were observed. The nicotine
gradually passed under the cover glass by mixing with the water, and as
quickly as it came in contact with the nematodes they began to
squirm vigorously, while the paramecia apparently were not affected. A
little later the nematodes formed themselves into spirals and lay ap¬
parently paresized; then suddenly the spirals unfolded. This kind of
behavior continued until the nematodes were no longer able to move.
By this time it was observed that the nicotine had passed into their
bodies, and later the tough cuticles were constricted and contained many

Oct. 16, 1916

Effects of Nicotine as an Insecticide

deep grooves. The paramecia were still alive when the nematodes be¬
came lifeless, but they finally died slowly and gradually and at no time
showed any reaction which could be attributed to a stimulation.

There are many papers dealing with the economic importance of
nicotine as an insecticide, but they contain nothing about the pharma¬
cological effects of nicotine and little about its physiological effects,
except that it is effective.

So far as known to the writer, only two authors have anything to say
about the pharmacological effects of nicotine on insects.

Del Guercio (10) sprayed silkworms with various dilutions of nicotine
and determined that this insecticide within a short time brings about con¬
vulsive movements in the caterpillars, causing them to fall from the plants
and resulting in death in most cases. He thinks that nicotine spray solu¬
tions affect insects by means of the vapors from the nicotine poisoning
them and that these vapors even in minute quantities cause irritation
and convulsive movements which result in death by total paralysis.
He made no histological study to ascertain what tissue is vitally affected,
and his view is based solely on the behavior of the caterpillars treated.

Shafer (20) ascertained that insects subjected to the vapors of nicotine
and other contact insecticides first pass through a stage of excitement,
then through a stage of depression in which the coordination of move¬
ments is uncertain, and finally through a stage in which there is total
loss of movement and sensibility. The last stage was followed more or
less rapidly by death. During the first stage the action of the heart was
increased and was irregular, then it became depressed, but the heart
action was one of the last visible signs of life to disappear. Secretions
were also observed to issue from the mouths. The value of the respira¬
tory ratio arose, showing that these vapors depress the activity of
oxygen absorption more than they do the ability of carbon-dioxid
excretion. Shafer fouhd that the insects used continued to give off
carbon dioxid when no oxygen was present to be taken up. Loeb (16)
cites similar experiments in which muscles deprived of oxygen continued
to give off carbon dioxid.

Before discussing the pharmacological effects of nicotine on the verte¬
brates, the physiological classification of this alkaloid as defined by toxi¬
cologists may be given. Blyth (3, p. 269-279) places nicotine in that
class of poisons affecting the nervous system which causes convulsive
movements and complex nervous phenomena. Kobert (12) places nico¬
tine in that class of poisons affecting the cerebrospinal system which is
able to kill without producing coarse anatomical changes. Brundage (4)
classifies nicotine as a neurotic which depresses the cerebrospinal system.

Blyth (3) says that small fish die within a few minutes from a milligram
of nicotine. They are first stimulated, then become less active, and are
rapidly paralyzed.

Journal of Agricultural Research

Vol. VII, No. 3

The successive stages of nicotine poisoning in the frog are briefly sum¬
marized by Langley and Dickinson (14) as follows: (1) Stage of excita¬
tion; (2) stage of spasms; (3) stage of quiescence; (4) stage of flacddity;
(5) stage of paralysis of the central nervous system; (6) stage of paralysis
of the motor-nerve endings.

Blyth (3, p. 273) says, “Birds. also show tetanic convulsions, followed
by paralysis and speedy death,” and Sollmann (23, p. 262) asserts that —

Nicotine is one of the most fatal and rapid of poisons; the vapor arising from a glass
rod moistened with it and brought near the beak of a small bird causes it to drop dead
at once, and two drops placed on the gums of a dog may cause a similar result.

According to Langley and Dickinson (14), the symptoms of nicotine
poisoning in rabbits, cats, and dogs are in a general way similar, and may
be briefly described as follows: There is a preliminary excitement; clonic
spasms; twitchings of the muscles in various parts of the body; stimula¬
tion of the central nervous system ; paralysis of the motor-nerve endings
in the skeletal muscles; quickening and deepening of the respiration,
followed by slowing and cessation; dilation of the pupils; paralysis of
the cervical sympathetic system; rise and fall of the blood pressure; rise
of temperature ; constriction of intestines, followed by dilation, and slight
vomiting in cats and dogs. If the doses are sufficiently large, the cerebro¬
spinal system is totally paralyzed.

According to Blyth (3), the symptoms witnessed in mammals poisoned
by nicotine are quite similar. With large doses, there is a cry, one or two
shuddering convulsions, and death; with smaller doses, there is trembling
of the limbs, excretion of feces and urine, stupor, a staggering gait, and
then the animal falls on one side. One or two drops of pure nicotine may
kill a rabbit, cat, or dog within five minutes. Vas (24) found that the
substance resulting after washing tobacco smoke affects the health of
rabbits; they lose weight, the number of blood corpuscles is decreased,
and the hemoglobin of the blood is diminished. According to Blyth,
nicotine also affects horses similarly to the smaller domestic animals.

Blyth says that Dragendorff ascertained that nicotine is absorbed into
the blood and is excreted unchanged, in part by the kidneys and in part
by the salivary glands.

Krocker (13) was among the first investigators to determine the phar¬
macological effects of nicotine on man. He found that it paralyzes the
nervous system and that death is caused by the rapid benumbing and
paralysis of the respiratory center, but not from heart paralysis, although
nicotine powerfully influences the action of the heart.

Holland (11) states that two or three drops of the alkaloid is fatally
poisonous to man when taken into the stomach, and that death is caused
by heart failure. In this latter statement other authorities do not agree
with him, for they say that death is due to asphyxia, on account of the
paralysis of the respiratory center. In the lower vertebrates the heart
still beats some time after life is extinct.

Oct. 16, 1916

Effects of Nicotine as an Insecticide

117

Sollmann (23) summarizes the symptoms of nicotine poisoning on man
as follows : The whole cerebrospinal axis is first stimulated, then depressed
from above downward; symptoms from large doses resemble those of
asphyxia or hydrocyanic acid. Action on the medullary centers is marked
and violent; respiration is at first increased, then markedly depressed;
paralysis of the respiratory center is the cause of death. Action on the
spinal cord consists in strong stimulation of the motor cells, producing
convulsions, passage of feces and urine. Nicotine acts on unstriped
muscles, paralyzing the ganglia after a brief stimulation. There is nausea
and vomiting, violent peristalsis, and even tetanic contraction of the
intestine and diarrhea. The respiration, heart strength, and blood pres¬
sure are increased; the heart rate is decreased. A strong nicotine solu¬
tion applied directly paralyzes the nerve fibers. “ Free nicotine is caustic
on account of its alkalinity.” The fatal dose for man is about 60 mgm. ;
one cigar contains enough nicotine to kill two persons, if it were directly
injected into the circulation. “It acts with a swiftness only equaled by
hydrocyanic acid.”

Cushny (6, p. 304-314) asserts that poisonous doses of nicotine adminis¬
tered to man or other mammals cause a hot, burning sensation in the
mouth which spreads down the esophagus to the stomach and is followed
by salivation, nausea, vomiting, and sometimes purging. Mental confu¬
sion, muscular weakness, giddiness, and restlessness are followed by loss
of coordination and partial or complete unconsciousness. Clonic con¬
vulsions set in later and eventually a tetanic spasm closes the scene
by arresting the respiration.

Autenrieth (1, p. 87) says that nicotine is absorbed from the tongue,
eye, and rectum within a few seconds, but from the stomach somewhat
more slowly. Its absorption is also possible from the outer skin, and it is
eliminated through the lungs and kidneys.

In concentrated form nicotine is a local irricant, though owing to the rapidity of
its toxic action, it does not behave like a true corrosive nor does it cause inflammation
of the mucous lining of the stomach after a lethal dose.

(3) PHYSICAL AND CHEMICAL EXPECTS ON THE} CHhhS

In the preceding discussion it has been pointed out that nicotine paral¬
yzes the respiratory center in the brain of vertebrates, causing death by
asphyxiation. This implies that, while the nervous system is benumbed
and rendered inactive, the lungs are prevented from functioning, and con¬
sequently the cells in the tissue die for want of oxygen. Since the in¬
vertebrates are differently organized, particularly in regard to their
respiratory system, an investigation will be made as to the period insects
can live without free oxygen.

Walling (25) states that grasshoppers confined in pure carbon dioxid
for 1 5 hours recover, and Shafer (20) determined that beetles ( Passalus

Journal of Agricultural Research

Vol. VII, No. 3

118

cornutus I?ab.) confined for 24 hours in pure hydrogen completely recover
when placed in fresh air. These experiments indicate that nicotine does
not kill insects merely on account of the paralysis of the respiratory cen¬
ters, because the tracheae and tissues of an insect contain enough oxygen to
keep the cells alive for several hours. Since the cells of insects are con¬
stantly surrounded by air containing oxygen, an investigation will be
made as to whether or not nicotine interferes with oxidation in the cells
and whether it kills physically or chemically.

Greenwood (9) observed that when simple animals die from the effects
of nicotine, death is often associated with injury to the cell contents so
that they tend to disintegrate. This is shown in the protozoan, Actino -
sphaerium sp., and in the coelenterates, Hydra spp. and Medusa spp.

Budgett (5) treated infusorians with a number of poisons, including
nicotine, and found that these protozoa become strongly vacuolated and
finally the membranes burst, allowing the protoplasm to flow out into the
water. The same structural changes occurred when he deprived them of
oxygen. He says (p. 214): “This indicates that either these poisons
prevent oxidation or that lack of oxygen produces toxic substances.”
He also believes that these poisons not only reduce the normal resistance
to the entrance of water but lead to the taking up of water, probably by
hastening the molecular breakdown and so increasing the osmotic pres¬
sure within the cell.

Toeb (16) and others have experimented extensively with amebae and
paramecia by depriving them of oxygen. They always observed the
same structural changes as already cited from Budgett. Loeb also
performed many experiments by depriving the eggs of a certain fish of
oxygen. He exposed the eggs to a current of hydrogen and observed —

The liquefaction of the cell walls and the formation of droplets began when the egg
was in the 8-cell stage (Fig. 2). These droplets fuse into larger drops and finally
nothing but these drops indicates the existence of the germinal disk.

The present writer placed living fat cells and cenocytes of the honeybee
on a slide in water under a cover glass. These cells live in tap water for
some time before any changes in their appearance can be observed;
but when a drop of pure nicotine is placed at the edge of the cover glass,
changes in their general appearance take place soon afterwards. The
globules in the cells sometimes dance about, resembling the Brownian
movement. The globules in the fat cells usually soon lose their rotun¬
dity, become massed together, and form a coarse, granular structure.
The refractive bodies in the cenocytes soon disappear, and then these
cells become opaque. After considerable time the cell walls of the fat
cells and cenocytes burst, and the cell contents disintegrate.

In the preceding pages it is shown that either lack of oxygen or the
presence of nicotine around simple animals brings about structural
changes resulting in death. A comparison, although a rough one, might

Oct. 16, i9x6 Effects of Nicotine as an Insecticide 1 19

also be made between the fish eggs used by Loeb and the fat cells and
cenocytes employed by the present writer. All these facts seem to indi¬
cate that either lack of oxygen or the presence of nicotine around the cells
kills physically rather than chemically. The following paragraphs will
considerably strengthen this statement.

It is well known that different fluids have different osmotic pressures, *
and this is also true for the blood of different animals. In order that
tissues removed from various animals might be kept alive for some time,
Lewis (15) has shown that they must be placed in fluids having different
osmotic pressures. Endeavoring to make a fluid having an osmotic
pressure equal to that of the blood in a grasshopper, Lewis used sea
water, distilled water, grasshopper bouillon, sodium bicarbonate, and
dextrose. The effect of osmotic pressure on cells is best illustrated by
using red corpuscles. According to Cushny (6, p. 304-314), water
passes into these cells readily and when placed into distilled water they
swell up and burst, but when placed into an aqueous solution of sodium
chlorid having an osmotic pressure greater than that of their contents,
they shrink because the contained salts are unable to retain water
against a higher concentration outside. A change brought about in
the osmotic pressure of the blood might be a probable explanation of
the death of the honeybees recently fed various salts by the present
writer (18).

There are several theories regarding the manner in which drugs and
powerful poisons affect the cells. Cushny has briefly summarized
them about as follows: (1) Some drugs enter into definite chemical com¬
binations with the constituent protoplasm; (2) some drugs act on the
cells by changing the relation of the cell constituents in which they
are dissolved; (3) some drugs alter the surface tension of the cells in
relation to the surrounding fluids; (4) a few powerful drugs may act by
altering the surfaces of the cells without penetrating into the interior;
(5) many drugs may change the intracellular membranes; and (6) other
drugs may reduce the permeability of the cellular membranes by altering
their electric charges. Cushny says (6) :

From the present confusion the only legitimate conclusion seems to be that the
activity of drugs depends on a large variety of factors and that pharmacological action
can not be brought under any one law, either chemical or physical.

Shafer (21) has added another view which should be classified with
the chemical ones, for it deals with the enzym-like cell constituents
which accomplish oxidation. He thinks that contact insecticides,
nicotine included, deleteriously affect the activities of the reductases,
catalases, and oxidases in an unequal degree, thereby disturbing the
natural or normal balance of the activities of these enzym-like factors.

120

Journal of Agricultural Research

Vol. VII, No. 3

SUMMARY .

(1) Nicotine spray solutions do not pass into the tracheae, nor do they
penetrate the integuments of insects.

(2) The fumes from nicotine used as a fumigant, the vapors from
nicotine spray solutions, and the odoriferous particles from evaporated
nicotine spray solutions or from powdered tobacco pass into the tracheae
and are widely distributed to all the tissues.

(3) Regardless of how it is applied, whenever nicotine kills insects,
as well as all other animals, it kills by paralysis, which in insects travels
along the ventral nerve cord from the abdomen to the brain.

(4) The writer does not know just how nicotine paralyzes the nervous
system, but he does know that it prevents the nerve cells from function¬
ing, and that in regard to the simplest animals its presence around the cells
causes the same structural changes resulting in death as observed when
other animals of the same kind are deprived of oxygen. In such cases
it seems to kill physically rather than chemically, but the evidence
presented does not conclusively prove this view. In the higher animals
it may kill by interfering with oxidation in the cells; whether this is
accomplished physically or chemically the writer does not know, but con¬
cluding from the properties of nicotine he is inclined to attribute more
to its physical effects than to its chemical effects.

LITERATURE CITED

(1) AutEnriEth, Wilhelm.

1915. Laboratory Manual for the Detection of Poisons and Powerful Drugs.
Ed. 4, translated by W. H, Warren. 320 p., 25 fig. Philadelphia.

(2) Baker, A. C.

1915. The woolly apple aphis. U. S. Dept. Agr. Office Sec. Rpt. 101, 55 p.,
3 fig., 15 pi. Literature referred to in the text, p. 53-55.

(3) Beyth, A. W.

1895. Poisons: their Effects and Detection. Ed. 3, 724 p., illus. London.

(4) Brundage, A. H.

1910. A Manual of Toxicology ... Ed. 7, 428 p., illus., pi. New York, Lon¬
don.

(5) BudgETT, S. P.

1898. On the similarity of structural changes produced by lack of oxygen and

certain poisons. In Amer. Jour. Physiol., v. 1, no. 2, p. 2 10-214,

9 fig-

(6) Cushny, A. R.

1915. A Text- Book of Pharmacology and Therapeutics ... Ed. 6, 708 p., 70
fig. Philadelphia, New York.

(7) DEwitz, J.

1912. The bearing of physiology on economic entomology. In Bui. Ent.
Research, v. 3, pt. 4, p. 343-354. Bibliography, p. 352-354.

(8) Fischer, Alfred.

1899. Fixirung, Farbung und Bau des Protoplasmas. 362 p., 21 fig., 1 col. pi.

Jena, Literatur, p. 341-348.

Oct. i6, 1916

Effects of Nicotine as an Insecticide

121

(9) Greenwood, M.

1890. On the action of nicotin upon certain invertebrates. In Jour. Physiol.,
v. 11, suppl. no., p. 573-605.

(10) GuERCio, Giacomo del.

1900. Sul potere mortifero dei liquidi alia nicotina e suir uso di essinella
distruzione degli insetti. In Nuove Relaz. R. Staz. Ent. Agr.
Firenze, s. 1, no. 3, p. 124-134.

(11) Holland, J. W.

1905. A Text-Book of Medical Chemistry and Toxicology. 592 p., 108 fig.,

8 pi. Philadelphia, London.

(12) Robert, Rudolf.

1906. Lehrbuch der Intoxikationen. Aufl. 2, Bd. 2. Stuttgart.

(13) KrockEr, Arthur.

1868. Ueber die Wirkung des Nikotins auf den thierischen Organismus.
39 p. Berlin.

(14) Langley, J. N., and Dickinson, W. L.

1890. Pituri and nicotin. In Jour. Physiol., v. 11, no. 4/5, p. 265-306.
Papers quoted, p. 304-306.

(15) Lewis, M. R.

1916. Sea water as a medium for tissue cultures. In Anat. Rec., v. 10, no. 4,
p. 287-299, illus. Literature cited, p. 299.

(16) Loeb, Jacques.

1906. The Dynamics of Living Matter. 233 p., 64 fig. New York.

(17) McIndoo, N. E.

1914. The olfactory sense of the honey bee. In Jour. Expt. Zool. , v. 16, no. 3,

p. 265-346, 24 fig. Literature cited, p. 345-346.

(18) —

1916. The sense organs on the mouth-parts of the honey bee. Smithsn. Misc.
Collect., v. 65, no. 14, 55 p., 10 fig. Literature cited, p. 52-54.

(19) Pictet, Am6, and Rotschy, A.

1904. Synthese des Nicotins. In Ber. Deut. Chem. Gesell., Jahrg. 37, Bd. 2,
p. 1225-1235.

(20) Sharer, G. D.

1911. How contact insecticides kill. I-II. Mich. Agr. Exp. Sta. Tech. Bui.
11, 65 p., 7 fig., 2 pi.

(21) —

1915. How contact insecticides kill. III. Mich. Agr. Exp. Sta. Tech. Bui.

21, 67 p., 3 figs., 1 pi.

(22) Snodgrass, R. E.

1910. The anatomy of the honey bee. U. S. Dept. Agr. Bur. Ent. Tech. Ser.
18, 162 p., 57 fig. Bibliography, p. 148-150.

(23) Sollmann, Torald.

1908. A Text- Book of Pharmacology . . . Ed. 2 , 1070 p.,127 fig. Philadelphia,
London. Bibliographic register, p. 1029-1042.

(24) Vas, Friedrich.

1894. Zur Kenntniss der chronischen Nikotin- und Alkoholvergiftung. In
Arch. Expt. Path. u. Pharmakol., Bd. 33, Heft 2/3, p. 141-154.

(25) Walling, Eulalia V.

1906. The influences of gases and temperature on the cardiac and respiratory
movements in the grasshopper. In Jour. Expt. Zool., v. 3, no. 4, p.
62 1-629.

55857°— 16— 3

PLATE 1

Fig. A. — Portion of the large longitudinal trachea of the house fly cut crosswise
obliquely, showing the carmine acid “ precipitate' * pr. X190. The fly had been
submerged for one hour in a pure nicotine solution (1 : 500) colored with carmine
acid.

Fig. B. — Combination drawing from two consecutive sections of a green peach
aphis, showing the indigo-carmine " precipitate ** pr in a trachea ir. X500. The
aphis had been submerged for 45 minutes in a pure nicotine solution (1 : 500) col¬
ored with indigo-carmine.

Fig. C. — Cross section of a large longitudinal trachea of larva of lesser wax moth,
showing the indigo-carmine “ precipitate” pr adhering to the tracheal wall trw.
X190. The larva had been submerged for 30 minutes in a pure nicotine solution
(1:500) colored with indigo-carmine.

Fig. D, — Longitudinal section of one of the smallest tracheae of the same larva
as in figure C; same treatment and same enlargement.

Fig. E. — Longitudinal section of a large trachea and one of its branches of a coccid,
showing the ' ‘ precipitate ” pr resulting from the union of pure nicotine and phospho-
molybdic acid. X500. The coccid had been submerged for 30 minutes in a pure
nicotine solution (1 : 100).

Fig. F. — Portion of a cross section of an aphid ( Aphis rumicis), showing the indigo-
carmine “ precipitate” pr in a spiracle sp. X320. The aphid had been heavily
sprayed with a pure nicotine solution (1 : 500) colored with indigo-carmine.

Fig. G. — Portion of a cross section of the same aphid as in figure F, showing no
precipitate in the trachea tr, but much on the outside of the integument int. X 320.

Fig. H-O. — Longitudinal sections of spiracles sp with connecting tracheae tr ,
showing how it is practically impossible for aqueous spray solutions to enter spiracles,
owing to hairs hr, a closing plate p, and a peculiar arrangement of rims r at mouths
of spiracles.

Fig. H. — Spiracle of a coccid ( Orthezia insignis). X 500.

Fig. I. — Spiracle of a caterpillar of Atteva aurea. X 190.

Fig. J. — Spiracle of a larva of lesser wax moth ( Achroia grisella). X 190.

Fig. K. — Spiracle of a caterpillar of Datana sp. X 500.

Fig. L. — Spiracle of a caterpillar of a catalpa sphinx ( Ceratomia catalpae.)
X 500.

Fig. M. — Spiracle of a larva of a Colorado potato beetle (Leptinoiarsa
decemlineata). X 500.

Fig. N. — Spiracle of fall webworm (caterpillar of Hyphantria cunea). X 320.

Fig. O. — Spiracle of the tomato worm (larva of Phlegethontius sexta ), show¬
ing the closing plate p. X 50.

(122)

PLATE 2

Fig. A to J. — Cross sections of portions of the alimentary canals and Malpighian
tubtiles of worker honeybees, showing “precipitated” indigo-carmine that had been
fed with pure nicotine and honey to bees three days before they were fixed in abso¬
lute alcohol. Owing to poor fixation, most of cells are drawn diagrammatically.

Fig. A. — Portion of the wall of the ventriculus, showing the “precipitate” pr
in inner ends of the epithelial cells ep. X320.

Fig. B. — Portion of the wall of the ventriculus, showing the “precipitate”
pr in the middle of the epithelial cells ep. X320.

Fig. C. — Portion of the wall of the ventriculus, showing the “precipitate”
in the outer ends of the epithelial cells ep and in the transverse muscle layer
tm. X320.

Fig. D. — Portion of the wall of the honey stomach joining the proven-
triculus, showing the “precipitate” pr in the chitinous chi and muscular m
layers. X320.

Fig. E. — Portion of the wall of the anterior part of the valve of the proven-
triculus, showing the tl precipitate ” pr in muscles m , tracheae tr, and epithelial
cells ep. X320.

Fig. F. — Section through the small intestine, showing the “precipitate” pr
in the center of the lumen l and lining epithelium ep, but none in the walls of
this organ nor in the Malpighian tubules mal by it. X 50.

Fig. G. — Section through two Malpighian tubules mal against the ventriculus,
showing the “ precipitate ” pr in their cells and lumens l. X320.

Fig. H. — Section through two Malpighian tubules mal near the ventricu¬
lus, tracheal branch tr and blood bl, showing the “precipitate” pr in these
tissues. X320.

Fig. I. — Section of one-third of the rectum in a compressed condition, showing
the “precipitate” pr in the lumen /, but none in the chitinous layer chi, nor in
the rectal glands rgl, nor in the muscular m layer. X 50.

Fig. J. — Section through the middle of the ventriculus, showing the distri¬
bution of the “ precipitate ” pr in the lumen /, between the peritrophic mem¬
branes pm, in the epithelial ep and muscular m layers of the ventriculus
and in the Malpighian tubules mal. X 50. The walls of the ventriculus were
drawn diagrammatically.

PLATE 3

Fig. A to I, L, M. — Drawings and diagrams representing the distribution of pre¬
cipitate resulting from the fumes of 40 per cent nicotine sulphate and phosphomolybdic
acid. Figures A, B, F to I are from the same green peach aphis, and figures L and M
are from the same house fly that had been fumigated. X 320. Figures Dand E are
diagrams, after Baker (2), representing the respiratory system of an aphid, Eriosoma
lanigerum.

Fig. A. — Transverse-longitudinal section of a trachea of an aphid, showing the
precipitate pr inside the trachea and in fat cells fc near by.

Fig. B. — Cross section of a portion of an aphid just molting, showing the
precipitate pr on the outer surfaces of the old intx and the new integuments int
and between them, but none in the fat cells fc.

Fig. C. — Combination drawing from six consecutive sections through thoracic
ganglion of an aphid, showing the precipitate pr in three tracheal branches tr
in the cortical layer cl and in the inner layer il of a ganglion.

Fig. D. — Diagram of the dorsal tracheal system of an aphid, showing the
dorsal trunk dt and dorsal arch da.

Fig. E. — Diagram of the ventral tracheal system of an aphid, showing the
anterior ventral arch ava, posterior ventral arch pva, and the ventral trunk vt.

Fig. F. — Combination drawing from five consecutive sections through the
thorax of an aphid, showing the precipitate pr on the outer surface of the
integument int, in the tracheae tr and in the subesophageal ganglion sg.

Fig. G. — Portion of cross section of an optic lobe of an aphid, showing the
precipitate pr inside and outside a tracheal branch tr.

Fig. H. — Cross section of the brain br and optic lobes opl of an aphid, showing
the precipitate pr in the tracheal branch tr and in the cortical’ layer of the
brain.

Fig. I. — Cross section of two ovaries ov of an aphid, showing the tracheal
branch tr containing precipitate passing between them.

Fig. L. — Cross section of two trachese tr and a fat cell/c of a house fly, showing
the precipitate pr in the tracheae and in the fat cell outside its nucleus.

Fig. M. — Longitudinal section through a spiracle sp and its connecting
trachea tr of a house fly, showing the precipitate pr in the neck of the spiracle
and along the tracheal wall.

Fig. J and K. — Cross sections of the small tracheae, showing the precipitate pr in
newly formed tracheal walls trw resulting from the union of pure nicotine and phospho¬
molybdic acid. These and other insects had been submerged in pure nicotine for 35
minutes. X 190. Figure J is from a house-fly larva and figure K from a lesser wax-
moth larva.

Fig. N and Q. — Cross sections, showing how well Camoy’s fluid passes through
hard cliitin, as indicated by remaining crystals pr of mercuric chlorid. Figure N is a
trachea from a lesser wax-moth larva. X 190. Figure Q shows a portion of the
integument int and the fat cells fc of an aphid, also showing the physical change in the
fat cells caused by a fixative. X 500.

Fig. O. — Cross section of a medium-sized trachea of a lesser wax-moth larva, showing
that pure nicotine did not pass into an older tracheal wall trw under the same condi¬
tions as stated for figure K. X 190.

Fig. P. — Cross section of portion of the integument int of an aphid ( Aphis rumicis),
showing that pure nicotine did not pass into chitin under same conditions as stated for
figures J, K, and O. X 500.

Effects of Nicotine

ACIDITY AND ADSORPTION IN SOILS AS MEASURED
BY THE HYDROGEN ELECTRODE

By L. T. Sharp and D. R. Hoagland,

Assistant Chemists, Agricultural Experiment Station of the University of California 1

INTRODUCTION

The problem of soil reaction has come to occupy an increasingly
important position in the realm of soil-fertility studies. This is evident
from the numerous papers recently appearing upon the subject of soil
acidity. The great diversity of opinion concerning the nature and
methods of measuring soil acidity has served thus far only to confuse the
matter. By adopting modern methods capable of measuring specifically
the hydrogen ion,2 data have been obtained by the writers which seem
to offer a clearer conception of the question.

Before considering the question of soil acidity, it seems desirable to
recall the fundamental significance of the terms “neutrality,” “acidity,”
and “alkalinity.” The dissociation of pure water produces H ions and
OH ions in equal concentration, denoting neutrality. The product of
the concentrations of these ions in any solution is a constant, approxi¬
mately i X io~14. When the H ions are present in a concentration greater
than the OH ions — that is, in a concentration greater than 1X10"7,
then the resulting solution is acid. Conversely, the presence of OH ions
in greater concentration than i X io-7 gives an alkaline solution. The
term “acidity” is often construed to mean the total quantity of H ions
which may be produced when the equilibrium is continually shifted by
the introduction of OH-ions. This total acidity is referred to as potential
acidity, while the H-ion concentration at any given moment determines
the intensity of acidity, to use Gillespie’s (13) 8 expression. In all
probability excessive concentrations of H or OH ions in soil solutions
exercise pronounced effects on plant growth and on the activities of
soil bacteria.

Potential acidity or alkalinity may be due to undissolved substances
or to soluble compounds only partly hydrolyzed or dissociated. An
obvious illustration of this is afforded by the titration of a suspension of
calcium carbonate in water. So long as solid calcium carbonate remains,

1 From the Divisions of Soil Chemistry and Bacteriology and Agricultural Chemistry in equal cooperation.
Acknowledgment is made to Mr. C. L. A. Schmidt for valuable suggestion during the progress of this work.

2 To be more exact, according to recent physical-chemical views, it is the activity of the H ion, rather
than the concentration, which is measured here. However, for the present purpose this distinction seems
unimportant.

8 Reference is made by number to “ Literature cited/’ p. 143-145.

(123)

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D, C.

Vol. VII, No. 3
Oct. 16, 1915

Cal.— 5

124

Journal of Agricultural Research

Vol. VII* No. 3

there will be a definite concentration of H and OH ions, while with the
addition of acid the reaction of H+ + 0H“ = H30 occurs and equilibrium
is restored by the solution of more calcium carbonate.

Soil acidity should not be set apart and considered as a phenomenon
unrelated to the ordinary concepts of acidity. The use of such terms
as “apparent acidity,” “real acidity/' “adsorption acidity” has led to a
confusion of ideas, and, hence, should be discarded. With this in view,
it seems desirable to emphasize the generally neglected point of view
that the equilibria resulting in the production of H ions are in their
essential nature the same for soil solutions as for other -simpler chemical
systems.

The methods now in vogue for estimating the lime requirement (soil
acidity) have often proved to be unsatisfactory and in any case do not
measure the intensity of acidity — that is, the H-ion concentration. In
order to secure reliable data upon these points, the method of determining
the H-ion concentration by means of a hydrogen electrode was used.
Some preliminary experiments led to the adoption of a modification of
Hildebrand's (17) apparatus, which will be described in detail later.
Few similar measurements with soils have been previously recorded.
The work of Saidel (29) was limited to a few alkaline soil extracts
whose reaction had been changed by boiling. For reasons discussed
later, data obtained on soil extracts are not convincing. Fischer (12)
also made a few hydrogen-ion determinations with soil. In a more
noteworthy investigation, Gillespie (13) has measured the H-ion concen¬
tration of 22 soils in water suspensions. The investigations reported in
this paper were begun before the appearance of Gillespie’s article. The
present authors have extended the work into various other phases, such
as titration with bases as a meafis of adjusting the soil reaction and of
studying the general phenomena grouped under the term “adsorption.”
Moreover, various factors affecting the H-ion concentration of soil
suspensions were also considered.

VARIOUS POINTS INVESTIGATED

In the course of these investigations several related questions were
studied, in addition to the determinations of the H-ion concentration in
various soil suspensions and in soil extracts. Experiments were under¬
taken in order to obtain further evidence regarding the influence of
varying proportions of soil to water, grinding the soil, heating it at vari¬
ous temperatures, and of the addition of salts on the H-ion concentration.
Consideration has also been given to the relation of HCO/, CO/, and
C02 to soil reaction as measured by the electrometric method. Experi¬
mental data have likewise been secured with respect to the lime require¬
ment of soils and the so-called “adsorption of bases.” Finally, the
apparatus and methods of procedure are described.

Oct. 16, 1916

Acidity and Adsorption in Soils

125

EXPERIMENTAL WORK
MATERIALS USED

Before taking up the detailed discussion of the data it appears desirable
to describe briefly the materials used. Table I contains a list and
description of all the soils referred to in this paper.

Table) I. — Description of soils used in experimentation

Laboratory No.

Source of soil.

General type.

California .

Silty clay loam .

. do .

Fine- sandy loam .

. do .

Sandy loam .

. do .

Gravelly loam .

. do .

Clay adobe .

.... .do .

Fine sandy loam .

Wisconsin .

Silty clay loam .

.... .do .

Sandy loam .

Pennsylvania .

Silty loam .

California .

Fine sandy loam .

Louisiana. . .

Silty loam .

California .

Peat .

Wisconsin .

. do .

California .

Sandy loam .

. do .

Fine sandy loam .

3
4-

10.

13
14.

1 6

20,

24-

Remarks.

Alkali soil.

Very infertile.

Alkali soil.
Do.

HYDROGEN-ION CONCENTRATION OF SOIL SUSPENSIONS

Table II gives the amounts of soil and water used for making the
suspensions. The soils were air-dried and passed through a i-mm. sieve.
The H-ion concentration is expressed in the customary units of gram
molecules of H ion per liter.

In accordance with our preliminary statements, Table II gives evi¬
dence that soils may give rise to acid solutions — that is, solutions con¬
taining a preponderance of H ions over OH ions. This conception is in
agreement with the conclusions which Truog (33) drew from his zinc-
sulphid method. The work of Gillespie (13), paralleled by that pre¬
sented in Table II, conclusively proves that there may be an excess of H
ions in the solution bathing the soil particles.

Out of 22 soils examined by Gillespie (13), 17 were found to yield
acid solutions. The writers experimented with 9 add soils. The soils
were of widely different types and origin. These facts do not support

126

Journal of Agricultural Research

Vol. VII, No. 3

the views of Cameron (5) concerning soil acidity, that it is generally due
to the selective absorption of bases. They also are in opposition to the
recent conclusions of Rice (26) and Harris (15), that water-soluble
acids are not characteristic of acid soils.

Table II. — Hydrogen-ion concentrations of suspensions oj unground soil

Soil No.

Quantity
of soil.

Water.

Readings on
voltmeter.

H-ion concentra¬
tion (gram-mole¬
cules per liter).

Gm.

C. c.

I .

2. O

0. 763

O. 4Xio“7

2 .

2. O

3°

• 7 S3

. 6 X io“7

3 .

2. O

• 759

. 5X10-7

4 .

2. O

• 763

. 4X10-7

5 .

2. O

• 759

.5X10-7

2. O

• 753

. 6X10-7

7 .

2. O

. 761

. 5X10-7

8 .

2. O

• 7S2

. 6X10 7

2. O

* 752

. 6Xio“7

2. 0

. 742

. 9Xio“7

2. O

. 760

• SXio-7

2. O

• 77°

. 3 X io~7

2. O

• 550

. 2X10-3

2. O

• 753

. 6Xio~7

2. O

50.

a. 590

.4X10-4

15 .

*5-0

• 565

. 1 X io“3

. I

. 627

.9X10-5

2. O

a . 648

.4X10-5

S-o

. 642

• 5X10-*

10. 0

.623

. 1X10-4

15.0

.628

.9X10-5

17* .

. 1

. 619

. 1X10-4

17 . .

2. 0

a. 605

■ 2XlO~*

17 .

5-o

5°

• 596

• 3XIO-4

10. 0

. 582

• SXlO-4

i7- .

25.0

. 582

.5XIO-4

18 .

2. 0

. 704

.4X10^

18 .

75-o

. 710

•3XIO-4

*9 . .

2. 0

. 638

.6XIO-5

. 1

.638

. 6XIO”6

2. 0

a. 630

.8XIO-5

S-o

• 633

. 7X10-5

10. 0

.628

.9X10-*

i5-o

• 633

• 7XlO-4

5-o

100

.605

• 2XIO-4

2. 0

5°

. 605

. 2XlO~'4

23 .

5-o

• 87s

• SXio-»

5-o

.898

.2X10— «

a Average of several determinations.

It is also evident from the data presented in Table II that there is a
considerable range in the H-ion concentration of the soil suspensions.
A number of soils known to be fertile show strikingly similar reactions,
slightly alkaline, as indicated by an H-ion concentration of slightly less
than 1 X io“7. Alkali soils, presumably containing sodium carbonate,
show alkalinity corresponding to an H-ion concentration of 0.2 X io-9.

Oct. 16, 19x6

Acidity and Adsorption in Soils

127

On the other hand, certain of the soils gave an H-ion concentration as
high as 0.2 X io-3, which indicates considerable intensity of acidity.

Any ultimate conclusions regarding soil fertility must presuppose a
knowledge of the composition of the soil solution. At present our
methods do not enable us to study the soil solution itself, and conse¬
quently all deductions in regard to it must be purely inferential. In
order to approximate the conditions existing in the soil solution, the
ratio of water to soil in several cases was reduced as far as the method
would permit. An inspection of the data indicates comparatively insig¬
nificant changes in the H-ion concentrations when widely varying pro¬
portions of water to soil are used. Most of these small fluctuations can
be ascribed to the limitations of the apparatus. Hence, it is reasonable
to assume that the H-ion concentrations of the soil suspensions approxi¬
mate those of the soil solutions. This argument is theoretically sound,
since in all the larger proportions of soil employed it is probable that the
solution is saturated with respect to the acid-forming constituents.

RELATION OF HCOI, C07, AND C02 TO SOIL REACTIONS

At present it has been found impracticable to simulate exactly the
COs equilibria existing under field conditions. In the case of several
acid soils the partial saturation of the soil suspensions with C03 gas did
not alter the reaction appreciably. Hence, it is quite possible that the
C03 content of the soil solution may not materially modify the above
conclusions regarding the magnitude of the H-ion concentration in acid
soils. In other words, the soil acids, whether organic acids or acid sili¬
cates, are the chief factors determining the reaction.

On the contrary, the reaction of alkaline soils depends in large measure
upon the equilibria between C03 gaseous, C02 dissolved, C03 ion, and
HC03 ion.1 From these considerations it might be predicted that the
reaction of most alkaline soils is the resultant of the equilibria existing
between HC03, C03, Ca ions and dissolved C02 in contact with the C03
of the soil atmosphere. From the work of Cameron and Bell (6) and
Johnson (19) it seems proper to infer that HCOJ in this class of soils
largely determines the reaction.

The application of the electrometric method to solutions or suspen¬
sions whose reaction depends upon the HC03 ion requires great precau¬
tion to prevent the decomposition of HCOJ, with a loss of C02. This
reaction is slow, but there will be a gradual increase in the alkalinity of
the solution as a result of the production of COs ions. In the present
work the shaking method by Gillespie (13) was employed and the decom¬
position of HCOJ largely avoided , although the results may tend to be
slightly high.

1 For a detailed discussion of this matter the reader is referred to Cameron and Bell (6) and to
Johnson (19).

128

Journal of Agricultural Research

Vol. VII, No. 3

As a means of comparison the OH-ion concentration of solutions sat¬
urated with Ca(HC03)2 and CaC03 were determined with an exactly
similar technique. Ca(HCOs)2 gave a value for OH“ of 0.5 X io-7 and for
CaC03 0.3 Xio-9. The figure for Ca(HC03)2 is almost identical with
those obtained for the alkaline soils. This is in keeping with the theo¬
retical considerations advanced above. T. Saidel (29) determined the
OH-ion concentration of several soil extracts after prolonged boiling.
His values approximate that given by CaC03, which is distinctly higher
than the normal OH concentration for soils of this class.

At first attempts were made to estimate the H-ion concentration of
filtered soil extracts. Serious difficulty was encountered in this pro¬
cedure, in that the nitrates present were slowly reduced by the hydrogen
gas in contact with platinum black. Obviously this reduction of nitrates
would result in the production of NH3 and a residue of fixed alkali,
with a corresponding increase in the concentration of OH ions. While
the reduction of nitrates does occur to some extent in soil suspensions,
yet in this case the results are not appreciably changed. This is accounted
for by the fact that the acid soils contain a large excess of potential acid¬
ity, while the absolute amounts of alkalinity produced are exceedingly
minute. Therefore, equilibrium would be immediately restored without
sensibly affecting the true H-ion concentration. As evidence thereof,
the voltmeter readings for acid soils became constant within a few
minutes and remained constant for an indefinite time.

In order to determine whether the reduction of nitrates gave rise to
appreciable errors in the readings for alkaline soils, the OH- concentra¬
tion of a Ca(HC03)2 solution was measured in the presence of large
amounts of NaN03. No disturbance of the equilibria was noted under
the conditions of the experiment. It has been suggested that the use of
palladium or iridium, instead of platinum, for coating the electrodes
would practically prevent the reduction of nitrates, but the authors have
no experimental data upon this point.

While the reduction of nitrates prevents the use of the hydrogen elec¬
trode with extracts of acid soils, it is not believed that this factor would
cause appreciable errors in the case of nutrient solutions containing
nitrates. A nutrient solution would usually have a considerable poten¬
tial acidity or alkalinity — for example, the unhydrolyzed or undissociated
fraction of alkaline phosphates — so that constant results could be
obtained, just as with soil suspensions. Moreover, by use of the shaking
method described by Michaelis (21) the reduction of NOJ may be reduced
to a minimum, but even this slight production of alkali causes serious
error in soil extracts, where the total quantity of acid present is exceed¬
ingly small.

Oct. 1 6, 1916

Acidity and Adsorption in Soils

129

EFFECT OF GRINDING SOILS ON THE HYDROGEN-ION CONCENTRA¬
TION OF THEIR SUSPENSIONS.

The determination of the lime requirement of soils by the usual
methods depends upon several factors. Among these factors the fine¬
ness of division has recently received some attention. Cook (9) found
that in certain soils the lime requirement by the Veitch method in¬
creased with grinding, while Brown and Johnson (4) obtained opposite
results working with another group of soils. Several of the soils already
described were ground to pass a 200-mesh sieve. Table III shows the
H-ion concentration of suspensions made from the ground soil.

Table III -Hydrogen-ion concentrations in suspensions of soil ground to pass through

a 200-mesh sieve

Soil No.

Quantity
of soil.

Water.

Readings on
voltmeter.

H-ion concentra¬
tion (gram mole¬
cules per liter).

Gm.

C. c.

IS .

0. 01

O. 629

0. 8X10-5

IS .

. IO

. 6x7

.1X10-4

IS .

■ 50

• 632

. 7Xio~3

IS- .

I. O

612

.2X10-4

IS .

2. O

So •

“• 593

.4X10-4

IS .

5* 0

a. 560

.1X10-3

IS .

10. 0

• 577

. 7X10-4

I. 0

a* 653

• 3X10-5

. 625

1. 0X10-5

17 .

. ox

a. 615

. iXio-4

17 .

I. 0

. 624

.iXio"4

17 .

5-o

S97

• 3X10 4

20 .

2. 0

a. 623

. 1X10-4

TQ . .

1. 0

. 646

. 4Xio-i 5

IQ .

2. 0

a . 634

. 7 X io~5

* 5

• 750

. 7Xio“7

18 .

2. 0

• 748

. 7Xio“7

18 .

5- 0

. 762

.4X10 7

a Average of several determinations.

By comparing Table III with Table II it will be seen that with the
exception of soil 18 grinding did not materially alter the H-ion con¬
centration of the soil suspensions. These remarks bear no reference to
the lime requirement, which will be discussed later, but apply only to
intensity of acidity. Apparently the anomalous behavior of soil 18
may be explained on the supposition that the interior cores of the soil
particles are of a different composition from the exterior, partially
weathered layers. This hypothesis may also account for the findings of
Brown and Johnson (4)*.

The data of Table III also corroborate in the main those presented
in Table II in reference to the slight fluctuations in H-ion concentra¬
tion due to varying the proportion of soil to water.

130

Journal of Agricultural Research

Vol. VII, No. 3

EFFECT OF HEATING ON THE HYDROGEN-ION CONCENTRATION

In order to ascertain the effect of heating upon the H-ion concentra¬
tion, several soils were subjected to the heating treatments indicated in
Table IV.

Tabi<e; IV. — Effect of heating on H~ion concentration of soil suspensions

Soil No.

17
i7

20
20

Quantity
of soil.

Water.

Method of heating.

Time of
heating.

Voltmeter

reading.

H-ion concentra¬
tion (gram mole¬
cules per liter).

Gm.

C. c.

Hours.

5°

Oven 140° C .

0. 578

0.6X10-4

5°

. do .

. 621

. iXict4

5°

. do .

.583

.5X10-4

Muffle below red

. 652

. 3Xxo_s

heat.

. do. . . .

• 730

• iXicf^

Blasted .

. 812

. 6X10""8

Oven 140° C .

. 603

. 2XlO~‘

• 598

.3XIO-4

. do . .

. 672

.2XIO-5

5°

Muffle below red

.678

.iXict*

heat.

Blasted .

. 803

.8X10-8

The data of Table IV confirm the views of Connor (8), in that the
intensity of acidity decreases when the soils are heated at high tempera¬
tures. The insufficiency of the data concerning heating at 140° C. does
not admit of positive conclusions, though there is indication that the
H-ion concentration may be slightly increased by this treatment.

ESTIMATION OF THE LIME REQUIREMENT BY THE ELECTROMETRIC

METHOD

The rational treatment of acid soils requires that sufficient lime be
added to bring the soil to a neutral or slightly alkaline reaction. The
attainment of this point may be definitely determined by the method
described in this paper. Many empirical methods have been suggested
for the determination of the lime requirement, but the inaccuracy of
these methods is indicated by the enormous variations in results, as
shown in the comparative tests reported by Ames and Schollenberger (1)
and others. These findings are also confirmed by data obtained in this
laboratory.

An attempt was made to determine more precisely the lime require¬
ment of soils by a method of electrometric titration with calcium hydroxid,
Ca(OH)2, in which a standard calcium-hydroxid solution was added to
the soil suspensions until a definite alkaline reaction was obtained. The
data were supplemented by pot and beaker studies. Tables V and VI
present the results of these experiments.

Oct. 16, 1916

Acidity and Adsorption in Soils

131

Table) V. — Titration of soil suspensions with calcium hydroxid f 07 lime requirement

Quantity of calcium
hydroxid added

H-ion concentration.

Soil No.

Quan¬
tity of
soil.

Water.

(calculated to cal¬
cium carbonate ) .

Original.

After titration.

Time
of titra¬
tion.

Entire

quantity.

Per gram
of soil.

Volt¬

meter

read¬

ing.

Gram
molecules
per liter.

Volt¬

meter

reading.

Gram
molecules
per liter.

15...'...

Gm.

C. c.

Gm.

0. 0085

Gm.

O. 00x7

0. 590

0.4X10-4

775

0. 2 X io~7

Hours.

15 .

5°

. 0128

. 0026

• 59°

.4X10-4

• 796

. iXxo"7

16 .

.0054

. 0027

.654

• 3X10 5

• Z61

. 4Xio~7

*7 .

5°

. 0061

. 0030

. 611

. 2XlO-4

• 759

.5X10-7

17 .

5°

. 0070

• 0035

•599

.3X1O-4

• 757

.5X10-7

19 .

5°

. 0081

. 0040

.638

. 6XlO 5

• 77°

.3X10 7

20 .

5°

•0053

. 0026

.618

. iXxo"4

•754

. 6X10 7

20 .

. 0047

. 0023

. 623

. IXIO-4

• 757

.5X10-7

18 .

. 00068

. OOO34

• 707

.4XIO-6

• 758

.5X10-7

Table VI. — Results of beaker and pot studies

Soil in
beaker or
pot.

Total
quantity
of calcium
carbonate
added.

Calcium

Treated

soil

tested.

Reaction of treated
soii.

Soil

No.

carbonate
per gram
of soil.

Water.

Read¬
ings on
volt¬
meter.

H ion (gram
molecules
per liter).

Remarks.

15. .

Gm.

200

Gm.

O. 200

Gm.

0. OOI

Gm.

C. c.

0. 664

0. 2X10"5

Incubated 7 days at

I5--

200

. 400

. 002

• 799

. 9X10-8

30° C.

Do.

15. •

11, 200

IO. 0

. OOO9

. 630

. 8X10-5

In jar 2 months.

15- *

II, 200

15-0

. OOI3

•733

. iXicr6

Do.

16. .

5°

• 135

. OO27

•752

. 6X10 7

Incubated 7 days at

17. .

' . 090

. OO36

5°

. 728

.2X1O-0

3°° £•

Do.

17. .

5°

. 250

.0050

5°

• 777

. 2X1 o— 7

Do.

18. .

5°

. 017

. OOO34

5°

• 740

1. 0X10-7

Do.

19. .

. IOO

. OO40

5°

•755

. 5X10-7

Do.

20. .

• 06$

. 0026

5°

. 780

. 2X10 7

Do.

* Such a method is logically adapted to obtain the information neces¬
sary for the proper adjustment of the soil reaction by the addition of
lime. There are, however, certain difficulties met with in its application
to soils. One of the chief obstacles is due to the relative insolubility of
the acid-forming constituents of soils, which prevents a rapid attain¬
ment of equilibrium. This drawback can probably be overcome by the
use of a shaking machine. Another but presumably minor source of
error lies in the loss of C02 from the soil suspension, as previously
mentioned.

In order to determine whether the titrations with calcium hydroxid
might serve as a guide for the application of lime necessary to produce a
neutral or slightly alkaline reaction several beaker and pot studies were
undertaken. A reference to Tables V and VI shows that in these soils
approximate neutrality resulted from the admixture of calcium carbon-

I32

Journal of Agricultural Research

Vol. VII, No. 3

ate with the soil in amounts indicated by the titrations. While these
data are not extensive, a valuable correlation is suggested.

So far as the writers are aware, this is the first time that this electro¬
metric titration has been applied to soil studies. It is believed that with
further work a valuable means may be developed for the more exact
determination and adjustment of soil reaction. Hence, this method may
be extremely useful in the accurate control of soil reaction in many field
and pot experiments. The somewhat complicated nature of the appara¬
tus and the time involved would doubtless militate against its general
adoption for routine analyses.

EFFECT OF THE* ADDITION OF NEUTRAL SALTS ON THE HYDROGEN-
ION CONCENTRATION OF SOIL SUSPENSIONS

The effect on the H-ion concentration of soil suspensions produced
by the addition of neutral salts is a matter of considerable theoretical
interest and of practical importance. Such data may be of significance
in their relation to the application of soluble fertilizing salts and to the
various lime-requirement methods dependent upon treatments with
solutions of potassium nitrate. The desirability of similar measurements
as correlated with the effects of soluble salts on the physical condition of
soils has already been suggested by one of the authors (31). The bear¬
ing of the results upon adsorption phenomena will be discussed in a
later section of this paper.

Table VII records the changes in H-ion concentration of various soil
suspensions when treated with different neutral salts:

Table VII. — Effects of neutral salts on H-ion concentration of soil suspensions

Quan¬
tity
of soil.

Quan¬
tity
of salt.

Original soil.

Treated soil.

Soil

No.

Water.

Salt added.

Volt¬

meter

read¬

ing.

H-ion (gram
molecules
per liter).

Volt¬

meter

read¬

ing.

H-ion (gram
molecules
per liter).

16. .

Gins.

. C.c.
30

Potassium chlorid.

Cm.

0. 628

a 9X10“*

0. 582

O. 5 X I6”4

16. .

Sodium chlorid . . .

.628

.9X10-5

• 572

.8X10-4

16. .

Barium chlorid. . .

.628

.9X10-6

•555

.2X10-3

20. .

Potassium chlorid.

■ 639

• 5Xicrs

*575

• 7XIO-4

20. .

Sodium chlorid . . .

•639

.5X10-5

.568

1. oX io“4

20. .

Barium chlorid . . .

•639

.5X10-5

. 564

. 1 X io“s

15 • •

3°

Potassium chlorid.

• 598

• 3X10-4

.548

. 2 X 10"3

IS • *

Sodium chlorid . , .

• 598

• 3X10 4

•55i

. 2 X io”3

i5- •

3°

Barium chlorid. . .

•598

• 3X10 4

•535

• 3Xio“3

18. .

3°

Potassium chlorid.

. 69O

• 7X10-6

. 614

. 2X10”4

18. .

3°

Sodium chlorid . . .

. 69O

• 7X10-8

.615

. 2 X io“4

18. .

Barium chlorid. . .

. 69O

.7X10-6

■ 590

.4X10"4

14. .

3°

Potassium chlorid.

* 753

. 6Xio-7

. 662

.2X10-5

14. .

Sodium chlorid . . .

*753

.6X10 7

. 672

.2X10-5

14. .

Barium chlorid. . .

• 753

. 6X10”7

.654

.3X10-5

1 . . .

Potassium chlorid.

• 763

.4X10-7

•763

.4X10"7

Oct. 16, 1916

Acidity and Adsorption in Soils

133

A study of the foregoing data makes it evident that in all cases there
has been a distinct increase in the H-ion concentration, when either
potassium chlorid, sodium chlorid, or barium chlorid was added to the
suspension in the quantities indicated. The increase does not vary
greatly for the three salts used, but on the whole the barium chlorid has
a somewhat greater effect. In soil 14 we have an interesting case in
which a change of reaction from alkaline to acid has taken place, as a
result of the addition of neutral salts of sodium, potassium, or barium.
It is obvious that such a soil would probably be adjudged acid by the
potassium-nitrate method, although its normal reaction is, in fact,
slightly alkaline. In a more normal type of soil (No. i), however,
no change in the reaction is found. The use of calcium chlorid in the
above experiment was found to be impracticable on account of the
difficulty in obtaining a perfectly neutral salt.

ADSORPTION OF OH IONS BY SOILS IN SUSPENSIONS OF VARIOUS

BASES

So far the H-ion concentration of soil suspensions, the factors affect¬
ing it, and the possible use of an electrometric titration method for
determining the lime requirement have been the chief topics considered.
We shall now consider another phase of the general problem, involving
the question of the adsorption of OH ions by the soil. The hydrogen
electrode has proved useful for this purpose.

Changes of the OH-ion concentration of soil suspensions were measured
when varying quantities of different hydrates were added. In addition
the removal of Ca from a solution of hydrate in contact with two of the
soils studied was noted. The data obtained from these experiments are
incorporated in Tables VIII and IX.

To suspensions of soils that pass through a 200-mesh sieve the hydrates
were added, a small portion at a time, until the neutral point was just
passed; then further additions of hydrate were made until an arbitrarily
selected OH-ion concentration was maintained over a considerable
period of time, as noted in Table VIII. After each addition of the
hydrate the soil suspension was given a prolonged shaking. The bases
added have been calculated for convenience of comparison to the equiva¬
lent OH expressed in grams. The above data enable us to estimate the
approximate quantity of OH ions removed from the solution by the soil,
and in two cases where calcium hydroxid was added the removal of Ca
has also been determined by the usual analytical method.

134

Journal of Agricultural Research

Vol. VII, No. 3

Table) VIII. — Results of titrations of ground ( 200-mesh sieve) soil with various bases

si?8

a 73

S’8

rt <u

£3

Soil No.

uantity of soil.

Base.

T eight of OH.

H per gram of soil.

ime of titration.

riginal OH-ion coi

tration (gram mole

OH per liter).

H-ion' concentratio

approximate neutr

(gram molecules

per liter).

ase added beyond

tral point (weigh

OH).

H per gram soil.

ime of titration.

inal OH-ion conce

tion (gram mole*

OH per liter).

ffl

Gm.

C.c.

Gm.

Hrs.

Gm.

Hr.

is--

Calcium hy-
droxid.

. do .

0. 00038

0. 00076

i*5

0.9X10-®

0. 8Xio-7

0. 00006

0. 00012

1. 9X10-7

is..

10. 0

.00798

. 00079

2.0

.9X10-10

2.oXio-7

.00088

.00009

2. 9X10-7

IS--

IS*.

. do .

3.9X10 10
3.6X10-10

“.00211
°. 00207

a. 00211
a . 00207

. 9X10-4
2. 2X10-4

1*0

Sodium hy¬
droxid.

15. .

Calcium hy-

.00163

.00081

2. 1 X 10-10

4-3X10-8

rwiT/i

2. 7X10-7

droxid.

IS-.

2.0

Sodium hy-

.00131

.00065

24.0

1. 5X10-10

. 7X10-7

.00026

.00013

2.4X10-7

droxid.

15..

S-o

Calcium hy-

.00429

.00085

6.0

.7X10-10

.9X10-7

.00029

.00006

1. iXio-7

droxid.

is- ■

5-o

Sodium hy-

. 00350

.00070

3-o

4. 2X10-11

i.iXio"7

.00052

• 00010

1.2X10-7

droxid.

16. .

5-o

. do .

.00560

.00112

20.0

.6X10-®

.6X10-7

.00105

. 00021

1. 7X10-7

16. .

1.0

Calcium hy-

.00103

24.0

2.0X10-0

.6X10-7

.00340

. 00340

4. 2X10-5

droxid.

16. .

1.0

Barium hy-

.00099

o-5

i.oXio-9

i.gXio”7

.00280

. 00280

3-8X10“*

droxid.

16..

1.0

Sodium hy-

.00105

3-o

2.0X10-®

1.5X10-7

. 00263

4. 2X10-5

droxid.

16. .

1. 0

Potassium hy¬
droxid.

1.9X10-®

00324

°. 00324

5. 7X10-5

17..

1-0

Calcium hy-

. 00088

.00088

2*5

5.8X10-10

1.9X10-7

. 00474

. 7X10-4

droxid.

. 0006b

17..

1.0

Barium hy-

.00066

I-S

6.3X10-10

1. 7X10-7

.00406

.9X10-4

droxid.

17..

5- 0

Calcium hy-

.00650

. 00130

1.0

2.0X10-10

.8X10-7

.00800

.00160

1.3X10-3

droxid.

17..

S*o

Sodium hy-

.00438

.00087

4-0

2. 2X10” !0

. 8Xio“7

.00526

.00105

1. 7X10-6

droxid.

19..

2.0

Calcium hy-

. 00306

.00153

7.0

.9X10-®

.9Xio~7

. OIOIO

.00505

4.2X10-8

droxid.

19. .

2.0

Barium hy-

.00330

.00165

21.0

.9X10“®

.8X10-7

. 00858

. 00429

3-sXio-®

droxid.

19. .

2.0

Sodium hy-

. 00207

.00103

7.0

.7XlO-®

1.0X10-7

.00350

•00175

4. iXio-5

droxid.

19..

2.0

Potassium hy-

.00198

.00099

24.0

I.3X1O-®

. 7X10-7

. 00524

. 00262

3.8X10-5

droxid.

20. .

2.0

Calcium hy-

.00118

.00059

4.0

5.8X10-10

5.9X10-8

.00446

.00223

4.8X10-8

droxid.

20. .

2.0

Sodium hy-

. 00091

. 00045

3-5

6.3X10-IO

. 7X10-7

. 00249

. 00125

3.1X10-8

droxid.

a Includes base added to neutralize acid.

Table IX. — Removal of calcium from solution of calcium hydroxid by soils

Experiment No.

Soil

No.

Weight
of soil.

Water.

Calcium

added.

Calcium
recovered
in solu¬
tion.

Calcium removed
from solution.

Gm.

C.c.

Gm.

P. ct.

I .

5°

O. 0003

1 7

•0055

. 0007

O. 0048

3 .

•0137

. OO23

. 0114

4 .

. 0192

•0053

•0139

5 .

5°

. 0220

. 0067

*0153

6 .

. 0247

. 0089

. 0158

. 0003

/ .

8 . . .

. 0022

. 0003

. 0019

9 .

. 0038

. 0003

•0035

10 .

■ 0055

. 0005

. 0050

11 .

„ So

. 0066

. 0005

. 0061

. 0082

. 0009

. 0073

Oct. 16, 1916

Acidity and Adsorption in Soils

135

When 0.1 mgm. of OH as calcium hydroxid is added to 50 c. c. of
distilled water, there is produced an OH-ion concentration of 3.8 X io“5
gram molecules per liter. JPo reach this same concentration of OH ions
in the presence of the soil requires the addition of a much greater quan¬
tity of hydrate, whether of sodium, calcium, barium, or potassium. The
difference between the quantity of hydrate necessary to add to distilled
water to obtain the concentration fixed upon and that required when the
soil is present gives an index of the amount of hydroxyl ions removed
from the ionic equilibrium. In the case of acid soils it is probable that
until the neutral point is reached the removal of OH ions can be accounted
for by the reaction between the added OH ions and H ions derived from
the soil acids. Beyond the neutral point, however, another type of
reaction must necessarily be involved. For example, soil 16 required the
addition of 1.0 mgm. of OH as calcium hydroxid per gram of soil to give
the suspension an OH-ion concentration of 0.6 X io“7, while the second
point, representing a concentration of 4.2 X io~5 gram molecules .of OH ions
per liter, required the further addition of 3.4 mgm, of OH per gram of
soil. By subtracting the quantity of OH required to reach the same
point when added to distilled water, it is evident that 3.3 mgm. of OH
have been removed by some mechanism not associated with the neutrali¬
zation of acid. The nature of the latter type of reaction has been the
subject of much conjecture; but before entering upon a detailed discus¬
sion of this matter it is desirable to point out certain other relations
which may be derived from a further study of the data presented above.
These are concerned with the comparison of the amounts of OH ions
removed from solutions of the various bases, when added in combination
with different positive ions. In general, it may be said that to bring
about this higher OH-ion concentration in the soil suspension requires
a larger equivalent of calcium hydroxid and barium hydroxid than of
sodium hydroxid or potassium hydroxid. As a striking instance of this,
soil 19 may be cited. In order to assure a reasonable degree of validity
for these comparisons, they have been made, in accordance with Han¬
ley’s (14) suggestions, under conditions such that the suspended soil
particles were in equilibrium with solutions of the same concentration.
It is difficult to draw definite conclusions in regard to the relative quan¬
tities of bases required to produce neutrality, but it may be said that the
equivalent quantities are of approximately the same magnitude. Exact
equivalents would hardly be expected on account of various side reac¬
tions involving the interchange of bases.

The interesting observation recorded in Table VII that barium chlorid
when added to soils brings about a slightly greater acidity than when
sodium or potassium chlorid is added may possibly be correlated with
the fact that a greater quantity of OH is removed by the soil when
added as barium hydroxid than when added as sodium or potassium
hydrate. This agrees with certain contentions of Parker (24).

55857°— 11 - 4

136

Journal of Agricultural Research

Vol. VII, No. 3

Many investigators consider soil acidity as being primarily related to
adsorptive phenomena. Cameron (5) has explained the reddening of
blue litmus paper by many soils as due to# selective absorption of the
base from the litmus base and does not correlate the reddening of the
litmus paper with the presence of soluble acid except in very rare cases.
Harris (15) also ascribes soil acidity to the selective adsorption of bases
by the soil. Bogue (3) and others have expressed a similar opinion
concerning soil acidity. The phenomena of adsorption, as is well known,
has also been invoked to explain the fixation of various ions by the soil.

Another view of soil acidity attributes it to an exchange of bases in
which a weak base, as aluminium, has been replaced by a strong base
such as potassium. The resulting hydrolysis of the aluminium salts
produces an acid reaction in the solution. This view has been advanced
most recently by Rice (26), Conner (8), Loew (20), Daikuhara (10), and
Veitch (35). This is a plausible explanation for the acidity of soils
which haye been treated with salts, although it does not indicate the
reaction of the soil previous to the salt treatments. The results already
presented in regard to the addition of salts show that in the case of acid
soils this treatment markedly increased the H-ion concentration in every
case. One slightly alkaline soil was also found to give an acid reaction
after the salt treatments. These data are in direct accord with the
accumulated evidence concerning the acidity developed when a strong
base reacts with the soil constituents replacing a weak base. In the
light of these observations it is quite unnecessary to have recourse to
adsorption theories to account for the acidity of many salt- treated
soils, even though the soils originally might have had an alkaline reac¬
tion. This is illustrated in the case of soil 14, already cited. Obviously,
the principal reactions involved in producing the acidity in such cases
are of a chemical rather than of a physical nature. The importance of
such an exchange of bases is emphasized by the work of Sullivan (32),
Rice (26), and Conner (8).

Truog (33), Hanley (14), Gillespie (13), and Loew (20) have recognized
that the acidity of a soil is due to the presence of soluble acids. The
experimental evidence reported in this paper has brought the writers
to a similar conclusion, in which case it is not necessary to associate
soil acidity with physical adsorption.

A great many obscure phenomena in the fields of biochemistry and
soil chemistry, as well as many others, have been classified under the
indefinite terms “absorption” and “adsorption.” The indiscriminate use
of these terms has not served to clarify the problem involved. Thus, the
fixation of plant foods by soils has been explained by some investigators
entirely on the basis of physical adsorption. Frequently, the importance
of the chemical exchange of bases has been disregarded. Moreover, the
possibility of addition compounds as suggested by Sullivan (32) has not
received sufficient consideration. The results given in Table VII show

Oct. 16, 1916

Acidity and Adsorption in Soils

137

that the OH ions of a solution of calcium hydroxid have been removed in
large measure from the solution by the soil, while those in Table VIII
demonstrate a simultaneous loss of calcium from the solution. Evidently
both the Ca and OH ions have been removed to a notable extent from
the solution by the soil. Two explanations of these phenomena suggest
themselves. One explanation would assume a condensation of the cal¬
cium hydroxid as a whole on the surfaces of the soil particles. The
other, in accordance with the ideas of Sullivan (32), Van Bemmelen (34),
and Lemberg (32, p. 20-23), considers that the calcium hydroxid has
been chemically united with some of the soil constituents, forming direct
addition compounds. In view of the evidence embodied in the literature,
the second theory seems the more tenable. In many ways the chemical
explanation appears to be more logical in accounting for the fixation of
bases by soils.

Moreover, the theories which regard adsorption in soils as a chemical
phenomenon have received an element of support in the proposed chemi¬
cal structure of the silicates. Clarke and Steiger (7) have shown that the
composition of the silicates is such that it frequently admits of an ex¬
change of bases. This has also been recognized by soil chemists. Like¬
wise, from the structural formulae for silicates proposed by the above
investigators, it is evident that acid salts of the various silicic acids might
contain replaceable hydrogen. Loew (20) attributes the acidity of certain
Porto Rican soils to the presence of acid silicates. As previously men¬
tioned, Sullivan (32) has also pointed out that a basic hydrate may form
van addition compound with silicates.

The theory of selective adsorption of a single ion which has been
advanced by various investigators is in its final analysis not entirely com¬
prehensible. One objection to this theory lies in the disregard of the
ionic equilibrium. For example, it is frequently assumed that from a
solution of a neutral salt one ion may be withdrawn by a colloid inde¬
pendently of its equivalent, oppositely charged ion. Thus if K+ be
selectively removed from a solution of potassium chlorid (KC1), then the
above assumption may be diagramatically represented as follows :

(Colloid — K) +C1 +H20

u _ .

H+OH

Obviously, such a system is electrically unbalanced. In order to
meet this difficulty, the electrical double-layer theory of Helmholtz has
been proposed. A modification of this theory has been especially urged
by Billitzer (2). In its simplest form the electrical double-layer theory
assumes that one ion may become more closely associated with a col¬
loidal particle than the oppositely charged ion. The force which binds
the ion to the colloid is not well understood, but there is some justifi¬
cation for believing that the most closely associated ion imparts its

138

Journal of Agricultural Research

Vol. VII, No. 3

charge to the colloidal particle. The charge thus produced on the
surface of the particle is balanced by oppositely charged ions in the
immediate sphere of attraction. In this manner the conditions of
electrical equilibrium are accounted for.

Parker (24) has avoided some of the difficulties involved in explaining
the electrical equilibria on the assumption that one ion alone is removed
by a colloid. He claims that potassium chlorid in aqueous solution is
hydrolyzed, yielding hydrochloric acid and potassium hydroxid. The
latter is then considered to be withdrawn from the solution as a whole.
Obviously, this does not explain the mechanism by which the potassium
hydroxid is combined with the colloid, although it does take into account
the electrical equilibria and the presence of acid in the solution. It
does not, however, exclude the possibility of a direct chemical addition
product of potassium hydroxid with some of the soil constituents.

Either of the last two theories may account for the presence of acid
in the filtrate or supernatant liquid derived from a colloid in contact
with a salt solution. Whichever explanation may be preferable, for
practical purposes the final result is the same — that is, the colloid has
retained equivalent quantities of negative and positive ions, and chem¬
ically equivalent quantities remain in solution.

Since these theories of adsorption necessitate the removal of both
positive and negative ions, the usefulness of the idea of selective adsorp¬
tion as applied to many types of soil reactions may be questioned,
especially in view of the possibility that chemical reactions, at least
to a considerable extent, may account for the observed phenomena.

This statement is not to be construed as denying the probability of
a partial condensation of a chemical compound on solid surfaces. A
clear exposition of this type of phenomena is given by Patten (25).
The magnitude of such condensations is exceedingly variable, for it de¬
pends upon the physical conditions of the particular system, and, hence,
it is difficult to estimate its significance in any specific instahce. In¬
deed, Robertson (27) is inclined to believe that surface condensation
accounts for only a small portion of the total amount of substance
combined with the adsorbing body. In his opinion most reactions
designated by the terms “absorption” and “adsorption” obey the usual
laws formulated for chemical reactions. The influence of surface in
accelerating and possibly changing the nature of these reactions is not
denied. This solution of the problem, as stated by Robertson, is not
admitted by certain other investigators, notably Wolfgang Ostwald (23).
Apparently, no final decision on this matter has been reached by physical
chemists.

It has been the aim of the present paper to set forth certain results and
methods of investigation which may serve to throw additional light on
some of the unsolved problems of soil fertility. Especially has the
method of attack made possible an investigation of the important ques-

Oct. 16, 1916

Acidity and Adsorption in Soils

139

tions of soil acidity and adsorption on a new basis. The evidence has
been confirmatory of the view that soil acidity is fundamentally de¬
pendent upon the equilibria of reactions yielding an excess of H ions,
and is not necessarily related to the various phenomena grouped under
the terms “absorption ” and “ adsorption.” Although the literature con¬
cerning soils constantly refers to u absorption ” and “ adsorption/7 yet no
very concise meaning has been attached to these terms. In fact, there
is quite as much evidence in favor of a chemical, as opposed to a physical,
interpretation of such phenomena. 3? or these reasons ,a more critical
examination of this field would be a welcome addition to agricultural
science.

DESCRIPTION OF EXPERIMENTAL APPARATUS

The apparatus used in determining the H-ion concentration is similar
to that described by Hildebrand (17), with such modifications as are
necessary or convenient for purposes of soil investigations (fig. 1).
Precise methods for determining small differences of potential have fre¬
quently been described, but the use of an elaborate potentiometer system
is neither practicable nor necessary in work with soils or nutrient solutions.
Obviously measurements of physical-chemical exactitude are useless
unless all the factors involved are capable of equally exact control, which
is not possible with most substances of agricultural interest. For prac¬
tical application, therefore, the voltmeter method of Hildebrand is
entirely adequate in point of accuracy and at the same time rapid and
convenient.

For information in regard to the physical-chemical principles under¬
lying the method, the reader is referred to Hildebrand (17), Michaelis (21),
Itano (18), or to textbooks on electrochemistry. Very convenient tables
for the transformation of voltmeter readings into H- and OH-ion con¬
centrations have been prepared by Schmidt (30). It is deemed desirable,
however, to present here certain details of the apparatus and method of
procedure, since these are not easily available to the general worker in
agricultural laboratories. Moreover, measurements with soil suspen¬
sions require special precautions, to avoid otherwise very misleading
results. ,

The arrangement of the apparatus and method of wiring are shown in
figure 1. The entire system includes the following pieces of apparatus:
Dry cell, two rheostats of 40 ohms resistance, with sliding contacts;
Weston voltmeter, o to 1.2 volts; Leeds and Northrup portable galva¬
nometer, sensitivity of 1 megohm; contact key; calomel cell, hydrogen-
electrode vessel; hydrogen generator, with purifying tube, wash bottles,
and rheostat or lamp board connected with direct current.

A convenient Cottrell hydrogen electrode, designed by the Department
of Chemistry, University of California, consists of a glass tube of about
1 cm. in diameter and 15 cm. in length, in the end of which is sealed a

14° Journal of Agricultural Research voi.vii,no.3

cylinder of fine platinum gauze approximately i cm. long. Near the
upper end of the electrode a branch tube permits the entrance of hydro¬
gen gas. A platinum wire is affixed to the gauze at one end and sealed
into a small glass tube at the other. The electrical connection is then

made by the use of a copper wire and a mercury cup, as shown in figure
i. The electrode must be gas-tight at the top. Platinum black is de¬
posited on the gauze, as described by Ostwald (22). The coating should
be renewed occasionally. It is sometimes necessary to clean the plat-

Oct. 16, 1916

Acidity and Adsorption in Soils

141

inum between determinations with a little diluted hydrochloric acid,
afterward rinsing in many changes of distilled water.

The electrode vessel must be closed to the air. A simple and easily
constructed cell may be made from a wide-mouthed bottle of 75 to 100
c. c. capacity. Through holes in the rubber stopper are fitted the
hydrogen electrode, an agar connecting tube, a small exit tube, and a
tube for saturating the solution with hydrogen when desired. A hole
to admit the tip of a burette should also be provided for purposes of
titration. The most convenient method of making the liquid connection
is by means of bent glass tubes, filled with agar jelly prepared with a
saturated solution of potassium chlorid. Only thoroughly washed agar
should be used in making the jelly. In order to avoid contact potentials
as far as possible connection with the calomel cell is made through a
beaker, containing saturated potassium-chlorid solution. The con¬
struction of a normal calomel cell is described by Ostwald (22). In
the present work N/10 potassium chlorid was used.

An adequate supply of pure hydrogen is of primary importance, and
this is most conveniently obtained from the decomposition of water by
means of a direct electric current. Such a generator may easily be pre¬
pared by using a large bottle and an inner cylinder made from a wide
glass tube as a means of separating the electrodes. The latter may be
of nickel, or pure iron and a 25 per cent (by weight) potassium hydroxid
solution is convenient as an electrolyte. In order to purify the hydrogen
from small quantities of oxygen, it is passed through a long glass tube
filled with platinized asbestos. This is heated by a fine nichrome wire
wound around the outside of the tube' and connected through a lamp
with a source of current. To provide a rapid stream of hydrogen re¬
quires the consumption of several amperes of current.

E^ERIMKNTAIv procedure

After placing the soil suspension in the bottle, hydrogen gas is per¬
mitted to flow through the electrode raised above the surface of the
liquid for several minutes. The electrode is now lowered until the plat¬
inum gauze is partially submerged in the liquid and the exit tube closed.
The bottle is now rotated back and forth for several minutes, as originally
described by Hasselbach and Gammeltoft (16). The agar tube (at
other times kept out of the liquid) is now lowered so that a connection
is made with the calomel cell through the beaker of potassium chlorid.
The rheostats are adjusted so that no deflection of the galvanometer
needle is noted when momentary connection is made by tapping the
key. The reading on the voltmeter is then recorded. The procedure
is repeated until constant readings are obtained. This occurs in the
case of acid s6i\s within a few minutes, but for soils approximately
neutral a slightly longer time will be required. In the case of titrations

I42

Journal of Agricultural Research

Vol. VII, No. 3

prolonged shaking is required after each addition of the titrating solu¬
tion, in order to obtain constant readings. This is due to the slow rate
of solubility possessed by the acid constituents of the soil. The use
of a mechanical shaking device would doubtless greatly facilitate the
operation.

Duplicate determinations on soil suspensions usually agree within
o.oi to 0.02 volt. In any one determination constancy of readings
may be obtained to within 0.002 volt. The larger sources of error
result from the change in the decomposition of the solution due to
reduction of nitrates or the breaking down of the HCOs ions. These
points have already been thoroughly considered. Robertson (28) and
later Desha and Acree (11) have shown the possible interference of cer¬
tain types of organic matter. Saturation of the electrode with hydrogen
before dipping seems to obviate the error, and no difficulty from this
source was experienced in the present work.

The use of the modification of Hildebrand's apparatus described in
this paper and due observance of the special precautions noted will, it
is believed, enable the investigator of soils and plants to obtain valuable
information in regard to H-ion concentrations without undue loss of
time. In many fields of biochemistry 1 similar methods have been
extensively employed during the last few years.

SUMMARY

(1) Soil acidity is due to the presence of an excess of hydrogen ions in
the soil solution.

(2) Direct evidence of this fact is given by hydrogen-electrode meas¬
urements.

(3) The hydrogen-ion concentration of different soil suspensions was
found to vary within wide limits, from a condi tipn of high acidity to one
of high alkalinity.

(4) Soils containing calcium in equilibrium with HCOJ and C02 have
a very slightly alkaline reaction.

(5) The effect of heating, grinding, and of varying the ratio of soil to
water on the hydrogen-ion concentration was studied.

(6) An electrometric method for the determination of the lime require¬
ment of soils is suggested.

(7) The addition of sodium chlorid, potassium chlorid, and barium
chlorid to certain soil suspensions was found to increase the hydrogen-ion
concentration.

1 For a review of the literature concerning the application of the hydrogen electrode to biochemistry
the reader is referred to the following:

Schmidt, C. L. A. Changes in the H+ and OH- concentration which take place in the formation
of certain protein compounds. In Jour. Biol. Chem., v. 25, no. 1, p. 63-79, 9 fig* 1916. Bibliographical
footnotes. «

Sorensen, S. F. I*. Uber die Messung und Bedeutung der Wasserstoffionenkonzentration bei biologis-
chen Prozessen. In Ergeb. Physiol., Jahrg. 12, p. 393-532, 12 fig* 1912. Eiteratur, p. 394-398.

Oct. i6f 1916

Acidity and Adsorption in Soils

143

(8) Several phases of the phenomena designated “adsorption” were
studied, with special reference to the removal of OH ions by the soil from
solutions of various hydrates.

(9) There appears to be a simultaneous removal of positive and nega¬
tive ions from solution by soils.

(10) Some general theoretical considerations with regard to the rela¬
tion of adsorption to chemical reactions in soils are presented.

(11) A convenient method of procedure for utilizing the hydrogen
electrode in soil studies is described.

literature cited

(1) Ames, J. W., and SchoeeenbErger, C. J.

1916. Comparison of lime requirement methods. In Jour. Indus, and Engin.
Chem., v. 8, no. 3, p. 243-246.

(2) Bieeitzer, Jean.

1903. Eine Theorie der Kolloide und Suspensionen. In Ztschr. Phys. Chem.,
Bd. 45, Heft 3, p. 307-330.

(3) BoguE, R. H.

1915. The adsorption of potassium and phosphate ions by typical soils of the
Connecticut Valley. In Jour. Phys. Chem., v. 19, no. 8, p. 665-695,

13 %*

(4) Brown, P. E., and Johnson, H. W.

1915. The effect of grinding the soil on its reaction as determined by the

Veitch method. In Jour. Amer. Soc. Agron., v. 7, no. 5, p. 216-220.

(5) Cameron, F. K.

1911. The Soil Solution . . . 136 p., 3 fig. Easton, Pa.

(6) - and Beee, 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.

(7) Cearke, F. W., and Steiger, George.

1902. The action of ammonium chloride upon silicates. U. S. Geof. Survey
Bui. 207, 57 p.

(8) 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. 1, p. 35-40, 2 fig.

(9) Cook, R. C.

1916. Effect of grinding on the lime requirement of soils. In Soil Science,
v. 1, no. 1, p. 95-98.

(10) Daieuhara, G.

1914. Ueber saure Mineralbdden. In Bui. Imp. Cent. Agr. Exp. Sta. Japan,
v. 2, no. 1, p. 1-40, 1 pi.

(11) Desha, L. J., and Acree, S. F.

19x1. On difficulties in the use of the hydrogen electrode in the measurement
of the concentration of hydrogen ions in the presence of organic com¬
pounds. In Amer. Chem. Jour., v. 46, no. 6, p. 638-648, 1 fig.

(12) Fischer, Gustav.

1914. Die Sauren und Kolloide des Humus. In Kuhn Archiv, V..4, p. 1-36/
4 fig. Literaturverzeichnis, p. 135-136.

(13) GieeEspie, L. J.

1916. The reaction of soil and measurements of hydrogen-ion concentration.
In Jour. Wash. Acad. Sci., v. 6, no. 1, p. 7-16, 2 fig.

144

Journal of Agricultural Research

Vol. VII, No. 3

(14) HaneEY, J. A.

1914. Estimation of the surface of soils. In Jour. Agr. Sci. , v. 6, pt. 1 , p. 58-62.

(15) Harris, J. E.

1914. Soil acidity. Mich. Agr. Exp. Sta. Tech. Bui. 19, p. 524-536.

(16) Hasselbach, K. A., and Gammeetoft, S. A.

1915. Die Neutralitatsregulation des graviden Organismus. In Biochem

Ztschr., Bd. 68, Heft 3/4, p. 206-264, 2 fig.

(17) Hiedebrand, J. H.

1913. Some applications of the hydrogen electrode in analysis, research and
teaching. In Jour. Amer. Chem. Soc., v. 35, no. 7, p. 847-871, 15 fig.

(18) Itano, Arao.

1916. The relation of hydrogen ion concentration of media to the proteolytic

activity of Bacillus subtilis. In Mass. Agr. Exp. Sta. Bui. 167, pt.
1, p. I39~I77> iUus.

(19) Johnson, John.

1916. The determination of carbonic acid, combined and free, in solution,
particularly in natural waters. In Jour. Amer. Chem. Soc., v. 38,
no. 5, p. 947-975-

(20) LoEw, Oscar.

1913. Studies on acid soils of Porto Rico. Porto Rico Agr. Exp. Sta. Bui. 13,

23 p., 1 fig.

(21) Michaeus, Leonor.

1914. Die Wasserstoffionenkonzentration. 210 p., 41 fig. Berlin. Litera-

turverzeichnis, p. 196-207.

(22) OSTWALD, W. F.

1910. Hand- und Hiilfsbuch zur Ausfiihrung physiko-chemischer Messungen.
Aufl. 3, hrsg. von R. Luther und K. Drucker. 573 p., 351 fig.
Leipzig.

(23) Ostwaed, Wolfgang.

1909. Grundriss der Kolloidchemie. 525 p., Dresden.

(24) Parker, E. G.

1913. Selective adsorption by soils. In Jour. Agr. Research, v. 1, no. 3,
p. 179-188, 2 fig.

(25) Patten, H. E.

1907. Energy changes accompanying absorption. In Trans. Amer. Elec-

trochem. Soc., v. n, p. 387-405, 1 fig. Discussion, p. 405-407.

(26) Rice, F. E.

1916. Studies on soils. I. Basic exchange. In Jour. Phys. Chem., v. 20, no. 3,
p. 214-227, 1 fig.

(27) Robertson, T. B.

1908. Einige kritische Bemerkungen zur Theorie der Adsorption. In Ztschr.

Chem. und Indus. Kolloide, Bd. 3, Heft 2, p. 49-76.

(28) - .

1910. Studies in the electrochemistry of the proteins. I. The dissociation of

potassium caseinate in solutions of varying alkalinity. In Jour. Phys.
Chem., v. 14, no. 6, p. 528-568, 2 fig.

(29) Saidee, Teodor.

1913. Quantitative Untersuchungen fiber die Reaktion wasseriger Boden-
auszfige. In Bui. Sect. Sci. Acad. Roumaine, ann. 2, no. 1, p. 38-44

3 fig-

(30) Schmidt, C. L. A.

1909. Table of H+ and OH“~ concentrations corresponding to electromotive

forces determined in gas-chain measurements. In Univ. Cal. Pub.
Phys., v. 3, no. 15, p. 101-113.

Oct. 16, 1916

Acidity and Adsorption in Soils

145

(31) Sharp, L. T.

1915. Salts, soil-colloids, and soils. In Proc. Nat. Acad. Sci., v. 1, no. 12,
p. 563”568.

(32) Sullivan, E. C.

1907. The interaction between minerals and water solutions. U. S. Geol.
Survey Bui. 312, 69 p.

Cites (p. 20-23) work by Lemberg.

(33) Truog, E.

1915. Soil acidity and methods for its detection. In Science, n. s. v. 42,
no. 1084, p. 505-507.

(34) Van Bemmelen, J. M.

1888. Die Absorptionsverbindungen und das Absorptionsvermogen der
Ackererde. Dritte Abhandlung. In Landw. Vers. Sta., Bd. 35, p.
69-136.

(35) Veitch, F. P.

1904. Comparison of methods for the estimation of soil acidity. In Jour.
Amer. Chem. Soc., v. 26, no. 6, p. 637-662.

ADDITIONAL COPIES

OF THIS PUBLICATION MAY BE PROCURED FROM
THE SUPERINTENDENT OF DOCUMENTS
GOVERNMENT PRINTING OFFICE
WASHINGTON, D. C.

15 CENTS PER COPY

Subscription Price, $3.00 Per Year

JOURNAL OF AGRKEFURAUESEARCH

DEPARTMENT OF AGRICULTURE

Vol. VII Washington, D. C., October 23, 1916 No. 4

LIFE HISTORY OF HABROCYTUS MEDICAGINIS, A
RECENTLY DESCRIBED PARASITE OF THE CHALCIS
FLY IN ALFALFA SEED

By Theodore D. Urbahns,

Entomological Assistant , Cereal and Forage Insect Investigations ,

Bureau of Entomology

INTRODUCTION

The following account of Habrocytus medicaginis Gahan is the result of
observations concerning its parasitism upon Bruchophagus funebris How.
inhabiting alfalfa seed (. Medicago sativa). The observations were begun
in the fall of 1912 and continued into the year 1915. The field observa¬
tions and collections extended over several of the States west of the
Rocky Mountains. The laboratory studies were conducted at Glendale
and Pasadena, Cal.

DISCOVERY OF THE PARASITE

This new hymenopterous parasite, H. medicaginis , was first found by
the writer on September 28, 1912. Several specimens had emerged from
alfalfa seeds infested by B. funebris collected at Yuma, Ariz., on August
30. Adults of the parasite continued to emerge from the seeds until
October 1 5 of the same year.

The parasite was again reared on November 4, 1912, from infested
alfalfa seeds taken at Chino, Cal., on September 24; from this lot of
seeds individuals continued to emerge until July 12, 1913. Additional
alfalfa seeds collected at Tulare, Cal., on October 1, 1912, yielded this
species at different times between June 11, 1913, and April 9, 1914.

Alfalfa seeds dissected on March 6, 1913, at Glendale, Cal., showed
larvae of this species feeding upon the dead larvae of B. funebris . Some
of the parasite larvae were reared to the adult stage as early as April 22.

The writer made many collections of alfalfa seeds throughout the
different alfalfa seed-growing districts between the Rocky Mountains
and the Pacific coast during the seasons of 1913 and 1914. Prom these
seeds adults of H . medicaginis emerged as follows: Dos Palos, Cal.,
October 11, 1913; Stockton, Cal., June 16, 1914; Brawley, Cal., June

(147)

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.
fv

Vol. VII, No. 4
Oct. 23, 1916

K — 44

148

Journal of Agricultural Research

Vol. VII, No. 4

17, 1914; Red Bluff, Cal., September 23, 1914; Pendleton, Oreg., Sep¬
tember 23, 1914; Twin Falls, Idaho, September 23, 1914; Blackfoot,
Idaho, September 29, 1914; Gunnison, Utah, October 10, 1914; Bishop,
Cal., April, 1915; Manti, Utah, June 5, 1915; Aberdeen, Idaho, June 16,
1915; and Nephi, Utah, June 16, 1915. The total rearings of this parasite
brought the number of adults up to several hundred specimens from these
different localities. Figure 1 shows the known distribution of the insect.

In recent years various members of the Bureau of Entomology, and
others, have reared B. funebris , together with miscellaneous insects
emerging from both clover and alfalfa seeds. These specimens were ex¬
amined and it was found that H. medicaginis was reared from alfalfa
seed pods infested by B . funebris as follows :

August 12, 190S, Mesilla Park, N. Mex., by C. N. Ainslie.

- 1908, Chico, Cal., by R. McKee.

- 1910, Sacramento, Cal., by O. E. Bremner.

September 24, 1910, Wellington, Kans., by E. O. G. Kelly.

- 1910, Wellington, Kans., by T. H. Parks.

September 9, 1910, Wellington, Kans., by H. Osborn.

- 1912, Cavite, S. Dak., S. Halvardgaard.

- 1913, Newell, S. Dak., C. N. Ainslie.

- 1914, Salt Bake City, Utah, by T. R. Chamberlin.

CLASSIFICATION AND DESCRIPTION

H . medicaginis belongs to the hymenopterous superfamily Chalci-
doidea, family Pteromalidae, subfamily Pteromalinae. Specimens reared

Fig. 1. — Map of the United States, showing the known distribution of Habrocyius medicaginis.

by the writer from B. funebris in alfalfa seeds at Yuma, Ariz., were
described by Gahan as a new species.1

The description of the female is as follows :

Length about 1.7 mm. Head and thorax closely punctate, the punctures on the
medial portion of the mesoscutum slightly larger than those on the scapulae and
scutellum; antennae with two ring- joints; pedicel and first funicle joint, excluding

1 Gahan, A. B. Descriptions of new genera and species, with notes on parasitica Hymenoptera. In
Proc. U, S. Nat. Mus., v. 48, p. 163. 1915.

Oct. 23, I916

Life History of Habrocytus Medicaginis

149

the ring- joints, about equal; following funicle joints a little longer than the first and
a trifle longer than broad; viewed from in front the head is broader than long, the
clypeal region with converging strise and a deep median sinus on the anterior margin;
viewed from above the head is slightly broader than the thorax, narrow antero-
posteriorly, the occiput slightly concave, the ocellocular line longer than the lateral
ocellar line, the lateral ocellar line not equal to half the postocellar line; pronotum
strongly transverse with a sharp margin anteriorly; propodeum short, without a neck,
with a median carina and lateral folds, the region between the lateral folds more or
less distinctly wrinkled and with a fovea-like depression at the base and another at
the apex of the fold; the region outside the lateral folds is usually more faintly sculp¬
tured with indistinct lines; propodeal spiracles elliptical; marginal and postmarginal
veins subequal, the stigmal one- third shorter; abdomen conic-ovate, about as long as
the head and thorax and nearly smooth, the dorsal segments beyond the first with
very faint transverse lines. Head and thorax aeneous; antennae brown, the scape
slightly paler beneath; wings hyaline; all coxae aeneous like the thorax, all tro¬
chanters and femora black with an aeneous tinge; tibiae and tarsi usually reddish
yellow, the former often brownish except at apex; apical joint of all tarsi dark; abdo¬
men polished aeneous.

LIFE HISTORY OF THE HOST

The host insect (B. funebris) of H. medicaginis completes its entire
life development within the growing seeds of alfalfa, red clover (Trifo¬
lium incarnatum), and wild species of Medicago. After reaching matu¬
rity the adult eats a hole through the seed wall and through the wall of
the seed pod to make its escape. B. funebris may pass through from
one to four or five generations in a single season.

METHOD OF STUDYING THE PARASITE

The fact that H . medicaginis completed its entire development within
the unbroken walls of an alfalfa seed made it necessary for the writer to
dissect many seeds under a microscope and remove this parasite in its
different Stages for special study. Small parasite larvae removed from
seeds were placed singly upon a larva of their host. The host and para¬
site were then placed in a small cavity made in a 7-mm. cork and covered
by a glass vial (PI. 4, fig. B). A most satisfactory method of observing
one of these larvae in its development was to place it upon a larva of
B. funebris and keep both host and its parasite in a cavity made between
two layers of sheet cork. The upper layer of cork could then be removed
to expose the parasite larva.

STAGES OF HOST SHOWING PARASITISM

H . medicaginis is parasitic upon the larval stage of its host with
possibly a few exceptions. Microscopic dissections of many infested
alfalfa seeds showed 77 larvae of this species, each feeding externally upon
the larval stage of its host (PI. 4, fig. D). This parasite was in no case
found to be attacking the pupa of B. funebris . Only a single parasite is
able to develop upon its host within the walls of a single infested alfalfa
seed.

Journal of Agricultural Research

Vol. VII, No. 4

ISO

The parasite larva in completing its development usually destroys
the larva of its host with the exception of the head and mandibles. If
two parasites chance to be upon a single host, one dies before develop¬
ment continues for any length of time.

APPEARANCE OF THE INSECT IN THE FIELDS

Throughout the Southwestern States the first adults make their
appearance in the fields as early as March and April, simultaneously
with the development of seed pods upon the earliest alfalfa plants. Out
of nearly 100 hibernating larvae kept under observation at the laboratory,
23 emerged as adults in March, 39 in April, 8 in May, and 4 in June.
They attack the first generation of larvae of B. funebris infesting the earliest
isolated plants and increase throughout the summer in accordance with
the abundance of their host insects.

OVIPOSITION

The adult female, frequently seen to be active on the blossoms and soft
green seed pods in the alfalfa fields, is apparently able to locate the pods
in which seeds have previously been infested by B . funebris . She selects
her position upon the green pod directly over an infested seed. During
oviposition the head is slightly elevated and the antennae are held directly
forward. The tip of the abdomen is lowered almost to the surface of the
seed pod. The ovipositor is forced through the soft walls of the pod and
into the watery seed. It is necessary for the egg of the parasite to be
placed within the infested seed and upon the larva of its host in order that
the newly hatched parasitic larva may secure food for its development.

THE LARVA

DEVELOPMENT

For several hours after emerging from the egg, the larva of the parasite
may move about on its host without feeding, but when it once begins to
attack its host and take food, its development follows rapidly. The
writer’s observations show that a growing larva of this species may
completely destroy its host and become fully developed within a minimum
period of five or six days after taking its first food.

DORMANT PERIOD

When the parasite larva has completed its development and con¬
sumed all of its available food, a period of rest frequently follows. The
occurrence and duration of this resting period depends upon the moisture
and temperature conditions to which the seed is subjected. A larva of
H . medicaginis completing its growth within a moist seed of a green and
growing pod will almost invariably transform to the pupal stage at once
and emerge as an adult in due time, but if the infested alfalfa seed has

Oct. S3. 19x6

Life History of Habrocytus Medicaginis

151

become dried from the hot desert winds before the parasitic larva has
completed its development, a prolonged resting period may follow.
This period may vary from a few weeks to a year. With combined
moisture and a warm temperature the insect resumes its development
toward the formation of the pupal stage.

description

The larva is grublike and almost white in color and averages 1.6 mm.
in length when fully developed. It is cylindrical in shape and rounded
anteriorly and posteriorly. Head medium-sized and slightly bilobed.
Mandibles almost invisible. It has a small inconspicuous tubercle on
each eye lobe. There are 13 body segments of approximately equal
length, except the first two, which are slightly longer; segmentation
medium. Body skin usually slightly wrinkled, but sometimes smoothj
glossy, and free from pubescence. Anal segment divided into a dorsal
and a ventral lobe. Three fine setae on dorsal lobe (PI. 4, fig. C).

LENGTH OF LARVAL STAGE

The length of the larval stage depends greatly, as has been previously
stated, upon the resting period following the development of the larva.
A small newly hatched larva began feeding Upon its host under observa¬
tion on April 23. It showed noticeable growth from day to day, and by
April 27 it had completely killed its host. On May 5 the larva trans¬
formed to the pupal stage. Another newly hatched larva dissected
from a green alfalfa seed began feeding upon its host on September 5
and by September 7 the host was killed. The larva was apparently fully
developed by September 11. Others removed from alfalfa seeds in
August did not pupate until the following April. The minimum length
of the larval stage as observed by the writer is normally about 12 days.
The maximum length is a year or more.

PREPUPAL FORM

Just before entering the pupal stage the larva of H. medicaginis dis¬
charges an excessive amount of excrement. This is followed by a
lengthening of the anterior body segments and the shaping of the pupa
within the larval skin. The prepupal form requires two or three days
unless retarded by unfavorable conditions.

THE PUPA
FORMATION

Pupation takes place after the pupal form has developed within the
larval skin. The larval skin breaks along the dorso-anterior margin and
is slowly worked back until the newly formed pupa is exposed.

152

Journal of Agricultural Research

Vol. VII, No. 4

DESCRIPTION

The newly formed pupa is white with salmon-colored eyes and Ocelli ;
before transforming to the adult stage it becomes almost black in color.
It averages 1.8 mm. in length. The head and tip of abdomen are bent
slightly forward. The wing pads, legs, and antennae, folded close to the
body, are visible through the thin pupal skin (PI. 4, fig. E).

length of pupal stage

The length of the pupal stage varies greatly even in midsummer.
This stage requires about 10 days under favorable field conditions in
midsummer. Under laboratory conditions approximating out-of-door
temperatures the pupal stage varied from 10 to 52 days, the long pupal
stages being recorded in the months of March and April.

Larvae, after hibernating through the winter, pupated in the spring
and showed the following average period in the pupal stage: In March,
23 pupae averaged 14 days; in April, 39 averaged 23 days; in May, 8
averaged 21 days; and in June, 4 averaged 18 days.

ADULT

The adult (PL 4, fig. A), upon emerging from the thin pupal skin,
finds itself completely inclosed within the alfalfa seed and within the seed
pod. It at once gnaws its way out, escaping by the small irregular open¬
ing which it makes.

CHOICE OF HOST PLANTS

H. medicagmis was not found to be present as a parasite of S. funebris
when the latter infested the seed of red clover. This was true even where
the red clover was taken near alfalfa fields and H . medicaginis was known
to be present.

RELATIVE PROPORTION OF SEXES

Some localities from which this species was reared showed apparently
no males, while in other localities a few males were found. The pro¬
portion of males is, however, very small to that of the females. Reared
adults were counted to get the proportion of sexes. It was found that
270 of these were females and 9 were males. This showed a ratio of 1
male to 30 females.

SEASONAL HISTORY

In western Arizona and southern California H, medicaginis appears in
the adult stage as early as the month of March. It is in its greatest
abundance during July and August on irrigated alfalfa fields. On drier
lands, where the seeds are subjected to desert conditions, many of the
larvae are driven to an early dormant period and the adults become less
abundant in the hot months. Under extremely variable conditions there
are from one to at least four generations in a single season. One larva

Oct. 23, 1916

Life History of Habrocytus Medicaginis

153

removed with its host from an alfalfa seed collected on December 19,
1913, transformed to the pupal stage in May, 1915, and emerged as an
adult on June 2, 1915. On the other hand, a newly hatched larva of
this species was placed upon its host on April 22, entered the pupal stage
on May 5, and emerged as an adult on May 19. Another newly hatched
larva placed under observation on September 5 had developed, pupated,
and emerged as an adult by September 24. Infested alfalfa seeds which
were collected on October 1, 1912, showed an adult of H. medicaginis
emerging as late as April 9, 1914. These observations show that a period
of from about 30 days to 1 year, and almost 2 years in exceptional cases,
may be required for the completion of a single generation.

HIBERNATION

H . medicaginis hibernates in the larval stage within the infested alfalfa
seeds which remain on the standing alfalfa, or on the ground when winter
approaches. The undeveloped larvae and those still in the pupal stage
are usually killed by the first severe frost. In the mild climate of southern
California occasional individuals of this species hibernate in the pupal
stage. Nearly 100 larvae of H . medicaginis were removed from their
natural inclosure within the alfalfa seeds and placed in cavities between
two layers of sheet cork. Of these larvae 74 lived throughout the winter,
entering the pupal stage in the months of March, April, and May at
Glendale, Cal.

RATE OF PARASITISM

While this species is generally distributed throughout the alfalfa seed¬
growing districts of the United States, the rate of parasitism is not so
large as might be expected. The comparative rearings of H . medicaginis
and their host (B. funebris) show parasitism by H. medicaginis in several
localities to be about as follows: Corcoran, Cal., 0.8 per cent; Tulare,
Cal., 2.8 per cent; Chino, Cal., 2.8 per cent; and Yuma, Ariz., 4.9 per
cent.

PLATE 4

Habrocytus medicaginis:

Fig. A —Adult.

Fig. B. — Cages for rearing parasite larvae.

Fig. C. — Larva.

Fig. D. — Larva destroying its host larva.

Fig. E.— Pupa.

DAILY TRANSPIRATION DURING THE NORMAL
GROWTH PERIOD AND ITS CORRELATION WITH
THE WEATHER

By Lyman J. Briggs, Biophysicist in Charge , Biophysical Investigations , and H. L.

Shantz, Plant Physiologist , Alkali and Drought Resistant Plant Investigations ,

Bureau of Plant Industry

This paper deals with the daily transpiration of a part of the crop
plants included in the water-requirement experiments at Akron, Colo.,
in 1914 and 1915.1 The principal objects of the measurements were the
determination of (1) the march of transpiration during the growth period,
and (2) the extent to which the daily transpiration is correlated with
various weather factors.

EXPERIMENTAL METHODS

The plants were grown in galvanized -iron pots, containing about 115
kgm. of soil and provided with tight-fitting covers with openings for
the stems of the plants. The annular space between the cover and the
stem of each plant was sealed with a plastic wax. Direct evaporation
from the soil was thus avoided and the loss of water limited to trans¬
piration.3

Twenty-two crops were included in the daily weighings in 1914 and
23 crops in 1915. Each crop was represented by six pots of plants
(PI. 5) weighed independently. The weighings were made with an
accuracy of 0.1 kgm. by means of a spring balance checked before and
after each series of weighings against a standard weight of 130 kgm.
The balance and weighing device are shown in Plate 6. The daily weights
served also as a basis for determining the quantity of water to be added
daily to each pot to insure an adequate water supply.

In differentiating between the transpiration of consecutive days it is
desirable that the weighings be made at a time when the plants are
losing very little water. Automatic records show that the transpira¬
tion is at a minimum just before sunrise.3 The daily weighings, which
required the time of three men for an hour, were accordingly begun in
the morning as soon as there was light enough to work and completed
before the transpiration response to sunlight had set in.

1 Acknowledgment is gratefully made of the efficient and valued assistance of Messrs. R. D. Piemeisel,
F. A. Cajori, P. N. Peter, J. D. Hird, G. Crawford, R. D. Rands, A. McG. Peter, H. W. Markward, H.
Shattyn, and T. R. Henault at Akron in 1914 and 1915. Mr. W. H. Heal has also aided very materially
in the reduction of the measurements.

2 Briggs, h. J., and Shantz, H. D. The water requirement of plants. I.— Investigations in the Great
Plains in 1910 and 1911. TT. S. Dept. Agr. Bur. Plant Indus. Bui. 284, p. 9. 19x3.

3 - Hourly transpiration rate on clear days as determined by cyclic environmental factors.

In Jour. Agr. Research, v. 5, no. 14, p. 583-650, 22 fig., pi. 53-55. 1916. literature cited, p. 648-649.

655)

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.
fx

Vol. VII, No. 4
Oct. 23, 1916

G— 95

Journal of Agricultural Research

Vol. VII, No. 4

156

The transpiration of the plants in each pot, as determined by the weigh¬
ings, was plotted daily as a check on the weighing and watering records.
The daily transpiration of the first five crops of the 1914 series, traced
directly from the original graphs, is shown in figure 1. These graphs are
typical of the series and show the proportional response of the individual
pots of plants to the fluctuations in weather factors. The daily trans¬
piration of each crop is represented by the mean value of the six indi¬
vidual determinations, which minimizes slight errors in the weights of
the individual pots, and abnormalities in the transpiration rate of indi¬
vidual plants. Inspection of figure 1 will show the close agreement of
the individual determinations. While the plants in some pots of a given
series transpire more than others, owing to differences in stand or size
of the plants, the daily fluctuations are very nearly proportional.

The weather factors measured included solar radiation, air temperature,
wet-bulb depression, and wind velocity. These factors, as well as
evaporation, were integrated for each day. The solar radiation and the
wet-bulb depression were measured by differential thermographs and the
air temperature by a standardized air thermograph. The wind velocity
was recorded by an anemometer located 3 feet above the ground. The
evaporation was measured by means of a shallow blackened tank 6,540
sq. cm. in area, exposed at the level of the plants, and also by means of
a large tank 8 feet in diameter and 2 feet deep, with the water surface
at the ground level.1

DAILY TRANSPIRATION AND THE DAILY INTENSITY OF THE
WEATHER FACTORS DURING THE GROWTH PERIOD

MEASUREMENTS IN 1914

The daily transpiration of 22 crops grown in 1914 is given in Table I,
the daily loss being expressed in kilograms per pot. The small grains
were well established before the daily weighings were begun, and had
lost during the previous month approximately 10 per cent of the total
water transpired during the entire growth period. The daily weighings
in the case of the other crops cover the entire growth period after the
daily loss had reached one-tenth of a kilogram or more per pot.

1 For a further description of the methods and apparatus employed, see Briggs and Shantz, op. cit., 1916,
p. 584-585, 625; and Briggs, E. J., and Belz, J. O. Dry farming in relation to rainfall and evaporation.
U. S. Dept. Agr. Bur. Plant Indus. Bui. 188, p. 17. 1910.

SL//7T OAT SWEDISH SELECT OAT HAA/A/CHEA/ BAALEy GALGAI.OS WHEAT HV&4NAC4 WHEAT

Oct. 23, 1916 Daily T ranspiration during Normal Growth Period

157

Fig. x,— Graphs showing the daily transpiration from the individual pots of plants which constituted the
first five sets in the transpiration measurements in 1914.

Table I.— Measurements of environmental factors and daily transpiration , June II to September 25, 1914

158

Journal of Agricultural Research voi. vii, No. 4

HOO O O'C CiK) H WlCQOHHIOiOOOOiOODU) tOO M N® H ^«00

W « « . . . . . . .

, . H H MO lO f' CO HI N M N H M H H H H M CO

lO tO . N h

'too N to t' O to

H H N . . . "

. . tox o O

N X , N

Isis'S

N to *" o NH'OOO'O H N O t' »"• O O O O N H t' O O Oi
H OQ CO COCOtOCOHWNHHHH H H * ’ H H N H OT

't'O O cox m N VO o

X CO N . . . O 't t-*

. . (O to t' H c

N 00 . to N

N to tt

H . .

N >0 N

NN^NHO«tOOiOOOOHN(<) 'tO O 00 O

tO'S‘tO^‘«tONHH«NHHH H HI W H lO

it) n h tn 't to in t".

O X to . , .00 't'O
. . N O t' O

O to . N H

(ONN

o . .

W to H

ONnt'OoOO«t'Oit'OiOiOON« >000 O n

COCOfOCONNNHHHH ' H ' * H M M H tO

O' t> t>

to . .

N N lO

'too O to to 'too toO 't h O O O vO N tot'-HOO to

tO^‘4,'^'NfONHNNWHMM * * H H N H O

to Ov N O O H rot' OO
O O 't . . . O to t' to
. . to to fOO 00

N 00 , to H N

to tooo

f' . .

H to N

too to H O N CO N O X O X t' 00 to N 't tooo to H
tO'ttO'tNfOMWWMM . hhhmio

t' t'O . . , 00 to t' to

, , n h t« O

0 to . to H «

00 N to

HI , .

H lflN

00 tooo to t' 00 t' O N N to to to to N O N H to t) ft)

N CO N CO H N H " H H H . HMHHlt

NWN Ot to Of too to
M H "t . . . Ot to t'

. . N ro to to

NO . [OH N

t"tO

00 . .

h rr «

O Ot to H H It) Of H to to t'O O to N *0 N 'J-'ttO
tOrOtO'tNtOHHMHH . HHHH'i-

00 Ntf H «o OtO 00 't
to tt Of . . . C- too 00
. . n o to o ot

0 t' . W H N

iooO o

H U) «

to OtO h t'O to t> CC 00 00 to N to H o H 00 Of 00 o
N N N CO H N H . H ’ ' ’to

to 't Of

t' . .

HOW

ION N 00 Ot O fOt'OOOOOitow NO 0 tooo O OtO

N to to to H to H . H HI ’ 't

O t'COO t't't'rOH
00 *0 Of . . . 00 to t'

. . N O H M C

O t' . to H N

NH to
o . .

H to N

O H 000 Of O too Of 0\ O to rj- 'f CO N to Ot 0 Of Ot

NtOtOtOHfOHH * H . H " H to

tO Ot H CO ctOfHOO tO't
H to Ot . . . Of to t> t'
. . N N 't to Ot

to l> . to H N

N tO '<■
00 . .
H t> CO

to h to Of 00 H coOt't'O'ttotON 0 to f'OO 00 O'

N tO tO CO H to H . . H ’ * (O

1 t'M t't'roOtO'frO
1 00 to . . . X in l> to
. N H hi CO Of

I to . to H N

X CO O

t' , .

H N H

H to X tox t' Of t' to to to N N N H o to tot© O N

NNNCOHN* . M ‘ * (J)

H 0 CO t'X O 0 tOX
X Ot Ot . , .X too
. .mono
X 'S' . N H N

OlOCONt'N N CO CO

t'O X . . . •£' too

. . H N H t>

O tj- . N H H

t' N H Tf

?£o dt

o n rt-t'XO x't

CO to . . . t' toO

, H Tj- H t'

ONHO'J'OHt'
H IN , , , t> tOO

. H O o o

to . N H H

OXt'NNOftO'l-NNCOOOHOOXCO'tcOtO

MHHNHH . * . . ’ N

O t'

O N

't'ttoOtotONOOHO • * H o to N N N Tt

H H H H

Xt' nnnnnnhhhmhhhhhOhhhOco

O H

O to

OMNOt'NO'tNNtOCOHNNOXNtOtOCO

NNNNHN . ’ N

ioXOOco*''t'tNNHOOOOOONtoNX

HI HI M N M HI* H

0 O

06 to

NNHCOOtOHH • - NHOOhONOONN

?N O n OX'tOO N N w OO'tcoX't'OON
H OO OO t'XX ON 'J-H HI H NO M N O to
H| | H | HMMHHNNNN'fCOHHIHICOH

t'Hitoiiiiiiiiiiriiiri
Ot'X to O to H t't'X tOH o OtOOtotot'
O OO t'X XhcoOhmhioOhON

hnnnn'S-cohhhn

Jo o O o o O U.-

« O'

> tfl

£3 tJ X

rt.n cd<

$ rt 4>
g W.T3

3 fl+i a g+j o'^
gagas ST3 ¥

^ j *3 +3 -P 4J

^ -HO

jy a3

4) *1 0> tf)

ri S3 4* S2

.a* :’SJi

^4-+-+-+-+) QJ 57
H»hHH«hH+J43

atj.:

'Bn

M a

ill*

*43,0 M'S

^5 hT 4)

Ilfrlil

as,&s

S1*"Jf5|llf4a4l98fi

h i-[ +-> T3 J? ^ 0*J3I3 S3 S 6 £f u T3

Cs. S ? t?0 0 0 0 0 3

vxwwvx

WWWh

c3 c3 ciS cS

ill5!

JUI*Y I TO 22

oct. 23. 1916 Daily T ranspiration during Normal Growth Period

159

July.

O' if H N H 00 « M ^ VO VO N O' t" OvO OO H to O' H00 <30 O 0 H

R If H H H H H *HH ‘ M H M

tO tfr . N H H

if CO if N O0 N 0 to N to H H to <00'0'0'00'HONt'*'Oco O' O'

® ^ *^00 n 6 4 cocococohnconnn ci co « n

CO O . CO H N H

H M

0 h TfH n ot 00 co to O' 00 if m if h 'cttj-toto'ooo h h t> <o v©

N Jsyj 4 . , 5n vO JT- J> *jr * , . . . • * « *

. , 01 H h <n 0V l-l 10 C* fO(OCOClHM«HH« « «

0 vO < fOH Cl

CO to to 0 O O 1" O' 00 O' 00 00 to 0,0000 CO 0 co to O O' H H if to H

* T h* to to <J to vo fl, ^ oi h hmhh*mmh*h h h m h

vO if . N M N

00 00 to N CO N N vO to to if to O' 0 O O 00 to 00 O' t" to 00 VO VO VO O'

, . W d co O sOCI HHH

lO CO . Cl H H

Jr-

vO O N O' if H t- O if CO vo co OOHOOO'OO'MHCO COCO O' 00

QOvOOO*- tOG lO VO tr> *0 * • • ■«•••*■** * • • *

. . >0 to O' t- H O' if NCOCONNHHHHHHHH

0 O' • N H H

’ H

t-- N t' N *- H H N O if O' N if tovo O 00 00 OvO 00 0 O' to w «

, # t^s© H M 00 to tO tO CO Cl H 01 W M H H« Cl Cl

« OV 1 Ot H Cl

vO 00 to 0 to H LO to O' Q O O 0 0 tovo 00 n to 00 O 00 O N N h co

. tTt?o to 06 vd ^ *" oo cf vo co <t44o« tococi « co coco co to

if O' . CO H N

O vo O0 00 CO if H if vo CO- if N CQ N if vO H O' t'* t"*vO H to N ^ if N

a n . . Q\ w , t «*•«••«**> * * * *

. , « M tt t»HVO N tO(t)(OCOHN«HN<l N N N 01

O' 0» , CO H N

vo to to to O O 01 Ov o 0t H 00 H O' tovo O H O to O' O' O' N O O'

qq n tj- , on to it— Jr— oQ * » ****■•*•#* * * • •

. . N to to to 00 H If N CO^-'t'tCI CO CO H N N « CO CO Cl

H vo . co W «

O' co O O' O' 00 CO t" CO CO CO CO to O'OOHC'.^co coQO O' O O' O' *"»

oo to 5, . . * O'tot'H to • • . •* • *

. . H to to « to H co Ht MCOCOCOHCINHHH «H H H

tf . to n «

'O h O' O' 00 h CO Ov H w vo H 'J-OvO'O^co'OO'NH wh O O

” to to c^ ^ ol « or e5 4444<ifot!jHW(o coco co to

« 00 . CO 11 «

o- ct OO OivO O to o « vO CO t'' 00 H O' O' to co to 00 vo 00 H O' O'

o VO H . . . O' VO to H . . . t. • •

.jTftotom O' « oo co 't'O Tj-o-tt totOH IN « CO « n «

to 0 . COH «

H M

o- r- Tf to O O' W O' to co oo Tf to 0 -o- j> to 00 vo h o h t-co tJ- rf-

m r» n . . . O' to t> 0 ^ i • •••••••••* i • • ■

, , Cttow to Owtico too UIOCI cocow coco coco CO to

rf O . CO H o<

H M M

N to CC O’ to O' Tt VO to t}- 00 CO OO vOWO'f'toco (OvO 10*0 H O' O'

Os Tt Jr— . . . Ov ir> t— J> <M , * . . * « * f * • • *

. . CO tr to to OiMvON 't'O ’t^-CttOOHCin CON N «

CO O' • CO H «

t'.

co too o co m to o O' to O' o 00r»O'O'H'ot>'tO'O co O' h O'

« Oh,. .00 to t'* to ■'}" . • . . *• ■ . •

. .COON H O' H ION tOtfrCOCQNNNHHN « H « H

H 00 . to H N

fO H J> H co H 00 vo O O' co VO 't ("to tocioo (t'O COH to N H N

"* 00 'f . . . 00 lo C» Ct C» , . . *• • *

; . ro H to H o. H t^eo CO’^-'t^-NNCOHNN «N « N

to 00 . CO H «

O HTfHCO N 00 VO .N O N N O' NvO H to O' "t ’Cl tf O' O Nt^OO 00

rr t"0 ., . oo to i» vo vo . . . *■ • *

. . ot H ro N 00 w 't H COWtCCOH M « H H M NH H H

0 VO . CO M «

oo tt Tf Ovoo h It IT) o « H VO it OC'ONC'Owcot'O' 00V) it *0

0 tJ-h,, .00 vj ti « h . . . .. . •

. .NOON H 00 H t-' CO CO CO tO CO M N Ct M H H H H H H

Ol to . N M N

« vO 0 0 't vO vO O' Ov It vo CO it to covo ONOOcoitOO H co to vO

co too,.. .oo itvo t>t>. . . . .. . .

. .NOOvO OVH ION CO'tfCO'fNNNHNN NH H H

N t' , to N

it CO to O' t- vo 't CO O' it 0 OVO it tovO O' N 00 00 tooO O' O' N it to

00 t-o .. .00 tovo H J> . . . . .. . .

. .NC0H o OtHOO co COltCOCONNHHMH HH H H

N vo . N H N

Ci to 0 vO N O' CO it 'O to 00 CO 00 It 00 0 0 rf H N it it 00 00 0

•« co <o , . . riiovO cico . . . .

. . N to N 00 OOHVON NCONCOMNHHHMH H

O' VO . N M H

! .... . . ! . . . . . VONvONOOOitOvON NN o vO

. . . . . . . . . i • • • | h O' O O' O r-oo oo O' « it h h

. . . - . . . . • ■ ■ • H(fH{HHHHH NN N N

. .... , . , * , . , . C^ H f tO 1 1 1 1 1 II 1 1

. .... , , . . . . . . ^ fi* 00 CO ^ to H ti t*oo to H

. .... . , , . . . . . O'0'Ot-.OOOOHroOH

. • . - • • . . . ■ • • • HMHHH NN N N

Evaporation (shallow
tank), kgm. per sq. meter
Evaporation (deep tank),

mm . . .

Evaporation . inches. .

ir temperature, max- °C .
.ir temperature, min. °C .
.ir temperature, inte¬
grated mean . °C. .

ir temperature, max.,

°F .

ir temperature, min.,

°F .

ir temperature, inte¬
grated , mean . °F . .

itegrated radiation,

cal. per sq. cm .

itegrated wet-bulb de¬
pression. _ _ _ deg. hr. .

iTind velocity,

miles per hr .

find velocity,

meters per sec .

>aily transpiration,
kgm.:

Wheat, Kubanka .

Wheat, Galgalos .

Oat, Swedish Select. . .

Oat, Burt .

Barley, Hannchen.

Rye, spring .

Cowpea .

I^upine .

Millet, Kursk .

Millet, Siberian .

Corn, Northwestern

Dent .

Com, Algeria .

Sorghum, Minnesota

Amber .

Sorghum, Dakota i
Amber .

TablK I. — Measurements of environmental factors and daily transpiration , fune n to September 15, 1914 — Continued

160

Journal of Agricultural Research

Vol. VII, No. 4

oct 23» 1916 Daily T ranspiration during Normal Growth Period

161

Table? I.— Measurements of environmental factors and daily transpiration , June n to September 25, IQ14 — Continued

AUG. 14 TO SEPT. 4

162

Journal of Agricultural Research voi. vn. No. 4

September.

Cl Tj- PO O ««t~ Cl vQ 0^ W H •• • •OOO'tN W>vO *0

Cl On PO . . . Ot no f- O h . • •• • . .

. . « PO CO « N© M ^ Cl •• • * H W Cl Cl Cl

H in . Cl ....

<•0

N 't'ONO vO 0 O Q\ O *" <4 O •• • ? NWifO « 'tfOO

H Cl Cl.. . 5\ to VO W t> , . . • •

. . N«00 OOH'td • • Cl Cl Cl Cl

0 >0 . (OH « ....

H ....

vO tJ- l> oO vO 00 n h co h co to .. . •vOt'.roOvHOvO

t' 0 t-- . . . 00 I/O N© ^ N© i . *• * . .

MC'OC' H C-» (O •• * • H Cl 01 Cl

O O' , Cl M H ....

O 0 ft « Vt it W> O 00 t-OOO 1/0 • ■ • . <r>vO H uo t' N If

CO *0 On . . .OOTfiOCl 0 .. • •• • ........

. . N CO Vf rf O H C> PO • ■ • • H l-l H Cl

ON C' » Cl H , , , ,

August.

t" PO vOTj-^t O loft 00 O 00 00 H •• CO HOOOrO*'"OHt'.

no i-» 00 . . . £30 if vo h t- . . •• . .

H ON On 0 00 H *J- W • - H H H H Cl Cl Cl

0 . ci ci . • •

0 00 'O no 0 0 C' « 00 0 h r(- 0 po On h OO n s i(- 10 no h

n Os On.. .00 C O « , . .

. . M 0 X) O NO H Cl HH H H H ON

00 't . CO ci

00 ft 00 ft n P0 'I" NO NO PO C" Tf 0 'Cj- H M H hMtXN 10 NO n

co .« 0 . . . 00 it 0 00 co . . .

. . PI 00 C" 00 l-' H rj- Cl HHH HHHC1

00 Vo . « H

On <t 00 Nf (o O t' h co to On po On co hnomihmpic.

no IOH., . 00 Tfr NO CO to . . .

. . N00 Nj ft C'.HTt'H HH HHHH

00 NO . d H

Nt NO 00 H O H Cl O H 10 NO PO On Oft N 00 C- 10 TfrOO <30 00 H

NO f> • . .C^VONO H no . . .

H (1 0 NO NO Pt H H

NO to . « H H

'to HOON Cl Cl NO PO NO O H NO Cl Cl . H . 1/3 to ioo0

t-' NO 00 . . .OOPtNO PO NO . . . . _ .

. . h *■■. r- t*. rt oo co O O'*

C" ^ . (1 H ■

oo O t'.t'.it 0 h On On "vo O O O' "t to h to - co co t}-no

PO O CP . . ,^*t NO NO NO . . .

♦ i H h Oi m m Ov^OO

t/> VI . N H

00 N 0 *-"0 H ON H o J>CO»000 NO Cl NO vooO On h Cl H Cl t'.

f' m oo . . . 00 »o o to On . . .. . . .

. . Cl H o H oqhoOpo hmh MHHM

N 1> , CO H Cl

n ci O ftN nO to On co CO 00 0 nO co 't Clvot'Cl On 00 h to

N H00.. ,00 X1 « TtVO. . . .

. . N00HO 00 H NO CO HHH HH

0 O' . « H «

On O CO O 00 0 no no 00 ci h On nO © po. to « no NO ci t-43 00 On

oo to 't.. .co t e 0 co . . .

. . Cl o C-* O NO H NO Cl HMH

C" NO . PO N

0 •'t O On O H Nj ft O 't O CO to No H ft C0 xi«C t CO no w

CO 00 CO.. . 00 NO 1> t" H . . .

. . Cl 00 NO H NOHt'.CO H o

On »o , ci h ci

00 00 NOcoco hcsnO O J> O Cl ON O N nO *tG0 © *}-'© NO lOOO

t- M , . . ON NO O NO O . . - . .

. Cl PO CO H NO H *4 H HHHH

00 NO . CO H «

PO H H H PO O OO NO H NO tt ON H NO CO CO PO t X) H POC1 POCO

POM O.. . 00 NO t-' On H , . ,. , . .

. . Cl H PO H j> H N© to HHH

On no . co h ci

to H Tj-N© ON H 1> C- o Cl to H « HH H On • ...

H t'.NO.. . 00 NO CO 't . • .. • •• ...

. . Cl O CO H '©Ht'.PO HH • O • • • O •

H N© , CO H « .....

w • • • • *

(30 00 CO to O t H ft N© o o N© to N't Nt PO NO It CO PO to CO

to N O.. . O NO On CO . . .

. . N 00 no *e}" H N© Cl NO ei H Cl Cl Cl O

Cl NO . co H Cl

g *5 H §n§ a 8 'S ■” ? ^ *r??

. , co no n© h on to no ci h co ci ci 0

J oo , tow «

N© ci o t"t N00O0 o O On H Ci C*-h ft » ci

i" ^ h . . . On *o 00 h h . . . . . ... .own >n

. . tJ-n© TtN© M to f to H CO N Cl UU U

NO O . CO H Cl

H H

PO ci cotN Tt 't Cl VO On *t O i> to 0 h h s n t-H On

oo On po . . . Onn© is co ^ . . . . . .

. . Cl 0-43 't 00 Cl NO Cl H Cl Cl Cl o Cl co^ «

H NO . CO H Cl

Pot No.

■ ■ * . • > • ■ • ■ • • • « N o N© Tt PO CO to O N

■ > • . . ■ • ■ ■ ■ . . . NTt H HHNNOHClOCO

■ • ■ ■ • • . . . " " ■ . Cl Cl Cl Cl'tCOMHHCOH

::::::::::::: U L t ^/okioiok

■ * ■ ... . . . . . H CO O HMHNOOHOiC!

■ * ... ... . . . , . «N Cl Cl'tPOHHMdH

Item.

Bvaporatou (shallow
tank), kgm. per sq.

meter .

Evaporation (deep tank),

mm . . . .

Evaporation (deep tank) ,

inches .

Air temperature, max., °C .
Air temperature, min. , °C .
Air temperature, inte¬
grated mean °C .

Air temperature , max . , °F .
Air temperature, min., °E.
Air temperature, inte¬
grated mean, °F .

Integrated radiation, cal.

per sq. cm .

Integrated wet-bulb de¬
pression, deg. hr .

Wind velocity, miles per

hr .

Wind velocity, meters per

sec. . . . . .

Daily transpiration, kgms:
Com, Northwestern

Dent .

Com, Algeria. . .

Sorghum, Minnesota

Amber .

Sorghum, Dakota

Amber .

Sudan grass . .

Sudan grass (in open) .

Amaranthus. . . .

Alfalfa, E23-20-52 .

Alfalfa, E23 .

Alfalfa, E23 (in open).
AJfalfa, 162-98A .

Oct. S3, i9ts Daily T ranspiration during Normal Growth Period

163

September.

00 ^0 ft « to ^ m 0 *0 to «

^ H N . . • CO ^ HO V> Cl . . . .

. . moo 00 00 N* « Oh CO Tl- ^1-

« 00 • « H

*r<g$ $ $ v***:*!?*

..Ct'OOO'O 00 H V « M co co co

00 wj • Ct M

O N H 0| "«*■ M CO O 00 to Ct Ct Tt N* OH u%

co'O 00.. . co *n "<*■«. .

..mmco« oOH^fH « c* « to

OH *t • Cl M

v© O CO 0 00 OhC© « O CO >fl « « 0 O ^<OH

Q ■<4- t- ■ • .0 (*} lO « . .

. . H 0 Ct CO t^HCOH « Ct «S *0

'O . Ct M

novO tfO'O O 00 h O « 't Ct h

m ou> . . . 0 co to Oh . . - • •

..MOt O « H Oh ^ Cl CO CO CO

0 SO • Ct 1 H

« 00 O h oO HO0 V) O 00 n£> 0O 00 00 wo ^ wi if

00 00 H . . . 00 to tH co Oh . • • • * • JL * •

. , «5 H « H C- M 0 ’4" HH CO 'S’ 'f *0

fO C* ■ fO M 01 H

'C'OHHCt tOOO't Oh to OlOt't Oh « if COO0 to

t. h s , . . 00 *0 vO t» tO » . .

. . CO H W O N M OMf H CO to

O 00 • CO H Ct

if ct ■o ct ho 00 O co 00 <g if ct 0 w 't f"0 00 C''

. et ct to ct 00 ct Oh 41 * h * to ^ ^ to

N C* . (O H Ct

JlVlH H00 H CO to O O’ O CO 0\ « « O

00 ot' . . • 00 >0 f- <0 if . . .

..HMMM O H tH Ct CO CO if

rr . CO 1-t ct

'fcMJiH h O O CO Oh h *0 00 h to if <1 " ho O

NOv)., . N- tr to co if • .

. . hhoo >0 c- h if ct « « co co

N« if . Ct H ^

OO W H ON N- Ct if H co Ct t-t if H H HO H OlOO V)

!HtfN , . . OH V fi IOVO . .

. . CO COVO M N* Ct Ct IO H H CO CO CO 't

if OH • CO Ct M

10 if 00 h f- 00 O vi if co cc co co ctcoetoococt'O

00 v, H . . . *>• co O Ct Ct .

• . « H H N 00 H 0 CO H Ct Ct Ct

Oh to . ct h

cowetoOvO oO^-h \o ct ho ct to ct ct ct to 00 0 co

hOhho.. . 0 « if a » .

C- . If to Ct M M H Cl

V)< . H 1

Ct O co 00 O CO H H to N. O H 0, if if to O Oh Oh

h to to . • . OH ^ HO O H . .

. . if ct ho oO fl h 10 h co co co ^

co H . CO H H

M M

N~ O Oh 00 co t'.Htd m co to co O Ct to Oi co tf Cl

■OON... .Onto N- to N... .

. , Ct Ct CO H N' H a tf H H CtCOCO'T

OhN . CO H Cl

to O' to co O' COCOOH to co 00 O' >0 NatfiflO h ^

H Ct 00 . . .00 t « M . .

. • ct 00 Oh 00 00 Htoct ct co co co

0 N . fl H

HO HD to 00 0 CO fl 0 to O' tJ- Ct co iohO Ct Oh rfnO h

>o a a . . • 00 »o \o 0 .

. . d t-~ 0 00 to tow H Ct ct CO

O' • fl M H

h 10 00 Oh tr fl N N- co HO to O' rtO H N. O ct N.

CO « 0 • • . OO to O N. to .. . .

. . fl n ro Oh to N* co h ct ct ct

NIO . fl H H

N-

ct to CO 't Oh 0 to 00 00 Ct HO0 V) HO N 10 O O' O' HO

vo hO 00 . . • 00 ^ HO O' 00 .

. . H Ov 00 O ho N*co Ct Ct ct ct

00 • Ct ct

CO HO VI N h O' 0 Ct N CO N> O HO to HO to H co tf 0

HOtJ-w.. . 00 to HO N . .

. . flHO H OH to H 00 co Ct Ct Ct CO

N tO 1 fl H H

CO HO "t ct O hOO O Ht HO CO 00 a O Hf VI U, 0 N

O ON . . .Onto W ct .

. . Ct Ct 0 w Ct HO ct H Ct Ct co CO

<n so ■ co h

Pot No.

***•• t • f • • »«•

+ *■*• • * * * • lit ^ f^oft ^ O O ci

i • * » * ##• * * * * * H H W 0 (O

• • * • » • ••• ^ H H H H

• * * • * • * * • ■

1 ♦ ♦ • * * * * • * ♦ » 1 • O w fO w ^ N

. * % * m • HH^OOHOvCI

. ... . i ... (^) H H H « h

!«*•* .** . | «.*

Evaporation (shallow tank),

kgm. per sq. cm .

Evaporation (deep tank), mm.
Evaporation (deep tank), inches.

Air temperature, max., °C .

Air temperature, min., °C .

Air temperature, integrated

mean, °C .

Air temperature, max., °F .

Air temperature, min., °E .

Air temperature, integrated

mean, °F .

Integrated radiation, cal. per

sq. cm .

Integrated wet-bulb depression,
deg. hr.. . .

Wind velocity, /""l® perhr"'

vciuHjuy , ^meters per sec

Daily transpiration, kgm.:

Sudan grass .

Sudan grass (in open) .

Amaranthus .

Alfalfa, E23-20-S2 .

Alfalfa, E23 . . .

Alfalfa, E23 (in open) .

Alfalfa, 162-98A .

55858°— 16 - 2

164

Journal of Agricultural Research

Vol. vn. No. 4

The evaporation from the shallow tank as given in Table I is expressed
in kilograms per square meter of water surface per day, calculated from
the observed evaporation from a shallow blackened tank 6,540 sq. cm. in
area and 2.5 cm. in depth, the depth of water being maintained auto¬
matically at about 1 cm. The evaporation from the deep tank (8 feet,
or 243 cm., in diameter; depth of water, 50 cm.) is expressed in terms of
the thickness of the layer of water evaporated each day, given both in
inches and in millimeters. Since the loss of 1 kgm. of water from an area
of 1 square meter represents a sheet of water 1 mm. in thickness, the
daily evaporation from the two tanks is easily compared.

The maximum, minimum, and mean temperature of each day is given
in Table I in both Fahrenheit and centigrade units. The mean daily
temperature was determined by integrating the area bounded by the
thermograph record with the aid of a planimeter, which gives a better
representation of the mean temperature than the mean of the maximum
and minimum values.

The daily radiation represents the total number of small calories
received during the day on a surface 1 sq. cm. in area kept normal to the
sun's rays. The radiation values given in Table I were computed from
the records of a differential thermograph calibrated by means of a stand¬
ardized Abbot silver-disk pyrheliometer.

The wet-bulb depression is expressed in hour degrees on both tempera¬
ture scales and represents the summation of the depression for each hour
of the day beginning at 5 a. m.

The mean wind velocity for the day as measured by a Robinson
anemometer 3 feet above the ground is given in miles per hour and
meters per second.

The daily values of the weather factors and the daily transpiration
during the growing season of 1914 as given in Table I are plotted in
figure 2. The graphs of the two evaporation tanks are seen to be similar
though not identical. This similarity is of special interest when the
difference in the hourly distribution of the evaporation from the two
tanks is considered. The loss from the small shallow tank is confined
almost wholly to the daylight hours, while the large deep tank shows a
marked evaporation at night, due to the heat stored during the day in
the large volume of water. The writers will show later that the daily
transpiration is more closely correlated with the daily evaporation from
the shallow tank than with that from the deep tank.

The daily transpiration is, of course, dependent not only upon the
environment but upon the relative water requirement of the various
species, the size of the plants, and the stage of growth. The daily trans¬
piration is therefore supplemented in Table II with a statement of the
period of growth, yield of dry matter, and water requirement of each
variety. To this table has also been added the variety of plant used,
the botanical name, and the pot numbers.

. /OA/S /&/# iSOLy' /41/&. «S2f7®7?

55858*— 16. (To tace page 164.)

Table) II. — Period of growth , yield, and water requirement of plants used in IQ14 transpiration measun

oct. 23. 1916 Daily T ranspiration during Normal Growth Period

ta' ’SI

isgsigg

-H-H-H-H-H-H-H-H-H-H-H-H-H-H

tO'O Tf MHO tj- fo 't H

-H-H-H -H-H-H -H-H-H-H -H-H-H-H
S SS o"8 S r?S<£"8 S':

fO 10 ro 't to to « to \Q

.*!

165

TABlyU II. — Period of growth , yield , and water requirement of plants used in IQ14 transpiration measurements — Continued

166

Journal of Agricultural Research

Vol.VII No. 4

Oct. 23, 1916 Daily T ranspiration during Normal Growth Period

167

While large daily fluctuations occur in all the physical factors, the
graphs do not show a marked “run” as the season advances. In other
words, the seasonal change is not large. The evaporation shows a slight
maximum in early August. The radiation is at its maximum in the
latter part of June, at which time the sun reaches its greatest altitude.
From this time onward the maximum intensity of the radiation (repre¬
senting cloudless days) gradually decreases. The wind velocity, on the
other hand, tends to increase slowly as the season advances. The maxi¬
mum temperature and maximum wet-bulb depression occur during the
middle of August.

MEASUREMENTS IN 1915

The daily transpiration measurements in 1915 were begun as soon as
the crops were established and include, therefore, the whole growth
period (Table III). The season was exceptionally rainy, which was
unfortunate from the standpoint of the transpiration measurements, as
the plants were often so wet in the morning that it was impossible to
determine the transpiration of the preceding day. In such cases the
mean transpiration has been given for the two or three days included in
the rainy period. Such breaks in the daily record are indicated by
dotted lines in figure 3, in which the data in Table III are presented
graphically. Supplementary data relative to the plants used in the 1915
measurements are given in Table IV.

Table III. — Measurements of environmental factors and daily transpiration, May 22 to September 2of igi$

MAY 22 TO JUNU 20

168

Journal of Agricultural Research

Vol. VII, No. 4

21 TO JULY II

Oct. 23, 1916 Daily T ranspiration during Normal Growth Period 1 69

H ci Oi O ^t O moo r- w
« oo « . . . O' m oog
. , ci m v m oi

oO m . co h ci

« Ulifl t-oG m o O' C-O t'' 00 H O *t H OcO N 'fifliflifliOH

h d eicieicococico^-cocococo ^■4' . h * ci ci

m . . . i

N'O'O H
00 VO . « H «

t^oo O HsiflOHtflO'flftniO'tcifloO'OcitooO'nO'N

. .

h\o « ctdBflcmionwweinnw hh

oo co ncicicoeicicicocicoeiHcoco

ci ci vO Oi C"- co ^t co m'O vO ci co oO'0'OC1'OoOOcoOH'Ot’.coO't'*Nfcoci coo h co o

oo <n \o . . .oo m'O w *o . • . .

. . HOO M00 'O m« WHHC1HH««WC1HHeiH HH

^ ^ « H H

OlOHOcf'tt'OOMi) -d- t- m
coo oo ... m\o h Ov . .

. . h m t* o\ o m ci

O 'f . Cl H H

t'.'O 'OCOt'OOTfHOt'O'COCICOCOHHCl'OCieiO
HMMCIMMCIHflCIHHWCI . HH

OO'OOoOOOlHO'OH Cl »H

ci co cf . . . O' m'O m co . .

. . ro ci o oo o hioco

O 00 . CO H H

OioO Oi^Oih rf T|- Cl ci O' t- m ^ co Cl « H t'H H
HMHciMBnciciciciHcici . hh

o h ci . . .oo mo o

. . H Oi CO O 0\

IO CO . Cl H Cl

HHC1HH«C1«<NHH

O O cq h ci OiO ic) n ci tj- O' co sHMOJcnONOi miO mom'tciHOO

Cl t"t . . . C' T}- lo CO o.. . . .

. . H M J> CO H H Cl H HHMHHHHHHMHHHM

IO co • Cl H

CO HHHHHMH

t'.'O O Ntf CO Cl 0"0 f'
Tf co . • . O if m'O
. . h lO C* CO o

Cl CO • H H

mo OioO OGhooOiOhoOh OiO 00 00

OcoOt'OiW O oo co m

h h ci . . .oo VO O

. . co O 00 N 00

O-CO . Cl H

co t- f'CO O O oo O cooo coci ci rj-H t^H

OOcO HHHNMMciB«W«HnH

'to N h mi « n lfl m tj- oo t'-r-~mj>o,cof'Ci to h o o t- m ci ci m

C' cm . . . ■t'. mo h V • • . .

. ci ci h so O m « hhhshm«ci«cicihcih

m . ci h m

t'* m'O ci co "too cimr'O'Oi^mMCiH

Cl H H H

moo c-> oo c>. h m w
ci ci ... c^. mo

, H H CIO

c~ m OMoO coOiOmcohujciooOiN

OMcihohciw

mm. . .oo m'O

. Cl N H CO

O H
l> CO

m'O *> m'O ooo^HOHcoifio(o«cici

MKMCiHMciciflciflHHM

co og oo 'O ©in ho
o o. ■ . .oo mo
. hnooo

oo t'- mo m h me* Oi ci ci m it m t'- m -d-

ci ci « ci

f't'Ci co co h t^m
t'' ci . . .oo rr'O

. Cl C-.Q0 oo

O O

oo co

^^moi^ij mM'O'O'Ooo O O'

hhmhhhhwh

It co ci it co CO

co ... ■C' mvo

O CO h

h co ei h co'tcom'tt-oO'O ci

T)-CltsN(IHCOCO

o ci . . . m'O

. Cl H H f>.

H CO

m ci

COOOOOOOOOhhhciO'O m'O ci

C' H t*t^t>iH COCI

oo co . . . mo

. ci H H O

m . ci h h

O' © O' h OOh ci h ci o mo m O

O ci 00 'too it ci -d- 0 00 O ci ci oG ci 00 O 'O ci'OOOCO -d"
T HHciOHco'd'mco mo OO ci ci h moo

H| ) | HHHHHHHHHHClndCICOCICIHCI

C* CO Oi I I I I I I I I i 11(11 111 I I

H H CO O' o m CO H c. c. (I) c. ro V) H f- M CO co O'

0 O ci co d co m moo o h « mto h iaooi

HC1CI«CICO0l«H

a -a

a. a

O o o o H’1"*

fa.

.a a

g 0*5

Sra

!*&

$ ?

a > a»

Sa-s

.3 2 3

p|!

if iiii.

t* Xi

•l-l

-.u >

U U U uZ +Jiri .H

ddddSfi is

I! | IlIHiflt I lliliilisilllllllllii

o o
M

w,Soofl

-V

tsss^^lia0*

r l

jtt'S

844

TablB III. — Measurements of environmental factors and daily transpiration, May 22 to September 20, IQI$ — Continued

JTJI.Y 12 to ATJG. 4

Journal of Agricultural Research

Vol. VII, No. 4

170

AUG. $ TO 27

oct. s3, 1916 Daily T ranspiration during Normal Growth Period

17 1

Table III. — Measurements of environmental factors and daily transpiration, May 22 to September 20, 1915 — Continued

AUG. 28 TO SEPT. 20

172

Journal of Agricultural Research

Vol. VII, No. 4

September.

M'Sv** S, S S’?? : :

B 00 : :

. • H H H t'. OV

* • M H

0000 J 0 0 COvO H T|- co co h 00 • I

WHO*. . 00 xj- VO O 00.. •*

..NOV) *> v2 H^H

N* »0 . CO H . .

• * M CO IO N IO

• • MM

CO M t- M CO \Q xf CtO H II

vO O (O • . . 00 to VO CO CO.. • •

. . PI N- CO 00 H In to ..

to O 1 O H H . .

• • N. 10 Ov t-~vO

• • MM

HVQXfCOOO « COO CO w CO to H II

‘O'O ^ .00 xfr VO CO VO . .

..HOOt^t^ N* H M H ..

vo to . n h . ,

• • CO M M O

• • O PI M

^ 0 0 O O O t' O Q Q to Ov N

Ov co co • . .t^tovO Ov h.. ..

■ . M U") O tO N H tOH ..

VO CO . « M H . .

. ■ • M H VO VO

H t-vo VO 00 VO O VO H 00 O»*»to ..

.VO^ to H 00.. • •

^^Htot'.O VO to Cl ..

« «

(-1

£ vt w ° H ONO O O CO O H II

0)00 Ov . . . P" Tf lo Q to . . . .

• M wtvO co vo ^ to • •

to ^ « H . .

^ « fO W to

H H

VO Ch CO 0 CO H t-> t-t 10 C.00 O II

O co C> . . . t'. x}- VO H O • ••

. . M to 00 vo to H vo CO * *

vo . N H . .

N W Cl <N 00 00
' * H H

CO COVO CO t" 00 M vo Tt- Ov Ov II

O O c* . . . J> x}- vo O oo>. ••

. . Pi to 00 VO N 6 xfr • -

OVO . Cl H M . .

00 PI W M co 00 H

H ci

H OvvO vo CO Ov H vo M 00 to to I

ON-.. . 0 to to co V..

^.OOOco 4 toMO;

H H H H H 0 O

H H

i>-

“ ?§g S, Ig £ : *? ” ? :

. CO H H 00 O • t'. CO

N . to H H

co vo n- 10 Tf 0

H * * * “ CO co

tOHC^fr* 00 O co Tf Ov CO t' to H I

00 a . . . 00 to vo m vo.. , •

. H vo H t>. i"x 10 M

V . Cl H H

VO M H M W VO CO

Cl M

M 00 Ov Ov V too© 00 W M co V II

• M COCO V to to M * -

Xfr . « H • •

M . H W W xt CO

‘ H H

V xfrOO CO tt N to Nx 10 H 00 O H ■

V H . . . CO to vO Tf M..

. M I'* M Ov vOhvOco <

to . M H M .

N coo H covO co

M M

O JOIN O xj- CO 00 CO 'tO'O M I

VOv.. . CO »0 vO to « . . ..

. M 00 H O J>HOV

N . M H M M

SO M M Xt W H O

ci ci

SU 3 % T i

^?ooo O * hooco :

H JCfCOS O O

M ' ’ * * CO CO

CO M CO M V M lO M CO 00 VO VO M vfvO ft) V)1 to N ct

h n . . • Ov V vo pi m.. .........

.fOfO^Os 00 GS Vj 01 WH W)fO

CO 1 fo H

August.

CO V V CO 00 WO CO vo vo aN co O Ov to co Pi h C-vO

Q\ Os • • 1 Os u* O'- . • «**•*,*,,

4 " *0 H S N HfOW w 01 «

V O H rf CO Ov Ov V to H 00 VO -to

coco.. .N't « O H * .

• Cl O Ol C" 00 H to M

to . M H

VtH H • . CO CO

* « H W

S'vg'J'S ^ 7'O^hhhcoVCI

« to co 00 H vO "CO HH

M M to t* Ov ttvo Ov 00 00 O x* «fr to 00 to to PI TfM CO XT

00 n cr . . . c> to to "tf Tt , , .........

^VO^N^H to H to N

Pot No.

• ■ * • ... . . ... PI 00 OO NVOCO00 xt

•••• ... . . ... mmVVhhloOvO

.... ... . . ... mpimmcomnhm

.... ... . . ... illlj|)|i

... . . ... NOUN NH (flCCO

* ... ... . , ... wmcoVQhioOvOv

.... ... . . ... MMMMCOCIMHH

Evaporation (shallow tank), kgm.

per sq. meter .

Evaporation (deep tank)|g^" ;

Air temperature, max. , °C .

Air temperature, min., °C .

Air temperature, integrated

mean, °C .

Air temperature, max., °F .

Air temperature, min., °E .

Air temperature, integrated

mean, °F . . .

Integrated radiation, cal. per sq.

cm .

Integrated wet-bulb depression,
hr. deg. C .

Wind velocity /miles per hour. .

; * (meters per sec. .
Daily transpiration, kgm.:

Cowpea .

Millet .

Com .

Potato .

Amaranthus. . .

Sudan grass .

Alfalfa, E-23 .

Alfalfa, 162-98A1 .

TabbE IV. — Period of growth , yield , and water requirement of plants used in 1915 transpiration measurements

Oct. 33. 1916 Daily Transpiration during Normal Growth Period

173

1 74 Journal of A gricultural Research voi. vii, No. 4

COMPARISON OF THE TRANSPIRATION OF THE DIFFERENT CROPS

The graphs in figures 2 and 3 illustrate in a striking manner the great
fluctuations in the water required daily by plants to maintain normal
growth. During the period from July 8 to 10, 1914, for example, the
wheats under observation required about three times as much water
each day as during the period from July 17 to 19. In 1914 the maximum
rate of transpiration of the series as a whole occurred about July 9, at
which period the grains were headed, the legumes were in bloom or coming
into bloom, and corn was beginning to tassel.

Alfalfa, Sudan grass, and amaranthus, of which several cuttings were
made from the same root system, showed a continuous increase in the
transpiration rate up to the time of cutting. The transpiration of Minne¬
sota Amber sorghum, Algeria com, and lupine was relatively uniform
from about July 9 to near the end of the growth season. Among the
small grains barley and rye showed the least change in the transpiration
rate. The grain crops were harvested at the stage when similar crops
in the field are cut with the binder, and it is interesting to note at this
time transpiration was approximately one-fourth to one-half the maxi¬
mum rate.

The season of 1915, like that of 1914, shows a nearly uniform evapora¬
tion rate throughout the more active growth period if the daily fluctua¬
tions are ignored. The season was very rainy. This is reflected in the
radiation graph which is far more irregular than in 1914, but shows the
same gradual decline as the season advances.

The fluctuations of the transpiration graphs give evidence again of the
great variation in the daily quantity of water required to maintain normal
growth, as is shown by comparing the transpiration during the period
from July 21 to 23 with the adjacent three-day periods.

While the fluctuations of the different crops from day to day are simi¬
lar, marked differences in the pitch of the graphs are noticeable as the
season advances. The short season crops in general show a gradual
increase in transpiration from seedtime to a little past the middle of the
growth period and then an equally gradual decline to harvest.

The idea is often advanced that wheat and similar crops increase their
water demand suddenly at or just before the time of heading. The
measurements of the writers, however, lend no support to this conclu¬
sion. (See transpiration-evaporation ratios.) Aside from fluctuations
due to weather the transpiration increases uniformly during this period.

WATER LOSS DURING PERIODS OF MAXIMUM TRANSPIRATION

It has already been shown in the case of annual crop plants that the
transpiration rises to a maximum near the middle of the growth period
and then decreases until the plants are harvested. This is especially
true of grain crops. On the other hand, the transpiration of perennial
forage crops, such as alfalfa, increases steadily to a maximum at the

55858°— 16. (To face page 174.)

Oct. 23, 1916 Daily T ranspiration during Normal Growth Period

175

time of cutting. Various crops show their individuality by departing
more or less from these types. By comparing the transpiration of two
crops during any period with their total transpiration the relative
demand of the crops for water during this period can be shown. Such
a comparison is given in Table V for the 10-day period in 1914 from
July 7 to 16, inclusive, which represents a period of maximum transpira¬
tion activity.

TablU V. — A comparison of transpiration during the 10-day period July 7 to 16, IQ14 »

with total transpiration of crop

Croj

Total trans¬
piration.

Transpiration
period Ji

Actual.

during to-day
lly 7-16.

Percentage of
total.

Kgm .

Kgm.

Wheat, Kubanka .

160

40. 0

Wheat, Galgalos .

186

46.9

Oat, Swedish . . .

158

43*4

Oat, Burt . . .

154

40. $

Barley .

22. 0

Rye . . . .

12 1

3°- s

Cowpea .

112

3*-7

Lupine .

17. 2

Millet, Kursk .

23-9

Millet, Siberian .

IOO

25-7

Com, Northwestern Dent .

112

27. <

Com, Algeria .

137

26. 9

Sorghum, Minnesota Amber .

I30

26.8

Sorghum, Dakota Amber .

126

26. 2

Sudan grass (in inclosure) .

29. 0

Sudan grass (in open) .

21. 9

The relative transpiration of the small grains during this period was
very nearly the same, ranging from 24 to 27 per cent of the total trans¬
piration. This group, including the millets, shows a uniformity in this
respect which is remarkable. Sudan grass, grown both outside and
inside the screened inclosure, showed a relative transpiration loss during
this period agreeing closely with the small grains. The same is true
with Northwestern Dent com, an early-maturing variety. Algeria com
and the two varieties of sorghum, on the other hand, show a lower rela¬
tive transpiration during this period, owing to the fact that they mature
later than the other crops considered.

Similar data covering a 10-day period in August are presented in
Table VI. All of the crops show a lower transpiration than during the
July period, except Sudan grass and Algeria com, which had not yet
begun to ripen.

Amaranthus and the alfalfas were cut the second time at the end of
this period. The ratios show that nearly one-half of the total water
used by the alfalfas in the production of the second crop was transpired
during this 10-day period. In the case of Amaranthus the percentage
was even greater.

176

Journal of Agricultural Research

Vol. VII, No. 4

Table VI. — A comparison of transpiration during the 10-day period August 2 to 11, 1914,

with total transpiration of crop

Crop.

Total transpi-

Transpiration during 10-day
period Aug. 2-1 1.

ration.

Actual.

Percentage of
total.

Cowpea . .

Kgm.

1X2

Kgm.

IO. 4

Lupine .

12. 4

Millet, Kursk . .

15. 2

Millet, Siberian .

IOO

16. $

Corn, Northwestern Dent .

112

14.4

Corn, Algeria .

137

27. 0

Sorghum, Minnesota Amber .

Sorghum, Dakota Amber .

130

126

23-4
21. 6

Sudan grass tin inclosure) . . .

XIO

26.8

Sudan grass (in open) .

24.4

Amaranthus .

9.4

Alfalfa, Grimm, E23-20-52 .

28. 3

Alfalfa, Grimm, E23 .

33-2

Alfalfa, Grimm, E23 (in open) .

32. 6

Alfalfa, Grimm, 162-98A .

118

S2- 9

Similar determinations of the transpiration of different crops in 1915
during a 10-day period from July 7 to 16 are given in Table VII. The
transpiration during this period ranged from 23 to 31 per cent of the
total. In other words, one-fourth or more of the total water used by
these crops was transpired during these 10 days. With the exception
of the flax varieties the crops showing the highest transpiration during
this period were those which matured earliest.

Table VII. — A comparison of transpiration during the 10-day period July 7 to 16, iqi$ ,

with total transpiration of crop

Crop.

Total transpi¬

Transpiration during 10-day
period July 7-16.

ration.

Actual.

Percentage of
total.

Wheat, Kubanka .

Kgm.

20. 5

Wheat, Galgalos. . .

21. 7

Wheat, Washington Bluestem .

24. I

Wheat, Turkestan .

29. O

Wheat, Marquis . . . . . .

22. 6

Wheat, Kubanka .

108

24. 6

Wheat, Preston .

126

28.6

Oat, Swedish Select .

148

38-3

Oat, Burt .

103

3°- 3

Barley, Hannchen .

106

3i- 5

Rye, Spring .

22. 8

Flax, North Dakota (C. I. 13) .

25. 2

Flax, North Dakota (C. I. 19) .

1x4

34-9

Flax, Smyrna .

121

33-4

Oct. 23, 1916 Daily T ranspiration during Normal Growth Period

177

Corresponding determinations for a later 10-day period in 19x5, July
27 to August 5, are given in Table VIII, which includes all crops in the
1915 daily transpiration measurements. The earlier maturing crops con¬
sidered in the preceding table (VII) show a transpiration during this
period amounting to 16 per cent of the total, compared with 26 per cent
during the July period. Many of the crops were harvested soon after
the termination of the July-August period, but the data given in Table
VIII show that they were still transpiring actively.

Table VIII. — A comparison of transpiration during the 10-day period July 27 to
August 5, 1915, with total transpiration of crop

Crop.

Total transpi¬
ration.

Transpiration
period July

Actual.

during 10-day
27-Aug. 5.

Percentage of
total.

Kilograms.

Wheat, Kubanka . . .

12. 6

Wheat, Galgalos .

10. 8

Wheat, Washington Bluestem .

105

17. 1

Wheat, Turkestan .

III

15. 1

Wheat, Marquis .

13. 2

Wheat, Kubanka .

108

17. 9

Wheat, Preston .

126

20. 4

Oat, Swedish Select .

148

25.8

Wheat, Burt .

103

15. 2

Barley, Hannchen .

106

13-8

Rye, spring . . . . .

i3-’i

Flax, North Dakota (C. I. 13) .

15. 6

Flax, North Dakota (C. I. 19) .

1 14

20. 2

Flax, Smyrna .

121

IQ. 8

Cowpea .

n*5

Millet .

11. 5

Sorghum . . .

8.4

Com .

5*6

Potato . . .

4. 1

Amaranthus .

9.0

Sudan grass .

11. 8

Alfalfa, E23 .

25* 2

Alfalfa, 162-98A1 .

23.8

The alfalfa and amaranthus measurements given in Table VIII repre¬
sent the first crops, which were harvested at the close of the period.
These plants, together with Sudan grass, show a markedly higher trans¬
piration loss than the small grains.

1 78

Journal of Agricultural Research

Vol. VII, No. 4

LOSS OF WATER DURING THE MAXIMUM TRANSPIRATION PERIOD
PER UNIT OF DRY MATTER HARVESTED

MEAN DAILY TRANSPIRATION PER GRAM OF DRY MATTER HARVESTED

It is not possible in the case of the grain crops to determine directly
the maximum transpiration per unit of dry matter without sacrificing
the crop in order to find the dry weight. Such calculations can, however,
be made upon the basis of the dry weight of the crop at maturity, which
from a practical standpoint is of more importance than the former
determination. Computations of this kind are presented in Tables IX
to XII for the io-day periods just considered.

Reference to Table IX will show that during the transpiration period
July 7 to 1 6, 1914, the small grains were transpiring from 12 to 16 gm.
of water per day per gram of dry matter harvested; cowpea and lupine,
19 gm.; and millet, sorghums, and com, 6 to 9 gm. These quantities
are approximately proportional to the water requirement of the crops.

Table IX. — Loss of water per unit of dry matter harvested during the maximum transpi¬
ration period July 7 to 16, 1914 a

Crop.

Dry

matter.

Mean daib
tion, Jn

Actual.

r transpira-
ily 7-16.

Per gram
of dry
matter.

Hourly
transpira¬
tion during
midday
per gram
of dry
matter.

Daily loss
of water
per ton of
dry matter
per acre.

Grams.

Kilograms.

Grams.

Acre-inches ,

Wheat, Kubanka .

306

4. 00

13- 1

1. 3

O. II

Wheat, Galgalos .

298

4. 69

T5- 7

l. 6

. 14

Oat, Swedish .

264

4- 34

16. 4

I. 6

• z5

Oat, Burt . .

251

4-05

16. I

i. 6

. 14

Barley .

182

2. 20

12. I

1. 2

. II

Rye .

195

3- 05

15.6

1. 6

. 14

Cowpea .

170

3- 17

18.6

i-9

. 16

Lupine .

1. 72

19. I

i-9

• z7

Millet, Kursk .

301

2- 39

7- 9

.07

Millet, Siberian .

318

2- 57

8. 1

. 07

Com, Northwestern Dent .

304

2- 75

9.0

-9

.08

Com, Algeria . . .

417

2. 69

6. 5

. 7

. 06

Sorghum, Minnesota Amber .

457

2.68

5-9

• 05

Sorghum, Dakota Amber .

427

2. 62

6. 1

•os

a Mean daily evaporation, 12.3 kgm. per square meter.

Sudan grass during a io-day period in August immediately preceding
the first cutting showed an average transpiration loss of from 9 to 1 1 gm.
per day per gram of dry matter harvested (Table X). Amaranthus and
alfalfa were cropped for the second time at the end of this period.
Amaranthus showed a transpiration loss of 17 gm. per day per gram of
dry matter produced, while the alfalfas transpired from 36 to 56 gm. per
day per gram of dry matter harvested. The Sudan grass and alfalfa
grown outside the screened inclosure showed a transpiration from 27 to
30 per cent higher than the same crops inside the inclosure. This cor-

Oct. 23t 19x6 Daily T ranspiration during Normal Growth Period

179

responds very closely with previous determinations of the influence of the
inclosure on the water requirement.1

Table X. — Loss of water per unit of dry matter harvested during the maximum transpi¬
ration period , August 2 to 11, IQ14 a

Dry

matter.

Mean daily transpira¬
tion Aug. 2-1 1.

Hourly
transpira¬
tion during

Daily loss
of water

Crop.

Actual.

Per gram
of dry
matter.

midday
per gram
of dry
matter.

per ton of
dry matter
per acre.

Suda^ grass (in inclosure) .

Grams,

Kilograms .

Grams.

Acre-inches .

3QI

2.68

8.9

0.9

0. 08

Sudan grass (in open) .

216

2. 44

ii- 3

1. I

. 10

Amaranthus . .

•94

17.4

i- 7

■ 15

Alfalfa, E23-20-52 .

2.83

42.9

4-3

• 38

Alfalfa, E23 .

3- 32

43- 1

4-3

• 38

Alfalfa, E23 (in open) .

3. 26

56. 2

5-6

• 50

Alfalfa, 162-98A .

146

5- 29

36. 2

3- 6

• 32

a Mean daily evaporation 12 kgm. per square meter.

Reference has already been made to the fact that the water require¬
ment in 1915 was much lower than in 1914. The daily transpiration
loss of the small grains during a 10-day period from July 7 to 16, 1915,
ranged from 9 to 13 gm. per gram of dry matter harvested (Table XI).
Flax, which has a much higher water requirement than the small grains,
showed during this period a daily loss of from 17 to 19 gm. of water per
gram of dry matter harvested.

Table XI. — Loss of water per unit of dry matter harvested during the maximum transpira¬
tion period July 7 to 16 , IQ1S 0

Crop.

Dry

matter.

Mean daily transpira¬
tion July 7-16.

Hourly
transpira¬
tion during

Daily loss
of water
per ton of
dry matter
per acre.

Actual.

Per gram
of dry
matter.

midday
per gram
of dry
matter.

Grams.

Kilograms.

Grams.

Acre-inches.

Wheat, Kubanka .

209

2. 05

9.8

I. O

O. 09

Wheat, Galgalos .

168

2. 17

12. 9

1-3

. II

Wheat, Washington Bluestem .

215

2. 41

II. 2

I. I

. IO

Wheat, Turkestan . . .

219

2. 90

13. 2

i- 3

. 12

Wheat, Marquis .

208

2. 26

10. 9

1. 1

. IO

Wheat, Kubanka .

267

2. 46

9. 2

•9

.08

Wheat, Preston .

279

2. 86

10. 2

1. 0

. 09

Oat, Swedish Select .

332

3- 83

II. 5

1. 2

. IO

Oat, Burt . .

233

3- 03

I3.O

i-3

. 12

Barley .

262

3- i5

12. O

1. 2

. II

Rye .

203

2. 28

II. 2

1. 1

• 10

Flax, North Dakota (C. I. 13) .

149

2. 52

l6. 9

1. 7

• 1S

Flax, North Dakota (C. I. 19) .

186

3- 49

18.8

1.9

• 17

Flax, Smyrna .

182

3* 34

l8. 4

1.8

. 16

a Mean daily evaporation, 7.3 kgm. per square meter.

1 Briggs, L. J., and Shantz, H. L. Relative water requirement of plants. In Jour. Agr. Research, v. 3,
no. 1, p. 1-64, 1 fig. pi. 1-7. 1914. Literature cited, p. 62-63.

55858°— 16 - 3

i8o

Journal of Agricultural Research

Vol. VII, No. 4

Similar measurements for other crops during a later io-day period in
1915 (Table XII) showed a marked reduction in the daily transpiration
rate compared with the preceding year. The evaporation rate during
the 1915 period was only about one-half that in 1914. The transpiration
rate of the alfalfas during the 1915 period was also approximately one-
half that observed in 1914.

Table XII. — Loss of water per unit of dry matter harvested during the maximum trans¬
piration period July 2J to August 5, IQ15

Crop.

Dry

Mean daily transpira¬
tion July 27-Aug. 5,
1915-

Hourly
transpira¬
tion during
midday
per gram
of dry
matter.

Daily loss
of water
per ton of
dry matter
per acre.

matter.

Actual.

Per gram
of dry
matter.

Cowpea .

Grams .
128

Kilograms.

I- 15

Grams.

9-0

Grams.

0.9

Acre-inches .
0. 08

Millet .

256

I- 15

4.5

• 5

.04

Sorghum .

203

.84

4. I

• 4

.04

Com .

1 12

.56

5-o

.04

Potato .

.41

6. 1

• 05

Amaranthus . .

129

.90

7.0

• 7

. 06

Sudan grass . .

I76

I. 18

6. 7

• 7

. 06

Alfalfa, E23 . .

133

2. 52

18. 9

1.9

• 17

Alfalfa, 162-98A1 .

”5

2. 38

20. 7

2. 1

. 18

°Mean daily evaporation, 6.2 kgm. per square meter.

HOURLY TRANSPIRATION DURING MIDDAY PER GRAM OR DRY MATTER

HARVESTED

An examination of the graphs of hourly transpiration on clear days 1
shows that the transpiration during one hour at or near midday in mid¬
summer is approximately one-tenth of the total transpiration for the
day. The hourly transpiration of different crops has been calculated
on this basis for the midday hours during the io-day maximum trans¬
piration periods considered above (Tables IX to XII). In 1914 the
midday transpiration of the small grains ranged from 1.2 gm. per hour
per gram of dry matter for barley to 1.6 gm. for wheat, oat, and rye
(Table IX). In other words, a crop of oat yielding 1 ton of dry matter
per acre would have lost 1.6 tons of water per acre per hour during the
midday hours of its maximum transpiration period.

Cowpea and lupine lost each hour during midday an amount of water
equal to 1.9 times the dry weight of the crop (Table IX). . Millet, com,
and sorghum transpired at a much slower rate than the other crops
here considered, the hourly loss of sorghum being 0.6, millet 0.8, and com
0.7 to 0.9 that of its dry weight.

1 Briggs, L. J., and Shantz, H. L. Hourly transpiration rate on clear days as determined by cyclic
environmental factors. In Jour. Agr. Research, v. 5, no. 14, p. 583-650, 22 fig., pi. 53-55. 1916, Literature
cited, p. 648-649.

Oct. 23, 1916 Daily Transpiration during Normal Growth Period 18 1

During the August transpiration period in 1914 (Table X) the alfalfas
showed an hourly transpiration loss near midday of from 3.6 to 5.6 gm.
per gram of dry matter. * In other words, during the last third of the
growth period the alfalfa crops lost during each midday hour an amount
of water ranging from 3.6 to 5.6 times the dry weight of the crop.

The conditions in 1915, as has already been mentioned, were less
severe than in 1914. The evaporation during the July maximum trans¬
piration period in 1915 was only 60 per cent of that during the corre¬
sponding period in 1914. The transpiration loss of the small grains
each hour during midday was approximately equal to the dry weight
of the crop (Table XI). The transpiration loss of the flax varieties was
nearly twice as great.

During the second transpiration period considered in 1915 (Table XII),
the midday loss was still further reduced, being approximately one-half
that observed in 1914.

DAILY LOSS OF WATER IN ACRE-INCHES PER TON OF DRY MATTER HAR¬
VESTED PER ACRE

From a practical standpoint it is desirable to express the daily trans¬
piration loss in terms of acre-inches of water per ton of dry matter pro¬
duced per acre. This is given in the last column of Tables IX to XII.
A wheat crop yielding a ton of dry matter (grain and straw combined)
would thrash approximately 12 bushels of wheat. Such a crop at
Akron in 1914 would have used approximately 0.13 acre-inch of water
each day during its maximum transpiration period. Millet, com,
and sorghum required approximately one-half this amount per ton of
dry matter, while the alfalfas during their maximum transpiration
period required the equivalent of a rainfall of from 0.3 to 0.5 inch per
day per ton of dry matter produced per acre.

LOSS OF WATER DURING MAXIMUM TRANSPIRATION PERIOD PER SQUARE

METER OF PLANT SURFACE

The surface area of the plant tissues of the different crops included in
the transpiration measurements was determined from a selected sample
of the plants in one pot of each set. The ratio of the portion selected to
the whole crop was found by comparing the green weight of the sample
(taken immediately after cutting) with the green weight of the whole
crop. The area of the selected portion was determined by one of the
following methods: (1) By direct measurement of the length and
breadth (or diameter) of the leaves or stems; (2) by pasting the leaf and
flower parts on surgeon's tape and determining the area by measure¬
ments; (3) by prints of the pasted leaves on squared photographic
paper, the area being determined either by direct counts of the squares
on the paper or by a planimeter. From the total area of the sample and

182

Journal of Agricultural Research

Vol. VII, No. 4

the ratio of the weights the mean surface area per pot was computed.
The area measurements were made at harvest time and include the area
of matured leaves as well as the fresh leaves and stems. The maximum
transpiring area is therefore represented in these measurements (Tables
XIII to XVI).

Table XIII. — Transpiration per square meter of plant surface during the io-day period

July y to 16, IQ14 »

Daily transpiration
July 7-16.

Hourly
transpira¬
tion per

Ratio of
transpira¬
tion per
square
meter to
evaporation
per square
meter.

Crop.

Plant

surface.

Actual.

Per square
meter of
plant
surface.

square
meter of
plant
surface
during
midday.

Wheat, Kubanka .

Sq. meters.

2. 45

Kgm .

4. OO

Kgm.

I. 63

Gm.

163

0. 13

Wheat, Galgalos . . .

2.68

4.69

I- 75

I75

. 14

Oat, Swedish .

3- 42

4- 34

I. 27

127

. 10

Oat, Burt .

2. 46

4-°s

I. 65

165

• 13

Barley .

i- 95

2. 20

n*3

IJ3

.09

Rye .

1. 99

3-05

1* 53

i53

. 12

Millet, Kursk .

3- I5

2.39

. 76

. 06

Millet, Siberian . : . .

4. 20

2. 57

. 61

• °5

Sorghum, Dakota Amber .

1. 76

2. 62

1.49

149

. 12

<* Mean daily evaporations 12.3 kgm. per square meter. Hourly evaporation during midday= 1,230 gm.
per square meter.

Table XIV. — Transpiration per square meter of plant surface during the 10-day period

August 2 to liy igi4a

Daily transpiration
Aug. 2-1 1.

Hourly
transpira¬
tion per

Ratio of
transpira¬
tion per
square
meter to
evaporation
per square
meter.

Crop.

Plant

surface.

Actual.

Per square
meter of
plant
surface.

square
meter of
plant
surface
during
midday.

Sudan grass (in inclosure) .

Sq. meters.

3* 27

Kgm.

2.68

Kgm.

0. 82

Gm.

0. 07

Sudan grass (in open) .

I. 90

2.44

1.28

128

. II

Amaranthus .

.67

•94

I. 40

140

. 12

Alfalfa, E23-20-52 . .

I. 71

2.83

i. 66

166

. 14

Alfalfa, E23 .

2. 08

3* 32

i*59

*59

•13

Alfalfa, E23 (in open) .

•95

3.26

3* 43

343

.29

Alfalfa, 1 62 -98 A .

3*37

5- 29

i*57

157

•13

a Mean daily evaporation** 12 kgm. per square meter. Hourly evaporation during midday** 1,200 gm.
per square meter.

Oct 23, 1916 Daily T ranspiration during Normal Growth Period

183

Table XV. — Transpiration per square meter of plant surface during the 10-day period

July y to i6y iqi$ &

Crop.

Plant

surface.

Daily transpiration
July 7-16.

Hourly
transpira¬
tion per
square
meter of
plant
surface
during
midday.

Ratio of
transpira¬
tion per
square
meter to
evaporation
per square
meter.

Actual.

Per square
meter of
plant
surface.

Sq. meters.

Kgm.

Kgm .

Gm.

Wheat, Kubanka .

I. 97

2.05

I. 04

104

0. 14

Wheat, Galgalos . .

i. 47

2. 17

I. 48

148

. 20

Wheat, Washington Bluestem .

2* 57

2. 41

* -94

• z3

Wheat, Turkestan .

2. 10

2. 90

1.38

138

. 19

Wheat, Marquis .

1. 78

2. 26

1. 27

127

. 17

Wheat, Kubanka .

2. 16

2. 46

1. 14

1 14

. 16

Wheat, Preston .

3- 7°

2.86

• 77

. 11

Oat, Swedish Select .

2. 82

3- 83

1. 36

136

. 19

Oat, Burt .

i- 95

3- °3

i55

. 21

Barley . .

1. 74

3- 15

1. 81

181

•25

Rye . . . . .

1. 50

2. 28

1. 52

x52

. 21

Flax, North Dakota (C. I. 13) .

1. 49

2. 52

* 1. 69

169

•23

Flax, North Dakota (C. I. 19) .

2* 73

3- 49

1. 28

128

.18

Flax, Smyrna .

3* 54

3- 34

•94

•*3

a Mean daily evaporation =7.3 kgm. per square meter. Hourly evaporation during midday 730 gm. per
square meter.

Table XVI. — Transpiration per square meter of plant surface during the 10-day period

July 2j to August 5, igisa

Daily transpiration
July 2 7- Aug. 5.

Hourly
transpira¬
tion per
square
meter of
plant
surface
during
midday.

Ratio
transpira¬
tion per
square
meter to
evaporation
per square
meter.

Crop.

Plant

surface.

Actual.

Per square
meter of
plant
surface.

Cowpea . .

Sq. meters.

2. 14

Kgm.

O. 54

Gm.

. 08

Millet, Kursk .

2. 20

I- 15

• S2

.08

Sorghum .

I. 09

.84

• 77

. 12

Com .

•97

•58

.09

Amaranthus . .

1.71

.90

• 53

.09

Sudan grass . k .

2. 41

I. 18

•49

.08

Alfalfa, E23 .

4. S3

2. 52

•56

.09

Alfalfa, 162-98A1 .

2.86

2. 38

.83

•13

« Mean daily evaporation=6.2 kgm. per square meter. Hourly evaporation during midday=62o gm.
per square meter.

184

Journal of Agricultural Research

Vol. VII, No. 4

The mean daily transpiration of the small-grain crops during the maxi¬
mum transpiration period in 1914 (Table XIII) was 1.49 ±0.08 kgm. per
square meter, and in 1915 (Table XV) 1.46 ±0.07 kgm. per square meter.
The evaporation during the 1915 period was only 59 per cent of that
observed during the corresponding period in 1914. While the transpira¬
tion periods selected were the same (July 7-16), the crops in 1914 were
relatively more advanced showing evidences of ripening in some cases.
This appears to be responsible, in part at least, for the observed agreement
in the transpiration rate during the two years, although it is of course
possible that under the weather conditions prevailing in 1915 a mor¬
phological adjustment took place, which increased the transpiration
coefficient per unit area.

During the August transpiration period of 1914 the transpiration rate
per square meter of plant surface (Table XIV) ranged from 81 gm. per
hour for Sudan grass to 166 gm. for a variety of Grimm alfalfa, these
plants being grown in the screened inclosure. The same plants grown in
the open showed a transpiration loss of from 128 to 343 gm. of water per
square meter of surface per hour. The water requirement of Sudan grass
in the inclosure was 10 per cent below that of the crop grown in the open,
while in the case of alfalfa the inclosure reduced the water requirement
22 per cent (Table II). The reduction in transpiration per unit area of
the crops grown inside the inclosure is over twice that indicated from the
water-requirement measurements. Here, again, there is an indication of
a morphological adjustment resulting from the difference in exposure.

Periods of maximum transpiration may be due to extreme weather
conditions or to the transpiration coefficient having reached its maximum
or to *both. It would appear that the maximum transpiration period of
the small grains in 1914 was determined largely by weather conditions,
since it falls rather late in the period of growth, and some of the grains
had already begun to ripen. In 1915 this period corresponds more nearly
to the period of maximum transpiration coefficient of the small grains.
During the July period in 1914 (Table XIII) the transpiration per square
meter per hour during midday ranged from 61 gm. for Siberian millet to
175 gm. for Galgalos wheat. The millets were just heading at this time,
and oats, barley, and rye had begun to ripen, although the transpiration
coefficient was still apparently near the maximum.

The loss of water from the plant surfaces during this period was 5 to 14
per cent of that from a water surface of equal area. In 1915 the plant
surfaces lost relatively more, ranging from 10 to 25 per cent of that from
a free water surface of the same area (Table XV).

It should be noted in this connection that the evaporation measure¬
ments were made in the open, while the transpiration measurements
for the most part were made within the screened inclosure. The data
just given regarding the transpiration of the plants relative to an equal
free water surface are consequently somewhat too low.

Oct. 23, 1916 Daily T ranspiration during Normal Growth Period 185

During the August period in 1914 the transpiration measurements for
alfalfa and Sudan grass were made both inside and outside the screened
inclosure (Table XIV). The hourly transpiration of Sudan grass in the
open was 11 per cent and that of alfalfa 29 per cent of the evaporation
loss from a water surface of equal area as compared with a loss within
the inclosure of 7 and 13 per cent, respectively.

The transpiration of the different plants per unit area of plant surface
shows less variation than the transpiration per unit weight of dry matter
produced. In other words, the greater efficiency exercised by some plants
in the use of water appears to be due more to a reduction in plant surface
than to a reduction of transpiration per unit area of surface. The latter
effect is, however, in evidence, the plants characterized by a low-water
requirement usually showing a somewhat lower transpiration rate per
unit area.1

The determination of the surface area of a large plant from measure¬
ments on a few shoots or branches is necessarily only an approximation.
For different sets of plants of the same crop the results were in satisfac¬
tory agreement. Thus, measurements of six sets of Kubanka wheat
gave 49, 48, 47/44, 52, and 49 sq. cm. of surface per gram (dry weight).
The measurements were less accurate in the case of plants which branch
freely, such as alfalfa. The greatest uncertainty, however, appears to
be in the assumption that the various surfaces presented by a plant
have the same transpiration loss per unit area.

COMPARISON OF THE ENERGY RECEIVED BY DIRECT RADIATION
WITH THE ENERGY DISSIPATED BY TRANSPIRATION

If the area of the shadow thrown by a plant on a plane normal to the
sun’s rays is known, the direct solar radiation received by the plant in
one hour, expressed in gram-calories, is equal to the product of the area
of the shadow in square centimeters and the direct radiation energy in
calories per square centimeter per hour.

The energy dissipated by the plant through transpiration during the
same period is equal to the product of the transpiration in grams and
the latent heat of vaporization (536 gm.-cal. per gram). The ratio of
these two quantities represents the part played by direct sunlight in
transpiration, assuming that all the radiation is absorbed. Such com¬
putations have been made for a number of plants employed in the trans¬
piration measurements of 1914, and are presented in Table XVII. The
measurements are based on the hourly transpiration and hourly radia¬
tion values at midday. For plants grown in the inclosure the radiation
values have been corrected for the shade of the wire screen.

1 The high value obtained for sorghum is due to the fact that the lower leaves were harvested early to
prevent loss and were not included in the area measurements made at the time the plants were cut. These
lower leaves were active during the transpiration period considered, so that the transpiring area was
greater than that finally measured. The transpiration recorded per unit area is therefore too high.

i86

Journal of Agricultural Research

Vol. VII, No. 4

Tabi.3 XVII. — A comparison of the energy received by direct radiation with' the energy
dissipated by transpiration during midday

Period and crop.

July 7-16, 1914.

Wheat, Kubanka .

Wheat, Galgalos .

Oat, Swedish Select .

Oat, Burt .

Barley .

Millet, Kursk .

Millet, Siberian .

Cowpea .

Aug. 2— 11, 1914.

Alfalfa, Grimm, E23-20-52 .

Alfalfa, Grimm, E23 .

Alfalfa, Grimm, E23 (in open). .
Alfalfa, Grimm, 162-98A. .

Hourly

transpiration.

Radiation (cal.
per sq. cm.
per hour in
shelter).

Cross section of
plant normal
to sun’s rays
(area of plant
shadow).

Ratio of radia¬
tion energy
received to
energy dissi¬
pated through
transpiration.

Gm.

400

Sq. cm.

2, OOO

0. 54

469

I, 920

• 44

434

2, 040

•51

4°s

1,940
I, 480

• 55

220

• 73

239

1,920
I, 980

•87

257

•83

3i7

2,230

.76

283

I, 840

.78

332

2,230

. 80

326

2, 200
2,870

1. 01

529

•6S

The area of the plant shadow in the case of the grain crops has been
determined by considering the plants in a pot to have the form of a

Fig. 4. — Determination of the area, on a plane
normal to the sun’s rays, of the shadow of a
cylinder of diameter d and height h in terms of
the angular departure (2) of the sun from the
vertical.

right cylinder, the diameter and
height of which were determined
from direct measurements and from
photographs. The method of com¬
puting this area is readily seen from
figure 4. Let x be the angle made by
the sun’s rays with the vertical at
midday (zenith distance). The pro¬
jection of the right cylinder of height
h and diameter d on a plane normal
to the sun’s rays would give a rec¬
tangular figure with elliptical ends
whose total length is d cos x + h sin x .

In this expression^ cos x represents

the minor semidiameter of each el¬
liptical portion, whose major semi¬
diameter is -. The area of the el-
2

• * 7T

liptical portion is consequently ~-<P cos x, while the area of the rectang¬
ular part of the figure is h d sin x. The total area (a) is therefore —

a — — d2 cos x 4- h d sin x.
4

Oct 23, 1916 Daily Transpiration during Normal Growth Period 187

The shadow areas of the other crops were computed from the pro¬
jection of the geometrical figures which they most closely approximated,
such as cylinders, spheres, or inverted truncated cones.

While measurements of this kind are at best approximations, the
results given in Table XVII are consistent in showing that the energy
received directly from the sun is insufficient to account for the energy
dissipated by the plants through transpiration during the midday
hours. Only in the case of alfalfa in the open is the direct solar radia¬
tion intercepted by the plant sufficient to account for the observed
transpiration. Even on bright days, therefore, other sources of energy,
such as the indirect radiation from the sky and from surrounding objects
and the heat energy received directly from the air, contribute mate¬
rially to the energy dissipated through transpiration. Comparative
transpiration measurements of shaded and unshaded plants also show
that the energy consumed in transpiration is only partially attributable
to direct radiation.

RELATION OF TRANSPIRATION TO THE WEATHER

If transpiration is determined absolutely by the intensity of any one
of the weather factors, the ratio of the daily transpiration to the daily
intensity of this factor should give a regular graph when plotted. This
graph, if the correlation were perfect, would be an expression of the
relative transpiration coefficient of the plant considered.

In figure 5 are plotted the ratios of the daily transpiration of Kubanka
wheat to the daily evaporation from the shallow tank. This graph
shows a gradual increase in the transpiration coefficient to a maximum
on July 13, followed by a somewhat more rapid decrease to harvest.
While the graph shows many irregularities a similar response in trans¬
piration and evaporation to changes in weather is indicated.

The second graph shows the ratio of transpiration to wet-bulb de¬
pression. Outstanding points in this graph indicate, of course, that on
certain days the transpiration values were influenced by some factor
other than the dryness of the air. Reference to the radiation graph
will show that on such day's the transpiration-radiation ratio is normal
or shows a departure in the opposite sense to the ratio of transpiration
to wet-bulb depression. In other words, both factors enter into the
determination of the transpiration.

Similar departures are in evidence in the transpiration-temperature
graph, and it will be noted that such departures are usually in an oppo¬
site sense to one or the other of the graphs just discussed. Tempera¬
ture, therefore, also enters into the determination of the transpiration.

The ratio of transpiration to wind velocity is plotted at the bottom of
figure 5. The change in the transpiration coefficient is evident in this
graph, but no marked correlation is indicated.

i88

Journal of Agricultural Research

Vol. VII, No. 4

*JOA/£ /S/4 *41/(3.

Fig. 5. — Ratios of the daily transpiration of Kubanka wheat to the daily intensity of various weather
factors, plotted with approximately the same amplitude.

UUMC/9/4 c /t/Ly /*(/<?.

Fig. 6. — Ratios of the daily transpiration of Minnesota Amber sorghum to the daily intensity of various
weather factors, plotted with approximately the same amplitude.

Oct. 23, 1916 Daily T ranspiration during Normal Growth Period

189

An inspection of the ratio graphs of sorghum and alfalfa (figs. 6 and 7)
shows, as in the case of Kubanka wheat, a dependence of transpiration
upon radiation, wet-bulb depression, and temperature. A marked de¬
parture in any one of the ratio graphs is usually, though not always,
accompanied by a departure in the opposite sense in one of the other
graphs. The correlation of transpiration with wind velocity is low in the
case of both of these crops.

MARCH OF TRANSPIRATION DURING THE GROWTH PERIOD AS SHOWN BY
THE RATIO OF DAILY TRANSPIRATION TO DAILY EVAPORATION

Although the association of transpiration with wet-bulb depression is
fully as marked as the transpiration-evaporation association (see correla¬
tion coefficients, p. 204), the writers have decided to employ the latter
ratio to represent the change in the transpiration coefficient of the dif¬
ferent plants, since it is already used extensively.1

The ratios of the daily transpiration of each crop to the daily evapora¬
tion (shallow tank) are plotted as percentages of the maximum in figures
8 and 9. By this method different crops may be easily compared.

The graphs show irregularities due to the lack of an exact correlation
between evaporation and transpiration, or to errors in the measurements.
On days when rain occurred the outstanding points are probably due to
inaccurate determinations of the transpiration.2 The outstanding ratios
on days without rain are not explainable on this basis and are to be re¬
garded as expressions of the inexact correlation of transpiration and
evaporation as here measured.

The general trend of the graphs indicates a gradual increase in the
transpiration coefficient from seedtime to a maximum which in the case
of annual plants comes just before they begin to ripen.

measurements in 1914

In 1914 (fig. 8) the small grains were well advanced before the trans¬
piration measurements were begun. They did not reach their maximum
transpiration rate, however, until about one month later.

1 Livingston, B. E. The relation of desert plants to soil moisture and to evaporation. 78 p.( illus.
Washington, D. C., 1906 (Carnegie Inst. Washington Pub. 50.) Literature cited, p. 77-78.

- The resistance offered by leaves to transpirational water loss. In Plant World, v. 16, no . 1 , p. 1-35,

illus. 1913.

2 Even a small amount of rain wets the plant thoroughly, a part of the water remaining on the surface of
the plants and a part being absorbed by the dry or living leaves or caught in the leaf sheaths or flower heads.
Water is also held on the surface of the pot and a small amount may find its way into the pot by suction
due to the change in temperature. If the morning weighing following a rain during the night is taken as
the basis of determining the transportation, on the subsequent day the transpiration is too high since some
of the water is merely evaporated from the surface of the plant and pot. If the two da£s are combined and
morning weighing discarded, the transpiration is too low since transpiration from wet plants is lower than
from dry plants, and since an equivalent of the water which was absorbed must be transpired before a loss
in weight can be recorded. Notwithstanding the errors of the second method, it seems best not to intro¬
duce the greater uncertainty involved in the first method and to regard the outstanding determinations
on rainy days as experimental errors which can not at present be successfully overcome without actually
protecting the plants from rain, which would change the conditions under which field crops are grown.

190

Journal of Agricultural Research

Vol. vn, No. 4

Fig. 7. Ratios of the daily transpiration of alfalfa, K-23, to the daily intensity of various weather factors, plotted with approximately the same amplitude. C indicates the

date of cutting.

Oct. 23. 1916 Daily T ranspiration during Normal Growth Period

191

Fig. 8. — The ratio of daily transpiration of different crops in 1914 to daily evaporation (shallow tank)
plotted in percentage of the maximum. The letter C in the graph indicates date of cutting, followed
by a new growth from the established root system. H signifies heading; F, flowering; R, ripening;
and T, tasseling.

Spongospora subterranea and Phoma tuberosa on Solanum tuberosum

Plate A

Journal of Agricultural Research a.hoen&co. Baltimore. Vol, VII, No. 5

192

Journal of Agricultural Research

Vol. VII, No. 4

The transpiration coefficient of Kubanka wheat had reached half the
maximum value when the daily weighings were begun, although the
total transpiration from seedtime to this date was only io per cent of
that for the whole season. A gradual increase in the transpiration coef¬
ficient is recorded from June 16 to a maximum on July 12. Although
the plants began to head on June 15, this produced no marked change
in the transpiration coefficient. The drop in the graph following the
period of heading is not significant in this connection since a similar drop
was recorded for the other crops, and it is evidently due to the failure of

Fig. 9.— The ratio of daily transpiration of different crops grown 4n 1915 to daily evaporation (shallow
tank) plotted in percentage of the maximum. 5 signifies that the plants were forming shoots; H,
heading; F, flowering; and D, lower leaves dying.

evaporation and transpiration to respond proportionately to the weather
conditions on these days. The gradual decrease in the transpiration
coefficient following the maximum is indicated by the graph and begins
almost one month before harvest. At harvest time the transpiration
coefficient of Kubanka was still 20 per cent of the maximum.

Galgalos wheat which began to head two days later than Kubanka
reached its maximum four days earlier. The period of gradual decline
was approximately one month long and the transpiration coefficient at
harvest was 20 per cent of the maximum. The depression in the curve
in the early part of the period of decline is probably not significant since

Oct. 23> 1916 Daily T ranspiration during Normal Growth Period

193

it is recorded for all crops and is to be attributed largely to the errors in
measuring transpiration during a period of rainy weather.

The transpiration coefficient of Swedish Select oat, which began to
head on June 28, did not reach its maximum until July 13, after which it
declined rather rapidly to about 25 per cent of the maximum at harvest
time. Burt oat began to head much earlier, but showed the same
gradual increase to the maximum. The value at harvest time was very
high, amounting to 50 per cent of the maximum. Hannchen barley and
Burt oat showed very little difference in the march of the' transpiration
coefficient.

Rye was beginning to head when the measurements were started. The
transpiration coefficient was unusually uniform throughout the season.
At harvest it was 65 per cent of the maximum and did not differ mate¬
rially from that at the period of heading.

The graph for cowpea is almost symmetrical, beginning and ending at
about 20 per cent of the maximum, which occurred just after the flower¬
ing period.

Lupine blossomed earlier than cowpea and showed no marked maxi¬
mum period, the rate remaining about the same from the 10th of July to
the 1st of August.

Kursk and Siberian millet showed no significant differences. The
graphs include practically the whole growth period. The increase was
rapid and uniform and the maximum was reached at the period of head¬
ing. The transpiration coefficient was about 40 per cent of the maximum
at the time of harvest.

The two varieties of Amber sorghum showed no marked differences.
Dakota Amber headed a little earlier than Minnesota Amber. The maxi¬
mum transpiration coefficient of both crops occurred about 10 days after
the plants began to head. The decline was gradual, the value at harvest
being over 40 per cent of the maximum, notwithstanding the fact that
the seeds were ripe at that time.

Sudan grass was grown in the screened inclosure and also in the open.
The transpiration coefficient of the plants in the inclosure increased some¬
what more rapidly at first than in the open, although the maximum was
reached at the same time. The second crop was much smaller than the
first and the transpiration coefficient reached only about one-fourth the
value attained during the first crop.

Algeria com reached its maximum much later than Northwestern Dent
and showed no marked decrease in its transpiration coefficient, the value
at harvest being 40 per cent of the maximum. Algeria corn did not
ripen at Akron, which accounts for the high transpiration coefficient
when the crop was harvested.

The graph for Amaranthus shows no marked change in the transpira¬
tion coefficient during the first crop. In other words, the young large-

194

Journal of Agricultural Research

Vol. VII, No. 4

leaved plants lost water almost as rapidly as the older plants on which
the lower leaves either drop off or become relatively inactive.

Four sets of alfalfa were included in these measurements. Three of
these sets were grown from seed, while the fourth alfalfa, 162-98A, was
grown from cuttings. The transpiration coefficient of the latter was
already 25 per cent of the maximum when the measurements were begun.
Some of the plants in the fir§t crop began to bloom when the transpira¬
tion coefficient was only 40 per cent of the maximum. In the second
and third crop's the plants were harvested shortly after they had developed
flowers. The march of the transpiration coefficient for each variety was
approximately the same, with the exception of the first crop of alfalfa,
1 63-98 A, which developed much more rapidly than the other varieties.
The set in the open developed more slowly than in the inclosure.

The relative loss of water at different periods in the growth of a crop,
the evaporation rate being uniform, can readily be determined from
the graphs representing the transpiration-evaporation ratio. The
weekly loss from different crops expressed in percentages of the total is
given in Table XVIII.

Table XVIII. — Weekly transpiration under uniform conditions of evaporation expressed
in terms of percentage of the total , IQ14

Crop.

1st.

2d.

3d*

4th.

5th.

6th.

7th.

8th.

9th.

10th.

nth.

Northwestern Dent com .

T *2

Algeria corn .

t e

Sorghum, Minnesota Amber . .

•r

T A

T 2

Sorghum. Dakota Amber .

T A

T C

2 2

Sudan grass (first crop) .

1 7

1 J
28

Sudan grass (first crop in open) . .

Alfalfa, E23-20-52 (second crop) . .

Alfalfa, E23 (second crop) .

Alfalfa, E23 (second crop in open)

Alfalfa, 162-98A (second crop). . . .

3°

The loss from com and sorghum during the week preceding the meas¬
urements was very slight, amounting to less than 1 per cent of the total.
The mean transpiration of the four crops for the second and succeeding
weeks was 2, 5, 10, 14, 13, 15, 14, 12, 1 1 , and 7 per cent of the total. The
drop in the sixth week is not significant and is probably due to imperfect
measurements of transpiration during days when rain occurred.

Sudan grass in its early stages of growth developed more rapidly in the
inclosure than in the open. Alfalfa produced its second crop in a period
of five weeks, two-thirds of the total transpiration occurring during the
last two weeks of the growth period. The transpiration coefficient
changed very rapidly, being approximately one-third the maximum in
the second week and two-thirds the maximum in the third week.

Oct. 23, 1916 Daily T ranspiration during Normdl Growth Period

195

MEASUREMENTS IN’ 1915

The daily weighings in 1915 were begun as soon as the plants had
started to grow, and the total transpiration period of wheat, oats, barley,
rye, and flax is included in the transpiration-evaporation graphs pre¬
sented in figure 9. These graphs show the march in the transpiration
independently of the fluctuations due to daily changes in the weather,
assuming the latter to be represented by the fluctuations in evaporation.
On the same assumption, the march of the weekly transpiration under
uniform weather conditions* is expressed numerically in Table XIX in
per cent of the total transpiration.

Table XIX. — Weekly transpiration under uniform conditions of evaporation expressed
in terms of percentage of the total, IQ15

Week of growth.

Crop.

Wheat, Kubanka 1440 .

Wheat, Galgalos .

WTieat, Washington Bluestem

Oat, Swedish Select .

Oat, Burt .

Barley, Hannchen .

Rye, spring .

Flax, North Dakota (C. 1. 13).
Flax, North Dakota (C. 1. 19).
Flax, Smyrna .

1st.

2d.

3d.

4th.

Sth.

6th.

7th.

8th.

9th.

10th.

nth.

12 th.

13th.

The wheats reached their maximum transpiration in the eighth week
following emergence and were harvested in the twelfth or thirteenth
week. Oats, barley, and rye, which were planted a little later, reached
their maximum in the seventh week of growth and were harvested in
the tenth or eleventh week. The flax varieties reached their maximum
in the sixth week, although they were not harvested until five weeks
later.

THE TRANSPIRATION COEFFICIENT DURING THE EARLY PERIODS OF GROWTH

The transpiration-evaporation graphs during the early development of
the crops either approximate a straight line, or curve upward as in the
case of Sudan grass. The latter form suggests an exponential relation¬
ship, which would mean that the rate of increase in the transpiration co¬
efficient is proportional to the transpiration coefficient itself.

Tet T represent the transpiration, E the evaporation, and — = & the

transpiration coefficient (referred to evaporation) at the time t in the
development of the crop. If the rate of increase of the transpira-
dk •

tion coefficient — is proportional to the transpiration coefficient, then

lit

in which o' is a constant of proportionality.

55858°— 16 - i

196

Journal of Agricultural Research

Vol. VII, No. 4

Integrating equation (1) and transforming to common logarithms, we
have

log10k = at + c (2)

in which c is the logarithm of k when t = 0 — that is, at the beginning of
the period under discussion. Expressing (2) exponentially, we have

k = 1 oat+c = /e0 . 1 oat (3)

Therefore if the logarithms of the daily transpiration when plotted
against the time form a straight line, the condition expressed in equa¬
tion (2) is satisfied, and the transpiration, coefficient increases in ac¬
cordance with the assumption made above.

^ The accompanying graphs (fig. 10 to 15) show that an approximate
linear relationship does exist between the logarithm of the transpira¬
tion coefficient and the time in the case of corn, sorghum, Sudan grass,

Fig. 10. — Graph showing a linear relation between the logarithm of the transpiration-evaporation ratio

of Sudan grass and the time.

and alfalfa. The transpiration coefficient of these plants during the
early stages of growth therefore changes exponentially.

The numerical value of the coefficient a may now be computed. This
is represented by the slope of the graphs in figures 10 to 15 and from
equation (2) it follows that

_ log kt - log kQ (4)

The significance of a can be readily seen by a comparison of equation
(3) with the compound interest law

= h(i + t>
\ 100/ ’

from which it is evident that

IO° — I 4;

IOO.

100

, or

Oct. 23, 1916 Daily T ranspiration during Normal Growth Period

197

in which is the interest rate. From this relationship the daily rate

of increase (daily interest) can be readily determined.

The rate of increase may also be expressed in terms of the time re¬
quired for k to double in value.

Let kx be the value of k at the time tv If k = 2kx,

- ioa - « - 2’

a (t - Q = log102 = 0.3010,

0.3010

Fig. ii.— Graph showing a linear relation between the logarithm of the transpiration-evaporation ratio
of Sudan grass (grown in the open) and the time.

The values of the coefficient a, the rate of increase r , and the time
(t — tt) required for k to double in value are given in Table XX for the
various crops considered in figures 10 to 15.

It will be seen from Table XX that the daily increase in the transpira¬
tion coefficient of Sudan grass during the early stages of growth was ex¬
tremely rapid, ranging from 16 to 21 per cent. Since this is compounded
daily, four days are required for the transpiration coefficient to double
in value during this active growth period. The growth rate during the
early stages of Algeria corn and sorghum was less rapid, about seven days
being required for k to double in value.

Journal of Agricultural Research

Vol. VII, No. 4

I98

The results obtained in case of alfalfa are of special interest, for the
plants were much more advanced than those just considered. In fact, in
two of the three alfalfa measurements given the periods were terminated
by the harvesting of the plants while in full bloom. The data show that
during the periods considered the transpiration coefficient of alfalfa was
compounding at the rate of 8 to 9 per cent per day, thus doubling in value
every 8 or 9 days.

-A 5

- A*

-A2

-A O

•2.6

•2.4

- 2.2

•2.0, . .

t6 20 22 24 26 26 20 2 4 6 6 /O /2 /4 /6

xJU/V£ /S/4 'JUl*'

Fig. is. — Graph showing a linear relation between the logarithm of the transpiration-evaporation ratio

of Algeria corn and the time.

Table XX. — Rate of increase in the transpiration coefficient of different crops in IQ14

Crop.

Observation period.

(a)

Daily rate
of increase.

Days re¬
quired for
k to double
in value.

Com, Northwestern Dent .

Corn, Algeria .

June 18-July 9 .

June 18-July 11

O. 026

.044

Per cent.

6. 2
10. 7

11. 6
6.8

Sorghum, Minnesota Amber .

June 18-July 9 -

. 041

9.9

7-3

Sudan grass (in inclosure) .

June 18-July 10 -

.066

16. 4

4. 6

Sudan grass (in open) .

June 24-July 11 -

. 082

20. 8

3* 7

Alfalfa, E23 (first crop in open) .

June 16-July 10. . . .

•033

7-9

9. 1

Alfalfa, E23-20-52 (first crop) .

June 19-July 9. ..

•°37

8.9

8. 1

Alfalfa, E23-20-52 (second crop). . . .

July 18-Aug. 5....

•037

. 8.9

8. 1

A test of the validity of the computations may be obtained by com¬
paring the observed graph of the transpiration coefficient with that com¬
puted from equation (3). The result of such a computation in the case
of Sudan grass (in the open) is given in figure 16, the computed graph
being represented by the smooth curve, while the observed values are
represented by circles.

oct. 23, 1916 Daily T ranspiration during Normal Growth Period

199

RELATIVE DAILY TRANSPIRATION OF DIFFERENT CROPS

In the preceding discussion the transpiration-evaporation ratio has *
been used in order to eliminate, so far as possible, the daily fluctuations

Fig. 13.— Graph showing a linear relation between the logarithm of the transpiration-evaporation ratio
of Northwestern Dent corn and the time.

in transpiration due to changes in weather. The relative march of the
transpiration of any two crops can, however, be determined directly

Fig. 14. — Graph showing a linear relation between the logarithm of the transpiration-evaporation ratio
of Minnesota Amber sorghum and the time.

by comparing the transpiration day by day; in other words, if two crops
are grown simultaneously and have approximately the same length of
growing season, the relative change in transpiration may be obtained by

200

Journal of Agricultural Research

Vol. Vtl, No. 4

Fig. 15. — Graph showing a linear relation between the logarithm of the transpiration-evaporation ratio
of Alfalfa E-23 (in the open) and the time.

c /(JA/£ /S/4 xJULy

Fig. 16. — Observed daily ratios of transpiration to evaporation during early stages of growth of Sudan
grass (shown by circles) compared with exponential graph computed from the relationship shown in fig¬
ure zi (solid line).

Oct. 23, 1916 Daily Transpiration during Normal Growth Period

201

plotting the ratio of the daily transpiration of one crop to that of the
other.

If such ratios depart from unity to any great extent, the difference be¬
tween the departure of the ratio and that of its reciprocal is so great that
the graphs resulting from the plotting of such ratios are not readily com-

/S/4 <40GUSr

Fig. 17.— Graphs of the daily ratios of the transpiration of the different crops grown in 1914 plotted loga¬
rithmically. F signifies flowering; C indicates the cutting of the crop, which was followed by a new
growth from the established root system; H, heading; and R, ripening.

parable. By plotting the logarithm of the ratio, however, the same
departure is obtained for a ratio and its reciprocal. The transpiration
ratios of a number of different crops grown in 1914 are shown in figure
17, plotted logarithmically.

The first graph represents the ratio of the daily transpiration of Gal-
galos wheat to the* daily transpiration of Kubanka wheat. The early

202

Journal of Agricultural Research

Vol. VII, No. 4

ratios are irregular, owing to rainy weather and to the small size of the
plants. After this period no difference in the transpiration of the two
crops was observable until after July io, at which time the plants began
to show signs of ripening through the dying back of the lower leaves.
There was then a sudden drop in the relative transpiration, followed after
about two weeks by a gradual increase to harvest time; in other words,
Galgalos lost proportionately more water than Kubanka in the final
stages of ripening but less in the early stages of ripening.

The ratio of the transpiration of Hannchen barley to Kubanka wheat
decreased gradually throughout the whole period. The same was true
of Burt oat and spring rye. Swedish Select oat and Kubanka wheat
showed only slight differences in the relative transpiration rate. Barley,
rye, and Burt oat made their greatest demand on the soil moisture early
in the growth season, while the other crops used relatively more water
near the latter part of the season.

The graph representing the ratio of Siberian to Kursk millet shows no
difference in the behavior of these crops. Dakota and Minnesota Amber
sorghum also show no differences except in the early stages of growth.
A marked difference is shown in the two varieties of com included in
the 1914 measurements. The ratio increased gradually throughout the
season, owing to the fact that Algeria is a late com compared with
Northwestern Dent. The first crop of Sudan grass grown in the open
gradually increased its transpiration coefficient with respect to Sudan in
the inclosure, but no difference was evident in the second crop.

Alfalfa E23-2 0-5 2 and alfalfa E2 3 showed no differences in their
transpiration response. Cuttings were used in alfalfa 162-98A. The
plants started more rapidly than the seedlings of alfalfa E23. Not until
the third cutting was this advantage fully overcome by the crop grown
from seed; in other words, during the period covered by the first two
cuttings the transpiration coefficient of alfalfa E23 was gradually in¬
creasing compared with that of alfalfa 162-98 A. The ratio graphs of
the same variety inside and outside the inclosure indicate that the plants
inside grew somewhat more rapidly, the graphs of the second and third
crops having a downward trend.

The measurements in 1915 included the whole life period of the plants
considered (fig. 18). At the beginning and end of the period the amount
of transpiration was very small, which results in irregularities in the
ratios. The graph representing the ratio of Galgalos wheat to Kubanka
wheat corresponds closely with that of 1914, although Kubanka trans¬
pired more rapidly than Galgalos in the ripening stages. In 1914 the
graph indicates a relatively rapid loss from Galgalos at the end of the
ripening season . Galgalos was rusted badly in 1 9 1 5 , and this may account
for the comparatively rapid decline in transpiration rate just before
harvest.

Oct. 23, 1916 Daily Transpiration during Normal Growth Period

203

Two sets of Kubanka wheat were included in the 1915 measurements,
which showed the same water requirement (405 ±6 and 406 ±3). The
second set (pots 109 to 114) produced a heavier crop and was three days
later in ripening, which accounts for the rise in the graph at the end of

Fig. 18.— Graphs of the daily ratios of the transpiration of the different crops grown in 1915 plotted loga¬
rithmically. H signifies heading; D , lower leaves dying; S, that the plants were forming shoots; F,
flowering; and C, the cutting of the crop, which was followed by a new growth from the established
root system.

the season. The rise of the ratio graph is even more marked in Wash¬
ington Bluestem and Preston, owing to the fact that they ripened five
days later than Kubanka. Turkestan showed a relatively greater
transpiration than Kubanka early in the season, while Marquis is nearly
identical with Kubanka throughout.

204

Journal of Agricultural Research

Vol. VII, No. 4

Barley and Burt oat increased in transpiration more rapidly and ripened
more rapidly than wheat. The ratio graph for Swedish Select oat shows a
steady increase from seedtime to heading, when there is a marked drop,
followed at first by a gradual and finally by a rapid rise as the wheat
ripened. The graph for rye shows a relatively greater transpiration than
Kubanka wheat early in the season, decreasing as the season advances.

No marked difference is shown in the flax varieties. C. I. 13 ripened
nine days ahead of the other varieties. Except for a gradual increase
in the ratio graph throughout the first crop, alfalfa E23 and 162-98 Ai
showed no marked differences.

CORRELATION OF DAILY TRANSPIRATION WITH WEATHER FACTORS

AND WITH EVAPORATION

The correlation of the daily transpiration of a plant with the intensity
of the weather factors presents difficulties, owing to the fact that the
transpiration coefficient undergoes a gradual change from seed time to
harvest. (See march of transpiration, p. 189.) It is consequently nec¬
essary in a correlation study to eliminate as far as possible effects due to
changes in the size of the plant and to ripening processes. This can be
accomplished by comparing the ratio of the transpiration on consecutive
days with the ratio of the intensity of a given weather factor for the cor¬
responding days. Since the fluctuations in transpiration from day to
day are large in comparison with the daily change in the transpiration
coefficient, the effect of the latter is thus minimized.

The use of direct ratios is not, however, wholly free from objection,
owing to the fact that the departure of the ratio from unity is not the same
in both directions. Transpiration ratios less than unity will be confined in
the transpiration table to classes lying between o and 1, while transpira¬
tion ratios greater than unity have infinity as their upper limit. This
can be avoided by correlating the logarithms of the ratios, instead of the
ratios themselves, the departure of the logarithm being independent of
the direction in which the ratio is taken. As an example, consider three
successive days during which the transpiration is 2, *8, and 2 kgm.,
respectively. The ratio of the first to the second is 0.25, while that of
the second to the third is 4. The departure of the ratio from unity is 0.75
in the first case and 3 in the second, while the logarithms of the ratios,
— 1.3979 (thatis, —0.6021) and 0.6021, respectively, show the same depar¬
ture from zero.

Since it was not practicable to determine the daily transpiration with
an accuracy greater than 0.1 kgm., the uncertainty of the ratio of the
transpiration on consecutive days increases as the daily transpiration
decreases. For this reason only pairs of terms in which at least one of
the pair showed a transpiration of 0.6 kgm. or more have been used in
the correlation tables.

In all cases where the available transpiration measurements were
sufficiently numerous to justify the procedure, the correlation has been

Oct. 23l 1916 Daily T ranspiration during Normal Growth Period

205

determined between the daily transpiration of each crop and the radia¬
tion, temperature, wet-bulb depression, wind velocity, and evaporation
from the shallow and from the deep tanks, making 6 correlation tables
for a crop, or about 200 tables in all. In order to avoid in the correla¬
tion studies the error arising from the change in the transpiration
coefficient as the crops develop, the writers have, as outlined above,
compared the transpiration ratio on consecutive days with the ratio of
the intensity of the- weather factor on corresponding days. The diffi¬
culty arising from the asymmetry of these ratios has been avoided by
taking the logarithm of the ratio in each case. This is, of course,
equivalent to taking the difference of the logarithms of the two terms
of the ratio, which was the procedure actually followed. In other words,
the data in Tables I and III were first converted to a logarithmic basis,
and the logarithmic differences then determined day by day. In brief,
then, the actual correlation is between the logarithm of the transpira¬
tion ratio on consecutive days and the logarithm of the ratio of the
intensity of the given weather factor on corresponding days. The
transpiration of a crop on a given day thus normally enters twice into
the correlation table for that crop, once as the numerator and once as the
denominator of the ratios involving the transpiration on consecutive
days. The results of the correlation studies are summarized in Tables
XXI to XXIII, inclusive.

Smaix Grains.— The small grains in Table XXI show such similarity
in the response of transpiration to various weather factors that the group
may profitably be considered as a whole. The correlation coefficients are
higher in 1914 than in 1915. It will be recalled that the latter season was
cooler, more cloudy, and included many more rainy days. The correla¬
tion of transpiration with air temperature is usually, in 9 cases out of 12,
slightly higher than with radiation, but the difference is usually less than
its probable error.

The correlation coefficients of transpiration of the small grains with
wet-bulb depression and with evaporation (shallow tank) show a strik¬
ing agreement. Considered as a group, they are markedly higher than
the correlations with either radiation or temperature, and constitute the
highest correlations in the series. Their agreement appears to be due,
in part at least, to the fact that the depression in the wet-bulb tempera¬
ture is dependent upon the rate of evaporation from the muslin covering.
It is of interest in this connection to recall the difference in exposure of
the wet-bulb instrument and the shallow tank, the former being shaded
from solar radiation, while the tank was blackened and fully exposed to
the sun’s rays.

The evaporation from the deep tank showed a lower correlation with
transpiration than the shallow tank. The deep-tank evaporation is corre¬
lated with transpiration approximately to the same extent as radiation
or temperature. In 1915 the correlation of the deep- tank evaporation
with transpiration was markedly lower for many of the small grains.

206

Journal of Agricultural Research

Vol. VII, No. 4

Tabi^eXXI. — Correlation of transpiration of small grains with intensity of weather factors

and with evaporation

Tempera¬

ture.

Wet-bulb

depression.

Evaporation.

Plant.

Year.

Radiation.

Shallow

tank.

Deep

tank.

Wind.

Wheat, Kubanka .

1914

0. 64 ±0. 06

O. 72±0wO5

0. 88±o. 02

0. 8s±o. 03

0. 76±o. 04

0. 26 ±0. 09

Do .

191s

•57± -08

. 66± .06

*77± *05

•73± .06

•5S± *07

. 26± .09

Wheat, Galgalos .

1914

. 63 ± .06

*73± -05

. 86± .03

• 86± .03

• 70± .05

• 29± .09

Do .

1915

•55± .08

. 64± . 06

.8s± .04

.8i± .04

.67± .06

.25± .09

Oat, Swedish Select ....

1914

. 66± .06

.69± .05

.86± .03

• 8s± .03

. ?8± .04

. I7± . 10

Do .

1915

• 62± .07

■ 55± •

.67± .06

■74± -05

•47± *07

.2I± .09

Oat, Burt .

1914

.69± .06

•74± -05

• 89± .02

. 90± . 02

•75± *05

, 2I± . to

Do .

1915

. 66± . 07

.64± .06

• 8o± .05

.89± .03

.64± .06

. I5± . to

Barley, Hannchen .

1914

. 61 ± .07

. 66± .06

.8s± .03

. 86± .03

. 72± .05

. I9± . II

Do .

1915

•49± .09

• 63 ± .06

• 7i± • 06

. 71 ± .06

• 49± .07

,I3± .10

Rye, spring .

1914

• 65± .06

•73± *05

. 94± *01

• 9I± -02

• 77± *04

. i9± . 10

Do .

I9i 5

.6 s± .06

• S5± • 06

• 73 i -05

• 79± .04

• 47± .07

.IS± -09

Mean of 1914 co-

efficients .

•65

•7i

.88

,87

• 75

. 22

Mean of all co-

efficients .

. 62

.66

.82

.65

. 21

Square of mean of

1914 coefficients.

.42

•SO

•77

• 76

• 56

•05

Square of mean of

all coefficients . .

.38

•44

.67

•67

.42

.04

Wind velocity showed during both years a very low correlation with
transpiration.

The relative influence of the various climatic factors on transpiration,
as expressed by these correlations, may now be considered. If these
factors are regarded as independent causative elements in transpiration,
the dependence is shown by the square of the correlation coefficient.
On this basis the transpiration of the small grains during the two sea¬
sons was determined by the different weather factors as follows: Radia¬
tion, 38 per cent; temperature, 44 per cent; wet-bulb depression, 67
per cent; wind, 4 per cent. The fact that the sum of these coefficients
exceeds unity is the result of intercorrelation among the weather factors.

The association of the transpiration of the small grains with evaporation
(shallow tank) is 67 per cent; or the same as with wet-bulb depression.
The evaporation from the deep tank shows an association of 42 per cent
with transpiration.

MiluET, Corn, and Sorghum. — The correlation of the transpiration of
millet, corn, and sorghum with the several climatic factors is given in
Table XXII. As in the case of the cereals, the coefficients are higher
for 1914 than for 1915. The correlation coefficients of the transpiration
of these crops with the intensity of the various weather factors show the
same relationships disclosed in the case of the cereals. The highest
correlation is obtained with wet-bulb depression and with the evapora¬
tion from the shallow tank. A somewhat lower correlation is obtained
in the case of temperature, radiation, and evaporation (deep tank),
which are correlated with transpiration nearly equally. Wind shows as
before a very low correlation with transpiration.

Oct. n, 1916 Daily Transpiration during Normal Growth Period

207

Table XXII. — Correlation of transpiration of millet , corn, sorghum, and Sudan grass
with intensity of weather factors and with evaporation

Plant.

Year.

Radiation.

Tempera¬

ture.

Wet-bulb

depression.

Evaporation.

Wind,

Shallow

tank.

Deep

tank.

Millet. Kursk .

1914

O. 54±0. 07

0. 61 ±0. 06

0. 77 ±0. 04

0. 68±o* 05

0.66 ±0.05

0. 23^0.09

Do .

1915

■58± .07

• 63 ± .06

. io± . 05

. 78 ± . 04

• 49± .08

. I9± • 10

Millet, Siberian .

1914

•S7± .06

.64± .06

. 78 ± .04

. 73 ± .04

.69± .05

• 3i± .08

Sorghum, Minnesota

Amber .

1914

.6i± .05

.64± .05

• 79± .03

. 7o± . 04

. 59± .05

.i6± .08

Do .

19 15

. 64 db .08

• 38± .12

•57± •*©

•75± -©6

• 48± .11

• o7± .14

Sorghum, Dakota Am-

ber .

1914

• 56± .06

• 72± .04

,8i± .03

• 74± *<>4

.$2± .06

.I7± .08

Sudan grass (in inclo-

sure) .

1914

*55± -©6

.84± .03

• 83 ± .03

•93± *oi

• 75± * °4

• 52± *©7

Do .

191 S

. 64± . 08

. 28 i . 12

. 64 ± .08

. 75± .06

• 37± -ii

• 04± .13

Sudan grass (in open) . . .

1914

• 52± .07

. 8i± .03

• 85± .03

. 82 i .03

.6od= .06

• 32± .08

Com, Northwestern

Dent .

1014

. 8o± .04

. 71 ± .04

.8i± .03

. 79 ± . 03

.69 h .05

.28± .08

Com, Algeria .

1914

.62± .06

• 79± .04

,88± .02

• 8s± .03

•7 5± • ©4

•33± *©9

Mean of 1914 co-

efficients .

. 60

* 72

.82

.78

.66

• 29

Mean of all coeffi-

cients .

.64

. 77

• 77

.60

• 23

Square of mean of

1914 coefficients .

■36

•52

* 67

.61

•44

.08

Square of mean of

all coefficients. .

.36

.41

*59

•59

•36

.05

The dependence of transpiration on the various factors, as expressed
by the squares of the mean correlation coefficients, are as follows: Radia¬
tion, 36 per cent; temperature, 41 per cent; wet-bulb depression, 59 per
cent; and wind, 5 per cent. The association of transpiration with
evaporation from the shallow tank is 59 per cent, and from the deep tank
36 per cent. As in the case of the cereals, the evaporation from the
shallow tank and the integrated wet-bulb depression show the same
degree of association with the transpiration.

Legumes. — The correlations of the transpiration of the legumes
(Table XXIII) with the various climatic factors and with evaporation
discloses the same relationships appearing in the other groups.1 Wet-
bulb depression and evaporation from the shallow tank gave, as before, the
highest coefficients. Radiation, temperature, and evaporation (deep
tank) show a somewhat lower correlation, while wind again shows a low
correlation.

The correlation of the transpiration with the intensity of the physical
factors of the legumes is lower than either group considered above, but
the different factors stand in about the same relationship. The depend¬
ence of transpiration on the several climatic factors and the association

1 The correlation coefficients between the transpiration
follows:

With radiation .

With temperature .

With wet-bulb depression . .

With evaporation (shallow tank) .

With evaporation (deep tank) .

With wind velocity .

of amaranthus and the weather factors were as
1914 1915

. o. 40±o. 09 o* 69±o. 07

. .4 s± .08 . 6o± .09

. . 6o± .07 .8o± .05

. 56± .06 . s8± .09

. . 47± .08 .62± .08

. 04± .10 .I5± .14

208

Journal of Agricultural Research

Vol. vn. No. 4

of transpiration with evaporation as shown by the square of the mean
correlation coefficients is as follows: Evaporation (shallow tank), 55 per
cent; wet-bulb depression, 50 per cent; evaporation (deep tank), 35 per
cent; temperature, 32 per cent; radiation, 27 per cent; and wind, 6 per
cent.

Table XXIII. — Correlation of transpiration of legumes with intensity of weather factors

and with evaporation

Tempera¬

ture.

Wet-bulb

depression.

Evaporation.

Plant.

Year.

Radiation.

Shallow

tank.

Deep

tank.

Wind.

Cowpea .

1914

0. s6±o. 06

0. 72 ±0. 04

o* 82 ±0.03

0. 79 ±0. 03

0. 73 ±0. 04

0.31 ±0.08

Do .

1915

1914

. 7S± .OS

• 38± • 09

. 6g± . 06

.82± .04

. 57± • 07

. 04 ± .11

Lupine .

• S8± .06

.63± .06

*75± -°4

. 76 ± . 04

.68± .05

• 36± .08

Alfalfa, E23-20-52 .

1914

• 40± .06

• 48± • 06

• 67± .04

.67± .04

•54± .05

.a8± .07

Alfalfa, 162-98A . .

1914

• 42± .06

• 4S± .06

.67± .04

. 69± . 04

•55± -05

•2S± .07

Do .

Alfalfa, E23 (in inclos-

1915

• S2± .07

■74± • °4

• 6s± .05

. 78 ± . 04

.6i± .06

• I9± -09

lire) .

1914

• 43 i *06

• 50± .06

. 70 ± . 04

•74± -03

.$6± .05

• 32± .07

Do .

1915

1914

• 52± .06
. 47± .06

. 7i± .04

. 66± .05

.69± .06

. 50± .06

. i8± . 08

Alfalfa, E23 (in open) . . .

•55± * °5

• 76=fc .03

. 7fl± .04

•58± *05

• 3i± -07

Mean of 1914 coeffi-

.48

*56

• 73

. 61

•31

Mean of all coeffi-

cients .

Square of mean of

■52

•57

.71

•74

•59

•25

1914 coefficients. . .
Square of mean of

*23

•3i

•53

• 53

•37

. 10

all coefficients ....

.27

•32

•50

•55

•35

. 06

CORRELATION OF DAILY TRANSPIRATION OF ALL CROPS, CONSIDERED AS A
SINGLE POPULATION, WITH THE INTENSITY OF WEATHER FACTORS

The degree of correlation of the various weather factors with the
transpiration ratios of all the plants considered as one population has
also been determined. The coefficients for the two years are given in
Table XXIV, together with their squares. They show the same rela¬
tionships as in the crop groups. The mean values of the squares of the
correlation coefficients for the two years are as follows : Wet-bulb depres¬
sion, 0.55; temperature, 0.38; radiation, 0.30; wind, 0.04; evaporation
(shallow tank), 0.56; and evaporation (deep tank), 0.33.

Table XXIV. — Correlation of transpiration of all crops , considered as a single popii -
lation, with weather f actors , based on logarithmic differences of consecutive days

Weather factor.

1914

1915

Correlation
coefficient. ,

Square of
correlation
coefficient.

Correlation

coefficient.

Square of
correlation
coefficient.

1. Radiation .

0. 50 ±0. 01

0. 25

0. 59±o. 02

0-35

2. Temperature .

. 64 ± . OI

.41

. 59± • 01

•35

3. Wet-bulb depression .

. 79 ± . 01

.62

.691b .01

.48

4. Evaporation (shallow tank). .

. 72± . 01

• 52

■ 75± - OI

• 59

5 . Evaporation (deep tank) .

. 63 zb • OI

.40

. 5i± .02

. 26

6. Wind velocity .

, 26± . 02

.07

. I4± • 02

. 02

Oct. 23> 1916 Daily T ranspiration during Normal Growth Period

209

A SECOND METHOD OF CORRECTING FOR THE CHANGE IN THE TRANSPIRA¬
TION COEFFICIENT IN DETERMINING THE CORRELATION OF TRANSPI¬
RATION WITH WEATHER FACTORS

The effect of the change in the transpiration coefficient can also be
largely avoided in correlation studies involving transpiration through
recourse to the transpiration-evaporation ratio. If transpiration and
evaporation show the same hourly response to weather factors, the
transpiration-evaporation graph for the season would represent the
progressive changes in the transpiration coefficient. But since the
transpiration-evaporation graph always presents irregularities owing to
experimental errors and to the fact that a one-to-one correspondence
does not exist between the two quantities, it is necessary to make use of
a smoothed curve through the observations in applying this reduction.
The ratios for each day of the ordinate of the smoothed graph to the
maximum ordinate will give the transpiration coefficient for each day
in terms of the maximum.1 If, now, we divide the transpiration
observed on each day by the corresponding ratio of the ordinates, we
obtain a series of daily transpiration quantities which are independent
of the size and degree of maturity of the plant, but which still retain
all the daily fluctuations due to environment. These corrected transpi¬
ration quantities can, therefore, be used directly in studying the cor¬
relation of transpiration with any given environmental factor.

In the case of Kubanka wheat, the smoothed transpiration-evaporation
ratio was represented by two straight lines meeting in a maximum on
July 12. The daily transpiration of Kubanka has been corrected on
this basis for the march in the transpiration coefficient and the corre¬
lation with the various physical factors determined. The results of the
computations are given in Table XXV. For comparison, the corre¬
lation coefficients based on the first method (ratios of values on consecu¬
tive days) have also been included in Table XXV. It will be seen that
the coefficients determined by the first method are slightly higher than
those based on the second method. This is to be expected, since the
second method assumes a one-to-one correspondence between trans¬
piration and evaporation, so that departures from such a relationship
distort the computed transpiration. The difference in the coefficients
computed by the two methods is, however, usually less than its probable

1 The daily ratios of the observed transpiration and evaporation can not be used directly, as the following
discussion will show:

Let ~ represent the ratio of the transpiration to evaporation at time i, and D^—=k, represent the
Et £max

maximum ratio.

The daily ratio at time t expressed in terms of the maximum is then

If, now, we divide the observed transpiration Tt by this ratio in order to free the daily observations
from the effect of the change in the transpiration coefficient, we obtain as the quotient simply kEt. In
other words, the observed transpiration is dropped from consideration, and the specific assumption is
made that transpiration is proportional to evaporation. The use of the smoothed graph avoids this as-
# sumption so far as the transpiration of any given day is concerned.

210

Journal of Agricultural Research

VoL VII, No. 4

error, and the correlations are substantially in accord with those pre¬
viously determined.

Tabl3 XXV. — Comparison of correlations obtained by the two methods of correcting for
the march in the transpiration coefficient , Kubanka wheat , 1914

Factor

First method

Second method

Transpiration and evaporation :

Shallow tank .

0. 85 ±0. 03
. j6± . 04

. 64 i « 06

. 72 i .05
. 88± . 02
. 26± . 09

O. 83 ±0. 03
. 67 ± .06
• 77± -03
. 62 ± .07
. 79± • 04
. 27± . 10

Deep tank .

Transpiration and radiation . .

Transpiration and temperature .

Transpiration and wet-bulb depression .

Transpiration and wind velocity .

SUMMARY

The transpiration studies included in this paper were made at Akron,
Colo., during the summers of 1914 and 1915.

The plants were grown in large pots (115 kgm. of soil) and sealed to
prevent evaporation from the soil surface.

The pots were weighed each morning before the transpiration response
to sunlight had set in.

Six pots of each crop were used in the determinations, and were weighed
too. 1 kgm.

Twenty-two crops (132 pots) were included in the 19x4 measurement
and 23 crops (138 pots) in 1915.

Continuous automatic records were also obtained of air temperature,
solar radiation, wet-bulb depression, wind velocity, evaporation from a
shallow tank, and evaporation from a deep tank.

The climatic conditions were exceptionally uniform throughout the
season of 1914. The summer of 1915 was unusually rainy.

During a 10-day period of maximum transpiration the annual crops
lost about one-fourth of the total water lost during the season. The
alfalfas lost during this period almost one-half of the total water trans¬
pired in the production of the second crop.

During a 10-day period of maximum transpiration the daily loss of
water from the small grains ranged from 12 to 16 times the dry weight
of the crop; millets, corn, and sorghums, 6 to 9 times; and alfalfas, 36 to
56 times the dry weight harvested. On the basis of a production of 1 ton
of dry matter per acre, this would correspond in the case of the small
grains to a daily loss of 0.11 to 0.14 acre-inch of water; corn, millet, and
sorghum, 0.05 to 0.08 acre-inch; and alfalfas, 0.32 to 0.49 acre-inch.

The loss of water from the small grains during the period of maximum
transpiration amounted to 1.5 kgm. per square meter of plant surface
per day; Sudan grass, 0.8 kgm; and alfalfa, 1.6 kgm. This is from 5 to 14
per cent of the loss during the same period from a free water surface of
equal area.

211

Oct. 23, 1916 Daily Transpiration during Normal Growth Period

The transpiration of the different crop plants per unit area of plant
surface shows less variation than the transpiration per unit weight of dry
matter. In other words, the greater efficiency shown by certain plants in
the use of water appears to be due more to a reduction in plant surface
than to a reduction in transpiration per unit area of surface. The direct
solar radiation received by the plants at Akron is usually not sufficient to
account for the observed transpiration during the midday hours. In
some of the small grains the energy dissipated through transpiration is
twice the amount received directly from the sun.

The march of the transpiration due to changes in the plant alone
(change in the transpiration coefficient) may be expressed by the ratio of
the daily transpiration to the daily evaporation if we assume the latter
to constitute a perfect summation of the weather conditions determining
transpiration. The transpiration of the annual crop plants (aside fiom
fluctuations due to weather) rises to a maximum a little beyond the
middle of the growth period and then decreases until the plants are
harvested. Perennial forage crops such as alfalfa increase steadily in
transpiration to a maximum at or near the time of cutting. Various
crops show their individuality by departing more or less from these types.

The transpiration coefficient of many of the crops increases expo¬
nentially during the early stages of growth. Sudan grass, for example,
doubled its transpiration coefficient every four days during the early
growth period. Alfalfa throughout practically the whole period between
cuttings doubled its transpiration every eight days.

The relative change in the transpiration coefficients of two crops may
be determined by taking the ratio of the transpiration of the two crops
day by day without the necessity of correcting for changes in weather. *

The correlation has been determined between the various physical
factors of environment and the transpiration of the different crops,
considered both individually and as one population. The correlation
coefficients in the latter case for the season of 1914 are as follows:

Transpiration with radiation, 0.50 ±0.01; with temperature, 0.64 ±
0.01; with wet-bulb depression, 0.79 ±0.01; with evaporation (shallow
tank), 0.72 ±0.01; with evaporation (deep tank), 0.63 ±0.01; and with
wind velocity, 0.26 ±0.01.

The small grains show individually a markedly higher correlation
between transpiration and the intensity of the various physical factors
than was observed when all the crops were combined in one population.
The mean correlation coefficients for the small grains (1914) are as follows :
Transpiration with radiation, 0.65; with temperature, 0.71; with wet-
bulb depression, 0.88; with evaporation (shallow tank), 0.87; with
evaporation (deep tank), 0.75; with wind velocity, 0.22.

The com, sorghum, and millet group and the legume group show a
somewhat lower correlation between transpiration and the intensity of
the physical factors of environment. The plants in the various groups,
55858°— 16 - 5

212

Journal of Agricultural Research

Vol. VII, No. 4

however, show the same relative dependence of transpiration upon the
physical factors. Wet-bulb depression and evaporation (shallow tank)
exhibit the highest correlation with transpiration in all cases, while wind
velocity is correlated with transpiration to a very slight extent at Akron.

The degree of dependence of transpiration of the small grains in 1914
upon radiation, temperature, wet-bulb depression and wind velocity,
considered as independent causative factors, as shown by the squares of
the correlation coefficients is as follows: Wet-bulb depression, 0.77;
temperature, 0.50; radiation, 0.42; and wind velocity, 0.05. Since the
sum of these squares exceeds unity, the physical factors are evidently
intercorrelated. The association of transpiration of the small grains
with evaporation (shallow tank) is 0.76, or the same as with wet-bulb
depression.

PLATE 5

Fig. A. — Six pots of alfalfa used in transpiration measurements. The daily trans¬
piration of each pot of plants was determined independently, and the mean of the six
determinations used as the daily transpiration of the crop.

Fig. B.— Six pots of com used in transpiration studies.

Plate 5

PLATE 6

The type of spring balance and lifting device used in the transpiration measurements.
The balance was checked against a standard weight of 130 kgm. each morning before
and after the weighings. The pot is 16 inches in diameter and 26 inches high. The
cover is provided with i-inch holes, as used with alfalfa and sorghum. The center
hole is used for watering.

JOURNAL OF AGRICULTURAL RESEARCH

DEPARTMENT OF AGRICULTURE

Vol. VII Washington, D. C., October 30, 1916 No. 5

SPONGOSPORA SUBTERRANEA AND PHOMA TUBEROSA
ON THE IRISH POTATO

By I. E. MelhuS, Pathologist, J. Rosenbaum, Mycologist, and E. S. Schultz, Expert ,
Cotton and Truck Disease Investigations , Bureau of Plant Industry

INTRODUCTION

After finding that Spongospora subterranea was well established within
our borders, the writers undertook to learn what its effect might be on
the American potato (Solanum tuberosum) industry, taking up for this
purpose the study of such questions as its geographical distribution and
the factors governing the same, relation of the fungus to the roots and
stems of the plant, and the possibility of its occurrence on other hosts.
Such points as its damage to the tubers, relation to soil types, moisture,
and control measures have received consideration. Some of these points
have not been fully settled, but the data available are published here
because of the widespread interest in this disease and its importance in
many parts of the United States.

GEOGRAPHICAL DISTRIBUTION OF SPONGOSPORA SUBTERRANEA IN

THE UNITED STATES •

When Spongospora subterranea was discovered in Canada and in Maine,
the question as to what would be its distribution in the United States
immediately arose. Although quite extensive, European literature
regarding the disease caused by this fungus contains but little informa¬
tion as to the factors that determine its geographical distribution. Owing
to this fact and to the varied soil and climatic conditions in the United
States, the writers were confronted with a new problem, which has been
studied by means of surveys, by planting infected seed in many different
localities, and by transplanting soil from various Southern States into
the infected section of northern Maine.

DISTRIBUTION AS DETERMINED BY SURVEYS

Since the publication in the spring of 1914 of Bulletin 82 of the Depart¬
ment of Agriculture (8)1, which showed that the disease was then known to

1 Reference is made by number to “ Literature cited,” p. 253.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.
ft

(213)

Vol. vn, No. S
Oct. 30, 1916
G — 96

214

Journal of Agricultural Research

Vol. VII, No. s

be generally distributed in northern and central Europe and the British
Isles and was present in northeastern Canada and northern Maine, the
writers, in cooperation with the Federal Horticultural Board and various
State agencies, have extended the survey and as a result have found that
the disease exists in isolated sections from the Atlantic to the Pacific.
Table I gives its distribution in the United States.

Table; I. — Present distribution of powdery-scab in the United States

First collection.

Collector.

Infected area.

Crop

season.

Date.

Location.

June 23, 1913

Presque Isle, Me ....

Oct. 24, 1914

Chateaugay, N. Y. .

Apr. 26, 1915

Nehalem, Oreg .

May 1, 1915

Hastings, Fla .

July 23,1915

Snohomish, Wash . . .

Sept. 18, 1915

Virginia, Minn .

I. E. Melhus

_ do .

F. D. Bailey. . .

I. E. Melhus. . .
F. D. Heald...

C. N. Frey .

E. C. Stakman

(Aroostook Comity
Washington Coun¬
ty.

Penobscot County
f Franklin Comity .
\ Clinton County. . .
f Clatsop County . . .
\Tillamook County
St. Johns County
Snohomish Coun¬

ty.

Carlton County. . .

Lake County .

St Louis County. .

1912

I9I4

1914

1915

1914

1915

As shown by Table I, there are six infested sections in the United
States, all northern, except the one in Florida. This distribution is
strikingly similar to that of Phytophthora infestans and lies wholly within
its geographical range, which is confined to the northern part of the
United States and to certain sections of the South in which a potato
crop is grown during the winter.

distribution as determined by infected seed in various

localities

In order to determine whether Spongospora subterranea would develop
south of infected sections in Maine and New York, infected seed was
planted in 12 different States along the Atlantic seaboard in the spring
of 1915, trials being made at more than one point in some States.
The experiment was carried on in cooperation with the State pathologists
and others interested in potato diseases, and the seed was planted in
each case in accordance with the general practice in the respective lo¬
cality. The results obtained are summarized in Table II.

As shown by Table II, no powdery-scab developed in any of these
experiments. In each case all of the suspicious-looking specimens were
sent to the writers for examination.

Oct. 3o, 1916 Spongospora subterranea and Phoma tuber osa

215

Table II. — Results of planting seed infected with Spongospora subterranea outside the

infected area

Locality.

Grower.

Number
of hills.

Result.

Arlington, Va .

Bureau of Plant In-

2,944

No powdery-scab.

dustry.

West Raleigh, N. C .

H. R. Fulton .

Do.

New Brunswick, N. J .

M. T. Cook .

2A2

Do.

College Park, Md . . .

J. B. S. Norton .

IOO

Do.

Gainesville, Fla . .

H. E. Stevens .

IOO

Do.

Ithaca, N. Y . .

M. F. Barrus .

Do.

East Marion, N. Y .

F. V. Rand .

Do.

Geneva, N. Y .

F. C. Stewart .

Do.

Newark, Del .

T. F. Manns .

Do.

Arrfhe rst } M p ss . .

A. V. Osmun .

Do.

Norfolk, Va .

T. C. Johnson .

200

Do.

Clemson College, S. C .

H. W.'Barre .

IOO

Do.

New Haven, Conn .

G. P. Clinton .

460

Do.

State College, Pa. . .

C. R. Orton .

200

Do.

Morgantown, W. Va .

N. J. Giddings .

Do.

As the season of 1915 was especially wet and still no Spongospora sub¬
terranea developed in any of the various localities named, it seems safe
to assume that its chances for development in seasons with normal
rainfall would be still more Unfavorable.

Each of the trials referred to was on a small scale and for one season
only, except that at Arlington, Va., which was carried through two
seasons. In this trial 13 rows were planted in 1914 and 10 in 1915, or
in all about 2,944 kills, with infected seed, and also, in 1915, 3 rows, or
384 hills, with clean seed in soil which had been planted with infected
seed in 1914. In none of these rows was a single tuber found to be
infected with Spongospora subterranea; consequently it seemed safe to
assume that the conditions were not favorable to the disease, and this
raised the question as to what constitutes the limiting conditions.

distribution as determined by transplanting soil from other
STATES INTO NORTHERN MAINE

The influence of climate was studied by obtaining 200-pound lots of
soil from 12 of the plots that had been planted with infected seed in
nine of the States mentioned in Table II, shipping it to northern Maine,
and planting it with infected seed of the Irish Cobbler variety. Naturally
these soils varied materially in texture and composition, some being
extremely light, while others were very heavy. These samples were
placed in boxes 2 feet square and 8 inches deep, and the boxes set down
into the soil in a field of virgin land where conditions were favorable for
the development of Spongospora subterranea.

Because of *the late arrival of most of the soils, planting was later
than is usual in this section — that is, on June 19 in eight cases and on

21 6

Journal of Agricultural Research

Vol. VII, No. s

July 3 in two cases — but otherwise the procedure was practically the
same as that generally followed in northern Maine, and the plants
received, so far as possible, the same cultural treatment as potatoes
growing under field conditions. Owing to the late planting and early
frost, the plants had a rather short growing season; and although not
harvested until October 15, the tubers were small and in most cases
immature. However, the conditions were sufficiently favorable to per¬
mit the development of the disease, which was the primary object of the
experiment, and the results from this standpoint are given in Table III.

Table III. — Results of planting potatoes infected with Spongospora suhterranea at
Presque Isle , Me. , in soil from the plots in which diseased seed had been grown in various
States

Source of soil.

Number

Number
of sound
tubers.

Tubers infected —

of hills.

Number.

Per cent.

Diamond Springs, Va .

Amherst, Mass .

40. OO
16. 66

College Park , ■ Md .

Morgantown, W.Va .

25. 00

New Brunswick, N. J .

T-

Unknown a .

2 1 • 40
9* 00
0

Geneva, N. Y .

Mount Carmel, Conn. . .

I A

11. 11
6. 66
27. 27

Gainesville, Fla .

Ithaca, N. Y .

-Lir

■Newark, Del .

Unknown a .

Presque Isle, Me .

20. 00

a In both of these cases the tags showing the source had been lost en route, which made it impossible
to say definitely whence they came.

From the fact that Spongospora suhterranea developed in 8 of the 12
soils tested, as shown by Table III, it seems safe to assume that it would
have resulted in all had it been possible to have more soil and to make
conditions more favorable for the growth and development of the po¬
tato plant. However, the experiment developed the important fact
that powdery-scab was produced under Maine conditions in soils in
which in their native States the disease was not produced.

This raised the question as to whether soil from infected sections
will produce an infected crop when transplanted into a noninfected sec¬
tion. To test this, two lots of soil of 200 pounds each were collected
from a field in northern Maine that produced an infected crop in 1914,
and one was shipped to Washington, D. C., and was used in growing
potatoes in the green house, and the other to Norfolk, Va., where it
was .placed in large iron cylinders sunk into the ground at the edge of a
potato field and planted with clean tubers. The former produced a
crop that became badly infected with powdery-scab, while the plants
produced in the latter developed unusually well and the progeny re-

Oct. 30, 1916 Spongospora subterranea and Phoma tuberosa

217

mained totally free from powdery-scab. This experiment indicates that
when transplanted outside of the infected section infected soil does not
yield an infected crop.

CLIMATIC CONDITIONS THIS CONTROLLING FACTOR IN DISTRIBUTION

Unless climatic conditions are suitable, Spongospor a subterranea will not
develop, as is shown in the preceding pages. The studies in Maine
show that periods of rainfall, followed by cool, damp, cloudy weather
during the growing season, are periods of infection and are highly es¬
sential to the development of the disease. Such periods prevail during
the growing season not only in the northern sections, but also in Florida,
where potatoes are planted in January and February and harvested in
April and May. Moreover, much of the infected area in Florida is in a
section in which irrigation is practiced, the water there being supplied
from artesian wells. The water level here is within a few inches of the
surface for days at a time. Table IV shows the rainfall in the six in¬
fected areas during one growing season.

Table IV. — Average monthly precipitation in powdery-scab-infected areas during the'

growing season of 1914 or 1915

State.

Locality.

Period.

Average

rainfall.

Florida .

Washington .

Hastings (St. Augus¬
tine).

Snohomish .

February-July, 1915 .

April-October, 1914 .

Inches.

4. 20

2. KO

3- 64
4. 70
2.47
2. 79
2. 80

2. OO

Oregon .

Astoria .

. do .

Do .

Glenora .

May-October, 1914 .

Minnesota .

Duluth .

April— September, 1915 .

Do .

Virginia .

. do .

New York .

Maine .

Chateaugay (Danne-

mora).

Presque Isle . . .

April-October, 1914 .

. do .

Do .

. do .

April-October, 1915 .

2. 08

Do .

Van Buren .

April-October, 1914 .

3- 46
3- 38

Do .

. do .

April-September , 1915 .

prevalence and period of existence of spongospora sub¬
terranea IN THIS COUNTRY

In Aroostook County, Me., Spongospora subterranea exists on many
farms. In most cases the disease caused by this organism occurs in
isolated spots, varying from a fraction of an acre to 5 acres; but in
other places— for example, sections north and northwest of Caribou,
including Perham, New Sweden, and Stockholm — infection is quite
general. The disease is always most prevalent on wet, poorly drained
land.

During the harvest season of 1914 several fields were examined; and
notwithstanding the fact that fairly clean seed which had been treated
with the usual strength of formaldehyde had been used, from 50 to 75 per

218

Journal of Agricultural Research

Vol. VII, No. 5

cent of the crop was found to be infected with Spongospora subterranea.
In the case of two barrels of tubers selected at random from a 15-acre
field near Caribou that had been planted to potatoes in 1914 after
they had been picked by laborers and the healthy and infected tubers
separated, 68 per cent by weight were found to be infected with the
disease. The field in question was no exception, so far as the prevalence
of the disease is concerned. Again, in the case of 117 hills which were
dug at the same time by hand at random over 2 acres of this field and the
infected and healthy tubers separated, 63 per cent of the progeny was
infected and only two hills were free from the disease. It is needless
to say that the crop thus infected could not be marketed for table use;
and as the only way to turn it into money was to sell it for starch pur¬
poses, the grower sustained considerable loss, potatoes for starch usually
selling for less than 1 5 cents per bushel.

The question as to how long land infected with Spongospora subterranea
will remain contaminated is an interesting one. Pethybridge (12, p. 352),
of Ireland, holds that it will remain infected for three years. In the
case of two fields which, after growing a crop of potatoes, were in
oats for one year and in meadow for four years and were planted to
potatoes in 1914, over 50 per cent of the tubers were infected with
5. subterranea , although there was every assurance that the seed used
was flee from the disease. Numerous other cases indicate that the
disease can live in the soil for more than three years, and from facts at
hand the writers believe that it can live for at least five years and proba¬
bly much longer. It is evident, therefore, that when a piece of land
once becomes infected, a very long rotation is necessary to rid it of
infection and its value for growing potatoes is materially diminished.

The facts that Spongospora subterranea can live in the soil for at least
five years and that it is prevalent in certain sections of Maine raise the
question as to what length of time it has existed in this country. Its
distribution in three counties of Maine, extending from the southern
coast to the northern boundary, and in areas scattered over thousands
of acres in the State naturally required some time; and consequently it
must have existed in this country for a considerable period.

Not only is the disease widely distributed in Maine but also in the
adjoining Province of New Brunswick, the St. John River Valley being
quite generally infected, and in Prince Edward Island, which has long
been settled. As the earlier settlers of northern Maine came from parts
of New England farther south and up the St. John River Valley and as
more than 60 per cent of the population of the State is of New Brunswick
origin it is evident that over half the inhabitants of Maine came from a
section generally infected with Spongospora subterranea . In view of this
fact and that the disease exists in this country and in Canada, it seems
reasonable to believe that it has existed in the infected sections of Maine
for at least 15 years. 4

Oct. 3o, 1916 Spongospora subterranea and Phoma tuberosa

219

susceptibility of roots, stolons, and stems of potato plants

TO SPONGOSPORA SUBTERRANEA

Although Spongospora subterranea is known only as a disease of the
tuber of the Irish potato, the fact that its causal organism is a slime
mold and that many other species of the family flourish on the root
systems of their hosts led the writers to suspect that it infects the
other underground parts of the potato. Very meager information on
this phase of the subject was obtained from a survey of the literature on
the disease. Johnson (5) incidentally mentions seeing pustules on
stolons and Pethybridge (12, p. 352) mentions evidence of the disease
on roots and sprouts, but in no way do these references show the preva¬
lence and actual relationship. In the fall of 1914 a lot of 200 pounds of
soil was collected in a field in which a crop of potatoes infected with
S. subterranea had just been grown. Some of the soil was placed in
1 2 -inch pots in the greenhouse and planted with potatoes known to be
free from powdery-scab. When the crop had been harvested, which
was before all the vines were dead, and the root systems carefully washed
so that they might be examined for any signs of infection, roots of all
sizes were found to be very generally infected with white galls (PI. 7,
fig. C, D) strongly resembling the well-known legume nodules. On
sectioning these galls (PI. 8, fig. A) when nearly mature, they were found
to contain a large number of immature spore balls of S. subterranea .
Similar galls were found on the stolons and main stems of the plant
(Pi. 7, fig. A, B), and in two of the pot cultures pustules were found on
the stems about 1 % inches above the surface of the soil, this latter infec¬
tion having taken place probably while the soil covered this portion of
the stem.

In one case in which the galls on a single plant were counted, 149 were
found on the roots, 19 on nine stolons, and 8 on three stems; besides,
some were doubtless lost in the process of disentangling the roots. This
plant produced nine tubers, four infected, and five free from Spongospora
subterranea .

The presence of Spongospora subterranea on the roots of potato plants
growing in pot cultures in the greenhouse naturally raised the question as
to whether infection is as prevalent under field conditions. Careful watch
was kept for evidence of the disease on the roots of plants growing in
the soil-treated plots on infected soil at Caribou during the summer
of 1915. The first infection was not found until August 5, although
examinations had been made weekly from the time the plants began to
come up. At this time the galls, which were on the small rootlets both
near the surface and deep down in the ground, were white and no larger
than pinheads; but day by day they became more pronounced until
finally they were comparable to those found in the greenhouse. No
infections were found on the tubers when first noted on the roots,

220

Journal of Agricultural Research

Vol. VII, No. s

although over 200, many as large as hen's eggs, from 47 hills were
examined.

On August 11, 100 hills were dug from one row, and on examination
galls were found on the roots in every case, and in many instances
were more numerous than on the roots of the plants grown in the green¬
house. From the progeny of these hundred hills five tubers were found
that showed under the epidermis of the tuber early stages of infection
with Spongospora subterranea , consisting of purplish brown, fimbriate,
bacterial-colony-like spots, some no larger than pinheads. Infections
on the roots progressed more rapidly than on the tubers, as shown by the
foregoing statements, but whether this was due to the former's having
been infected earlier or whether the fungus developed faster on the root¬
lets is not known.

It is not unusual — in fact, it is very common — for the root system of
a potato plant to be quite generally infected while the tubers remain
totally free. This occurred in the case of the Green Mountain and Irish
Cobbler varieties, which are generally grown in the vicinity of Caribou.
It was very common in the variety plots to find the root system badly
infected and the tubers absolutely clean. This leads to the natural con¬
clusion that the critical test for the presence of the disease in a field or
section is the freedom of the root system, and that the roots and not the
tubers are the organs of the plant which determine the resistance of
potato plants to the disease.

The development of new galls on the roots in the field ceased before the
vines died or were killed by frost. There was an outbreak of root infec¬
tion on August 5, and this lasted about two weeks; but after that no new
galls could be found, and those present matured and broke up into a mass
of spore balls of Spongospora subterranea in exactly the same way as does
the content of the sorus on the tuber. Why new infections did not
continue to develop is not well understood, as there were plenty of young
tender rootlets and the soil conditions were apparently comparable with
those which existed earlier. It may well be that there are host relations
and environmental influences that check the disease after it has reached
a given stage of development or time of the year.

When infected seed was planted in virgin soil, root infection took place
in 57 days; where clean seed was planted in infested soil, it took place in
69 days. The first galls found were on rootlets close to the diseased
parent tuber, but later others showed infection. Although not as
numerous on the plants grown in infected soil, it was not uncommon to
find from 30 to 50 galls on a single plant grown in virgin soil. The
periods of infection mentioned do not, of course, represent the incubation
period of the disease, because of the fact that some time elapsed before
the plants threw out roots. A better idea of the time required for infec¬
tion and development sufficient for galls to become conspicuous can be
formed from the following experiment.

oct. 3o, 19x6 Spongospora subterranea and Phoma tuber osa

221

Twelve plants with strong root systems and from 8 to 10 inches high
were washed and transplanted from clean to infected land in the field on
August 5, the day on which root infection was first discovered, and for
three or four days they were well watered and shaded. On August 19
three of the plants were taken up and the roots examined, but only one
gall was found. However, when the nine others were dug, on September
3, and the roots examined, from 3 to 1 1 galls each were found on 6 plants,
which suggests that from 14 to 34 days must elapse after the plant reaches
a certain stage of development before infection with Spongospora sub -
terranea takes place.

A similar experiment was begun on September 7, plants of approxi¬
mately the same age as those used on August 5 being transplanted from
healthy to diseased soil. The plants were dug on October 1 and the root
systems and tubers carefully examined, but no infection was found.

The small amount of infection in the case of the first experiment and
absence of infection in the second is explained by the fact that very few
infections took place on any of the plants on the plots after August 15,
as previously noted.

With a view to determine whether the plant must be a certain age
before infection can take place, 200 seed pieces were planted on July 26
in infected soil in the field. On August 20, when the plants were 3 inches
high and had extensive root systems, 100 hills were dug and examined,
but no infection was found, although in an adjoining row planted on May
26 the roots were generally infected, infection having occurred between
August 1 and 15. On September 24 the remaining hundred hills were
dug and examined, but no infection was found, although doubtless the
plants had some roots by August 15, which tends to indicate that the
host tissue had not reached the susceptible stage.

Very little is known about the factors that favor infection. Moisture,
however, is doubtless an important limiting factor. Lime increases the
amount of powdery-scab on the roots and tubers. Root galls were
especially large and abundant on the plots that received lime at the rate
of 3,000 pounds per acre. That injuries also increase the tendency to
infection was indicated by the finding of numerous galls on lesions
caused by fungi other than Spongospora subterranea. In one case four
galls were found on an injured portion of a plant that had been partially
broken off and had recovered, the indications being that infection had
taken place after the injury.

NEW HOSTS OF SPONGOSPORA SUBTERRANEA

As soon as it was found that Spongospora subterranea infects the root
system of the potato plant, investigations were undertaken to determine
whether it infects the roots of other solanaceous plants. Fifty-three
species of solanaceous hosts were planted at Caribou on June 2, 1915, in
powdery-scab-infected soil in adjoining beds about 3 feet square, potatoes

222

Journal of Agricultural Research

Vol. VII, No. s

having already been planted on three sides of the experimental plot.
Of the 53* species planted, only 16 grew; but this is not surprising, in
view of the fact that many were from tropical sections of the world,
while the summer season at Caribou is distinctly temperate and short.
The solanaceous plants which became infected are :

Solatium warscewiczii.
Solatium haematoclodum.
Solatium mammosum.
Solatium marginatum.

Solanum ciliatum.
Solanum commersoni.
Lycop ersico n escu ten turn .

Those which remained free from infection are :

Solanum nigrum.
Solanum mauritianum.
Solanum duplosumatum,
Solanum labelii.
Solanum heteracanthum

Solanum seaforthianum.
Solanum laciniatum.
Solanum torvum.
Solanum sp.

As shown by the lists, infection resulted on 7 of the 16 species that
grew, and all that grew were species of the genus Solanum except Lyco-
persicon esculentum (tomato) .

In the case of Solanum commersoni , a tuber-bearing plant very closely
related to 5. tuberosum , the cultivated potato, 50 hills were planted and
46 grew. On August 20 six of these hills were dug and examined.
Numerous galls, similar to those on the potato plant, were found on the
roots, but the tubers were free from infection, as were also the tubers
from the remaining hills when harvested, on October 10.

In the case of Lycopersicon esculentum , the seed was sown rather thick ;
consequently the plants, none of which grew to be more than a foot
high, were crowded in the bed and the soil was a solid mass of roots.
Infection was first found on August 20, at which time the galls were
quite pronounced and were present on roots of all sizes (PI. 9, fig. B, C).
The galls were examined for evidence of immature spore balls; but, in
marked contrast to the potato galls, in which immature spore balls were
very evident, none were found. Even as late as September 24 the galls
contained no spore balls. This is believed to be due to the fact that
the host plant continued rapid growth until killed by frost; for as long
as the growth of the host continues the fungus penetrates deeper into
the tissues, without any marked tendency to form spore balls.

In the case of Solanum warscewiczii , a subtropical ornamental plant
and probably a native of South America, 22 of the 66 plants growing in
the bed were infected. Two very interesting points in the reaction of
this host are the large size of the galls and their formation in a ring
around the taproot (PI. 9, fig. A). Although there seemed to be no
difference in general vigor between the infected and noninfected plants,
nevertheless this girdling of the taproot doubtless interferes materially
with the natural processes of the plant. As in the case of Lycopersicon

223

Oct. 3of 1916 Spongospora subterranea and Phoma tuberosa

esculentum, no spore balls could be found even after the plants had been
killed by frost.

The other four hosts did not do as well as Lycopersicon esculentum and
Solanum warscewiczii, but their root systems were quite extensive and
showed numerous galls having superficial characteristics common to
those in the potato and tomato. The phloem was vigorously attacked
by the parasite, which caused the xylem and vessels to be twisted out
of their normal course in many instances, and the cells of the phloem
were hypertrophied and contained a considerable quantity of starch
grains.

The most significant fact brought out in these tests is that infection
took place in the root system of the tomato, and this is important in
view of the fact that the tomato, is very extensively grown in the United
States, and often on land used for potatoes. The roots are very gener¬
ally infected and the distortions and malformations more conspicuous
and destructive than those on the potato (PI. 7, fig. C, D).

When Spongospora subterranea was found to thrive on hosts other
than the potato, examinations were made of the root systems of weeds
common in and about potato fields, including members of the Cruciferae,
Labiatae, Scrophulariaceae, and many other families, and also of culti¬
vated plants, such as clover, oats, wheat, and barley; but in no case
were signs of infection found. No wild species of Solanum were found
growing as a weed in the potato fields or in vacant lots near by.

• HISTOLOGY OF THE GALLS

The galls on the roots of the potato are simple in structure and may be
termed kataplasmatic galls (PI. 7, fig. C, D). The funghs or exciting
agent flourishes only in the phloem or meristematic tissue (PI. 8, fig. A),
as explained by Woronin (17) in the case of rootlets of cabbage seedlings.
Occasionally amebse are seen in the xylem, but they have never been
found to be numerous ofi to show any signs of stimulating the cells to
further growth. The plasmodium of the fungus enters the cell and causes
marked hypertrophy, which distends the cell until it is from 80 to 140/i
in length. Later this divides by cross walls until instead of one large
cell there may be as many as six normal-sized cells, all of which have a
nucleus and are surrounded by a goodly number of amebse, which would
indicate that there is some sort of reaction between these bodies. Occa¬
sionally, however, one or more of the amebae are distantly removed from
the host nucleus. As the gall enlarges, this hyperplastic growth of the
phloem tissues often pushes the vascular system out of its normal posi¬
tion. The infected cells always contain numerous large and small starch
grains, which do not wholly disappear until the spore balls are mature.
Osborn (11) claimed that the fungus feeds on starch; but were this the
case, such an abundant supply would not always be present. It seems
clear, however, that the fungus stimulates the protoplast to produce

224

Journal of Agricultural Research

Vol. VII, No. 5

abundant starch. The amebae finally bunch up around the host nucleus,
which disappears before the spore ball is mature. The amebae vary from
2.5 to 3ju in diameter, are uninucleate and spherical, and contain a single
nucleus, which stains heavily with safranin (PI. 8, fig. A). As in the case
of the tuber, the diseased root cells differ from the healthy in that they
show a large number of various-sized starch grains.

In Solanum warscewiczii the infection is likewise confined to the
phloem and the infected cells occur in groups (PI. 8 , fig. B) . These groups
originate by continual division of one or more cells in both L. esculentum
(PI. 8, fig. C) and 5. warscewiczii (PI. 8, fig. B), which has been described
by Nawaschin (10) for “clubroot” and called by him “ Krankheitsherde.”
However, the starch grains are not as abundant nor as large in the cells
of £. warscewiczii as in those of the potato or tomato; in fact, numerous
infected cells without a trace of starch are often found. The host nucleus
in the infected cells does not differ in size from those found in diseased
cells, but the cells proper are increased in size and elongate and form the
giant cells referred to by Kunkel (6) in the case of the potato. Such cells
are also referred to in the case of Ostenfeldiella attacking Diplanthera
wrightii . Two such giant cells from the potato partially surrounded by
smaller healthy cells are shown in Plate 8, figure B. In one of these giant
cells it will be noticed that there is only a single cross wall. In another
case one of the cells after dividing into three cells showed early stages of
another division. Each of the three cells measured approximately 22.3/z
in diameter and was almost square, while the cells immediately surround¬
ing these and belonging to the same histological tissue measured only
11.25/*. The amebae do not differ in size from those in the tomato, and
as in the case of that host are found grouped around the nucleus.

It will be noticed that spore balls were not produced on either the
tomato or Solanum warscewiczii . This is due to the fact that the host
plants had not reached maturity before they were killed by frost, the
tomato being about 1 foot high and 5. warscewiczii just blossoming.
There can be no question, however, as to the causal agent in either case,
as the amebae in both were uninucleate and showed the same measure¬
ments, staining reaction, position in the host cell, and presence in the
phloem, which reacts alike in all of the different hosts.

SPONGOSPORA SUBTERRANEA ON THE TUBER

The first evidence of Spongospora infection on the tubers consists of
faint, brownish purple, fimbricate, discolored areas about the size of pin¬
heads and resembling in outline very small bacterial colonies (PI. 10,
fig. B). These spots indicate that the causal organism has entered and
destroyed the cells immediately under the epidermal layer. At this stage
the infection is comparatively superficial ; but in from six to eight days the
spot may increase in diameter to cm., lose its brownish color, and be
replaced by a somewhat jellylike protuberance, consisting of metaplastic

oct. 30, 1916 Spongospora subterranea and Phoma tuberosa

225

tuber tissue filled with spore balls of the fungus. The diseased tissue
gradually dies, disintegrates, and liberates the spore balls, the sorus at
this time being mature. The earliest infections on the tuber usually occur
about the stem end, but as the host matures pustules may develop about
the eye end.

The symptoms do not always agree with the foregoing, but may vary
in accordance with external conditions. For instance, it would seem that
if a tuber were detached from the mother plant at the time infection is
becoming visible and immediately placed in a moist chamber the devel¬
opment of the pustule should continue; but this is not the case, for while
the infected area may increase a little, its developmentis seriously checked.
Again, if the stems are cut off at the surface of the soil in the early stages
of infection, the formation of a pustule is checked, but the discolored
area may increase and the epidermis may be raised slightly, which symp¬
toms suggest that the plasmodium continues to vegetate rather than to
break up and form spore balls. Two rows from which the tops were
removed on August 1 1 , six days after the first infection was found on one
of the plots, showed only 3 per cent infection, while the checks showed 24
per cent, but the former showed a reduction of about 75 per cent in yield.
In like manner, if plants are attacked by lateblight shortly after infection
with Spongospora subterranea begins to appear, further infection is
stopped, and that which has already begun makes but little progress.
This occurred in many fields in Maine during the season of 1915. The
severe outbreak of lateblight checked the development of the potato
crop and of powdery-scab. On the other hand, if the potato plant
continues to grow after infection has once taken place and external
conditions are favorable for the development of 5. subterranea , the sori
increase in depth and diameter (PI. 10, fig. C), which suggests the canker
stage of 5. subterranea known in Ireland. The tuber shown in Plate 10,
figure C, is the progeny of a plant that grew in the greenhouse and did not
mature and die down until 152 days after planting, although its tubers
showed infection after 76 days. The unusually large size of the sori in
this case is attributed to the long growing period of the plant, which was
about 40 days longer than usual. Such sori are common where potato
plants are grown in wet, infected soil in the greenhouse and would doubt¬
less result in the field if the crop had a longer growing season. In most
potato-growing sections along the Atlantic seaboard only about 100 days
are allowed, this being especially true in northern Maine.

These greenhouse experiments suggest that the absence of the canker
stage in the United States may be due to the short growing period for the
potato plant in districts in which Spongospora subterranea thrives.

A careful examination of material showing the canker stage as it exists
in Ireland 1 (PL 10, fig. E) proved that it differs from the prevailing type
of scab caused by Spongospora subterranea prevalent in this country.

1 This material was obtained through the kindness of Dr. George H. Pethybridge.

226

Journal of Agricultural Research

Vol. VII, No. s

The symptoms are not fully comparable with those produced in the
greenhouse, although they had some points in common, as is shown in
Plate 10, figures C, D, and E. The diseased surface of this material pre¬
sented the appearance of having been gnawed and chewed by insects
after S. subterranea had made its appearance. The wide variation in the
symptoms of this disease emphasizes its dependence on environmental
conditions.

The resemblance of the injury caused by the flea beetle to that caused
by the sori of Spongospora subterranea is very striking, as evidenced by
the fact that specimens showing the former were received from several
pathologists with an inquiry in each case as to whether the injury was due
to powdery-scab. There is no doubt that insects play a r61e in the de¬
struction of the diseased portions of the potato plant. The injury prob¬
ably caused by the flea beetle and other subterraneous larvae (PI. io, fig.
A) can be distinguished from that due to 5. subterranea by the fact that
in its earlier stages there is a minute central opening caused by the punc¬
ture of the insect, and in its later stages the affected area may show defi¬
nite splitting, which, however, does not extend very deep into the tissues.
Frequently the central portion of the raised area is crumbled, and it is at
this stage that the trouble is most often confused with the open pustules
of 5. subterranea . At this stage also it shows another similarity to 5.
subterranea — that is, the sunken area immediately surrounding the sorus,
which is due either to what is designated and described later in this paper
as physical drying out (Pl. n, fig. F, H) or to the action of the plasmo-
dium (PI. n, fig. D).

In. the course of the work the writers also received many specimens
showing enlarged lenticels (PI. io, fig. F), resembling superficially the
early stages of powdery-scab, with an inquiry in each case as to whether
the injury was due to Spongospora subterranea. Intumescence associated
with the lenticels develops when the tubers are held in moist chambers
for io days. The proliferation of the cells below and around the lenticels
at the end of that time resembles the protruding hyperplastic tissue of
the sorus in early stages of formation. A microscopic examination and
the identification of spore balls within the sorus, however, is the cri¬
terion for determining the disease, but, as will be pointed out in the
following paragraphs, care should be taken not to confuse the spore form
of other fungi with the spore balls of S. subterranea.

CONFUSION IN EARLIER WRITINGS DUE TO THE FUNGI ASSOCIATED
WITH SPONGOSPORA SUBTERRANEA ON THE POTATO TUBER

Wallroth (15) and Berkeley (1) described fungi associated with a
potato disease which bear a striking resemblance to the spore balls of
Spongospora subterranea , and earlier investigators observed numerous
fungus threads in the sori, all of which resulted in much confusion and
error. The writers believe that the “ bulbils” of a species of Papulospora

Oct. 30, 1916

Spongospora subterranea and Phoma tuber osa

227

which they isolated from the sori of S. subterranea were largely respon¬
sible for this confusion and error.

Wallroth saw in the spore balls of Spongospora subterranea a striking,
resemblance to the smuts and accordingly described the organism as
Erysibe subterranea. Berkeley (1) more than 20 years after Wallroth’s
observations published a drawing of two spore forms in connection with
an article on the potato murrain and refers these two forms to the spores
of Tubercinia sp. Fr., which cause a scab. The spore forms shown in
Berkeley's drawings are certainly not the spore balls of 5. subterranea , as
the former show a distinct pedicel, but Berkeley’s description of their
effect on the tuber is highly suggestive of the latter fungus. The surface
of the potato, he says, is “covered with pustules, which at length become
cup-shaped and are powdered within with an olive-yellow meal consisting
of the spores of a fungus.” The drawings and description of the fungus
create some doubt as to whether he referred to 5. subterranea. In view
of his illustration and statement that he saw the various stages of growth
attached to flocci, it is possible he was working with material which in
addition to being infected with 5. subterranea was contaminated with the
“bulbils” of Papulospora sp. or with the forms referred to by Home (3)

The presence of fungus hyphae in the sorus and the consequent hypoth¬
esis that there are two kinds of powdery-scab is shown by the following
statement by Johnson (5, p. 172) :

In some Scotch material I examined, hyphae were clearly present in the scab areas
outside of the cellular tissue of the tuber, and though some could be accounted for as
the mycelial hyphae of Rhizoctonia scab there were others not so explicable. The
hyphae are swollen, septate, and branching; their contents abundant and granular.
In some case chains or masses of spores may be seen arising from the protoplasmic
contents of the hyphae. Spongy spore-balls, very like those of Spongospora, arise,
and ultimately the inclosing walls of the hyphae disappear and leave the balls lying
free. The more external balls lose their compactness and break up into single spores
or small groups of spores, so that they form a finer powder than Spongospora, whose
spore-balls remain intact to the end. The mode of origin of the spores is not unlike
that met with in the Ustilaginea , so far as they have been studied ; and it thus appears
as if . . . there are two kinds of potato scab characterized by powdery spore-balls, the
one with a plasmodium — Spongospora subterranea Wallr. — the other a Hyphomycete,
possibly one of the Ustilaginaceae.

Horne (3, p. 377, 380) says:

If spore-balls are present, they are frequently associated with the hyphae of various
fungi — the association is so close in some instances that it is difficult to convince one¬
self that the spore-balls are not the reproductive bodies of the fungus. ... In late
stages of the disease, and even in the powdery-scab stage, the spore-balls . . . are
frequently intimately associated with the hyphae of various fungi. The spore-balls
appear sometimes to be attached to hyphae, or hyphae twine around them and link
them together.

These statements are indeed significant and show that extreme care
must be taken in identifying the spore balls of Spongospora subterranea ,
55859°— 16 - 2 .

228

Journal of Agricultural Research

Vol. VII, No. 5

which are never attached to hyphae. Plate 12, fig. G, shows figures given
by Home of spores of fungi associated with S. subterranea , and referred
by him to species of Verticillium and Stysanus, and also drawings of the
“bulbils" (PI. 12, fig. E) of Papulospora sp. found by the writers; the
similarity between these is very striking.

As early as 1883 Eidam (2, p. 41 1-414) called attention to the similarity
between a species of Papulospora and one of the smuts. It is the “bul¬
bils" of Papulospora sp. that bear such a striking resemblance to the
smuts and also to the spore balls of Spongospora subterranea . Hotson (4)
published a paper that treats in part of the genus Papulospora and
includes a good description of the “bulbil." In this paper he says (p.
299):

Lastly, among the structures which bear a striking resemblance to bulbils, the
peculiar spore-balls of Spongospora subterranea (Vallr.) Johnson, should be mentioned;
which although they might readily be taken for a species of Papulospora, have been
shown to belong to the life-cycle of one of the Mycetozoa.

In view of the recent isolation of Papulospora sp. by one of the writers
from tubers affected with Spongospora subterranea this statement is very
significant. A culture of this species of Papulospora submitted to
Dr. J. W. Hotson was identified by him as Papulospora coprophila
(Zukal) Hotson. All inoculations on the potato tuber with this fungus
gave negative results.

CONTROL MEASURES

Although it has been known for more than 50 years that Spongospora
subterranea occurs in most northern and central countries of Europe, very
little effort has been concentrated on its control ; and such studies as have
been made apply only to local conditions in Ireland, where cultural, soil,
and weather conditions are markedly different from those in the potato¬
growing sections of the United States. As soon, therefore, as it was
found that the disease had become established in some of the leading
potato-growing sections of this country, the study of its control became
imperative and was undertaken by the writers along four lines : (1) Early
harvesting, (2) seed treatment, (3) varietal response, and (4) soil
treatment.

Early harvesting

That Spongospora subterranea develops on the tubers only after they
are partially mature has already been shown. Although carefully sought
for during the entire season of 1914 in the plots in which infected seed
was planted, no infections were found on the crop until August 20; and
not until about three weeks later did it become common or conspicuous.
The first case of infection in the warehouse at Caribou was found on
September 12 by the Maine Potato Inspection Service, which employed

Oct. 30, 1916 Spongospora subterranea and Phoma tuberosa

229

about 50 inspectors. By that time 12 per cent of the crop had been dug
and much of it doubtless thereby saved from infection. Were infection
to appear as late* every season, early harvesting would be a very simple
and effective means of avoiding the disease.

The results obtained in 1914 suggested early harvesting as a means of
wholly avoiding the disease where the crop was growing on land known to
be infected. To test this four rows 16 rods long were planted on infected
land in 1915. Two of these rows were harvested on August 15, when
about two-thirds mature, or about two weeks earlier than usual, the har¬
vesting season here beginning about September 1 and extending to Octo¬
ber 10. The remaining two rows were dug on October 10. The two rows
dug in August produced 323 pounds; of this lot 1 53 tubers were infected.
The two dug in October produced 412 pounds, of which 167 tubers were
infected. These figures show that practically all the infection in 1915
made its appearance before August 15, and that from the standpoint of
control nothing was gained by harvesting on August 15, while 89 pounds
of potatoes were lost in the two rows harvested at that time. However,
had infection taken place as late in 1915 as it did in 1914, the crop on a
given farm might have been harvested before it became infected ; conse¬
quently it is believed that this line of attack has not been exhausted and
that it has greater possibilities than the results obtained in 1915 indicate.

SEED TREATMENT

In the spring of 1914 an experiment to control the disease by disin¬
fecting the seed potatoes badly infected with Spongospora subterranea
was begun, Green Mountain, the variety commonly grown in northern
Maine, being used. The seed was treated as shown in Table VI.

Table VI. — Results of experiments in controlling powdery-scab by seed treatment in igi 4

Plot

No.

Treatment.

Average

weight.

Total

num¬

ber.

Hills.

Num¬

ber

sound.

Num¬
ber in¬
fected.

Per¬
cent¬
age of
infec¬
tion.

Pounds.

I :

Check, infected seed .

i-39

200

137

6S.0

2 ;

Seed wet and rolled in sulphur . .

1.48

152

44.0

Formaldehyde (1:30), hours . . .

i- 27

164

138

15-8

Atomic sulphur <5 per cent), iM hours .

i- IS

202

148

26. 7

5 i

Formaldehyde (2:30), hours .

1. 04

136

116

14. 7

Mercuric chlorid (2:15), x% hours .

1.05

189

184

2. 0

Check, infected seed . . .

164

42.0

Copper sulphate (5 per cent), iJ4 hours .

•94

142

137

3- 5

Mercuric chlorid (4:15), 53°-*54°C., 5 minutes .

•93

6.6

Formaldehyde (2:30), 46-50° C., 5 minutes .

.90

1.8

Mercuric chlorid (4:15), 44°-45°C., 5 minutes .

1. 14

8.0

Check, healthy seed .

1.25

200

198

1.0

230

Journal of Agricultural Research

Vol. VII, No. 5

Tabl3 VI.— Results of experiments in controlling powdery-scab by seed treatment in IQI4 —

Continued

Plot

No.

Sound tubers.

Infected tubers.

Treatment.

Num¬

ber.

Weight.

Num¬

ber.

Weight.

Per

cent.

Check, infected seed . .

1,790

Pounds.

222

53i

Pounds.

56-5

22. 87

Seed wet and rolled in sulphur .

1,613

200

221

26.0

12. 00

Formaldehyde (1:30), 1 X/Z hours .

1,613

200

9.0

3-30

Atomic sulphur (5 per cent), 1 lA hours .

i,750

217

118

16.5

6.30

Formaldehyde (2:30), hours .

1,097

136

6-5

3-68

Mercuric clilorid (2:15), hours .

1, 600

198-5

• 75

*37

Check, infected seed .

1, 260

i55

170

2.4

11. 80

Copper sulphate (5 per cent), 1% hours .

1,068

556

132-5

1.0

•83

Mercuric chlorid (4:15), 53°-S4° C., 5 minutes .

•75

1.24

iol

Formaldehyde (2:30), 46°-'>o'sC., 5 minutes .

Mercuric chlorid (4:15), 44 *“45° C., 5 minutes .

397

49-25

•5

•50

681

84-5

i- 5

i- 73

Check, healthy seed .

1, 640

204

•5

. 20

After treatment the seed was spread out to dry on a laboratory floor
which had been previously covered with paper. When dry, the potatoes
were cut and placed at once in i-peck sacks that had been previously
soaked for three hours in a 1 per cent solution of copper sulphate or in
sacks that had never contained potatoes. In this way they were kept
apart from other potatoes and other sources of infection until planted,
June 13. The seed used in the clean checks was carefully selected from
several barrels believed to be free from powdery-scab, treated with the
usual strength of mercuric chlorid, dried, cut, and bagged in the man¬
ner just described. Except in the case of clean check tubers, which
were wholly free, each seed piece bore a considerable number of powdery-
scab pustules.

The plot selected for the experiment was an old orchard which sloped
gently toward the east and which had been in sod the preceding 10 years.
The soil had been ridged up toward the apple trees; and consequently
the ground was somewhat uneven. The soil was a black gravelly loam,
but too wet and heavy for ideal potato land. A complete commercial
fertilizer was used at the rate of 1,200 pounds per acre. Because of the
irregularity of the ground and shading from the orchard trees, no
attempt was made to record the germination of the seed or the develop¬
ment of the potato vines during the season. The relative position of the
plots with reference to each other was the same as in the table, and the
crop was harvested on October 15.

As shown by Table VI, the progeny in the plots adjoining plot 11, with
the exception of one tuber in each of two hills, was free from infection
by Spongospora subterranean which showed that the land was not infected
previous to the planting in 1914. Notwithstanding this nominal infec¬
tion in the healthy check tubers, the amount of powdery-scab that devel¬
oped on the various plots would represent the relative efficiency of the
various treatments if other conditions were equal, which was not the case,

231

Oct. 3of 1916 Spongospora subterranea and Phoma tuberosa

for, as already stated, the land was uneven on account of the ridging,
and plots 1 and 2 were at the edge of the orchard and had more sunlight
and consequently gave a larger average yield per hill.

The hill was considered a unit, each being harvested separately. The
tubers were examined individually, being washed before examination
if not clean when taken out of the soil, and a record made of the number
of infected and the number of sound tubers in each hill and also of the
gross weight of each.

As shown by Table VI, some of the progeny of the checks planted
with diseased seed gave from 42 to 68 per cent of infection, the varia¬
tion depending doubtless on the soil and water conditions in different
parts of the field.

In drawing conclusions from the results given in Table VI it must be
borne in mind that in each case only a comparatively small number of
hills were used and that variation in the soil and moisture contents
materially influenced the results. Notwithstanding these facts, how¬
ever, it is perfectly obvious from the tests that several chemicals have
a deleterious effect on the development of Spongospora subterranea , the
most active being mercuric chlorid and formaldehyde, hot solutions being
more effective than cold.

During the season of 1915 the most promising of the experiments
made in 1914 were duplicated, the land used being cleared in the spring
and put in condition for planting. This land was much lower than that
used in 1914 and not so well drained. The soil was a rather heavy gray
silt with considerable humus at the surface. The soil of parts of this
plot was of the type on which much of the infection in northern Maine
occurs. The land was planted by hand on June 10, the seed being
handled in the same way as that used in the experiment of 1914. The
arrangement of the plots with regard to each other and the results
obtained when harvested on October 10 are given in Table VII.

The most striking result of this experiment is the infection of the
control plots planted with healthy seed. One of these check plots was
on each side of the field, which consisted of about one-fourth of an acre,
and one in the middle, the former being numbered 1 and 9 and the latter,
in which there were four rows, 18. Plots 1 and 9 received the same
treatment as the plots used in the experiment in 1914, but plot 18 re¬
ceived special care in preparation and planting; moreover, the seed used
in this plot was rigidly selected, inspected, and treated with double¬
strength mercuric chlorid. Everything with which it came in contact
until planted was also disinfected with mercuric chlorid.

232

Journal of Agricultural Research

Vol.'VII, No. 5

Table VII. — Results of experiments in controlling powdery-scab by seed treatment in IQ15

Plot

No.

Treatment.

Hills.

Tubers.

Total

num¬

ber.

Num¬

ber

sound.

Num¬

ber

in¬

fected.

Per¬
cent¬
age of
infec¬
tion.

Total

num¬

ber.

Num¬

ber

sound.

Num¬

ber

in¬

fected.

Per¬
cent¬
age of
infec¬
tion.

Check, clean seed .

16.0

282

263

6. 7

Check, infected cut seed .

65. 6

443

349

21. 0

Formaldehyde (2:30), hours .

25- 7

392

364

7*i

Mercuric chlorid + ethyl alcohol (2:15),

hours; 1,000 c. c. of alcohol in

gallons of water .

18.0

363

346

4.6

Check, infected seed. .

28.6

Atomic sulphur (5 per cent), iK hours.

53-5

469

388

17.2

Mercuric chlorid (4:15), so6 C., 5 min-

utes .

15-3

332

321

3-3

Check, infected cut seed .

83-9

428

304

124

28. 9

Check, clean seed .

16. 4

449

43i

4.0

Wet and rolled in sulphur .

.57

35*o

366

33 2

9. 2

Formaldehyde (1:30), 1%, hours .

11. 8

486

474

2.4

Check, infected cut seed .

31-0

374

329

.12. 0

Formaldehyde^:^), 50° C., 5 min-

utes .

7- 7

366

359

1.9

Mercuric chlorid (1:15), hours .

8.7

386

380

1* 5

Copper sulphate (5 per cent), 1 % hours

18.3

327

300

8. 2

Check, infected whole seed .

78.4

780

506

274

35- 1

Mercuric chlorid (2:15), 1 XA hours .

100

15.0

39i

37°

5-3

Check, clean seed .

138

124

10. 1

643

629

2. 1

As will be seen in Table VII, 10. 1 per cent of the progeny of the hills in
plot 18 became infected, and as in all probability the seed was free from
infection and the disease is not indigenous to virgin land, the question
arises as to the source of the infection. In the case of the healthy control
plots the probable source of infection was the adjoining infected controls,
soil water, animal life, and cultural methods being the probable agencies
by which the disease was spread. Considering the results from each
plot individually and in relation to those in adjoining plots, it would
seem reasonable to believe that in the case of the healthy checks adjoin¬
ing those planted with untreated seed, the latter served as centers of
infection. In 8 out of 10 cases more infection occurred in the rows next
to the untreated checks than in the rows farther away.

Although the infection of the healthy checks diminishes the value of
the seed-treatment experiment, it serves to emphasize the infectiousness
of the disease and to some extent indicates the rate and means of spread
where conditions are favorable for its growth. In the light of the devel¬
opments on the healthy plots in 1915 it is easily understood that the
two cases of infection in the clean control plots in 1914 were doubtless
carried from the infected to the healthy controls.

As in the case of the experiments in 1914, those in 1915 show, in addi¬
tion to the spread of the organism in the soil, that seed disinfection has
a beneficial effect in diminishing the amount of infection and that treat¬
ment with mercuric chlorid and formaldehyde are the most effective,

233

oct. 30, 1916 Spongospora subterranea and Phoma tuber osa

treatments with a hot solution of these for a short time being more effi¬
cient than with a cold solution for a longer time.

The relative efficiency of the various disinfectants, according to the
data obtained for the two years, is shown in Table VIII.

Table VIII. — Average results in controlling powdery-scab obtained from different dis¬
infectants used in the tests of 1914 and 1915

Treatment.

Average
percent¬
age of
hills in¬
fected.

Average
percent¬
age of
tubers
infected.

Formaldehyde (2:30), 46°-5o° C., 5 minutes . .

4- 75

7- 58

8- 00

1. 20

Check, clean seed . 7 .

Mercuric chlorid (4:15), 44°-45° C., 5 minutes . . .

*•73

2.83

1-50

Mercuric chlorid (2:15), 1% hours . . .

8. 50
8. 70
10. 90
10. 95
13. 80

18. 00

Mercuric chlorid (1:15), 1 y* hours .

Copper sulphate (5 per cent), hours .

Mercuric chlorid (4:15), so°-s4° C., 5 minutes .

Formaldehyde (1:30), 1% hours. . .7 .

2.85

Mercuric chlorid +~ethyl alcohol (2: 1 5), hours; i,oooc. c.of alcohol in 3K gallons of

water .

Formaldehyde (2:30), hours .

20* 20

5-39

10.60

Cut and rolled in sulphur .

Atomic sulphur (5 per cent), 1^ hours .

Checks, powdery-scab seed .

64. 60

25* 12

varietae response

Tor a number of years the Office of Cotton and Truck Disease Inves¬
tigations has carried on experiments on varietal susceptibility of potatoes
to the lateblight fungus ( Phytophfhora infestans) and in this connection
has made an extensive collection of European and American varieties
reputed to be more or less resistant to lateblight. The study of powdery-
scab also suggested the selection of resistant varieties as a possible means
of control, and accordingly it was decided to use the above collection in
the experimental plots infected with powdery-scab. In addition to this
collection of standard varieties, a collection of seedlings also was used in
the experiment.1

A piece of ground with a uniform soil type and infected with Spongos¬
pora subterranea was selected and 25 hills each of the different varieties
of seedlings were planted, the Green Mountain variety being planted in
alternate hills as controls. Table VIII gives the varieties tested and
the results of the tests, including the percentage ratio of infection in the
varieties tested and the checks.

1 These seedlings, which have not yet been distributed to farmers and seedsmen, were obtained through
the courtesy of Prof. William Stuart, of Horticultural and Pomological Investigations, Bureau of Plant
Industry, by whom they were developed.

234

Journal of Agricultural Research

Vol. VII, No. s

TablH IX. — Results of tests of potatoes f or resistance to powdery-scab in iqi$

Variety.

Tubers.

Control tubers (Green
Mountain).

Infection®
in terms
of p. ct.
var. X100
p. ct. G. M.

Num¬

ber

plant¬

ed.

Num¬
ber in¬
fected.

Per¬
centage
of in¬
fection.

Num¬

ber

plant¬

ed.

Num¬
ber in¬
fected.

Per¬
centage
of in¬
fection.

Eldorado .

4.8

Farys .

18. I

Prof. Wohltman .

26. 3

Senator .

215

159

5- 3

Ursus .

178

I. I

*5-

24. 6

4. 473

Pearl. . . .

256

O. 7

12. 1

5- 785

Aldona . ‘ .

194

2. O

32- 3

6. 192

Gryf .

452

4-4

27. 2

16. 176

Bonar .

8.8

35- 0

25- 143

Cimbals .

4. 2

12. 5

33. 600

Constantia .

Gracya. . . .

323

12.4

20. 1

42. 612

Gedymin .

215

7* 7

103

14. 1

54. 610

Prof. Wohltman . . .

41. 0

41. 8

98. 086

Kalif .

r93

188

49-3

5°. 0

98. 600

Topaz .

12. 0

144

8.8

■136. 364

Soliman .

147

29. 6

52'

17.4

170. 115

a In order to get a direct comparison between the different varieties, the percentage of infection in any
variety was divided by the percentage of infection of the Green Mountain variety planted in the same row
and the result multiplied by 100.

As shown in Table IX, 4 of the varieties tested escaped the disease,
but none of the checks were entirely free, the percentage of infected tubers
in the latter varying from 4.8 to 50 per cent. Out of the 16 rows, 3
showed over 15 per cent infection in the case of the varieties tested and
10 in the case of the checks, and the latter fact naturally raised the
question as to why the percentage of infection varied so in the checks,
all of which were planted to the same variety.

While it may be that the soil was not generally infected or that condi¬
tions in small isolated spots were unfavorable for the development of the
disease, the fact that there was often a wide difference in the amount of
infection in alternate hills of the control and the other variety in the same
row is significant. A striking example of this latter is shown in the row
planted to Farys and the row of Ursus, the former showing no infection
and the latter 1.1 per cent of infection, while the Green Mountain, the
control variety planted in alternate hills in these rows, showed 18. 1 and
26.6 per cent of infection, respectively. Another interesting reaction was
that in the Soliman variety, which showed 29 per cent, while its check
showed only 1 1 per cent of infection.

Thirty different selections of seedlings were planted in infected soil,
and the checks in this case also were planted with the Green Mountain
variety in alternate hills in each row. Table X gives the details of the
experiment.

Oct. 30, 1916

235

Tabl£ X. — Results of tests of potato seedlings for resistance to powdery-scab in 1915

Seedlings.

Control tubers (Green
Mountain).

Infection
in terms
of p. ct.
seed X 100
p. ct. G. M.

Collection No.

Num¬

ber

plant¬

ed.

Num¬
ber in¬
fected.

Per¬
centage
of in¬
fection.

Num¬

ber

plant¬

ed.

Num¬
ber in¬
fected.

Per¬
centage
of in¬
fection.

1357 .

130

12. 1

1488 .

IOI

3°

40. 5

992 . . .

3- 7

2892 .

12 5

7*9

628 .

5*2

23r5 .

8.4

2387 .

128

8.7

2193 .

I. I

22. 2

4- 955

I522 .

129

2. 2

31* 5

6. 984

1124 . .* .

160

4. 2

43*3

9. 700

22402 .

151

*■3

148

13*4

9. 702

2426 .

141

3-o

17. 1

n. 696

42 59 .

1 19

4. 0

31* 7

12. 618

1429 .

1055 .

4755 .

i74

TI3

112

3-8
*3- 7
I4- 5

i5- 1

48. 2
41.9

15. 166
28. 423
24. 606

3760 .

II3

2. 5

6.9

36. 232

2294 .

116

0. 8

103

1.9

42. 105

295P .

126

22. 7

4i.3

54. 964

1212 .

109

°* 9

194

i* 5

60. 000

1034 .

144

12. 7

18. 9

67. 196

2870 .

171

10. 4

x3* 1

79- 389

13660 .

105

16. 6

20. 7

80. I93

4927 .

22. 5

23- 3

96. 567

4227 .

213

29. 2

25. 2

ii5- 873

968I .

12. 2

8.6

141. 86l

15284 .

44- 7

28. 5

156. 842

14329 .

107

117

52. 2

108

32.9

158. 663

I3896 .

IIO

9*7

6. 1

159. 0l6

1449 .

1 14

55-8

109

16. 8

332- 143

As will be seen by a comparison of Tables IX and X, the response to the
disease was very similar in the standard varieties and the seedlings, except
that the extremes of infection were greater in the latter, ranging from o
to 55.8. Seven of the seedlings showed no infection, while the checks
were infected in every case, the infection varying from 1.5 to 48.2; or in
the rows of the nine cases referred to, from 3.7 to 40.5. Although the
tubers of these nine varieties were free, the roots showed a goodly number
of galls, which indicates either that the roots and tubers do not resist the
disease to the same degree or that the tubers merely escaped the disease.
As already explained, infection in 1915 took place only for about two
weeks during the growing season — that is, between August 1 and 15 —
. and in case the tubers had not set by this time or had not reached a sus¬
ceptible stage, infection was avoided. It was also observed that, so far
as the amount of infection is concerned, it made no difference whether
the variety was early or late. That there is a close correlation between

236

Journal of Agricultural Research

Vol. VII, No. 5

the development of the cork cambium and susceptibility is not improbable,
and this is strongly indicated by the fact that probably 90 per cent of the
pustules on the infected tubers from the experimental plots were about
the stem ends rather than the eye ends. While this preponderance of
infection about the stem end was true in 1915 on the experimental plots,
the condition in this respect in other fields was not determined, and no
such preponderance was noted during the previous season, when infection
occurred much later, or between August 20 and about September 5. A
critical examination of some of the varieties and of the seedlings attacked
showed that not only the percentage of tubers infected but also the
macroscopic character of the sori differ.

Different varieties of potatoes respond differently to the attacks of the
disease, as shown in Plate A, which represents various kinds of sori. It
was possible to arrange a series from mature sori, varying from the size
of pin pricks to that of 1 or 2 cm. as they occur on such standard varieties
as Green Mountain and Irish Cobbler. The reaction of the host showing
the smaller sori differed markedly from that of the host showing the
larger sori. In some varieties the sorus attains a considerable size
before it bursts, in others it breaks open very early, while in still others
the infected area corresponding to the sorus shows only discoloration and
gives the impression that the plasmodium is spread throughout this dis¬
colored area, but at no point does it cause sufficient proliferation of the
host cells to form a definite open sorus.

The data regarding the response of the different varieties lead the
writers to believe that the response is not due entirely to resistance but
rather to the fact that the tubers of certain varieties have escaped infec¬
tion, and furthermore that a variety which will show much less infection
than the Green Mountain may be found.

SOII, TREATMENT

Five acres on one of the farms first cleared in the vicinity of Caribou
and said to have been under cultivation for at least 30 years were used in
1915 for the study of effects of soil type and soil treatment on the disease.
Because of its close proximity to the village, this land has been cropped
rather heavily to potatoes during the previous 15 years.

Topographically the tract occupies the position of a high glacial river
terrace sloping from west to east. Along the northeast side and in part
forming the northern boundary is a well-marked drainage way or ravine
in which considerable erosion has taken place, leaving the soil exceedingly
stony and with less fine earth than elsewhere in the field. With the
exception of this and a small area at the southwest comer, the slopes are
not steep and there is little or no noticeable erosion. In the central part
of the field the slope appears inconsiderable, but the instrumental deter¬
mination showed not less than 5 feet elevation in 100 horizontal feet, a

Oct. 30, 1916 Spongospora subterranea and Phoma tuberosa

237

slope sufficient for the rapid drainage of surface waters. On the tract were
found certain fairly well marked differences of color, texture, and other
physical characteristics of the soil. These soils, which are here desig¬
nated by a set of arbitrary numbers, are as follows : 1

Soil, 1.0. — The surface is grayish brown to ashy gray, depending on the
moisture content. The texture is silty loam, fairly friable when dry, but
somewhat plastic when wet. Numerous small stones and gravel occur
in the soil and subsoil, the latter being a light ashy-gray or mottled gray
and brown and rather heavy silty loam. The drainage of this soil is suf¬
ficient for surface run-off, but appears to be deficient in the subsoil,
probably due to seepage of water from the higher lying upland. This
type comprises the greater area of the tract.

Son, 1. 1. — The surface of this soil is light brown and is underlain by a
subsoil which is gray and much like that of No. 1.0. The texture is
about the same as that of the foregoing type, though the structure of the
surface is more open and friable, stone and gravel are not so conspicuous
a feature, the surface is more sloping, and the subsoil drainage is freer.

Son, 1.2. — The color of the surface is slightly darker than that of No.
1 .0, though the subsoil is about the same and the texture is a silty loam.
In the area in the northeastern corner of the tract the content of small
stone and gravel is extremely high, so that the amount of interstitial soil
material is considerably reduced. The two small areas on the southeast¬
ern corner of the tract are less stony, the percentage being about the same
as in soil 1 .0. The texture of the subsoil is rather finer than that of the
surface and more compact, the color being a dark grayish brown or gray
slightly mottled with brown to the depth of 3 feet or more. The soil
occupies relative depressions along natural drainage ways, and in the
larger area there is considerable erosion. While the surface drainage
seems adequate, there is so much seepage apparently that the internal
drainage is actually poorer than in soil 1 .0.

Soil, 2.0. — This soil is a medium- to light-brown silty loam and con¬
tains considerable gravel of a smaller size generally than is common to the
other types. The subsoil is reddish brown to yellowish brown, with a
somewhat larger amount of small gravel than in the surface. There is no
mottling, oxidation being quite uniform in both soil and subsoil. This
type differs from all other soils in that there is no gray mottling in the sub¬
soil. The topography is abrupt or rolling and both surface drainage and
subdrainage are free.

Figure 1 2 shows the distribution of the soil types on the portion of the
field used.

1 The description of soils here given was very courteously contributed by Mr. J. E, Dapham, Scientist in
Soil Survey, Bureau of Soils.

2 This map was made by Dr. Oswald Schreiner and Messrs. J. E. Eapham and H. L. Westover, of the
Bureau of Soils, through whose kindness it is published. It was of great value in interpreting results of the
experiments.

238

Journal, of Agricultural Research

Vol. VII, No. s

As there will be seen from this map, there are five different types of
soil, which occur very irregularly in both the upper and the lower block.
The diagram superimposed on the map shows the arrangement of the
plots, the distribution of Spongospora subterranea, the percentage by
weight of the progeny infected with the disease, and the yield per acre.
The seed used was of the Green Mountain variety and was carefully
selected and treated with the usual strength of mercuric chlorid. The
plots, which were numbered i to 15, inclusive, received treatment as
follows : 1

1. — Sodium nitrate, 20 pounds.

2. — Control, with commercial fertilizer, 150 pounds.

3. — Old horse manure, 2 ,400 pounds.

4. — Control; no treatment.

5. — New horse manure, 2,400 pounds.

6. — Phosphoric acid, 24 pounds.

7. — Ammonium sulphate, 20 pounds, and phosphoric acid, 24 pounds.

9. — Potassium chlorid, 30 pounds.

10. — Ammonium sulphate, 20 pounds.

11. — Flowers of sulphur in drills, 90 pounds.

12. — Flowers of sulphur broadcast, 90 pounds.

13. — Calcium carbonate, 300 pounds.

ib. — Sodium nitrate, 20 pounds.

2b. — Control with commercial fertilizer, 150 pounds.

4b. — Control; no treatment.

7b. — Ammonium sulphate, 20 pounds, and phosphoric acid, 24 pounds.

13b. — Ammonium sulphate, 20 pounds.

14b. — Flowers of sulphur, 90 pounds.

15b. — Phosphoric acid, 24 pounds.

As will be seen from the foregoing outline of treatments, the effect of the
common commercial-fertilizer ingredients, as well as of some of the well-
known soil disinfectants, were studied on the plots either with or without
commercial fertilizer. The treatments in all the plots were applied with
the potato planter and at time of planting, except in the case of those
marked “b,” which were applied on August 12, a few days after the first
infections were noted, and each treatment was duplicated on what was
known as the upper and lower blocks.

Seven of the control plots received at the rate of 1,500 pounds of com¬
mercial fertilizer per acre, or 150 pounds each, and an equal number
received no treatment. In the former the percentage of infection ranged
from 1 5. 1 to 35.4 and averaged 24.7 per cent, while the latter showed a
greater variation, ranging from 4.9 to 36.5 and averaging 23.4, or a total
average of 24.05 for the 14 control plots. None of the treated plots
gave as high a percentage of infection as the average of the control plots,
and those treated with sulphur at the rate of 900 pounds per acre gave the
lowest percentage, the average of the four thus treated being 8.7 per cent.

1 All plots marked "b" received treatment on August 12.

Oct. 3o, 1916 Spongospora subterranea and Phoma tuber osa

239

Lime at the rate of 3,000 pounds per acre increased the amount of infec¬
tion 13.2 per cent over its nearest control in the upper block, but dimin¬
ished it as compared with the control plot in the lower block, and in this
case also reduced the yield 28.3 bushels per acre. The difference in the
reaction of lime on the two blocks, however, was probably due to the
difference in the soil types in the two cases, which are plainly shown in
figure 1.

When the crop was harvested, every other row in each plot was dug
with the machine and the tubers examined and sorted before being picked
up. A stake was placed at the last case of infection in each row, which was
determined and plotted as shown by the shaded area in figure 1. In the

Fig. i.— Map of the experimental plots at Caribou, Me., showing their arrangement, distribution of
Spongospora subterranea, percentage by weight of the progeny infected with the disease, and the yield
per acre.

lower block the boundary line between the infected and the noninfected
portions of the plots was more difficult to mark, the infections being few
and scattered south of the lightly-shaded area in the block, especially
in the area between plots 7 and 3. The fact that soil 1 . 1 is partly infected
in the corner is doubtless due to seepage from soil 1 .0 lying above it.

The significant point brought out in this study is the close relationship
of the development of Spongospora subterranea to the Washburn silt-loam
type of soil, which is marked “1.0” on the map. It was repeatedly
observed that where infection was bad the soil had a grayish surface,
which showed plainly that it was of the 1 .0 silt-loam type. Exceptions to
this have been found, but in general there is a close correlation.

240

Journal of Agricultural Research

Vol. VII, No. s

Table XI shows the comparative value of the treatments on plots
similar in other respects :

Tabi<£ XI. — Comparative effect of different soil treatments on powdery-scab and on the

yield of potatoes

Plat

No.

Treatment.

Percent¬
age of
powdery-
scab.

Bushels
per acre.

Sodium nitrate .

16. 95
24. 7

i38-3
205. 8

Control with commercial fertilizer .

Old horse manure .

l8. 2

165. 9

Control with no fertilizer .

23. 8

123. 2

New horse manure .

II. 05

108. 8

Acid phosphate .

20. ' I

143. 2
204. 8

’ 114.9

Ammonium sulphate and acid phosphate .

Potassium chlorid .

17. I

11. 5

Ammonium sulphate .

12. 7

I37- 5

Sulphur drill .

10. 1

140.9
13a 0

Sulphur broadcast .

7-35

Calcium carbonate .

J7‘ 3

104. 4

As shown by Table XI, ammonium sulphate and acid phosphate gave
nearly the same yield as the checks fertilized, and in addition dimin¬
ished the amount of infection by Spongospora subterranea 7.6 per cent.
In the plots treated with potassium chlorid there was less infection than
in any of the plats receiving 'other fertilizer ingredients, or only 11.5
per cent. This is attributed to the slow growth of the plants in the
early part of the season and their continued growth until killed by frost.
The potassium chlorid apparently prolonged the growing season, and it
may be that the crop in a measure escaped the infection period, which,
as already shown, was in August.

While these experiments extended through only one season and conse¬
quently only tentative conclusions can be drawn, they demonstrated
that sulphur at the rate of 900 pounds per acre applied broadcast reduced
the amount of infection by Spongospora subterranea , and all of the fer¬
tilizer ingredients tested reduced the amount of infection from 5 to 12
per cent when applied alone.

DRY-ROT ASSOCIATED WITH SPONGOSPORA SUBTERRANEA

Although Spongospora subterranea has been known in Kurope since
early in the forties of the last century, no mention has been made of a
dry rot commonly associated with and following the disease. This rot
differs markedly from the many rots which have already been described
and which are common to the potato tuber, and a discussion of it neces¬
sarily involves the description of several types of rot not heretofore
distinguished and a study of their causes. This is the rot to which the
senior writer in his brief mention (7) of the shriveling and shrinking

oct. 3o, 1916 Spongospora subterranea and Phoma tuber osa

241

which occur around some of the sori, and the importance of the part
played in this connection by the wound parasites which enter through
the injury caused by 5. subterranea . In view of the fact that the rot was
first found associated with 5. subterranea and is most common on pota¬
toes infected with the disease caused by that organism, it seems desirable
to call it “powdery-scab dryrot.”

HISTORY AND DISTRIBUTION

Powdery-scab dryrot was first observed by the senior writer on pota¬
toes infected with Spongospora subterranea collected in New Brunswick,
Canada, in the fall of 1913, and held in storage in Washington, D. C.
Later it was collected in Aroostook County, Me., and in the infected
section of New York State.

Although first found in America, the rot is not confined to this coun¬
try. A typical case of it was found in a shipment from Ireland by
Mr. H. B. Shaw, Pathological Inspector of the Port of New York. About
a bushel of the infected tubers collected and sent by Mr. Shaw to one
of the writers for examination as to the causal organism showed pro¬
nounced discoloration and shrinkage and later on a rot. In the fall of
1913 three i-barrel sacks were taken at the port of New York from a
shipment of potatoes from the Netherlands which showed a considerable
percentage of infection by Spongospora subterranea and was shipped to
Washington, where the diseased tubers were separated from the healthy
ones and both lots placed in storage. When these were examined two
months later, 2 1 per cent showed shrinkage and rotten spots.

A dryrot has been found on infected tubers collected in Chile, South
America, but in this case the rot was not marked, as in the case of the
European tubers, and may well have been accentuated by the long
period of transit under poor storage conditions. Be this as it may,
however, it is perfectly clear that to some extent at least a dryrot follow¬
ing infection by Spongospora subterranea develops on potatoes, no matter
where they are grown.

PREVALENCE) OP AND LOSSES FROM POWDERY-SCAB DRYROT

Powdery-scab dryrot develops on potatoes after they have been held
in storage for some time. It is accelerated by poor storage conditions,
but even in good storage from 30 to 75 per cent of the tubers become
partially or wholly decayed and consequently worthless for seed or table
use. Morse (9) refers very briefly to a rot connected with Spongospora
subterranea which is doubtless the same as the one under consideration
here. This rot, he states, develops on potatoes held in good storage
and is hastened when infected potatoes are subjected to ordinary room
temperature for a few days.

242

Journal of Agricultural Research

Vol. VII, No. 5

While it is not an uncommon thing to find powdery-scab-infected
tubers entirely decayed, the rot generally occurs in spots. The spots vary
from i to io cm. in diameter. They may be only slight depressions in
the superficial layers, or they may extend into the center of the tuber.
In this respect powdery-scab differs markedly from the scab caused by
Oospora scabies , with which it may be easily confused before the dryrot
sets in. The rotting of the tubers following infection by S. subterraned
not only distinguishes this disease from that caused from O. scabies , but
emphasizes its more destructive nature.

Some idea of the prevalence and destructiveness of powdery-scab dry-
rot can be formed from the percentage of this rot on tubers grown during
the season of 1913 infected by Spongospora subterranea. Infected
potatoes were collected from individual growers at different points in
Aroostook County between May 2 and June 15, 1914. These were washed
and separated into two lots, those showing dryrot and those free from
the rot but infected with powdery-scab. Later the lots were weighed,
and the percentage proportion by weight of those showing the rot is given
in Table XII, together with other data regarding the tubers used.

Table XII. — Percentage of powdery-scab dryrot on potatoes of the IQI3 crop infected

with Spongospora subterranea

Date of
collection.

Place of collection.

Variety.

Quantity

collected.

Percent¬
age of
powdery-
scab dry-
rot.

May 2
May 5
May 5
May 21
June 10
June 15

Patten , Me .

Green Mountain .

Bushels.

Presque Isle, Me .

Irish Cobbler .

Caribou, Me .

Green Mountain .

Patten, Me .

. do .

Caribou, Me .

. do .

Ashland, Me .

Irish Cobbler, Green

Mountain.

As will be seen from Table XII, nearly 50 per cent of the tubers showed
the dry-rot stage, the percentage of rot on the two varieties ranging from
30 to 73.

Although limited, these observations indicate that the Green Mountain
is more severely attacked than the Irish Cobbler. This may be due to
varietal differences, but it seems more logical to believe that it is due to
the fact that the Green Mountain is a later-maturing variety.

The rot was most prevalent on the tubers infected with the scabby
stage of Spongospora subterranea , and this explains the low percentage
found on the two lots collected at Presque Isle and Caribou on May 5.
A very high percentage of the tubers badly infected with the scabby
stage showed the dryrot in some form. Not only does the scabby stage
of powdery-scab mar and deface the tuber and render it objectionable in

oct. 30, 1916 Spongospora subterranea and Phoma tuber osa

243

the market, but it may be followed by a rot which will render the tuber
worthless for either table or seed purposes.

Naturally the question arose as to whether tubers infected with Spon¬
gospora subterranea harvested while immature will rot more than such
tubers harvested when mature ; and in order to get light on this phase of
the subject, experiments were carried on with the crop of 1914. About
two weeks before the harvest season, or on September 9, 91 tubers were
dug from the experimental plots and examined. Out of the total num¬
ber, 67 showed various stages of infection, ranging from a few small,
immature pustules to a generous sprinkling with sori, while 24 were free
from the disease. After this the diseased tubers were placed in one sack
and the healthy tubers in another, and both sacks were placed in a potato-
storage cellar, such as is commonly used in northern Maine. On June 1 5,
1915, both sacks were examined, and it was found that of the 61 tubers
in the sack containing the infected potatoes 31 were two-thirds or wholly
rotten, 4 had rotten spots from 2 to 4 cm. in diameter, and 32 had rotten
spots from 1 to 2 cm. in diameter. With one exception, which was per¬
fectly sound, the sori apparently having corked over perfectly and pre¬
vented the entrance of fungi, all the tubers generally infected with
sori of S. subterranea were rotten. The 32 tubers with small dryrot spots
consisted of those which showed few small sori of 5. subterranea . Of the
24 tubers in the sack containing the healthy potatoes, 3 were two-thirds
or wholly rotten, 1 had a rotten spot 2 cm. in diameter, and 20 were
sound, except in the case of 2 which were infected with common scab.

In another case 1 bushel of powdery-scab-infected tubers was col¬
lected from a field at the beginning of harvest, or on September 24. At
this time the tubers were not fully mature, and the sori in most cases
had not ruptured the epidermis; but the badly infected specimens
showed brownish purple areas. These potatoes, together with others,
were placed in storage. When examined for the development of the
rot, on June 10, 1915, 63.9 per cent of the tubers were found to be
decayed two-thirds or more, and the collection was in such a condition
that it could scarcely be handled. In every case the sound tubers were
those which showed but little powdery-scab.

From these results it is evident that powdery-scab dryrot becomes
much worse on potatoes harvested early than on those harvested after
the tubers are fully mature.

In another case 620 tubers, the progeny of no hills in a field infected
with powdery-scab, were harvested on November 7, 1914, the potatoes
being dug by hand. Each tuber was carefully examined, and it was
found that 63 per cent of the lot by weight was infected and that the
progeny of only two hills was free from the powdery-scab. The healthy
tubers, numbering 295, and the tubers infected with S. subterranea ,
numbering 325, were put in separate sacks and placed in storage on
November 9, 1914. When taken out and examined, on June 14, 1915,
55859°— 16 - 3

244.

Journal of Agricultural Research

Vol. VII, No. s

only 32 of the sound tubers, or about 11 per cent, showed signs of decay,
while 188 of the infected tubers, or about 58 per cent, showed the rot.
About 3 per cent of the former and about 58 per cent of the latter were
worthless for table use.

Although interesting, the data gathered from the crop of 1913 was
not definite, owing to the fact that the conditions under which the
potatoes were grown and stored were unknown. More exact data were
obtained from the crop of 1914, which was grown on 15 acres of land
located in a field in which 75 per cent of the hills had produced tubers
infected with Spongospora subterranea. From this 15 -acre plot 16
barrels were collected on October 7, as they were being harvested, and
placed in good storage at Caribou the following day. The amount of
infection shown by these potatoes varied from only a few sori to enough
literally to cover the tuber. On June 15, 1915, parts of 7 of the 16
barrels were sorted over, and the potatoes showing powdery-scab but
no rot were separated from those showing rot. Table XIV gives the
results of this study.

TablK XIV. — Occurrence of powdery-scab dryrot on the crop of IQ14

Barrel No.

Number
of bushels
examined.

Number
of tubers
free from
powdery-
scab.

Number
of tubers
showing
powdery-
scab.

Percentage
of tubers
showing
powdery-
scab.

124

63-9

170

173

50. 1

2 X

190

368

65- 9

124

59- 6

I %

238

149

40. I

I75

55-9

As will be seen from Table XIV, from 38 to 65.9 per cent of the tubers
infected with powdery-scab rotted to such an extent as to be unfit for
table use, and this notwithstanding the fact that these potatoes were
harvested when fully mature and placed in good storage at once, as
already stated.

SYMPTOMS OF POWDERY-SCAB DRYROT

The relation of Spongospora subterranea to powdery-scab dryrot is in
a way comparable to the rot which follows Phytophthora infestans. S.
subterranea causes little rot, but it leaves an open wound, through which
wound parasites may enter. It is well known that P. infestans causes
little rot unless it is followed by bacteria and wound parasites.

The symptoms of the dryrot following infection by Spongospora sub¬
terranea may vary greatly, according to the time of year, storage condi¬
tions, state of the tuber when harvested, and the stage of development
of the sorus when the infected tuber is removed from the soil. In

oct. 3o( 1916 Spongospora subterranea and Phoma tuberosa

245

general, however, the symptoms may be divided into three groups: (1)
Desiccation or loss of moisture from the wound caused by S. subterranea ,
(2) those caused by the germination of spore balls in the bottom of the
sorus and the formation of the plasmodium which destroys the adjoining
cells, and (3) those caused by the entrance of wound parasites into the
sori. Often, however, all three causes work together, and in many cases
the fungi begin their work after some desiccation and plasmodium injury
have set in. It should also be borne in mind that dryrot does not occur
about every sorus of a tuber infected with 5. subterranea . A tuber may
be well sprinkled with sori and yet show no dryrot; or the rot may
appear about a few of the sori, this latter being true in the case of
infected tubers harvested both when immature and when quite ripe and
filled with countless spore balls.

Powdery-scab dryrot due to the desiccation of the tissues adjoining
the sorus results in discoloration, shriveling, and shrinkage. When
infested tubers are harvested, the sori are virtually open wounds, and
when placed in warm, dry storage the temperatures incident to which
often prevail in early fall, desiccation of the living cells bordering on the
wound takes place. When the storage temperature drops with the
advance of the season, this type of dryrot is retarded. It can be induced
readily, however, by taking tubers from storage and holding them for
a fortnight at ordinary room temperature. The cause of the desicca¬
tion of the cells about a sorus is readily seen in sections of the sori made
before and after shrinkage takes place. Plate 12 , figure A, shows the con¬
dition in the bottom of a sorus before there is any appreciable amount
of desiccation. It should be especially noted that no cork has formed
in the bottom of the pit, the adjoining living cells being protected by
the dead debris and spore balls of the sorus. By referring to Plate 14,
figure A, it will be seen that abundant cork cells are formed in connec¬
tion with a sorus of common scab. This explains why dryrot is very
seldom or never associated with this disease,

A comparison of figures A in Plates 12 and 14 shows clearly why desic¬
cation might take place in one case and not in the other.

The condition of a sorus after some dryrot due to desiccation has
taken place is shown in Plate 12, figure B. When the material was
dehydrated, the spore balls of the fungus were washed out. It is espe¬
cially interesting to note the number of dead cells that are still partly
intact. There are also signs of the formation of some wound cork.
Where this was found, the underlying cells were still alive. There was
no evidence in this sorus of a plasmodium, such as occurs in connection
with dryrot due to plasmodium injury. The type of dryrot due to desic¬
cation never causes serious damage, but it does further mar the appear¬
ance of tubers already injured by the sori of Spongospora subterranea.

The second type of dryrot is that caused by the plasmodium of Spongo¬
spora subterranea . The spore balls in the bottom of the sorus germinate

246

Journal of Agricultural Research

Vol. VII, No. 5

and form plasmodia, which move about in the immediate vicinity of the
sorus and destroy the host cells. This type of dryrot has recently been
described by Kunkel (6). The destruction of the cells is followed by a
brownish discoloration and shrinkage, which in the early stages is often
most conspicuous on one side of the sorus, but later surrounds the pustule
and a small depressed area appears (PI. 1 1 , fig. D) . The plasmodium, which
has never been found except in the parenchyma, works largely in the
superficial tissues, producing a dry, hard spot from % to 1 cm. in diameter
and % to X cm. in depth. Isolations made from this type of rot often
give a variety of fungi, most of which are saprophytes; but in some of
these isolations wound parasites exist and probably find such injured
spots excellent points at which to begin the complete destruction of the
tuber.

The entrance of wound parasites through the open sori marks the
beginning of the third and most destructive type of powdery-scab
dryrot (PI. 11, fig. B, C, E). In the bottom of many of the sori there is
little or no wound cork (PI. 12, fig. A, B), and the pit is filled with dead
tissues and numerous spore balls in the same way as open cavities in the
tuber are filled with small masses of culture media on which fungi
naturally flourish.

The most common of the wound parasites found associated with the
early stages of rot is a species of Phoma, producing brownish to gray
lesions (PI. A and PI. 11, fig. A). As these lesions progress they become
more sunken, darker, and often hard and bony; and when removed,
which can often be easily done, they leave a clean and smooth cavity
in the tissues of the tuber. The shape and texture of the spots removed
give the impression of a button; hence the name “ button-rot,” by which
the trouble is known among farmers. Eater stages of the lesions (PI.
13, fig. A, B) vary from 2 mm. to 5 cm. in diameter and often reach a
depth of 2 to 4 cm. (PL 11, fig. C). The diseased tissues are sharply
defined on the surface, where the pustules are numerous and where
infection may take place through each of the pustules; in this way
large areas may become discolored and later depressed (PL 11, fig. E).

Often, after the above-described lesions have formed, other wound-rot
and decay organisms enter, in which case the symptoms are somewhat
confusing (Pl. 11, fig. B, C, E). The more common of these organisms
are Fusarium coertdeum (Lib.) Sacc., F. discolor , var. sulphur eum Schlect. ,
and various bacteria. The symptoms of each of these are typical of
the particular species, as described by Wollenweber (16). These fungi
have been repeatedly isolated and identified from single-spore cultures
as well as from typical rots produced in potato tubers artificially inocu¬
lated. Generally when any of these wound- rot organisms are present,
the lesion is soggy and less firm to the touch, and its surface is often
cracked and broken.

Oct. 30, 1916 Spongospom subterranea and Phoma tuberosa

247

ISOLATION STUDIES

During the past three years a large number of isolations were made
from the dryrot lesions, and a considerable number of parasitic and
saprophytic organisms, including species of Coniothyrium, Ramularia,
Periola, Fusarium, Phoma, Rhizoctonia, Vermicularia, Papulospora, and
Bacteria, were studied as regards their relation to dryrot.

A type of spot, which at first had brownish to gray lesions and later
became hard and dark (Pi. 13, fig. A, B), was found to be very common,
not only in connection with Spongospora subterranea but also on tubers
free from this disease, the senior writer (7) having found it in Maine
before the presence of 5. subterranea had been reported.

When not associated with any of the fungi above mentioned, the
rot caused by the species of Phoma under consideration in section is
slate-colored, dry, and powdery; but when other fungi are associated the
tissues of the tuber show cavities or chambers and the color is character¬
istic of the mycelium or spores of the fungus present — that is, white and
blue in the case of Fusarium coeruleum , light green or sulphur-colored in
the case of F. discolor , var. sulphureum , and dark brown when Papulo¬
spora coprophila is the secondary saprophyte. The following studies
indicate the distinctive characteristics of the rot caused by the species of
Phoma under consideration.

Twenty tubers showing typical symptoms were selected for these
studies. These were washed and immersed in mercuric chlorid (1 : 1,000)
for 10 minutes, after which the surface tissue was peeled off and plant¬
ings made from the newly exposed tissue. In the majority of cases the
plantings gave a pure culture of a species of Phoma. The isolations
were made from button-shaped spots associated with and apart from
Spongospora subterranea , and similar results were obtained in each case.

In order to get more accurate data regarding the association of the
fungus with the lesion, a more detailed study was made. Twenty tubers
having typical lesions were selected on May 17, 1915, an effort being
made to secure tubers in which no other wound fungi had entered. The
lesions, which varied from 6 to 25 mm. in diameter, were firm to the
touch, dark gray, and had the appearance of typical button-shaped spots.
The isolation from each tuber was made on a separate plate, and 4 plant¬
ings were made on each plate, or a total of 80. These were examined on
May 20, and the number and kind of colonies found on each are shown
in Table XIV.

As shown by Table XIV, the 80 plantings produced 6 bacterial and 46
fungus colonies. Microscopic examination showed that the latter in¬
cluded at least three different groups of fungi: One containing 2 colonies,
one 38 colonies, and one 6 colonies.

In order to check up the identity of these colonies, 2 transfers were
made from each of the 46 colonies, to test tubes of potato hard agar and

248

Journal of Agricultural Research

Vol, VII, No. 5

of sterilized sweet-clover stems. The results obtained confirmed the
examinations made direct from the plates. The 38 colonies on both the
agar and the sterilized sweet clover in all the test tubes proved to be a
species of Phoma. Inoculations were also made with one or more of the
fungi from each group, which showed that only the fungus included in
the 38 colonies was pathogenic.

Table XIV. — Results of isolations f rom tubers showing typical lesions caused by the rot

of Phoma sp.

12
T3

Plate No.

Number
of colo¬
nies

Number of growths
produced on plate.

growing
on plate.

Bacterial.

Fungus.

•

Total

INOCULATION OF TUBERS WITH PHOMA SP.

As the fungus was obtained from the tuber, a series of inoculations was
made on tubers, which were selected, washed, and immersed in mercuric
chlorid (1:1 ,000) for 10 minutes, or in some cases in 85 per cent alcohol and
burned over the surface. By means of a flamed scalpel the tuber was cut to
a depth of 1 to 3 cm., a piece of a pure culture from a transfer 4 days old
inserted, and the point of inoculation marked with india ink, after which
the potatoes were placed in a moist chamber at ordinary room temper¬
atures. At the end of four days the first indications of the disease became
apparent. Ten days after inoculation the lesions reached 6 mm. in diam¬
eter. The diseased spots developed until they were from 12 to 25 mm. in
diameter, but after this made little or no progress. The control tubers,
treated in a similar manner but with no fungus inserted in the wound,
remained healthy throughout the entire experiment. The writers have

Oct. 30, 1916 Spongospora subterranea and Phoma tuber osa

249

never been able to produce larger lesions, which is significant, as in nature
larger lesions are rarely found unless other wound parasites have entered.
A tuber from one series of inoculations in which lesions are only 6 mm.
in diameter are shown in Plate 13, figure E, and a series which produced
much larger lesions in Plate 13, figure C, and for comparison tubers
naturally infected are shown in Plate 13, figures A and B, and Plate 11,
figure A. A tuber injured but not inoculated is also shown in Plate 13,
figure D.

Table XV gives the results of a series of inoculations made during the
course of the work.

Tabi^E XV. — Results of inoculating potato tubers with Phoma sp. in 1915

Date of inoculation.

Number
of inocu¬
lations.

Percent¬
age of in¬
fection.

Condition of
control.

March 26 .

Healthy.

Do.

April 3 . .

IOO

April 13 . . .

Do.

April 23 . .

9°

66%

Do.

May 1 .

Do.

May 10 .

Do.

May 15 .

Do.

May 27 .

Do.

In the series of inoculations made on May 1 new potatoes were used.
The fungus was able to produce very slight infection on these new pota¬
toes, but the lesion never exceeded 3 mm. in diameter. The tubers inocu¬
lated May 10 showed only a slight infection. These tubers were held in
the moist chamber until May 25, when they were examined. The in¬
fected area at this time measured 3 mm. in diameter, and in one inocula¬
tion scattered pycnidia could be seen with the naked eye and were as
abundant near the edge of the spot as in the center. Reisolations were
made on poured plates of potato agar from the inoculations made on
March 26, April 3, and May 27, and in every case a pure culture of a
fungus, macroscopically and microscopically identical with the culture
used for inoculation, was obtained. Plate 14, fig. B, shows a section of
one of the inoculated tubers. Most of the mycelium of Phoma tuberosa
is beneath the epidermal cells. This mat later gives rise to the pycnidia,
which grow towards the surface of the tuber.

CULTURAL CHARACTERISTICS

Tests made with the fungus on a number of media showed that it
grows well on sterilized sweet-clover (Melilotus alba) stems, potato hard
agar, potato cylinders, corn meal, and Beyerinck’s agar. The luxuri¬
ance of its growth on all of these leaves no doubt that it is able to grow

250

Journal of Agricultural Research

Vol. VII, No. s

on practically all of the common media. Since its growth does not differ
in any essential characteristic from that of other species of Phoma, only
the following brief cultural characters need to be mentioned : On sweet-
clover stems the growth is fluffy, profuse, and white, turning gray with
age; the pycnidial development is good, especially when little moisture
is in the tube; the pycnospores often ooze to the surface. On potato
hard agar its growth is profuse; the mycelial growth becomes evident
within three or four days, is whitish at first, and at the end of a week
begins to darken to the characteristic gray of Phoma spp.; the pycnidial
development is very scarce. On potato cylinders the growth is fluffy
and white, turning to gray within eight or nine days, the entire culture
turning dark after the development of numerous pycnidia. On corn
meal there is a profuse mycelial growth, at first light gray, turning
darker with age; numerous pycnidia. On Beyerinck’s agar the mycelial
growth is scarce and white, and this medium is especially favorable for
the production of pycnidial bodies (Pi. 12, fig. C, D).

By actual measurements it was found that there is no difference in
size between the pycnospores from the host and those produced in cul¬
ture, those from both sources, with occasional exceptions, varying from
3.7 to 6.0011 in length and 1.8 to 3.7// in width (Pi. 12, fig. E).

TAXONOMY OF THE FUNGUS

Saccardo lists four species of Phoma (14) as occurring on Solarium tu¬
berosum: Phoma nebulosa (Pers.) Mont., P. eupyrena , P. solani Cook and
Harkn., and P. solanicola Prill, and Delacr. The original description
of P. solanicola (13) states that it was found on the stems, but it was
impossible to determine from the literature whether the three others
were associated with the aerial part of the plant or with the tuber. The
measurements for each of these four species and for the one under dis¬
cussion are given in Table XVI.

Table XVI. — Size of different species of Phoma occurring on Solanum tuberosum

Species.

Size of pycnidia.

Size of spore.

Phoma nebulosa .

135-145x110-115
25° (f'k mm-”)

Minute.

7- 5-3
4-1/?

Phoma eupyrena .

Phoma solani .

7-8 X lK-2
6-7X4

3. 78-6. 10 X I. 8-3. 7

Phoma so lanico la .

Minute.

Phoma sp. under consideration. .

90-260 X 80-160

Unfortunately it was impossible to obtain authentic material of the
four species described. The original descriptions are too meager to enable
one to identify a species with any degree of accuracy, but they show that
in size of the spores the species differ markedly from the Phoma sp. under

Oct. 30, 1916

Spongospora subterranea and Phoma tuberosa

251

consideration. These meager descriptions and difference in size of spores,
coupled with the pathogenicity of this fungus on the tuber, led the
writers to designate the organism “ Phoma tuberosa , n. sp.”

Phoma tuberosa, n. sp.

Lesions on tubers of Solanum tuberosum; brownish to dark gray or black; 6 to 2 5 mm.
in diameter; sunken, membraneous, with an irregular and sharply defined margin.
Pycnidia black, generally scattered over entire surface, subcuticular, irregular,
subglobose to spherical, majority provided with a single well-defined ostiole, some¬
times breaking at several points for exudation of spore mass, varying in size from 80 to
160 by 90 to i6oju. When placed in water the pycnospores are seen to ooze out in a
shiny string, which soon breaks up into the individual spores. Pycnospores i-celled,
hyalin, subglobose, 3.7 to 6 by 1.8 to 3.7 ji. Hyphae septate, dark brown in the tissues
of the host. A definite stroma absent.

Habitat. — Wound parasite on tubers of Solanum tuberosum often associated with the
sori of Spongospora subterranea . First found as a storage-rot in Maine .

SUMMARY

(1) Spongospora subterranea exists in six different potato-growing
sections of the United States, all northern except one.

(2) No infections resulted on the progeny of powdery-scab-infected
seed potatoes planted in 15 different localities along the Atlantic seaboard.
However, 8 lots of soil out of 12 shipped from as many of these localities
to northern Maine and planted with infected seed produced a crop showing
powdery-scab.

(3) Periods of damp, rainy, and cloudy weather, coupled with poor
drainage, favor the development of S. subterranea.

(4) Infection develops earlier on the roots than on the tubers. In 1915
jn northern Maine 57 days elapsed between planting of infected tubers in
virgin soil and the first signs of root infection. Infection on the tubers
appears about the stem end first. All underground portions of the
potato plant may become infected with S. subterranea . Galls are often
very numerous on the root system of potato plants growing in infected
soil, while the tubers are absolutely free from infection; hence, a clean
root system is the criterion for determining the absence of the disease.

(5) It is not unusual to find parts of fields in northern Maine in which
some of the progeny of over 90 per cent of the hills are infected with
powdery-scab. Several cases were found in which from 50 to 75 per cent
of the 1914 crop was infected. It was found that S. subterranea may be
spread when infected tubers are planted in virgin soil. Cultural practices
and soil water are probably the most important agents in spreading the
disease.

(6) Besides the potato, there are seven other solanaceous hosts of
S. subterranea , including the tomato, as determined by the writers.
The disease manifests itself on these hosts in the form of large destructive
galls on the roots, these being fully as injurious as those on the potato

252

Journal of Agricultural Research

Vol. VII, No. s

plant. The histology of the galls on all the hosts is very similar and has
many points in common with Plasmodiophora brassiceae on cabbage.

(7) The absence of the canker stage of S. subterranea in the United
States may be due to the short growing period afforded the potato crop
in infected districts.

(8) Flea-beetle injury, intumescence associated with the lenticels, and
certain forms of common scab on the tuber are often mistaken for stages
of S. subterranea .

(9) Among the saprophytic fungi found associated with the sori of
S. subterranea is a species of Papulospora. The “bulbils” of the latter
are strikingly similar to the spore balls of the former fungus, and this
similarity may account for the confusion in earlier writings as to its
identity on the potato tuber.

(10) A study of early harvesting, seed treatment, varietal response,
and soil treatment as control measures for the disease was made. This
suggests that (a) early harvesting may be beneficial certain seasons in
Maine, but can not be relied on every year; (b) seed treatment with
certain chemicals will reduce the disease, this being especially true of
mercuric chlorid and formaldehyde, the hot solutions for short periods
being probably as efficient as the cold for longer periods; (c) certain
varieties may escape infection; this may be due not to disease resistance
but to differences in development at the time infection is most likely to
take place; (d) the possibility of finding a resistant variety has not yet
been exhausted; (e) no soil treatment will eradicate the disease, but
sulphur at the rate of 900 pounds per acre applied broadcast reduces the
amount of infection by S. subterranea .

(11) Several types of dryrot follow 5. subterranea . These, designated
according to cause, are desiccation, plasmodium, and wound-parasite
injury. The percentage of these secondary rots as found in nature
in infected tubers varied from 30 to 73.

(12) There is a close relation between certain soil types and the
development of fungus. (See fig. 1 .) From the type of soil and its drain¬
age it is possible to predict what the development of the disease will be
in any particular field.

(13) The dryrot due to a species of Phoma and other wound parasites
is the most serious of the rots. After a comparison with earlier descrip¬
tions of various species of Phoma on the potato, the writers designated
the species here discussed as “Phoma tuber osa9 n. sp.”

253

Oct. 30, 1916

Spongospora subterranea and Phoma tuber osa

literature cited

(1) Berkeley, M. J. %

1846. Observations, botanical and physiological, on the potato murrain. In
Jour. Hort. Soc., London, v. 1, p. 9-34, 2 fig.

(2) Eidam, Eduard.

1883. Zur Kenntniss der Entwicklung bei den Ascomyceten. In Beitr. Biol.
Pflanzen., Bd. 3, p. 377-433* P1- I9“23*

(3) Horne, A. S.

1911. On tumour and canker in potato. In Jour. Roy. Hort. Soc. [London],

v. 37, pt. 2, p. 362-389, fig. 96-106.

(4) Hotson, J. W.

1912. Culture studies of fungi producing bulbils and similar propagative bodies.

In Proc. Amer. Acad. Arts and Sci., v. 48, no. 8, p. 227-306, 12 pi.
Literature cited, p. 303-306.

(5) Johnson, Thomas.

1909. Further observations on powdery-scab, Spongospora subterranea (Wallr.).
In Sci. Proc. Roy. Dublin Soc., n. s. v. 12, no. 16, p. 165-174, pi. 12-14.

(6) KunkEl, L. O.

1915. A contribution to the life history of Spongospora subterranea. In Jour.
Agr. Research, v. 4, no. 3, p. 265-278, pi. 39-43-

(7) Melhus, I. E.

1914. A phoma rot of Irish potatoes. (Abstract.) In Phytopathology, v. 4>
no. 1, p. 41.

(8) -

1914. Powdery scab (Spongospora subterranea) of potatoes. U. S. Dept. Agr.
Bui. 82. 16 p., 3 pi. Bibliography, p. 15-16.

(9) Morse, W. J.

1914. Powdery scab of potatoes. Maine Agr. Exp. Sta. Bui. 227, p. 87-104,
fig- 44“52*

(10) Nawaschin, S.

1899. Beobachtungen fiber den feineren Bau und Umwaldlungen von Plas-
modiophora Brassicae, Woron. im Laufe ihres intracellelaren Lebens.
In Flora, Bd. 86, Heft 5, p. 404-427* Pl- 20 (col.).

(11) Osborn, T. G. B.

1911. Spongospora subterranea (Wallroth) Johnson. In Ann. Bot., v. 25,

no. 98, p. 327-341, pl. 27.

(12) PethybridgE, G. H.

1912. Investigations of potato diseases. In Dept. Agr. and Tech. Instr.

Ireland Jour., v. 12, no. 12, p. 334-36°* 5 fig-

(13) Prillieux, E. E., and Delacroix, Georges.

1890. Sur quelques champignons parasites nouveaux. In Bui. Soc. Mycol.
France, t. 6, fasc. 3, p. 178-180.

(14) Saccardo, P. A.

1879-92. Sylloge Fungorum ... v. 1, 1879; v. 3, 1884; v. 10, 1892. Patavii.

(15) Wallroth, F. W.

1842. Die Naturgeschichte der Erysibe subterranea Wallr. In his Beitrage
zur Botanik, Bd. 1, p. 118-123, pl. 2, fig. 12-15. Leipzig.

(16) WOLLENWEBER, H. W.

1913. Studies on the Fusarium problem. In Phytopathology, v. 3, no. 1,

p. 24-59, 1 fig., pl. 5. Literature, p. 46-48.

(17) WORONIN, M. S.

1878. Plasmodiophora brassicae, Urheber der Kohlpflanzen-Hemie. In Jahrb.
Wiss. Bot. [Pringsheim], Bd. 11, p. 548-574, pl. 29-34.

PLATE A

Spongospora subterranea and Phoma tuberosa on Solatium tuberosum :

Fig. 1-5. — Spongospora subterranea as found on different varietiesof the Irish potato.
Figure 2 shows the sori on the Green Mountain variety. The sori shown in figures 1,
3,4; and 5 are from seedling varieties not yet distributed.

Fig. 6, 7. — Stages in the development of dryrot caused by Phoma tuberosa. Figure
6 shows a very early stage and figure 7 a stage commonly found in April in storage
houses.

(254)

PLATE 7

Spongospora subterranea on Solatium tuberosum:

Fig. A. — Stem of a potato showing formation of a gall caused by Spongospora sub¬
terranea . Note the discoloration and shrinkage near the point of formation . Although
sometimes present, galls rarely occur on the stems.

Fig. B. — Part of a stolon showing galls caused by Spongospora subterranea.

Fig. C. — Discoloration so often found on the root near the point where the galls
form.

Fig. D. — Spongospora subterranea as found on the root system of the potato. The
galls resemble nematode root galls or nitrogen nodules. Generally these galls appear
on the roots before any sign of the disease is seen on the tuber.

PLATE 8

Spongospora subterranea in the roots of various hosts (fixed in Flemming ’s solution

and stained with triple stain) :

Fig. A. — Section through a potato root affected with Spongospora subterranea . The
portion to the left is healthy, while that to the right shows the increase in number
and size of the host cells. Amebae in the cells are very numerous. X no.

Fig. B. — Several cells from Solanum warscewiczii , showing the formation of “giant
cells * ' and their division into daughter cells. Note the tendency of the amebae to
cluster around the host nucleus. X 2,600.

Fig. C. — Section through a tomato root ( Lycoperstcon esculentum , showing effects of
infection by Spongospora subterranea. Note the abnormal increase in number and
size of the cells. The parasite is confined to the cortex, being entirely absent from
the xylem. The amebae are numerous and generally clustered around the host
nucleus. X 325.

PLATE 9

Spongospora subterranea on the roots of various hosts:

Fig. A. — Galls caused by Spongospora subterranea on the roots of Solanum warsce -
* wiczii . In this plant the galls have a tendency to girdle the root.

Fig. B, C. — Galls caused by Spongospora subterranea formed on the roots of the
tomato. These galls, owing to the age and size of the plant, show a variety of forms.

No. 5

PLATE io

Injuries caused by Spongospora subterranea and other agencies:

Fig. A.— Tuber showing the effect of flea-beetle injury. Note the similarity of this
injury and the young sori of Spongospora subterranea. After being placed in storage
a shrinkage occurs around the sori-like injuries similar to the desiccation injury around
the sori of Spongospora subterranea.

Fig. B. — Tuber showing a very early stage of infection by Spongospora subterranea.
The points of infection show brown, fimbriate colonies. This condition is found in
potatoes at time of harvesting. Spore balls are absent.

Fig. C, D. — Tubers grown in infected soil in the greenhouse under exceptionally
moist conditions and allowed a long growing season. These lesions resemble very
materially the cankerous stage.

Fig. E- — A potato from Ireland showing the cankerous stage . The powdery material
is almost gone from the tuber, which appears to have been eaten by insects.

Fig. F. — A tuber showing enlargement of the lenticels. This tuber was kept in a
moist chamber for io days. Similar enlargements are often found in Florida-grown
potatoes, and have also been seen in Delaware and New Jersey. Such conditions are
often mistaken for the early stages of infection caused by Spongospora subterranea.

55859°— 16 - i

PLATE ii

Dryrots associated with Spongospora subterranea:

Fig. A. — A potato tuber showing natural infection with Phoma sp . The color of the
spot is much darker than that of the healthy tissues. No pycnidia are present as yet.

Fig. B, C. — Sections through tubers showing more advanced stages of a rot caused
by a species of Phoma. A variety of other fungi have entered, causing further decay
of the tuber.

Fig. D. — A potato tuber showing injury immediately around the sori, due partially
to the work of the plasmodium. The lower side of the tuber also shows the beginning
of the rot caused by Phoma sp.

Fig. E- — Infection due to Phoma sp. on a potato tuber infected with Spongospora
subterranea , followed by another, due probably to Fusarium coeruleum.

Fig. F, H. — Potato tubers infected with Spongospora subterranea about three weeks
after harvesting, showing the effects of desiccation injury.

Fig. G.^Section through a tuber, showing the depth to which rot caused by Phoma
sp. extends. In this case no other fungi had entered.

PLATE 12

Spongospora subterranea and Phoma tuberosa:

Fig. A. — Section of a potato tuber through a sorus around which no dryrot has
as yet set in. Spore balls are numerous and cork cells absent at the base of the
sorus. Xno.

Fig. B. — Section of a potato tuber made through a sorus of Spongospora subterranea
after the tuber had been held in storage and some dryrot due to desiccation had devel¬
oped. A tendency toward the laying down of cork at the base is shown, but the
cork cells have not yet thickened. X i io.

Fig. C, D. — Two views of the pycnidia of Phoma tuberosa as grown in pure culture.
Figure C shows the pycnospores emerging from the ill-defined ostiole. X650.

Fig. E. — Pycnospores. X2,6oo.

Fig. F. — Mature “bulbils” of Papulospora coprophila (Zukal) Hotson, which in the
tissues of potato tubers may be mistaken for spore balls of Spongospora subterranea .
X650.

Fig. G. — Spores of fungi associated with Spongospora subterranea and referred to
Verticillium sp. and Stysanus sp. by Horne, of whose drawing this figure is a reproduc¬
tion. Note the similarity to those shown in figure F.

PLATE 13

Phoma tuberosa on Solatium tuberosum:

Fig. A, B. — Stages of the rot caused by Phoma tuberosa on the Irish potato. The
size and appearance of the lesion as shown in figure B is the most common. The
abundance and size of the pycnidia vary.

Fig. C, D, E. — Results of artificial inoculation with pure cultures of Phoma tuberosa.
Figure D shows an injured uninoculated tuber. Figure E shows the buttonhole lesion
produced 10 days after inoculation. Figure C shows a tuber from another series three
weeks after inoculation. Note the similarity of this lesion and that shown in Plate
12, figure A.

PLATE 14

Scab caused by Phoma tuberosa and Oospora scabies on Solatium tuberosum:

Fig. A.— Section through a tuber affected with common scab. The portion to the
right is healthy tissue. Note the thickened cortex cells formed immediately below
the epidermis. These cells prevent the desiccation of the tissues about the scab
sorus. Generally such a condition is absent in tubers affected with scab caused by
Spongospora subterranea. X325.

Fig. B. — Section through a tuber affected with the rot caused by Phoma tuberosa.
Note the mat of mycelium immediately below the epidermal cells. It is this mat
that gives the lesion its hard, bony texture. The cells below the mat are broken
down . It is in this cavity that other fungi begin their work . X 32 5 .

GROWTH OF PARASITIC FUNGI IN CONCENTRATED

SOLUTIONS

By Lon A. Hawkins,

Plant Physiologist , Drug-Plantf Poisonous-Plant , Physiological , and Fermentation
Investigations , Bureau of Plant Industry

The mycelium of a fungus growing parasitically is frequently in con¬
tact with the cell sap of its host plant. This cell sap is capable of an
osmotic pressure which would vary, of course, with the amount and nature
of the compounds in solution. The hyphae of the invading parasite, then,
may grow in a medium which has more or less high osmotic pressure. In
order to grow in this medium, they must, of course, be able to withdraw
water from it. It is apparent, then, that plant parasites must quite
commonly possess the ability to withdraw water from more or less highly
concentrated solutions and to grow in them.

The relative concentration of the cell sap of parasite and host has
received some attention in the case of phanerogamic parasites.
MacDougal (9, io)1 and MacDougal and Cannon (11) reached the con¬
clusion that for their “Xeno parasites' ’ the osmotic pressure of the
parasite must be greater than the plant into which it is transplanted.
Senn (14), in a recent investigation, has shown for certain phanerogamic
parasites that the osmotic pressure of the parasite, as measured by the
plasmolytic method, is invariably higher than that of its host. This
writer seems not to have seen the work of MacDougal. Not so much
attention has been paid to fungus parasites, though some work has been
done in growing fungi in concentrated solutions. Eschenhagen (4)
grew Aspergillus niger, Penicillium glaucum , and Botrytis cinerea on
rather highly concentrated solutions of various substances. He found
that these fungi grew in a saturated solution of potassium nitrate at
ordinary temperatures and that the concentration of glucose which lim¬
ited growth was above 50 per cent. He came to the conclusion that the
ability of the fungi to live and grow when transferred to a higher con¬
centration was due to a heightened osmotic pressure within the cell
produced by an actual increase in the osmotically active substance
therein.

Raciborski (13) also has shown that some fungi can live in exceedingly
concentrated solutions. He grew Torula sp. in a saturated solution of
lithium chlorid and Aspergillus glaucus in a similar solution of sodium
chlorid. He considered the osmotic pressure in the cell to be greater
than that of the outer medium and attempted to calculate the molecular

1 Reference is made by number to “ literature cited,” pp. 259-260.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.
fw

(255)

Vol. VII, No. 5
Oct. 30, 1916
G— 97

256

Journal of Agricultural Research

Vol. VII, No. s

weight of a carbohydrate that would produce the necessary osmotic
pressure. The present writer (7) germinated the conidia of Glomerella
cingulata and grew the fungus in concentrations of calcium nitrate, potas¬
sium nitrate, and sucrose, which had diffusion tensions of 29.1, 39.3,
and 47.3 atmospheres, respectively.

Dorn (3), in his study on the penetration of plant membranes by fun¬
gus hyphae, mentions that there is an osmotic pressure of about 50 atmos¬
pheres in fungus hyphae.

It seems probable from a consideration of the question and from the
work that has been done that fungus parasites should be able to live
and grow in solutions of a considerably higher concentration than the
total concentration of the cell sap of their host plants. It was considered
worth while, however, to obtain more evidence on this point. Ten com¬
mon parasitic fungi were grown in solutions of salts and sugars of rather
high concentrations. The total diffusion tensions of the dissolved mate¬
rials in the expressed juice of some of their host plants was determined
by the freezing-point method. The present paper deals with this work.

The fungi 1 studied were Fusarium radicicola Wollenw. and F. oxy -
sporum Schlecht., two potato-rotting fungi (2); Plenodomus destruens
Harter, Diplodia tubericola (E. and E.) Taub., Sphaeronema fimbriatum
(E. and H.) Sacc., and Rhizopus nigricans Ehrenb., which are parasitic
upon sweet potato (6), and Botrytis cinerea Pers., Sclerotinia cinerea
(Bon.) Schroter, and Sphaeropsis malorum Peck, well-known parasites
on the apple fruit. A strain of Rhizopus nigricans Ehrenb., which causes
a serious rot of the strawberry (16), was also used. The data on the dif¬
fusion tension of the solutions in which Glomerella cingulata (Stonem.)
S. and v. S. was grown are taken from an earlier paper by the present
writer. Several of these fungi are parasitic on more than one of the hosts
mentioned. The lowering of the freezing point of the expressed juice
of apples (Malus sylvestris ), sweet potatoes ( Ipomoea batatas) , potatoes
(Solanum tuberosum ), and strawberries (Fragaria spp.) was determined.

In the experiments with the fungi two methods were followed for the
determination of the highest concentrations of the various substances
used in which the fungi could grow. In the one method hanging-drop
cultures were made of the spores which had been sown in salt or sugar
solution of varying concentrations. These cultures were examined by
means of a miscroscope, and the highest concentration in which the
germination was apparent was noted. The other method was similar in
principle to the one just outlined. In this the procedure was to sow the
spores in sterilized tubes of the salt (sodium chlorid) or sugar solution and
after about a week determine the growth or lack of growth by observation.

1 The writer’s thanks are due Mr. C. W. Carpenter for cultures of the species of Fusarium, Mr. L. L.
Harter for cultures of the fungi from sweet potato, Dr. J. S. Cooley for the apple-rot fungi, and Dr. Neil E.
Stevens for cultures of the species of Rhizopus from strawberry.

Oct 30, 1916 Growth of Parasitic Fungi in Concentrated Solutions 257

The spores were germinated and the fungi grown in solutions of cal¬
cium nitrate, potassium nitrate, sucrose, and glucose. Concentrated so¬
lutions were made of these substances, which were diluted down to the
concentration desired in the experiments by the addition of distilled
water to which a very little potato extract had been added.

The weight normal method of Morse and Frazer (12) was followed in
making up the solutions.

The diffusion tension of the highest concentrations used in which the
spores germinated and grew was calculated for the various compounds,
taking into account, of course, the ionization of the salts. The data for
calculating the percentage of ionization of the two salts were obtained
from Jones's tables (8). The calculations of the percentage of ionization
give probably only approximate values as the calculations are based on
interpolations in most cases, and, moreover, some other substances were
present which might influence the dissociation of the salts. The data thus
obtained, however, probably offer a better basis for the comparison of the
diffusion tensions of the two electrolytes and the two nonelectrolytes used
in this study than the molecular concentrations given in the adjoining col¬
umns. The diffusion tension of the cane-sugar solutions was calculated
from Morse and Frazer's determination of the osmotic pressure of a molec¬
ular solution of this substance at 250 C. From their work and from the
determinations of Berkeley and Hartley (1) it seems quite probable that
the values given are too low. The results obtained in growing fungi in
the concentrated solutions of salts and sugars are given in Table I.

Tabi^E I. — Highest concentrations (molecular) of calcium nitratef potassium nitrate ,
sucrose , and glucose in which the fungi grew and the calculated diffusion tensions in
atmospheres of these solutions

Fungus.

Glucose,

Sucrose.

Potassium

nitrate.

Calcium nitrate.

Con¬

centra¬

tion

(molec¬

ular).

Diffu¬

sion

tension

(atmos¬

pheres).

Con¬

centra¬

tion

(molec¬

ular).

Diffu¬

sion

tension

(atmos¬

pheres).

Con¬

centra¬

tion

(molec¬

ular).

Diffu¬

sion

tension

(atmos¬

pheres).

Con¬

centra¬

tion

(molec¬

ular).

Diffu¬

sion

tension

(atmos¬

pheres).

Fusarium radicicola .

1. 6

038.9

1.8

47-4

1. 6

54- 5

O. 6

27. 7

Fusarium oxysporum .

I. 6

038. 9

1.8

47-4

1. 6

54-5

27.7

Plenodomus destruens .

2.4

“58-3

1.8

47-4

1. 6

54- 5

* 7

33*6

Sphaeronema fimbriatum . . . .

2. 6

a 63. 2

1.8

47-4

1. 6

54*5

•4

19*5

Diplodia tubericola .

2. 6

a 63. 2

1. 6

42. 1

1. 8

58.8

• 7

33*6

Rhizopus nigricans (from

strawberry) .

1. 6

a 63. 2

27*5

15*9

Rhizopus nigricans (from

sweet potato) .

1. 6

a 63. 2

1. 6

42. 1

27*5

•3

15* 9

Botrytis cinerea .

2. 6

a 63. 2

1.8

47*4

1. 6

54*5

27.7

Sclerotinia cinerea .

a KS.

I. A

47. 6

. 6

27. 7

Sphaeropsis malorum .

2. 6

a 63. 2

1.8

47-4

i. 6

54-5

27.7

Glomerella cingulata .

4i.3

“39-3

29. I

a No higher concentrations used.

Journal of Agricultural Research

Vol. VII, No. s

258

The freezing points of the expressed juice of apples, sweet potatoes,
potatoes, and strawberries were determined by means of a Beckman
freezing-point apparatus, and the total diffusion tension for the material
in these solutions was calculated from the lowering of the freezing
point (15).

The results obtained by the calculation of these data are given in
Table II.

Table) II. — Diffusion tension in atmospheres of the juice of certain hosts of the fungi
studied , as calculated from the lowering of the freezing point

Host.

Diffusion
tension |
(atmospheres.)

Strawberry .

8. 27
17* 85
10. 25
6. 52

Apple (Blacktwig) .

Sweet potato (Jersey Big Stem) .

Potato (Irish Cobbler) .

From Table I it is evident that the fungi used in these experiments
are able to grow in relatively high concentrations of salts and sugars.
The highest diffusion tension of any solution in which growth was
evident was in the concentrated solutions of glucose. Growth occurred
in all concentrations of this sugar used.

With potassium nitrate growth was inhibited in all cases when the
diffusion tension of the solution was about 59 atmospheres; and in only
one case was a fungus able to grow in a solution so concentrated. All
the fungi except the two strains of Rhizopus nigricans and Glomerella
cingulata grew in concentrations which have a calculated diffusion ten¬
sion of 47 atmospheres.

The fungi also grew in solutions of sucrose of rather high concentra¬
tions, the two strains of Rhizopus nigricans and Diplodia tuhericola being
the only fungi which were unable to grow in sucrose solutions having a
diffusion tension of 40 atmospheres. It is noticeable that growth in the
case of R, nigricans is inhibited also at lower concentrations of potassium
and calcium nitrates than with the other fungi.

Calcium nitrate inhibited germination and growth always at concen¬
trations considerably lower than those required to produce the same
effect with sucrose, glucose, and potassium nitrate. Nevertheless a
comparison of the diffusion tensions of the highest concentrations of
calcium-nitrate solutions in which the fungi grew and the diffusion ten¬
sion of the juice of the host plant as calculated from the lowering of the
freezing points shows that the parasite is in all cases able to grow in
considerably higher concentrations than are present in the cell sap of
its host plant. Whether or not the protoplasm of the fungi used is
impermeable to the two salts and two sugars used in these experiments

oct 30, 1916 Growth of Parasitic Fungi in Concentrated Solutions

259

is a question which can not be answered from the data now at hand.
Potassium nitrate has been used extensively as a plasmolyzing agent for
determining osmotic pressure in plants.

Fischer (5) reaches the conclusion, however, that the protoplasts of
bacteria are readily permeable to this salt. True (17) used cane sugar in
his work on species of Spirogyra, and it has been used to a considerable
extent by other investigators. Calcium nitrate is not commonly used
as a plasmolyzing agent.

From the concentrations of glucose which were found to be favorable to
the growth of fungi it seems probable that this substance penetrated the
protoplasm, perhaps quite readily. It is, of course, probable that all
the substances penetrate into the cell to some extent. It is possible that
the protoplasm of the fungi here used is readily permeable to the salts
and sugars employed in this study and that they can pass into the cell
until the concentration within is the same as without. Then the presence
within the cell of some substance to which the protoplasm was imperme¬
able would raise the osmotic pressure within above that of the surround¬
ing solution, and the fungus could grow.

The importance of this ability of parasitic fungi to grow in solutions
which are capable of exerting a high osmotic pressure is evident.
Whether this ability is due to the osmotic pressure in the fungus being
originally higher or whether it becomes higher through a diffusion of
substance into the hyphae, or whether there is an actual increase in the
osmotically active substances within the cell, as Eschenhagen concludes,
or whether other factors than a high osmotic pressure enable fungi to
remain turgid and grow in concentrated solutions are questions which
need further investigation.

In these experiments in which fungi were grown in solutions of potas¬
sium and calcium nitrate, sucrose, and glucose it was found that in every
case the fungi grew readily in solutions in which the diffusion tensions
were much higher than the total diffusion tensions of the dissolved
substances in the juices of their host plants.

literature CITED

(1) Berkeley, Earl op, and Hartley, E. C. J.

1906. On the osmotic pressures of some concentrated aqueous solutions. In
Phil. Trans. Roy. Soc. London, s. A, v. 206, p. 481-507, 3 diagr.

(2) Carpenter, C. W.

1915. Some potato tuber-rots caused by species of Fusarium. In Jour. Agr.
Research, v. 5, no. 5, p. 183-210, pi. A-B (col.), 14-19. Literature
cited, p. 208-209.

(3) Dorn, O.

1914. Beitrage zur Kenntnis von der Durchbohrung pflanzlicher Membranen
durch Pilzhyphen. 48 p. Leipzig. Inaug. Diss. Abstract in Zentbl.
Biochem. u. Biophys., Bd. 18, No. 1/2, p. 9-10. 1915. Original not

seen.

26o

Journal of Agricultural Research

Vol. VII, No. s

(4) Eschenhagen, Franz.

1889. Uber den Einfluss von Ldsungen verschiedener Concentration auf das
Wachstum von Schimmelpilzen. 56 p. Stolp. Inaug. Diss. Leipzig.

(5) Fischer, Alfred.

1895. Untersuchungen fiber Bakterien. In Jahrb. Wiss. Bot., Bd. 27, Heft 1,
p. 1-163, pl- I_5-

(6) Harter, L. L.

1916. Sweet-potato diseases. U. S. Dept. Agr. Farmers' Bui. 714, 26 p., 21 fig.

(7) Hawkins, L. A.

1913. The influence of calcium, magnesium, and potassium upon the toxicity
of certain heavy metals toward fungus spores. In Physiol. Researches,
v. 1, no. 2, p. 57-92, 6 fig- Literature cited, 91-92.

(8) Jones, H. C., CeovER, A. M., and others.

1912. The electrical conductivity, dissociation, and temperature coefficients of

conductivity from zero to sixty-five degrees, of aqueous solutions of a
number of salts and organic acids. 148 p., illus., tab. Washington,
D. C. (Carnegie Inst. Washington Pub. 170.)

(9) Macdougae, D. T.

1911. An attempted analysis of parasitism. In Bot. Gaz., v. 52, no. 4, p.
249-260, 6 fig.

(10) -

1911. Induced and occasional parasitism. In Bui. Torrey Bot. Club., v. 38,
no. 10, p. 473-480, pl. 23-25.

(n) - and Cannon, W. A.

1910. The condition of parasitism in plants. 60 p., illus., 10 pl. (partly col.).
Washington, D. C. (Carnegie Inst. Washington Pub. 129.)

(12) Morse, H. N., and Frazer, J. C. W.

1905. The osmotic pressure and freezing points of solutions of cane-sugar. In
Amer. Chem. Jour., v. 34, no. 1, p. 1-99, 5 fig.

(13) Raciborski, Maryan.

1905. Pr6ba okre&enia g6mej granicy cisnienia osmotycznego umozliwiaj^cgo
zycie. (Uber die obere Grenze des osmotischen Druckes der lebenden
Zelle.) [In German.] In Bui. Intemat. Acad. Sci. Cracovie Cl. Sci.
Math. et. Nat., 1905, no. 7, p. 461-471.

(14) Senn, Gustave.

1913. Der osmotische Druck einiger Epiphyten und Parasiten. In Verhandl.

Naturf. Gesell. Basel, Bd. 24, p. 179-183.

(15) Skive, J. W.

1914. The freezing points of Tottingham's nutrient solutions. In Plant World,

v. 17, no. 12, p. 345-353-

(16) Stevens, N. E.

1916. Pathological histology of strawberries affected by species of Botrytis and
Rhizopus. In Jour. Agr. Research, v. 6, no. 10, p. 361-366, pl. 49-50.

(17) True, R. H.

1898. The physiological action of certain plasmolyzing agents. In Bot. Gaz.,
v. 26, no. 6, p. 407-415.

JOURNAL OF AGRMIim RESEARCH

DEPARTMENT OF AGRICULTURE

Von. VII Washington, D. C., November 6, 1916 No. 6

FREEZING-POINT LOWERING OF THE LEAF SAP OF THE
HORTICULTURAL TYPES OF PERSEA AMERICANA

By J. Arthur Harris, Investigator, Station for Experimental Evolution, and Wilson
PopEnoE, Agricultural Explorer, Bureau of Plant Industry

INTRODUCTION

In the introduction of tropical economic plants into the warmer por¬
tions of the United States, which for the most part are not free from at
least occasional frosts, ability to survive transient low temperature is a
characteristic of such fundamental importance that any quantitative
contribution to our knowledge of its physiology must be of value. Among
the factors to which frost resistance in plants is due, the magnitude of
the depression of the freezing point of the cell sap has been suggested as
one of importance.

That the freezing-point lowering of the cell sap is not in itself a suffi¬
cient explanation of differences in frost resistances should be quite
obvious from the narrowness of its range in plants of economic importance.
To what degree the osmotic pressure of the sap may be correlated with
other characteristics which are of significance in cold resistance remains
to be investigated.

A discussion of the several possible factors or references to the very
extensive literature of the subject fall quite outside the scope of a note
which is designed only to present a series of constants pertinent to the
problem of hardiness in a particular tropical fruit.

The Plant Introduction Field Station at Miami, Fla., has seemed to
the writers to afford particularly advantageous materials for a test of
the existence of a relationship between capacity for cold resistance and
the freezing-point lowering of the extracted sap, since it contains,
assembled under the same environmental conditions, a wider range of
varieties than can be found elsewhere.

The avocado, Per sea americana Miller (Per sea gratissima Gaertn. f.),
while introduced to Florida and California a good many years ago, has
only been propagated asexually since the beginning of the present
century. Hence, the number of horticultural varieties is not over¬
whelmingly large. Some of these varieties have been under observation

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C

(261)

Vol. VII, No. 6
Nov. 6, 1916
G-98

26 2

Journal of Agricultural Research

Vol. VII, No. 6

by horticulturists for a period of 5 to 15 years, and records as to compara¬
tive hardiness have been obtained. These records show that the horti¬
cultural types differ significantly in regard to the amount of frost to
which they can be subjected without injury.

The cultivated avocados fall naturally into three groups, or types,
each of which possesses distinguishing characteristics which are quite
constant, though occasional forms are seen which appear to be interme¬
diate or to belong to distinct groups not yet well known in cultivation.
The characteristics of the three cultivated types are briefly as follows:

Mexican Type. — Foliage and sometimes the fruit distinctly anise-
scented. Leaves usually smaller than those of the Guatemalan and
West Indian types. Fruit commonly 4 to 8 ounces in weight; skin
thin, often membranous, usually smooth and glossy on the surface.
Seed coats thin; closely united and adhering to the cotyledons, or
separating, as in the West Indian type. Flowers heavily pubescent,
appearing in early spring, from January to March in California and
Florida. Ripe fruits from June to October. Occasionally a second
crop, from later bloom, ripens in winter and spring.

Guatemalan type. — Foliage not anise-scented, deep green, the new
growth bronze-red, commonly, but not always, deeper colored than that
of the West Indian type. Fruits commonly from 12 to 18 ounces in
weight; skin usually verrucose or tuberculate on the surface, one-six¬
teenth to three-sixteenths of an inch thick, woody, brittle, and coarsely
granular in texture, sharply differentiated from the flesh. Seed, as a
rule, comparatively small; cotyledons smooth; the seed coats thin,
closely united, and adhering to the cotyledons throughout. Flowers
more finely pubescent than in the Mexican type, appearing in late
spring — March to May in Florida. Fruits matured in the winter or
spring of the following year.1

West Indian type. — Foliage not anise-scented; generally similar to
the Guatemalan, but the young branchlets and leaves often lighter in
color. Fruits variable in form and size, as in the other types, compara¬
tively large, averaging 14 to 20 ounces, but sometimes weighing 3 pounds
or more; surface nearly always smooth, the skin rarely more than one-
sixteenth of an inch thick, pliable and leathery, and scarcely so well
differentiated from the flesh as in the Guatemalan. Seed usually large
in proportion to the size of the fruit; cotyledons more or less rough;
the two seed coats frequently thick and separated, at least over the
distal end of the seed, one adhering to the cotyledons and the other loose
or adhering to the wall of the seed cavity. Flowers usually as pubescent
as those of the Guatemalan type, occasionally glabrate; in Florida and
the West Indies they appear from February to April. Matured fruits
from July to November.

1 Some of the varieties of this type have the skin nearly smooth and about the same thickness as that of
the West Indian. For this reason they have been considered by a few horticulturists as forming a distinct
class; but inasmuch as they seem to differ only in these two points, which are both variable characters
at best, it seems safe to retain them in the Guatemalan type.

Nov. 6, 1916 Freezing Point of Leaf Sap of Per sea americana 263

EXPERIMENTAL METHODS

The technique employed in determining freezing-point lowering was
very simple. Fully matured but still normally green leaves were col¬
lected and frozen in an ice and salt mixture to facilitate the extraction
of sap by pressing1 in a small heavily tinned bowl under a powerful
screw. The freezing-point lowering was determined by mesons of a
mercury thermometer graduated to 0.01 of a degree centigrade in subdi¬
visions sufficiently large to allow of fairly accurate estimation to smaller
fractions. The vaporization of ether or carbon bisulphid in a vacuum
jacket was used in determining the freezing-point lowering. The results
are expressed in terms of freezing-point lowering in degrees centigrade (A),
corrected for undercooling by the usual formula.

For those who prefer to think in terms of osmotic pressure the values
of P are given from a published table.2 Finally, for the convenience of
those who wish to know the actual freezing point of the saps on the
Fahrenheit scale, these values have been added (F).

PRESENTATION OF CONSTANTS FOR THE THREE TYPES3

Mexican type. — Throughout the highlands of central and northern
Mexico this type is very common. . Because of its superior hardiness, it
has been extensively planted in several subtropical regions, most notably
in California and Chile. It is known to have been planted in California
as early as 1870. In Florida it has fruited as far north as Gainesville,
but is not generally cultivated in any part of the State. It has been
planted along the Riviera in southern Europe and has fruited in the open
at Rome. In Algeria it has also been planted, though to a very limited
extent. In the West Indies it is scarcely known.

During the cold weather of January, 1913, in California trees of this
type were reported to have withstood temperatures of 160 to 20° F.
without injury.

1. Quer6taro . . . A=i.ii, P= 13.3, F=3o.oo°

Buds received from California, where the variety was introduced from Quer6taro,

Mexico, in 1911. Cion 1 year old, on one branch of No. 4, Mexican stock.

2. San Sebastian . . . A=i.2o, P=i4.5, 29.84°

Buds received from California, where the variety was introduced from Quer£taro,

Mexico in 1911. One-year-old cion on Mexican stock, a branch of No. 4.

3. Harman . . . A=i.24, P=i5.o, F= 29.77°

Originated at Sherman, Cal., budded on a limb of 26713, West Indian type. Cion

2 years old.

4. Seedling . A-1.27, P=i5.2, F= 29.710

Grown from seed of unknown origin.

1 Gortner, R. A., and Harris, J. A. Notes on the technique of the determination of the depression of the
freezing point of vegetable saps. In Plant World, v. 17, No. 2, p. 49-53, 1914.

2 Harris, J. A., and Gortner, R. A. Notes on the calculation of the osmotic pressure of expressed vege¬
table saps from the depression of the freezing point, with a table for the values of P for A=o.ooi * to A-2.9990.
In Amer. Jour. Bot., v. 1, no. 2, p. 75-78. 1914.

3 The numbers following the varietal names refer to the Inventory of the Office of Foreign Seed and Plant
Introduction, Bureau of Plant Industry.

264

Journal of Agricultural Research

Vol. VII, No. 6

5. Harman . A=i.29, P— 15,6, F= 29.68°

Originated at Sherman, Cal. Two-year-old cion on one branch of No. 4, Mexican

stock.

6. Seedling (19206) . A=i.3o, P=i5.7, F= 29.66°

Grown from seed received from Coahuila, Mexico.

7. Seedling . . . . A=i.3o, P—15.6, F— 29.66°

Grown from seed of unknown origin.

8. Seedling . A=i.3o, P=x$.6, F= 29.66°

Grown from seed of unknown origin.

9. Seedling . A* *i.36, P«i6.3, F— 29.550

Grown from seed of unknown origin.

10. Seedling (34831) . A=i.37, P—16.5, F= 29.53°

Budded tree from a seedling growing in Rome, Italy.

11. Seedling (32400) . A=i.39, P=i6.7, F=29.5o°

Originated at Orange, Cal. Budded tree.

12. Fuerte . A=i.4i, P=i6.9, 1^=29.46°

Buds received from California, where the variety was introduced in 19 11 from

State of Puebla, Mexico. One-year-old cion on branch of No. 33, West Indian stock.

13. Seedling . A=i.43, P=i'j.2, F= 29.430

Grown from seed of unknown origin.

Guatemalan type. — This group was first called to the attention of
horticulturists by Collins in 1905.1 Within the last few years it has
been extensively planted in California and is now becoming known in
Florida. It is found commonly in th'e mountainous parts of Guatemala
and northward into southern Mexico, whence have come many of the
cultivated varieties now being propagated in the United States. It
was introduced to Hawaii about 20 years ago, according to Higgins,
Hunn, and Holt,2 while it appears to have been first planted in California
about 1885. In Florida it was probably not introduced earlier than
1900. It has not been observed in the West Indies,, with the exception
of a few trees recently planted, and its distribution in other countries
is quite limited.

In California, where it is best known horticulturally, it has been
found considerably hardier than the West Indian type, but somewhat
more tender, as a rule, than the Mexican.

14. Seedling (10978) . A=i.09, P=i3.i, F=3o.o4°

Grown from seed introduced from Guatemala in 1904.

15. Colorado . A=i.i6, P=i3-9, F= 29.91°

Budded on one limb of No. 24, West Indian type. Originated at Los Angeles, Cal.,

the seed having been sent from the State of Puebla, Mexico. Cion 1 year old.

16. Taft . A— 1.25, P=is.o, F=29.75°

Originated at Orange, Cal. Two-year-old cion on limb of No. 4, Mexican stock.

17. Seedling (38549) . A=i-34, P=i6.2, F=29.59°

Originated at Antigua, Guatemala. Budded on one limb of (26694), West Indian

type.

18. Sinaloa . A=i.35, P=i6.2, F= 29.57°

Buds obtained from California, where the variety was introduced in 19 11 from the

State of Puebla, Mexico. Cion 1 year old, on limb of No. 27, West Indian type.

1 Collins, G! N. The avocado, a salad fruit from the Tropics. U.-S. Dept. Agr. Bur. Plant Indus. Bui.
77. 53 P-, 8 pi. 1905*

* Higgins, J. E., Hunn, C. j., and Holt, V. S. The avocado in Hawaii. Hawaii Agr. Exp. Sta. Bui.
25, 48 P-, illus., 7 pl* 1911*

Nov. a, 1916 Freezing Point of Leaf Sap of Persea americana 265

19. Nutmeg . A=i.38, P= 16.5, ^=29.52°

Buds obtained from Hawaii, where the variety originated from a seed received from

Guatemala. Two-year-old cion on West Indian stock.

20. Seedling (36603) . A=i.38, P=i6.6, F— 29. 520

Budded tree. Original grown at Honolulu, Hawaii, from a seed imported from

Guatemala.

21. Seedling (19080) . A=i.39, P—16.8, F— 29.50°

Tree introduced as a plant from Guatemala in 1906.

22. Seedling (19058) . A=i.40, P—16.8, F— 29.48°

Grown from seed introduced from Guatemala in 1906.

West Indian type. — This is the commonest type on tropical American
seacoasts. Its precise origin is not known, but since it has been demon¬
strated by Collins 1 that the avocado was not cultivated in the West
Indies in pre-Columbian times it must have been introduced to the
islands from some point on the mainland. At the present day it is
practically the only type grown in the West Indies, and it is known to
be common in Colombia, Venezuela, Brazil, and Peru. It occurs in the
Mexican lowlands and in Yucatan. Most of the avocados grown in
Hawaii belong to this type, as do those of Tahiti, some of which are
occasionally shipped to San Francisco. In south Florida it is the prin¬
cipal type cultivated, having been introduced probably from Cuba at
an early day and planted on both the east and west coasts. In Cali¬
fornia its cultivation is very limited. It is found in the Canary Islands
and has been introduced to India, but in this latter country avocado
culture is of no importance.

Horticulturally the most important feature of this type is its suscep¬
tibility to cold, which prevents its culture in any but the warmest
sections of the United States. Experience has shown it to be severely
injured by a temperature at least 4 or 5 degrees Fahrenheit higher than
that required to injure the Guatemalan type.

23. Seedling (26707) . A=o.96, P=n.6, F=3 0.270

Originated at Fulford, Fla. Budded tree.

24. Seedling (26692) . A=i.oo, P= 12.0, F= 30.20°

Originated at Santiago de las Vegas, Cuba. Budded tree.

25. Seedling (26693) . A=i.o3, P=i2.4, F=3o.i5°

Originated at Cocoanut Grove, Fla. Budded tree.

26. Seedling (26704). . . . . A=i.o3, P=i2.4, F=3o.i5°

Originated at Miami, Fla. Budded tree.

27. Largo (18730) . A— 1.04, P?=i2.5, F=3o.i3°

Originated at Nassau, Bahama Islands. Budded tree.

28. Seedling (26698) . . . A=i.o4, P= 12.5, F=30.i3°

Originated at Fort Myers, Fla. Budded tree.

29. Seedling (19297) . A=i.o6, P= 12.8, F=3o.09°

Originated at Cocoanut Grove, Fla. Budded tree.

30. Seedling (36270) . A—1.07, P—12.9, F=3 0.07°

Grown from seed of unknown origin.

31. Seedling (26703) . .A=i.o8, P= 12.9, F=30.o6°

Originated at Buena Vista, Fla. Budded tree.

1 Collins, G. N. Op. cit.

266

Journal of Agricultural Research

Vol. VII, No. 6

32. Seedling (26694) . . .A=i.o9, P= 13.2, F=30.04°

Originated at Marco, Fla. Budded tree. *

33. Seedling . A=i.o9, P—13.1, P=3o.o4°

Grown from seed of unknown origin.

34. Butler (26690). . A=i.io, P= 13.3* F— 30.02 0

Originated at St. Petersburg, Fla. Budded tree.

35. Seedling (26691) . A=i.i5, P«i3-9, F=* 29.93 0

Originated at Buena Vista, Fla. Budded tree.

36. Mitchell (18120) . A=»i.i9, P—14.3, F=29.86°

Originated at Bayamon, Porto Rico. Budded tree.

37. Seedling (26713) . A=*i.23, P«i4.8, F— 29.790

Originated at Cocoanut Grove, Fla. Budded tree.

38. Seedling (19379) . A=»i.23, P= 14.8, F= 29.790

Grown from a seed received from Hawaii.

39. Pollock . . . . .A==i,24, P= 14.9, F=®29.77°

Originated at Miami, Fla. Budded tree.

DISCUSSION OF CONSTANTS

All these types show considerable variation in freezing-point lowering.
For the whole series the range is from A=o.96° to A=i.43°, or 0.470.
On the Fahrenheit scale the range in the freezing point of the sap is
from 29.430 to 30.27°. This variation is doubtless due to many causes.
In addition to errors of sampling in the collection of the leaves from the
individual trees and the unavoidable errors of measurement involved in
the manipulation of the sap samples, there are the fluctuations attrib¬
utable to the uncontrollable differences in the physiological states of the
different trees at the time of collecting the materials.

In view of these various factors, one would not expect to find transi¬
tion types entirely wanting, unless the differentiation of the types be
very clearly marked indeed. For more convenient comparison the values
obtained from the trees of the three types may be serially arranged in
intervals of five-hundredths of a degree (Table I).

Tabi.3 I. — Comparison of the freezing-point lowering values of three types of Persea

americana

Nov. 6, 1916 Freezing Point of Leaf Sap of Persea americana 267

An inspection of the values in Table I shows no clear difference between
the plants of the Guatemalan and Mexican types. From both of these
the West Indian type seems to be differentiated by a distinctly slighter
freezing-point lowering. All but 3 of the 17 West Indian determinations
are lower than 1.200, whereas all but 4 of the 22 based on plants of the
Guatemalan and Mexican types show a freezing-point lowering of 1.2 1°
or more.

If the averages of the freezing-point lowerings (A) with their probable
errors be determined as a more exact means of comparison from the
ungrouped depressions given for the individual varieties above, the
following constants are obtained (Table II).

Table II. — Freezing-point lowering constants of three types of Persea americana

Type.

Mean.

Standard

deviation.

. Coefficient
of variation.

Guatemalan .

1. 304 ±0. 024
1. 3°5 ± -0i6
1* 305 ± .014
1. 096 ± . o'i4

O. 106 ±0, 017
. 087 ± • 012
. 095 ± .014
. 0831b . 010

8. 16

Mexican .

6. 68

Guatemalan and Mexican .

7-2$
7- 55

West Indian .

The average freezing-point lowering in the Guatemalan and Mexican
types is practically the same. The difference is only 0.001 ±0.029 °f a
degree. The West Indian type is characterized by a distinctly lower
average than either of the other types. The differences are :

West Indian and Guatemalan type . — o. 209 ±0. 02 7 0

West Indian and Mexican type . — o. 2 10 ±0. 02 1 0

West Indian and Guatemalan and Mexican types . — o. 209 ±0. 019°

These differences are 7.6, 9.9, and n times as large as their probable
errors. Thus, notwithstanding the relatively small number of observa¬
tions upon which this study has of necessity been based, the differences
seem to be quite trustworthy in comparison with their probable errors.

Within the type the absolute variation in freezing-point lowering is
very slight, amounting to one- tenth of 1 degree or less. The relative
variability as expressed in terms of the coefficient of variation is also
low for a plant character.

SUMMARY

The constants presented in this paper prove that in a tropical fruit
of relatively recent introduction to North American horticulture, the
avocado, one of the groups of varieties, the so-called West Indian type,
is characterized by tissue fluids which freeze at a distinctly higher tem¬
perature than in the two other groups of varieties (Guatemalan and
Mexican). In the conventional terms of physical chemistry adopted by
physiologists, the expressed leaf sap of West Indian type varieties is
characterized by a slighter depression of the freezing point or by a
slighter freezing-point lowering than is that of the two other groups of

268

Journal of Agricultural Research

Vol. VII, No. 6

varieties. This differentiation seems to hold with remarkable constancy
notwithstanding the wide geographic origin — West Indian, Bahaman,
Central American, Mexican, and Hawaiian — of the seeds or budwood
from which the tissues dealt with originated.

The type which is characterized by the slightest freezing-point lowering
of its extracted sap — that is, the type in which the expressed sap freezes
at the highest temperature — is the one which has been shown by horti¬
cultural experience to be the least capable of enduring cold. That
capacity to withstand low temperatures is not solely due to differences
in the freezing point of the sap is evident from the slightness of the dif¬
ferences in the cryoscopic constants of the West Indian as compared
with the Mexican and Guatemalan types. Furthermore, horticulturists
believe that the plants of the Guatemalan type are intermediate in hardi¬
ness between those of the Mexican and West Indian types. There is, so
far as our data go, no discernible difference in the freezing point of the
sap of these types.

The problem is evidently one of considerable complexity. To what
extent other characteristics contributing to the capacity of the organism
to withstand low temperatures are correlated with sap properties remains
to be investigated.

It seems highly probable from the evidences presented in the paper
that in the case of the tropical perennials, a knowledge of the freezing-
point lowering of the sap would be of some service in predicting ability
to withstand cold. At least the subject is one deserving of more extensive
investigation. We would have been glad to carry out the present study
on a far more extensive scale, but the determinations given practically
exhaust the trees of flowering age in the collection of Guatemalan and
Mexican types at the Miami Plant Introduction Field Station, and it
will probably be several years before a better series is available.

GRAIN OF THE TOBACCO LEAF

By Charles S. Ridgway,1

Assistant , Tobacco and Plant-Nutrition Investigations , Bureau of Plant Industry

INTRODUCTION

When the buyer examines a sample of cigar leaf, he takes into account
such factors as texture, elasticity, color, luster, thickness, and grain. He
rolls a trial cigar and considers the bum, flavor, aroma, and character
of the ash. These he judges through long experience and by comparison
with an arbitrary standard for the crop of that particular year and the
type of leaf in determining the price he will pay for the tobacco of which
the sample is representative. Though the character of the grain itself,
whether it be fine or coarse, close or open, may be a relatively minor
consideration in fixing the value of a given sample, it is undoubtedly
closely correlated with some of the other factors considered and is there¬
fore of more importance than has hitherto been supposed.

OCCURRENCE AND GENERAL MACROSCOPIC APPEARANCE OF THE

GRAIN

The usual form of the grain of tobacco appears to the unaided eye as
minute pimples or papillae slightly raised above the general surface of the
leaf tissue. They are more prominent on the upper surface and vary in
size from about i mm. in diameter down to bodies so small as to be
entirely invisible without the aid of the microscope. When a well-cased
leaf is stretched over the ball of the finger, the grain becomes more con¬
spicuous (PI. 15, fig. A and B), and, indeed, it is in this way that the
tobacco buyer or grader determines the size and distribution of these
bodies. If a leaf with a coarse grain be pressed between the thumb
and finger, it feels as though particles of sand were adhering to it. Some
of the larger grain bodies can be picked out with the point of a knife;
and when such a grain is pressed between two hard substances, it crushes
in the same manner as would a particle of crystal, for the grain itself
is a crystalline body. On burning, these grain bodies swell and cause
the pearl-like pimples so frequently seen on the ashes of cigars (PI. 15,
fig. C and D). Grain is a characteristic of cigar tobacco, both domestic
and imported (7).2 In the flue-cured types macroscopically visible crys¬
talline deposits seem to be entirely absent. However, it is reasonable
to suppose that all air-cured types possess grain in some degree.

1 The writer wishes to express his gratitude for helpful suggestions and kind criticisms made by Dr.
W. W. Gamer, for inorganic and organic analyses made by Mr. C. L. Foubert and Dr. C. W. Bacon, re¬
spectively, and for kind cooperation on the part of Mr. Otto Olson in the matter of obtaining material.

2 Reference is made by number to '‘Literature cited," p. 273.

Journal of Agricultural Research.

Dept, of Agriculture, Washington, D. C.
fu

(269)

Vol. vn, No.- 4

Nov. 6, 1916
G— 99

270

Journal of Agricultural Research

Vol. VII, No. 6

With two notable exceptions the grain particles are entirely embedded
in the tissue of the leaf, being distributed more or less evenly in the web
and along the veins. One of the superficial types appears as raised
black dots, which are found chiefly on dark-colored leaves of a heavy
texture. The other is composed of those which give somewhat the
appearance of minute disks situated on or immediately beneath the
surface on either side of the leaf. These latter seem most conspicuous
on leaves of a medium or light color.

MICROSCOPIC CHARACTERS OF THE GRAIN

When a cased leaf of tobacco containing grain is stretched over the
stage of the microscope and examined in ordinary transmitted light with
a low magnification, most of the larger grain particles appear as more
or less rounded, highly refractive bodies of compact structure and bright
reddish brown color. Indeed, the ground tissue appears uniformly light
in color in comparison with that of the leaf observed macroscopically,
and it is seen that by far the larger part of the brown color of the leaf is
localized in the grain and in some of the small veins. Even when one
looks with a hand lens through a moist leaf against a strong source of
light, this localization of color in the grain is evident; the effect is that
of minute garnets embedded in a yellowish brown matrix.

In material mounted in Canada balsam, after having been dehydrated
in alcohol and cleared in xylol, the grain bodies become less conspicuous
in ordinary transmitted light, inasmuch as their refractive index more
nearly approaches that of the balsam than that of the air. The bodies
may be located, however, because of their difference in color from the
surrounding ground tissue, owing to the concentration in them of the
brown coloring matter and their slight power of refraction of light in
balsam (PI. 16, fig. A).

It is in polarized light that the grain bodies become most conspicuous,
since, as stated before, they are of crystalline structure. Favorable ma¬
terial mounted in balsam then exhibits clearly their size, shape, and
structure, as well as showing the degree to which the grain material has
been brought together into definite aggregates, a point which will be
mentioned later in this paper.

VARIOUS FORMS OF GRAIN

There are five general types of grain which merge insensibly into one
another and which may be briefly described in their order of abundance,
as follows: The first and most common type consists of more or less
spherical masses each composed chiefly of a group of palisade cells dis¬
tended with minute, radiating, needle-shaped crystals of a brown color
(PI. 16, fig. B). The second type includes flat or roughly hemispherical
bodies composed of cells of the mesophyll and epidermis, which are filled
with light brown to almost colorless, comparatively large plates ar¬
ranged somewhat regularly (PL 16, fig. C). The third is similar to the

Nov. 6, 1916

Grain of the Tobacco Leaf

271

first type, in that the bodies are usually spherical; but here the surface is
decidedly nodular and the cells included may be either the palisade or
the spongy parenchyma, or both. The crystals are radially arranged in
small groups the individuals of which appear to be thin, narrow plates,
and the color is more gray than brown (PI. 16, fig. D).

The fourth type attains the largest size and always has one surface
in common with the surface of the leaf. The particles consist of a num¬
ber of cells, palisade and epidermal, filled with a mass of dark-brown or
black substance which in the unbroken particle is inactive in polarized
light. When crushed, however, the fragments between crossed Nicol
prisms show the presence of crystalline material the form of which is
not apparent. These dark bodies are very striking, in that the epidermal
surface is usually craterlike in appearance, having a concavity which
frequently contains a few minute, colorless crystals surrounded by a
raised black ring. In a few of this type the central portion of the upper
surface shows no concavity, but radiating lines extending from a centrally
located spot beneath which the substance of the particle is soft and
easily crumbled suggest that this spot marks the location of the base of
one of the large trichomes or, possibly, the position of a stoma which
failed to close during the curing process (PI. 16, fig. E).

The fifth type of grain is composed entirely of microscopic sphere
crystals which are very active in polarized light and plainly show the four
extinction bars which rotate upon revolving the analyzer. Unlike the
other types, they usually do not fill the cells in which they occur. They
are colorless or light brown in ordinary transmitted light and possess no
visible differentiation into individual crystals (Pi. 16, fig. F). Tunmann
(8, p. 147) indicates that these spherites, which he believes to be malic-
acid salts, separate out when dried tobacco is placed in alcohol. The
writer has observed that they are formed in the leaf during the process
of curing and that in some instances they are visible under the micro¬
scope in the untreated, dry leaf with the aid of strong polarized light,
even after fermentation has been completed.

OTHER CRYSTALLINE MATERIAL OF THE LEAF

Aside from the grain, two other types of crystalline material are found
in the tissue of the cured and fermented tobacco leaf. One of these is the
cryptocrystalline or sand crystals of calcium oxalate contained in certain
cells (idioblasts) which are always present in all tobacco, even in the
green leaf while it is still attached to the growing plant, and which show
no appreciable change during the process of curing or fermentation. The
other is that which appears in nearly every cell of the leaf in both green
and cured tobacco and consists of small, single prisms. Some of the
properties of both of these types will be mentioned in connection with a
consideration of some of the chemical characteristics of the grain (PI.
16, fig. A, and PI. 17, fig. A).

272

Journal of Agricultural Research ,

Vol. VII? No. 6

METHOD OF SEPARATION OF GRAIN FROM THE DEAF AND MECHANICAL

ANALYSIS OF THE LEAF

Preliminary investigations indicated that a considerable supply of the
grain bodies free from the surrounding tissues of the leaf would be de¬
sirable. A method of mechanical separation was therefore worked out,
which, in brief, is as follows : A quantity of air-dry cigar leaf tobacco was
rubbed through a series of sieves (from io to 150 mesh) with a flat pestle
made from a rubber stopper. This process removed a large amount of
the soft web of the leaf from the veins as well as from the hard grain
particles, at the same time breaking the veins into short lengths, but
not crushing the grain. The mixture of grain and vein remaining in
each sieve was then ground in an unglazed porcelain mortar with a
rounded rubber pestle. This served to remove any soft tissue still
adhering to the particles of grain and vein. After sifting out with the
appropriate sieve the small amount of fine material thus rubbed off, the
mixture of grain and vein remaining was slowly poured upon a piece of
smooth paper inclined at an appropriate angle, determined by experiment.
The more or less spherical particles of grain rolled down the inclined paper
more rapidly than the cylindrically-shaped pieces of vein, resulting in
a partial separation of the former from the latter. It was found, however,
that a large amount of the grain belonging to the flatter types remained
with the vein, and it was therefore necessary to pick this grain out of the
mixture with the aid of a binocular microscope. By these tedious means
about 30 or 40 gm. of practically pure grain were obtained for subse¬
quent detailed study.

The separation of the grain and vein from the soft tissues of the leaf in
this manner in one instance also resulted in an approximate mechanical
analysis of the tobacco leaf. It was found that of a 70-gm. sample the
midribs represented about 33 per cent of the weight, while the soft tissue
(that which passed through a 150-mesh sieve) constituted 48 per cent; the
veins (other than the ribs), 8 per cent; and the grain, n per cent. Ex¬
cluding the ribs, the soft tissue represented 70 per cent; the veins, 12 per
cent ; and the grain, 1 7 per cent by weight. These results are only approxi¬
mate, since the primary object of the separation was to obtain a quantity
of the pure grain, and the determination of the proportion of various
tissues was an afterthought. It seems probable, however, that an im¬
proved method for the accurate mechanical analyses of tobacco would be
highly desirable, inasmuch as its various properties doubtless depend in
a large part upon the proportion of these three components of the leaf —
namely, soft tissue, vein, and grain.

CHEMICAL NATURE OF GRAIN

Chemically, the grain has been supposed by Sturgis (7) and by Loew
(5, p. 38-39) to be calcium oxalate, while from the fact that the grain
particles produce minute explosions as fire reaches them in the process of

Nov. 6, 1916

Grain of the Tobacco Leaf

273

smoking, they have been considered by many to be composed of potassium
nitrate. However, it has been shown that the grain is slowly soluble in
water. This was done by allowing free-hand sections of a fermented leaf
to remain in water for 4 hours and also by permitting a piece of a similar
leaf to soak in water on a slide for 17% hours. In order to show the posi¬
tion of the grain bodies and the idioblasts containing calcium oxalate,
camera-lucida sketches were made with the aid of polarized light at the be¬
ginning of the treatment with water. In both cases, at the end of 4 and
17 hours, respectively, no crystalline material remained in the positions
of the grain bodies, where at the beginning of the experiment the bodies
were highly active in polarized light. In the positions of the bodies,
however, both in the sections and in the piece of leaf, the cells that had
contained the crystalline material were found to be markedly distended,
although upon the application of the slightest pressure their walls imme¬
diately collapsed, giving additional proof that the substances which had
formerly filled the cells and held them rigid had been dissolved out. The
calcium oxalate sand crystals and -the scattered, single crystals referred
to above remained unchanged.

In a subsequent experiment a piece of fermented leaf tobacco was
dehydrated in alcohol and cleared and examined in xylol to determine the
character and distribution of the grain bodies and at the same time the
position and abundance of the idioblasts and the single, scattered crystals.
The xylol was then removed from the piece with alcohol and the latter in
turn displaced by distilled water, in a relatively large volume of which the
material was allowed to remain for 24 hours at room temperature. An
examination with polarized light was then made with the specimen
mounted in water. In this case also it was noted that the crystalline
material had disappeared from the grain bodies, but it was also evident
that neither the calcium oxalate sand crystals nor the single, scattered
crystals had been affected by the treatment. This was verified by
dehydrating, clearing, and mounting the tissue as already described.
The piece of leaf was then run back through water and allowed to stand in
50 per cent acetic acid for 48 hours and mounted and examined in the
same medium. Both at this examination and after washing in water
and again dehydrating and clearing, no scattered, single crystals could be
found — that is to say, the only crystalline material remaining in the
tissue after successive treatments with distilled water and with 50 per
cent acetic acid was the sand crystals of calcium oxalate. However,
these last all disappeared, leaving no optically active crystalline sub¬
stance whatever in the cells after the tissue had been subjected to treat¬
ment with 50 per cent hydrochloric acid for about 17 hours.

When separated from the surrounding tissues, however, the grain bodies
seemed to be less readily soluble in distilled water than when they were
treated in situ . This is probably due to the presence of some substance
in the tissues of the leaf which affects the solubility of the grain. In

274

Journal of Agricultural Research

Vol. VII, No. 6

every case, however, the grain-forming material was dissolved upon
treatment with water for 24 hours. There were left in the cells which
had formerly contained the solid crystalline substance the scattered,
single crystals and also frequently calcium oxalate sand crystals an
idioblast of which had been included in the group of cells which became
petrified through the deposit in them of the grain-forming substance.

HYGROSCOPIC PROPERTIES

In connection with the physical and chemical characteristics of the
grain may also be mentioned the results of some determinations of the
hygroscopidty of some of the component parts of the leaf procured by
the mechanical process already described. For these determinations
four classes of air-dry material were used: (1) the leaf web— that is,
the soft, parenchymatous tissue which passed through a 200-mesh
sieve; (2) large veins — that is, those which, excluding the midrib, would
not pass through a 20-mesh sieve; (3) smaller veins, which passed through
a 30 — but not an 80-mesh sieve; and (4) grain, a mixture of various sizes.
Two-gm. samples of each of these classes were exposed for 50 hours side
by side in open, tared Petri dishes to an atmosphere containing moisture
derived from 125 gm. of granulated cigar-filler tobacco containing 26
per cent of moisture, a tight desiccator kept at about 30° C. being used as
a container. The percentages of moisture absorbed by the samples were
determined in the usual way, with the following results: Leaf web, 20 per
cent; large vein, 17 per cent; small vein, 25 per cent; and grain, 14 per
cent.

The granulated tobacco was used as a source of moisture in order to
produce, as nearly as possible, natural conditions and the greatest com¬
petition between the samples and also between these and the source of
moisture, at the same time insuring against the condensation of moisture
within the desiccator through changes in temperature. While these
results are products of work of a preliminary nature, enough has been
done to indicate that the grain is the least and the small veins the most
hygroscopic of the kinds of material studied. In connection with the
latter it is believed that their power to absorb so large an amount of
moisture may be due, at least in part, to a colloidal substance present in
some of their xylem or phloem elements. This substance has the property
of swelling markedly and protruding from the ends of pieces of small
veins when the latter are submerged in water.

qualitative tests

Qualitative tests of the pure grain and its ash showed an abundance
of calcium, some potash and magnesia, and a little ammonia, but only
traces of oxalic, nitric, and sulphuric acids, the last of which was proved
to arise chiefly from the combustion of proteids in the process of ashing.

Nov. 6, 1916

Grain of the Tobacco Leaf

275

The ash of the grain, however, contained a large amount of carbonate and
therefore indicated the presence of some organic acids other than oxalic.

Aside from the solubility of the grain material in water and the detec¬
tion of only a small amount of oxalic acid by ordinary qualitative means,
Borodin’s method (2), which consists of treating the substance to be
tested with a saturated solution of the suspected substance, showed the
grain to be composed of a salt or salts other than calcium oxalate.

Further, through the interest and kind assistance of Dr. F. E. Wright,
of the Geophysical Laboratory of the Carnegie Institution of Washington,1
the application of the petrographic microscope and methods to a study of
the crystalline substances of the grain revealed the presence of normal
calcium malate in these bodies. These methods also showed the presence
of other crystalline substances, the identity of which has not yet been
established. As is evident from the detailed quantitative analyses
hereinafter recorded, the grain is a mixture of salts some of which may
be double or even triple combinations, and the difficulties of procuring
these various possible compounds in crystalline form for a comparison
of their optical properties with those of the salts entering into the com¬
position of the grain have not been surmounted at this stage of the
investigation.

QUANTITATIVE ANALYSES

The data given in Table I represent quantitative determinations made
on portions of the same material as that from which samples were taken
for a study of the hygroscopic properties. The mineral components were
determined by the official methods 2 and the organic acids by the method
of Kissling.3

Table I. — Composition of the leaf web , vein, and grain of Pennsylvania tobacco , on

basis of sand-free, air-dry material

Constituents determined.

Leaf web pass¬
ing through
200-mesh sieve.

Large vein
(excluding
midrib) not
passing through
20-mesh sieve.

Small, vein
passing through
30-mesh but
not 80-mesh
sieve.

Grain of
various sizes.

Moisture . .

Pure ash .

Potassium oxid .

Calcium oxid .

Magnesium oxid .

Oxalic acid .

Citric acid . . .

Malic acid . .

Total determined a .

Per cent.

7. 80
16. 48
4- n
6. 70
i- 33

3. 16
3. 26
3- 46

Per cent.

7. II
17. 80
5- 72

6. 65

1. 46

2. 75

4*09

- Per cent.

6. 25
16. 91

5- 29
6.6
*■ 59
3-o

2. 91

5- 24

Per cent.

8.06
40. 26
3* 42
26. 34
3- 13
.82
22. 38

13. 58

29. 82

28.93

30. 88

77- 73

1 The writer wishes to express his deep appreciation of Dr. Wright's assistance.

2 Wiley, H. W., ed. Official 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.), 272 p., 13 fig. 1908. Reprinted, 1912.

3 Kissling, Richard. Beitrage zur Chemie des Tabaks. Zur Tabakanalyse. In Chem. Ztg., Jahrg. 28,
No. 66, p. 77S-776, 1 fig. 1904*

« Pure ash not included.

276

Journal of Agricultural Research

Vol. VII, No. 6

The analyses recorded in Table I show that the grain contains a very
high percentage of ash, that calcium is by far the most abundant con¬
stituent, that nearly as much magnesium is present as is contained in the
three other parts of the leaf taken together, but that a smaller amount of
potash is found in the bodies than in any other class of leaf material
studied. Of the organic acids, citric and malic are abundant in the grain,
while oxalic acid is present in a much smaller percentage than in any of
the other classes of material. These data indicate that the grain ma¬
terial must be a mixture of citrates and malates, chiefly of calcium, with
some magnesium and a little potassium. Nicotin seems to enter into
the composition of the grain also, since about 1.5 per cent of that sub¬
stance was found in an analysis of one sample of grain. The totals of
the percentages of substances determined indicate that the grain (77.73
per cent total determined) contains little material other than these sub¬
stances. There remains in this case only about 22 per cent undetermined,
and this must account for cellulose, nicotin, and other nitrogenous com¬
pounds, sulphur compounds, and phosphoric acid, as well as small
quantities of sodium, iron, aluminum, etc., while in the three other classes
these undetermined substances constitute in each about 70 per cent.

CORRELATION OF THE GRAIN WITH BURNING QUALITY

Although persons experienced in the handling of tobacco consider that
a well-developed grain is an indication of good quality, particularly with
reference to the bum, it seems that this is a matter of practical experi¬
ence with them rather than a factor permitting definite discussion or
explanation.1 They assert that a “close-grained” leaf will bum poorly,
while one possessing an “open grain” will have a greater fire-holding
capacity. Recently, in the course of interviews with packers and manu¬
facturers in Lancaster County, Pa., these observations have been found
capable of substantiation by microscopic examination. The writer
requested each of several practical tobacco men to select, on the character
of the grain alone, what he considered to be a good-burning and a poor-
burning leaf. In making this selection the leaf was stretched and while
taut was allowed to pass slowly through the fingers, usually accompanied
by the remark that the leaf showed an open grain or that it was close-
grained. The leaves designated as poor-burning (close-grained), invari¬
ably possessed a hard texture — that is, something of the nature of the
softer grades of paper — always showed poor elasticity, and rarely ex¬
hibited grain bodies on the surface. On the other hand, the leaves judged
to be good-burning (showing an open grain) were comparatively soft in
texture, elastic, and usually possessed grain bodies sufficiently large to

1 See also Hayes, H. K., East, E. M., and Beinhart, E. G. (4, p. 28). The authors used seven classes to
indicate the prominence of grain in considering the quality of the strains studied.

Nov. 6, 1916

Grain of the Tobacco Leaf

277

be visible to the unaided eye. It may be mentioned, however, that
occasionally a leaf was found which was classed among the “open¬
grained” the grain of which was not apparent on the surface of the leaf,
though evidently the experienced hand could detect its presence and
determine its character.

The bum of the leaves thus selected was tested, always with the result
predicted. A small piece was then cut from the unbumed portion of
each leaf at about a quarter of an inch from the point at which the fire
was extinguished. These pieces were dehydrated in absolute alcohol,
cleared and mounted in cedar oil, and examined microscopically with polar¬
ized light. Without exception, the good-burning leaf showed that the
grain material had become well aggregated into definite bodies separated
one from the other by a band of tissue free from crystalline material,
save for the scattered, single crystals and occasional idioblasts containing
calcium oxalate. This condition was evident even in the open-grained
leaves mentioned above in which no grain was macroscopieally visible on
the leaf, for in those cases the bodies, though too small to cause a swell¬
ing on the surface, were definite in form and sufficiently separated to
allow a zone of comparatively empty cells around each.

In the pieces from the poor-burning (close-grained) tobacco, on the
contrary, the crystalline grain material proved to be scattered more or
less evenly throughout the tissue, without any considerable degree of
aggregation into definite bodies large enough to leave an encircling zone
of empty cells around each particle. A few poor-burning samples have
been found, however, in which some aggregation of grain substance
had occurred, but in these the intervening tissue was filled with a mass
of grain material which, for some reason, had failed to form definite
bodies, thereby producing the same condition found in the poor-burning
samples which were without appreciable aggregation.

From these facts it would seem that a certain degree of aggregation,
with intervening tissue free from grain substance, is necessary in order
that good fire-holding capacity may be assured. This suggests that the
substance composing the grain bodies may have in reality a retarding
effect upon the advance of fire in the tissue and that zones of grain-free
cells must be present in order that a sufficient degree of heat may be
generated, by their more rapid combustion, to ignite the solid grain
particles. Verification of this suggestion has partially been obtained
by watching through a binocular microscope the progress of the fire in
the tissue. In a piece of leaf with prominent grain it was found that in
the process of burning the fire line passed part way around and in some
cases even beyond the centers of large grain bodies before they became
ignited. It seems reasonable to suppose that the same process takes
place in a leaf in which the grain particles are too small to admit of in¬
vestigation in this manner and, further, that in a close-grained leaf there
64310°— 16 - 2

278

Journal of Agricultural Research

Vol. VII, No. 6

is not enough grain-free tissue to act as kindling material and produce the
temperature required to ignite the almost solid mass of grain substance
characteristic of poor-burning tobacco.

While several theories have been set forth1 concerning the bum of to¬
bacco, for the most part chemical and differing from the hypothesis just
advanced, an instance in favor of the present explanation is found in the
case of the notoriously poor-burning 1909 crop of Pennsylvania. In this
tobacco the grain substance is abundant, but it occurs in the tissue as an
almost solid sheet, showing very poor aggregation into definite grain
bodies (PI. 17, fig. B). Again, flue-cured tobacco, which is always con¬
sidered a poor-burning type, is entirely without grain aggregates. In this
tobacco there is a small amount of grain substance (PI. 17, fig. C) com¬
pared with cigar types, and it appears to be deposited very rapidly in
the form of a haze of minute crystals throughout the tissues of the leaf.
It is thought that the rapid-curing method employed with this type is
responsible for this condition, inasmuch as the same kind of tobacco,
grown in the same region and air-cured instead of being subjected to
heat, shows to some degree the aggregation of the grain substance into
definite bodies. In this connection it may be said also that Connecticut
tobacco (PI. 17, fig. D), which as a rule shows an excellent bum, also
possesses a high degree of aggregation of the grain substance.

In connection with some investigations concerning the cause of the
defects in the burning quality of York County tobaccos jointly pursued
by the Office of Tobacco and Plant Nutrition Investigations and the
Pennsylvania Agricultural Experiment Station2 working cooperatively,
a certain degree of correlation of grain formation and burning quality
has been found. The tobacco was grown at Red Lion, Pa., during the
seasons of 1913 and 1914 in duplicate plots upon which nine different
fertilizer treatments, identical for the two years, were used to determine
their influence upon the burning quality of the product. The material
used for microscopical investigations was comparable to that upon which
the bum and other factors were determined. In preparing the material
for examination with the microscope, disks were cut from the leaves

t Gamer discusses (3) the more important contributions concerning #the bum of tobacco. Of the
two physical theories thus far advanced, the one sought (1) to correlate the burning quality of to¬
bacco, at least in part, with such anatomical features of the leaf as the number of rows of cells, the size of
intercellular spaces, etc., while the other attributes (6) a beneficial influence to the action of potash salts
of the organic acids in swelling to many times their original bulk and thereby yielding a porous mass of
finely divided carbon when decomposed by heat. The chemical data given herewith do not support the
latter theory, since the organic acids are largely localized in the grain, and only a small portion of the potas¬
sium is present in these bodies. This indicates that the greater portion of the potassium in the leaf is not
in combination with the organic acids. No experimental work has been done by the writer concerning the
former theory, though it is believed that in view of the results published herein and the fact that in cigar
tobaccos the intercellular spaces are obliterated and the cell walls collapsed during the fermentation process
little ground would be found for its substantiation.

2 A report of this cooperative work by Dr. William Frear, Chemist and Vice Director of the Pennsylvania
Agricultural Experiment Station, will appear in the annual reports of that institution for 1913-14 and
1914-15,

Nov. 6, 1916

Grain of the Tobacco Leaf

279

after fermentation and these were dehydrated in absolute alcohol, cleared
in xylol, and mounted in Canada balsam.1

In ordinary transmitted light the samples showed no marked varia¬
tions one from the other. The use of polarized light, however, revealed
striking differences in the degree to which the grain substance had
become aggregated, and it was found that these differences seemed to
remain fairly constant between the good-burning samples on the one
hand and the poor-burning samples on the other. In order that the
condition found might be expressed in some form more concise than by
descriptive terms, a system of scoring the degree of aggregation was
adopted. By using a maximum perfect score of 10 points, convenient
minor classes occurred as follows: “Good,” 8 points; “fair,” 6 points;
“poor,” 4 points; and “very poor,” 2 points. A detailed score (Table
II) of a poor-burning sample will best serve to illustrate the method
used.

Table II. — Detailed score of poor-burning sample from Plot I, Red Lion, Pa,, 1913

Disk No.

Grain.

Score.

Poor aggregation; small amount; haze of single and cryptocrys¬
talline crystals .

Poor aggregation; small aggregates; flaky spherites, cryptocrys¬
talline haze .

Very poor aggregation; small amount; fine cryptocrystalline

haze with larger singles .

Very poor aggregation; dense, hazy mass .

Poor aggregation ; similar to disk 4, though less dense .

No aggregation ; spherites and cryptocrystalline haze . .

Very poor aggregation; similar to disk 4 .

No aggregation; fine, thin cryptocrystalline haze .

Poor aggregation; dense; a few flaky and a few black superficial
aggregates . .

o- 5
2. o

i9- 5

Divided by 9, score is

2. 16

It will be noted in the score that the word “spherites” is used in the
description of some of the samples. These are of much more frequent

1 Of the 1913 material only the two best and the two poorest-burning samples were studied in detail,
while the samples from all the fertilizer treatments used in 1914 were subjected to an equally thorough
investigation. The microscopical data here presented represent, for the 1913 material, detailed exami¬
nations of disks cut from each of 18 representative leaves from each of the four fertilizer treatments
after the fermentation process was completed. In 1914, however, the leaves were selected in the field
before the plants were harvested. In this case three leaves from each of six average plants in each treat¬
ment were sampled in the field by removing a small disk of tissue from each leaf selected at a point equi¬
distant from the midrib and the margin and midway between the base and apex. These disks were
killed in absolute alcohol and preserved for future study in the same liquid. The leaves thus sampled
were the eleventh, twelfth, and thirteenth below the point of topping, which point in all of the plants bore,
as nearly as possible, the same relation to the developing flowers. These same leaves, having been tagged
in the field, were used later in a preliminary study of the development of the grain and, after fermenta¬
tion, for determining the character of the fully developed grain together with the fire-holding capacity and
certain chemical analyses.

28o

Journal of Agricultural Research

Vol. VII, No. 6

occurrence in poor-burning than in good-burning tobacco, as are also
large, single crystals of grain substance and aggregates composed of large,
platelike or flaky crystals. The cryptocrystalline haze is also much
more prominent in the tissues between the relatively few definite grain
bodies in the poor-burning tobacco. It will also be noted that no
account is taken of the calcium oxalate or the single, scattered, acetic-
acid-soluble crystals. The result of this method and the correlation of
this result with apparently related factors are seen in Tables III and IV
and figures i and 2. Corresponding data relative to the potash and
chlorin content of the tobaccos are also included in the tables.

Table III. — Data relative to the burning quality of tobacco from fertilizer treatments at
Red Lion, Pa., crops of IQ13 and IQI4

1913

Treatment.

Bum
score
(maxi-
mum=> 20
points).0

Fire-holding
capacity in
cigar.®

Grain
aggrega¬
tion
(Maxi-
mum= 20
points.
Original
score
Xa).«

Potash

content.®

Chlorin

content.®

Potash-

chlorin

ratio.®

Burning

quality.

Min. sec.

Per cent.

I .

13-3

4 25

5-4

3- 7

I. 78

2*3

Poor.

Ill .

18. 0

5 37

13. 2

4. 01

•

7* 1

Good.

VII .

17-5

6 7

12. 8

4. 14

. 64

6. 7

Do.

IX .

11. 0

4 22

7*3

3*44

2. 66

i*3

Poor.

1914

I .

14- 0

5*7

3. 61

1. 38

2. 6

Poor.

Ill .

18. 0

11. 1

3* 42

. 26

r3* 1 5

Good.

VII .

18. 0

9*7

3* 57

•23

I5*5

Do.

IX .

12. 0

4*3

3* 34

2. 44

1.36

Poor.

a Data furnished from report of cooperative work by Dr. William Frear, Chemist and Vice Director of the
Pennsylvania Agricultural Experiment Station. The burn score and fire-holding capacity in the cigar
were determined by Mr. Otto Olson, Assistant in Tobacco Investigations, Bureau of Plant Industry, and
the determinations of the potash ana chlorin content and the potash-chlorin ratio were made by Mr. E.
S. Erb, Assistant Chemist, Pennsylvania Agricultural Experiment Station.

6 Data obtained by the writer and included in the report of Dr. W. W. Gamer on the portion of the
cooperative work undertaken by the Office of Tobacco Investigations, Bureau of Plant Industry.

In Tables III and IV and figures 1 and 2 the fire-holding capacity of the
cigar has reference to the length of time a cigar made entirely of the
tobacco in question remains lighted. The bum score includes the fire¬
holding capacity, amount of charring, character of ash, etc.

In determining the fire-holding capacity of the open leaf the writer
ignited separately each leaf of the sample with the glowing end of a piece
of punk, and the number of seconds during which the fire glowed in the
larger portion of the burning area was taken as the fire-holding capacity.

Nov. 6, 1916

Grain of the Tobacco Leaf

281

Table IV. — Data relative to the burning quality of the tobacco grown on fertilizer plots

at Red Lion , Pa., in IQ14

Fertilizer applied per acre.

Fire-

Grain
aggre¬
gation
score
(maxi¬
mum 20
points.
Origi¬
nal
score
X2,“)

Treat¬

ment.

Kind.

Quantity.

Bum
score
(maxi¬
mum 20
points)/1

Fire¬
holding
capacity
in cigar.

holding

capac¬

ity,

punk-

stick

meth¬

ods

Potash

con¬

tent.*1

Chlorin

con¬

tent.®

Potash-

chlorin

ratio.®

Manure .

Pounds.
20, OOO

14. 0

Min.

sec.

Sec.

8. O

5- 7

Per ct.
3.61

Per ct.

1. 38

2. 6

[Manure .

20, OOO

| Dissolved rock. . . .
Potassium sul-
[ phate .

343

100

I16. 0

6. 2

6. 0

3- 85

• 91

4. 2

III

[Cottonseed meal. .
Potassium sul-
[ phate .

1,485

2§8

i>48S

190

258

1,485

372

258

1,485

129

1X2

18.0

16. 8

11. 1

3- 42

. 26

13- 15

Cottonseed meal. .
Dissolved rock. . . .
Potassium sul¬
phate . .

17-5

14. 8

7- 1

3* i-7

. 18

17. 6l

Cottonseed meal. .
Dissolved rock. . . .
Potassium sul¬
phate .

‘17. O

14. 2

8. 1

3-75

. 19

T9* 73

Cottonseed meal. .
Potassium sul¬
phate .

17- 5

19. 8

8. 1

4. 06

•23

i7- 7

Potassium car¬
bonate .

VII

Cottonseed meal. .
Precipitated bone .
Potassium sul¬
phate .

1,485

136

258

1,485

224

190

1,485

372

.258

18. 0

20. 5

9-7

3- 57

*23

15- 5

VIII

[Cottonseed meal. .

1 Potassium car-
| bonate .

17. 0

10. 1

7.0

3- 47

•23

15.08

[Dissolved rock. . . .
[Cottonseed meal. .
< Dissolved rock. . . .
1 Potassium chlorid .

‘12. O

5-4

4.3

3- 34

2.44

1. 36

Aver

II]

age for treatments
II, IX .

14. O

12.5

6- 5

16. 0

5-3

8-5

7.. 6

1. 58

0. 22

2. 7

age for treatments
[ to VIII, inclusive

O' w

3- 57

16. 46

a Same as for Table III.

As is illustrated by figures 1 and 2, representing the data included in
Tables III and IV, respectively, the curves for the burn, the fire-holding
capacity in the cigar, and the grain aggregation show a marked parallel¬
ism, although in the fire-holding capacity determined on single leaves the
second highest point in the curve does not correspond to either of the
principal maxima in the three other curves.

While the relative values representing the physical factors determined
for the poor-burning samples correspond in a general way with Dr. Frear's

282

Journal of Agricultural Research

Vol. VII, No. 6

Fio. i.— Curves plotting data relative to the burning quality of tobacco from fertilizer treatment at Red
Lion, Pa., for crops of 1913 and 1914. Data taken from Table III.

Nov. 6, 1916

Grain of the Tobacco Leaf

283

Fig. 2.— Curves plotting data relative to the burning quality of the tobacco grown on fertilizer plots at
Red I<iont Pa.f in 1914. Data taken from Table IV.

Journal of Agricultural Research

Vol. VII, No. 6

284

potash-chlorin ratios and show an inverse relation to his chlorin-content
figures, this relation does not hold for the good-burning samples, except
in the case of the Treatment VI previously referred to. The relative
inferiority of the poor-burning samples in this series is doubtless traceable
in part to their high content of chlorin, which element is commonly re¬
puted to be antagonistic to combustion.

Though a more or less definite correlation seems to exist in the tobacco
studied between the burning qualities and grain aggregation, for which a
tentative explanation has been advanced, no reason is evident at this
time for the apparent relation between the low degree of aggregation and
the high chlorin content of the poor-burning samples. It is hoped that
future work may throw some light upon this phase of the problem as well
as upon other possible factors affecting grain formation in tobacco.

DEVELOPMENT OF THE GRAIN

While the opportunity has not been offered for a thorough study of the
process of formation and aggregation of the grain in the leaf, preliminary
investigations upon material referred to in footnote on page 279, as well as
upon various kinds of tobacco observed in the laboratory, indicate that
the grain substance is not present in crystalline form in the living, green
leaf. These observations prove that the formation and development of
the grain bodies are concomitant with the process of curing, and there
are indications that the aggregation of the grain substance continues
during fermentation (compare PI. 17, fig. A, E, G).

At Red Lion, Pa., on September 22, 1914, four days after harvesting,
samples of tobacco were becoming flecked near the margins of the leaves
with the yellow mottling characteristic of the early stages of curing.

In this condition microscopic examination showed the absence of grain
bodies and also that the protoplasmic contents of the cells were still intact.
In the later stages of yellowing and with the apparent death and disorgan¬
ization of the protoplasm, just before the development of the brown
color, highly refractive droplets appeared in the cells. Upon attempting
to dehydrate a piece of such a leaf in alcohol minute crystals appeared in
these droplets. In the next stage — that is, at about the first indication
of a brown color in the leaf — the droplets had enlarged and become de¬
cidedly viscid, and each showed upon examination with polarized direct
sunlight conspicuous crossed extinction bands and presented the appear¬
ance of spherites. Upon dehydration at this stage they developed no
small crystals. As the browning progressed, certain of the bodies seemed
to differentiate within themselves, without artificial dehydration, a
nucleus of crystals of definite though not identifiable form. The sub¬
stance of the droplets in which no crystals had formed seemed to migrate
toward those which had produced crystal nuclei, increasing by accretion
the crystalline mass until groups of several cells each had become literally

Nov. 6, 1916

Grain of the Tobacco Leaf

285

petrified and formed a solid grain body. Definite grain was formed in
leaves of average maturity in nine days after the tobacco was harvested.

It is thought that the migration of the still viscid substance was
brought about by diffusion, through the dead and permeable proto¬
plasm, set up by a lowering of concentration in the immediate vicinity
of the developing crystal nuclei. Under ideal curing conditions it is prob¬
able that this process results in the aggregation of all, or nearly all,
of the grain-forming substance into definite bodies, leaving around each
a zone of cells largely free from that material. It is believed that under
these conditions the number of grain bodies developed in a given area is
dependent upon the number of crystal nuclei formed. The factors which
determine the number and distribution of the latter are not known.
It is thought, however, that the rate of desiccation and the degree of
maturity of the curing leaves may be of importance in this respect.

The behavior of the grain material in the flue-cured tobacco seems to
support the idea that the rate of drying is a factor in grain development.
In the completely cured leaf of this type the substance which would have
formed more or less definite grain bodies under ordinary air-curing con¬
ditions apparently has been thrown down in crystalline form by the
rapid curing and drying characteristic of the method, in the first droplet
stage referred to above, resulting in a haze of minutely crystalline mate¬
rial in all the cells. Even Connecticut tobacco, when cured very rap¬
idly (“hayed-down”), frequently shows poor burning qualities. Again,
in the poor-burning crop of tobacco produced in Pennsylvania in 1909,
which was a “dry- weather crop,” the grain material failed to become
well aggregated. It seems probable that either a low percentage of
water in the tissues or an abnormally large quantity of grain-forming
substance in solution in the cell sap at the time of harvesting, or both,
must have resulted in the formation of an unusually large number of
crystal nuclei in a given area, especially if weather conditions were such
as to cause rapid dessication in the early stages of curing. In the later
stages, then, opportunity for the development of zones of cells free from
grain substance would have been limited by the closeness together of
the crystal nuclei. Indeed, samples of this crop have been seen in
which the grain substance seemed to have been thrown down in prac¬
tically every cell, much in the manner of that of flue-cured tobacco,
though greater in amount. Of interest to note in passing is the fact that
the burning properties of this crop, much of which is still in storage, are
very gradually improving. It seems probable that this may be due
in part to the slow aggregation of grain material rather than to the loss
through aging of some substance injurious to the burning quality.

The leaves upon which the observations were made at Red Lion, Pa.,
were sampled after the curing had been completed, and it was found
that the grain throughout the leaf was in practically the same condition

286

Journal of Agricultural Research

Vol. VII, No. 6

as had obtained near the outer edges of the leaves on the ninth day of
curing. Examination after the same leaves had been fermented, how¬
ever, showed that the grain was much more pronounced on the surface,
and the microscope revealed a greater degree of aggregation than had
existed at the end of the curing process. It is believed that with the
high water content of the fermenting tissue the grain substance is gradu¬
ally put in solution, particularly in the case of the smaller bodies, and
that further grain aggregation and development of intervening zones of
comparatively empty cells is thereby made possible.

SUMMARY

(1) The grain of cigar tobacco consists of hard bodies which, if suf¬
ficiently large, cause the surface of the leaf containing them to present a
pimply appearance. This grain, in connection with other intimately
related properties of the leaf, constitutes an important factor in deter¬
mining the value of cigar tobacco.

(2) Each body consists of from one to several leaf cells distended with
a mass of crystalline substance. They are most prominent microscopi¬
cally when examined with polarized light. Although visible in ordinary
transmitted light, owing to the concentration in them of a large part of
the brown coloring matter of the leaf, the details of their structure are
more apparent when the former method of examination is used.

(3) Based upon microscopic features, five forms or types of grain are
recognized, though their significance is a matter still to be investigated.

(4) Two other kinds of crystalline material are found in the tobacco —
namely, cryptocrystalline calcium oxalate, contained in certain cells in
the various tissues of the leaf; and single, small, prismatic crystals
scattered evenly throughout the leaf, one in nearly every cell of the
mesophyll and epidermis.

(5) A mechanical method resulted in the separation of the grain bodies
from the other portions of the leaf and gave an approximate mechanical
analysis. This analysis showed roughly the percentages by weight of
leaf web, veins, and grain.

(6) Chemical analyses proved that the grain is composed chiefly of
calcium, with a little magnesium and potassium, in combination with
citric and malic acids rather than with oxalic acid. One of the salts,
normal calcium malate, was identified by petrographic methods. De¬
terminations of the hygroscopic properties of the component parts of
the leaf separated by mechanical means indicate that the grain is not
responsible for the marked hygroscopic properties of tobacco, since it
absorbed the least water from a moist atmosphere. The small veins
showed the greatest hygroscopicity.

(7) The grain bodies of tobacco are developed in the course of post¬
mortem changes which take place during the process of curing and con-

Nov. 6, 1916

Grain of the Tobacco Leaf

287

tinue during fermentation. A microscopically visible change consists
of a more or less complete aggregation of the grain-forming substance
of all the cells into certain groups of cells. The factors determining the
location of these groups are unknown.

(8) In the tobacco studied a correlation was found between the grain
and burning properties. It is believed that the substances contained
in the grain bodies are injurious to the burn and that the quality of the
latter is dependent upon the degree to which the former are aggregated
into definite bodies sufficiently separated, one from the other, to permit
a considerable fire-carrying zone of cells, emptied of grain material,
around each. The influence of the degree of aggregation of the grain
substance upon the color, texture, and elasticity of the leaf has not yet
been thoroughly investigated.

LITERATURE CITED

(1) Behrens, Johannes.

1894. Weitere Beitr&ge zur Kenntnis der Tabakpflanze. V. Der anatomische
Ban nnd die Bestandteile des Tabakblatts in ihrer Beziehung zur Brenn-
barkeit. In Landw, Vers. Stat., Bd. 43, p. 271-301.

(2) Borodin, Johann,

1878. Ueber die physiologische Rolle und die Verbreitung des Asparagins im
Pflanzenreiche. In Bot. Ztg., Jahrg. 36, No. 51, p. 801-816; No. 52, p.
817-831.

(3) Garner, W. W.

1907. The relation of the composition of the leaf to the burning qualities of tobacco.
U. S. Dept. Agr. Bur. Plant Indus. Bui. 105, 25 p.

(4) Hayes, H. K., East, E. M., and Beinhart, E. G.

1913. Tobacco breeding in Connecticut. Conn. Agr. Exp. Sta. Bui. 176, 68 p.,
12 pi. _

(5) LoEw, Oscar.

1900. Physiological studies of Connecticut leaf tobacco. U. S. Dept. Agr. Rpt.
65. 57 P-

(6) SCHEOESING.

i860. Chimie appliqu6e a la vegetation. — Nouvelles recherches sur le tabac.
In Compt. Rend. Acad. Sci. [Paris], t. 50, no. 13, p. 642-644.

(7) Sturgis, W. C.

1900. On the so-called grain” of wrapper tobacco. In Conn. Agr. Exp. Sta.
23d Ann. Rpt., 1899, p. 262-264, pi. 2.

(8) Tunmann, Otto.

1913. Pflanzenmikrochemie ... 631 p., illus. Berlin.

PLATE is

Fig. A, B. — Well-cased tobacco leaves stretched over the closed end of a test tube;
showing very pronounced grain development. X 2.

Fig. C. — A portion of a cigar wrapped with a leaf containing very coarse grain. X 2.
Fig. D. — The same as figure C, but after a portion of the cigar had been smoked,
showing the white pimples in the ash produced by the burning and swelling of the
grain bodies. X 2.

(288)

PLATE 16

Fig. A. — Grain bodies of Connecticut Broadleaf tobacco as seen in ordinary trans¬
mitted light. a, Idioblasts containing sand crystals of calcium oxalate. X about 300.

Fig. B. — Representative grain bodies of class 1. X 19.5.

Fig. C. — Representative grain bodies of class 2 . The high lights show the positions
of the distended epidermal cells. X 19.5.

Fig. D. — Representative grain bodies of class 3. X 19.5.

Fig. E. — Representative grain bodies of class 4. X 19.5.

Fig. F. — Grain substance in the form of minute spherites. Photographed with polar¬
ized light. X 280.

PLATE 17

Fig. A. — Green tobacco leaf from Treatment I, Red Lion, Pa., 1914, killed in absolute
alcohol and showing idioblasts of calcium oxalate and minute, scattered, single crystals
of an undetermined substance, but no grain. Photographed with polarized light.

X 65.5.

Fig. B. — Representative sample of the poor burning 1909 Pennsylvania tobacco.
Photographed with polarized light. X 65.5.

Fig. C. — Flue-cured tobacco. Poor-burning. Photographed with polarized light.
X 65.5.

Fig. D. — Connecticut Broadleaf tobacco. Good burning. The brown coloring mat¬
ter concentrated in the grain partially masks their structure in some cases. Photo¬
graphed with polarized light. X 65.5.

Fig. E. — Fermented tobacco from Treatment I , Red Lion, Pa. , 1914. Poor burning.
Photographed with polarized light. X 65.5.

Fig. F. — Fermented tobacco from Treatment III, Red Lion, Pa., 1914. Good burn¬
ing. The brown coloring matter concentrated in the grain partially masks their
structure in some cases. Photographed with polarized light. X 65.5.

Fig. G. — Tobacco from Treatment I, Red Lion, Pa., 1914. Cured only. Photo¬
graphed with polarized light. X 65.5.

HOST PLANTS OF THIELAVIA BASICOLA

By James Johnson,

Assistant Horticulturist , Wisconsin Agricultural Experiment Station 1

The increasing economic importance of Thielavia basicola Zopf as a
root parasite of certain cultivated plants has led to a desire for more com¬
plete information regarding its range of host plants. This knowledge is
important not only from the standpoint of the effect upon the hosts
themselves but in the relation of the use of these host plants in rotation
with other crops susceptible to attack by the fungus. T. basicola appears
to be primarily a parasite of leguminous plants, and the common use of
these in rotation may prove to be unprofitable for certain crops. The
benefit derived from the added fertility in the soil may be entirely offset
in certain instances by the injury done as a result of maintaining the
fungus in the soil through the use of host plants in rotation.

From a mycological point of view, considerable interest may also be
attached to the hosts of T. basicola . The earlier botanists who observed
this fungus were in some doubt as to the parasitic nature of the organism.
This view has persisted to some extent, and the fungus is still believed by
some to be purely superficial in its mode of life and not ordinarily injurious
when present. On the other hand, some of the later investigators have
not only convinced themselves of the parasitism of the organism but also
of a high degree of pathogenecity, while some have gone so far as to sup¬
pose the fungus capable of attacking almost any plant under favorable
conditions (n).2 It is fairly clear to the writer that this difference in
opinion is due in a large measure to not only the species but the variety of
host plant under observation. Considerable emphasis has also been laid
upon the conditions necessary for disease to be developed by this fungus,
leading to the supposition that certain unknown special conditions are
necessary and that infection is difficult to obtain. Although this may
be true within certain limits, it has been found that infection could be
repeatedly and easily produced with the more susceptible plants. Where
the writer has experienced difficulty in obtaining infection on certain host
plants, it has been attributed rather to resistance or immunity in the

1 This investigation was carried on in cooperation with the Bureau of Plant Industry, Department of
Agriculture.

2 Reference is made by number to “Literature cited,” p. 299-300.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.
iz

(289)

Vol. VII, No. 6
Nov. 6, 1916
Wis. — 6

290

Journal of Agricultural Research

Vol. VII, No. 6

plant itself than to lack of favorable conditions for fungus attack. Plants
of known susceptibility were at all times grown as controls in the infec¬
tion experiments and served as a good basis for such conclusions.

HISTORICAL REVIEW

Berkeley and Broome (2, p. 461) in 1850 first described the fungus now
known as T. basicola. These authors found it at the base of stems of
Pisum sativum and NemophUa auriculata . The fungus was apparently
not again noted until 1876, when Zopf (23) reported it from Ger¬
many on Senecio elegans and Sorokin (20) from Russia on horse¬
radish (1 Cochlearia armoracia ). Massee (10) described a new fungus in
1884 which he found on decaying leaves of Blysmus compressus and named
it Milowia nivea. This fungus he considers in a later paper (1 1 ) as having
been T. basicola. Prom descriptions and drawings of this earlier de¬
scribed form, however, it may be seen that there is some room for doubt
as to the identity of these two forms.

In 1891 Zopf (24) again published upon the occurrence of Thielavia
basicola and noted that it was especially common on leguminous plants,
adding the following new hosts: Lupinus angustifolius, L. albus , L.
thermis , L. luteus , Trigonella coeruleay and Onobrychis cristargalli. In the
same year Thaxter (22) made the first report of the occurrence of T.
basicola in America, finding it on the violet. Sorauer (19) in 1895
reported the fungus as causing a disease of the roots of the cyclamen.
Peglion (13), working in Italy, in 1897 was first to record the parasitism
of T. basicola on Nicotiana tabacum . Killebrew (8, p. 162), however,
as early as 1884, described the symptoms of a root disease of tobacco in
Pennsylvania, which was undoubtedly due to T. basicola9 though no
causal organism was named in the description. Selby (15, p. 228) has
noted T. basicola as occurring upon the roots of Begonia rubra , following
the nematode disease, and later found it causing a rootrot of Catalpa
speciosa (17, p. 384, 447). Smith (18, p. 35-38) in 1899 added two new
hosts, Gossypium herbaceum and Vigna sinensis. Van Hook (16, p. 96)
was first to note T. basicola upon Panax (Ar alia) quinquefolium. Ader-
hold (1) in 1905 found T. basicola on begonia and carried on infection
experiments with pure cultures of the organism. He obtained slight
infections on Scorzonera hispanica, Daucos carotaf Apium graveolens, and
Beta vulgaris and better development on Lupinus angustifolius and
Phaseolus vulgaris. He concluded from his experiments that T. basicola
was only a weak parasite.

Kirchner (9) includes two previously unreported species as hosts,
Phaseolus multiflorus and Nicotiana rustica. Since no reference to per¬
sonal observation or infection experiments could be found, it is pre¬
sumable that these are not authentic host plants, especially in view of
the facts noted later.

Nov. 6, 1916

Host Plants of Thielavia basicola

291

Gilbert (7) in 1909 added three new hosts: Linaria canadensis , Oxalis
corniculata , var. stricia , and Trifolium repens . Chittenden (5) in 1912
reported T. basicola as a parasite of sweet peas. Massee (1 1) in the same
year added to the list of host plants a species of Cypridpedium, Aster spp.,
and Capsella bursa-pastoris.

Rosenbaum (14) published in 1912 on infection experiments with
T. basicola , and by cross-inoculation experiments found that the species
of Thielavia on tobacco, cotton, and ginseng were identical.

O’Gara (12) in 1915 found T. basicola causing a disease of Citridlus
vulgaris in Utah. This is the first report of the fungus on a member of
the cucurbit family. Recently Burkholder (3) has noted the fungus on
Trifolium pratense , Trtfolium hybridum , and Medicago sativa .

METHOD OF WORK

The investigation of the host plants of 7\ basicola was undertaken
primarily to corroborate, so far as possible* the hosts reported by earlier
investigators, as in the majority of instances this had not been done. As
it was deemed important to know something of the possible relation of
the fungus to our agricultural plants, these were for the most part included
in the earlier tests; but later the trials were made to include as many
species as obtainable of the more susceptible families of plants. In this
way about 200 species of plants have been grown on infected soil. The
work was carried on almost entirely in the greenhouse. The seeds or
plants were sown or transplanted into the infected soil, and optimum
conditions maintained, so far as possible, especially as regarded the mois¬
ture content of the soil. The work was done partly at Arlington, Va., in
cooperation with the Office of Tobacco Investigations, Bureau of Plant
Industry, and partly at the Wisconsin Agricultural Experiment Station
at Madison. In the Arlington greenhouses a fairly heavy clay loam soil
from a tobacco field on the farm was used. This soil was taken from a
spot in the field known to be badly infected with the rootrot caused by
T. basicola . At Madison a greenhouse soil containing considerable vege¬
table matter was infected with soil from a tobacco field, where this
rootrot had been occurring annually in recent years. On the whole,
the conditions were such as might occur outdoors during a season of
high precipitation. Recording soil and air thermometers showed tem¬
peratures ranging ordinarily between 180 and 250 C. in the soil and 20°
to 30° C. in the air. The data were taken largely from infection on
seedlings, as this was naturally an advantage in examining minutely the
root system of a large number of plants. Since the age of the individual
roots is considered a greater factor than that of the entire plant in deter¬
mining the occurrence of infection, it is believed the results would be
comparable if older plants were used.

64310°— 16 - 3

292

Journal of Agricultural Research

Vol. VII, No. 6

At intervals of about five days after the plants were well above the
ground one or more plants were carefully removed from the soil with as
much root system as possible. The soil was then washed off the roots,
after which they were usually placed immediately in 80 per cent alcohol
and transferred to the laboratory, where microscopical examination of any
spots or lesions occurring on the roots were made. Where T . basicola was
found without difficulty, three to five examinations only were made;
but where it could not be found on reported or suspected hosts, the roots
of io to 15 plants were carefully examined, except in a few rare instances
where this number of plants could not be obtained.

EXPERIMENTAL RESULTS

The results obtained are presented largely in the following tables. The
host plants are separated into those reported by earlier workers (Table I)
and the new hosts (Table II). For convenience in reference the authority
for the first report and the country and year in which reported are given.
These columns are followed by the results obtained with the various
plants in the present experiments, giving at the same time the approxi¬
mate degree of susceptibility, as nearly as could be determined by the
amount of infection obtained. In the same way these results are given
for the new hosts reported here. From Table I it may be seen that out of
the 39 host plants previously reported the parasitism of T . basicola on 25
of these has been corroborated. Of the remaining 14 species infection
could not be obtained on 7 species. Seeds or plants of 7 species have
not been secured up to this time, but infection in most of these cases was
obtained upon closely related species, indicating at least that the unob¬
tainable species are for the most part probably susceptible to attack.
Sixty-six new species of plants have been added to those already reported
and corroborated as host plants of T . basicola . As will be seen from the
list, these are largely in the leguminous, solanaceous, and cucurbitaceous
families. Although the plants tested were largely representatives of these
families, a number of species of other families have been included in the
tests, especially species of the Compositae, Gramineae, and Rosaceae. It
is fairly safe to conclude from these tests that the latter families are
generally immune from attack by T . basicola .

Nov. 6, 1916

Host Plants of Thielavia basicola

293

Table) I. — Host plants of Thielavia basicola reported before the present investigation

Host plant.

Authority.

Locality.

Year.

Susceptibility.

England .

1850

Tow.

Germany .

1891

Medium.

1891

Do.

1891

Plants not
obtained.

1891

Do.

1891

Slight.

1891

Medium.

United States . .

1899

Do.

Germany .

1905

Low.

1906

None.

United States . .

1909

Low.

England .

1911

Slight.

United States . .

1916

Low.

1916

Do.

1916

Do.

Italy .

1897

High; low.

Germany .

1906

None.

United States . .

I9I5

Low.

1891

High.

1904

Medium.

1899

Do.

England .

1850

Plants not
obtained.

United States . .

x9°9

Low.

. do .

1896

Plants not
obtained.

Germany .

J905

Do.

United States . .

1909

Slight.

1910

Do.

Germany .

1876

Medium.

England .

1912

Low.

Germany .

1905

None.

United States . .

19H

Do.

Russia .

1876

Slight.

England .

19x2

1912

Plants not
obtained.

Germany. .....

189s

Slight.

England .

1912

Plants not
obtained.

Germany .

i9°s

None.

i9°s

Do.

190s

Do.

1. Leguminosae:
Pisum sativum ,

Onobrychis crista -
galli .

Trigonella coerulea . .
Lupinus angustifol-
tus .

Lupinus thermis ....

Lupinus albus .

Lupinus luteus .

Vigna sinensis .

Phaseolus vulgaris.
Phaseolus multiflorus

Trifolium repens .

Lathyrus odoratus . . .

Medicago sativa .

Trifolium pratense. . .
Tnfolium hybridum.

2. Solanaceae:

Nicotiana tabacum . .
Nicotiana rustica . . . .

3. Cucurbitaceae:

Citrullus vulgaris. . .

4. Miscellaneous fami¬

lies:

Viola odorata .

Aralia quinquefolia. .
Gossypium herbaceum
Nemopkila auriculata

Linaria canadensis. . .
Begonia rubra .

Berkeley and
Broome (2).

Zopf (24) .

do.

Begonia (tuberhy-
bridaf).

Oxalis corniculata,
var. stricta .

Catalpa speciosa .

Senecio elegans .

Aster sp .

Scorzonera hispanica .

Pastinica sativa .

Cochlearia armoracia .
Capsella bursa-pas-
toris.

Cypripedium sp .

Cyclamen sp .

Blysmus compressus .

Apium graveolens . . .

Daucus carota .

Beta vulgaris .

- do .

. do .

_ do .

Smith (18) .

Aderhold (1). ..
Kirchner (9). . .

Gilbert (7) .

Chittenden (4).
Burkholder (3).

_ do .

Peglion(i3).
Kirchner (9).

O’Gara (12) .

Thaxter (22) .

Van Hook (16) .

Smith (18) .

Berkeley and
Broome (2).

Gilbert (7) .

Sett»y (15) .

Aderhold (i).
Gilbert (7). . .
Selby (17). .

- •

Zopf (23).

Massee (11) .

Aderhold (1) .

Taubenhaus (21) . .

Sorokin (20) .

Massee (n) _

do.

Sorauer (19).
Massee (11) .

Aderhold (1).

. do .

. do......

294

Journal of Agricultural Research

Vol. VII, No. 6

Table II. — New host plants of Thielavia hasicola

Host plant.

Suscepti¬

bility.

Host plant.

Suscepti¬

bility.

i. Beguminosae :

A rachis hypogaea .

Astragalus sinicus .

Cassia chamaecrista .

Cytisus scoparius .

Desmodium tortuosum . .

Dolichos lab lab .

Galactia sp .

Glycine kispida . .

Lens esculenta .

Lespedeza striata .

Lotus corniculatus .

Lotus villostis .

Lupinus hirsutus . .....
Medicago denticulata. .. .

Melilotus alba .

Melilotus indica .

Ornithopsus sativus .

Onobrychis viciaefolia. . .
Pkaseolus acutifolius . . .
Robinia pseudoacacia . . .

Sclotis chinensis .

Strophostyles helvola. . . .
Trif o Hum incar natum . .

T rigonella f oenum-graecum
Tephrosia virginiana. , . .

Ulex europaeus .

Vida villosa .

Host plant.

Medium*

Bo. 2.
Bo.

Bow.

Slight.

Bo.

Bow.

Bo.

Medium.

Bow.

Bo.

Medium.

Bow.

Bo.

Medium.

Bow.

Slight.

Medium.

Bo.

Bow.

Medium.

Bo.

Bow.

Bo.

Slight.

Bo.

Suscepti¬

bility.

Beguminosae — Continued:

Viciafaba .

Solanaceae:

Datura metel .

Datura stramonium .

Datura tatula .

Datura cornucopia .

Nicoiiana glauca (Gratz). .
Nicotiana silvestris (Speg.)
Nicoiiana sanderae (San¬
der.)

Nicoiiana repanda (Willd. )
Nicotiana atropurpurea . . .
Nicotiana lan gsdorffii
(Weinm.)

Nicotiana chinensis (Fisch)
Nicoiiana macro phylla
(Behm.)

Nicotiana glutinosa (Binn.)
Nicotiana calydfloraa . . . .
Nicotiana latterrima (Mill.)
Nicotiana alta (Bink and
Otto).

Nicotiana angusiifolia
(R. and P.)

Nicotiana longiflora (Cav.)

Petunia ( hybridia ?) .

Solanum carolinense .

Host plant.

Slight.

Medium.

Bow.

Slight.

Bo.

Bow.

Medium.

Bow.

Bo.

High.

Bow.

High.

Bow.

High.

Medium.

High.

Medium.

Bo.

Bow.

Slight.

Bo.

Suscepti¬

bility.

3. Cucurbitaceae:

Cucurbita maxima. . . .

Cucurbita pepo .

Cucumis melo . ......

Cucumis sativus .

Cucurbita moschata . .
Cucumis acutangulus.
* Cucumis flexuosus ... .

4. Miscellaneous families:

Viola tricolor .

Begonia semperflorens .

Bow.

Bo.

Slight.

Bow.

Bo.

Slight.

Bo.

. Miscellaneous families —
Continued:

Nemophilia aurita .

Nemophilia insignis .

Linaria cymbalaria .

Linaria maroccana .

Paphiopedilum grossianum

Portulaca oleracea .

Ipomoea coccinea .

Phlox drummondii .

Papaver nudicaule .

Medium.

Bow.

Bo.

Slight.

Bo.

Bow.

•Slight.

Bo.

a The validity of these species may be questioned.

Negative results have also been obtained with a number of species in
the most susceptible families. In the legume family these apparently
immune species may be represented by Phaseolus multiflorusy Vicia sativa,
and Hedysarum coronarium. In the Solanaceae there are varieties at
least of Capsicum annuum , Solanum melongena , Solanum tuberosum,
Hyoscyamus niger, and certain species of Nicotiana which are apparently
immune to attack. Benincasa cerifera is the only species of the cucurbit
family tested, which appears to be entirely immune from infection by
Thielavia hasicola when grown in infected soil. The common agricultural
plants upon which infection could not be obtained were principally the

Nov. 6, 1916

Host Plants of Thielavia basicola

295

cereals — wheat (Triiicum spp.), oats (Avena sativa) , barley ( Hordeum
spp.), and rye (Secale cereale) — com (Zea mays), potatoes (Solanum
tuberosum), hemp (1 Cannabis sativa), flax (Linum usitatissimum), and
sweet potatoes ( Ipomoea batatas ). With the vegetables the cabbage
( Brassica oleracea), onion (Allium cepa), parsnip (Pastinaca sativa),
carrot (Daucus carota), beet (Beta vulgaris), lettuce (Lactuca sativa) ,
eggplant (Solanum melongena), and peppers (Capsicum annuum) appeared
to be free from attack by the fungus; and of the fruits no infection could
be obtained on strawberries (Fragaria spp.), raspberries (Rubus spp.), or
blackberries (Rubus spp.).

Infection could not be obtained upon the following plants, which
were previously recorded by others as being attacked by T . basicola:
Nicotiana rustica, Phaseolus multiflorus, Pastinica sativa , Scorzonera
hispanica, Daucus carota , Apium graveolens, and Beta vulgaris . The
failure to obtain infection on these hosts may be attributed to one or more
of several obscure causes. The writer is inclined to believe, however,
that these species which have been mentioned as hosts of T. basicola
should not be included in the list of host plants. There is a small proba¬
bility of immune varieties or strains being used in the studies reported
here. In an attempt to clear up this point different varieties of some of
these species were used, but only negative results in infection were
obtained.

Phaseolus multiflorus and Nicotiana rustica are included as hosts of
T. basicola by Kirchner (9). A search through the literature has failed
to reveal earlier reports of these species as hosts. It is evident that
Kirchner included these species because of their relation to known
susceptible species, as no mention of any original observations of his
own could be found. This conclusion is further substantiated by the
fact that after repeated tests by the writer, these species appear to be
immune to attack.

Aderhold obtained very slight infections on Scorzonera hispanica ,
Daucus carota , Apium graveolens , and Beta vulgaris . His results were
obtained, however, under quite artificial conditions and no doubt
resulted from one form or other of injury to the roots. The conclusion
was drawn that T. basicola was only a very weak parasite. All attempts
to infect the above-named species, under the conditions of these experi¬
ments, show that these forms are immune or at least extremely resistant
to attack by T. basicola . Aderhold’s conclusion (1, p. 465) —

Der Verlauf der Erscheinung zeigte, dass der Pilz kein heftiger Parasit ist. Ich
schliesse mich auf Grund dieser Versuche der Auffassung Sorauers an, der Zufolge
besondere Verhaltnisse geboten sein mtissen, tun ihn zu einen wirklichen Schadiger
zu machen —

should, therefore, certainly not be taken to apply to other host plants of
T, basicola .

296

Journal of Agricultural Research

Vol. VII, No. 6

Massee (11) includes Blysmus compressus as a host of T. basicola,
apparently on the basis of an early report (10) of MUowia nivea on
decaying leaves of B. compressus. Unless this observation was corrobo¬
rated by later observation, the writer believes some doubt may be
entertained in regard to this species as a host, on account of the point of
attack and question of the identity of this fungus with T. basicola
Zopf. The description and figures in Massee's first article bear only a
slight resemblance to T. basicola . Unfortunately the writer has been
unable to obtain this species of Blysmus for trial in these experiments.

It is also interesting to note that certain species commonly cited as
hosts of T. basicola are strikingly resistant or practically immune.
Chittenden (4-5), Massee (n), and Taubenhaus (21) report T. basicola
as a serious disease of sweet peas (Lathyrus odoraius ), though they differ
in their opinions as to the symptoms of the disease. Infection of sweet
peas in the writer's experiments was obtained only with great difficulty
and then only slightly. According to Taubenhaus, the organisms causing
this infection of sweet peas and tobacco are interchangeable and no
physiological race difference exists in the fungus. It seems plausible,
therefore, that more resistant varieties of L. odoraius were used in the
present trials than were used by Chittenden and Taubenhaus. It may
be, therefore, that the sweet-pea disease may be controlled by selection
for disease resistance. Taubenhaus also mentioned obtaining a culture
of T. basicola from parsnip, although he does not include it in his list of
hosts. No infection could be obtained upon that vegetable in the
present trials.

Aderhold (1) failed to get infection on Begonia semperflorensy and it
was only after repeated examinations that the writer found the fungus
on this host and then only on nematode galls, as reported by Selby for
Begonia rubra . Cochlearia armoracia , Cyclamen spp., Pisum sativum ,
Lupinus albuSy Catalpa speciosa , and certain species of orchids were
found to be very difficult to infect.

Plants upon which very slight infections were secured in one or two
instances but are not included in the new list as hosts are Lycopersicon
esculentum, Tropaeolum majus , Fagopyrum esculentum , and Solanum
nigrum.

As a result of these studies, it is concluded, therefore, that certain
members of the following families of plants are likely to be attacked
by Thielavia basicola under favorable conditions for the development
of the fungus: Araliaceae, Bignoniaceae, Compositae, Convolvulaceae,
Crucifereae, Cucurbitaceae, Hydrophyllaceae, Ueguminosae, Malvaceae,
Orchidacae, Oxalidaceae, Papaveraceae, Polemoniaceae, Portulaceae,
Primulaceae, Scrophulariaceae, Solanaceae, and Violaceae.

The occurrence of T. basicola on the various hosts studied differed
principally in two respects: (1) The point of attack by the fungus and
(2) the character of the sporulation. The occurrence of T. basicola in

Nov. 6, 1916

Host Plants of Thielavia basicola

297

certain cases on stems above the surface of the ground has been noted
by others (11, 14), but is relatively rarely found. The lesions obtained
on Portulaca oleracea were almost wholly on the low succulent stems,
apparently irrespective of any infection at the base of the stem. Ordi¬
narily, however, infection occurs only on the roots of the host plants or
upon the base of the stem just at or below the surface of the soil (PL 18).
In the case of Nicotiana spp., infection is ordinarily found on the sec¬
ondary roots, although in some species the collar of the plants may be
most markedly injured, whereas in the cucurbits the infection is usually
on the stem just at or below the surface of the soil (PL 18, fig. A). In
species of Pisum, Phaseolus, and Lupinus the fungus was ordinarily
found first upon the base of the stem or on the primary root (Pl. 18,
fig. C). On pansy ( Viola tricolor) and phlox, where only very slight
infections occurred, they were invariably found at the ends of the smallest
fibrous roots.

The character of sporulation seemingly differs mostly in the variation
of the appearance of the perithecial stage. Many workers have expe¬
rienced failure or considerable difficulty in locating this stage and have
therefore questioned the connection of the perithecia as described by
Zopf with the chlamydospores of T. basicola . The association of the
perithecia upon a large number of the different host plants observed in
these tests with the chlamydospore stage of T . basicola is fairly con¬
vincing as to the correctness of Zopf’s conclusions. The perithecia
were found to be especially abundant upon Cucumis maxima , Robinia
pseudoacacia , Cytisus scoporius , Nicotiana tabacum , and to a lesser
extent upon a number of other hosts. Although it can not be so stated
with certainty, the perithecial stage is apparently never produced in
the same way on some host plants, as it is lacking in pure cultures of
the fungus. The size, shape, number, and color of chlamydospores
produced upon the various hosts differed to some extent. These differ¬
ences appear to be determined in part by the location of these spore
chains. When formed within the host cells, as is common in certain
hosts, they were often restricted in their growth and were malformed.
In other hosts such as Cucumis spp. and Linaria spp. they are com¬
monly formed outside of the host cells and are large and uniformly
shaped. The color of the chlamydospores varies from a deep blue-black,
as on species of Cucumis, to a light brown, as on species of Nicotiana.
The conidial spore form is only rarely seen on the living host, although
it is produced early and profusely in culture media.

The addition of a large number of new host plants of T. basicola ,
while working with a comparatively small number of species, is taken to
mean that the range of host species may perhaps be again doubled in
time. Although its pathogenicity was questioned by earlier workers, it
is shown that these conclusions were drawn from limited data and that
these investigators used species which were either immune or relatively

298

Journal of Agricultural Research

Vol. VII, No. 6

resistant to attack. The vigorous pathogenicity of this organism on
the more susceptible species and varieties can not be questioned. The
vigor of the attack on certain susceptible varieties of tobacco at least is
as striking as that of any known root disease of plants (PI. 19). Since the
fungus apparently spreads relatively slowly through the cell tissue,
however, it seems apparent that soils must be abundantly infected and
that a large number of local infections must occur before plants become
badly diseased. Considering the wide range of hosts possessed by this
fungus, it is possible that large economic losses, which are as yet rela¬
tively unknown, may be due to this fungus. In the case of tobacco
alone it is certain that an average annual loss amounting to several
millions of dollars occurs. Large losses also occur in the culture of
violets (Viola spp.) and ginseng ( Panax quinque folium). Little or
nothing is known about the extent of the damage which may be done to
the various cultivated leguminous crops under field conditions. It is
believed that the fungus will become a serious disease of the peanut
(Arachis hypogaea) and cotton (Gossypium spp.), if it has not already
become one.

SUMMARY

(1) Thielavia basicola Zopf is a fungus parasite attacking primarily
members of the Leguminosae, Solanaceae, and Cucurbitaceae. Other
families containing hosts of this fungus are Araliaceae, Bignoniaceae,
Compositae, Convulvulaceae, Cruciferae, Hydrophyllaceae, Malvaceae,
Orchidaceae, Oxalidaceae, Papaveraceae, Polemoniaceae, Portulaceae,
Primulaceae, Scrophulariaceae, and Violaceae.

(2) Infection could not be obtained upon the following species of
plants reported by others as hosts of T. basicola: Phaseolus midtiflorus ,
Nicoiiana rustica , Scorzonera hispanica, Daucus carota , Apium graveolens ,
Beta vulgaris , and Pastinica saliva.

(3) Thirty-nine species of plants have been reported by earlier investi¬
gators as hosts of T. basicola . Thirty-two of the thirty-nine reported hosts
plants have been grown in soil infected with T. basicola , and infection
obtained upon twenty-five of these plants. Of the seven upon which
negative results in infection were secured, it is believed that all should
be excluded from the list of hosts until further corroboratory evidence
of infection is obtained. The remaining seven species could not be tested,
owing to the difficulty of getting seeds or plants.

(4) Sixty-six new species of plants are added as hosts of T. basicola ,
of which twenty-eight are legumes, twenty are solanaceous plants, seven
are cucurbits, and eleven belong to miscellaneous families.

(5) A great difference in the susceptibility of the various species
exists; and where earlier workers have been inclined to doubt the
parasitism of T. basicola , it appears to have been due to the fact that
infection experiments were carried on with what are now known to be
immune or very resistant plants.

Nov. 6, 1916

Host Plants of T hielavia basicola

299

(6) Some differences in point of attack and character of sporulation
of T. basicola on different hosts have been noted. The common occur¬
rence of the perithecial stage in close association with the chlamy-
dospores of a number on different hosts is taken as good indirect cor¬
roboratory evidence of Zopf ’s connection of this form with T. basicola .

(7) The infection of nearly 100 different species of plants with T.
basicola from tobacco is further evidence that no specialized races of
this fungus appear to exist.

LITERATURE CITED

(1) Aderhold, Rudolf.

1965. Impfversuche mit Thielavia basicola Zopf. In Arb. K. Gsndhtsamt., Biol.
Abt., Bd. 4, Heft 5, p. 463-465.

(2) Berkeley, M. J., and Broome, C. E.

1850. Notices of British fungi. In Ann. and Mag. Nat. Hist., s. 2, v. 5, p. 455-466,
pi. 11-12.

(3) Burkholder, W. H.

1916. Some root diseases of the bean. (Abstract.) In Phytopathology, v. 6, no.
1, p. 104.

(4) Chittenden, F. J.

[1911.] Report [on investigations of sweet pea diseases]. In Sweet Pea Ann. 1911,
P* 35-39*

(s) -

1912. On some plant diseases new to, or little known in Britain. In Jour. Roy.
Hort. Soc. [London], v. 37, pt. 3, p. 541-55°*

(6) Clinton, G. P.

1907. Root rot of tobacco, Thielavia basicola, (B. & Br.) Zopf. In Conn. Agr. Exp.
Sta. 30th Ann. Rpt. [1905] /06, p. 342-368, fig. 14, pi. 29-32.

(7) Gilbert, W. W.

1909. The root-rot of tobacco caused by Thielavia basicola. U. S. Dept. Agr. Bur.
Plant Indus. Bui. 158, 55 p., 5 pi. Bibliography, p. 44-48.

(8) KjllEbrew, J. B.

1884. Report on the Culture and Curing of Tobacco in the United States. 286 p.,
illus. Washington, D. C. Published by U. S. Dept. Int. Census Office.

(9) Kirchner, Oskar.

1906. Die Krankheiten und Beschadigungen unserer landwirtschaftlichen Kultur-
pflanzen ... Aufl. 2, 675 p. Stuttgart.

(10) MassEE, G. E.

1884. Description and life-history of a new fungus, Milowia nivea. In Jour. Roy.
Micros. Soc. [London], s. 2, v. 4, p. 841-845.

(11) —

1912. A disease of sweet peas, asters, and other plants. In Roy. Gard. Kew, Bui.
Misc. Inform., 1912, no. 1, p. 44-52.

(12) O’Gara, P. J.

1915. Occurrence of Thielavia basicola as a root parasite of watermelons in the
Salt Lake Valley of Utah. In Science, n. s. v. 42, no. 1079, p. 314.

(13) Peglion, Vittorio.

1897. Marciume radicale delle piantine di tabacco causato dalla Thielavia basicola,
Zopf. In Centbl. Bakt. [etc.], Abt. 2, Bd. 3, p. 580-584.

(14) Rosenbaum, Joseph.

1912. Infection experiments with Thielavia basicola on ginseng. In Phytopath¬
ology, v. 2, no. 5, p. 191-196, pi. 18-19.

300

Journal of Agricultural Research

Vol. VII, No. 6

(is) Si&by, A. D.

1896. Investigations of plant diseases in forcing house and garden. Ohio Agr.
Exp. Sta. Bui. 73, p. 221-246, illus.

(16) -

1904. Tobacco diseases and tobacco breeding. Ohio Agr. Exp. Sta. Bui. 156, p.
87-114, 3 fig- » 8 pi.

(17) -

1910. A brief handbook of the diseases of cultivated plants in Ohio. Ohio Agr.
Exp. Sta. Bui. 214, p. 307-456. Literature referred to, 7 p. at end of
bulletin.

(18) Smith, Erwin F.

1899. Wilt disease of cotton, watermelon, and cowpea (Neocosmospora nov. gen.)
U. S. Dept. Agr. Div. Veg. Physiol, and Path. Bui. 17, 72 p., 10 pi. (1 col.).

(19) SorauBR, Paul.

1895. iiber die Wurzelbraune der Cyclamen. In Ztschr. Pflanzenkrank., Bd. 5,
Heft 1, p. 18-20.

(20) Sorokin, Nicolas.

1876. Ueber Helminthosporium fragile sp. n. In Hedwigia, Bd. 15, No. 8, p. 113.

(21) Taubenhaus, J. J.

1914. The diseases of the sweet pea. Del. Agr. Exp. Sta. Bui. 106, 93 p., tab., 43
fig. References, p. 88-93.

(22) Thaxter, Roland.

1892. Fungus in violet roots. In Conn. Agr. Exp. Sta. Ann. Rpt. 1891, p. 166^167.

(23) ZopR, Wilhelm.

1876. [Ueber Thielavia basicola, einen endophytischen Parasiten in den Wurzeln
des Senecio elegans.] In Verhandl. Bot. Ver. Brandb., Jahrg. 18, p.
101-105.

(24) — rr

1891. Uber die Wurzelbraune der Lupinen, eine neue Pilzkrankheit. In Ztschr.
Pflanzenkrank., Bd. 1, p. 72-76, illus.

PLATE 18

Fairly typical diseased spots and lesions caused by Thielavia basicola on various
host plants. Some of the infected areas are indicated by arrows and braces:

Fig. A. — Citrullus vulgaris (citron).

Fig. B. — Onobrychis viciaefolia (sainfoin).

Fig. C. — Lupinus luteus (yellow lupine).

Fig. D. — Arachis hypogaea (peanut).

Fig. E. — Robinia pseudoacacia (black locust).

Fig. F. — Sclotis chinensis (wistaria).

PLATE 19

Fig. A. — Part of a field infected with Thielavia basicola in foreground, with newer
soil planted to tobacco in the background, illustrating the marked pathogenic
powers of this organism.

Fig. B. — A tobacco plant showing diseased roots from infected soil.

Fig. C. — Healthy roots from uninfected soil of a semiresistant type of tobacco (Con¬
necticut Havana). Figures B and C show the relative growth of plants and amount
of root system after equal care in removing roots from the soil.

DEPARTMENT OF AGRICULTURE

Voiv. VII Washington, D. C., November 13, 1916 No. 7

CHEMICAL COMPOSITION, DIGESTIBILITY, AND FEED¬
ING VALUE OF VEGETABLE-IVORY MEAL1

By C. L. Beals and J. B. Lindsey,

Massachusetts Agricultural Experiment Station

INTRODUCTION

Vegetable ivory, or the corozo nut, as it is commonly known in com¬
merce, is the seed or nut of the palmlike plant Phytelephas macrocarpa. It
is a native of the Latin American countries, being found in great quanti¬
ties along the banks of the Magdalena, in Colombia, where it is known as
“tagua.” It is also found in Peru and in the forests of northern Ecua¬
dor. In appearance the plant itself is a stemless palm, bearing its fruit
in conglomerate heads, often weighing 25 to 30 pounds apiece. These
heads are made up of 30 to 50 seeds or nuts, varying in size from half
an inch to several inches in diameter. In the earlier stages of growth
the seed contains a clear, insipid liquid, which later changes to a sweet,
milky paste, and finally hardens into the white horny substance from
which it derives its name “vegetable ivory.”

Large quantities of the nuts are imported annually by Great Britain
and Germany, principally for the manufacture of buttons. The United
States uses about 10,000 tons annually, costing $1,500,000 (1, p. 200). 2
Beneath the brown outer coating the dried nut has the appearance of
dentine ivory, and can easily be sawed, carved, and turned into all sizes
and shapes of buttons, while the texture is such that it readily absorbs
dyes and will take a high polish.

In the process of manufacture a considerable portion of the nut is
wasted in the form of sawdust, chips, and turnings. In foreign countries
this waste has been mixed with other ingredients to be used as a cattle
food. German writers state that vegetable-ivory meal has been used

1 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 to “ Literature cited,” p. 320.

(301)

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C,
gb

Vol. VII, No. 7
Nov. 13, 1916
Mass.— -2

3°2

Journal of Agricultural Research

Vol. VII, No. 7

as an adulterant in the manufacture of so-called concentrated feeds.
Some instances are given where as much as 50 per cent of this material has
been added.

In the last few years considerable attention has been attracted in this
country to the enormous amount of waste material produced by ivory-
button factories, and many attempts have been made to discover a
practical use for the material aside from fuel.

The material experimented with had the appearance of a mediufn-fine
meal, white in color, though flecked here and there with particles of the
brown outer coating of the nut. It was tasteless, odorless, and very hard,
being almost gritty to the touch.

CHEMICAL COMPOSITION OF VEGETABLE-IVORY MEAL

Considerable work has been done on the chemistry of vegetable ivory
both in this country and in Europe. Tollens (14, p. 113), Fischer and
Hirschberger (7), Liebscher (11), and others have worked with ivory
meal as a source of mannose and have otherwise investigated the material,

A considerable amount of corroborative chemical investigation of vege¬
table ivory has been undertaken and an attempt made to determine
quantitatively the mannose present. The results are presented with as
little detail as possible, the fodder analyses being given in Table I.

Tabi<3 I. — Fodder analyses of vegetable ivory

Constituent.

Maximum.

Minimum.

Average.®

German anal¬
yses for com¬
parison.

Moisture .

12. 64
2. 30
5. 26

I. 18

7-75
77- 56

6. 13
.80

3- 94
. 60
6. 13
74- 17

H- 39

I. 08
4-63
.92

6. 89
75-«>9

18. 30-13. 20
I. 30- I. 10
4. 60- 4. 00
I. IO- O. 80

I79. 80-75. So

Ash .

Protein .

Fat .

Fiber .

Nitrogen-free extract .

a Average of nine samples.

At a glance Table I shows not only the variations met with in different
samples but that by far the greater part of the material is carbohydrate
in nature. The protein rarely exceeds 5 per cent and was found to con¬
tain about one-third of its nitrogen in the amido form. The fat or ether
extract had the appearance of a heavy light-colored oil and possessed a
pleasant nutty odor.

The fiber in all cases was fairly uniform in amount, being about 7 per
cent of the dry matter. It was noticed while making the determinations
that the vegetable ivory acted as an indicator, the change from light
buff with acid to a deep wine color with alkali being quite abrupt. Both
the residue from the fiber determinations and the original material were

Nov. 13, 1916

Vegetable-Ivory Meal

303

tested for the presence of lignin, but neither phloroglucinol nor anilin
sulphate produced any color reaction whatever. As three-fourths of the
vegetable ivory was found to be nitrogeh-free extract, it was to this
portion that the most attention was given.

It has long been known that the greater part of the carbohydrate
material consists of mannose, or, more accurately speaking, mannan, its
anhydrid condensation product.

The isolation of mannose was carried out practically as described by
Fischer and Hirschberger (7). One hundred gm. of vegetable-ivory meal
were digested on the water bath with reflux condenser for six hours with
200 c. c. of 6 per cent hydrochloric acid. The liquid was then filtered
off, the filtrate and washings neutralized with sodium hydroxid, and
shaken out several times with carbon black. After filtration, phenyl-
hydrazin (dissolved in acetic acid) was added at the rate of 0.3 gm. for
every gram of ivory meal used. The mannose phenylhydrazone sepa¬
rated out on standing for 24 hours in the cold as a heavy, fine-grained,
buff-colored precipitate. This was washed with cold water and dried
in a vacuum at room temperature. Particles of this impure hydrazone
when placed in a capillary tube and heated slowly in a sulphuric-acid
bath melted at 183° C.

A portion of the precipitate was purified by boiling for a long time with
a large volume of 95 per cent alcohol, filtering, and again boiling with
fresh alcohol until at the end of two days an almost snow-white hydrazone
resulted. This melted at 196° C., demonstrating the existence of man¬
nose or its polymer mannan in vegetable ivory.

To liberate mannose from its phenylhydrazone, a portion of the latter
was digested with benzaldehyde and alcohol until crystals of benzaldehyde
hydrazone formed. The mannose containing filtrate from these, after
clarifying and evaporating to a sirup, was treated with absolute alcohol
and set aside to crystallize. The mannose crystals obtained had a melting
point of 1320 C.

Pentosans were determined by the hydrochloric-acid distillation
method and precipitation with phloroglucinol. The average of three
determinations was 2.43 per cent of the dry matter.

Repeated attempts to produce mucic acid by oxidation with nitric
acid proved futile. The exact method for the detection and determi¬
nation of galactan was carried out always with negative results.

Microscopic examination with iodin failed to give the slightest evidence
of starch, either in the white fleshy part of the nut or in the brown outer
coating.

Dextrose (or dextran) was shown to be absent in vegetable ivory by
its inability to form saccharic acid. As a check on the method used, a
sample of pure glucose was treated exactly as was the ivory meal. No
difficulty was experienced in obtaining the saccharic acid from the check.

304

Journal of Agricultural Research

Vol. VII, No. 7

It was found by boiling a little of the vegetable-ivory meal in water,
filtering out the insoluble portion, and then adding the dear filtrate
to a large volume of strong alcohol that a predpitate would form after
standing for some time. Since this process is similar to that employed
in separating pectin (plant mucilage) from fruits, it was first supposed to
be the same product. The amount present was found to be 2.78 per
cent on a dry-matter basis.

According to the best authorities, pectin, supposed by many to be an
oxygen or add derivative of cellulose, is readily oxidized to mucic acid
by proper treatment with nitric acid. The product from vegetable-ivory
meal when so treated produced no mudc acid. That it could not be of
pentose character was demonstrated by making determinations on the
filtered predpitate. Not the slightest trace of phlorogludd formed,
showing the absence of five carbon sugars.

From various authorities and from actual observation, pectin derived
from fruit is known to reduce Fehling's solution. Indications of such
reduction were not noted in the case of this precipitate.

On the supposition that it might be of a nitrogenous nature, a nitrogen
determination was made with negative results.

Considering these results, it would seem that the alcoholic precipitate
from vegetable-ivory meal is distinctly different from the so-called
“plant mudlage.”

In attempting the determination of the sugars present in vegetable-
ivory meal by Fehling’s gravimetric method, many difficulties were
encountered. However, a brief summary of the results obtained seems
worthy of note. Water extracts of the material without inversion
gave about 0.5 per cent of reducing material. The same solution after
hydrolysis with hydrochloric add at 20° C. for 24 hours gave an average
of 2 per cent of redudng material. From this it was evident that the
mannose, or, more properly speaking, the mannan, existed as a hemicellu-
lose, since otherwise the total sugars would have been in an enormous
excess of 2 per cent. Consequently the hydrolysis was made more drastic
by boiling, and it was found that with an increase in the length of the
boiling period the percentages of sugars increased (Table II).

Table II. — Relation of the length of the boiling period to the percentage of sugars in

vegetable-ivory meal

Hours boiled.

Percentage of
sugar (as
dextrose).

2K . - .

A .

47* 40
65.00
73-40
73*40
73*40

t . . . . .

6 .

7 . . .

Nov. 13, 1916

Vegetable-Ivory Meal

305

It was noticed that five hours’ boiling in an acid solution was necessary
to hydrolyze completely the mannose and other reducing materials and
that more than five hours’ boiling produced no increased percentage.

The percentages were calculated not as mannose but as dextrose,
since no table for the determination of mannose by Fehling’s method
has been found.

The pentosans present no doubt had a somewhat different reducing
capacity than the mannan. However, when appropriate allowance had
been made for moisture and pectin (previously determined), the total
carbohydrates estimated in this fashion approached to within less than
1 per cent of the amount estimated as nitrogen-free extract in the original
fodder analysis. It is evident, therefore, that the so-called nitrogen-
free extract, comprising fully 75 per cent of the vegetable-ivory meal,
was composed principally of mannan with small amounts of pentosans
and of a substance insoluble in alcohol but not identical with the pectin
substances as usually found in plants.

CALORIFIC VALUE OF VEGETABLE-IVORY MEAL

To determine the calorific value of this substance a number of bomb-
calorimeter determinations were made, the average of which is given in
Table III, together with representative figures for other common sub¬
stances used as food.

Table III. — Comparative calorific values of vegetable-ivory mealt corn meal, sugar, and

cornstarch

Material.

Small calories,
per gram.

Large calories,
per pound.

Vegetable-ivory meal .

3,78s

3,549

3,958

3,692

h 717
1, 610

I» 753

1,675

Com meal (0, p. 401;, 420)1 .

Sugar (guaranteed) .

Cornstarch (q. p. aok. 420) . .

W* UWVVM \y 1 I J * T V / . . tttt*..

1 H. P. Armsby (2, p. 13) reports com meal as having a chemical energy of 170.9 therms per 100 pounds,
the equivalent of 3,766 small calories per gram, or 1,709 large calories per pound.

In button factories, where the largest amount of ivory waste, or meal,
is produced, the material is used under the boilers as fuel, and it has been
authoritatively stated1 that it produces about half as much heat as
coal. It is interesting to note how accurately this statement is borne
out scientifically. The average of 20 samples of soft coal recently ana¬
lyzed at this station was 14,074 B. T. U., which, expressed in large calories
per pound, equals 3,546. This figure is approximately twice that of
the vegetable-ivory meal, 1,717.

1 Courtesy of Mr. C. J. Spill, Superintendent of the United Button Co., Springfield, Mass.

3°6

Journal of Agricultural Research

Vol. VII, No. 7

DIGESTION EXPERIMENTS

Experiment I. — The determination of digestibility was carried out
with two sheep in the usual manner (12). The sheep were fed a ration
consisting of 500 gm. of English hay, 150 gm. of gluten feed, 200 gm. of
finely ground vegetable-ivory meal, and 10 gm. of salt with water ad
libitum . Rations for the entire test were weighed out at the beginning
and samples sent to the laboratory and immediately analyzed. The
results are given in Table IV.

Table IV. — Percentage composition of feeds tuffs used in Experiment I

MOISTURE

Constituent.

Hay.

Gluten feed.

Vegetable-
ivory meal.

Moisture . .

95
88. 05

10. 07 ■

89- 93

12. 64
87. 36

Dry matter . .

Total .

IOO. 00

DRY MATTER

Ash . .

5. 81

9- 34
2. 88
31* 7o
50-78

1. 11
26. 96

3- 94
8. 70

59- 29

i-37

6. 02
.67

7. 02
84. 90

Protein .

Fat . .

Fiber .

Nitrogen-free extract . .

Total .

IOO. 00

The experiment lasted 14 days, 7 of which were preliminary; and
during that time no disturbances in digestion were observed. The
sheep ate the vegetable-ivory mixture readily.

One-tenth of the daily manure excreted by each sheep was carefully
dried and preserved. Eater these portions were composited and analyzed,
the results being given in Tables V and VI.

Nov. 13, 1916

Vegetable-Ivory Meal

307

Table V. — Quantities of manure and urine excreted and water consumed daily by sheep
fed vegetable-ivory meal , gluten feed, and hay

SHEEP 5

Date.

Manure.

Urine.

Total

weight.

Weight
of the
one-tenth
preserved.

Air-dry

weight.

Weight.®

Nitrogen
in urine.

Nov. 16 .

Nov. 17 .

Nov. 18 . — .

Nov. 19 .

Nov. 20 .

Nov. 21.. .

Nov. 22 .

Average .

Gm.

711

626

684

476

536
61 5

Gm.

71. I
62. 6

68.4
47.6
47.6
53-6
61. 5

Gm.

29. 25
28. 20
28. 63
21. 20
20. 45
23-37
25-37

Gm.

2,15!

*>75
1,397
2, 201

I, 135

i,459

914

Per cent,

0. 48

■ 83
.58
' • 64
.68

• 57
.80

589

S8-9

25. 21

1,461

•6S

SHEEP 6

Nov. 16 .

535

53- 5

25. 21

621

i- 73

Nov. 17. . .

527

52. 7

25- 45

665

i- 34

Nov. 18 .

495

49-5

23. 21

681

1. 25

Nov. 19 .

483

48.3

23- 39

547

1. 42

Nov. 20 .

545

54- 5

25. 89

643

1. 40

Nov. 21 .

407

40. 7

19- 93

543

1. 08

Nov. 22 .

512

51.2

26. 28

647

i- 3i

Average .

507

50- 7

24. 19

621

1. 36

a The quantity of urine in each case was increased by 100 c. c. of carbolic disinfectant and wash water
used at the bam.

Table VI. — Percentage composition of feces of sheep fed vegetable-ivory meal , gluten

feed, and hay

MOISTURE

Constituent.

Sheep 5.

Sheep 6.

Moisture .

5- 70
94- 30

5- 50
94- 50

Dry matter .

Total . '. . . .

100. 00

DRY MATTER

Ash .

9. 16
15. 60

3-96
25.96
45- 32

9. 80
14. 70
3-44
25- 34
46. 72

Protein .

Fat .

Fiber .

Nitrogen-free extract .

Total . .

100. do

100. 00

308

Journal of Agricultural Research

Vol. VII, No. 7

The sheep were weighed on the first two and the last two days of the
digestion period and the average taken to determine gain or loss in body
weight (Table VII).

Table VII. — Gain or loss in weight ( pounds ) hy sheep fed vegetable-ivory mealy gluten

feed , and hay

Sheep No.

Weight at
beginning.

Weight at
end.

Gain.

Eoss.

r .

I39- 25
160. 13

I37-25
163. 83

2. 00

6 .

3- 70

- - , -

As a supplementary check on the metabolism of the sheep, the urine
was collected, weighed, and sampled daily, and the nitrogen determined.
A study of Table VIII shows that sheep 5 excreted more nitrogen than
was supplied in its food. This sheep lost in weight. Sheep 6, however,
gained in. body weight, and it will be noted that less nitrogen was given
off than was consumed.

Table VIII. — Nitrogen balance of sheep fed vegetable-ivory meal , gluten feed, and hay

[Estimated in grams of protein.]

Sheep No.

Consumed,

Excreted.

Gain.

Loss.

e .

647. 92
647. 92

659. 07
606. 69

11. 15

6 .

41. 23

By applying the analyses in Table IV to the total rations fed, the total
amounts of dry matter and food constituents are obtained. From these
the amounts of the several constituents of the manure (calculated by the
use of Table VI) are subtracted. The remainder is the quantity of hay,
gluten, and vegetable-ivory meal digested. By subtracting from this
the amount of hay and gluten digested 1 the amount and percentage of
the vegetable ivory digested is obtained (Table IX).

Obtained by applying the digestion coefficients of hay and gluten alone to the quantity fed.

Nov. 13, 1916

Vegetable-Ivory Meal

309

Table IX. — Daily consumption and excretion {in grams) and the digestion coefficients of
sheep fed vegetable-ivory meal , gluten feedt and hay

SHEEP 5

Item.

Dry

matter.

Protein.

Pat.

Fiber.

Ash.

Nitrogen-

free

extract.

550 gm. of English hay .

484. 28

45- 67

II. 04

153- 52

28. 14

245. 91

150 gm. of gluten feed .

134. 90

36.37

5- 32

11. 74

I. 50

79- 97

200 gm. of vegetable ivory .

174. 72

IO. 52

I. 21

12. 27

2* 39

148- 33

Amount consumed .

793-9°

92. 56

i7* 57

x77* 53

32*03

474. 21

Minus 252.10 gm. of ma-

nure . .

237- 73

37*09

9. 41

61. 71

21. 78

107. 74

Hay, gluten, and vegetable ivory

digested .

556- 17

55* 47

8. 16

1 15. 82

10. 25

366. 47

Minus hay and gluten di-

gested .

408. 66

55* 79

9. 16

109. 07

9. 19

228. 12

Vegetable ivory digested .

Percentage digested .

147- 5i
84*43

6- 75
55* 01

I. 06
44*35

138- 35

93- 27

SHEEP 6

Amount consumed (as per sheep

5) . . . ; .

Minus 241.94 gm. of ma¬

793- 9°

92. 56

i7* 57

177- 53

32. 03

474* 21

nure .

228. 63

33- 61

7.86

57-93

22. 41

106. 82

Hay, gluten, and vegetable ivory

digested .

Minus hay and gluten di¬

565- 27

58.95

9.71

1 19. 60

9. 62

267. 39

gested .

408. 66

55- 79

9. 16

IO9. 07

9. 19

228. 12

Vegetable ivory digested .

156. 61

3. l6

•55

10. 53

• 43

139- 27

Percentage digested .

89.63

30.04

45* 45

85. 82

17.99

9389

Average percentage for both

87-03

sheep .

Percentage digestibility of Eng¬
lish hay and gluten as previ¬

a 30. 04

“45-45

70. 42

31* 17

93- 58

ously determined & .

a One sheep only.

& Obtained from previous digestion experiments similar to the one under discussion, in which a part
of the same lot of hay and gluten feed was used.

Table IX shows that the two digestion trials for nitrogen-free extract
agree very closely. It is to be noted further that this extrafct matter
constituted about 85 per cent of the total dry matter of the vegetable-
ivory meal and that it had a digestibility of 94 per cent. In the case of
the fat and ash the results are uncertain, but this is not surprising because
so very little of these two ingredients is present. The digestibility of the
fiber is not very satisfactory, and the same may be said of the protein.
The percentages of these two ingredients, however, in the vegetable-
ivory meal are relatively small.

3io

Journal of Agricultural Research

Vol. VII, No. 7

Experiment II. — In another similar experiment conducted with three
sheep and a different sample of vegetable-ivory meal the following diges¬
tion coefficients were obtained. With them are compared the results of
Experiment I as well as the average coefficients for corn meal (Table X).

Table X. — Comparison of the digestion coefficients obtained in Experiments I and II

Experiment No.

Feed.

Number
of sheep.

Dry

matter.

Pro¬

tein.

Fat.

Fiber.

Nitrogen-

free

extract.

I . . .

Vegetable ivory . .

. 3°

“45

II. . .

6 41

Average . . .

Com meal (13, p.

291;) .

a One sheep only. & Different sheep showed variable results.

Of these figures the first and last two columns demand the most atten¬
tion. Corn meal contains nearly as much nitrogen-free extract as the
vegetable ivory, and it would appear that the percentage digestibility of
this ingredient in each feed is approximately the same.

Applying the average coefficients to the composition of the dry matter
of the vegetable-ivory meal and the average coefficients for corn meal to
the dry matter contained in the latter1 and multiplying by 2,000, one
obtains the following amounts of digestible matter in 1 ton of each of the
two feeds (Table XI).

Table XI. — Digestible nutrients ( in pounds ) in vegetable-ivory meal and corn meal per

ton

Feed,

Protein.

Fat.

Fiber.

Nitrogen-

free

extract.

Total.

Vegetable ivory .

Com meal .

42. 34
147- 52

6.83

78.84

IOI. 08

I, 582. 20
I, 486. 40

17 732. 45
1, 712. 76

On the basis of total digestible organic matter the results indicate that
the vegetable-ivory meal is equal in feeding value to corn meal. Kellner
(10) and Armsby and Fries (3, 4, 5, 6) have shown, however, that it is
not possible to estimate with accuracy, by means of digestion experiments,
the relative value of different feedstuffs. In view of the excess of fiber
in the vegetable-ivory meal over that of the corn meal (7 per cent in
ivory meal v. 2 per cent in corn meal), of the tough horny nature of the
ivory nut, of the uncertainty of the nutritive value of the mannan as
compared with starch, and of the unknown influence of the two feedstuffs

1 Tfce average composition of corn meal on a dry-matter basis in Lindsey’s compilations (12) is protein

xi.oi per cent, fat 4.38 per cent, fiber 2.25 per cent, and nitrogen-free extract 80.79 per cent.

Nov. 13, 1916

Vegetable-Ivory Meal

3ii

on metabolism, one is justified in assuming that the vegetable-ivory meal
can not have as high a nutritive effect as has the corn meal.

Experiment III. — As another means of determining how completely
vegetable-ivory meal was digested, Experiment III was undertaken, feed¬
ing the same amounts of the several feeds as in Experiment I. The
basal ration consisted of hay and gluten feed, and the ration proper
of the same feeds in like quantities, plus 200 gm. of vegetable-ivory
meal. Each ration was fed for 14 consecutive days, the feces being
collected for the last 7 days in each period, and aliquots preserved.
In this experiment the hay, gluten feed, and vegetable-ivory meal ration
preceded the basal ration of hay and gluten feed. The feces were tested
for total sugar after acid hydrolysis, to note whether the percentage of
sugar was higher in the ivory-meal period than in the period without the
meal. It is understood that little or no sugar should appear as such in
normal feces, and the relatively large amounts which are reported below
are accounted for as a result of the hydrolysis of pentosans and other
hemicelluloses, largely from the hay. It was necessary to hydrolyze
with strong acid and boiling in order to include completely the sugar
of the vegetable ivory, if any, which might have passed through the
animal unchanged.

On a dry-matter basis it was found that the average carbohydrate
content, estimated as dextrose, for the feces of the hay, gluten, and
ivory-meal period was 25.46 per cent and that for the hay and gluten
period was 24.68 per cent. In other words, the total amount of carbo¬
hydrates, so called, found in the feces when vegetable ivory had been
included in the ration was only 0.78 per cent more than was found when
it had not been included. This is relatively such a small amount that
it seems safe to conclude that very little, if any, of the carbohydrate
of the vegetable ivory escaped undigested. The mannan therefore ap¬
pears to have been quite thoroughly hydrolyzed and assimilated by the
sheep.

FEEDING EXPERIMENTS

Experiment I. — During March, April, and May, 1914, an experi¬
ment to compare the relative feeding value of vegetable-ivory meal and
corn meal was carried out and may be described as follows:

Three pairs of cows were fed for periods of five weeks each, exclusive
of preliminary periods of 10 days, on basal rations consisting daily for
each cow of substantially 2.5 pounds of wheat bran, 2.5 pounds of cotton¬
seed meal, and what hay the animals would eat clean (about 20 pounds).
Either 3 pounds of vegetable-ivory meal or 3 pounds of corn meal were
fed in addition. The experiment was conducted on the reversal plan —
that is, one cow of each pair was fed the basal ration plus the vegetable
ivory for five weeks, while the other received corn meal; then the ration
was reversed for five weeks (Table XII).

312

Journal of Agricultural Research

Vol. VII, No. 7

Table XII. — Average daily ration (in pounds ) consumed per cow in Feeding Experi¬
ment I

Character of ration.

Hay.

Bran.

Cottonseed

meal.

Com meal.

Vegetable-
ivory meal.

Com meal .

Vegetable-ivory meal .

20. 58
20. 46

2. 36
2. 36

2. 28
2. 28

3- 36

Daily samples of the grain and ivory meal were taken and preserved
in glass-stoppered bottles. These were brought to the laboratory for
analysis at the end of each half of the trial. The hay was sampled three
times during each half and determinations of moisture made and ali¬
quots preserved for analysis.

Milk samples were taken for five consecutive days in the first, third, and
fifth weeks of each half of the test and analyzed for fat and total solids.

The cows were weighed on two consecutive days at the beginning and
end of each half of the experiment.

The cows averaged 0.12 of a pound more hay daily while on the corn-
meal ration. The 1 pound of com meal fed during the vegetable-ivory
period applied only to two of the cows for periods of 35 days each.

From the analyses of the materials fed, their digestibility coefficients,
and the amounts daily consumed, the quantity of digestible organic nutri¬
ents received daily per cow was estimated, the results being given in
Table XIII.

Table XIII. — Quantity of digestible organic nutrients in the average daily ration of

cows in Feeding Experiment I

Character of ration.

Protein.

Fat.

Fiber.

Nitrogen-
free extract.

Total.

Nutritive

ratio.

Com meal .

Vegetable-ivory meal . . .

2, 03

I. 88

O- 53
•43

4. 07
4. 29

8. 86
9. 17

15- 47
I5* 77

1 : 6. 9
1 : 7. 6

It will be noticed that the total digestible carbohydrates were slightly
greater for the vegetable-ivory ration than for the corn -meal ration. On
the other hand, protein and fat show a favorable balance for the corn-
meal ration.

At the completion of the experiment a distinct gain in herd weight
was noticed when the corn-meal ration was fed over that of the vegetable-
ivory ration (Table XIV).

Table XIV. — Gain or loss (in pounds) in herd weight in Feeding Experiment I

Character of ration.

Gain.

Loss.

Average per cow.

Gain.

Loss.

Com meal .

15.6

I. 0

Vegetable-ivory meal .

Nov. 13, 1916

Vegetable-Ivory Meal

3*3

Table XV records the total yield of milk and milk ingredients per cow
for each ration, as well as the total yield of the herd.

Table XV. — Total yield of milk and milk ingredients from different rations in Feeding

Experiment I

corn-meal ration

Name of cow.

Time.

Milk,

Solids.

Fat.

Butter.®

Samantha II . .
Fancy III ....

Betty .

Betty II .

Amy .

Amy II .

Total. . .

Weeks.

Pounds .

1, 184. 6
1, 131.8
761. 6

933-4
673. 8
558-3

Per cent.

12. 57
12. 90
13- 76
13- 76

13- 45

14- 75

Pounds.

148.9
146. O
104. 8
128.8
90. 6
82. 3

Per cent.
4.03
4- 33

4- 77

4- 75
4. 68

5- 45

Pounds.

47-7

49.0

3<5.3
44-3
31- 5
3°- 4

Pounds.

54-9

57-2

42.3
51- 7

36- 7

35-5

5. 243- 5

6 i3- 37

701. 0

c 4- 56

239. 2

278.3

vegetable-ivory-meal ration

Samantha II . .

1, 200. 6

12.

652- 5

49. 6

57-9

Fancy III ....

1, 001. 6

x3-

x30-3

45*4

53-0

Betty .

805. 0

x3-

hi. 1

39- 7

46.3

Betty II .

897. 6

x3-

122. 2

40. 3

47.0

Amy .

620. 7

x3-

82. 3

29. 4

34-8

Amy II .

547-2

1 5-

82. 9

3i* 7

37-o

Total...

5, 072. 7

h T3-

681. 3

c4-6s

236. 1

276. 2

a Butter equals fat plus one- sixth.

& Averages obtained by dividing the total weight of solids by the total weight of milk.
c Averages obtained by dividing the total weight of fat by the total weight of milk.

One hundred and seventy pounds more milk were produced by the corn-
meal ration than by the vegetable-ivory-meal ration. This excess is not
pronounced; and while it is possible that the difference may be within
the limit of experimental error when taken together with the fact that
the corn-meal ration increased the live weight of the cows, it indicates
at least that the com meal was somewhat superior to the vegetable-
ivory meal as a source of nutrition.

Table XVI. — Average percentage composition of milk of the herd in Experiment I on

each ration

Character of ration.

Total

solids.

Fat.

Solids
not fat.

Corn meal . ; .

x3- 37
x3-43

4* 56
4-65

8. 81

Vegetable-ivory meal .

8.74

The very concisely stated data of Table XVI of this experiment indi¬
cate that vegetable-ivory meal possesses a distinct feeding value and
that, while somewhat inferior to com meal, the difference is not marked.

As greater difficulty was met with in hydrolizing the carbohydrate of
the vegetable ivory than the carbohydrate of other feeding materials, it

Journal of Agricultural Research

Vol. VII, No. 7

3*4

would seem reasonable to suppose that the same relation would hold
true in the digestive processes of animals. This being the case, more of
the total energy of the material would be used up in digestion and assimi¬
lation, which otherwise might be used for milk production.

Experiment II. — A short experiment with three cows, supplementary
to Experiment I, was carried out during January and February of 1915.
A basal ration somewhat below what the cows needed for maintenance
and milk was fed for two weeks. Then the cows were given 3 pounds of
vegetable-ivory meal in addition for two weeks. During the fifth and
sixth weeks the cows were again given only the basal ration. The object
was to see whether the animals would show any increase in weight or milk
production in response to the addition of the ivory meal.

Table XVII. — Average daily ration {in pounds) consumed per cow in Feeding Experi¬
ment II

Character of ration.

Hay.

Wheat

bran.

Cottonseed

meal.

Hominy.

Vegetable-
ivory meal.

Basal .

18. 67
18. 67

2. 34
2.34

Vegetable-ivory meal .

The quantity and quality of the milk was determined for the second week
of each section of the experiment. Samples were taken in the usual way
and the animals were weighed on two consecutive days at the beginning
and end of each second week.

Little can be said about body weight, as the periods were of too short
duration.

Table XVIII. — Yield of milk and milk ingredients from different rations in Feeding

Experiment II

BASAL RATION, JANUARY 20-26

Name of cow.

Milk.

Fat.

Solids.

Amy .

Pounds.
196. s
137- 6
121. 9

Per cent .

5* So
4. 70

5-38

Pounds.
IO. 8l

6.47

6.56

Per cent.

i3- 77
13- 23
13. 86

Pounds .
27. 06

l8. 20
l6. 90

Betty III .

Red III .

Total . ...

456. 0

a 5. 22

23.84

a 13- 63

62. l6

VEGETABLE-IVORY- MEAL RATION, FEBRUARY 3-9

Atny . .

199.2

4.90

9. 76

13- i4

26. 17

Betty III .

158. 0

4- 58

7. 24

13-25

20. 94

Red III .

125. 1

5. 28

6. 61

i3-7i

17* IS

Total .

482.3

a 4. 90

23. 61

a 13. 30

64. 26

BASAL RATION, FEBRUARY 2 1-2 7

Amy .

190. 3
146. 4
125. 4

4. 80
4. 60
S- 38

9- 13
6. 73
6- 75

13. 12
*3- 31
i3- 93

24.97

19. 49

17- 47

Betty III .

Red III .

Total .

462. 1

« 4. 89

22. 6l

a 13- 21

61.93

® Average obtained by dividing the total number of pounds of fat or solids by the total number of pounds
of milk.

Nov. 13, 1916

Vegetable-Ivory Meal

3i5

Though of short duration, the experiment shows the favorable effects
of the addition of 3 pounds per cow daily of the ivory meal to the basal
ration. This addition increased the milk flow 5.7 per cent, and its
removal caused a decrease of 4.2 per cent.

Another experiment in which a basal ration with and without vegetable-
ivory meal was fed three cows was carried on for a period of 81 days.
The addition of the meal was followed by an increase in milk yield and
its removal resulted in a milk shrinkage. This trial, together with the
one just described, shows that the ivory meal possesses a distinct nutri¬
tive value.

Experiment III. — In Experiment I a definite quantity of vegetable-,
ivory meal was compared with an equal quantity of com meal. In
order to demonstrate more fully the effect of the vegetable ivory, a
herd of six cows was put on a basal ration of hay, bran, cottonseed meal,
and hominy meal for four weeks, exclusive of a preliminary period of 10
days, during which tithe three received in addition a quantity of the
ivory meal. After this period, which will be known as the first half of
the experiment, conditions were reversed, and during the four weeks of
the second half the first three cows went without the vegetable-ivory
meal while the others received it.

Each cow was weighed before watering and feeding on two consecu¬
tive days at the beginning and end of each half of the trial. Samples of
hay were taken at the beginning, middle, and end of each half, while the
grains were sampled daily. The milk of each cow was sampled in the
usual manner on the first, third, and fourth week of each half. The
results are given in Tables XIX and XX.

Table XIX. — Average daily ration (in pounds) consumed per cow in Feeding Experi¬
ment III

Character of ration.

Hay.

Wheat

bran.

Cottonseed

meal.

Hominy.

Vegetable-
ivory meal.

Basal plus vegetable-ivory meal. . .
Basal minus vegetable-ivory meal .

18. 32
18.31

2. 30
2. 31

2. 22

2. 23

i- 57
57

2. 79

Table XX. — Quantity of digestible organic nutrients in the average daily ration of

Feeding Experiment III

Character oi ration.

Protein.

Fat.

Fiber.

Nitrogen-
free ex¬
tract.

Total.

Nutritive

ratio.

Basal plus vegetable-
ivory meal .

2. JO

O. 61

3- 04

io-37

16. 52

1:5. 90

Basal minus vegetable-
ivory meal .

2. II

. 60

2. 92

8. 36

13- 99

^5- 97

316

Journal of Agricultural Research

Vol. VII, No. 7

In order to avoid excessive feeding, a basal ration was fed which was
rather below what each animal needed for maintenance and normal
milk production. This was calculated on the basis of the writers'
knowledge of the individual animal and with the aid of Haecker's stand¬
ards (8). The addition of the vegetable-ivory meal to this basal ration
should therefore prove its distinctive feeding value.

In Table XXI will be found the total yield of milk, fat, and solids
produced by each ration.

Table XXI. — Total yield of milk and milk ingredients in Feeding Experiment III
BASAL RATION PLUS VEGETABLE-IVORY MEAL

Name of cow.

Milk.

Solids.

Pat.

Butter.®

Pounds,

Per cent.

Pounds .

Per cent.

Pounds .

Pounds.

Fancy III . . .

I, 057. 2

13. OO

137* 43

4. 67

49*37

57-6

Betty III .

599*0

13* 55

81. 16

4. 72

28. 27

33-o

Ida II .

608. 4

14.38

87.49

5*33

32.43

37-8

Betty .

676. O

12. 96

87. 61

4*39

29. 68

34-6

Red III .

312. 5

15. 12

47-25

5* 29

16. S3

18.3

Amy .

705*9

I3* 27

93- 67

4. 78

33- 74

39-4

Total .

3 > 959*o

6 13- S°

534- 61

6 4- 79

I90. 02

220. 7

BASAL RATION MINUS VEGETABLE-IVORY MEAL

Fancy III .

970. 1

12. 52

II3- $6

4. 26

41* 33

48. 2

Betty III .

477* 1

12.95

6l. 78

4. 49

21. 42

25.0

Ida II .

531* 7

12. 95

68. 86

5. 16

27.44

32.0

Betty .

706. 7

13. 16

93.00

4. 62

32.65

38.1

Red III .

433*2

14. 38

62. 29

5*25

22. 74

26. s

Amy .

707. 8

13* 14

93.00

4. 84

34. 26

40. 0

Total . . . .

3, 826. 6

a 12. 87

492. 49

0 4. 70

179. 84

209. 8

a Butter equals fat plus one-sixth.

6 Averages obtained by dividing the total weight of fat or solids by the total weight of milk.

It will be noticed here that the ivory-meal ration produced 132.4
pounds more milk than the other ration, an increase of 3.46 per cent.
Inasmuch as the total feed consumed was identical in each ration with
the exception of 470 pounds of the vegetable-ivory meal, one may con¬
clude in the case of this particular experiment that the 132.4 pounds of
milk were produced by the 470 pounds of ivory meal, or that it required
3.56 pounds to produce 1 pound of milk.

If rather less of the basal ration had been fed and more of the vege-
table^ivory meal, it is probable that the effect of the latter would have
been more pronounced.

Experiment IV. — Another feeding experiment comparing vegetable-
ivory meal and corn meal was carried out during January, February,
and March, 1916, with eight cows. In this case, as in Experiment T,

Nov. 13. 19*6

Vegetable-Ivory Meal

317

the reversal method was employed, each period continuing five weeks,
in addition to the preliminary period. Hay, bran, and cottonseed meal
composed the basal ration, to which were added like amounts of dry
matter in the form of corn meal or ivory meal (Tables XXII and XXIII).
Only a summary is here presented.

All the customary precautions were taken to make the experiment as
accurate and representative as possible. The milking was done at the
same time each morning and evening, and the animals were weighed at
regular intervals. All feed and milk samples were taken in the usual
manner.

Table XXII.— Average daily ration (in pounds) consumed per cow in Feeding Experi¬
ment IV

Character of ration.

Hay.

Bran.

Cottonseed

meal.

Com meal.

Vegetable-
ivory meal.

Vegetable-ivory meal .

Pnrn mpfll .

18. 5

18. s

2. 38
2. 38

2. 19
2. 19

4. OI

3- 75

Table XXIII. — Quantity of digestible organic nutrients in the average daily ration in
" " Feeding Experiment IV

Character of ration.

Protein.

Fat.

Fiber.

Nitrogen-
free extract.

Total.

Nutritive

ratio.

Vegetable-ivory meal . . .
Corn meal. . .

1. 88

2. 07

0.44
• 55

3- 58
3- 38

9. 06
9. OI

14. 96

15- 01

17-24
1:6. 57

The calculations indicate that the herd consumed substantially like
amounts of digestible nutrients in each of the two rations. One would
expect a like effect on body weight and milk production. As in Experi¬
ment III, care was taken to feed less digestible nutrients than was
required for maintenance and milk yield according to the Haecker stand¬
ard, in order to secure the maximum effect of each ration. The results
indicate that this object was in a measure at least achieved. The herds
uffered a loss of weight on each ration, the loss being the greater on the
one containing vegetable-ivory meal, 95 pounds, as against 38 pounds for
the corn-meal ration.

64311°— 16 - 2

Journal of Agricultural Research

Vol. VII, No. 7

318

Tabi,K XXIV. — Total yield of milk and milk ingredients in Feeding Experiment IV

vegetable-ivory-meal ration

Solids.

Fat.

Name of cow.

Total milk.

Average

percentage.

Pounds.

Average

percentage.

Pounds.

Amy .

Pounds .
852. I

13. 60

115. 89

5- 14

43. 80

Betty III . .

798. 2

13* 30

106. 16

4. 76

37- 99

Cecile II .

658. 2

14. 76

97- 15

5- 58

36- 73

White .

676. O

13- 54

91- 53

5- 19

35- 08

Betty II .

Red III .

775- 8
576- 1

i3- 75
13. 69

106. 67
78. 87

4.94
5- 33

38-32
3°- 71

Red IV .

723- 5

13. 29

96. 15

4- 60

33- 28

Samantha II .

1, 343- 4

12. 64

169. 8l

4. 46

59- 92

Total .

6, 403- 3

° 13- 47

862. 23

a 4* 93

315- 83

CORN-MEAL RATION

Amy .

I, 001. 0

J3- 23

132. 43

4-93

49- 25

Betty III .

851.0

12. 83

109. 18

4 43

37- 7°

Cecile II .

744-4

14. 28

106. 30

5. 22

38.86

White .

704. 8

14. 10

99-38

5*09

35-87

Betty II .

840. 6

13-83

116. 25

4. 94

4i- 53

Red III .

599* 5

13- 88

83. M-

' .5-39

32-31

Red IV .

736.0

14. 29

I05. 17

5. 20

38. 27

Samantha II .

1. 454- 0

12. 65

i83- 93

4-39

63-83

Total .

6,93r-3

a 13- 5°

935- 85 1

“4.87

337- 72

0 Averages obtained by dividing the total weight of fat or solids by the total weight of milk.

Table XXIV shows that the corn-meal ration produced an increase of
528 pounds of milk, or approximately 8 per cent over that produced by
the ivory-meal ration. The total solids were also increased 73.62 pounds
and the milk fat 21.89 pounds. The cows in this experiment were in an
earlier stage of lactation than those used in the previous trials, and the
results are to be regarded as the most satisfactory. It is evident that
while the vegetable-ivory meal possesses a distinct feeding value, a given
amount has not the feeding equivalent of the same amount of corn meal.
The methods of experimentation with milch cows are not sufficiently
sharp to enable one to draw accurate deductions as to the exact relative
feeding effect of the two materials.

Nov. 13, 1916

Vegetable-Ivory Meal

3i9

CONCLUSIONS

(I) Analyses show vegetable ivory to be carbohydrate in nature,
containing about 5 per cent of protein and 75 per cent of nitrogen-free
extract. Fat and mineral matter are negligible, while crude fiber averages
7 per cent.

(2.) Ninety- two and one-half per cent of the nitrogen-free extract is
mannan, a polymer of mannose sugar.

(3) Pentosans are present to the extent of 2.5 per cent.

(4) Lignin, galactan, starch, and dextran are not found in vegetable
ivory.

(5) A nonnitrogenous “alcoholic precipitate” amounting to about 2.5
per cent is present which is not pentose in nature. It differs from fruit
“pectin” in that it does not form mucic acid and does not reduce copper.

(6) By the use of Fehling’s solution about 0.5 per cent of water-soluble
reducing material and 2 per cent of so-called total sugars are shown to be
present after inversion with hydrochloric acid in the cold.

(7) The mannan in vegetable ivory is not entirely hydrolyzed without
at least 4 % hours' boiling in an acid solution. The characteristic “acid ”
color of the solution bleaches out at the completion of hydrolysis.

(8) With continued acid boiling the use of Fehling's solution gives
results which, when estimated as dextrose, agree closely with the per¬
centage of nitrogen-free extract minus the percentage “pectin” present.
Otherwise stated, practically the entire nitrogen-free extract is accounted
for in the form of a hexose sugar or its condensation product, except a
small percentage of pentoses and pectin.

(9) The energy equivalent of the material ranks well with other
carbohydrate foods, and it possesses a fuel value equal to one-half that of
soft coal.

(10) Sheep ate vegetable-ivory meal readily when it was mixed with
other grains and digested it very thoroughly. Eighty-four per cent of
the dry matter and ninety-two per cent of the nitrogen-free extract were
digested.

(II) All the carbohydrates appeared to have been hydrolyzed and
absorbed in the digestive tract.

(12) Cows ate the material when mixed with other feed, without
evidence of digestive disturbances. They refused to eat it if fed by
itself.

(13) When fed as an addition to a basal ration, the increase in milk
was sufficient to indicate its positive value as a productive feed.

(14) Though the methods of feeding necessarily followed were not
such that exact relative values could be shown, it seems certain that
vegetable-ivory meal does not fully equal com meal for milk production.

320

Journal of Agricultural Research

Vol. VII, No. v

LITERATURE CITED

(1) Axbes, Edward.

1913. Tagua — vegetable ivory. In Bui. Pan Amer. Union, v. 37, no. 2, p.
192-208, illus.

(2) Armsby, H. P.

1909. The computation of rations for farm animals by the use of energy values.
U. S. Dept. Agr. Farmers' Bui. 346, 32 p.

(3) - and Fries, J. A.

1903. The available energy of timothy hay. U. S. Dept. Agr. Bur. Anim.
Indus. Bui. 51, 77 p., 2 pi.

(4) -

1905. Energy values of red clover hay and maize meal. U. S. Dept. Agr. Bur.
Anim. Indus. Bui. 74, 64 p.

(s) -

1908. The available energy of red clover hay. U. S. Dept. Agr. Bur. Anim.
Indus. Bui. 101, 64 p.

(6) -

1915. Net energy value of feeding stuffs for cattle. In Jour. Agr. Research, v.
3, no. 6, p. 435-491, 2 fig.

(7) Fischer, Emil, and Hirschberger, Josef.

1889. Ueber Mannose IV. In Ber. Deut. Chem. Gesell., Jahrg. 22, p. 3218-3224.

(8) Haecker, T. L.

1913. Feeding dairy cows. Minn. Agr. Exp. Sta. Bui. 130, 43 p.

(9) Jordan, W. H.

1912. The Principles of Human Nutrition . . . 450 p., illus., pi., tab. New
York.

(10) Keener, Oskar.

1912. Die Emahrung der landwirtschaftlichen Nutztiere . . , Aufl. 6, 640 p.,
front. Berlin.

(11) Liebscher, G.

1890. Der Nahrwerth der Steinnusspahne. Nach Versuchen von Dr. Schuster
und Prof. Dr. Liebscher. In Landw. Jahrb., Bd. 19, p. 143-148.

(12) Lindsey, J. B.

1894. Digestion experiments with sheep. In Mass. Agr. Exp. Sta. nth Ann.

Rpt. 1893, p. 146-178, 1 pi.

(13) - and Smith, P. H.

1911. Coefficients of digestibility of American fodder articles. Experiments
made in the United States. In Mass. Agr. Exp. Sta. 23d Ann. Rpt.
[1910], pt. 1, p. 273-303. Literature, p. 363.

(14) ToixEns, B.

1895. Kurzes Handbuch der Kohlenhydrate. Bd. 2. Breslau.

ROSY APPLE APHIS

By A. C. Baker and W. F. Turner,

Entomological Assistants , Deciduous Fruit Insect Investigations , Bureau of Entomology

CONTENTS

Page

Introduction . 321

Nomenclature of rosy apple aphis . 321

History and distribution of the species . 325

Methods of study . 325

The egg . 326

Stem mother . 326

Spring forms . . . 328

Spring wingless viviparous female . 328

Intermediate form . 330

Spring migrant . 331

Migrations of the species . 333

Page

Summer forms . 335

Summer wingless viviparous female . 335

Summer winged viviparous female . 336

D imorphic reproduction . 337

Pali forms . 337

Fall migrant . 337

Male . 339

Oviparous female . 339

Feeding habits . 340

Summary of life history . 342

Literature cited . 342

INTRODUCTION

The rosy apple aphis is undoubtedly the most injurious leaf-feeding
apple aphis. Its attacks not only injure the foliage and deform the
growing apple trees {Mains sylvestris) , but when abundant or unchecked
it deforms the fruit, causing the production of “aphis apples/ ’ which are
unfit for sale. The experiments on which the present paper is based were
conducted during the seasons of 1914 and 1915 and the manuscript
prepared for publication during the winter of 1915-16. Besides the work
recorded in the present paper a study of the embryology was undertaken.
This work still remains to be completed, the present paper recording the
life history only after the hatching of the egg.1

NOMENCLATURE OF ROSY APPLE APHIS

The proper scientific name for the rosy apple aphis has for some years
been in doubt. Pergande, as shown by his manuscript notes, always
considered it to be Aphis malifoliae Fitch. Other American writers
usually adopted the view that this name became a synonym of A . soy hi
Kalt. In studying European specimens and the literature carefully
the writers have come to the conclusion that A . malifoliae is the only name
to apply to the species. Their reasons for this view are pointed out under
the different names following.

Aphis malifoliae Fitch, 1854. — Fitch (4, p. 760-761 ; Repr. p. 56-57) 2
separated our rosy aphis from his mali Fab. in connection with his

1 During the course of the study the writers were assisted by Miss Dorothy Walton and by Mr. James
Luckett. Mr. Luckett handled a large number of the experiments during the summer of 1915.

* Reference is made by number to “Literature cited,” p. 342-343*

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

(321)

Vol. VII, No. 7
Nov. 13, 1916

K-45

322

Journal of Agricultural Research

Vol. VH, No. 7

discussion of that species. No complete description was given but enough
was recorded to leave no doubt that he had our well-known rosy aphis.
The fact that he referred to the winged form as almost entirely black is
given by Gillette as sufficient evidence for rejecting the name as not
applicable to the insect herein considered. It must be remembered,
however, that the insects to which he referred were fall migrants (taken
on October 4). These fall migrants are usually much darker than the
spring migrants and the writers have reared many at Vienna, Va., which
were a uniform jet-black. There seems very good reason, then, for
accepting Fitch’s name as referring to the species, and since A. sorbi
and A. pyri prove to be distinct, A. malifoliae Fitch is the name that
■ must stand for the species.

Aphis pyri Boyer de Fonscolombe, r84i. — This name (r) is the earliest
one that has been applied to our rosy aphis. Gillette and Taylor (7,
p. 31-32) referred the species to it when they found themselves unable
to accept A. sorbi Kalt. The original description, however, will not fit
our insect, and, although later descriptions given under this name might
do so, it will for this reason be impossible to use the name for our common
rosy aphis. It may be that Gillette and Taylor based their identification
upon the description referred to Boyer’s species by Koch (3).

In redescribing what he thought to be the A. crataegii of Kaltenbach (2)
as well as the A. pyri of Boyer, Koch based his descriptions entirely
upon apple forms. In this he was describing the forms doubtfully
referred to A. crataegii by Kaltenbach, but seemingly not the true
A. crataegii of that author. The description given agrees very closely
with our apple insect. It would seem, therefore, that the descriptions
of A. crataegii Kalt. and A. pyri Boyer refer to another insect, but the
A. pyri Boyer of Koch (3, p. 108-110, fig. 145, 146) is the A. malifoliae
of Fitch. This insect, however, must be distinguished from the A. pyri
of Koch (3, p. 60), which is quite a distinct species and has been renamed
.4, kochii by Schouteden.

Aphis crataegii Kaltenbach, 1843. — Kaltenbach (2, p. 66-67) described
a species under this name from the white thorn. The description as
given does not agree with specimens of the insects under consideration.
It seems evident, then, that this name can not be applied to the rosy
aphis. In his description, however, Kaltenbach refers to specimens
taken on apple which he believed might be the same as those on thorn.
It seems to the writers that these forms are the same as ours upon apple
and that Kaltenbach was in error in considering them to be the same
species as his specimens on thorn. This is indicated by his comparison
between the two.

Aphis sorbi Kaltenbach, 1843— This species (PI. 20, A) was described
(2, p. 70-71) from specimens on Sorbus. This is the name now most
uniformly applied to the rosy apple aphis in this country. The first

Nov. 13, 1916

Rosy Apple Aphis

323

application of the name here appears to have been made by Sanderson
(5, p. 189-191), who went rather fully into the history of the species.

In some characters our rosy apple aphis does not fit the description
given by Kaltenbach, and these characters cause the writers to believe
that they are dealing with another insect. This belief is strengthened
after examination of the European insects. Collections of A. sorbi from
Sorbus, taken by Mr. J. F. Strauss in Germany, being from the same
region and the same host as the original specimens, can with some cer¬
tainty be considered typical. Moreover, they fit very closely the original
description of Kaltenbach.,

Although these specimens are in a general way very close indeed to
the rosy apple aphis, a careful comparison shows that they represent a
distinct species (PI. 22, C). In the wingless form the cornicles are con¬
siderably longer than are those of the rosy apple aphis, and the lateral
tubercles are more prominent. The antennae, too, show a considerable
difference, the length of IV, as compared with V, being much less in A.
sorbi than in the rosy aphis. These characters are well illustrated in
Table I.

Table I. — Relative proportions of antennce and cornicles of wingless forms of Aphis

sorbi and A. malifoliae

Aphis sorbi .

Aphis malifoliae .

Segment

IV.

Segment

VI.

Cornicle.

Segment

III.

Segment

IV.

Segment

VI.

Cornicle.

8 ~24

6. 5-27

6. 5-22

19- 5

6. 5-28

6- 5"25

6. 5-27

IS-

7 ”23

6- 5-32

The winged forms of the two species, while very much alike in general
appearance and color characters, can be separated quite easily by meas¬
urements of the antennal segments. The specimens of A. sorbi show
the base of Segment VI considerably longer and the unguis considerably
shorter than the same portions of Segment VI of the rosy aphis. The
large number of sensoria on the antennae of the two species cause them
to resemble each other very closely.

Measurements of the antennae of six specimens chosen at random
from the two species are given in Table II, together with the number
of sensoria on the different segments.

A glance at these data will show a fairly constant difference between
the spring migrants of these two forms, and this structural difference is
borne out by the writers' experiments. They have been unable to rear
the rosy aphis on the host plant of the European A . sorbi . The European
form they have been unable to test on apples on account of the lack of
live material.

324

Journal of Agricultural Research

Vol. VTI, No. 7

Table II. — Relative proportions of antennce and number of sensoria of winged forms of

Aphis sorbi and A. malifoliae

Aphis sorbi.

Aphis malifoliae.

Segment

in.

Segment

IV.

Segment

VI.

Segment

III.

Segment
. IV.

Segment

VI.

Bast

3°

Aphis kochii Schouteden, 1903 (PI. 20, B). — This name was given
by Schouteden (6, p. 185) to the species described as A. pyri by Koch,
since Boyer's species already was called by that name. It was found
by Koch curling the leaves of Pyrus pyraster . The description given
by Koch does not agree with our apple insect, but it does agree with
another species occurring on apple in Germany, of which the writers
have specimens. Two species were collected from the apple in Ger¬
many at the same time, one which seems undoubtedly to be A . malifoliae
and another which is very similar to it but having short cornicles.

In describing his A . pyri Koch says (3, p. 60): “ Honigrohrchen sehr
kurz, etwas walzenformig.” His figures also show short cornicles which
are very unlike those of the rosy aphis, but very much like those of the
other species. The writers believe, therefore, that A . kochii is quite a
distinct species from A . malifoliae }

Myzus plantaginis Passerini? (PI. 24, A ). — A species of Myzus occur¬
ring commonly on the broad-leaved plantain in this country must be
distinguished from the rosy apple aphis occurring on plants of the same
genus. This is a very simple process where winged forms of the species
are present. These have the wing veins bordered with dusky, giving
the venation a much heavier appearance than it has in the rosy aphis.
The third segment, moreover, of the antennae has one simple row of
sensoria, whereas the same segment of the rosy aphis is crowded with
sensoria. When, however, only wingless forms are present, the two
species look remarkably alike and there would seem to be almost as
much reason for calling the one a species of Myzus as the other. The
measurements, too, are very similar, but Segment V of M. plantaginis
nearly always averages a little longer than Segment IV, whereas in the

1 Since the account just given was written, Theobald (io. p. 202-210) places A . malifoliae Fitch, 1856, as •
a synonym of A. kochii Schouteden, 1903. As shown herein, the two are distinct, but A. pyri Boyer of
Koch is A . malifoliae Fitch.

Nov. 13, 1916

Rosy Apple Aphis

325

rosy aphis the reverse is true, Segment IV being usually the longer.
Segment III of M. plantaginis is also relatively shorter than is the same
segment in the rosy apple aphis.

The synonymy of the rosy apple aphis will thus stand as follows:

Aphis malifoliae Fitch

Aphis pyri Boyer of Koch (but not A. pyri Boyer nor A. pyri Koch).

Aphis sorbi Kaltenbach of recent European and American authors (but not A . sorbt
Kaltenbach).

Aphis pyri Boyer of Gillette and Taylor (but not A. pyri Boyer).

Aphis kochii Schouteden of Theobald (but not A. kochii Schouteden).

HISTORY AND DISTRIBUTION OF THE SPECIES

The descriptions given by Koch and Kaltenbach indicate that this
species was present in Europe at an early date. It now appears to be
well distributed in the apple-growing regions. In America Fitch’s
description makes its early presence known. It would seem, however,
that it was not until about 1900 that the insect assumed the importance
of a leading pest. The first reports of its occurrence in injurious numbers
came from the Eastern States, but in the few years following it had been
observed over a wide area. At present the species occurs over nearly
all the apple-growing regions of the country, and in some sections it is
very abundant. In some local areas the insect may be not at all common,
even though in that general region it is abundant. This seems to be due
to the comparative scarcity in some places of its secondary host, and in
areas where the insect assumes considerable importance plantain is
usually found in great abundance. Although the species occurs abun¬
dantly in some of the Northern States, it does not seem as yet to have
penetrated very far into Canada. It occurs in Quebec, Ontario, and
British Columbia, but nowhere in Canada does it seem to have assumed
such importance as in some sections of this country.

METHODS OF STUDY

Experiments. — The experiments on which this paper is based were
conducted along much the same line as recorded by the writers (n) for
the green apple aphis. Eggs were allowed to hatch and the stem
mothers and their offspring to grow on young seedlings in pots. The
same methods of transfer adopted for the green aphis were employed
with this species. As soon as spring migrants were produced, these
were transferred to rib-grass plants grown in pots, and covered with
lantern globes. In this manner the species was grown throughout the
summer, and in the fall the fall migrants and males were returned to the
apple. All insects possible were reared to maturity in order that the
percentage of winged forms occurring throughout the summer might be
ascertained. Winged forms during the summer proved to be rare and
this simplified considerably the handling of the insects, since it reduced

326

Journal of Agricultural Research

Vol. VII, No. 7

the number of lines carried. It is quite possible that lines from stem
mothers taken in different localities or possibly different rearing condi¬
tions might have resulted in a higher percentage of summer winged forms.
All molts and specimens in every generation in each line were mounted
for study and these many forms gave an excellent opportunity for
determination of variations.

Technique. — The technique employed for the study of insects of this
species was the same as that recorded in the writers' paper (n) on the
green apple aphis.

THE EGG
DESCRIPTION

Size 0.550 by 0.272 mm. The largest were 0.608 mm. long and the shortest 0.480
mm. The width varied from 0.256 to 0.288 mm. The longest eggs were not neces¬
sarily the widest. In fact, this was seldom the case.

The newly laid egg is light yellow in color, changing through greenish yellow and
yellowish green to black in about four days.

LOCATION ON TREE

The eggs are laid mostly on the small twigs, under buds, or in crevices
in the bark. They may, however, be laid on the small branches, or even
occasionally on the large branches and trunk. They are seldom laid on
the water sprouts, though- this also may occur. Usually oviparous
females bom and reared on water sprouts lay their eggs at the base of
such sprouts or on the trunk of the tree.

The small plants used in the experiments were frequently potted with
portions of the highest roots exposed. In such cases some oviparous
females invariably laid eggs upon these exposed roots.

HATCHING

The eggs of this species commenced hatching, in 1914, about April 8,
at the same time that the eggs of A. pomi began to hatch, and about 10
days later than A . avenae . In 1915 hatching also commenced about
April 8. Eggs of A. pomi commenced hatching on the same date. At
this time the stem mothers of A. avenae were in the second instar.

STEM MOTHER
DESCRIPTION

First instar. — Morphological characters. Antennal segments as follows: III, 0.12
to 0.144 mm., average 0.133 mm.; IV, (0.048 plus 0.12 mm.) to (0.064 plus 0.136mm.).
Eyes with few facets. Cornicles short and thick, 0.088 to 0.096 mm., average
0.089 mm.

Color characters: Dark green; appendages and crown black. Insects covered after
a short time with a mealy bloom.

Second instar.— Morphological characters. Antennal segments as follows: III,
0.112 to 0.114 mm., average 0.128 mm.; IV, 0.072 to 0.096 mm., average 0.081 mm.;
V, (0.064 plus 0.136 mm.) to (0.072 plus 0.168 mm.), average (0.069 plus 0.150 mm.).

Nov. 13, 1916

Rosy Apple Aphis

327

Segments IV and V quite distinctly imbricated. Eyes with about 3 5 facets; cornicles
0.12 to 0.128 mm., average 0.123 mm., rather thick.

Color characters: Considerably lighter than specimens of the first instar, although
still of a somewhat greenish shade.

Third instar. — Morphological characters. Antennal segments as follows: III,
0.2 to 0.224 mm., average 0.213 mm.; IV, 0.096 to 0.112 mm., average 0.106 mm.;
V, (0.072 plus 0.16 mm.) to (0.088 plus 0.208 mm.), average (0.08 plus 0.177 mm.). In
this instar Segment III sometimes shows division, in which case Segments III and IV
have about the following proportions: III, 0.144 mm.; IV, 0.112 mm. Eyes with a
large number of facets; cornicles still stout and measuring 0.144 to 0.176 mm., average
0.16 mm.

Color characters: Approaching those of the adult insect.

Fourth instar, — Morphological characters. Antennal segments as follows: III,
0.192 to 0.24 mm., average 0.216 mm.; IV, 0.112 to 0.16 mm., average 0.128 mm.;
V, 0.128 to 0.152 mm., average 0.136 mm.; VI, (0.08 plus 0.184 mm.) to (0.112 plus
0.24 mm.), average (0.096 plus 0.213 mm.). Cornicles 0.1224 to 0.256 mm., average
0.227 mm*

Color characters: Almost those of the adult insect.

Fifth instar. — Morphological characters. Antennal segments as follows: III,
0.336 to 0.384 mm., average 0.352 mm.; IV, 0.192 to 0.224 mm., average 0.2 mm.;
V, 0.144 to 0.176 mm., average 0.16 mm.; VI, base, 0.096 to 0.112 mm., average 0.099
mm.; unguis o 192 to 0.224 mm., average 0.208 mm. Segments imbricated, the first
segment being distinctly ridged on its inner margin and armed with a number of
slightly capitate hairs. Antennae on frontal tubercles, these being armed on their
inner edge with slightly capitate hairs; vertex slightly protruding; crown armed with
numerous hairs and near its caudal margin with a pair of distinct tubercles (in many
specimens only one of this pair present and in other specimens neither); eyes promi¬
nent, their tubercles small ; pfothoracic tubercles not as prominent as in the later forms;
abdomen with distinct lateral tubercles and fine hairs; the last two abdominal seg¬
ments each with a pair of distinct dorsal tubercles; cornicles subcylindric, 0.288 to
0.32 mm., average 0.305 mm., considerably broader at the base than at the apex,
slightly flanged and strongly imbricated; cauda conical, short, densely setose and
covered with numerous long hairs; anal plate rounded and similarly armed; legs
with short stiff hairs, femora rough and covered with sensory-like markings, hind
tibiae 0.83 mm. long; form very globose, the abdomen not showing segmentation;
length from vertex to tip of cauda 1.68 mm; width of abdomen 1.44 mm.

Color characters: General color reddish or purplish brown, dusted with a bluish
white powder. Antennae, with the exception of the basal portion of Segment III and
sometimes Segment II, black; head and prothorax dark brown or blackish; abdomen
varying shades of brown or purplish with a few minute dark markings which include
the lateral tubercles; cauda and anal and genital plates black; cornicles black; legs
entirely black with the exception of a light ring at the base of each femur; color
between and surrounding the cornicles rusty.

Location: Found within the curled leaves of the apple, usually entirely hidden by
the tightly rolled leaf.

LENGTH OF NYMPHAL LIFE

When newly hatched, the stem mother wanders about on the twig until
a bud is reached. Here she settles and commences feeding, crowding
down into the bursting bud. Before feeding she has a wrinkled appear¬
ance, but begins to fill out in a day or two.

The duration of the first instar is considerably longer than that of the
following ones, but this depends a good deal on weather conditions. If

328

Journal of Agricultural Research

Vol. VII, No. 7

the weather is favorable, the first instar will be much shortened. The
total nymphal life averaged in the experiments 1 5 days, and this was the
average in spite of the variation in the length of the instars.

REPRODUCTION

The stem moth.ers began reproducing in the experiments during the
24 hours following the last molt. Considerable variation was noted
amongst individuals, both as to the total number of young produced and
the number produced daily. Reproduction in groups, as has been noted
by the authors for Aphis pomi , is very common with this species also.
The greatest number of young produced by any individual stem mother
was 260, and these were produced during a period of 20 days. The
smallest number produced by one adult was 81 in 24 days. The average
number for 12 individuals was 7 1.1 young each, the average reproductive
period being 26.3 days. The average daily production was 6.3 per insect
and the greatest record was 14.6, one female giving birth to 44 young
in 3 days.

LONGEVITY

The greatest length of life observed was 45 days. Many other stem
mothers which died from accident before an equally long period of time
were in very good condition at the time of their death. It would
seem, therefore, that this period is not very far above the average.

SPRING FORMS

Several generations of aphids, in which both wingless and winged
forms occur, follow the stem mother upon apple. In 1914 five apple-
infesting generations occurred in the experiments, while in 1915 seven
such generations were observed. In both cases the first stem-mother
generation is not included in the figures given.

The first of these generations, the second generation from the egg,
appears to be composed entirely of wingless insects — at least there were
no winged forms in the experiments, either in 1 9 1 4 or 1 9 1 5 . In the third
generation, however, a few winged forms occurred and the percentage of
winged to wingless insects increased rapidly up to the sixth generation
in 1914 and the eighth in 1915, in which all the adults had wings. Two
intermediates were also reared with the spring migrants.

SPRING WINGLESS VIVIPAROUS FEMALE
description

First instar. — Morphological characters. Antennae with the following measure¬
ments: Segment III, 0.16 to 0.208 mm., average 0.182 mm.; IV, (0.048 plus 0.176 mm.)
to (0.072 plus 0.224 mm.), average (0.059 plus 0.204 nun.). Cornicles about 0.088
mm., rather short and thick.

Color characters: General color light yellow; about the base of each cornicle a small
reddish patch. Eyes dark red or brownish; cornicles dusky. Legs pale yellow with
the tarsi and possibly the distal extremities of the tibiae dusky to black. Antennae

Nov. 13, 1916

Rosy Apple Aphis

* 329

with the proximal half pale yellow and the distal half dusky or black. Labium pale
yellow tipped with dusky brown; body somewhat pulverulent.

Second instar. — Morphological characters. Antennae with the following measure¬
ments: Segment III, 0.192 to 0.232 mm., average 0.212 mm.; IV, 0.096 to 0.112 mm.,
average 0.104 nun.; V, (0.064 plus 0.256 mm.) to (0.08 plus 0.288 mm.), average (0.067
plus 0.272 mm.). Cornicles about 0.112 mm.

Color characters: Somewhat similar in color to the insects of the first instar, but
somewhat more pinkish as compared to the distinct yellow of the first instar.

Third instar. — Morphological characters. Antennae with the following measure¬
ments: Segment III, 0.208 to 0.24 mm., average 0.22 mm.; IV, 0.160 to 0.192 mm.,
average 0.177 mm.; V, 0.128 to 0.150 mm., average 0.14 mm.; VI, (0.064 plus 0.288
mm.) to (0.088 plus 0.384 mm.), average (0.083 plus 0.342 mm.). Cornicles 0.176 to
0.224 mm., average 0.198 mm.

Color characters: Much darker than the earlier instars, taking on the rosy tint of the
adult insects and the gray slaty powdering met with on the adult.

Fourth instar. — Morphological characters. Antennae as follows: Segment III,
0.304 to 0.4 mm., average 0.352 mm.; IV, 0.208 to 0.328 mm., average 0.271 mm.; V,
0.144 to 0.208 mm., average 0.187 mm.; VI, (0.08 plus 0.352 mm.) to (0.102 plus 0.504
mm.), average (0.094 plus 0.433 mm-)* Cornicles 0.24 to 0.288 mm., average
0.259 mm.

Color characters: Similar to the previous instar, although darker and more nearly
like the adult. .

Fifth instar (adult) (PI. 21, C). — Morphological characters. Antennae with the
following measurements: Segment III, 0.43 to 0.656 mm., average 0.52 mm.; IV,
0.256 to 0.48 mm., average 0.348 mm.; V, 0.192 to 0.288 mm., average 0.235 mm.;
VI, (0.096 plus 0.4 mm.) to (0.128 plus 0.592 mm.), average (0.108 plus 0.486 mm.).
Cornicles 0.336 to 0.496 mm. Body rotund; upper surface of the head often with two
rather prominent tubercles and a similar pair on the dorsum of the last two abdominal
segments. Cauda short, abruptly conical, setose, cornicles distinctly flanged and
slightly curved, imbricated. Antennae not distinctly imbricated, except the distal
segments. Labium extending to between the second and third pairs of coxae. Lateral
tubercles distinct.

Color characters: General color rosy brown, having a pinkish cast, owing to a
powdery covering. Some of the older specimens are almost of a purplish color whereas
younger specimens are decidedly reddish pink. Antennae yellowish, with the distal
extremity black. Legs yellowish, the tarsi and the distal extremities of the tibiae
black or brown. Labium tipped with brown. Cornicles and eyes black.

Location: Occurring in colonies within the curled leaves of the apple.

duration of nymphal stages

The average length of nymphal life of the wingless spring form is
9 to 10 days. This period is divided about equally among the four
immature stages. Occasionally one stage will be a day or even more
longer than the others, and such retardation of growth may occur in any
one of the four stages. These variations are apparently due almost en¬
tirely to temperature conditions, although in the experiments occasional
effects were due to poor food conditions.

REPRODUCTION

Reproduction commenced almost invariably about 24 hours after the
insect had attained maturity and continued for a period of from 12 to 26

330

Journal of Agricultural Research

Vol. VII, No. 7

days. In the experiments, with one exception to be noted later, the
adults died within one or two days after depositing their last young.
The average number of young deposited by 14 females was 121.8 per
insect. The greatest number deposited by one female was 180 young,
the smallest 66 young. It should be noted that the female which de¬
posited only 66 young was abnormal. The average daily production
for the 14 insects was 5.7 per female. The greatest daily average for
one female was 14.3 young, the mother producing 43 young in 3 days.
One other aphid produced 60 young in 5 days. In a general way the
aphids of the earlier generations produced more young than did those
which came later. No exact statement can be made, however, as there
was much variation in the matter.

LONGEVITY

The average total length of life of 13 of the aphids used in obtaining the
data on reproduction was 31 .4 days.

As has been previously stated, all but one of the insects in question
died within one or two days after depositing their last young. One insect,
however, gave birth to only 66 young and then lived for 1 1 days more
without reproducing. This aphid was killed on the eleventh day and
sectioned. When killed, the insect was apparently perfectly normal,
except for the fact that the abdomen was darker in color than that of the
other aphids. This dark color was diffused irregularly over the abdomen
and appeared to be produced by some change within the body of the
insect, rather than to be simply a case of melanism in the hypodermal
coloring. The insect was very plump; in fact, much plumper than were
other aphids which had produced even less young.

On sectioning it was found that there was an almost complete dis¬
integration of the reproductive system, only small isolated portions being
present. Moreover, there were the remnants of two half -grown embryos
lying free in the coeloma. The fat body had made an excessive growth,
almost completely filling the abdomen, and being abundant in the thorax.
The digestive canal was apparently perfectly normal, as were the other
organs of the body.

It is very interesting to note in this connection that three examples of
a similar nature were found in experiments on the life history of Aphis
avenae.

INTERMEDIATE FORM

* During the spring of 1914 two intermediates (PI. 21, D) of this species
were reared. These two insects are of particular interest since the other
intermediates reared have all occurred among the summer forms, the
winged insects of which migrate, if at all, to other plants of the same
species as those upon which the wingless insects feed. These two insects,
however, were reared in the spring upon apple. Had they become

Nov. 13, 1916

Rosy Apple Aphis

33i

winged they would have been unable to exist upon apple, but must have
flown to plantain. Any young which they might possibly have deposited
upon apple would have died soon after birth.

The two intermediates in question continued feeding upon apple after
becoming adult, produced their young normally, and these, in turn,
produced spring migrants, which left the apple. In other words, the two
intermediates not only evince a change of form from winged toward
wingless, but also a like change in habit.

In discussing the intermediates of Aphis pomi the writers (9) have
advanced the suggestion that the winged form and bisexual reproduc¬
tion represent the more primitive condition among aphids and that
these insects are at present in an active state of variation toward a
wingless form and parthenogenetic reproduction. If this supposition is
correct, the present examples would indicate that the alternation of
hosts is a more primitive condition, even possibly that the aphids were
originally general feeders and that some of them are varying toward
forms which will feed only upon one host.

DESCRIPTION

Morphological characters: Antennae of about equal proportions in both specimens.
Segment III, 0.576 mm.; IV, 0.352 mm.; V, 0.288 mm.; VI, baseo.112 mm.; unguis,
0.576 mm. One specimen has 6 sensoria near the distal extremity of III, 15 on IV,
and 3 on V. The other specimen has both antennae with the following sensoria;
32 and 38 on Segment III, 17 and 21 on IV, 3 and 4 on V. Cornicles in one specimen
0.4 mm. long and in the other 0.448 mm. Winged fhoracic characters absent. Wings
represented by small padlike structures 0.32 mm. long in one specimen and 0.16 mm.
long in the other.

Color characters: General appearance and color resembling those of the wingless
form.

SPRING MIGRANT
DESCRIPTION

Since the forms of the first three instars show little difference between
those which become wingless and those which become winged no descrip¬
tion is here given of the first three instars of the spring migrant.

Fourth instar (pupa). — Morphological characters. Antennae with the following
measurements: Segment III, 0.384 to 0.432 mm., average 0.414 mm.; IV, 0.24 to
0.32 mm., average 0.28 mm.; V, 0.176 to 0.216 mm., average 0.2 mm.; VI, (0.08
plus 0.4 mm.) to (0.112 plus 0.512 mm.), average (0.96 plus 0.433 mm.). Cornicles
0.256 to 0.32 mm., average 0.272 mm.

Color characters: Thorax and wing pads pink, shaded with dusky at the tips of the
pads. Top of head and first two antennal segments bluish dusky; Segments III and
IV of antennae whitish, distal segments black. Abdomen slaty blue, the embryos show¬
ing through as yellowish white patches. The covering of mealy wax gives a grayish cast
to the abdomen. Between the cornicles and caudad of them a dull rusty area, not a
bright rusty area as in the wingless individuals. Lateral and caudal tubercles
showing as minute dark brown spots. Cornicles brownish black. Legs whitish
with the exception of the tarsi and the distal tips of the tibiae. Labium tipped with
black or dark brown. Byes dark brown.

332

Journal of Agricultural Research

Vol. VII, No. 7

Fifth instar (adult) (PI. 20, C). — Morphological characters. Antennae as follows:
Segment I, 0.08 mm.; II, 0.064 mm.; Ill, 0.56 to 0.72 mm., average, 0.676 mm.; IV,
0.368 to 0.464 mm., average, 0.419 mm.; V, 0.240 to 0.32 mm., average, 0.272 mm.;
VI, (0.112 plus 0.504 mm.) to (0.128 plus 0.68 mm.), average (0.116 plus 0.582 mm.).
Segments I and II strongly imbricated on their inner margins and armed with a few
short spines. Segment III imbricated and armed with 53 to 60 oval, double rimmed,
protruding sensoria which give the segment a knotted appearance. These are dis¬
tributed over the entire segment; and a number of short spinelike hairs are also
present. Segment IV similar to segment III but with from 22 to 29 sensoria. Seg¬
ment V also similar, but the smaller number of sensoria (4 to 10) causing the imbrica¬
tions to appear more distinct. The distal sensorium on this segment is the usual
fringed one and not similar to the others on the segment. Segment VI strongly
imbricated throughout and with the usual sensory group at the base of the unguis.
Antennae placed on small frontal tubercles which are notched within and armed
with a few capitate hairs; vertex almost straight in some specimens, while the median
ocellus protrudes in others; crown with a pair of tubercles placed between the com¬
pound eyes. This character is, however, not a constant one, as the writers have
many specimens in which these tubercles are lacking. Prothorax with a prominent
lateral tubercle on each side. In some specimens there are two of these tubercles
on one side, and in the writers* specimens this seems always to be the left side. Fore¬
wing 2.48 mm. long and 1.008 mm. wide at its broadest part. Venation usually
normal, first branch of media slightly nearer its insertion than it is to the tip, radial
sector considerably curved, stigma 0.56 mm. long on the costal margin; hind wing
1.6 mm. long, hamuli 0.56 mm. from the distal extremity. Abdomen with distinct
lateral tubercles and with prominent dorsal tubercles on the two caudal segments;
cauda short, conical, setose, and armed with usually three pairs of long curved spine¬
like hairs; anal plate bluntly conical or rounded, densely setose, and armed with many
curved spinelike hairs. Cornicles cylindric, slightly curved, faintly imbricated and
distinctly flanged; length 0.288 to 0.384 mm.; legs slender, hind tibiae 1.07 to 1.1
mm. ; hind tarsus about 0.096 mm. long.

Color characters: General color brownish green, marked with black as follows: Head
above uniform black, the prothorax and the thoracic shield black, the margins, how¬
ever, near the wing insertions without black markings. Abdomen with a large black
quadrate patch on middle of dorsum surrounded by a narrow unmarked area; cephalad
of this patch there are often a number of small transverse markings, and caudad of it a
semicircular black marking which includes the insertions of the cornicles; this is
sometimes fused with the quadrate patch; margins of the abdomen with three large
rounded black spots cephalad of the insertion of the cornicles and with sometimes a few
smaller markings, lateral margins of the thorax with a large black spot. Below marked
with black as follows: Vertex, trophic tubercle, and margins and tip of labium;
a small spot on the prothorax; sternal plate; a band between and encircling the
hind coxse ; cauda an d anal and genital plates. Antennae black ; coxae black , trochanter
and proximal extremity of femur black; remainder of femur, the distal extremity
of the tibia, the tarsus, and claws black; remainder of tibia yellowish brown; cornicles
black. Wings clear, veins thin, stigma dusky.

Location : Found in the curled leaves of the apple and on the leaves and stems of
the plantain reproducing.

LENGTH of nymphal life

The winged form requires about two days more than the wingless one
for its immature stages, though this period may vary in exceptional cases
from 1 to 3 days. Thus, the total length of the period is from 11 to 13
days, the usual length being 12. The two extra days are spent in the
fourth or pupal stage, the pupae being from 2 to 3 days old when the
wingless form becomes adult.

Nov. 13, 19x6

Rosy Apple Aphis

333

REPRODUCTION

The migrants appear to commence reproduction in about 1 to 2 days
after settling upon the rib grass. The period for those transferred in the
experiments varied from 1 to 5 days. It is very probable, however, that
those insects which required the longer period were transferred too soon
after becoming winged, since in the field the migrants usually remain on
the apple for one or two days at least after the last molt.

Ten migrants produced an average of 18 young, varying from 10 to 29.
The average reproduction period for these individuals was 5.8 days,
varying from 3 to 9 days. The daily average was 3.2 young per female.
It is very interesting to note that without exception the mothers brought
forth more young on the first day of reproduction than on any of the
following days, the numbers of young ranging, with one or two excep¬
tions, from 7 to 10 for that day. It will be noted later that the repro¬
duction of fall migrants resembles that of the spring winged form.

longevity

The total average length of life for the 10 insects which have just been
mentioned was 25^ days. As stated when discussing the length of life
of immature forms, the period spent on apple averages 12 days, plus 1 or 2
days as adults. The average length of life on rib grass was 12% days,
varying from 3 to 29. All but two of the insects observed lived for at
least 2 days after producing their last young, while one lived in this way
for 15 and another for 22 days. These conditions also will be found to
exist in an exaggerated form among the fall migrants.

MIGRATIONS OF THE) SPECIES

For many years entomologists have recognized the importance in the
life history of this species of its secondary host. The actual discovery
of this host was first made by Theodore Pergande, although no record
was published. It would appear that Pergande observed the aphids
under greenhouse conditions, for his first date referring to the plantain
is January 25, 1882. His manuscript note reads as follows: “Large
numbers of this aphid are noticed to-day on leaves of the narrow-leaved
plantain. * * * They are found in all stages; larva, wingless females,
pupae and winged insects.” Several of the specimens referred to were
mounted on slides and these have been examined by the senior writer
and are without a doubt A. imlifoliae.

Since Pergande published no account of his findings, it had been
believed by entomologists that the alternate host of the species had
never been located. This is the view set forth in all the publications
on the species. Ross (8, p. 23), however, reported his transfers made
to Plantago major and P. lanceolata; and since his note appeared, these
plants have become fairly well known as the alternate hosts of the
species. The writers’ observations on the species in connection with
plantains were begun in 1913 when the senior writer made transfers to
64311° — 16 - 3

334

Journal of Agricultural Research

Vol. VII, No. 7

broad-leaved plantain. Other transfers were made during 1914, but it
was not until the spring of 1915 that a large and definite migration to
plantains was noted in the field. Winged specimens were then observed
alighting in numbers and reproducing upon the underside of the leaves
and on the long flower stems of the rib grass (PI. 23). Ross (8) reported
that he was able to rear the species throughout the summer on apple.
This the writers were unable to do even with a very few insects on a
plant; all of the lines carried ultimately produced winged forms and
migrated. It must be remembered also in this connection that the
writers selected wingless and winged forms from each mother in order
to obtain offspring. It is quite possible, however, that the lines carried
by Ross had less of a tendency toward winged forms throughout than
had those of the writers. Since the migration does not occur in any
definite generation but is scattered throughout the apple life of the insect
from the second generation onward, such a condition of affairs might
easily occur. The following fact also is df importance here: The inter¬
mediates discovered by the writers remained upon apple and there
reproduced, thus taking on the nature of the primary wingless forms.
This would indicate that ultimately the species may become a permanent
apple species like Aphis pomi , and the fact that Ross had carried it
through the season upon apple would seem to show that this tendency is
further advanced in the Ontario region where Ross conducted his experi¬
ments than in the Virginia region where the writers reared their insects.

The fact that spring migrants occurred in every generation on apple
in the writers' experiments from the third to the eighth caused this
migration to spread over a long period of time, from May 20 to July 1
at Vienna, Va. The pupae of the spring migrants very often do not
feed but may be found under loose bark and in the crevices of the limbs
and trunk. Many, however, remain within the curled leaves. In these
places the winged form is produced and in a few days it migrates to
the rib grass and settles upon the underside of the leaves and on the
flower stems. Here it produces its young. Occasionally some winged
forms reproduce upon the broad-leaved plantain, but at Vienna this is
rare as compared to the large number that migrate to the rib grass.

The fall migrants leave the plantains about 1 or 2 days after becoming
adult, and fly to apple trees, where they settle upon the under surfaces
of the leaves and commence feeding. Reproduction usually begins
within 24 hours after the insect reaches the apple, though the migrants
may feed for two or three days before producing young.

When mature, the males leave the plantain, flying to apple. If the
oviparous females are not fully mature when the males arrive, the
latter settle down beside them and feed until such time as copulation
can take place.

Recently, since this paper was prepared, Brittain (12, p. 16, fig. 2)
has reported his generation experiments on plantain.

Nov. 13, 1916

Rosy Apple Aphis

335

SUMMER FORMS

The usual summer form of this species is wingless. In fact, among over
1,000 individuals reared to maturity during the summer only 6 specimens
of the winged form were observed. One of these died when a pupa. This
latter insect belonged to the sixth generation from the spring migrant.
The other 5 insects were apparently members of the third generation on
plantain, although this was ascertained with certainty for only one indi¬
vidual, the other form occurring in experiments which had been set aside
merely to maintain a surplus stock of material. Only 2 of these winged
insects reproduced, their progeny being normal summer wingless aphids.
Thus, while the winged insects may occur on plantain they are of no par¬
ticular importance in the life history of the species.

It should be noted in connection with this form that many wingless
insects were reared on the plants in the generations with these winged
aphids and in the generations preceding and following them, the ratio
being over 100 to 1. The production of this form can hardly have been
due, therefore, to food conditions. •

A maximum of 14 generations of the summer form were reared on
plantain, with a possible theoretical minimum of 4 generations.

SUMMER WINGLESS VIVIPAROUS FEMALE
DESCRIPTION

First instar. — Morphological characters. Antennae with the following measure¬
ments: Segment III, 0.176 to 0.224 mm., average 0.196 mm.; IV, (0.048 plus 0.208
mm.) to (0.064 plus 0.248 mm.), average (0.056 plus 0.238 mm.). Cornicles 0.08 to
0.096 mm., average 0.084 mm.

Color characters: Pale greenish yellow with ultimately a rusty color across the
abdomen between the cornicles. The color sometimes extends into the cornicles.

Second instar. — Morphological characters. Antennae with the following measure¬
ments: Segment III, 0.16 to 0.24 mm., average 0.209 mm.; IV, 0.102 to 0.128 mm.,
average 0.107 mm.; V, (0.064 plus 0-24 mm.) to (0.072 plus 0.32 mm.), average (0.065
plus 0.288 mm.). Cornicles 0.112 to 0.144 mm., average 0.136 mm.

Color characters: Similar to the specimens of the first instar.

Third instar. — Morphological characters. Antennae with the following measure¬
ments: Segment III, 0.128 to 0.176 mm., average 0.164 mm.; IV, 0.128 to 0.184 mm.,
average 0.156 mm.; V, 0.128 to 0.144 mm., average 0.136 mm.; VI, (0.064 plus 0.288
mm.) to (0.088 plus 0.392 mm.), average (0.076 plus 0.342 mm.). Cornicles 0.0176 to
0.208 mm., average 0.19 mm.

Color characters: Similar to those of the second instar.

Fourth instar. — Morphological characters. Antennae with the following meas¬
urements: Segment III, 0.24 to 0.288 mm., average 0.264 mm. ; IV, 0.176 to 0.24 mm.,
average 0.201 mm.; V, 0.144 to 0.192 mm., average 0.166 mm.; VI, (0.08 plus 0.36
mm.) to (0.096 plus 0.424 mm.), average (0.086 plus 0.39 mm.). Cornicles 0.24 to
0.272 mm., average 0.265 mm.

Color characters: The same as those of the adult form.

Fifth instar (adult) (PI. 22, A). — Morphological characters. Antennae slender,
extending beyond the tips of the cornicles; measurements as follows: Segment III,
0.416 to 0.512 mm., average 0.454 mm.; IV, 0.256 to 0.368 mm., average 0.305 mm.;

336

Journal of Agricultural Research

Vol. VII, No. 7

V, 0.192 to 0.256 mm., average 0.225 mm-; VI, (0.088 plus 0.368 mm.) to (0.112 plus
0.512 mm.), average (0.097 to 0.441 mm.); segments imbricated, armed with a few
hairs, but without sensoria. Antennal tubercles distinct; head without dorsal tuber¬
cles met with in the wingless form from apple. Legs slender ; hind tibia 0.656 to o. 784
mm. long. Cornicles very slender, slightly curved, distinctly flanged and imbricated ;
length, 0.384 to 0.464 mm., average 0.443 mm- Abdomen without the prominent
caudal tubercles met with in the apple form; cauda conical, rather elongate, more
slender than that of the apple form, armed with three pairs of hairs and imbricated
by rows of minute setae. Form more elongate than the spring wingless forms, 1 .36 by
0.8 mm.

Color characters: Color creamy-yellow with a slight brownish or even purplish cast;
eyes dark brown; distal portions of antennae, tips of cornicles, tip of labium, and
tarsi brown or dusky. Between the cornicles and inclosing their insertions is a band
of rusty red, this color sometimes also extending into the proximal portion of the cor¬
nicles; the red eyes of the embryos showing through the abdomen of the adult.

Location : Found in colonies on the under surface of the leaves of rib grass ( Plantago
lanceolate) and also on the flower stems; rarely also on other species of the genus.

DURATION OR NYMPHAL STAGES

This form of the species became adult in from 8 to 12 days, the period
varying with the prevailing temperature. Insects born from the 10th
to the 20th of June required 10 days for this period, while those bom in
July matured in 8 to 9 days. Later the period increased again to 12
days in September.

REPRODUCTION

The average number of young produced by 35 adults was 65.2, the
maximum being 108 (produced by two insects) and the minimum 12.
The average length of the reproductive period was 19 days, thus varying
from 5 to 35 days. The average daily production of these 35 adults
was 3.4 young per mother. The greatest average daily production for
one adult was 5.4, this insect producing 43 young in 8 days; the lowest
was 2.2, 77 young in 35 days. One insect brought forth 14 young in one
day while 8 produced 10 or more in a like period. Of the two winged
insects which reproduced, one brought forth 21 young in 7 days, and the
other 13 young in 3 days. The first of the two lived for a few days after
reproduction ceased.

longevity

•

The average total length of life for this form was 28 days, varying
from 14 to 45 days. In general the insects bom during the early part
of the summer, particularly in June, were longer lived than any which
followed, the latter all equaling or exceeding the average for the entire
summer.

SUMMER WINGED VIVIPAROUS FEMALE

The early instars of this form are very similar to those of the summer
wingless viviparous female and those of the fall migrant. A description
of the fifth instar (adult) only will be given here. Since the form was

Nov. 13, 1916

Rosy Apple Aphis

337

of very rare occurrence throughout the summer, it is not worth while
to give averages for these very few insects.

Fifth instar (adult). — Morphological characters. Antennae with the following
measurements: Segment III, 0.688 mm.; IV, 0.496 mm.; V, 0.304 mm.; VI (0.128
plus 0.648 mm.). Cornicles 0.352 mm. Segment III with about 58 circular sensoria,
IV with 34 similar ones, V with 6 or 8 arranged evenly along the segment. Head with
tubercles above and the last two segments of the abdomen with a pair of dorsal
tubercles.

Color characters: Similar to those of the spring migrant, the abdomen above nearly
uniform brownish fclack or with a distinct patch.

DIMORPHIC REPRODUCTION

The writers have already noted in two other species of aphids, A. pomi
and A . avenae , that the late summer and early fall individuals of the summer
form may produce two forms of young. In A. malifoliae both summer
wingless and fall migrants, or fall migrants and males, may be produced
by one mother. In the experiments the former combination occurred,
in a general way, earlier than did the latter. Records of the production
of all three forms by one mother were not obtained.

As we found to be the case with A. avenae , when one mother produced
both fall migrants and males, the latter were usually her last progeny.

FALL FORMS

Among the fall forms may be included the fall migrant which leaves the
plantains and flies to apple, the oviparous females which are deposited
upon the apple leaves by these fall migrants, and the males which fly
from the plantains to fertilize the oviparous females.

FALL MIGRANT
DESCRIPTION

The early instars of the fall migrant are practically the same as those
of the summer wingless form. It is therefore unnecessary to give descrip¬
tions of any instars excepting those following:

Fourth instar (pupa). — Morphological characters. Antennae with the following
measurements: Segment III, 0.4 to 0.464 mm., average 0.427 mm.; IV, 0.288 to 0.344
mm., average 0.315 mm.; V, 0.208 to 0.24 mm., average 0.227 mm.; VI, (0.096 plus
0.488 mm.) to (0.112 plus 0.56 mm.), average (0.107 plus 0.518 mm.). Cornicles 0.288
to 0.320 mm., average 0.305 mm.

Color characters: Similar to the pupae of the spring migrant.

Fifth instar (adult) (PI. 21, A). — Morphological characters. Antennae with the
following measurements: Segment III, 0.624 to 0.800 mm., average 0.689 mm.; IV,
0.376 to 0.528 mm., average 0.472 mm.; V, 0.256 to 0.336 mm., average 0.308 mm,;
VI, (0.112 plus 0.56 mm.) to (0.152 plus 0.712 mm.), average (0.132 plus 0.632 mm.).
Segment III with about 60 circular elevated sensoria, IV with about 27, V with about
6 sensoria. Head without tubercles above as is usual in the spring migrant.
Prothorax with very small lateral tubercles. Abdomen also with very small ones.
Cornicles 0.304 to 0.384 mm., average 0.348 mm. Cauda conical, short; last two seg¬
ments of the abdomen without tubercles above as in the spring migrant. Otherwise
as in that form.

338

Journal of Agricultural Research

Vol. VII, No. y

Color characters: Similar to those of the spring migrant. Abdomen usually with
one large central black patch, three or four marginal patches, and a transverse band
caudad of the cornicles, this band extending cephalad at its edges to touch the base of
the cornicles. In some specimens the whole of the abdomen appears a uniform black
and the legs also do not show the yellow regions usually met with..

Location: Upon the leaves of the plantain, usually upon the underside, and upon the
underside of the leaves of the apple, depositing young oviparous females.

LENGTH OF NYMPHAL PERIOD

The length of the nymphal period varied from 13 or 14 days in early
September to as much as 24 days for aphids bom in October. The
average period was 16 to 18 days. No figures for the various instars can
be given since they varied greatly with the temperature, so that, while in
one case the first instar might be the longest, in others the third or fourth
would be.

REPRODUCTION

Sixteen fall migrants produced 114 yoting, an average of 7.1 young per
mother. The greatest number produced by one mother was 13 (or
more), two insects together producing 25; the smallest number was 3.
The average length of the reproductive period of 25 insects was 5 days,
varying from 1 to 10 days. In every case all but one or two of the young
were produced within a period of 1 to at most 2 days. Later at periods
varying from 3 to 8 days one or two more young might be produced.

LONGEVITY

The adults of this form usually lived for a considerable period after
reproduction ceased, this period varying from 2 to (in one case) over
40 days.1 The average length of this post-reproductive period for 15
insects was 23.6 days.

The longest total life recorded was 62 days, while the average was
about 45 days.

The conditions with regard to the rate of opposition found to exist
among both spring and fall migrants are very interesting, since they
are the exact opposite of the theoretical conditions for insects. The
general statement is commonly made that those females which produce
all their eggs (01 progeny) in a short period die very soon afterwards,
while the females which live for a long time are those which produce
only a few eggs (or young) daily. In this case, however, the adults
produce their young in a short period and then proceed to live for many
days. The theory is based, in part at least, on the proposition that the
insects which produce all their offspring or progeny within a short
period become exhausted, and, of course, with these migrants only a
small number of young are produced. Still, in so far as the general
theory endeavors to explain the causes of a long adult life, it fails for
these forms.

1 Experiment closed before insect died.

Nov. 13, 1916

Rosy Apple Aphis

339

MALE

DESCRIPTION

As with the fall migrant, the earlier instars of the male are similar
to those of the wingless viviparous form ; only the fourth and fifth instars,
therefore, are here described.

Fourth instar (pupa). — Morphological characters. Antennae with the following
measurements: Segment III, 0.416 mm.; IV, 0.32 mm.; V, 0.216 mm.; VI, (0.092
plus 0.504 mm.). Segment III with about 58 sensoria, IV with about 27, V with
about 7. Cornicles about 0.284 mm.

Color characters: Somewhat similar to those of the spring pupa.

Firth instar (adult) (PI. 21, B). — Morphological characters. Antennae with the
following average measurements: Segment III, 0.728 mm.; IV, 0.462 mm.; V, 0.302
mm.; VI, (0.113 plus 0.657 mm.). Cornicles 0.304 mm., slightly thicker than those
of the migrants. Head without tubercles above, prothorax with very minute lateral
tubercles, sometimes with none visible; lateral abdominal tubercles very small;
last two segments of abdomen without tubercles above as in the spring migrant.
In this respect the male is like the fall migrant.

Color characters: Similar to those of the fall migrant. In many cases the large
central abdominal spot is broken into a number of irregular bands and the crossband
caudad of the cornicles is often divided, the total forming a series of cross stripes on
the abdomen.

production

As has been noted previously, the males, when produced by a mother
also bearing fall migrants, were usually the last progeny of the mother.
The few mothers which produced only males were sisters of fall migrants.
In this way the production of males commenced at about the same time
as did that of the oviparous females.

nymphal stages

. The period spent in the immature stages varied from about 20 to 25
days. No general statement can be made as to the duration of the
various instars, since these varied greatly among the individual insects.

migration

When mature, the males leave the plantain, flying to apple. If the
oviparous females are not fully mature the males usually settle down
beside them and feed. *

OVIPAROUS FEMALE
description

First instar. — Morphological characters. Antennae with the following measure¬
ments: Segment III, 0.144 to 0.16 mm., average 0.148 mm.; IV, (0.048 plus 0.184
mm.) to (0.056 plus 0.216 mm.), average (0.049 plus 0*196 mm.). Cornicles 0.056 mm.

Color characters: Very pale greenish yellow with brown eyes.

Second instar. — Morphological characters. Antennae with the following measure¬
ments: Segment III, 0.136 to 0.16 mm., average 0.142 mm.; IV, 0.064 to 0.08 mm.,
average 0.07 mm.; V, (0.056 plus 0.216 mm.) to (0.064 plus 0.248 mm.), average
(0.059 plus 0.232 mm.). Cornicles 0.08 mm.

Color characters: Slightly darker than in the first instar.

340

Journal of Agricultural Research

Vol. VII, No. 7

Third instar. — Morphological characters. Antennae with the following measure¬
ments. Segment III, 0.192 to 0.224 mm., average 0.206 mm.; IV, 0.088 to 0.104 mm.,
average 0.096 mm. ; V, (0.064 plus 0.264 mm.) to (0.072 plus 0.296 mm.), average (0.065
plus 0.273 mm.). Cornicles 0.112 mm.

Color characters: Somewhat more orange-yellow than the earlier instars, with more
or less distinct brownish areas about the bases of the cornicles.

Fourth instar. — Morphological characters. Antennae with the following measure¬
ments. Segment III, 0.144 to 0.184 mm., average 0.169 mm. ; IV, 0.112 to 0.144 mm.,
average 0.128 mm.; V, 0.112 to 0.128 mm., average 0.12 mm.; VI, (0.064 plus 0.296mm.)
to (0.08 plus 0.344 mm.), average (0.075 phis 0-3r mm.). Cornicles 0.144 mm.

Color characters : General color yellow-orange, with brownish red patches around the
bases of the cornicles; head grayish, eyes brown to black.

Fifth instar, adult (PI. 22, B). — Morphological characters. Antennae with the
following measurements: Segment III, 0.208 to 0.256 mm., average 0.241 mm.; IV,
0.152 to 0.192 mm., average 0.169 mm.; V, 0.128 to 0.16 mm., average 0.142 mm.; VI,
(0.08 plus 0.304 mm.) to (0.088 plus 0.36 mm.), average (0.081 plus 0.329 mm.).
Cornicles 0.192 to 0.216, average 0.204 mm. Antennae without sensoria. Head
occasionally with a pair of tubercles above but the last two segments of the abdomen
without such tubercles. Cauda short and abruptly conical.

Color characters: Similar to those given for the previous instar excepting that the
colors are a little more distinct. The eggs show through the body as dark areas.

Location: Found on the underside of the apple leaves feeding, or on the twigs or in
the axils of the buds depositing eggs.

nymphal stages

The oviparous females required a period of from 20 to about 28 days
for the immature stages. As with the males, the duration of the various
instars varied greatly with the individuals, the principal cause of this
being the variation of the prevailing temperatures.

MATING

Mating occurred mostly on the twigs. Males mated with several
females and in some cases the oviparous females mated at least twice.
As was the case with A . avenae , the male endeavors to copulate with
every adult oviparous female he meets; and unless the female has just
mated, she does not endeavor to hinder the male.

OVIPOSITION

Nineteen ovipara laid a total of 120 eggs, an average of 6.3 eggs per
insect. Five of these oviparous females laid 36 eggs, an average of a
little over 7 eggs each. The highest record for one oviparous female is,
therefore, 8 eggs.

FEEDING HABITS

The results of the feeding of the rosy aphis are very noticeable, as the
leaves are much curled thereby. (Pi. 24, B.) The young stem mothers
crowd into the opening buds, and as the leaves grow they curl and twist
about the insects. As young are produced these reach other leaves and

Nov. 13, 1916

Rosy Apple Aphis

341

soon there is a large cluster of twisted leaves which is very conspicuous.
In orchards composed of large trees these curled bunches are usually met
with on the lower half of the trees and very often the small fruit §purs
on the larger limbs in the body of the tree are affected. On very young
trees, however, the feeding habits seem to differ in that the growing tips
of the branches are attacked. This causes them, as they develop, to
become twisted (Pi. 25, B), and as this growth hardens the limb is perma¬
nently deformed by being looped upon itself (Pi. 25, C). A branch will
sometimes become looped several times, thus causing a much deformed
tree unless the affected branches are cut out. In very young trees that
are badly infested this is often a good portion of the tree.

The insects may also be found feeding upon the fruit, and this they
cause to be reduced in size, irregular in shape, and somewhat gnarled,
or more or less pitted (PI. 25, A, D). In orchards composed of large
trees this damage to the fruit is often a factor of considerable importance.

As has already been stated, the normal summer host of the species
appears to be rib grass. During the season of 1915, however, the insects
were reared successfully on broad-leaved plantain {Plantago major) for
a period of two months. No record of the generations was kept in this
case and no transfers were made. The plant finally died. At that time
a strong healthy colony was living upon it and apparently under normal
conditions would have survived until fall. In several other experiments,
however, much difficulty was experienced in procuring successful trans¬
fers from the apple to this species of plantain. Rib grass appears to be
the host preferred. This is borne out by an examination of the occurrence
of the insects in orchard regions. Observations were made over a large
extent of orchard territory. Wherever the rosy aphis was found to be
very abundant the narrow-leaved plantain was common in and about the
orchards, the worst infested orchards being full of the growing plantain.

The writers* observations upon the insects, both in the experiments and
in the field, indicate that all parts of the rib grass are subject to attack.
The insects appear to feed with equal readiness upon the leaves, the stem,
and the flower stalks, and they may be found upon both sides of the
leaves. The greater number of the insects, however, fed on the under
surface of the leaves, especially along the veins.

In the fall the migrants alight on the underside of the apple leaves, and
when the oviparous females are produced these feed also upon the under¬
side of the leaves. The males, after migration from the plantains, often
may be found feeding with immature oviparous forms. So far as the
writers have been able to observe, these fall forms do not cause the leaves
to curl and twist as do the spring forms.

342

Journal of Agricultural Research

Vol. VII, No. 7

SUMMARY OF LIFE HISTORY

The eggs of this species begin hatching early in April (about April 8 in
1915) and hatching ceases in about a week (April 16 in 1915). The first
stem mothers begin reproduction about April 25. From 5 to 7 genera¬
tions of the spring forms occur on apple in Virginia, although Ross reports
the species all summer on apple in Ontario. The first, generation is wing¬
less. A few winged forms appear in the next generation and their per¬
centage to the wingless insects increases steadily in each generation until
finally all the insects produced become winged. Intermediates may also
occur, these acquiring the wingless habits and behaving like wingless
insects.

Migration to plantain commences about May 20, and most of the insects
have left the apple by about June 20. A few may continue on apple till
about July 1.

From 4 to 14 generations of the summer form occur at Vienna, Va.
These insects are practically all wingless, only a few occasional winged
insects appearing.

The first fall migrants become adult about the second week of Septem¬
ber (Sept. 13 in 1915; these insects were bom on Aug. 31). They
remain on the trees until after November 1. (In the writers’ experi¬
ments they were produced till a much later period, but in the field they
succumb to prevailing low temperatures more quickly than do either
oviparous females or males.)

Production of oviparous females commences about the middle to the
20th of September, but very few are produced till early in October and
their production is at its height about the middle of that month. Males
begin to appear early in October, at the time the oviparous females
begin to become adults, and the males also are most numerous about the
last of October and early in November.

Oviposition commences the middle of October and continues till the
oviparous females are all dead. Some oviparous females may oviposit
as late as the latter part of December in case excessive low temperatures
have not occurred before that time.

The life history as summarized is for the vicinity of Washington, D. C.

LITERATURE CITED

(1) Fonscoi«ombe, Boyer de.

1841. Description des pucerons qui se trouvent aus environs d’Aix. In Ann.
Soc. Ent. France, t. 10, p. 157-198. •

(2) Kai*tsnbach, J. H.

1843. Monographic der Familien der Pfianzenlause (Phytophthires). ... 222
p., 1 pi. Aachen.

(3) Koch, C. L.

1854. Die Pfianzenlause Aphiden. ... Heft 1. Niimberg.

Nov. 13, 1916

Rosy Apple Aphis

343

(4) Fitch, Asa.

1855. [Report on the noxious and other insects of the state of New York.]
In Trans. N. Y. State Agr. Soc., v. 14, 1854, p. 705-880, illus. Re¬
printed, Albany, N. Y., 1856.

(5) Sanderson, E. D. '

1901. Report of the entomologist. In Del. Agr. Exp. Sta. 12th Ann. Rpt.,
1900, p. 142-2 1 1.

(6) SchoutEdEN, H.

1903. Les aphidoc£cidies pal6arctiques. In Ann. Soc. Ent. Belg., t. 47, p.
167-195.

(7) Gillette, C. P., and Taylor, E. P.

1908. A few orchard plant lice. Colo. Agr. Exp, Sta. Bui. 133, 48 p., 1 fig.,
4 pi. (2 col.).

(8) Ross, W. A.

1915. Orchard insects. In 45th Ann. Rpt. Ent. Soc. Ontario, 1914, p. 22-24.

(9) Turner, W. F., and Baker, A. C.

1915. On an occurrence of an intermediate in Aphis pomi De Geer, In Proc.

Ent. Soc. Wash., v. 17, no. 1, p. 42-51, pi. 10.

(10) Theobald, F. V.

1916. Aphididse found on the apple in Britain and the description of a new

species from Africa. In Canad. Ent., v. 48, p. 202-213.

(11) Baker, A. C., and Turner, W. F.

1916. Morphology and biology of the green apple aphis. In Jour. Agr. Re¬
search., v. 5, p. 955-993, fig. i-4, pi. 67-75.

(12) Brittain, W. H.

1916. Some Hemiptera attacking the apple. In Proc. Ent. Soc. Nova Scotia,
1915, no. 1, p. 7-47, fig. 1-6.

PLATE 20

A. — Aphis sorbi : Spring migrant.

B. — Aphis kochii: Spring migrant.

C. — Aphis malifoliae: Spring migrant.

(344)

PLATE 2i

Aphis malifoliae:

A. — Fall migrant.

B. — Male.

C. — Spring wingless female.

D. — Intermediate form.

PLATE 22

A. — Aphis malijoliae : Summer wingless form.

B. — Aphis malifoliae: Oviparous female.

C. — Structural details of Aphis malifoliae, A, sorbi , and A. kochii.

a, A. sorbi: Segment VI of antenna of winged form.

b, A. malifoliae: Cornicle of spring wingless form.

c, A. malifoliae: Cornicle of summer wingless form.

d, A. malifoliae: Cauda of summer wingless form.

e, A. malifoliae: Cauda of spring wingless form.

f, A. malifoliae: Segment VI of antenna of winged form.

g, A. malifoliae: Segment, VI of antenna of stem mother.

h, A. kochii: Segment VI of antenna of winged form.

i, A. sorbi: Cauda of winged form.

j, A . kochii: Cornicle of spring migrant.

k , A . sorbi: Cornicle of spring migrant.

l, A. malifoliae: Cornicle of spring migrant.

m , A. sorbi: Segment III of antenna of spring migrant.

n, A. malifoliae : Segment III of antenna of spring migrant.

PLATE 23

Aphis malifoUae on its alternate host, Plantago lanceolata.

PLATE 24

A. — Broad-leaved plantain showing the effect of an attack by Myzus plantaginis.

B. — Apple leaves curled by colonies of Aphis malifoliae.

Plate 24

PLATE 25

A. — Rhode Island Greening apples deformed by Aphis malifoliae.

B. — Apple twigs twisted by colonies of Aphis malifoliae: Beginning of twisting.

C. — Apple twigs twisted by colonies of Aphis malifoliae: Twisted twig.

D. — Winesap apples deformed by A Mis malifoliae .

64311° — 16 - i

JMtNAl OF AfMILTIM RESEARCH

DEPARTMENT OF AGRICULTURE

Vol. VII Washington, D. C., November 20, 1916 No. 8

USE OF TWO INDIRECT METHODS FOR THE DETERMI¬
NATION OF THE HYGROSCOPIC COEFFICIENTS OF
SOILS 1

By Frederick J. Alway, Chief , Division of Soils , Agricultural Experiment Station,
University of Minnesota, and Verne L. Clark, formerly Assistant in Chemistry ,
Nebraska Agricultural Experiment Station

INTRODUCTION

The relative advantages and disadvantages of the moisture-equivalent
method for the indirect determination of the hygroscopic coefficients of
soils have been discussed in a previous article G).2 While in connection
with soil-survey reports the moisture equivalents may conveniently be
used to indicate the relative fineness of texture as a single-valued expres¬
sion,3 even the determination of these consumes far more time than is
desirable in connection with many soil studies.

A very simple, rapid method seemed to be offered by a formula derived
by Briggs and Shantz (4, p. 66) for the estimation of the hygroscopic
coefficient from the maximum water capacity, as determined by Hil-
gard’s method. Such an indirect method would prove extremely useful
if it could be relied upon to give results in at least fair accord with those
obtained by direct determination. The determination of the maximum
water capacity by Hilgard’s method requires only very simple apparatus,
consumes but little time, and can be carried out in the most poorly
equipped laboratory and doubtless even in an ordinary farm kitchen
(14). To test the reliability of this proposed method, we have made
determinations of the water capacity of 53 soils of which the hygroscopic
coefficients had previously been carefully determined.

1 The work reported in this paper was carried out in 1912 at the Nebraska Agricultural Experiment
Station, where the authors were, respectively, Chemist and Assistant in Chemistry.

3 Reference is made by number to “ Literature dted,” p. 359.

* The advantages of such a single-valued expression of the “different degrees of ‘heaviness* * of soils”
(7, p. 440) appears to have first been suggested by Hilgard in connection with his introduction of the deter¬
mination of the hygroscopic coefficient (6, p. xi).

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

*c . .

(345)

Vol. VII, No. 8
Nov, 20, 1916
Nebr.— 1

64312°— 16

346

Journal of Agricultural Research

Vol. VII, No. 8

A method which will give results sufficiently reliable and accurate for
many purposes, such as soil surveys, and yet be far more simple even
than the preceding and more economical of time where any very large
number of samples is to be handled, seems capable of development.
This method would be an indirect one based upon the determination of
the hygroscopic moisture in air-dried samples. As our work on this was
interrupted several years ago by the removal from Nebraska first of the
junior author and later of the senior author and there appears little
probability of either of them being able to continue the work, at least for
some time, the results are reported in the hope that in some other labo¬
ratory the limitations of the method may soon be determined.

HISTORICAL REVIEW

Hilgard’s method for the determination of the maximum water
capacity first proposed in 1893 (8, p. 256; 9, p. 74) has later been described
in various publications (12, p. 82; 10, p. 15; n, p. 209). From a com¬
parison of the hygroscopic coefficient with the wilting coefficient in the
case of a series of 17 soils and of the maximum water capacity with the
wilting coefficient in the case of another series of some 15 soils Briggs
and Shantz (4) have derived the formula:

Hygroscopic coefficient = (maximum water capacity— 21) X 0.234

Unfortunately only four of the samples appear in both series. The
data on these are reported in Table I. The calculated values agree
fairly satisfactorily with those directly determined.

Table .1. — Relation of the maximum water capacity to the hygroscopic coefficient as found

by Briggs and Shantz

Soil No.

Type of soil.

Maximum

water

capacity.

Hygroscopic coefficient.

Difference.

Determined.

Calculated.

7 .

Coarse sand .

23. 2
29.9

o- 5

0. s

2. I

0. O

2 .

Fine sand .

+ .6

8 .

. do .

28. 5

31-4

2. 3

I* 7

- .6

9 .

2. 3

2.4

+ • 1

Briggs and Shantz appear to have overlooked the work of Eough-
ridge (12), who, some 20 years previously, using some 40 California
soils ranging in texture from clays to sands, made a critical study of the
relation of both the maximum water capacity and the hygroscopic
coefficient to the mechanical composition. In the case of all of these he
determined the hygroscopic coefficient, the maximum water capacity,
both by weight and by volume, and the mechanical composition. In the
first three columns of Table II we give the portions of his data dealing
with the present subject, arranging the soils in the order of their hygro-

Nov. so, 1916 Determination of Hygroscopic Coefficients of Soils

347

scopic coefficients. In the fourth column are shown the hygroscopic
coefficients calculated from the maximum water capacities by the
Briggs-Shantz formula, and in the last the differences between these and
the directly determined values. While with about half the soils there is
a fair agreement, in the case of the others the divergence is so wide
that the ’calculated values would be quite misleading as to the relative
hygroscopicity.

Tabi^ II. — Concordance of the determined hygroscopic coefficient with that calculated by
the Briggs-Shantz formula from the maximum water capacity in the case of California
soils reported by Loughridge

Soil No.

Maximum

water

capacity.

Hygroscopi

Deter¬

mined.

c coefficient.

Calculated.®

Departure.

6 . .

54-9

14.53

7- 93

-6. 6o*

863... .

68.8

14. 20

11. 18

-3.02

188 .

59-6

13- 70

9.03

-4.67

643 . . •

64. 6

I3-5I

10. 20

“3- 31

1679 .

46. 0

II. 98

S- «S

-6. 13

68 . . .

59-8

II. I9

9. 08

—2. 11

676 .

54-4

II, II

7. 82

“3- 29

77 . .

68.9

10.38

II. 21

• 83

52- 4

10. 32

7-35

-2. 97

67 . ' . .

5°- 7

10. 26

6-95

-3-31

789 .

57-8

9- 74

8.6l

-1. 13

$06 . . .

55-6

9. 26

8. 10

— 1. 16

8 .

52.1

9. 18

7. 28

— 1. 90

82.3

9. 18

14-34

5. 16

63-5

7- 52

9-9S

2-43

1678 .

34-2

6. 41

' 3-09

“3- 32

m3- .

52- 3

6. 00

7*32

1. 32

ms .

S3- 8

5- 93

7. 68

1284 .

58.8

5.81

8. 85

3-<H

168 .

38.7

5-49

4- 14

-i- 35

1167 . .

41.7

5-38

4.84

“ • 54

1645 .

58* 4

5- 27

8.7s

3-48

4. • ■ ■ ' .

46. 6

5- 18

5-99

.81

52* 4

5- 14

7- 3S

2. 21

1647 .

49.9

5- 14

6. 76

— 1. 62

42. 0

5. 10

4.91

- - i9

1663 .

48. 0

5.06

6. 32

I. 26i

586.., .

47*3

4.86

6. 15

1. 29

1655 .

44-3

4. 20

5- 45

XI59 .

37-4

4. 00

3- 84

— . 10

i°55 . .

36-9

3. 62

3- 72

. 10

1197 .

34-8

3-48

3- 23

- *25

1148 .

. 41. s

3- 43

4. 80

I- 37

51 .

49-5

3- 17

6. 70

3- 53

9 . . .

58-9

2. 63

8. 87

6. 24

13° .

40. 7

2. 30

4. 61

2.31

1281 .

37-9

1. 98

3- 95

1.97

1147 .

28. 7

1. 84

I. 80

- .04.

1595 . . .

29- 7

1. i5

2. 04

.89

233 .

23. 0

• 79

•47

- .3 2.

a By the Briggs-Shantz formula: Hygroscopic coefficient “(maximum water capacity— 21) X0.234.

348

Journal of Agricultural Research

Vol. VII, No. 8

Caldwell (5, p. 15) in connection with a study of the permanent wilting
in plants has determined both the hygroscopic coefficients and the
water capacities of “a very pure sand, an ‘adobe’ clay loam containing
less than 0.1 per cent of humus, and nineteen artificial mixtures of these
two soils varying by increments of 5 per cent of loam from 95 per cent
sand with 5 per cent loam to 95 per cent loam with 5 per cent sand.”
The observed hygroscopic coefficients of the two soils were 2.41 and 7.51
and the corresponding values calculated from the water capacities 2.30
and 7.40, respectively. The calculated coefficients of the , artificial
mixtures agreed equally well with the directly determined values.
Clearly the concordance is here well within the limits of experimental
error; and if the same held true for a series of 21 natural soils with a range
in hygroscopic coefficient from 2.5 to 7.5, this indirect method would
leave little to be desired in respect to simplicity and reliability.

, It must, however, be emphasized that such artificial mixtures can
not take the place of natural soils in testing such a method, as, if a
formula holds true on each of two components separately, it should
hold equally true for all mixtures of these. Thus, to illustrate, in the
case of the soils studied by Loughridge (12), No. 77 and 1147, with
determined hygroscopic coefficients of 10.38 and 1.84 and calculated
values of 11.21 and 1.80, respectively, might be employed to prepare
artificial mixtures and all the latter should show water capacities such
that the calculated hygroscopic coefficients should agree with those
directly determined, while with similar mixtures of No. 643 and 51, with
determined values of 13.51 and 3.17 and calculated values of 10.20 and
6.70, respectively, the intermediate mixtures should be expected to
show a concordance, although neither of the two soils by itself did.

ESTIMATION OF THE HYGROSCOPIC COEFFICIENT FROM THE
MAXIMUM WATER CAPACITY

The determinations of the maximum water capacity were made in
triplicate or quadruplicate, the concordance of the determinations
being shown in Table III.

Table III. — Concordance of quadruplicate determination of the maximum water capacity

Soil No.

Individual determinations.

Aver¬

age.

A .

60. 5
62. 2
72.9

61. s

60.8

62.3

73- 4
62.3

61. 8

62. 5

l3-1

63-4

63.0
63-7
74*7
61. 9

61. s
02. 7

73*5
02. 2

B . . . . . . .

c .

D .

Nov. 2o, 1916 Determination of Hygroscopic Coefficients of Soils

349

In Table IV are reported the data on a series of 36 loess soils, including
the maximum water capacity, the hygroscopic coefficient calculated
from the preceding by the Briggs-Shantz formula, and the departure of
the calculated from the determined value, previously reported (1, p. 216).
The samples were collected from 30 virgin prairie fields in Nebraska, 5
near each of the 6 towns mentioned at the head of the columns. All
were taken from fields classified by the Federal Bureau of Soils as
“Marshall silt loam” or “Colby silt loam.” In each field 10 borings
were made to a depth of 6 feet and composite samples prepared of each
foot section, thus giving 6 samples from each field. From these were
prepared the samples used in this work, each of the latter being prepared
by mixing equal weights of the corresponding five field samples (1, p.
204). Each value for the directly determined hygroscopic coefficients
represents the average of 10 determinations (1, p. 214).

Table XV. — Maximum water capacities of loess soils in Nebraska and the hygroscopic
coefficients calculated from these by the Briggs-Shantz formula

MAXIMUM WATER CAPACITY

Depth.

Wauneta.

McCook.

Holdrege.

Hastings.

Lincoln.

Weeping

Water.

Average.

Foot.

I .

2. .

66.0

64.4

65-4

63. 0

72. 0
68.8

69.7

75-6

69-5

69.9

73-5

66.8

69 -3
68. 1

3 . . .

61. 6

60.6

67.4

73- 5

63-4

62. 5

64. 8

4 .

62. 2

59-7

61. 0

70. I

64. 2

63- 7

63- 5

5 .

59-4

58.1

61. 5

67. O

63- 4

62. 3

61. 9

6 .

55-8

57-8

62. 7

63. O

66. 2

60. 2

60. 9

Average . .

6l.6

60.8

65.6

69. 8

66. 1

64.8

CALCULATED HYGROSCOPIC COEFFICIENT

1 .

11. 3

11. 9

11. 4

11- 3

12. 3

11. 4

2 .

10. 1

9.8

11. 2

12. 8

11. 4

10. 7

11. 0

3 .

9* 5

9.2

10. 8

12. 3

9- 5

9- 7

10. 2

4 .

9.6

9.0

9-4

n- 5

10. 1

10. 0

9.9

5 .

8.9

8.8

9* 5

10. 6

9.9

9.6

9-5

6 .

8. 1

.8-3

9- 7

9.8

10. 6

9. 2

9-4

Average . .

9.6

9-2

10. 4

11. 4

10. 5

10. 2

DEPARTURE OF CALCULATED FROM DETERMINED HYGROSCOPIC COEFFICIENT

1 .

2. 2

o-3

1.8

2. 8

-0. 7

0. 2

1. 1

2 .

• 5

— 1. 1

. 0

1. 2

-3-°

“3* 0

3 . .

— . 2

-i-S

- • 5

— . 1

-4. 1

-4. 2

-1.7

4 . .

- -3

~ • 7

- .8

- 4

-2.9

-3.0

— 1. 2

5 .

— . 1

- -3

— . 1

-2.9

-3-o

— 1. 1

6 .

— . 2

- .8

•3

- -9

—2. 1

-3-3

— 1. 2

Average . .

•3

“ -7

. 1

• 5

—2. 6

-2.8

- .8

350

Journal of Agricultural Research

Vol. VII, No. 8

In the case of rather more than half of the samples the calculated
values agree satisfactorily with the directly determined hygroscopic
coefficients. In the case of the Lincoln and Weeping Water samples,
which are richest in clay and silt, and correspondingly poorest in very fine
sand (2, p. 407), the calculated values are much too low for all except
those samples from the surface foot. As an index of texture the calcu¬
lated hygroscopic coefficients would be very misleading. For instance,
the Wauneta samples, with an average of 9.6, would appear almost as
fine-textured as the Weeping Water soils with an average of 10.2, whereas
in fact actual mechanical analyses have shown the latter to contain
nearly three times as much clay, more than half as much again silt, and
only about one-fourth as much very fine sand (Table V). It is of interest
that in the case of those members of this series with which satisfactory
results are obtained by the Briggs-Shantz formula for the calculation of
the hygroscopic coefficient from the mechanical analysis the formula of
the same authors for the calculation of this value from the maximum
water capacity fails, and vice versa.

Table V Relation of calculated hygroscopic coefficients to the texture of Nebraska soils.a

Item,

Wauneta.

McCook.

Holdrege.

Hastings.

Lincoln.

Weeping

Water.

Maximum water ca¬

Per cent.

Per cent .

Per cent.

pacity .

Calculated hygroscopic

61. 6

60.8

65.6

69.8

66. I

64.8

coefficient . . .

Determined hygroscopic

9.6

9.2

IO. 4

II. 4

IO. 2

coefficient .

9.2

9.9

10.3

II. 0

J3* 1

13.0

Gravel, above 1.0 mm. . .
Gsarse sand, 1. 0—0.5

. 0

• 0

. O

. 1

. O

mm .

Medium sand, 0.5—0.25

. 1

. I

. 2

mm .

Fine sand, 0.25—0.10

•3

• 4

•3

•5

* 7

•3

mm .

Very fine sand, 0.10—

1.9

i-5

i-3

1.9

2.4

1. 2

0.05 mm .

49-3

3& 6

27.9

21. 6

10.3

12. 7

Silt, 0.05-0.005 mm -

42. 2

5i. 6

62.6

64- 5

68.5

Clay, 0.005—0.000 mm. .

6-3

7.8

7*7

II. 3

17. 6

17*3

a The mechanical analyses are from Alway and Rost (2, p. 407.)

The series of samples reported in Table VI includes, in addition to loess
soils from Nebraska, residual soils from the same State and a few samples
from the Southwestern States. The data upon both the hygroscopic
coefficient and the water capacity are the means of five or more, concordant
determinations. The range in texture is much wider than that of those in
Table IV and quite similar to that of the soils dealt with by Briggs and
Shantz (4, p. 67). Except in a few cases, the calculated value agrees fairly
weM with the directly determined value. The widest departures are shown
by the eastern Nebraska loess subsoil, similar in texture to the Lincoln and
Weeping Water subsoils reported in Table IV and by two surface soils,

Nov. so, 1916 Determination of Hygroscopic Coefficients of Soils

35i

one from New Mexico and the other from Arizona. Three residual soils
from Nebraska, No. 4, 5, and 6, which showed very similar hygroscopic
coefficients, differ distinctly in water capacity.

Table VI. — Comparison of the determined hygroscopic coefficient with that calculated from

the maximum water capacity

Soil

No.

Description of soil.

Maximum

water

capacity.®

Hygr

Directly

deter¬

mined.^

oscopic coeffi

Calculated

from

maximum

water

capacity.®

dent.

Departure
of calcu¬
lated coeffi¬
cient from
that found.

Dune sand, western Nebraska. .

25. 8

0. 6

I. I

a 5

Desert sand, Palm Springs, Cal .

28. 9

• 9

1.9

1. 0

Sandy subsoil, Palm Springs, Cal .

27. O

1. 1

1-4

-3

Desert sand, Orogrande, N. Mex .

27. I

i- 7

1.4

“ -3

Sandy surface soil, .western Nebraska. .

34-2

3*3

3- 1

— . 2

Sandy subsoil, A, western Nebraska. . .

31- 0

3-4

2-3

— 1. 1

Sandy subsoil, B, western Nebraska. . .

36. 0

3- 4

3- 5

. 1

Sandy loam subsoil, western Nebraska.

46. 3

5- 6

5- 9

. 3

Sandy loam surface, western Nebraska.

53- 4

7* 1

7. 6

. 5

Silt loam subsoil, A, western Nebraska.

57- 2

7. 6

8-5

• 9

Silt loam subsoil, B, western Nebraska.

55- 4

8. 2

8. 1

— . 1

Red loam surface, Cuervo, N. Mex _

49. 0

10. 0

6.6

“3-4

Silt loam surface, A, western Nebraska.

56.8

10. 1

8.4

-1. 7

Silt loam surface, eastern Nebraska. . . .

60. 9

10. 2

9- 7

— . 5

Silt loam surface, B, western Nebraska.

^3- 7

10. 5

10. 0

— . 5

Adobe surface soil, McNeal, Ariz .

60. 3

12. 9

9. 0

-3-9

Silt loam subsoil, eastern Nebraska -

6S- 7

13-3

—2. 8

® Determined by Mr. J. C. Russel.

6 Determined by Mr. G. R. McDole.

c Using Briggs-Shantz formula: Hygroscopic coefficient™ (maximum water capacity— 2i)Xo.234.

Thus, while the Briggs-Shantz formula with many soils gives values
fully in accord with those directly determined, with many others it
gives results so widely divergent that it can not be regarded as suf¬
ficiently reliable for studies of available soil moisture, or even for soil-
survey purposes.

ESTIMATION OF THE HYGROSCOPIC COEFFICIENT FROM THE HYGRO¬
SCOPIC MOISTURE

The hygroscopic coefficient indicates the maximum of hygroscopic
moisture, the amount found when a more or less completely dried soil
has been kept in contact with a saturated atmosphere at a constant
temperature until the moisture in the soil is in approximate equilibrium
with that in the atmosphere. Theoretically, actual equilibrium would
not be attained until the moisture content of the soil equaled that of the
same soil in actual contact with water (13, p. 448), but the time required
for this is so great that this theoretical consideration does not affect the
present discussion.

If two soils, A and B, be allowed in one case to reach equilibrium with
a saturated atmosphere and the hygroscopic coefficients thus found be a
and b , respectively, and in another case the same soils be allowed to reach

352

Journal of Agricultural Research

Vol. VII, No. 8

equilibrium with a partially saturated atmosphere, the amounts of hy¬
groscopic water present in the latter case, a1 and b\ should bear the
same relation to one another as do the hygroscopic coefficients —

= or b=a
a '

If the soils can conveniently be brought into equilibrium with the
partially saturated atmosphere, it would simply be necessary to deter¬
mine accurately the hygroscopic coefficients of a few typical soils of
which large quantities of thoroughly mixed, air-dried samples have been
prepared and then expose to a partially saturated atmosphere portions of
some of these along with the samples of which the hygroscopic coeffi¬
cients are desired. From the found amounts of hygroscopic moisture the
hygroscopic coefficients could be calculated by the above formula. This
would obviate many of the inconveniences connected with the deter¬
mination of hygroscopic coefficients, including the difficulty of obtaining
a fully saturated atmosphere and of preventing dew formation through
fluctuations of temperature.

As we had a series of foot samples of loess which had already been
subjected to careful hygroscopic-coefficient determinations (i, p. 215-216),
they were the first to be tried. The samples, after being brought from
the fields in cloth sacks, had been stored for several months in the un¬
heated but well-ventilated attic of the Nebraska Experiment Station
building and later reduced to the desired degree of fineness, thoroughly
mixed, placed in sealed jars, and again stored in the attic. About a
year previous the moisture had been determined in some 50 of the field
samples before preparing the composites and had been found to lie
between the limits of 2.5 and 4.9, practically the same as found in this
experiment. In shallow aluminum trays 5 by 7 inches with edges 0.75
inch high the samples were exposed in triplicate on the shelves of the
attic storeroom mentioned above. Each tray carried about 10 gm. of
soil. All the samples were exposed for seven days, those from Wauneta
and McCook from March 15 to 22 and 'the others from March 25 to
April 1 . With the first set the range of temperature in the air of the room
was i° to 140 C.; and that of the humidity at the Lincoln (Nebr.) station
of the Weather Bureau, 2 miles distant, was 67 to 94 per cent, while
with the second the corresponding data were 90 to 230 C. and 42 to 93
per cent. Three of the samples which had been exposed in the first set
were exposed again with the second and were found to have the same
amount of hygroscopic moisture in both cases. Accordingly we may
assume that all of the former contained the same amount of moisture
that they would have shown if exposed with the second set, with which
there were exposed in duplicate two soils, H and S, which had been
repeatedly used as control soils in the determination of hygroscopic co¬
efficients and for which, accordingly, we had a great many concordant
determinations, the average of which was 5.6 for H and 22 for S. The

Nov. ao, 1916 Determination of Hygroscopic Coefficients of Soils

353

hygroscopic moisture is reported in Table VII. The triplicate determi¬
nations were concordant, in most instances differing by less than 0.2 per
cent. The ratio of the hygroscopic coefficient to the found hygroscopic
moisture (Table VII) shows an average of 2.71 and, with the exception
of the first foot at Hastings and the fourth at Wauneta, lies between 2.38
and 2.96, a range of less than 25 per cent.

TABLE VII. — Hygroscopic moisture in a series of air-dried loess soils and its relation to
the hygroscopic coefficient in these soils

HYGROSCOPIC MOISTURE

Depth.

Wauneta.

McCook.

Holdredge.

Hastings.

Lincoln.

Weeping

Water.

Average.

Foot.

P.ct.

P. a.

P. ct.

P.ct.

P. cf.

I .

3- I

3*5

3*8

3* 0

4.2

4. I

3*6

2 .

3*4

4.0

4-7

4.2

5*5

4*9

4.4

3* * * .

3*4

3*7

4-7

4. 6

5*4

5*o

4.4

4 .

3*3

3*4

4-3

5* 0

4*9

4.2

5 .

3*2

3-2

4.0

4.1

4*9

4*8

4.0

6 .

2. 8

3*2

3-8

4.2

4.8

4*7

3*9

Average . .

3*2

3*5

4- 2

4. 1

5*o

4* 7

4. 1

RATIO OF HYGROSCOPIC COEFFICIENT TO HYGROSCOPIC

MOISTURE

I .

2. 93

2. 86

2. 66

3. 20

2. 86

2. 95

2. 91

2 .

2. 82

2. 72

2. 38

2. 76

2. 62

2. 80

2.68

3 .

2. 85

2. 89

2. 40

2. 70

2. 52

2. 78

2. 69

4 . .

3. 00

2.85

2.38

2.58

2. 60

2. 65

2. 68

5 .

2. 81

2. 93

2. 40

2. 61

2. 62

2. 66

6.. .

2. 96

2. 84

2.47

2. 60

2. 64

2. 66

2. 69

Average . .

2. 91

2. 45

2. 70

2. 62

2. 76

2. 7L

CALCULATED HYGROSCOPIC COEFFICIENT a

1 .

8.4

9* 5

10.3

8. 1

11. 4

11. 1

9.8

2 . . .

9.2

10. 8

12. 7

11. 4

14.9

i3*3

12. 0

3 .

9.2

10. 0

12. 7

12. 5

14. 6

13. 6

12. 1

4 .

8.9

9.2

11. y

11. 7

13. 6

13*3

11. 4

5 .

8.7

10. 8

11. 1

I3- 3

13.°

10. 9

6 .

7*3

8*7

10.3

11. 4

13. 0

12. 7

10. 6

Average. .

8.7

9r 5

11. 4

11. 1

13. 6

12. 7

11. 2

DEPARTURE OF CALCULATED FROM DETERMINED

HYGROSCOPIC COEFFICIENT

I .

-0. 7

— 0. 2

0. 2

“1*5

— O. 6

— 1. 0

-0.7

2. . .

- *4

— . I

i* 5

— . 2

* 5

- .4

. 1

3 .

- *5

- *7

1.4

. 1

1. 0

•4

. 2

4 .

— 1. 0

~ • 5

i*5

- *3

5 .

~ *3

- *4

1. 2

•4

•5

“ *4

•3

6 .

— 1. 0

- *4

•9

• 7

•4

— . 2

. 1

Average . .

- .6

- *4

1. 1

. 1

• 5

“ *3

. 1

a Using ratio found with control soils H and S.

354

Journal of Agricultural Research

Vol. VII, No. 8

The control soils H and S, exposed in the second set, contained 2.3 and
7,4 per cent of hygroscopic moisture, respectively, thus giving ratios of
2.44 and 2.98, respectively, with an average of 2.71. The hygroscopic
coefficients calculated from the hygroscopic moisture, using this average
ratio, and the departures of these from the directly determined values
are reported in Table VII. In general, the departure is slight and on
the average is negligible. However, it should be pointed out that the
latter circumstance is due to the average of the ratios for the two control
soils being the same as that for those of the 36 samples under considera¬
tion. If we had used only H as a control, the calculated values would
have been one-tenth higher, while if S alone had been employed, they
would have been one-tenth lower. In either of these two cases, however,
the calculated values would have been much nearer the directly de¬
termined hygroscopic coefficients than would the values computed' from
the moisture equivalents by the Briggs-Shantz formula (3, p. 839).

We similarly exposed several other series of soils of which the hygro¬
scopic coefficients had already been determined. The concordance of
the calculated with the found values was much alike in all. The first of
these series, which was strictly typical, is reported in Table VIII. It
consisted of 24 samples, part surface soils and part subsoils, but all
derived from residual material in western Nebraska. The soils were ex¬
posed on metal trays on the shelves of an inclosed basement room for two
weeks, April 29 to May 13. The maximum temperature recorded in the
room during this period was 2 2 0 C. , and the minimum 1 8° ; observations in
this room, extending over three years, had shown that there was but rarely
a daily range exceeding one degree. In the table the soils are arranged
in order of texture. The ratio for the two control soils exposed in triplicate
was 2.9. The data on the determined hygroscopic coefficients are the
means of duplicate determinations. The calculated values in nearly all
cases agree satisfactorily with those directly determined. The greatest
divergences are shown by soils 13 and 23, the calculated hygroscopic
coefficient being one-sixth too low for the former and one-seventh too
high for the latter.

Finally we exposed on paper pie plates 145 soils of which the hygro¬
scopic coefficients had not been determined. Each soil was exposed in
duplicate and there were also 15 plates of each of two control soils,
La and S. In the case of both of the control soils triplicate samples
were exposed in different parts of the room in order to determine whether
the position in the small room exerted a marked influence upon the amount
of hygroscopic moisture absorbed. During the eight days of exposure,
June 5 to 13, the temperature ranged from 180 to 230 C.

The position in the room wag found to have a slight but appreciable
influence upon the amount of moisture absorbed, the extremes being
shown by two shelves, on one of which the samples of S and La were
found to have 9.3 and 3.8 per cent, respectively, and on the other 8.7

Nov. 20, 1916 Determination of Hygroscopic Coefficients of Soils 355

and 3.3. The averages for the 15 samples of each were 8.9 for S and 3.5
for La. The ratio was accordingly 2.47 for S and 2.74 for La, the hygro¬
scopic coefficients being 22.0 and 9.6 per cent, respectively.

Table} VIII. — Comparison of the determined hygroscopic coefficients with those computed
from the hygroscopic moisture in a series of residual soils from southwestern Nebraska

Soil.

Hygroscopic

moisture.

Hygroscopic coefficient.

Ratio of
hygroscopic
coefficient
to hygro¬
scopic
moisture.

Determined.

Computed.

Departure.

Per cent.

H (control) .

I. 0

5-6

S (control) . .

10. 2

O v)

1 . , .

•3

0.9

0. I

2.7

2 .

•3

•9

. I

2. 7

3 .

. -3

•9

. I

2.7

4 .

•3

■9

•9

. O

3-o

5 .

•4

1. 1

* I. 2

. I

2. 8

6 .

• 4

1. 1

I. 2

. I

2.8

7 . .

1.4

I. 2

— . 2

3- 5

8 .

* 5

1.4

I.4

. O

2.8

. 7

2. 2

2. 0

— . 2

1. T

10 .

2. 2

2-3

. I

O' A
2. 8

11 .

i* 7

4.4

4*9

•5

2. 6

12 . : .

i*3

4-5

3-8

- -7

3-5

13 .

1.8

5-4

5*2

— . 2

3*o

14 .

i-7

5-9

4.9

— 1. 0

3* 5

15 .

2.4

6. 6

7.0

•4

2. 7

16 . .

2-5

8.0

7.2

- .8

3- 2

17 .

2.9

8-3

8.4

. 1

2.8

18 .

2.7

8.4

7.8

- .6

3*i

19 .

3- 5

9-7

10. 1

•4

2.8

3-6

9.8

10. 4

2. 7

3-6

10. 7

10. 4

- -3

3*o

3-6

10. 8

10. 4

- -4

3*o

23 .

5-7

14.4

16. 5

2. 1

2. 6

24 .

5-2

14.9

i5- 1

. 2

2.9

At this point the removal of the junior author to California interrupted
the work, but during the following year Mr. A. Skudma made single
determinations of the hygroscopic coefficients in the usual manner,
using also the same control samples. In Table IX are shown the ratios
of typical sets for various fields, the others being similar. It will be
seen that the ratio is not affected by the presence of organic matter, as
in the surface 6 inches it is similar to that in the subsoil.

This set serves well to illustrate the advantage of the indirect method
of determination. During the early part of May the samples had been
collected from 23 fields in western Nebraska for the study of field mois¬
ture conditions, which happened to be of unusual interest at that time,
a wet winter and early spring having succeeded an exceptionally dry
summer and autumn. The moisture samples had been weighed into
light cotton sacks and dried at no0 C. to constant weight, and then left
exposed to the air for a week or two before being placed on the plates
in the so-called “ constant temperature room.” By the middle of June,
as it later proved, reliable data on the unusual field moisture conditions

356

Journal of Agricultural Research

Vol. VII, No. 8

were at hand. The ratio varied in the case of the individual soil sam¬
ples from 2.0 to 2.7, and the average ratio for the different fields only
from 2.1 to 2.6. While the ratio tended to be higher in the coarse-
textured soils there were many exceptions to this generalization.

Table IX. — Ratio of hygroscopic coefficient to hygroscopic water in air-dried soils from
typical fields in western Nebraska , May, IQI2

Depth.

Prairies.

Corn stubble.

Winter wheat.

Average
of all.

II.

III.

IV.

V!.

VII.

VIII.

Foot.

o-K .

2. 7

2.4

2- S

2.3

2. 5

2. 2

2. I

2.4

K-i .

2. 7

2. 4

2. 6

2- 5

2. 6

2. 2

2. I

2.4

2. 3

2.4

2. 7

2. I

2. 4

2.4

2. O

2.4

3 .

2. 7

2-3

2. 6

2. I

2.4

2. 2

2- 3

4 .

2. 5

2. 2

2* 5

2. 2

2. I

2.4

2. 2

2. I

2.3

e .

2. 6

2. 2

2. 6

2. O

2. 7

2. I

2. 7

6 .

2. 6

. 2. I

2.3

2. 3

2. 5

2* 6

2. I

2. O

2.3

Average .

2. 6

2.3

2. 5

2-3

2. 4

2* 3

2. I

2-3

Table X and figure 1 show the concordance of the data on free water,
using in the one case the computed and in the other the directly deter¬
mined hygroscopic coefficients. Whichever set, B or C, is used, the
same general moisture relations are shown, and for the purposes of the

Pra[r/e Corn Wheat

Field f F/e/c/ 2 Field 3 Fields Fieidd Fiefd6 Field 7 Fieid8
t Using determined hyg. coeffs

► /

0 0 10 20 c

> IP 20 _ <

> 4? £

f7 i

? 10 2

2_ 5

1 JO 2

O 0 JO 20 c

> JO 20

_ j

1 m !

■

H 1

i 1

jjL

■D

11 1

■J

Us/ng computed hyg. coeffs.

Si: 1

■

■Ii 1

H 'I

|||

1 ■ 1

1 JHJ

§p

pi,

Pif

■lii

ii 1

■ill

■B

IlL

■

L egenct: Hyg. moisfure mzm
Free wafer mtm

Fig. i. — Diagram showing the amounts of free water at different levels in eight fields, illustrating the
concordance of the values obtained for the hygroscopic coefficient by calculation from the hygroscopic
moisture with those directly determined.

field moisture study they are almost equally satisfactory, but the data
in C were all at hand a month after the field work had been completed,
while those in D were not obtainable within half a year without dealing
with this particular series out of its regular order.

Nov. 20, 1916 Determination of Hygroscopic Coefficients of Soils

357

TabuS X. — Moisture conditions in typical fields of western Nebraska in May, IQ12,
showing the applicability of the data obtained by computing the hygroscopic coefficient
from the hygroscopic water in the air-dried soil

A. — TOTAL WATER IN SOIL IN FIELD

Depth.

Prairies.

Com stubble.

Winter wheat.

ill

VII

VIII

Foot.

P. ct.

o-K .

3*6

19.4

17. 1

16. 3

7*3

18. 9

20. 4

20. I

K-i .

4. 5

21. 6

20. 2

18.7

10. 0

21. 6

21. 5

22. O

5*4

14-3

I9.9

17. 8

8.7

19-3

16. 0

IS* i

3 . .

7*9

9.4

14-3

i3- 1

8.5

13.8

12.3

I°- 5

4 .

6.9

8. 0

9.6

12. 0

10. 0

9.8

10. 7

9.6

5 .

5- 9

8. 1

9. I

10. 5

15.6

9.2

9*4

6 .

6. 1

8.8

9. I

6.9

21.3

9*4

8.8

9.6

Average a .

6. 0

ii- 5

13-4

13. 0

12. 1

13.6

12* 5

B. — FREE WATER (USING DETERMINED HYGROSCOPIC COEFFICIENTS)

0 -K .

2. 0

10. 8

8.2

10. 4

5*3

10. 4

12; 2

12.8

K-I .

2.9

11. 0

10. 7

i°* 5

7* 1

ii* 5

11. 2

i3*4

3*8

4.0

9* 5

9*5

6. 1

9* 1

6.2

7*3

3 .

6. 0

1. 1

4*3

3* 7

5*9

5*3

3* 9

1. 1

4 .

5*4

•5

. 0

2. 0

7.0

i*7

2.9

5 .

4.6

•3

• 7

2.7

10. 0

i*5

1.9

4. 8

1. 2

i* 5

2.4

12.4

•9

1.4

2. 1

Average .

4. 2

3* 0

4. 0

5* 1

7*9

4.9

4*7

4. 2

C. — FREE WATER (USING COMPUTED HYGROSCOPIC COEFFICIENTS)

o-}4 .

2. 1

10. 5

7*9

9*7

5*3

9.0

10. 7

12. 6

K-i .

3* 0

10. 2

10. 7

10.3

7.2

10. 1

9.8

12.8

3*6

3* 1

10. 2

7.8

5*9

8.6

3* 1

3*6

3 . . •*••

6. 1

. 0

4.6

i*9

5*7

4. 1

2.4

- .4

5*4

- *7

— . 1

■3

6.4

1. 1

2. 7

- .8

< . .

4. 6

- .8

. 7

8. 2

1. 3

1. 4

1. 6

6 .

4.8

“ *3

• 7

2. 6

12. 1

1* 5

I. O

1.8

Average .

4. 2

1.9

4. 2

& 4. 2

7*4

4*4

3* 5

3* 1

» In estimating both the average hygroscopic coefficient and the average moisture content of the 6 feet
of soil we have employed the mean for the first foot instead of averaging the seven data in each column.

& In estimating the average the missing datum is replaced by that from the preceding part of the table.

It should be pointed out that in the case of all the soils with which
we have employed the simpler method , so satisfactorily the samples
either had been dried at no° C. or for many months had been exposed
to the dry air of the storeroom and so had attained an air-dried con¬
dition, As it would in many instances be more convenient if the moist
samples could be exposed on paper trays or plates and left until they
came into equilibrium with the atmosphere, we exposed for varying
periods — 2 to 21 days — 10 to 15 gm. samples of the control soils, H,
Ta, and S, all containing the maximum hygroscopic moisture, together

358

Journal of Agricultural Research

Vol. VII, No. 8

with samples of the same soils dried at ioo° C. and still others that
were in an air-dried condition. At the end of from 3 to 7 days those
of the second and third sets were alike in moisture content, while those
of the first were distinctly moister. A difference was frequently shown
even after an exposure of 21 days.

Using 5-, 10-, 1 5-, and 20-gm. samples of the same three control soils
we compared paper with aluminum trays. Both gave the same results
with H and La, but in some cases S, which had a coefficient of 22.0, gave
a somewhat lower result on the paper trays, 7.7 as compared with 8.0, as
though, following an increase in the relative humidity of the air, the
material of the trays, being in itself hygroscopic, had competed with
the contained soil for moisture from the atmosphere.

From the above statements it would appear that once having accu¬
rately determined the hygroscopic coefficient of suitable control soils
it would only be necessary to expose for a few weeks on trays in a closed
room thin layers of the air-dried or oven-dried soils under investigation,
together with a sufficient number of trays containing the control soils.
After sufficient exposure the mere determination of the hygroscopic
moisture in all would permit of the calculation of the hygroscopic coeffi¬
cients. Thus, the hundreds or thousands of samples which might be
collected during the summer in connection with a soil survey might be
placed in trays on the shelves as they reached the laboratory, the drying
being left until winter. There could then be obtained that single¬
valued factor expressing texture which appears to us the most desirable
of all those so far proposed (1, p. 214).

In connection with the method there remains to be determined, among
other things, the minimum time of exposure necessary, the most suitable
material for trays, the desirability of providing for the agitation of the
air in the exposure room and maintaining the humidity of the air in the
room between definite limits.

This indirect method of estimating the hygroscopic coefficient appears
to give more reliable results than those to be obtained by the use of a single
formula applied to either the mechanical composition (2, p. 41 1) or the
moisture equivalent (3, p. 842), while at the same time requiring only the
simplest equipment, as well as a minimum of skill on the part of the
operator, and being economical of time.

SUMMARY

The estimation of the hygroscopic coefficient from the maximum water
capacity, while with many soils giving values in full accord with those
directly determined, with so many soils gives such erroneous results that
it is to be regarded as too unreliable for use in connection with either
studies of available soil moisture or for soil-survey purposes.

From the studies reported it appears that the hygroscopic coefficient
may be calculated from the hygroscopic moisture found in a soil which

Nov. 20, 1916 Determination of Hygroscopic Coefficients of Soils 359

has been allowed to come into equilibrium with an only partially saturated
atmosphere and that this method will require only simple equipment, a
minimum of skill on the part of the operator, and be so economical of time
as to recommend it wherever a very large number of samples have to be
dealt with.

LITERATURE CITED

(1) Alway, F. J., and McDole, C. R.

1916. The loess soils of the Nebraska portion of the Transition Region. I. Hy¬
groscopic ity, nitrogen and organic carbon. 'In Soil Sci., v. 1, no. 3, p.
197-238, 2 fig., 3 pi. Literature cited, p. 236-238.

(2) - and Rost, C. O.

1916. The loess soils of the Nebraska portion of the Transition Region: IV.
Mechanical composition and inorganic constituents. In Soil Sci., v. 1,
no. 5, p. 405-436, 2 fig. Literature cited, p. 435-436.

(3) - and Russel, J. C.

1916. Use of the moisture equivalent for the indirect determination of the hygro¬
scopic coefficient. In Jour. Agr. Research, v. 6, no. 22, p. 833-846.
Literature cited, p. 845-846.

(4) Briggs, L. J., and Shantz, H. L.

1912. The wilting coefficient for different plants and its indirect determination.

U. S. Dept. Agr. Bur. Plant Indus. Bui. 230, 83 p., 9 fig., 2 pi.

(5) Caldwell, J. S.

1913. The relation of environmental conditions to the phenomenon of permanent

wilting in plants. In Physiol. Researches, v. 1, no. 1, p. 1-56. Litera¬
ture cited, p. 55-56.

(6) Hilgard, E. W.

i860. Report on the Geology and Agriculture of the State of Mississippi.
XXIV+391 p., 6 fig., 2 pi., 1 map. Jackson, Miss.

(7) -

1872. On soil analyses and their utility. In Amer. Jour. Sci., s. 3, v. 4, no. 24,
p. 434-445-

(8) -

1893. Methods of physical and chemical soil analysis. In Cal. Agr. Exp. Sta.
Rpt. 1891/92, p. 241-257, 1 fig.

(9) -

1893. Report on the methods of physical and chemical soil analysis. In U. S.
Dept. Agr. Bur. Chem. Bui. 38, p. 60-82.

(10) -

1903. Methods of physical and chemical soil analysis. Cal. Agr. Exp. Sta.
Circ. 6, 23 p., illus.

(11) -

1906. Soils, their Formation, Properties, Composition, and Relations to Cli¬
mate and Plant Growth in the Humid and Arid Regions. 593 p.,
illus., diagr. New York, London.

(12) Loughridge, R. H.

1894. Investigations in soil physics. In Cal. Agr. Exp. Sta. Rpt. 1892/94,
p. 70-100.

(13) Thomson, William.

1871. On the equilibrium of vapour at a curved surface of liquid. In Phil.
Mag. and Jour. Sci., s. 4, v. 42, p. 448-452, 1 fig.

(14) Trumbull, R. S.

1912. On the determination of soil moisture. In Nebr. Agr. Exp. Sta. 25th
Ann. Rpt. [1911J, p. 71-73.

CORRELATION BETWEEN THE SIZE OP CANNON
BONE IN THE OFFSPRING AND THE AGE OF THE
PARENTS 1

By Christian Wriedt,2
Fellow of the A merican-Scandinavian Foundation

Jensen (3) 3 and the writer (12, 13) have both found that in the Jut¬
land and Gudbrandsdal horses the young and relatively young animals
have given the best offspring and that old animals have very seldom
produced high-class stallions. Both of these breeds are selected for
heavy bones. On the other hand, in the English thoroughbred, which
is selected for speed, old horses produce high-class animals for that pur¬
pose. It is, then, naturally suggested that an investigation be made
as to whether young parents are producing heavier cannon bones than
old parents. Von Oettingen (7) noticed this in numerous cases and stated
that the offspring of the halfbred stallions Halm, Hamish, and Optimus
(in Gudwallen and Trakehnen, Germany), with relatively heavy bones,
tended to become slender in the cannon bones when the stallions became
old. Jensen made a similar statement in a letter to the author about
the offspring of the famous Jutland stallion Aldrup Munkedal. The
writer has also had the same experience with Gudbrandsdal stallions.
In Percheron horses bred in France, where young stallions and mares
are much used for breeding purposes, it is recognized that the breed
shows a big increase in cannon-bone size. This can not be substantiated
by measurement, but it seems to be a fact. Of statements concerning
other mammals that show a similar tendency, it is worth while to cite
Stonehenge’s (11) observation concerning dogs: “When, however, the
produce is desired to be very small, the older both animals are the
more likely this result is.”

When these views are considered, although they are not supported
by any scientific measurements, there seems to be good reason for
undertaking an investigation of the correlation between the age of the
parents and the measurements of the cannon bone of the offspring.
Fortunately the Gudbrandsdal stud book4 contains material that is suit¬
able for such an investigation. The writer selected as material for the
investigation all mares measured in the “South of Norway” district

1 Contribution from the laboratory of Genetics, Agricultural Experiment Station of the University of
Illinois.

2 The author wishes to express his thanks to Prof. J. A. Detlefsen and Prof. H. L. Rietz, of the Illinois
Station, for many suggestions and criticisms in connection with this investigation.

8 Reference is made by number to “Literature cited,” p. 370-371.

4 Stambog over Heste av gudbranddalsk race. Bd. 3, 1906; Bd. 5, 1909; Bd. 6, 1913. Kristiania.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.
gd

Vol. vn. No 8
Nov. 20, 1916
Ill. — S

643 1 2 0 — 1 6 - 2

(361)

362

Journal of Agricultural Research

Vol. VII, No. 8

(Sondenfjeldske). These mares are all measured in the same way, and
the measurements are registered in the stud book. The measurement is
taken near the middle of the cannon bone, at its narrowest point. In
the data the writer found five mares whose dams were 1 year old at the
time of service. Both because this number is too small to give average
values whose significance can be estimated, and because of the abnormal
character of these extremely young dams, they have been excluded in
the general treatment of the problem. Two other mares are excluded, one
for the reason that the writer knows that her cannon bone is abnormally
thickened, and the other because the measurement of the side breadth
is abnormally small compared with the circumference.

Along with age, there are other causes that have an influence on the
size of the cannon bone of the offspring. Heredity is the first of these
causes to be considered, because it is probably of great importance in
this particular investigation, since the modem Gudbrandsdal horses
include blood lines that represent a cross between the original horse
type of eastern Norway (the most northern branch of the occidental
horse) and different light-horse types. In the seventeenth and eight¬
eenth centuries there were imported into Norway stallions of the Fred-
ricsborgian horse or a closely related type. The Fredricsborgian horse
was a light-horse type that descended from crosses of Seeland country
horses, Spanish, and oriental horses. In the nineteenth century there
was imported into Norway an English thoroughbred stallion, Odin.
This stallion became by chance the founder of one of the most impor¬
tant stallion lines of the Gudbrandsdal breed. In the nineteenth century
the blood of the other Norwegian breed, the Fjord horse, was mixed
with the Gudbrandsdal horse by the use of several mares, and thus be¬
came an influence on the latter breed (12, 13). If cannon-bone size is
inherited, it is perfectly clear that the Gudbrandsdal horse is not a pure¬
bred in a genetic sense, since extreme heterozygosis must result from
the numerous infusions of blood from such diverse sources.

One should consider the influence of nourishment on the size of the
cannon bone, but the number of individuals in this investigation is so
large and the conditions of nourishment are such that the effects of
different nourishments would probably eliminate each other, or at least
would be insignificant. Although nutrition undoubtedly has a pro¬
nounced effect upon size and growth, the writer has assumed that
under good conditions the adult size of the cannon bone represents in
a fair way the inherent possibilities of an animal. Such has been the
usual assumption in genetic investigations of size characters.

The first task was to calculate the correlation between the age of
the sires and the size of the cannon bone of female offspring. Table I
gives the number and kinds of mares sired by stallions of different
ages. Measurements under 18 cm. and over 21 cm. are arranged in
the subclass headed 18 and 21, respectively. The averages of arrays

Nov. 20, 1916

Size of Cannon Bone

363

are calculated from the original figures before they were grouped in a
correlation table. The age of the stallions is arranged in classes with
3-year intervals in each class — that is, class 1 is composed of stallions
2 to 4 years old; class 2 is
composed of stallions 5 to 7
years old, etc. In this no¬
tation 2 to 4 means under 5,

5 to 7 means under 8, and
so on. The correlation co¬
efficient is —0.061 ±0.012.

Although the coefficient of
correlation is five times as
large as the probable error,
the negative correlation is
so small that it indicates only
a very slight tendency for
older sires to beget daughters with smaller cannon bones. The average
size of cannon bone decreases slightly as the age of the sires increases.
The difference between size of cannon bone for the mares whose sires
are 5 to 7 years old and the mares whose sires are 17 years of age or
older is 0.1 1 1 ±0.036.

A9.S 20 2G*S 2/CM.

Fig. i. —Curve showing the percentages of mares with
various-sized cannon bones, sired by stallions under n
years old.

Table) I. — Correlation between the age of the sires and the measurement of the cannon

bones of the female offspring

Age of stallion.

Number of female offspring having a camion bone of a given
size (cm.).

Average size of

18. s

19- S

20. 5

Total.

cannon bone.

, Years.

2 tO 4 .

. 58

143

415

Cm.

19. 273 ±0. 021

5 to 7 .

*37

387

182

178

h 023
691

19. 272±0. 014

8 to 10 .

252

107

135

19. 244 ±0. OI7

11 to 13 .

166

426

10. 2I9±0. 021

14 to 16 .

102

256

19. i75±o. 026

17 and older .

140

19. i6i±o, 033

Total .

226

4i$

1, 108

533

5*5

101

2> 951

r= — 0.061 ±0.01 2

Figure 1 shows the percentages of all mares of various size classes sired
by stallions 10 years and younger. To be sure, more mares of all classes
were sired by stallions under 11 years old, but it is interesting to note
that gradually increasing percentages of the larger mares were sired by
these younger stallions. To state this differently, mares whose cannon
bone measures more than 19.5 cm. are more likely to come from sires
under 1 1 years of age than are the mares whose cannon bone measures
under 19.5 cm. There is a perceptible rise in the curve beginning at the
19.5 cm. Figure 1 was plotted from Table I by taking the ratio of the

364

Journal of Agricultural Research

Vol. VII, No. 8

number of mares in any given class whose sires were under 1 1 years to
the total number of mares in that class. It would seem from the curve
that a decided negative correlation existed, but one must remember
that rapid rise in the right-hand part of the curve (20 cm. and above)
is based upon 669 cases, whereas the left-hand portion is based upon
2,281 cases.

To get sufficiently large numbers and as homogeneous material as pos¬
sible, the author has taken out for investigation daughters of five Govern¬
ment and five association stallions, each of which has more than 25 reg¬
istered daughters, of which at least 10 are sired before or after the stallion
was 10 years old (Table II). The following observations may be made
from the table in regard to the average size of cannon bone for the
daughters of these stallions.

Tabl3 II. — Average size (in centimeters) of cannon bone in the daughters of 10 selected
males , grouped according to the age of the males

Name of sire.

Age of sire.

Total

number

Number of daughters having a cannon bone of a given
size (cm.).

daugh¬

ters.

18. S

19-5

20. 5

Average.

Difference.

Digreaaa. ...

Bamsen 254. .

Bolen 260. . . .

Sverre 270 .. .

Sindre 297. . .

B jame 301

Galde 372. .. .

Kongen 3 76. .

Gimle 425.,,.

Dalegud-
brand 466..

fio years and younger. . . .

\Older than 10 years .

/ 10 years and younger. . . .

\Older than 10 years .

f 10 years and younger. . . .

V Older than 10 years .

fio years and younger. . . .

1 Older than 10 years .

fio years and younger. . . .

lOlder than 10 years .

ho years and younger. . . .

\ Older than 10 years .

ho years and younger. . . .

\ Older than 10 years .

ho years and younger. . . .

\ Older than 10 years .

ho years and younger. . . .

\ Older than 10 years .

fio years and younger. . . .
\Older than 10 years .

13
2 S
13

to-

36
31
16
29
60

37
25
13

38
47
13

4
0

5
0

3
0

5
0

5
5
3
0

7
IS

8
4

11
8
8

6
14

4
0

5
8
8

4
9

2
19
12

3
6
2

3
0

3
2

3
I
6
0

19-33 ±0.099
19.09 ±0.074
19-52 ±0.085
19- n ±0. 082
19-39 ±0.14
19.00 ±0.092
19-30 ±0.096
19.24 ±0.058
19-35 ±0.058
19-24 ±0.07
19-55 ±0.13
19-03 ±0.068
19.49 ±0.061
19-52 ±0.057
19- 17 ±0. 11
19-38 ±0.11
19-51 ±0.051
19-51 ±0.074
19- 57 ±0.086
19-38 ±0.073

J- 0. 24±o. ia
| o.4i±o. ia
j 0. 39±0. 16
| o.o6±o- 11
| 0. n±o. 10
| 0. 52±o. 15
j— 0.03 ±0.083
j— 0. 2I±0. 16

I 0. 90±0.09

I O. I9±o. II

Total. .

f 10 years and younger. . . .
(Older than 10 years .

300

262

22.
34 j

100

l9-448 ±0.O26

19- 232±o. 027

J- 0. 2i6±o. 037

Either the average for daughters sired when the stallion was 10 years
old or younger, or the average when the stallion was older than 10 years,
or the average in both these cases, is larger than the average for the
breed.

Table II shows that for offspring of 7 of the 10 stallions the average
size of the cannon bone is larger when the sire was 10 years old or
younger at the time of service. The difference is not so large that it is
significant in every case considered singly. The offspring of two, Galde
and Gimle, of the remaining three sires show practically the same
average for the two classes. The third stallion, Kongen, has a larger
average for the daughters that he sired when he w*as older than 10 years.
With such small numbers as 25 and 13 this deviation of this result from

Nov. 20, 1916

Size of Cannon Bone

365

our other results may be a chance fluctuation. In fact, the apparent
lack of harmony of this case with the others is largely due to the size
of the cannon bone of one mare — 21 cm. This mare was sired when
the stallion, Kongen, was 12

/3 /3-S /S> /3.S3Q 2073/ CM-

Fig. 2. — The percentages of mares with various-sized cannon
bones, sired by io selected stallions when these were under
11 years old.

years old. The mother was
6 years old. With all 10
stallions it is remarkable how
few there are of the daughters
with a circumference of 20.5
and 21 cm. in that class
which was sired when the
stallions were older than
10 years. Taking the two
classes as a whole, the writer
finds for all daughters sired
when the stallions were 10
years old and younger at
time of service that the aver¬
age cannon bone is 1 9.448 ±

0.026 and for the other class
the average is 19.232 ±0.027.

The difference, 0.2 16 ±0.037,
is significant, f or thedifference
is nearly 6 times the proba¬
ble error. Figure 2 shows the percentage of the daughters of the 10
stallions that correspond to different sizes of cannon bone when the
stallions were 10 years and younger at the time of service. Just as in

the case of figure 1, there is
a perceptible rise, the curve
beginning at 19.5 cm. In
other words, gradually in¬
creasing percentages of the
larger mares (those over
19.5 cm.) were sired by these
stallions when 10 years old
or younger.

The correlation between
the age of the dams and the
size of the cannon bone of
their female offspring was cal¬
culated to ascertain whether
a similar relationship held as in the case of the sires in Table I. Curiously
enough, the coefficient ( — 0.062 ±0.017) is practically the same as that
given in Table I. The total number of variates is not as large as that
given in Table I, for the age of the dam is not as frequently registered

Z3 /3S /sP ZS.S 30 30.S 3/ CM

Fig. 3. — The percentages of mares of various-sized camion
bones bred from dams under 11 years old.

366

Journal of Agricultural Research

Vol. VII, No. 8

as that of the sire. Table III gives the number of mares bred by dams
in the different age classes. The averages show that the young dams 2 to 4
years old have given offspring with the heaviest cannon bone. The dif¬
ference between these and
the class that has the lowest
average is 0.159^0.039.
The average for the class
with dams 14 to 16 years
old is 19.341 ±0.035. This
average is practically the
same as the average for the
mares in the classes 5 to 7
and 8 to 10 year old class.
This large average is due to
the 41 mares with a circum¬
ference of 20 cm. Of these, *
20 are sired by stallions 2
to 7 years old. Figure 3
shows the percentages of
mares bred from dams 10
years old and younger. In this case also the rise begins in the neigh¬
borhood of 19.5 cm., and agrees with figures 1 and 2 in this respect.

Fig. 4. — The percentages of mares of various-sized cannon bones
bred from both parents under ir years old.

Table III. — Correlation between the age of the dams and the size of the cannon bone of the

offspring

Age of dam.

Number of offspring having a cannon bone of a
given size (cm.).

Average size of
cannon bone.

18.5

19.5

20.5

Total.

Years.

2 tO 4. . .

5 to 7 . .

8 to 10. . . .

11 to 13 . . . .

£4 to 16 .

17 and older .

Total . 1

*55

1 77
103
57
34

18
14

276

427

332

260

171

117

Cm.

19. 404 ±0. 028
19* 344±°. o23
19. 340±o. 025
19. 245 ±0. 027
19. 341 ±0. 035
19. 27I±0. 037

1S3

546

347

308

1.583

r= —0.062 ±0.017

In order to get a general view of the relation of the age of both parents
to the size of the cannon bone of their offspring, the writer has computed
in Table IV the average size of the cannon bone of the female offspring
grouped according to whether both, either one, or neither of the parents
was under 11 years of age. The differences between the averages in
these groups are not striking, except when both parents are under 11
years old.

Nov. 20, 1916

Size of Cannon Bone

367

Figure 4 shows a rise beginning at 19.5 cm., which indicates that
matings of parents both under 1 1 years old have given gradually increas¬
ing percentages of the larger size classes.

Table IV. — Average size of the cannon bone in the females in relation to the age of both

parents

Age of parents.

Number of females having a camion bone of a
given size (cm.).

Average size of
cfltmnn bone.

i8.s

19-5

20.5

Total.

Both parents 10 years old and

younger .

Sire 10 years old and younger ,
the dams older than 10

years .

Sire older than 10 years,
the dam 10 years old and

younger . .

Both parents older than 10
years .

Total . .

251

144

IOI

166

i65

779

412

256

136

Cm.

19. 386^:0. 020

19. 277 ±0. 021

•

19. 275 ±0. 026

19. 291 ±0. 036

*83

546

347

308

i>583

Table V. — Averages of the deviations of the daughters from their damst grouped according

to the age of both parents

Age of parents.

Number of daughters having a given deviation in the size of
the cannon bone (cm.).

Average

-2. 5

— 2

— 1. 5

— 1

— 0. 5

0-5

i-5

2. t

3.5

Total.

deviation.

Both parents io years old
and younger .

104. 5

133

98. 5

97* 5

10. s

S68

Cm.

0. i3i±o. 024

The sire io years old and
younger. The dams
older than io years .

0. 5

5-5

8-5

24*5

60.5

12.5

2. s

0. s

218

0. 046±o. 040

The sire older than 10
years. The dam io
years old and younger. .
Both parents older than
10 years .

18. s

i*5

186

—0. oi6±o. 039

i-5

12. s

10.5

5*5

0. o67±o. 072

To show how the measurements of the cannon bones of the offspring
deviate from those of the parents the writer has tabulated (Table V) the
deviations of the daughters from the dams — that is, when the measure¬
ment of the cannon bone of the dam is 20 cm. and the cannon bone of the
daughter is 19 cm., the daughter is entered in Table V under — 1. In
this table the mares that have both parents 10 years old and younger
give an average deviation of 0.131 ±0.024 above the dam. This is
times as large as the probable error. For the mares that have the sire
10 years old and younger, the dam being older than 10 years, and the
mares with both parents older than 10 years, the difference is practically
the same as the probable error. For the mares whose sires are older
than 10 years and the dam younger, the difference is slightly negative

368

Journal of Agricultural Research

Vol. VII, No. 8

It is noteworthy that the daughters deviate from the dams by a signifi¬
cant amount only when both parents are io years old and younger, and
in this case the daughters are larger than the dams in cannon-bone cir¬
cumference. Figure 5 shows the percentages of mares in the various
classes with both parents 10 years of age and younger. The curve was
constructed from Table V. The offspring of parents under 1 1 years of
age constitute about 55 per cent of the total number of individuals en¬
tered in the table, but the curve shows that the percentage of the daugh¬
ters which average higher than their dams is greater than 55 per cent

when both parents are
under 1 1 years of age.

It takes large num¬
bers to establish the
significance of the dif¬
ferences with which we
are concerned in this
paper, and it is to be
hoped that further
data will be obtained
in order to test this
matter for other heavy
breeds. However, the
present study points to
the following conclu¬
sions :

(1) The age of the
parent has an influence
on the circumference
of the cannon bone of
the offspring.

(2) Immature par¬
ents 2 to 4 years old
give offspring with the

the cannon bone as parents as old as 5 to 7

2C/r.

Fig, 5. — The percentages of mares in various classes deviating from
their dams when both parents were under n years old.

same measurement of
years.

(3) Parents older than 10 years considered as a class give offspring
with lighter cannon bones than parents 10 years old and younger. In the
breed examined there was found a larger percentage of individuals over
average size whose parents were 10 years old or younger. On the other
hand, the average individuals and those smaller have parents which are
just as frequently under 10 years old as they are over. In other words,
the lighter classes of cannon bone come as frequently from young as from
old parents, but the heavier classes seem to come more frequently from
younger parents.

NOV. 20, 19x6

Size of Cannon Bone

369

(4) There seems to be some basis for the current opinion among breed¬
ers of Gudbrandsdal and other heavy breeds that young parents give
better offspring than older parents.

It is extremely difficult to connect the data obtained with current
genetic hypotheses and conceptions. The history of the Gudbrandsdal
heavy horses leads one to expect heterozygosis. That the heavy regis¬
tered mares and stallions should give a range of types is not surprising
in itself, for such a recent breed can hardly be expected to breed true.
However, in genetic investigations on size, segregation is supposed to be
independent of age, and an increased proportion of large-sized offspring
from the younger parents is hardly expected. It will naturally occur to
the critical that circumstances entirely independent of heredity underlie
these peculiar frequencies of the size classes. For example, one wonders
whether more large offspring from young stallions and young mares are
not registered in order to establish their reputation; or may not young
stallions be mated to high-grade mares in order to make a better showing
as sires, whereas this selection in mating would not be as rigorous after
the sire had proven his worth? Many years of familiarity with this
breed and with the methods of registration lead the writer to attach but
little value to these considerations. The best mares and the best sires
are those which are kept the longest for breeding purposes. The fact
that measurements must accompany registration has prompted careful
selection at all ages.

Available measurements or investigations that throw light on the
inheritance of size in horses are scarce. It is not known whether cannon-
bone size in horses is due to multiple factors such as are postulated in
recent investigations on size in poultry by Punnett and Bailey (10), in
rabbits by MacDowell (5), in ducks by Phillips (8, 9), in guinea pigs
by Detlefsen (1), in com by Emerson and East (2), and the like. The
investigation of Landman (4) may throw some light on the inheritance
of cannon-bone size in horses. East Prussian country mares, with rela¬
tively small cannon bones, were graded up by crossing them with Belgian
stallions. Although the total numbers recorded are smaller than is
desired as a basis for conclusions, they indicate the gradual rise in the
hybrid animals approaching the average for Belgian mares (Table VI).

Table VI. — Sizes of the cannon hone in the East Prussian country mares , Belgian marest

and three hybrid generations

Animal.

Number of horses having a cannon bone of a given size (cm.).

Average
size of
cannon
bone.

i7- S

18.5

*9- 5

20. s

21.5

22.5

23- S

24*5

East Prussian coun¬
try mares .

Cm.

18. 83

20. 17

21. 38

22. IO

23. 53

Ft generation .

F2 back cross .

F3 back cross .

Belgian mares .

370

Journal of Agricultural Research

Vol. VII, No. 8

The measurements for the Belgian mares are those given by Von
Nathusius (6). The inheritance is what is usually called blending, but it
was shown by Detlefsen (i) and others that such back crosses with appar¬
ently blending inheritance are susceptible of a Mendelian interpretation,
on the basis of multiple factors incompletely dominant. If this interpre¬
tation of Landman's data (4) is correct, then heavy-horse types may con¬
tain factors for lighter bone. The heavy but heterozygous Gudbrandsdal
horses may well give light cannon bone, for their history shows the infu¬
sion of much light-horse blood. The writer hesitates to advance the ten¬
tative hypothesis that segregation of size factors in horses may be influ¬
enced by age, but some such hypothesis is necessary to account for
the poor performance of older stallions when compared with young.
Race-horse stallions have been known to give excellent results when
very old, for they are not bred for the dominant heavy-bone characters.
Draft horses of the Gudbrandsdal and other heavy-horse breeds have not
given similar satisfactory results. The reason may lie in the possibility
that older stallions and mares are responsible for more recessives,
whereas the’ younger give more dominants. Any such hypothesis, how¬
ever suggestive, must be recast in the light of future investigations.

LITERATURE CITED

(1) DETl,EFSEN, J. A.

1914. Genetic Studies on a Cavy Species Cross. 134 p., 10 pi. (1 col.), tab.
Washington, D. C. (Carnegie Inst. Washington Pub. 205.) Bibli¬
ography, p. 129-132.

(2) Emerson, R. A., and East, E. M.

1913. The inheritance of quantitative characters in maize. Nebr. Agr. Exp.

Sta. Research Bui. 2, 120 p., 21 fig. Literature cited, p. 118-120.

(3) Jensen, J.

1887. Nogle Forhold, der have Indflydelse paa Nedarvingsevnen hos Hesten.
In Tidsskr. Landokonom., R. 5, Bd. 6, Hefte 3/4, p. 234-256. Discus¬
sion, p. 256-272.

(4) Landmann, Adolf.

1914. Die Zucht eines schweren Arbeitspferdes in der Provinz Ostpreussen.

In Kiihn Arch., Bd. 4, p. 137-293, 11 fig., 2 pi.

(5) MacDoweix, E. C.

1914. Size Inheritance in Rabbits, with a Prefatory Note and Appendix by
W. E. Castle. 55 p., illus. (Carnegie Inst. Washington Pub. 196.)
Bibliography, p. 47-49.

(6) Nathusius, Simon von.

1912. Messungen an 1460 Zuchtpferden und 590 Soldatenpferden . . . 247 p.
Berlin. (Arb. Deut. Landw. Gesell., Heft 205.)

(7) OettingEn, B. von.

1908. Die Zucht des edlen Pferdes in Theorie und Praxis. 639 p., 1 pi. Berlin.

(8) Philips, J. C.

1912. Size inheritance in ducks. In Jour. Exp. Zool., v. 12, no. 3, p. 369-380.

(9) -

1914. A further study of size inheritance in ducks with observations on the sex
ratio of hybrid birds. In Jour. Exp. Zool., v. 16, no. 1, p. 131-148.

Nov. 20, 1916

Size of Cannon Bone

3 7i

(10) Punnett, R. C., and Bailey, P. G.

1914. On inheritance of weight in poultry. In Jour. Genetics,, v. 4, no. i,
p. 23-39, 2 pl- 4-

(11) Walsh, J. H.

1879. The Dogs of Great Britain, America, and other' Countries ... by Stone¬
henge [pseud.] 384 p., illus., pl. New York.

(12) WriEdt, Christian.

1912. Fjordblodet i Gudbrandsdalshesten. In Tidsskr. Norske Landbr.,
Aarg. 19, p. 306-309.

(13) -

1914. Avlsdyrenes alder hos dolehesten sammeniignet med andre raser. In
Tidsskr. Norske Landbr., Aarg. 21, Hefte 2, p. 82-87.

LASPEYRESIA MOLESTA, AN IMPORTANT NEW INSECT
ENEMY OF THE PEACH

[PRELIMINARY PAPER]

By A. L. Quaintance, Entomologist in Charge, and W. B. Wood, Entomological
Assistant, Deciduous Fruit Insect Investigations, Bureau of Entomology

INTRODUCTION

Attention is called to the discovery in the District of Columbia and
environs of an important insect enemy of the peach believed to be new
to the United States and apparently not heretofore known to science.
Observations on this species by the writers during the summer and fall
of 1916 warrant the fear that another formidable insect enemy of the
peach and other deciduous fruits has become established in America.
The insect is a moth belonging to the tortricid genus Laspeyresia, which
contains numerous species of prime importance as pests in different parts
of the world. Thus, Laspeyresia funebrana Tr. is the common plum worm
or plum maggot of Europe and is said to be plentiful in plum pies. L.
woeberiana Schiff. in Europe bores the bark of peach, cherry, plum, and
apple trees. L. nebritana is the common pea moth, and L. schistaceana Sn.
is a sugar-cane pest of importance in Java. In America the most impor¬
tant species is L. pomonella E., as yet better known under the generic
name “ Carpocapsa.” The lesser apple worm, L. prumvora Walsh;
the pecan moth, L. caryana Fitch; and L. pyricolana Murtfeldt are other
familiar examples of the genus.

DESCRIPTION OF THE MOTH

Mr. August Busck, of the Bureau of Entomology, has prepared the fol¬
lowing description of the species with comment on its relationships and
possible origin :

Laspeyresia molesta, n. sp.

Head dark, smoky fuscous; face a shade darker, nearly black; labial palpi a shade
lighter fuscous; antennae simple, rather stout, half as long as the forewings, dark
fuscous with thin, indistinct, whitish annulations. Thorax blackish fuscous; patagia
faintly irrorated with white , each scale being slightly white-tipped . Forewings normal
in form; termen with slight sinuation below apex; dark fuscous, obscurely irrorated
by white-tipped scales; costal edge blackish, strigulated with obscure, geminate, white
dashes, four very faint pairs on basal half and three more distinct on outer half besides
two single white dashes before apex; from the black costal intervals run very obscure,
wavy, dark lines across the wing, all with a strong outwardly directed wave on the middle
of the wing; on the middle of the dorsal edge the spaces between three of these lines are
more strongly irrorated with white than is the rest of the wing, so as to constitute two
faint and poorly defined, white dorsal streaks. All these markings are only discernible
in perfect specimens and under a lens; ocellus strongly irrorated with white, edged
by two broad, perpendicular, faint bluish metallic lines and containing several small,

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

(373)

Vol. VII, No. 8
Nov. 20, 1916
K— 46

374

Journal of Agricultural Research

Vol. VII, No. 8 ■

deep black, irregular dashes, of which the fourth from tomus is the longest and placed
farther outward, so as to break the outer metallic edge of ocellus; the line of black
dashes as well as the adjoining bluish metallic lines are continued faintly above the
ocellus in a curve to the last geminate costal spots; there is an indistinct, black apical
spot and two or three small black dots below it; a thin but distinct, deep black, termi¬
nal line before the cilia ; cilia dark bronzy fuscous. Hind wings dark brown with costal
edge broadly white ; cilia whitish ; underside of wings lighter fuscous with strong irides¬
cent sheen; abdomen dark fuscous with silvery white underside; legs dark fuscous
with inner sides silvery; tarsi blackish with narrow, yellowish white annulations.

Alar expanse: io to 15 mm.

United States National Museum type 20664.

The present species is very similar to the European Laspeyresia fune-
brana, which is an important enemy of stone fruits in Europe, and it was
at first supposed that it might be this European species which had been
accidentally introduced into America, but several minor discrepancies
both in the ornamentation of the moth and in the biology of the larva
made this determination uncertain, and specimens were therefore sub¬
mitted to the European specialists, Messrs. Edward Meyrick and J. H.
Durrant, both of whom pronounced the species distinct from L. funebrana
and unknown to them.

There are several American species closely allied to Laspeyresia molesta ,
but it is unlikely that the species is a native of this country; it has more
probably been accidentally introduced from Japan, where closely allied
species also occur, though the present species has not hitherto been
reported. The theory of the Japanese origin is strengthened by a single
specimen of a species of Laspeyresia which was reared from a shipment
of pears from Japan to Seattle, Wash. The writers are unable to differ¬
entiate this specimen from those reared from peach in the East, and
believe it to be the identical species.

Among the American species Laspeyresia molesta may easily be con¬
fused with ( Epinotia ) Laspeyresia pyricolana Murtfeldt, which not only is
very similar both in adult and larval stages but which has similar bio¬
logical habits and has also been reared from peach in the vicinity of
Washington, D. C.

Laspeyresia molesta is, however, a larger and less mottled species,
without the dark-brown transverse facia on the forewing found in
L. pyricolana; the hind wings are more rounded, especially in the males,
and not so triangular as in L. pyricolana . The males of L. pyricolana can
at once be distinguished by a large patch of black scales on the upper
surface near the base of the hindwings and by a similar black patch on
the underside of the forewings; no such ornamentation is found in the
males of L. molesta .

FULL-GROWN LARVA

Thirteen to fifteen mm. long; whitish suffused with pink; tubercles minute, black.
Head light brown with darker brown markings; hind margin, ocellar area, and the
tips of the trophi black. Thoracic shield light yellow, edged with brown. Spiracles
small, circular, dark brown. Anal plate blackish fuscous. Legs and prolegs normal.

Nov. 20, 1916

Laspeyresia molesta

375

DISTRIBUTION OF THE SPECIES

So far as known to the writers, the insect in the United States is still
confined to the general region of the District of Columbia. It is very
generally present on peach trees in yards and elsewhere in the city of
Washington and adjacent towns in Virginia and Maryland within a
radius of 15 or 18 miles. Examples of injury to the peach by what is
believed to be this -insect have, however, been seen in the environs of
Baltimore. The insect is thought to have been present in the District
of Columbia for four or five years, or perhaps somewhat longer. Speci¬
mens of injured twigs were received at the Bureau of Entomology in the
fall of 1913, and the work attributed to an unknown lepidopterous larva,
although they are now believed to have been injured by Laspeyresia
molesta . A few examples of injured twigs were received or collected during
1914 and 1915, but it was not until the fall of 1915 that its injuries were
at all common. The writers were, unfortunately, not successful in obtaining
adults from the larvae until the spring of 1916, and the single specimen
then obtained did not prove sufficient for identification purposes. During
the summer of 1916, however, an abundance of adults were reared and cer¬
tain observations made concerning the biology and injuries of the insect.

CHARACTER OF INJURY AND HABITS

The larvae have been found injuring twigs of the peach (Amygdalus
persica ), plum (Prunus spp.), and cherry (Prunus spp.) and the fruit of the
peach. The scarcity of the plum and cherry during 1916 in the infested
area prevented observations as to the extent to which these fruits are
attacked. The plum and cherry, however, have not shown such general
infestation as observed for the peach, and it would appear that this
latter is the insect's preferred food plant. It should be stated, however,
that flowering cherries growing here and there in parks in Washington,
especially the extensive plantings of Japanese flowering cherries in
Potomac Park, are very generally infested. The twig injury to the
cherry and plum is essentially the same as for the peach, though it is less
conspicuous, due to less gum exudation (PI. 26, A, B).

TWIG INJURY

In one peach orchard under observation an examination in mid-Sep¬
tember showed that from 80 to 90 per cent of the twigs had been injured,
and an even higher percentage of twigs of adjacent peach nursery stock
had been attacked. Its injuries to the twigs of bearing orchards, while
important as interfering with normal growth, are of less significance than
the injuries of the caterpillars to the fruit. Twig injury in nurseries,
however, is of much more importance, as the destruction of the terminal
growing shoots results in the pushing out of shoots from lateral buds,
producing a much-branched and bushy plant unsuitable for nursery
stock (PI. 27). Twig injury to newly planted orchards and to replants

376

Journal of Agricultural Research

Vol. VII, No. 8

in bearing orchards is also quite important, and aside from the actual
injury inflicted would interfere a good deal with the proper shaping of
the tree.

Attack on the twigs begins in the spring when the shoots are from 6 to 8
inches long and continues until active growth of the trees ceases in the
fall. Many twigs injured in the latter part of the season present the
appearance shown in Plate 28, B. As the twig hardens, the larva may
leave its burrow and feed more or less on the exterior of the twig, cutting
holes and pits into the bark and causing a copious exudation of gum,
rendering the injury quite conspicuous. The more typical injury to
twigs in the fall, however, is that represented in Plate 28, A.

The larvae prefer tender, actively growing shoots, and their injury to
these (PI. 29, A, JB) is scarcely distinguishable from that of the com¬
mon peach-twig borer, or peach moth (A narsia lineatella Zell) . The cater¬
pillars pass from one shoot to another in their search for appropriate
food, and several shoots may thus be injured by a larva in the course of
its growth. A striking illustration of this preference for tender growth
was noted in an orchard near by in Virginia. Here the orchard trees had
practically ceased growth, and although a large percentage of the twigs
showed injury a careful search of these resulted in finding no larvae. In
an adjoining block of seedling nursery trees still growing vigorously
larvae in all stages were found very abundant. Injury to the shoots is
apparently continuous during the active period of growth of the trees,
even in the presence of fruit. The writers' observations are not con¬
clusive as to whether the fruit is preferred to the twigs.

INJURY TO TRUITS

The fruit may be attacked while quite green, the infestation increasing
as it approaches maturity. Larvae of all sizes have been found abun¬
dantly in peaches during the ripening stage from midsummer on. Mid¬
season and early fall varieties have been noted as being worse infested,
owing probably to the concentration of larvae on the fruit by reason
of the cessation of active growth of the twigs. Thus, in the case of some
Sal way and Smock trees and certain varieties of clingstone peaches, of
similar season, practically all the fruit on the trees was infested with from
one to three or four larvae.

In attacking the fruit the young caterpillars rather generally eat
through the skin at or near the point of attachment of the fruit stem, the
place being indicated by more or less frass adhering to the surface of
the fruit (PI. 30, B). Entrance is also made at other places, especially
where the fruit has been punctured by the curculio or abrased by limbs or
branches or other causes, as by hail. If the fruit is ripe, or nearly so,
the entrance point of the larva may soon be invaded by the brownrot
fungus, the larva continuing its development, in frequent instances, in
the fungus-invaded and decaying flesh of the peach. Owing to the com-

Nov. 20, 1916

Laspeyresia molesta

377

bined effect of the caterpillar and brownrot fungus, a good deal of fruit
may fall to the ground, though the majority of the fruit infested by the
caterpillars will remain hanging on the trees, especially if the fruit was
invaded when nearly mature. If the peach be entered at the stem end,
the larva as it grows makes its way to the pit, where it feeds on the flesh,
which soon becomes much discolored and more or less slimy (PI. 31).
Larvae entering at the side of the fruit are more likely to eat out pockets
or cavities in the flesh, as shown in Plate 30, A. The inconspicuous
entrance holes of the young larvae, especially at the stem end, often render
it difficult to detect wormy fruit by exterior examination. In numerous
cases apparently sound fruit when cut open has been found infested with
one or more larvae.

PUPATION AND HIBERNATION

The caterpillar when full grown seeks some protected place where a
cocoon of whitish silk is made preparatory to pupation. Cocoons in
summer have been found in the cavity at the stem end of the fruit (PI.
30, B), between fruits in contact, on or between mummified peaches, in
leaves gummed to the twigs, or similar situations. * It is probable that
many larvae find protected places on the twigs, in cracks, under bark
scales on the trunk and branches, and in debris on the soil. During
September larvae were frequently observed making their way into the
cracks in the bark of the trunk and larger limbs of the peach, evidently
seeking winter quarters. Winter cocoons have been found in a few
instances in little cavities eaten into the bark at the tips of injured twigs
and more or less protected by the dried exuded gum and attached leaf
fragments. The larvae in general appear to be rather indiscriminate in
their choice of pupation quarters and may be expected to choose any
place on the trees where protection is afforded. Many larvae have been
collected under bands of burlap wrapped around the trunk and larger
limbs of the trees. In the case of nursery stock the absence of rough bark
and other protection on the young trees probably forces the larvae to the
ground, though a few individuals might find protection here and there on
the plants. The insect hibernates in the full-grown larval condition in
silken cocoons, pupation occurring in the spring. Owing to its manner
of hibernating, the detection of the insect on nursery stock and young
trees would be extremely difficult, and the disinfection of trees from the
pest could be insured only by adequate fumigation with hydrocyanic-
acid gas or other suitable substance.

EMERGENCE OF MOTHS AND NUMBER OF GENERATIONS.

Moths are out egg laying in the spring by the time the shoots of the
peach are well out, as the work of the larvae is in evidence when the shoots
are 6 or 8 inches long. It would appear that there are two and probably
three broods of larvae each year, since injury begins early in the season,
and larvae in various stages of growth are to be found in late fall.

64312° — 16 - 3

PLATE 26
Laspeyresia molesta:

A. — Injury to shoot of a Domestica plum.

B. — Injury by larva to cherry.

(378)

PLATE 27
Laspeyresia molesta:

One-year budded peach nursery tree, showing injury of caterpillars.

PLATE 28
Laspeyresia molesta:

A. — Typical appearance of peach twigs in fall injured by larva

B. — Peach twig, showing large mass of dried gum and leaf fragments due to attack
by the caterpillar.

Plate 28

PLATE 29

Laspeyresia molesta:

A .^-Typical exterior appearance of larval injury to peach shoot.
B. — The same shoot cut open, showing the larva in its burrow.

PLATE 30
Laspeyresia molesta:

A. — Cavity excavated in peach by larva entering at the side.

B. — Larval injury at stem end of peach; also the summer cocoon of the insect.

PLATO 31

Laspeyresia molesta;

Peach cut open to show larval injury at the pit.

join (IF Alfflfflim RESEARCH

DEPARTMENT OF AGRICULTURE

Vol. VII Washington, D. C., November 27, 1916 No. 9

ENERGY VALUES OF RED-CLOVER HAY AND MAIZE

MEAL

By Henry Prentiss Armsby, Director , J. August Fries, Assistant Director, and
Winfred Waite Braman, Associate , Institute of Animal Nutrition of The Pennsyl¬
vania State College

COOPERATIVE INVESTIGATIONS BETWEEN THE BUREAU OF ANIMAL, INDUSTRY
OF THE UNITED STATES DEPARTMENT OF AGRICULTURE AND THE INSTITUTE
OF ANIMAL NUTRITION OF THE PENNSYLVANIA STATE COLLEGE

INTRODUCTION

The results of determinations at this institute of the net energy values
for cattle of 10 different feeding stuffs or mixtures were reported by Arms¬
by and Fries in a previous paper.1 Attention was there called to the
discordant results obtained for red-clover hay and for maize meal, and
certain of them were tentatively rejected, for reasons stated, in making
up the final averages. The present experiments were undertaken in
order to obtain additional data concerning the energy values of these
feeding stuffs. They were conducted in 1915 along the lines of the
previous experiments just referred to.

GENERAL DESCRIPTION OF EXPERIMENTS

The general plan of the experiments was to feed five different rations.
The first two consisted of two different amounts of red-clover hay alone,
one a submaintenance ration, and the other a heavy ration. The remain¬
ing rations contained different amounts of a mixture of one-third clover
hay and two- thirds maize meal, one amount being much below main¬
tenance, one approximately maintenance, and one a heavy ration. Ten
gm. of salt were added to each day's ration.

The animal used was a pure-bred Shorthorn steer, 2 years old, weighing
at the beginning of the experiments a little over 500 kgm.

As in previous experiments, each feeding period covered 21 days, the
first 11 being preliminary and the last 10 the digestion period proper.
Table I gives the dates and the rations for the several periods and also
the live weights of the animals.

1 Armsby, H. P., and Fries, J. A. Net energy values of feeding stuffs for cattle. In Jour. Agr. Research,
v. 3, no. 6, p. 435-491. 1915.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.
ge

Vol. VII, No. 9
Nov. 27, 1916
Pa.— 1

(379)

38o

Journal of Agricultural Research

Vol. VII, No. 9

Table I. — Duration of experiments in igi$ , rations fed , and live weight of the animals

Period

No.

Preliminary period.

Digestion

period.

I. .

II.
Ill

IV.

Jan. 3-13. . . .
Tan. 24-Feb.
Feb. 21-Mar.

Mar. 21-31
Apr. 1 1-2 r

Jan. 14-23 . . .
Feb. 4-13- • • •

Mar. 4-13 .

Apr. 1-10 .

Apr. 22-May 1. .

Ration.

Clover

Maize •

hay.

meal.

Gm.

Gm,

7, OOO

4. COO

*r, 0

1,50°

3,000

2,500

5, OOO

I, OOO

2, OOO

Live

weight of
steer.

Kgm.

5*3-5
497-2
4^9-5
5*3- 5
491.2

The hay used was grown on the college farm and cut when in full
bloom. It graded as “good hay.” It was fed cut in lengths of 5 to 10
cm. The maize mpal was ground from No. 2 yellow com.

EXPERIMENTAL METHODS

The experimental and analytical methods in these experiments were
the same as those previously given in detail,1 with the exception of the
determinations of carbon and hydrogen. The total carbon in the urines
was determined directly by combustion of the liquid urine by a method
worked out in this laboratory 2 and that of the feed and feces by com¬
bustion in a bomb calorimeter as described by Fries.3 The organic
hydrogen was not determined, as it has been found that the error resulting
from omitting it entirely from the computation is very small.

Table II shows the average composition of the dry matter of the
feeding stuffs used.

Table II. — Composition of the dry matter of the feeding stuffs

Feeding stuff and period No.

Ash.

Pro¬

tein,

Non¬

protein.

Crude

fiber.

Nitrogen-

free

extract.

Ether

extract.

Heat of
combus¬
tion per
kilogram.

Clover hay:

I and II . . . .

P. ct.

5- 3i
6. 20
S- 79

P . ct.
8.26
9-49
9- 94

P . ct.

I. II
1-2$
1-03

P. ct.

34.34
31* 4°
30.

P.ct.
48. 54
49- 33
50. 14

P.ct.

1- 95
2.34

2- 55

Calories.

4,367
4, 403
4, 407

Ill .

IV and V .

Average .

5- 93

9- 23

i- *3

32. 10

. 49- 34

2. 28

4, 393

Maize meal:

III .

1.44

i-37

9- 74
9. 82

-35
. 20

2. 04
2. 04

82.- 13
82. 21

4-3°

4-36

4, 5I5
4,496

IV and V . .

Average .

1. 41

9. 78

.28

2. 04

82. 17

4-33

■ 4,505

1 Armsby, H. P., and Fries, J. A. The influence of type and of age upon the utilization of feed by cattle.
U. S. Dept. Agr. Bur. Anim. Indus. Bui. 128, p. 203. 1911.

•2 Braman, W. W. A study in drying urine for chemical analysis. In Jour. Biol. Chem., v. 19, no. i, p.
X05-113. 1914.

2 Fries, J. A. The determination of carbon by means of the bomb calorimeter. In Jour. Amer, Chem.
Soc., v, 31, no. 2, p. 272-278, 1 fig. 1909.

Nov. a?, 1916 Energy Values of Red-Clover Hay and Maize Meal 381

PERCENTAGE DIGESTIBILITY OF RATIONS

From the daily records of feed and excreta and their chemical com¬
position the percentage digestibility of the several rations has been
computed in the usual manner, with the results shown in Table III.

It was assumed that in the mixed rations the hay had the percentage
digestibility shown by the average of the periods when hay was fed
alone, and the percentage digestibility of the maize meal in Periods
III, IV, and V has also been computed, with the results shown in the
last three columns of the table.

Table? III. — Digestibility of the rations

Percentage digestibility of rations.

Constituent.

Clover hay.

Clover hay and maize
meal.

Aver¬
age per*
centage
digesti¬
bility of
clover
hay.

Computed percentage
digestibility of maize
meal.

Period

II.

Period

III.

Period

IV.

Period

Periods
I and II.

Period

III.

Period

IV.

Period

Dry matter .

55-88

59-26

78. 78

73-22

79. 20

57- 23

89.78

81. 54

90. 63

Ash . .

28.57

31. 62

41.94

27. 22

27. 94

29. 78

68. 72

21. 54

23. 75

Organic matter .

57- 56

6a 97

79-94

74- 58

8a 72

58. 92

90. 15

82.37

9i. 55

Protein .

35- 29

41. 24

62.44

55-86

63.00

37- 66

74- 74

65. 46

76. 34

Crude fiber .

SO- 23

53- 19

54-22

47- 54

46.55

5i. 4i

76. 27

17. 42

8.68

Nitrogen-free extract .

66. 24

69. 76

87. 19

82.04

89.13

67. 64

93* 19

86. 61

95* 95

Ether extract .

49- 14

54- 29

78. 10

76.30

81.49

Si* 19

85. 52

83. 96

90. 68

Total nitrogen . .

42. 87

46.33

60. 82

53- 85

62. 01

44- 25

69.68

59*. 2 1

71. 77

Carbon .

54. 10

57. 50

77. 85

72. 41

78.32

55.46

89.32

81. 25

90. 23

Energy .

53*8o

57. 20

77.28

71. 87

77- 65

55- 15

88. 29

80. 40

89. 12

METABOLIZABLE ENERGY

The difference between the chemical energy of the feed and that lost
in the excreta shows how much of the former is capable of transformation
in the animal body. This has been called metabolizable energy.

Computed in the same manner as in the earlier paper, the losses of
chemical energy per kilogram of dry matter consumed and the metab¬
olizable energy remaining were as shown in Table IV, which includes
also the percentage distribution of the feed energy between the various
excreta on the one hand and the metabolizable energy on the other.

The average results for the metabolizable energy per kilogram of dry
matter and per kilogram of digestible organic matter are brought together
for convenience in Table V.

382

Journal of Agricultural Research

Vol. VII, No. 9

Table IV. — Losses of energy and their percentage distribution

Feed and period No.

Dry matter
eaten per
head and
per day.

Energy per kilogram of
dry matter.

Metabolizable energy.

Metabolizable energy

per kilogram of digest¬

ible organic matter.

Percentage losses.

Percentage metaboliza¬

ble. 1

Tosses.

§2
a a

<3*

ag
a 5

Clover hay:

I .

II .

True average .

Gm.
5> 952
3)942

Gm.

Cal .

4*367

Cal.

2,018

1,869

Cal.

153

172

Cal.

287

304

Cal.

1,909

2,022

Cal.

3*522

46. 20
42. 80

3-51

3-93

6- 57
6. 96

43-72

46.31

4*367

1*958

161

294

i*954

3*522

44- 85

3- 68

6. 72

44- 75

Clover hay and maize meal:

III. ... .

IV. . .

V . . .

True average .

Maize meal computed:

III .

IV .

V .

True average _ _

1)328

2,272

909

2,602

4>363

1,748

4*477
4* 46s
4* 46s

1,017

1,256

998

173

136

192

413

340

495

2,874

2*733
2, 780

3*7o8

3*733

3*546

22. 72
28. 13
22.35

3- 86

3- 05

4- 30

9- 23
7. 61
11. 09

64. 19

61. 21

62. 26

4,469

1*133

i59

393

2,764

3*705

25. 26

3-55

8. 79

62. 30

1,328

2,272

909

2,602

4*363

1,748

4* 515
4*496
4*496

529

881

489

179

133

208

473

363

599

3*334

3*129

3*200

3*753

3*851

3*543

11. 71
19. 60
10. 88

3*96

3- 73

4- 63

10. 48
8- 07
13-32

73.85
69. 60
71. 17

4*5<5i

697

isx

443

3*204

3*755

15-49

3-4°

9- 85

71. 18

Table V. — Average losses of chemical energy and metabolizable energy

Gross energy
per kilogram
of dry matter.

Tosses of
chemical .
energy per
kilogram of
dry matter.

Metabolizable energy.

Feed and period No.

Per kilogram
of dry matter.

Per kilogram
of digestible
organic matter.

Clover hay:

Calories.

Calories .

I .

4, 367

2,458

*, 9°9

3>522

II. .

4,367

2,345

2, 022

3j 522

True average .

4,367

2,413

*,954 i

37522

Maize meal computed:

I, l8l

1, 367

III .

4,5*5

3j 334

3, 753

IV .

4, 496

3? I29

3,85!

V .

•4* 496

1,296

3,200

3, 543

True average .

4, 5°*

1,297

3>2°4

3,755

Clover hay and maize meal:

3,7°8

III .

4, 477

1,603

2, 874

IV .

4,46s

i,732
1, 685

2,733

2, 780

3,773

3,546

V . .

4,465

True average .

4, 469

1,685

2,784

3,705

A comparison of the metabolizable energy per kilogram of digested
organic matter as given in Table IV with the previous results shows a
very close agreement, the figures being for clover hay 3.52 therms as
compared with 3.49 therms, and that for maize meal 3.76 therms as
compared with 3.80 therms.

Nov. 27f 1916 Energy V dines of Red-Clover Hay and Maize Meal 383

INFLUENCE OF QUANTITY OF FEED CONSUMED ON LOSSES OF

CHEMICAL ENERGY

A study of the percentage losses of chemical energy substantially
confirms the earlier results regarding the influence of the quantity of
feed upon these losses. If Periods I and II, in which a large and a much
smaller hay ration were fed under similar conditions, are compared, it
is seen that with the smaller ration the losses in the urine and in the
methane were decidedly greater and the loss in the feces less than with
the larger ration.

Periods III, IV, and V are similarly comparable, the rations of hay
and maize meal being fed under the same conditions but in varying
quantities. Here, as in the periods when hay alone was fed, the lightest
ration shows the greatest loss in urine and methane and the least in the
feces, while with the heaviest ration the reverse was true. The losses
computed for the maize meal alone show differences in the same direction.

As regards variation in the percentage of total energy which was me¬
tabolizable there were slight differences. In the case of both the hay
rations and the mixed rations the largest feed gave the smallest per¬
centage, but with the mixed ration the smallest feed did not give the
largest percentage. As in the experiments previously reported, the
quantity of feed failed to show any definite effect upon the percentage of
energy metabolized.

METHANE PRODUCTION

The relation of the methane to the digestible carbohydrates has been
found to be fairly constant, so that an average figure may be used to
estimate the combustible gases in the absence of the costly apparatus
necessary for their actual determination.

Table VI gives this relation as actually found in these experiments.

TablK VI. — Quantity of methane per 100 gm. of digestible carbohydrates

Feed and period No.

Carbo¬

hydrates.

Methane.

MeUiane
per icAgm. of
digestible
carbohydrates.

Average.

Clover hay:

I .

Gm.

2, 940. 6

2,054.4

2, 689. 2
4; 249- 3

1, 832. 4

Gm.

127.9

89.8

12 1. 7

Gm.

4* 35

Gm.

} 4-36

II .

4* 37

Hay and maize meal:

III .

4- 53

]

IV .

169. I
98.6

3- 98

4.63

V .

S' 38

If these figures are compared with those of the previous experiments
it is seen that while Periods I and II agree very closely, the average is
somewhat lower than the earlier average for clover hay, 4.6 gm. The
results from Periods III, IV, and V vary rather widely, although their
average, 4.63 gm., is only a little lower than the average of the previous
experiments, 4.8 gm.

384

Journal of Agricultural Research

Vol. VII, No. 9

HEAT PRODUCTION

The daily heat production measured and that computed in the usual
manner from the balance of carbon and nitrogen are compared in
Table VII.

Table) VII. — Observed and computed daily heat production

Period.

Observed.

Computed.

Error.

Computed
a- observed.

I, first day . .

Calories.

12, 238. 7

Calories.

12, 128. 7

11,634. 3

Calories.

— no. 0

Per cent.

99. IO

I, second day .

12, 008. 4

IO, 389. I

“374* 1

96.88

II, first day .

10, 068. 9
10, 148. 5

-320. 2

96. 92

II, second day .

IO, 187. I

io, 761. 9

- 38.6

99. 62

Ill, first day .

10, 848. 7

+ 86.8

IOO. 81

Ill, second day .

10, 746. 5

10, 770. 6

+ 24, 1

IOO. 22

IV, first day . . .

13, 757- 7

*3, 574- 3

-183.4

98. 67

IV, second day .

*3, 930- 3

13,474*3
. 10, 076. 8

—456. 0

96. 73

V, first day . - . .

9, 910. 2

+ 166. 6

101. 68

V, second day .

10, 201. 4

10, 195. 8

“ 5-6

99- 95

ANALYSIS OF HEAT PRODUCTION

Standing and lying have been found to exert such an influence on the
heat production of animals that in order to make comparisons the
observed results must be corrected to a uniform proportion of time
standing and lying. The total heat production for each day of the
2-day periods has therefore been corrected to 12 hours' standing and
12 hours' lying in the manner described in the previous paper 1 and the
two days averaged, and the distribution of this corrected heat produc¬
tion also has been computed by the method explained on page 468 of
the publication just referred to.1 The results of these computations are
recorded in Table VIII.

Table VIII. — Heat production per day per head corrected to 12 hours' standing

— r -

Dry matter eaten.

Total heat

Distribution of heat production.

Period No.

Hay.

Grain.

production
(average of
24 hours).

Standing.

Rising
and lying
down.

Fermen¬

tation.

Remain¬

der.

I.... .

II .

Gm.

5, 952* 3
3,94i* 5

Gm.

Cal.

12,251.3
10, 332. 9

Cal.

1.584

1,287

Cal.

777

545

Cal.

9, 818
8, 391

Ill .

i,327*9

2, 601. 7

11, 100. 8

1,411

1,879

1 12

739

8,839

IV .

2,271. 7
908. 7

4, 36 3* O

14, 129. 0

127

1, 026

11,097

V. .

747* 5

9> 854- 7

I. 183

113

598

7, 961

1 Annsby, H. P., and Fries, J. A. Net energy values of feeding stuffs for cattle. In Jour. Agr. Research,
v. 3. no. 6, p. 454. 1915.

Nov. 27l 19x6 Energy Values of Red-Clover Hay and Maize Meal 385

ENERGY EXPENDITURE PER KILOGRAM OF DRY MATTER

A comparison of Periods I and II shows how much each additional
kilogram of dry matter of the hay consumed increased the total heat
production and its several factors (Table IX).

Table IX. — Computation of energy expenditure per kilogram of clover hay

Period No.

Quantity of
dry matter
eaten.

Total heat
production

Distribution of heat production.

Standing.

Rising
and lying
down.

Fermen¬

tation.

Remain¬

der.

Gm.

Cal.

5> 952- 3

12,251.3

1,584

, 72

777

9, 818

II . .

3> 941- 5

10, 332. 9

1,287

545

8, 39i

Difference .

2, 010. 8

1, 918. 4

297

-38

232

1,427

Difference per kilo-

gram of dry mat-

ter .

954-0

148

-19

IXS

710

In making the computation for the maize meal fed in Periods III, IV,
and V, when hay and meal were fed, the increase in the heat production
due to the differences in the quantity of hay consumed, computed by the
use of the value per kilogram of dry matter just obtained, has to be sub¬
tracted from the total increment in the manner shown in the following
example (Table X) :

Table X. — Computation of energy expenditure per kilogram of maize meal

Period No.

Quantity of dry matter
eaten.

Total

heat

produc¬

tion.

Distribution of heat production.

Hay.

Grain.

Stand¬

ing.

Rising

and

lying

down.

Fer¬

menta¬

tion.

Re¬

main¬

der.

Gm.

Cal.

Period IV . . .

2,271. 7

4, 363- 0

14, 129. 0

I, 879

127

I, 026

11,097

Period V .

908. 7

1, 747- 5

97 854- 7

I, 183

xx3

598

7» 961

Difference .

1, 363- 0

2, 615. S

4, 274. 3

696

428

3, 136

Difference due to

i,363.ogm.ofhay.

i, 300. 3

202

—26

x57

968

Difference due to

2,615.5 gm. of

grain .

2, 974. 0

494

+40

271

2, 168

Difference per kilo¬

gram of grain ....

D I37- 1

189

104

829

Six comparisons according to this method are possible affording the
results given in Table XI.

386

Journal of Agricultural Research

Vol. VII, No. 9

Table XI. — Increment of heat production per kilogram of dry matter

Total

incre¬

ment.

Analysis of heat increment.

Feeding and period No.

Standing
12 hours.

Rising
and lying
down.

Methane

fermen¬

tation.

Remain¬

der.

Clover hay:

Cal.

I-II .

9S4

148

-19

Maize meal:

III-II .

1, 2 53

196

-18

190

885

IV-II .

i»235

1,382

192

“ 3

154

892

V-II .

197

“31

230

986

IV-III .

I, 208

186

+ 19

IOI

902

III-V .

990

194

+ 8

I09

679

IV-V .

T37

189

+15

104

829

Average of all .

1, 201

192

— 2

148

862

Average, omitting Periods III-II

and II-V .

143

190

117

826

The total energy expenditure per kilogram of dry matter of the clover
hay, 954 Calories, agrees well with the value 992 Calories previously
obtained for clover hay in experiment 179, and indicates that the very
low value of 453 Calories obtained in experiment 186 was, as was sus¬
pected, erroneous.1 It would appear that the mean of the two, 973
Calories, may be taken as the average value for red-clover hay, particu¬
larly as it is only slightly higher than that of 932 Calories computed from
one of Kellner s experiments.2

The average figure for the total increment per kilogram of dry matter
of maize meal eaten is 1,201 Calories, somewhat lower than the value of
1,434 Calories previously published. Another earlier experiment gave a
value of 952 calories; but this was discarded, since the increment during
lying (the “remainder” of Table XI) was only 393 Calories, as compared
with 906 Calories in the experiment reported and 863 Calories, the aver¬
age obtained in the present series.

Of the comparisons tabulated, however, Periods III— II and Periods
II-V, are based on comparatively small differences in total heat produc¬
tion, so that the deduction for the energy expenditure due to the hay
enters as a relatively large factor. If these two comparisons are omitted,
the average of the remaining four is 1,143 Calories, which agrees closely
with that of 1 ,137 Calories obtained from a comparison of the lightest and
heaviest mixed rations (Periods IV-V), so that we are inclined to attach
greater weight to this figure. The mean of this and the earlier experi¬
ment is 1,289 Calories, which may be taken as the corrected value for
the heat production caused by the consumption of 1 kgm. of dry matter
of maize rneafby cattle.

1 Armsby, H. P., and Fries, J. A. Op. cit., 1915, p. 473, 482.

2 Idem., p. 478.

Nov. 27, 1916 Energy Values of Red-Clover Hay and Maize Meal 387

NET ENERGY VALUES

In computing, finally, the net energy values, by subtracting the sum
of the losses of chemical energy and the energy expended in feed con¬
sumption from the gross energy Qf the feed, we have used first the values
for the energy expended in feed consumption obtained in these experi¬
ments and then have made a second computation, using instead the
average between these values and those obtained in previous experiments
as computed in the foregoing paragraph. Table XII gives the final
results.

Table XII. — Net energy values of feeding stuffs per kilogram of dry matter

Feed.

Gross

energy.

Losses of
chemical
energy.

Energy ex¬
pended in
feed con¬
sumption.

Net energy
values.

Clover hay:

Cal .

Cal.

These experiments .

4,367

2,413

954

I, OOO

981

Mean . .

4,367

2,413

973

Maize meal:

These experiments .

4, 501

I, 297

1, 143

2,061

Mean . .

4,501

1,297

1,289

L9I3

SUMMARY

Results are here reported of five feeding periods with cattle, two with
differing amounts of clover hay alone and three with clover hay and
maize meal in differing quantities.

(1) The metabolizable energy per kilogram of digested organic matter
was found to be 3.52 therms for the clover hay and 3.76 therms for the
meize meal as compared with 3.49 therms and 3.80 therms, respectively,
as previously reported.

(2) The average increment in heat production caused by the con¬
sumption of 1 kgm. of dry matter was as follows:

Calories

(a) For clover hay . 954

(b) For maize meal . 1, 143

(3) When these results are combined with those of previous experi¬
ments, the following corrected values for the average heat increment
per kilogram dry matter are computed:

Calories

(a) For clover hay . 973

(b) For maize meal . 1, 289

(4) The average net energy values per kilogram of dry matter ob¬
tained by the use of the foregoing averages were:

Calories

(а) For clover hay . 981

(б) For maize meal . . 1, 91

RELATIONSHIP BETWEEN THE WETTING POWER AND
EFFICIENCY OF NICOTINE -SULPHATE AND FISH-OIL-
SOAP SPRAYS

By Loren B. Smith,1

Assistant State Entomologist , and Entomologist, Virginia Truck Experiment Station

INTRODUCTION

The influence of the wetting power upon the efficiency of a contact
insecticide has never been entirely determined, although it has long been
realized that this quality is an important limiting factor in the efficacy
of certain sprays in killing insects, especially aphids, whose bodies are
more or less covered with a waxy secretion. The difficulty of determin¬
ing the wetting power of a solution has in the past precluded its con¬
sideration in the comparison of contact sprays in the laboratory. The
present work upon the relationship of wetting power to the efficiency of
nicotine and soap solutions developed from experiments performed during
1914 and 1915 in spraying garden peas (Pisum sativum) for the control of
the green-pea aphid [(Macrosiphum) Acyrtho'siphum pisi Kalt.], spinach
( Spinacia oleraceae) for the control of the spinach aphid (Myzus persicae
Sulz.), and strawberries ( Fragaria sp.) for the control of red spiders
( Tetranychus sp.). The results of the experiments demonstrated that
the optimum efficiency of sprays containing nicotine sulphate and fish-
oil soap was reached with a definite degree of concentration 2 and that
solutions which were more concentrated and also those of lower concen¬
tration were less effective in killing the insects. In order to avoid com¬
plicated conditions, the following charts of the efficiency of the sprays are
based altogether on the results of the pea-spraying experiments. The
proportional efficiency of the sprays against the spinach aphids and red
spiders was almost identical with the results obtained on the pea aphids.

When nicotine sulphate and fish-oil soap are mixed before they are
diluted, under ordinary conditions of temperature, a precipitate may be
formed. This fact has previously been noted by other authors.3 The
composition of the precipitate is probably unknown, although it is gen¬
erally supposed that the resulting solutions are less effective killing agents
than when the materials are mixed in dilute solutions. From observa-

1 Detailed, by the Virginia Crop Pest Commission for the investigation of insects affecting truck crops.

2 The terms "concentration” and "concentrated solutions” as used in this paper refer to the amounts of
materials called for by the various formulae, and not to the original insecticides previous to their dilution,
unless so stated.

3 Parker, W. B. The hop aphis in the Pacific region. U. S. Dept. Agr. Bur. Ent. Bui. in, p. 27. 1913.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.
gg

(389)

Vol. VII, No. 9
Nov. 27, 1916
Va.— 1

39°

Journal of Agricultural Research

Vol. VII, No. 9

tions in the field it was noticed that certain sprays of high concentrations
of soap apparently did not wet the insects as thoroughly or spread over
the glaucous leaves of the pea vines as well as less concentrated soap
and nicotine solutions. It was also noted that the solutions of high
concentrations were not as effective insecticides as were some of the
more dilute mixtures.

methods of determining the efficiency and wetting powers

OF THE SOLUTIONS

The efficiency, or the killing power, of the solutions was determined
from actual field-spraying experiments. Peas were sprayed five times,
spinach once, and strawberries once. The experiments were performed
on 27 one-twentieth-acre plots. A power sprayer which maintained a
pressure of 75 to 125 pounds was used to apply the materials. Three
nozzles per row were employed, two lateral and one vertical. Before the
spray was applied, the number of live aphids on the vines were counted
for a certain distance in the center of the plot. Two hours after spraying,
another determination of the number of live insects was made on the
same vines as before, and from these figures the percentage of the insects
killed was computed. When spraying strawberries to control the red
spider, the number of live and dead mites on several leaves from each
plant were counted, and the efficiency determined in this way. De¬
tailed results of this work are shown in Table I.

The comparative wetting powers of the spray solutions were deter¬
mined by the method recommended by Cooper and Nuttall.1 In this
method a standard paraffin oil having a density of 0.8690 is run from a
pipette through the solution to be tested, and the number of drops
formed from a definite volume of oil are counted. The wetting power
is directly prop6rtional to the drop number. It is advisable to use
distilled water as a standard liquid, as there is variation in different
samples of oil; hence, the wetting power of the solutions is expressed
as the ratio of the drop number of the solution to that of distilled water
multiplied by 100.

The determination of the percentage of nicotine in the solutions was
made by a test which was approved and adopted by the Bureau of Animal
Industry on March 1, 1915. The method is well known and needs no
description here.

1 Cooper, W. F. , and Nuttall, W. H. The theory of wetting, and the determination of the wetting power
of dipping and spraying fluids containing a soap basis. In Jour. Agr. Sci., v. 7, pt. 2, p. 235. 1915.

Nov. 11, 1916 Nicotine-Sulphate and Fish-Oil-Soap Sprays

39i

Table I. — Combined data of spraying experiments with nicotine sulphate and fish-oil
soap in IQ14 and IQ15 at Norfolk , Va.

Group No.

Plot.

(Macrosiphum) Acyrthosi -
phum pisi (5 sprayings).

Num¬
ber of
Myzus
persicae
killed.

Num¬
ber of

T etra-
nychus
sp.

killed.

Wetting
power of
solutions.

Nicotine
in the
sprays.

Number

alive

before

spraying.

Number

alive

after

spraying.

Num¬

ber

killed.

Per ct.

I .

1,343

336

75- 0

72.5

79. 6

103

O. 0650

I .

1,614

308

80. 9

82. O

83-3

193

. 0650

i, 567

144

90. 8

88.9

90. 8

615

.0650

1,830

242

86.8

85.2

86.9

628

. 0650

!> 353

199

85-3

8s-9

86.9

743

.0650

i,359

270

80. 1

79.8

82.8

75°

. 0650

1,289

320

75*2

75*i

80. 0

788

. 0215

' 7

1, 191

355

70. 2

66.5

76. 8

754

. 0260

1, 282

93-5

93-8

95- 0

743

.0425

1, 246

95*9

95* 1

95*o

732

•0575

1,830

242

86.8

85.2

86.9

628

. 0650

2 .

1, 462

934

36. 1

34. 9

60, 5

181

• 0215

3 .

741

51.6

52- 1

66. 4

. 175

. 0260

3 .

1,391

522

62. 5

59*9

71* 4

163

• °325

3 .

i>435

445

69. 0

69. 7

76.9

154

.0425

3 .

1,406

404

71*3

7o*3

76.9

125

•0575

3 .

i,343

336

75* 0

72.5

79.6

103

. 0650

4 .

i,499

955

36- 3

34* 7

60. 6

307

4 .

1, 579

846

46.4

45*3

68.8

363

4 .

1, 681

891

47.0

44*9

69. 0

45°

4 .

1, 406

600

57*3

55-2

64. 4

645

4 .

564

575

63. 2

60. 1

74- 5

689

4- .

1, 344

497

63. 0

60. 3

7i* 5

1, 067

4 . .

1,487

39i

73* 7

75* 0

78. 0

1, 080

4 .

1,334

33i

75* 2

74.2

82. 0

1, 106

4 .

1, 466

365

75* 1

74. 8

82. 0

1, 112

FORMULAE TESTED

In the preparation of contact sprays to be applied to tender plants
the concentration of the solutions is limited between a minimum which
is a strength sufficient to kill all the insects which it strikes and a maxi¬
mum which is the greatest concentration that can be applied without
injury to the foliage. In the experiments on peas or young spinach the
greatest concentration which could be used without injuring the plants
was 8 pounds of soap 1 to 50 gallons of water, or a i-to-534 concentration
of nicotine sulphate. The minimum concentration was not so sharply
defined.

It was the writer’s endeavor to try, so far as possible, such practical
combinations of the two materials as were at all likely to give satisfac¬
tory results. In order to make the formulae more comprehensive, they

1 A standard caustic-potash fish-oil soap was used throughout these experiments.

392 Journal of Agricultural Research voi. vii, no. 9

are placed below in their logical groups, 50 gallons of water being used
in each case.

Group i. — A constant amount of nicotine sulphate, with which fish-oil
soap was used in varying quantities.

Plot.

Nicotine sulphate.

Fish-oil

soap.

17 .

Ounces.

Ratio.

1:630

Pounds.

q . ; . . . .

4 . . .

10 . . .

11 .

12 .

Group 2. — A constant amount of fish-oil soap, with which nicotine
sulphate was used in varying quantities.

Plot.

Fish-oil

soap.

Nicotine sulphate.

6 . . .

Pounds .

Ounces .

Ratio.

1:1,938

i:i>575
1:969
1 : 720
1:630

7 . ■ .

7 . . . . .

8 . ; .

IO .

Group 3. — A minimum constant amount of soap, to which nicotine
sulphate was added in varying quantities.

Plot.

Fish-oil

soap.

Nicotine sulphate.

18 . . .

Pounds.

Ounces.

Hi'

Ratio.

1:1,938

i:i>575

1:1,260

1:969

1:720

1:630

17 . . . .

16 .

IC .

14 . . .

12 . .

Group 4. — Fish-oil soap used alone.

Plot.

Fish-oil

soap.

Plot.

Fish-oil

soap.

19 .

Pounds.

24 .

Pounds.

20 . .

2 C .

. . .

26 .

27 .

23 . .

Nov. 27, 1916 Nicotine-Sulphate and Fish-Oil-Soap Sprays

393

EXPERIMENTAL WORK
GROUP I

The effectiveness of the sprays containing nicotine sulphate in the
ratio of 1 to 630 with various amounts of soap to 50 gallons of solution
is represented by the efficiency curve for this group in figure 1. The
curve begins with 1 pound of soap to 50 gallons of spray, at 75 per cent
efficiency. It gradually rises with the increased amounts of soap in
the formulae to 90.8 per cent, which is the efficacy of 4 pounds of soap
plus 50 gallons of a i-to-630 Solution of nicotine sulphate. The solutions
containing greater concentrations of soap than the above lose effective¬
ness, and the curve drops to 85.3 per cent at 6 pounds and to 80.1 per
cent for the formula which contained 7 pounds of soap.

The degree of wetting of the solutions in group 1 is shown by a curve
in figure 1. This is based on the arbitrary comparative values in the
column at the right. The wetting power of the formula containing
1 pound of soap to 50 gallons of i-to-630 nicotine solution is 103, the
curve then rises to 193 for the formula containing 3 pounds of soap.
A sudden increase in the wetting power takes place at this point, the
curve going to 615 for 4 pounds, 628 for 5 pounds, 743 for 6 pounds, and
750 for 7 pounds of soap to the 50 gallons of i-to-630 nicotine-sulphate
solution.

GROUP 4

Group 4 is one of eight formulae for fish-oil-soap solutions at ratios
between 2 and 8 pounds to 50 gallons of water. The efficiency curve
for group 4 is given in figure 1. The efficacy of a solution of 2 pounds of
fish-oil soap to 50 gallons of water is 36.3 per cent. From this the curve
rises to 73.7 per cent for a solution which contained 6 pounds of soap to
50 gallons. Greater concentration of the solutions gave but slight
increase in the effectiveness, as is shown by the curve, which remains
only a fraction above 75 per cent for the solutions containing 7 and 8
pounds of soap.

The wetting-power curve of the solutions in group 4 is shown in figure 1 .
The curve begins at 307 for the solution containing 2 pounds of soap and
rises gradually to 363 for 3 pounds of soap. From this point the wetting
power being greatly increased by further additions of soap, the curve
rises to 645 for 4 pounds and 1,067 f°r 5 pounds of soap. Further con¬
centration of the solutions increased the wetting power very little, a
solution of 8 pounds of soap having the wetting value of 1,112.

DISCUSSION OR GROUPS I AND 4

The results of these experiments indicate (i) that the addition of
nicotine sulphate to fish-oil-soap solutions decidedly increases their
effectiveness in destroying aphids; (2) that the efficiency of nicotine

394

Journal of Agricultural Research

Vol. VII, No. 9

sprays can be increased to a considerable degree by the addition of soap,
but when more than 4 pounds of soap are used to 50 gallons of i-to-630
nicotine-sulphate solution the effectiveness of these solutions decreases;

Fig. i. — Efficiency and wetting power graphs for sprays in group i, containing io ounces of nicotine sul¬
phate and varying quantities of soap, and group 4, containing various amounts of soap with no nicotine.
Wetting values are given in the column at the right.

(3) that if soap is used without the nicotine, 6 pounds to 50 gallons of
water is all that can be used economically, since more concentrated
solutions do not have an appreciably greater efficiency; (4) that when

Nov. 27, 1916 Nicotine-Sulphate and Fish-Oil-Soap Sprays

395

nicotine sulphate at the rate of 1 to 630 is present in a soap solution the
wetting power of this solution is less than that of one which contains
an equal amount of soap without the nicotine; and (5) that when 4
pounds of soap or less are added to a 1 -to-630 nicotine-sulphate solution
the wetting power is but slightly affected by the presence of the nicotine.
When more than 4 pounds of soap are added, the wetting powers of the
subsequent solutions are greatly reduced from those of similar soap
solutions containing no nicotine. The efficiency of the combination
sprays likewise decreases from the point where the wetting power is
influenced the least by the concentration.

From the foregoing statements it is evident that the addition of nicotine
sulphate to soap solutions reduces the wretting power. A comparison of
the wetting-power determinations of the soap solutions containing nicotine
with those of the soap solutions without nicotine shows that the loss of wet¬
ting power is not by any means in direct ratio to the quantity of soap in
the solution; therefore the loss is probably not entirely due to a physical
effect of the nicotine upon the solution, for in that case the loss of wetting
power would be proportional to the amount of soap contained in the
solution before the nicotine was added. The wetting-power curves of
the two groups of solutions indicate that a chemical change takes place
when a certain degree of concentration is reached, which affects the
physical properties of the solutions containing nicotine, and also that the
effect is greater after a definite degree of concentration of soap is reached.
Since all the sprays in group 1 have an efficiency of 75 per cent or more,
depending on the amount of soap contained in the formula, and the
highest efficiency of any of the sprays in group 4 was only slightly above
75 per cent, it is evident that the chemical reaction affects the soap and
not the active nicotine sulphate. In support of this it may be stated
that the solutions in groups 1,2, and 3 were tested and it was found that
the percentage of nicotine in them agreed with the amount of nicotine
sulphate contained in the formulae, and that the amount of soap present
apparently had no influence upon the nicotine content of the solutions,
for if any nicotine was set free by a reaction with the soap, less than
0.005 Per cent was l°st- This was not sufficient to cause any appreciable
variation in their efficiency.

groups 2 and 3

The formulae in group 2 contain 5 pounds of soap, which is constant
for the group, plus varying quantities of nicotine sulphate to 50 gallons
of water. Group 3 contains formulae for 1 pound of soap, with similar
amounts of nicotine sulphate as above, to 50 gallons of water. The
efficiency and wetting-power curves for these groups appear in figure 2.
The efficiency curve for group 2 rises gradually with the increased amounts
of nicotine in the solution from 70 per cent at 3X ounces of nicotine to
93.5 per cent at 6% ounces of nicotine in the solution. The efficiency of
64313°— 16— 2

396

Journal of Agricultural Research

Vol. VII, No. 9

8% ounces of nicotine was 95.9 per cent, the highest of any of the for¬
mulae used in these experiments. This point of concentration is appar¬
ently the optimum, as the efficiency dropped to 86.8 per cent when 10
ounces of nicotine sulphate were used. The curve for the sprays in
group 3 shows that the efficiency increases from 36.1 per cent for the
formula containing ounces to 69 per cent for the formula containing
6X ounces of nicotine sulphate. From this point the efficiency rises
gradually to 75 per cent for the formula containing 10 ounces of nicotine
sulphate.

The wetting powers of the solutions in group 2 fall gradually from 788
for the formula containing 3^ ounces to 732 for the formula containing
8$i ounces of nicotine sulphate. Further concentration of the nicotine
in the solution causes a considerable loss of wetting power, and the
formula containing 10 ounces of nicotine has a wetting power of only 628.
The wetting-power curve of the solutions in group 3 falls gradually from
1 81 for the formula containing 3^ ounces to 103 for the formula con¬
taining 10 ounces of nicotine sulphate to 50 gallons of solution.

DISCUSSION OF GROUPS 2 AND 3

The main facts to be noted from the results given in figure 2 are: (i)The
addition of 5 pounds of soap to the 50 gallons of nicotine solution in¬
creased the efficiency from 20 to 30 per cent more than that of similar
nicotine solutions which contained only 1 pound of soap to 50 gallons
of water; (2) the most efficient results were obtained with formulae con¬
taining 5 pounds of soap, 6 % to 8 % ounces of nicotine sulphate, and 50
gallons of water; (3) when more than 8$4 ounces of nicotine were added
to the 5-to-5o soap solution there was a loss of efficiency and likewise a
corresponding loss of wetting power; (4) while the quantities of soap in
the solutions remained constant through both groups of formulae, there
was a gradual loss of wetting power, as the quantity of nicotine was in¬
creased in the solutions. The results derived from the formulae in
group 2 support the deductions already drawn from the results obtained
with the formulae in group 1 — namely, that when certain concentrations
of the soap and nicotine are reached, not only is there a decided loss of
wetting power but there is also a corresponding loss in the insecticidal
efficacy of the sprays. If nicotine sulphate is used at the rate of 1 to
630, the optimum efficiency is obtained with 4 pounds of soap to 50
gallons of water. By reducing the concentration of nicotine sulphate
to 1 to 720, 5 pounds of soap to 50 gallons of solution gives the greater
efficiency. The effect which a loss of wetting power may have upon
the efficiency of a solution is indeterminable, since soap, as well as nico¬
tine, has insecticidal properties. Thus, it is probable that a reaction
which would cause a loss of wetting power would also reduce the insec¬
ticidal properties of the soap, ending in a loss of efficiency. There is

Nov. 27, 1916 Nicotine-Sulphate and Fish-Oil-Soap Sprays

397

another condition which must be considered before judgment is made
concerning the importance of wetting power. In case the insects are
thoroughly drenched with the solution, a much higher percentage of

Fig. 2. —Efficiency and wetting-power gsaphs for group a, containing 5 pounds of soap, and group 3, con¬
taining 1 pound of soap plus varying amounts of nicotine sulphate. Wetting values are in the column
on the right.

mortality occurs than when the insects are struck by a few minute drops,
as is usually the case in field spraying where the materials are applied
at high pressure. Under the latter conditions wetting power is highly

398

Journal of Agricultural Research

Vol. VII, No. 9

important, for from our present knowledge the insects are killed by the
absorption of the materials through the trachea, and in this case it would
be necessary for the solution to spread over the body of the insect in
order to gain entrance to the spiracles. A slight loss of wetting power of
solutions which have an efficiency of 75 per cent or less would probably
not cause an appreciable change in their effectiveness. This is shown
in the case of group 3, in which the wetting power becomes less as the
concentration of the solution is increased, but the efficiency rises in a
normal curve. If a solution with an efficiency of 85 per cent or more
upon further concentration loses wetting power, there is an appreciably
greater corresponding loss of efficiency than would occur by increasing
the concentrations of a solution whose efficiency is less than 75 per cent.
The reason for this is obvious. A certain percentage of the insects are
completely covered by the spray, so that the wetting-power influence
on the efficiency is negligible; but when the insects which are struck by
only a small quantity of the spray are considered, it is evident that the
wetting power, as well as the strength of the solution, is an important
factor governing its efficiency.

It is evident from the discussion in the preceding paragraphs that the
loss of wetting power in the more concentrated mixtures is not due to a
physical effect, but to chemical reactions caused by the nicotine sulphate
in a soap solution. Since, if the nicotine, without causing chemical reac¬
tions, did exert an influence on the physical properties of the solutions, the
loss of wetting power would be directly proportional to the amount of soap
in the solution and also to the wetting powers of the soap solutions which
contained no nicotine. Likewise, the loss of efficiency in the concentrated
solutions is not due to a reaction which would cause a portion of the nico¬
tine to be liberated as free nicotine, for the nicotine contents of the solu¬
tions were determined and the percentages were found to remain constant
irrespective of the amounts of soap added to the solutions. From these
facts we are led to assume that either a direct loss of wetting power or a
reduction of the insecticidal value of the soap or both are contributing
factors in the loss of efficiency of the more concentrated nicotine-sulphate
and fish-oil-soap solutions.

SUMMARY

(1) When using combination fish-oil-soap and nicotine-sulphate sprays
for the control of insects affecting truck crops, it was found that certain
concentrated mixtures did not give as satisfactory results as did some of
lower concentration. In connection with these results it was noticed
during the spraying operations that some of the more concentrated
solutions did not possess as high wetting or spreading powers as other
mixtures which contained less soap.

(2) The spraying operations were performed on peas, spinach, and
strawberries against the pea aphid, spinach aphid, and red spider,

Nov. 21, 1916 ' Nicotine-Sulphate and Fish-Oil-Soap Sprays

399

respectively. The proportional efficiency of the sprays proved to be
similar for each species. The efficiency of the sprays was determined
by counting the number of live insects on a portion of the plot previous
to the application of the sprays, and again determining the number two
hours after treatment. The experiments were performed on 27 one-
twentieth-acre plots. Peas were sprayed five times, spinach and straw¬
berries once. The wetting powers as well as the nicotine content of the
solutions were determined.

(3) When more than 4 pounds of soap were used with 10 ounces of
nicotine sulphate to 50 gallons of water, there was a loss of both wetting
power and efficiency.

(4) When more than 8^ ounces of nicotine sulphate were combined
with 5 pounds of fish-oil soap to 50 gallons, a loss occurred in both the
wetting power and the efficiency.

(5) When nicotine sulphate was used in quantities up to 10 ounces, to
a i-to-50 fish-oil-soap solution, none of the resultant sprays had an effi¬
ciency of more than 75 per cent. Also, when fish-oil soap was used alone
in quantities not exceeding 8 pounds to 50 gallons, the highest efficiency
of any of the formulas was only a fraction over 75 per cent.

(6) It was found that the nicotine content of the solutions remained
the same irrespective of the amount of soap used.

(7) The loss of efficiency due to increasing the concentrations of the
solutions is probably caused by a loss of both wetting power and insecti¬
cidal value of the soap.

(8) The loss of wetting power which occurs when the concentration
of the solutions is increased has a stronger tendency to reduce the effi¬
ciency of the subsequent solutions, if the original solution has an effi¬
ciency of 85 per cent or more, than it does if the original efficiency is
below 75 per cent.

(9) The actual importance of wetting power is difficult to determine
in this case, as the fish-oil soap has insecticidal properties in itself.
Where the wetting power is affected, it is probable that the soap is also
broken down sufficiently to lose some of its value as an insecticide; hence,
both factors must be considered as the cause of the loss of efficiency of
some of the more concentrated mixtures.

LIFE HISTORY AND POISONOUS PROPERTIES OF
CLAVICEPS PASPALI

By H. B. Brown, 1

Plant Breeder , Mississippi Agricultural Experiment Station
INTRODUCTION

During the last decade Paspalum dilatatum Poir. has attained con¬
siderable prominence as a forage grass in various parts of the South.
One serious objection to its use, however, is that forage poisoning fre¬
quently results among cattle feeding on it. Brown and Ranck 2 showed
that the poisonous property is due to Claviceps paspali Stevens and
Hall, a fungus that infects the grass very generally. This species was
described by Stevens and Hall3 in 1910. Norton4 observed this fun¬
gus on P. dilatatum in Maryland in 1902. He suspected that it was
poisonous, but carried on no feeding experiments to determine this.

Since September, 1914, the writer has been making a study of the
life history of Claviceps paspali and its growth and distribution in the
region about the Mississippi Agricultural College. In this region the
fungus infects Paspalum dilatatum very generally, a few weeks after the
grass heads out at least 90 per cent of the old heads showing infection.

LIFE HISTORY OF THE FUNGUS

Sclerotia produced during the summer and autumn (PI. 32, F) drop
to the ground when the old grass head sheds its spikelets, and lie
on the ground until spring. They may be found at any time during
the winter and spring by searching in the litter on the ground where
infected Paspalum dilatatum grew the season before. Sclerotia gathered
during the winter and placed in moist chambers kept at room temper¬
ature will germinate in 20 to 30 days, but it is the writer's experience
that sclerotia forced in this way do not produce as many nor as large
and vigorous stromata as those that germinate in the normal way.
After a few days of rainy weather about the middle of May. sclerotia
germinating on the ground may be expected. They were first found

1 1 wish to express my obligation to Dr. Charles F. Briscoe and to Prof. J. M. Beal, of the Mississippi
Experiment Station, for the use of their laboratories in carrying on this work, and for other courtesies
extended to me,

2 Brown, H. B,, and Ranck, E. M. Forage poisoning due to Claviceps paspali on Paspalum. Miss.
Agr. Exp. Sta. Tech. Bui. no. 6, 35 p., 18 fig. 1915.

3 Stevens, F. E-, and Hall, J. G. Three interesting species of Claviceps. In Bot. Gaz., v. 50, no. 6,
p. 460-463, 8 fig. 1910.

4 Norton, J. B. S. Plant diseases in Maryland in 1902. In Rpt. Md. State Hort. Soc., v. 5, 1902,
P. 90-99. Ii902.]

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.
gf

Vol. VII, No. 9
Nov. 27, 1916
Miss. — 1

402

Journal of Agricultural Research

Vol. VII, No. 9

on May io in 1915 and on May 21 in 1916. In each case this was
just after the host plant had begun to flower.

The sclerotia of Claviceps paspali when mature are globular in shape,
2 to 4 mm. in diameter, irregularly roughened on the surface, and yel¬
lowish gray in color; the interior is homogeneous in structure and con¬
tains a considerable quantity of oil. ’ Germinating sclerotia produce
from one to several stromata, usually two or three, with slender whitish
stalks 3 to 15 mm. in length, and heads about 1 mm. in diameter (PI.
32, E). The heads are roughened over the surface owing to project¬
ing perithecial necks (PI. 32, A, E), and are at first whitish in color, later
becoming rather bright yellow, and finally brownish.

A vertical section of a stromatic head (PI. 32, A) shows numerous
flask-shaped perithecia embedded in the outer part of the head. The
neck of each perithecium projects a short distance beyond the surface,
thus forming small pimple-like projections. Each
perithecium contains numerous slender, cylindrical
asci, 150 to 170M in length (fig. 1, a) ; at the outer end
of each ascus there is a thimble-like knob fitting over
the end. The wall of the ascus is so thin that it can
not be distinguished clearly. The ascospores are fili¬
form and hyalin, being a little less than iju in diameter
and 70 to ioofi in length (fig. 1, c). There are
probably eight spores in an ascus, although not more
than seven were counted with certainty. It was not
possible to count the spores when inside an ascus, as
they are hyalin and packed together closely, and it
was a rather difficult matter to count them as the
ascus disintegrated.

Mature stromatic heads from sclerotia just gathered
from the field when allowed to dry slightly and then
moistened exuded asci very freely. The asci go to
pieces quickly after escaping from the perithecia and liberate the spores.
A change of moisture conditions in the field will cause spores to be
deposited on the surface of the stromatic head, where they are in
position to be picked up by insects that chance to rub against the
head. The stromata are somewhat tough and leathery and last for sev¬
eral days. If the ground becomes dry during their regular period they
dry out, but revive with the coming of moisture and again shed spores.
No stromata were found in the field after July 2.

Flowers of Paspalum dilaiatum inoculated with ascospores by rubbing
stromatic heads against stigmas and spikelets of the grass heads showed
abundant evidence of infection in seven days. Flowers on control plants
showed no infection. (Both inoculated plants and controls were kept
under bell jars.) In the field, infected heads are not found for several
days after the sclerotia germinate. They were first noticed on June 8 in

Fig . 1 . — Claviceps paspali:
a, Mature ascus; 6, ascus
breaking up to liberate
spores; c, ascospore.

Nov. 27, 1916

Claviceps paspali

403

1915 and on June 12 in 1916, being, respectively, 29 and 22 days after
germinating sclerotia were first found. In 1915, infected or diseased
heads were not plentiful in the fields until about July 12. Preceding
this date there were several days of rainy weather. In 1916, similar
observations were made. Diseased heads became very common during
July, following several weeks of rain. On August 1, 1916, they were
more plentiful than since the autumn of 1914.

In the fields the first infection of the season is doubtless carried by
insects. Running over the ground, they are likely to rub against the
stromatic heads, which are covered with ascospores, and, climbing up the
grass culms to take flight, may carry ascospores to the grass flowers and
produce infection. That infection does not take place often is evidenced
by the fact that the disease is slow in
getting a start after the sclerotia ger¬
minate.

The infecting fungus attacks the pistil
of the grass flower, and in a few days
the ovary is almost entirely destroyed,
a mass of fungus tissue filling the space
it occupied. Plate 32, D, shows a sec¬
tion of the mass of fungus tissue between
the glumes of a grass spikelet a week
after infection. The two spots in the
central part of the figure represent
remnants of the grass flower. The rest
of the central part of the section is
homogeneous tissue, while around the
edge are numerous tufts of hyphge stand¬
ing at right angles to the central mass. Figure C of Plate 32 shows
the tufts enlarged. Each tuft contains a number of hyphge. The
digital ends of these hyphge, or certain of them, enlarge and form
conidia or sphacelia spores. Figure 2 shows the tip of a tuft of
hyphge. The spores are hyalin but show granules when stained,
oblong, about 5^ wide and 15 /x long. They are produced in great abun¬
dance and are carried from the hyphge on which they were produced by
a droplet of honeydew, a sticky, sweetish exudation of the fungus tissue.
Insects of many kinds feed on this honeydew and carry infection by
means of the spores clinging to their bodies. Hand inoculations, which
were made by smearing honeydew containing sphacelia spores on flower
stigmas, produced infections that were exuding honeydew and sphacelia
spores freely within the space of a wreek. This result was obtained
in the case of plants kept under bell jars, and also with plants inoculated
in the field. Sphacelia spores frequently germinate in the droplet of
honeydew and give it a whitish appearance.

Fig. 2. — Claviceps paspali : Tip of tuft of
hyphae, showing the production of spha¬
celia spores.

404 Journal of Agricultural Research voi. vn, no. 9

The sphacelia stage in which honeydew is exuded lasts but a few days.
If the weather is dry, the whole grass head is likely to become dry and
dead, and no further development occurs. Or, again, honeydew may
become infected with a species of Fusarium or Cladosporium and growth
be stopped. If weather conditions are favorable, the solid mass of
fungus tissue, constituting the bulk of the sphacelia tissue, continues to
enlarge and soon forces the glumes of the spikelets apart. These masses
are young sclerotia. In some cases within a week after the sphacelia
stage was at its height the young sclerotia were projecting from between
the glumes of the spikelet and were i to 2 mm. in diameter. Following
this, some of the sclerotia continue to enlarge, attaining a maximum
diameter of about 4 mm. and characters as outlined above. During
September and October the largest sclerotia are to be found ; sclerotia are
also most plentiful then.

OTHER FUNGI INFECTING PASPAEUM DILATATUM

As was mentioned above, Fusarium heterosporum Nees. and Clado¬
sporium sp. are two other fungi found infecting heads of Paspalum
dilataium. While these fungi have not been studied carefully, they seem
to be largely in the nature of molds growing on Claviceps paspali and parts
of the diseased grass heads. The inoculation of healthy grass heads with
spores from pure cultures of each of these fungi produced no infection.
They are probably of no great consequence.

POISONOUS PROPERTIES OF CLAVICEPS PASPALI

As was shown by Brown and Ranck,1 Claviceps paspali is poisonous
to certain animals, especially to cattle and guinea pigs. It produces a
peculiar nervousness, resembling considerably that shown in certain
stages of rabies, and if eaten in quantity may cause death. A gram of
extract made from this fungus, although probably containing other sub¬
stances in addition to the poisonous element, will, if fed to a guinea pig,
cause death within a few hours. Many cattle running on pastures in
which the diseased grass is plentiful perish when under the influence of
the poison by getting down in the pasture out of reach of water and feed.
A good many others, too, perish by drowning in pools or ponds of shallow'
water. They fall into the water in a nervous paroxysm and drown
before getting over it.

Guinea pigs used in feeding experiments showed nervousness after
being fed 50 sclerotia that had been picked from old heads of Paspalum
dilatatum . Continued feeding produced death within a week or less.
In most cases the sclerotia were given in doses of 25 a day.

In feeding experiments carried on during the summer of 1915 it was
found that sclerotia that had been in the laboratory for about 10 months

1 Brown, H. B., and Ranck, E. M. Forage poisoning due to Claviceps paspali on Paspalum. Miss. Agr.
Exp. Sta. Tech. Bui. 6, 35 p„ 18 fig. 1915.

Nov. 27, 1916

Claviceps paspali

405

were still poisonous and that a small amount of the extract of Claviceps
paspali exposed to air and hot summer temperature was still active after
a period of about 10 months.

A guinea pig fed 40 grass spikelets daily for seven days, each con¬
taining a mass of Claviceps paspali tissue in the sphacelia, or honeydew,
stagehand 60 each day for the next 36 days, showed no bad effects, but
gained in weight. Another pig fed 25 young sclerotia daily for 7 days
and 40 daily for the next 21 days showed no bad effects, but gained in
weight slightly. This feeding was started on July 16. The last two
experiments seem to indicate that it is only the old sclerotia that are
poisonous. The experience of farmers with cattle on pastures indicates
the same.

Mowing pastures one or more times during the late summer or autumn,
or as often as mature sclerotia become abundant, is an effective method
of preventing poisoning, and is a measure of practical value in most
places.

PLATE 32

A. — Section through, a mature stromatic head of Claviceps paspali , showing peri-
thecia containing asci . X 45 •

B. — Spike of P asp alum dilatatum with mature sclerotia attached. Nearly natural
size.

C. — Tufts of hyphae producing sphacelial spores. X150.

D. — Section of mass of tissue within grass spikelet during sphacelia stage of Clavi¬
ceps paspali ; spores are produced by tufts of hyphae along edge of section. X 50.

E. — Sclerotium of Claviceps paspali with stromata. X 5.

F. — Spikes of Paspalum dilatatum , showing a number of sclerotia attached. About
one-half natural size.

(406)

il of Agricultural Research

EFFECT OF SODIUM SALTS IN WATER CULTURES ON
THE ABSORPTION OF PLANT FOOD BY WHEAT
SEEDLINGS

By J. F. BreazealE,

Laboratory Assistant , Biophysical Investigations , Bureau of Plant Industry

INTRODUCTION

The following experiments were undertaken to determine the extent
to which the presence of the various sodium salts commonly found in
alkali soils affects the absorption of plant -food elements by wheat seed¬
lings. Sodium chlorid, sodium sulphate, and sodium carbonate in con¬
centrations ranging from 50 to 1,000 p. p. m.1 were employed in connec¬
tion with a standard nutrient solution, consisting of 200 p. p. m. of N03
as sodium nitrate, 200 p. p. m. of K20 as potassium chlorid, and 1 30 p. p. m.
of P205 as sodium phosphate, together with calcium carbonate (CaC03) in
excess. The same variety of hard wheat, Minnesota Bluestem C. I. 169
( Triticum vulgar e) , was used in all the measurements.

CULTURE METHOD

The enameled culture pans each contained 2,500 c. c. of the nutrient
solution. Each pan was provided with a perforated aluminum disk, sup¬
ported on sealed glass buoys, so as to float at the surface of the solution.
Wheat seeds were sprinkled over the disks in numbers sufficient to pro¬
vide about 1,000 seedlings in each pan.

The nutrient solution in each pan was changed every two days and
during the intervening period was kept approximately at the original vol¬
ume by the addition of water. The analyses showed that with this
method of procedure there was always an abundance of plant food at the
disposal of the seedlings.

During the first two days of the experiment the seedlings were grown
in the nutrient solution alone. At the end of the second day sodium
chlorid, sodium sulphate, and sodium carbonate were added to the nutri¬
ent solution in concentrations varying from 50 to 1,000 p. p. m., as shown
in Tables I to IV. The sodium carbonate in the lower concentrations
gradually changed to sodium bicarbonate, owing to the absorption of
carbon dioxid from the atmosphere and to its evolution fronj the roots of
the growing seedlings. Where the original concentration was 300 p. p. m.
and above, sodium carbonate was still present after the plants had been
grown in the culture solution for two days.

1 Parts per million in solution by weight.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.
gh

(407)

yoi. vn, No. 9

Nov. 27, 1916
G— 100

408

Journal of Agricultural Research

Vol. VII, No. 9

The cultures were carried on for 21 days in the sunshine at Riverside,
Cal. At the end of this period the green and dry weights of 100 repre¬
sentative plants from each culture pan were determined, and the plants
were analyzed by official methods for nitrogen, phosphoric acid, and
potash. The results are given in Tables I, II, III, and IV, the weight in
each instance being based upon 100 plants.

SODIUM CHLORlD

The experimental data obtained with nutrient solutions containing
graduated amounts of sodium chlorid are given in Table I, and in figure 1
the quantities of potash, phosphoric acid, and nitrogen contained in
100 plants are plotted separately against the sodium-chlorid concentra¬
tion. The presence of sodium chlorid in the nutrient solution appears to

►

0 '

- - Tt - :

\ A/

k X

*2 O $

A/aC/ /^/-yvc/r/?/£r/vT solc/t/oa/^

Fig. 1. — Graphs showing the effect of sodium chlorid in nutrient solutions on the nitrogen, potash, and
phosphorie-acid content of wheat seedlings.

diminish very slightly the potash and nitrogen content of the wheat
plants, but the effect is so small as to be comparable with experimental
errors. The phosphoric-acid content of the wheat seedlings appears
to be quite independent of the amount of sodium chlorid in the culture
solution for the range in concentration here employed.

The nitrogen, potash, and phosphoric-acid content of the seedling
wheat plants grown in the presence of sodium chlorid are also expressed
in Table I in percentage of the dry weight of the plants. The results
in this form are not as concordant as those already discussed, but lead
to the same conclusion — namely, that the presence of sodium chlorid in
culture solutions in graduated concentrations up to 1,000 p.p. m. has
very little effect upon the total nitrogen, potash, and phosphoric-acid
content of young wheat plants.

Nov. 27, 19x6 Effect of Sodium on Absorption of Plant Food 409

Table I. — Effect of sodium chlorid on the weight and composition of wheat seedlings

Culture

Parts per million of
sodium chlorid
added to nutrient
solution.

Green

weight

Dry-
weight
of 100
plants.

Weight of element in ioo plants.

Percentage of dry
weight of plants.

of IOO
plants.

KaO.

P2O5.

K20.

P2O5.

O .

Gm.

7i- 5

Gm.

6.21

Gm.

O. 240

Gm.

0. 41 X

Gm.

O. 125

3-9

6.6

2. O

50 . ; .

59- 5

5- 7°

. 236

•334

. 119

4- 1

5-9

2. I

IOO .

58-5

5- 70

. 252

•343

• 125

4.4

6. 0

2. 2

200 .

61. s

5- 70

.249

•345

. 124

4. 4

6. 1

2. 2

. 5

3 00 . . .

63-5

5- 70

•251

•332

• 125

4.4

s-f

2. 2

400 .

64- 5

5- 96

• 230

.288

•US

3-9

4.8

1.9

5 00 . . •

59-o

5- 29

.231

• 348

• 133

4.4

6.6

2- 5

1,000 .

56. 6

5* 55

. 216

.322

. 121

3-9

5-8

2. 2

Fig. 2. — Graphs showing the effect of sodium sulphate in nutrient solutions on the nitrogen, potash, and
phosphoric-add content of wheat seedlings.

SODIUM SULPHATE

The data obtained with culture solutions containing graduated amounts
of sodium sulphate are given in Table II. The results of the analyses
of the plants for nitrogen, potash, and phosphoric acid are given both in
terms of the actual amounts found in 100 plants from each culture and
also in percentage of the dry weight of the plants. The dry weight is
nearly uniform, so that the percentage relationship does not differ
materially from that represented by the absolute amounts of nitrogen,
potash, and phosphoric acid found. - The latter determinations are plotted
as ordinates in figure 2 and the concentration of sodium sulphate in the
culture solutions as abscissae. The results show that the addition of
sodium sulphate to the nutrient solution in concentrations up to 1,000
p. p.m. has practically no effect on the total nitrogen content of the
young wheat plants. In concentrations greater than 400 p. p. m. the
sodium sulphate depressed the potash content slightly. In the case of
phosphoric acid the plants show a very slight but steady decrease in the

4io

Journal of Agricultural Research

Vol. VII , No. 9

total phosphoric-acid content as the concentration of the sodium sulphate
increases. This effect is in evidence throughout the range of sodium-
sulphate concentrations employed.

■Table II. — Effect of sodium sulphate on the weight and composition of wheat seedlings

Cul¬

ture

Parts per million of so¬
dium sulphate add¬
ed to nutrient solu¬
tion.

Green

weight

of IOO

plants.

Dry

weight

of IOO

plants.

Weight of element in

100 plants.

Percentage of dry-
weight of plants.

No.

K2O.

P2O5.

K2O.

P2O5.

0 .

Gm.
63. I

Gm .

5- 85

Gm.

O. 206

Gm.

O. 369

Gm.

O. 141

3- 5

6-3

2. 4

5° . .

67. 0

6. 07

. 178

• 367

•ISO

2.9

6. 0

2. 4

100 .

69.4

6. 23

. 163

• 367

. 140

2. 6

5*9

2. 2

200 .

66.6

5* 75

.181

• 385

• 133

3- 1

6. 7

2*3

300 .

67. 2

6. 08

• J94

• 370

• 125

3-2

6. 1

2. 1

400 .

65- 5

6. 13

. 184

• 367

. 120

3*o

6. 0

2. 0

s°° .

62. 5

5-75

. 194

•31*

. 120

3*4

5*4

2. 1

1 ,000 .

59-o

5* 75

. 176

. 248

. I06

3*i

4-3

1.8

SODIUM CARBONATE

The data obtained with culture solutions containing sodium carbonate
are presented in Table III. The absolute quantities of potash, nitrogen,
and phosphoric acid found in ioo seedlings are plotted in figure 3 against

Fig. 3. — Graphs showing the effect of sodium carbonate on the nitrogen, potash, and phosphoric-acid con¬
tent of wheat seedlings. First series.

the sodium-carbonate concentration. The results show a marked
reduction in the amount of potash and phosphoric acid in the seedlings
as the concentration of the sodium carbonate increases, the total potash
or phosphoric-acid content of 100 wheat plants grown in the presence of
1,000 p. p. m. of sodium carbonate being only one-third that of the
plants grown in the control solutions. The total nitrogen content of
the wheat plants is also slightly decreased.

Nov. 27. 1916 Effect of Sodium on Absorption of Plant Food

41 1

Table III. — Effect of sodium carbonate on the weight and composition of wheat seedlings

FIRST SERIES

Cul¬

ture

Parts per million of so¬
dium carbonate add¬
ed to nutrient solu¬
tion.

Green
weight
of IOO
plants.

Dry-
weight
of IOO
plants.

Weight of element in ioo plants.

Percentage of dry
weight of phtnts.

No.

K2O.

P2O5.

K2O.

P2O5.

. I

0 .

Gm.
60. 3

Gm.

5- 59

Gm.

O. 128

Gm.

O. 329

Gm.

O. 130

2-3

5-8

2-3

50 .

61. 0

5- 65

• 139

•327

. 118

2. 5

5-8

2. 1

IOO .

S3- 6

4.89

. 156

. 287

. 108

3- 2

5-9

2. 2

200 .

53-2

4- 93

. 164

•307

• i°5

3-3

6. 2

2. 1

300 .

50.5

4. 92

. 140

•275

•093

2.9

5-6

1.9

400 .

32. 7

3- 36

. 148

.158

.050

4. 4

4*7

i-5

500 .

28. 0

3.81

. 098

• 113

. 049

2. 6

2.9

i-3

1,000 .

i5- 5

2. 77

. 087

. 047

•037

3-2

i-7

i-3

SECOND SERIES

O. . . .

50.. .

IOO. .
200. .
30°. .

400. .

500. .

1,000

71- s

57- 5
57- 5
52- 5
40. 5

31-3

36. s

20. 5

6. 21
5. 61
5. 42
5- 11
4. 42
4-25
4. 11
3. 16

o. 240
. 230

*257

. 209
. 160

• 125
■ 144

. no

o. 41 1

• 332

• 331

•359

. 291

• 230

• 144

. 061

o. 125

• 115

. 116
. 101
.097
. 067

• 057

. 046

3- 9

4. 1

4- 7

4. 1

3*^

2.9

3- 5
3- 5

6.6

5-9
6. 1
7.0
6. 6
5-4
3- 5

1.9

2. o
2. 1
2. 1
2. o
2. 2
1. 6
1.4

Fig. 4. — Graphs showing the effect of sodium carbonate on the nitrogen, potash, and phosphoric-acid con¬
tent of wheat seedlings. Second series.

These results are in such striking contrast with the effects obtained
with the other sodium salts that the sodium-carbonate series was re¬
peated. The results of the second series of determinations are also given
in Table III and are presented graphically in figure 4. A marked reduc-
64313°— 16 - 3

412

Journal of Agricuimral Research

Vol. VII, No. 9

tion in the potash and phosphoric-acid content is again shown as the
sodium-carbonate concentration increases. In this series also there is a
decided reduction in the total nitrogen content with increasing concentra¬
tion of the sodium carbonate.

Reference to Table III will show that the weight of the seedlings de¬
creased markedly as the concentration of the sodium carbonate in¬
creased. It is consequently of interest to express the nitrogen, potash,

Fig. 5. — Graphs of the mean values of the first and second series showing the effect of sodium carbonate
on the nitrogen, potash, and phosphoric-acid content expressed in percentage of the dry weight of wheat
seedlings.

and phosphoric-acid content of the seedling plants in percentage of
their dry weight. The results computed on this basis will be found in
the last three columns of the table. The mean values for both series of
determinations are plotted in figure 5. It will be seen that the percentage
of nitrogen does not show any consistent change as the concentration of
the sodium carbonate increases. The percentages of potash and phos¬
phoric acid, on the other hand, decrease markedly with increasing con¬
centration of the sodium carbonate.

Nov. 27, 1916 Effect of Sodium on Absorption of Plant Food 413

NITROGEN, POTASH, AND PHOSPHORIC ACID ABSORBED FROM THE
NUTRIENT SOLUTION BY THE VARIOUS CULTURES

The data so far given include the nitrogen, potash, and phosphoric
acid stored in the seed. The analysis of 100 seeds for these substances

gave the following results:

Gram.

Weight of 100 seeds (dry) . 2. 45

Nitrogen . 0486

Potash (K20) . 0185

Phosphoric acid (P206) . 0242

If it is assumed that each lot of 100 seeds is of uniform composition
and weight, the amount of nitrogen, potash, and phosphoric acid ab¬
sorbed from the nutrient solutions by the various cultures can be deter¬
mined by deducting the above quantities from those found in the plants
grown in the culture solutions. The data obtained from the various
cultures, reduced to this basis, are given in Tables V and VI, and are
presented graphically in figures 6 to 8, the last figure representing the
mean of the two sodium-carbonate series.

TabbE V. — Effect of sodium chlorid and sodium sulphate on the absorption of nutrients by

wheat seedlings

Sodium chlorid.

Sodium sulphate.

Culture No.

Sodium
chlorid
added to
nutrient

Elements absorbed from solu¬
tion (in percentage of dry-
weight of plants).

Sodium
sulphate
added to
nutrient

Elements absorbed
from solution (in
percentage of dry
weight of plants).

solution.

KsO.

P2O5.

solution.

K2O.

P2O5.

P. p . m.

3* 1

6-3

i-.7

P. p. m.

2. 8

6. 0

2. 0

3*3

. 5-5

i-7

5°

2. 1

5- 7

2. 1

3 .

IOO

3- 6

5-7

1.9

IOO

1. 8

5- 6

i-9

4 .

200

3*5

5-9

1. 8

200

2-3

6. 4

1.9

5 .

300

3-6

5-5

1.9

300

2.4

5*8

1.8

6 .

400

3-o

4* S

i- 5

400

2. 2

$• 7

1. 6

7 .

500

3-5

6. 6

2. 1

500

2. s

5- 1 .

1. 7

8 .

I, OOO

3*o

5- 5

I* 7

I, OOO

2. 2

4. 0

x*4

The percentage of nitrogen absorbed by the young wheat plants does
not appear to be measurably modified by the presence of any of the
sodium salts investigated in concentrations up to 1,000 p. p. m. Sodium
chlorid in this concentration does not affect the absorption of phosphoric
acid measurably (fig. 6), but depresses slightly the percentage of potajsh
absorbed. Sodium sulphate depresses the absorption of potash decidedly
and of phosphoric acid slightly (fig. 7), while with sodium carbonate the
depression in the absorption of both potash and phosphoric acid is very
marked (fig. 8). The depressing effect *>f sodium carbonate on the

414

Journal of Agricultural Research

Vol. VII, No. 9

absorption of potash and phosphoric acid is in evidence even in con¬
centrations of sodium carbonate as low as 100 p. p. m. It is evident

Fig. 6. — Graphs showing the effect of sodium chlorid on the absorption of nutrients by wheat seedlings.

Fig. 7. — Graphs showing the effect of sodium sulphate on the absorption of nutrients by wheat seedlings.

from these measurements that the presence of sodium carbonate, even
in these minute amounts, may have a markedly deleterious effect upon
the metabolism of the small-grain crops.

Nov. 27, 1916 Effect of Sodium on Absorption of Plant Food

4i5

Table VI. — Effect of sodium carbonate on the absorption of nutrients by wheat seedlings

FIRST SERIES

Culture No.

Sodium
carbonate
added to

Elements absorbed from solution (in per¬
centage of dry weight of plants).

nutrient

solution.

KaO.

Pa05.

P. p. m.

1*4

*5*6

1.9

5°

1. 6

5*6

I* 7

3 .

IOO

2. 2

5* 5

i*7

4 .

200

2*3

5*9

1. 6

5 . .

6 .

3° 0
400

1.9
3- 0

5- 2

4. 2

1.4

7 .

500

i*3

2* 5

• 7

8 .

I, 000

1. 4

1. 0

•4

Fig. 8. — Graphs showing the effect of sodium carbonate on the absorption of nutrients by wheat seedlings.

It has been shown by Le Clerc and Breazeale1 that there is a very
marked absorption of potash by sprouting grain seedlings. As potash
appears to be vitally concerned in the decomposition and translocation

1 Le Clerc, J. A., and Breazeale, J. F. Translocation of plant food and elaboration of organic plant
material in wheat seedlings. U. S. Dept. Agr. Bur. Chem. Bui. 138, 32 p., 2 fig. 1911.

4i 6 Journal of A gricultural Research voi. vn, no. 9

of carbohydrates, any salt such as sodium carbonate, which would
interfere with the absorption of potash §it this stage of growth would
seriously handicap the development of the plant.

It will be recalled that calcium carbonate was present in the culture
solutions in the solid phase. It will be shown in another paper that
appreciable quantities of sodium carbonate are formed through the
reaction of sodium chlorid and sodium sulphate with calcium carbonate,
and the resulting hydrolysis is greater with sodium sulphate than with
sodium chlorid. This is in harmony with the greater activity shown
by sodium sulphate in depressing the absorption of potash and phos¬
phoric acid and suggests that the effect observed in the case of sodium
chlorid and sodium sulphate may be in part due to the small amounts
of sodium carbonate formed through reaction with the calcium carbonate.

CONCLUSIONS

Sodium chlorid, sodium sulphate, and sodium carbonate added to
nutrient solutions in concentrations up to 1 ,000 p. p. m. do not measur¬
ably affect the nitrogen absorbed from culture solutions by young wheat
plants. ,

Sodium chlorid in concentrations up to 1,000 p. p. m. does not affect
the absorption of phosphoric acid, but decreases slightly the absorption
of potash.

Sodium sulphate in concentrations of 1,000 p. p. m. depresses the
absorption of potash and phosphoric acid to approximately 70 per cent
of that of the control cultures, expressed in percentage of dry weight of
the plants.

Sodium carbonate in concentrations of 1,000 p. p. m. reduces the
absorption of potash to 20 per cent of that of the control and the absorp¬
tion of phosphoric acid to 30 per cent of that of the control. The depress¬
ing effect of sodium carbonate is in evidence in concentrations as low as
100 p. p. m., and is marked in concentration of 300 p. p. m.

The relative effect of sodium sulphate and sodium chlorid in depressing
the absorption of potash is directionally the same as the relative hydroly¬
sis resulting from the reaction of the twTo salts with the calcium car¬
bonate present in the culture solution. This suggests that the observed
effects in the case of sodium sulphate and sodium chlorid may be due
in part to the accumulative action of the slight amounts of sodium
carbonate formed in this reaction.

JOURNAL OF ACRKETffiAL RESEARCH

DEPARTMENT OF AGRICULTURE

Vol. VII Washington, D. C., December 4, 1916 No. 10

NITRIFICATION IN SEMIARID SOILS — I.1

By W. P. Kelley,2

Professor of Agricultural Chemistry , Graduate School of Tropical Agriculture and Citrus
Experiment Station , University of California

HISTORICAL INTRODUCTION

The distribution and amounts of nitrogen in the humus of the arid and
semiarid soils of America and the activities of the microorganisms con¬
tained therein have been discussed at length by a number of writers. It
has been especially emphasized that, as a rule, the humus and nitrogen
are distributed more uniformly to a greater depth in the subsoil of the
arid than of the humid regions. The soils are commonly very deep and
are not sharply separated from the subsoils, but the actual percentage
of nitrogen in the surface soil is frequently low and usually decreases in
passing downward into the subsoil. The soils and subsoils in many parts
of California, for example, contain less than 0.05 per cent of nitrogen
and in many localities even less than 0.03 per cent.

Hilgard (10), 3 Loughridge (26), Lipman (16), and others have directed
attention to the great depth of root penetration in the semiarid region,
and Loughridge has suggested that the apparent deficiency of nitrogen
in the surface soils may be compensated for by the distribution of nitro¬
gen in the deep subsoils. He finds, for example, that the total nitrogen
in the zone occupied by plant roots commonly compares favorably with
that in humid regions. But, as is well known, the application of nitroge¬
nous fertilizers commonly results in marked stimulation to crops, and
in many localities successful crop production depends upon the use of
nitrogenous fertilizers. As much a£ 1,000 to 1,500 pounds of dried
blood per acre has been applied annually to some of the Citrus groves
of southern California, and in some cases as much as 2,000 pounds per
acre has been applied. Corresponding amounts of other nitrogenous
materials have been used.

1 Paper 35, Citrus Experiment Station, College of Agriculture, University of California, Riverside, Cal.

2 The writer acknowledges the valuable analytical assistance of Mr. A. B. Cummins.

3 Reference is made by number to “ Literature cited,” p. 436-437.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.
gi

(417)

Vol. VII, No. 10
Dec. 4, 1916
Cal. — 7

4i 8 Journal of Agricultural Research voi. vn. No. »

In the course of investigations on the origin of the so-called ‘'niter
spots” in Colorado, Sackett (28) concluded that nitrification is unusually
active in Colorado soils and more so than in soils from certain other
Western States. Stewart, Greaves, and coworkers (31, 32, 33, 34) in
Utah concluded, on the other hand, that the excessive accumulation of
nitrates in the surface soils of certain localities is to be accounted for
mainly by the capillary rise of nitrates from deeper strata rather than
from unusual nitrifying activity in the soils at the present time.

Lipman (16), by the use of the modified Remy solution method, found
that ammonification and nitrification are most active in the first foot of
California soils, but in some cases these processes were also found to take
place quite actively in the subsoils down to a depth of 8 to 10 feet. He
concluded that the deep penetration of the roots of cultivated plants may
be accounted for in part by the active formation of available nitrogen in
the -deep substrata. McBeth and Smith (27) also found from experiments
with the use of actual soil as the culture medium that dried blood and
ammonium sulphate undergo active nitrification in the first foot of
certain Utah soils, but the activity was found to decrease markedly in the
successive sections below the first foot until at a depth of 4 or more
feet it almost ceased. In their experiments about 90 per cent of the
total nitrate produced in the 5-foot sections studied was formed in the
first foot, from which it would seem that nitrification is not particularly
active in the substrata below the first foot.

Greaves (8) found that, as a rule, cultivation brings about an increase
in the numbers of organisms and in the rates of nitrification and nitrogen
fixation in certain soils of Utah ; he has also studied the effects of different
arsenic compounds on the biological activities in soils (7, 9). Lipman
and Burgess (17, 18, 19, 22, 23) have devoted considerable study to the
effects of alkali salts and small amounts of copper, zinc, iron, and lead
compounds on ammonification and nitrification in California soils.

Recently Lipman and Burgess (24) published experiments on the rates
of nitrification of different fertilizers in 29 different soils from California.
Briefly, it was found that, on the whole, ammonium sulphate was most
actively and quite vigorously nitrified in most of the soils studied. The
rates at which the different organic substances underwent nitrification
varied widely. In certain soils low in organic matter little or no nitrate
was formed from dried blood or high-grade tankage, while at the same
time cottonseed meal, bone meal, garbage tankage, and other low-
nitrogen-containing materials were quite vigorously nitrified. They
concluded as follows :

In all soils of our interior arid valleys which are not very close to stream channels
or those which for other reasons are markedly deficient in organic matter, the proper
bacteriological and perhaps other conditions do not obtain to render into nitrates
most economically and quickly the nitrogen of high-grade organic nitrogenous fer¬
tilizers. On the other hand, conditions in those same soils are much more favorable
for the nitrification of nitrogen of the low grade nitrogenous fertilizers. Similar con-

Dec. 4, 1916

Nitrification in Semiarid Soils

419

ditions prevail in the humus-poor soils of our coast valleys and of other valleys in the
state, in which either the soils have always been deficient in organic matter or have
become depleted in that respect through excessive oxidation under favorable cli¬
matic conditions assisted by constant summer cultivation.

In connection with the above and other investigations, Lipman has
found that ammonification is active in California soils generally and that
in some cases greater amounts of ammonia than nitrate occur in soil
samples freshly drawn from the field. This condition was especially
noted in soils on which the plant diseases known as dieback and mottle-
leaf of Citrus spp. occur. In these soils Lipman (20, 21) found nitrifi¬
cation to be inactive. He suggested a causal relationship between the
active ammonification and inactive nitrification on the one hand and the
abnormal plant growth on the other, the latter being attributed to .
enforced ammonia absorption occasioned by the inability of the soil to
transform ammonium nitrogen into nitrates.

As contrasted with the above hypothesis, Kellerman and Wright (11)
have pointed out that the physiological disease known as mottle-leaf of
Citrus trees may be caused by excessive accumulations of nitrates in the
soil; but the source of the nitrate, whether being actively formed at the
present time or otherwise, is not clear from their publication.

Beckwith, Vass, and Robinson (3) have studied the effects of lime on
the ammonification of dried blood and peptone, and the nitrification of
dried blood and ammonium sulphate in six soils from Oregon. In every
case they found that active ammonification took place, but in two of the
soils less nitrate was found after four weeks’ incubation where either
dried blood or ammonium sulphate had been added than in the portions
to which no nitrogenous material was added. In one case the further
addition of lime failed to induce the nitrification of these materials. An
increase in the nitrate content in the check portions was found in every
case, indicating the presence of the nitrifying organisms in the soil.

From the foregoing partial review of the literature on this subject it
is apparent that radical differences of opinion are held with reference
both to the formation and the movement of nitrates in the soils of the
semiarid region.

In view of the economic importance of nitrogen and the scientific
interest attached to nitrification, the writer has for some time been
engaged in a series of studies on this subject at the University of Cali¬
fornia Citrus Experiment Station, at Riverside, Cal. At this place a
fertilizer experiment with Citrus trees has been maintained during the
past nine years. The plots of this experiment and other semiarid soils
near by have been used in these investigations and have made it possible
to compare the data obtained in the laboratory studies with the effects
produced in the field.

The relative rates of nitrification in the field and laboratory, the
effects of soil treatments including different fertilizers, manure, and cover

420

Journal of Agricultural' Research

Vol. VII, No. io

crops, the effects of alkali salts, the relative nitrifiability of different fer¬
tilizing substances, seasonable variation in nitrification, the movement
of nitrates, and other phases of this question are being studied. The
results obtained in the early stages of this work strongly emphasized the
need for further study of the methods to be used.

In the previous studies on nitrification in semiarid soil the Remy-
solution method, with certain modifications, has been used to a limited
extent; but usually the direct-soil method has been used, in which actual
soil is employed as the culture medium.1 Different investigators, how¬
ever, have modified the details to suit their own ideas. These modi¬
fications have to do mainly with variations in temperature, moisture
content, and periods of incubation on the one hand, and differences in
the ratio of soil to nitrogenous materials on the other. Regarding the
latter it is noteworthy that the percentage of nitrogenous, organic mate¬
rial employed has been varied from about 0.7 to 2 per cent. In the
case of ammonium sulphate the variations have ranged from 0.1 to 1
per cent.

Likewise, widely variant percentages of actual nitrogen from different
sources have been added by one and the same investigator. For example,
Lipman and Burgess (24) employed equal weights (1 per cent) of calcium
cyanamid, dried blood, bone meal, high-grade tankage, cottonseed meal,
manure, etc. ; but since the nitrogen content of these materials ranged
from 2.46 per cent to 16.55 Per cent, the actual quantities of nitrogen
added* must have varied accordingly. Ammonium sulphate was added
at the rate of 0.2 per cent. On the basis of the data thus obtained,
deductions were made concerning the relative nitrifiability of these mate¬
rials. Likewise, McBeth and Smith (27) added dried blood and ammo¬
nium sulphate at the rates of 1 and 0.08 per cent, respectively; and
while the absolute amounts of nitrate formed from the latter were con¬
siderably larger than from the former, no particular notice was given to
it, an average of the results from the two forms being recorded in many
cases.

Before presenting the full data bearing upon the specific subjects
named above, the present paper will be devoted to a discussion of the
methods commonly employed in laboratory studies on nitrification with
special reference to the concentration of nitrogenous materials and period
of incubation used. In most cases the phenol-disulphonic-acid method
was used for the determination of nitrate. At frequent intervals through¬
out this investigation the aluminum reduction method as outlined by
Burgess (5) was also used for the purpose of checking the results obtained
by the colorimetric method. The results by the two methods were found
to agree closely in all cases, except where high concentrations of nitroge-

1 It is not deemed necessary to discuss in detail the advantages and disadvantages of these two methods.
The reader is referred to a paper by Tohnis and Green (25) in which a critical review of the subject is given.
Some of the points emphasized below were also strongly emphasized by them two years previously.

Dec. 4, 1916

Nitrification in Semiarid Soils

421

nous materials or other abnormal conditions were employed. In such
cases the results by the reduction method were frequently much higher
than by the colorimetric method. Further reference will be made to
this point later. It is recognized that the colorimetric method is not
accurate where high concentrations of nitrate occur, but the results are
believed to be sufficiently accurate for the purposes of this paper.

EXPERIMENTAL WORK

The soil used was for the most part drawn from the fertilizer plots 1
referred to above. This soil has been derived from the disintegration
of monzonite and is a light, sandy loam, very low in organic matter and
nitrogen. It is underlain with a deep subsoil similar in nature to the
surface soil. A composite sample composed of about 20 borings was
obtained from each plot sampled and also from the virgin soil 2 near by.
The samples were taken to a depth of 6 inches with a King soil tube and
were then immediately brought to the laboratory and spread out on clean
paper to dry. After becoming air-dry and being thoroughly mixed,
duplicate portions of 150 gm. were mixed in tumblers with 1 .5 gm. of dried
blood, and the moisture content made up to 15 per cent with distilled
water, after which the samples were incubated at 250 to 28° C.

In order to ascertain the formation of ammonia, 50-gm. portions were
withdrawn at the end of 7 and 28 days, and the ammonia determined by
distilling with an excess of magnesium oxid, and at the end of 28 days
the nitric nitrogen was determined. The average of closely agreeing
duplicates is recorded in Table I.

Tabu$ I. — Ammonificaiion and nitrification of I per cent dried bfood (in parts per

million)

Soil.

Ammonia nitrogen.

Nitric nitrogen.

After 7
days.

After 28
days.

Original

soil.

Gain in 28
days.

Virgin .

584

38s

I. 2

5- 5

Control plot .

503

400

2. I

2. 5

Manured plot .

497

287

8.4

241. 6

The* foregoing data show that active ammonification took place in the
soil from each plot studied and that a relatively high concentration of
ammonia still occurred at the end of 28 days. The lesser amount of
ammonia found in each case at the end of 28 days was probably due
in part to the loss of ammonia by volatilization. Strong odors of ammonia
were detected, especially in the tumblers containing the virgin soil and

1 It is not deemed necessary at this point to discuss in detail the different treatments that have been
applied in the field experiments. A more complete discussion will be presented in a subsequent paper. '

2 The term "virgin soil’' as used in this paper signifies uncultivated soil still bearing native vegetation.

422

Journal of Agricultural Research

Vol. VII, No. 10

that from the control plot. It is not deemed necessary to dwell further
on the ammonification of organic materials in these soils. Suffice it to
say that the ammonia has been determined in a large number of instances
throughout this work and without exception ammonification has been
found to be active in every soil studied. The above data on ammonifi¬
cation are submitted merely to show that nitrification was not limited
by inactive ammonification and not as a special contribution to the
study of ammonification.

Considering the amounts of nitric nitrogen found, it will be noted
that little or no nitrification took place except in the soil previously
treated with manure. In this case quite active nitrification took place*
It is of special interest that only the most enfeebled nitrification of dried
blood took place in the virgin soil; and although slightly more nitrate
was formed in the virgin soil than in the check plot, the difference is too
small to be noteworthy.

Two of the plots (C and S) in the field experiments from which the
above samples were drawn have been annually fertilized with dried blood
for the past nine years (1907-1915). During the past two years (1914-
15) the application has been made at the rate of 1,080 pounds per acre.
Notable stimulation in the growth and yield of fruit has been produced.
One of these plots (C) lies adjacent to the check plot used in the foregoing
experiments. Soil samples drawn at frequent intervals during the past
two years from this and other plots fertilized with dried blood have
consistently shown a well-defined increase in nitric nitrogen over that in
the unfertilized plots. Furthermore, considerable increases in the
nitrate content have been found following each application of dried blood.
It would seem, therefore, that dried blood undergoes active nitrification
in the field where no other form of organic matter has been applied,
notwithstanding the fact that the above data indicate that both the virgin
soil and control plot are unable to nitrify dried blood.

Two questions presented themselves: First, why does dried blood
undergo active nitrification in the field but not in the laboratory?
Second, why is it that dried blood undergoes active nitrification in the
soil from the manured plot but apparently not in the control plot?
Considerations arising out of these questions have led to an extended
study of the factors affecting nitrification in the field and laboratory.

NITRIFICATION AS AFFECTED BY VARYING CONCENTRATIONS OF

MATERIALS

As stated above, 1,080 pounds of dried blood per acre have been ap¬
plied annually for the past two years (1914-15) to plots C and S. Until
the present year (1916) only one-third of this quantity was applied at
one time, the remaining two- thirds being applied at intervals of about
two months each. But assuming that the entire amount becomes thor¬
oughly mixed with the soil to a depth of 6 inches and estimating that the

Dec. 4, 1916

Nitrification in Semiarid Soils

423

soil weighs 2,000,000 pounds per acre of 6 inches, the average concentra¬
tion of dried blood that obtains in the field would be 0.054 Per cent. On
the other hand, a concentration of 1 per cent was employed in the pre¬
ceding laboratory experiments, which is 18.5 times that which obtains
in the field.

It was at once suggested that an excessive concentration had been used
in the laboratory. Accordingly a preliminary set of experiments was
made, in which widely different concentrations of dried blood were added.
The result was that the soil from the plots which had previously failed to
nitrify a 1 per cent concentration of dried blood was found to support
active nitrification of this material when added in low concentrations.

An extended study has been made on the rates of nitrification of differ¬
ent nitrogenous materials when used in varying concentrations. Fresh
samples were drawn from the same plots as in the preceding series. The
nitrogeneous materials used were dried blood, bone meal, and ammonium
sulphate, representing a high-grade and a low-grade organic form and an
inorganic compound, respectively. The dried blood (13.20 per cent of
nitrogen) was added in quantities ranging from 1 to 0.0625 Per cent. The
bone meal (4.25 per cent of nitrogen) was varied from 4 to 0.25 per cent.
Ammonium sulphate (21 per cent of nitrogen) was varied from 0.6 to
0.0375 per cent. The experiments were made in duplicate. Control por¬
tions of each soil without the addition of nitrogenous material were also
incubated. The incubation period was four weeks. The results are
given in Table II.

Table; II. — Nitrification as affected by different concentrations of nitrogenous materials

Virgin soil.

Control plot.

Manured plot.

Nitrogen

Materials added.

added per
100 gm.

Nitric

Percent¬

Nitric

Percent¬

Nitric

Percent¬

of soil.

nitrogen

age nitri¬

nitrogen

age nitri¬

nitrogen

age nitri¬

found.

fied.

found.

fied.

found.

fied.

Mgm.

P. p. m.

T'Jnne .

34.O

19. 6

38-7

Dried blood:

1.0 per cent .

132. 0
66.0

22. O

3-8

21. O

3i6-°

248. O
140. O

21. 0

0.5 per cent .

42. O

57*5

I. 2

0. 2

3**7
3°* 7

0.25 per cent .

33* 0

7* 1

120. O

30*4

0.125 per cent.

rf-5

hi. 0

46.6

99-0

48. I

96. O

34- 8

0.0625 per cent. .*.

! 8. 25

73*o

47*3

75*o

67. I

82. O

52* 5

Bone meal :

4.0 per cent .

! 170. 0

24- 5

18. 9

l88. O

8.8

2.0 per cent .

: 85.0

2. 1

27* 5

0.9

240. O

23*7

1.0 per cent .

! 42. 5

109. 0

17.7

96. 0

17.9

220. O

42. 7

0.5 per cent .

21.25

99* 5

30*9

104. 0

39*7

150.0

52* 4

0.25 per cent .

10. 62

80. 0

43*3

69. 0

46. 5

91. O

49*3

Ammonium sulphate :

126. 0

0.6 per cent.* .

15*4

31.0

0.9

94.O

4.4

0.3 per cent .

63.0

53*o

3. 0

63.0

6.9

176. O

21. 8

9.8

0.15 per cent .

31- s

65.0

82. 0

19. 8

158.O

37*9

0.075 per cent. . . .

I5- 7

78. 0

28. 0

' 88.0

43-6

148. 0

69. 6

0.0375 per cent. . .
Original soil .

7-85

69. 0
1. 8

44. 6

74.0

4.S

69.3

112. O

I3* 6

93*4

424

Journal of Agricultural Research

Vol. VII, No. io

As shown above, i per cent of dried blood failed to be nitrified in the
virgin soil and that from the control plot, but underwent active nitrifica¬
tion in the previously manured soil as in the preceding series. The
larger quantities of bone meal and ammonium sulphate also failed to be
nitrified, but the lower concentrations of each of these substances under¬
went active nitrification in every case. Dried blood in a concentration
of 0.625 Per cent, which corresponds closely with that used in the field,
underwent active nitrification in all cases. The percentages of nitri¬
fication, calculated after subtracting the amounts of nitric nitrogen
found in the control portions, show that when corresponding amounts
of actual nitrogen from the different sources are compared the rates of
nitrification of dried blood, bone meal, and ammonium sulphate were
quite similar in all cases with the single exception of ammonium sulphate
in the manured soil. In this case ammonium sulphate was oxidized
the most completely of any of the materials studied.1

Soils from other localities have also been studied. Two samples were
obtained from the lemon groves of a ranch in Ventura County. One of
these (A) is a light sandy soil; the other (B), a heavy adobe soil high in
organic matter. A sample was taken from a young lemon grove on
another ranch in Ventura County and is a heavy clay soil, containing
considerable organic matter. Another sample of a light sandy character
was taken from an orange grove opposite the Tark Ellen station near
Covina. A sandy loam containing considerable gravel and organic
matter was obtained from a 24-year-old orange grove in the Ta Verne
section in southern California.

Studies were made in duplicate with the use of the same materials as in
the preceding series. Since a 2 per cent concentration of dried blood (8,
13,18) has been previously used to some extent in studies on nitrification,
this proportion was added in certain cases. The concentrations of bone
meal were varied from 4 to 0.5 per cent, and of ammonium sulphate,
from 0.3 to 0.075 per cent. The results are given in Table III.

Table III. — Nitrification in soils from different localities

Materials added.

Nitro¬
gen
added
per
100
gm.
of soil.

Sespe soil
(A).

Sespe soil
(B).

Limoneira

soil.

Lark Ellen
soil.

La Verne
soil.

Nitric

nitro¬

gen

found.

Per¬

cent¬

age

nitri¬

fied.

Nitric

nitro¬

gen

found.

Per¬

cent¬

age

nitri¬

fied.

Nitric

nitro¬

gen

found.

Per¬

cent¬

age

nitri¬

fied.

Nitric

nitro¬

gen

found.

Per¬

cent¬

age

nitri¬

fied.

Nitric

nitro¬

gen

found.

Per¬

cent¬

age

nitri¬

fied.

Mgm.

P.p.m

P.p.m.

P.p.m,

None. .

61. 0

38.5

49.0

86.0

34*o

Dried blood:

2.0 per cent .

264. 0

6. 5

2 A* <

x.o per cent .

132.0

736.0

51- 1

360. 0

24-3

468. 0

3i- 7

IO7.O

i* 6

442* 5

30-9

0.125 per cent .

16. 5

170.0

66. 1

129.0

54*9

163.0

69. 1

I9I- O

63.6

144.0

66.6

Bone meal:

4.0 percent .

170. 0

220. 0

9.4

224.0

10. 9

332. 0

16. 6

76. O

1.0 per cent .

42. 5

196. 0

34**6

< v w

222* O

0.5 per cent .

21. 25

186. 0

58.8

136.0

45- 9

152.0

48. 5

168. 0

38.6

140. 0

49*8

Ammonium sulphate:

0.3 per cent .

63.0

328.0

26. 5

hi. 5

XI. 6

128.0

12. 5

162.0

12. X

0. 75 per cent .

15- 75

206. 0

'92. 1

106. 0

42.9

148. 0

62. 9

172. 0

54*6

124.0

57* 1

Original soil .

37-o

25-5

3i-5

45*5

23. 0

1 It should be stated that the addition of calcium carbonate exerts almost no effect on the nitrification
of dried blood in this soil. A preliminary report on the above phases of this investigation has previously
been issued (15).

Dec. 4, 1916

Nitrification in Sentiarid Soils

425

Again, it was found that the lower concentrations of these materials
were actively nitrified in every case; and when the concentration was
increased, the percentage of nitrification decreased. One per cent
dried blood was actively nitrified in every soil except that from Lark
Ellen, but 2 per cent was toxic in each case. The percentage of dried
blood nitrified was somewhat greater than that of bone meal; otherwise
the degrees of nitrification of the different materials were similar. The
results, therefore, are in harmony with those of the preceding series.

The preceding data show that each of the soils studied, representing
quite a wide range of soil conditions, is capable of supporting active
nitrification of dried blood, bone meal, or ammonium sulphate, provided
these materials be added in low concentrations. They also indicate that
the results obtained with the use of such high concentration of dried
blood as 1 and 2 per cent, or 0.3 and 0.6 per cent of ammonium sulphate,
do not form a reliable criterion upon which to base practical conclusions.

NITRIFICATION AT DIFFERENT DEPTHS AS AFFECTED BY CONCEN¬
TRATION

As already stated, differences of opinion are held regarding nitrifica¬
tion in the substrata of semiarid soils. The following data are of interest
in this connection. The soil used was taken from an orange grove near
Woodlake, in Tulare County. It is a dark-colored clay loam, high in
organic matter. The subsoil contains considerably less organic matter
than the surface soil and closely resembles adobe. The samples were
drawn in foot sections down to 5 feet in depth. Only a few concentra¬
tions of nitrogenous materials were employed, owing to the smallness of
the samples. The incubation period Avas four weeks, as in the previous
series. (Table IV).

Table IV. — Nitrification in soil from different depths

Materials added.

Nitro¬
gen
added
per
100
gm.
of soil.

First foot.

Second foot.

Third foot.

Fourth foot.

Fifth foot.

Nitric

nitro¬

gen

found.

Per¬

cent¬

age

nitri¬

fied.

Nitric

nitro¬

gen

found.

Per¬

cent¬

age

nitri¬

fied.

Nitric

nitro¬

gen

found.

Per¬

cent¬

age

nitri¬

fied.

Nitric

nitro¬

gen

found.

Per¬

cent¬

age

nitri¬

fied.

Nitric

nitro¬

gen

found.

Per¬

cent¬

age

nitric

fied.

None . .

Mgm.

132- 0
13- 2

13- 2

P.p.m.

60. 0
32. 0
132. 0

130- 0

1. i

P.p.m.
64. 0
2. 6
95-0

92.0

4-5

P.p.m.
46.0
x. 8
100.0

74-0
2. 5

P.p.m.

36.0

0. X
40.0

68.0
3- 6

P.p.m.

35-o
9. 2
34-0

68.0
0. 6

Dried blood, r per cent. .
Dried blood, o.i per cent
Ammonium sulphate,

0.062 s per cent.... _

Original soil .

54- S

53-o

23-5

21. 2

40-9

21. 2

3*o

24. 2

25.0

Active nitrate formation took place in the check portions from each
of the 5 -foot sections studied, showing that the nitrifying organisms are
not only present down to 5 feet in depth but that the chemical, physical,
and biological conditions ensuing in the samples were favorable for nitri-

426

Journal of Agricultural Research

Vol. VII, No. 10

fication. The results show, furthermore, that 1 per cent of dried blood
may be an excessive concentration, even in a highly organic soil; less
nitrate was formed in the samples for each foot where 1 per cent of
dried blood had been added than in the control portions.

A concentration of 0.1 per cent of dried blood, or an equivalent amount
of nitrogen in the form of ammonium sulphate, underwent vigorous and
practically equal nitrification in the soil from the first foot and was also
actively nitrified in the subsoils from the second and third feet. Am¬
monium sulphate likewise underwent considerable nitrification in the
samples from the fourth and fifth feet; but the amounts of nitrate pro¬
duced with a 0.1 per cent concentration of dried blood in soil from these
depths were approximately the same as in the checks, indicating that a
concentration of o. 1 per cent of dried blood may be excessive.

Considering the fact that nitrification took place in the check portions
and where ammonium sulphate was added, the conclusion seems war¬
ranted that the subsoil from this orchard at least possesses the potential
capacity of producing nitrates down to a depth of 5 feet.

However, the writer does not consider it safe to conclude from the
preceding data that active nitrification takes place in the field in the
subsoils of the orchard from which the above sample was drawn, since
much more thorough aeration took place after the samples were drawn
than ordinarily takes place in the subsoil in situ.

The results, as a whole, again emphasize the importance of employing
low concentrations of nitrogenous materials and show that the inability
to nitrify a concentration of 1 per cent of dried blood is not confined to
humus-poor soil, as suggested by Lipman and Burgess (24).

EFFECTS OF ALKALI SALTS ON NITRIFICATION AS MODIFIED BY THE
CONCENTRATION OF NITROGENOUS MATERIALS

One of the most important soil questions in the semiarid region relates
to the effects of alkali salts, particularly the carbonate, chlorid, and
sulphate of sodium. As already stated, Lipman and his coworkers (17,
18, 19, 22) have devoted considerable study to the biochemical effects
of these salts. But the conclusions which were drawn relative to nitri¬
fication were based on the effects produced with a concentration of 1 or
2 per cent of dried blood. In the light of the results presented above,
it becomes a matter of interest to study the effects of alkali salts with
the use of varying concentrations of nitrogenous materials.

The soils used were drawn from the check and manured plots; the
required amounts of the salts were added in solution after the dried
blood or ammonium sulphate had previously been mixed with the soil.
The same percentage of moisture and an incubation period of four weeks
were employed, as in the previous series. Table V gives the results of
this series.

Dec. 4, 1916

Nitrification in Semiarid Soils

427

Tabi,E V.— Effect of alkali salts on the nitrification of dried blood and ammonium sulphate

Nitric nitrogen (per million).

Alkali salt added.

Control plot.

Manured plot.

0.1 per
cent of
dried
blood.

0.0625 per
cent of
ammonium
sulphate.

0.15 per
cent of
ammonium
sulphate.

1.0 per
cent of
dried
blood.

0.1 per
cent of
dried
blood.

0.15 per
cent of
ammonium
sulphate.

None .

108. 0

98. O

89. 0

172. 0

106. O

170. 0

Sodium carbonate :

0.05 per cent .

0.1 per cent . .

104. 0

160. 0

s6. 0
33* 5

31.0
19. O

108. O
102. 0

129. O
62. O

0. 5 per cent .

i* 5

1.9

1. 1

7.6

a no. 0

5* 6

Sodium sulphate:

0.1 per cent .

115. 0

96. O

60. 0

296. O

103.0

174.0

0.5 per cent .

98. 0

95*o

41. 0

48. 0

102. 0

134.0

a 0.4 per cent of sodium carbonate was used in this case.

Considering the results from the control plot, it is interesting to note
that 0.1 per cent of sodium carbonate produced no effect upon the
nitrification of a concentration of 0.1 per cent of dried blood, was dis¬
tinctly toxic to the nitrification of a concentration of 0.15 per cent of
ammonium sulphate, and markedly stimulating to the nitrification of
a concentration of 0.0625 per cent of ammonium sulphate. A concen¬
tration of 0.5 per cent of sodium carbonate was toxic in all cases.1

With a 1 per cent concentration of dried blood in the soil from the
manured plot, the addition of a concentration of 0.05 per cent of sodium
carbonate was quite toxic to nitrification, causing a reduction in the
yield of the nitrate from 172 to 31 p. p. m. A 0.1 per cent concentration
of sodium carbonate was still more toxic, while a concentration of 0.5
per cent totally inhibited nitrification. (The original soil contained
6.7 p. p. m.) With the use of 0.1 per cent dried blood no effects were
produced by any of the concentrations of sodium carbonate employed.
With the use of a concentration of 0.15 per cent of ammonium sulphate
the addition of 0.05 or 0.1 percent sodium carbonate retarded nitrifica¬
tion considerably, especially in the case of the latter, but not so markedly
as was found in the case of 1 per cent dried blood.

The addition of a concentration of 0.1 per cent of sodium sulphate was
without effect on the nitrification of 0.1 per cent dried blood or an equiv¬
alent amount of ammonium sulphate in the control plot, and upon the
nitrification of concentrations of 0.1 per cent of dried blood or 0.15 per
cent of ammonium sulphate in the manured plot ; but it produced marked
stimulation with 1 per cent dried blood in the manured plot and was
toxic with 0.15 per cent ammonium sulphate in the control plot. A con¬
centration of 0.5 per cent of sodium sulphate was somewhat toxic with 1

1 The chemistry of the action of sodium carbonate and other sodium salts on this soil will be discussed in
a subsequent paper. Suffice it to say that considerable light has been thrown on the above results from
a study of the pure chemistry involved.

428

Journal of Agricultural Research

Vol. VII, No. 10

per cent dried blood and 0.15 per cent ammonium sulphate; but it pro¬
duced no effects on the nitrification of a concentration of 0.1 per cent of
dried blood.

With one exception it is noteworthy that the effects produced by either
sodium carbonate or sodium sulphate were quite similar when equal
amounts of nitrogen in the form of dried blood and ammonium sulphate
were employed in low concentrations.

The above results are in harmony with those of Lipman (18) in showing
that sodium carbonate is extremely toxic to the nitrification of a high
concentration of dried blood and far more toxic than sodium sulphate.

NITRIFICATION DURING DIFFERENT LENGTHS OF TIME

In the preceding studies the samples were incubated for four weeks,
and, as already stated, conclusions on the relative rates of nitrification
of different materials and in different soils have frequently been drawn
from data obtained in this way. In the light of the preceding results it
becomes a matter of interest to study nitrate formation at different
intervals of time. In a preliminary study with the use of 1 per cent of
dried blood in soil from the control plot it was found that no nitrifi¬
cation took place during a period of 68 days. In another series with the
use of the same soil the incubation period was extended to 105 days, with
the same result. With still other soils in which a concentration of 1 per
cent of dried blood failed to be nitrified in four weeks, it has been found,
however, that active nitrification may set in later, and in some cases
eventually becomes quite as active as in soils which have the power of
nitrifying 1 per cent vigorously within four weeks.

For the purpose of studying nitrification progressively with low con¬
centrations of materials, 2,000 gm. of the fresh soil from the check and
manured plots were kept in large jars. Dried blood and ammonium
sulphate were added in quantities supplying 10 mgm. of actual nitrogen
per 100 gm. of soil. The mositure content was brought up to 15 per cent
and maintained near this point by the occasional addition of distilled
water as evaporation took place. The soils were incubated as before.
One-hundred-gm. portions were withdrawn at intervals and the nitrate
determined, as shown in Table VI.

Table VI. — Nitrification during different intervals of time

Nitric nitrogen (parts per million).

Soil.

Nitrogenous materials
added.

Origi¬

nal

soil.

After

4 •
days.

After

days.

After

days.

After

days.

After

days.

After

a9A

days.

Control plot . . .
Manured plot. .
Control plot. . .

None .

I. 2

4.0

7.0

7*4

9. O

7* 5

16. 5

. do . .

Dried blood .

4.9

I. 2

9.8
2. 2

10. 0
10. 4

12. 0

33* 5

13. 2
66. 0

14. 0
60. 0

23.O
72. 0

Manured plot. .
Control plot. . .

. do .

Ammonium sul¬

4*9

1. 2

10. 4
4-4

44- 0
16. 5

58. 0
37* 0

73*o
65. 0

64. 0
76. 0

64. 0
88.0

Manured plot. .

phate.

4-9

14.4

60. 0

72. 0

96. 0

96. O

Dec. 4, 1916

Nitrification in Semiarid Soils

429

The above data show that nitrification set in within the first four days
and continued in the untreated portions throughout the 94 days of the
experiment. When allowance is made for the nitrate originally present,
it will be seen that almost as active nitrification took place in the un¬
treated soil from the control plot as in that from the manured plot.
But with the addition of iomgm. of nitrogen in the form of dried blood,
the more active nitrification took place during the first nine days in the
soil from the manured plot. After this time nitrate formation took place
the more vigorously in the soil from the control plot.

Ammonium sulphate was most actively nitrified in the manured soil
during the first 31 days. After this time the rate in the control plot
exceeded that in the manured plot. At every time interval, with the
exception of the 15-day period, ammonium sulphate was found to undergo
more active nitrification than dried blood.

If the rates of nitrification in the two soils be compared on the basis of
the data obtained upon the ninth day,1 it would seem reasonable to con¬
clude that the manured plot is capable of supporting more active nitrifi¬
cation of either dried blood or ammonium sulphate than the check plot.
If a later period be chosen the inference seems equally reasonable that
the two soils are about equal in ability to nitrify dried blood. The data
obtained from the untreated portions, however, would seem to indicate
that the floras of the two soils are quite similar, so far as their ability to
produce nitrate is concerned. In a subsequent paper this point will be
more fully discussed.

The preceding data strongly emphasize the importance of studying
the formation of nitrates in laboratory studies during different intervals
of time and in the presence of varying concentrations of different nitroge¬
nous materials. Just as different concentrations of nitrogenous materials,
as already stated, may lead to widely different conclusions, the above
results show that almost any conclusion may be drawn regarding the
relative rates of nitrification of dried blood and ammonium sulphate in
a given soil or of dried blood in different soils, provided the incubation
periods be carefully chosen.

ACCUMULATION OF NITRITES IN LABORATORY EXPERIMENTS ON

NITRIFICATION

From the classical experiments of Winogradsky it is generally con¬
sidered that nitrification proper begins with ammonia and takes place in
two stages, each stage being brought about by a different set of bacteria.
The nitrite bacteria oxidize the ammonia to nitrous acid, and the nitrate
bacteria complete the oxidation to nitric acid. In field soils, however,
the activity of the latter is usually sufficiently great to complete the oxida¬
tion of nitrite almost, if not quite, as fast as it is fothied. Hence, nitrite
rarely accumulates in notable amounts in arable soils.

1 Such comparisons have previously been made (8, 27) upon the basis of a 10-day incubation period.

430

Journal of Agricultural Research

Vol. VII, No. 10

However, nitrite may accumulate to a considerable extent under poorly
aerated conditions, especially when artificial applications of nitrate fer¬
tilizers are made, but under such conditions it is highly probable that the
nitrite is formed in part at least through the reduction of nitrates rather
than from the incomplete oxidation of ammonia. The writer (12) has
shown, for example, that the application of sodium nitrate to rice soils
immediately preceding or during the time of submergence may result in
an accumulation of considerable amounts of nitrite. The addition of
large amounts of carbohydrates may also bring about a similar reduction
of nitrates, even under aerobic conditions. In general, it may be said
that the accumulation of notable amounts of nitrite in soils is an indica¬
tion of the existence of unfavorable soil conditions.

Notable amounts of nitrite have previously been found in laboratory
incubation experiments on nitrification. In experiments with the use of
asparagin Withers (35) found considerable amounts of nitrite in certain
soils in North Carolina, while only slight nitrate formation took place.
On the other hand, ammonium sulphate was oxidized to nitrate without
the accumulation of more than a trace of nitrite. In sterilized portions
of this soil, which were later exposed to reinoculation, notable amounts
of nitrite were formed within four weeks’ time from both asparagin and
ammonium sulphate, but practically no nitrate was formed from either.
The amounts of nitrogenous materials used in these experiments were
not stated.

Sackett (28) also found considerable amounts of nitrite in laboratory
experiments. He used 100 mgm. of actual nitrogen in the form of ammo¬
nium sulphate, ammonium chlorid, ammonium carbonate, and dried
blood per 100 gm. of soil, which corresponds, in the case of ammonium
sulphate, to a concentration of about 0.5 per cent, and in that of dried
blood approximately to 0.75 per cent. The incubation period was six
weeks. It is notable that in certain soils he found that nitrite formation
took place much more rapidly than nitrate formation, and in other soils
there was evidence of nitrite formation through the reduction of nitrate.
In the control portions to which no nitrogenous materials were added
the concentrations of nitrite did not amount to more than 1 to 2 p. p. m.
It is probable that the results obtained by Sackett would have been
greatly different had he employed a lower concentration of the nitroge¬
nous materials.

The writer (13) has also previously found considerable amounts of
nitrite in laboratory experiments with the use of excessive amounts of
magnesium carbonate. In these experiments a concentration of 2 per
cent of dried blood was employed with the light, sandy soil from Anaheim,
Cal. In discussing this point it was tentatively suggested that the
nitrite found arose from the reduction of nitrate, and that the magne¬
sium carbonate was more toxic to the nitrifying bacteria than to the
denitrifiers. In the light of evidence obtained more recently it seems

Dec. 4, 1916

Nitrification in Semiarid Soils

43i

more probable, however, that the magnesium carbonate in low concen¬
trations was toxic to the nitrate formers but not to the nitrite formers,
whereas in still higher concentrations it was toxic to both groups.

In the course of some studies on the effects of concentration on the
nitrification of ammonium carbonate it was observed that notable amounts
of nitrite began to accumulate as the concentration was increased above
15 mgm. of nitrogen per 100 gm. of soil; with a concentration of 30 mgm.
the nitrite content, after four weeks’ incubation, was found to be 268.5
p. p. m., while at the same time no nitrate was formed. From these
observations it would seem that the nitrate bacteria are more sensitive
to high concentrations of ammonium carbonate than the nitrite group,
as has been definitely shown to be the case by Boullanger and Massol (4).

The effects on nitrite accumulation brought about by the concentra¬
tion of different nitrogenous substances and in different soils, the effects
produced by the addition of alkali salts, organic matter, etc., have been
studied at varying intervals of time. The full data will be presented
in a later paper. Briefly, it may be stated that not more than a few
tenths p. p. m. of nitrite have been found where low concentrations of
nitrogenous materials have been used, but that as the conditions become
increasingly abnormal, either through the use of excessive amounts of
nitrogenous substances, the addition of alkali salts, or by other means, a
point is usually reached where nitrite formation proceeds more vigor¬
ously than nitrate formation, with a consequent accumulation of con¬
siderable amounts of nitrite. In addition, it has been found that under
certain conditions nitrite formation may proceed vigorously without
nitrate formation taking place at all, even in a soil where the nitrate
bacteria are present in abundance.

The occurrence of notable amounts of nitrite necessitates some depar¬
ture from the methods usually employed in the determination of nitrate
in soils. As already stated, the results obtained by the phenol-disulphonic-
acid method were found to agree closely with those obtained by the
aluminum-reduction method except where high concentrations of ni¬
trogenous materials had been employed. In such cases the reduction
method frequently gave much higher results.1 As shown below, the use
of the reduction method effects a conversion of nitrite into ammonia,
just as is the case with nitrates; and consequently the results found
represent the total of the nitrite and nitrate nitrogen present, although
it is recorded as nitrate.

If a solution containing nitrite is evaporated and the residue then
treated with the phenol-disulphonic-acid reagent, small amounts of nitrite
may also be converted into nitrate, thus introducing a slight error.

After considerable experimentation the method adopted for the deter¬
mination of nitrate in the presence of nitrite was as follows: The water

1 As shown by Allen (i), certain soluble organic forms of nitrogen also become reduced to ammonia under
the conditions employed in this method.

43^

Journal of Agricultural Research

Vol. VII, No. 10

solution of the soil was made up as usual. To an aliquot portion, io c. c.
of a i per cent solution of ammonium sulphate was added, then evap¬
orated on the water bath, and the determination completed as usual
with the phenol-disulphonic-acid method. During the evaporation the
ammonium sulphate brings about complete decomposition of the nitrite
through the formation of ammonium nitrite, which decomposes at the
temperature employed.1 The nitrite was determined by the Griess-
Ilosvay method.

The following data (Table VII) will show the wide range of results
obtained by the use of different methods :

Tabu® VII. — Effects of nitrite on the determination of nitrate (in parts per million )

Nitrite

nitrogen.

Nitrate
nitrogen by
modified
coloro-
metric
method.

Nitrate
nitrogen by
the usual
coloro-
metric
method.

Nitrate
nitrogen by
aluminum
reduction
method.

Soil plus i per cent of dried blood incu¬
bated for 46 days .

275

283

The above data show that nitrite becomes reduced to ammonia under
the conditions employed in the aluminum reduction method. It is
reasonable to infer, therefore, that unless allowance be made for the
nitrite present the results obtained by the reduction method from incu¬
bations with high concentrations of nitrogenous materials will represent
the total nitrite and nitrate present rather than the nitrate only (22).

It has been found that with the use of 1 per cent of dried blood nitrites
may accumulate in large amounts in soils of various types and that the
nitrite may persist without undergoing further oxidation, at least for
a period of 105 days, as shown in Table VIII.

Table VIII. — Accumulation of nitrite as affected by concentration (in parts per million)

Original

soil.

After

28 days.

After

42 days.

After

56 days.

After

71 days.

After
ios days.

Soil.

Ni¬

trite

nitro¬

gen.

Ni¬

trate

nitro¬

gen.

Ni¬

trite

nitro¬

gen.

Ni¬

trate

nitro¬

gen.

Ni¬

trite

nitro¬

gen.

Ni¬

trate

nitro¬

gen.

Ni¬

trite

nitro¬

gen.

Ni¬

trate

nitro¬

gen.

Ni¬

trite

nitro¬

gen.

Ni¬

trate

nitro¬

gen.

Ni¬

trite

nitro¬

gen.

Ni¬

trate

nitro¬

gen.

Control plot untreated .

1.4

12.0

11. 2

Control plot plus 0.0625 per

cent of dried blood .

1.4

Trace.

60.0

70.0'

Control plot plus 1 per

cent of dried blood .

1.4

19*5

87*5

26. 0

is©

iso

265

Manured plot untreated. . . .

3* 2

14.0

17. 0

Manured plot plus 0.0625

per cent of dried blood, . .

3*2

Trace.

So-o

60.0

Manured plot plus 1 per

cent of dried blood .

3*2

107

160.0

56*0

272.0

390

46s

416

1 This procedure was suggested by the method formerly employed for the determination of nitrate in
the presence of nitrite by Frankland (6).

Dec, 4, 1916

Nitrification in Semiarid Soils

433

The above shows that large amounts of nitrite may accumulate wh$n
high concentration of dried blood have been used, but with a low concen¬
tration such is not the case. It is noteworthy that high concentration
of other nitrogenous materials, such as bone meal, ammonium sulphate,
and ammonium carbonate, all have been found to promote the accumu¬
lation of large amounts of nitrite in incubation studies.

Alkali salts also exert marked effects upon the accumulation of nitrites
as shown by Table IX,

Table IX. — Effects of sodium carbonate on the accumulation of nitrite

After two weeks.

After four weeks.

. Nitrogenous material added.

Sodium

carbonate

added.

Nitrite

nitrogen.

Nitrate

nitrogen.

Nitrite

nitrogen.

Nitrate

nitrogen.

None . . . .

Per cent.

P. p. m.

P . p. m.

None.

l6. 2

48. O

0.1 per cent of dried blood .

. . .do _

84. O

1 10. 0

Do .

0. I

Trace.

86.0

100. 0

Do . * . t .

0.25

*5-o

19. 2

25.O

102. O

Do . .

0.0625 per cent of ammonium sul¬

0. 5

Trace.

2. O

phate .

None.

68.0

104. 0

Do .

0. 1

Trace.

106. 0

l6o. O

Do .

0. 25

62. 5

10. 0

93-8

79.0

Do . . . .

°-5

J* 7

i- 7

o-S

I.9

Not more than a mere trace of nitrite was found where a concentration
of 0.1 per cent of sodium carbonate had been added, and no effect was
produced on nitrate formation; but considerably greater amounts of
nitrite than nitrate occurred after .two weeks’ incubation where a con¬
centration of 0.25 per cent of sodium carbonate had been added. How¬
ever, after four weeks most of the nitrite formed from dried blood had
become oxidized to nitrate; while in the case of ammonium sulphate,
the concentration of nitrite after four wTeeks still exceeded that of nitrate.
The addition of 0.5 per cent of sodium carbonate entirely inhibited the
formation of either nitrite or nitrate.

It has been found that a concentration of from 30 to 40 mgm. of nitrogen
in the form of sodium nitrite per 100 gm. of soil entirely inhibits the for¬
mation of nitrate in the soil from the experimental plots. With lower
concentrations the nitrite was completely oxidized in four weeks’ time.

, These results taken in connection with the preceding indicate that when
abnormal soil conditions are brought about, the concentration of nitrite
produced in the oxidation of nitrogenous materials may become so high
as in itself to inhibit nitrate formation.

CONCLUSIONS

In the preceding investigations it has been shown that the amounts of
nitrate formed from dried blood, bone meal, or ammonium sulphate dur¬
ing four weeks’ incubation varied enormously when different concentra¬
tions were employed. This is true in regard both to the absolute amount
66846°— 16- - 2

434

Journal of Agricultural Research

Vol. VII, No. 10

of nitrate formed and the percentage of the nitrogen added that was
nitrified.

When i per cent of dried blood was used, the nitrifying activity was
found to be feeble or even negative in certain soils in which i per cent of
bone meal and 0.2 to 0.3 per cent of ammonium sulphate underwent
active nitrification, as was previously found by Lipman and Burgess
(24). However, when low concentrations of dried blood were employed,
such as are used in the field, active nitrification took place in every case;
and when equal amounts of actual nitrogen were added, it was found
that the yields of nitrate were quite similar, whether the nitrate had been
derived from dried blood, bone meal, or ammonium sulphate. High
concentrations of bone meal with a nitrogen content corresponding to
that furnished by 1 per cent of dried blood were also toxic to nitrifica¬
tion, very much as was the case with 1 per cent of dried blood.

Experiments were made with widely different types of soil from a
number of localities in southern California. It was found that the
inability to nitrify 1 per cent of dried blood is not confined to any one
type of soil nor to soils low in organic matter. The results as a whole
seem to warrant the conclusion, however, that the soils of southern
California in general are capable of supporting active nitrification of
dried blood, provided it be added in concentrations corresponding to
field practice.

It wTas found that the effects produced by the addition of alkali salts
varied greatly when different concentrations of nitrogenous materials
were employed. In a given soil a concentration of 0.05 per cent of
sodium carbonate was distinctly toxic to the nitrification of 1 per cent
of dried blood, while as high a concentration as 0.4 per cent produced
no effects on the nitrification of 0.1 per cent of dried blood. Likewise,
0.1 per cent of sodium carbonate was toxic to the nitrification of 0.15
per cent of ammonium sulphate, and markedly stimulating when 0.0625
per cent of ammonium sulphate was used. Similar statements may be
made with regard to the effects of sodium sulphate.

The results also show that widely different conclusions may be drawn
from laboratory experiments when different periods of incubation are used.

Nitrites were found to accumulate in large amounts where excessive
amounts of nitrogenous materials were employed. In some cases
the nitrite content greatly exceeded the nitrate content after an in¬
cubation period of several weeks. Likewise, the addition of alkali salts
may suppress nitrate formation, while at the same time permitting nitrite
formation to proceed actively.

It is necessary to make allowance for the nitrite present in the de¬
termination of nitrate by the aluminum reduction or phenol-disulphonic-
acid methods, but the error introduced by nitrite is far greater with the
former method than with the latter.

It is not intended to give the impression from the above that all
nitrogenous materials will undergo nitrification at equal rates when

Dec. 4, 1916

Nitrification in Semiarid Soils

435

present in low concentrations. It is recognized that different organic
fertilizers undergo biochemical decomposition in varying degrees in a
given soil, and this for a number of reasons, some of which will be dis¬
cussed in a later paper. The writer holds, however, that the methods
now employed by many students of nitrification, in which high con¬
centrations of nitrogenous materials are added and the nitrate determined
at a fixed interval of time, are not only unsatisfactory but that the re¬
sults thus obtained are likely to be more misleading than informing.

In the light of the results obtained in this investigation, it seems highly
probable that at least some of the peculiarities that have been noted in
previous nitrification studies will be found to disappear under the more
rational procedure of studying the activities of the organisms in an
environment as nearly similar to that of the field as possible. The writer,
while criticizing the methods in common use, freely admits that some of
the conclusions previously drawn by him from studies with Hawaiian
soils are open to serious question because of the methods employed (14).

The nitrate merely represents one of the end products formed ; and in
the case of an organic substance the intermediate products that are pro¬
duced may, either directly or indirectly through the effect upon other
organisms, exert much influence upon the oxidation of ammonia. In
the presence of large amounts of materials it is highly probable that the
relations of the different groups of organisms present become greatly
changed, with a consequent effect on the oxidizing activity of the nitri¬
fying organisms.

It is, of course, a matter of scientific interest that certain soils are
capable of supporting active nitrification of 1 per cent of dried blood,
while others are not, and the reasons underlying these differences are
matters deserving further study, but so far as the practical side of
nitrification is concerned the writer holds that laboratory experiments
should be conducted under conditions as nearly comparable with those
that obtain in the field as possible, and that at the present time nothing
more than scientific interest can safely be attached to the results obtained
with the use of such abnormally high concentrations of nitrogenous
materials as are commonly used in laboratory experiments on this subject.

Many American bacteriologists have apparently accepted the conclu¬
sions of Stevens and Withers (29, 30) and have multiplied laboratory
tests in a conventional way without seriously questioning the method.
The result has been that the practical aspects of nitrification studies have
become extremely empirical. While it has frequently been stated that
nitrification in the laboratory is not strictly comparable with that in
the field, the conditions obtaining in the nitrification of 1 per cent dried
blood in the laboratory have been referred to as optimum conditions
(27). The preceding data indicate, however, that such may be far from
the case. The writer thoroughly agrees with the position taken by
Lohnis and Green (2 5) and Allen and Bonazzi (2) in their recent discus¬
sions of this subject.

436

Journal of Agricultural Research

Vol. VII, No. 10

LITERATURE CITED

(1) Aluen, E. R.

1915. The determination of nitric nitrogen in soils. In Jour. Indus, and Engin.
Chem., v. 7, no. 6, p. 521-529, 1 fig.

(2) - and Bonazzi, A.

1915. On nitrification. Ohio Agr. Exp. Sta. Tech. Ser., Bui. 7, 42 p. Biblio¬

graphy, p. 41-42.

(3) Beckwith, T. D., Vass, A. F., and Robinson, R. H.

1914. Ammonification and nitrification studies of certain types of Oregon soils.
Oreg. Agr. Exp. Sta. Bui. 118, 40 p., 30 fig. References, p. 40.

(4) Bouluanger, E., and Massoe, L.

1904. Etudes sur les microbes nitrificateurs. (Deuxi&me m6moire.) Ann.
Inst. Pasteur, t. 18, no. 3, p. 181-196.

(5) Burgess, P. S.

1913. The aluminum reduction method as applied to the determination of
nitrates in “alkali*' soils. In Univ. Cal. Pub. Agr. Sci., v. 1, no. 4,
p. 51-62, illus.

(6) Frankuand, P. F.

1888. A gasometric method of determining nitrous acid. In Jour. Chem. Soc.
[London], v. 53, p. 364-373, 1 fig.

(7) Greaves, J. E.

1913. The influence of arsenic upon the biological transformation of nitrogen in

soils. In Biochem. Bui., v. 3, no. 9, p. 2-16.

(8) — -

1914. A study of the bacterial activities of virgin and cultivated soils. In

Centbl. Bakt. [etc.], Abt. 2, Bd. 41, No. 11/17, p. 444-459.

(9) —

1916. Stimulating influence of arsenic upon the nitrogen-fixing organisms of the

soil. In Jour. Agr. Research, v. 6, no, 11, p. 389-416, 5 fig. Litera¬
ture cited, p. 414-416.

(10) Hilgard, E. W.

1906. Soils . . . 593 p.,89 fig. New York.

(11) Keluerman, K. F., and Wright, R. C.

1914. Relation of bacterial transformations of soil nitrogen to nutrition of

citrous plants. In Jour. Agr. Research, v. 2, no. 2, p. 101-113, 7 fig.
, Literature cited,, p. 113.

(12) Keuuey, W. P.

1911. The assimilation of nitrogen by rice. Hawaii Agr. Exp. Bui. 24, 20 p.

(13) —

1912. The effects of calcium and magnesium carbonates on some biological

transformations of nitrogen in soils. In Univ. Cal. Pub. Agr. Sci.,
v. 1, no. 3, p. 39-49*

(14) -

1915. Ammonification and nitrification in Hawaiian soils. Hawaii Agr. Exp.

Sta. Bui. 37, 52 p.

(15) -

1916. Some suggestions on methods for the study of nitrification. In Science,

n. s., v. 43, no. 1097, p. 30-33.

(16) Lipman, C. B.

1911. Toxic effects of “ alkali salts’ ' in soils on soil bacteria. I. Ammonifica¬

tion. In Centbl*. Bakt. [etc.], Abt. 2, Bd. 32, No. 1/2, p. 58-64, 1 fig.

(17) -

1912. The distribution and activities of bacteria in soils of the arid regions.

In Univ. Cal. Pub. Agr. Sci., v. 1, no. 1, p. 1-20.

(18) — -

1912 . Toxic effects of * ‘alkali salts ' ’ in soils on soil bacteria. 1 1 . Nitrification.
In Centbl. Bakt. [etc.], Abt. 2, Bd. 33, No. 11/14, p. 305-313, 2 fig.

Dec. 4, 1916

Nitrification in Semiarid Soils

437

(19) Lipman, C. B.

1913. Antagonism between anions as affecting ammonification in soils. In

Centbl. Bakt. [etc.], Abt. 2, Bd. 36, No. 15/18, p. 382-394, 3 tig.

(20) -

1914. The poor nitrifying power of soils a possible cause of “die back” (exan¬

thema) in lemons. In Science, n. s., v. 39, No. you, p. 728-730.

(21) - —

1915. The nitrogen problem in arid soils. In Proc. Nat. Acad. Sci., v. i,

no. 9, p. 477-480.

(22) - and Burgess, P. S.

1914. Antagonism between anions as affecting soil bacteria. II. Nitrifica¬
tion. In Centbl. feakt. [etc.], Abt. 2, Bd. 41, no. n/17, p. 43°”444» 6

fig*

(23) -

1914. The effect of copper, zinc, iron, and lead salts on ammonification and

nitrification in soils. In Univ. Cal. Pub. Agr. Sci., v. 1, no. 6, p.
127-139.

(24) - -

1915. The determination of availability of nitrogenous fertilizers in various

California soil types by their nitrifiability. Cal. Agr. Exp. Sta. Bui.
260, p. 107-127.

(25) EOhnis, Feux, and GrEEn, H. H.

1914. Methods in soil bacteriology. VII. Ammonification and nitrification
in soil and soil solution. In Centbl. Bakt. [etc.], Abt. 2, Bd. 40,
No. 19/21, p. 457-479*

(26) Loughridge, R. H.

1914. Humus and humus-nitrogen in California soil columns. In Univ. Cal.
Pub. Agr. Sci., v. 1, no. 8, p. 173-274. '

(27) McBeth, I. G.» and Smith, N. R.

1914. The influence of irrigation and crop production on soil nitrification. In
Centbl. Bakt, [etc.], Abt. 2, Bd. 40, No. 1/8, p. 24-51, 6 fig.

(28) Sackett, W. G.

1914. The nitrifying efficiency of certain Colorado soils. Col. Agr. Exp. Sta.
Bui. 193, 38 p.

(29) Stevens, F. L. and Withers, W. A.

1909. Studies in soil bacteriology. I. Nitrification in soils and in solutions.
In Centbl. Bakt. [etc.], Abt. 2, Bd. 23, No. 10/13, P* 355“373-

(30) -

1909. Studies in soil bacteriology. III. Concerning methods for determina¬
tion of nitrifying and ammonifying powers. In Centbl. Bakt. [etc.],
Abt. 2, Bd. 25, No. 1/4, p. 64-80.

(31) Stewart, Robert.

1913. The intensity of nitrification in arid soils. In Centbl. Bakt. [etc.],

Abt. 2, Bd. 36, No. 15/18, p. 477-490.

(32) - and Greaves, J. E.

1911. The movement of nitric nitrogen in soil and its relation to “nitrogen

fixation.” Utah Agr. Exp. Sta. Bui. 114, p. 181-194.

(33) -

1912. The production and movement of nitric nitrogen in soil. In Centbl.

Bakt. [etc.], Abt. 2, Bd. 34, No. 4/7, p. iis-i47> * fig* ,

(34) - - — } and Paterson, William.

1914. The origin of the “nitre spots” in certain western soils. In Jour. Amer.

Soc. Agron. v. 6, no. 6, p. 241-248.

(35) Withers, W. A.

1907. Report of the chemical division. [Studies on the changes which take
place in nitrogen when added to soils in different forms of combina¬
tion.] In N. C. Agr. Exp. Sta. 29th Ann. Rpt. [i905]/o6, p. 15-18.

FACTORS AFFECTING THE EVAPORATION OF
MOISTURE FROM THE SOIL 1

By F. S. Harris, Director and Agronomist, and J. S. Robinson, Fellow in Agronomy,
Utah Agricultural Experiment Station

INTRODUCTION

The importance of soil moisture in crop production is well understood.
No plant can grow unless moisture is present to help make food available
and furnish the water necessary to carry on regular plant functions. In
arid regions the growth of crops is limited more by a lack of moisture
than by any other factor, and even in regions of high rainfall crop yields
are often materially reduced by droughts.

In sections where only a small amount of rain falls, practically all
that sinks into the soil returns to the surface and is evaporated directly
or passes through plants, from which it is evaporated. In humid re¬
gions also, where some of the soil moisture percolates to great depths,
there is considerable loss by evaporation from the surface of the soil.

Moisture evaporated from the soil is completely lost and is of no
value to crops; hence, it is important to reduce evaporation to a mini¬
mum, particularly where the supply of moisture is limited. The best
condition would be to have no evaporation of moisture except that pass¬
ing through the plant and assisting in its functions. Any information,
therefore, that will lead to a better understanding of the factors involved
in evaporation and a fuller knowledge of methods of controlling these
factors will be of considerable practical importance as well as scientific
interest.

Surface losses are due to two factors: (i) Capillarity, by which the
moisture is brought to the surface, and (2) evaporation. In many of the
soil-moisture studies that have been made these two factors have not
been clearly separated, but have been considered together in determining
loss. This has led to considerable confusion, since a difference in two
losses might, in one case, be due to a difference in the rate at which
moisture was supplied to the surface by capillarity and, in another case,
to the evaporation factors.

In the experiments reported in this paper an attempt has been made
to eliminate the factor of capillarity and to confine the studies entirely
to evaporation in order to determine, as nearly as possible, the effect of
a number of the factors involved in the evaporation of moisture from
the soil.

iThe writers wish to acknowledge their indebtedness to Messrs. George Stewart and N. I. Butt, of
the Utah Experiment Station, for their assistance with certain experiments and in preparing this paper
for publication.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.
gj

Vol. VII, No. 10
Dec. 4. 1916
Utah— 4

(439)

440

Journal of Agricultural Research

Vol. VII, No. io

REVIEW OF THE LITERATURE

Considerable work has been done on various phases of the evaporation
problem. One of the important factors influencing water losses is the
wetness of the soil, or the initial percentage.

Widtsoe (19, p. 35, Table 20)1 gives the results of four years’ experi¬
mental work, as shown in Table I.

Tabl3 I. — Total evaporation of water from bare College loam

Moisture.

Loss in pounds per square foot.

1902

1903

1904

1905

Average.

► Per cent.

10 .

. 9
23
51

I c .

20 .

Loss increased with the percentage of water up to 20 per cent, which
was the highest degree of wetness used. He (19) says: “The evaporation
of water from bare soils increased with the increased saturation of the
soil. The increase in loss was usually much larger than the increase in
saturation. In another treatise (20) he found that “the wetter the soil
at the surface, the more rapid is the evaporation of water from it.”

This is confirmed by Whitney and Cameron (18) and by Fortier (5).
In Whitney and Cameron’s work, 26 per cent was the highest humidity,
that being used in but one set of experiments. Fortier’s highest per¬
centage of water was 17.5. In his conclusions he says, “The rate of
evaporation from soils varies directly with the amount of moisture in the
top layer.”

The work of Cameron and Gallagher (3, p. 45-49) indicates that, after a
certain wetness is reached, there is little if any increase in water loss. In
their work the soils of different degrees of wetness were placed over sul¬
phuric acid of different concentrations in order to control the humidity.
Over 95 per cent sulphuric acid in desiccators, Podunk fine sandy loam
gave the most rapid loss by evaporation up to 4 per cent moisture, with a
very slight increase in loss up to 28 per cent. Miami black clay loam
lost most rapidly up to 22 per cent, with a small increase to 41 per cent,
where the loss was highest.

The color of the soil is claimed by King (10) to affect evaporation
greatly, since the darker the soil the more heat it absorbs and radiates.
He found that the rise in temperature, due to the darker color, is the im¬
portant factor.

Concerning winds, King (9, p. 16) shows that up to 300 feet from woods
the loss gradually increased with the distance. He (1 1) gives the following

1 Reference is made by number to "Literature cited,” p. 460-461.

Dec. 4, 1916 Evaporation of Moisture from the Soil 441

data on evaporation for an hour from a wet soil with a surface of 27
square inches:

20 feet from hedgerow the evaporation was 10.3 c. c.

150 feet from hedgerow the evaporation was 12.5 c. c.

300 feet from hedgerow the evaporation was 13.4 c. c.

At 300 feet the evaporation was 30 per cent greater than at 20 feet and
6.7 per cent greater than at 150 feet, due largely to the vessels closer to
the hedge having protection from the wind.

McDonald (12) gives this terse summary:

Evaporation depends upon the temperature of the evaporating surface, the dryness
of the air, and the velocity of the wind. The hotter the day, the greater the evapora¬
tion; the drier, the greater the evaporation — the ceaseless sucking up of moisture.

Bowie (1) claims that loss due to wind is caused by the more intimate
contact of the air with the moist soil surface. He says:

With average wind velocities of from 2.4 to 4.0 miles an hour, and with an aver¬
age water temperature of 70 F.,the increased evaporation rate due to wind was
about 0.5 per cent a day for each mile of wind.

Payne (13), in giving the advantage of windbreaks in retarding
evaporation, shows that a sod wall 4 feet high and 20 feet long, running
east and west, reduces the loss from buckets placed in the ground level
with the field. On the north side, buckets 1, 3, 5, 7, and 10 rods dis¬
tant lost in 62 days moisture the equivalent of 677, 633, 700, 703, and
712 tons to the acre, respectively; on the south side, buckets 1, 3, 5, 7,
and 8 rods distant gave losses of 647, 686, 738, 764, and 761 tons to the
acre, respectively.

In work done by Principi (14) the conclusions reached were —

That evaporation is most rapid from the materials which have the largest pore
Spaces, and that it remains almost the same whether it arises from a free water sur¬
face or from thin films covering the particles of wet material.

Woolny (21) claims that capillarity ceases when the diameter of the
particles is more than 2 mm. and that it varies with smaller particles
in proportion to their fineness — the finer the particles, the greater the
lifting power, but the slower the movement.

In regard to moisture movement through a column of dry soil not in
direct contact with the moisture but with a saturated atmosphere be¬
tween, Buckingham (2, p. 9-18) says that moisture escapes probably by
pure diffusion and that the loss in this way is proportional to the square
of the porosity, following the same law as the diffusion of air and carbon-
dioxid gas through soils.

Whitney and Cameron (18) showed that loams over a saturated
atmosphere gave a greater diffusion of moisture through the pore spaces
than through those of clay.

442

Journal of Agricultural Research

Vol. VII, No, io

Fortier (5) shows, in the measurement of losses from water surfaces
kept at different temperature, that reducing the temperature from 88°
to 80.4° F. lessened evaporation 20 per cent; reducing it to 73. 50 de¬
creased evaporation 40 per cent; to 61. 30 decreased it to 67 per cent; and
to 53. 40 reduced it to 85 per cent.

Shade, which is a great protection from heat, aids in preventing
evaporation. Seelhorst (16) shows that a loam soil shaded by dry rye
plants evaporated 13.9 per cent less of the rainfall than the same kind of
soil unshaded.

Widtsoe’s (20) work on the influence of shade on evaporation gives a
loss of 22 pounds to the square foot as against 32 pounds for the un¬
shaded portion; that is, there was 29 per dbnt greater loss in sunshine
than in shade.

Fortier (6) shows that in actual experiments on Mount Whitney evapo¬
ration decreased with altitude. The decrease was rather regular, except
at the summit (14,502 feet), where greater evaporation took place than
at either 10,000 or 12,000 feet.

According to Carpenter (4), diminished barometric pressure tends to
increase evaporation to the extent of 14 per cent at 9,000 feet and 18
per cent at 10,000 feet over that at 5,000 feet.

Mulching the surface of the soil a few inches by stirring it is the most
common practice in use for the preservation of moisture under field
conditions. Ridgaway (15) shows that stirring the surface to depths
of 2, 4, and 6 inches in different tanks once a week, with the water level
kept 22 inches below the surface of the soil, lessened evaporation by 19,
23, and 45 per cent, respectively, of the amount lost from unstirred
soil. His work also shows that where water is maintained at 6, 12, 18,
and 22 inches below the surface, the losses were 95, 70, 45, and 35 per cent
of evaporation from a free- water surface. This bears out Wollny’s (21)
statement that “if there is water underneath the soil the evaporation
decreases as the distance between the surface of this water and that of
the soil increases.”

Fortier (7) sums up the results when dry soil coverings were substi¬
tuted for stirred surfaces, as shown in Table II.

Table; II. — Losses by evaporation from soil surfaces variously treated

Locality.

Loss from
unmulched
soil.

Loss from
3-inch
mulch on
soil.

Loss from
6-inch
mulch on
soil.

Loss from
9-inch
mulch on
soil.

Number
of days
run.

•

Per cent.

Per cent .

Per cent.

Davis, Cal. .

35-oo

14. 71

5- 94

O. 78

Wenatchee, Wash . .

14- 33

3- 98

2. 10

I. 06

Reno, Nev . .

20. 39

8. 26

2. 74

1. 96

Average .

23. 24

8. 98

3- 59

I. 27

24. 7

Dec. 4, 1916

Evaporation of Moisture from the Soil

443

Using enough water to cover the surface with 3.14 inches, this author
(5) found that no mulch gave a 0.72-inch loss; a 4-inch mulch, 0.21
inch; an 8-inch mulch, 0.1 inch; and a 10-inch mulch, 0.03 inch in 14
days— that is, the various mulches saved in 14 days 16.24, 19.75, and
21.97 per cent of the amount applied. #

The advantage of deep furrows as a saving in furrow irrigation is
brought out by Fortier (5). When the same quantities of water were
applied to tanks in furrows 3, 6, 9, and 12 inches deep, he found that the
losses at the end of 10 days were 25, 18, 10.2, and 6 per cent of the
total water applied. The loss was most rapid for the first 2 days, and at
the end of 5 days 77 per cent of the total loss for 10 days had occurred.

Widtsoe’s (19) work shows that in the two treatments, sand contain¬
ing 15 per cent, sandy loam containing 20 per cent, and Sanpete clay
containing 25 per cent moisture, the average loss from the soil receiving
surface irrigation was more than three times as great in the same length
of time as that from a subirrigated soil. Fortier (5) shows that soil
subirrigated 2 feet underground by pipes lost only 25 per cent as much
water in 20 days as when irrigated by surface flooding.

According to Stigell (17), bacterial growth retards evaporation. He
says that this is attributed to utilization of moisture by the organisms
in their metabolic products, and reduction of the porosity of the medium
by the metabolic products of the organisms. Hoffman's (8) work shows
that after the bacteria in various culture media were added to the soil,
evaporation was increased except in the case of manure. In drawing
his conclusions he leads one to believe that if the experiment is of long
duration the results may be reversed, due to the accumulation of carbon
dioxid from the organism being taken up by the soil moisture, thereby
increasing the surface tension of the water. Gelatin was found to retard
evaporation and for that reason could not be used as a culture media.

EXPERIMENTAL WORK
INITIAL QUANTITY OF SOIL MOISTURE

To find, if possible, the specific influence of varying percentages of water
in soils, a series of experiments was conducted at the Utah Agricultural
Experiment Station in the years 1912 to 1916, inclusive. The pre¬
liminary tests show that evaporation losses increased rapidly with the
increased wetness of the soil, or with the initial quantity. The latter
studies were so arranged as to locate as many points of variation in the
losses as possible.

Preliminary Study

Late in June, 1913, a study of the effect of initial quantity of moisture
in the soil was begun. One hundred gm. of dry Greenville loam were
put in small, weighed tin plates about 5 inches in diameter, and puddled

444

Journal of Agricultural Research

Vol. VII, No. 10

with water. The soil was then set aside to dry. When the condition
of dryness was approached, the moisture was made up to the desired
content. Twelve percentages— 5, 10, 12^, 15, 20, 25, 30, 35, 40,

45, and 50 — were run in triplicate. The pans were set on a table in a
large laboratory room. Weighings were taken daily to the nearest
one-tenth of 1 gm. and water was added to make up the loss. This
experiment continued for 42 days, at the end of which time the losses
were computed for each week and for the entire period. Table III gives
the losses for each period. *

Table III. — Effect of initial quantity of water on evaporation

Soil mois¬
ture.

Water evap- j
orated. ]

Soil mois¬
ture.

Water evap¬
orated.

Per cent .

5 ,

7 X
10

Gm.

96.4
229. 2

3^3* °
484. 8

594 °
822. 2

Per cent.

5°

Gm.

956-1
h 013- 3

X, 048. I
1,074. 5

1, 148. 1

1, 185. 8

-An examination of Table III shows that evaporation increased rapidly
to about 20 per cent, less rapidly from 20 to 30 per cent, and slowly
from 30 to 50 per cent.

Later Experiments

Various later studies were made with loam, sand, clay, and muck. .
The methods were similar to those used in the preliminary work, but
the experiments were much more detailed and thorough. Most of the
work was done in small tin plates and copper vessels 8 inches in diam¬
eter and 4 inches deep, though some tests were made in Petri dishes,
some in long galvanized-iron tanks and some in deep cylindrical gal¬
vanized cans.

LOAM

The test with loam was conducted in the manner already described
with 100 gm. of dry soil in tin plates. The percentages of moisture ran
from hygroscopic water, which was about 1.8 per cent, to 50 per cent.
The experiment was conducted for 13 days, from August 29 to Septem¬
ber 12, the pans being made up to the original weight each day and
shifted on the table to eliminate, in part, the influence of air currents.

A similar test was made during the winter in a steam-heated laboratory.

. All percentages from 1 to 40 were run in triplicate with 1 per cent inter¬
vals. Weighings were made each day and losses were considerably
greater than those for the same number of days in the other trial. A
possible explanation for this is that the laboratory under artificial heat

Dec. 4,1916

Evaporation of Moisture from the Soil

445

was slightly warmer and changed air much more frequently than in
summer. The hygrometer showed somewhat lower humidity, but it
could not show the influence of thorough ventilation.

The next study was made between June 4 and June 17, 1914. It was
conducted in the same fashion, except that weighings were made twice
daily, between 6 and 8 o'clock morning and evening. Four pans, instead
of three, were used with each percentage from 1 to 50. The pans for
1 per cent gained up to about 1 .8 per cent and then neither lost nor gained
noticeably.

In addition to the pans containing wet soil, a pan of free water was
exposed at each corner of the table on which the test was made. The
two pans on the east and nearer the doors lost somewhat more than the
two at the west end. All soils that approached saturation lost more
than the free water. The losses from the 20 per cent pans were about
equal to the average of
those from the free¬
water pans.

An almost identical
test was made in July
which practically dupli¬
cated the former results.

As 12 weighings were
made, the length of this
study was equivalent to
6 days.

These various experi¬
ments are combined in
figure 1, which gives
results for a total of 51
days with moisture up
to 40 per cent, and 36
days with moisture up to 50 per cent. The curve for the 36-day results
shows a rapid increase in evaporation with a higher initial percentage
of moisture in the soil up to 7 or 8 per cent; then a less rapid increase
up to 18 or 20 per cent, from which point the increase is small. In the
tests continued for 51 days, the same general changes in the curve are
noted, although the total loss is decidedly more than for the 36-day test.

In some of the trials the free water lost more by evaporation than any
of the wet soils, but usually there was a greater loss from soil which was
completely saturated than from the free-water surface.

In each of these tests with Greenville loam, there seemed to be a
number of more or less definite breaking points in the curves of loss.
These indicate critical points where the moisture relations of the soil
made rather sudden changes. A great deal of work will need to be done

Fig.

Initial Per Cent Moisture In Greenville L 01m

1. — Evaporation from Greenville loam containing different
initial percentages of moisture.

446

Journal of Agricultural Research

Vol. VII, No. io

under favorable conditions before these exact points of change can be
determined.

In the summer of 1912 nine large galvanized-iron pans, or tanks,
5H feet long by 1 foot wide by 3 inches deep were employed to give
larger surfaces and deeper soil. The equivalent of 10 kgm. of dry Green¬
ville loam was put into each tank and puddled with excess water to
firm the soil and smooth its surface. Enough water in addition to the'
water already present — 5.4 per cent — was added to make a series with
5 per cent intervals from 5.4 to 35.4 per cent. These tanks were set on
the floor of the laboratory and were weighed on Monday, Wednesday,
and Friday of each week. The losses were made up at the time of
weighing. This study ran from June 12 to July 15, a period of 34 days.

The next year seven tanks were set up and run during the 81 -day
period from June 18 to September 6. This time 7.2 kgm. of Greenville
loam, containing 2.6 per cent moisture, were used. Intervals of 5 per
cent were again made, bringing the moisture content up to 7.6, 12.6,
etc., as high as 37.6 per cent. Weighings were taken three times a
week as in the first trial and water added to make up the evaporation
loss. Table IV contains the results of the two trials.

Table IV. Effect of initial percentage of soil moisture on evaporation from Greenville
loam in galvanized-iron tanks feet long and I foot wide

1912

1913

Average of 2 tests.

Moisture.

Loss in 34
days.

Moisture.

Loss in 34
days.

Average

moisture.

Loss in iis
days.

Per cent.

Gm.

Per cent.

Gm.

Per cent.

Gm.

5* 4

3, 744

7. 6

10, 710

6* 5

14, 454

10. 4

7,575

9.518

12. 6

17, 530

n* 5

25, 105

IS* 4

17. 6

19,275

16. 5

28, 793

20. 4

9, 200

22. 6

20, 395

21. 5

29, 595

25*4

10, 128

27. 6

20, 300
19, 870

26. 5

30, 428

30* 4

10, 340

3 2. 6

31* 5

30,210

35*4

11,045

37* 5

19, 990

36. 5

31,035

Table IV shows that the loss was rapid to 16.5 per cent and then slow.
Of course, the wide intervals prevented locating exact points, but in
general this test corroborates rather closely those already reported for
loam.

SAND, CLAY, AND MUCK

During the period from December 26, 1913, to January 3, 1914, similar
tests were made with sand, clay, and muck. One hundred gm., dry weight,
of sand and clay and 50 gm. of dry muck were used in the same kind of
tin plates as those used in the other trials. All percentages were run in
triplicate, the sand for each percentage from 1 to 33, the clay for each
percentage from 1 to 55, and the muck for each 20 per cent interval from

Dec. 4, 1916

Evaporation of Moisture from the Soil

447

20 to 240 per cent. The muck consisted almost entirely of vegetable
mold gathered from accumulated pockets in the brush swales of Logan
River. It was known to have high water-holding capacity from previous
experiments.

The pans were kept several inches back from the edge of the tables on
which they rested. This almost entirely eliminated the effect of air
currents which had, in previous tests, caused some variations in the rows
of pans set close to the edge. The loss was made up each day after the
weighing. The winter losses were much higher in this case than with
the loam previously reported.

Figure 2 gives the results for sand and shows an increase in evaporation
as the initial moisture is increased, up to 33 per cent. The most rapid
increase is up to about
7 per cent. Above this
point the increase in
loss is not so great.

Figures 3 and 4 give
the curves for clay and
muck. The results for
clay show a more grad¬
ual ascent in the curve,
and a higher point
before there is any
break, than do the re¬
sults for sand. This
might have been ex¬
pected from the great
water-holding power
of the clay. Muck with high percentages and differences at greater
intervals produced about the same kind of curve as did loam.

More tests are necessary, however, to establish points as nearly exact
as was done in the case of loam.

HUMIDITY OF THE AIR

To study the effect of a saturated atmosphere on evaporation, a set of
wet soils in Petri dishes was placed in an air-tight copper germinator.
Forty gm. of dry soil were put in each of these dishes, which were about
3 inches in diameter. Duplicate vessels were made up with soil for each
1 per cent from 1 to 35, and for each 2 per cent from 35 to 45 per cent of
moisture. They were then set on the shelves of the germinator, which
is a hollow box 22 inches wide, 37X inches deep, and 47^ inches long.
The whole was surrounded with a water jacket, except at the doors,
which were made of two panes of glass inclosing a dead-air space about
half an inch thick. The perforated copper shelves were covered with

Fig. 2. — Evaporation from sand containing different initial per¬
centages of moisture.

448

Journal of Agricultural Research

Vol. VII, No. io

Initi&l Percent Moisture In Cl&y

Fig. 3.— Evaporation from clay containing different initial per¬
centages of moisture.

cheesecloth and connected by a perpendicular wick to a water container
on top. The wick kept all the cheesecloth covers saturated. To make
sure no vapor escaped at the door, pieces of cloth kept constantly drip¬
ping were hung in front of the door. Tested thermometers showed

throughout the experi¬
ment an almost constant
temperature of i9.5°C.,
about one degree lower
than that of the labora¬
tory. Each day for 10
days the vessels were
weighed to 0.01 gm.
and made up by
adding the water lost.
The lower percent¬
ages, however, gained
and were left at the
wetness reached, which
was about 7 per cent.
Fairly comparable
with this test was another in which the same vessels, made up as before,
were surrounded with cheesecloth 24 inches high to prevent drafts and
to maintain a uniform temperature. Possible variations were elimi¬
nated by shifting the vessels in such a way that each occupied all
sections of the cloth
box some time during
the 20 days of the
experiment. The top
was left open to permit
vapor to escape upward.

The relative humidity
of the air was almost
constantly at 76 per
cent, whereas it was
about 68 in the open
laboratory. The tem¬
perature averaged 1 90
C. during the 20 days.

Completely dried
soil gained to 4 per
cent, then remained constant. Weighings were made each day to the
nearest 0.01 gm. and the losses made up as with the test in the germi-
nator.

Figure 5 shows the losses for each 1 per cent in the saturated atmosphere
of the germinator and for the slightly overraoist air of the cloth boxes.

40 €0 to too tzo 140 160

Initi&l Per Cent Moisture InMucfe

Fig. 4.— Evaporation from muck containing different initial per¬
centages of moisture.

Dec. 4, 1916

Evaporation of Moisture from the Soil

449

Since the latter ran 20 days, the total losses have been divided by 2 in
order to get the total loss for 10 days to make the figures comparable.
r There is a decided difference in the results of the two experiments.
The losses in the cloth boxes were nearly 20 times as great as those in the
germinator, showing the enormous retarding effect of high humidity.
Moreover, the open laboratory was somewhat dri$r and about one degree
warmer than the air in the cloth boxes. No test was made in the open
laboratory, but the losses would have been somewhat higher, since both
a drier air and a higher temperature prevailed. This would, of course,
further accentuate the already enormous differences.

WIND VELOCITY

As was suggested in the literature on wind velocity, King (9) has shown
that vessels 20 feet from a windbreak lost 30 per cent less moisture by
evaporation than ves¬
sels 30 feet distant.

This was because the
outer vessels were
more exposed to air
movement. Payne
(13) shows nearly
similar results, while
Bowie (1) reports
that with light winds
and normal temper¬
ature there is an
evaporation loss of
0.5 per cent for each
mile of wind.

In order to deter¬
mine the effect of wind velocity on the rate of evaporation, a series of
alleys 8% inches wide and 70 inches long were arranged in such a way
that the air could be made to pass through them at different velocities.
The alleys were separated by oilcloth partitions 2 feet high; the air
currents were made by electric fans placed in such positions that the
desired velocities could be obtained. The velocity of air in each alley ■
was measured with an anemometer placed at the end of the alley away
from the fans.

The soils were contained in copper evaporimeters 6 inches in diameter.
The loss each day was made up by adding water through a tube entering
the evaporimeter at one side of the soil. In each alley there were five
evaporimeters — one containing distilled water, one Greenville loam, and
one each of quartz sand 0.25 mm., 0.5 tnm., and 0.8 mm. in diameter.

Two tests were made in the experiment, one running continuously for
16 days with three wind velocities and another running continuously for
66846° — 16- - 3

Fig. 5.— Loss of moisture from Petri dishes containing different per¬
centages of soil moisture and kept in a saturated and unsaturated
atmosphere.

450

Journal of Agricultural Research

Vol. VII, No. io

20 days with seven velocities. In each case there was an alley where the
air was kept quiet. The combined results of the two tests are shown in
figure 6. There is a rapid increase in evaporation with increased wind
velocity at first ; but after a velocity of about io miles per hour is reached,
the increase in evaporation is slight. The water loss with the highest
velocity was over four times that of the calm for the 1 6-day trial and
nearly six times that of the calm in the 20-day test.

sunshine

The general effect of sunshine in increasing the evaporation of soil
moisture has long been known. The work of Seelhorst and Widtsoe on
this subject has already been reviewed. In order to get more data an
experiment was begun in the summer of 1913 at the Utah Agricultural

Experiment Station.
A spot on the college
lawn just west of the
main building was se¬
lected, because it was
level and exposed to
sunshine most of the
day.

Small tin plates con¬
taining 100 gm. dry
weight of Greenville
loam were prepared by
puddling the soil and
then drying, in order
to get uniformity.
They were then made
up to the desired de¬
gree of wetness — 5, 10, 15, 20, and 25 per cent, respectively. Each
treatment was run in triplicate, making 15 pans to the set. One set
was exposed to open sunshine, another was placed under a shade of
cheesecloth 8 inches above the soil, and a third was shaded by a tight
board cover also 8 inches above the soil. In each case the pans of soil
were placed on a floor of boards and air was allowed to circulate freely
over the soil beneath the cover.

5 to 15 T3

Wind Velocity In Miles Per Hour

Fig. 6.— Evaporation of water from wet soils with different wind
velocities.

The pans were each morning made up in the laboratory to the proper
wetness and carried outside. Temperature readings were taken in the
morning when the pans were carried out by laying a tested thermometer
for a few minutes between the pans. A reading was taken at noon and
another in the afternoon just before the pans were carried back into the
laboratory. Another weighing of the pans showed the recorded loss due
to evaporation. The pans were then left over night to be made up in
the morning before placing outside.

Dec. 4, 1916

Evaporation of Moisture from the Soil

45i

Table V shows the average temperature for morning, noon, and after¬
noon. It is worth noting that the shade caused a reduction in tempera¬
ture, complete shade causing a greater reduction than part shade.

for 13 days

Treatment.

Time.

Aver¬

age.

Evapora¬
tion loss.

Morn¬

ing.

Noon.

Even¬

ing.

Sunshine .

3°. 2
25. 6
19.7

41-3

32. 2
23. 2

34.2

28. 7
24. 6

35- 2
28.8
22. 5

Gw.

554-8

513-4

407.8

Half shade (cheesecloth) .

Shade (wood) .

Table V shows the total loss in grams for each treatment. Figure 7 shows
by graph the average temperature and the total loss with sunshine for
part and for complete
shade. It is notice¬
able that the evapo¬
ration losses decrease
as shade increases.

TEMPERATURE

Temperature, rela¬
tive humidity, and
initial quantity of mois¬
ture are usually con¬
sidered to be the most
important factors in
determining the in¬
tensity of evaporation.
Perhaps temperature
is most active. Fortier
(5) shows that a reduc¬
tion from 88° to 530 F.
causes a corresponding
reduction in evapora¬
tion loss of 85 per cent.
These figures suggest

Fig. 7. — Loss' of water from soil and temperatures in the sun and
under cheesecloth and board shade.

the close relationship of evaporation to temperature. This important
effect of temperature was made clear in a number of experiments where
other factors were being studied.

To get some specific effects of temperature on the rate of evaporation
of moisture from the soil, a number of large water baths were so arranged

452

Journal of Agricultural Research

Vol. VII, No. io

that they could be kept uniform in temperature to within about one
degree. Twenty-five gm. of soil were moistened and then put in flat-
bottomed aluminum cans 7 cm. in diameter and 3 cm. deep, the cans
being about one-fourth full. In order to bring the temperature of the
soil quickly to the temperature of the water, these cans were allowed to
float on water in the water baths. The baths were maintained at 20,
30, 40, 50, 60, 70, 80, and 90° C. It was almost impossible to maintain
a uniform temperature when the hath was hotter than 90°; and when it

was colder than 20° the

\o-

Vr>

\3>

Y !

—

\«

loff

evaporation was too
slow to give noticeable
results.

Two soils were in¬
vestigated. The first
was Greenville loam
with an initial moisture
content of 12 per cent
of the dry weight of
the soil and the second
was a coarse sand with
20 per cent of initial
moisture on the dry
basis.

Hundreds of weigh¬
ings were made to de¬
termine the rate of loss
with each temperature.
90 c These weighings gave
the results contained in
figures 8 and 9.

The loam containing
12 per cent of moisture required 265 minutes at 20° C. to lose half of
its moisture and 510 minutes to become practically dry. At 30°, 89
and 312 minutes were required to make it half and completely dry,
respectively; at 40°, 46 and 143; at 50°, 23 and 88; at 6o°, 17 and 56; at
70°, 12% and 45 ; at 8o°, and 38; and at 90°, 7 and 27.

In sand containing 20 per cent of moisture, the number of minutes re¬
quired to become half and completely dry respectively at the various
temperatures was as follows: At 20°, 315 and 819; at 30°, 90 and 240;
at 40°, 45 and 100; at 50°, 30 and 72; at 6o°, 13 and 35; at 70°, 9 and 24;
at 8o°, 6 and 18; and at 90°, and 13 minutes.

In the loam containing 12 per cent of water as an average, it required
nearly three times as long to drive off the last 6 per cent of w ater as it did
for the first 6. In the sand containing 20 per cent, it required more than

50 40 SO 60

TEMPERATURE

Fig . 8 .—Time required at different temperatures to drive off half and
all the water from Greenville loam containing 12 per cent moisture.

Dec. 4. *9*6

453

Evaporation of Moisture from the Soil

twice as long to evaporate the last io per cent as it did the first. In
each case the last water that was driven off was the hygroscopic moisture.
This probably accounts for the greater time required.

SIZE OF SOII, PARTICLES

Since evaporation takes place almost entirely at the surface, the rate
of capillary movement directly affects this form of water loss. Because
the size of soil particles and the porosity of the soil influence capillary
movement, they indirectly affect evaporation. Principi (14) says that
materials having the greatest pore space permit greatest evaporation.
Wollny (21) reports no
capillarity with particles
more than 2 mm. in
diameter, and also an
increase as the particles
get finer, though in clay
the movement is slow.

Losses from below the
surface take place, ac¬
cording to Buckingham
(2) by diffusion and
vary with the square of
the porosity of the soil.

On account of the
great difficulty in in¬
terpreting the results
connected with this
phase of the evapora¬
tion problem when or¬
dinary soils are used,
most of our work has
been done with sands
having grains of
different diameters. The early experiments were conducted with river
sands and gravels; but because of greater ease in obtaining uniform size
of particles and because of lesser influence due to the composition of the
materials, most of the later work was done with graded quartz sand.

In all trials except those in saturated atmosphere beneath the soil,
these tests were made in the copper evaporimeters with free water main¬
tained from about 1 to 3 cm. below the surface of the soil, the distance
varying in different experiments. At weighings, which occurred about
a week apart, the vessels were made up to their original weight.

The results of these tests are presented in figures 10, 11, and 12.

Fig. 9.— Time required at different temperatures to drive off half
and all the water from sand containing 20 per cent moisture.

454

Journal of Agricultural Research

Vol. VII, No. io

An examination of figure 19 shows a gradual increase in evaporation as
the size of particles decrease. There are a number of irregularities prob¬
ably caused by the difficulty in getting a uniform surface on all the
evaporimeters. It will be noted that in some cases the evaporation was

greater from the wet
sand than from water.

In figure 11 results
for three grades of
pure quartz sand, for
Greenville loam, and
for water are shown.
Here, as in figure 10,
the finer grades lose
more than the coarser.

Figure 12 includes
five sizes of quartz
sand, three sizes of
river sand, and water.
The differences are not
marked, but are suffi¬
cient to bear out previous results in showing the greater evaporation
from the surface of the smaller particles.

mulches

Wherever water storage in field soils is important, mulches are used
to decrease the evaporation loss. Ridgaway (15) and Fortier (5) both
indicate the great sav-

River Sand

Fig. io. — Evaporation of water in 66 days from sand of different sizes
with a water table maintained i cm. below the surface.

ing due to mulches
made by stirring the
topsoil or by adding
covers of dry soil.

A number of labora¬
tory experiments to
study the effectiveness
of different mulches
when the effect of ca¬
pillarity has been elim¬
inated were conducted .

In these experiments
the different mulches
were suspended above
the water. The mulch
was placed on wire gauze covered with cheesecloth or on perforated
sheet metal to keep it about i cm. from the surface of the water in the
lower part of the vessels. No water could evaporate except through
the mulches.

Fig. ii. — Evaporation of water in 36 days from loam and sand of
different sizes with a water table maintained 3 cm. below the sur¬
face.

Dec, 4, 1916

Evaporation of Moisture from the Soil

455

Fig. 12.— Evaporation of water in 115 days from quartz and river
sand of different sizes with a water table maintained 3 cm. below
the surface.

A study of two sand mulches inch and i inch deep, respectively,
when placed where the sun would shine on them for half the day for 32
days shows a loss of 57 gm. for a mulch 1 inch deep, 60 gm. for a mulch
% inch deep, and 1 55 gm. for the cheesecloth with no dry soil over it.
Thus the shallower mulch lost but little more than 5 per cent more than
the deeper. The check
in which the water
evaporated through
the cloth and gauze
lost nearly three times
as much water as that
from the mulches.

This bears out For¬
tier’s findings in regard
to the effectiveness of
mulches, but not in
regard to the relative
value of deep and
shallow mulches.

Figure 13 gives the
results of the evapo¬
ration from 1 -cm. mulches of river sands varying in size from 0.1 mm.
to 7 mm. for 40 days. The loss is somewhat greater through the
smaller sands. When, however, the sizes are larger than 1 mm., the
variations in loss are irregular and inconclusive. The mulches pre¬
vented over half the
evaporation that oc¬
curred from free water.
* With perforated
aluminum lids <> cm.
in diameter and with a
mulch nearly 2 cm.
thick, a more exhaust¬
ive experiment was
conducted. It was simi¬
lar to the above, except
that quartz sands were
used for the smaller
sizes and that tests with

Fig. 13.— Loss of water from glasses having dry mulches of sand of niUCk, day, loam, and
various sizes suspended above free water. i

straw were also run.

This experiment continued for a period equivalent to 180 days, the weigh¬
ings being made at about 10-day intervals. Results of this experiment
are shown in figure 14. Of the mulches the greatest loss was through the
muck, followed by loam, clay, and straw in order, with the sand mulches

456

Journal of Agricultural Research

Vol. VII, No. 10

having the least evaporation. The sands show somewhat the same
results as were found in the experiments mentioned above, except that
there is a tendency for the losses to increase with increasing size of soil
particles when the particles are more than 2 mm. in diameter.

COMPACTING THE) SOIL

A set of six galvanized-iron cans 1 3 inches deep and 1 1 inches in diameter
with an opening at the bottom’ through which water could be added
from below to maintain the soil at a constant moisture content were
filled with soil. Two sections of the soil at various depths were com¬
pacted in order to determine the effect of compacting on evaporation.
The cans contained 12 inches of Greenville loam made up to about

15 per cent moisture.
In can 1 the surface 2
inches were compacted;
in can 2 the second 2
inches; and so on, until
in can 6 the bottom 2
inches were compacted.
The packed layers con¬
tained 20 per cent more
soil in a 2 -inch layer
than a similar volume
of the loose soil. These
cans were weighed
weekly for seven weeks
and the loss made up
through the side tubes
at the bottom, which «vere kept closely stoppered except while water
was being added.

Table VI shows that cans 1 and 2, compacted in the first and second
2-inch layers, respectively, lost much more heavily than the cans in
which the packed layers were farther from the surface.

Table VI.— Loss of moisture from cans of soil containing a 2-inch section compacted

at various depths

Can No.

2-inch section
compacted.

Total loss.

Top

Gm.

2 .

Second ....
Third

I, 205
1,045
885
870
880

Fourth

tfifth

6 .

Bottom ....

505

Fig. 14. — Loss of water in 180 days from glasses having dry mulches
of various kinds suspended above free water.

Dec. 4, 19x6

Evaporation of Moisture from the Soil

457

Compacting the surface caused a marked increase in the loss; packing
the second 2-inch section also increased the loss, but only about half as
much. Compacting below 4 inches affected evaporation little if any.

method of applying water

An important thing to know where irrigation's practiced is the effect
on evaporation of applying water in different ways. Fortier (5) and
Widtsoe (19) indicate that a great saving results from applying water in
deep furrows or by subirrigation in which the water is added some distance
below the surface.

In the summer of 1912 a study with soils 12 inches deep was conducted.
Cylindrical vessels 11 inches in diameter and 13 inches deep were filled
to within an inch of the top with Greenville loam. An equivalent of
10 kgm. of dry soil was
used and made up with
moisture ranging from
5.4 to 35.4 per cent in
5 per cent intervals.

Weighings were made
on three days weekly —
usually Monday,

Wednesday, and Fri¬
day. The losses were
made up by adding
water through spouts
which entered the bot¬
tom of the cans and
which .were kept closed
except while water was
being added. Thus, the water had to move through 12 inches of soil,
rather compact and unstirred, and evaporated from a small surface.
Parallel to this test and used as a companion were the large galvanized-
iron pans already described under the initial-quantity study. Here the
soil was only about 1 % inches deep. To these shallow tanks the water was
applied at the surface. The two trials were parallel throughout. The
same percentages of moisture were used, weighings were made at the
same time, they were run the same period, and the same kind of soil
was used. A comparison of the two sets of results may be interesting,
as they show the effect of different methods of applying water.

Table VII gives the comparative data. In this table it may be noted
that the losses at low percentages were somewhat more rapid from the
shallow tanks, but that the deeper cans tended to lose more at the higher
percentages. These had no free water exposed on top, while the wetter
soils in shallow pans did. As already pointed out, the wet soils often

Fig. 15— Evaporation from distilled water and from sodium-chlorid
solutions of different concentrations.

458

Journal of Agricultural Research

Vol. VII, No. io

lost more than the free-water surfaces. The deeper soils were also more
uniformly wet. Capillarity doubtless played a part in. this experiment.

Table VII. — Comparison of evaporation from large areas and shallow soil with small
areas and deep soil, io kgm. of Greenville loam being used in each case

Soil

moisture.

Shallow tanks.

Deep cans.

Surface

area.

Total

evapora¬

tion.

Evapora¬
tion
for each
square foot
of surface.

Surface

area.

Total

evapora¬

tion.

Evapora¬
tion
for each
square foot
of surface.

Per cent.

5-4
io. 4
15-4
20. 4
25-4
30-4

35-4

Sq.feet.

5*42

5* 42

5- 42

5- 42
5*42
5*42

5* 42

Gnt.

2, 935

6, 100

7, 770
7,405
8,299

8,41s

8, 98s

Gw.

542

1,125

1,434

1,366

1,531

i> 5S3
1,658

Sq. feet.

0. 66
.66
.66
.66
.66
.66
.66

Gw.

270

440

700

920

I, OIO

995

1,270

Gw.

409

666

1,066

U395

1,530

1,500

i,939

SOLUBLE SALTS

The effect of dissolved salts in reducing the vapor tension, and conse¬
quently the evaporation of solutions, is well known. The action of
these salts in the soil on evaporation, however, is not so clear, since

secondary factors may
be introduced. In the
ordinary agricultural
soil the concentration
of soluble salts is not
sufficient to have any
marked effect on
evaporation, but in the
alkali soils of arid re¬
gions salts may be pres¬
ent in sufficiently high
concentrations to affect
the loss of moisture
materially.

With a view to deter¬
mining some of the ef¬
fects of salts on evaporation, a number of experiments were conducted.
In the first, solutions of sodium chlorid of various concentrations without
soil were investigated. The solutions were placed in glass tumblers, two
tumblers being used for each treatment, and set in the open laboratory
where evaporation could go on freely. The tumblers were weighted every
few days and the loss made up with distilled water.

Fig. 16. — Evaporation from sand wet with distilled water and with
sodium-nitrate solutions of differ ent concentrations.

Dec. 4, 1916

Evaporation of Moisture from the Soil

459

The results of this test are shown in figure 15, which brings out clearly
the fact that as the concentration of the solution increases the evapo¬
ration decreases.

The second test was conducted in porcelain crucibles, each containing
10 gtn. of quartz sand which had been wet with 4 c. c. of solutions of
sodium nitrate ranging in concentration from a check solution containing
no salt up to 5 times a normal solution. There were two crucibles for
each concentration. The crucibles were placed under a bell jar in order
to avoid air currents and to keep the humidity as uniform as possible
over all the crucibles.

Weighings were made each day at first, and every two or three
days later. The experiment was begun January 12 and continued till
January 27, making a period of 15 days.

The results are given in figure 16, which shows a decrease of evapo¬
ration from the sand the same as when the solution of sodium nitrate is
added. A third test
was conducted in gal-
vanized-iron cans, 1 1
inches in diameter and
13 inches deep, partly
filled with Greenville
loam to which sodium
chlorid was added in
different quantities,
rangingf rom the control
containing nothing to 7
per cent of the dry soil.

The quantity of moist¬
ure that evaporated was
added every three or
four days through a
tube near the bottom of the cans. In this way the surface of the soil
was never disturbed, but there wTas a gradual accumulation of salts at the
surface. The experiment ran from August 19 to September 25.

The results of the experiment are given in figure 17. There is a
gradual decrease in the evaporation as the salt content of the soil is in¬
creased. The can with 7 per cent of salt lost slightly more than that
with 6 per cent. This irregularity was doubtless due to the fact that
considerable salt was crystallized at the surface of the soil in this can,
consequently the real concentration of the solution was decreased; and
it is the salt actually in solution that affects vapor tension.

From these experiments it seems clear that soluble salts in the soil
decidedly decrease the evaporation of moisture if the concentrations are
high, but the reduction is only slight for the solutions found in ordinary
soils.

Percent Sa.it

Soil

Fig. 17. — Evaporation of water from Greenville loam containing dif¬
ferent quantities of sodium chlorid.

460 Journal of Agricultural Research voi. vn. No. »

SUMMARY

(1) The conservation of soil moisture is one of the most important
problems of agriculture, particularly in arid regions.

(2) One of the important factors involved in water conservation is
evaporation.

(3) Jn this paper a study has been made of a number of the factors
having to do with evaporation.

(4) Evaporation of moisture increases with the initial quantity in the
soil. The increase is not so great with the higher percentages as with
the lower,’ and there seems to be a number of critical points where the
rate of loss changes rapidly.

(5) The rate of evaporation from a moist soil is very rapidly decreased
as the humidity of the air is increased.

(6) Air currents greatly increase evaporation; but after about a certain
wind velocity is reached, the rate of evaporation is only slightly increased
by increasing the wind velocity.

(7) For the sizes investigated, evaporation was higher from the finer
soil particles than from the coarser when both are completely saturated.

(8) Reducing the intensity of sunshine greatly reduces the rate of
evaporation.

(9) Slight changes in temperature have a marked effect on evaporation.

(10) A thin mulch, if kept dry, is effective in reducing evaporation.
Dry mulches, composed of fine particles, seem to be less effective than
if composed of coarser particles.

(11) Compacting the surface of the soil increases evaporation.

(12) Dissolved salts in high concentrations reduce the evaporation of
moisture from soils.

LITERATURE CITED

(1) Bowie, A. J., jr.

1908. Practical Irrigation. . . 232 p., 53 fig. New York.

(2) Buckingham, Edgar.

1907. Studies on the movement of soil moisture. U. S. Dept. Agr. Bur. Soils

Bui. 38, 61 p., 23 fig.

(3) Cameron, F. K., and Galuagher, F. E.

1908. Moisture content and physical condition of soils, U. S. Dept. Agr. Bur.

Soils Bui. 50, 70 p., 33 fig.

(4) Carpenter, L. G.

1898. The loss of water from reservoirs by seepage and evaporation. Colo. Agr.
Exp. Sta. Bui. 45> 32 P*> 1

(5) Fortier, Samuel.

1907. Evaporation losses in irrigation and water requirements of crops. U. S.
Dept. Agr. Office Exp. Stas. Bui. 177, 64 p.( 19 fig.

(6) -

1907. Evaporation losses in irrigation. In Engin. News, v. 58, no. 12,
P- 304-307, 14 fig-

(7)

1909. Soil mulches for checking evaporation. In U. S. Dept. Agr. Yearbook
1908, p. 465-472 , fig. 22-27.

Dec. 4. 1916

Evaporation of Moisture from the Soil

461

(8) Hoffmann, Conrad.

1912. Relation of soil bacteria to evaporation. Wis. Agr. Exp. Sta. Research
Bui. 23, p. 183-215, 1 fig.

(9) King, F. H. * •

1894. Destructive effects of winds on sandy soils and light sandy loams, with
methods of prevention. Wis. Agr. Exp. Sta. Bui. 42, 29 p., 16 fig.

(10) -

1901. A Textbook of the Physics of Agriculture, ed. 2, 604 P-> 276 fig.

Madison, Wis.

(n) -

1913. Irrigation and Drainage, ed. 8, 502 p., 163 fig. New York.

(12) MacDonald, William.

1909. Dry Farming: Its Principles and Practice. 290 p., illus. New York.

(13) Payne, J. E.

1899. Influence of a wind-break upon evaporation from soil surface. In Colo.
Agr. Exp. Sta. 20th Ann. Rpt.; 1898, p. 214-215.

(14) Principi, Paolo.

1912. Alcune osservazioni sul comportamento rispetto air evaporazione dei
principali costituenti minerali del terreno agrario. In Gior. Geol.
Prat., ann. 10, pt. 1, p. 14-20; 4 fig. Abstract in Exp. Sta, Rec.,
v. 28, p. 812. Original not seen.

(15) Ridgaway, C. B.

1902. Experiments in evaporation. Wyo. Agr. Exp. Sta. Bui. 52, p. 43~55»

illus.

(16) Seelhorst, C. v.

1910. Uber den Einfluss der Beschattung auf die Wasserverdunstung des

Bodens. In Jour. Landw., Bd. 58, Heft 3, p. 221-228.

(17) STiGEix, R.

1908. Uber die Einwirkung der Bakterien auf die Verdungstungsverhaltnisse

im Boden. In Centbl. Bakt. [etc.] Abt. 2, Bd. 21, No. 1/3, p. 60-61.

(18) Whitney, Milton, and Cameron, F. K.

1904. Investigations in soil fertility. U. S. Dept. Agr. Bur. Soils Bui. 23,
p. 6-21, fig. 1-3.

(19) WidtsoE, J. A.

1909. Irrigation investigations: factors influencing evaporation and transpira¬

tion. Utah Agr. Exp. Sta. Bui. 105, 64 p., 7 fig.

\*SJJ -

1914. The Principles of Irrigation Practice. 496 p., 179 fig. New York,
London.

(21) Woixny, Ewald.

1895. The physical properties of the soil. Part 2. In Exp. Sta. Rec., v. 6,
p. 853-863.

ADDITIONAL COPIES

OF THIS PUBLICATION MAY BE PROCURED FROM
THE SUPERINTENDENT OF DOCUMENTS
GOVERNMENT PRINTING OFFICE
WASHINGTON, D. C.

10 CENTS PER COPY
SUBSCRIPTION PRICE, $3.00 PER YEAR

JOURNAL OF AGRKULTtML RESEARCH

DEPARTMENT OF AGRICULTURE

Vol. VII Washington, D. C., December ii, 1916 No. 11

MACROSIPHUM GRANARIUM, THE ENGLISH GRAIN

APHIS

By W. J. Phillips,1

Entomological Assistant , Cereal and Forage Insect Investigations , Bureau of Entomology

Although the English grain aphis (Macrosiphum granarium Kirby) is
widely disseminated throughout the United States and is a familiar pest
of long standing, there are some interesting facts connected with its life
history that heretofore have escaped observation. The object of this
brief paper is primarily to put on record some details of life history
and to discuss the interesting color variations in relation to the sexes.

SYNONYMY

Macrosiphum granarium seems to have been described first by William
Kirby (1, p. 238, footnote)2 in 1798. The description verbatim and
complete is as follows:

Possibly this may be the Aphis avenae of Fabricius: but as he has given no descrip¬
tion of it, I cannot be positive; I shall therefore describe it under the name of A.
Granaria , viridis, cauda biseta, setis geniculisque pedum nigris.

Aphis avenae, Fab. Sp. Ins. ii. p. 386. n. 17. Gmel. tom. 1. part IV. p. 2206. n. 52.
Vill. Ent. Eur. i. p. 551. n. 50?

Caput falvidu, uti antennarum articulus primus. Oculi nigri. Abdomen obova-
turn cauda aculeata. Pedes lividi, tarsis geniculisque nigris.

Habitat in tritici et hordei spicis, aveneque paniculis.

Although meager, the foregoing description agrees with the species
known to entomologists by this name. It can hardly be construed as a
description of Aphis avenae Fab., as has been supposed by some authors.
Curtis (3, p. 504) redescribed what he considered Kirby’s species in the

1 The writer wishes to acknowledge his indebtedness to Mr. T. H. Parks, lately of the Bureau of Ento¬
mology, for his assistance in conducting breeding experiments in 1909 at La Fayette, Ind., The observa¬
tions upon which this paper is based were made a* Richmond, Ind. (1907-8), La Fayette, Ind. (1909-1912),
and Charlottesville, Va. (1915). Mr. J. J. Davis, of the Bureau of Entomology, kindly consented to prepare
the synonymy.

2 Reference is made by number to “ Literature cited,” p. 480.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.
gk

(463)

VoL VII, No. iz
Dec. xxt 1916
K— 47

464

Journal of Agricultural Research

Vol. VII, No. 11

Journal of the Royal Agricultural Society and again in his “Farm In¬
sects” (6, p. 289). While there are minor points in Curtis’s description
and figures which seem to disagree, as a whole they apply quite well for
M. granarium auct., and it is reasonably certain that he had this spe¬
cies before him when he made his description.

In 1843 Kaltenbach (2, p. 16) described Aphis cerealis. Pergande
(11, p. 13-23) considered this to be distinct from M. granarium , basing
his opinion largely on the presence or absence of abdominal macula-
tions. However, this character is unreliable, as has been proved in
breeding experiments where individuals showing all degrees of abdom¬
inal markings and some without the faintest trace of maculations were
reared from the same mother. Most European authors now consider the
two species M. granarium and A . cerealis as synonyms, and on inquiry
the following replies have been received from the respective eminent
European aphidologists. Under date of February 15, 1913, Prof. Feed
W. Theobald writes: “I look upon cerealis and granarium as the same.
I can see no difference.” Under date of January 17, 1913, Mr. P. van
der Goot writes as follows: “M. granarium and M. cerealis I must con¬
sider as one species.” Dr. G. del Guercio has the following to say in a
letter dated December 13, 1911:

I have examined the specimens of Macrosiphum or Siphonopkora granariae. They
show some differences on which it may be possible for ns to distinguish certain forms,
which, however, as far as I am concerned, could never be considered varieties, let
alone species. Fundamentally the Siph. granariae there (in America) is the Siph.
cerealis here, and both in fact, secondary differences aside, are the same species.
Your specimens have the antennae a little longer than the body, while in our forms,
at least those of Italy and of the European basin of the Mediterranean, the antennae
are shorter than the body. Buckton [7, p. 114-119, pi. 6], in his first volume, gives a
good representation of Siphonopkora granariae . (Free translation from the Italian.)

In 1849 Walker (4, p. 45-46) described this species as Aphis avenae
Fab. The species was transferred to the genus Siphonophora by Koch
(5, p. 186-187) in 1857, to Nectarophora by Oestlund (8, p. 82) in 1887,
and finally to the genus Macrosiphum by Schouteden (10, p. 113-117)
in 1901. In 1905 Kirkaldy (12, p. 132) proposed the name M. aveni-
vorum for M. granaria Buckton, nec Kirby, and this must now be con¬
sidered a synonym of M . granarium.

The synonymy as it now stands is as follows:

Macrosiphum granarium Kirby.

Aphis granaria Kirby, 1798, in Trans. Ihm. Soc. [London], v. 4, p. 238.

t Aphis hordei Kyber, 1815, in Mag. Eat. [Germar], 1798, in v. 1, pt. 2, p. 211, nomen nudum.

Aphis cerealis, Kaltenb., 1843, Monog. Fam. Pflanzenlause, p. 16.

Aphis granaria Curtis, 1845?, i860. Jour. Roy. Agr. Soc., England, v. 6, p. 504; Farm Insects, p. 289.

Aphis avenae Walker (nec Fab.), 1849, in Ann. and Mag. Nat. Hist., s. 2, v. 3. P- 4S“46.

Bromaphis Amyot, 1847, in Ann. Soc. Ent. France, s. 2, v. 5. P- 479-

Siphonophora cerealis Koch, 1857, Monog. Pflanzenlause, p. 186-187.

Siphonophora granaria Buckton, 1876, Monog. Brit. Aphides, v. 1, p. 114-119, pi. 6.

Nectarophora granaria Oestl., 1887, in Geol. and Nat. Hist. Survey Minn. Bui. 4, p. 82.

Macrosiphum granarium Schout., 1901, in Ann. Soc. Ent. Belg., t. 45, p. 113-117.

Macrosiphum avenivorum Kirkaldy, 1905, in Entomologist, v. 38, p. 132.

Dec. ii, 1916

Macrosiphum granarium

465

DISTRIBUTION IN THE UNITED STATES

M. granarium undoubtedly occurs throughout the United States
wherever the small grains are cultivated. The map (fig. 1) indicates
localities from which the Bureau of Entomology has records of occurrence.
It will be noted that there are 10 States from which the Bureau has no
records, though M. granarium undoubtedly occurs in those States.

FOOD PLANTS OF THE APHID

This aphid does not confine itself exclusively to its well-known host
plants, the small grains, but will live and thrive on a number of the wild
and cultivated grasses.

Riley (9) listed A grostis vulgaris [alba], Bromus secalinus , Dactylis
glomerata, and Poa pratensis as host plants. Besides the plants just

Fig. i. — Map showing the distribution of Macrosiphum granarium in the United States as indicated by
records on file in the Bureau of Entomology, 1916.

mentioned, the late F. M. Webster recorded it, in notes on file in the
Bureau of Entomology, as breeding on heads of timothy (. Phleum pra-
tense) at Mitchell and La Fayette, Ind., in 1889, and in 1890 he recorded
it as feeding on com (Zea mays) at La Fayette, Ind. In 1904 Pergande
(11) recorded Elymus sp. as a host.

A series of experiments conducted at La Fayette, Ind., in 1909, showed
that M. granarium will breed and thrive in confinement upon the follow¬
ing grasses: Bromus commuiatus (?) [racemosus], B. secalinus, Elymus sp.,
Festuca duriuscula [ovina], ‘ F. heterophylla , F. • pratensis [elatior ], F.
tectorum , J uncus tenuis, Lolium italicum, Poa compressa, and P. pra¬
tensis. M . granarium was found to breed freely in confinement upon
Eleusine indica and foxtail (probably Chaetochloa glauca) at Richmond,

466

Journal of Agricultural Research

Vol. VII, No. ii

Ind., in 1908. Most of the grasses in the experiments at La Fayette,
Ind., were obtained from the experimental plots of the Purdue Experi¬
ment Station.

Other members of the Office of Cereal and Forage Insect Investiga¬
tions have recorded the following hosts : Bursa bursa-pastorisf Nashville,
Tenn. (G. G. Ainslie); Syntherisma sanguinale, North Vernon, Ind.,
1908; Echinochloa crus-galli, Princeton, Ind., 1908; and Hordeum pusil -
lum} Salisbury, N. C., 1909 (R. A. Vickery).

DESCRIPTIONS OF THE FORMS
STEM MOTHER

A number of stem mothers were secured at La Fayette, Ind., in the
spring of 1911 ; but as the writer was absent at the time, the description
was written by Mr. J. J. Davis. There was some slight variation in
color of the adults, some having a somewhat yellowish tinge.

Body, including thorax, slightly darker than apple green on the dorsum; venter
very slightly pruinose, giving it a silvery or whitish color. Eyes dark red. Head
pale brownish. Antennae (PL 34, F): I and II concolorous with head, but with slight
duskiness; III, IV, V, and VI black. Beak pale brownish at base and the last two
segments black. Tip of beak just reaching second coxae. Legs: Basal half of femora
pale greenish, distal half dusky to black; tibiae pale, with slight brownish tint, the
distal end black; tarsi black. Cornicles black; style pale whitish, with slight greenish
tint.

Measurements made from four individuals immediately after mounting in balsam:

Length of body.

With style.

Width.

Mm.

2. 27

2. 59

1. 21

2. 63

3. 02

i- 47

2. 39

2.39

1. 12

2. 22

2. 50

1. 24

Cornicles, 0.39, 0.515, 0.408, and 0.51 mm., respectively. The antennal measure¬
ments are given in Table I.

Tabus I. — Length (in millimeters) of antennce of the stem mothers of Macrosiphum

granarium

No.

Specimen No.

I . . .

0. 106

O. 106

0. 12

0. 11

O. IO

155

o- *55

II .

.07

.08

.07

.066

. 06

Ill .

•43

•44

•53

•435

.586

IV .

• m

.i8S

•30

.28

•15

•31

•3°

V .

• 24

• 23

. 28

• SO

•23

•337

VI base .

. n

. 12

•13

. 12

> 14

. 14

VI filament .

• 31

•34

•39

•32

.42

.44

Total .

i- 571

1. 491

1. 77

1.81

1. 421

2. 008

2. 018

Dec. ii, 1916

Macrosiphum granarium

467

SUMMER FORMS

The following description is taken from Pergande (1 1) :

Apterous female. [PI. B, 2.] Length 2.4 to 2.8 mm.; fusiform, broadest near the
base of the abdomen. Frontal tubercles large, diverging at the apex, as usual, in this
genus; antennae bristle-shaped, as long or slightly longer than the abdomen; joint
six, including the spur, longer than joint three; generally there are one or two small,
circular and projecting sensoria near the base of the third joint; all of the joints are
very sparsely beset with short and stiff bristles which are rarely slightly clavate.
The nectaries are long and reach beyond the tip of the abdomen, though rarely beyond
the tip of the tail; they are cylindrical, tapering, becoming again slightly stouter
toward the end. The tail is rather long and stout, curved upward, and about two-
thirds the length of the nectaries, lanceolate, and more or less distinctly constricted
about the middle; it is densely covered with acute, minute points and furnished
each side of its terminal half with three, backward-curved, long bristles. The legs
are long and provided with short, stiff, and simple hairs.

The color of the apterous female is yellowish-green, often slightly pruinose; fre¬
quently darker toward the end of the body; the head varying from yellow to brownish-
yellow. The eyes are red to brown, while the tail varies from white to a distinct
yellow. The antennae, as a rule, are black, though sometimes the first joint may be
yellow or the first three joints dusky. The terminal half or more of the femora, apex
of the tibiae, the tarsi, and the nectaries brown to black; the rest of the leg is yellow.
The body is frequently marked with a brownish puncture or spot each side of the
prothorax; sometimes there is a narrow dusky or black line, composed of minute spots,
each side of the mesothorax and a dorso-lateral row of about five linear or rounded,
blackish or dusky spots each side 6f the abdomen, which sometimes are extremely
faint or even wanting. Occasionally there are also two additional small black or
dusky spots between the nectaries. Lateral spots in front of nectaries black.

Winged migrant. [PI. 33, A.] Expanse of wings 9 to 9.4 mm. ; length of body
1.4 to 2,6 mm. Antennae long, generally about one-third longer than the body; the
third joint about one-third shorter than the sixth and provided along its exterior or
posterior edge with from six to eleven more or less elevated, round sensoria along
its basal third. The hairs of the various joints are similar to those of the apterous
female, though sometimes one or the other may be distinctly clavate. The nectaries,
tail, and legs in general appearance and size are very similar to those of the apterous
form. The wings are almost twice the length of the body, while the venation cor¬
responds very much to that of Aphis.

Color yellowish green to green; the mesothorax yellow and its lobes brown to black.
Sometimes a small, oblique, dusky, subdorsal spot and a transverse pale dusky band
may be observed on the prothorax. Head brown or brownish-yellow; eyes red to
brown. Antennae black, the first joint sometimes brownish-yellow externally.
Nectaries black, the tail yellowish or greenish-yellow; sternal plate and lateral spot
in front of wings black. The abdomen is marked with four or five small, transverse,
blackish dorso-lateral spots and four black lateral spots in front of nectaries; the colora¬
tion of the legs is similar to that of the apterous female. Wings clear, the costa dusky,
and the subcosta yellow; stigma yellowish, its inner margin dusky; veins yellowish-
brown, changing to black toward the end.

It is probably well to state in this connection that the maculation of
the abdomen is quite a variable character and that the cornicles are
reticulate at the tip.

The following measurements (Table II) were made from specimens
that had been mounted in balsam for over a year.

468 Journal of Agricultural Research voi. vn, no. »

Table II. — Length ( in millimeters) of the antennae of summer forms of Macrosiphum

granarium

WINGED VIVIPAROUS FEMALES (PL. 34, B)

No.

Specimen No.

I .

II .

III .

IV. . .

V. .

VI . .

Total . . .

Cornicles .

Cauda .

O. 09653
.06895
. 66192

• 55849

. 38612
/ .12411

l • 77224

0. 09653
. 06895
. 60676
• 55849
. 38612
. 13100

0. 09633
. 06895
. 66192
.57128
. 44128
. 12411
. 66192

0. 09653

• 07584

• 64813

• 56539

. 44128
■ 13790

• 64813

0. 09653
. 06895

■ 59986

■ 47575

. 38612
. 12411

• 75845

0. 09653
. 08274

• 63434
. 46886

* 358S4
. 11721
. 64813

2. 66836

2, 62594

2. 61320

2. 50977

2. 40635

f .42749
l * 41370
• 3°338

- 43438
. 44128
• 30338

• 37233
. 38612
. 27580

WINGLESS VIVIPAROUS FEMALES (PL. 34, D)

I . . .

0. 10342
, 06895

• 57918

• 35854

. 31027
r . 11712

I • 56539

0. 11032
. 07584
. 75228

• 35854

• 28959
. 12411

• 53091

0. 1 1032

• 07584

• 56539

• 42438

• 33785

• 13790

. 68260

0. 1 1032
. 08274
. 71708
. 52402

• 37233
. 12411

• 70329

0. 1 103 2
. 08274
. 68950
. 52402
. 38612
. 12411
. 68950

II .

III .

IV .

V .

VI .

Total...

0. 62055
. 42749

•3*7*7
. 12411
. 60676

0. 62055
. 44128
.33096
. 12411
. 62055

2. 00296

2. 06159

2. 33428

2. 63389

2. 60631

/ .42749
l -42749

.33096

. 46196
.44817

• 33096

* 59297

• 59297
- 41370

. 46886
. 48265
•34475

v^omicies .

Cauda .

SEXES

The following description of the sexes is from Sanderson (13) :

Apterous oviparous female. [PI. B, 5.] One specimen, 1.9 mm. long by 1 mm.
wide; antennae [PI. 34, E] 2 mm., segments, 3, 0.50 mm.; 4, 0.35 mm.; 5, 0.30 mm.;
6, 0.10 mm.; 7, 0.50 mm.; cornicle, 0.43 mm.; cauda, 0.21 mm. Somewhat smaller
than viviparous form. At first yellow, then turning green and darker green. Head
light brown. ■ Distal two-thirds of femora, tip of tibia, tarsi and cornicles black,
antennae black. Conspicuous horizontal black marking in pit of connexivum on
either side, these being more or less connected by black lines on the sutures of the
first six abdominal segments and coalescing to form a faint but distinct black spot
on abdominal segments 4-6. Meta-tibia with numerous pores. [PI. 34, C.]

Winged male. [PI. 33, B.] Antennae [PI. 34, A] 2.8 mm.; segments, 3, 0.68 min.;
4, 0.50 mm.; 5, 0.46 mm.; 6, 0.14 mm.; 7, 0.78 mm.; cornicle, 0.14 mm.; cauda,
0,14 mm.; wing 3.35 mm. long. The third antennal segment with 35 to 50 sensoria,
the fourth segment with a row of 10 to 12 on basal two-thirds, about 10 large sensoria

Pec. xr, 1916

Macrosiphum granarium

469

on distal two-thirds of fifth segment, and usual large sensoria at tip of sixth and seventh
segments. Similar to winged viviparous female, but reddish to reddish brown, with
black markings on either side of dorso-meson of abdominal segments, especially on the
seventh segment where the marking converges on the meson.

INTERMEDIATE FORM

One individual was found that contained only eggs. Her hind tibiae
were not swollen; nor did they have sensoria. The general color was the
same as that of the oviparous female. It is not known whether she
produced young previous to being mounted.

EGG

The egg is elliptical, 0.3 mm. in diameter .and 0.7 mm, long. It is a
pale yellow when first deposited, changing in a few days through different
shades of green to black.

LIFE HISTORY AND HABITS

Eggs begin to hatch during the last week in March in the latitude of La
Fayette, Ind., and continue hatching through the first week in April.

Eggs were obtained in Richmond, Ind., in the fall of 1908, but none
hatched the following spring. Eggs were again secured in the fall of

1909 at LaFayette, Ind., though only one hatched from this lot. The
mortality of the eggs is very high, but no definite cause can be assigned
for this at present. The eggs of this species were placed in hibernation
under apparently the same conditions as those of several other species of
Aphididae, the latter hatching readily and M. granarium hatching very
sparingly or not at all. From hundreds of eggs secured in the fall of

1910 only about 15 or 20 hatched. The eggs would remain plump
until about time to hatch and would then shrivel. Eggs began hatching
on March 24 in 1911.

As is common with Aphididae in general in this latitude, this species
at La Fayette, Ind., reproduces parthenogenetically until October, when
the sexes appear and eggs are deposited. The writer took adult males
in the field on bluegrass at La Fayette in November, 1909, and young
males were observed on rye and volunteer oats on the Purdue Uni¬
versity farm in November, 1911. No oviparous females have been
observed in the fields as yet, but the presence of the males indicates
that the sexes occur normally on the small grains and on blue grass in
the fall.

Mr. R. A. Vickery, of the Bureau of Entomology, stated that he has
taken the sexes on wheat in Minnesota, but he made no mention of having
obtained eggs or stem mothers. Sanderson (13) reared adults of both
sexes indoors in Texas in 1903, although he made no record relative to
the egg. It is doubtful whether eggs occur normally that far south.

470

Journal of Agricultural Research

Vol. VII, No. n

Viviparous females have been carried through the winter out of doors
in breeding cages at La Fayette, Ind., and at Charlottesville, Va. , and have
been found on the small grains throughout the fall, in the winter, and
again in the early spring, so they doubtless pass the winter both in the
egg and as viviparous females in the Northern States. It is doubtful
whether eggs and stem mothers normally occur much south of latitude
350 unless it is in higher altitudes.

The aphids remain on the leaves of wheat and other small grains until
the heads are formed and then cluster around the tender kernels, sucking
the rich sap. Just before harvest, when the plant tissues become hard
and tough, all immature individuals become winged and migrate to some
of the grasses, where they remain until volunteer grain and fall wheat
put in their appearance.

REARING CAGES

The same type of shelter and rearing cages were used as those pre¬
viously described and figured by the writer (14).

GENERATION SERIES

The generation series were not started with stem mothers in any
instance, as no eggs hatched until the spring of 1911. Since they had
been carried through consecutive generation series for each of the three
preceding years, it was thought unnecessary to continue longer. The
series were started each year with the progeny of individuals that had
survived the winter. In fact, they were started at the time the eggs
of Toxoptera graminum began to hatch. It was found later that this
was approximately the date of hatching for M. granarium.

The usual method that has been followed in the past by the Office of Cereal
and Forage Insect Investigations in the generation rearing was adopted for
this species — that is, the first born from each first born and the last born
from each last born were isolated and daily records made as far as possi¬
ble. Two partial generation series were carried through in 1907 and four
complete ones in 1908 at Richmond, Ind. — that is, they were run either
until the sexes appeared in the fall or until the work was interrupted by
cold weather. Another generation series was carried through at La Fay¬
ette, Ind., in 1909 and one in 1915 at Charlottesville, Va. The writer
thus has observations covering nearly 120 individuals. This is a com¬
paratively small number, but since the observations cover practically
four years the data should prove reliable.

Tables III and IV give in detail consecutive generations from one indi¬
vidual hatched March 27, 1908, at Richmond, Ind.

Dec. ii, 1916

Maerosiphum granarium

47i

Tablr III. — Line of generations of Maerosiphum granarium at Richmond , Ind., in igo8

Date.

Temper¬

ature.

First-born generation series.

Last-born genera¬
tion series.

•SS

! 55

pc,

’I

Mar. 27
Apr. 18

May 1

June 1

°F.

70
68
68

71
74

67
54

6 S

76
80

68
76

85
92

88
90
90

76
70

87
90
86
78

72
82

88
70

74
82

°F,

53
42
38

36
33

37
4i
49

60
SS
57

59
52

61
54

55
55
66
57

62
62
55

43
43
52
61

50
55

60
64

64
54
43

58
45

30
47

66
6c

<>B

}■

} 4

1 2

/ 3

•

t>D

}■

}«'

}«

....

■ X

{°

2 i

89 65
9i i 64

472

Journal of Agricultural Research

Vol. VII, No. «

Table III. — Line of generatio ns of Macrosiphum granarium at Richmond , Ind. , in igo8 —

Continued.

Date.

Temper¬

ature.

First-born generation series.

Last-born genera¬
tion series.

jjj

■I

■i

-3

(ft

flj

June 23

July 1

Aug- 1

°F.

94
79

79
88
89

92
86

80
77
75

93
86

85
87

91
91

86
93
93
93

82
86

79
75

82
88

80
90

86
86

°F.

53
60

52
57
64
66
60
66

64
48
47

51
54
67

66
7i

59
06

61
66

64
67

65
67

51
45

€

...

. . i

....

. ..

:::

...

}’

3
l

■

...

. ..

(

pec. ii, 191.6

Macrosiphum granarium

473

Table III. — Line of generations of Macrosiphum granarium at Richmond , Ind. , in iq8o —

Continued.

Temper¬

ature.

First-born generation series.

Last-born genera¬
tion series.

Bate.

««

t/i

•g

j Fourteenth generation.

j Seventeenth generation.

b£

’8

&
t n

Aug. 29

Sept. 1

°F.

~ i

81
86
89
74

...I ..

. . J

....

::::

... .

/ 1

..J

...!...

85
87
91

49
56
52

50
48

' } ‘

...i...

J 2
L

5°

...

74
79
78
81
Ra

45
39

47
47
45

40
5®
55
32

}

f°

....

1°

....

i ■

. J. . .

...i...

I 0

j 0

■ 18

/...

1 B

1 0

!...

L...

Total . . . .

1 The “B ” at the head of each column shows that the aphid was bom on the date indicated.

2 The. “ D * ’ immediately following each column of fi gures shows that the aphid died on the date indicated.
The total number of young for each female is given at the foot of each column.

474

Journal of Agricultural Research

Vol. VII, No. ii

Table IV. — Line of generations of Macrosiphum granarium at Richmond, Ind., in iqo8

Generation.

Date of birth.

Date of first young.

Age at birth of

first young.

Date of last young.

Productive

period.

Life after last

young.

. Number of young.

Average young per

day of produc¬

tive period.

Largest number of

young in one day.

Date of death or

disappearance.

Total length of

life.

First-bom genera-

tion series:

.Days.

Days.

Days .

Days.

1 .

Mar. 27

Apr.

May 26

1.0+

June 9

2 .

Apr. 19

May

May 17

1. s

May 18

3 .

May 9

May

May 27

1. 7+

May 31

4 .

May 18

May

June 25

1. 6-f-

June 27

5 .

May 27

June

June 20

1. 1+

June 22

6 .

June 7

June

July s

July 8

7 .

June is

June

July 4

2. 6+

July s

8 .

June 23

July

July 18

2. 6+

July 20

9 .

July 3

July

July 19

i- 5

. . .do _

10 .

July 11

July

July 26

1. 5+

July 27

11 .

July 19

July

July 29

Aug. 3

12 .

July 29

Aug.

Aug. 17

I. o-l-

Aug. 17

13 .

Aug. 6

Aug.

Sept. 9

I- 5+

Sept. 9

14 .

Aug. 18

Aug.

Oct. 13

•5+

Oct. 26

15 .

Aug. 30

Sept.

Sept. 25

1. 0+

Oet. 19

16 .

17 . .

Sept. 11
Sept. 29

Sept.

Oct. 17

• 5+

Oct. 26

Last-bom genera-

tion series:

2 .

May 27

June

July 1

.4+

July 4

3 .

July 1

July

July 16

2.6+

July 19

4 .

July 16

Aug.

Aug. 6

1. 3+

Aug. 8

5 .

Aug. 6

Aug.

Sept. 5

Sept. 9

6 .

Sept, s

Sept.

Oct. 19

•7+

Oct. 22

7 . . .

Oct. 19

The other generation series are not tabulated as there do not seem
to be sufficiently striking differences to justify it. One generation
series ran to 18 in the direct line of first born and to 8 in the last
bom. Other generation series ran below the one tabulated. One female
in the series not tabulated began producing young at the age of 7 days;
one female produced 52 young and another lived 79 days. The writer
has complete records of the number of young produced by 117 females.
They produced 2,333 young, or an average of 19.9+ young each. The
average length of time from birth to the production of young for 91
individuals was 12.6+ days. The average productive period for 99
individuals was 16.7 days. The productive period for one female (not
in the tabulated series) was 48 days. The average length of life for 89
individuals was 34.3 days.

COLOR VARIATION IN RELATION TO THE SEXES

All who are familiar with this species of Aphididae will probably recall
having seen them clustered on heads of wheat just before harvest, and
noted distinct variation in color. The majority of them at this time are
strongly tinged with pink. There seems to be no satisfactory explana¬
tion for the occurrence of the pink forms at this time. The pink forms
occur again in October, and this is the signal for the appearance of the

Dec. ii, 1916

Macrosiphum granarium

475

sexes. The writer has taken both pink and green individuals in the
summer and kept the progeny of each isolated in rearing cages until
fall. The sexes appeared among the descendants from the pink indi¬
viduals, but very sparingly from the descendants of the green ones.
This may not hold in every instance, but it has been the experience of
the writer that the sexes can be obtained with certainty by starting
with the pink summer forms.

If the cages are examined closely when the sexes begin to appear in the
fall, two distinct types of adults will be noted. One is the usual green
form (PI. B, 2) and the other will have a pinkish tinge (PL B, 1). If
the pink wingless individuals are isolated it will be found that they
produce two kinds of young, one slightly tinged with pink (PL B, 4)
and another a. deep pink (PL B, 3). The color of the mother after
she begins producing young is due in the main to the pinkish young
showing through the body wall. The slightly pink individuals are
usually produced first. The offspring of the pink wingless individuals
usually all become winged, the deep-pink individuals developing to
winged males, while the pale-pink winged individuals are viviparous
and produce the wingless, yellow, oviparous females (Pl. B, 5). This
fact was not known definitely by the writer until the fall of 1909, when he
isolated a few pink wingless viviparous females in order to learn what sex
their offspring would be. In every case the results were as just stated.

In the fall of 1910 pink wingless individuals were again isolated for
observation on their progeny. The results obtained entirely corrobo¬
rated the data of 1909. In the fall of 1911 a large series was isolated
as in 1909 and 1910. A heavy storm accompanied by very low tempera¬
tures put an end to the observations before the data were complete,
killing all individuals under observation, since the rearing cages offered
little natural protection from cold.

Since 1909 and 1910 the writer has found that the winged viviparous
females of this series 1 may produce only viviparous individuals in some
cases or both oviparous and viviparous, or may produce only the ovip¬
arous females. The males are produced only by the wingless pink
viviparous females and the oviparous females are produced only by the
winged adults that develop from the slightly pinkish young. In other
words, the males are sons of the pink viviparous females and the ovip¬
arous females are the granddaughters. In no case are oviparous
females and males produced by the same mother. The oviparous
females might be termed nieces of the males; they are never sisters of
the males.

The following outline will illustrate more clearly the sequence of the
sexes :

1 The offspring of the pinkish wingless viviparous females of the autumn forms.

476

Journal of Agricultural Research

Vol. VII, No. ii

Outlines showing methods of sequence of the autumn forms of Macrosiphum granarium
First method:

Parental t e J^orms °f ^rst generation off-JForms of second generation
^ spring. | offspring.

Pinkish nymphs, becoming f Yellow, wingless oviparous fe-
winged viviparous females. I males only.

Deep-pink nymphs, becoming
winged males only.

Pinkish wingless
: viviparous fe¬
males.

Second method :

Parental type. . .

Pinkish wingless
viviparous fe¬
males.

(Forms of first generation off-JForms of second generation
I spring. | offspring.

a. Yellow, wingless oviparous
females only;

b. Wingless viviparous females
only;

c. Winged and wingless vivip¬
arous females;

d. Yellow, wingless oviparous
females and winged and
wingless viviparous fe¬
males.

B. Pinkish nymphs, becominglwinged males and winged and

A. Pinkish nymphs, becoming
winged viviparous fe¬
males.

wingless viviparous fe- :
males. J

C. Deep-pink nymphs, becom¬
ing winged males only.

wingless viviparous females.

It will be seen, therefore, that in a single cage there may be green,
slightly pink, deep-pink, and pale-yellow individuals — quite a wide
range in color for a single species.

INFLUENCE OF TEMPERATURE ON PRODUCTION OF SEXES

During the first week in October, 1912, pinkish forms were plentiful
in the stock cages and the sexes had begun to appear sparingly. A
large number of pinkish individuals that showed promise of producing
the sexes were isolated and placed in one of the greenhouses of the
Purdue Experiment Station to hasten the production of the sexes and
that more data on the progeny of the pink wingless females that appear
at this time might be gathered. The greenhouse was kept at a tem¬
perature between 50° and 70° F., and the writer thought that since the
sexes had begun to appear their numbers could be rapidly increased by
placing them in a warmer temperature. Almost the opposite effect was
produced. A number of males but only a very few oyiparous females
appeared. The ones that were obtained were probably born just before
or very soon after they were placed in the greenhouse. A stock cage
that was left outdoors produced quite a number of oviparous females and

Dec. ii, 1916

Macrosiphum granarium

477

males. This is not conclusive proof, but it certainly indicates that tem¬
peratures below 50° F. for a daily minimum in some way exert an in¬
fluence on the normal production of the sexes.

OCCURRENCE OR THE SEXES AND THE PROPORTION OP MAIES TO FEMAIES

The first published record on the sexes is by Sanderson (13), who
secured them from indoor rearing cages in April. These were the
progeny of individuals taken in the fields in January. Notes on file in
the Bureau of Entomology show that the late E. M. Webster made observa¬
tions on the sexes as early as 1884 at Oxford, Ind. He records the males
appearing as early as September and females in October and November.
Although the egg was observed, no record was made relative to the
stem mothers.

These observations in regard to the sexes agree very well with observa¬
tions made by the writer. The young males first made their appearance
during the last week in September or the first week in October. The
females usually appear a little later. The males are likely to appear any
time during the winter if kept in breeding cages indoors. In rearing
cages indoors they appeared sparingly from September to April, inclusive.
No oviparous females occurred during the winter. Males undoubtedly
outnumber the females from the very fact of their occurrence both in and
out of season. During the breeding season (October and November),
however, the oviparous females usually outnumber the males, as there
are from a fourth to a half as many pink individuals (mothers of ovipa¬
rous females) as there are males. Each of the slightly pink females may
produce from 6 to 20 oviparous females, and that would bring the num¬
bers of the oviparous females far ahead of those of the males.

MATING

Mating occurs sometimes during the first two or three days after the
female becomes adult, and oviposition begins in ^ to 5 or 6 days,
depending upon the temperature. Females refuse to deposit eggs
before mating. If the male is not present the bodies of the females
become almost twice the normal size. In one case 5 females were
isolated from males for about 10 days or more. Their bodies increased
greatly in size, but no eggs were deposited. At the end of that period
males were .placed in the cage. Mating soon took place. In 10 days
there were 13 eggs in the cage but all were infertile.

AGE OP OVIPAROUS FEMALES AT OVIPOSITION

The length of time for maturity of oviparous as for viviparous females
depends largely upon temperature. Under the same conditions the ovip¬
arous females develop in about the same length of time as the vivip¬
arous. Targe numbers of both sexes never reach maturity because of

478

Journal of Agricultural Research ‘

Vol. VII, No. n

low temperature. One oviparous female became adult in 9 days indoors.
Another, bom the same day and kept outdoors, developed in 12 days.
The age at oviposition, therefore, would be from 14 to 20 days, depend¬
ing upon the temperature and the presence of the males. The eggs are
deposited on the leaves of the plant and on the sides of the cage.

FECUNDITY OF OVIPAROUS FEMALES

The largest number of eggs produced by a single female is 18, the pro*
ductive period lasting from November 2 to December 1 (1909). Com¬
plete records on 20 individuals give an average of 8.4 eggs. The duration
of the productive period is from 8 to 29 days.

length of life of the sexes

The majority of the oviparous females observed by the writer lived
until killed by very low temperatures in November or December. The
males do not live quite a month. The female just mentioned, that pro¬
duced 18 eggs, lived over a month after oviposition began, and was then
killed by a severe freeze. Add to this her developmental period and she
would be at least months old. This would probably be high for an
average.

molting

Molting experiments have been conducted with each form and it
was found that the stem mothers, winged and wingless forms, males, and
oviparous females, without exception, molt four times.

NATURAL ENEMIES

APHIDIUS NIGRIPES

The most efficient enemy of M. granarium is undoubtedly Aphidius
nigripes Ashmead. As soon as A . nigripes becomes abundant,. the brown
leather-like, almost circular bodies of the aphids will be noticed firmly
attached to the plant. These contain the immature stage of the parasite.
Just before harvest, if the infestation of M. granarium is heavy, the heads
of grain will be almost covered with their brown, dead bodies.

In the fall of 1908 sufficient data were secured to establish the parthe-
nogenetic habits of this parasite. It produces only males under these
conditions.

On October 7, 1907, at Richmond, Ind., two virgin females were intro¬
duced into a cage with a number of M. granarium that had been grown in
confinement and had not been parasitized previously. They began ovi¬
position at once. They would approach the aphid cautiously, bend the
abdomen under until the tip extended beyond the head, then quickly
stab the aphid. There does not seem to be any favorite point of attack,
the parasite thrusting at the nearest point. The cage was kept out of
doors in the rearing shelter and on November 1 the aphids began to

Dec. ix, 1916

Macrosiphum granarium

479

turn brown. On November 4 they had the usual leather-like appearance
and each one was firmly glued to the leaf. On November 23 part of
these old bodies were taken indoors and kept at the ordinary room tem¬
perature. The temperature went much lower at night than during the
day, as the fire was allowed to go out. On December 2, male A. nigripes
began to emerge and on the 4th all had emerged that were brought in¬
doors. The ones that were left outside were still in the larval stage.
This is probably the stage in which they pass the winter.

OTHER INSECT ENEMIES

The late F. M. Webster made more observations on the parasites and
predacious enemies of M. granarium than any other entomologist. His
notes from 1884 to 1890 that are on file in the Bureau of Entomology
record the following insects as attacking this aphid :

CODEOPTERA

HYMENOPTERA

Podabrus tomentosus Say.
Coccinella g-notata Herbst.
Hippodamia parenthesis Say.

H. convergens Guerin.

H. iy punctata Linnaeus.

H. glacialis Fabricius.

Anatis i$-punctata Olivier.
Megilla maculata De Geer.

DIPTERA

Allograpta obliqua Say.
Sphaerophoria cylindrica Say.
Xanthogramma emarginata Say.

Aphtdius avenaphis Fitch.

(Dioeretus) Praon americanus Ashmead.
(D.) Praon brunneiventris Ashmead (=
Praon americanus) .

(D.) Praon ferruginipes Ashmead (—
Praon americanus ).

Isocratus vulgaris Walker.

Encyrtus websteri Howard.

Pachyneuron micans Howard.

Allotria tritici Fitch.

Riley (9) records in addition the following :

COLEOPTERA

HYMENOPTERA

Coccinella sanguinea Linnaeus.

DIPTERA

Syrphus americanus Wiedemann.

Aphidius granariaphis Cook.

Tetrasiichus ingratus Howard [no men
nudum ]

Megaspilus niger Curtis.

All of the parasites listed, however, are not primary. In recent years
two species, Pachyneuron sp. and Allotria sp., have been definitely proved
to be secondary parasites. It is very probable that others in the list
will be proved secondary upon further study.

FUNGUS ENEMIES

This aphid seems to be very susceptible to fungus attack. During
warm, moist weather rearing cages have to be carefully watched or fun¬
gus will soon gain control. It undoubtedly destroys many aphids in the
fields also.

66847°— 16 - 2

480

Journal of Agricultural Research

Vol. VII, No. 11

LITERATURE CITED

(1) Kirby, William.

1798. History of Tipula tritici, and Ichneumon tipulae with some observations
upon other insects that attend the wheat, in a letter to Thomas Marsham,
Esq. In Trans. Linn. Soc. [London], v. 4, p. 230-239.

(2) Kaltenbach, J. H.

1843. Monographic der Familien Pflanzenlause (Phytophthires). 222 p.,
1 pi. Aachen.

(3) Curtis, John.

1845? Observations on the natural history and economy of various insects, etc.,
affecting the com crops ... X. Cecidomvia tritici — the British
wheat-midge. In Jour. Roy. Agr. Soc., England, v. 6, p. 493-518.

(4) Walker, Francis.

1849. Descriptions of aphides. In Ann. and Mag. Nat. Hist., s. 2, v. 3, no. 13,

P* 43-53*

(5) Koch, C. L.

1857. Die Pfianzenlause Aphiden. 334 p., 54 pi. Niimberg.

(6) Curtis, John.

i860. Farm Insects. 528 p., 69 fig., 16 pi, London.

(7) Buckton, G. B.

1876. Monograph of the British aphides, v. 1, 193 p., 38 pi. London.

(8) Oestlund, O. W.

1887. Synopsis of the Aphididae of Minnesota. In Geol. and Nat. Hist, Survey
Minn. Bui. 4, 100 p.

(9) Riley, C. V.

1890. Report of the entomologist. In U. S. Dept. Agr. Ann. Rpt., 1889,

P* 33I_361* P1- I~6*

(10) Schouteden, H.

1901. Le Genre Siphonophora C. In Ann. Soc, Ent. Belg., v. 45, no. 4,
p. m-117.

(11) Pergande, Theodore.

1904. On some of the aphides affecting grains and grasses of the United States.

In U. S. Dept. Agr. Div. Ent. [Bui.] 44, p. 5-23, fig. 1-4.

(12) Kirkaldy, G. W.

1905. Current notes. In Entomologist, v. 38, no. 504, p. 127-132.

(13) Sanderson, E. D.

1906. Texas notes — II. In Ent. News, v. 17, no. 9, p. 327-328.

(14) Phillips, W. J., and Davis, J. J.

1912. Studies on a new species of Toxoptera, with an analytical key to the
genus and notes on rearing methods. U. S. Dept. Agr. Bur. Ent.
[Bui.] Tech. ser. 25, pt. 1, 16 p., 9 fig., 1 pi.

(15) Webster, F. M., and Phillips, W. J.

1912. The spring grain-aphis or "green bug.” U. S. Dept. Agr. Bur. Ent.
Bui. no, 153 p., 48 fig., 9 pi.

PLATE B

Forms of Macrosiphum granarium:

1. — Mother of males and grandmother of oviparous females.

2. — Typical green viviparous female.

3. — Pupa of male.

4. — Pupa of the mother of oviparous females.

5. — Oviparous female.

Henry Fox, artist.

Macrosiphum granarium

Plate B

Journal of Agricultural Research

Vol. VII, No. 11

PLATE 33

Macrostphum grananum:

A. — Winged viviparous female : a, cornicle.

B* — Winged male.

Plate 33

Journal of Agricultural Research

Vol. VII, No. 11

PLATE 34

Macrosiphum granarium:

A. — Antenna of male.

B. — Antenna of winged viviparous female.

C. — Hind tibia of oviparous female.

D. — Antenna of wingless viviparous female.

E. — Antenna of wingless oviparous female.

F. — Antenna of stem mother

A SPECIFIC MOSAIC DISEASE IN NICOTIANA VISCOSUM
DISTINCT FROM THE MOSAIC DISEASE OF TOBACCO

By H. A. Allard,

Assistant Physiologist, Tobacco and Plant-Nutrition Investigations ,

Bureau of Plant Industry

During the summer of 1 91 5 many plants of Nicotiana viscosum and first-
generation plants of the cross N . tabacum 9 X N. viscosum were grown
in the field at Arlington, Va. Late in the season three plants of N.
viscosum and one of the hybrid plants showed unmistakable symptoms
of a typical mosaic disease. From the fact that the species viscosum
and its hybrids had never before shown symptoms of disease from inocu¬
lations made with the virus of the ordinary mosaic disease of tobacco,
these affected plants were taken into the greenhouse for further study.
It has now been established that this mosaic disease affecting N. viscosum
and its hybrids is biologically very different from the ordinary form of
mosaic disease affecting varieties of N. tabacum , tomatoes (. Ly coper sicon
esculentum ), etc. Ordinary tobacco and also tomatoes appear to be quite
immune from the type of mosaic disease in N. viscosum . Experiments
have shown that this mosaic disease is infectious to plants of N. viscosum ,
although it appears that the disease is not as readily transferred by needle
inoculations as the ordinary form of the mosaic disease, and longer periods
of time are usually required before the disease comes into evidence.

A number of distinct varieties of N. tabacum have been crossed with
N. viscosum , including Maryland Mammoth, White Burley, and Connecti¬
cut Broadleaf. In these crosses the pollen of N. viscosum has been trans¬
ferred to the pistils of N. tabacum. In size, general appearance, and habit
of growth first-generation plants of these crosses resemble much more
closely the female parent ( N . tabacum) than the male parent ( N . vis¬
cosum) . In general appearance the leaves and blossoms also resemble very
closely the leaves and blossoms of the female parent. These first-genera¬
tion plants inherit more strongly the visible physical characteristics of
the female parent. They possess, however, certain physiological character¬
istics peculiar to the male parent (N. viscosum). This is indicated by the
fact that they, like N. viscosum , appear to be immune to that form of
mosaic disease which affects varieties of N. tabacum , but are susceptible
to the mosaic disease affecting N. viscosum. The disease is readily ob¬
tained in these hybrids by grafting upon them scions taken from plants of
N. viscosum. It is much more difficult to obtain the disease by needle
inoculations. All phases of catacorolla in the blossoms and mottling and
distortions in the leaves are shown in these hybrids affected with the

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.
gl

Vol. VII, No. n
Dec. 11, 1916
G — 101

482

Journal of Agricultural Research

Vol. VII. No. 1%

mosaic disease of N . viscosum as in ordinary tobacco plants affected with
the common form of the mosaic disease (Pis. 35 and 36.) The mosaic
disease of N, viscosum produces more or less mottling and distortion in
the blossoms of these plants. The abnormality known as catacorolla,
however, has never appeared in connection with the disease.

Table I. — Inoculations made with the expressed sap of scions of N. viscosum grafted
upon ordinary tobacco ( N . tabacum)

Number of
plants
inoculated
(Connecticut
Broadleaf).

Date of
inocula¬
tions.

Material used.

Symptoms of
mosaic disease
in scion of
N. viscosum.

Results.

1915-
Dec. 16

10 16

Sap of scion of N. viscosum
grafted on mosaic stock of N.
tabacum several weeks.

_ do .

Tap water (control) .

None

All healthy.

3 mosaic.

All healthy.

1916.
Jan. 5

Sap of scion of N. viscosum grafted
upon mosaic stock of N. taba¬
cum till scion was in bloom.

Sap of stock upon which above
scion was grafted (symptoms
in stock severe).

Sap of scion of N. viscosum
grafted upon mosaic stock of
N. tabacum till scion was in
bloom.

Tap water (control) . . . .

None

None .

Sap of scion of AT. viscosum

None .

grafted on stock of N. tabacum
several weeks.

Tap water (control). . . . .

Sap of scion of N. viscosum

None .

grafted several weeks on stock
of N. tabacum .

. do .

... do .

. . . . .do .

Sap of scion of N. viscosum

. . .do .

grafted upon stock of N , taba¬
cum till scion was in bloom.

. do .

. . .do .

..... do .

. . .do .

. do . . .

... do .

. do. . . .

. . .do .

Sap of scion of N. viscosum

... do .

grafted upon stock of N . taba¬
cum several weeks.

Tap water and healthy sap .

7 mosaic.
10 mosaic,
2 mosaic.

All healthy.
Do.

Do.

Although the species N. viscosum is susceptible to a mosaic disease
peculiar to itself, this species of Nicotiana appears to be immune to the
ordinary form of mosaic disease affecting N. tabacum . Likewise, first-
generation plants of the cross N. tabacum 9 X N. viscosum <? appear to
be quite as immune from the disease as the species N. viscosum . All

Dec. ii, 1916 Specific Mosaic Disease in Nicotiana viscosum

483

methods of inoculation which have been attempted with these plants
have been without success. It has been shown that the virus was not
present in these plants by extracting the sap of all parts of the plants
and testing its infectivity by making inoculations into young tobacco
plants. These inoculations have never produced infection. Further¬
more, many successful grafts have been made between N. tabacum and
N. viscosum , using N. tabacum as the stock. As soon as the N. viscosum
scion had started to grow, the stock (N. tabacum) was inoculated with
the ordinary form of the mosaic disease. Scions of N. viscosum in many
instances remained upon the mosaic stocks for many weeks and finally
blossomed, yet symptoms of the mosaic disease never appeared in the
blossoms or leaves. In all instances inoculation tests have been made
to determine if the infective principle of the disease was present in the sap
of the immune scions. As shown in Table I, these scions in many in¬
stances appeared to be entirely free from infection. In other instances
the sap proved to be more or less infectious to tobacco plants. Why the
sap of the scion should carry the infective principle at one time and not
at another can not at present be explained.

The mosaic disease affecting N. viscosum appears to be identical in
all its symptoms with the mosaic disease of tobacco (N. tabacum ). The
virus of the disease, however, has behaved very differently from the virus
of the mosaic disease of tobacco in all inoculation tests. With the
exception of Datura fastuosa (Golden Queen variety), and Datura stra¬
monium , no other plants of the solanaceous family have been found
susceptible to the virus of the mosaic disease affecting N. viscosum .
Although peppers and tomatoes are very susceptible to the virus of the
mosaic disease of tobacco, these plants appear to be immune from the
virus of the mosaic disease affecting N. viscosum , or at least highly
resistant to it, since the most persistent and rigorous needle inoculations
have failed to produce infection. The most rigorous methods of inocu¬
lation have also failed to produce either the mosaic disease of tobacco
or the mosaic disease of N. viscosum in the Irish potato (Solanum
tuberosum) .

Datura stramonium is the only solanaceous plant which has given
evidence of being susceptible to both mosaic diseases. Inoculations
made at different times with different lots of virus producing the mosaic
disease in N. tabacum have given very different results. In some tests
the plants were highly resistant to infection. In other tests similar
methods of inoculation gave a high percentage of mosaic-diseased plants.
It has not been determined whether this variability indicates differences
in the infective properties of the virus or differences in the relative
resistance of different lots of plants. In one experiment 18 young
vigorous plants of Datura stramonium were divided into two lots of 9
plants each. One lot was inoculated at many points in the stems and
leaves with the virus of the mosaic disease of N. viscosum . The remain-

4«4

Journal of Agricultural Research

Vol. VII, No. ii

in g 9 plants were inoculated in the same manner with the virus of the
mosaic disease of N. tabacum . For a period of several weeks numerous
inoculations were made from time to time in each lot of plants. The
plants of each lot were also cut back severely several times and the virus
inoculated into all cut surfaces. The plants were kept under observa¬
tion for several months. Every plant in the series inoculated with the
virus of the mosaic disease of N. viscosum developed the disease, the
first observable symptoms appearing 21 days after the first inoculation.
In this experiment the datura plants proved to be highly resistant to the
virus of the mosaic disease of N. tabacum , as none became diseased. In
those plants affected with the mosaic disease of N. viscosum the symptoms
were particularly malignant. The leaves became greatly curled, wrinkled,
and depauperate. Mottling of the leaves, however, was less marked
than in those instances where Datura stramonium has been affected with
the mosaic disease of N> tabacum .

The virus of the mosaic disease affecting N. viscosum differs from the
virus of the mosaic disease of tobacco as follows :

CHARACTERISTICS OR THE VIRUS OP THE

MOSAIC DISEASE OP TOBACCO (n.

tabacum)

(1) Transmission through the seed has
never occurred.

(2) Incubation period short (minimum
6 days).

(3) Needle inoculations readily pro¬
duce the disease.

(4) All attempts to infect belladonna
(Atropa belladonna) and Solanum tubero¬
sum have been unsuccessful.

, (5) All attempts to infect poke weed
(. Phytolacca decandra) have been unsuc¬
cessful.

(6) All attempts to infect the hybrid
N. tabacum 9 X N. viscosum <£ have been
unsuccessful.

(7) Highly infectious to tomatoes.

(8) Infectious to the pepper ( Capsicum
cerasiforme).

(9) All attempts to infect sweet peas
have been unsuccessful.

(10) All attempts to infect Datura fas-
tuosa (Golden Queen variety) have been
unsuccessful.

(11) Affects Jimson weed ( Datura stra¬
monium) producing typical symptoms.

characteristics of the virus op the

MOSAIC DISEASE AFFECTING N. VIS¬
COSUM

(1) Transmission through the seed has
never occurred.

(2) Incubation period in N. viscosum
rather long (minimum may be several
weeks) .

(3) Needle inoculation rather uncer¬
tain. Grafts of mosaic-diseased shoots of
N. viscosum upon susceptible plants
readily produce infection.

(4) All attempts to infect belladonna
and Solanum tuberosum have been un¬
successful.

(5) All attempts to infect poke weed
have been unsuccessful.

(6) The hybrid N. tabacum 9 X N.
viscosum $ is susceptible, manifesting
typical symptoms of the disease.

(7) All attempts to infect tomatoes have
been unsuccessful.

(8) All attempts to inoculate the pepper
have been unsuccessful.

(9) All attempts to infect sweet peas
have been unsuccessful.

(10) Datura fastuosa (Golden Queen
variety) is susceptible, manifesting symp¬
toms more or less typical of the disease.

(11) Affects Jimson weed, producing
symptoms very similar to those produced

Dec. it, 1916 Specific Mosaic Disease in Nicotiana viscosum

485

Jimson weed, however, sometimes shows
considerable resistance to the mosaic dis¬
ease affecting N. tabacum .

(12) Highly infectious and particularly
malignant to N. rustica.

by the ordinary mosaic disease of
tobacco.

(12) All attempts to infect N. rustica
have been unsuccessful.

The writer is of the opinion that this distinctive type of mosaic disease
affecting N. viscosum has in some manner originated from the ordinary
form of mosaic disease, possibly through the agency of insect transmission
in the field. This does not seem improbable, since practically every sus¬
ceptible plant in a half -acre field of* ordinary tobacco in which the N..
viscosum plants were grown became mosaic; and throughout the season
both species were infested with great numbers of flea beetles. It is
possible that insects may become efficient transmitters of disease where
ordinary methods of artificial inoculation fail.

During the same season the writer's attention was called to the occur¬
rence of typical symptoms of the mosaic disease in peppers grown in a
field near by. To all outward appearances the plants were affected with
a severe mosaic disease which gradually spread over the field and per¬
sisted in all affected plants. Tomato plants in adjoining rows, however,
were unaffected. The expressed sap from the most severely attacked
pepper plants failed to produce the mosaic disease in young tobacco
plants (N. tabacum). Whether this mosaic disease was infectious to
healthy pepper plants or might have been in any way related to the
mosaic disease affecting N . viscosum was not determined.

In this connection it is interesting to note that various European in¬
vestigators have reported that they were unable to inoculate other
species of solanaceous plants with the virus of the mosaic disease of
tobacco with which they worked. Thus, Mayer 1 failed to produce the
disease in other solanaceous plants.

Iwanowski2 has stated that the mosaic disease of tobacco does not
occur upon Datura stramonium or Hyoscyamus niger .

Iwanowski,3 in a later publication, stated that he had never known
Nicotiana rustica to be affected by the mosaic disease.

Koning 4 also failed to communicate the mosaic disease of tobacco to
Datura stramonium , Hyoscyamus niger , Solanum tuberosum , and Petunia
nyctagini folia.

Westerdijk,5 working with a mosaic disease which was infectious to
tomatoes, reported that she could not communicate this disease to

1 Mayer, Adolf. Ueber die Mosaikkrankheit des Tabaks. In Landw. Vers. Stat., Bd. 32, p. 450-467,

pi. 3. 1886.

2 Iwanowski, D, fiber die Mosaikkrankheit der Tabakspflanze. In Bui. Acad. Imp. Sci. St. Petersb.,
n. s. v. 3 (v. 35)1 *10. 1. P- 67-70. 1892.

3 Iwanowski, D. fiber die Mosaikkrankheit der Tabakspflanze. In Centbl. Bakt. [etc.] Abt. 2, Bd. 5,
No. 8, p. 250^254, 2 fig. 1899-

4 Koning, C. J. Der Tabak ... p. 71-86, fig. 13-15. Amsterdam, 1900.

3 Westerdijk, Johanna. Die Mosaikkrankheit der Tomaten. 19 p., 3. pi. Amsterdam, 1910. [Meded.
Phytopath. Lab. ** Wille Commelin Scholten.” Amsterdam.]

486

Journal of Agricultural Research

VoX. VII, No. IX

tobacco. Likewise, she could not infect tomato plants with the sap
of a mosaic disease of tobacco with which she worked.

The constancy of these negative results is rather striking. It is
possible that the type of mosaic disease with which European investi¬
gators worked may not have been quite so readily communicable to
plants of other species and genera of the solanaceous family as the type
in the writer’s possession. It has been more or less generally believed
in Europe that N. rustica was even immune to the mosaic disease affecting
tobacco. In the writer’s experience the virus of the common form of the
mosaic disease is not only very infectious but particularly malignant to
plants of N. rustica . Likewise, the disease is readily communicable to
all the more distinct varieties of tomatoes, petunia, Datura stramonium ,
and is highly infectious to Hyoscyamus niger.

PLATE 35

Leaves of Nicoiiana viscosum affected with the mosaic disease. This mosaic disease
does not affect ordinary tobacco ( N . tabacum); nor does the mosaic disease affecting
ordinary tobacco affect N. viscosum.

PLATE 36

A. — Normal blossoms from healthy plants of Nicotiana viscosum.

B. — Depauperate blossoms from mosaic plants affected with the mosaic disease
peculiar to N. viscosum. This disease is distinct from the ordinary form of the mosaic
disease affecting varieties of N. tabacum and does not affect them.

C. D.— Blossoms showing catacorolla, etc,, as a result of the mosaic disease affecting
Nicotiana viscosum. These are from first-generation plants of the cross Connecticut
Broadleaf tobacco $ XN. viscosum S . This hybrid appears to be immune from the
ordinary mosaic disease affecting the female parent, but is susceptible to the mosaic
disease affecting the male parent, N. viscosum. Although this mosaic disease has never
produced instances of catacorolla in N. viscosum , all phases of catacorolla are produced
in the hybrid. Catacorolla is a common malformation in varieties of N. tabacum as a
result of the ordinary form of the mosaic disease.

SYNTOMASPIS DRUPARUM, THE APPLE-SEED CHALCID

By R. A. Cushman,

Entomological Assistant , Deciduous Fruit Insect Investigations , Bureau of Entomology

INTRODUCTION

Since the publication by Crosby (7 *, p. 369) of his paper on the apple-
seed chalcid ( Syntomaspis druparum Boh.) this insect has attracted more
and more attention among those associated with the apple industry, and
numerous letters relating to it have been received at the Bureau of Ento¬
mology. The frequency and wide distribution of these inquiries and com¬
plaints seemed to warrant a rather detailed investigation of the insect,
and the writer has spent portions of the past two seasons (1914 and 1915)
in such an investigation. The biological work was done at the field
laboratory for the investigation of deciduous-fruit insects of the Bureau
of Entomology at North East, Pa., while the field observations have been
conducted throughout the northern tier of States from Vermont to
Michigan.

DESCRIPTION OF THE ADULT INSECT

The adult insect is somewhat wasplike in appearance, bright green,
with coppery or bronzy metallic reflections, brownish yellow legs, and
clear hyalin wings. The female (PI. 37, A) is normally about 4 mm. in
length and is provided with a slender ovipositor slightly longer than the
body. The male (PI. 37, B) is somewhat smaller than the female.

DISTRIBUTION IN THE UNITED STATES

The apple-seed chalcid apparently occurs throughout the northern tier
of States, at least from Vermont to Michigan. It has not been found in
Ohio or Indiana. At the time of the writer's visit to those States there
was a very small crop of apples, and none especially suitable for the attack
of the insect were found. But in the same season the chalcid was found
in the seeds of a wild seedling at Benton Harbor, Mich. The writer has also
found it as far south as Clearfield, Pa., and some years earlier what was
almost undoubtedly the larva of this species was found in a crab apple
(Malus sp.) at Vienna, Va. It is probably distributed throughout the
eastern part of the country wherever small seedling apples (Malus sylvestris)
are to be found.

HISTORICAL REVIEW

Crosby (7, p. 369; 9) has given a nearly complete r£sum£ of the his¬
tory of the apple-seed chalcid in Europe, where it has been well treated

1 Reference is made by number to “literature cited," p. 501.

Journal of Agricultural Research, Vol. VII, No. n

Dept, of Agriculture, Washington, D. C. Dec. ir, 1916

go K— 47

66847°— 16 - 3

(487)

488

Journal of Agricultural Research

Vol. VII, No. ir

recently by Mokrzecki (4) . His own papers record the only original
observation^ on the species in America. The insect was first discovered
by Prof. Crosby in July, 1906, at Ithaca, N. Y., when he found the seeds
of crab apples to contain the partly grown larvae. His first report of his
discovery appeared in 1908 (6, p. 38), and the following year he published
his full account (7, p. 369). In the latter paper he summarizes most
of the previously published accounts of the species and records in detail
his own observations in regard to life history, habits, distribution, and
host fruits and gives descriptions of the stages. His 1912 paper (9)
consists of further r6sum£s of European literature.

From the wide distribution of the species it is evident that it must have
been present though undiscovered in America for a long time, but any
statement as to the time of its introduction can be nothing more than
speculation. However, that there have been many opportunities for its
introduction in the past and that it has been repeatedly introduced in
fruit from Europe can not be doubted. It may even have been brought
to America before its discovery in Europe, and its establishment here may
have been effected at that early time; for it is a historical fact that in
the early days of American history apples were imported and their seeds
planted by the colonists. Much of the early spread of the apple to the
West was due to the Indians, who planted in favorable spots the seeds
from apples given them by the settlers. These trees, planted mostly
along the trails to the West, would form easy avenues of distribution,
and it is quite likely 'that they and their progeny have aided in the
spread of the insect.

EFFECT UPON FRUIT

The only externally visible effect of infestation is caused by the op¬
position puncture, which, after a few days, appears as a minute scar
situated in a small, shallow dimple. From this scar to the seed extends
a discolored line. Under ordinary circumstances of growth and infes¬
tation the fruit apparently is able to outgrow both of these manifesta¬
tions of injury. But occasionally, especially when fruit is scarce or the
insects very abundant, the gross injury due to repeated puncturing at
nearly the same spot causes permanent and deep dimpling, together with
corky, discolored streaks in the flesh. However, even in 1915, when suit¬
able fruit was rather scarce, the season cold, and the chaldds abundant
in the region of North East, Pa., such injury was the exception rather than
the rule; and in 1914, when the converse of these conditions prevailed,
no single case of severe distortion that could be attributed to this species
was found. Distorted fruit is shown in Plate 38, A, B, C.

Frequently injury caused by other insects is attributed to
the apple-seed chalcid because at the time the injury is noted this
species is the only one present. As an example of this, the case of an
orchard near Clearfield, Pa., may be cited. The bulk of the fruit in this

Dec. ix, 1916

Syntomaspis druparum

489

orchard in 1914 was very badly distorted, and specimens sent to various
entomologists were pronounced to be the work of Syntomaspis druparum
for the reason that larvae of this species were found in the seed. The
writer visited this orchard in October and examined large numbers of
the fruits, but found the chalcid larvae in comparatively few. Obviously
the chalcid was not responsible for such extensive injury, especially in
view of the fact that the fruit of wild seedlings almost within the boun¬
daries of the orchard was heavily infested by the chalcid and showed
no sign of distortion. Observation in the same orchard the following
spring disclosed the fact that it was grossly infested by both species of
apple red bugs (Lygidea mendax Reut. and Heterocordylus malinus Reut.),
which had come from Crataegus sp. and wild crab in the surrounding
woods. These were the insects responsible for the injury to the apples,
and the chalcids were able to infest the seeds because of the stunting
due to the red-bug injury. It should be stated, in justification of this
mistaken determination, that both of the insects concerned are of com¬
paratively recent discovery, and their work is familiar to but few ento¬
mologists.

When first infested, the seeds show the laceration caused by the ovi¬
positor surrounded by a brownish area; but as they darken, the injured
area heals and ultimately appears as a lighter area, a repeatedly punc¬
tured seed having a mottled appearance. At full growth infested seeds
are less plump and more irregular than normal seeds. Infested and
sound seeds are shown in Plate 39.

Crosby (7, p. 369) states that in the Lady apple the texture of the
flesh is considerably injured. This has not been apparent to the writer,
for. on visiting an orchard containing trees of this variety, from which
the owner had picked what he termed an “ unusually fine crop,” fully
two-thirds of the apples examined were heavily infested by the chalcid;
but it was impossible to tell whether an apple was infested without ex¬
amining the seeds or making an almost microscopic examination of the
skin for the minute oviposition scars. Special attention was paid to
apples of commercial size and color,, and a very large percentage was
.(found infested. Moreover, fruit of this variety has been purchased on
the Washington market 50 per cent of the seeds of which contained
larvae of the chalcid.

Horvath records failure in Budapest of apple seed to produce a good
stand on account of infestation by the chalcid.

VARIETIES AND SPECIES OF FRUIT ATTACKED

The apple-seed chalcid has been found to infest a great variety of
fruits. The original description (1, p. 361-362) was based on specimens
reared from the seeds of Sorbus scandica. The species was redescribed
by Thomson (2, p. 76) from seeds of Sorbus sp. Rodzianko (5, p. 593-
602) reared it from Sorbus aria , Pyrus baccata , and Malus sylvestris . In

490

Journal of Agricultural Research

Vol. VII, No. ii

Europe it has been mentioned a number of times in connection with the
apple, but frequently without any statement as to the nature of the fruit.
Porchinsky (3) records the rearing of a species of Torymus from the
seeds of wild pear ( Pyrus communis ), but gives no specific determi¬
nation of the insect. It may have been Syntomaspis druparum , but not
certainly so. Crosby (7, p. 369) lists the Lady apple, natural fruit, the
wild crab ( Pyrus \Malus] coronaria ), and the following cultivated crab
apples: Pyrus [Mains'] sibricia var. striata , Pyrus [Malus] fioribunda ,
Pyrus [Malus] prunifoliae , and Pyrus [Malus] ioensis. He also states
that larvae, apparently the same, were found in the seeds of Sorbus lati-
folia , but that the adults were not reared. In correspondence with the
Bureau of Entomology Mr. M. E. Benn, of Coudersport, Pa., states that
he has found infestation by the seed chalcid in Northern Spy, Baldwin,
Fameuse, Wagener, Russet, Tolman Sweet, and two seedlings. Mr. G.
McE. Stevens, of Orwell, Vt., reported it as attacking Lady apples at
Orwell, Vt., and natural fruit at Peru, N. Y., while Mr. A. E. Stene re¬
ports it from Kingston, R. I., in the seeds of crab apple.

The writer’s observations on the species began a number of years ago
at Vienna, Va., where what was undoubtedly the larva of the seed chalcid
was found in a seed of a crab apple.

Since the beginning of the work on the species, many varieties of apples
have been examined under many conditions and in widely separated
localities. At practically every point visited nearly every variety of
natural fruit, except the largest; has been found to be more or less gen¬
erally infested.

Among cultivated varieties the Lady apple only is apparently subject
to very serious attack, this variety being frequently very heavily in¬
fested. The ordinary commercial varieties are never infested except in
neglected and run-down orchards or when fruit is stunted by the over¬
loading of trees or by the attack of some other insect or disease. The
reason for the immunity of the ordinary apples of commerce from attack
is purely mechanical, in that, at the time the chalcids are ovipositing, such
fruit is so large that the ovipositor will not reach to the seeds. How¬
ever, under the circumstances enumerated above, such varieties are occa$
siorially more or less infested, though never very heavily so. Larvae have
been found by the writer in neglected orchards at North East, Pa., in
the following varieties: French Russet, Northern Spy, and Baldwin.
In a large orchard near Clearfield, Pa., which in 1914 was very badly in¬
fested by red bugs (Lygidea mendax Reut. and Heterocordylus malinus
Reut.), and the fruit much distorted and stunted thereby, only Grimes
Golden, Ben Davis, and Missouri of the many varieties examined were
infested. Of these Grimes Golden showed about 25 per cent of the fruit
infested, from one to four seeds in the infested apples containing larvae
of the chalcid. Of the two other varieties only one apple each was found
to be infested.

Dec. ii, 1916

Syntomaspis druparum

491

Crab apples, both cultivated and wild, are very frequently infested,
but invariably to a less extent than the small wild seedlings of the true
apple, and it is evident that the latter is the natural host of the insect.
On one occasion the opportunity offered to compare the infestation in
these two classes of fruit where wild crabs and wild seedlings were found
growing side by side. Only about 50 per cent of the crabs were infested,
and rarely more than one seed to the fruit contained larvae, while the
infestation in the seedling apples was practically 100 per cent, unin¬
fested seeds being scarce.

Although the fruit of the common mountain ash ( Sorbus americana)
has been repeatedly and extensively examined, the writer has never
found any trace of infestation by this or any other chalcid.

Neither pears nor the fruit of Crataegus sp. exposed to the attack of
the chalcid in cages were infested, although attempts at oviposition on
the latter were repeatedly observed and many fruits were exposed to
attack and later examined.

LIFE HISTORY OF THE CHALCID

The life-history data given below were obtained very largely by propa¬
gation of the apple-seed chalcid on wild seedling fruit at North East, Pa.,
and involved the examination of many hundreds of apples. The female
insects were caged on fruit for one day in cages constructed of mica
lamp chimneys and cheesecloth (PI. 40, D).

EMERGENCE IN SPRING

The insects reared in 1914 were from a lot of apples that had been kept
in Washington, D. C., during the previous winter and until about May
15, when they were shipped to North East, Pa. The earlier spring of.
Washington undoubtedly hastened somewhat the emergence of some of
the earlier reared adults, for they began to emerge from the seeds on May
26, which was some time before the apples at North East were at the
proper stage for oviposition. However, they did not begin to appear in
numbers until after the middle of June, the heaviest emergence occurring
during the week of June 22-29 anc^ the last on July 5. The adults
reared in 1915 were from apples that passed the winter in an unprotected
wire cage at North East. The first to emerge appeared on June 16 and
the last on July 16, with the heaviest emergence, as in 1914, during about
the last week of June.

There appears to be very little, if any, difference in the time of emerg¬
ence of the sexes. In the more normal emergence of 1915 a few males
appeared before any females, and the few belated individuals that emerged
after the first few days of July were all females. But during every other
day of the emergence season some individuals of each sex appeared.

492

Journal of Agricultural Research

Vol. VII, No. ii

RELATIVE ABUNDANCE OF SEXES

The rearings during 1914 consisted of 254 females and 85 males, 74.9
and 25.1 per cent, respectively. In 1915, 316 females and 100 males
were reared, 75.9 and 24.1 per cent, respectively. These figures show a
ratio of about 3 females to each male.

oviposition

Age at beginning. — The female chalcids become mature and able to
deposit eggs within a very short time after emergence, for they have
been repeatedly observed in the act of oviposition within two days after
issuing from the seed.

Age of fruit. — At the time of the heaviest emergence of the chalcids
apples have grown, depending on the variety, to a diameter of from a
half inch to somewhat over an inch. The seeds have attained nearly full
growth, but have not begun to harden. Most of the space within the
seed is occupied by a jelly-like mass, with the small embryo at one end.
Between this and the outer seed coat is a rather thick mucilaginous layer.

Method and time required. — In ovipositing, the female chalcid first
feels carefully over the surface of the apple with her antennae; then,
when she has located a place to her liking, she raises the abdomen, at the
same time releasing the ovipositor from its sheath and lowering it
until its tip is against the surface of the apple directly beneath the pos¬
terior end of the thorax. The abdomen is now perpendicular to its
normal axis. With pressure accompanied by a slow swinging of the abdo¬
men from side to side the ovipositor is forced slowly into the apple until
inserted to its full length. At the end of this time the abdomen has
resumed nearly its normal position except that the hypopygidium is
directed downward with the ovipositor, making a triangular projection
below the abdomen. Now the ovipositor is several times partially with¬
drawn and thrust back until the insect is apparently satisfied that it has
been properly inserted, when she remains perfectly quiet for a consider¬
able period, during which the egg is deposited. When this is finished the
ovipositor is withdrawn and swung back into its sheath. The whole
process occupies, on the average, somewhat in excess of five minutes.
Living chalcids in various phases of the act of oviposition are shown in
Plate 40, B, C, while in A one is shown attempting oviposition in the
fruit of Crataegus sp.

Point of attack. — When oviposition first begins, most of the punc¬
tures are made around the middle of the apple, but later in the season the
attack is shifted nearer to the calyx end. This is apparently made
necessary by the fact that the growth of the apples makes it impossible
for the ovipositor to reach the seed from the side. Figure 1 shows the
position of punctures in fruit and seed.

Dec. ii, 1916

Syntomaspis druparum

493

Relation between punctures made and eggs deposited. — That
the instinct of the ovipositing female in locating the seed is not so strong
and unerring as might at first be supposed, when the frequent very high
percentage of infestation of the seeds is considered, is indicated by the
number of punctures made in the seed compared with the number show¬
ing on the surface of the apple. The puncture is much more conspicuous
on the white seed than on the skin of the apple; yet one fruit that had
been punctured 36 times had only five punctures on its seeds. It is unlikely
that each puncture made represents an egg deposited, but rather that
many punctures represent unsuccessful attempts at finding seed. This
is borne out by the observations on ovipositing females, which frequently
inserted their ovipositors repeatedly at almost the same point before
ultimately going through all the
phases of the act of oviposition. It
is not even probable that every punc¬
ture in the seed represents the depo¬
sition of an egg. No definite assertion
on this point can be made, since the
eggs are rather difficult to find.

Place of deposition of egg. —

Apparently it is the aim of the insect
to place its egg in the central gelat¬
inous mass of the seed, and from the
position of many of the punctures it
is impossible that through them the
ovipositor could have reached this
body. Many punctures are on the
side of the seed, in such position that the ovipositor must have been
nearly tangent to the surface of the seed.

Sometimes eggs are deposited in the mucilaginous layer next to the
seed coat, but the resulting larvae apparently never mature, for many
dead larvae of the first instar have been found in this situation, and living
larvae found there have always been in the first instar and far behind, in
growth, the larvae of the same age in the more favorable jelly-like body.

Oviposition period. — The longest period during which any of the
caged females were ovipositing in 1914 was from June 25 to July 21, a
period of 26 days, two insects in the same cage having died on the same
date. During this time 48 apples were exposed, and all were more or less
infested. Others lived for periods ranging from 3 to 24 days, the quicker
deaths being due apparently to the sun striking the cages.

EGG

Description. — The egg (fig. 2) is elongate oval, roundly pointed at
the caudal end, and prolonged at the cephalic end into a slender, twisted
pedicle about one-fourth the diameter and nearly as long as the body

Fig. i. — Syntomaspis druparum: Apple, natural
size, and seed, enlarged, showing oviposition
punctures. (Original.)

494

Journal of Agricultural Research

Vol. VII, No. «

of the egg. Exclusive of this appendage, the egg is about 0.55 mm. in
length by about a fourth as thick in the middle. It is yellowish white
and without sculpture. Within two days after oviposition the embryo
can be seen to have drawn away from the poles,
and in some the cephalic constriction can be seen.

Incubation period. — In 1914 the eggs began
to hatch on the sixth day after deposition, and by
the eighth day all had hatched. In 1915 hatching
commenced on the seventh day and continued until
the tenth day.

LARVA

Number and description op instars. — The
Pig. 2.—s yntomaspis dmpa - newly hatched larva (fig. 3) is about 0.4 mm. in
nifild^^origSln^ ma8: *engtli by about a fourth as thick at the thickest
point, which is at the junction of the thoracic
and abdominal segments. From this point it tapers in both directions,
but is much smaller at the caudal end. The body, including the head,
consists of 14 segments; the 3 thoracic segments are about equal in
length, and the abdominal segments gradually decrease in length toward
the caudal end. The head is nearly hem¬
ispherical and rather heavily chitinized.

The mouth opening is nearly circular and
surrounded by a raised rim. Owing to the
minute- size and delicacy of the mouth
parts, except the strong mandibles, it is
difficult to determine definitely their ex¬
act relation to each other, but they appear
to be about as in the illustration (fig. 4) . The mandibles are long, strongly
curved, and dark colored. They cross in the middle of the mouth opening.
The head is from 9.108 to 0.123 mm. in breadth and the mandibles
0.021 mm. in length. At full growth the larva of the first instar is

slightly less than 1 mm. in length.

The larva of the second instar is very
similar in general appearance to that of
the first instar, having the same tapering
form though being somewhat stouter. It
can, however, be easily distinguished by
the weaker chitinization of . the oral region
and the change in the form of the man¬
dibles, which at this molt assume a form
more similar to those of the full-grown
larva. The brown color of the mandibles is confined to their tips,
and they are very stout at the base and much less strongly curved
(fig. 5, a). The head in this instar is from 0.184 to 0.215 broad,

Pig. 3. — Syntomaspis dfuparunt: Newly
hatched larva. Highly magnified.
(Original.)

Fig. 4. — Syntomaspis druparum:

Mouth parts of larva of first instar.
a.Labrum; 6, mandible; c, maxilla;
d, maxillary palpus; e, labium; /,
labial palpus. Highly magnified.
(Original.)

Dec. zx, 19x6

Syntomaspis druparum

495

and the mandibles from 0.036 to 0.039 mm* long. At full growth the
second instar is about 1.5 mm. long.

With each succeeding molt the larva becomes gradually stouter and
less tapering behind until at full growth it is more than a third as thick
as long, with the caudal end but little more tapering than the head end,
and the head becomes relatively smaller and more retracted within the
thorax. With each molt the
head and mandibles increase
markedly in size, and the lat¬
ter change somewhat in form.

These changes in measure¬
ments and form constitute
the only real differentiating
characters until in the last
instar the spiracles become
open and visible.

The head of the third-instar
larva varies in breadth from
0.277 0.308 mm. and the

length of the mandibles from
0.050 to 0.057 mm- The lat¬
ter (fig. 5, b) are curved to¬
ward the apex. In the fourth
instar the head is from 0.415
to 0.461 mm. broad and the
mandibles (fig. 5, c) 0.079 to 0.086 mm. long and nearly straight at the
apex.

The full-grown or fifth-instar larva (fig. 6) is of the typical chalcid form,
rather spindle-shaped but somewhat curved toward the ventral side,
with the head short and flattened and partially retracted within the first

thoracic segment. The mesothoracic
and metathoracic segments and the
first seven segments of the abdomen
bears each a pair of minute spiracles.
Fully fed larvae are mostly from 4.5
to 5 mm. long, but a few, which
develop in small seeds, are much
smaller, the smallest measured being
3 mm. in length. The head is from
0.554 °-6 mm- broad and the mandible (fig. 5, d), the blade of which

is rather slender and nearly straight, is from o.ni to 0.129 long.
The arrangement of the mouth parts is shown in figure 7.

Place of feeding. — The earlier feeding of the larva is done in the
gelatinous portion of the developing seed. In the meantime the embryo
is developing and the gelatinous body is being absorbed. Before the

Fig. Syntomaspis druparum: Mandibles of larvae of vari¬
ous instars. o, Second; 6, third; c, fourth; d, fifth. Highly
magnified. (Original,)

Fig. — Syntomaspis druparum: Full-grown larva.
Much enlarged . ( Original . )

496

Journal of Agricultural Research

Vol. VII, No. ii

latter entirely disappears the larva begins to feed on the cotyledons,
eating out a pit on one edge or one side and ultimately devouring the
last traces of the embryo.

Duration of instars and feeding period. — The two seasons during
which observations on the life history of the apple-seed chalcid were
made were quite different, and the development of the larvae was con¬
sequently quite different in point of time required. The summer of 1915
was unusually cold, and the larvae required about a week longer to com¬
plete their development than those of 1914.

The observations of 1914 were complicated and rendered somewhat diffi¬
cult of interpretation because of an unexpected infestation from natural

sources. The presence of the species
in the locality was not discovered un¬
til too late to escape the infestation
from that source. However, the nat¬
ural infestation was, as a rule, either
considerably earlier or considerably
later than that in the cages, and some
information of value can be secured
from the data obtained.

In 1915 these difficulties were elim¬
inated by the expedient of bagging
all fruit to be used in the work both before and after it was exposed to
the attack of the insects in the cages.

In 1914 the first individual of each instar was found about four days
after the first of the immediately preceding instar, all transformations
from one instar to the next taking place within a period of about a
week. In about 45 days from oviposition practically all the larvae had
finished feeding. Table I gives the data on the development of the
seed chalcid during 1914. From this table are excluded all individuals
the presence of which was obviously due to natural infestation.

Fig. 7 .—Sytiiomaspis druparum: Mouth parts of
full-grown larva, c, Labrum; b, mandible; c,
maxilla; d, maxillary palpus; e, labium; /, la¬
bial palpus. Highly magnified. (Original.)

Table I. — Development of Syntomaspis druparum at North East, Pa., in IQ14

Period from infestation to
examination.

Date of
infestation.

Date of
examina¬
tion.

Stages of insect found.

Eggs.

First Second
larval, larval.

Third

larval.

Fourth

larval.

Fifth

larval.

Days.

3—4 . , . „ „ .

June 20-21..
July 3-4 . . . .
June 18-19. •
June 25 ..

June 24
July 8
June 24
July 1
July 8
July is
July 1
...do .

4
17

4-5 .

S-6 .

6 .

2 . .

7 .

July a .

July 9 .

O * . ■

A .

July 1 . ...

*T * * * * * * * *

2 .

8 . . .

jime 23 .

J * * . .

6 .

8-9 .

July 7 .

July is
July 8
July 15
July 16

12 ...

June 29-30. .
July 6 .

3 . .

July 7 .

' 4 .

Dec. ii, 1916

Syntomaspis druparum

497

Table I. — Development of Syntomapsis druparum at North East , Pa., in IQ14 — Contd.

Date of

Stages of insect found.

Period from infestation to
examination.

Date of
infestation.

examina¬

tion.

Eggs.

First

larval.

Second

larval.

Third

larval.

Fourth

larval.

Fifth

larval.

Days.

8-10 .

July 12-14. •

July 6 .

June 18-19. .
June 27-28. .
July 5 .

July 22
July 16
June 29
July 8
July 16
July 8
July 16
July 8
July 22
July 8
July 16
July 22
July 8
. .do .

11 .

June 26 .

12-13 . .

July 3-4-...
June 24 .

julv p .

June 24 .

July 2 .

July fi .

June 23 .

16 .

June 22 . . . .

17 .

July 5 . .

July 22
July 8
July 11
July 16
July 22
July 11
July 16
July 22
July 16
. .do .

17-18 .

June 20-21..
June 23 ....

18-19 .

June 27-28..
July 3-4....
June 22 .

20 . . .

Tune 26 .

July 2 .

Tune 24 .

June 23

Time 22 .

. . . do .

24-25. . .

June 27-28..
June 20-21..
June 26 .

July 22
July 16
July 22
. . .do .

25-26. . .

June 25 .

June 23

. . .do .

•

31 . . .

June 22 .

July 23
. . .do .

32-33 . . .

June 20-21. .
. do .

38-39 .

July 29
_ do .

40-41 .

June 18-19..

In 1915 the earliest hatching took place on the seventh day after
oviposition, the earliest first molt on the sixteenth day, the earliest sec¬
ond molt on the twenty-first day, the earliest third molt on the twenty-
fifth day, and the earliest last molt on the twenty-ninth day, and the
first larva to consume the entire seed contents had done so on the forty-
ninth day. The last larva to finish feeding required 57 days.

Table II shows all the life-history data obtained during 1915 from
cage-infested apples. As will be noted, all apples used in this work
were infested during a period of 5 days from June 28 to July 2. Thus
all individuals were developing under practically identical conditions.
It will also be noted that many of the belated first-instar larvae and eggs
were found in the mucilaginous tissue surrounding the central gelati¬
nous body.

498

Journal of Agricultural Research

Vol. VH, No. n

Table II. — Development of Syntomaspis druparum at North East t Pa.t in 1915

Period
from ovi-

Date of

Stage and number of individuals found.

Remarks.

to ex ami
nation.

tion.

nation.

Bgg.

First

larva]

Seconc
. larval

Third

larval

Fourtl

larval

1 Fifth
. larval

Days.

June 29

Jnly s

June 28

. . .do....

July 2

July 9

July 1

. . .do .

July 2

July 10

June 28

July 7

. . .do .

July 8
July 9

. . .do .

July 10
July 14

July 2

June 28

July 11

July 1

July 14

June 28

July 12

. . .do .

July 13
July 14

. . .do .

July is
July 16

. . .do _

. . .do .

July 17

Bgg and two first-stage lar¬
vae were in mucilaginous
tissue at large end of seed.

July 1

July 20

June 28

July 18

. . .do .

July 19

First-stage larva in mucilagi¬
nous tissue.

Bgg and first-stage larva in
mucilaginous tissue, seeds
considerably hardened.

First-stage larvae in muci¬
laginous tissue.

Do.

. . .do .

July 20

• 1

June 29

July 21

June 28

. . .do .

July 22

June 29

July 23*

June 28

. . .do .

July 24
July 25
July 26

. .do .

. . .do .

July 27
July 31
July 28

July 2

June 28

July 1

July 31
Aug. 1

July 2

July 1

. . .do .

June 29

July 31

Aug. 3

First-stage larva in mucilagi¬
nous tissue.

July 2

July 1

. . .do .

July 2

Aug. 4

July 1

... .do .

. . .do .

Aug. 5

. . .do .

Aug. 6

. . .do .

Aug. 7

July 2

Aug. 9
. . .do .

July 1

' 8
20

July 3

Aug. 10

July 1

. . .do .

July 2

Aug. 11

. . .do .

Aug. 12

June 29

Aug. 10

June 28

. .do .

June 29

Aug. 11

. . .do .

Aug. 12

. .do.,...

Aug. 13
Aug. 14
Aug. 15

. .do .

Aug. 16

...do .

Aug. 17

Aug. 21

2 had consumed entire seed
contents.

July 2

. .do .

Aug. 22

3 had consumed entire seed
contents.

Do.

4 had consumed entire seed
contents.

5 had consumed entire seed
contents.

. .do .

Aug. 23

. .do .

Aug. 24

. .do .

Aug. 25

. .do .

Aug. 26

. .do .

Aug. 27

7 had consumed entire seed
contents.

Do.

. .do _ _

Aug. 28

Dec. n, 1916

Syntomaspis druparum

499

On August 30, 1915, two days after the last cage-infested apple had
been examined, 165 seeds infested naturally were examined to determine
whether all larvae had by that time finished feeding. Of the larvae found,
132, or exactly 80 per cent, had consumed the entire contents of the seed
and the rest had practically done so. Of 50 larvae examined on Septem¬
ber 2, all had finished feeding. In other words, by the last of August all
the larvae had reached full growth.

Number maturing in a single SEED, — In removing larvae from apple
seeds the fact has been observed that as they increase in size and age the
likelihood of finding more than one in a seed decreases. It is not at all
uncommon to find 6 or 7 very young larvae in a single seed, even in an
apple naturally infested; but on only one occasion has more than 1 of
the fifth instar been found within a single seed. In this case there were
2. The number is usually reduced to 1 before the fourth instar is reached.
This reduction in number is brought about by the actual killing and eating
of the surplus larvae by the one which ultimately matures. On a number
of occasions this cannibalistic habit has been observed, the larvae con¬
cerned being usually in the second or third instar.

Hibernating larva. — When the larva has consumed its total supply
of food it very shortly assumes what may be called the hibernating form
(PI. 38, D). This does not involve a molting of the skin but consists
merely in longitudinal contraction of the body,, the head* and caudal
segments being drawn in and the body becoming relatively thicker and
more deeply wrinkled. In this condition it remains until the following
spring.

Biennial brood. — Not all of the larvae from eggs of a given season
finish their development and emerge as adults the following spring, but
a large percentage of them remain as larvae within the seeds until the
second spring. This was suspected during the summer of 1914, when,
on July 23, the writer, in examining some seeds infested in 1913, found
some that still contained living larvae. One hundred seeds were selected
at random to determine roughly what percentage of the larvae were
likely to live over until the next spring. Of these 100 seeds, 54 contained
dead larvae, 26 living larvae, and from 14 the adult insects had emerged.
Of the living insects 65 per cent had not emerged. This lot of seed was
kept until the summer of 1915 and count kept of the emerging adults.
A total of 416 insects were reared in the second spring as against 339 in
1914. In other words, 55.1 per cent of the insects lived over two winters
as larvae.

It would appear that this curious habit serves to prevent extermination
of the species by a season of no fruit.

Pupation. — The larvae begin to pupate during the latter half of May,
the latest pupation, judging from the emergence of the adults, probably
taking place from three weeks to a month later.

500

Journal of Agricultural Research

Vol. VII, No. ii

PUPA

Description . — The pupa (fig. 8) is normally, depending on the sex,
from 3 to 4 mm. long, females being the larger. It is at first white, but
later those parts that are chitinized in the adult become first brownish and
later dark greenish; this color being really on the body of the adult,

developing within the pupal skin and
showing through the latter. The
legs, wing pads, antennae, and palpi
are folded along the sides and venter,
and in the female the ovipositor ex¬
tends over the back, reaching nearly
to the head.

Pupae period. — The pupal period
is of about four weeks* duration, some
ndividuals requiring slightly less and some slightly more than this period.

ECONOMIC IMPORTANCE

As has been pointed out on an earlier page, the only commercial fruits
that are, under conditions of ordinary care, at all heavily infested by
the seed chalcid are the Lady apple and, occasionally, crab apples, both
varieties with very limited markets. Also, under normal conditions of
growth distortion of fruit to such extent as to render it unmarketable is
rather rare, and infestation by the chalcid apparently has no effect on
the color of fruit. As pointed out by Crosby and as proven by the obser¬
vations of the writer, Lady apples are apparently practically immune to
the distortion of oviposition. These things being true, it is apparent
that economically the seed chalcid is of little importance.

CONTROL OF THE CHALCID

Natural control. — Apparently the apple-seed chalcid has no specific
enemies. No records of such are to be found in European or American
literature, and none has come under the observation of the writer.
Other apple insects, such as the codling moth, which sometimes devour
the seeds, undoubtedly destroy a limited number of chalcid larvse, and
others, which feed in the fallen apples, account for the death of a few
more. Some adult chalcids doubtless fall prey to birds, spiders, and
other predators. But all of these together constitute only a very small
measure of control.

Mortality among the hibernating larvse is apparently very small
also; for of 115 larvse found in apples that had lain under the tree all
through the winter of 19 14-15 only three were dead, and each of these
was in a seed that had been eaten into by some other insect. The mor¬
tality in seeds that become separated from the pulp may be higher, but
as it is almost impossible to find such seeds no data on the point are
available.

Fig. 8. — Syntomaspis druparum: Pupa of female.
Much enlarged. (Original.)

Dec. ii, 1916

Syntomaspis druparum

5oi

Artificial, control. — Inasmuch as the seed chalcid attacks normally
only varieties that are grown on a very small scale it can be controlled
with comparatively little effort by purely mechanical means. In the
first place, all wild seedling apples and wild crab apples in the neighbor¬
hood of such varieties should be destroyed. This would not only elimi¬
nate the outside source of this insect, but also of many other much more
serious pests. This should be done in the spring or summer, preferably in
August after oviposition has ceased, to insure the destruction of the chal¬
cid larvae of the season. From the seed crop of the previous season there
will still be the chalcids of the biennial brood for the next following
season to contend with, but in two years this source of infestation will be
entirely eliminated. In addition to the foregoing the careful destruc¬
tion of all drop fruit and culls for two seasons will practically extermi¬
nate the chalcids. If the waste fruit is converted into cider, the pomace
should be destroyed.

LITERATURE CITED

(1) Boheman, C. H.

1834. Skandinaviska Pteromaliner. In K. Svenska Vetensk. Akad. Handl.,
1833. P- 329-38o.

(2) Thompson, C. G.

1875. Skandinaviens Hymenoptera. Delen 4. Lund.

(3) Porchinskii, I. A.

1893. [Torymus bred from pear seeds.] (In Russian.) In Horae Soc. Ent.,
Rosst., t. 27, 1892/93, Bui. Ent., p. ix.

(4) Mokrzecki, S.

1906. Beitrag zur Kenntnis der Lebensweise von Syntomaspis pubescens
Forst., druparum (Boh.) Thoms., (Hymenoptera, Chalcididae). In
Ztschr. Wiss. Insektenbiol., Bd. 2, Heft 12, p. 390-392, 2 fig.

(5) Rodzianko, W. N.

1908. Commentatio de Torymidis, quarum larvae in seminibus Pomacearum
vitam agunt. In Bui. Soc. Imp. Nat. Moscou, n. s. t. 21, 1907, p.
592-611.

6) Crosby, C. R.

1908. Notes on a chalcid (Syntomaspis druparum Boh.) infesting apple seeds.
(Abstract.) In Ann. Ent. Soc. Amer., v. 1, no. r, p. 38.

(7) -

1908. On certain seed-infesting Chalcis flies. N. Y. Cornell Agr. Exp. Sta.
Bui. 265, p. 367-388, fig. 73-98, 2 pi.

(8) Horvath, Geza.

1911. A magyar entomologusok tomoriilese. In Rov. Lapok, k6tet 18, fuzet 1,

P* 1-3*

(9) Crosby, C. R.

1912. Notes on Syntomaspis druparum Boh. and Ichneumon nigricomis
Berger. In Canad. Ent., v. 44, no. 12, p. 365-366.

PLATE 37

Syniomaspis druparum:

A. — Adult female, a, greatly enlarged, b, X 3- (Original.)

B. — Adult male; outline of abdomen, lateral view, at right. Gteatly enlarged.

(Original.)

(5<»)

PLATE 38

Syntomaspis druparum: Apple injury and hibernating larvae

A. — Usual type of injury resulting from oviposition. Natural size.

B, C. — Extreme type of injury resulting from oviposition. Natural size.
D. — Hibernating larvae within seeds of an apple. Greatly enlarged.

66847°— 16 - 4

PIRATE 39

Syniomaspis druparum; Infested and sound seeds of apples

A. — Infested seeds. Much enlarged.

B. — Sound seeds. Much enlarged.

Plate 40

PLATE 4®

Syniomaspis druparum; Oviposition

A. — Female ovipositing in fruit of Crataegus sp. Photographed from life, X 2

B, C. — Oviposition in apples. Photographed from life. X 2.

D. — Mica cage used in the life-history studies of Syniomaspis druparum . XK-

JOURNAL OF ACRKULTURAL RESEARCH

DEPARTMENT OF AGRICULTURE

Vol. VII Washington, D. C., December 18, 1916 No. 12

ASSIMILATION OF IRON BY RICE FROM CERTAIN
NUTRIENT SOLUTIONS

By P. L. Gile, Chemist , and J. O. Carrero, Assistant Chemist , Porto Rico Agri¬
cultural Experiment Station

INTRODUCTION

It has long been recognized that on calcareous soils certain plants do
not make a normal growth and are often affected by chlorosis This
has been variously ascribed to the physical condition of the soil; an
increasing assimilation of lime by the plant; a diminished assimilation of
potash, iron, or phosphoric acid ; or a diminished assimilation of all mineral
nutrients due to the neutralization of acid root excretions. As already
noted, work at this station with pineapples and upland rice indicated an
insufficient assimilation of iron as the principal cause (4, 5)1

As calcareous soils have a slightly alkaline reaction it is important to
know whether plants intolerant of calcareous soils are sensitive to an
acid or alkaline reaction per se, and whether the reaction of the soil has
any effect on the assimilation of iron. Various experiments with upland
rice {Oryza saliva) in nutrient solutions and soil cultures have been con¬
ducted to gain information on this subject. The experiments with nutrient
solutions reported here show the effect of the quantity and form of iron
and reaction of the nutrient medium on the asssimilation of iron by rice.

PLAN AND EXPERIMENTAL METHODS

The plan of the work was to measure the growth of rice in acid, neutral,
and alkaline nutrient solutions when supplied with 0.002 and 0.008 gm.
of iron (Fe) per liter from ferrous sulphate, ferric chlorid, dialyzed iron,
ferric citrate, and ferric tartrate ; to determine the amount of iron taken up
by the plants from these solutions ; and to determine the amount of soluble
iron actually present in the solutions at different times. A few other ex¬
periments were conducted to explain certain results obtained. Three
preliminary experiments were also conducted before the general plan and
exact detail of the experiments were decided upon.

1 Reference is made by number to 'Literature cited/* p. 528.

Journal of Agricultural Research,

Depth of Agriculture. Washington, D. C,
gm

Vol. VII, No. is
Dec. x8, 19x6
B — 12

(503)

504

Vol. VII, No. 12

Journal of Agricultural Research

Upland rice seedlings germinated over distilled water were placed in
the nutrient solutions when the plumules were about 2 inches long.
Water distilled from a cast-iron still with a block-tin condenser and stored
in a tin-lined copper tank was used in the nutrient solutions. The plants
were grown for 40 days in Erlenmeyer flasks of Jena or Nonsol glass.
Flasks of 200 c. c. capacity were used for the first 25 days of each experi¬
ment and flasks of 500 c. c. capacity for the remaining 15 days. Tran¬
spired water was replaced with distilled water daily, and the plants were
changed to fresh solutions every 4 days, except they were left in the solu¬
tion 6 days before the first change. . The plants were kept in a wire in¬
closure (five meshes to the inch) during fair weather and in a glass house
during rains.

The nutrient solutions used were of the following compositions:

Acid solution. Gm.

Potassium nitrate (KN03) 10. 71

Monobasic potassium
phosphate (KH2P04). . . 7. 14

Sodium nitrate (NaN 03 ) . . 2 1 . 43

Sodium sulphate (Na2S04) 3. 15

Calcium ‘chlorid (CaCl2) . . 2. o

Magnesium chlorid

(MgCl2) . 2.0

Sulphuric acid (H2S04) ... o. 245
Distilled water . 100, 000

N eutral solution . Gm.

Potassium nitrate (KN03) 10. 71

Monobasic potassium
phosphate (KH2P04) . . 3. 57

Dibasic potassium phos¬
phate (K2HP04) . 3. 57

Sodium nitrate (NaN03).. 21. 43

Sodium sulphate (N^SOJ 3. 15

Calcium chlorid (CaCl2). . 2. o

Magnesium chlorid

(MgCl2) . . 2.0

Distilled water . . . 100, 000

The solution alkaline with carbonate of lime was the same as the neutral
solution except for the addition of 0.41 gm. of precipitated calcium car¬
bonate per liter. In the various experiments different quantities and
kinds of iron were added to these solutions.

It is possible that the above solutions are not ideal for rice, although
plants equal in size to exceptionally large field plants were grown in them.
Previous and subsequent work with these solutions showed that the growth
of rice was not increased by increasing the quantity of calcium chlorid
or magnesium chlorid, by doubling the quantities of dibasic pdtassium
phosphate and monobasic potassium phosphate, or by doubling the
quantities of potassium nitrate and sodium nitrate. The concentration
of total salts in the solutions was lower than usual, but afforded sufficient
nutrients because of the frequent changes. There is an advantage in
using dilute solutions in that imperfect balancing of the salts (antago¬
nistic effects of the ions) may produce no injury in low concentrations.

Before each change the nutrient solutions were made up fresh from a
stock solution and the iron added 18 hours before the plants were inserted.
After emptying out the old solution, the flasks and roots were rinsed once
with a small quantity of distilled water. These details seem unimportant,
but, as will be seen later, alteration in such apparently trivial details may
affect the results.

Dec. 18, 1916

Assimilation of Iron by Rice

505

In analyzing the plants for iron the substance was ashed over a low
flame without the addition of calcium acetate and iron determined colori-
metrically by the method of Stokes and Cain (14). In attempting to
estimate the soluble iron in the nutrient solutions, the solutions, were
filtered, the filtrate concentrated when necessary, and iron determined
colorimetrically by the above method. The ordinary colorimetric method
with potassium sulphocyanate (KSCN) was used in some of the pre¬
liminary work, but all the determinations reported were made by the
method of Stokes and Cain unless otherwise specified. Blanks were run
for iron with the acid and materials employed.

PRELIMINARY EXPERIMENTS

Following are the data of three preliminary experiments conducted
before a uniform method was adopted :

Experiment i (Source of iron: Ferrous sulphate. Plants grown in
double flasks). — It was at first thought it would be advantageous to
grow the plants with their roots equally divided between two flasks, one
flask to contain the acid solution and the other the neutral. By adding
iron only to the acid solution in one lot and only to the neutral solution
in another lot it was thought that decisive results would be obtained
on the effect of the reaction of the solution on the assimilation of iron,
as the only difference between the two lots would be in the solution in
which iron occurred. The results, however, did not bear out the assump¬
tion.

Two plants were grown in each twin flask (A and B). Four flasks were
taken as a unit and the units triplicated for each treatment. All solu¬
tions containing iron contained the same quantity, but in this experiment
the quantity of iron was not kept constant during the 40 days of growth,
as it was thought that at first so much iron was added as to obscure the
effect of the reaction on assimilation. During the first 10 days 0.008 gm.
of iron per liter was used, during the second 10 days 0.004 gm- of iron,
and during the last 20 days 0.002 gm. of iron. In Table I are given the
weights of the plants grown in the different solutions, together with the
percentages of nitrogen and iron in the dry substance.

The color of the plants during growth was as follows: At 10 days
plants 1 to 12 were strongly chlorotic, plants 25 to 36 were chlorotic,
plants 13 to 24 and 49 to 60 were green, and plants 37 to 48 were dark
green; from the twentieth to the fortieth day plants 1 to 12 were strongly
chlorotic and all others were about the same normal green.

From the weights of the stalks and leaves it appears that with the
quantities of iron used the reaction of the solution had little effect on
the assimilation of iron. It also seems that there is no advantage in
growing the plants with their roots divided between two flasks, as one-
half the roots are not able to absorb sufficient iron for the full needs of

506

Journal of Agricultural Research

Vol. VII, No. xa

the plants.1 This is evident from a comparison of the weights of plants
13 to 24, with 37 to 48, and of plants 25 to 36 with 49 to 60.

Table I. — Comparative weights of rice plants grown in double flasks with ferrous sulphate
* in acid and neutral solutions and percentage of nitrogen and iron

Nutrient solution in flask —

Neutral. . . .

Acid .

Do .

Acid-f-iron. . .

Neutral+

iron.

Acid .

Acid-}- iron. .

Acid+iron. . .

Neutral-f

Neutral 4*

iron.

Flasks

No.

1-4

9-12

13-16

17-20

21-24

25-28

29-32

33-36

37-40

41-44

45-48

49-52

53-56

57-6o

Green

weight

stalks

and

leaves.

Gm.
I. 42

I- 39
1.32
36. 10
38. 55

28. 85

29. 62

33- 80

3i- 19

48.42
48. 70
S1, 97
49- 2S
S3- 67
S°- 49

Oven-

dry

weight

stalks

and

leaves.

Gm
O. 26
. 26
. 26
4. 78
S- °9
3- 93

4. IO
4. 78
4.27

5- 99

6. 22
6.61

6- 34

7. OO

6.48

Oven-

dry

weight

roots.

Gm,
O. 10
.09
. 10
1. 29
i- 39
1. 10
1. 14
1. 23
1. 17

1.87

1.88
1. 91
1. 67
1. 91
1. 70

Aver¬

age

oven-

dry

weight

stalks

and

leaves.

Gm.

o. 26

4. 60

4- 38

6. 27
6. 61

Aver¬

age

oven-

dry

weight

whole

plant.

Gm.

O. 36

5* 86

5- 56

8. 16
8-37

Iron
(EeaOs)
in dry
stalks
and
leaves.

Per ct.

o. 040

038

023

038

Nitro¬
gen (N)
in dry
stalks
and
leaves.

Per ct.

3 - 90

3- 72

3- 90

3- 70
3.80

Experiment 2 (Source of iron: Ferrous sulphate. Three different
quantities of iron in acid and neutral solutions).— Since it was thought
that in experiment 1 so much iron was used as to obscure partially the
effect of the reaction on the assimilation of iron, an experiment was
conducted using different quantities of iron in acid and neutral solutions.
As a check on the previous results, two lots received iron at the same
rate as in experiment 1 — that is, 0,008 gm. of iron per liter for the first
10 days, 0.004 g**1- for the second 10 days, and 0.002 gm. for the last
20 days of the experiment. The other lots received either 0.008 or 0.002
gm. per liter during the whole 40 days. Ferrous sulphate was used as
the source of iron. One plant was grown in each flask, six flasks were
taken as a unit, and the units triplicated for each treatment. The
weights of the plants with the percentages of nitrogen, phosphoric acid,
and iron in the dry substance are given in Table II.

The color of the plants during growth was as follows: After 7 days1
growth plants 55 to 72 were yellowish green, plants 91 to 108 and 73 to 90
were light green, plants 1 to 54 were dark green; after 12 days’ growth
plants 55 to 72 were light green and all others were of a good green color.

It is apparent from this experiment that the effect of the reaction of the
solution depends somewhat on the quantity of iron supplied, this reaction
being more evident in solutions containing a small amount of iron.

1 This is not definitely proved by this experiment alone, but further work substantiated it.

Dec, x8, 1916

Assimilation of Iron by Rice

507

Table II. — Comparative weights of rice plants grown in acid and neutral solutions with
three quantities of iron from ferrous sulphate

Nutrient solution.

Quantity of iron per
liter.

Acid .

Gw.

0.002 .

Do .

.008, O.OO4, 0.002

Do .

.008 .

Neutral .

.002 .

Do .

.008, O.OO4, 0.002

Do .

.008 .

1- 6
7-12
13- 18
19- 24
25- 30

3i- 3634-38

37^ 42
43- 48
49 “ 54
55- 60
61- 66
67- 72

Gm.
32. 05
32.98
31.40
30.48
30. 00

34- 72
S3- 10
51. 18
28.

26. 71
27. 72

483

73“ 7836-42
79- 8437- 75
85- 90 28. 28
91- 9659. l8
97-102
103-108

59- 45

Gm.

4.04
4. 12
3-93
3- 90
3- 78
4. 18

4*37
6. 58
6. 12
73
3* 34
3- 56

T «

Gw.
I. 04
1. 08
.92
.96
•95

• 91
1.32
1. 92
1. 80

.85

•79

• 79

4* 40; 1. 03

4.69
3. 62
7- 54

7- 49

i- 15
.84
2. 17

2. 00

Average
oven-dry
weight
of —

In dry substances
of . stalks and
leaves, percent¬
ages of —

Stalks and
leaves.

jwhole plant.

Nitrogen

(N)

O *

|<5

•SfC

Gw.

P. c.

4*03

5-04

4* 34

2.31

0. 037

j. .

3- 95

4.89

4* 55

2. 20

•037

5-69

7-37

4. 40 2. 02

* 034

_ i _

3- 54

4*35

4- 56

2. 06

. 024

4.24

5* 24

4.40

2. 04

. 020

7- 52

9. 60

4.28

I. 80

. 028

Experiment 3 (Sources of iron in acid, neutral, and alkaline solutions:
Ferrous sulphate and ferric citrate). — This experiment was designed to
observe the assimilation of iron in a solution containing calcium car¬
bonate — that is, a slightly alkaline solution — and to compare the assim-
ilability of ferric citrate with that of ferrous sulphate. The method used
in the conduct of this experiment differed from that of other experiments
only in the making of the nutrient solutions. In all other experiments
reported the nutrient solutions were made up fresh 18 hours before the
plants were inserted, and any residue remaining in the bottles from the
previous lot was thrown away. In this experiment the nutrient solutions
were made up fresh 18 hours beforehand, but whatever nutrient solution
remained from the previous change was left in the bottles and added to the
fresh solutions. This residue increased at times but never amounted to
more than a quarter of the whole solution. It nevertheless affected the
results.

Iron from both ferric citrate and ferrous sulphate was used at the rate of
0.002 gm. of iron per liter in all the cultures. The alkaline solutions with
calcium carbonate contained 0.41 gm. of precipitated calcium carbonate
per liter. One plant was grown in each flash. Six flasks were taken as
a unit, and the units were triplicated for each treatment. The growths
of the plants in the different solutions and a partial ash analysis of the
stalks and leaves are given in Table III.

5°8

Journal of Agricultural Research

VoL VII, No. 12

Table III. — Comparative weights of rice plants grown in acid, neutral, and alkaline
solutions with 0.002 gm. of iron per liter from ferrous sulphate or ferric citrate

1 J

•s

•s .

Average
oven-dry
weight of —

Composition of stalks and leaves.

Nutrient

solution.

Source of iron.

8 M
>

Oven-dry w

roots.

Stalks and

leaves.

Whole plant.

Silica (SiOa).

■ a

Magnesia

(MgO).

Phospho r i c

acid (P2O5).

t-4

fFerrous 1
\ sulphate!

fFer r i c I
\ citrate. . J

fFerrous 1
\ sulphate]

fFerr i c f
\ citrate . . |

fFerrous 1
\ sulphate]

fFer r i c [
\ citrate . . j

1-6

Gm.

38. 44

27. 22

Gm.
4- 57

A. AO

Gm.
I. 12

Gm.

P. Ct.

P. ct.

P. CL P. ct.

P. ct.

P. Ct.

Acid ....

7-12

I. 07
I - 03

I. 21

13-18

19-24

25-30

31-36

37-42

43-48

49-54

S5-6o

61-66

32. 604. 05
46. 02 5. 13
40. 24I4. 63
43- 64 5- !3
25- 25:2- 99
29. 67I3. 48
26. 66J3. 20
38. I4|4. 45
3i- 78:3- 79

38. 78)4. 45
13. 201. 76
21. 05I2. 60

19. 13 2. 34
33- 3° 3- 98

39. 72 4. 78
33. 644. 09

4-37

5-44

13. 90

O. 26

0. 74

I. 18

2. 03

0. 044

Do. .

I. 14
I. 26
•74

■ 83
• 76
1. 06

4.96

6. 16

14.41

• 13

• 75

I. 18

I.97

• 039

Neutral . .

3. 22

3- 99

14. I9

.28

•79

I.44

2. 03

. 028

Do. .

.92
1. 02
•70
1. 07
.85
1. 25

.67-72

73-78

79-84

85-90

91-96

97-102

103-108

4-23

5-23

14. 23

. 22

.78

I- 15

I.94

. 024

Alkaline

2. 23

3- 10

13.64

•39

.96

I. 23

2. 17

. 026

Do. .

i*43

4. 28

5- 68

14. 22

. 22

• 87

I. 16

2. 02

. 021

The color of the plants during growth was as follows: After 8 days
growth plants 1 to 18 were a good green, and all others were strongly
chlorotic; after 15 days growth plants 1 to 36 were a good green, 91 to 108
were slightly chlorotic, 55 to 72 were chlorotic, 73 to 90 and 37 to 54 were
strongly chlorotic; after 23 days growth plants 1 to 36 were dark green,
91 to 108 and 55 to 72 were light green, 37 to 54 were a lighter green, and
73 to 90 were chlorotic.

In this experiment there was a greater difference in growth between
the acid and neutral solutions with 0.002 gm. of iron per liter from ferrous
sulphate than in the preceding experiment, probably due to the different
method of preparing the solutions. The residual solutions not only con¬
tained less soluble iron than the freshly prepared solutions, but the pre¬
cipitate already formed in the residual solutions probably affected the
rate at which the iron in the freshly prepared solutions became insoluble.

RESULTS OF PRELIMINARY EXPERIMENTS

The three preliminary experiments showed that the effect of the reac¬
tion of the solution on growth and assimilation of iron depended some¬
what on the quantity of iron added. With 0.002 gm. of iron per liter
from ferrous sulphate, growth was greatest in the acid solution, but with
0.008 gm. of iron per liter, growth was greatest in the neutral solution.

Dec. xS, 19x6

Assimilation of Iron by Rice

509

Increasing the quantity of iron from 0,002 to 0.008 gm. per liter greatly
increased the growth in both the acid and neutral solutions.

Analyses showed' that plahts grown in the different solutions did not
differ appreciably in their percentages of nitrogen, phosphoric acid, lime,
magnesia, or carbon-free ash, but did differ materially in the quantity
of iron they contained. Plants grown in the acid solutions contained
about 50 per cent more iron than plants grown in corresponding neutral
or alkaline solutions.

In the course of this and other work it has been found that the color of
the leaves is a pretty good indication of whether or not the plant is ob¬
taining sufficient iron, although a lack of green is often attributable to
other causes than a lack of iron. During growth, plants in the acid
solution were the darkest green, in accordance with the percentages of
iron found in the plants.

Used at the rate of 0.002 gm. of iron per liter ferric citrate was plainly
a better source of iron than ferrous sulphate, especially in neutral and
alkaline solutions.

That the growth in the different solutions was largely controlled by
the supply of available iron seems proved by the color and analyses of
the plants, and by the increased growth following an increase in quantity
or change in the kind of iron added to the solutions.

FINAL EXPERIMENTS

Experiment 4 (Sources of iron in acid, neutral, and alkaline solutions:
Ferrous sulphate and ferric citrate). — In this experiment 0.002 and 0.008
gm. of iron per liter from ferrous sulphate and 0.002 gm. of iron from
ferric citrate were tested in the three different nutrient solutions. Two
plants were grown in each flask and six flasks were taken as a unit, the
units being duplicated for each treatment. The data of growth and the
percentages of iron in the dry substance of the stalks and leaves are given
in Table IV.

The color of the plants during growth was as follows : At 10 days plants
1 to 24 and 49 to 60 ^Vere a good green, all others were a yellowish green
except plants 85 to 96, which were strongly chlorotic; at 20 days plants
1 to 24 and 49 to 72 were a good green, plants 25 to 48, 97 to 108, 73 to
84, and 97 to 108 were a slightly poorer color, and plants 85 to 96 were
almost white.

The results of this experiment confirm those of the preliminary experi¬
ments in respect to (1) the superiority of ferric citrate to ferrous sulphate
as a source of iron, (2) the relative growths in acid and neutral solutions
with the two quantities of iron, and (3) the markedly higher percentages
of iron in plants grown in the acid solutions. The new fact established
was that increasing the iron from 0.002 to 0.008 gm. per liter in the alka¬
line solution depressed the growth of plants to a surprising degree. That
the lack of growth in this solution was due to a lack of available iron is

5io

Journal of Agricultural Research

Vol. VII, No. xa

proved by the pronounced chlorosis of the plants and by the results of
experiment io. Here, as well as in experiment i, plants 85 to 96, which
made practically no growth because of a lack of irori, contained a high
percentage of iron in the dry substance. This point will be discussed
later.

Table IV. — Comparative weights of plants grown in acid , neutral , and alkaline solutions,
with ferrous sulphate or ferric citrate as the source of iron

Nutrient

solution.

Source of iron.

Acid .......

Ferrous sulphate . .
. do .

Do .

. do .

Do .

. do . .

Do .

Ferric citrate .

. do .

Neutral .

Do .

Ferrous sulphate. . .
. do .

Do .

. do .

Do .

. do .

Do .

Ferric citrate .

Do .

. do .

Alkaline.. . .
Do .

Ferrous sulphate . .
. do .

Do.......

. do. . . .* .

Do.......

. . . . .do .

Do .

Ferric citrate .

Do .

. do .

Quantity of iron
per liter.

Flasks No.

ight of

s and

V •

Average oven-
dry weight of —

<5j3

Green we

stalk;

leaves.

Oven-dry

of stall

leaves.

Stalks

and

leaves.

Whole

plant.

Iron (Fes

dry stal

leaves.

Gm.

Per cU

002

1-6

48.45

6. 66

1.47

......

- 4 * • •

002

008

7-12

° 13-18

45-85

91. 88

6. 42
12. 41

x, 46
3. 10

6-54

.8. 01

.036

•

008

002

19-24

23-30

83. 29

6547

10. 94
8. 70

2. 65
1. 82

11. 68

14. 56

*032

. 002

31-36

37-42

60. 07
47* 19

8. 47
6. 51

1.83

8- 59

10. 42

. 026

. 002
. 008

43-48

49-54

42. 39
r 00. 22

5- 92
12. 93

i*35
2. 97

6. 21

7. 61

. 017

. 008
. 002

5S-6o

61-66

100. 31
62. 07
58. 60
45- *8

r3- 05
8. 32
7. 85
6. 17

2. 90
1.83
1. 79
1. 63

12. 99

r5* 93

.023

. 002

67-72

73-78

8. 09

9. 90

.015

. 002
. 008

79-84

85-90

42. 13
•93

6. 22
. 21

1. 68
. 11

6. 20

7. 86

. 017

. 008
. 002

91-96

97-102

103-108

1. 07
63. 28
61. 85

•23

8.33

8.29

. 11

2. 20

. 22

*33

*049

. 002

2. 11

8.31

10.47

. 012

a Results calculated from five flasks.

Experiment 5 (Ferrous sulphate in neutral solution with and without
carbon black). — Data of growth in the different solutions seemed to point
overwhelmingly to the iron supply as the factor controlling growth;
nevertheless, it was possible that in some cases the precipitate in the
solutions affected growth aside from any influence of the precipitate on
the amount of soluble iron. After the addition of iron salts to the
nutrient solutions a precipitate was formed which varied according to
the source of the iron and the nutrient solution, but was greater as the
amount of iron added became larger. In the preceding experiments a
marked increase in growth was produced by increasing the iron from
0.002 to 0.008 gm. per liter in the neutral solution. It was possible that
the greater precipitate in the solution with 0.008 gm. of iron favored
growth in removing by absorption traces of heavy metals or other sub¬
stances present in the distilled water. As carbon black and ferric hydroxid
have been found to improve the quality of distilled water in this way, the
following test was conducted.

Dec. x8, 19x6

Assimilation of Iron by Rice

5ii

To two lots of solutions carbon black was added at the rate of 0.0432 gm.
per liter. This is about double the weight of the precipitate in the
ordinary neutral solution with 0.008 gm. of iron per liter from ferrous
sulphate. Two other solutions were made up using distilled water that
had previously been treated with 0.086 gm. of carbon black per liter and
then filtered. The neutral nutrient solution was used in all sets, and
ferrous sulphate was the source of the iron. Two plants were grown in
each flask, six flasks were taken as a unit, and the units duplicated for
each treatment. The results of the test are given in Table V.

Table V. — Comparative weights of rice plants grown in neutral nutrient solution with
iron from ferrous sulphate and also with distilled water treated with carbon block

Kind of distilled 'water used in
nutrient solution.

Ordinary distilled wdter . . .

Do .

Ordinary distilled water +
carbon black.

Do .

Do . :

| .
« Ih'
O 4>

Gm,

+0. 002
+ • 008
+ . 002
-f“ . 008
4- • 002
-j- • 008

’ll

Green weight of
stalks and leaves.

Oven-dry weight of

stalks and leaves.

f .

Average oven-
dry weight of —

Stalks and

leaves.

Gm,

Gm.

f 1—6

62. 19

8. 27

1. 86

{ 7-“

65.07

8.48

1. 92

8. 38

10. 27

/ 13-18

128. 62

16. 31

3- 73

\ 19-24

137- 5°

16- 73

4. 01

16. 52

20.39

{23-3°

58. 56

7- 65

1. 69

131-36

57- 01

7-55

1. 71

7. 60

9- 30

{37-42

139. 40

17- 54

3- 54

143-48

127. 70

I5-9I

3- 47

16. 73'

20. 24

{49-54

88. 20

II. 10

2, 56

Iss— 60

88. 10

10. 97

2.49

11. 04

13- 57

{61-66

148. 53

18. 49

4*42

167-72

147- 87

18. 62

4- 65

18. 56

23. IO

During growth lots with 0.002 gm. of iron per liter were of a slightly
poorer color than lots with 0.008 gm., but no differences in color were
apparent between the lots with different distilled waters or with carbon
black.

It is evident that the addition of carbon black to the nutrient solutions
with 0.002 gm. of iron slightly depressed rather than increased the yield
while carbon in the solution with 0.008 gm. of iron had no effect on the
yield. Treating the distilled water with carbon and then filtering in¬
creased the yield in the solution with 0.002 gm. of iron by 32 per cent and
in the solution with 0.008 gm. by 12 per cent.

Previously treating the distilled water with carbon increased the growth
less in the solution with 0.008 gm. of iron than in the solution with 0.002
gm., possibly because the formation of the larger flocculent precipitate
of iron in the former solution acted similarly to carbon in removing injuri¬
ous substances from the distilled water. Probably the adsorption of the
other salts prevented the carbon when added directly to the nutrient

512

Journal of Agricultural Research

Vol. VII, No. 12

solution from adsorbing to any extent these injurious substances. The
slight depression in growth produced by carbon in the solution with 0.002
gm. of iron may well be due to slightly diminishing the amount of soluble
iron in the solution by adsorption.

It thus seems possible that the larger flocculent iron precipitate which
existed in the solution with 0.008 gm. of iron per liter may have increased
growth to some extent by improving the distilled water; but only a small
part of the greater growth produced by increasing the iron from 0.002 to
0.008 gm. can be attributed to that cause.

Experiment 6 (Source of iron in acid, neutral, and alkaline solutions:
Ferric chlorid). — The results with ferrous sulphate as the source of iron
may well be complicated by the fact that both ferrous and ferric iron
doubtless existed in the solution unless the ferrous iron was oxidized by
the plants' roots (13). In the very dilute solutions it was, of course,
impossible to tell how much of each kind of iron was present. In the
following experiment with ferric chlorid as the source of iron naturally
ferric iron only was present.

Ferric chlorid was added to the acid, neutral, and alkaline solutions so
as to furnish 0.002 and 0.008 gm. of iron per liter. Two seedlings were
grown in each flask, six flasks were taken as a unit, and the units tripli¬
cated for each treatment. The growth of plants and percentages of iron
in the dry substance are given in Table VI .

Table VI. — Comparative weights of rice plants grown in acid, neutral, and alkaline solu¬
tions, with ferric chlorid as the source of iron

Nutrient solution.

Quantity
of iron
per liter.

Flasks No.

Green
weight
of stalks
and
leaves.

Oven-
dry
weight
of stalks
and
leaves.

Oven-

dry

weight

roots.

Average
oven-dry
weight of—

Stalks

and

leaves.

Whole

plant.

Iron
(FeaOa)
in dry
stalks
and
leaves.

Gm.

Per ct.

Acid

Neutral

Do..-

Alkaline

o. 002

. 008

. 002

. 008

. 002

. 008

I- 6

7- 12
13- 18
19' 24
25' 30
31' 36
37' 42
43- 48
49' 54
55' 60
61- 66
67- 72
73' 78
79- 84
85- 90
91- 96
97-102
103-108

26. 50
32.37
24. 96
37- 67
39-43
34.62
29. 72
28. 22
29. 56
38. 12
41. 76
40. 70

6.96

6.97
6- 33

10. 03
9- 26
7.81

3- 89

4- 59
3- 46

5- 32
5- 49
4.78

4. l6

3- 81

3 - 83
5- 17

5. 82
S- 65

I. 09
I. IO

•97
i. 5i
1. 36
1. 22

o. 75
.87
.68

■97
1. 10

•94

.83

. 80

.85

1.03
1. 21
1. 16
*32
•33
•3i
.42
•37
•33

3-98

4-75

0. 025

5. 20

6. 20

. 026

3-93

4. 76

. 022

5- SS

6. 68

. 026

I- 05

r* 37

. 022

I. 36

. 023

Dec. i8, 1916

Assimilation of Iron by Rice

5i3

The color of the plants during growth was as follows : At 10 days plants
19 to 36 and 55 to 72 were a good green, 1 to 18 were a lighter green,
37 to 54 were a yellowish green, 73 to 108 had a still poorer color; at 20
days plants 19 to 36 and 55 to 72 were a fair green, 1 to 18 and 37 to 54
were markedly chlorotic, 73 to 108 were more chlorotic; at 30 days plants
19 to 36 and 55 to 72 were of slightly poor color, in 1 to 18 and 37 to 54
color was improved though still chlorotic, 73 to 108 were strongly chlo¬
rotic. The color and growth of plants were, on the whole, noticeably
inferior to that of plants grown with ferrous sulphate as the source of iron.

The growths made in the acid and neutral solutions were approxi¬
mately equal and much superior to those in the alkaline solution. Growth
was markedly increased in all three solutions by increasing the iron from
0.002 to 0.008 gm. per liter.

The results differ from those of previous experiments with ferrous
sulphate because in acid and neutral solutions growth was increased less
by increasing the iron from 0.002 to 0.008 gm. ; with 0.002 gm. of iron
growth was much less in the alkaline solution relative to growth in neutral
and acid solutions; increasing the iron from 0.002 to 0.008 gm. in the
alkaline solution measurably increased growth instead of markedly
depressing it.

The .percentages of iron in the dry substance of stalks and leaves varied
very little between plants from the different solutions. In each case
plants grown in solutions with 0.008 gm. of iron per liter contained very
slightly more iron than those grown in similar solutions with 0.002 gm.
of iron. It will be noted that there was not such a marked increase in the
assimilation of iron from the acid solution as in tests with ferrous sulphate.

Experiment 7 (Source of iron in acid, neutral, and alkaline solutions:
Ferric citrate). — Ferric citrate as a source of iron was compared with
ferrous sulphate in experiments 3 and 4, but only 0.002 gm. of iron per
liter was used. In this test both 0.002 and 0.008 gm. of iron per liter
were compared in the different solutions.

Three seedlings were grown in each flask, but 200 c. c. flasks were used
only during the first 15 days, 500 c. c. flasks during the next 15 days,
and 1,000 c. c. flasks during the last 10 days of growth. Six flasks were
taken as a unit and the units duplicated for each treatment. The growth
of plants and percentages of iron in dry stalks and leaves are shown in
Table VII.

During the first 30 days all plants were of a good green color, except
No. 61 to 72, which were yellowish green. All plants had a good color
during the last 10 days.

As in the previous tests with ferric citrate, the growth was equal in
acid, neutral, and alkaline solutions with 0.002 gm. of iron, per liter.
With 0.008 gm. of iron per liter growth was practically equal in the acid
and neutral solutions, but much less in the alkaline. Growth ’was

5*4

Vol. VII, No. 13

Journal of Agricultural Research

markedly increased in acid and neutral solutions, but was unaffected
in the alkaline solution by increasing the iron from 0.002 to 0.008 gm.
per liter.

Table VII. — Comparative weights of rice plants grown in acid , neutral , and alkaline
solutions , with ferric citrate as the source of iron

Nutrient solution.

Quantity
of iron per
liter.

Gm.

Acid .

Do. ..
Neutral. .

Do...
Alkaline .
Do...

. 008 |

Average oven-
dry weight of—

Flasks

1-6

7-12

13-18

19-24

25-30

31-36

37-42

43-48

49-54

55-6°

61-66

67-72

Green
weight
ot stalks
and
leaves.

Gm,
87. go

87. OI

124. 45
143. 40

87- 3 1
84.43
13°. 95

no. 12
91. 16

88. 65
70. 30
91. 64

Oven-
dry
weight
of stalks
and
leaves.

Gm.
11. 02

10. 96

17. 42

18. 08

11. 19
11. 08
16.35
13- 73
ii- 55
11. 28

8. 82
ii- 75

Oven-
dry
weight
of roots.

Gm.

2. 13

2- 15

3. 28

3- 65
2. 23
2. 28
3-27
2. 60
2. 34
2. 55
2. 11
2. 82

Stalks

and

leaves.

Gm.

II. 42

10. 29

Whole

plant.

Gm.

13- 87

12. 76

Iron
(FeaOa)
in dry
stalks
and
leaves.

Per ct.

. Ol6

. 020

No.

10.99

13- 13

0. 019

i7- 75

21. 27

. 025

II. 14

13- 40

. 016

15. 04

17.98

. 020

The percentages of iron in plants grown in acid solutions were slightly
greater than in plants grown in neutral and alkaline solutions. The
iron contents of plants from all three solutions were slightly increased by
increasing the iron from 0.002 to 0.008 gm. per liter.

Experiment 8 (Source of iron, in acid, neutral, and alkaline solu¬
tions: Ferric tartrate). — This experiment was conducted to see whether
other organic iron compounds would show the same availability as ferric
citrate. The two quantities of iron from ferric tartrate were used in the
three nutrient solutions. Two plants were grown in each flask, six
flasks were taken as a unit, and the units triplicated for each treatment.
The growth of plants and percentages of iron in the stalks and leaves
are given in Table VIII.

During the growth the plants were all of good color except No. 37 to 54
and 73 to 90, which were slightly yellowish from the fifteenth to thirtieth
day, but of good color the rest of the time.

With 0.002 gm. of iron per liter growth was considerably better in the
acid solution than in the neutral or alkaline. With 0.008 gm. of iron
per liter growth was practically equal in all three solutions, though
possibly there was a slight depression in the neutral solution. Increasing
the iron did not increase growth in the acid solution, but did increase
it in the neutral and alkaline solutions to a small extent.

The percentages of iron in plants grown in the acid solutions were very
slightly greater than in plants grown in the neutral and alkaline solutions.

Dec. i8» 1916

Assimilation of Iron by Rice

5i5

The iron contents of all plants except those grown in the alkaline solution
were increased by increasing the iron from 0.002 to 0.008 gm. of iron per
liter.

Table VIII. — Comparative weight of rice plants grown in acid, neutral, and alkaline
solutions , with ferric tartrate as the source of iron

Nutrient solution.

Quantity
of iron
per liter.

Flasks No.

Green
weight
of stalks
and
leaves.

Oven-

dry

weight

stalks

and

leaves.

Oven-

dry

weight

roots.

Average oven-
dry weight of —

Stalks

and

leaves.

Whole

plant.

(FejOa)
in dry
stalks
and
leaves.

Gm.

Per ct.

Acid ....

Do..

Neutral.

Do..
Alkaline
Do. .

o. 002

. 008

. 002

. 008

. 002

. 008

1- 6
7- 12
13- 18
19- 24

25“ 30
31- 36
37“ 42
43“ 48
49“ 54
55“ 60
61- 66
67- 72

73- 78
79- 84
85- 90

91- 96
97-102
IO3-IO8

67. 8l
56. 6l

62. 48
52. 05

67. 27

68. 46
S3- 58
47- 9S
51. 80
57-36
56. 86

63. 70
so. 22
48. 17
43. 01
S8.9S
63- 13
62. 90

8. 97

7. 92

8. 65
6.89
8. 91
9. 14
7. 01
6. 41

6. 91
7- 58

7. 72
8.60
6- 75
6. 52

6.06

8. 04
8. 56
8. 46

2. 20
i- 93
2. 04

1. 90

2. 19
2. 26
1. 52

1. 47

1. 58
1. 81

1. 78

2. 01
1. 84
1. 92

1. 70

2. 40
2. 32
2. 17

8. 51

10- 57

0. 022

8.31

10.43

. 030

6. 78

8.30

. 020

7-97

9.84

. 025

6.44

8.26

. 019

8-35

10. 65

. 022

Experiment 9 (Source of iron in acid and neutral solutions : Dialyzed
iron). — Dialyzed iron was tested as a source of iron in acid and neutral
solutions. From the previous experiments it seemed evident that this
form of iron would be of very low availability, but as it has been recom¬
mended as a source of iron in certain nutrient solutions it was thought
advisable to make the test. The comparison between dialyzed iron and
ferrous sulphate in the neutral solution was carried out at one time and
the comparison between the neutral and acid solutions with dialyzed
iron at another time. The two tests are combined in one table for
conciseness.

Two plants were grown in each flask, six flasks were taken as a unit,
and the unit duplicated for each treatment. Plants in flasks 49 to 84,
the comparison between acid and neutral solutions, were grown but 25
days; all others were grown 40 days. The data of growth are given in
Table IX.

All plants except No. 25 to 48 were markedly chlorotic at all times.

Dialyzed iron apparently afforded practically no available iron when
used in the neutral solution at the rate of either 0.002 or 0.008 gm. of iron
per liter. In the acid solution it was slightly more available than in the
neutral, although utterly inadequate for the needs of the plant.

Journal of Agricultural Research

Vol. VII, No. i2

516

Table IX. — Comparative weights of rice plants grown in acid and neutral solutions , with

dialyzed iron as the source of iron

Nutrient

solution.

Neutral... .

Do .

Do......

Do .

Acid .

Quantity
of iron,
per liter.

O. 002
. 008
. 002
. 008
. 008
. 008

Source of iron.

Dialyzed iron,.

. . .do .

f Ferrous sul-

\ phate.

. . . .do .

Dialyzed iron..
_ do .

Flasks

No.

Green
weight
of stalks
and
leaves.

Oven-
dry
weight
of stalks
and
leaves.

Oven-
dry
weight
of roots.

Average oven-dry
weight of —

Stalks. •
and
leaves.

Whole

plant.

Gm.

1-6

O. 67

O. 15

O.054

. 7-12

. 61

. 16

•054

O. 16

0. 21

13-18

•63

• 17

• 054

.19-24

. 61

. 16

• 054

. 16

. 22

25-3°

62. 19

8. 27

i. 86

.31-36

65.07

8.48

1. 92

8. 38

IO. 27

37-42

128. 62

16.31

3-73

.43-48

*37- 50

16. 73

4. 01

16. 52

20. 39

49-54

.90

.16

. 062

,55-6°

•99

. 19

. 062

. 18

. 24

73-78

4-38

.67

. 182

.79-84

4. 26

.64

. 176

.66

.84

Experiment io (Effect of applying ferrous sulphate to leaves of plants
grown in solutions where iron was markedly unavailable). — In experiment
4 rice failed to grow appreciably in the alkaline solution when 0.008
gm. of iron per liter from ferrous sulphate was used, although it made
a fair growth with 0.002 gm. of iron, and in the previous experiment no
perceptible growth was made in the neutral solution with dialyzed iron.
In order to be certain that the inability of the plants to grow in these
solutions was due to a lack of available iron, the following test was made
of applying iron to the leaves.

The leaves of plants in these two solutions were brushed with ferrous
sulphate, and two control lots were not brushed. The specially treated
plants were brushed once with a 0.1 per cent solution of ferrous sulphate,
twice with a 0.2 per cent solution, and three times with a 0.4 per cent
solution. This was done so that no iron could get from the leaves to the
solution. Two plants were grown in each flask, six flasks were taken as a
unit, and the units triplicated for each treatment. The plants were
grown but 25 days, as results were then decisive, the unbrushed plants
having ceased to grow perceptibly some time before. The data on
growth are given in Table X.

The leaves brushed with the ferrous-sulphate solution became green in
two or three days. New leaves appearing at intervals between the
brushings were strongly chlorotic, but became green quickly when
treated with the solution. Plants 1 to 12 and 25 to 36, ones that were not
brushed at any time, were always strongly chlorotic and made no appre¬
ciable growth. The comparative growths of the brushed and unbrushed
plants show decisively that the inability of plants to grow in these solu¬
tions was due to a lack of available iron.

Dec. 18, 1916

Assimilation of Iron by Rice

5i7

Table X. — Comparative weights of rice plants grown in two solutions where the iron was
unavailable, but with the leaves treated with ferrous sulphate

Nutrient

solution.

Quantity of iron per
liter.

Source
of iron.

Treatment.

Green weight of stalks

and leaves.

Oven-dry weight of

stalks and leaves.

Oven-dry weight of

roots.

Alkaline . .

Gm,

0. 008

[Ferrous
< sul-
[ phate.

| Leaves not brushed . . j

1-6

7-12

Gm.

2. 47
2. 48

Gm.

0-89

•39

Gm.
O. 13 1
. 126

Do .

. 008

. .do .

/heaves brushed withf
l ferrous sulphate. \

13-18

19-24

14. 63
i5- 95

2. 14
2. 32

.630
. 602

Neutral ...

. 008

jDialyzed
\ iron.

jheaves not brushed . . j

25-30

3i_36

.90

•99

. 16
. 19

. 062
. 062

Do .

.. 008

. . .do .

f Leaves brushed withf
\ ferrous sulphate. \

37-42

43-48

11.44

9.40

1. 69
1. 40

.463
• 383

Average
oven-dry
weight of —

ilks and

leaves.

Gm.

°* 39

O. 52

2. 23

2. 85

. 18

. 24

i- 50

I. 92

Experiment ii (Effect of increasing the phosphates in the neutral
solution). — Since there is always more or less iron in the nutrient solu¬
tions precipitated as phosphates, it was important to see whether in¬
creasing the phosphates in the solution would affect the growth of plants.
In this test the neutral solution used in the preceding experiments was
compared with a solution which contained double the quantity of mono-
and dibasic potassium phosphate, but which was otherwise similar.
Two tests were run at different times, one using ferric chlorid as the
source of iron and one using ferric citrate. For the sake of conciseness
they are combined in one table.

In the test with ferric chlorid two plants were grown in each flask, 6
flasks were taken as a unit and the units triplicated for each treatment.
In the test with ferric citrate one plant was grown in each flask, 12 flasks
were taken as a unit, and the units duplicated for each treatment. The
data on the growth of these plants are given in Table XI.

The ferric-chlorid plants were all of rather poor color, those with
p.002 gm. of iron per liter being poorer than those with 0.008 gm., but
no differences were apparent between plants in the ordinary and double¬
phosphate solutions. The ferric-citrate plants in both ordinary and
double-phosphate solutions were of good color at all times.

In regard to growth and iron content, there were no appreciable dif¬
ferences between plants in the ordinary and double-phosphate solu¬
tions, although the very slight differences in growth were always in favor
of the ordinary solution.

5i8

Journal of Agricultural Research

Vol. VII, No. 12

Table; XI. — Comparative weights of rice plants grown in neutral solution with phosphates
doubled , with ferric chlorid or citrate source of iron

Green weight of stalks and

leaves.

Oven-dry weight of stalks

and leaves.

Gm,

Gm.

' 1-6

28. 22

3* 54

■ 7-12

34* 30

4* 24

28. 76

3- 63

19-24

29. 84

3- 78

25-30

27*43

3-49

.3*-36

29. 42

3-89

37-42

42. 12

5-21

‘ 43-48

37* 24

4- 56

:49-54

38. 16

4.76

55-6o

39* 23

4.88

- 61-66

37* 30

4. 72

.67-72

34* 51

4-39

1-12

202. 36

29-95

>13-24

192. 20

28. 72

/25-36

181. 78

27*33

137-48

189. 34

28.85

Nutrient solution.

Neutral.

Neutral, with dou¬
ble phosphates.

Neutral.

Neutral, with dou¬
ble phosphates.

Neutral .

Neutral, with dou¬
ble phosphates.

Source of iron.

O. 002

002

}■

008

. 008
►. 008

Ferric chlorid.

....do .

. .do .

— do .

Ferric citrate.

do.

o.68

.84

.69

* 74
. 6S

.78

I. 20\
1. 03
I. IO

•99
•93
.90
7. 41
7. 11

6. 74

7. 21

Average
oven-dry
weight of—

eft El

Gm.

3. 80

3- 72

4- 84

4.66
29- 34.
28.* * ’

Gm,

4- 54

4*45

5- 95

0935

5. 60

36-65
07

£{8
.3 1

Per ct.

O. 020

021

023

022

OI9

021

SOLUBILITY OF IRON IN THE} NUTRIENT SOLUTIONS

An attempt was made to estimate the soluble iron in some of the
nutrient solutions by filtering the solutions to remove precipitated iron
and determining iron in the filtrate. Solutions 1 to 9 (used in experi¬
ment 4) were analyzed after they had stood various lengths of time both
with and without plants growing in them. Iron in all the solutions was
determined by the colorimetric method with potassium sulphocyanate,
but solutions 1, 2, 4, 5, 7, and 8 without the growth of plants were also
analyzed by the method of Stokes and Cain. Practically the same re¬
sults were obtained by the two methods. The results are shown in
Table XII. Figures for iron in the solutions without plants are the
averages of two or more determinations, while the figures for iron in solu¬
tions in which plants had grown represent single determinations. The
results for solution 5 without plants are not trustworthy, as this solution
could not be filtered absolutely clear by any device the writers considered
permissible to employ.

In a comparison of the growth of plants in the different solutions with
the amount of iron in the filtered solutions, almost as many discrepancies
as agreements are apparent. One striking discrepancy is the small
amount of iron in the filtrate of No. 3 and the good growth of plants in
this solution. Repeated analyses of the precipitate left on the filter

Dec. 18, 1916

519

Assimilation of Iron by Rice

showed that practically all the iron added to this solution was precipitated.
As this solution evidently had less precipitate than some others, the
character of the precipitate probably varied somewhat in the different
solutions. The amount of iron present in the filtered solutions in which
plants had grown also failed to correspond to the growth of plants.

Table XII. — Quantity of iron (in grams) in filtered solutions at different times after
addition of iron to solution, with and without growth of plants in the solution

Quantity of iron in 1,000 liters of filtered
solution.

Nutrient

solution.

Source of iron.

No plants in solu¬
tions.

Plants grown in solutions.

Acid. . . .

. do. .

Neutral.

. do. .

Alkaline

. do. .

Ferrous sulphate . . .

_ do .

Ferric citrate .

Ferrous sulphate . . .

_ do .

Ferric citrate .

Ferrous sulphate . . .

... .do .

Ferric citrate .

o. 07
•30

•03
•49
•03
.87
. 40
•05
• 57

o. 05

•43
.07
1. 27
• 17
1. 63
i- 33
. 18
1. 00

o. 05
• 23
•05
i- i3
. 10
i- 33
.86

0.03
.07
•03
.80
.07
.80
. 60

• 53

The growth of plants did not agree with the quantity of iron found in
the filtered solutions because, without doubt, colloidal iron, as well as
truly soluble iron, was present. Previous work having shown that
colloidal iron was not available for rice (6) , the determinations in Table XII
did not represent the available iron. At this great dilution it was, of
course, impossible to distinguish analytically between colloidal and soluble
iron or between ferrous and ferric iron. While the existence of colloidal
iron could not be definitely demonstrated in these solutions, it is well
known that most ferric salts in dilute solutions are more or less com¬
pletely hydrolytically dissociated into colloidal ferric hydroxid and the
acid. Moreover, a test showed that dialyzed iron could not be distin¬
guished from distilled water with regard to color or filtration when used
in somewhat greater concentration than iron was present in the above
filtrates.

In solutions where ferric iron was used iron was probably present in the
following forms : (i) As precipitated ferric phosphate and hydroxid, (2)

66848°— 16- — 2

520

Journal of Agricultural Research

Vol. VII, No. 12

as colloidal ferric hydroxid, (3) as soluble undissociated iron compounds,
and (4) as ionized iron. There was probably a balance between these
forms of iron, as after filtering the nutrient solutions more precipitate
formed on standing a short time.1 From determinations of iron in the
filtered nutrient solutions it is evident that more or less half the iron was
precipitated as phosphate and hydroxid. The amount and composition
of this precipitate probably varied somewhat in the different solutions (9).
The greater part of the remaining iron was probably present as colloidal
ferric hydroxid. The available iron, which included the soluble undis¬
sociated and ionized iron, was undoubtedly extremely small and was gov¬
erned chiefly by the completeness of the hydrolysis of the dissolved iron.

The amount of iron hydrolized would depend on the reaction of the
solution, being less in acid solutions, and would also depend on the form
in which iron was added, being less with the less ionized organic salts.
The effect of the form of iron and the reaction of the solution on the assimi¬
lation of iron by rice is thus easily comprehensible.

While the amount of soluble or available iron could not be determined
analytically, some idea of the very small amount present could be obtained
by comparing the amount of iron absorbed by the plants with the total
amount of iron added in the volume of solution available during growth.
This calculation showed that the plants absorbed only one-fifteenth to
one two-hundredth part of the iron supplied, even in solutions where
growth was obviously inhibited by lack of iron. If it is assumed that one
* fifteenth to one two-hundredth part of the iron added to the solutions
was in a soluble condition, the concentration of the soluble iron in some
cases would have been from 0.13 to 0.01 part per million. A more prob¬
able assumption is that the concentration of soluble iron present at any
one time was even lower and that as iron was removed by the plant more
went into solution. It thus appears that in some cases the amount of
iron in true solution must have been too small for slight differences to be
accurately determined. It is evident that in certain nutrient solutions
rice can assimilate sufficient iron’ when the concentration of soluble iron
is probably less than 1 part in 10,000,000.

Comparative analyses in Table XII of solutions in which plants had and
had not grown show that at least the amount of colloidal iron was notably
diminished by the growth of plants in the solution. The colloidal iron
was evidently precipitated.

SUMMARY OF EXPERIMENTAL RESULTS

The results of the previous culture experiments showed plainly that
rice in nutrient solutions was not particularly sensitive to an acid or alka¬
line reaction per se. Apparently the reaction of the nutrient solution
affected the growth of rice only through influencing the availability of the

1 The precipitates in these dilute nutrient solutions were iron compounds, as before the addition of iron
salts the solutions were perfectly clear.

Dec, 18, 191^

Assimilation of Iron by Rice

521

iron, since the relative growths made in acid, neutral, and alkaline solu¬
tions depended on the kind and quantity of iron supplied. This is well
shown by Table XIII, which summarizes the relative growths of stalks
and leaves made in the three solutions with different kinds and quantities
of iron. The growth of stalks and leaves 1 made in the acid solution is
always taken as 100; growths in the neutral and alkaline solutions are
expressed relative to 100.

Table XIII. — Relative growths of rice plants in acid , neutral , and alkaline solutions
with different sources and amounts of iron

Source of iron in nutrient solutions.

Iron per
liter added
to nutri¬
ent solu¬
tions.

Relative growths in —

Table

from

which

data

were

calcu¬

lated.

Acid

solu¬

tion.

Neutral

solu¬

tion.

Alka¬
line so¬
lution.

Gm.

Ferrous sulphate. . .

0. 002

100

Do .

100

7 A

III

Do. . . . .

100

0 ■

Do . . .

. 008

Do .

. 004

f 100

105

Do . ■. .

. 002

100

107

Do . . .

008

100

132

Do . . .

I OO

III

Ferric chlorid .

. 002

100

Do . .

. 008

IOO

107

Ferric citrate . .

. 002

IOO

III

Do .

IOO

Do . . .

ICO

Vt-

IOI

104

VII

Do .

. 008

IOO

VII

Ferric tartrate .

. 002

IOO

VIII

Do .

. 008

IOO

VIII

Dialyzed iron .

. 008

IOO

It is evident that with 0.002 gm. of iron per liter, growth was more
or less best in the acid solution, while with 0.008 gm. of iron per liter
growth was best in the neutral solution with some forms of iron. Growth
with most, but not all, forms and quantities of iron was strikingly inferior
in the alkaline solution.

The effect on growth of increasing the quantity of iron in the different
solutions is significant in throwing light on the availability of the iron
in these solutions. In Table XIV are shown the extents to which growth
was increased or decreased by increasing the iron in the three solutions
from 0.002 to 0.008 gm. per liter. The growth made in each solution
(acid, neutral, or alkaline) with 0.002 gm. of iron per liter is taken as 100,
and the growth made with 0.008 gm. of iron per liter in these solutions
is expressed relative to 100.

1 In this and previous comparisons more significance was attached to the weight of stalks and leaves
than to the weight of the whole plant. The weights given for the roots must have been very slightly in
excess of the true values, as the roots were always more or less contaminated with a precipitate that could
not be removed by washing. Also, the root growth relative to top growth is markedly influenced in some
cases by an insufficiency of a mineral nutrient.

522

Journal of Agricultural Research voi.v1x.No.12

Table XIV. — Relative growths of rice plants with 0.002 and 0.008 gm. of iron per liter

in the different solutions

Source of iron in nutrient solution.

Growth 0 with 0.008 gm.
ot iron per liter in —

Table

from

which

data

were

calcu¬

lated.

Acid

solu¬

tion.

Neu¬

tral

solu¬

tion.

Alka¬

line

solu¬

tion.

Ferrous sulphate .

141

179

212

209

197

141

127

100

13s

118

VII

VIII

Do . . .

Do. . . .

Ferric chlorid . .

131

140

Do .

Dialyzed iron . ■ . / . . .

Ferric citrate .

162

130

Ferric tartrate . .

a Growth with 0.002 gm. of iron per liter =100.

In most cases an increase in the amount of iron added to the solution
produced a marked increase in growth. With ferric tartrate in the acid
solution an increase in iron produced no increase in growth, as apparently
the smaller quantity of iron was adequate. In the alkaline solution
with ferric citrate and with ferrous sulphate there were, respectively,
no increase and a striking depression in growth, following an increase
in the amount of iron added to the solution. Increasing the ferric
citrate in this solution apparently did not increase the soluble iron and
increasing the ferrous sulphate must have decreased it, as the smaller
amounts of iron were also inadequate in these solutions.

From the growth of the plants it appears that ferrous sulphate, ferric
citrate, or ferric tartrate used in proper quantities afforded sufficient
iron in acid or neutral solutions. With the two quantities of iron used
ferric chlorid was inferior as a source of iron, and dialyzed iron was utterly
inadequate. Ferric tartrate was the only form of iron tried which
appeared to afford sufficient iron in the solution alkaline with calcium
carbonate.

The addition of carbon black to the neutral nutrient solution with
0.002 gm. of iron per liter very slightly depressed growth, but carbon
black in the solution with 0.008 gm. of iron did not affect the yield.
Treating the distilled water with carbon and filtering previous to its use
in the nutrient solution increased growth considerably over that in the
ordinary nutrient solution.

Doubling the phosphates in the neutral nutrient solution did not
measurably affect growth when either ferric chlorid or citrate was used
as the source of iron.

In all the cultural tests, plants grown in acid solutions contained more
iron than those grown in corresponding neutral or alkaline solutions.
Plants grown in neutral solutions contained more iron than those grown

Dec. 18, 1916

Assimilation of Iron by Rice

523

in corresponding alkaline solutions when ferrous sulphate or ferric chlorid
was the source of iron, but about equal percentages when ferric citrate
or tartrate was the source of iron. With ferric chlorid, citrate, or
tartrate an increase in the quantity of iron from 0.002 to 0.008 gm. of iron
per liter in acid, neutral, or alkaline solution raised the percentage of iron
in the plant. Increasing the ferrous sulphate in the acid solution did not
increase the percentage of iron in the plants, while in the neutral solution
it did have this effect. The percentages of nitrogen, phosphoric acid,
lime, magnesia, and carbon-free ash in plants grown in six different
solutions did not vary appreciably. The relative percentages of iron in
the plants thus agreed with relative growths in showing that the amount
of available iron in most solutions was the main factor controlling
growth.

In regard to the percentages of iron in the plants two anomalous fea¬
tures are apparent in the preceding tables: (1) The percentages of iron
in plants supplied with ferrous sulphate were higher than in plants
supplied with ferric citrate, and (2) in two cases plants which made no
growth because of a lack of iron contained as high a percentage of iron as
plants of good growth.

Experiments 3 and 4 each afford a comparison of ferric citrate and
ferrous sulphate in the three different nutrient solutions. In all cases
0.002 gm. of iron per liter from ferric citrate produced a better growth of
plants with a lower percentage of iron than did the same quantity of iron
from ferrous sulphate. As tests 2, 3, 4, and 7 show, neither of these forms
of iron, supplied at this rate, furnished sufficient iron for the maximum
needs of the plant. Therefore the higher percentages of iron in the
ferrous-sulphate plants could not have been due to excessive consumption.
About the only explanation of the anomaly that occurs to the writers is
that the ferrous-sulphate plants contained a certain amount of iron
which was ineffective in the metabolism of the plant. In the solutions
to which ferrous sulphate was added undoubtedly both ferrous and ferric
iron existed in solution, while in the solutions with ferric citrate there was
only ferric iron. It is possible that in the solutions with ferrous sulphate
the plants absorbed both ferrous and ferric iron and that the ferrous iron
was not so effective in the plant as the ferric iron. The ferrous-sulphate
plants might therefore contain a certain amount of effective and non*
effective iron. This explanation is somewhat doubtful, as it is not, to the
knowledge of the writers, supported by similar well-established facts.

The second irregularity is the high percentage of iron in plants 1 to 12,
experiment 1, and plants 85 to 96, experiment 4, the former being grown
without the addition of any iron to the solution, and the latter obtaining
practically no iron from the alkaline solution, as proved by experiment
10. . It is, of course, possible that the high percentages of iron were due
to contamination, but this is thought not to be the cause. The plants
made practically no growth and were so strongly chlorotic that they

524

Journal of Agricultural Research

Vol. Vlf, No. 12

could have elaborated scarcely any organic matter; the leaves were
especially thin and often withered as soon as formed. It is possible that
the high percentage of iron in the dry substance might have been due to
the fact that enough iron was not present in the plant to start the pro¬
duction of carbohydrates, which would have lowered the percentage.
While this explanation opens up several points not covered by the present
investigation, it is supported somewhat by the following test (Table XV,
first test), where plants grown for 13 days with iron and then for 13 days
without iron contained a lower percentage of iron than plants grown the
full 26 days without iron.

Table XV. — Percentages of iron in rice plants grown with and without iron

Test and treatment of plants.

Average
dry weight
of stalks
and leaves
per plant.

Percentage
of iron
(F2O3). in
dry stalks
and leaves.

First test:

Grown for 13 days, without the addition of iron to the solu¬
tion . .

Grown for 26 days, without the addition of iron to the solu¬
tion .

Grown for 13 days, without the addition of iron to the solu¬
tion and then for 13 days with iron .

Grown for 26 days, with the addition of iron to the solution. .

Second test:

Grown 13 days, without the addition of iron to the solution. .
Grown 26 days, without the addition of iron to the solution. .
Grown 40 days, without the addition of iron to the solution. .
Grown 40 days, with the addition of iron to the solution .

Gm.

0. 017

0. 020

. 026

.097

. 020

* 311

. 027

. 016

.015

. 019

. 024

. 026

.049

547

. 030

In the first test plants grown 26 days without iron contained a higher
percentage of iron than those grown 13 days, although there was practi¬
cally no increase in growth. This peculiarity was repeated in a second
test (Table XV).

From the regularity of these results it appears they were due to pecu¬
liarities in the metabolism of plants grown without iron. That plants
grown for 40 days in the iron-free solution made only slightly more growth
than those grown for 13 days, although they contained much more iron,
is probably due in part at least to the immobility of iron in the plant (7).
Considering the extremely low concentration from which plants absorb
their iron in solutions supplied with iron, it is probable that the plants
obtained traces of iron even from the “iron-free” solutions, as it
would be very difficult to exclude one part of iron in several hundred mil¬
lion of solution.1

The attempt to determine the soluble iron in the various solutions was
unsuccessful because of the impossibility of distinguishing between col-

1 If the difference between the Iron present in plants grown 13 days and 40 days (Table XV) is due to iron
absorbed from the solution, 0.0000055 gm. of ferric oxid (FetOs) was absorbed between the thirteenth and
fortieth day by the three plants grown in each flask. As 1,600 c. c. of solution were supplied per flask during
this interval, the concentration of soluble iron, expressed as Fe*Oj, would have been less than parts in

1,000,000,000.

Dec. x8, 1916

Assimilation of Iron by Rice

525

loidal and soluble iron at relatively great dilutions. From the amount of
iron supplied and the amount absorbed it appeared that there could have
been not more than approximately 1 part in 10,000,000 in solution at one
time and very probably much less. The much larger quantities of iron
found in the filtrate of the nutrient solutions were attributed to colloidal
iron, which a previous test showed to be unavailable to the plant.

DISCUSSION OF RESULTS

The reactions of the solutions used in this work could not be exactly
measured by titration because of the presence of interfering ions. It is
evident, however, that the acid solution was relatively quite strongly acid
as it contained only monopotassium phosphate besides a trace of sulphuric
add, the neutral solution was nearly neutral as it contained a mixture of
mono and di-potassium phosphates, and the alkaline solution was slightly
alkaline from the presence of calcium carbonate and its reaction with the
phosphates.

The preceding tests demonstrated that the growth of rice was markedly
dependent on the quantity and form of iron added to these nutrient solu¬
tions and apparently dependent on the reaction of the solution only so far
as it affected the availability of the iron. These facts are important in bear¬
ing on the nature and cause of lime-induced chlorosis, rice bdng markedly
affected with this nutritional disturbance. Previous work in soil cultures
showed pretty decisively that lime-induced chlorosis was not caused by
lack of any mineral nutrient except possibly iron, but did not show
whether the reaction of the soil in itself affected the plants. The present
study seems to substantiate the previous work and show, moreover, that
rice is not particularly sensitive to the reaction per se, provided the iron
supply is maintained. While in soils there are many more factors affect¬
ing the availability of iron than in nutrient solutions, the preceding results
point strongly to calcium carbonate diminishing the quantity of available
iron in a soil through affecting the reaction. Direct evidence on this latter
point is afforded by the results of Morse and Curry (11) and Ruprecht and
Morse (12). The results also point to different iron compounds, particu¬
larly the organic and inorganic compounds, varying greatly in their availa¬
bility in calcareous soils. The existence of organic iron compounds in the
soil has been pointed out by Hartwell and Kellogg (8).

While all the tests reported here were carried out with rice, certain re¬
sults are of general interest in relation to the proper composition of plant-
nutrient solutions. The extent to which growth was dependent on the
iron supply in the previous tests shows how important it may be to con¬
sider the form and quantity of iron used in the nutrient solution. Evi¬
dently the addition of a few drops of a dilute iron solution, as recom¬
mended in most plant physiologies, may not insure an adequate supply
of iron. While the color of the leaves will indicate a marked deficiency
of iron, a slight deficiency may materially diminish the yield without mate-

526

Journal of Agricultural Research

Vol. VII, No. 12

rially affecting the appearance of the plants. It is, therefore, not suffi¬
cient to judge the adequacy of the iron supply by the mere color of the
leaves. Also a sufficiency of iron can not be insured by simply increasing
the quantity of iron added to the solution as this may even diminish the
amount of available iron.

Although this work demonstrated chiefly the reaction of the solution
and the source of iron as factors influencing the availability of iron, the
importance of other conditions was indicated. Undoubtedly the fre¬
quency with which the nutrient solution is changed influences appreciably
the availability of the iron. The results in Table XII showed that the
precipitated iron increased with the age of the solution and also with the
growth of plants in the solution. This increase in precipitated iron prob¬
ably accompanied a certain decrease in soluble iron, as there was doubt¬
less some balance between the precipitated, colloidal, and soluble iron.

Doubtless absorption also affected the quantity of soluble iron in certain
cases. The diminution in growth following the addition of carbon black
to the solution with 0.002 gm. of iron per liter in experiment 10 was prob¬
ably caused by a decrease in soluble iron through simple adsorption. In
experiment 4 the decrease in available iron following an increase in the
quantity of iron added to the alkaline solution may have been partially
due to the adsorption of soluble iron by the larger precipitate of iron.

On the basis that hydrolysis was the chief factor determining the
amount of soluble iron in the solutions, one would not expect that increas¬
ing the phosphate^ in the neutral solution would appreciably affect the
iron assimilated by the plants. Experiment 1 1 confirmed this.1

This work furnished no evidence of rice being able to assimilate other
than soluble iron, but tended to confirm previous work showing that even
colloidal iron is unavailable (6). It did show, however, that rice used
iron, which must have been present in exceedingly low concentrations.

The facts established concerning the availability of iron in these nutrient
solutions help explain results obtained by certain investigators with other
nutrient solutions. It is realized, however, that the availability of iron
in each nutrient solution probably varies according to the concentration
and composition of the solution as well as according to the method of con¬
ducting the cultural test.

The chlorosis of peas in certain nutrient solutions observed by Maze et
al. (10) was evidently not due to the excretion of calcium carbonate by
the roots, but to salts in the solution depressing the availability of the
iron. Probably the potassium silicate, which is strongly hydrolyzed into
potassium hydroxid, was chiefly responsible for the nonavailability of
the iron. The chlorosis observed by Von Crone (3) with plants in cer¬
tain nutrient solutions was not due to the soluble phosphates, but prob¬
ably to a deficiency of iron. In certain solutions the lack of iron was

1 If the concentration of phosphate ions had been very low, such as would be afforded by tricalcium or
ferric phosphate, doubtless the addition of soluble phosphates would have depressed the available iron.

Dec. 18, 1916

Assimilation of Iron by Rice

527

doubtless due to the reaction of the solution, calcium carbonate being
present. In other solutions where ferric phosphate was the source of
iron, addition of soluble phosphates produced chlorosis. The soluble
phosphates here evidently precipitated the very small amount of iron that
went into solution from the decomposition of ferric phosphate. Von
Crone made the mistake of assuming that because ferric or ferrous phos¬
phate furnished sufficient iron in some solutions that it did in all solu¬
tions. Doubtless the reason that ferrous phosphate furnished sufficient
iron in Von Crone’s solution was due to the fact that the concentration of
phosphate ions was also particularly low.1

The results of this work, as well as that of Takeuchi (15) and Benecke
(1), showed that soluble phosphates do not in themselves produce chlo¬
rosis. Benecke in his criticism of Von Crone’s solution failed to take
into account that part of the iron he determined in his tests of the solu¬
bility of iron phosphates was colloidal iron.

SUMMARY

Rice was grown in acid, neutral, and alkaline solutions with different
forms and quantities of iron to determine whether rice is particularly
sensitive to the reaction of the solution and whether the reaction of the
solution influences the assimilation of iron.

In nearly all cases growth was much better in the nutrient solutions
employed with 0.008 gm. of iron per liter than with 0.002 gm. When
judged by the growth of plants ferrous sulphate, ferric citrate, and ferric
tartrate afforded sufficient iron when used in proper quantities in the
acid and neutral solutions. Ferric chlorid was an inferior source of
iron, and dialyzed iron utterly inadequate. Only ferric tartrate fur¬
nished sufficient iron in the alkaline solution.

Plants grown in the acid solutions contained the highest percentages
of iron. Plants grown in the neutral solutions contained higher per¬
centages of iron than those grown in the alkaline solutions when some
forms of iron were used, but equal percentages when other forms of iron
were used. The percentages of nitrogen, phosphoric acid, lime, mag¬
nesia, and carbon-free ash in plants grown in six different solutions did
not vary appreciably when compared with the iron content.

It was evident that rice was not particularly sensitive to the reaction
of the solution, except as the reaction influenced the availability of the
iron. This substantiates previous work in showing that lime-induced
chlorosis is caused by a lack of iron and indicates strongly that the only
action of carbonate of lime in inducing chlorosis lies in diminishing the
availability of the iron.

The amount of available iron in the different solutions could not be
determined analytically, because of the impossibility of distinguishing

1 Ferrous and ferric phosphate evidently afford iron and phosphoric acid not through the dissolving action
of plant roots, as Crone believed, but through the hydrolytic decomposition of these compounds. As a
result of this decomposition, colloidal iron hydroxid is formed, as well as phosphate ions and a very small
amount of soluble iron <9, 2).

528

Journal of Agricultural Research

Vol. VII, No. 12

between colloidal and soluble iron. Calculations showed, however, that
the concentration of available iron in many cases must have been less
than one part in 10,000,000 of solution

Reference was made to the bearing of these results on the proper com¬
position of plant nutrient solutions.

LITERATURE CITED

(1) Benecke, Wilhelm.

1909. Die von der Cronesche Nahrsalzlosung. In Ztschr. Bot., Bd. 1, p. 235-252.

(2) 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.

(3) Crone, Gustave von der.

1904. Ergebnisse von Untersuchungen fiber die Wirkung der Phosphorsaure auf die
hfihere Pflanze und eine neue Nahrlosung... 46 p. Bonn. Inaug. Diss.

(4) Gile, P. L.

1911. Relation of calcareous soils to pineapple chlorosis. Porto Rico Agr. Exp.
Sta. Bui. 11, 45 p., 2 pi. ( 1 col.).

(5) - and Ageton, C. N.

1914. The effect of strongly calcareous soils on the growth and ash composition
of certain plants. Porto Rico Agr. Exp. Sta. Bui. 16, 45 p., 4 pi.

(6) - and Carrero, J. O.

1914. Assimilation of colloidal iron by rice. In Jour. Agr. Research, v. 3, no. 3,

p. 205-210.

m - •

1916. Immobility of iron in the plant. In Jour. Agr. Research, v. 7, no. 2, p.
83-87. Literature cited, p. 87.

(8) Hartwell, B. L., and Kellogg, J. W.

1906. On the effect of liming upon certain constituents of a soil. In R. I. Agr.

Exp. Sta. Ann. Rpt. 1904/05, p. 242-252,

(9) Lachowicz, B.

1892. Uber die Dissociation der Ferriphosphat durch Wasser und Salzldsungen.
In Monatsh. Chem., Bd. 13, p. 357.

(10) MazE, P., Ruot, and Lemoigne.

1913. Chlorose calcaire des plantes vertes. R61e des excretions des racines dans
1 'absorption du fer des sols calcaires. In Compt. Rend. Acad. Sci.
[Paris], t. 157, no. 12, p. 495-498.

(11) Morse, F. W., and Curry, B. E.

1908. A study of the reactions between the manurial salts and clays, mucks and

soils. In N. H. Agr. Exp. Sta. Ann. Rpt. 19/20, t9o6/o8, p. 271-293,
4 hg-

(12) Ruprecht, R. W., and Morse, F. W.

1915. The effect of sulfate of ammonia on soil. Mass. Agr. Exp. Sta. Bui. 165

p. 73-90.

(13) Schreiner, Oswald, and Reed, H. S.

1909. The r61e of oxidation in soil fertility. U. S. Dept. Agr. Bur. Soils Bui.
561 52 P-

(14) Stokes, H. N., and Cain, J. R.

1907. On the colorimetric determination of iron with special reference to chem¬

ical reagents. In U. S. Dept. Com. Bur. Standards Bui., v. 3, no. 1,
p. 115-156, 5 fig.

(15) Takeuchi, T.

1907. Kdnnen Phosphate Chlorose erzeugen? In Bui. Col. Agr., Tokyo Imp.
Univ., v. 7, no. 3, p. 425-428.

INFLUENCE OF BORDEAUX MIXTURE ON THE RATES
OF TRANSPIRATION FROM ABSCISED LEAVES AND
FROM POTTED PLANTS

By William H. Martin,1

Assistant in the Department of Plant Pathology, New Jersey Agricultural Experiment

Station

INTRODUCTION

Since the introduction of Bordeaux mixture much work has been
done on the effects which this fungicide has on the growth of normal
plants. For more than 20 years this work has occupied the attention
of pathologists. During this time progress has been made, but there
still remains much to be done. Many observations have been made as
to the effect which an application of this spray has upon transpiration
rates, and a review of the literature brings out the fact that the conclu¬
sions drawn from these observations are conflicting.

Rumm (15)2 found that when abscised leaves were placed in water
the unsprayed ones wilted first; from this he concluded that there is a
decrease in the rate of transpiration following an application of the spray.
Clinton (4) expressed the view that the water pores and stomata of potato
leaves are clogged by the spray and as a result transpiration is decreased.
Schander (16), Bayer (2), and Miiller-Thurgau (14) each expressed the
opinion that decreased transpiration rates follow spraying with Bordeaux
mixture.

As a result of extensive investigations, Frank and Kruger (8, 9) con¬
clude that the water loss from sprayed plants is greater than from plants
not so treated. Bain (1) likewise found that an increased rate of trans¬
piration occurs in peach seedlings as a result of spraying. He arrived at
this conclusion from the fact that he found it necessary to supply water
more frequently to the roots of seedlings that had been sprayed. More
recently, however, Duggar and Cooley (6, 7) have furnished direct evi¬
dence bearing upon this question. In a series of very carefully per¬
formed experiments they have demonstrated that not only does a film
of Bordeaux mixture on the leaves of castor beans, tomatoes, and potatoes
increase their rates of transpiration but that other surface films have a
similar effect. They have brought out the fact that certain specific
qualities of the films applied are definitely related to the phenomenon of
increased transpiration. They further state that the color of the film
applied is also a factor to be considered in this connection.

1 Thanks are due to Dr. M. T. Cook for suggesting the problem and to Dr. J. W. Shive for valuable sug¬
gestions during the progress of the work and for aid and criticism in the preparation of the manuscript.

% Reference is made by number to “ Literature cited,” p. 547-548.

Journal of Agricultural Research, Vol. VII, No. 12

Department of Agriculture, Washington, D, C. Dec. 18, 1916

gn N.J.— 4

(529)

530

Journal of Agricultural Research

Vol. VII, No.

While the last-named authors have definitely established the fact that
surface films with certain specific characteristics have an accelerating
influence on rates of transpiration when applied to the leaves of castor
bean, tomato, and potato plants, the problem here involved seemed to
be of sufficient importance to warrant further investigation in a different
locality and under different conditions.

It was the purpose of the experiments here reported to determine the
influence of Bordeaux mixture on the rates of transpiration of abscised
leaves of several species, as well as to determine the effect of this spray
material upon the rates of water loss from a variety of potted plants.
The experiments were carried out in the greenhouse of the Department
of Plant Pathology of the New Jersey Agricultural Experiment Station.

EXPERIMENTS WITH ABSCISED LEAVES

Experimental methods. — Abscised leaves of radish (Raphanus
saiivus L.), bean ( Phaseolus vulgaris L.)> Swiss chard {Beta cycla L.)>
Hibiscus cardinalis, Clerodendrum balfouri , Caladium sp., Datura mete -
loides , and castor bean (Ricinus communis L.) were used. The leaves,
together with portions of the stems, were severed from the plants and
the cut ends immediately placed in water. They were then taken to the
greenhouse room, where the experiments were carried out. The leaves
were now cut off under water and the petioles were inserted into
Erlenmeyer flasks having a capacity of 180 c. c. ; the flasks were nearly
filled with water. A layer of cotton was then placed tightly around the
leaf petiole just at the surface of the water; this served to hold the leaf
in place. The flasks were sealed by pouring melted wax over the cotton
around the leaf petioles. This wax was prepared according to the for¬
mula of Briggs and Shantz (3) and consisted of a mixture of about 80
per cent of paraffin and 20 per cent of petrolatum. The mixture had a
melting point of about 450 C. By pouring a layer of this wax about 1 cm.
thick on the layer of cotton around the leaf petioles the leaves were held
firmly in place. To permit the entrance of air into the flasks as water
was removed by transpiration from the leaf a small hole was made in
the wax with a pin.

In testing the effect of Bordeaux mixture on the rates of transpiration
of abscised leaves, six leaves of each species were employed. These were
chosen from larger groups of leaves which had been mounted as above
described and allowed to stand for some time in order to become adjusted
to the new conditions. Leaves of the different species used were chosen
with special reference to equality of surface exposed and also with
reference to similarity of general appearance. The six leaves chosen
were divided into two groups of three each. For facility in comparison
and for ease of reference in discussion, one of these groups will be desig¬
nated “series A” for periods before treatment and “series A'” for periods
after treatment; the other group, remaining untreated throughout the
experiment, will be designated “series B.”

Dec. 18, 1916 Influence of Bordeaux Mixture on Transpiration

53i

It is, of course, not possible in experiments of this nature to subject
each leaf to precisely the same changes in aerial conditions. Special
precautions were taken, however, to arrange the leaves involved in a
single experiment in such a manner with reference to each other and to
their surroundings that each might experience, as nearly as possible, the
same changes in environmental conditions. This was accomplished by
placing the flasks in rows 2 feet apart on a table centrally located in the
greenhouse, where air currents from ventilators or open doors would
affect the plants similarly. The arrangement decided upon for each
experiment was maintained until the experiment was terminated. The
leaves were removed from their positions only for the purpose of weighing,
and each leaf was returned to the position previously occupied as soon as
the weighing was completed.

Each flask with its leaf (in series A and series B) was weighed to 0.01
gm., allowed to stand for a definite time period, and again weighed. The
difference between the two readings, of course, gave the absolute
transpiration for the time period. The first period of exposure, before
spraying the leaves of series A with Bordeaux mixture, may be regarded
as the standardization period, and the leaves of the control series B may
be regarded as the standard leaves for comparison. At the close of the
standardization period the leaves of series A, which then became series
A', were sprayed with Bordeaux mixture. This spray was prepared in
the usual way and contained 12 gm. of copper sulphate and 12 gm. of
lime in 1 liter of the mixture, approximating the 5-5-50 formula of
agricultural practice. Both upper and lower surfaces of the leaves were
sprayed, the spray being applied with an atomizer. This method yielded
a very uniform film of spray over the leaf surfaces.

A comparison of the ratios between the transpiration quantities of
series A and those of series B for the time period before spraying with
similar ratios for the time period after spraying will determine whether
the transpiration rates of the leaves for the periods after spraying have
changed. For the sake of convenience in the treatment of ratios, tran¬
spiration quantities for series A will be termed “A” for periods before
treatment and “A'” for periods after treatment. The corresponding
transpiration quantities for series B will be designated “B” throughout,
since the leaves of series B were not treated with Bordeaux mixture. If
then the ratio of A to B is greater than the ratio of A' to B, it follows that
the rates of transpiration of the sprayed leaves have suffered a decrease
relative to the rates of the unsprayed leaves. If, however, the ratio of
A to B is less than the ratio of A' to B , then the rates of transpiration of the
sprayed leaves have increased, relative to the rates of the unspraved
leaves. In making such comparisons it must be assumed, of course, that
the ratio of A to B for successive time periods would remain unchanged if
the leaves of series A were not treated. It could scarcely be expected
that the ratio of A to B would remain constant for successive intervals, since
St is not possible to subject all the leaves to precisely the same changes in

532

Journal of Agricultural Research

Vol. VII, No. 12

aerial conditions. Furthermore, internal changes continually takingplace
in the leaves would tend to cause some variations in the ratio of A to B,
since the degree of these changes would certainly vary from leaf to leaf,
even supposing the nature of these changes to be the same in all the
leaves. It may reasonably be supposed, however, that variations in the
ratio of A to B for successive intervals, due to internal changes and to
differences in the changes of environmental conditions experienced by
the different leaves of a series, are comparatively small. This was,
indeed, found to be the case in a series of preliminary experiments. The
time peiiod during which an experiment may be conducted with relatively
very small variations in the ratio of A to B, owing to conditions other than
the treatment with spray, varies, of course, with the different species.

Experimental results. — The experimental data showing the effect
of Bordeaux mixture on the transpiration of abscised leaves are pre¬
sented in Table I. The first column of this table gives the names of the
various species dealt with and the time at the beginning and end of
each experimental period. This is followed, under “ Periods before treat¬
ment,' # by three columns giving transpiration quantities, in grams, of
the three leaves of series A; then are given three columns presenting
transpiration data for series B for the same time periods. The last
column of this section gives ratios obtained by summing the values of
the transpiration quantities of the three leaves (on the same horizontal
line) of series A and dividing this summed value by the summed value
of the corresponding transpiration quantities for series B. The second
section, under “Periods after treatment," presents in the same way as
the first the data comparing series A' with series B. The table is further
divided into a number of horizontal sections, each section presenting all
of the data for a single species.

The data included in the first horizontal section of Table I represent
an experiment extending over a total time period of more than five days.
The transpiration data for datura leaves presented in the second hori¬
zontal section of the table represent an experiment extending over a
total time period of more than four days. During these two experiments
weighings were made each day, as indicated. The data presented in the
remaining sections of the table represent experiments conducted mainly
to determine whether the spray becomes effective in its influence on
transpiration, immediately after drying on the leaves, or whether modified
rates of water loss begin at some later period. All of these experiments
were conducted during the same day, extending over a total time period
of a little more than io hours. Water loss from each leaf was determined
at intervals of two hours. The containers used in these experiments
were small, so that weighings could be made to o.oi gm. At the end of
the two standardization periods the leaves of the A series were sprayed
on the upper and lower foliar surfaces, and weighings were made soon
after the spray had completely dried on the leaves.

Dec. 18, 1916 Influence of Bordeaux Mixture on Transpiration

533

Table I. — Effect of Bordeaux mixture on the rates of transpiration of abscised leaves for

periods before and after treatment

PERIODS BEFORE TREATMENT

Transpiration quantities.

Ratio
A : B.

Series A.

Series B (control).

Leaf 1.

Leaf 2.

Leaf 3.

Leaf 1.

Leaf 2.

Leaf 3.

Gm.

Gm .

Gm.

} 6.40

3* 60

2. 60

i- 30

4-50

6. 20

1. 05

} 1-40

2. 40

4.00

2. 70

4.60

•?*,

} i- 70

4. 20

3-50

2. 80

4. 80

5-90

. 70

9- 50

10. 20

10. 10

6. 80

12. 00

16. 70

. 84

| 10. 90

8. 20

1. 50

9- 50

5* 80

7. 90

.89

} 4-70

4. 20

2. 20

4. 80

4. 10

4. 60

•85

15. 60

12. 40

3- 70

14.30

9. 90

12. 50

.86

} .os

• 14

.18

. 08

■43

. 21

• 56

} .07

. 14

. 21

•13

.18

. 20

.82

• 15

.28

•39

. 21

. 61

.41

.67

} -

• 27

• 23

•23

•45

•25

• 72

} -07

. 20

. 21

■ 19

•3i

.27

.62

.24

•47

•44

.42

. 76

• 52

.68

} -

• 25

. 10

•30

•17

• 72

} .16

•33

. 21

■ 27

. 20

• 72

•33

•58

•31

• 57

•37

. 72

} '30

.28

•30

•33

■ 29

. 21

1. 06

} -* 2 3 * * * * * * * * * * * * * * * * * * * * * 2S

. 20

•25

. 20

.14

1. 29

•SS

.48

• 55

•53

■49

•35

i- 15

} *04

.07

.08

. 12

. 10

•05

• 73

. 08

•05

.18

. 10

1 .67

• 13

• 15

■ 13

• 17

.28

1 -5

| .68

} 09

• 13

. 10

• 13

. 10

.07

!' ■

! 1. 06

} ■«

• 15

• 13

• 15

. 10

.14

| 1. 00

. 20

.28

• 23

.28

. 20

| .21

| 1. 02

} .66

.48

.92

1.03

• 27

•44

j 1. 18

} .45

.48

1.09

■ 75

■ 25

•39

j i-45

1. 11

.96

2. 01

1. 78

•52

| .80

i- 3i

Plant and period.

Ricinus communis:
Feb. 1, 8.20 a.m.
t, 3.30 p.m.
2,10.00 a. m.

2, 6.00 p.m.

'* 3, 8.30 a.m.

3, 4.00 p.m.

Total.

Datura metaloides:

Mar. 16, 4.00 p.m.
17, 1.25 p.m.
17, 1.25 p.m.

18, 11.30 a. m.

Total.

Phaseolus vulgaris:
Mar. 22, 8.50 a.m.
22, 10.40 a. m,
22, 10.40 a. m.
22, 12.40 p. m.

Total.

Beta cyela:

Mar. 22, 9.45 a.m.
22, 10.35 a. *** ■
22, 10.35 a. m.
22, 1.35 p. m.

Total.

Raphanus sativus:

Mar. 22, 9.10a.m.
22, 11. 10 a. m.
22, 11. 10 a. m.
22, 12.10 p.m.

Total.

Caladium sp. (small variety):

Mar. 22, 9.05 a. m .

22, 10.55 a. m .

22, 12.55 p.m .

Total.

Hibiscus cardinalis:
Mar. 22, 9.30 a.m.
22, 11.20 a. m.
22, 11.20 a. m.
22, 1.25 p.m.

Total.

Clerodendron balfouri:
Mar. 22, 8.35 a.m.

22. 10.30 a. m.

22. 10.30 a. m.
22, 12.30 p. m.

Total.

Datura metaloides:

Mar. 22, 8.15 a. m.
22, 10.15 a. m.
22, 10.15 a. m.
22, 12,10 p. m.

Total

534

Journal of Agricultural Research

Vol. VII, No. u

Tabib I. — Effect of Bordeaux mixture on the rates of transpiration of abscised leaves for
periods before and after treatment — Continued

PERIODS AFTER TREATMENT

Transpiration quantities.

Plant and period.

Series A'.

Series B (control).

Ratio

A':B.

Eeaf 1.

Eeaf 2.

Eeaf 3.

Deaf 1.

Deaf 2.

Leaf 3.

Ricmus communis:

Feb. 4, 9.30 a. m . .

\ Gm.

Gm.

Gm,

4, 3.30 p.m .

/ 3*50

5*50

4*70

3-oo

2.30

4- 10

i- 45

S, 8.30 a. m .

} 3-6o

* 5t 4-3o p. m .

3*40

4.40

3-io

3.80

4-00

i.tS

6, 12.00 a. m .

6, 5. 30 p.m .

/ a. 30

1.30

2.40

z.90

1.70

2.30

1. 01

Total .

n a^

it. 50

10.40

10' 20

0* oO

1.23

Datura metaloides:

Mar. 18, 4.00 p.m .

19, 9*oo a. m . .

j 5-oo

710

300

3-io

1.30

a. 70

2. 13

19, 9.00 a. m .

{

2o, 4.00 p.m .

/ 3-30

4.90

5.00

1.30

6.00

1-53

Total .

8.30

4.40

8. 70

1.82

12.00

H.OO

Phaseolus vulgaris:

Mar. 22, 3.00 p.m .

—

22, 4.50 p.m .

/ 13

•24

.19

. IX

.06

.14

1.80

22, 4.50 p.m .

22, 6.50 p.m .

| .09

.16

. 12

.09

. 10

.08

i-37

Total .

•31

.22

1. 60

— —

• 40

. 10

Beta cyda:

Mar. 22, 3.55 p.m .

22, 5.55 p. m. .

/ 19

.19

. 22

•19

. 22

•15

1.07

22, S.55 P.m .

{

a2, 7.55 p. m .

} *“

•17

.16

. 12

•15

*14

1.07

Total .

•31

• 37

•29

* 30

*3®

1.07

Raphanus sativus:

Mar. 22, 3.40 p.m .

22, 5.35 p. m .

} -33

•52

. 19

• 33

• 31

i* 33

*2, 5.25 p. m .

aa, 7.35 p.m . . . .

} *2r

*38

•13

. 20

.36

1. 00

Total .

_ -S3

1*17

•32

■ 43*

•47

C alodium sp. (small variety):

Mar. 22, 3.30 p.m .

22, 5,20 p. m .

/ -34

. 10

•34

.06

•IS

.02

3-39

22, 5.20 p.m .

22, 7.25 p.m .

.14

• 16

. 10

•13

• IO

1. 18

Total .

•50

. 12

2.08

• 24

• 28

Hibiscus cardinalis:

Mar. 22, 3.4s p.m .

22, 5.45 p.m . .

\ 14

.14

•05

•07

.04

2.63

22, 5.45 p, m .

22, 7.40 p.m .

f -09

. 12

.18

.06

.04

2.42

Total .

•32

•13

.08

a-53

Clerodendron balfouri:

Mar. 22, 3.10 p.m . . . , ,1

\ .08

22, 5.05 p. m .

. 12

. 10

.09

•07

1.23

aa, 5.05 p. m .

22, 7.00 p. m .

' -°9

• 10

. 10

.08

•05

.04

1. 70

Total... . .

.14

1. 46

* 22

• l8

. 11

Datura metaloides:

Mar. 22, 3. 15 p. m .

22, 5.15 P- m . j

22, 5.15 p. m . \

•72

.81

.18

•53

.28

•35

1.47

22, 7.10 p.m . j

.53

•59

•79

•37

•25

. 22

2.26

Total .

•97

.90

•53

•57

1.80

Dec. 18, 1916

Influence of Bordeaux Mixture on Transpiration 535

A study of the last column of Table I brings out the fact that the first
2-hour period after spraying yielded higher ratio values than the second
period after spraying. A notable exception to this, however, is the
experiment with Datura metaloides conducted on March 22. Here the
second period after spraying yields a ratio value considerably higher
than the first, while in the experiment with Swiss chard the ratios for
the two periods after spraying have the same value. The experiments
with Datura sp. and with castor-bean leaves, extending over time periods
of four days and five days, respectively, show a decrease in the ratio
values for the time periods of successive days. It appears, therefore,
that in these experiments the highest rate of water loss from the leaves
due solely to the influence of the spray occurred soon after the spray had
dried upon the leaves, certainly within the first 2-hour period, after
which there was a gradual decrease in the transpiration rates as influenced
by the surface film of the spray. As is clearly brought out in the experi¬
ment with castor-bean leaves, this gradual decrease in the rates of
water loss from the leaves, as indicated by the ratio values, extends
through a period of three days after spraying. The gradual decrease in
the ratio values during the period after spraying is undoubtedly caused
by the gradual disintegration and wearing away of the surface films. It
may therefore be assumed that whatever may be the physical basis for
the increased evaporation rates from plant surfaces following the appli¬
cation of Bordeaux mixture, a modification of leaf structure in response
to a chemical stimulus induced by the spray can not be responsible for
the increased transpiration rates.

In Table II is given a brief summary of the average data presented in
Table I. The ratio values of total transpiration quantities for standard¬
ization periods are here considered as 1.00, while the corresponding ratio
values for periods after spraying are expressed in terms of these.

Tabi<S II. — Summary of the average data from Table /. Ratio values of A' to B for
periods of ter spraying are relative to the ratio values of A to B for standardization periods
expressed as 1.00

Plant.

Duration of
experiment.

Transpiration
(ratio A':B).

Ricinus communis .

6 days .

1. 5©

2. 12

2*39

I-I7

I.63

I. 8l

3- 72

T-37

Datura metaloides . ,

4 days

Phaseolus vulgaris .

10 hours... .

. . .do. .....

Beta cycla .

Raphanus sativus .

. . .do .

C alodium sp .

. . .do .

Hibiscus cardinalis .

. . .do. . .

Clerodendron balfouri .

. . .do. . .

Datura metaloides .

11 hours. . .

Average .

1. 99

66848° — 16 — —3

536 Journal of Agricultural Research vd. vn, no. X3

Prom the brief summary in Table II it will be observed that the lowest
ratio value for any period after treatment is 1.37 times the ratio value for
the corresponding standardization period, while the highest ratio value
is 3.72 times the ratio value for the corresponding standardization period.
The average value of the ratios for periods after spraying is 1.99 times
the average ratio for the standardization periods. Expressed in another
way, the leaves sprayed with Bordeaux mixture showed an increase in
the rates of water loss, relative to their respective controls in each case,
varying from 37 per cent for the lowest increase to 272 per cent for the
highest. The average rate of water loss by transpiration of the species
here employed shows an increase of 99 per cent over the average rate for
the standardization period. The degree of this accelerating influence of
Bordeaux mixture on the rates of transpiration varies considerably with
the different species of leaves, as is indicated by this wide range of varia¬
tion in the ratio values for the different leaves.

The leaves employed in these experiments were allowed to remain in
the places occupied during the experiments for some time after* the
experiments had been terminated and were kept under observation to
determine whether the sprayed or the unsprayed leaves should first show
signs of wilting. It was observed that in every instance the sprayed
leaves showed signs of wilting at some period preceding the time at which
the unsprayed leaves of the same species began to wilt. This is only what
would be expected in view of the fact that the average rate of water loss
from the sprayed leaves is nearly double that from the unsprayed leaves
during the experimental time period.

EXPERIMENTS WITH POTTED PLANTS

Experimental methods. — In the experiments with potted plants the
method of procedure was similar to that followed in the experiments with
abscised leaves. The plants employed consisted of tomato (Lycopersicon
esculentum Mill.), cabbage (Brassica oleracea L.), pepper (Capsicum
annuum L.), egg plant (Solanum melongena L.), and soy bean (Glycine
hispida M.). These plants, excepting the soy beans, were grown in beds
of soil in the greenhouse until they had attained a size suitable for experi¬
mentation of this character. They were then transplanted to earthen¬
ware pots glazed inside and outside and having a capacity of approxi¬
mately 1.5 liters. In order to prevent evaporation from the surface of
the soil, melted wax prepared according to the Briggs and Shantz formula
was poured over the surface of the soil in each pot, thus making a perfect
seal.

Water was automatically supplied to the roots of the plants by means
of the autoirrigator. This instrument has been described by Livingston
(12) and later by Hawkins (11). It consists of a porous clay cup closed
by a rubber stopper through which extend two glass tubes. One of these
tubes is bent into the form of an inverted U having one arm considerably

Dec. iS, 1916

Influence of Bordeaux Mixture on Transpiration 537

longer than the other. The short arm extends through the rubber stopper
into the cup. The bend of the U turns over the edge of the container and
the end of the longer arm dips into a reservoir of water below. The
second glass tube passes through the rubber stopper and ends just at its
lower surface. To the upper end of this glass tube a short piece of rubber
tubing is attached. This is closed by means of a pinchcock or screw
clamp when the instrument is installed. In order to install the instru¬
ment, the rubber stopper with its glass tubes is pressed firmly into the
mouth of the cup. The cup is then buried in the soil of the container
among the roots of the plant. The end of the long arm of the U-tube is
dipped into a suitable reservoir of water below. Suction is now applied
to the short rubber tube. This causes water to rise through the U-tube
into the cup, filling the latter. When the water has risen and filled the cup
and the short rubber tube, a pinchcock is applied, thus closing the tube.
In installing the instrument, care must be taken that all connections are
tight and that all air is removed from the system. As water is removed
from the soil by the plant roots, more water passes through the porous
walls of the clay cup, replacing that absorbed by the roots. In this way
an approximately constant soil-moisture content may be maintained.
Any desired moisture content of the soil may be obtained and maintained
approximately by simply increasing or decreasing the distance between
the surface of the water in the reservoir and the surface of the water in
the porous clay cup.

In experiments such as are here reported, a constant moisture content
of the soil in which the plants are rooted is of considerable importance,
and the old method of supplying water to the roots at stated intervals has
proved unsatisfactory.

After installing the autoirrigators and sealing the pots the plants
were allowed to stand a week in order to become adjusted to the new
conditions before the experiments were begun.

In each experiment 12 plants were employed. These were chosen
from a much larger number and were selected with special reference to
uniformity of size and vigor. The 12 plants were divided into two
groups, each group constituting a series of 6 plants. In order to expose
the plants included in a single experiment to the constantly changing
aerial conditions in such a way that all might be affected in a somewhat
similar manner, they were arranged in two rows, 6 plants to a row, on
opposite sides of a greenhouse bench, the rows being placed near the
edge of the bench. The bench was centrally located and was not in the
direct path of air currents from open doors or ventilators. As a further
precaution, the plants were shifted in their positions in the rows each
day according to a definite plan previously decided upon.

The plants were weighed each day and the water loss from each plant
for the time period immediately preceding was determined. The weigh¬
ings were made in the order in which the plants were numbered, from 1

538

Journal of Agricultural Research

Vol, VII, No. 12

to 6, and this plan was maintained throughout the experiment, so that
the time periods for each plant between any two weighings were approxi¬
mately the same. At the end of the standardization period the plants
of series A were sprayed. This series now became series A'. The
Bordeaux mixture employed was prepared according to the same formula
as that used in the experiments with abscised leaves. The spray was
applied to the leaves of the plants with an atomizer. This method gives
a very uniform film of the spray over the leaf surfaces. The spray was
applied to both the upper and lower surfaces of the leaves, except in the
experiments with cabbage plants.

At the close of an experiment the tops were severed from the roots
just at the surface of the wax seal, and the green weights were imme¬
diately obtained. The plants were then placed in weighing bottles and
dried in an oven at a temperature of from 76° to ioo° C. for a period of
28 hours, after which they were dried to constant weight at a temperature
of from 1020 to 105° C. The bottles were then transferred to a large
desiccator and were allowed to cool to room temperature before weighing.

Measurements were taken of the evaporating power of the air in the
greenhouse room where the experiments were carried out. These measure¬
ments were made by means of standardized porous-cup atmometers (13) ;
the instruments were placed among the plants on the greenhouse bench
and readings were taken each day at the time when the plants were
weighed. The readings were corrected to the Livingston cylindrical
standard by multiplying by the coefficient of correction of the cup used.
In the time during which the experiments were conducted the water
loss from the porous-cup atmometer gave a daily mean of 10.4 c. c. ; a
maximum daily rate of 24.0 c. c. (on March 1) and a minimum daily
rate of 6.0 c. c. (on March 6).

As in the experiments with abscised leaves, one of the two groups of
plants comprised in a single experiment will be designated “series A.* *
The other group, remaining untreated throughout the experiment, will
be designated “series B.” In the experiments with potted plants,
transpiration quantities for series A and for series B will be denoted
“A” and “B” in their respective series. This method of notation is
precisely the same as that adopted in the experiments with abscised
leaves.

In presenting the data for the experiments with potted plants the
ratios between the water loss per gram of green substance for periods
before and after spraying, as well as the corresponding water loss per
gram of dry substance, will also be treated. Green- and dry-weight
values could, of course, not be obtained until the close of an experiment.
These values must serve, therefore, for the calculation of ratios for the
standardization periods as well as for the periods after spraying. This,
however, could make no material difference in the ratio values for periods
before spraying, since any increase in the weight of plant substance due

Dec. is, 1916 Influence of Bordeaux Mixture on Transpiration

539

.to the growth of the plants during the periods after spraying would affect
approximately alike the green- and dry-weight values of both series A'
and series B. If, in the treatment of these ratios the green- weight
values for series A and for series A' (the values for these two series being
the same) are denoted by “ G,” and the corresponding values for series B
are denoted by * *G',” it follows that the ratios of A to G, A' to G, and B to G'
must represent the water loss per gram of green substance for series A,
series A', and series B, respectively. Now, the ratio between the average
water loss per gram of green substance for series A and for series B

A*G AG'

(periods before spraying) is expressed by g-^,,or The correspond¬

ing ratio value between the average water loss per gram of green sub¬
stance for series A' and for series B (periods after spraying) is expressed

A':G

B:G,J

A'G'
BG *

If, therefore, the ratio of AG' to BG is less than that

of A'G' to BG, the water loss per gram of green substance must be greater
for the sprayed plants than for the unsprayed.

The dry-weight values for series A' and for series B will be denoted
by “D” and “D',” respectively, and the ratios between the water loss
per gram of dry substance for periods before spraying and the corre¬
sponding ratio for periods after spraying, derived in the same manner
as the ratios between the water loss per gram of green substance, are
expressed by the ratios of AD' to BD and A'D' to BD, for standardiza¬
tion periods and for periods after spraying, respectively.

CABBAGE PLANTS

In this experiment the plants used were vigorous and fairly uniform
in size. The standardization period continued from March 1 to March 4.
The period after spraying began on March 4 and continued to March 17.

Considerable difficulty was encountered in attempts to spray the
leaves of the cabbage plants. The bloom on the surface of the leaves
prevented the spray material from adhering. Various substances were
added to the spray material in an attempt to find something which
would cause the spray to adhere to the leaves. Rosin-fish-oil soap in
the proportion of 2 pounds to 50 gallons of Bordeaux mixture, as rec¬
ommended by Hawkins (10), proved to be the most effective; but even
this did not give the desired results. It was found that by lightly brush¬
ing the surface of the leaves with a wad of dry absorbent cotton, and by
immediately afterwards applying the spray, a surface film of the spray
could be obtained. This film was by no means ideal, but it was better than
that secured by using the rosin-fish-oil soap in connection with the Bor¬
deaux mixture. The leaves of the control series B were also lightly
brushed with a wad of absorbent cotton at the same time that the leaves
of series A' were thus treated. Only the upper surfaces of the leaves
were thus treated and sprayed.

54°

Journal of Agricultural Research

VoL VII, No. la

In Table III are given the data for the experiment with potted cab-,
bage plants. This table presents first in a horizontal section the summed
transpiration quantities for each plant and the average for the plants
in each series. It also gives the green and dry weight values for each
plant and the average for the plants in series A' and in series B. This
section is followed below and in the order here given by (i) the ratios
between the average transpiration quantities of the two series for the pe¬
riods before and after spraying; (2) the ratios between the quantities
representing the average water loss per gram of green substance, of the
two series in periods before and after spraying; and (3) the ratio between
quantities representing the average water loss per gram of dry substance,
of the two series in the periods before and after spraying. The differ¬
ence between the two ratios in a pair, on the same horizontal line, is
given to the right. The plus ( + ) sign following the difference indicates
that the ratio for the period after spraying is higher than the correspond¬
ing ratio for the standardization period.

Table; III. — Effect of Bordeaux mixture on the rate of transpiration of potted cabbage

plants

[Period before treatment, Mar. i to Mar. 3; period after treatment, Mar. 4 to Mar. 17, 1916]

Plant No.

Transpiration quantities.

Green weight
of tops.

Dry weight of
tops.

Period before
treatment.

Period after
treatment.

Series

A'.

1 Series

1 B-

Series

A'.

Series

A*.

Series

(A)

<5)

(A')

(B)

(C)

«?)

(£>)

(&)

67.6

33-4

IOIS. 2

342. 1

56.48

36.92

10.48

6. 15

46.5

32-7

709.O

523*9

58. 28

39* 18

8.98

7* 70

71. 1

36.1

1041. 9

493*3

58. 81

40.72

10.90

6. 20

4 .

42. 6

30. 7

642.4

585*5

42. 82

38. 11

7* 74

6.99

5 .

6o*s

33-4

1008. 8

685. s

69.90

57-46

9.92

10. 04

32. 2

38*3

516. 8

497*9

42. XI

49*27

9. ii

8.76

Average .

53-4

34* I

»3S* 7

521.4

54* 72

43*6i

9* 36

7*64

Transpiration . . . ratio. .

A:B=i,s6

A': B

~i.6o

Difference . .

. a 04+

Water loss per gram of

green weight . ratio . .

AG : BG=*i.24

A G': BG=i.28

Difference .

• 04-f

Water loss per gram of

dry weight, . ratio. .

AD': BD-x.27

A'D': BD=i.3i

Difference .

. .04+

It will be observed from Table III that the ratio values for the period
after treatment are, in each case, slightly higher than the corresponding
ratio values for the period before treatment. This indicates that the
average rate of water loss by transpiration from the sprayed plants
(series A') was slightly higher, relative to the average rates from the
control plants, than was the corresponding ratio of water loss from those
same plants (series A) for the period before treatment. It is to be noted,
however, that while the ratios between average quantities for the period
after spraying are all higher than the corresponding ratios for the period
before treatment, the individual ratios for the period after treatment

Dec. 18, 1916 Influence of Bordeaux Mixture on T ranspiration 541

vary on either side of the average ratios for the standardization period.
In this experiment with cabbage plants it may not, therefore, be stated
with entire certainty that an application of Bordeaux mixture has an
accelerating influence on the rates of water loss, since the difference be¬
tween the ratio values for the periods before and after treatment are no
greater than might be expected to lie within the limits of experimental
error. On the other hand, the spray was applied only to the upper sur¬
faces of the leaves, so that only one-half of the transpiring surface was
affected by the spray material. Furthermore, stomatal transpiration
from the upper surfaces of the leaves should be somewhat less than from
the lower surfaces, considering the number of stomata per square milli¬
meter on the upper leaf surface to be 219 and that on the lower surface
to be 301, according to Duggar (5, p. 91). This would still further tend
to minimize any accelerating influence which the spray might have on the
rates of water loss.

PEPPER PlyANTS

The plants used in the experiment with peppers were very uniform in
size, about 8 inches tall, vigorous, and were just beginning to bloom when
the experiment was begun. The experiment continued during a time
period of 16 days. The standardization period extended from February
26 to March 3; the period after treatment continued from March 3 to
March 14. The results of this experiment are given in Table IV. The
arrangement of this table and subsequent ones is precisely the same as
that employed in the presentation of the data for the experiment with
cabbage plants.

Table IV. — Effect of Bordeaux mixture on the rate of transpiration of potted pepper

plants

[Period before treatment, Feb. 26 to Mar. 3; period after treatment, Mar. 3 to Mar. 14, 1916J

Plant No.

Transpiration quantities.

Green weight
of tops.

Dry weight
of tops.

Period before
treatment.

Period after
treatment.

Series

A'.

Series

A'.

Series

A'.

Series

(A)

(B)

(AO

(B)

(G)

(GO

(D)

(DO

38.x

35-7

109.3

43*0

3* 70

5* 02

0*75

0.83

37*9

17. 0

91*4

37*7

2. 08

2-57

■45

•31

3*- 4

27.9

63- 3

42. s

4* 57

2.25

.70

.40

37-3

40.0

102. 8

103*3

4* 4°

4-87

•65

* 73

29* 8

32*9

65. 2

54* 7

4. 20

4-73

•69

*78

6 .

S7* 9

20. 9

114. 8

37*9

6.80

2* 75

I. 12

*47

Average . .

38.7

29.0

91. 1 |

53* 1

4. 29

3* 69

.72

.58

Transpiration . ratio. .

AB=

= 1.33

A'B=i.7i

Difference .

. 0. 38+

Water loss per gram of green

substance . ratio. .

AG' BG=i.i4

A G BG=x.47

Difference .

. .33+

Water loss per gram of dry sub¬

stance . ratio. .

AD' BD=i.o7

A D BD==i,37

Difference . .

. .30+

542

Journal of Agricultural Research

Vol. VII, No. ta

A comparison of the ratio values in Table IV shows very clearly the
accelerating influence of a film of Bordeaux mixture on the rates of
transpiration of pepper plants. If each of the three ratio values for the
period before spraying is made equal to i.oo, the corresponding ratio
values for the period after spraying are, in every case, i .29— that is, 29
per cent higher than the values for the standardization period.

SOY-BI$AN PLANTS

The soy-bean plants here employed were about 20 cm. tall when the
experiment was begun. Soy-bean seedlings were transplanted directly
into the containers when they were about 5 cm. tall. Three seedlings
constituted a culture. Four series of cultures were employed, each series
comprising six cultures and a total of 18 plants. The four series will be
designated series A, series B, series C, and series D, respectively, for the
standardization period. Series D is here considered the control series and
remained untreated throughout the experiment. The plants of the first
three series were treated at the end of the standardization period and are
designated “series A',” “series B',” and “series C',” in the order given,
for the period after treatment. As in all the experiments here reported,
the leaves of the plants of series A' were covered with a film of Bordeaux
mixture. The leaves of series B' were sprayed with a suspension of
barium sulphate in water. The mixture consisted of 28 gm. of barium
sulphate in 1 liter of water. This was applied by means of an atomizer
in precisely the same manner in which the Bordeaux mixture was applied.
The leaves of series C' were treated with dry copper sulphate in the form
of a fine powder. The copper sulphate was prepared by gently heating
the crystals in a porcelain crucible until all the water of crystallization had
been driven off. After the salt had thus been dried to constant weight, it
was ground to a fine powder in a mortar and kept in a desiccator until
used. This powder was dusted on the upper surfaces of the leaves until
a thin but fairly uniform covering was obtained. During the time period
of the experiment the copper sulphate produced no injurious effects upon
the leaves, though during this period care was taken to keep the green¬
house room as dry as possible.

For the sake of convenience in presenting the data for this experiment,
the transpiration quantities for series A, series B, series C, and series D
are denoted by “A,” “B,” “C,” and “D,” respectively, for the stand¬
ardization period; the transpiration quantities for the corresponding
series for the period after spraying are denoted by “A',” “B'f” “O',”
and “D,” respectively.

The period of standardization extended from February 29 to March 4;
the period after treatment continued from March 4 to Match 11. The
data for this experiment are presented in Table V.

Dee. is. i»i6 Influence of Bordeaux Mixture on Transpiration

543

Tabi^S V. — Effect of Bordeaux mixture , copper sulphate , and barium sulphate on the
rate of transpiration of potted soy-bean plants

[Period before treatment, Feb. 29 to Mar. 4: period after treatment, Mar. 4 to Mar. xx, 1916I

Transpiration quantities.

Plant No.

Period before treatment.

Series

(Bor¬

deaux

mix¬

ture).

Series

(ba¬

rium

sul¬

phate).

Series

(copper

sul¬

phate).

Series

(con¬

trol).

Period after treatment.

Series

(Bor¬

deaux

mix¬

ture).

Series

(ba¬

rium

sul¬

phate).

Series

(copper

sul¬

phate).

Series

(con¬

trol).

o»

177.9
130. s
76.4
142.3
94.8
8s- 7

Average .

Transpiration ratio ,
Difference .

34*2
64. o
77* S

65.8

67.9
55- 6

<B>

53-0

34-0

45*8

38.7

46.5

37- S

92.3

( D )
53- 1
58. 2
33-5

65-3

46.6

43*9

<ri')

128. 1
254-7

304. 8

217. 8

265. 1
227. 8

(Bf)

131-3

77-4

103.5
88. 7

135-6
87. 7

(C')
•195- 6
198. 6
124.9
75*6
252.0
253* 2

60.8

42- 5

58. 8

50. 1

233-0

104. o

183.3

117- 8

B D=
0.85

C D=*
x.17

fA# D=*
l 1.98

B' D=
0.88

a d=
1.55

o. 77+

0.034-

o. 384-

Prom the data of Table V it appears that of the three kinds of mate¬
rials applied to the leaves of soy-bean plants Bordeaux mixture is the
most effective in bringing about * an increased rate of water loss. The
plants sprayed with Bordeaux mixture here showed, relative to the
control plants, an average rate of water loss 63 per cent higher than
the average rates of the same plants before spraying. The leaves of the
plants dusted with copper sulphate showed a corresponding increase of
37 per cent, while the plants sprayed with barium sulphate show an
increase in the average rate of water loss, relative to the control plants,
of only 3 per cent over the average rate for the standardization period.

The results here obtained are in entire accord with the conclusions
reached by Duggar and Cooley (6) that certain specific characters of the
films are important factors, considered in relation to the modified rates
of transpiration. The further suggestion of these authors that modified
leaf temperature induced by heat absorption due to the color of the sur¬
face films is an important factor might also be considered as effective
in bringing about the results obtained here, since a film of Bordeaux
mixture, being darker in color than either of the two other surface cov¬
erings used may be assumed, without any other direct evidence, to have
a higher heat-absorbing power than either copper sulphate or barium
sulphate. It is scarcely probable, however, that modified leaf tem¬
peratures, induced by differences in the heat-absorbing powers of the
surface films here used, could account for any great portion of the differ¬
ences in the rates of water loss observed.

544

Journal of Agricultural Research

Vol.Vn, No. 12

EGGPLANTS

In the experiment with eggplants the standardization period extended
two days from Aprils to April 5. The period after spraying continued
from April 5 to April 18.

The results of this experiment are presented in Table VI.

Table VI. — Effect of Bordeaux mixture on the rate of transpiration of potted eggplants

{Period before treatment, Apr. 3 to Apr. 5; period after treatment, Apr. 5 to Apr. 18, 1916]

Plant No.

Transpiration quantities.

Green weight of
tops.

Dry weight of
tops.

Period before
treatment.

Period after
treatment.

Series

A'.

Series

A'.

Series

B.,

Series

A'.

Series

(A)

< B )

(A')

C B )

<G)

(S')

(/>)

(DO

20.9

17-4

400. 1

326. 2

12.02

H-73

1.67

1.68

27.9

23*6

336.2

316.6

II. 12

12.63

1.87

1.83

33-4

40.0

295-9

426. 2

n-35

18. 10

1.58

2. 72

27.4

26. 5

445-7

367-0

15- 11

14. 62

2.14

1-77

30.I

33-7

494- 1

394-2

19.62

14. 19

1.97

1.99

18.0

IS- 9

292. i

334-4

12. 40

11.60

1-52

1-55

Average .

24.6 ;

| 26. 1

377-3

360.7

13.60

13-81

i-79 :

1.92

Transpiration . ratio , .

A : B

=0.94

A' : B

1 = 1.04

Difterei

ace .

0. io4

Water loss per gram of green

substance . ratio. .

AGr : BG“*o.95

A'G' : BG=i.o6

Difference .

, _

.11+

Water loss per gram of dry sub¬

stance . ratio. .

AD' ; RD=*i.oo

A'TV : TVD = t ,T5

Difference _

. 124-

The results given in Table VI show a pronounced increase in the rate
of water loss for the sprayed plants relative to the rates from the control
plants, as is indicated by the higher ratio values for the period after
spraying. The value of the ratio between transpiration quantities for
the period after treatment is here 1 1 per cent higher than the correspond¬
ing ratio value for the period before treatment. Ratios between quanti¬
ties representing water loss per gram of green weight and water loss per
gram of dry weight are each 12 per cent higher for the period after
spraying than for the standardization period.

TOMATO PLANTS

It has been previously stated that the Bordeaux mixture employed
throughout these experiments was prepared according to the 5-5-50
formula of agricultural practice. When this mixture was applied to
tomato plants, however, injury to the leaves resulted. In this experi¬
ment, therefore, the strength of the mixture was reduced to conform
approximately to the 4-4-50 formula commonly used. This mixture
produced no injurious effects upon the plants here employed.

In this experiment the period of standardization extended from May 26
to June 1 ; the period after treatment continued from June 1 to June 12.
The experimental data are presented in Table VII.

Dec. x8, 19x6 Influence of Bordeaux Mixture on Transpiration

545

Table VII. — Effect of Bordeaux mixture on the rate of transpiration of potted tomato

plants

{Period before treatment, May 26 to June x; period after treatment, June x to June 12, 19x6]

Plant No.

Transpiration quantities.

Green weight
of tops.

Dry weight
of tops.

Period before
treatment.

Period after
treatment.

Series

A'.

Series

A'.

Series

A'.

Series

B'.

(A)

<B)

i'A)

(B)

(G)

(G')

(D)

(/>')

831.2

210.6

245.0

231-9

17* 7o

16. 76

2. 40

2* X3

260.8

182. 2

269.9

268.9

17-50

14.65

2.20

1*77

3 . .

143*0

170.6

*38.3

203.2

19.15

14.72

2. OX

1.87

171*7

17*5

226. 5

207.0

X$‘ 12

*5- os

2.02

2.00

215.8

171*0

291. 2

142*4

21.03

15.52

2.68

1.98

184.5

2x2.0

202. s

214. 1

13*95

X4*oi

x.92

1.92

Average .

201. i

X86.4

*45*5

2ix. 2

17.40

15. XI

2.20

1*94

Transpiration . ratio . .

AB«i

.07

A'B=

*1.16

Differea

ice .

Water loss per gram of green

substance . ratio . ,

AG' BG— 0.03

A'G' BG-x.oi

Difference .

Water loss per gram of dry

substance . ratio. .

AD' BD=*o.95

A'D' BD=*x.o2

Difference .

From Table VII it will be, observed that the ratios between transpira¬
tion qualities and between quantities representing water loss per gram
of green substance for the periods after spraying are 8 per cent and 9 per
cent higher, respectively, than the corresponding ratio values for the
periods before spraying, while the ratio between quantities representing
water loss per gram of dry substance is 7 per cent higher for the period
after treatment. The average transpiration rate for the sprayed plants,
relative to the control plants, is therefore 8 per cent higher than the
average relative rate for the same plants before treating them with
Bordeaux mixture.

Table VIII presents a brief summary of the average data contained in
Tables III, IV, V, VI, and VII. The values of the ratios between aver¬
age transpiration quantities, between average water loss per gram of
green substance and between average water loss per gram of dry sub¬
stance, for the periods after spraying, are here given in terms of the cor¬
responding ratio values for the standardization periods considered as
unity. The ratio values for the standardization periods are therefore
omitted from the table, since all have the same relative value (1.00).

A comparison of the ratio values presented in this brief summary
shows that the treated plants of each species here dealt with giave higher
average rates of transpiration, relative to the rates from their respective
controls, than did the same plants during the standardization periods.
The lowest average rates of water loss for periods after treatment occurred
with cabbage plants, showing an average increased rate of 2 per cent
over the corresponding rate for the standardization period. It has
already been pointed out that the low increased rates of water loss

546

Journal of Agricultural Research

Vol. vn. No. 12

recorded for cabbage plants may be due to the fact that the leaves were
sprayed only on the upper surfaces, and that at best only a very imperfect
film was obtained.

Table VIII. — Average ratio values for periods after treatment , relative to the corre¬
sponding ratio values for the standardization periods taken as unity, being a summary
of the average data from Tables III , IV, V, VI, and VII

Plant.

Transpiration
(ratio A':B).

Water loss per
gram of dry
substance
(ratio A'G':BG).

Water loss per
gram of green
substance
(ratio A'D':BD).

Cabbage . . .

I.03

Eggplant . . . .

I. II

I. 12

Pepper .

I. 29

Tomato .

I. 08

I.09

1.07

Sav hpfin . .

I. 64

It is further to be noted that with each species employed in these
experiments the three ratios, each derived from a single set of the three
kinds of measurements here dealt with — transpiration, water loss per
gram of green substance, and water loss per gram of dry substance- —
are in very close agreement. The greatest variation in the values of
the three ratios occurred with tomato plants. The values of the ratios
in question are 1.08, 1.09, and 1.07 for transpiration quantities, water loss
per gram of green substance, and water loss per gram of dry substance,
respectively. The influence of Bordeaux mixture in bringing about
increased rates of transpiration varies with the different species of plants,
as is indicated by the variation in the ratio values for the different species.
This has already been observed in the experiments with abscised leaves.

A comparison of the results obtained from abscised leaves with those
obtained from potted plants shows very clearly that the influence of
Bordeaux mixture in bringing about increased rates of transpiration
is much more pronounced when the spray is applied to abscised leaves
than when applied to the leaves of potted plants. Thus, the ratio values
representing the highest and lowest average increased rates of water
loss from abscised leaves, due to an application of Bordeaux mixture,
are 3.71 and 1.73, respectively, relative to the corresponding ratio
values for the standardization periods taken as unity. The ratio values
representing the highest and lowest average increased rates of transpi¬
ration from potted plants, following an application of Bordeaux mixture,
are 1.64 and 1.02, respectively, relative to the corresponding ratio values
for the time period before spraying.

SUMMARY

The results of the experiments presented above substantiate the
general principle already established by Duggar and Cooley (6, 7) that
the rates of transpiration from abscised leaves and also from the leaves

Doc. 18, 1916

Influence of Bordeaux Mixture on Transpiration 547

of potted plants are materially increased by an application of Bordeaux
mixture.

A surface covering of dry, powdered copper sulphate was less effective
in accelerating rates of transpiration than was a surface film of Bordeaux
mixture, but was more effective than was a film of barium sulphate.

The accelerating influence of Bordeaux mixture on the rates of trans¬
piration is much more pronounced when the spray is applied to abscised
leaves than when applied to the leaves of potted plants.

The influence of Bordeaux mixture in increasing the rates of water
loss from abscised leaves becomes effective immediately after the spray
dries upon the leaves. The highest average increased rates of trans¬
piration occurred during the first 2-hour period following an application
of the spray.

The effectiveness of a film of Bordeaux mixture for inducing increased
rates of water loss from abscised leaves varies considerably with the
different species. This, to a lesser degree, is true also with potted plants.

LITERATURE CITED

(1) Bain, S. M.

1902. The action of copper on leaves ... Tenn. Agr. Exp. Sta. Bill., v. 15,
no. 2, 108 p., 8 pi.

(2) Bayer, Ludwig.

1902. Beitrag zur pflanzenphysiologischen Bedeutung des Kupfers in der
Bordeaux-briihe . 56 p. KOnigsberg. Inaug. Diss.

(3) Briggs, L- J., and Shantz, H. L.

1912. The wilting coefficient for different plants and its indirect determination.
U. S. Dept. Agr. Bur. Plant Indus. Bui. 230, 83 p., 9 fig.

(4) Clinton, G. P.

1911. Spraying potatoes in dry seasons. In Conn. Agr. Exp. Sta. Bien. Rpt.,
1909/10, pt. 10, p. 73ST752-

(5) Dugar, B. M.

1911. Plant Physiology ... 516 p. illus., New York.

(6) - and CoouiY, J. S.

1914. The effect of surface films and dust on the rate of transpiration. In Ann .
Mo. Bot. Gard., v. 1, no. 1, p. 1-22, pi. 1. Bibliography, p. 20-21.

1914. The effect of surface films on the rate of transpiration: Experiments with
potted potatoes. In Ann. Mo. Bot. Gard., v. 1, no. 3, p. 3S1— 35^* P1* l8-

(8) Frank, A. B., and Kroger, Friedrich.

1894. Uber den direkten Einfluss der Kupfervitriolkalkbrfihe auf die Kar-
toffelpflanze. 46 p., 1 col. pi. (Arb. Deut. Landw. Gesell., Heft 2.)

1894. Ueber den Reiz, welchen die Behandlung mit Kupfer auf die Kartoffel,
pflanze hervorbrmgt. In Ber. Deut. Bot. Gesell. , Bd. 12 , Heft 1 , p. 8-1 1
(10) Hawkins, L. A.

1912. Some factors influencing the efficiency of Bordeaux mixture. U. S.
Dept. Agr. Bur. Plant Indus. Bui. 265, 29 p., 4 fig.

(11)

1910. The porous clay cup for the automatic watering of plants. In Plant
World, v. 13, no. 9, p. 220-227, 3 fig.

548

Journal of Agricultural Research

Vol. vn, No. 12

(12) Livingston, B. E.

1908. A method for controlling plant moisture. In Plant World, v. 11, no. 2,
p. 39-40.

(13) —

1915. Atometry and the porous cup atometer. In Plant World, v. 18, no. 2,
p. 21-30; no. 3, p. 51-74, 8 fig; no. 4, p. 95-m; no. 5, p. 143-149-

(14) MthxBR-THURGAU, Hermann.

1894. Einfluss der aufgepritzten Kupferpraparate auf die Rebenblatter.
In Jarhresber. Deut. Schweiz. Vers. Sta. Schule Obst-, Wein* u.
Gartenbau Wadensweil, 1892/93, p. 58-59.

(15) Rumm, C.

1893 . Ueber die Wirkung der Kupferpraparate bei Bekampfung der sogenannte n
Blattfallkrankheit der Weinrebe. In Ber. Deut. Bot. Gesell., Jahrg.
11, Heft. 2, p. 79-93.

(16) Schander, Richard.

1904. Uber die physiologische Wirkung der Kupf ervitriolkalkbriihe . In.
Landw. Jahrb., Bd. 33, Heft 4/5, p. 517-584-

INDEX

Achroia grisella — Page

effect of nicotine sulphate on . 94

nicotine in the tissues of . ioi, 104

Acidity and Adsorption in Soils as Measured

by the Hydrogen Electrode (paper) . 123-145

Adsorption and acidity in soils as measured

by the hydrogen electrode . 123-145

Allard, H. A. (paper): A Specific Mosaic Dis¬
ease in Nicotiana viscosum Distinct from

the Mosaic Disease of Tobacco . 481-486

Alway, F. J., and Clark, V. L. (paper): Useof
Two Indirect Methods for the Determina¬
tion of the Hygroscopic Coefficients of

Soils . 345-359

Alfalfa. See Medicago saliva.

Ammonium sulphate, effect of alkali salts on

the nitrification of . 426-428

Amygdalus persica, host plant of Laspeyresia

molesia . 373“378

Ananas saiivus , chlorosis of . 84

Aphidius nigripes, enemy of Macrosiphum

gtanarium . 478-479

Aphis —

avenae, syn. Macrosiphum granarium.

brassicae, nicotine in the tissues of . 103

cereaiis, syn . Macrosiphum granarium.

English grain. See Macrosiphum granar
rium.

granar ia, syn. Macrosiphum granarium.

hordei, syn. Macrosiphum granarium.

Kochii, syn. Aphis malifoliae.
malifoliae —

dimorphic reproduction of . 337

egg of —

description of . 326

hatching of . 326

fall forms of . 337”34°

feeding habits of . 340-341

history and distribution of . 32s

life history of . 342

longevity of . 328

methods of studying . 325-326

nymphal life of . 327-328

reproduction of . 328

spring forms of . 328-334

stem mother of, description of . 326-327

summer forms of, description and life his¬
tory of . . . ; 335“337

synonymy of . . . *. . . 325

mellificus, nicotine in the tissues of . 101-

102, 106-109

populifoliae —
effect of —

fumigation of, with nicotine . 94-95

nicotine dip on . 92

nicotine spray on . 92-94

pyri, syn. Aphis malifoliae.

A phis — Continued. Page

rosy apple. See Aphis malifoliae.
rumicis —

effect of fumigation with nicotine on . 95

nicotine in the system of . 103

sorbi, syn. A phis malifoliae.

Apis mellifica, effect of nicotine-sulphate solu¬
tion on . 95

Apium graveolens, host plant of Thielavia basi-

cola . 293

Apple-seed chalcid. See Syntomaspis dru-
Parum.

Apple. See Malus spp.

Arackis hypogaea , host plant of Thielavia basi-

cola . 294

Aralia quinquefolia , host plant of Thielavia

basicola . 293

Armsby, H. P.( Fries, J. A., and Braman,

W. W. (paper): Energy Values of Red-

Clover Hay and Maize Meal . 379-387

Aspergillus niger Group (paper) . 1-15

Aspergillus —
niger group-

bibliography of variant forms of . 10-15

colony characters of . 6

group characterization of . 6-1 5

morphology of . 6-15

spp., oxalic-add production of . 1-5

Assimilation of Iron by Rice from Certain

Nutrient Solutions (paper) . 503-528

Aster sp., host plant of Thielavia basicola . 293

Astragalus sinicus, host plant of Thielavia

basicola . 294

Atteva aurea —

effect ofmicotine sulphate upon the larvae of. 93

nicotine in the system of . 104

Autolysis, formation of hematoporphyrin in
ox muscle during . 41-45

Baker, A. C., and Turner, W. F. (paper): Rosy

Apple Aphis . 321-344

Beals.C. L-, and Lindsey, J. B. (paper): Chem¬
ical Composition, Digestibility, and Feed¬
ing Value of Vegetable-Ivory Meal . 301-320

Bean. See Phaseolus vulgaris.

Bee, honey. See A pis mellifica.

Beetle, blister. See Epicauta Pennsylvania.

Beetle, potato. See Leptinotarsa decendineata.

Begonia —

rubra, host plant of Thielavia basicola . 293

semperflorens , host plant of Thielavia basi¬
cola . 294

tuberhybrtda, host plant of Thielavia basicola. 293
Beta —

cycla, transpiration experiments with ab¬
scised leaves of . 530-536

vulgaris, host plant of Thielavia basicola - 293

66849°— 17 - 2 (549)

550

Journal of Agricultural Research

Vol. VII

Page

Blackrot fungus, Sphaeropsis malorum, some
effects of, upon the chemical composition of

the apple . 17-40

BlatteUa germanica, nicotine in the tissues of . . 100

Blister beetle. See Epicauia pennsylvanica .

Blood, dried, effect of alkali salts on the nitri¬
fication of . 426-428

Blysmus compressus, host plant of Thielavia

basicola . 293

Bone, cannon, correlation between size of, in
the offspring, and the age of the parents . . 361-371
Bordeaux mixture —
effect of —

on rates of transpiration from abscised

leaves . 530-536

on transpiration of potted plants . 536-546

Botrytis cinerea , growth of, in concentrated

solution... . 257

Braman, W. W., Arinsby, H. P„ and Fries,

J. A. (paper): Energy Values of Red-Clover

Hay and Maize Meal . 379-387

Brassica oleracea , transpiration experiments

with potted plants of . 536-541

Breazeale.J.F. (paper): Effect of Sodium Salts
in Water Cultures on the Absorption of

Plant Food by Wheat Seedlings . 407-416

Briggs, E. J., and Shantz, H.E. (paper): Daily
Transpiration During the Normal Growth
Period and Its Correlation with the

Weather . 155-212

Bromaphis sp.,syn. Macrosipkum granarium.
Bromus —

commutatus (?) [racemosus], food plant of

Macrosipkum. granarium . 465

secalmus , food plant of Macrosipkum grana¬
rium . 465

Brown, H. B. (paper) : Life History and
Poisonous Properties of Claviceps paspali . 401-406
BruchoPhagus funebris —

host insect of Habrocytus medicaginis . 147-154

life history of . 149

Burgess, P. S., Klein, M. A., and Eipman,

C. B. (paper): Comparison of Nitrifying
Powers of Some Humid and Some Arid

Soils . 47-82

Bursa bur so- pastor is, food plant of Macro¬
sipkum granarium . 465

Cabbage. See Brassica oleracea .

Caladium sp. , transpiration experiments with

abscised leaves of . 53 0-536

Caldwell, J. S., Culpepper, C. W., and Foster,

A. C. (paper): Some Effects of the Blackrot
Fungus, Sphaeropsis malorum, upon the

Chemical Composition of the Apple . 1 7-40

Cataphora vomitoria, nicotine in the tissues of . 100

Cannon bone, correlation between the size of,

and the age of the parents . . . 361-371

Capsella bursa- pas tor is, host plant of Thielavia

basicola . 293

Capsicum annuum, transpiration experiments

with potted plants of . 536-539* 541-54®

Carrero, J. O., and Gile, P. E. (paper):
Assimilation of Iron by Rice from Certain

Nutrient Solutions . 503-528

Immobility of Iron in the Plant . 83-87

Page

Cassia chamaecrista, host plant of Thielavia

basicola . 294

Castor bean. See Ricinus communis.

Catalpa speciosa, host plant of Thielavia basi¬
cola . , . 293

Catalpa sphinx. See Ceratomia catalpae.

Ceratomia catalpae —

effect of spray of nicotine sulphate on . 93-94

nicotine in the system of . 104

Chaetochloa glauca , food plant of Macrosipkum

granarium . 465

Chalcid, apple-seed. See Syntomaspis dru-
parum.

Chard, Swiss. See Beta cycla.

Chemical Composition, Digestibility, and

Feeding Value of Vegetable-Ivory Meal

(paper) . 301-320

Cherry. See Prunus spp.

Chlorosis, effect of, on Ananas saiivus and.

Oryza saliva . . 84

Citrullus vulgaris, host plant of Thielavia basi¬
cola . 293

Citrus Umonum, analyses of leaves of, for iron . 85-86
Cladosporium sp., parasite of Paspalum dilata-

tum . 404

Clark, V. E., and Alway, F. J. (paper): Use of
Two Indirect Methods for the Determina¬
tion of the Hygroscopic Coefficients of

Soils . 345-359

Claviceps paspalt —

life history of . 401-406

parasite of Paspalum dilatatum . 401-406

poisonous properties of . 401-406

Clerodendrum Balfouri, transpiration experi¬
ments with abscised leaves of . 530-536

Clover, red. See Tri folium incarnaium.

Cocci d. See Orthezia insignis.

Cochlear ia armor acta, host plant of Thielavia

basicola . 293

Comparison of the Nitrifying Powers of Some

Humid and Some Arid Soils (paper) . 47-82

Com. See Zea mays.

Corozo nut. See Phytelephas macrocarpa.
Correlation Between the Size of Cannon Bone
in the Offspring and the Age of the Parents

(paper) . 361-371

Croton bug. See BlatteUa germanica.

Cucumis spp., host plants of Thielavia basicola 294
Cucurbita spp., host plants of Thielavia basi¬
cola . 294

Culpepper, C. W., Foster, A. C., and Caldwell,

J. S. (paper): Some Effects of the Blackrot
Fungus, Sphaeropsis malorum, upon the

Chemical Composition of the Apple . 1 7-40

Currie, J. N., and Thom, C. (paper): Aspergil¬
lus niger Group . 1-1 5

Cushmail, R. A, (paper): Syntomaspis dru-

parum, the Apple-Seed Chalcid . . 487-502

Cyclamen sp. , host plant of T hielavia basicola . 293

Cypripedium sp., host plant of Thielavia basi¬
cola . 293

Cytisus sw Partus, host plant of Thielavia basi¬
cola . 294

Daily Transpiration During the Normal
Growth Period and Its Correlation with the
Weather (paper) . . 155-aia

Oct. 2-Dec. 26, 1916

Index

551

Datana sp. — Page

effect of nicotine sulphate on . 93

nicotine in the system of . 104

Datura —

fastuosa, host plant of the mosaic disease ... 483

meteloides, transpiration experiments with

abscised leaves of . 530-536

spp., host plant of Thielavia basicola . 294

stramonium , host plant of mosaic disease . . . 483

Daucus carota, host plant of Thielavia basicola. 293
Desmodium tortuosum, host plant of Thielavia

basicola . . . 294

Diplodia tubericola, growth of, in concentrated

solutions . 357

Dolichos lablab , host plant of Thielavia basicola 294
Dryrot, association with Spongospora subter-
ranea . 240-251

Echinochloa crus-galli , food plant of Macrosi-

phum granarium . 466

Effect of Nicotine as an Insecticide (paper) . . 89-122
Effect of Sodium Salts in Water Cultures on
the Absorption of Plant Food by Wheat

Seedlings (paper) . 407-416

Eggplant. See Solanum melongena.

Eleusine indica , food plant of Macrosiphum

granarium . 465

Elymus sp., food plant of Macrosiphum gra¬
narium . 465

Energy Values of Red-Clover Hay and Maize

Meal (paper) . 379-387

Epicauia Pennsylvania, effect of nicotine sul¬
phate on . 94

Errata . v

Factors Affecting the Evaporation of Moist¬
ure from the Soil (paper) . 439-461

Festuca —

duriuscula [ovina], food plant of Macrosi¬
phum granarium . 465

heterophylla, food plant of Macrosiphum

granarium . 465

praiensis [eliator], food plant of Macrosi¬
phum granarium . 465

tectorum, food plant of Macrosiphum gran¬
arium . 465

Fish-oil-soap sprays, relationship between the

wetting power and efficiency of . 389-399

Fly¬
blow. See Calliphora vomitoria.
house. See Musca domestica.

Food value of vegetable-ivory meal . 301-320

Formation of Henlatoporphyrin in Ox Muscle

During Autolysis (paper) . 41-45

Foster, A, C., Caldwell, J. S., and Culpepper,

C. W. (paper); Some Effects of the Blackrot
Fungus, Sphaeropsis malorum, upon the

Chemical Composition of the Apple . 17-40

Freezing-point lowering, determination of . . . 263

Freezing-Point Dowering of the Leaf Sap of
the Horticultural Types of Persea americana

(paper) . 261,268

Fries, J. A., Braman, W. W., and Armsby,

H. P. (paper): Energy Values of Red-Clover

Hay and Maize Meal . 3 79-387

Fungus, parasitic, growth of, in concentrated
solutions . 255-260

Fusarium — Page

heterosporum , parasite of Paspalum dilata-

tum . 404

spp., growth of, in concentrated solutions. . 257

Galactia sp. , host plant of Thielavia basicola . . . 294

Gile, P. L-, and Carrero, J. O. (paper);

Assimilation of Iron by Rice from Certain

- Nutrient Solutions . 503-528

Immobility of Iron in the Plant . 83-87

Glomerelta cingulata, growth of, in concen¬
trated solution . 257

Glycine hispida—

host plant of Thielavia basicola . 294

transpiration experiments with abscised

leaves of . 536-539*542-543

Gossypium herbaceum , host plant of Thielavia

basicola . 293

Grain of the Tobacco Leaf (paper) . 269-288

Grasshopper. See Melanoplus femoratus.

Growth of Parasitic Fungi in Concentrated
Solutions (paper) . 255-260

Habrocytus medicaginis —

adult stage of . 152

appearance of, in fields . 150

choice of host plants of . 152

classification and description of . 148-149

discovery of . . . *. 147-148

hibernation of . 153

larval stage of . 151

life history of . 147-154

method of studying . 149

oviposition of . 150

pupal stage of . 151-152

rate of parasitism of . 153

relative proportion of sexes of . 152

seasonal history of . 1 5 2-1 53

Harris, F. S., and Robinson, J. S. (paper):
Factors Affecting the Evaporation of

Moisture from the Soil . 439-461

Harris, J. A., and Popenoe, W. (paper):
Freezing-Point of the Leaf Sap of the Horti¬
cultural Varieties of Persea americana. . . 261-268
Hawkins, L . A. (paper): Growth of Parasitic

Fungi in Concentrated Solutions . 255-260

Hay, red-clover —

composition of . 380

energy value of . . 379-387

loss of energy in feeding . . . 382-383

metabolizable energy in . 381-382

net energy value of . 387

percentage digestibility of . 381

Heat-

effect of, on hydrogen-ion concentration in

soil . 130

production of, in animals fed with red-clover

hay and maize meal . 384

Hematoporphyrin —

experiments on the formation of, during

autolysis . ■ . 41-44

f ormation of, in ox muscle during autolysis . 4145
significance of the formation of, during au¬
tolysis . 44-45

Hibiscus cardinalis, transpiration experiments
with abscised leaves of . 530-536

552

Journal of Agricultural Research

Vol. VII

Page

Hoagland, D. R., and Sharp, L. T. (paper):
Acidity and Adsorption in Soils as Meas¬
ured by the Hydrogen Electrode . 123-145

Hoagland, Ralph (paper): Formation of He-
matoporphyrin in Ox Muscle During

Autolysis . 4I-45

Honeybee. See Apis mellifica.

Hordeum pusUlum , food plant of M acrosiphum

QTanarium . 466

Host Plants of Thielavia basicola (paper) . . 289-300
Hydrogen electrode, acidity and adsorption

in soils as measured by the . 123-145

Hygroscopic coefficient —

concordance of . 347

estimation of—

f rom hygroscopic moisture . 3 5 1-388

from maximum water capacity . 348-351

of soils, use of two indirect methods for the

determination of . 345“359

Hyphantrea cunea , effect of nicotine sulphate
on . 95

Immobility of Iron in the Plant (paper) . 83-87

Influence of Bordeaux Mixture on the Rates
of Transpiration from Abscised Leaves and

from Potted Plants (paper) . 529-548

Insecticide, effect of nicotine as an . 89-122

Ipomoea coccinea, host plant of Thielavia

basicola . 294

Iron-

analysis of leaves of Citrus limonum for . . 85-86

assimilation of , by rice . 503-5 ^ 28

immobility of, in the plant . 83-87

Johnson, J. (paper): Host Plants of Thielavia

basicola . 289-300

Juncus tenuis , food plant of Macrosiphum
granarium . 465

Kelley, W. P. (paper): Nitrification in Semi-

arid Soils . 417-437

Klein, M. A., Lipman, C. B., and Burgess,

P. S. (paper): Comparison of the Nitrifying
Powers of Some Humid and Some Arid
Soils . 47-82

Laspeyresia molesta, an Important New

Insect Enemy of the Peach (paper) . 373"378

Laspeyresia molesta —

character of injury by . 375~377

description of.. . 373“374

Lathyrus odoratus , host plant of Thielavia

basicola . 293

Lemon, rough. See Citrus limonum.

Lens esculenta , host plant of T hielavia basicola . 294

Leptinotarsa decemlineata , effect of nicotine

sulphate on . 95

Lespedeza striata, host plant of Thielavia

basicola . 294

Life History and Poisonous Properties of

Claviceps paspali (paper) . 401-406

Life History of Habroeytus medicaginis, a
Recently Described Parasite of the Chalcis

Fly in Alfalfa Seed (paper) . 147-154

Lime requirement, estimate of, by electro¬
metric method . 130-13 2

Page

Linar ia spp., host plants of Thielavia

basicola . 293-294

Lindsey, J. B., and Beals, C. L. (paper):
Chemical Composition, Digestibility, and
Feeding Value of Vegetable-Ivory Meal. . 301-320
Lipman, C. B., Burgess, P. S., and Klein,

M. A. (paper): Comparison of the Nitrifying
Powers of Some Humid and Some Arid

Soils . 47-82

Lolium italicum , food plant of Macrosiphum

granarium . 465

Lotus —

corniculatus , host p lant of T hielav ia basicola . 2 94

viUosus , host plant of Thielavia basicola - 294

Lupinus —

albus, host plant of Thielavia basicola . 293

angustifolius, host plant of Thielavia basi¬
cola . 293

hirsuius, host plant of Thielavia basicola - 294

luteus, host plant of Thielavia basicola . . 293

thermis , host plant of Thielavia basicola . 293

Lycopersicon esculentum —

host plant of Spongospora subterranea - 222-223

transpiration experiments with potted
plants of . 536-539; 544-546

Mclndoo, N. E. (paper): Effects of Nicotine as

an Insecticide . 89-122

Macrosiphum granarium, the English Grain

Aphis (paper) . 463-480

Macrosiphum —

avenivorum, syn. Macrosiphum granarium.

granarium —
description of —

egg of . 469

forms of . 466-469

sexes of . 468-469

distribution of . 465

food plants of . 465-466

fungus enemies of . 479

life history and habits of . 469-478

natural enemies of . 478-479

synonymy of . 463-464

sanborni —

nicotine in the system of . 103

Maize. See Zea mays.

Malus —
sp —

acidity of normal and diseased . 34-36

alcohol determinations in sound and dis¬
eased . 36-37

host plant of A phis malifoliae . 321-344

sylvestris, some effects of Sphaeropsis ma-
lorum upon the chemical composition of . . 1 7-40
Martin, W. H. (paper): Influence of Bordeaux
Mixture on the Rates of Transpiration from
Abscised Leaves and From Potted Plants 529-548
Medicago —

denticulata , host plant of Thielavia basicola . 294

sativa —

host plant of Thielavia basicola . 293

life history of Habroeytus medicaginis , a
recently described parasite of Brucho-

phagus funebris in the seed of . 147-154

Melanoplus femoratus, nicotine in the sys¬
tem of

ios

Oct. 2-Dec. 26, 1916

Index

553

Page

Melhus, I. E., Rosenbaum, J., and Schultz,

E. S. (paper): Spongospora subterranea and

Phoma tuberosa of the Irish Potato . 213-254

Melilotus—

alba , host plant of Thielavia basicola . 294

indica, host plant of Thielavia basicola . 294

Methane, relation of, to carbohydrates in red-

clover hay and maize meal . . . 383

Mosaic disease-

characteristics of . 484-485

of Nicotiana iabacum, comparison with mo¬
saic disease of Nicotiana glutinosa . 481-486

Musca domesiica, effect of nicotine-sulphate

solution on . 95

Muscle, ox, formation of hematoporphyrin in,

during autolysis of, significance of . 44-45

Myzus persicae —

effect of nicotine-sulphate fumes on . 95

passage of nicotine through the tissues of . . . 100-

101,109-113

Nasturtium. See TroPaeolus majus.

Nectarophora granaria, syn. Macrosipkum
granarium.

Nemopkilia spp., host plants of Thielavia basi¬

cola . 283-294

New species . 251,373-374

Nicotiana —
glutinosa —

inoculations of, with mosaic factors . 482

mosaic disease in, comparison with mo¬
saic disease of N. Iabacum . . . 481-486

rusttca , host plant of Thielavia basicola .... 293

fabacum —

analysis of the leaf of . 272

chemical nature of grain in the leaf of. . 272-276
correlation of grain with burning quality

in the leaf of . 276-284

crystalline matter in the leaf of . 271

development of grain in the leaf of . 284-286

forms of grain in the leaf of . 270-271

grain of the leaf of . 269-288

host plant of Thielavia basicola . 293

macroscopic appearance of grain in the

leaf of . 269-270

microscopic characters of grain, the leaf of . 270

mosaic disease of, comparison with mo¬
saic disease of Nicotiana viscosum _ 481-486

occurrence of grain in the leaf of . 269-270

spp.—

characteristics of the mosaic disease of. 484-485

host plants of Thielavia basicola . 294

viscosum. See Errata, p. v.

Nicotine-^
effect of—

as a fumigant . 94-95

as an insecticide. . 89-122

as a stomach poison . 90^2

odor and vapor of, on insects . 95-98

on bees . 90-92

occurrence of, in insect tissues . 98-113

physiological effects of, on insects . 90-98

spray solutions, effect of . 92-94

Nicotine-sulphate sprays, relationship be¬
tween the wetting power and efficiency of . 389-399
Nitrification in Semiarid Soils (paper) . 417-437

Page

Nitrification, soil —

effect of concentrations of nitrogenous ma¬
terials on . 422-425

at varying depths as affected by concentra¬
tion. . 425-426

during different lengths of time, . 428-429

in some humid and some arid soils, com¬
parison of . 47-82

Nitrites, accumulation of, in nitrification. . 429-433

Onobrychis—

crista-galli, host plant of Thielavia basicola.. 293
viciae folia, host plant of Thielavia basicola. . . 494

Ornithopsus sativus, host plant of Thielavia

basicola . 494

Orthezia insignis —

effect of nicotine-sulphate solution on . 95

nicotine in the tissues of . 102-103

Oryza sativa—

assimilation of iron by . 503-525

chlorosis of . 84

growth of, in nutrient solutions . 83-84

immobility of iron in . 83-87

leaves of, effect of brushing, with iron salts . 84-85
Ox muscle, formation of hematoporphyrin

in, during autolysis . 41-45

Oxalis corniculata , var. stricta, host plant of
T hielavia basicola . 293

Papaver nudicaule , host plant of Thielavia

basicola . 294

Paphiopedilum grossianum , host plant of

Thielavia basicola . 294

Parasitic fungi, diffusion tension of juice of

hosts of . 258

P a$P alum dilatatum —
host plant of —

Cladosporium sp . 404

Claviceps paspali . 401-406

Fusarium heterosporum . 404

Pastinica sativa , host plant of Thielavia basi¬
cola . 293

Pepper. See Capsicum annuum.

Periplaneta americana , passage of nicotine

through the tissues of . 100

Per sea americana —

freezing-point lowering of the leaf sap of the

horticultural varieties of . 261-268

presentation of constants for . 263-266

Petunia ( hybridaf ), host plant of Thielavia

basicola . 294

Phaseolus —

acutifolius, host plant of Thielavia basicola. . 294

muUiflorus , hos t plant of T hielavia basicola . . 293

vulgaris —

host plant of Thielavia basicola . 293

transpiration experiments with abscised

leaves of . 53°"536

Phillips, W. J. (paper): Macrosiphum grana¬
rium, the English Grain Aphis . 463-480

Phlox drummondii, host plant of Thielavia

basicola . 294

Phoma tuberosa —

description of . 251

occurrence on Solanum tuberosum . 213-254

554

Journal of Agricultural Research

Vol. VII

Page

Pkytelephas macrocar pa, meal of the seed of —

calorific value of . 305

digestion experiments with . 306-311

feeding experiments with . 311-318

Pineapple. See Ananas sativus.

Pisum sativum, host plant of Thielavia bast -

cola .

Plenodomus destruens, growth of, in concen¬
trated solutions . .

Plum . See Prunus spp .

Poa compressa, food plant of Macrosiphum

granarium .

Poa pratensis, food plant of Macrosiphum

granarium .

Popenoe, W., and Harris, J. A. (paper): Freez¬
ing-Point lowering of the Leaf Sap of the
Horticultural Varieties of Persea ameri-

cana. . . . . 261-268

Poplar, Carolina. See Populus deltoides.

Populus deltoides, host plant of Aphis populi-

foliae . 92“93

Portulaca oleracea, host plant of Thielavia

basicola .

Potato. See Solanum tuberosum .

Potato beetle. See Lepiinotarsa decemlineata.
Prunus spp., host plants of Laspeyresia mo-
lesta .

Quaintance, A L., and Wood, W. B. (paper) :
Laspeyresia molesta, an Important New
Insect Enemy of the Peach . 373-378

Radish. See Raphanus sativus.

Raphanus sativus, transpiration in abscised

leaves of . 530-536

Relation Between the Wetting Power and
Efficiency of Nicotine-Sulphate and Fish-

Oil-Soap Sprays (paper) . 389-399

Rhizopus nigricans , growth of, in concentrated

solution . 257

Rice. See Oryza saliva.

Ricinus communis, transpiration in abscised

leaves of . 530-536

Ridgway, C. S. (paper): Grain of the Tobacco

Leaf . 269-288

Roach. See Periplaneta americana.

Robinia pseudoacacia, host plant of Thielavia

basicola . 294

Robinson, J, S., and Harris, F. S. (paper):
Factors Affecting the Evaporation of Mois¬
ture from the Soil . 439-461

Rosenbaum, J., Schultz, E. S., and Melhus,

I. E. (paper): Spongospora subterranea and

Phoma tuberosa of the Irish Potato . 213-254

Rosy Apple Aphis (paper) . 321-344

294

375

293

255

465

Schultz, E. S., Melhus, I. E., and Rosen¬
baum, J. (paper): Spongospora subterranea
and Phoma tuberosa of the Irish Potato . . 213-254

Sclerotinia cinerea, growth of, in concentrated

solution . . . 257

Sclotis chinensis, host plant of Thielavia basi¬
cola . 294

Scorzonera htspanica , host plant of Thielavia
basicola . . ... 293

Page

Senecio elegans, host plant of Thielavia basi¬
cola . 293

Shantz, H. L., and Briggs, L. J. (paper):

Daily Transpiration During the Normal
Growth Period and Its Correlation with

the Weather . 155-212

Sharp, L. T., and Hoagland, D. R. (paper):
Acidity and Adsorption in Soils as Meas-

■ ured by the Hydrogen Electrode . 1 23-145

Siphonophora —

cerealis , syn. Macrosiphum granarium.
granaria, syn, Macrosiphum granarium.

Smith, L. B. (paper): Relationship Between
the Wetting Power and Efficiency of Nico¬
tine-Sulphate and Fish-Oil-Soap Sprays . . 389-399
Sodium-
carbonate —
effect of—

on absorption of nutrients by seedlings

of Triticum vulgare . . . 415

on composition and weight of seedlings

of Triticum sp . 410-412

chlorid—
effect of—

in solutions, on plant food of Triticum

vulgare . 408-409

on absorption of nutrients by seedlings

of Triticum sp . 413-414

salts in water cultures, effect of, on the ab¬
sorption of plant food by wheat seed¬
lings . 407-416

sulphate-
effect of —

in solutions, on plant food in Triticum

vulgare . 409-410

on absorption of nutrients by seedlings
of Triticum vulgare . 414

Soil-

acidity and adsorption in, as measured by

the hydrogen electrode . 123-145

adsorption of OH ions by . 133-143

comparative nitrification in . 72-76

comparison of the nitrifying powers of

humid and arid . 47-82

effect of grinding, on the hydrogen-ion con¬
centration of their suspension . 1 29-130

evaporation of moisture from —
effect of —

compacting the soil on . 456-457

humidity on . 447-449

method of applying water on . 457-458

mulches on . 454-455

size of soil particles on . 453-454

soluble salts on . 458-459

sunshine on . 45<>"453

wind velocity on . 449-450

experiments on nitrification in, in Cali¬
fornia . 55-72

factors affecting the evaporation of moisture

from . 439-461

from various States, comparison of nitri¬
fication in . 50-55

hygroscopic coefficients of, two indirect
methods for the determination of. . 345-359

Oct. 2-Dec. 26, 1916

Index

555

Soil— Continued. Page

reactions, relation of equilibria to . 127-128

semiarid, nitrification in . 417-437

effect of neutral salts on the H-ion con¬
centration of . ■ . 132-133

hydrogen-ion concentration of . 12 5-1 2 7

used in adsorption and acidity experiments,

description of . 125

Solatium —

carolinense, host plant of T hielavia basicola . . 294

melongena, transpiration in abscised leaves

of . 53<5~539>544

spp., host plants of Spongospora sub-

terranea . 221-323

tuberosum —
host plant of—

Myzus persicae . . 95

Spongospora subterranea . 213-254

Some Effects of the Blackrot Fungus, Sphae-
ropsis malorum, upon the Chemical Com¬
position of the Apple (paper) . 17-40

Soybean. See Glycine hispida.

Species, new . 251,373-374

Specific Mosaic Disease in Nicotiana viscosum
Distinct from the Mosaic Disease of Tobacco

(paper).. . 481-486

Sphaeronema fimbriatum , growth of, in con¬
centrated solution . 257

Sphaeropsis malorum —

analyses of tissues affected with . 34-3 2 .

changes produced by, on Malus sp. in

artificial culture . 32-34

growth of, in concentrated solution . 357

methods of analysis of tissues affected with. 18-34
some effects of, upon the chemical com¬
position of Malus sylvestris . 17-40

Spongospora subterranea and* Phoma tube-

rosa on the Irish Potato (paper) . 2x3-254

Spongospora subterranea —

control measures taken with . 228r-24o

distribution of, in the United States . 213-217

dryrot associated with. . . : . 240-251

histology of the galls of . 223-224

host plants of . . . 221-233

on the tuber of Solanum tuberosum . . — 224-226
prevalence and period of existence of, in the

United States . 2x7-218

susceptibility of roots, stolons, and stems of

Solanum tuberosum tek . 219-221

symptoms of . 244-246

Spray-

contact, formulae for, tested . 391-392

fish-oil soap, relationship between wetting

power and efficiency of . 389-399

methods of determining the wetting power

and efficiency of . 390-391

nicotine-sulphate, relationship between the

wetting power and efficiency of . 389-399

Slropkostyles helvola, host plant of Tkielavia

basicola . 394

suspensions—

Syntherisma sanguinale, food plant of Macrosi-

phum granarium . 466

Syntomaspis druparum, the Apple-Seed Chal-
cid (paper) . 487-502

Syntomaspis druparum — Page

control of . 500-501

description of adult of . 487

distribution in the United States . 487

economic importance of . 500

effect of , upon fruit . 488-489

fruits attacked by . 489^-491

life history of . 491-500

Tepkosia Virginia, host plant of Tkielavia

basicola . 294

Tkielavia basicola, host plants of . 289-300

Thom, C., and Currie, J. N. (paper): Aspergil¬
lus niger Group . x-15

Thyridopteryx epkemeraeformis, effect of nico¬
tine sulphate on . 94

Tobacco. See Nicotiana tabacum.

Tomato. See Lycopersicon esculentum.
Transpiration-
daily —

correlation of, with weather and evapora¬
tion . 204-310

during the normal growth period and its

correlation with the weather . 155-212

comparison of energy received and dissi¬
pated in . 185-187

from abscised leaves and from potted plants,
influence of Bordeaux mixture on the

rates of . 529-548

measurements of plants . 156-173

of different crops, comparison of . 174

period, maximum loss of water during. . . 178-185

relation of, to the weather . 187-304

water loss during periods of . 1 74-177

Trifolium—

hybridum, host plant of Tkielavia basicola . . . 293

incarnatum —

host plant of Tkielavia basicola . 294

See also Hay, red-clover.
pratense, host plant of Tkielavia basicola, ... 293

repens, host plant of Tkielavia basicola . 293

Trigonella —

coerulea , host plant of Tkielavia basicola — 293

foenum-graecum , host plant of Tkielavia

basicola . 294

Triticum vulgare —

effect of salts on absorption of nutrients by 407-416
seedlings of —

determination of plant-food absorption

by . . — 407-408

effect of sodium salts in water cultures on

the absorption of plant food by . 407-416

Tropaeolus majus, host plant of A pkis spp ... 95

Turner, W. F„ and Baker, A.C. (paper): Rosy
Apple Aphis . 321-344

Ulex euroPaeus, host plant of T kielavia basicola 294
Urbahns, T. D. (paper): Life History of Hab-
rocytus medicaginis, a Recently Described
Parasite of the Chalcis Fly in Alfalfa Seed. 147-154
Use of Two Indirect Methods for the Determi¬
nation of the Hygroscopic Coefficients of
Soils (paper) . 345-359-

556

Journal of Agricultural Research

Vol. VII

Page

Vegetable ivory-

chemical analysis of . 302-305, 310-3 1 a

meal, food value of . 301-320

nut. See Phytelephas macrocarpa.

Vicia —

faba, host plant of Thielavia basicola . 39 4

villosa, host plant of Thielavia basicola . 294

Vigna sinensis, host plant of Thielavia basicola 293
Viola—

odorata, host plant of Thielavia basicola . 293

tricolor, host plant of Thielavia basicola . 294

Wax moth. See A chroia grisella.

Weather, correlation of daily transpiration

during the normal growth period . 155-2 12

Webworms, fall. See Hyphantrea cunea .

Wheat. See Triticum vulgare.

Page

Wood, W. B„ and Quaintance, A. I,, (paper):
Easpeyresia molesta, an Important New

Insect Enemy of the Peach . 373-378

Wriedt, Christian (paper): Correlation Be¬
tween the Size of Cannon Bone in the Off¬
spring and the Age of the Parents . 361-371

Zea mays —

and red-clover hay, loss of energy in feed¬
ing . 382-383

meal of—

composition of . 380

energy expenditure per kilogram of . 385

energy value of . 379-387

metabolizable energy in . 383

net energy values of . 387

percentage digestibility of . 381

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Journal of agricultural research"

See other formats