lIllillliliiB
iiii: , .iiiiiiiii^
' "iliiiiiiiiiiiiil
iiiiiiiiiiiii;:;;;^^;:;^!;:^^^;:^
JOURNAL OF
AGRICULTURAL
RESEARCH
Volume XVII
APRIL i5~SEPTEMBER 15, 191 9
PUBLISHED BY AUTHORITY OF THE SECRETARY OF AGRICULTURE
WITH THE COOPERATION OF THE ASSOCIATION OF AMERICAN
AGRICULTURAL COLLEGES AND EXPERIMENT STATIONS
WASHINGTON, D. C.
EDITORIAL COMMITTEE OF THE
UNITED STATES DEPARTMENT OF AGRICULTURE AND
THE ASSOCIATION OF AMERICAN AGRICULTURAL
COLLEGES AND EXPERIMENT STATIONS
FOR THE DEPARTMENT
KARL F. KELLERMAN, Chairman
Physiologist and Associate Chief, Bureau
of Plant Industry
EDWIN W. ALLEN
Chief, Office of Experiment Stations
CHARLES L. MARLATT
Entomologist and Assistant Chief, Bureau
of Entomology
FOR THE ASSOCIATION
H. P. ARMSBY
Director, Institute of Animal Nutrition, The
Pennsylvania State College
J. G. LIPMAN
Director, New Jersey Agricultural Eiperimen
Station, Rutgers College
W. A. RILEY
Entomologist and Chief , Division of Entomology
and Economic Zoology, Agricultural Experiment
Station of the University of Minnesota.
All correspondence regarding articles from the Department of Agriculture should be
addressed to Karl F. Kellerman, Journal of Agricultural Research, Washington, D. C.
All correspondence regarding articles from State Experiment Stations should be
addressed to H. P. Armsby, Institute of Animal Nutrition, State College, Pa.
CONTENTS
Page
Meat Extracts, Their Composition and Identification. James A.
Emery and Robert R. Henley i
Quantity and Composition of Ewes' Milk: Its Relation to the
Growth of Eambs. Ray E. Neidig and E. J. Iddings 19
Seed Disinfection by Formaldehyde Vapor. Cecil C. Thomas. . 33
Influence of Soil Environment on the Rootrot of Tobacco. James
Johnson and R. E. Hartman 41
Relation of Sulphates to Plant Growth and Composition. H. G.
Miller 87
Relation of Weather to Fruitfulness in the Plum. M. J. Dorsey. 103
Structure of the Maize Ear as Indicated in Zea-Euchlaena Hy-
brids. G. N. Collins 127
Carbohydrate Metabolism in Green Sweet Corn during Storage
at Different Temperatures. Charles O. Appleman and John
M. Arthur 137
Certain Relationships between the Flowers and Fruits of the
Eemon. Howard vS. Reed 153
Ultra-Microscopic Examination of Disperse Colloids Present in.
Bituminous Road Materials. E. C. E. Eord 167
Derris as an Insecticide. N. E. McIndoo, A. F. SeivERS, and
W. S. Abbott 177
Effects of Heat on Trichinae. B. H. Ransom and Benjamin
Schwartz 201
Effect of Removing the Pulp from Camphor Seed on Germination
and the Subsequent Growth of the Seedlings. G. A. RussELL. 223
Bacterium abortus Infection of Bulls. J. M. Buck, G. T. CreEch,
and H. H. Ladson 239
Investigations on the Mosaic Disease of the Irish Potato. E. S.
Schultz, Donald Folsom, F. Merrill Hildebrandt, and
LoN A. Hawkins 247
Temperature in Relation to Quality of Sweetcorn. Neil E.
Stevens and C. H. Higgins 275
Variation of Ayrshire Cows in the Quantity and Fat Content of
Their Milk. Raymond Pearl and John Rice Miner 285
Index 323
ERRATA AND AUTHORS' EMENDATIONS
Page 82, TableIL"Gni." should read "Pounds."
Page 158, line 2, "variability" should read "viability."
Page 160, line 23, "0.183" should read "—0.183."
Page 162, Table VII, " Months required for maturity " should stand above the table. The legend at
the left of the table should be "Month in which fruit set."
Page 179, line 30, ''petroleum, ether" should read " petroleum ether."
(ni)
ILLUSTRATIONS
PLATES
Influence of Son, Environment on the Rootrot op Tobacco
Page
Plate i. I. — Influence of amount of infestation on injury from tobacco rootrot:
A, All tminfested soil; B, three-fourths iminfested soil; C, one-half unin-
fested soil; D, one-fovirth uninfested soil; E, all infested soil. II, III. —
Influence of moisture content of soil on the amoirnt of injury done by tlie
tobacco rootrot; II, infested soil; III, iminfested soil (control series) — lA,
one-fourth saturation infested soil; 2 A, one-half saturation infested soil;
3A, three-fourths saturation infested soil; 4A, full saturation infested soil;
I B , one-fourth saturation uninfested soil ; 2 B , one-half saturation uninfested
soil; 3B, three-fourths saturation uninfested soil; 4B, full saturation unin-
fested soil. IV. — Influence of soil temperature on the growth of tobacco in
infested soil (jars to left of temperature labels) and in uninfested soil (jars to
right of temperatm-e labels) at temperatures of approximately 13°, 17°, 23°,
26°, and 36° C ' 86
Plate 2. I. — Soil temperature tanks used in the temperature experiments.
II, III. — Influence of soil temperature on the growth of tobacco: lA, in-
fested soil, i7°-i8° C. ; 2A, infested soil, 2o°-2i° C; 3A, infested soil, 23°-
24° C; 4A, infested soil, 25°-26° C; 5A, infested soil, 28°-29° C; 6A,
infested soil, 3i°-32° C; iB, uninfested soil, i7°-i8° C; 2B, tminfested
soil, 2o°-2i° C; 3B, uninfested soil, 23°-24'' C.;4B, iminfested soil, 25°-26°
C; 5B, tminfested soil, 28°-29° C; 6B, uninfested soil, 3i°-23° C. IV.—
Influence of different soil temperatiueson root development: lA, luiinfested
soil, i7°-i8° C; iB, infested soil, i7°-i8° C; 2A, uninfested soil, 2o°-2i°
C; ^B, infested soil, 2o°-2i° C; 3A, uninfested soil, 23°-24° C; 3B, in-
fested soil, 23^-24° C; 4A, uninfested soil, 2 5°-26° C; 4B, infested soil,
25°-26° C; 5A, uninfested soil, 28°-29° C; 5E, infested soil, 28°-29° C;
6A, iminfested soil, 3i°-32° C; 6B, infested soil, 3i°-32° C 86
Plate 3. Influence of high (30° C.) and low (20" C.) soil -temperature on
recovery of plants in infested soil 86
Plate 4. I, II. — Influence of soil reaction on extent of damage by tobacco
rootrot: I, Infested soil; II, uninfested soil — lA, infested soil, lime require-
ment 9.38 tons per acre; 2A, infested soil, lime requirement 7.19 tons per
acre; 3A, infested soil, lime requirement 4.60 tons per acre; 4A, infested
soil, lime requirement 2.62 tons per acre; 5A, infested soil, lime require-
ment 0.72 ton per acre; 6A, infested soil, slightly alkaline; 7A, infested
soil, strongly alkaline; iB, uninfested soil, lime requirement 9.38 tons per
acre; 2B, uninfested soil, lime requirement 7.19 tons per acre; 3B, unin-
fested soil, lime requirement 4.60 tons per acre; 4B, uninfested soil, lime
requirement 2.62 tons per acre; 5B, uninfested soil, lime requirement
0.72 ton per acre; 6B, tminfested soil, slightly alkaline; 7B, tminfested soil,
strongly alkaline. Ill, IV. — Influence of the amount of organic matter
in the soil on injury by tobacco rootrot: III, 1A-6A, Influence of gradually
increasing amotmts of organic matter in infested soil from lA, no organic
matter, to 6A, all leaf mold. IV. — 1B-6B, Influence of gradually increasing
amounts of organic matter in uninfested siol from iB, no organic matter, to
(V)
VI Journal of Agricultural Research voi. xvii
Page
6B, all leaf mold (control series). V, VI. — Influence of the amount of
organic matter in the soil on injury by tobacco rootrot: V, 1A-5A, Influence
of gradually increasing amounts of organic matter in tminf ested soil from i A,
no organic matter, to 5A, all leaf mold (control series); VI, 1B-5B, Influ-
ence of gradually increasing amounts of organic matter in infested soil from
iB,noorganicmatter, to 5B, all leaf mold 86
Plate 5. I. — Influence of relative amount of sand and clay on tobacco rootrot:
A, uninfested series: B, infested series — lA, iminfested soil, three-fotirths
clay and one-foiorth sand; iB, infested soil, three-fourths clay and one-
fourth sand; 2A, uninfested soil, one-half clay and one-half sand; 2B,
infested soil, one-half clay and one-half sand; 3A, uninfested soil, three-
fourths sand; 3B, infested soil, three-fourths sand; 4A, uninfested soil, all
sand; 4B, infested soil, all sand. II, III. — Influence of soil fertility on
amount of tobacco rootrot: II, infested series; III, uninfested series —
lA, infested soil, no treatment; 2A, infested soil, 3.5 gms. of nutrient salts;
3A, infested soil, 7.Q gms. of nutrient salts; 4A, infested soil, 14.00 gms. of
nutrient salts; 5A, infested soil, 28 gms. of nutrient salts; 6A, infested soil,
56 gms. of nutrient salts. Note increasing injury from nutrient salts
beginning at pot 3 A. iB, iminfested soil, no treatment; 2B, iminfested
soil, 3.5 gms. of nutrient salts; 3B, iminfested soil, 7.0 gms. of nutrient
salts; 4B, iminfested soil, 14.00 gms. of nutrient salts; 6B, uninfested soil,
28 gms. of nutrient salts; 6B, uninfested soil, 56 gms. of nutrient salts.
Note injury from nutrient in pots 5B and 6B. IV. — Relation of com-
pactness of soil to injury caused by Thielavia basicola: lA, infested soil,
loosely packed; iB, uninfested soil, loosely packed; 2A, infested soil,
very compact; 2B, iminfested soil, very compact. V. — Influence of
transplanting infected seedlings in healthy soil: A, Pennsylvania
Broadleaf infected seedlings; B, Pennsylvania Broadleaf healthy seedlings
C, WTiite Burley infected seedlings; D. White Burley healthy seedlings;
E, Northern Hybrid (a resistant type) infected seedlings; F, Northern
Hybrid (a resistant type) healthy seedlings 86
Plate 6. Soil temperature graphs for the month of June, 1915-1918, inclusive,
at depths of 2, 4, and 8 inches 86
Plate 7. Soil temperature graphs for the month of July, 1915-1918, inclusive,
at depths of 2, 4, and 8 inches 86
Plate 8. Soil temperature graphs for the m.onth of August, 1915-191S, in-
clusive, at depths of 2, 4, and 8 inches 86
Relatio.nt op Sulphates to I^lant Growth and Composition'
Plate 9. A. — Clover on soil A. The top row, reading from left to right, shows
the soil pots \\hich received the following fertilizers: Pot i, calcium
sulphate, sodium nitrate; pot 2, sodium sulphate, sodium nitrate; pot 3,
sulphur, sodium nitrate, calcium carbonate; pot 4, sodium nitrate; pot 5,
sodium nitrate, calcium carbonate; pot 6, no fertilizer; pot 7, calcium
sulphate, sodium nitrate; pot 8, sodium sulphate, sodium nitrate; pot 9,
sulphur, calcium carbonate, sodium nitrate; pot 10, sodium nitrate.
B. — Clover on soil B. C. — Clover on soil C 103
Plate 10. A. — Rape on soil A. B. — Rape on soil B. C. Rape on soil C. . . . 102
Pl.-\te II. A. — OatsonsoilA. B. — OatsonsoilB. C. — OatsonsoilC 102
Plate 12. A. — Oats on sand cultures from soil A. B. — Oats on sand cultures
from soil B. C. — Oats on sand cultures from soil C 102
Apr. 15-Sept. IS, 1919
Illustrations vn
Relation ok Weather to Fruitfxjuness in the Plum
Page
Plate 13. Plum tree and fruiting branch showing difference between number
of flowers borne and quantity of fruit set: A.— The appearance of a plum
tree bearing a normal crop of bloom. B.— A single fruiting branch 2 years
old showing the contrast to A 126
Plate 14. A. — Stigma of Minnesota No. 21, a greenhouse tree, 24 hours after
being selfed, showing the condition of papillate cells in the stigma, pollen
tubes, and also traces of the stigmatic fluid. B.— Stigma of Minnesota No.
35, open to cross pollination, showing the condition of a stigma three days
after bloom, having withstood a rain of 0.87 inch which fell in the two days
previous, lasting in all 18 hours. C. — The tiurgid papillate cells in Sapa
before receptiveness. D. — Opata. Same as C. E. — Abscission layer Min-
nesota No. 35, showing the cells of the layer 11 days after bloom. F. — The
surface at the abscission layer of Assiniboin after the style has fallen, 12
days after bloom 126
Plate 15. Graphic analysis of the weather from the standpoint of wind, stm-
shine, rain, and temperature for seven years from 1912 to 1918 126
Structure op the Maize Ear as Indicated in Zea-Euchlaena Hybrids
Plate 16. Intermediate stages between a simple spike of the pistillate inflores-
cence of Euchlaena and an ear of maize : A. — Spike of pure Florida teosinte.
B. — Spike with slightly shortened axis. C. — A still more compact spike
with an increased number of seeds. A-C have single spikelets and separate
two-ranked alicoles. D. — Spike with single spikelets and yoked alicoles,
irregularly fotu-rowed. E. — Compact spike with two-ranked separate
alicoles and single spikelets. F. — Spike with paired spikelets and four
ranks of yoked alicoles. G. — Transition stage between four-rowed and
eight-rowed ear. H. — Ear of maize with eight rather poorly defined rows
of seeds 136
Plate 17. Pistillate inflorescences of hybrid between Euchlaena and maize:
A. — Showing pedicelled staminate spikelets with sessile pistillate spikelets.
B. — Closely compacted inflorescense with two rows of alicoles and four
rows of seeds. C-E. — Spirally twisted inflorescences, with three rows of
alicoles 136
Plate 18. Pistillate inflorescences of hybrid between Euchlaena and maize,
showing yoked alicoles: A-C. — The alicoles are in four rows corresponding
to an eight-rowed ear. D. — The alicoles are in five rows, corresponding
to a ten-rowed ear * 136
Ultra-Microscopic Examlnation op Disperse Colloids Present in Bitu-
minous Road Materials
Plate 19. A. — Microscope with ray filter and arc lamp for dark field illumina-
tion. B. — Photomicrograph of cross-line micrometer scale, showing col-
loids in dark field. X320. Taken by E. A. Shuster, jr.. Photographic
Laboratory, United States Geological Survey 176
Effect of Removing the Pulp from Camphor Seed on Germination and
Subsequent Growth op the Seedlings
Plate 20. A camphor seed bed, showing the growth of seedlings from pulped
and unpulped camphor seed planted in alternate rows 238
Plate 21. A. — Camphor seedlings at the time of transplanting. B. — Camphor
seedlings cut back and trimmed ready for transplanting 328
122502°— 19 7
VIII Journal of Agricultural Research voi. xvn
Bacterium abortus Infection of Bulls
Page
PivATE 22. Photograph of normal and diseased seminal vesicles of bull 98, show-
ing the marked increase in size and the gross pathological changes of one of
the organs 246
Plate 23. A. — Photomicrograph of a section from a normal seminal vesicle of
bull. B. — Photomicrograph of section from seminal vesicle of bull 409,
showing inflammatory changes 246
Plate 24. A. — Photomicrograph of section from seminal vesicle of bull 98,
showing tissue proliferation and exfoliation of epithelium lining acini.
B. — Photomicrograph of section from seminal vesicle of bull 98, showing
advanced pathological changes with cell degeneration and necrosis 246
Investigations on the Mosaic Disease of the Irish Potato
Plate A. Foliage of Irish potato, Green Mountain variety 274
Plate B. Foliage of potato, Bliss Triumph variety 274
Plate 25. Leaf of Irish potato, Green Mountain variety, infected with mosaic.
Meditim stage of disease 274
Plate 26. A. — Healthy scion grafted upon diseased stock. B. — An illustration
of a method used for introducing aphids 274
Plate 27. A. — Leaves from graft shown in B of this plate: At right, from
healthy parent of scion; at left, from mosaic stock; in center, from mosaic
• scion. B. — At left, healthy scion grafted to diseased stock. Green Moun-
tain variety; at right, two mosaic shoots of stock. C. — Leaves from cor-
responding parts of plants shown in Plate 29, B 274
Plate 28. A. — 49IX, inoculated artificially with unfiltered juices from mosiac
plant February 22 to March 22, 1919. B. — 473y, inoculated in same way as
49IX, but with juices from healthy plant. 458y, also inoculated with juices
from healthy plant 274
Plate 29. A. — Mosaic of potato transmitted by aphids. 142a, infected plant.
Green Mountain variety. B. — Two plants from the same tuber treated
alike, except that about 200 aphids were introduced upon one when it was
2 inches high 274
Plate 30. A. — Inoculated by means of artificial transfers of aphids from dis-
eased plants. Green Mountain variety. B. — Plants inoculated in same
way as those in A of this plate, but with aphids taken from healthy plants. . 274
TEXT FIGURES
Page
Seed Disinfection by Formaldehyde Vapor
Fig. I. Formaldehyde-vapor disinfecting apparatus 34
Influence of Soil Environment on the Rot op Tobacco
Fig. I. Soil-thermograph, records showing the influence on soil temperature of
the shading of soil by growing tobacco 68
2. Soil-thlrmograph records given comparison of a typical record of
regulated soil temperature in tanks with a typical record from the
field at a depth of 4 inches 69
Relation of Weather to Fruitfulness in the Plum
Fig. I. An outline drawing of an anther of Minnesota No. 12, showing the ad-
justment which takes place as a result of taking up or giving off water:
A, an anther which has been open in the orchard for three days; B,
the same with the anthers pushed up to show the dead area at the
upper end of the filament; C, the appearance of the anther after two
minutes in water iii
Structure of the Maize Ear as Indicated in Zea-Euchlaena Hybrids
Fig. I. Diagram showing arrangement of pedicelled and sessile spikelets in A,
undifferentiated four-rowed branch; B, eight-rowed ear, the result of
the fasciation of two undifferentiated branches; C, eight-rowed ear
the result of twisting a single undifferentiated branch; D, i6-rowed
ear, the result of fasciation; E, i6-rowed ear, the result of a further
twisting of " C" 129
Carbohydrate Metabolism in GrEEn Sweet Corn During Storage at
Different Temperatures
Fig. I. Depletion of total sugars in green sweet com during consecutive 24-hour
periods of storage at different temperatures 146
2. Depletion of sucrose in green sweet corn during consecutive 24-hour
periods of storage, expressed as percentages of the initial sucrose in
the corn, which was 3.87 per cent, wet weight 147
Certain Relationships between The Flowers and Fruits of the Lemon
Fig. I. Average monthly production of lemon buds during the year 155
Ultra-Microscopic Examination of Disperse Colloids Present in Bitu-
minous Road Materials
Fig. I. Glass slide with ultra-microscope cell drawn to natiu^al scale 170
(IX)
X Journal of Agricultural Research Voi. xvii
Effect of Removing the Pulp from Camphor Seed on Germination and
THE Subsequent Growth of the Seedings
Page
Fig. I. Diagram showing percentage of germination of camphor seed secured
from parent tree A under varying conditions 227
2. Graphs showing time required for pulped and unpulped camphor seed
to reach maximum germination 229
3. Graph showing time required for camphor seed secured from parent
tree A at various times and under various conditions to reach
maximum germination 230
4. Diagram showing percentage of total germination of pulped and un-
pulped camphor seed from 10 parent trees 233
5. Graphs showing rates and percentage of germination of pulped and
unpulped camphor seed from 10 parent trees 234
Temperature in Relation to Quality of Sweetcorn
Fig. I. Mean hourlj^ temperature for August at Baltimore, Md., and for Sep-
tember, 1918, at Portland, Me 282
Variation of A\'rshire Cows in the Quantity and Fat Content of Their
Milk
Fig. I. Histograms and fitted curves for variation in mean weekly milk yield
of Ayrshire cows of ages 3 to 7 years 306
2. Histograms and fitted curves for variation in mean weekly milk yield
of Ayrshire cows of ages 8 to 12 years 307
3. Histograms and fitted curves for variation in fat percentage of milk of
Ayrshire cows of ages 3 to 7 years 308
4. Histograms and fitted ciu-ves for variation in fat percentage of milk of
Ayrshire cows of ages 8 to 12 years 309
5. Showing the change in mean weekly yield of milk in Ayrshire cows.
The smooth curve is of the form yKa=bx=cx~=d log x 316
6. Showing the observed (zigzag line) and calculated (straight line) changes
in the mean fat percentage of the milk of Ayrshire cows with advanc-
ing age 318
ammmmmammtmaamaa
Vol. XVI I AP*K 11. 1 5, 1 9 I Q No. 1
JOURNAL OP
AGRICULTURAL
COMTKNXS
Pace
Meat Extracts, Their Composition and Identification - 1
JAMES A. EMERY and ROBERT R. HENLEY
( Contribu{i<m (rom Bureau ot Animal Industry)
Quantity and Composition of Ewes' Milk: Its Relation to
the Growth of Lambs -_--__ ig
RAY E. NEIDIG and E. J. IDDINGS
< Contribution Jrom Idaho Agricultural Kxperimant Statlou)
Seed Disinfection by Formaldehyde Vapor - - * 33
CECIL C. THOMAS
( Contributioti from Federal Horticultural Board )
PUBUSHED BY AUTHORITY OF THE SECRETARY OF AGRICULTURE.
WITH THE COOPERATION OF THE ASSOCIATION OF AMERICAN
AGRICULTURAL COLLEGES AND EXPERIMENT STATIONS
WASHINOXON, O. C.
WASHINarOHSCOVERKMBNT rinNTIN« OFFICE : 1311
EDITORIAL COMMITTEE OF THE
UNITED STATES DEPARTMENT OF AGRICULTURE AND
THE ASSOCIATION OF AMERICAN AGRICULTURAL
COLLEGES AND EXPERIMENT STATIONS
FOR THE DEPARTMENT
KARL F. KELLERMAN, Chairman
Physiologist and A ssociale Chief, Bureau
of PCoHl Industry
EDWIN W. A.LLBN
Chief, 0£ice of Experiment Stations
CHARLES L. ALA.RLATT
Entomologist and Assistant Chief, Bureau
ofEniomotogy
FOR THE ASSOCIATIOH
H. P. ARMSBY
Director, Institute of Animal Sutrition, The
Pennsylvania State CoUege
j. G. UP^LJLN
Director. New Jersey A gricuUural E%p*rims>tt
Station, Rutoers CoUege
W. A. RILEY
Entomologist and Citief, Diyidon^ of Ento-
mology ftnd Economic Zoology, AgriaU-
turat Ezperimenl Station of the Uhrvertity
of Minnesota
All correspondence regarding articles from tJie Department of Agriculttire shotdd be
addressed to Karl F. Kellerman, Journal of Agricultitral Research, Washington, D. C.
All correspondence regarding articles from State Experiment Stations should be
addressed to H. P. Armsby, Institute of Animal Nutrition, State College, Pa.
JOMALOFACRICIIMAIRESEARCH
Vol. XVII Washington, D. C, April 15, 1919 No. i
MEAT EXTRACTS, THEIR COMPOSITION AND IDENTI-
FICATION
By James A. Emery, Senior Biochemist, and Robert R. HenlEy, Biochemist, Bio-
chemic Division, Bureau of Animal Industry, United States Department of Agriculture
INTRODUCTION
The historical aspect of meat extract has been presented so extensively
in the numerous articles which from time to time have appeared in the
literature that it is not considered necessary in this paper more than to
refer to that phase of the question. As is well known, this product, now
so generally used, owes its origin to I^iebig, the chemist whose process
for its preparation, as modified by Pettenkofer, has been in use in one
of the large commercial houses ever since 1864.
tn the method of preparation as originally described, muscle tissue
alone was used for extraction, but in more recent years various influential
factors, the foremost being the utiHzation of waste products, have caused
many of the manufacturers to adapt the principles of the original process
to the preparation of extracts from edible portions of the carcass other
than true muscle tissue. Livers, spleens, hearts, cured-meat cook water,*
roast-beef soak water, and bones to which more or less meat is adherent,
are among the materials now employed, and the food analyst of to-day
is confronted with many difficulties in his attempts to establish the
identity of an extract under examination.
This investigation, therefore, was undertaken with the view of obtain-
ing information regarding possible differences in composition of the
various extracts that might be applied in formulating methods for their
identification.
PREPARATION OF EXTRACTS
COMMERCIAL METHOD
Extracts of the various tissues and organs, such as chuck and plate
(representing true muscle tissue), cured meat, bones (with and without
adherent meat), hearts, livers, spleens, etc., were prepared, under the
direct supervision of one of the authors, in the meat-extract department
of one of the large commercial estabhshments. The method of prepara-
tion in each instance was that ordinarily used in the establishment, and
• Extracts were also prepared from the pickle in which the meats were cured, but the use of this material
has been discontinued.
Journal of Agricidtural Research. Vol. XVII. No. i
Washington, D. C. Apr. is, 1919
rq Key No. A-47
(l)
2 Journal of Agricultural Research voi.xvii. No. r
to all intents and purposes was practically the same in its essential
features as that in use in the general commercial preparation of these
articles. For the purpose of clarification "roast-beef soak water,"
" defibrinated blood," and "blood water," were added during the process
of manufacture in all cases with the exception of the extracts prepared
from cured meat. The comparatively large quantities employed of
these agents necessarily influenced the composition of certain of the
extracts, particularly those prepared from livers and spleens, and extracts
of the various organs and tissues, therefore, were prepared in the labo-
ratory, the method followed being nearly identical with the commercial
process. Practically the only exception was the replacement of the
materials commercially used in clarifying the extracts with those of a like
composition, equally efficient, but derived from the specific tissue or
organ under investigation. A detailed description of the laboratory
process follows.
LABORATORY METHOD
The finely minced material from which the extract was prepared was
placed in a large tin-lined box and iced water added until the minced
meat was well covered. The box with its contents was then placed in
the refrigerator where it was allowed to remain overnight, when the
resulting "soak water" was drawn off and reserved for clarifying pur-
poses. The partially extracted minced meat was then transferred to a
large open kettle provided with a perforated steam coil, an equal weight
of water added, and steam slowly applied, the temperature being grad-
ually raised to 95° to 97° C, and the liquid kept in constant agitation
by the entrance of the steam from the perforated pipe.
This extraction was continued for 45 minutes, after which the liquid
was drawn off, cooled, and transferred to an evaporating kettle provided
with a closed-coil steam pipe. The "soak water" obtained as above
was then added, the whole brought to a boil, and the evaporation con-
tinued until the liquid was reduced to two-thirds of its original volume,
the coagulable proteids which form a scum upon the surface of the liquid
being removed from time to time. After this concentration the liquid
was filtered and transferred to a vacuum kettle where it was evaporated
under reduced pressure until the extract was of the desired consistence.
This method yielded extracts identical in physical appearance and organo-
leptic properties with those obtained by the commercial process.
As it was also considered desirable to obtain data regarding possible
differences in extracts prepared from cold and hot water extractions,
the process described above was modified in the case of chuck and plate
extracts prepared in the laboratory. Chuck and plate extract 29 was
prepared by repeatedly exhausting the minced meat with large quan-
tities of cold water and then concentrating the extract. Chuck and
plate extract 30 was prepared by placing the minced meat in an equal
Apr. IS. 1919 Meat Extracts, their Composition and Identification 3
quantity of cold water, bringing the whole rapidly to a temperature of
95° to 97° C, where it was kept for 45 minutes, after which the liquor
was drawn off and reduced by evaporation to the desired concentration.
It may be noted here that the two laboratory-prepared bone extracts,
Nos. 27 and 28, were made by long-continued boiling of bones from
which all meat had been removed.
List of extracts prepared
Commercially.
In the laboratory-.
No.
10.
Beef spleens.
No. 21. Beef spleens.
No.
II.
Hog spleens.
No. 22. Beef spleens.
No.
12.
Roast-beef soak water.
No. 23. Hog liver.
No.
13-
Hog livers.
No. 24. Beef spleens.
No.
14.
Bare beef bones.
No. 25. Hog liver.
No.
15-
Regular bones.
No. 26. Beef hearts.
No.
16.
Beef livers.
No. 27. Bones.
No.
17-
Pickle.
No. 28. Bones.
No.
18.
Beef hearts.
No. 29. Chuck and plate
No.
19.
Chuck and plate.
No. 30. Chuck and plate
No.
20.
Corned-beef cook liquor.
QUANTITATIVE INVESTIGATION OF EXTRACTS
METHODS USED
In the analysis of the foregoing extracts the methods used were essen-
tially those described by Street {Sy and, in brief, were as follows:
A 10 per cent solution of solid extract or a 20 per cent solution of
liquid extract was used for the following determinations:
1. Water. — The water representing the degree of concentration of
the extract was determined by placing 20 cc. of the solution in a 100
cc. glass-stoppered weighing bottle containing 20 gm. of asbestos, and
drying to constant weight in a vacuum of 30 inches at a temperature of
60° to 65° C.
2. Ash. — Ten cc. of the solution in a tared porcelain dish ^ were
evaporated to dryness upon the steam bath, thoroughly carbonized at
a low red heat, macerated with water, filtered, and the residue thoroughly
washed and ignited. The filtrate was then added to the ignited residue
in the dish, the whole evaporated to dryness upon the steam bath,
ignited at a low red heat, and weighed.
3. Sodium chlorid.^ — After weighing, the ash obtained was dissolved
in water with the aid of a few drops of nitric acid, diluted to 100 cc,
an aliquot taken, and chlorin determined by the Volhard method.*
1 Reference is made by number (italic) to " Literature cited," p. 17.
' Porcelain was used instead of platinum in order that the possibility of volatilization of chlorin would
be reduced to a minimum, as the ash was later utihzed in the chlorin determination.
'Chlorin may be determined separately according to the method adopted by, the Association of Official
Agricultural Chemists. (2).
* Only a small portion of the chlorin of the ash of meat extracts is due to sodium chlorid, the greater
ixirtion being combined as chlorid of potassium (8). Allen (i) makes an allowance of 0.06 per cent sodium
chlorid for evcrj' unit per cent of dry matter present, considering the excess as added salt.
4 Journal of Agricultural Research voi. xvii, no. t
4. Total phosphoric acid. — Five cc. of the solution were digested
with 15 cc. each of sulphuric and nitric acids until colorless (nitric acid
was added from time to time when necessary), 20 cc. of water were
added, and the solution boiled in order to expel any oxids of nitrogen.
It was then diluted with water, a slight excess of ammonium hydroxid
added, after which it was rendered slightly acid with nitric acid, and
phosphorus determined (2).
5. Inorganic phosphoric acid. — Ten cc. of tlie solution were
diluted with from 20 to 30 cc. of water, boiled three minutes, two drops
of acetic acid added, the boiling continued for a minute, cooled, and
diluted to 100 cc. The solution was then filtered, a 50 cc. portion
was made faintly alkaline with ammonium hydroxid, and the phosphoric
acid precipitated in the usual manner with magnesia mixture. After
standing for two hours or longer the precipitate was filtered off, washed
with water containing 2.5 per cent of ammonia, and dissolved in dilute
nitric acid. The phosphoric acid was then determined as in total phos-
phoric acid.
6. Total nitrogen. — Nitrogen was determined by the Gunning
method, using 10 cc. of the solution.
7. Soluble nitrogen. — A portion of about 15 cc. of the solution
was centrifuged until clear, the clear liquid poured off, and the nitrogen
determined in a 10 cc. portion.
8. CoagulablE nitrogen. — Fifty cc. of the solution in a glass evapor-
ating dish to which 50 cc. of water were added were evaporated on the
steam bath to one-half volume; 0.5 cc. of a 10 per cent solution of acetic
acid was added, heating was continued for 15 minutes, the coagulable
albumen was filtered, washed, and nitrogen determined in the residue
on the filter.
9. Ammonia nitrogen. — The ammonia nitrogen in these extracts
was determined by the magnesium-oxid method, but the more recent
and exact Folin method (6) is recommended.
10. Nitrogen precipitated by zinc sulphate. — Twenty-five cc.
of the original solution were placed in a 50 cc. graduated flask, i cc.
of a 50 per cent sulphuric-acid solution was added, with zinc sulphate
enough to saturate the solution, after which the flask was filled to the
mark with a saturated solution of zinc sulphate. After 18 hours it was
filtered and the nitrogen determined by the Gunning method in 20 cc.
of the filtrate, corresponding to 10 cc. of the original. The total nitrogen
of the extract, less the sum of the coagulable, insoluble, and zinc-sulphate-
filtrate nitrogen represents the nitrogen of the zinc-sulphate precipitate.
A control determination of the nitrogen of the precipitate was also
made.
11. Nitrogen precipitated by tannic- acid-salt solution. —
Twenty cc. of the original solution were placed in a loo-cc. graduated
Apr. IS, 1919 Meat Extracts, their Composition and Identification 5
flask, 50 cc. of a saturated sodium-chlorid solution were added, and the
flask filled to the mark with a 24 per cent solution of tannic acid. After
a thorough mixing it was placed in the ice box and allowed to stand over-
night; any loss in volume due to contraction was corrected by the
addition of the tannic-acid solution. On the following day it was filtered,
the solution being kept in the ice box during filtration, and 50 cc. of the
filtrate, corresponding to 10 cc. of the original, were transferred to a
Kjeldahl flask and evaporated to dryness on the steam bath with the
aid of a current of air. The nitrogen in the dried residue was determined
by the Gunning method and control determinations made on the re-
agents used.
Nitrogen in the tannic-acid-salt precipitate was obtained by subtracting
the sum of the tannic-acid-salt filtrate and the coagulable and insoluble
nitrogen from the total nitrogen.
12. "Meat-base" nitrogen. — This was obtained by subtracting
the sum of the coagulable, insoluble, ammonia, and tannic-acid-salt
precipitate nitrogen from the total nitrogen.
13. Nitrogen due to peptone-uke bodies. — This was found by
deducting the proteose nitrogen obtained by precipitation with zinc
sulphate from the total quantity of nitrogen precipitated by the tannic-
acid-salt reagent,
14. NonnitrogEnous organic matter. — This was determined by
difference. From the ash-free total solids was deducted the sum of
the products of the " meat-base " nitrogen X 3.12 and the nonmeat-
base nitrogen X 6.25.
15. Purins (j). — Three gm. of the sample were dissolved in 500 cc.
of a I per cent solution of sulphuric acid and heated for four hours in an
open dish on the steam bath. (At the end of this time about 75 cc. should
remain.) It was then neutralized with caustic soda, with litmus paper
as an indicator, transferred to a beaker, and 15 cc. of a 15 per cent solu-
tion of sodium bisulphite and 15 to 20 cc. of a 15 per cent solution of
copper-sulphate solution were added. This was allowed to stand over-
night, filtered, w^ashed with dilute copper-sulphate solution, and the
precipitate then washed with hot water from the paper into the original
beaker. The contents of the beaker were brought to the boiling point
and sodium sulphid added to precipitate all of the copper. It was then
placed upon the steam bath for several minutes, made acid with acetic
acid, and allowed to settle thoroughly, after which the precipitate was
filtered off, washed with hot water, 10 cc. of 10 per cent hydrochloric
acid added to the filtrate washings, and the solution evaporated to dry-
ness on the steam bath. Ten cc. more of 10 per cent hydrochloric acid
were added and digestion was continued until the bases in the residue
were dissolved. It was then filtered, washed, the filtrate made alkaline
with 25 cc. of concentrated ammonium hydroxid, 10 cc. of a 3 per cent
6 Journal of Agricultural Research voi. xvii, no. r
ammoniacal silver-nitrate solution added, allowed to stand overnight,
filtered on the following morning, the residue on the paper washed until
all traces of ammonia were removed, and its nitrogen content determined.
16. CrEatinin, — ^The method of Folin as modified by Kmmett and
Grindley (5) was used. An aliquot free from coagulable and insoluble
nitrogen and containing from 7 to 1 5 mgm. of creatinin was placed in a
500 cc. flask, 15 cc. of picric acid and 10 cc. of a 10 per cent solution of
sodium hydrate added, allowed to stand for five minutes, being agitated
several times in the interim, and then diluted to 500 cc. After mixing,
a portion of the solution was poured into one tube of a Duboscq color-
imeter and compared with NI2 potassium-bichromate solution contained
in the other tube, the scale of which was set at 8.0.
Creatinin was calculated by the following formula :
f 8.1 ^ Volume"! ^, _ milligrams of creatinin in the
i Reading 500""/ ''^ ^° ~ ^ aliquot taken.
17. CrEatin. — To 5 cc. of the extract in a 50 cc. graduated flask,
10 cc. of Nil hydrochloric acid and 5 cc. of water were added, and the
solution heated in an autoclave at 135° C. for 30 minutes. It was then
cooled, 10 cc. of Nji sodium hydroxid added and the solution made to
volume with water. An aliquot was taken and creatinin determined as
above, with 30 cc. of 1.2 per cent picric acid and 10 cc. of a 10 per cent
solution of sodium hydroxid as suggested by Emmett and Grindley (5),
the result so obtained representing the total creatinin — creatinin due
to creatin and to preformed creatinin. The difference between the total
creatinin and the preformed creatinin multiplied by 1.16 represents the
creatin.
1 8. Nitrates. — To a few drops on a porcelain spot plate of a reagent
containing o.i to 0.2 gm. of diphenylamin {4) in 100 cc. concentrated
sulphuric acid were added a few drops of the extract solution. In the
presence of nitrates a blue color developed. They were then quantita-
tively estimated by the Schlossing-Wagner method (9).
DISCUSSION OF quantitative RESULTS
The results of the quantitative chemical examination of the extracts
are presented in Table I and, calculated to a water-free basis, in Table II.
In Table III differences in the forms of nitrogen are shown. The per-
centages of creatin and creatinin appear in Table IV together with the
ratio between total nitrogen and the sum of the creatin and creatinin.
The percentages in this table are also calculated on a water-free basis.
Apr. 15, 1919 Meat Extracts, their Composition and Identification
^
5
<
g
«
d
w
8
<
S
U
w
S
>
s
0
43
ij
s
^
•mnijEaio pamiopjj;
•p3uiin
- J 3 :> 3 p n fi
•(00 naiaj
satpoq ajjti-auojda J
a^-Bjidpajd iiBS-orancx
DiB^idpajd a^Bqdjns oinz
vmonnnv
aiqBrnS-EO^
•snunj
-aa p3niJoj3j<£
a,5j6-«-««-H««H
•nriBoJO
^ HcO^n'Oi-^MfOr^'NH.NO
IB^OX
Si's -0 005
UTD
-IBqrmdionaqj
■snnnii
IB^o; o^ orneajom jo 01 ^b^
•pixoj
-uad stuoqdsbqd otireSioni
■pixoiirad snjoqdsoqd ib^ox
oCgo-o X
spqos iB^oj) spuoiqo iwn^Bf^
X Buojqa) puopjo uinipog
•uuoiqo
'i3))cni 3ino3jo snon33anmuo>i
•i9XV3va oin-B3io
•qsy
spqos lE^oX
uajCj^
^SOOOOOOO-VOOJO
^^OOCT.OO-OtHrov/-)g"ro
. ■ M M fo 0*00 ^ O O 1/1 r- Tf
^'■co^coo-^oocS--
^^ ci H 0'-''0 O :hcOO0 moo
^ j»0 O r-ooo rO'Tj'O li^MCO
V *IcO lOfOOoO »OM uiM r-o
^ g o\ o WTO a> 10 o -^00 wi M
^■OMMMroOOOt*l-*<N
^^ fOOoo 0\M M r-r-c* ^r^
ci,§«M ' 'hmm ■*;«►;
~~I~ nO O"^0t^f00\r^0 O O
^^6
. O00i/^0<NOU^OOO
^^00000000000
a.'-d
V TfOOOMOvOfOOvOMOO
5-^"00o>-i"Ooooo
0,^6
. --O-O -Tmio-o 00 O OvO
5j g^ mosoo OiJi'tO •-« 00
D^ ^ t^t^t^-^r^j^ too t^oo «-•
OOi-tp^Nr-wr-^OTOO*
■ rt lOMCO 'tvnOVfn^r^o
/^c^r*M O\On00m tJ-o Coo
^gi-.iNo>0'rot>o-<}-o
h^O>-'OCiOOroi-.*t'^0
r\ St *^^^c>o OiCTiO ovoooo r*
"^"6
5jgiHsOCO OOu-irofO-O m 'J-
■ O r-O t~*(-iTfvofi OvO r^
^ _j M iH ^\0 »o M w 10 -^ -1 00
. .^ O po 0\0 moo M t^ M M CO
5—000 000 O ONC>fO t^CO Ot
5^'^f^oow ci o M <v)so mr-
. • r^oo mO toO-O M r-.(«-jr^
*j j*0Ol'*M 0\»0'^coo win p»
^ -O tOOif^O »O00 u-iO <^»^,
Q, ^ 't 6 t^sd t^ d\ a M w <> «
w -^ 2^ nS'S "^ moo O O M Q fT"
. -^o;^ ^co in<N*o 000 000
^ g <*i O O r^ 100 *^ w% ^ « r*
ft, ^ 6 po *o 6 <>o6 M t^ ^^ f^ ,^1
5 j*r»ioroO 0*000 Ovr-M o
G<§o6o6m rijf^iMOOOO 6 A
'•^COOOOOOOOOOQOO r^QOOO
.-^0»O000Q000000
^^CI-tl^OMOMMfOOOO
ft^ 5i '^ ^ "^ ^"^ *^°° w M ^ (A.
8 8 '■3 2 «- .Q
•38
2^
00
dfldfladaaaa
0000000000
00000
(7,>0 Oi "-<00 O
OOOOOOWM
6
?j
:?
?
?
0
CT.
:^
0
-I
;?
"
«
"
"
"
<^
w
■^
0 ^o
0
IT
r^
r^
g
-J-
0
"
"
"
"
"
«
'^
C(
•^
■^
Z
(^
li?
0
r^
S
K
GO -^00 '*^ O CO TtoO
o o rr f*^ omx o
t^OO 10
JO « 00 0 « t- vo
6
?8^
0 t^ 0 0 O* -n »/^
■^ 0 0 ■* 0 0 0
6 ••►.■■•
2i:8
8?2^8 : :
6 .
z:^:^
00 -^ M -rr t^ sO
ro n 0 »oo 1^0
• r-o O »n r^ ro Too O*
<^ 00 -fi M c^ t?»o oco
O t^ t^ M »/l00 o
^ ^a" fO r^ 10
M o o> o o
I fOCO M CO CO
T^T't'^^CI CO
f*^M MM u^■^o^M^^ co
O 00 viO CO O
-•O Tj-sO O <
f< t^O 000 Tf OnO
« u^mw ovw oco
di-d M d wifnoooo
0>0 O O O fO r^
3 00 ^ moQ 00
CO O* tooo O M m r* po
t- O r- O ro 0\ rO OO
'jod di d d» d» »^ 3 M
pi nl O
r»o r»oo o o
Journal of Agricultural Research
Vol. XVII, No. X
s>
I K
■a ■M
o ^
•2 J
5 o
'?^
■UpE3J3
•inin5B3J3
-jaiapun
snunj
■inin?B3i3
•(30U3
-•«3J}!P
•satpoq
aJUI-anoidaj;
•a^Fjidp
,00 »^ O. M 00 0» ^00 vO
w •.* 00 r^oo O '^ »noo « w
ft! S
t^MOO O ^OoOOO o »o»o
■-too f^oo >o r^ O O
. »; t» w-o o
r^ « so >o ^o ■* r^
•aiHjidpajd
a loildins-oinz
•ajqBjnSBog
•aiqnjosui
I'nox
-JOUl JO opB^
•pixo'jiiad stuoqd
-SOqd 3IUB3JOBI
■pixo-jusd siu
-oqdsoqd l^^ox
fJg'XnrJoiqo)
puojqa nmipog
•nuoiqo
■isxyeva
ovweSio snoiuSoj)iiino>i
'J3))«ni DjtreSjo
■qsv
^
.•*i '^O OvMOO 0\w O Os t^ {
< u 6
;*ooooooooooo
^6
V ^ OO O Oi »0
O O O »H n O
■ Os O* O « CO »*>
a>CT»oo'0 otOi-o t^ 0\ o o^
g O "H I
St t- r- .
§t O -«*■ »ft -^ o
O #D *- ** ^O
O OvO 0*00 00 «^
TO O O" -^ o
> O ■* r- O^O
(J ^>0 t^w^O'O'O wso r*(
U 00 t^ 'O O
o o o a>o o
^ O r^ O ♦-< O
I ^00 *0 "O « « «
(^ fO p ^O 1^ <
' Oi rf ^ i-« O O* 0» ^00
t>- tr"G r*«0^ t^vC^O t>-r*
OsOO OV)OO0Q0«OO
8«
3
V b <u S «3 . S ■ M 9 «3
r>^ r*oO O* O
OOOOO-OOO'd-'*
»0 rO O* M -^ ■* <
^0 i-l 1-1 f- «
O M M <0 IH M
r^oo o M 00
rj- r- M CO
) 00 *^ O TfvO O -^
T) ^ O OiO ^ r* 00
« ti M hi ' * M ■>*
00 «^ f*l « ^ fOOO
O O w T}- o
00 r- «*■ r>. « r-o "0 0*^*
6>6o6 oiooooooo d>)-'
I Oi*o ^ r* M - 00
I f*^«o O r»O0
■d- Tf i>^ O, r^ r- )
1 W t^ f^ w CO
•* f*) 0» ■^00 C* •-< CO 00 ^
• Oi ©i Oi « -O 0\ <
'f o. 't o>
■^•o fo 6 b « 00
^0 f*lM M MOO QOOOOO
V)0 "^OOO 0> <^ O'^O to
»^ ♦*i'0 6> (>o6 00 ^n 6 00
a a > a > 2
Hgo Jlotigg
mnKnXnncq
ft) V
ao,
•o-o
a a
d CO
MM
u u
3 3
Apr. 15. 1919 Meat Extracts, their Composition and Identification
Table III. — Distribution of nitrogen in meat extracts
Extract No.
Method of
preparation.
Commercial .
Laboratory. .
....do
Commercial .
....do
....do
....do
19. Chuck and plate
29. Chuck and plate
30. Chuck and plate
12. Roast-beef soak water .
20. Corn-beef cook liquor .
14. Beef bones
1 5. Beef bones
17. Pickle i do
18. Beef hearts
26. Beef hearts
10. Beef spleens
11. Hog spleens
21. Hog spleens
22. Hog spleens
24. Hog spleens
13. Hog liver
16. Beef liver
23. Hog liver
25. Hog liver
Averages:
Chuck and plate, bones, liquors
Hearts
Spleens
Livers
....do
Laboratory. . .
Commercial . .
....do
Laboratory. ..
....do
....do
Commercial . .
....do
Laboratory. . .
....do
Non-
nitrog-
enous
matter.
Per
cent.
24. 20
28. 02
14. 70
19.70
28.21
19.17
21. 62
29.30
29.74
32-57
24.07
25-23
24. 79
23- 51
24. II
44.96
40. 62
30-54
48.79
22. 23
31-65
24-34
41. 22
Total
nitrogen
Per
cent.
10. 08
69
67
99
23
47
59
60
02
77
77
Total nitrogen in —
Zinc
sulphate
precip-
itate.
Per
cent.
17-75
21-57
10.34
17. 21
13-58
15. 00
II. 31
17.17
II. 17
30.07
26- 53
23-74
22. 36
21. 14
18.66
32- 23
24. 19
9-35
15.90
14. 17
24.76
Tannjc-
salt
precii)-
itate.
Per
cent.
44- 13
27.86
" 87
OS
04
96
49
71
99
76
81
45
10
63
54
33
29
37
30
63
33
10
32
"Meat
base."
Per
cent.
50.98
63-56
43-74
49-94
50- 13
43-90
50. 01
57.20
53-62
55-97
41. 68
41.68
30.87
40.37
36.91
43- 16
41. 72
31.06
41. 10
50-32
54-79
38-30
39. 26
In consulting these tables it will be noted that the percentage quan-
tities of certain constituents show marked and characteristic differences,
depending upon the nature of the extract. The most striking variations
are the figures representing total nitrogen, "meat-base" nitrogen,
creatinin, and nonnitrogenous organic matter. Differences in the
amounts of the other constituents, with the exception of the ratio of
total phosphorus to inorganic phosphorus, are not considered sufficiently
marked to justify their being used, and attention is directed to the
following results:
I. Total nitrogen. — This was found to be very low in liver extracts,
as compared with other extracts. The percentage of total nitrogen in
one of the liver extracts (No. 23) is much higher than that of the re-
maining three, but is, nevertheless, lower than that of any of the other
extracts with the exception of the pickle extract. Chuck and plate
extracts contain the largest quantity of nitrogen, with spleen extracts
next. The other extracts vary between 9 and 10 per cent total nitrogen
with the exception of the pickle extract, which is very low (7.60 per
cent).
108121°-19 2
lO
Journal of Agricultural Research
Vol. XVII, No. I
Table IV. — Distribution of creatin and creatinin {water-free basis)
Extract No
Method of
preparation.
Total
nitrogen.
Creatin.
Crea-
tinin.
Per ct.
Perct.
Perct.
lO. 08
2. 00
5- 69
9.69
.41
7-32
II. 67
3-43
2.97
8.99
.89
6.94
9-23
1.36
4. 02
9-47
1.38
6.18
9-59
1-59
6.60
7. 60
.28
3-48
9. 02
I. 26
4- 63
8.77
I. 60
6.64
9.98
.88
1.70
9-38
•73
1.50
9.98
. 01
•23
10.77
. 01
•23
9. 07
•03
•31
6. 00
.61
1-59
6. 42
.89
1-54
8. 14
.04
•39
6. 52
.04
.24
9.81
8.89
9.68
9.94
6. 21
7-33
Total
creatin
plus
crea-
tinin.
Ratio of
creatin
plus
crea-
tinin
to ni-
trogen.
19. Chuck and plate Commercial .
29. Chuck and plate Laboratory. . . .
30. Chuck and plate do.
12. Roast-beef soak water Commercial .
Corn-beef cook liquor do
Beef bones 1 do
20.
14.
15-
17-
18.
26.
Beef bones do
Pickle do
Beef hearts do
Beef hearts Laboratory. .
10. Beef spleens Commercial .
11. Hog spleens do
21. Hog spleens Laboratory. ,
22 . Hog spleens do
Hog spleens do
Hog liver Commercial .
Beef liver do
Hog liver Laboratory.
Hog liver I do
Averages'.
Chuck and plate, bones, liquors
Hearts
Commercial spleens
Laboratory spleens
Commercial livers
Laboratory livers
Per ct.
7.69
7-73
6. 40
7-83
3-70
5-89
8.24
2.58
2. 23
.24
.24
•34
2. 20
2-43
•43
.28
7-25
7. 06
2. 40
•27
2.31
Perct.
763
797
463
870
583
798
854
493
653
939
258
237
024
022
037
366
368
053
043
732
796
848
031
367
048
2. NoNNiTROGENOus ORGANIC MATTER. — As will be noted in Table III,
liver extracts, as compared with the other extracts, are extraordinarily
high in nonnitrogenous organic matter, containing, with the exception
of extract 23, more than 40 per cent. Heart extracts (containing 30 per
cent) more nearly resemble liver extracts.
3. "Meat-base" nitrogen. — The determination of total nitrogen and
of "meat-base" nitrogen enables one with a fair degree of certainty to
dififerentiate liver extracts and spleen extracts from each other and from
other extracts as well. While liver extracts and spleen extracts differ
from other extracts in showing a low percentage of "meat -base" nitrogen,
they differ from each other in that the spleen extracts show a high total
nitrogen, whereas the total nitrogen of liver extracts is low.
In liver extracts the "meat-base" nitrogen constitutes only about 40
per cent of the total nitrogen, while in other extracts, with the exception
of spleen extracts, the figure is nearer 50 per cent. Particular attention
is called to liver extract 23, which contains 8.14 per cent total nitrogen
and pickle extract 17, containing 7.60 per cent total nitrogen. Although
the pickle extract contains much less total nitrogen than the liver extract,
a much larger proportion of this nitrogen (57.23 per cent) is "meat-base"
Apr. 15, 1919 Aleat Extracts, their Composition and Identification 1 1
nitrogen. In the liver extract the "meat-base " nitrogen constitutes only
31.08 per cent of the total nitrogen.
Chuck and plate extract 30 is much lower in "meat-base" nitrogen
than the other chuck and plate extracts. This is exceptional and is
undoubtedly due to the laboratory process used in its preparation, the
extraction having been made entirely with hot water. (See p. 2.)
4. Proteose nitrogen (zinc-sulphate precipitate, Table III). — Al-
though the quantity of the proteose nitrogen varies from 9.35 to 32.23
per cent of the total nitrogen, the amounts in any one kind of extract are
not sufficiently constant to render the figure of any value in the identifica-
tion of extracts. On the whole, however, liver and spleen extracts are
somewhat higher in that constituent than other extracts. This factor is
probably influenced more by the precess used in the preparation of the
extract than by the material from which the extract is made.
5. Creatin and crEatinin. — It is in the total creatinin content of the
various extracts that the greatest and most uniform differences occur.
The sum of the quantities of creatin and creatinin, together with the
ratio between this total and the total nitrogen of the extracts, is shown
in Table IV. It appears from these results that a determination of the
total creatinin will suffice in any case to classify an extract, if pure, as a
liver or spleen extract, on the one hand, or as a true meat extract on
the other. ^
It will be noticed at once that the liver and spleen extracts prepared
under commercial conditions contain about lo times as much total crea-
tinin as the laboratory extracts, attributable to the creatinin of the roast-
beef soak water, defibrinated blood, and blood water used in clarifying
these extracts. However, even though these commercially prepared
liver and spleen extracts are relatively high in creatinin, they are, never-
theless, much lower than any of the other extracts. The greatest quan-
tity of creatinin found in any of theliver and spleen extracts is 2.58 per
cent and the highest total creatinin — total nitrogen ratio^-0.37, while the
smallest amount of creatinin in the other extracts (except .the pickle
extract) is 5.38 per cent and the lowest ratio 0.46. From these results
it appears that all extracts of fresh flesh, with the exception of extracts
of liver and spleen, contain more than 5 per cent of total creatinin.
6. Phosphorus. — Rather marked differences occur in the amounts of
phosphorus found in the extracts as well as in the relation existing
between the inorganic and total phosphorus present.
The extracts of pickle and of corned-beef cook liquor contain very little
phosphorus, about 2 per cent; none of the other extracts contain less than
5 per cent with the exception of spleen extract 10. The laboratory liver
extracts are noticeably high in phosphorus.
' Extracts from bones as made commercially will show a relatively high total creatinin. This is dvie
to the fact that the commercial bone extracts are essentially meat extracts, mobt of the extractives in
them being derived from the adherent meat and the clarifying agents which are used, and not from the
bones themselves.
12 Journal of Agricultural Research voi. xvii, ko. i
Diflferences which are highly characteristic are found in the rela-
tion of total and inorganic phosphorus, the ratio of inorganic phos-
phorus pentoxid to total phosphorus pentoxid being much lower for
liver than for other extracts, the next higher being that of spleens.
Grouping these ratios, livers have a ratio lower than 0.65; spleens a
ratio near 0.7; hearts, chuck, and plate, and corned-beef cook liquor
run above 0.75 and nearer 0.8; and the ratio in the remaining extracts
is 0.90 or higher.
SUMMARY OF QUANTITATIVE DIFFERENCES
Liver extracts are low both in total nitrogen and "meat-base" nitro-
gen; have a low inorganic phosphorus to total phosphorus ratio, are very
low in total creatinin, and as a rule are very high in nonnitrogenous
organic matter.
Spleen extracts are high in total nitrogen, low in "meat-base" nitro-
gen, very low in creatinin, and lower than other extracts, liver excepted,
in the inorganic-phosphorus to total-phosphorus ratio.
Heart extracts are low in total nitrogen as compared with chuck and
plate extracts, but much higher than liver. They contain considerable
nonnitrogenous organic matter, being next to liver extracts in this
respect. Heart extracts differ from liver and spleen extracts in total
creatinin and in "meat-base" nitrogen, the latter comprising at least
50 per cent of the total nitrogen in heart extracts.
Pickle and cured-meat extracts are readily identified by the presence
of nitrates, which are always present in such extracts. The quantity of
total phosphorus present in such extracts is very small. In other respects
cured-meat extracts are found to resemble true-meat extracts. Pickle
extracts contain rather less creatinin than true-meat extracts.
Chuck and plate extracts run high in total nitrogen, "meat-base"
nitrogen, and total creatinin and have a high inorganic-phosphorus to
total-phosphorus ratio.
The bone extracts prepared commercially ^ and the extract prepared
from roast-beef soak water resemble chuck and plate extract.
PHYSICAL CHARACTERISTICS OP^ EXTRACTS
In addition to the chemical differences which have been discussed
above marked physical characteristics of the extracts exist which in
many cases are so decided that workmen engaged in their manufacture
become very expert in identifying meat extracts solely by their physical
appearance. The properties upon which their judgment is based are
color, texture, and "shortness," an extract being termed "short"
when it quickly and easily breaks upon testing its elasticity.
1 The bone extracts prepared in the laboratory have not been discussed, as they do not in the least re-
semble commercial bone extracts, and are included in Tables I and II merely as a matter of general interest.
Apr. IS. 1919 Meat Extracts, their Composition and Identification 13
Liver extracts are very dark brown, almost black, in color, are very
gummy — that is, they are not "short," and their solution in water is
dark red, with a trace of fluorescence.
Spleen extracts are light-chocolate to light yellow-brown in color
have a smooth texture, and are very "short."
Bone extracts closely resemble spleen extracts.
Other extracts, including heart extract, are darker than spleen extract,
but not so dark as liver extract. They are usually very "short," and
their solutions are dark, but are not fluorescent.
QUALITATIVE INVESTIGATION OF EXTRACTS
In addition to the quantitative differences in extracts qualitative
differences have been noted and, based on these differences, qualitative
tests have been devised for the identification of liver and spleen extracts
either when pure or in the absence of any considerable proportions of
true-meat extracts. In mixtures in which liver or spleen extracts are
present in only small amounts the tests are not entirely dependable,
although in such instances they are as reliable as any other known method.
ACETIC-ACID TEST
A qualitative test for the identification of spleen extracts was sug-
gested by Robert M. Chapin, of the Biochemic Division, who noticed
that the addition of an excess of acetic acid to a spleen extract resulted
in the formation of an abundant precipitate. Confirmation of this
observation was found in the literature, Hammarsten (7) stating that
spleens are characterized by a peculiar protein which is soluble in boiling
water, but which is precipitated by an excess of acetic acid.
Acting upon this information, the writers tested all the commercially
prepared spleen extracts (the laboratory-prepared extracts having been
exhausted in the quantitative investigation) with acetic acid in the
manner described below.
About 30 cc. of a 10 per cent solution of the extract under examination
are boiled, filtered, the filtrate cooled, and an equal quantity of a 10 per
cent solution of acetic acid added.
Extract No. Effect of acetic acid.
10. Beef spleens A dense, white precipitate.
11. Hog spleens A dense, white precipitate.
12. Roast-beef soak water No effect.
13. Hog livers A slight, dark precipitate.
14. Bare beef bones No effect.
15. Regular bones No effect
16. Beef livers A slight, dark precipitate.
18. Beef hearts No effect.
19. Chuck and plate No effect.
20. Corned-beef cook liquor A very slight yellowish pre-
cipitate.
14
Journal of Agricultural Research
Vol. XVII, No. I
With spleen extracts only was a decided precipitate obtained. The
precipitate was very bulky and yellowish white in color, easily distin-
guishable from both the slight, dark precipitate yielded by the liver ex-
tracts and the slight precipitate obtained with the corned-beef cook liquor.
Mixtures of varying amounts of hog spleens (No. ii) and chuck and
plate extract (No. 19) were prepared and tested with acetic acid in the
manner shown above, in order to determine the delicacy of the reaction.
A precipitate was obtained in mixtures containing 5 per cent of spleen
extract, but the reaction was faint, and it was not until the mixed extract
contained a quantity approximating 20 per cent that a decided reaction
was obtained.
Since the above tests were made an extract has been received and
examined under the meat-inspection regulations which gave a positive
reaction with the acetic-acid test, although not resembling spleen extracts
in other respects.
An investigation by the field service of the Meat Inspection Division
disclosed that this extract was prepared from bones through long-con-
tinued extraction with boiling water. Inasmuch as bone extract
previously examined had not given the reaction, a laboratory investiga-
tion of this unusual feature was made, three bone extracts being pre-
pared by extraction with boiling water for at least three hours. Upon
testing the finished products the reaction, which heretofore had been
limited to spleen extracts, was obtained. The precipitate from both
the spleen and these bone extracts appeared to be a mucin. From the
standpoint of the food analyst the value of the test, however, is not
lessened, as the diflferentiation of bone extracts from other extracts,
including spleen, is readily made, which may be noted by comparing
the analyses given in Table V with those of other extracts previously
tabulated (see Table II).
Table V. — Analyses of bone extracts
Constituent.
Total solids
Ash
Sodium chlorid
Phosphorus pentoxid (total)
Nitrogen
Tannic-acid salt filtrate nitrogen
Zinc-sulphate filtrate nitrogen . .
Total creatinin
Acetic-acid test
Molisch test
Per cent.
57-54
3-42
.81
.29
8.54
.35
3-23
None.
Positive.
Negative.
Extract
°er cent.
62. 90
13. 80
7-58
.49
7.90
1. 46
2. 21
None.
Positive.
Negative.
Extract J.
Per cent.
63-53
14. 81
7-32
-59
10. 40
1-37
3.60
None.
Positive.
Negative.
While not applying to the real purpose of this paper, it may be of
interest to note that an extract of bone marrow failed to afford a precipi-
tate with acetic acid. '
Apr. IS, 1919 Meat Extracts, their Composition and Identification
15
MOI.ISCH TEST
By determining the total creatinin content of an extract, the water
content being known, it can be definitely classified as either a true-meat
extract or as an extract of liver or spleen, and further, an extract identified
as having been prepared from either livers or spleens may be further
classified as either a spleen or a liver extract, depending upon the reaction
in the acetic-acid test. While the identity of a liver extract may thus be
established, qualitative tests confirmatory of the conclusions arrived at
through the quantitative examination were applied. A reaction based
on the occurrence in liver extracts of comparatively large amounts of
carbohydrates was suggested and the Molisch test was employed.
Method of using Moi^isch test. — One cc. of a 10 per cent solution
of solid or of a 20 per cent solution of fluid extract was placed in a grad-
uated glass-stoppered cylinder of 25 cc. capacity, 9 cc. of concentrated
sulphuric acid were allowed to flow gently down the sides of the cylinder,
and 6 to 10 drops of a 20 per cent alcoholic solution of alphanaphthol
were then added. The stopper was inserted, and the contents of the
cylinder were thoroughly mixed. In the presence of carbohydrates a
persistent and intense reddish-purple to deep-violet color developed
immediately.
In the initial tests the color of the mixture in the cylinder was noted
one minute after shaking, and the contents of the cylinder were then
poured into 200 cc. of water, the color again being observed. It was
later found that a more satisfactory color test was obtained when the
mixture in the cylinder was allowed to stand overnight and observed
directly.
Extracts of knovv'n origin were tested with the results given in Table
VI.
Table VI. — Results of Molisch test of meat extracts of known origin
Extract.
Method of prepa-
ration.
Color on shaking."
Color in water."
Color after stand-
ing."
Roast-beef soak water. .
Beef bones
Commercial . .
do
No color
do
No color
do
No color.
Do.
Beef chuck
do
do
do
Do.
Beef spleens
do
Very faint
purple.
. do . . . .
Very faint
purple.
.do. ..
Do.
Hog spleens
do
Do.
Beef liver
do
Strong
Pronounced. .
Very faint . . .
Pronounced. .
Very faint . . .
Pronounced . .
do
Strong
Pronounced. .
Very faint. . . .
Pronounced. .
Very faint . . .
Pronounced. .
do
Strong.
Pronounced.
Do
do
Beef chuck
Laboratory . . .
Commercial . .
Laboratory . . .
do
No color.
Hog liver
Pronounced.
Beef melts (spleens). . . .
Beef liver
Very faint.
Pronounced.
Do
do
Do.
Do
do
do
. . . do . .
Do.
"In cases marked "no color" there was no trace of the characteristic purple-red color, but there was
usually a greenish-brown color.
1 6 Journal of Agricultural Research voi. xvii. No. i
Mixtures of liver extracts with chuck and plate extracts in varying
proportions were prepared for the purpose of determining the delicacy
of the reaction, and it was found that the characteristic color was recog-
nizable in an extract containing 20 per cent of liver extract; smaller
quantities gave a purple color, but it was not sufficiently distinct to be of
value.
In all cases where a positive reaction is obtained and in which the
other factors, such as creatinin, nitrogen, etc., indicate the absence of
liver extract, the sample should be examined for starch and cane sugar.
It may be stated that, aside from its value in indicating the presence
of liver extract, the Molisch test is necessary in a routine examination
for the rapid detection of carbohydrates which may have been added to
meat extracts.^ In the analyses of some hundreds of extracts a true-
meat extract has never yielded a positive reaction with this test, and
whenever a positive reaction is obtained in an extract which can be
shown to contain no liver extract it is due to added carbohydrate. In
such instances the test should be supplemented with a more complete
examination to identify the carbohydrate thus indicated.
COPPER TEST
During the course of the investigation it was also noted that the ash
of liver extracts in every instance exhibited a more or less pronounced
greenish color, which was not observed in the ash of any of the large
number of other kinds of extracts examined. As the presence of copper
in livers has been demonstrated, it having been found even in the liver
of the fetus, the presence of copper in the ash of extracts other than
those prepared from livers should be considered. As practically all the
commercial extracts prepared from other tissues and organs underwent
the same course of preparation, and no copper was thus indicated in
the ash of any of the resulting products, it woul4 seem that this test
would apply alone to the ash of extracts obtained from livers. If,
however, copper utensils are used in the preparation of an extract its
presence in limited amounts could probably be readily demonstrated by
chemical methods, but a greenish-tinted ash should always lead one to
suspect the presence of liver extract.
Procedure suggested in the identification of an extract :
1. Total solids. 9. Creatin.
2. Ash. 10. Molisch test.
3. Sodium chlorid. 11. Acetic-acid test.
4. Total phosphoric pentoxid. 12. Test for starch and sugar if a
5. Inorganic phosphoric pentoxid. positive Molisch test is given.
6. Total nitrogen. 13. Test fornitrates.
7. "Meat-base" nitrogen. 14. Test of ash for copper.
8. Preformed creatinin.
' Asan illustration of thevalueof this test, regardless of its value in detecting the presence of liver, several
extracts have been examined which conformed in every respect to pure-meat extracts with the exception.
that they gave a positive Molisch test. Upon investigation the presence of sucrose was demonstrated-
Apr. IS, I9I9 Meat Extracts, their Composition and Identification 17
The quantities of insoluble, coagulable, and ammonia nitrogen are
so small in all ordinary extracts that they are determined only in case
they are indicated in extraordinary amounts. Should an extract show
an unusually high nitrogen content a study of the various forms of
nitrogen present is essential.
After an examination of an extract as suggested its accurate classifi-
cation as a true-meat extract, as a cured-meat extract, as a compound
extract, or as an extract of liver or spleen is possible, and at the same
time the addition of foreign material, such as sugar, starch, or salt, will
be established.
LITERATURE CITED
(i) Allen, Alfred H.
1S98. COMMERCIAL ORGANIC ANALYSIS . . . ed. 2, V. 4. Lnndoil.
(2) Association of Official Agricultural Chemists.
1916. report of the committee ON editing tentative and official methods
OF analysis. 381 p. Jour. Assoc, off. Agr. Chera., v. i, no. 4, pt. 2; v.
2, no. I, pt. 2; V. 2, no. 2, pt. 2; V. 3, no. 3, pt. 2.
(3) BiGELOW, W. D.
1905. report on the separation of meat proteids. In U. S. Dept. Agr. Bur.
Chem. Bui. 90, p. 126-130.
(4) Egger, E.
1884. UEBER EIN NEU^S UNTERSCHiJiDUNGSMERKMAL REINER NATURWEINE VON
WEINEN, die UNTER ZUHILFENAHME VON WASSER VERBESSERT WORDEN
siND. In Arch. Hyg., Bd. 2, Heft. 3, p. 373-380.
(5) EmmETT, a. D., and Grindley, H. S.
1907. chemistry OF FLESH. (SIXTH PAPER.) FURTHER STUDIES ON THE APPLICA-
TION OF FOLIN'S CREATIN AND CREATININ METHOD TO MEATS AND MEAT
EXTRACTS, /n Jour. Biol. Chem., v. 3, no. 6, p. 491-516.
(6) FOLIN, Otto.
I910. NOTE ON THE DETERMINATION OF .\MMONIA IN URINE. In JoUT. Biol. Chem.,
V. 8, no. 6, p. 497-498.
(7) Hammarsten, Olof.
1904. A TE.xTBOOK OF PHYSIOLOGICAL CHEMISTRY. Translation from . . . 5th
German edition by John A. Mandel. ed. 4, 703 p., i pi. New York, London.
(8) Street, John Phillips.
1908. thirteenth report on FOOD PRODUCTS FOR 1908. MEAT EXTRACTS AND
MEAT PREPARATIONS. In Conn. Agr. Exp. Sta. Bien. Rpt. 1907/1908. p.
606-672. Bibliography, p. 664-672.
(9) Wiley, H. W., ed.
1908. OFFICIAL AND PROVISIONAL METHODS OF ANALYSIS, ASSOCIATION OF OFFICIAL
AGRICULTURAL CHEMISTS. AS COMPILED BY THE COMMITTEE ON REVISION
oi>- METHODS. U. S. Dept. Agr. Bur. Chem. Bul. 107 (rev.), 272 p., 13 fig.
Reprinted in 1912.
108121°— 19 3
QUANTITY AND COMPOSITION OF HWHS' MILK: ITS
RELATION TO THE GROWTH OF LAMBS
By Ray E. Neidig, Chemist, and E. J. Iddings, Dean arid Director, Idaho Agricultural
Experiment Station^
INTRODUCTION
During the progress of an investigation upon dififerent breeds of sheep
at the Idaho Agricultural Experiment Station observations were made
upon the rate of growth of lambs from five breeds of ewes that are com-
monly found in this section of the country. The results indicated that
lambs from some of the breeds studied made a decided gain over others
in the same period of time. Inasmuch as the sheep industry is of such
economic importance to the Nation, the rate of gro^\i:h of the lambs
assumes more than ordinary interest, and an effort is being made to
ascertain the relation of the quantity and composition of ewes' milk of
each breed to the growth of the lambs.
FACTORS ESSENTIAL IN GROWTH
The chief factors in growth are inherited capacity and a sufficient
quantity of nutritious food. The second factor only will be studied,
since without sufficient food inherited capacity for growth will be more
or less inhibited.
Growth depends upon nutritious foods, and recent investigators have
demonstrated that these foods must contain a sufficient quantity of inor-
ganic salts, certain amino acids, lipoids, fats or oils of a peculiar nature,
and vitamines. The absence of any of these substances is detrimental
to growth. It is obvious that a well-balanced food is essential, and in
milk we have the highest type of such food. The fact, however, remains
that milk from ewes of dififerent breeds has been found to vary in quan-
tity and composition, and this may account to some extent for dififer-
ences in growth.
REVIEW OF LITERATURE
As early as 1850 data were collected giving the analysis of ewes' milk.
Since that time many investigators ^ have contributed to our knowledge
of the composition of ewes' milk. Their results, however, have dealt
mainly with the high-milk-producing ewes of foreign countries, where
• The authors wish to acknowledge with thanks the careful work of the foUowing men whose assistance
made it possible to carry on this work: To Messrs. Grover V>. Tumbow, R. R. Groninger, and Ronald Wood
credit is due for the cheniical analyses; to Messrs. O. W. Johnson, C. H. Ficke, and W. H. Booth (killed in
service, France) for the careful determination on yield of milk and growth of lambs.
• K6nIG, J. CHSMIB DER MENSCHUCHEN NaHRXTOGS- UND GENUSSMITTEL. Aufl. 4, Bd. 1, p. 365-271.
Berlin, 1903.
Journal of Agricultural Research. Vol. XVII, No. i
Washington, D. C. Apr. 15, 1919
rt Key No. Idaho — 2
20 Journal of Agricultural Research voi. xvii. no. t
yield of milk and the butter-fat content was of greatest consequence.
These results are of little benefit in the solution of this problem, other
,than to show the great variation that occurs between different breeds
and within the breeds. This problem will include only such breeds as
are common to this section of the country. Among the above-mentioned
investigations, only two give figures upon breeds that will be included
in this work. Filhol and Joly * give figures upon the Southdown, and
Hucho ^ upon the Hampshire breed. These results serve only to verify
the results of other investigators, and show the variation between breeds.
Fuller and Kleinheinz,^ of the Wisconsin Station, made a study of the
yield, fat, and total solids of the milk of five breeds of sheep; the Oxford,
Southdown, Dorset, Shropshire, Merino, and the Montana grade. They
included two ewes of each breed in their study, and took the average of
the two results as the average of the breed. In determining the milk
yield, the lamb was weighed before and after sucking the mother ewe.
This was repeated at frequent interv'als during a 48-hour period, from
which the yield of milk for 24 hours was calculated. They observed
that, when the ewes were milked by hand, only about one-half the
quantity of milk was obtained as when the first method was used. The
results on the two ewes of each breed show a wide variation in milk
yield and percentage of fat between the breeds.
Ritzman,* of the New Hampshire Station, in a recent publication has
made a valuable contribution to the present knowledge of ewes' milk.
His work dealt especially with the fat content and its relation to growth
of lambs. A summary of his results on the fat content of 6 distinct
breeds and 11 crossbreeds over a considerable period of years showed a
great variation in the percentage of fat. The outstanding feature was
that not only did breeds differ in fat content of milk, but individual
ewes within the breed differed greatly. Moreover, these individual ewes
showed marked differences in fat percentage at different lactation periods. '
This fact was observed by the writers during a preliminary investigation
of ewes' milk carried on a year previous to this present investigation.
Ritzman concluded that the growth of the lamb was not dependent
upon the percentage of fat, but he was of the opinion that it depended
mainly on the quantity of milk. No actual milk yields were obtained
by him, but an estimation of the yields made by observ^ation was tabu-
lated as "high-", "good-", "fair-", and "poor-milking" ewes.
From a revievr of the literature it is evident that an accurate estima-
tion of the quantity and composition of ewes' milk is necessary in order
' FiLHOL, and Joly. aralysbs du lait de brebis APPARTB^fANT A DIFPERE^fTEs races. In Compt.
Rend. Acad. Sci. [Paris] t. 47, no. 35. p. 1013-1014. 1858.
' Hucho, Hermann, xtntersucetungen uber scHAPMacH mit bbesonderBR ERUCKSiomoCNa
DSR OSTFRIESISCHEN MiLCHSCUAFE. In Landw. Jahrb., Bd. 36, Heft 3/3, p. 496-547. 1897.
■^ Fuller, J. G., and Kleinheinz, Frank, on the daily yield and composition of milk prom ewes
OP VARIOUS BREEDS. In Wis. Agr. Exp. Sta. 31st Ann. Rpt. 1903/04, p. 48-50. 1904.
* Ritzman, E. O. B'ags' milk: its fat content and relation to the growth of lambs. In Jour.
Agr. Research, v. 8, 110. 2, p. 2';-36, i fig. 1917. Literature cited, p. 35-36.
Apr. 15, 1919 Quantity and Composition of Ewes' Milk 21
to ascertain the factors which influence growth, since analyses of the
milk of individual ewes differ widely. In the first year's work, which
was preliminary in nature, five breeds of ewes were studied, an estima-
tion of the quantity of milk given by each ewe was made every seven
days, on two ewes of each breed. Chemical analyses of the samples of
milk taken in lo-day periods after lambing were made for a period of
70 days. The gain of the lamb was recorded every seven days. The
chief objections showing up in the preliminary work were as follows:
It became evident that samples of milk for analysis and total quantities
of milk ought to be taken at the same period, or as near thereto as pos-
sible. The experiment included only two ewes in each breed, and in
some cases one might give an abnormally high or low milk yield, which
would show unfair averages in the breed. Still another factor entered
into the work. A period of 70 days proved too long, for lambs need
access to grain early in their life, and as grain was fed to them this
made any correlation of composition of milk and growth futile. All
the above difficulties were eliminated by the following procedure adopted
in this work :
PLAN OF INVESTIGATION
It vv^as realized that any work on the study of the milk of ewes must
include a number of ewes before a fair average of the milk constituents
could be obtained. However, in this work the difficulty becomes very
evident, for with a great number of ewes the work becomes so labo-
rious that the use of a great number in the experiment is prohibitive.
The aim was to choose three ewes which showed characteristics of the
average ewe of the particular breed. This was done by starting with
four ewes of each breed and continuing with the three that showed
the nearest to the normal milk yield for the breed. Six breeds of ewes
and three ewes from each breed were used in this experiment. The
period of investigation continued for 50 days. Every 10 days after
lambing the total quantity of milk was recorded, and samples of milk
were taken. The weight of the lamb was taken at birth and every
10 days thereafter, from which the gain was calculated.
METHODS USED IN OBTAINING MILK SAMPLES
In determining the total milk yield of each ewe the lamb was separated
from the mother ewe at 6 o'clock in the morning. At 7 it was allowed
to suckle the ewe. This was done in order to start all ewes on a
uniform basis. At frequent intervals during the 24-hour period, which
began after the lamb suckled the ewe at 7 o'clock, the lamb was weighed,
allowed to suckle, and reweighed, on a balance weighing accurately to
I gm. The sum of the differences in the lamb's weight before and
after suckling the ewe during the 24-hour period gave the total yield
of milk. In this manner all the milk was obtained without causing
any nervousness on the part of the ewe, and the results gave a good
22 Journal of Agrictdtural Research voi. xvu. No. i
representative total peld of milk. The milk samples for the analysis
were obtained as follows: After the 24-hour period was concluded for
the total >neld of milk the lamb was kept away from the ewe until a
sufficient quantity of milk was in the udder; then the lamb '..^as allowed
to suckle one side, while the other was milked dry. In this manner a
uniform sample was obtained without causing undue nervousness on the
part of the ewe.
CONSTITUENTS DETERMINED IN THE MILK
The samples of milk were analyzed for the following constituents:
Total nitrogen, casein, albumin, fat, lactose, specific gravity, and ash.
The ash was then analyzed for the calcium and phosphorus content.
METHODS USED
Total nitrogen. — A quantity of milk (approximately 5 gm.) was weighed accu-
rately and the nitrogen determined by the Kjeldahl method.
Casein. — Casein was precipitated by acetic acid on a weighed quantity of milk
according to the official method. The nitrogen deuermined by the Kjeldahl method
and the results multiplied by tlie factor 6.38.
Albumin. — After neutralizing the filtrate obtained after removing the casein, with
sodium hydroxid, and adding acetic acid of the proper strength and quantity, accord-
ing to the official methods,' the nitrogen was determined by the Kjeldahl method
and the result multiplied by 6.38.
Nonprotein nitrogen. — The sum of the nitrogen of the casein and albimiin was
subtracted from the total nitrogen. The result gave the nonprotein nitrogen.
Fat. — The fat was determined by the Babcock method.
Lactose. — A portion of milk (approximately 10 gm.) was weighed accurately
in a flask and 25 cc. of distilled water were added. The proteins were precipitated
with a sufficient quantity of colloidal ferric hydroxid as described b)^ Hill.- They
were then filtered off and the clear filtrate collected in a volumetric flask. The
proteins were washed ^vith distilled water until free from lactose. The combined
filtrate and washings were made up to a definite volume and the lactose determined
by the volumetric method of Benedict.* The colloidal ferric hydroxid proved to
be a very efficacious clarifier, as it is very simple to use and insures thorough clarifi-
cation and a clear solution.
Specific GRA\^TY. — Specific gravity was determined by the Westphal balance.
Ash. — The ash was made upon composite samples of the four samples of milk by
the official methods.
Calcium and phosphorus. — Calcium and phosphorus were determined from the
ash residues by the methods described by Richmond.*
DISCUSSION OF RESULTS
In Table I is found the percentage composition and total yield of
roiik of each ewe for the entire series taken every 10 days during a
period of 24 hours. In all cases the first results upon the total weight
'Association of Opfioai. Agricctltdrai, Chemists, peport op committee on' editino methods
OP ANALYSIS, p. 287-Z89. Baltimore, Md., 1916. (Jour. Assoc. Off. Agr. Chem., v. 2, no. 3, pt. 2.)
- Hni,, Reuben L. note on the use op colloidai, iron in the determination op lactose in
MTLK. /rt Jour. Biol. Chem., V. JO, no. 3, p. 175. 176. 1915.
' Benedict, Stanley R. the detection and estimation of glucose in itrine. In Jour. .\mer.
Mtd. Assoc., V. 57, no. 15, p. 1193-1194. 1911.
^ Richmond, Henry Droop, dairy chemistry, p. 8i-8a. London. iS<^.
Apr. 15, 1919
Quantity and Composition of Ewes' Milk
23
of milk were secured lo days after the birth of the lamb. This duration
of time was allowed to elapse in order to allow the milk of the ewe to
become normal. Analyses of ewes' milk, made by Weiske and Ken-
nepohP at different periods, varying from i}4 hours to several days
after the birth of the lamb, show that 10 days is ample time for the
milk flow to assume its normal composition.
The results of the table indicate, as would naturally be assumed,
that there is a decrease in the milk flow of the ewes in the 50-day period.
In only one instance was this not true; that was in the case of Cotswold
ewe, No. 753, which maintained not only a constant milk flow throughout
the experiment, but actually showed a slight increase at the end of the
50-day period.
Table I. — Quantity and composition oj ewes' milk
Breed and No. of
d
"a
a
Date of sampling.
&
"0
1
Ii
§1
3
0
1
U
0.
c3
<
i
u
"S
1
1
§
2:
a
0
<
Ash percent-
age of cal-
cium and
phosphorus
in ash.
ewes.
s
3
0
1-6
11
COT.SWOLD
( I
2
1 3
4
I s
Feb. 18
Feb. 28
Mar. 10
Mar. 20
Mar. 30
Lbs.
190
1S4
189
193
188
Gvi.
I-93I
1,980
I. 80s
I. 122
I, 176
i-SSs
I. 029
1-033
1.031
I- 03s
I- 033
P. ci.
2-44
2.47
3-60
3-S9
3-52
P.ci.
0.91
.88
.64
.78
•83
P.ct.
0. 076
-07s
.087
.062
.067
P.ct.
7. 2
8.1
8.6
7^4
5^2
p.ct.
4-79
4-83
4.00
4-93
5-12
p.ct.
0.87
P.ct.
15^ 15
p.ct,
lg.92
2Si8
1
1
|.. .
188
1.034
i-32
.81
.06s
1-i
4-73
.87
i
2
1 ^
4
S
Feb. 21
Mar. 3
Mar. 13
Mar. 23
Apr. 2
164
159
161
157
159
i,9S6
2, 141
2,IOS
I '637
1.986
1-033
1.032
1-033
1-033
I- 032
2.86
3- 01
3- 07
3- 10
3- 19
.81
.78
•45
•83
•54
.070
. 040
.081
.051
• 051
10.4
8.0
7.8
6.4
6.0
4-81
S-07
5^i6
4- 79
5-20
.77
IS- 24 28. 24
753
159
1,96s
I- 033
3-04
.68
.058
7^7
5-00
•77
15^24 t 28.24
I
3
4
5
Feb. 28
Mar. 10
Mar. 20
Mar. 30
Apr. 9
159
ISO
147
145
149
I.SSS
1.302
1,113
838
816
1-035
1.028
1-033
I- 03s
1.034
2-82
3-04
2-94
3- 04
2.98
•72
•52
.64
.8s
.88
.088
.087
.062
.067
• 059
7.8
9.2
7.6
7^8
8.2
4. 76
4-93
4-03
4.60
4. 22
-34
11-33
18.77
S097
149
1. 124.8
I- 033
2. 96
•72
. 072
8.1
4. 62
.84
"•33
18.77
I
2
3
4
5
Feb. 17
Feb. 27
Mar. 9
Mar. 19
Mar. 29
HAMPSHIKE
189
177
176
16s
164
2,477
2.487
3,328
1,845
1,328
1.029
1-034
1-032
I- 03s
I- 033
3.84
2-73
2.89
3-62
3-41
•72
.81
.48
.48
.48
.081
.072
.081
.062
-051
10.3
6.2
6.0
8.2
4-s3
4.80
4-.=;o
4.88
4-95
.76
12.32
22. 12
30
174
2,093
1.032
3- 09
•59
. 069
7-6
4- 74
.76
12.32 22.12
I
2
3
4
I S
Feb. 20
Mar. 2
Mar. 12
Mar. 22
Apr. I
172
152
151
136
139
3,439
2.273
2' 534
2,300
1,848
1.030
1,031
1.030
I. 031
I-03S
2.62
3-77
2.84
3- 36
3-07
.91
•43
•36
1-36
-52
.078
• 059
.054
. 056
.039
8-25
6.8s
62
5.0
3-9
4. 61
4-53
4-86
4-97
4-73
.81
14.22 24.65
SO
ISO
2,479-8
I. 031
3- 13
•71
.061 {
6.0 1
4- 74
.81
1 f.
1
' Weiske, H., and Kennepohi,, G. ttntersuchungbn Cber scH.\FMUca u>rr8R vsrschisdsnbm
VRRHAI.TNISSBN. /» Jour. L^ndw., Jahrg. 29, p. 431-472. 1881.
24
Jour7ial of Agricultural Research
Vol. XVII. No. I
Table I. — Quantity and composition of ewes' milk — Continued
Breed and No. of
ewes.
HAMPSHIRB— con
Average.
SOUTHDOWN
Average.
Average .
Average .
SHROPSHIRB
Average .
366346 .
Average .
Average.
UNCOUI
Average .
Feb. 23
Mar. 5
Mar. IS
Mar. 25
Apr. 4
Feb. 23
Mar. 8
Mar. IS
Mar. 2$
Apr. 4
Feb. 26
Mar. s
Mar. 18
Mar. 28
Apr. 7
Mar. 10
Mar. 20
Mar. 30
Apr. 9
Apr. 19
Mar. 12
Mar. 22
Apr. I
Apr. II
Apr. 21
Mar. IS
Mar. 2S
Apr. 4
Apr. 14
Apr. 24
Mar. 14
Mar. 24
Apr. 3
Apr. 13
Apr. 23
Feb. 21
Mar. 3
Mar. 13
Mar. 23
Apr. 2
Lbs.
189
169
176
16S
169
176
156
'B-i
" o
"o'g
>.n
— I u
3
O
H
Gm.
3.103
2.IS9
1.352
I.. SOS
I. 573
1.938-4
^Z^i
863
753
1. 146. 8
.368
•393
■317
1,470. 6
1,417
1.596
1.468
1.037
018
3,602
144
148
1,836
524
2,050. 8
■499
,701
,oS8
996
924
I, 241. 8
1.528
1,456
1,193
1, 191
1,180
1,309- 6
I. 030
I- 033
1-030
1.032
I- 033
P.ct.
2.83
3-04
3-33
1-032 2.97
I- 013
1.036
I- 027
I- 02s
I- 026
1-033
3- II
3-24
4-03
3- 18
3-60
3-25
3-72
3- 76
3-74
3-66
3-97
2-83
3-64
3-66
3-77
3-93
3-97
■36
P.ct.
I. 27
I. II
•59
.81
.82
P.ct.
.094
.070
.067
.090
.030
.064
.0.^4
.064
. os6
.078
69
3-08
3-28
3.98
2-83
2-97
. 070
. 070
.030
. 050
79 .08
P.ci.\P.cl.
4-59
6.0
6.8
6.8
7.6
4-83
4-93
4.86
11-35
8.0
.5-6
8.8
7-4
4-94
4.98
4-97
4-98
4-73
8. 2 I 4. 92
5-2
5-3
8-0
S-6
7.0
5.02
4.86
4-97
4.96
4-59
4- 88
7-6
8-0
67
.066
88
.042
89
.092
92
.09
96
.09
. 076 9- I
.06
•043
. 076
.087
•073
.067 ! 8.8
4. 82
4-73
4-83
4-97
4.07
4.48
4.88
4.89
4-73
4.12
3-82
S-IO
4.90
4-73
4.72
4, 71
4.89
4- 8a
4-73 I
4-77 I
P.ct.
0.78
.76
.76
•78
.84
.87
.87
Ash percent
age of cal-
citun and
phosphorus
in ash.
S3
a*'
p. ct.
15-19
P.ct.
27.68
15. 19 27.68
15-39 23-78
IS- 39 23- 78
14. 43 28. 68
14. 4s I 28. 68
12. 22 20. 96
12.22 20.96
21-78 I 34.73
19-04 ! 30.57
19- 04 I 30- 57
16.93 28.62
16- 95 I aS. 6j
Apr. IS, 1919
Quantity and Composition of Ewes' Milk
25
Table I. — Quantity and composition of ewes' milk — Continued
v
S
ti
d
a
i
"o
D
P
Si
>. 0.
H
>
60
a
a
■53
a
a
<
g
M
2
1
1
0
a
h4
<
Ash percent-
age of cal-
cium and
phosphorus
in ash.
ewes.
■5
5
ll
0.2
UKC01.N — contd .
I
3
3
4
I 5
Feb. 27
Mar. 9
Mar. 19
Mar. 29
Apr. 8
Z,6f.
187
179
176
176
176
Gvi.
I.95S
1 . 569
1,441
1.274
1,482
I- 033
1-033
I- 033
1-035
I- 03s
P.ct.
2.92
3-22
2.98
3-39
3-38
P.ct.
I. 12
.84
•S8
•9S
.91
P.ct.
. lOI
.062
•059
.077
.078
P.ct.
7-6
6-0
6.9
6.4
7-4
P.ct.
4-74
4-72
4.87
4. 60
4.70
P.ct.
0.67
p.ct.
IS- 20
p.ct.
27-56
179
1,544-2
1-033
3-17
.88
•07s
6.8
4-72
-67
15-20
27- 56
I
a
3
4
5
Feb. 24
Mar. 6
Mar. 16
Mar. 26
Apr. 5
193
180
182
i8s
175
1-574
I. 313
735
481
S06
1-032
1-032
1.029
1.030
1-032
2.60
2. 67
3-66
3-25
3-35
I. 19
1.26
•45
.40
•39
. 100
.0S4
• 037
. on
.032
9.6
7.2
II. 4
8.2
8.2
4-75
4-85
4-93
5-12
4-3°
.80
14-73
29-13
183
921.8
I- 031
3.10
•73
-053
8.9
4- 79
.80
14-73
29- 13
I
2
■ 3
4
5
Feb. 12
Feb. 22
Mar. 4
Mar. 14
Mar. 24
ItAMBOT7II,LET
136
131
130
130
131
I. 918
I. 391
1.347
I. 131
I. 112
1-033
1-037
1-032
1-034
1-033
3-63
3.62
4-75
4.22
4-58
.86
•75
• 52
1-23
.84
.062
.029
.048
• 037
.070
ro. 05
II. 9
9.6
9.6
8.2
4.82
4-78
5-00
4. 60
4-63
.91
18.49
29.38
«6
132
I. 379- 8
I- 034
4. 16
.84
• 059
9.8
4-77
-91
18. 49
39. ?8
I
3
1 ^
4
I S
Feb. 26
Mar. 8
Mar. 18
Mar. 28
Apr. 7
is6
151
149
147
144
2,582
2, 113
1,766
1,706
1,758
1.039
1.032
1. 040
1.034
1-033
3-42
3-05
3-40
3-17
3-59
.98
•95
.48
•74
•51
. 110
. 019
.064
.062
.029
7-4
6.8
6.4
5-9
7.0
5.00
4-97
5- 08
5-07
4-56
:8o
16.64
31.08
I
149
1,98s
I- 03s
3-32
•73
.056
6-7
4-93
.80
16. 64
31-08
( I
a
■ 3
4
I S
Feb. 28
Mar. 10
Mar. 20
Mar. 30
Apr. 9
169
167
160
164
i6s
1,525
1.262
1, 140
880
802
1-042
1-030
1.034
I- 037
1.034
3-64
3-45
3-21
2. 96
3-45
-94
1-27
•72
•58
.86
.067
• 054
.091
.057
.062
3-4
9-6
6-4
8-0
8.2
4.71
4-76
4-79
4-75
4- 00
.86
19.64
32-47
i6s
1,121.8
1-035
3-34
.87
.066
7-1
4. 60
.86
19.64
A study of Table I brings out the fact that there is a great variability
in the percentages of the constituents of ewes' milk. Not only is this tru e
among the different breeds, but also during the lactation period of the indi-
vidual. The most constant constituent in the milk of all breeds exam-
ined appears to be lactose, while fat seems to be the most variable. The
difference in the percentages of fat is very marked, not only between the
breeds, but during the lactation period of the individual. These observa-
tions are in harmony with the results secured by Ritzman,^ who also
found that the fat varied at different lactation periods of individual ewes
and who concluded that —
No great reliance can be placed on single tests of an individual, and that a test must
either cover a larger number of periods during one lactation of an individual or that
I RiTZMAN, E. G., 1917. OP. CIT., p. 31.
26
Journal of Agricultural Research voi. xvn.No. r
it must cover an average of a large number of individuals at one period, in order to be
representative.
When the average percentages of fat for the five lactation periods of
each ewe are determined and compared, the variation of fat content is
not so marked, which indicates clearly the value of a number of tests
rather than one single test on an individual.
Table II. — Average qtianiity and composition of milk for each ewe and for each breed
Average
total
quantity
of milk
for 24-
hour
periods.
Specific
gravity.
Composition of milk.
Breed and No. of ewe.
Caseiu.
AI-
bum in.
Non-
protein.
Fat.
Lactose.
Ash.
COTSWOLD
2518
Gm.
1.585
1,96s
1, 124. 8
1.034
1-032
1-033
Per cent.
3-32
3-04
2.9t>
Per cent.
o.8i
.68
-72
Per cent.
0.065
.05S
.073
Per cent.
7-3
7-7
8.1
Per cent.
4-73
5- 00
4. 62
Per cent.
0.87
7J5
•77
.84
Average
i,5S8
1-033
3-10
- 74
.065
7- 7
4- 78
.82
HAMPSHIRE
2,479
1,938.4
2.093
I- 031
1.032
1.032
3- 13
2.97
3-09
• 71
.92
•59
.061
.080
.069
6.0
7.6
7-6
4-74
4. 72
4-74
.81
•78
.76
Average
2,170
1-032
3-06
- 74
.070
7-1
4-73
-78
UNCOLN
1,309-6
921. 8
1,544-2
1-023
1-033
1.025
2.97
3- 17
3.10
-72
.S8
•73
.067
.077
•053
8.8
6.8
8.9
4-77
4-72
4-79
.82
1996
.67
.80
Average
1.258
1.027
3-08
-77
.06s
8.1
4.76
.76
RAMBOUILLET
36
Ij379-8
1,985
1,121.8
1.039
I- 03s
I- 035
4. 16
3-32
3-34
.84
-73
.87
.039
. 056
.066
9-8
6-7
7-1
4-77
4-93
4. 60
.91
.So
.86
Average
1-495
1.036
3-60
.81
.078
7-8
4-77
•8s
SOUTHDOWN
89
1,146.8
1,470.6
I, 100. 4
1-031
1.034
I- 03s
3-43
3-26
3-66
.86
.69
.82
. 070
.060
.068
8-2
6.2
8.0
4.92
4.88
4-31
.76
128
-78
Average
1,238
1-033
3-45
•79
.066
7-5
4-70
.91
SHROPSHIRE
1,307. 2
2,030, 8
I, 241. S
I- 035
I.OJ4
1.032
3-57
3- 72
3- 12
•56
- 79
.86
.06
.08
.07
8.1
7.2
9.1
4-48
4.48
4- 57
.84
.87
Average
1,532
1033
3-47
•77
■ 07
8.1
4- SO
.88
Tables II and III are given for convenience of comparison of the aver-
age yield and the average analysis of the milk for the 50-day period.
Table II gives the averages for the three individual ewes of each breed,
and the average of these averages is represented in Table III as the
average for the breed.
Apr. 15, 1919
Qtiantity and Composition of Ewes' Milk
27
Table III. — Average quantity and composition of milk for each breed
Average
toUl
quantity
of milk
for 24-
hour
periods.
Specific
gravity.
Composition of milk.
Name of breed.
Casein.
Al-
bumin.
Non-
protein.
Fat.
Lactose.
Ash.
Gm.
Hampshire 3,170
Cotswold 1,558
Shropshire 1,533
Rambouillet 1,495
Lincoln i, 258
Southdown i, 238
I.ojr
I- 033
I- 033
1.036
1.027
'•033
Per cent.
3.06
3-10
3-47
3^6o
3^o8
3-45
Per cent.
0.74
•74
•77
.81
•77
.79
Per cent.
0.070
.065
.007
.078
.065
.066
Per cent.
7- I
7- 7
8. I
7-8
8. I
7-5
Per cent.
4- 73
4-45
4-5°
4- 77
4.76
4- 70
Per cent.
0. 78
.81
.88
•8s
.76
•91
Table III brings out clearly the differences in milk yields for the
different breeds. The Hampshire ewes in this experiment easily ranked
first in quantity of milk produced, while the differences in the other
five breeds were not so great.
Table IV. — Initial weight of lambs and their gain during each lo-day period
Breed and No. of ewe.
Num-
ber of
lambs.
Initial weight of
lambs.
Amount of weight gained by
lambs each lo-day period.
First.
Second.
First.
Second.
Total.
SHROPSHIRE
7c;2
I
Gm.
4,294
Gm.
Gm.
2,392
2,468
2, 902
2,032
1,957
Gm..
Gm.
2,392
2,468
2, 902
2,032
1,957
Total gain 50 days
",751
11,751
366346
3,955
4, 407
2,535
2, no
1,372
1,644
1,522
2,518
2,304
1,452
1.637
1,374
5,053
4,414
2,824
3,281
2,896
Total gain 50 days
18, 468
40
2
3,503
3,277
1,682
1,139
1-395
1,360
751
1,888
1,085
1,023
977
761
3,570
2, 224
2,418
2,337
1,5^1
Total gain co days
12, 060
I.INCOLN-
1040
I
5.311
2,507
2, 269
2, 296
2, 089
2, III
2,507
2, 269
2, 296
2,089
2, III
Total gain i;o days
II, 272
II, 272
28
Journal of Agricultural Research
Vol. XVIX. No. I
Table IV. — Initial weight of lambs and their gain during each lo-day period — Contd.
Breed and No. of ewe.
Num-
ber of
lajnbs.
Initial weight of
lambs.
Amount of weight gained by
lambs each lo-day period.
First.
Second.
First.
Second.
Total.
LINCOLN — continued
lOI^
I
Gm.
5,650
Gm.
Gm.
3, 0C3
2, 754
2, 256
2, 286
1,782
Gm.
Gm.
3,003
2,754
2,256
2,286
1,782
Total gain 50 days
12, 081
12, 081
1006
I
5,424
2,463
2,078
"844
696
702
2,463
2,078
a 844
696
702
Total gain 50 days
6,783
6,783
RAMBOUILLET
^6
I
5,085
2,189
3,114
2,117
2, 395
1,681
2, 189
3, "4
2,117
2,395
i,68i
Total gain 50 days
II, 496
II, 496
CQ
2
4, 181
4, 181
2,526
249
I, 140
770
I, 612
2,230
1,969
I, 722
1,681
I, 423
4, 756
2,218
2,862
2,451
3,03s
Total gain 50 days
16, 092
74
2
3,616
3," 390
1,427
542
992
755
I, 012
937
598
780
795
1,227
2,364
I, 140
1,772
1,550
2,239
Total gain 50 days
0, 065
COTSWOLD
21:18
2
4,633
4,520
2,431
1,527
I, 305
I, 260
I, 222
2,318
1,587
1,566
1,363
I, 152
4, 749
3, 114
2,871
2,623
2,374
Total gain 50 days
...
15, 731
1
71:^
I
4,068
3,696
3,549
2,976
2,995
2, 191
3, 696
3,549
2,976
2,995
2, 191
Total gain 50 days
IS, 407
15, 407
a Lamb sick, did not thrive.
Apr. 15, 1919
Quantity and Compositioyi of Ewes' Milk
29
Table IV. — Initial weight of lambs and their gain during each lo-day period — Contd.
Breed and Xd. of ewe. 1
1
Num-
ber of
lapibs.
Initial weight of
lajnbs.
Amount of weight gained by
lambs each lo-day period.
First.
Second.
First.
Second.
Gm.
Total.
2097
COTSWOLD — continued
I
Gm.
5,650
Gm.
Gm.
2,945
1,861
1,561
1,646
1,796
Gm.
2,945
1
1
1,861
1,561
I, 646
1,796
9,809
9,809
HAMPSHIRE
1
2
5.424
3.503
1,932
2,919
1,937
2,308
1,934
1,679
2,479
1,149
1,527
822
3. 611
5.398
3,086
3.83s
2.756
18, 686
2
4.407
4,859
3, 735
2,213
1,598
2, 061
I, 416
3-338
2, 202
2.073
2.275
2,315
7,073
i"-"' • •
4,415
3,681
4,336
3.731
23, 236
I
4,294
6,349
3.814
2,769
2,83s
2,317
6,349
Total srain 50 davs
3,814
2,769
2,835
2,317
18, 084
18, 084
SOUTHDOWN
80
I
4,294
2,764
2,752
2,239
1,864
982
2,764
Total gain 50 days
2, 752
2,239
1,864
982
10, 601
10, 601
128.
I
4,068
2,995
2,443
2,592
1,950
1,789
2, 995
Total gain t;o days
2,443
2,592
1,950
1,789
11,769
II, 769
3
3,616
3.503
1,503
983
81S
886
1,247
I, 555
1,280
1,207
1,524
1,483
3.058
2,263
2,022
3,410
3,730
i«i 483
1
30
Journal of Agricultural Research voi. xvii, no. i
In Table IV data are given on the initial weight of the lamb or lambs
and the gain in weight every lo days during the period of the experiment.
The total gain is also included.
Table V is a combination of the results on total inilk yield and the
total constituents of the milk, expressed in grams, calculated from
the average percentages secured on the 50-day period, and also data
on the total gain in weight of the lambs.
Table VI gives the averages of the above constituents for each breed.
TablB V. — Relation of Milk Constituents of Individtial Ewes to Growth of Lambs
Breed and No. of
ewe.
Total
quantity
of milk.
Total
casein
Total
albu-
min.
Total
non-
pro-
teids.
Total
fat.
Total
lac-
tose
Total
ash.
Weight at
birth.
Num-
ber of
Total
Lamb
No. I.
Lamb
No. 2.
lambs growth,
to ewe. j
HAMPSHIKE
Gm.
104, 650
123.950
96, 920
Gm.
3,237
3,897
2,878
Gm.
617
880
891
Gm.
72
76
77
Gm.
7,953
7,437
7,375
Gm.
4,960
S. 974
4-574
Gm.
795
1,004
756
Gm.
5-424
4,407
4.294
Gm.
3,50s
4,859
1
Gm.
"■2 21,764
2 23,236
33
I 18,084
Average
108, 506. 6
3,331
796
75
7,585
5,126
852
21,028
COTSWOLD
98, 250
79,250
56, 240
2,987
2,631
1,664
668
642
405
57
52
40
7. 565
5,785
4-555
4,912
3.748
2,588
7S6
689
472
4.068
4.633
5.650
4,520
I
15-407
I
17,731
3097
9.809
Average
77.913-3
2,427
572
SO
5,968
3,749
639
14,318
RAMBOUnXET
36
68,990
99,250
56,090
2,870
3.295
1,873
579
70s
487
41
S6
37
6,761
6,650
3.982
3,280
4.913
2,680
627
794
482
5.08s
4,181
3-616
3.842
4,181
3.390
2
13,078
IS. 322
74
9.06s
Average .;....
75-110
2,677
590
45
5,797
3,624
634
12,488
UNCOLN
65,450
77,210
46,090
1,944
2,393
1,461
471
564
406
44
41
35
5.650
6,871
3.135
3-122
3,698
2, 176
537
618
309
5.311
5-424
5,650
I 11,272
I 12,081
1996
I 6, 783
Average
63,250
1,933
480
40
5.219
2,999
388
10, 04s
SOUTHDOWN
89
57,340
73,530
55,020
1,967
2,397
2,013
493
507
451
40
51
37
4,702
5-559
4.401
2,821
3,SS8
2.371
435
573
654
4.294
4.068
3.616
3,503
I
I
a
10, 6or
128
11,769
307
12,483
Average
61,963-3
2,136
484
43
4,554
2,926
554
II. 618
,
SHROPSHIRE
65,360
102,540
62,090
2,333
3.814
1,937
366
810
534
39
82
5-294
7.382
2.928
4.594
2,837
621
4-294
i
I II. 751
366346
861 3-955 4-407
540 3-503 3.277
3 18.468
49
43
5.650
2 12,060
Average
1
570
1 f.
3.452
67s
1 j
1 14,093
1
' '
n Ewe had triplets, one was taken away on ninth day.
b Lamb had leg broken on ninth day and was removed.
In a comparison of the total quantity of milk constituents and the
total growth of the lambs there is one disturbing factor. In all breeds,
with the exception of the Lincoln, twins were born to one or more ewes
in each breed, and in one case triplets. The three Lincoln ewes all
gave birth to single lambs. It is obvious that in comparison of quantity
Apr. 15, 19.9 Quantity and Composition 0} Ewes' Milk
31
of milk and growth of lambs the best experimental results in this investi-
gation would have been obtained if all ewes were allowed to raise only
one lamb. In future work it is hoped that this condition may be ful-
filled. However, many factors prevented such an arrangement. At
the time of this investigation, which is an outgrowth of a more extended
investigation on sheep, it was desired to make the work correspond as
closely as possible to the actual conditions found in sheep husbandry,
and other data were collected besides those included in this paper.
Table VI. — Relation of Milk Constituents of Breeds to Growth of Lambs
Breed.
Hampshire.
CotswoH ...
Shropshire. .
Rajnbouiilet
Lincobi
Southdo-wn .
Total
quantity
of milk.
Total
casein
Gm.
108, 506. 6
77.913-3
76,636.3
75.110
63,250
61,963.3
Gm.
3'33I
2,427
2.69s
2.677
1-933
2. 136
Total
albu-
min.
Gm.
796
572
570
590
480
484
Total
non-
pro-
teids.
Gm.
Total
fat.
Gm.
1, 585
S-968
6, 107
5.797
5,219
4-554
Total
lactose.
Gm.
5, 126
3,749
3.452
3.624
2.999
2,926
Total
ash.
Gm.
852
639
674
634
388
554
Num-
ber of
lambs
to
ewes.
Total
growth.
Gm,.
21,028
14,318
14,093
12.488
io,04S
11,618
It is quite evident that twin lambs, given a sufficient quantity of milk,
will make a greater total gain than a single lamb, provided their initial
weights correspond and they are equally strong at birth. A certain
amount of milk is essential for the growiih of a lamb, but on the other
hand there is a limit to the amount of milk that an animal can assimilate.
Therefore, two lambs, given a sufficient quantity of milk, will have an
advantage in total gain over a single lamb. The single lamb, however,
is generally larger than either of the twin lambs at birth, but from an
economical standpoint it is obvious that twins are more desirable in the
flock than singles.
A compilation of the data on the Hampshire breed shows the single
lamb of ewe No. 33 gained nearly as much in the same period of time
as the twin lambs of ewe No, 30. A comparison of the total yield of
milk shows ewe No. 33 produced slightly less than ewe No. 30.
In the Cotswold breed we have ewe No. 753 giving more than either
of the other two ewes, and the single lamb has made a gain almost equal
to the gain of the twin lambs of ewe No. 2518. Ewe No. 2518 has twin
lambs, and their total gain is only slightly greater than the single lamb
of ewe No. 753. The third ewe. No. 2097, shows a smaller milk yield
than ewe No. 2518 with the twin lambs, and the gain of her single lamb
is a little more than one-half as much as the total gain of the twin lambs.
In the Rambouillet and Southdown breeds we find the total gain in
weight of the lambs is proportional to the amount of milk consumed.
In the Lincoln breed, the only breed where there are three single lambs,
their gain in weight is also proportional to the quantity of milk consumed.
However, the lamb belonging to the Lincoln ewe, No. 1996, became sick
at the end of 20 days and did not thrive thereafter.
32 Journal of Agricultural Research voi. xvii. no. i
It appears from this experiment that the greatest factor in growth is
quantity of milk ; hence, a high-milk-producing ewe is more valuable than
a low one. The inherited capacity for growth, however, must not be
overlooked. As to the relative merits of the breeds, it is not the purposeof
this investigation to enter upon a discussion. To draw conclusions
upon such a small number of ewes in each breed would be unfair. It
was the aim of this experiment to make the investigation as fair as possi-
ble to all breeds studied, and the authors desire to emphasize clearly
the fact that results upon the different breeds are given wholly as an
attempt to correlate milk yields, their composition, and their relation to
growth. The results are not given with an idea of comparing the desira-
bility or undesirability of the breeds included in this experiment, but
rather for the purpose of presenting to the farmer and student information
in regard to features of certain well-known breeds that have to do with
utility and adaptation to certain specific purposes. For example, the
man interested in the growiih of lambs for early marketing would be inter-
ested in a breed that by its yield of milk, and possibly certain other fac-
tors, made the greatest average growth of lambs. Another purpose of
the experiment is to stimulate the interest of investigators and students of
animal breeding in the field for the improvement of certain breeds with
reference to factors having to do with profit for the grower. There
might even be room for a new breed that would possess all the desirable
and highly useful factors of some of the breeds included in this experiment.
SEED DISINFECTION BY FORMALDEHYDE VAPOR
[PRELIMINARY REPORT]
By Cecil C. Thomas
Pathological Inspector, Federal Horticultural Board, United States Department of
Agriculture
INTRODUCTION
The continual introduction of plants by the Department of Agriculture,
chiefly by means of seeds, from all parts of the world, and the constant
danger of allowing little-known or serious diseases to enter thereon,
emphasize the necessity for a study of the methods of seed disinfection.
There are few data on this subject except in the case of cereals and for
a small number of seeds used in physiological experiments.
The pathological inspectors of the Federal Horticultural Board have
encountered many difficulties in treating hundreds of lots of seeds of
\videly varying types and quantities with the various liquid treatments
in common use. Most of the treatments recommended and used at the
present time require dipping or soaking in a water solution of some
fungicide or germicide. The seeds, therefore, remain wet for a longer
or shorter period, depending on the treatment given and the method of
drying.
Some seeds like wheat and rye absorb water slowly and can be dried
without much injury, while seeds like the various members of the mustard
family absorb water very readily and with even a very brief treatment
swell sufficiently to break the seed coat and allow the cotyledons to fall
apart, thus destroying the seed.
Light seeds such as are found in many of the grasses present another
problem for the wet treatment. It is very difficult to give them anything
like a uniform treatment because of the difficulty of wetting them and
keeping them under the liquid. Seeds such as flax, which have a muci-
laginous covering, present still another difficulty for wet treatments.
The large number of shipments and the great variety of seeds passing
through the quarantine inspection house of the Federal Horticultural
Board, United States Department of Agriculture, that need to be treated
render desirable the adoption of a method of treatment which will
ob\nate wetting and drying. An attempt, therefore, is being made to
develop a treatment of this type with formaldehyde vapor. While it is
far from being perfected, it seems desirable to make a preliminary report
on some of the results obtained.
Journal of Agricultural Research, Vol. XVII, No. t
Washmgton, D. C. Apr. is. 1919
rr Key No. O— 3
(33)
34
Journal of Agricultural Research
Vol. XVII. No. I
The pathological inspectors of the Federal Horticultural Board have
treated several hundred lots of seeds each year in the liquid treatments
commonly recommended and have found the formaldehyde solutions
the best for the greatest number of cases. Formaldehyde is also known
to be a very efficient germicide when used in the form of a vapor as a
disinfectant for contagious human diseases.
For the above reasons formaldehyde vapor has been selected for this
work.
APPARATUS
A galvanized iron can (fig. i, A) having a capacity of approximately
130,000 cc. was used. After introducing the seeds and organisms.
Fig. I— Formaldehyde-vapor disinfecting apparatus
Steam was added through a rubber tube (B) from an autoclave (C).
The formaldehyde solution (Shoemaker and Busch, U. S. P. VIII, 40
per cent by volume of formic aldehyde) was diluted one to one with
water to give a greater volume of liquid. This formaldehyde solution
was introduced through an atomizer (D) with the aid of compressed air
(F), the nozzle of the atomizer being inserted in the rubber tube (B),
through which the steam passes as it enters the can. This insertion
was made as close to the can as possible. Steam was first introduced
and then the compressed-air tube was attached to the atomizer and the
formaldehyde solution was forced in while the steam was still entering.
April 15. I9I9 • Seed Disinfection by Formaldehyde Vapoi
35
The finely atomized formaldehyde solution thus enters the can and is
carried to all parts of it with the steam. Condensation takes place on
the surface of the seeds, forming a thin film of moisttire about each
seed in which the formaldehyde may act, and as this film evaporates the
gas is freed.
A frame (E) containing three wire shelves was used inside the can,
and the seeds were placed in porcelain dishes ®n these shelves.
EXPERIMENTS
There are two distinct phases of this problem: (i) The effect on the
seeds and (2), the effect on the fungi and bacteria.
Table I gives the result of a series of treatments of a number of differ-
ent seeds. The formaldehyde solution, before dilution, was used at the
rate of 10 ounces per 1,000 cubic feet to procure the results given in
the second, third, and fourth columns and for the time indicated. The
results given in the sixth column are for formaldehyde used at the rate
of 30 ounces per i ,000 cubic feet for 2 hours.
The germination percentages given in this table are an average of the
results obtained by the Seed lyaboratories, Bureau of Plant Industry,
United States Department of Agriculture, from germinating two samples
of 100 seeds in each case.
Table I.
-Effect of formaldehyde vapor of different strengths for varying lengths of time
on the gcrminatioti of seed
Seed.
Alfalfa (C. I. 44)'
Barley (C. I. 25)'
Beet
Carrot, Oxheart
Clover, Crimson
Com, Miner's Yellow Dent. . .
Field Pea
Flax
Lettuce
Millet
Muskmelon, Rocky Ford. . . .
Natal Grass
Oats(C. I. 541-4)*
Orchard Grass
Radish, Icicle
Rice(C. I. 1561)'
Rye (C. I. 138)'
Soy Bean
Sudan Grass
Wheat, Blue Stem (C. I. 1912-
"V
Checks.
95-5
92-5
71-5
69- 5
70-5
96-5
93-5
94-5
97-5
94- S
87
I- 5
98
64- 5
97
93-5
83-5
95-5
84
57-5
Strength formalin.
10 ounces
per 1,000 cubic feet.
I hour.
3 hours.
3 hours.
Per cent.
Per cent.
Per cent.
92
93-5
93- 5
94-5
93-5
94
58-5
64
66
72.5
66.5
83-5
69
75- 5
65
95-5
96-5
97
94-5
89-5
93
92
93-5
93
97
98
99
92-5
93
92-5
82.5
88
95- S
3-5
3
4-5
98
99
97
63.5
72-5
73-5
98.5
97
98
94
96
92
84. 5
89-5
87
96
96-5
98
90- S
87-5
84
61.5
59-5
55-5
Checks.
96
90.
66
72.
69
97
93-
86
99
51-5
95
76.
96
93-
83
99
80
62
30 ounces
per 1,000
cubic feet
(3 hours).
Per cent.
89-5
92.5
68.5
79
64
85
99
88.
93-
o
98
61.
96.
95
86
97-
84.
61. S
• These are accession numbers of the OfiSce of Cereal Investigations, Bureau of Plant Industry, U. S.
Department of .\griculture, from which some of the seeds were obtained.
36 Journal of Agricultural Research voi. xvii. No. i
Apparently there was little or no injury in any case. As shown by
the percentage of germination in the checks the killing of all seeds in
the case of Natal grass where 30 ounces per thousand cubic feet for 2
hours was used probably was due to the low vitality of the seeds.
A number of experiments have been conducted with fungi and bac-
teria in which they were treated with various amounts of the formaldehyde
vapor and for different lengths of time. Five different organisms, Monilia
fructigena, Colleiotrichum gloeosporioides, Fusarium vasinfectum, Asco-
chyta sp., and Bacillus caratovorus were used in the following experiments.
The spores were exposed to the treatment in four different ways:
(i) Three drops from a cloudy water suspension were placed on the bot-
tom of a sterile petri dish with a sterile platinum loop and dried before
treating. (2) Three drops of the suspension were placed in a dish as
above and the dish was placed in the treating chamber before the drops
had dried. (3) The drops from the suspension were placed on sterile
cover glasses and these were then placed in sterile petri dishes and
treated. After treatment the cover glasses were removed to another
petri dish in order to avoid a chance of getting any great amount of
formaldehyde into the culture medium when the plates were poured.
(4) Masses of dry spores were used. The masses of spores were placed
on cover glasses by smearing with a platinum loop containing an abund-
ance of spores taken from the surface of a pure culture. These cover
glasses were then handled as described under the third method.
The checks were made in the same way as the plates used in the first
method except that they were not treated in any way. After the treat-
ments all plates were poured, using potato agar. They were kept under
observation for from a week to 10 days.
The use of different amounts of formaldehyde solution and changes in
duration of the treatment show that under the conditions described 10
ounces of standard formaldehyde solution per i ,000 cubic feet for i hour
will kill the organisms used when they are exposed in a thin film. When
a mass of spores is used, more time is necessary to kill them.
The masses of organisms or spores in the case of Bacillus caratovorus
and Monilia were killed when formaldehyde was used at the rate of 10
ounces per 1,000 cubic feet for 2 hours; Ascochyta spores in mass were
killed when formaldehyde was used at the rate of 20 ounces per 1,000
cubic feet for i hour; but 20 ounces per 1,000 cubic feet for 2 hours was
necessary to kill the masses of spores of Colletotrichum. Fusarium
proved to be the most resistant, and a test was made using the spores of
four different species of Fusarium in masses. Formaldehyde was used
at the rate of 30 ounces per 1,000 cubic feet for 2 hours, and in all cases
growth occurred.
Following the above experiments some work was undertaken to deter-
mine the effect of formaldehyde vapor on the fungous spores and bacteria
borne on the surface of seeds. Five seeds of each of the various kinds
April IS. X9I9 5"^^^ Disinfection by Formaldehyde Vapor 37
used above were placed in a series of sterile petri dishes, two sets of each
kind of seeds being used. One set of dishes was held untreated as a
check and the other set was given the vapor, using formaldehyde solution
at the rate of 20 ounces per 1,000 cubic feet for 2 hours. The plates
were all poured in the usual way and observations were made for several
days. This experiment was repeated three times, and very promising
results were secured. Alfalfa, carrot, clover, field pea, flax, lettuce,
millet, muskmelon, radish, and soy beans were free from fungi when
treated, but fungi were present in abundance in all the checks, except
flax.
In practically all cases, whether treated or untreated, a few bacteria
developed on the plates, but the treated plates showed very few colonies,
while the checks showed a great many.
In the case of barley, oats, corn, rye, rice, and wheat no growth
appeared for two or three days in the treated plates, while the untreated
plates had an abundance of growth within a day or two. This difference
may be due to an inhibiting effect on the part of the vapor, but it seems
more probable that it is due to the fact that the surface spores and
mycelia were killed in the case of the treated seeds and that the appear-
ance of fungi a day or two later is due to the growth of mycelium from
within. The fungi appearing in such cases were species of Fusarium and
Altemaria.
It seemed desirable to determine in so far as possible what fungi are
present on the seeds passing through the inspection house and at the
same time get some additional information as to what effect the vapor
treatment would have on these organisms under actual working condi-
tions. A chance was also afforded to study the effect of the vapor treat-
ment, in a very limited way, on germination.
In this work five seeds of the material to be tested were removed
before treatment and five after treatment. These seeds were placed in
sterile petri dishes and treated in the usual way. The plants were under
observation for several days. Bean seeds were used more than any other,
inasmuch as a large number of shipments of beans happened to be
coming in from South America.
Out of the 86 different samples of beans treated and studied 9 seemed
to be retarded from one to two days in germination, while 8 were accel-
erated slightly, but in no case was there any apparent injury. The
remainder of the samples did not seem to be affected one way or the
other, so far as their germination was concerned. There was very
marked reduction in the number of fungi and bacteria present in the
treated samples as compared with the untreated. In taking samples of
this kind, average seeds were selected, and as a result some of the beans
were diseased and probably had internal mycelium, as in the case of
CoUetotrichura, thus making it impossible to render them absolutely free
from fungi without killing them.
^8 Journal of Agricultural Research voi. xvn, No. i
Sixteen different fungi were found in these samples, including Fusa-
rium, Altemaria, and Colletotrichum, species of all three of which are
known to cause serious bean diseases.
In all of the experiments set forth above, only a few seeds were included
in each sample and inasmuch as formaldehyde vapor is known to be
lacking in penetration it seemed desirable to try the treatment of a larger
quantity of seeds.
A shipment of poppy and Cryptotaenia seed afforded an opportunity
to try the treatment of a larger quantity of seeds than previously had
been attempted. The samples used were of sufficient size to cover the
bottom of the dish in which the seeds were treated to a depth of three-
fourths of an inch. After treating the seeds a sample was taken from the
surface and then the seeds were carefully removed from the surface to a
depth of about one-half inch where another sample was taken. These
samples were plated out and the samples taken one-half inch below the
surface showed fully as many colonies of fungi and bacteria as did the
untreated samples while those taken from the surface showed no fungi
and a marked reduction in bacteria. This experiment shows the lack
of penetration of formaldehyde vapor.
A comparison of the formaldehyde vapor and a 2 per cent formalin
solution was made. Ten different kinds of seeds were used and three
samples, each consisting of five seeds, of each of the different kinds of
seeds were made and placed in sterile petri dishes. The first set of each
was retained as a check; the second was treated with 2 per cent formalin
for 10 minutes and then washed with sterilized water twice, while the third
was given formaldehyde gas at the rate of 20 ounces of formalin per 1,000
cubic feet for 2 hours. After treatment all the plates, including the
checks, were poured and kept under observation for several days. In
all cases the 2 per cent formalin sample stood intermediate between the
checks and those treated with formaldehyde vapor. The vapor-treated
samples were remarkably free from fungi and bacteria. In fact only the
wheat, rice, and rye samples had any fungi present, and there was much
less growth in these than in the checks or in those treated with 2 per cent
formalin. One of the noticeable things in this experiment was that in all
the plates treated with the vapor there was a very marked reduction in
the number of bacterial colonies, as compared with the checks, while'
the samples treated with 2 per cent formalin showed little or no reduction
in the number of bacterial colonies, as compared with the checks It
should be stated in connection with these experiments that different lots
of seeds or a change in any one of the many factors concerned in all
probability would bring about a change in the results obtained.
Several hundred lots of seeds have been treated wnth 20 ounces of
formaldehyde per 1,000 cubic feet. Subsequent plating in agar has
shown that molds and other fungi rarely appear in these plates if the
April IS, I9I9 Seed Disinfection by Formaldehyde Vapor 39
seeds are sound, whereas the untreated checks seldom fail to develop
several colonies.
The work thus far necessarily has been limited to a few fungi and a
few seeds, but there is an almost unlimited field here that needs investiga-
tion if efficient and satisfactory results are to be obtained in the disinfec-
tion of seeds.
CONCLUSION
1. The use of liquids for disinfection is unsatisfactory for many kinds
of seeds.
2. A number of species of fungi and bacteria are killed when treated
for 2 hours with 20 ounces of formaldehyde vaporized under the condi-
tions described.
3. This same treatment is not injunous to any of the seeds tested.
4. The experiments completed indicate that the formaldehyde gas
treatment described is a very efficient means of seed disinfection.
ADDITIONAL COPIES
OF THI3 PUBUCATION MAY BE PROCURED FROM
THE SUPERINTENDENT OF DOCUMENTS
GOVERNMENT PRINTING OFFICE
WASHINOTON, D. C.
AT
10 CENTS PER COPY
StJBSCEIPTION PEICE,$3.00 PER YEAR
A
Vol. XVII IVTAY 15, 1919 No. 2
JOURNAL OP
AGRICUUTURAL
RESEARCH
CONTKNXS
Page
Influence of Soil Environment on the Rootrot of Tobacco 41
JAMES JOHNSON and R. E. HARTMAN
(Contribution from Wisconsin Agricultural Experiment Station)
PUBLISHED BY AUTHORITY OF THE SECRETARY OF AGRICULTURE.
WTTH THE COOPERATION OF THE ASSOCMTION OF AMERICAN
AGRICULTURAL COLLEGES AND EXPERIMENT STATIONS
^VSTASHINaXON, D. C.
WASHINQTON 3 OOVERNMCNT PRINTINO Of PJOE : tttt
EDITORIAL COMMITTEE OF THE
UNITED STATES DEPARTMENT OF AGRICULTURE AND
THE ASSOCLA.TION OF AMERICAN AGRICULTURAL
COLLEGES AND EXPERIMENT STATIONS
FOR THE DEPARTMENT
KARL F. KELLERMAN, Chairman
Physiolooist and Associate Chief, Bureau
of Plant Industry
EDWIN W. ALLEN
Chief, Office of Experiment Stalions
CHARLES L. MARLATT
Entovwlogist and Assistant Chief, Bureau
of Entomology
FOR THE ASSOCIATION
H. P. ARMSBY
Director, histitute of Ammal NutriU'n, The
PennsyhcKiz State College
J. G. LIPMAN \
Director, New Jersey AgricvlluralEiperiment
Station, Rutgers College
W. A. RILEY
Entomologist atid Chief, Division of Ento-
mology and Economic Zoology, Agricul-
tural Experiment Station of the University
of Minnesota
All correspondence regarding articles from the Department of Agriculture should be
addressed to Karl F. Kellerman/journal of Agricultural Research, Washington, D. C.
AU correspondence regarding articles from State Experiment Stations should be
addressed to H. P. Armsby, Institute of Animal Nutrition, State College, Pa.
JOMAL OF AGRlClllTDRAlESEARCH
Vol. XVII Washington, D. C, May 15, 1919 No. 2
''^>H^
INFLUENCE OF SOIL ENVIRONMENT ON THE ROOT- '^IZlT^''^
ROT OF TOBACCO '''^'
By James Johnson, Assistant Professor of Horticulture, College of Agriculture, Uni-
versity of Wisconsin, and R. E. Hartman, Agent, Office of Tobacco Investigations,
Bureau of Plant Industry, United States Department of Agriculture^
COOPERATIVE INVESTIGATIONS OF THE OFFICE OF TOBACCO INVESTIGATIONS
BUREAU OF PLANT INDUSTRY, UNITED STATES DEPARTMENT OF AGRICULTURE
AND THE WISCONSIN AGRICULTURAL EXPERIMENT STATION
INTRODUCTION
The foremost considerations in connection with the study of disease
in plants are the pathogenicity of the parasite, the susceptibility of the
host, and the environmental conditions favoring the infection and prog-
ress of the parasite. It is well known, however, that the relative patho-
genicity of the parasite and susceptibility of the host are not always
easily distinguishable one from the other in disease, and that they are
largely influenced by environmental conditions. To the practical grower
environmental conditions have been considered as all important, to the
exclusion of the parasite, while, on the other hand, the tendency in the
past on the part of pathologists and botanists has been to devote a great
deal of energy to the study of the parasite, with only passing interest
being given to the influence of the environment on disease, as recently
emphasized by Jones {16).^ This is especially true of plant diseases
having their origin or region of attack on underground portions of
plants. The literature upon actual experimental data with reference to
the influence of soil conditions upon a soil-infesting parasite is frag-
mentary, and for the most part concerned with only one or two variable
factors, so that the conclusions can not always be relied upon because
of failure to give due consideration to other factors perhaps even more
influential in the end result obtained. The Thielavia-rootrot of tobacco
{Nicotiana tabacum) forms a relationship of host and parasite appar-
ently admirably adapted for such experimental work in that it permits
quantitative determination of the influence of the disease upon the host ;
' The writers are indebted to Dr. L. R. Jones, of the Department of Plant Patholojry, Wisconsin Agri-
cultural Experiment Station, for helpful suggestions, and to Dr. W. W. Gamer, of the Office of Tobacco
Investigations, Bureau of Plant Industry, United States Department of Agriculture, for critical reading
of the manuscript.
2 Reference is made by number {italic) to "Literature cited," p. 85-86.
Journal of Agricultural Research, Vol. XVII, No. 7.
Washington, D. C. May 15, 1919.
rs Key No. Wis. —is
(41)
42 Journal of Agricultural Research voj. xvii, No. a
the fungus is readily recognizable, and both the parasite and the host
are easily manipulated under widely varying environmental conditions.
With reference to this disease alone no problem was seemingly more in
need of investigation from a practical standpoint than the great varia-
bility in the occurrence of the disease observed both in general and
local areas, and the influence of external conditions on the application
of prophylactic measures. The literature, furthermore, abounds in state-
ments intended to explain the epidemics of this disease, which are greatly
in need of modification and correction.
Accordingly, a study was undertaken with the view of covering prac-
tically all phases of the environmental conditions which might influence
the tobacco rootrot. Although it is felt that the problem is still in need
of further study, it is believed that the evidence here presented will serv^e
to show the relations of the more important factors concerned.
SYMPTOMS OF ROOTROT
The rootrot of tobacco and other plants, caused by Thielavia hasicola
(B. and Br.) Zopf, is the most serious disease with which the tobacco
growers in most producing sections have to contend. Its importance
is especially evident in Kentucky, Wisconsin, Ohio, Connecticut, and
Petmsylvania. The aboveground symptoms are much the same as those
produced by the usual unfavorable soil or weather condition which may
stunt the growth of tobacco; hence, as a rule, its effects are not recog-
nized by the growers as having a parasitic origin. Where infection is
abundant, however, the signs of the disease on the roots are sufficiently
specific to leave no doubt as to the causal organism. It is difficult, how-
ever, even for the pathologist to judge adequately the relative amount
of damage done by T. hasicola and by other causes which may reduce
yield, even when the roots are carefully removed from the soil and
washed before examination. The relative importance can be deter-
mined with considerable accuracy, nevertheless, by comparing the plants
especially the roots, which have been grown in infested and uninfested or
sterilized soil ; or by comparing both resistant and susceptible strains grown
on infested soil. In this way one may find what appears to be a compara-
tively unimportant amount of infection is in reality a controlling factor;
or, on the other hand, that a seemingly heavy infection is of compara-
tively small importance. In this way casual judgment may be replaced
by definite experimental evidence.
The effects of the rootrot may range from a complete checking of the
plants, or even death when infection occurs in the early stages of growth,
to only slight signs of reduced yield. It is indeed highly probable that
under certain conditions considerable infection may be present without
appreciably affecting the yield. Furthermore, it appears to be equally
certain that in some instances infection by T. hasicola has markedly in-
creased yields as a result of temporarily delaying growth during a period un-
PLATE 8
Soil temperature graphs for the month of August, 1915-1918, inclusive, at depths
of 2, 4, and 8 inches.
ADDITIONAL COPIES
OF TEUa PtTBUCATION MAY BE PECXItmED FROM
THE SUPEP.tNTENDENT OF DOCtrMENT3
GOVERNMENT PRINTING OFFICE
WASHINGTON, D. C.
AT
15 CENTS PER COPY
Subscription Price, $3.00 Pee Yeae
May IS. 1919 Influence of Soil Environment on Roetrot of Tobacco 43
favorable for normal plant growth, hence preventing maturity of the crop
until seasonal conditions develop more favorable for the normal growth
of the host, but at the same time unfavorable for the development of the
parasite. On the other hand, the parasite has produced through this
indirect action heavy losses due to hail or frost injuries, or has reduced
the quality of the product as a result of extending the ripening and cur-
ing process into unfavorable seasons. No above-ground symptoms of
rootrot are more common than the failure of tobacco to grow appi^ciably
during the first month or six iX^eeks after transplanting to the field, fol-
lowed by a period of relatively rapid growth and development caused by
a change of conditions which have up to the present remained more or
less obscure.
No detailed description of the disease on the roots need be given here,
as this phase of the subject has been frequently presented and will become
more or less evident in the progress of the present discussion.
REVIEW OF THE LITERATURE
Peglion (20) was the first to describe T. hasicola as a parasite of tobacco
when he reported it from Italy in 1897. There is considerable evidence,
however, that this disease had occurred upon tobacco in America many
decades prior to that time, although it was not recognized as a disease.
When Jones, according to Tatham {24), as early as 1724, wrote with
reference to agriculture in Virginia —
when land is tired of tobacco, it will bear Indian corn or English wheat or any other
European grain or feed with wonderful increase —
he made a statement which is being annually "rediscovered" by hun-
dreds of tobacco growers, but which is an established principle with
thousands of other growers. It is now quite certain that parasitism
explains the majority of the modern tobacco growers' difficulties of the
nature referred to above, and no hypothesis yet formulated will explain
with equal satisfaction the observations of the early Virginia planters.
Antedating the first report of the parasitic origin of the rootrot by 13
years, Killebrew (77) in 1884 wrote:
In some years the plants both in the seed bed and after being set out are affected
by a disease known as the " black-root. " The plants so affected do not die, but after
standing comparatively still for a long time revive later in the season, but do not make
a good quality of tobacco. It is not known what the agencies are producing this
disease, nor has there been a remedy discovered for it. By some it is believed to be
the result of sowing seed continuously in old beds. Seed beds in newly cleared
groimd are said to be entirely free from it.
There can be no doubt that this is the description of the rootrot, or
blackrot, of tobacco caused by T. hasicola. This brief description of
the disease is given in full, since it is probably not only the first authentic
report of the disease but also because it describes the common behavior
of infected plants, as follows:
after standing comparatively still for a long time revive later in the season.
44 Journal of Agricultural Research voi. xvii, no. »
This observation may be repeatedly noted in infested soils, and it was
with the idea of explaining this condition particularly that the investi-
gation on the influence of environmental conditions upon the disease was
begun in 191 4.
Sorauer (25) in 1895 made some observations upon the rootrot on
cyclamens, and concluded that heavy manuring, too abundant watering,
and too high temperatures favored the attack by the parasite.
Peglion (20), who first reported the disease on tobacco in 1897, also
believed that too much manure and water were controlling factors in
producing the disease.
Campbell (5) believes excessive quantity of humus in the soil predis-
poses the tobacco plant to disease, and also that an acid condition of the
soil weakens the plant and predisposes it to disease.
Buttaro (4), probably following the lead of other European authors,
also writes that the disease on tobacco is favored by abundant organic
matter, excessive humidity, and high temperatures.
Benincasa (/) concludes that in some years the disease appears only
slightly or not at all, and states that its development is favored by too
much organic matter, excessive v/atering, and generally damp weather.
Benincasa, at about this time, began to study the relation of moisture
and different kinds of soils to the development of the disease. He con-
cluded at this time that porous soils give the best results and in 1911
he discussed the subject in more detail.
Cappelluti-Altomare (6) concluded that the disease could be checked
by limiting the amount of watering and by not reducing too greatly the
light and air supply of the seed beds. He also advises against sowing
the seed too thickly.
Galloway (11, pp. 174-178) reporting on the wilt of violets, caused by
T. hasicola, advises against the use of decaying vegetable matter in the
propagating beds. His statement that —
plants affected may make a good growth in summer and show no evidence of trouble
until September or October, when they will wilt more or less during the day and
revive at night.
is especially pertinent to the investigations in this paper.
Clinton and Jenkins (9) suggest that excessive fertilization, soil reac-
tion, and soil moisture, the latter in particular, may be important sec-
ondary factors determining the extent of the injury by T. hasicola.
They also state that the cold, wet weather of early spring helps along
the trouble in the seed beds, particularly when they are not properly
ventilated.
Clinton {8) is quite convinced that the character of the season,
especially the moisture and possibly unusually cold wet spring weather,
and the character of the soil and subsoil — fineness, liability to become
water-soaked, drainage, amount of humus, especially in the shape of
manure — have much to do with determining whether or not the fungus
May IS. I9J9 Influence of Soil Environment on Rooirot of Tobacco 45
does much damage. No definite experimental data are given, however,
in support of these views. The following year (1908) drouth is said to
have reduced the injury due to the disease.
Briggs (j) reported, upon evidence obtained from Connecticut soils
that the fungus attacks are most severe on soils made alkaline by large
applications of lime, ashes, or fertilizers containing carbonate of potash,
and that the alkaline condition in infested soils should be corrected by
the use of acid fertilizers in order to obviate the damage by T. hasicola.
This advice was received favorably by both practical growers and sci-
entists, and many recommendations were based upon it.
Gilbert {12) concludes that an abundance of humus, a considerable
percentage of clay, high fertilization either with chemicals or manure
(expecially nitrogenous fertilizers), excessive water, and high tempera-
tures favor the disease. In an experiment to determine the influence
of the amount of watering he found that excessive water increased the
disease, although 62 per cent of the plants in the scantily watered beds
were diseased. He also compared the yield as a result of transplanting
diseased and healthy plants in the field, using a "Havana Broadleaf "
variety. He obtained as good yield from the diseased plants as from the
healthy ones.
Whetzel and Osner (27) recommended acid-phosphate fertilization for
the control of T. hasicola which causes fiber-rot on ginseng.
Benincasa (2) reporting on results obtained in comparing different
"soils" for growing tobacco plants, recommends sand or "pozzolano"
a volcanic ash for this purpose, since favorable conditions for disease
are said to be absent in these . He also states that T. hasicola is a weak
parasite under certain conditions.
Martinazzoli (18), however, reported that he obtained T. hasicola
from beds where pozzolano was used, infection probably having come from
soil.
Massee (jp) concludes that T. hasicola can not infect host plants in
pure sand, since the fungus is able to infect only in the presence of
organic matter which will permit the mycelium to exist for some time
as a saprophyte.
Chittenden (7) had difficulty in obtaining infection with T. hasicola
until overwatering of the soil was practiced.
Rosenbaum (22) believes that such external conditions as excessive
water, lack of aeration, and heavy manuring favor infection.
Reddick (21) reported unsatisfactory results for the control of
Thielavia of violets by acidifying the soil with acid phosphate as recom-
mended by Briggs. Stable manure apparently did not act deleteriously
on infested soil. The experiments were not carried far enough, however,
to be entirely conclusive.
The present writer (14), as a result of field observation, also believed
soil moisture to be the main controlling factor in determining the severity
of the disease.
46 Journal of Agricultural Research voi. xvn. No. a
GENERAL CONSIDERATION OF FACTORS CONCERNED
It is evident at the outset that any attempt at a separate analysis of
each factor concerned in disease occurrence is practically impossible.
Varying one single factor of the environment to the total exclusion of
variability in all others is an ideal to be kept in mind in experimental
work of the nature to be described. Failure to reach this ideal in prac-
tice, however, need not necessarily reduce the value of the result, pro-
vided the effect of other variables on such a result is properly considered.
As an illustration of a complication of factors of this sort, there may be
cited the maintenance of two pots of soil at two different temperatures,
say 30° and 10° C, respectively, in order to compare the effects of these
temperatures on the occurrence of T. hasicola on the roots of tobacco.
By means of proper controls in uninfested soil the influence of many
factors involved may be eliminated, but it does not seem possible to sepa-
rate clearly the factor of soil temperature from that of soil moisture.
The soil and plants at 30° will require several times as much water as will
the soil and plants at 10° because of the increased evaporation and trans-
piration at the higher temperature. The correct replacement of this
water for maintaining like moisture relations for the host and parasite is
uncertain no matter how frequently and carefully it may be done either
by weight or by the use of an auto-irrigator. If, however, the moisture
relations have previously been studied and the range of the effects to be
expected from this factor are known, it may be possible to carry on
soil-temperature studies with only moderate attention to the moisture
relations.
A study of the factors concerned in the development of the Thielavia
rootrot has served to bring out clearly the fact that all the factors con-
cerned are inseparably connected with one another, and that the amount
of disease occurring is the product of a number of plus and minus factors,
but that, nevertheless, in an analytical study of this nature, it is possible
to arrive at the relative importance of these various influences.
It is important, furthermore, that not only the true environmental
factors be taken into account, but also that such inheritable factors as the
relative degree of susceptibility of the host plants used and the virulence
of the parasite concerned receive proper consideration. In addition, the
amount of infection and the time of its' occurrence may greatly modify
the results both as regards the readiness with which the host may become
infected and the effect of a "mass action" upon the measurable end result
of disease.
The purely environmental conditions to which the roots of the host or
the parasite harbored by the soil are subjected may be conveniently con-
sidered under the following subjects: (i) Amount of infestation present;
(2) percentage of moisture; (3) temperature; (4) soil reaction; (5) physical
May IS. 1919 Influence of Soil Environment on Rootrot of Tobacco 47
structure of the soil, including the relative amount, the state of the vege-
table matter, and the size of the mineral particles; (6) available chemical
fertility; (7) state of cultivation as regards compactness and aeration.
To the soil physicist, chemist, and biologist it will appear that the
entire scope of soil science may be concerned in the production of disease
in the roots of plants, and such seems to be the case. On consulting the
branches of soil science it is at the same time both encouraging and dis-
couraging to find many of these factors influenced by a number of other
interrelated factors under normal conditions. For instance, soil tem-
perature as such can not be thought of without also considering the
influences of the air temperature, specific heat, moisture content, exposure,
and color of the soil upon such temperature. With soil moisture it
becomes essential to regard moisture-holding capacities, rainfall, drain-
age, cultivation, humidity, and temperature; or, in the case of soil fer-
tility, to consider along with the natural fertility, its cropping history,
applied fertility, and various other modifying factors.
With these things in mind, however, it has become increasingly possible
to account for, if not to explain fully, seeming contradictions and lack of
accord with established principles of infection which have come under
the writer's attention during the past five years in the case of the rootrot
of tobacco. The occurrence and economic importance of the disease in
one State and not in another, on one farm and not on the neighboring
farm, or on the hilltop in one field and in the low spots of another, as
well as the total failure of a crop in a field one year followed by a com-
plete success the following year, or the change of crop prospects from
failure to 100 per cent yield within the period of two weeks, are all more
or less subject to scientific interpretation from this viewpoint.
With respect to those factors, aside from environmental conditions
which may influence experimental results, it should be said that as far
as evidence from literature, or as far as the observation of the writer is
concerned, there is nothing to indicate that specialized races of T. basi-
cola occur, or that the fungus varies in any way in virulence owing to
differences in strain or age of cultures. It may be said with considerable
certainty, therefore, that we are dealing with a relatively constant organ-
ism as to pathogenicity. With respec+ to host differences it has been
shown (13, 14) that very decided differences in susceptibility in host
plants, and in varieties and strains of tobacco occur. By using pure
strains of seed experimental error from this source may be eliminated.
It should be remembered, however, as will be shown in the data here
presented, that because of these differences in susceptibility the critical
points in disease occurrence and severity may be shifted in one direction
or another to some extent, a fact which makes it important that the sus-
ceptibility of the variety used for experimental work be taken into
account in any interpretation of results.
48 Journal of Agricvltural Research voi. xvii. no. 3
AMOUNT OF INFESTATION
With diseases in which the injury from infection is confined to rela-
tively small local areas on the host, it is to be expected that the amount
of damage done is in some measure proportional {o the number of infected
areas which occur. Although it is evident that under the most favorable
conditions, T. hasicola may spread a considerable distance over a single
root from one source of infection, more commonly the diseased portions
are confined to areas of from i to 5 mm. In either case it is evident
that relatively few infections may cause no appreciable stunting in
growth, where a larger number of infections in proportion to the size of
the root system may produce a rapid check in growth, owing largely to
the mechanical reduction of the feeding area, and possibly also in some
degree to the formation by the fungus of toxic substances injurious to
normal metabolism. The latter possibility appears not to hold, however,
since the host is rarely killed by the parasite. In fact, the plant appears
to receive a stimulus toward the formation of new roots to replace those
lost by disease, and in many instances the youngest leaves of infected
plants possess a deeper green color than healthy plants. The new roots
are at a greater disadvantage than the original ones, if they are formed
at or near the region of disease, since frequently they are obliged to pass
through small local areas of high infestation before reaching the deeper
layers of soil.
It is a well -recognized fact that infestation increases under field con-
ditions at an enormous rate once it is present or introduced to a soil
which is being cropped to a highly susceptible host. This is illustrated
in the practice in certain tobacco-growing areas of growing only one or
two crops of tobacco and then laying the land aside for other crops for
a long time. That this practice has been empirically developed, in the
Burley section at least, as a result of the rootrot, seems unquestionable,
in view of some unpublished results obtained in that section during
the last three years.
The influence of the amount of infestation on the amount of disease
might be illustrated in a number of ways. The simplest method appeared
to be the mixing of varying quantities of infested and uninfested soil
and transplanting into this mixture a susceptible variety of tobacco.
The soil selected for this purpose was from the old tobacco field on the
Station farm at Madison, on which tobacco had been grown continu-
ously for 10 to 12 years, together with soil of the same type from a neigh-
boring plot located not more than 2 rods away, but never having been
used for growing tobacco. The fertility of the two soils naturally would
not be the same, but the soil from the tobacco field because of heavy
applications of fertilizers was the more fertile of the two from a chemi-
cal standpoint. The soils, after having been thoroughly screened, were
weighed out and mixed in the proportion shown in Table 1. Two sepa-
May IS, 1919 Infitience of Soil Environment on Rootrot of Tobacco 49
rate experiments of this nature were carried out. In a third experi-
ment, steam-sterilized tobacco-field soil was used for mixing with the
untreated tobacco soil in the same way as before. A gradual falling off
in growth will be noted with the increase of amount of infested soil
(PI. I, I). In the case of No. 4, in experiment 3, the lowered yield, as
compared with all infested soil is no doubt due to the frequently observed
fact that reinfested sterilized soil favors the growth of fungi and conse-
quently the increased development of disease in the soil. In a mixture
of three-fourths infested and one-fourth steamed soil the balance of
infestation and conditions apparently was such as to cause greater
infection than in all-infested soil.
Table I. — Influence of amount of infestation of soil by Thielavia basicola on the yield
of tobacco
Pot
No.
Soil mixture.
Average air-dry weight of duplicates.
Infested.
Umnfested.
Experi-
ment I.
Experi-
ment 2.
Gm.
Gm.
3-25
1.65
5-37
3.66
I. 40
.70
2-73
2. 64
. 01
1.32
Experi-
ments."
None
One-fourth. .
One-half
Three-fourths
All
All....
Three-foturths
One-half
One-fourth . . .
None
Gm.
6-93
4. 20
2. 65
a. SI
^■33
0 Steam-sterilized soil was used as uniafested soil in Experiment 3. The low yield in pot 4 was probably
due to steamed infested soil favoring disease more than unsteamed infested soil.
The results obtained in Table I are considered to be due only to the
fact that more spores exist per unit of soil in the higher proportion of
infested soil, and therefore the roots are more likely to come in contact
with infecting material. This fact has an important bearing upon
results obtained in a study of environmental conditions. However, since
any deleterious or beneficial action to which the parasite is subjected
may merely reduce or increase the amount of infestation, the results
will be proportional in some measure to the amount of infestation present.
INFLUENCE OF MOISTURE CONTENT OF SOIL
A review of the literature concerning factors influencing the severity
of the rootrot of tobacco shows that soil water has been considered the
primary limiting factor by practically all observers and investigators of
this disease. Most of the conclusions drawn in regard to this, however,
have been based on observations in the greenhouse or in plant beds,
where artificial watering is resorted to and where it is relatively common
in many instances to overwater the soil. A study of the disease under
field conditions with reference to moisture, though equally indefinite and
inconclusive owing to the number of other variable factors, is at any
rate convincing that an oversupply of moisture is not necessary for heavy
infection and severe attacks by the parasite.
108122°— 19 2
50 Journal of Agricultural Research voi. xvii. No. 2
In order to get more accurate evidence on the influence of soil moisture,
a series of pot experiments, with the moisture supply controlled as
closely as possible by weight, were carried out.
Several difficulties, not readily overcome, exist in such an experiment,
the principal one being that it is practically impossible to maintain a
uniform moisture content throughout the soil. No doubt the use of
Livingston auto-irrigators would have made possible more uniform results,
but these were not available at the time. Two-gallon crocks, perforated
at the base for drainage and holding about 10 kgm. of soil, were used.
The naturally infested soil from the old tobacco field on the Station
Farm was used. After a large quantity of this soil had been dried,
thoroughly mixed, and screened, its moisture content and water-holding
capacity were determined in the ordinary manner. Ten kgm. of the soil
were then placed in each of twenty 2-gallon crocks. The soil in 8 of
these crocks was sterilized by steam at about 100° C. for the purpose of
destroying all the T. hasicola present in order to provide disease-free
controls in the experiments. The water relations, as well as the food
relations, were, of course, changed in some degree by the sterilization,
and an absolute comparison between the sterilized and infested series
was therefore not permissible, although it is believed that the results
are not altered appreciably by this fact.
The crocks of soil were then divided into four series, each containing
three infested and two uninf ested crocks of soil. Two glass tubes, X inch
in diameter, one being inserted to a depth of 2 inches and the other to a
depth of 6 inches, were placed in each crock for the purpose of permitting
a more uniform distribution of water in the soil. Of the four series, one
was now made up to one-fourth its full water-holding capacity, and the
others to one-half, three-fourths, and full water-holding capacity. After
the water had been allowed to distribute itself fairly evenly, one plant
of the White Burley variety grown in sterilized soil was transplanted to
each of 20 crocks. The loss of moisture from the crocks was very slow
when the plants were small, especially during the winter in the green-
house. Usually it was not necessary to make the pots up to the required
weights oftener than once every three days, but later in the tests daily
attention was usually necessary. In an experiment begun on February
13, 1917, with the White Burley variety, it was noted at the end of one
week that in the infested series the plants at one-fourth saturation wilted
during days of high transpiration and showed the poorest growth. The
plants at three-fourths saturation got the best start, while those at full
saturation were already yellowing and apparently diseased, since no
such condition was observed in the sterile controls. On March 5 the
conditions were about the same in relative growth except that the diseased
condition of the plants at full saturation in infested soil was greatly
increased, and the controls in sterilized soils were now beginning to forge
rapidly ahead of those in infested soil. On March 14 it seemed quite
May IS. 1919 Influence of Soil Environment on Rooirot of Tobacco 5 1
evident that in uninf ested soil the optimum moisture content of this soil
type for the growth gf tobacco lay close to three-fourths saturation and
that full saturation was more favorable than the one-half and one-fourth
saturation. Although still no great differences existed between the
infested and uninf ested soils at one-fourth, one-half, and three-fourths
saturation, the plants in the uninf ested soil at full saturation were about
10 times as large as those in the infested soil at the same saturation.
The data taken upon the growth of the plants in these experiments are
recorded for the most part as the total leaf area of each plant taken at
intervals of about one week. These determinations were made by plac-
ing the leaves over a standard leaf -area chart on which areas for varying
sizes and shapes of leaves had been previously determined with a plani-
meter. These areas, expressed in square inches, were determined at the
end of this experiment, on March 28, and are given under experiment 2,
Table II. The results appear to justify the conclusion that a fairly
constant ratio exists between the growth on infested and uninfested soil
at the three lower saturations. This ratio is approximately i to 3^.
On the soils at full saturation, however, the ratio of growth on infested
soil to that on uninfested soil is about i to 40. The evidence from
this experiment therefore shows that a very considerable amount of
disease can occur in a relatively very dry soil and that it does not appear
to be proportionately increased in a relatively moist soil, but that a wet
or saturated soil, which still permits a good growth of tobacco when
uninfested, causes a rapid decrease in yield when infested with T. basicola.
Table; 2. — Influence of the moisture content of the soil on the amount of tobacco rootrot
Saturation.
Approxi-
mate per-
centage of
moisture.
Experiment 2: Average
leaf area (square inches).
Experiment 4: Average
air-dry weight (gm.).
Uninfested
soil.
Infested
soil.
Uninfested
soil.
Infested
soil.
One-fourth
7-7
15-5
23-3
31.0
34
81
301
239
9
22
89
7
2. 0
9.8
19.7
9-3
1.5
4.5
5-4
I. I
One-half
Three-fourths
Full
In a following experiment, which was conducted on the same soil in
the greenhouse, the above results were practically duplicated so that
the data will not be presented here. A third experiment was conducted
during the growing season in a shelter out of doors, permitting atmos-
pheric relations more nearly normal than those occurring in the green-
house during the winter season. The experiment w^as run in the same
manner as the foregoing one, except that a change of soil w^as made,
another batch from the same source was used in order to avoid nematode
injury. The crocks were set to White Burley on July 5. On July 12
all the plants appeared to have a good start except those at one-fourth
52 Journal of Agricultural Research voi. xvii, no. a
saturation, which usually wilted during the daytime owing to lack of
moisture. By July 21 marked signs of heavy infection of all the plants
in the infested soils was shown by reduced growth and yellowing of the
lower leaves; this condition was most marked at full saturation. The
uninfested soil showed the optimum growth at three-fourths satura-
tion. On August 14 the experiment was discontinued, the plants photo-
graphed (Pi. I, II-III), and then cut and dried. The air-dry weights
are given in Table II under experiment 4.
While the results of this experiment as shown by air-dry weight in
comparison with the leaf area given in experiment 2 do not conform in
all details with those of experiment 2 , they are believed to agree in general
in that the greatest amount of injury from disease resulted in the satu-
rated soil; the ratio of the yield on infested soil to that on uninfested
soil was again considerably greater than in the other cases.
It is realized that further experimental evidence could be profitably
obtained as to the relation of moisture to the disease. The difficulties
already referred to, however, together with complication of other factors
such as temperature, and some of the more obscure factors such as
aeration and compactness of the soil, have rather discouraged further
tests until more accurate technic can be devised. It is certain, how-
ever, that T. bascicola has a wide range of action as regards actual
percentage of moisture present in the soil. It has been found, for
instance, that in water culture containing a spore suspension, good
infection occurs on roots and that it will occur in soils too dry to permit
anything like normal growth of tobacco. Whether there is a direct
increase in amount of infection and severity of the disease with per-
centage increase of moisture in the soil may not be exactly clear from
the data here presented. From the majority of the data obtained,
however, some of which is not given here, it seems fairly certain that
such direct proportionality does not exist, but that a fairly constant
relation is maintained in soils v/ith moisture content ranging from those
sufficient only for poor plant growth to those approaching saturation,
followed by a very rapid increase of disease from this latter point up
to full saturation.
At any rate it may be said that poorly drained infested soils which
are likely to remain saturated for a period of three or four days, or
any inlested soil kept near saturation for a period of days due to excessive
rainfall will undoubtedly show higher infection than well drained soils
or soils not affected by an excessively wet season. On the other hand
it appears that, as a rule, soil moisture is not an important controlling
factor in the prevalence of the rootrot of tobacco. Relatively dry or
relatively wet seasons, in so far as they affect soil moisture alone, are
not especially to be feared nor to be relied upon for holding the disease
in check. For the same reason a careful check has not been kept upon
the moisture content of the soils in the various experiments described
May IS, 1919 Influence oj Soil Environment on Rooiroi of Tobacco 53
in this paper with other environmental conditions. The soils have
been kept as nearly as possible uniformly watered, never approaching
saturation for any considerable period of time and never relatively
dry, so that it is not believed that the results have been vitiated by
this factor, though it is admitted that they may have been responsible
for many of the individual variations in results always occurring in
experiments of this sort, and which is planned to be overcome by mass
of data rather than by the most careful attention to a single experiment.
If the above-ground symptoms of the rootrot are considered, it is at
once realized that the reduction of the water supply is probably the
most important one, partly as a result of reduced food supply to the
plant brought about by the gradual but effective depletion of the root
system. It is therefore usually quite impossible to judge from the
above-ground portion of plants alone as to whether lack of available soil
moisture or lack of roots or both are responsible for a reduced yield.
It is only when growers obtain a greatly reduced yield on land known
to be in a high state of fertility that they begin to suspect other troubles.
It is said with confidence, therefore, that fully nine-tenths of the damage
by the rootrot is attributed by the growers either to a deficiency or to
an excess of soil moisture. Sixteen moisture determinations of the field
plots (the same soil as used in the pot experiments) at about 3-day
intervals between June 19 and August 6 in the summer of 1917 showed
a range of moisture content from 25.3 per cent on June 25 to 16.2 per
cent on August 2. These determinations showed that during the entire
season the moisture content was practically between the limits of one-
half to three-fourths saturation. In these plots White Burley tobacco
made no growth whatever during this time because of T. basicola, and
Connecticut Havana made only half a crop. It is clear that the moisture
content was not excessive for the best growth of tobacco, and yet the
parasite was almost at its maximum of activity.
INFLUENCE OF SOIL REACTION ON ROOTROT
The reaction of the soil has been considered to play a part in the
severity of parasitism in practically all plant diseases having their origin
in the soil. The reasons for these rather widespread calculations are
perhaps manifold. Among the earliest chemical agents applied to soil
with the hope of checking plant pests was lime, and experiments too
numerous to mention here have since been conducted with it in the
hope of checking the diseases and insects attacking plants. Where lime
has proved efficacious, however, pathologists have considered it both as
influential as a sterilizing agent against the parasite and as a neutralizer
of soil acidity favorable to parasitic action. The beneficial action of
lime to the growth of green plants and to bacterial activity in soils has
also no doubt served to stimulate its use in phytopathological problems.
No advantage is to be gained by reviewing the rather extensive study of
54 Journal of Agrictdtural Research voi. xvii, no. a
the value of lime in the control of numerous plant parasites harbored in
the soil because the results obtained depend altogether upon the disease
concerned. Experiments in its use have been most complete in relation
to the control of potato-scab, clubroot of crucifers, and nematodes.
Limed soils favor scab, whereas clubroot and nematode injury are much
reduced by its use.
With the appearance of a paper by Briggs (j), based on field experi-
ments in Connecticut, a great deal of interest was revived on the influ-
ence of soil reaction on plant diseases having their origin in the soil.
Briggs concluded briefly that materials applied to the soil which tended
to make it alkaline in reaction favored Thielavia-rootrot of tobacco,
whereas materials applied which made the soil acid reduced the disease.
The actual change in soil reaction apparently was not determined. On
the basis of these results the use of acid fertilizers came to be recom-
mended in both scientific and popular literature and the use of lime
cautioned against. Considerable experimental work also v/as under-
taken at various places with this and other diseases, some of which
apparently verified the results of Briggs, while others showed no favor-
able results. Thus, the problem has remained in a more or less uncer-
tain state. Clearly it is not one which is easily solved. Changing the
reaction of the soil from acidity to alkalinity, and especially from alka-
linity to acidity by the application of different chemicals, is open to
many difficulties not clearly analyzable. The problem of soil reaction
as influencing disease presents two aspects: First, to determine the actual
influence of the reaction of the soil medium upon the disease; and second,
to determine in how far this influence may be utilized in a practical
manner by actually changing the reaction of large areas of soil to a
sufficient degree to modify the severity of the disease. The latter prob-
lem is complicated by a number of factors, the most evident of which is
the naturally or normally attained reaction of the soil, since this must
have considerable bearing upon the amount of acid-producing materials
which must be applied to get the desired result. Aside from the final
influence of such treatment on the soil itself, in a system of economical
and permanent agriculture a more discouraging feature of the problem
is that from the standpoint of the disease, for, as will be shown, many
other factors must be taken into consideration, such as the suscepti-
bility of the variety of tobacco grown, the amount of infestation, and
the temperature of the soil. By varying these factors markedly differ-
ent results may be secured on the influence of soil reaction in relation to
disease.
The experiments carried out in the investigation presented here v/ere
of two kinds, pot tests and field plot tests. The former were carried on
for the most part in the greenhouse during the winter season and the
latter on an old heavily infested tobacco field on the Station farm at
Madison.
May 15, 1919 Influence of Soil Enmronmeni on Rootrot of Tobacco 55
POT EXPERIMENTS
As has already been suggested, it is especially difficult to render an
alkaline soil acid in various degrees by the application of a theoretical
quantity of an acid or acid salt. The alteration produced in the soil by
either treatment is likely to alter it so fundamentally that comparison
with another soil treated in a dissimilar manner tends to complicate the
results to an unnecessary degree. What seems to be a considerably
better plan is to select a naturally highly acid soil and to change its
acidity to various degrees of alkalinity by the application of the theo-
retically correct quantities of lime. Accordingly, this plan was followed.
The soil selected was a very acid Sparta sand from a field at Lavalle,
Wis. A total acidity determination of this soil by the Truog method {26)
showed that its lime requirement was 9.38 tons per acre. The strength
of acidity was found to be 108 on the basis of acetic acid at 1,000. The
soil after being finely screened was placed into 2 -gallon stoneware crocks,
perforated at the base for drainage. Ten kgm. of soil were weighed into
each of 36 crocks. These crocks were then divided into 9 sets of 4 crocks
each. The calculated quantity of precipitated calcium hydroxid of the
highest purity was thoroughly incorporated in the soil of each of the 4
crocks in each series with a view to reducing the acidity to fairly definite
degrees. In Table III are given the quantities of calcium hydroxid
applied, together with the determinations of total acidity by the Truog
method made several months later, when it was considered that the full
effect of the treatment on the soil had occurred.
. Table III. — Influence of soil reaction on developtnent of rootrot
Series.
Quantity
of lime
added
to 10
kilograms
of
soil.
Lime
require-
ment
per
acre.
Average air-dry weight.
Experiment I (White Burley).
Experilnent II
(Connecticut
Havana).
In-
Unin-
fested
fested
soil.
soil,
Gtti.
Gm.
2. 46
2. 72
I. OQ
4. 09
■33
4-65
■ 17
4.40
.68
I. 49
• 24
1-53
. 40
•79
•3«
I. 24
•23
■73
Amount of infection on
roots (infested soil se-
ries).
In-
fested
soil.
Unin-
fested
soil.
Experiment III
(Maryland
Broadleaf).'*
In-
fested
soil.
Unin-
fested
soil.
3
4
s
6
7
8
9
Gm.
O. O
17.80
35- 60
53-40
71-25
89.05
106. 85
124. 60
142. 40
Tons.
9-38
Very slight . .
Considerable .
Heavy
....do
....do
....do
....do
....do
Gm.
6-35
2.65
2. 22
•75
2.06
•57
1.99
•43
•5S
Gm.
6.45
Gm.
O. II
14
17
14
18
24
16
Gm.
6.80
5.00
8. IS
8.67
IO-35
7. 12
7. 10
5.22
4. go
» Heavy infestation.
Two pots of each series were inoculated with pure cultures of T.
basicola, and young seedlings of the susceptible White Burley variety
were transplanted into them. The first test of plant growth failed.
56 Journal of Agricultural Research voi. xvii. No. a
owing, probably in large measure, to poor infestation; and the second
test was ruined by a heavy infection of nematodes. All the soils were
then sterilized by steam, and two of each series again inoculated, this
time by the application of equal quantities of chopped-up, air-dried,
heavily infected roots which had been taken in the fall from the tobacco
field. The pots were again set to young seedlings of White Burley from
steamed soils. The infestation now proved to be good, but complica-
tions arose in the limed end especially, owing either to the influence of
the lime itself or to the sterilization alone or more likely to the two
combined. It seems most probable that the lime was concerned in
rendering the phosphates unavailable to the plants, but, as was expected,
this injury was probably not a factor in the following experiment. How-
ever, this test, which has been designated as experiment I, in Table III,
produced some fairly striking results in the infested series.
It was evident during the early growth of the plants that those in
the infested soil of highest acidity were making considerably better
growth than those at the alkaline end. It also appeared early that the
line of demarcation between heavy infection and reduced infection in
this series did not lie near the point of neutrality but well into the acid
end and so the soil requiring 4.6 tons of lime per acre was as productive
of disease as any at the alkaline end. Growth of all the plants was
slow as the soil was not very fertile and the light poor. The plants,
therefore, were harvested when still relatively far removed from the
blossoming stage. They were dried at about 80° C. for several days,
and then allowed to come to air-dry weight. The data given under ex-
periment I, Table III, sufficiently illustrates the results obtained. The
roots were carefully removed and examined for lesions of the disease,
and, as was expected, these were correlated with the growth of the plants.
In the soil requiring 9.38 tons lime per acre only a most careful search
revealed any T. hasicola at all. In the next lowest series (7.19 tons per
acre), although considerable disease was present, plainly its development
was markedly checked, whereas in all the series below this practically
no uninfected portions of roots existed.
The crocks were now replanted with Connecticut Havana tobacco, a
semi- resistant type. These were allowed to grow for about 50 days be-
fore being harvested. The air-dry weights are given under experiment
II in Table III, It will be noted that the soil at the alkaline end prac-
tically recovered from the injurious properties previously described in
the control series. In the infested series it ma}' be noted that the point
of effectiveness of the acid reaction in reducing the disease shifted to a
somewhat lower degree of acidity, undoubtedly due to the greater re-
sistance of the variety (PI. 4, I-II). For some unexplainable reason
the plant in one of the pots of series 5 and also one in series 7 failed to
become as seriously infected as those in the neighboring pots. The in-
creased yields in series 5 and 7, however, are not considered as inter-
May J 5, 1919 Influence of Soil Environment on Rooirot of Tobacco 57
fering with the general conclusions to be drawn from the experiment,
and the writers have again shown that the highest acidity practically
eliminated damage from rootrot, but that heavy infection still occurred
in fairly acid soil.
It was believed that the abnormal behavior of the two soils in series
5 and 7 might be due to reduced infestation. All the crocks, therefore,
were resterilized and the two of each series reinfested with 100 cc. of a
fairly heavy suspension of endoconidia of T. hasicola in water. This
was thoroughly incorporated throughout the 10 kgm. of soil of each
crock. Maryland Broadleaf tobacco, a variety almost as susceptible as
White Burley, was then transplanted into them soon after inoculation.
The results of 65 days of growth are shown in the air-dry weight under
experiment III in Table III. A heavy infestation apparently reduced
the efficacy of the acid soils to nothing, at least in the presence of a
susceptible variety. The disease appeared, in fact, more virulent in
the most acid soil.
The soils were now again replanted to Connecticut Havana, the semi-
resistant type. The actual amount of infestation was also probably
somewhat reduced, as many of the spores originally introduced must
have spent themselves, although it is probable that the fungus was liv-
ing in the soil as a saprophyte. Results similar to those obtained in
experiment I were now secured, indicating that partial recovery from
infestation had occurred in series i and 2.
The question arises as to just what effect soil reaction has upon the
occurrence of the disease. High acidity may increase the resistance of
the host plant; or it may act deleteriously upon the germination of the
spores or the growth of the parasite. If we assume that acidity in-
creases the acidity of the cell sap and, hence, the resistance to disease,
as suggested by Comes {10) for cereal diseases, we have a working hy-
pothesis which is, however, difficult to establish definitely. It has been
shown that T. hasicola (as do most fungi) grows best on an acid medium
(about I per cent). Water extracts of the soils from the various series
were made which represented approximately the concentration of the
soil solutions. Germination tests of endoconidia in these extracts showed
better germination in the acid end than in the alkaline end. Other soil
extracts tubed with agar showed better growth of T. hasicola at the
acid than at the alkaline end. Although the acidity from the higher
acid series was such as would not permit the growth of bacteria, yet
tests of this nature probably fall short of resembling the actual acidity
in the soil. The results in experiment III show, of course, that spore
germination and fungus growth are not completely inhibited by an
acidity requiring 9 to 10 tons of lime per acre. The writer is inclined to
believe, however, that the beneficial action of soil acidity in reducing
infection by T. hasicola is due to a gradual depressing effect upon the
fungus.
108122°— 19 3
58 Journal of Agricultural Research voi. xvii, no. 1
FIELD EXPERIMENTS WITH SOIL REACTION
The field plots were located on the Experiment Station farm at Madison
on a tobacco field which had grown 10 and possibly 12 successive crops
of tobacco, and on a neighboring field which had previously grown only
I crop of tobacco. The infested field had for three or four years previous
to this experiment shown itself to be heavily infested and would grow
only half a crop of Coimecticut Havana tobacco, while White Burley
would make no growth whatever on this soil, especially during relatively
cool growing seasons. This soil had had heavy annual applications of
barnyard manure and was in a good state of fertility as shown by corn
and cereals growing in adjacent plots. The soil reaction at the beginning
of the experiments was practically neutral.
A control field across the road on uninfested soil was started for a
double purpose. In the first place it made it possible to check up the
beneficial or injurious action of the fertilizer and lime applied, aside from
infection from disease. In the second place it has been considered that,
although the apphcation of acid fertilizer might not remedy the condition
in a badly infested field, it might serve to hold down the rate of infesta-
tion of new soil to a considerable degree. This soil is equally as fertile
as the infested field, but shows a slightly greater degree of natural acidity,
being classed as slightly acid according to the Truog color chart. Since
it was found in the pot experiments that a very considerable range of
reaction was required to make any appreciable difference in amount of
infection by T. hasicola, it was decided to use the more simple though
fairly accurate comparative test of Truog {26) with lead-acetate paper.
The reference to the degree of acidity, therefore, will be based on the
standard color chart accompanying the description of this test.
The plots used were one-fortieth acre in size. The applications were
made in two different amounts, a heavy application and a light applica-
tion, also referred to as a full application and a half application, respec-
tively. The original plans of the experiment called for the use of alkaline
fertilizers — that is, potassium carbonate, basic slag, and nitrate of soda,
with equivalent amounts of sulphate of potash, acid phosphate, and sul-
phate of ammonia for the acid fertilizers. On account of the apparent
impossibility of obtaining all of the alkaline fertilizers, it was decided
to use the acid fertilizers and heavy applications of lime to produce the
alkaline condition. The rates of applications, in pounds per acre, then,
are as follows:
Acid plots. Full amount. Half amount.
Sulphate of ammonia i, 200 600
Sulphate of potash i, 200 600
Acid phosphate 2, 400 i, 200
Alkaline plots.
Slaked lime 12, 000 6, 000
Sulphate of ammonia i, 200 600
Sulphate of potash i, 200 600
Acid phosphate 2, 400 i, 200
May IS. 1919 Influence of Soil Environment on Rootrot oj Tobacco 59
The first application was made on June 18, 1917. The appUcations
were made by hand, disked, and harrowed in. The lime was previously
allowed to slake in the field. Connecticut Havana tobacco was trans-
planted on all the plots on June 27. Acidity determinations made one
month after the applications showed slightly increased acidity for the
acid plots and slightly decreased acidity for the alkaline plots. Determin-
ations unfortunately were not made at the end of the season, but the
tests for the following year served to indicate that, although the changes
were not great in degree, they were decidedly effective in bringing about
a marked change in reaction between the alkaline and acid plots.
In the first year's tests the acid plots in the infested soil showed up
decidedly the poorest throughout most of the growing period, while the
heavily limed plot was decidedly the best in the series. On the new field
the fertilizers both with and without lime gave somewhat better results
than the controls. So far as can be judged by the results, the acid ferti-
lizers were not injurious to the crop on this soil although it is possible, of
course, that these materials might have had some direct injurious action
on the tobacco in the case of the infested soil. Apparently, such a condi-
tion did not occur on the uninfested soil, nor on the infested soil when
the plots treated in the same manner were limed.
The plots were harvested and cured separately; the yield of cured
leaves for 191 7 are given in Table IV.
TablB IV. — Yield of tobacco on soil with acid fertilizers , with and without lime, igiy-lS
Plot.
Application on is acre.
Yield of cured leaf on A acre
(pounds).
Treatment.
Acid fertilizers.
Lime.
1917-
1918.
In-
fested
soil.
Unin-
fested
soil.
In-
fested
soil.o
Unin-
fested
soil.a
In-
fested
soil.
Alkaline ....
A...
B...
A...
B...
A...
B...
[Sulphate of ammonia, 30 pounds
1 Lbs.
> 300
[ 150
None,
None.
>None.
[•None.
38. s
26; 0
24.0
22. 0
iS-o
IS- 5
44- S
40.0
33-5
33-0
44- S
38-5
31-3
29. 0
38.5
36.0
30-5
29-3
34- 0
(Acid phosphate, 60 pounds
Do
(Sulphate of ammonia, 15 pounds
■J Sulphate of potash, 15 pounds
Control
Do
None
28. 0
Acid
[Sulphate of ammonia, 30 pounds
Acid phosphate, 60 pounds
Do
[Sulphate of ammonia, 15 pounds
sSulphate of potash, 15 pounds
41.5 29.3
" 1917 series. These plots received a second application of same amounts in 1918 and had, therefore, the
residual effect of the 191 7 appUcations.
The results were sufficiently interesting to warrant repetition the fol-
lowing season (1918) on a slightly increased scale. All the plots were
again given an additional application, the same amounts as in 1 917 being
used. In the infested field six plots were added, these being given
6o Journal of Agricultural Research voi. xvir, No. a
the same treatment as the others, the essential difference being that these
did not have the residual effects of the previous season's applications and
would therefore be more directly comparable with the plots in 191 7.
The applications of lime were made on May 21 and fertilizers applied
on June 3. On June 12 all the plots were planted to Connecticut Havana
tobacco. On June 21 samples of soil were taken from each of the plots
and tested for reaction in the ordinary manner.
In the infested soil the control plots showed very slight acidity. The
acid plot of last year (full amount) showed medium to strong acidity;
the half-amount plot showed slight acidity. In the same way the plots
which had received the acid fertilization for the first time in 191 8 showed
nearly medium acidity for the full application and slight acidity for the
half amount. None of the alkaline plots showed acidity and presumably
were considerably below the neutral point, though this could not be
shown by the test used. On the uninfested plots the change in acidity
due to the application of the fertilizers were not so marked, probably only
a slight change having been produced. The limed plots, however,
showed no signs of acid reaction.
On June 18 it already appeared that on the full-limed plots the lime
was acting injuriously upon the seedlings, both in the uninfested and in
the infested soils. This may have been due in part to the fact that the
lime was not well air slaked and was, hence, not thoroughly incorporated
in the soil. The action of the lime was, therefore, probably toxic and
probably vitiated the results, so far as lime was concerned, although the
plants appeared to recover later in the season. The outstanding feature
of the results in 191 8 was again that the plots made acid with heavy
applications of fertilizers under field conditions were on the average little
or no better than the untreated plots (Table IV) . In fact, the untreated
plots of the 191 7 series were considerably better than the acid-treated
plots of 1917 or 1918, although the plots treated with acid fertilizers for
the first time in 191 8 were slightly better than their controls for this year.
There is no question as to the extent of infection on this soil this season,
since resistant and susceptible types planted at the same time behaved
in the expected manner. In interpreting the results from the field plots
it should be recognized that the tests are not exhaustive, and that on
account of the complexity of the problem the conclusions drawn may not
apply under all conditions. For Wisconsin conditions, however, it ap-
pears that the application of acid fertilizers to soils, alkaline or neutral in
reaction, will not reduce infection by T. hasicola.
SOIL TEMPERATURE AS A FACTOR IN ROOTROT
A review of the more important literature concerning the influence of
soil temperature on diseases of plants and the importance of such studies
has been presented by Jones (16). With respect to the influence of
this factor on infection and severity of the rootrot of tobacco caused by
May 15, J919 Influence of Soil Environment on Rooirot of Tobacco 61
T. hasicola practically nothing of a definite nature exists. Rather
obscure statements that high temperatures favor the disease have been
published by Gilbert {12), while, on the other hand, Clinton {8) states
that possibly unusually cold, wet spring weather has something to do
with determining whether or not the fungus does much damage. Gallo-
way (11, p. 174-17S) found that in the greenhouse the disease was appar-
ently more severe on violets on the approach of fall than in^ summer,
indicating a temperature relation. In Italy where very considerable
observation has been made on the disease, it is agreed that weather
conditions have much to do with its occurrence and severity. That
such was the case in Wisconsin was evident during the first season of
observation. The recovery of badly infected plants in large areas
during the course of only two or three weeks led to the desire to study
in more detail the environmental conditions bringing this about. It
was at first suspected that the moisture relations were the all-important
factor; but in connection with its study, temperature records of the
soil under field conditions were taken beginning in the spring of 191 5,
and continued for the seasons of 1916, 1917, and 191 8.
In the fall of 191 6, following some interesting results by Tisdale (25)
on the influence of soil temperature on flaxwilt (caused by Fusarium
lini), the writers, under the advice and support of Dr. L. R. Jones,
undertook to have a large tank (PI. 2, I) constructed in which soil could
be held fairly constant at several different temperatures. This tank
has already been described and illustrated in some detail by Jones {16).
Further detailed description of the mechanical part of the apparatus
seems unnecessary, especially in view of the fact that improvements are
being gradually made on these tanks which will no doubt necessitate
further description of similar apparatus developed in the Department of
Plant Pathology of the University of Wisconsin. It should be said,
however, that by means of proper insulation of the compartments it
has been possible to maintain a fairly constant temperature of water
at any selected temperatures between approximately 5° and 40° C.
This has been done by the inflow of cold water from the taps in the winter
time and by heating the water to the higher temperatures with electric
bulbs or with steam. The expense of automatic temperature regulation
in a large number of chambers has discouraged the use of such apparatus
up to the present time, but personal attention and regulation two and
three times every 24 hours, in combination with good insulation, has been
found to give results sufficiently accurate for most needs. It was found
that although considerable ranges of temperature occurred at the
extremes (below 15° and above 30°) the temperatures between 15° and
30° could be held quite constantly within i degree.
The soils used were placed in i -gallon battery jars and set on boards
suspended in the water in the tanks. Four jars could be placed in each
compartment with displacement of only a relatively small amount of
62 Journal of Agricultural Research voi. xvii, No. »
water. Two jars in each compartxnent containing sterilized or uninfested
soil were used as controls^for plant growth alongside two jars containing
infested soil. Naturally infested soil from the old tobacco field on the
Station farm, previously referred to, was used in most of the experi-
ments. After being given a good application of well-rotted manure,
the soil was thoroughly mixed and screened before weighing equal
quantities into the jars. The sterilized soil used in the earlier experi-
ments was sterilized by steam to destroy the infestation by T. basicola.
Considerable difficulty was experienced, however, as a result of the toxic
action of the heated soils on plant growth at the lower temperatures,
which interfered to some extent with the reliability and uniformity of
the data obtained by leaf measurements.
In later experiments the employment of soil steamed two or three
weeks previous to being used and allowed to stand in a moist condition
at a fairly high room temperature reduced this action to a minimum.
In still other tests formalin-sterilized soil was used with equal success,
and in the final experiment another uninfested soil was used and artificial
innoculation resorted to for the infested series.
The data taken in the earlier experiments were mostly in the form of
measurements of leaf area in square inches as determined by a standard
chart of various leaf sizes whose areas had previously been determined by
the use of a solar planimeter. In later experiments air-dry weight de-
terminations of the stalks and leaves were made.
The determination of the actual amount of disease on the roots is, of
course, the final criterion for judgment, and in the last experiments it was
found that with care the greater part of the roots could be washed out
from the soil, examined for disease, dried, and weighed; these weights are
closely correlated with growth aboveground, so that either the area of
the leaves, weight of the leaves and stalks, or weight of the roots alone
give a good index of the extent of the disease. A preliminary report of
the results obtained has been given and an abstract published (15).
Eight separate experiments have now been made upon the influence of
soil temperature on the extent of the root disease, four determinations
being made in the winter and spring of 191 7, and four during the fall and
winter of 191 7-1 8. The first experiments were made over a range of
about 35° C, but, as these were found to be beyond the ranges of normal
growth and infection, the later experiments usually included a tempera-
ture range of about 15°. Three of the experiments failed more or less
to give uniform results; one due to nematode infection at the higher
temperatures, another to toxic action of the heated soil, and a third to the
accidental use of infected seedlings.
The procedure in each experiment consisted merely in filling the re-
quired number of jars with soil; one-half with uninfested and one-half
with infested soil. Glass tubes 2)4 inches long, were inserted into each
jar to permit watering part of the soil at about half its total depth. After
May 15, 1919 Influence of Soil Environment on Rootroi 0} Tobacco 63
being brought up to about three-fourths saturation, the jars- were set in
the tanks at the different temperatures and allowed to remain there for
three to five days to permit the necessary changes of temperature. One
young seedling of tobacco, usually the susceptible White Burley variety,
was then transplanted into each jar. Subsequent attention then con-
sisted only in taking the temperature records twice a day, in maintaining
the proper temperature, and in watering the plants as required. In the
first experiments, when the temperature range was determined, 12
difi'erent soil temperatures were run at one time ; but in the latter experi-
ments, when a closer approximation of the critical temperature was
necessary, only 6 or 7 temperatures were used.
Experiment I. — Twelve temperatures were used, ranging from 7° to
40° C, and the white burley variety was transplanted into the jars.
The plants in the uninfested or sterilized soil series grew best at tempera-
tures of 29° and 31°. Practically no growth occurred below 13°, and
again there was poor growth at 40°. iProm a physiological standpoint it
was interesting to note that a marked effect upon the shape of the plants
occurred especially at the higher temperatures. While the plants grew
low and stocky with broad but rather pointed leaves at the optimum
temperature for growth, the plants became tall and spindly, with short
and rounded leaves at a temperature of about 36° to 40°. '
In the infested soils at temperatures above 26° the plant growth ap-
peared to be almost as good as that in uninfested soil (Pi. 2, I). At the
temperatures 23°, 21°, 19°, and 17°, however, a very decided reduction in
growth occurred as compared with the uninfested soils at the same tem-
perature.
Upon removal of the roots from the infested soil series it was found that
those at temperatures between 23° and 17° were heavily attacked by T.
basicola and that slight infection occurred at 7°, while at 26° relatively
few lesions occurred. The lesions were still less common at 29°, while at
31 ° only one lesion could be found. At the higher temperatures, 35° and
approximately 40°, no signs of Thielavia infection were found.
Experiment II. — In this experiment the temperature range was 9° to
40° C. The toxic action of the heated soils at temperatures of 17° to 25°
became quite marked early in the experiment, and no doubt affected the
results. The total leaf area of each plant was determined at four different
times during the course of the experiment. The results lack uniformity,
however, owing to the toxic action of the heated soils. The infested soils
gave the best growth at 35°, with an average of 251 square inches, as com-
pared with 289 square inches for the sterilized soil at the same tempera-
ture. The poorest growth was at 19°, where a leaf area of only 19.2 square
inches was obtained in the infested soil, as compared with 206.4 square
inches in the sterilized soil. While the disease was quite marked at 24.5°,
71 square inches on infested soil as against 205 square inches on sterilized
soil, decided indication of reduced severity again appeared at 26^, 169
64
Journal of Agricultural Research
Vol. XVII. No. a
square inches in infested soil as against 203 square inches on sterile soil.
Examination of the roots in the infested series showed a relatively reduced
amount of infection at 9° and 13°, heavy infection between 17° and 24.5^,
much less infection again at. 26° and 29°, and no infection at 31°, 35°,
and 40°.
The results of Experiment II are in accord with the results of Experi-
ment I, and apparently show in addition that the optimum temperature
for the disease lies around 19° and 21°, although heavy infection still oc-
curs as high as 24.5°.
Experiment III. — The same soil was used as in Experiments I and II.
This soil was now so heavily infested with nematodes at the higher tem-
peratures that the results with T. hasicola were vitiated, and no data were
taken.
Experiment IV. — New soil from the same infested field was used in
this experiment, the proper care being taken to sterilize thoroughly the
battery jars before filling them with soil. To reduce the harmful effect
of the sterilized soils, the pots, after being filled, were allowed to stand
moist for a week at room temperature before being placed in the tanks.
In this experiment only seven different temperatures were used, which
permitted the use of four jars of infested soil and four controls at each of
five temperatures, but only two of each at the extremes. The final results
are given in Table V. It may be again noted that the greatest amount
of injury from disease occurred at the temperatures from 19° to 22° C, less
occurred at 24° to 25°, while at 26° to 27° the injury was much reduced.
Table V. — Influence of soil temperature on development of rootrot
Temperature.
Average leaf area (in square inches) of duplicates.
Experiment III.
Uninfested
soil.
Infested
soil.
Experiment IV.
Uninfested
soil.
Infested
soil.
13-15
15-17
19-20
21-22
24-25
26-27
30-31
63
106
206
212
205
202
256
28. 5
24.9
19. 2
26. 7
71. I
168.7
243.6
71
87
137
304
430
306
339
41.7
13.0
9-5
21.4
197.6
280. 4
327.1
Experiment V. — ^This experiment was largely a failure, owing to the
use of plants that apparently were slightly infected by T. hasicola and
also by nematodes. No plants from sterilized soil were available at the
time. The results were interesting, however, in that an examination of
the roots showed that at the lower temperatures — that is, those favorable
to infection — the heaviest infection occurred in the sterilized soil. This
May 15, 1919 Influence of Soil Environment on Rootroi of Tobacco 65
is in line with the frequently observed fact that sterilized soil reinfested
is a very favorable medium for the progress of disease. At 31° to
32° C. a trace of infection was found in the sterilized soil, but no infection
occurred in the naturally infested soil. Nematode injury was most serious
at the higher temperatures.
Experiment VI. — Six different temperatures ranging between 17°
and 32° C. were used. The best growth of the controls in uninfested
soil occurred at 31° to 32°, and the poorest at 17° to 18°. The best
growth in the infested series was also obtained from 31° to 32°, which
was practically equal to that of the controls. Almost equally good
growth occurred at 28° to 29°, but at lower temperatures the results
were again unfortunately interfered with by the toxic action of sterilized
soil, which, though it had been treated for the purpose of reducing the
toxicity, had not apparently sufficiently reduced the toxicity. Exami-
nation of the roots, however, which were carefully washed out, yielded
results in line with the previous experiments.
Experiment VII. — In this experiment the soil in the uninfested series
was sterilized with a i to 50 formalin drench three weeks prior to its use,
in order to avoid further interference by the toxic action of the steam-
sterilized soils. Six temperatures ranging from 17° to 32° C. were again
used. This experiment was begun on January 18, 1918, using the "White
Burley variety in the same soil as previously used, and concluded on
February 26. Marked differences in growth on the uninfested and
infested soils w^ere already noticeable on Februar}/ 5; the plants in the
sterilized soil 20° to 21° and 23° to 24° were twice as large as those in the
infested soil; whereas the plants in the invested and uninfested soil at
31° to 32° were practically equal in size. The final results are shown in
Table VI A, in terms of air-dry weight of the stalks, leaves, and roots in the
infested and uninfested series, together w4th the amount of infection on
the roots. It may be seen that the best temperature for growth in this
case was apparently 28° to 29° for the above-ground portions of the plant,
but that the best root development took place in the cooler soil at 23° to
24°. In 'the infested soil a gradual increase in growth from the lowest
•to the highest temperature is evident. (PI. 2, II-III.) Practically the
same is true for root development. (PI. 2, IV.)
Experiment VIII. — In this experiment an ordinary greenhouse soil
mixture, free from T. hasicola was used. No sterilization, therefore,
was used, and infestation with T. hasicola was produced by thoroughly
incorporating a heavy spore suspension of endoconidia of the fungus
from young cultures on agar. The experiment was now run as before,
except that the Connecticut Havana variety, which is relatively much
more resistant to T. hasicola than the White Burley, was used. The
young seedlings were transplanted on March 6, 191 8. On March 17
the plants in the inoculated pots already showed signs of heavy infec-
tion at the lower temperatures. On days of high transpiration the
108122°— 19 4
66
Journal of Agricultural Research
Vol. XVII. No. 1
plants in the inoculated series wilted first at from 22° to 23° C, but no
wilting occurred at 31° to 32° or in the inoculated soil. On April 15
the experiment was terminated, the stalks and roots were cut and dried,
and the roots washed out as carefully as possible and dried. The air-
dry weights are shown in Table VI B.
Table VI. — Influence of soil temperature on the rootrot of tobacco
A. WHITE BURLEY VARIETY, NORMALLY IISTFESTED SOIL
Tempera-
ture of
Average air-dry weight per
plant.
Series.
Stalk and leaves.
Roots.
Amount of infection in infested soil.
„„'!
Unin-
Infested
Unin-
Infested
fested soil.
soil.
fested soil.
soil.
"C.
Gm.
Gm.
Gm.
G)n.
I.». . .
17-18
3-6
0-3S
0.32
0 03
Heavy; roots all black.
2
20-21
5-4
I. 20
.66
.07
Do.
3
23-24
7-1
1.70
.86
. 10
Not quite as heavy as in series
I and 2.
4
25-26
6.9
■ 2.7s
•79
•17
Considerable, but much less than
in series 3.
5
28-29
7.6
3-8
.70
.16
Slight infection.
6
31-32
5-9
5-7
•25
.28
No definite signs of disease.
B. CONNECTICUT HAVANA VARIETY, ARTIFICIALLY INFESTED SOtI,
I
12-13
1.8
0-53
0. 26
0. 16
Heavy; very few white roots.
2
17-18
7-9
•65
.70
■ 15
Heavy; about same as in series i.
3
22-23
9. I
I- 15
. 1.05
• 19
Heavy but less than in series i
and 2.
4
26-27
10. 6
3-75
I- 13
•38
Much less than in series 3.
5
28-29
10. 8
8. 10
I- 13
.90
Very slight.
6
31-32
lo- S
7. 20
■75
.60
No sign of infection.
The largest yield of the above-ground portions of the plants in the
uninoculated series occurred at 28° to 29°, but was only slightly
larger than at 26° to 27° or 31° to 32°. The largest root 'develop-
ment occurred at 26° to 27° and 28° to 29°, with a decided falling
off at 31° to 32°. In the inoculated soil the largest yield of the above-
ground parts was at 28° to 29°, with some falling off at 31° to 32°,
though not due to infection. It should be noted here that the greater
reduction in yield is at 17° to 18°, the disease apparently less marked
at 12° to 13° and at 22° to 23°. Practically this same relation
holds for the roots. This, together with other experiments, seems to
indicate with considerable certainty that the amount of infection and
severity of the rootrot are most marked at temperatures ranging between
17° to 23° C. At temperatures below about 15° the extent of the disease
is reduced, but this temperature also is too low to permit any growth
of tobacco, and consequently is of little practical importance. On the
May 15, 1919 Influence of Soil Environment on Roolrot of Tobacco 67
other hand, at temperatures of 26° and above, the amount of infection
and the extent of the injury done are gradually reduced until at about
30° no appreciable injury results, and at at 31° to 32° it is permissible
to say that practically no infection whatever occurs.
The results having shown that the rootrot can be practically con-
trolled by high soil temperatures, which at the same time are favorable
for the growth of tobacco, the question naturally arises as to how far
a plant may recover from serious root infection, provided a change of
soil temperature from one favorable to disease to one unfavorable to
disease is brought about. Eight White Burley plants which had been
planted to the infested tobacco field in June but which had made no
appreciable growth during the entire season in the field were taken up
late in September with their adhering soil and transplanted into the
jars with the infested soil. Four of these were then set in the tem-
perature tanks at a low temperature (20° to 21°) and four at a high
temperature (30° to 31°). After remaining at these temperatures for
a month the roots were dug out as carefully as possible, and the results
are illustrated in Plate 3. The plants had almost no roots when placed
in the tank, and one must marvel at the wonderful persistence of tobacco
plants in maintaining themselves with an almost complete lack of root
system. At the higher temperature, however, new roots were forced
©ut through the blackened bases of the stalks and remained uniformly
clean, white, and free from disease. This experiment was repeated
with even more striking results by mo-^ng jars with badly diseased
plants from the low temperatures to the high temperatures in the tanks.
In the space of three or four days the plants seemed to have taken
on renewed vigor and growth. These experiments seem to prove
beyond doubt that similar conditions may happen in the field under
practical conditions, and that the phenomena of recovery of a badly
diseased crop, so frequently noted in the field within a short period of
time, is no doubt due in large measure to natural changes in tempera-
ture relations of the soil.
SOIL TEMPERATURES IN THE FIELD
It now remains to ascertain how far the soil temperatures occurring
under normal conditions in the field may influence the actual amount
of infection and damage from rootrot. It is necessary, therefore, to
determine the actual soil temperatures occurring during the growing
season in order that a knowledge may be obtained of the change occur-
ring at different times in the same season and during different seasons
taken as a whole. Unfortunately not a great many reliable data upon
soil temperatures for summer months in various sections of the country
exist. Such as do exist, however, may have a bearing upon future
studies of the influence of soil temperatures upon the occurrence of
disease. It is to be expected that soil temperatures have a fairly con-
68
Journal of Agricultural Research
Vol. xvir, No. 2
stant correlation with air temperatures, and it is highly probable that
a fairly constant ratio may be calculated which will enable the exten-
sive data on air temperatures to
be used in considering relations of
soil temperatures to disease.
The data taken in connection
with the studies presented in this
paper were started in the spring of
1 91 5. For this purpose electrical
resistance thermometers were used.
These were buried in the soil in
tobacco fields at the Station farm
at depths of 2, 4, and 8 inches.
Some of the thermometers were
buried in such a way that they
would record the temperature of
soil becoming gradually shaded by
the growing tobacco, while with
others the soil was exposed con-
tinuously to the full sunshine.
The latter temperatures are the
ones upon which conclusions were
drawn, since in a badly infested
field shading of soil would be rela-
tively small, owing to the poor
growth of the crop. On the other
hand, where very heavy infestation
does not occur or a relatively re-
sistant variety is used, the relative
importance of shading must be
considered (fig. i).
The temperature readings were
taken with duplicate thermometers
each day at i p. m. This hour
was selected as it was the most con-
venient time of the day to take
the readings. No great im-por-
tance, however, can be attached to
the time of taking daily readings,
on account of the great daily varia-
tion which occurs especially near
the surface of the soil. It would
be most desirable to record the minimum and maximum tempera-
ture for each day, but in using electrical thermometers this would
May IS. 1919 Influence of Soil Environment on Rooirot of Tobacco 69
entail too many readings. Temperature readings taken at 7 a. m, and
at I and 5 p. m. for one week (July 6 to 13, 191 6) showed that the tem-
perature was anywhere from i to
5 degrees lower at 7 a. m. than at
I p. m. and to average slightly
higher at 5 p. m. than at i p. m.,
indicating that the maximum
perhaps was reached at some
time between i and 5 p. m., and
on clear days at about 3 p. m.
A recording soil thermograph
was also used at a depth of 4
inches in 191 7. From these rec-
ords it may be noted that the
highest temperature usually oc-
curred about 4 p. m. (fig. 2).
A more important consider-
ation, however, is the general
rise or fall of temperature during
extended periods of a week or
more, or the general trend of the
temperature for one season as
compared with another.
In the northerly latitudes the
growing period of tobacco,
practically speaking, lies within
the months June, July, and Au-
gust. Although much tobacco
remains in the field during the
month of September, practically
all the growth must be made be-
fore that time. July undoubt-
edly is the critical month in
which most of the growth should
be manifested, although if
growth is retarded until August,
and a warm fall follows, with ab-
sence of frost until late into
September, a marketable crop
may often be produced. In the
northern districts nearly all to-
bacco is planted in June. From
the standpoint of temperature this is the most favorable month for the
rootrot. Heavy infestation in June followed by a warm July, however,
yo Journal of Agricultural Research voi. xvii, no. 2
may overcome the disease. If the warm period is delayed until late
July or August, recovery may still be made and a late crop of good
yield produced, provided the balance has not swung to the other ex-
treme— that is, forced maturity.
No condition is more commonly seen in infested tobacco fields than
that of plants budded out and ready for topping two to three weeks
before the normal date when the plants have obtained only one-fourth
to one-half their normal growth. This is a direct result of the starvation
of the plants caused by disease. A drouth may bring on the same
condition. The plants then must be topped when this stage is reached,
and although much spread of leaf may subsequently occur, owing to the
arrival of more favorable conditions for growth, yet the yield is almost
certain to be light.
Several years of practical observation of infested fields have shown
that heavy infection almost always occurs in June. Every tobacco
grower of experience, at least in Wisconsin, can cite cases where during
the first two or three weeks after planting the crop prospects have been
excellent, followed by a like period of uncertainty, when the condition
of the crop has apparently made no progress or has gone slightly back-
ward, and finally, for no apparent reason, where the crop has taken on a
new lease of life, or, on the contrary, has remained to the end more or
less of a failure. In Wisconsin a large percentage of poor crops in the
years 1913, 1915, and 1917 was due either to poor yield or delayed
maturity directly traceable to the rootrot. In the years 191 4 and 191 6
fairly good yields were obtained, and not much root disease occurred
even on infested soils.
It is believed that an examination of the summarized soil temperature
records for these years in Table VII, or a glance at the temperature
curv'es in Plates 6-8, will furnish in a large measure an explanation for
the results obtained with tobacco in 1915, 1916, 1917, and 1918. The
year 191 5 was an especially cold season; according to weather bureau
records at Madison it was the coldest on record, and also a comparatively
wet one. The studies of the writers on the influence of soil moisture,
however, have now convinced them that its importance as a controlling
factor under field conditions is small as compared with temperature.
In 1 91 5 the loss from the rootrot of tobacco was estimated at from
$10,000,000 to $20,000,000 in the United States alone. The year 191 8
showed very poor prospects of a good crop for a period of several weeks
in July and early August. In the latter half of August, however, the
Wisconsin crop made a remarkable growth even in the most heavily in-
fested fields; this growth was unquestionably a direct result of the in-
crfeased soil temperatures during this month.
May IS. 1919 Influence of Soil Environmeyit on Rootrot of Tobacco
TabliJ VII. — Average monthly and seasonal soil temperatures for tobacco-growing periods,
IQIS-IQ18, at different depths of soil
Season.
1915-
1916.
1917.
1918.
Depth
of soil.
Temperatures during month of—
June.
July.
August.
Inches.
\ 1
°C.
20. 9
20. 4
18. 0
°C.
20. 9
20. 6
19. 2
19. 0
19. I
18.3
1 i
27. 6
23.0
18.6
31.8
27.8
24-3
23.8
22. 7
22.4
1 I
21.7
18.7
16.3
28. 2
24- 5
21. 6
27. 6
24.8
21-5
1 I
23.8
21. 0
17.9
27. 0
23.6
20. 5
29. 0
25-4
21.8
Average
for
growing
period.
C.
20.3
20. o
18.5
27.7
24- 5
21.8
25.8
22. 7
19.8
26.6
23.0
The practical bearing of this problem is manifold. In so far as sea-
sonal temperatures can be judged and predicted, crop prospects on in-
fested soils can be predicted, and in the northern tobacco-growing sec-
tions the infested soils usually constitute anywhere from one-half to
three-fourths of the acreage grown. In so far as "warm" soils can be
selected — that is, sandy, dark soils with good drainage and a southerly
exposure — in preference to "cold" soils, the possible extent of the damage
from disease has been reduced. If the crop is planted early on infested
soils, heavy infection is more Ukely to occur in the early stages of plant
growth, and the plants will find it more difficult to recover. One of the
most common beliefs of the Wisconsin grower, based on observations
of several years, is that early planting means plants budding out
in July, and an early, light-weight crop. From a purely physiological
standpoint there could be only one possible explanation for such behavior
of early-set tobacco, namely, the more or less common occurrence of a
drouth in July. The inadequacy of such an explanation, however, is
shown by the follomng observations: The vigor of growth of corn and
other cultivated crops has remained practically unchecked during many
of these frequently recurring so-called drouths in July; likewise, the
growth of ordinary tobacco on new soil of a neighboring farm or of a
resistant variety in the adjoining row on an infested field has not been
greatly interfered with by these weather conditions; finally, exceptionally
poor crops of tobacco were grown in Dane County, Wis., in the years
1 91 3 and 1 91 5, whereas the July rainfall was 8.47 inches in 191 3 and 5.04
inches in 191 5, both greatly in excess of the normal.
It would seem that some value could be attached to late planting on
soil infested by T. hasicola, in view of the low temperatures in June.
72 Journal of Agricultural Research voi. xvii, no. 2
On the other hand, the practical application of such a recommendation
is doubtful on account of the variations in seasons as to temperature and
general growing conditions. In general the farmers must transplant to
the fields when the seedlings are of proper size, a matter which usually
can not be predetermined very effectively for more than two or three
weeks. With steam-sterilized seed beds closer approximations can be
made, and seed may be sown two to three weeks later than normally,
with fair certainty of obtaining plants by June 20 to 30. Planting later
than July i, however, is no more certain of giving satisfactory final
results than early planting.
During the season of 191 7 a planting experiment was carried out,
with the hope of getting some data on this subject. Seedlings were
transplanted at intervals of one week from June 1 1 to July 23 on infested
and uninfested soil. Unfortunately, the White Burley variety was used
on heavily infested soil, and the season being relatively cool throughout,
no appreciable difference in yield occurred on the infested soil. On the
uninfested soil, however, the advantage was all with the early-set tobacco;
a gradual decrease in size and value occurred in the later plantings. A
wide range of obser^'^ation has convinced us that, other conditions being
alike, early-planted tobacco on uninfested soil usually is considerably
safer than late-set tobacco on either infested or uninfested soil.
Before leaving this subject another point of more scientific interest
should be considered: Why are tobacco roots most seriously attacked
by T. basicola at from 17° to 23° C. and practically not at all at a tem-
perature of 30° C. ? Several hypotheses may be formulated. The sim-
plest explanation would be that the resistance of the roots to the parasite
is modified at different temperatures, high susceptibility occurring from
17° to 23° and practical immunity at 30°. At first sight a tenable
theory seemed to be that the increased vigor of root formation at higher
temperatures sufficed to overcome the destructive effects of the disease.
On the other hand, the action of temperature variation may be regarded
as modifying the ability of the fungus to grow in the soil or to attack
the host. On the basis of some preliminary experimental results, the
latter theory seems to be the most probable.
It should be noted, however, that the behavior of the parasite in cul-
ture does not correspond entirely with its behavior on the host as regards
temperature relations. Gilbert (12) found the following critical tem-
peratures for growth: Minimum 7° to 8° C, optimum 30°, maximum
34° to 37°. The determinations of the writers have given figures very
much the same as these. The temperature most favorable for infection
does not therefore agree with the optimum for growth in culture.
On the other hand, the optimum growth in culture is obtained at 29°
to 30°, where the organism is apparently ineffective as a parasite. It
is not possible, therefore, to draw any decisive conclusions as to the
behavior of the fungus as a parasite from its behavior in artificial culture
May 15, 1919 Influence of Soil Enviro7iment on Rootrot of Tobacco 73
media. At temperatures of only 3 or 4 degrees above the optimum in
culture, however, the fungus, though making some growth, behaves
quite normally, and it is not difficult to conceive of no infection or
growth occurring on the host at temperatures above 30°. The results
are probably in line with the relation of temperature to infection with
other parasites, where it is known that the fundamental factor concerned
is that of spore germination.
The following brief description of some experiments may be of inter-
est : The roots of tobacco plants in 7-inch pots were forced to grow out
through the perforation in the bottom of the pots by setting them in
battery jars partly filled with water. After the roots had made a good
start, the jars containing the plants were set in the temperature tanks,
at high and low temperatures, 31° to 32° C. and 17° to 18°, respec-
tively. They were allowed to remain there for a week to 10 days. Dur-
ing this time many fresh roots were formed. Two jars at each tempera-
ture were now removed to a temperature of 23° to 24°, and young endo-
conidia of T. hasicola introduced into the water surrounding the roots.
Other plants remaining at 31° to 32° were also inoculated in a similar
manner. Good visible infection occurred in 3 to 4 days in all plants at
23° to 24°. No difference was obsen/ed at this time or later in the
roots which had formed, either at a high or at a low temperature. No
infection occurred at 31° to 32° after 8 to 10 days, but infection did
occur when the pots were removed to a lower temperature without rein-
oculation. This showed that the fungus had not been destroyed.
This test at least demonstrated that the increased growth of the host
at higher temperatures is not due to the overcoming of the effects of the
disease by increased root development, but is due to the inability of the
fungus to infect the host. It also tends to show that any resistance or
susceptibility at high or low temperatures which the roots have developed
is rapidly lost, since infection must have occurred within 24 to 36 hours
after changing from the extremes to the medium temperature. There is
room for a great deal of investigation, however, upon the intimate environ-
mental relations of host and parasite in this disease, and it is expected
that this subject v/ill be treated in more detail in another paper.
INFLUENCE OF ORGANIC MATTER IN THE SOIL
The content of organic matter and humus in the soil has been ascribed
by most investigators of tobacco rootrot as being a very influential
factor in determining the amount of disease. Practically all the writers
agree that the addition to the soil of vegetable matter in the form of green
manures or barnyard manures increases the extent of the disease.
Massee {ig) has gone so far as to state that the disease can not occur at
all in the total absence of organic matter, since he believes the fungus
must gain some stimulus v/hile living as a saprophyte before being able
to penetrate the host. Nearly all of these conclusions, however, have
74
Journal of Agricultural Research
Vol. XVII, No. 3
been based on observation rather than on experimental data. The
question is an important one from a practical standpoint. Will the
selection of soils low in organic matter or avoidance of the use of green
or barnyard manures materially aid in reducing the disease?
This is one of the most difficult problems to subject to experimental
test in such a way that wholly reliable conclusions can be drawn. It
illustrates equally well the fallacy of drawing far-reaching conclusions
from mere field observation. It is evident that the organic matter of
the soil has profound influence upon a large number of other factors
such as water-holding capacity, food supply, temperature, reaction,
texture, aeration, and saprophytic growth of microorganisms in the
soil. To eliminate all these factors, even in the most carefully con-
trolled experiments, is impossible. To judge of their relative impor-
tance in the results obtained in an experiment, however, on the basis of
the behavior of such factors from other experimental evidence, is quite
likely to yield fairly reliable results.
T.\BLE VIII. — Influence of amount of organic matter in soil en rootrof of tobacco
Rela-
tive
percent-
age of
leaf
mold to
ground
quartz
by
weight.
Experiment I.
Experiment II.
Series.
Loss on
ignition.
Average air-dry
weight per plant.
Ratio.
Loss on
ignition.
Average air-dry
weight per plant.
Unin-
fested
soil.
Infested
soil.
Unin-
fested
soil.
Infested-
soil.
Ratio.
I
O
lO
20
40
60
80
100
0 100
Per cent.
0. 22
1.65
Gm.
0.88
5-07
Gm.
0. 014
. 020
1:63
I :253
Per cent.
0. 24
1.05
2.8s
7-45
12.7
23-5
52-7
Gm.
3-35
3-35
4.90
5. 60
5-55
7. 00
7.27
Gm.
I : l^
T -8
2
I
42
60
32
29
85
3
4
I : 8
5-85
9.67
24. I
40.8
40. 8
5-93
7. II
6.17
8.01
7.19
• 052
.107
.179
.156
. 072
I : 114
I :66
I -35
I : 51
I : 99
I : 17
I • 18
c
6
I .4
7
8
• /-
o Leaf mold heated to 110° C. before infestation.
An attempt was therefore made to arrive at such conclusions by study-
ing the behavior of the disease in the purest ground quartz sand avail-
able and also in pure leaf mold, together with mixtures of the two in
various proportions (Table VIII). The chief difficulty met with at once
in such a combination is to obtain an approximately uniform supply
of plant food in these various media. The leaf mold was found to con-
tain sufficient plant food to support normal growth, though after the
third crop the plants showed potash hunger. To the pure quartz cul-
tures a nutrient solution sufficient for plant growiih was added, and
decreasing amounts were added to the various mixtures of sand and leaf
mold with a rough estimate of the amount of nutrient salts required.
Two pots of each series were then inoculated with T. hasicola and two
May IS. 1919 Influence of Soil Environment on Rootrot of Tobacco 75
left uninoculated as controls. All were then transplanted with the White
Burley variety from sterilized soil. It became evident at once that on
uninfested soil the rate of growth of the plants in the extremes of the
series was markedly different; the leaf mold was much more favorable
for growth than the sand with nutrient solution. Several tests were
run on these pots, and also on another series made up in a similar manner
(Experiment II, Table VIII). Most of the data concerning them exists
as notes and estimates of relative growth during the progress of the
experiment. Much reliance can not be placed on the weights, owing
to the large variation in fertility, although the ratios given of the growth
on infested to that on uninfested soil indicate the general trend of the
results.
The experiments, of which there were a considerable number, can not
profitably be discussed here in detail. Ownng to the variation in results
obtained in growth, much reliance was placed on estimates of actual
infection on the roots themselves, estimates difficult to express in figures.
After summarizing the results of all the tests run (nine in number), it
can be stated with considerable confidence that the importance of organic
matter in the soil is relatively small, so far as infection and severity of
the disease are concerned. It seems, however, that heavy infestation
is more rapid, and is more likely to be maintained through unfavorable
periods for the parasites in soils high in organic matter rather than in
those low in organic matter. Given a uniformly heavy inoculation of
the soil with endoconidia of T. hasicola, the rate and severity of infection
is apparently practically the same in pure sand as in the pure leaf mold.
Massee's conclusion (jp) that T. hasicola is a weak parasite and unable to
infect the host except in the presence of organic matter seems entirely
unwarranted. This has been further shown by infections obtained from
spore suspensions in pure water or spores alone placed directly upon roots
grown in a moist atmosphere. After the lapse of a considerable period
of time from the date of inoculation, however, it seems certain that
T. hasicola becomes more finnly established in pure leaf mold than it
does in pure sand, although this is apparently a difference of amount of
infestation and not one of virulence (Pi. 4, V-VI).
With regard to the various mixtures of sand and organic matter, the
conclusion seems justified that, so far as infection following inoculation
is concerned, it takes place wth equal ease in all (PI. 4, III-IV). The
development of infestation of the soil, however, has not given quite the
expected results. Doubling or tripling the content of organic matter
apparentl)^ has not increased infestation, and in some cases increasing
the ratio up to 80 parts of leaf mold seemed actually to reduce it. The
results, however, have not been sufficiently uniform in this respect to
warrant a final conclusion, and it is not certain that factors other than the
organic matter do not play a part here. Nevertheless, the fact that
increasing the organic content of the soil, two, four, six, and eight times,
76
Journal of Agricultural Research
Vol. XVII. No. 2
on the basis of percentage loss on ignition, has not consistently increased
the amount of disease seems to warrant the conclusion that the growers
have little to fear in the way of increased infestation of the soil as a result
of plowing under green manures or applying lo, 20, or 40 tons of manure
to the acre.
INFLUENCE OF THE CLAY AND SAND CONTENT OF THE SOIL
The value of sand or sandy soils in reducing the severity of the rootrot
and its increased severity in clay soils has been especially suggested by
Benincasa {2) and Gilbert (12).
A pot experiment with pure quartz sand and with the purest clay
obtainable was carried out with the hope of throwing more light upon
this subject. Theonly factor which it is desirable to vary in such an experi-
ment is the size of the soil particles. Although this is not practicable,
the relative proportion of sand and clay particles no doubt resembles soil
conditions equally well. Superior red clay was obtained from the sub-
station of the State experiment station located at Ashland, Wis. This
is a very "heavy" pure clay soil containing very little organic matter and
is low in fertility. The sand used was a medium to coarse ground quartz.
The mixtures of sand and clay made were those shown in Table IX.
Table IX. — Influence of relative amount of clay and sand on rootrot of tobacco
Mixttire.
Average air-dry weight (in grams) of plants.
Series.
Sand.
Clay.
Experiment IV.
Experiment V.
Uninfested
soil
Infested
soil.
Uninfested
soil.
Infested
soil.
All
None
I. 60
I. 0
.90
•75
I. 10
0. 40
•25
. 20
•25
• 15
0-75
.61
•52
• 50
•30
0. 26
2
3
4
5
Three-fourths. .. .
One-half
One-fourth
One-half
. 20
. 16
One-fotirth
Three-fourths ....
All
•25
■ 13
The experiment in this case was confronted with practically the same
difficulties and complications as occurred in the tests with organic mat-
ter. The clay soil alone, or in mixtures with sand, seemed to have an
"injurious" action upon the growth of tobacco which was not remediable
with nutrient solutions applied. The yield, therefore, was low in all
cases. Two pots of each series were inoculated with equal volumes of
spore suspension from cultures of T. hasicola on agar, which were thor-
oughly mixed with the soils. White Burley tobacco was then trans-
planted into them. The first experiment was started November 18, 19x6.
The result of this experiment was not recorded by weight, but some of
the crocks were photographed (Pi. 5, I), and serve to illustrate the re-
sults obtained. The conclusion drawn from this experim.ent was that
May 15, 1919 Infiiience of Soil Environment on Rooirot of Tobacco 77
sand was considerably more favorable to infection of roots with T.
basicola than was clay.
Another test with the Maryland broadleaf variety, started on October
i3» 19171 gave practically the same results; root examination showed
the greatest infection with sand, less with an admixture of one-fourth
clay, and almost none with one-half clay, and still less with larger
amounts of clay. However, in one pot heavy infection occurred at the
base of one plant, and the results were interfered with somewhat by
nematode injury.
All the soils were then resterilized with steam and two pots of each
series heavily infested with a suspension of young endoconidia of T.
basicola from agar slants. White Burley was again replanted into all
the crocks. The results obtained in this case differed somewhat from
the preceding, owing most probably to heavier infestation, infection
seemed to occur most rapidly and severely in the one-half and three-
fourths clay mixtures, but in a few days the plants in the infested series
were practically identical in appearance, and after about four weeks,
nearly all were killed.
These were now removed, and a more resistant variety, Connecticut
Havana, transplanted into the pots. After 18 days all these also were
practically equally diseased in the infested series, and were not quite
half the size of those in the uninfested series. After about six weeks'
growth these plants were cut and the air-dry weights determined as
shown in Table IX, Experiment IV. The pots were again planted to
tobacco with similar results (Table IX, Experiment V). The results of
the last experiments seem to indicate that in the presence of heavy infesta-
tion of the soil very little difference exists between clay and sand mix-
tures in the severity of infection of tobacco by T. basicola.
The results obtained in the first experiments are believed to be due to
the fact that the parasite found clay soils unfavorable for growth and
multiplication, and especially for penetration of mycelium as compared
with the sand, and therefore less actual infection occurred. On the
other hand, with heavy infestation sufficient spores were in close prox-
imity to the roots to produce good infection at once.
With respect to the persistence of T. basicola in soil, and its gradually
increasing infestation in spite of unfavorable conditions, it is believed
that clay soils may eventually be more injurious than sandy soils, but
the results seem to justify the conclusion that from the standpoint of
texture alone, the selection of loose sandy soils, or the use of clay soils
does not necessarily predetermine to any important degree what the
injury from T. basicola will be. It should be added that clay soils
draining more poorly and warming up more slowly, undoubtedly may
be considerably more haraiful than sandy soils, because of the influence
of saturated soils and low temperature upon the severity of the rootrot.
It is also believed that the tendency of the clay soils toward greater
compactness may also favor somewhat the occurrence of the disease.
78 Journal of Agricultural Research Voi. x\'t:i. no. 2
INFLUENCE OF SOIL FERTILITY
In the case of soil-infesting parasites which cause the loss of large por-
tions of the root system in such a way that it can no longer function nor-
mally for the benefit of the plant, it seemed probable that the quantity of
available plant food would influence growth in infested soil. It may be
supposed, for instance, that, if in a soil low in fertility one-half of the
roots are destroyed by disease, doubling the quantity of available plant
food would materially reduce the actual amount of damage done in yield
of crop. This is in accord with the conclusions of Briggs (j) on this
subject. On the other hand, there are the views expressed by many pa-
thologists with respect to various diseases, and also by Gilbert {12) and
others for the Thielavia rootrot, that an increase of fertilizers, especially
those of nitrogenous nature, renders the plant more susceptible to attack.
Aside from these theoretical conclusions, we are confronted with the
facts that the tobacco rootrot as it occurs in the field is not confined espe-
cially to soils in a low or high state of fertility, and that the application of
fertilizers, whether as barn manure or commercial fertilizer, seems to have
no marked effect upon the relative amount of disease, or on the growth of
the plants in infested soil in seasons favorable to rootrot. These con-
clusions are based on four years of fertilizer plot experiments carried on
at Edgerton, Wis., during the years igioto 1913, inclusive, the detailed
results of which can not be given here. These experiments were planned
to determine if it is possible to remedy the "worn-out" or "deteriorated"
condition of old tobacco soils by the use of fertilizers of various sorts,
although the full significance of T. hasicola as the cause of this condition
was not recognized at the time the experiments were started, and it was
not until after three years of failure to obtain any marked results with
fertilizer treatments on a wide variety of old soils, coupled with highly
beneficial results on soil new to tobacco that the importance of T. hasicola
in crop production was fully realized.
An experiment to determine more carefully the efifect of plant food
applied in the form of pure salts on the severity of the rootrot was carried
out in pot tests in the greenhouse during the winter of 1917-18. Twenty-
four 2 -gallon crocks with a drainage perforation at the base were each
filled with 10 kgm. of soil infested with T. hasicola from the old tobacco
field on the station farm. Twelve of these were now steriUzed by steam
to destroy the fungus. The cultures were divided into six series of four
pots each, two containing infested and two uninfested soil. A complete
fertilizer was made up from chemically pure salts according to a formula
used for nutrient water cultures, as follows:
Gtn.
Calcium nitrate 40
Potassium chlorid 10
Magnesium sulphate ." 10
Tribasic potassium sulpnate 10
May 15. 1919 Influence of Soil Environment on Rootrof of Tobacco 79
This fertilizer was added to each pot in each series and thoroughly mixed
with the soil in the amounts shown in Table X; the application ranged
from one which was considered only light, to one which was so heavy as
to decrease materially the yield. In these pots three successive crops
were grown; the first being the susceptible White Burley;the second the
semi- resistant Connecticut Havana; and the third the susceptible Mary-
land Broadleaf variety. The average air-dry weights for the plants in the
infested and the uninfested soil for the three crops are given in Table X:
Table X. — Influence of fertilizer on roof rot of tobacco
Series.
Fertilizer
added.
Gm.
3-5
7.0
14. o
20. o
56.0
Average air-dry weight (gm.) per plant.
First crop (White
Burley).
Uninfest-
ed soil.
Infested
soil.
0-95
2. 25
•57
• 17
Second crop (Con-
necticut Havana).
Uninfest-
ed soil.
7-97
12. 05
11-75
12. 17
II. 00
6. 10
Infested
soil.
4. 62
3-87
3.60
3-75
1.50
•59
Third crop (Mary-
land Broadleaf).
Uninfest
ed soil.
5.60
5-25
5-85
9.40
16.80
6. 42
Infested
soil.
O. 52
60
45
40
31
15
It may be observed at once that the uninfested soil responded to the
fertilizer treatment; the maximum yield for the first and second crops
was in the pots which received 14 gm. of fertilizer. Doubling the amount
of salt, however, decreased the yield, and quadrupling it acted ver>'
injuriously, presumably owing to increased concentration of the soil
solution. Very poor growth was made on the infested soil in all cases
(PI. 5, II-III). In the first experiment, though, the lowest application
of fertilizer apparently increased the growth, followed, however, by a
decrease at higher applications. The most striking fact was that the
most beneficial application on the uninfested soil showed no signs of
such beneficial action on the infested soil.
The results obtained with a more resistant variety as a second crop
are believed to be more representative. In this case there is a gradual
decrease in yield with the application of the nutrient salts on infested
soil; no increase from the application of 3.5 gm. of fertilizer occurred as
in the first crop grown. This influence of a light application of fertilizer
to infested soils is apparently in need of further investigation on a wide
range of soils and with different vaiieties. From a practical standpoint
it seems safe to conclude that fertilizer as such is wasted when applied
for tobacco on soils badly infested with T. basicola, and that it may, in
fact, do more harm than good.
8o Journal of Agricultural Research voi. xvii, no. 2
Theoretically we are concerned with the reasons for the injurious
action of nutrient salts on the growth of tobacco in infested soil, or
more directly, the increased severity of the disease in the presence of
increased supply of nutrient salts. A root system reduced by T. hasicola
evidently is not able to increase its functions in the presence of increased
fertility, in substitution for the lost roots; but it still seems as though
this should be possible, provided other factors do not interfere.
There are at least three plausible explanations for the observed be-
havior: (i) the increased concentration of the soil solution may favor
fungus growth; (2) increased food suppl}^ may favor increased suscepti-
bility to disease ; (3) the reduced root system in the presence of increased
concentration of soil solution may not have been able to furnish a suffi-
cient supply of water to the plants. The first explanation seems most
plausible and yet seemingly can not wiiolly account for the results
obtained. The second, that of increased susceptibility, seems least
plausible, since no one has yet satisfactorily shown that actual suscepti-
bility to disease is increased by increased fertilization. There is some
reason to suppose that the third hypothesis is a factor. In the unin-
fested soil a good illustration of the injurious action of high concentration
of soil solution on plant growth is found. This is explained as an osmotic
relation, the entrance of water to the plant being reduced, owing to the
high concentration of the soil solution. It seems probable, therefore,
that in the presence of a greatly reduced root system this condition
would be exaggerated with a resultant reduced growth. The water-
relation theory is strengthened by the observations on the relative time
and extent of wilting of plants on days favoring high transpiration.
Wilting of tobacco plants during periods of high transpiration on infested
soils and rapid recovery is quite common and indicates a delicate water
relation existing between the plant and the soil.
INFLUENCE OF COMPACTNESS OF SOIL
Field observations have seemed to indicate that in many instances where
the soil in infested fields is hard or compact, owing to poor preparation
of the soil or to other causes, the damage from T. basicola is more marked
than in loose soils. In fact, many farmers have been found who have
attributed poor yields to compact soils alone, when, as a matter of fact,
rootrot was undoubtedly the primary cause. It is, however, very difficult
to say just how much injury is due directly to the hard compact or baked
condition of the soil and how much is due to its influence on the progress
of the rootrot when present.
A simple experiment to determine this point was carried out. The
naturally infested soil from the field was carefully screened and mixed
in a relatively moist condition. A 6-inch clay pot was filled mth the soil
in as loose a condition as possible. This held 2,000 gm. Another pot
was then filled with the same soil, with as heavy tamping as possible.
May 15. 1919 Influence of Soil Environment on Rootrot of Tobacco 81
It contained 3,200 gm. Other pots were now filled with 2,900, 2,600,
and 2,300 gm. of soil, four pots of each degree of compactness being used.
Two pots of each series were steam-sterilized. Connecticut Havana
tobacco was then transplanted into them. To avoid any abnormal con-
ditions due to transplanting, especially in the compacted soils, which
were so hard that a knife could scarcely be inserted into them, a small
amount of soil was taken out of the center of each pot with a cork borer.
The hole was filled with uninfested loose soil, and the young plants were
transplanted into it so as to give all an equal chance to start.
The results were very interesting. The plants in the sterile com-
pacted soils did very much better than was expected, although the
loose soils were much more favorable to growth. In the infested soil,
however, the plants in the compacted soil made no growth whatever
(PI. 5, IV). Examination of the roots showed that in the loose infested
soil comparatively few lesions occurred, and the taproot was present,
while with increasing compactness the taproot was lost, and gradually
increasing numbers of lesions occurred.
The experiment has an important practical bearing on the preparation
and cultivation of tobacco soils infested with T. hasicola. Anything
which can be done to avoid, or remedy compactness or baking of soils
will no doubt lessen the disease even in badly infested soils. In other
words, soil in good tilth is less likely to be heavily damaged by T. hasicola
than soil in poor tilth.
TRANSPLANTING DISEASED SEEDLINGS
The influence of the use of diseased seedlings, for transplanting to the
field is not strictly an environmental feature of the problem. It relates,
however, to the amount of infestation in the soil and has a very impor-
tant practical bearing on results obtained under field conditions. It is,
furthermore, a point upon which some contradictory evidence has been
obtained by various experimenters, especially Benincasa (j), Gilbert
{12), and Clinton {8). The results already presented, especially in
regard to the influence of soil temperature, may serve to explain the
variation in results from year to year, or of the recovery of infected
transplanted seedlings. This is, however, apparently not the only
explanation. The writer has shown that varieties of tobacco and even
strains, vary in their resistance to rootrot (14). Transplanting healthy
plants to infested soil under favorable enviro^imental conditions for dis-
ease is shown to result in marked differences in yield of the different
types used.
It is known that the tendency of diseased plants is to send out new
roots to replace those lost by disease. Transplanting diseased plants
consists practically in infesting a small area of soil surrounding the base
of the plant with T. hasicola. Part of the new roots, especially those
82
Journal of Agricultural Research
Vol. XVII. No. 2
in the early stages of growth, must penetrate this infested soil before
reaching large areas of uninfested soil. The ability to resist the disease
will therefore determine roughly the number of roots and rapidity with
which they pass through this infested area and become established, and
should be roughly proportional to the resistance of the different varieties
under similar environmental conditions.
To determine the influence of transplanting varieties differing widely
in relative resistance to the rootrot, ii such varieties were sown in a
seed bed infested with T. basicola, and also in a sterilized bed as controls.
The relative resistance in the seed bed is about the same as that in the
field. The susceptible varieties especially did very poorly in the infested
beds, but most of them reached a sufficient size for transplanting. About
40 plants of each variety from infested soil and the same number from
uninfested soil, were then transplanted side by side in uninfested soil
(PI. 5, V). The results are given in Table XI in which the green weights
of 25 healthy plants and 25 diseased plants of each variety are shown,
together with the decreases in weight due to the use of infested seedlings.
It will be noted that a reduced yield occurred in all cases, but whereas
the disease was small in the case of varieties known to be resistant to
T. basicola it was relatively very high in those varieties which are sus-
ceptible. The results are not exactly comparable on this basis on account
of the difference in yield of varieties under normal conditions but serve
to illustrate the point in question.
Table XI. — Influence of transplanting diseased tobacco plants in uninfested soil
Variety.
White Burley
Maryland Broadleaf
Big Oronoco
Yellow Pryor
Pennsylvania Broadleaf
Kentucky Greenleaf. .. .
Italia Kentucky
"Pease Seed"
Ohio Seedleaf
"Northern Hybrid". . .
Brasile Beneventano. . .
Weight of 25 green plants.
Healthy
plants.
Gm.
66. 50
65. 00
57-75
59.00
82. 50
49-75
60. GO
49. GO
70. 50
65. GO
56.50
Infected
plants.
G7n.
25. 00
38.00
32- 50
35- 50
27.25
33- 50
49. 00
40. 00
51. 00
55- 00
53- 00
Decrease
due to
disease.
Gm.
41. 50
27. 00
25-25
23- 50
55-25
16. 25
II. 00
9. CO
19-50
10. 00
3-50
Gilbert {12) reports an experiment in which "Havana Broadleaf"
tobacco was used and in which the yield from infected and healthy
plants was practically identical. By "Havana" Broadleaf was meant,
it is presumed, the relatively resistant variety better known as "Con-
necticut Broadleaf. ' ' If environmental conditions were favorable for the
May IS. r9i9 Influence of Soil Environment on Rootrot of Tobacco 83
occurrence of disease, different results no doubt would have occurred had
Gilbert used a more susceptible variety.
Practical advice on the use of infected seedlings will, then, vary with
the variety used. Injected seedlings should never be used if it is possible
to avoid it, especially on soils which are not infested, since this will only
hasten the time when all the soil will become so thoroughly infested as
to make a change to newer soils necessary. On the other hand, it fre-
quently is necessary to risk infected plants, as others may be unobtain-
able. In such instances it is much less likely that serious injury will
result if the infection is on a resistant variety. Again, it should be re-
membered that infected seedlings of even a susceptible variety trans-
planted into a heavily infested soil may produce a normal crop under
favorable conditions, such as a very warm season and a relatively high
soil temperature persisting for a long time.
SUMMARY
(i) The rootrot of tobacco, caused by Thielavia hasicola, is marked by
the stunting of plants in various degrees due to a reduced root system.
The extent of the damage is determined in a large measure by the environ-
mental conditions surrounding the roots of the host.
(2) A study of these environmental conditions is essential to the proper
understanding of the occurrence and distribution of the disease in general
and local areas, and to good judgment in recommendation for control
measures.
(3) There seems to be no variation in the pathogenicity of the rootrot
fungus secured from different sources. The amount of disease is deter-
mined entirely by the susceptibility of the host, the amount of infection,
and the soil environmental factors surrounding the roots of the host.
(4) The factors especially studied were the amount of infestation in the
soil, the soil moisture, soil temperature, soil reaction, physical structure,
and fertility. An analysis of these f|actors separately as related to rootrot
frequently is very difficult, if not impossible. Under normal conditions
the end result in injur}^ by rootrot is the sum total of the favorable and
unfavorable action of these factors on the disease. Some of these factors
are much more important than others.
(5) Other factors aside, the extent of infection and injury from tobacco
rootrot is directly proportional to the amount of infestation of the soil.
(6) Rootrot is seemingly capable of developing in relatively dry soils.
Increasing the moisture content of the soil up to three-fourths of its
water-holding capacity does not materially increase rootrot. Saturated
soils are, however, considerably more favorable for the disease than
unsaturated ones.
(7) The temperature of the soil is undoubtedly the most important
factor determining the extent of the rootrot of tobacco, other factors
84 Journal of Agricultural Research voi. x\ai, No. 2
being equal. The most favorable temperature for the disease ranges
from 17° to 23° C. Below 15° the disease is less marked, and above
26° the severity is gradually reduced, until at about 29° or 30° it has
little or no influence. At 32° practically no infection occurs even in
the most heavily infested soils. Soil temperature records in the field
for four seasons indicate that occurrence of the disease under practical
conditions is determined primarily by soil temperature.
(8) The disease is checked by very high soil acidity. Heavy infection
can occur, however, in soils showing a considerable acid reaction. The
results depend a great deal upon the susceptibility of the variety used
in the test, the amount of infection, the soil temperature, and on other
factors. The results of tests of Wisconsin tobacco soils indicate that the
use of acid fertilizers will not reduce infection by T. basicola. Although
alkaline soils are more favorable to disease than very acid ones, the use
of lime on infested soils may not necessarily reduce the yield due to
increased infection from T. basicola.
(9) The amount of organic matter present or introduced into the soil
does not play a very important part in the amount of infection. High
organic matter content, however, no doubt favors increased infestation
and aids the fungus to persist in the soil. Where heavy inoculation is
made, infection apparently occurs more readily in pure sand than in the
presence of organic matter, but under conditions unfavorable for the
parasite the amount of infestation is more rapidly reduced in soils lacking
in organic matter.
(10) Clay soils as such seemingly are no more favorable for infection
than sand, and under certain conditions possibly less so. Clay may,
however, favor the persistence of the parasite in the soil, and may
actually favor infection because of increased danger of saturation with
water and because of the occurrence of lower temperatures than in sandy
soils.
(11) Increasing the fertility of infested soil by pure chemicals is likely
to cause increased stunting of growth rather than increased growth,
especially if too high a concentration of soil solution results. Fertilizers
applied to heavily infested soils under practical conditions seem to be
largely wasted except in seasons in which such high temperatures result
that the disease is held in check.
(12) Field observations and limited laboratory experiments seem to
show that infested soils when compacted are more favorable for the
disease than loose, open soil.
(13) Transplanting infected seedlings to an uninfested field is a bad
practice, although recovery from the disease may occur. Such recovery,
environmental conditions aside, is proportional to the resistance of the
type used.
May IS. 1919 Influence of Soil Environmenl on Rootrot of Tobacco 85
LITERATURE CITED
(i) Benincasa, M.
1902. RICERCHE SUI MEZZI PER DIKENDERE I SEMENZAI DI TABACCO DAI, " MAR.
CIUME RADICALE" CAUSATO DALLA THIELAVIA BASICOLA ZOPP. In Bol.
Tecnico Coltiv. Tab., ann. i, no. i, p. 24-33.
(3)
1911. I SEMENZAI DI SABBIA CONSIDERATI QUALE MEZZO DI DIFESA CONTRO H
MARCIUME RADICALS CAUSATO DALLA THIELAVIA BASICOLA ZOPP. In
Bol. Tecnico Coltiv. Tab., ann. 10, no. i, p. 3-22, 7 fig.
(3) Briggs, Lyman J.
1908. THE FIELD TREATMENT OF TOBACCO ROOT-ROT. U. S. Dept. AgT. BUT.
Plant Indus. Circ. 7, 8 p.
(4) BUTTARO.
1902. NOTIZIE SULL 'aNDAMENTO DELLA COLTIVAZIONI E CURS DEI TABACCH I
PONTECORVO. In Bol. Tecnico Coltiv. Tab., ann. i, no. 2, p. 94-95.
(5) Campbell, C.
1901. MORIA DELLE PIANTINE DI TABACCO NEI SEMENZAI. In Italia AgT., V. 38,
P- 540-542.
(6) Cappelluti-Altomare, G.
1902. i semenzai di tabacco e la "thielavia basicola zopf." in bol.
Tecnico Coltiv. Tab., ann. i, no. 3, p. 137-146.
(7) Chittenden, F. J.
1912. ON SOME PLANT DISEASES NEW TO, OR LITTLE KNOWN IN, BRITAIN. In
Jour. Roy. Hort. Soc, v. 37, p. 541-550. Bibliographical footnotes.
(8) CUNTON, G. p.
1907. ROOT ROT OF TOBACCO, THIELAVIA BASICOLA (b. & BR.) ZOPF. In Conn.
Agr. Exp. Sta. 30th Ann. Rpt. [i905]/o6, p. 342-368, illus., pi. 29-32.
(9) and Jenkins, E. H.
1906. ROOT-ROT OF TOBACCO. Conn. Agr. Exp. Sta. Bui. Immed. Inform. 4,
II p., 2 pi.
(10) Comes, Orazio.
I913. DELLA RESISTENZA DEI FRUMENTI ALLE RUGGINI STATO ATTUALE DELLA
QUISTIONE E PROVVEDIMENTI. In Atti R. 1st. Incorag. Napoli, v. 64,
1912, p. 419-441. Letturaturae note, p. 437-440. Abstract in Intemsit.
Inst. Agr. Rome, Mo. Bui. Agr. Intell. and Plant Diseases, year 4,
no. 7, p. 1117-1119. 1913. Original not seen.
(11) Galloway, B. T.
1903. COMMERCIAL VIOLET CULTURE. Ed. 2, 239 p., front., illus. New York.
(12) 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. 44-4S.
(13) Johnson, James.
1916. host plants of THIELAVIA BASICOLA. In Jour. Agr. Research, v. 7, no.
6, p. 2S9-300, pi. 18-19.
(14)
1916. RESISTANCE IN TOBACCO TO THE ROOT-ROT DISEASE. In Phytopathology,
V. 6, no. 2, p. 167-181, 6 fig.
(15) and Hartm.\n, R. E.
1918. INFLUENCE OF SOIL TEMPERATURE ON THIELAVIA ROOT-ROT. (Abstract.)
In Phytopathology, v. 8, no. 2, p. 77.
(16) Jones, L. R.
1917. SOIL TEMPERATURES AS A FACTOR IN PHYTOPATHOLOGY. In Plant
World, V. 20, p. 229-237, illus. Literature cited, p. 236-237.
86 Journal of Agricultural Research voi. xvii, No. 2
(17) KiiLEBREW, 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.
(18) Martinazzoli, G.
191 1. AU:UNE NOTIZIE SUI SEMENZAI Dl POZZOLANA. In Bol. Tecnico Coltiv.
Tab., ann. 10, no. 6, p. 367-369.
(19) Masses, G. E.
1912. A DISEASE OP SWEET PEAS, ASTERS AND OTHER PLANTS. In Roy. Gard.
Kew, Bui. Misc. Inform., 1912, no. i, p. 44-52.
(20) Peguon, V.
1897. MARCIUME RADICALS DELLE PIANTINE DI TABACCO CAUSATO DALLA
THIELAVIA BASICOLA, ZOPF. In Cent. Bakt. [etc.], Abt. 2, Bd. 3,
No. 21/22, p. 580-584. Bibliografia, p. 584.
(21) Reddick, Donald.
1913. THE DISEASES OF THE VIOLET. In Trans. Mass. Hort. Soc, 1913, p.
85-102 pi. 1-2.
(22) RoSENBAUM, Joseph.
I912. INFECTION EXPERIMENTS WITH THIELAVIA BASICOLA ON GINSENG. In
Phytopathology, v. 2, no. 5, p. 191-196, pi. 18-19.
(23) SORAUER. P.
1895. tTBER DIE WURZELBRAUNE DER CYCLAMEN. In Ztschr. Pflanzcnkrank.,
Bd. 5, Heft I, p. 18-20.
(24) TaTham, William.
1800. AN HISTORICAL AND PRACTICAL ESSAY ON THE CULTURE AND COMMERCE
OF TOBACCO. 330 p., 4 pi. (2 col.). London.
(25) TiSDALE, W. H.
1916. RELATION OF SOIL TEMPERATURE TO INFECTION OF FLAX BY FUSARIUM
LINI. In Phytopathology, v. 6, no. 5, p. 412-413.
(26) Truog, E.
I915. A NEW TEST FOR SOIL ACIDITY. Wis. AgT. Exp. Sta. Bul. 249, 16 p.,
3 fig-. I pi-
(27) Whetzel, H. H., and OsnEr, George.
I9IO. THE FIBER ROT OF GINSENG AND ITS CONTROL. In SpCC. CropS, n. S. V.
9, no. 97, p. 411-416, 4 fig.
PLATE I
I. — Influence of amount of infestation on injury from tobacco rootrot: A, All unin-
fested soil;" B, three-fourths uninfested soil; C, one-half uninfested soil; D, one-
fourth uninfested soil; E, all infested soil.
II, III. — Influence of moisture content of soil on the amount of injury done by the
tobacco rootrot; II, infested soil; III, iminfested soil (control series) —
I A, one-fourth saturation infested soil;
2 A, one-half saturation infested soil ;
3 A, three-fourths saturation infested soil ;
4A, full saturation infested soil;
iB, one-fourth saturation uninfested soil;
2B, one-half saturation uninfested soil;
3B, three-fourths saturation uninfested soil;
4B, full saturation uninfested soil.
IV. — Influence of soil temperature on the growth of tobacco in infested soil (jars
to left of temperature labels) and in uninfested soil (jars to right of temperature labels)
at temperatures of approximately 13°, 17°, 23°, 26°, and 36° C.
Influence of Soil En vironment on Rootrot of Tobacco
Plate I
J •
pi^
I-
■ — " ^
D
Ift ^ • 2A ■ 3A
w
V^'^B Wfn"^^ H^J'cf^ H
26c
Journal of Agricultural Research
4A
Vol. XVII, No. 2
Influence of Soil Environment on Rootrot of Tobacco
Plate 2
Journal of Agricultural Research
Vol. XVII. No. 2
PLATE 2
I . — Soil temperature tanks used in the temperature experiments. The water surface
was later covered with sheet metal and asbestos board.
II, III. — Influence of soil temperature on the growth of tobacco:
lA, infested soil, i7°-i8° C.
2 A, infested soil, 2o°-2i° C.
3A, infested soil, 23°-24° C;
4A, infested soil, 25°-26° C.
5A, infested soil, 28°-29° C;
6A, infested soil, 3i°-32° C;
iB, uninfested soil, i7°-i8° C;
2B, uninfested soil, 2o°-2i° C;
3B, uninfested soil, 23°-24° C;
4B, uninfested soil, 25°-26° C;
5B, uninfested soil, 28°-29° C;
6B, uninfested soil, 31 "-32 ° C.
IV. — Influence of different soil temperatures on root development:
lA, uninfested soil, i7°-i8° C;
iB, infested soil, i7°-i8° C;
2 A, uninfested soil, 2o°-2i° C;
2B, infested soil, 20°-2i° C;
3A, uninfested soil, 23°-24° C;
3B, infested soil, 23°-24° C;
4A, uninfested soil, 25°-26° C;
4B, infested soil, 25°-26° C;
5A, uninfested soil, 28°-29° C;
5B, infested soil, 28°-29° C;
6A, uninfested soil, 3i°-32° C;
6B, infested soil, 3i°-32° C.
PLATE 3
Influence of high (30" C.) and low (20° C.) soil-temperature on recovery of plants in
infested soil. Both plants were taken from the field where they had made very little
growth during the season and placed in temperature control tanks, the roots at 30"
being like those at 20° at the beginning of the experiment.
Influence of Soil Environment on Rootrot of Tobacco
Plate 3
Journal of Agricultural Research
Vol. XVII, No. 2
Influence of Soil Environment on Rootrot of Tobacco
Plate 4
Journai of Agricultural Research
Vol. XVII, No. 2
PLATE 4
I, II.— Influence of soil reaction on extent of damage by tobacco rootrot: I, Infested
soil; II, tininfested soil —
lA, infested soil, lime requirement 9.38 tons per acre;
2A, infested soil, lime requirement 7.19 tons per acre;
3A, infested soil, lime requirement 4.60 tons per acre;
4A, infested soil, lime requirement 2.62 tons per acre;
SA, infested soil, lime requirement 0.72 ton per acre;
6A, infested soil, slightly alkaline;
7A, infested soil, strongly alkaline;
iB, uninfested soil, lime requirement 9.38 tons per acre;
2B, uninfested soil, lime requirement 7.19 tons per acre;
3B, uninfested soil, lime requirement 4.60 tons per acre;
4B, uninfested soil, lime requirement 2.62 tons per acre;
5B, uninfested soil, lime requirement 0.72 ton per acre;
6B, uninfested soil, slightly alkaline;
7B, uninfested soil, strongly alkaline.
Ill, IV.— Influence of the amount of organic matter in the soil on injury by tobacco
rootrot: III, 1A-6A, Influence, of gradually increasing amounts of organic matter
in infested soil from lA, no organic matter, to 6A, all leaf mold.
Planted soon after heavy inoculation.
IV.— 1B-6B, Influence of gradually increasing amounts of organic matter in unin-
fested soil from iB, no organic matter, to 6B, all leaf mold (control series).
Planted soon after heavy inoculation.
V, VI.— Influence of the amount of organic matter in the soil on injury by tobacco
rootrot: V, 1A-5A, Influence of gradually increasing amounts of organic matter in
uninfested soil from lA, no organic matter, to 5 A, all leaf mold (control series) ;
VI, 1B-5B, Influence of gradually increasing amounts of organic matter in infested
soil from iB, no organic matter, to 5B, all leaf mold.
Planted some months after moderate inoculation.
PLATE 5
I. — Influence of relative amount of sand and clay on lohacco rootrot: A, unin-
fested series: B, infested series —
I A, uninfested soil, three-fourths clay ajid one-fourth siuTid;
1 B, infested soil, three-fourths clay and'one-fourth sand;
2 A, uninfested soil, one-half cla}-- and one-half sand;
2B, infested soil, one-half clay and one-half sand;
3 A, uninfested soil, three-fourths sand;
3B, infested soil, three-fourths sand;
4A, uninfested soil, all sand;
4B, infested soil, all sand.
II, III. — Influence of soil fertility on amount of tobacco rootrot: II, infested
series; III, uninfested series —
I A, infested soil, no treatment;
2A, infested soil, 3.5 gms. of nutrient salts;
3A, infested soil, 7.0 gms. of nutrient salts;
4A, infested soil, 14.00 gms. of nutrient salts;
5A, im'ested soil, 28 gms. of nutrient salts;
6A, infested soil, 56 gms. of nutrient salts.
Note increasing injur\'^ from nutrient salts beginning at pot 3A.
iB, uninfested soil, no treatment; »
2B, uninfested soil, 3.5 gms. of nutrient salts;
3B, uninfested soil, 7.0 gms. of nutrient salts;
4B, uninfested soil, 14.00 gms. of nutrient salts;
5B, tminfested soil, 28 gms. of nutrient salts;
6B, uninfested soil, 56 gms. of nutrient salts.
Note injury from nutrient salts in pots 5B and 6B.
IV. — Relation of compactness of soil to injury caused by Thielavia basicola:
lA, infested soil, loosely packed;
iB, uninfested soil, loosely packed;
2A, infested soil, very compact;
2B, uninfested soil, very compact.
V. — Influence of transplanting infected seedlings in healthy soil:
A, Pennsylvania Broadleaf infected geedlings;
B, Pennsylvania Broadleaf healthy seedlings;
C, White Eurley infected seedlings;
D, WTiite Burley healthy seedlings;
E, Northern Hybrid (a resistant type) infected seedlings,
F, Northern Hybrid (a resistant type) healthy seedlings.
Influence of Soil Environment on Rootrot of Tobacco
Plate 5
n __.
— ~—
Journal of Agricultural Research
Vol. XVII, No. 2
Influence of Soil Environment on Rootrot of Tobacco
Plate 6
June / 3 5 7 9 W I3 15 I7 19 21 23 25 27 29 3/
36
34
32
30
ZQ
'^^
i^20
16
"■
^
^
^
1 1
1915
1916
1917
IQIR
1
'n
i
I
1
"'
1
^
1
1
!
1
1
1
1
/
/
>
1
\ ^
;
u
1
i
\
1
.\-<
.• /
l' 1
i /
/
/
Ov^
V
\
!\
\
1
/
X
••'l
1
i
'
y
^
;;
V
\
/
/
\
1
V
/'
\
>
,'
{
/
1
1
\
y
\
/,
^■-
;
\
1
/
1
''•
f'
^
k
'
-r
j
\
//
—
\,
/\
/
/
\
1
1
\
i
\
■J-
/
\
/
>-.f
1
\
—
-.,
10
JUN£ I 3 5 7 9 II IS J5 17 19 21 2Z 25 27 29 31
32
30
2d
1/) Vg /26
^S24
10
June » 3 579 n 13 n n 19 21 23 zs 27 29 31
W
r
1
1
<
/
\
"m
1
-•
/
/
1^^
/
/
^
\
_.
^.
,•'/
\
\\
^
;
V
' ■•
f'
\
...
N
n
\
1
\
1
\
/
v
r
.-V
/
'
/
x'
\
l
^
t
>
'
1
1
\
V
'
/
\
/
1
\
/
—
1 1
'">
/
/
'^
1
...
~i
'
'"
i
n
n
f
\
\/
y
f'
\
\
>
/
\
V
»,
•^
^\^
i
C
4
/"
\
• r"
r-
--
,
J
f
'
"
\
f-
■■-
■?
\
■/
i
N
'
- -.
f
)
V
^/
"
^
/
-J
--
/
\
f
')
-1
r
Journal of Agricultural Research
Vol. XVII, No. 2
PLATE 6
Soil temperature graphs for the month of June, 1915-1918, inclusive, at depths of
2, 4, and 8 inches.
PLATE 7
Soil temperature graphs for the month of July, 1915-1918, inclusive, at depths of
2, 4, and 8 inches.
Influence of Soil Environment on Rootrot of Tobacco
Plate 7
JuLr\ 3 5 7 9 II 13 15 17 l9 21 23 25 27 29 31
3b
34,
32
30
26
Z6
^24
f^^-2
^^
$>:-?o
^ 16
Q.J3 M
/
/
...■
'
/
\
\
/
>
1
1
1
\
^ .'
•-.'
^>
v'
y
...
"■
T-
s.
.i-r^
-.'■•
--
/
/
,
/
\.'
/'
\
/j
' \
••■J
^
/
/
\
__
--
/
/
\
7\
N
^'
I
\
\
/
/
\')
V
.'/
s
l^
/
(
\
/
/
\
;
'
\
\
1
1
\i
A
■ \
V,
.'/■
""•
s,
/
\
, *
/
/
;
f
\
/
\
1/
\
\
1
■\
/
'M
/
X
/
\
/
\
s
^
/
\
/
\
V
/
\
/
\
/
^
y'
^
/
\
July i 3579/1/3/5/7 /9 21 21 zs 21 29 31
32
30
26 h
26
fi^24
S k
1^/6
"-12/6
10
Jl/LYl 3
/
..-
■
•'
1'
--
'"
«
....
••/
....
/
,,
/
V
.■1
■,\
/
...
1
/
>
~/i
'•"
'■~
— -
-•'
'■'\
—
r-
^
v
',
,<f
y
^
/
/
.
'
^^
\
/
'"
/
,
<
>\
\
/
^
f1\
\
\
\
L
/.
\
/
K
/
/
'
1
\
V
>■
/
\
,^
-f
, ^
•,1
/
\
z
,
^1 —
f
*\
\'
■^
/
^
\
•
s
/
\
^
'
\
A
/
\^
/
1
\
\,
/
\
/
Vw
/
*s
// 13 15 17 19 21 23 25 27 29 21
ujUj
/
\.-
>'
y
-.
••..
*
-■/
''
Y
■v
/
.^
...
'_.
J.v
•
--
''
(
\
s\
/
>.
-/
y
~~-
■^.
>..
...
-•
k//
\
V
,-\
^-
_..
■>
/
^
• -
'".
/
\
y
\
^
— -
N
/
>
ITI
...
1 \
\
/
\
/
1
,,'
\
/
fOfi«
1913
1916
1917
"
IS
»lt
\
Journal of Agricultural Research
Vol. XVII, No. 2
Influence of Soil Environment en Rootrot of Tobacco
Plate 8
AuGUiTl 3 5 7 9 // 13 15 17 19 21 ^3 25 27 29 31
3b
34
32
30
28
2b
20
1^/6
*-v^
-^ ''J
Ih
-tkj
M
UjUj
titri
12
,
__.-
/
\
\
I
•
^-
--
--
^
/
m
\
\
1
\
/
I
\
/
\
\
—
1
1
'
/
\
\
1
1
\
/
>>>
^
/
\
\
\
/
[
/
\
\
1
^•'
\;
^':
^
;
1
\
■'V
1'
1
■\-
/
\
\
1
\
,
>^
/
N
/
\
/
\
'\
I
"r._
\
/..
1
...
1
I
\
'
\
\
/
;
\
/
;'
,'
1 1
I 1
\
—
'
\
\
I
\
\
/
/
\
\
/
\
/
\
I 1
I 1
[
X
~
\,
/
/
\
f
\...
i
^
\
l|
/
^
/
\
/
-\
—
\
/
/
\
1
\
/
\
! •
^/
to
AuGl/^Tt 3 5 7 9 II 13 15 n 19 21 23 25 27 29 31
32
30
26
^'^/^
10
AUCi/bTl 3 5 7 9 II 13 15 17 l9 21 23 25 27 29 31
2b
/
\
,
\
/
/
\
/
\
/
\
—
I
/
'•-
--
-/-
I
\
,■
\
/
\
/
—
\'
\
\
/
V
l-
V
-/■
/
V
/
v
.>s
/
v
'
'^
p-
\
1
\
**
\
-A-
/
\
/■
-v.
/'
\
--
'
\
1
_..
\
1
\
^'j
\
\
_y
J
■•A
~
\
/
/
\
\
1 \
I
\
...
1
1
1
/
\
/
—
/
\
/
--\
'^
\
/
\
N
\
/
s
I
\
~^
/
•
/
\
,.-
/•.
/
'."■
~-^
/
^^
/■'
—
v
^
-.,
/
■~
K
- >
,
/
'
\
,.•
-.,
,.
/'
A
-■
^
'\
\
\
N
"
/
X
\
N
•
-
'
^
_.
/'
\
N
/
y
N
—
/
1919
1916 .
1917
"~
~
-
19
1(
1
20
Id
76
/4
12
10
Journal of Agricultural Research
Vol. XVII, No. 2
Vol. XVII JUNE 16, 1919 No. 3
JOURNAL OF
AGRICULTURAL
RESEARCH
CONXEMSTTS
Page
Relation of Sulphates to Plant Growth and Composition - 87
H. G. MILLER
( Contribution from Oregon Agricultural Experiment Station )
Relation of Weather to Fruitfulness in the Plum - - 103
M. J. DORSEY
( Contribution from Minnesota Agricultural Experiment Station )
Structure of the Maize Ear as Indicated in Zea-Euchlaena
Hybrids •• 127
G. N. COLLINS
(Contribution from Bureau of Plant Industry)
PUBLISHED BY AUTHORITY OF THE SECRETARY OF AGRICULTURE,
WITH THE COOPERATION OF THE ASSOCIATION OF AMERICAN
AGRICULTURAL COLLEGES AND EXPERIMENT STATIONS
WASHINGTON, T>, C.
WASHINGTON : GOVERNMENT PRINTING OFFICE : 1(18
EDITORIAL COMMITTEE OF THE
UNITED STATES DEPARTMENT OF AGRICULTURE AND
THE ASSOCIATION OF AMERICAN AGRICULTURAL
COLLEGES AND EXPERIMENT STATIONS
FOR THE DEPARTMENT
KARL F. KELLERMAN, Chairman
Physiologist and Associate Chief, Bureau
of Plant Industry
EDWIN W. ALLEN
Chief, Office of Experiment Stations
CHARLES L. MARLATT
Entomologist and Assistant Chief, Bureau
of Entomology
FOR THE ASSOCIATION
H. P. ARMSBY
Director, Institute of Animal Nutrition, The
Pennsylvania State College
J. G. LIPMAN
Director, New Jersey Agricultural Experiment
Station, Rutgers College
W. A. RILEY
Entomologist and Chief, Division of Ento-
mology and Economic Zoology, Agricul-
tural Experiment Station of the University
of Minnesota
All correspondence regarding articles from the Department of Agriculture should be
addressed to Karl F. Kellerman, Journal of Agricultiu-al Research, Washington, D. C.
All correspondence regarding articles from State Experiment Stations should be
addressed to H. P. Armsby, Institute of Animal Nutrition, State College, Pa.
.|0M£ OFAGBIQinmffiSEARCH
Vol. XVII Washington, D. C, June i6, 1919 No. 3
RELATION OF SULPHATES TO PLANT GROWTH AND
COMPOSITION
By H. G. Miller,
Assistant Chemist, Oregon Agricultural Experiment Station
HISTORICAL REVIEW
The oxidation of sulphur in the soil and the relation of the products
formed is plant growth, bacterial development and activity, and to the
release of other forms of plant food have been reported upon by many
investigators. In reviewing the work, many of the writers have reported
beneficial results from the use of sulphur fertilizers, especially with those
plants high in protein and other sulphur-containing compounds. Various
views are given as to how sulphur functions in producing these increased
yields. Analysis of soils reported by Hart and Peterson (11),^ Shedd
(22, 23) , Brown and Kellogg (7) , and Swanson and Miller {26) , show a lower
sulphur content in the cultivated soil as compared to the phosphorus,
while many of the cultivated plants show a larger content of sulphur
than phosphorus. These results indicate that sulphur v/ould become a
limiting factor before phosphorus.
It is generally concluded that sulphur to be available for plant food
must be in the ^Iphate form, so that a soil having a high sulphur
content may not necessarily supply enough sulphate sulphur for maxi-
mum growth. Brown and Kellogg (6) have shown that different soils
have unlike sulphofying powers and some of the factors influencing the
change of elemental sulphur and sulphides to sulphate form. In
lysimeter experiments at Cornell, Lyon and Bizzell (77) report that
the sulphate sulphur in the drainage water was from three to six times
as great as in the crops and the sulphur content of the drainage water
from the unplanted soil was about equal to the sulphur content of the
crop and drainage water from the planted soil. Swanson and Miller (26)
conclude from an investigation on sulphur in Kansas soils that —
the loss in sulphur due to the amount taken up by the crop is insignificant as compared
with the total amount which has disappeared from the soil. This means that sulpho-
fication has been in excess of the needs of the crop, and the sulphates produced have
been leached out of the ground.
' Reference is made by number (italic) to "Literature cited," pp. 100-102.
Journal of Agricultural Research, (87) Vol. XVII, No. 3
Washington, D. C. June 16, 1919
rw Key No. Oreg.-4
88 Journal of Agricultural Research voi. xvii, No. 3
They report no increased yield when sulphur was applied as potassium
sulphate (KjSOJ, but state that the loss of sulphur can not continue
without affecting crop yields. Hart and Peterson (11) calculate from
data obtained at the Rothamsted Experiment Station, Harpenden,
England, and the Wisconsin Experiment Station that the loss of sulphur
in drainage water is three times that brought to an acre surface from
the atmosphere.
Results of the investigations mentioned above show that the soil is
capable of producing sulphate sulphur and that there is a tremendous
loss of such sulphur in the drainage water. In certain cases no benefxial
results have been obtained from sulphur fertilizers, which is not sur-
prising, but in many instances sulphur application has caused increased
yields. No doubt, in many soils, if the supply of sulphate sulphur
formed was retained for plant food, they would not respond to sulphate
treatment but the continual loss of sulphate sulphur and the large
amount needed by some plants require that additional sulphate sulphur
be present during the growing period to obtain better growth. A soil
with a high sulphur content may not supply enough sulphate sulphur in
comparison to the other plant food to obtain the maximum growth,
while conditions in a soil of lower sulphur content may be such as to
supply an adequate amount of sulphate sulphur.
In addition to sulphur acting as a plant food, several other explana-
tions have been given as to its action in the soil. Certain investigators,
Bemhard (j) , Chancrin and Desriot (8) , say that it functions as a partial
sterilizer, others that the sulphuric acid produced acts upon the mineral
matter of the soil, rendering it more soluble. Lipman (15, 16) and his
coworkers have shown (j) that the oxidation of sulphur in sand and soils
has acted upon the raw-rock-phosphate so as to increase the water
and ammonium-citrate-soluble phosphorus (2) ; that the formation of
sulphate sulphur paralleled the increase of available phosphoric acid (j) ;
and that the sulphur-floats-soil compost could be employed as a sub-
stitute for acid phosphate for plant growth. Brown and Gwinn (5)
have found that the addition of sulphur to the soil increased the avail-
ability of raw-rockphosphate, the gain being greatest where manure and
sulphur were used together. McLean (18) in a number of experiments
has shown the conditions which are more favorable for the oxidation
of sulphur by microorganisms and production of available phosphorus.
Lipman (14) suggests that the sulphuric acid may act in making alkali
areas productive by converting sodium carbonate into sodium sulphate.
The favorable results obtained by adding gypsum have often been
attributed to the calcium liberating potassium, but the experiments of
Hart and Tottingham (12) show that a complete fertilizer plus calcium
sulphate gave increased yields over those obtained with a complete fer-
tilizer containing potassium chlorid, and that here the action of calcium
sulphate must have been direct.
June i6, 1919 Sulphates Affecting Plant Growth and Composition
89
The action of sulphates upon soil bacteria has also been studied. Fred
and Hart (9) have shown small increases in the number of soil bacteria
and a slight increase in ammonification and carbon-dioxid evolution by
adding certain sulphates to the soil. A bacteriological investigation by
Janicaud (ij) indicated that sulphur had a favorable influence on the
development of bacteria in the soil. Pitz (20) reports no marked effect
on the number of bacteria found on agar plates, but he does report an
increase in legume bacteria from the use of calcium sulphate. Elemental
sulphur caused a decrease in the total number of bacteria that grow on
agar plates, but an increase in ammonification was accompanied by a
parallel decrease in nitrate formation. Boullanger and Dugardin (4)
state that the presence of small amounts of sulphur materially increases
the activity of the ammonifying bacteria. Ames and Richmond {2) con-
clude from experimenting on relation of sulphofication to nitrogen trans-
formation that the increase in ammonia which accompanied the decrease
in yield of nitrates when sulphur was oxidized can not be considered as
indicative of sulphofication having exerted a stimulating effect on am-
monification. A deficiency of base in the soil allows the ammonia
formed to neutralize the sulphur and it remains as ammonium sulphate.
Certain soils in Oregon have responded greatly to the use of sulphur
and its compounds. In this locality greater crop production in many
cases has resulted from using gypsum than in using lime. Reimer {21)
of the Southern Oregon Experiment Station has obtained large increases
in alfalfa yield by the use of elemental sulphur. This marked effect
from the use of sulphur fertilizers suggested that it would be of interest
and practical value to carry on some greenhouse experiments in order to
study the effect of sulphur on early growth and composition of the
plants, and also to attempt to determine whether it acted directly in
supplying the plant with food or functioned in some other way.
PLAN AND OBJECT OF THE EXPERIMENT
For this work three Oregon soils, classified as a beaverdam, a Med-
ford loam, and an antelope-clay-adobe, and designated as A, B, and C,
respectively, in the tables, were chosen. The first was taken mainly for
its high sulphur content, the second one because it did not respond to
sulphur treatment in the field, while the third did respond to elemental
sulphur fertilizer. The results obtained on analysis of these soils are
given in Table I. The plants chosen were red clover, oats, and rape.
Table I. — Percentage of constituents foun in soils
Constituent.
Potassium oxid (KjO)
Phosphorus pentoxid (P2O5).
Sulphur
Calcium oxid (CaO) «
Calcium carbonate (CaCOg)"
Soil A. 6 SoilB.6
349
183
1.25
. 160
•034
4. 02
None.
Soil C.6
o. 60
. 027
3-24
. II
o Taken from analyses of soil sample when the field experiments were carried on.
'Soil A=beaverdam; soil B=Medford loam; soil C= antelope-clay-adobe.
90
Journal of Agricultural Research
Vol. XVII, No. 3
It was decided to apply sulphur in the form of sodium sulphate
(Na2S04), calcium sulphate (CaS04), and elemental sulphur. The ele-
mental sulphur was mixed with the soil at the time of planting, but the
sulphates were added daily in the form of a solution. This daily addi-
tion of sulphates maintained a continuous supply of sulphates for the
plant and it was thought that through the growth of the plants in the
pots receiving the different forms of sulphur one could ascertain whether
the elemental sulphur was able to supply the necessary sulphate. In
order to eliminate nitrogen as a limiting factor, sodium nitrate (NaNOg)
in solution was added daily. By keeping up an available supply of
nitrogen, a study could be made of the influence of sulphur fertilizer on
the amount of nitrogen taken up by the plant. Nitrogen and sulphur
enter into the composition of proteins and mustard oils so that an abund-
ant supply of sulphates and nitrates in the plant may increase the elabo-
ration of those organic compounds containing sulphur and nitrogen.
In addition to growing the plants on soil, they were also grown in sand
pots receiving extract from the soil plus any nutrient that was added to
the soil. For example, for a soil receiving a solution containing calcium
sulphate and sodium nitrate, there was a corresponding sand pot re-
ceiving a water extract of the soil plus calcium sulphate and sodium
nitrate. The foUowdng statement shows what each pot growing clover
received :
SOIL
Pot I :
Calcium sulphate.
Sodium nitrate.
Pot 2:
Sodium sulphate.
Sodium nitrate.
Pots:
Sodium nitrate.
Calcium carbonate.
Sulphur.
Pot 7:
Soil solution.
Calcium sulphate.
Sodium nitrate.
Pot 8:
Soil solution.
Sodium sulphate.
Sodium nitrate.
Pot 4: Sodium nitrate.
Pots:
Sodium nitrate.
Calcium carbonate.
Pot 6: No fertilizer.
SAND
Pot 9:
Soil solution.
Sodium nitrate.
Calcium carbonate.
Sulphur.
Pot 10:
Soil solution.
Sodium nitrate.
Pot 4 (PI. 9, A, B, and C) is a control to compare with i and 2, and
5 is a control on pot 3. Pot 10 receives no sulphur except that in the
original soil extract. The pots growing oats and rape were treated in
the same way, and this was repeated for each soil, making in all 90 pots
(PI. 9-12).
June i6, 1919 Sulphates Affecting Plant Growth and Composition 91
Hall, Brenchley, and Underwood (lo) at the Rothamsted Experiment
Station, in some experimental work in support of the theory of the direct
nutrition of plants by fertilizers, found that wheat and barley showed
parallel growth in the soil, in the soil extracts, and in artificial solutions
of the same phosphorus-pentoxid (P2O5) and potassium-oxid (KjO) con-
tent. The soil solutions corresponded to the natural drainage water,
depending upon the past fertilizing treatment and present composition of
the soil. The growth in extracts from poorly fertilized soils could be
made equal by direct addition of suitable phosphate and potassium
salts. Boiling did not affect the nutritive value of the solutions, and the
diffusion of the nutrient solution over particles of sand did not interfere
with the growth, although proper aeration of the roots was found to be
essential.
By growing the plants on sand the possibility of the sulphates acting
on minerals in the soil would be eliminated, and boiling the soil extract
would destroy the bacteria originally in the solution. So in these experi-
ments, if an increase in growth is observed in the soil pots from the appli-
cation of sulphur fertilizer, and a corresponding increase is also obtained
in the sand pots containing soil extract to which sulphur had been added,
this same order of growth in the soil and sand would indicate in all
probability that sulphur had acted directly in promoting the growth.
The object of this experiment was to make a study of the influence of
an available supply of sulphates on the early growth of the plants and
determine whether the elemental sulphur was capable also of supplying
the necessary sulphates; to see what effect sulphates would have on the
nitrogen content of the plant and if beneficial results are obtained whether
the sulphur acted directly as a plant food in producing them.
EXPERIMENTAL WORK
The pots used were ordinary clay flowerpots which had been paraffined
on the inside, and each contained about 700 gm. of soil. The sand was
of a fine quality, obtained from Eimer and Amend. It was washed with
dilute hydorchloric acid until no potassium, phorphorus, or sulphur was
detected in the acid extract. Larger pots were not used on account of
the beginning of this type of experiment, the number of pots needed and
the individual attention required. The growing period was about two
months with the exception of the oats which were allowed to ripen. The
seed was sown on March 15, and the work was carried on in the green-
house. At the end of a month the plants were thinned out so that there
were 10 clover plants, 6 oat plants, and 3 rape plants per pot. At this
time the sulphur and nitrogen were determined in the clover plants
taken from soil A, The clover and rape were cut on June i, the dry
weight taken, and the total sulphur and nitrogen determined. Twenty-
five cc. of nutrient solutions containing compounds as given in the state-
ment on p. 90 were added daily to each pot after growth had started.
92
Journal of Agricultural Research
Vol. XVII, No. 3
Where elemental sulphur and calcium carbonate (CaCOa) were added,
0.3 gm. and i gm., respectively, of the above substances were mixed
with the soil or sand in each pot before planting. In the control receiv-
ing no added nutrients, 25 cc. of distilled water was added. The con-
centration of the salts per liter of solution were as follows :
Gm. per
liter.
Sodium nitrate 0-25
Sodium sulphate 20
Calcium sulphate 25
The salts were dissolved in the same solution when more than one salt
was added to a pot. The sand cultures received the soil extract which
contained the additional nutrients as required. When the cultures
required further moisture, the same amount of water was added to each
of the pots. The soil extract was prepared by thoroughly mixing one
part of soil with two of water. The solution was allowed to stand over
night and then filtered through a porcelain filter. The clear filtrate was
sterilized by boiling for 15 minutes. Analysis of the soil solution as
given in Table II was made according to a method given by Stewart (25).
TabIvB II. — Soil constituents found in soil extracts expressed as parts per million of the
soil extract
vSoil.
Potas-
sium oxid.
Phosphor-
us pen-
toxid.
Sulphate
sulphur.
Total
sulphur.
Calciuta
oxid.
Aa
10. 0
19.8
7.2
3-2
3-4
I. 0
7-4
3-6
3-2
9.6
3-6
3-2
23. 0
Ba
15.0
22. 0
Co
» Soil A=beaverdam; soil B=Medford loam; soil C=aatelope-clay-adobe.
DISCUSSION OF RESULTS
On examining the data in Table III it is noted that the weight of the
straw grown on soils B and C and receiving sulphur fertilizer is greater
than where no sulphur was used. An increase in weight of the crops
is also observed in the sand pots receiving additional sulphur over those
receiving soil extract plus sodium nitrate. The absence of plant food
in the original sand and the use of sterilized soil extract shows undoubt-
edly that sulphur acted directly in promoting this growth. The same
response of the soil to sulphur leads one to conclude that the sulphur
here too has acted directly in promoting the growth. This increase in
growth is also accompanied by an increase in sulphur content of the oat
straw. In soil A this increase from sulphur application is not obtained.
The weight of straw from the pots receiving calcium sulphate, sodium
sulphate, and sulphur in addition to sodium nitrate is no greater than
from the one receiving sodium nitrate only. In the sand pots receiving
added sulphur we find no increase in weight of straw over the one re-
ceiving no extra sulphur. This is not surprising when the sulphur content
June i6, 1919 Sulphates Affecting Plant Growth and Composition
93
of the soil extracts is compared. The soil solution from A contains
twice as much sulphate sulphur as the soil extracts from B and C and the
total sulphur is three times as great. Apparently there is enough sulphur
in the soil solution compared to the other elements for straw production.
The development of the oat seed as shown in Table IV, agrees very well
with the weight increases of the straw as shown in Table III. In com-
paring the growth of the oats on the soils receiving sodium nitrate plus
calcium carbonate to those receiving sodium nitrate only, the calcium
carbonate appears to have an injurious effect upon growth, but in com-
paring the calcium carbonate-sodium-nitrate-treated soils to those receiv-
ing sulphur, sodium nitrate, and calcium carbonate, the sulphur has
caused increased growth in all cases.
Table III. — Weight of oat straw and its percentage of sulphur and nitrogen on the different
soils and sand cultures obtained fro^rn the different fertilizer treatment
Soil."
A. B.
C.
Treatment.
Weight.
Per-
cent-
age of
sul-
phur.
Per-
cent-
age of
nitro-
gen.
Weight.
Per-
cent-
age of
sul-
phur.
Per-
cent-
age of
nitro-
gen.
Weight.
Per-
cent-
age of
sul-
phur.
Per-
cent-
age of
nitro-
gen.
Calcium sulphate
Sodium nitrate .....
Sodium sulphate
Sodium nitrate
Sulphur
Sodium nitrate
Calcium carbonate ....
Sodium nitrate
Sodium nitrate
Gw.
>2. 21
}2-I3
[2. 22
2.23
li.8^
0.346
.282
•370
. 016
•143
0.28
•33
.42
•17
.29
.66
Gm.
I. 92
I. 90
I. 96
I. 41
1-34
1.67
0-4S
.42
■58
0. 21
.266
•336
.476
. 672
.806
Gm.
2. 01
2. 01
1-75
1.68
•63
•54
©•175
•113
•30
• 005
0.32
■33
. 21
•54
No fertilizer 97
•49
Sand.a
Extract A.
Extract B.
Extract C.
Weight.
Per-
cent-
age of
sul-
phur.
Per-
cent-
age of
nitro-
gen.
Weight.
Per-
cent-
age of
sul-
phur.
Per-
cent-
age of
nitro-
gen.
Weight.
Per-
cent-
age of
sul-
phur.
Per-
cent-
age of
nitro-
gen.
Calcium sulphate
Sodium nitrate
Sodium sulphate
Sodium nitrate
Sulphur
Gm.
}i.69
}i-63
1.48
1-59
0. 16
•30
.42
• 13
1.30
I. 10
I- 54
I- 50
Gm.
1.72
1-57
1.36
I. 06
0. 21
•13
•32
•015
0. 92
.98
1. 29
I. 46
Gm.
I. 62
1-53
I. 76
1.08
0-33
•34
•39
•05
0.80
.96
Sodium nitrate
Calcium carbonate ....
Sodium nitrate
•73
1.70
o A=beaverdam soil or sand; B=Medford loam; C= antelope-clay-adobe.
94
Journal of Agricultural Research
Vol. XVII, No. 3
Table IV. — Weight of oat seed grown on soil and sand cultures
[Percentage of nitrogen is given on oats grown in soils A and B]
Treatment.
Soil.o
Weight.
Calcium sulpjiate . .
Sodium nitrate
Sodium sulphate. .
Sodium nitrate
Sulphur
Sodium nitrate
Calcium carbonate .
Sodium nitrate
Sodium nitrate . . . . ,
Calcium carbonate .
No fertilizer
Gm.
[1-63
I-I-44
1-45
I. 62
I. 06
•34
Per-
cent-
age of
nitro-
gen.
Weight.
I. 72
1.79
1. 46
1-73
1.77
2. 29
Gm.
1.25
I- 13
.96
.76
•52
.27
Per-
cent-
age of
nitro-
gen.
Weight.
1-73
I. 71
1.79
2-53
2-33
Ex-
tract A.
Gm.
0.85
.92
.82
.61
.08
. 10
Sand.o
Weight.
Gm.
O. 67
•77
.67
•55
Ex-
tract B.
Ex-
tract C.
Gm.
O. 71
•55
•38
•36
Gnt.
O. 64
.48
■72
.26
a A=beaverdain soil or sand; B=Medford loam; C= antelope-clay-adobe.
The results on clover as given in Table V show increased yields in all
cases where sulphur was added to the soils. This increase is also seen on
the sand cultures receiving soil solutions from soils B and C, but not on
the sand receiving nutrients as soil A. The oats grown on soil A, as on
the sand receiving extract from A, did not respond to sulphur applica-
tion, but the clover did respond to sulphur treatment of the soil. How-
ever, the clover growing on the sand and receiving its plant food from
the soil extract did not show increased growth where sulphur was added.
Naturally one may attribute this difference to some other factor present
in the soil which was favorably influenced by the sulphur. Omitting
this important factor and observing the weights of the dry material grown
in the soil solution, there appears to be enough available sulphur in
soil A and in the soil extract to eliminate sulphur as being directly a
limiting factor as a plant food, while in the other soils the supply of
available sulphur seems to be limited in comparison to the other plant
food available.
It is probable that the other factors acting here are the legume bacteria
which are present in the soil but not in the sand. The data in Table VI
show the increase in weight of the roots where sulphur was used, and
when the roots were examined the number of nodules, according to esti-
mate, varied directly as the weight of the roots. The roots of those plants
grown in the sand, of course, contained no nodules. Another reason why
the bacteria appear to be favorably influenced by the sulphur is the
noticeable percentage of increase of nitrogen in those plants grown on
soil receiving sulphur while the plants grown on sand do not show this
increase in nitrogen content. For this short period of growth the sulphur
apparently has a marked influence on the nitrogen content of the clover.
June i6. 1919 Sulphates Affecting Plant Growth and Composition
95
Table V. — Weight of clover and its percentage of sulphur and nitrogen on the different
soils and sand cultures obtained from the different fertilizer treatments
Soil."
A.
B.
c.
Weight.
Per-
centage
of sul-
phur.
Per-
centage
of nitro-
gen.
Weight.
Per-
centage
of sul-
phur.
Per-
centage
of nitro-
gen.
Weight.
Per-
centage
of sul-
phur.
Per-
centage
of nitro-
gen.
Calcium sulphate
Sodium nitrate
Sodium sulphate
Sodium nitrate
Sulphur
Calcium carbonate
Sodium nitrate
Sodium nitrate
Calcium carbonate ....
Sodium nitrate
No fertilizer
Gm.
}3-98
U. 06
}3-89
1.76
}i.48
1.80
0.197
.181
. 19S
.097
. 012
•055
2.94
2-93
3.10
2.32
1.98
I. 91
Gm.
2. 2Q
1-37
1.87
•99
.61
•65
0. 227
.280
•234
.056
.032
•037
2. 78
2-95
3-64
2. 19
I. 92
2.58
Gm.
I. 69
1.66
•65
.62
.42
•49
0. 240
.205
2. 67
3-33
1.28
2. 10
1.79
Treatment.
Sand .a
Extract A.
Weight.
Per-
centage
of ni-
trogen.
Extract B.
Weight.
Per-
centage
of ni-
trogen
Extract C.
Weight.6
Per-
centage
of sul-
phur.
Per-
centage
of ni-
trogen.
Calcium sulphate. .
Sodium nitrate. . . .
Sodium sulphate. . .
Sodium nitrate. . . .
Sulphur
Calcium carbonate.
Sodium nitrate. . . .
Sodium nitrate. . . .
Calcium carbonate.
Sodium nitrate. . . .
No fertilizer
Gm.
k54
49
63
2.38
2.56
2- 54
2-35
Gm..
0.38
■36
•24
. 22
Gm.
I. o
0.467
2. 14
I. 92
I. 90
493
52
2. 42
1.86
o A=beaverdam soil or sand; B=Medford loam; C=anteIope-clay-adobe.
b Plants were not thinned out. There was no growth in sodium-sulphate sodium-nitrate pot.
It appears that the sulphur increases the nitrog"en content by stimulating
the activity of the legume bacteria causing greater nitrogen fixation.
The total nitrogen removed from the sulphured soils is three times as
great as from the unsulphured soils. These plants, of course, have grown
for only two months and whether the mature plant would show this
same ratio will have to be decided by further experiments. In comparing,
the nitrogen and sulphur contents of the clover grown in soil A at two
different periods as given in Table VII, there is a decrease in percentage
of nitrogen and sulphur from May i to June i . Perhaps the percentage
96
Journal of Agricultural Research
Vol. XVII, No. 3
of total sulphur and nitrogen would grow less as the plant developed,
until, at maturity, the nitrogen content would average about the same
for all the clover grown under the diflferent fertilizer treatments. In
certain pots, maturity, undoubtedly, would be reached sooner, but allow-
ing each group to grov/ until they all reached the same stage of develop-
ment, it would be of importance to know whether the sulphur had affected
the nitrogen content and the character of the compounds containing
nitrogen. Samples of alfalfa grown on sulphur-fertilized soils in Oregon,^
as shown in Table IX, have shown a higher nitrogen content than those
grown on the same soil without sulphur application. Shedd (24) reports
increase in protein content of soybeans from ammonium-sulphate fertili-
zer and Ames and Boltz (i) report larger protein content in rape where
sulphates were in the fertilizer used. In this experiment available
nitrogen was present in the form of nitrates. It would be of interest
to know whether by maintaining the sulphate supply an increase in
nitrogen assimilation from the air could be brought about.
Table VI. — Weightof clover roots expressed in gms. grown on the different soils receiving
various treatments
Treatment.
Soil A.o
Soil B.»
Soil C. 1
Calcium sulphate.
Sodium nitrate. .. .
Sodium sulphate .
Sodium nitrate. .. .
Sulphur
Calcium carbonate
Sodium nitrate. .. .
Sodium nitrate. .. .
Sodium nitrate. .. .
Calcium carbonate
No fertilizer
I- 15
•95
•93
.60
• 51
0.85
•36
.49
•23
. 21
0.44
.28
. 10
13
"Soil A=beaverdam; soil B=Medford loam; soil C= antelope-clay-adobe.
Table VII. — Sulphur and nitrogen content at different stages of growth in clover
grown on soil A
Treatment.
Percentage
of sulphur.
May I.
Percentage
of sulphur,
June I.
Percentage
of nitrogen,
May I.
Percentage
of nitrogen,
Jtme 1.
Calcium sulphate
} 0.285
> . 260
.360
.086
} -070
• 130
0.197
.181
.198
.097
. 012
•055
3-14
3^57
3^24
2.80
2.13
2. 67
Sodium nitrate
2.94
Sodium sulphate
Sodium nitrate
2^93
Sulphur
Sodium nitrate
3- 10
Calcium carbonate
Sodium nitrate .
2.3a
Sodium nitrate
Calcium carbonate
1.98
No fertilizer
\. QI
'Reimer and Tartar. Unpublished data, Oregon Agricultiu-al Experiment Station.
June i6, 1919 Sulphates Affecting Plant Growth and Composition
97
The rape plant did not show this general response to sulphur treatment
like the other plants, for in several instances the growth is greater in those
pots receiving no sulphur. However, if a comparison is made between
the soil and sand pots receiving sodium nitrate and those receiving sodium
nitrate plus sodium sulphate, it is observed from the data in Table VIII
that increased growth has resulted from the addition of sodium sulphate
on both the sand and the soil. The growths of rape on the soils and
their extracts parallel each other very well. The rape grew very poorly
on the extract from soil C, so no data are given. Where a comparison
is made on the growths of the crops on the different soils they do not
follow the same order, and the sulphur and nitrogen content do not
show the same change from the different fertilizer treatments; but the
rape grown in pots receiving sulphur, sodium nitrate, and calcium carbo-
nate has a higher percentage of sulphur than that grown in the other
pots, yet the total sulphur removed is not much larger. The plants
grown on the sand have higher percentages of nitrogen and sulphur
but the total sulphur and nitrogen removed is no greater than for' those
plants grown in the soil.
The total sulphur present in plants is far greater where sulphur ferti-
lizer was used. On account of the small amount of material, the sul-
phate sulphur was not determined so that it is not possible to tell whether
the organic sulphur was increased. Analysis, in this laboratory, for
organic sulphur and sulphate sulphur in alfalfa hay grown on soils receiv-
ing 300 pounds of sulphur per acre and on the same soils receiving no
sulphur fertilizer as given in Table IX shows that the organic sulphur was
increased by the application of sulphur. Shedd (24) found an increase in
organic sulphur in soybeans from the use of ammonium sulphate fertilizer.
Table VIII.-
-Weight of rape and its percentage of sulphur and nitrogen on the different
soils and sand cultures
Soil.«
Treatment.
A.
B.
c.
Weight.
centage
of sul-
phur.
Pe,.
centage
of nitro-
gen.
Weight.
Per-
centage
of sul-
phur.
Per-
centage
of nitro-
gen.
Weight.
Per-
centage
of sul-
phur.
Per-
centage
of nitro-
gen.
Calcium sulphate
Sodium nitrate
Sodium sulphate
Sodium nitrate
Sulphur
Gm,.
|i.8i
}2.92
[1-03
1.46
}i.89
•71
0-7S
.50
1. 19
• 057
.051
.236
1-39
1.65
I. 29
1-57
.84
I. 18
Gm.
1-57
2. 26
1.89
1.79
i.6i
•75
0.66
•55
.80
. 024
.017
. 022
0.903
.990
1.47
1-43
1-45
I. 24
Gm.
1.98
2.31
1-73
1.98
•59
• 30
0. 61
•65
.81
.02
I. 06
•99
Sodium nitrate
Calcium carbonate ....
Sodium nitrate
Sodium nitrate
Calcium carbonate ....
No fertilizer
1.74
1.36
3.00
I. 20
a A=beaverdanisoiland sand; B = Medford loam; C=antelope-clay-adobe.
98
Journal of Agricultural Research
Vol. XVII, No. 3
Table VIII.
-Weight of rape and its percentage of sulphur and nitrogen on the different
soils and sand cultures— Continued.
Treatment.
Sand."
Extract A.
Weight.
Calcium sulphate . ,
Sodium nitrate
Sodium sulphate . .
Sodium nitrate
Sulphur
Sodium nitrate
Calcium carbonate.
Sodium nitrate
Sodium nitrate
Calcium carbonate.
No fertilizer
Gm.
h
1.03
50
59
Per- I Per-
centage j centage
of sul- lOf nitro-
phur. I gen
I
Weight.
1-13
•93
I. 29
.07
3-53
2-35
8.29
4-31
Extract B.
Gm.
0-73
.62
•35
•25
Per-
centage
of sul-
phur.
Per-
centage
of nitro-
gen.
0. 90
1. 40
I- 51
3-75
3-38
5-58
4. 00
a A=beaverdani soil and sand; B=Medford loam; C= antelope-clay-adobe.
Petersen (19) in an analysis showing different forms of sulphur in plants
found more organic sulphur in clover, rape, and radish where sulphur was
present in the fertilizers used, and Ames and Boltz (j) report increase of
organic sulphur in rape where sulphates were applied to the soil. These
results and the increase in nitrogen content support the idea that main-
taining a sufficient supply of sulphate sulphur and available nitrogen in the
soil vvould tend toward more protein or other sulphur organic-compounds
being formed in the plant. The sulphur content is generally increased
wherever sulphur fertilizer is added. The sulphate radical is in combi- ,
nation with some other radical and the question arises whether the mineral
content or ash of the plant is not increased by this noticeable increase
of sulphate sulphur. If sulphur is applied as sodium sulphate will the
sodium content of the plant be increased or if calcium sulphate is used
will the calcium be absorbed by the plant ?
Table IX. — Percentage of total sulphur, sulphate sulphur, organic sulphur, and total
nitrogen in alfalfa grown on sulphured and unsulphured portions of three different
Oregon soils
Soil.
I. . .
2. . .
3---
Treatment.
Total
sulphur.
Sulphur. . .
No sulphur.
Sulphur . . .
No sulphur.
Sulphur. . .
No sulphur
o. 227
. 127
. 167
.118
. 200
.118
Sulphate
sulphur.
o. 0603
None.
• 0356
None.
None.
Organic
sulphur.
167
127
131
118
141
118
Total
nitrogen.
2.51
2. 22
2. 16
2. 01
2.38
2.09
June 16, 1919 Sulphates Affecting Plant Growth and Composition
99
Soil B responds to sulphur treatment in these pot tests while in the
field elemental sulphur caused no increase in production. In the analy-
sis of the soils in Table I soil C contains calcium carbonate, while B
does not. It may be that the sulphur was oxidized in the field as
in these pot tests, but as no base was present to combine with the sul-
phuric acid, the latter interfered with the growth. In C, calcium carbo-
nate was present which neutralized the acidity and provided sulphates
which produced the beneficial effects. Data in Table X show the differ-
ence in sulphate content between the soils receiving calcium carbonate and
sodium nitrate compared to those receiving the above named compounds
plus sulphur. The results show that the elemental sulphur was oxidized
to the sulphate form. Furthermore, no weighable quantities of barium
sulphate were obtained from, the water extracts of the unsulphured soils,
showing a deficiency of sulphate sulphur for immediate plant use. The
rate of sulphofication appears to be greater in the beaverdam soil than
in the other soils containing less organic material. While soil A has a
high sulphur content and also readily oxidizes elemental sulphur, it gave
a noticeable response to sulphate teatment when clover was grov/n.
Table X. — Sulphur as sulphate in the water extract from 40 gm. of soil A and 80 gm.
each of B and C after growth of plants
[Weight in milligrams]
Treatment.
Soil A.o
Soil B.o
Soil C.o
Clover.
Oats.
Rape.
Clover.
Oats.
Clover.
Oats.
Sulphur
1 21.9
I None.
34-4
None.
25-9
None.
4.1
None.
10.8
None.
10.3
None.
Calcium carbonate . . .
Sodium nitrate
Calcium carbonate. . . .
Sodium nitrate
II. 2
None.
« Soil A = beaverdam; soil B=Medford loam; soilC=anteIope-clay-adobe.
It is realized that the experiments conducted here have not been on a
large scale and the conditions are not comparable to those in the field.
No general conclusions can be made, but what conclusions are drawn
apply only to the limits of this experiment and based upon conditions
of this work where each individual case can be compared to the other.
This work will be repeated on a larger scale and expanded so as to
answer some of the questions which have arisen during this experiment.
SUMMARY
I . Pot experiments to show the effect of sulphur fertilizers — namely,
sodium sulphate, calcium sulphate and sulphur on red clover, rape, and
oats were carried with three different soils, including one with a high
sulphur content, one that did not respond to elemental sulphur in the
field, and one that did.
lOO Journal of Agricultural Research voi. xvii. No. 3
2. To eliminate the sulphur compounds acting upon the insoluble
plant food and soil organisms, these plants were also grown on sand
receiving the sterilized soil extract and certain pots received the addi-
tional sulphur fertilizers as the soil.
3. Sodium sulphate and calcium sulphate were added daily in solu-
tion form. The elemental sulphur was mixed with the soil and calcium
carbonate at the time of sowing the seed.
4. Sodium nitrate solution was added daily to eliminate available
nitrogen as a limiting factor of growth and also to determine what effect
sulphates would have on nitrogen assimilation by the plant.
5. The plants were grown for two and one-half months and the dry
weights of the tops were recorded. The total sulphur and nitrogen was
determined in the majority of cases.
CONCLUSIONvS
1. Addition of sulphate and elemental sulphur enhanced the growth
of the plants grown in pots in the greenhouse.
2. The corresponding increases obtained on the soil extracts indi-
cated that sulphur acted directly in promoting this growth.
3. The great increase in the nitrogen content of the clover grown on
the soil where sulphates had been added is the result in all probability
of the sulphates stimulating the action of the legume bacteria.
4. Sulphates caused increased root development and number of
nodules on the clover grown in the soil pots.
LITERATURE CITED
(i) Ames, J. W., AND BoLTz, G. E.
1916. SULPHUR IN RELATION TO SOILS AND CROPS. Ohio Agr. Exp. Bul. 292,
p. 219-256. References, p. 255-256.
(2) and Richmond, T. E.
1918. SLTLFOFICATION IN RELATION TO NITROGEN TRANSFORMATIONS. In Soil
Sci., V. 5, no. 4, p. 311-321.
(3) Bernhard, a. D.
1910. EXPERIMENTS ON CONTROL OF" POTATO SCAB. (Abstract.) In Chem.
Abstracts, v. 5, no. 13, p. 2295. 1911. Original article (Versuche zur
Bekampfung dcs Kartoffelschorfes) in Deut. Landw. Presse, Jahrg. 37,
No. 18, p. 204-205. 1910. Not seen.
(4) BOULLANGER, E., and DUGARDIN, M.
1912. M^CANISME DE l'action FERTIlisante du souFrE. In Compt. Rend.,
Acad. Sci. [Paris], t. 155, no. 4, p. 327-329.
(5) Brown, P. E-, and Gwinn, A. R.
1917. EFFECT OF SULFUR AND MANURE ON AVAILABILITY OP ROCK PHOSPHATE
IN SOIL. la. Agr. Exp. Sta. Research Bul. 43, p. 367-389. Bibliogra-
phy, p. 389.
(6) and Kellogg, E. H.
1914. SULFOFICATION IN SOILS. la. Agr. Exp. Sta. Research Bul. 18, p. 49-111.
(7)
191 5. SULFUR AND PERMANENT SOIL FERTILITY IN lOWA. In Jour. Amer.
Soc. Agron., v. 7, no. 3, p. 97-108.
junei6, I9I9 Sulpkates AffecUng Plant Gvowtk Gud Compositiou , loi
(8) Chancrin, E., and Desriot, A.
191 1. ACTION OF SULFUR AS A FERTILIZER. (Abstract.) In Cliem. Ab-
stracts, V. 6, no. 6, p. 789. 1912. Original article (Action du soufre
comme engrais sur le developpement des pommes de terre et des
betteraves) in Jour. Agr. Prat., ann. 75 (n. s. t. 21), no. 14, p. 427-429.
1911. Not seen.
(9) Fred, E. B., and Hart, E. B.
191 5. THE comparative EFFECT OF PHOSPHATES AND SLTLPHATES ON SOIL
BACTERIA. Wis. Agr. Exp. Sta. Research Bui. 35, p. 35-66, 6 fig.
(10) Hall, Alfred Daniel, BrenchlEy, Winifred, Elsie, and Underwood, Lilian
Marion.
1914. THE soil solution and the MINERAL CONSTITUENTS OF THE SOIL.
In Jour. Agr. Sci., v. 6, pt. 3, p. 278-300, pi. 4-8.
(11) Hart, E. B., and Peterson, W. H.
1 911. sulphur requirements of farm crops in relation to the soil
AND air supply. Wis. Agr. Exp. Sta. Research Bui. 14, 21 p.
(12) and Tottingham, W. E.
1915. relation of sulphur compounds to plant nutrition. In Jour,
Agr. Research, v. 5, no. 6, p. 233-250, pi. 20-22. Literature cited,
p. 249.
(13) Janicaud, W.
I914. HAS SULPHUR A DIRECT GROWTH EFFECT ON PLANTS? (Abstract.)
In Chem. Abstracts, v. 8, no. 14, p. 2592. 1914. Original article
(Wirkt Schwefeldiingung Wachstumsfordemd ?) in Gartenwelt, Jahrg.
18, No. 3, p. 29-32, illus. 1914. Not seen.
(14) LiPMAN, Jacob G.
1916. SULPHLTR ON ALKALI SOILS. In Soil Sci., V. 2, no. 3, p. 205.
(15) and McLean, H. C.
1918. experiments WITH SULPHUR-PHOSPHATE COMPOSTS CONDUCTED UNDER
FIELD CONDITIONS. In Soil Sci., v. 5, no. 3, p. 243-250.
(16) and Lint, H. Clay.
1916. SULPHUR OXIDATION IN SOILS AND ITS EFFECT ON THE AVAILABILITY
OF MINERAL PHOSPHATES. In Soil Sci., V. 2, no. 6, p. 499-538, 5 fig.
Literature cited, p. 535-538.
(17) Lyon, T. Littleton, and Bizzell, James A.
1918. lysimeter experiments ... N. Y. Cornell Agr, Exp. Sta. Mem. 12,
115 p., 4 pi. Bibliography, p. 82-84.
(18) McLean, Harry C.
1918. oxidation of sulphur by microorganisms, /n Soil Sci., v. 5, no. 4,
p. 251-290. References, p. 287-290.
(19) Peterson, W. H.
I914. forms of sulfur in plant materials and THEIR VARIATION WITH
THE SOIL SUTPLY. In Jour. Amer. Chem. Soc, v. 36, no. 6, p. 1290-
1300. Bibliography, p. 1300.
(20) PiTZ, W.
1916. EFFECT OF ELEMENTAL SULPHUR AND OF CALCIUM SULPHATE ON
CERTAIN OF THE HIGHER AND LOWER FORMS OP PLANT LIFE.
In Jour. Agr. Research, v. 5, no. 16, p. 771-780, pi. 56.
(21) REiMER, F. C.
1914. SULPHUR FERTILIZER FOR ALFALFA. In Pacific Rural Press, v. 87^
no. 26, p. 717.
I02 Journal of Agricultural Research voi. xvii. No. 3
(22) Shedd, O. M.
1913. SUIvPilUR CONTENT OF SOME TYPICAL KENTUCKY SOILS. Ky. Agr.
Exp. Sta. Bui. 174, p. 267-306. References, p. 306.
(23)
1914. The RELATION OP SULPHUR TO SOIL FERTILITY. Ky. Agr. Exp. Sta.
Bui. 188, p. 593-630. Bibliography, p. 629-630.
(24)
1917. EFFECT OF SULPHUR ON DIFFERENT CROPS AND SOILS. In Jour. Agr.
Research, v. 11, no. 4, p. 91-103. Literature cited, p. 103.
(25) Stewart, Guy R.
1918. EFFECT OF SEASON AND CROP GROWTH IN MODIFYING THE SOIL
EXTRACT, /n Jour. Agr. Research, V. 12, no. 6, p. 311-368, 24%., pi. 14.
Literature cited, p. 364-368.
(26) Sw ANSON, C. O., and Miller, R. W.
1917. THE SULPHUR CONTENT OF SOME TYPICAL KANSAS SOILS, AND , THE
LOSS OF SULPHUR DUE TO CULTIVATION. In Soil Sci., V. 3, no. 2,
p. 1 3 9-1 48. Literature cited, p. 147-148.
108123°— 19 2
PLATE 9
A. — Clover on soil A. The top row, reading from left to right, shows the soil pots
which received the following fertilizers:
Pot I, calcium sulphate, sodium nitrate; pot 2, sodium sulphate, sodium nitrate;
pot 3, sulphur, sodium nitrate, calcium carbonate; pot 4, sodium nitrate; pot 5,
sodium nitrate, calcium carbonate; pot 6, no fertilizer.
The lower row, reading from left to right, shows the sand pots which received the
follo^ving fertilizers:
Pot 7, calcium sulphate, sodium nitrate; pot 8, sodium sulphate, sodium nitrate;
pot 9, sulphur, calcium carbonate, sodium nitrate; pot 10, sodium nitrate.
B. — Clover on soil B. The top row, reading from left to right, shows the soil pots
which received the same fertilizers as in series A above. The lower row, reading from
left to right, shows the sand pots which received the same fertilizers as in series A.
C. — Clover on soil C. The top row, reading from left to right, shows the soil pots
which received the same fertilizers as in series A. The lower row, reading from left
to right, shows the sand pots which received the same fertilizers as in series A.
Sulphates Affecting Plant Growth and Composition
Plate 9
Journal of Agricultural Research
Vul. XVII, No, 3
Sulphates Affecting Plant Growth and Composition
Plate 10
Journal of Agricultural Research
Vol. XVII, No. 3
PLATE lo
A. — Rape on soil A. The top row, reading from left to right, shows the soil pots
which received the same fertilizers as in Plate g, series A. The lower row, reading
from left to right, shows the sand pots which received the same fertilizers as in pots
in Plate 9, series A.
B. — Rape on soil B. The top row, reading from left to right, shows the sand pots
which received the same fertilizers as in pots in Plate 9, series A.
C. — Rape on soil C. The soil pots, reading from left to right, received the same
fertilizers as in pots in Plate 9, series A.
PLATE II
A. — Oats on soil A. The soil pots received the same fertilizers as in pots showTi in
Plate 9, series A.
B. — Oats on soil B. The soil pots received the same fertilizers as in pots shown in
Plate 9, series A.
C. — Oats on soil C. The soil pots received the same fertilizers as in pots shown in
Plate 9, series A.
Sulphates Affecting Plant Growth and Composition
Plate 1 1
Journal of Agricultural Research
Vol. XVII, No. 3
Sulphates Affecting Plant Growth and Composition
Plate 12
Journal of Agricultural Research
Vol. XVII, No. 3
PLATE
A. — Oats on sand cultures from soil A.
as in pots shown in Plate 9, series A.
B. — Oats on sand cultures from soil B.
as in pots shown in Plate 9, series A.
C. — Oats on sand cultures from soil C.
as in pots shown in Plate 9, series A,
The sand pots received the same fertilizers
The sand pots received the same fertilizers
The sand pots received the same fertilizers
RELATION OF WEATHER TO FRUITFULNESS IN
THE PLUM*
By M. J. DORSEY
Head of Section of Fruit Breeding, Agricultural Experiment Station of the University
of Minnesota
Under suitable growing conditions the plum tree is remarkable for the
uniformity with which it annually produces a crop of flower buds.
Bearing a full crop of flower buds annually, however, does not insure a
full crop of fruit annually; therefore, it is evident that a considerable
number of flowers fail to set fruit. ^ From the standpoint of fruit pro-
duction, thinning, up to three-fourths of the bloom, is actually beneficial,
but beyond this the margin is approached where the thinning process
reduces the yield and there is economic loss. The status of setting in
controlled crosses known to be fertile under tents was similar to that
in the orchard generally. This general condition led to an attempt to
isolate those factors of the weather influencing the setting of fruit which
result in such great differences as a complete crop failure one year
and an overproduction of fruit another.
The elements of what is commonly known as "weather" which have
a bearing upon pollination and fertilization are wind, temperature,
sunshine, and rain. The combinations of these most favorable to the
setting of fruit are sunshine, a relatively high temperature, slight or no
wind, and an absence of rain. It is apparent that certain weather con-
ditions, good and bad, go together, but temperature and rain are undoubt-
edly the most important elements considered from the standpoint of
the setting of fruit and will be given greatest emphasis.
The following statements may be regarded as fairly typical of the
conception of the influence of unfavorable conditions at bloom. Cold
weather, rain, poor locality, and severe cold winter weather are men-
tioned by Goff (4) ^ as inhibiting fruitfulness. Bad weather at flowering
time has an "injurious influence on fruitage" by keeping away insect
visitors and affecting the fecundation of the flowers (13). Damage to
flowers by wind, hail, rain, insects, and fungi are commonly mentioned.
Lord (11) states that all varieties when in bloom are extremely sensitive
to cold or wet weather. Heideman (9) notes that ample cross-fertiliza-
1 Published, with the approval of the Director, as Paper 162 of the Journal Series of the Minnesota Agri-
cultural Experiment Station.
2 "Setting of fruit" is a term in common use among fruit growers. In general, it is used in referring
to the number of pistils which are swelling or "setting" six weeks or so after bloom. A distinction is
made in common usage between the percentage of fruit to set and the percentage of a crop, in that the *
latter is used in speaking of mature fruit.
'Reference is made by number (italic) to "Literature cited," p. 125-126.
Journal of Agricultural Research, Vol. XVII, No. 3-
Washington, D. C, June 16, 1919.
ru Key No. Minn. — 38.
(103)
I04 Journal of Agricultural Research voi. xvii, no. 3
tion does not prevent great differences in the crop from year to year.
Some growers hold that there is a good fruit crop only during seasons
ynih. favorable weather for bees at blooming time. Hedrick (7) analyzed
the weather records of New York with respect to fruit production and
showed that in general unfavorable weather is the dominant factor in
crop failures. In fact, for a long time fruit growers have recognized
certain weather combinations as detrimental to or prohibiting the
setting of fruit.
If weather is to be assigned such an important role in relation to fruit
fulness, the question arises as to the significance of the great variation
in the time of bloom from year to year. For instance, plums have
varied nearly one month in the time of flowering at the Fruit- Breeding
Farm in the last seven years, the earliest bloom in this period beginning
April 24, 1915, and the latest May 20, 1916. The cause for such a varia-
tion in time of bloom should not be assigned entirely to the weather
of early spring, because Sandsten {13) found, upon analyzing the bloom-
ing records at Madison, Wis., that the time of flowering was influenced
more by the growing conditions of the preceding summer and fall than
by those of the spring. In Plate 15 the prevailing weather of early
spring when plums are in flower is presented in some detail. It will be
seen from the analysis presented in these graphs that cool weather and
frequent rains can be expected in Minnesota for a period of even greater
length than that covered by the greatest extremes in the time of bloom.
Therefore, inasmuch as a range in blooming time of as much as one
month has not meant an escape from periods of unfavorable weather,
early or late blooming does not necessarily have a constant relation to
fruitfulness.
The period of 10 days after bloom was selected (PI. 15) because it covers
for the most part the time of fertilization. In only 10 instances out of
142 did the minimum temperature occur in the day and the maximum
at night, so that the curve of maximum temperature may be considered
as the day temperature and that of the minimum as the night tempera-
ture. In the graph for each season the period of bloom is indicated by
the lighter shaded portion between the maximum and minimum temper-
ature curves. In the case of wind and the character of the day (sun-
shine or cloudiness) a 12-hour day was taken because of the bearing of
wind and sunshine on bee flight. The date in the graph is located in the
midpoint, which is 12 m. The short, broken-line curves indicate the
wind velocity during the daytime only, i. e., from 6 a. m. to 6 p. m.
The legend is at the right of the graph. The character of the day is
' shown by the shading at the base of each graph ; the dark bar represents
the portion of the day which was cloudy, the cross bar that which was
partly cloudy, and the white the time of sunshine. A dotted line is
drawn through each graph at the 40° and 51° F. points, the former
junei6, I9I9 Relation of Weather to Fruitfulness in Plum 105
being the point Goff (5) found that plum pollen did not germinate and
the latter the temperature of slow tube growth.
Since the weather at the Fruit-Breeding Farm has not been recorded,
this analysis is made from the records furnished by Mr. U. G. Purssell,
of the United States Weather Bureau, at Minneapolis.
EFFECT OF UNFAVORABLE WEATHER ON THE SETTING OF FRUIT
It has been a matter of common observation among fruit growers that
when the blooming period is accompanied by a prolonged rain there is
generally a light setting of fruit. Halsted (d), in an attempt to deter-
mine the cause of this, performed an experim.ent in which an apple tree
was kept wet with a spray of water for six days while in bloom. The
weather was fair during the experiment. The sprayed tree failed to set
any fruit, except in a few instances on the upper branches, while the
surrounding trees of the same variety set full.
Beach and Fairchild (j) performed a similar experiment with a Mount
Vernon pear tree and a Duchess grapevine. The pear tree subjected to
a spray for nine days bore a single fruit. Pollen taken from "fresh
anthers" on the fifth day and placed in a sugar solution proved to be
"perfectly capable" of germination. Many of the stigmas examined 24
hours after the experiment began were found to be "dusted with pollen,"
although no insects had been seen near the tree. After the close of the
experiment many anthers opened and shed an abundance of pollen.
In the case of the Duchess grape, although the 12 days' treatment did
not cover the entire period of bloom, the treated vines bore many aborted
berries, but on none of the clusters were all of the berries aborted. Also,
the average size of the fruit was reduced approximately one-half.
In these experiments the conditions which generally accompany a pro-
longed rain were not duplicated exactly, and consequently other factors
may have entered into the results obtained. However, a constant spray
was effective in preventing fruitfulness in the apple and pear, and even
in the case of the grape sufficient pollination to account for the setting of
fruit which took place may have occurred after the water was turned off.
It will be of interest here, after a review of the experiments of Halsted
(6) and of Beach and Fairchild {3), to include a statement concerning the
percentage of fruit to set in a plot of Surprise seedlings at the University
Farm in order to show the general effect of unfavorable weather. All
trees bloomed heavily during the seasons of 191 7 and 191 8 and for this
reason present an excellent illustration of the effect of weather upon the
setting of fruit. These seedlings are about 13 years old, fairly uniform
in size, and are growing under clean cultivation. It would appear that
ample pollination would take place if the weather were favorable, be-
cause these seedlings are located within less than a quarter of a mile of
the University apiary of about 100 colonics. In general it may be
io6
Journal of Agricultural Research
Vol. X\ai. No. 3
stated that during both seasons conditions were unfavorable for insect
flight. The weather conditions at time of blooming for these two seasons
are shown in Plate 15.
TablC I. — Comparison of fruit setting in an orchard of 226 Surprise seedlings during
the two relatively unfavorable seasons of igij and igi8 ^
Range in percentage of fruit to set in 1918.
Total
"Rtmzt in percentage of fruit to set in 1917.
0
1
5
10
20
30
number
of trees.
0
Number
of trees.
3
5
6
4
17
12
3
Number
of trees.
I
8
9
13
22
9
3
Number
of trees.
Nu mber
of trees.
Number
of trees.
Number
of trees.
4
1
II
6
12
2
4
14
9
13
3
3
6
14
3
4
3
3
I
6
50
35
78
32
7
10
20
30
40
Total number of trees
50
65
31
43
26
II
226
' The percentage set is based upon the total number of flowers borne. Each tree is placed in the table
with reference to the percentage of fruit set in 1917 compared with that in 1918. For instance, in 1917 there
were 78 trees in which 20 per cent of the flowers set, but in 1918 the set on these same trees ranged from o to
20 per cent.
The data are presented in the form of a correlation table in order to
show the influence of heavy fruiting during one year upon the crop the
succeeding year. Accordingly, each tree is placed in the table with
reference to the percentage of fruit set in 191 7 compared with that set
in 1918.
Three things are outstanding in Table I: (i) The heavy setting or
bearing of 191 7 was shown to have no distinct influence on the succeeding
crop in 1918; (2) there was a heavier setting in 1917 than in 1918, the
relative number of trees setting below 20 per cent being 109 and 189,
respectively; and (3) since by actual count it was determined in the
6-weeks period after blooming that only one pistil in four set or persisted
on those trees bearing what was arbitrarily regarded as a "full set," it
will be seen that many of the trees set an unusually small number of
fruits, too few, in fact, to produce a full crop after allowing for subse-
quent loss. This condition is not unusual in the plum when blooming
time is accompanied by unfavorable weather. The light set in those
trees which produced normal flowers in abundance presents a condition
quite similar to that which prevailed both seasons in a number of standard
varieties and other seedlings under cultivation. In Plate 13, A and B,
the contrast between the number of flowers borne and the fruit to set is
shown.
junei6, I9I9 Relation of Weather to Fruttfulfiess in Plum 107
ANALYSIS OF WEATHER AT BLOOMING TIME
With weather apparently having such an important bearing upon the
setting of fruit, as is indicated in the spraying experiment and in Table I,
a more detailed analysis of weather has been made during blooming time
and for 10 days after, with the object of determining whether there are
certain conditions each season which can be singled out as prohibiting
a set of fruit. At the outset it should be stated that there are factors
which operate beyond the 20-day period to reduce the crop. Neverthe-
less, there are influences entering during blooming time which do not
operate in the same manner anywhere else in the life cycle. As a result
of the sum total of these influences a sufficient number of pistils have or
have not set, as the case may be, at the 5- or 6-week period to determine
definitely the prospect of a crop.
WIND
The experiments of Waugh (16) show that no fruit set from wind-
carried pollen when insects were excluded by a covering of coarse mos-
quito netting. Further tests (18) with microscopic slides covered with
vaseline, to which pollen adheres, showed that when the slides were
placed at various heights and distances from trees in full bloom on bright
sunny days even a direct wind did not carry sufficient pollen to bring
about effective pollination at a distance equal to that from one tree to
another. Wind pollination, therefore, may be regarded as insufficient,
even under the most favorable conditions.
Pollination under orchard conditions is affected by windy weather,
however, especially when prolonged, if insect visits are prevented. Dur-
ing a strong wind, rain, cold, or cloudy weather, conditions are such that
insect activity is reduced to a minimum. Waugh {16, 17) shows that
honey bees, of the 30 or more species of insects found to visit the plum, are
(16, p. 247) "nearly always the most active workers, and the ones which,
by the character of their operations in the flower, may be held chiefly
responsible for the proper distribution of pollen." These results are con-
firmed by Backhouse (j). Wind, therefore, may be regarded as having
more of an indirect than direct bearing upon the setting of fruit, since
pollen is not wind-carried in quantities sufficient for ample pollination.
The influence upon bee flight, however, may be serious at certain times.
The curve for wind in Plate 15 runs through the point of hourly wind
movement from 6 a. m. to 6 p. m. While the average wind movement
considered aside from sunshine and the character of the day is of little
significance, it shows what may be expected at this time of year in Min-
nesota. The average wind movement per hour, within the above limits,
for 7 years was approximately 15 miles, while the average of the extreme
wind movement recorded, within the same limits, for the 7-year period
was near 19. The extreme movement recorded was 38 miles. Assum-
ing that a wind of 25 miles per hour approaches a condition where bee
io8 Journal of Agricultural Research voi. xvii, no. 3
flight is hindered, it will be seen from Plate 15 that wind alone is not gen-
erally prohibitive of bee blight, but that at certain critical times, as on
April 28 and 29, 191 5, following a period of cloudy weather with fiequent
rains, it may become important — more so, in fact, from the standpoint
of insect flight than from that of mechanical injury to flowers.
In addition to the considerations noted above, wind has a general dry-
ing effect upon the flower parts. Dehiscence is quickened and petals
drop earlier. There is, however, no marked drying noticeable in the
stigma during early receptiveness, but late in the receptive period stigmas
can be found which appear distinctly dry even before the stigmatic cells
are dead. Since the absorption of stigmatic fluid is no doubt the first
act in germination the dr3ang effect of wind upon stigmas may be re-
garded as much more critical late in receptiveness than earlier, especially
in view of the more unfavorable conditions for tube growth, if pollination
has been delayed.
TEMPERATURE
Temperature is primarily of interest in this connection from three stand-
points: (i) Its direct effect upon pollen or pistil, (2) its influence upon
pollen-tube growth, and (3) its interference with bee flight. Krom Plate 1 5
it will be seen that there are many periods of low temperature during
blooming time which are occasionally accompanied by frost. With ref-
erence to direct injury, it will be interesting to record here the damage
to flowers at two distinct stages of growth.
On the night of April 19, 191 8, a freeze occurred at the Fruit-Breeding
Farm, when the petals were just showing in the earliest blooming varie-
ties. There was no injury to pollen or pistil, but as many as one-half of
the petals were killed on some of the varieties. These bloomed, however,
at the usual time, and the small dead petals persisted, while those not
killed underwent the usual enlargement.
This freeze was follovv^ed by another on May 12, one week after bloom-
ing, when the flowers were further advanced. But this time all stigmas
were dead on the varieties which had bloomed earlier. The calyx tube
was still persistent, as there had not as yet been sufficient pistil growth
to break it except in two varieties of Prioius nigra. Although generally
there was little injury to pistils at this stage, different varieties showed
considerable differences in the degree of injury. On Stella, growing in a
low location, approximately 65 per cent of the pistils were killed, and on
Minnesota No. 21 (Burbank X Wolf), adjacent, there was less than i per
cent. Where injury occurred the entire pistil was killed, and in two days
it turned black, dried rapidly, and dropped a few days later at the pedicel
base. On the higher locations there was no injury to any of the varieties.
Compared with the region in Utah in which -Ballantyne (2) studied frost
injury, frosts do not appear to bear such a vital relation to fruitfulness in
Minnesota.
juneie, I9I9 Relation of Weather to Fruitjulness in Plum 109
Pollen taken from flowers in which the pistils were killed appeared
normal in color and in content when observed in a mount of lactic acid.
Its viability, however, was not tested, but judging from appearances
this freeze injured pollen much, less than pistils.
Goff (5) shows that plum pollen was not destroyed by a short exposure
to freezing temperatures. Sandsten {14) tested this point further and
found that when plum pollen was exposed to a temperature of 29.3° F.,
56 pef cent germinated, compared with 62 per cent in the check, Vv^hich
was not exposed to the freezing temperature. He also found that the
time required for germination was increased one-half as a result of the
influence of the low temperature. On the other hand, 21 plum pistils
exposed to the same temperature for six hours were all killed except two.
The action of low temperatures in retarding pollen-tube growth is no
doubt one of the primary factors in the failure of fruit to set. The exper-
iments of Goff (5) show that plum pollen does not germinate at tem-
peratures below 40° F., and even at temperatures as high as 51° F. that
there is slow pollen-tube growth. A dotted line is drawn through the
graph for each year in Plate 1 5 at these two points. The extent to which
the curve for the minimum temperature extends below the line where
pollen-tube growth does not take place shows that in some seasons, as in
1 91 5, a prolonged cool period following blooming may be the principal
cause of the failure of fruit to set.
With reference to the influence of temperature upon insect flight, it
appears that a definite point can not be selected below which activity
ceases. Furthermore, temperature can not be considered separate from
wind, rain, and sunshine. Recent investigations upon the honey bee,
which is the chief pollinizer of the plum, however, show something of its
response to temperature. Phillips {12) states that 57° F. is "the lowest
temperature which normal bees ever experience in the hive." At air
temperatures below this immediately surrounding the bees in cold
weather, they begin to cluster. Kenoyer {10) in reporting the data col-
lected over a 29-year period by J. ly. Strong at Clarinda, Iowa, shows that
only I per cent of the total honey produced for that period was collected
when the temperature was below 70° F. compared with 53 per cent when
the temperature ranged between 80° and 90° F. Nevertheless, this does
not deal directly with the point as to what temperature prevents the
pollinating activity of bees on plums in early spring.
The opinions of two bee men regarding the lower temperature in which
bees will fly are as follows:
The normal temperature for bees to take flight is 46° F. This temperature is i
degree to 2 degrees lower for Camiolan races and up to 3 degrees lower after long
confinement. ^ The individual bee can continue muscular movement only so
long as the temperature of the body does not fall below 45° F., but at this
temperature it loses its power of movement. {12, p. 59.) In general bees will not
' Letter from Prof. Frances Jaeger, University Farm, Dec. 31, 1918.
no Journal of Agricultural Research voi. xvii. no. 3
fly from the hive until the temperature is about 60° F. unless they are inpelled to fly
by a long period of confinement resulting in an accumulation of feces. ^
The minimum temperature curves in Plate 15 show that there are
only relatively short intervals when the temperature is below 50° F.
It would appear that if bees were present in sufficient numbers, other
conditions being suitable, ample pollination would undoubtedly take
place, at even short intervals of favorable weather.
SUNSHINE
Sandsten {14) showed that while sunshine had a direct influence upon
fertilization in the tomato, it had none in the plum. Judging from his
experiments, sunshine appears to have its chief bearing in this connection
upon such factors as insect flight and general plant activity, particularly
nectar secretion. Kenoyer states {10, p. 21) that "clear days are
preeminently the days for honey production." From general observa-
tion of bee activity on plum bloom, the same may be said regarding
pollination. As will be seen later, however, pollen is most readily avail-
able for dissemination in dry, sunny weather when bees are most active.
The total hours of sunshine during blooming are less than might be
expected. The character of the day is indicated in Plate 15 at the base
of the graph for each year by the shading. For the 7-year period there
has been, while plums were in bloom, an average of only 49 hours of
sunshine each season, compared with an average of 56 hours of cloudiness.
The minimum was reached in 191 6, when there were only 27 hours of
sunshine. Alone, however, the absence of sunshine does not prohibit
the setting of fruit.
RAIN
On account of the nature of the processes taking place at blooming
time, rain has the most immediate action of all of the factors of weather.
The fact that the period of pollination is so limited in the plum makes
it possible for rain to delay normal functioning to an injurious extent.
Furthermore, the stigma is exposed to weather during the limited time
it functions. It will be seen, therefore, that rain may influence processes
which, on account of the structure of the organs concerned, must function
when more or less exposed.
EFFECT OF RAIN UPON DEmSCENCE
A study of the bloom in the orchard during a heavy and prolonged
rain showed that the stamens were drawn together and held in a cluster
about the pistil by a large drop of water. This was typically the con-
dition in the absence of wind and in protected locations. The added
weight of water held in this way resulted in a drooping of the branches,
» Personal correspondence with E. F- Phillips of the U. S. Department of Agriculture, Bureau of Ento-
mology, Dec. 26, 1918.
June i6, 1919
Relation of Weather to Fruitjulness in Plum
III
and a large part of the water dripping from the tree fell immediately
from the stamen cluster. When the style was the same length or shorter
than the stamens, the stigma was completely immersed in water. In
cases where the style was considerably longer than the stamens, the
stigma projected from the drop, especially in positions where the pistils
pointed upward.
During the period of drying after a rain, when the water holding the
stamens and pistils is partly evaporated, the anthers break up into
groups, each group, however, being still held in water. Gradually, upon
further drying, the groups break up, and the anthers assume their normal
position in the flower.
In order to study anther action more in detail at the time of rain, a
limb which had been in bloom for three days was cut from a tree during
a heavy rain and brought into the laboratory, the temperature of which
was about 68° F. All anthers were closed when first brought in, but some
Fig. I. — An outline drawing of an anther of Minnesota No. 12, showing the adjustment which takes place as
a result of taking up or giving off water: A, an anther which has been open in the orchard for three days;
B, the same with the anthers pushed up to show the dead area at the upper end of the filament; C, the
appearance of the anther after two minutes in water. The anthers are completely closed and have
reached their usual size; D, the degree of opening of one suture of the same anthers in 8 minutes when
exposed in the laboratory at a temperature of 70° F.; E, the same anther at the end of 12 minutes' drying.
opened completely in lo minutes under the conditions in the laboratory.
When these anthers which had opened were again placed in water they
closed in two to three minutes.
Furthermore, anthers which had been open for approximately 3 days
and from which all of the pollen had been shed, when placed in water,
closed up and in some trials swelled to the usual size in as short a time
as 2 minutes (fig. i ) . Other tests showed that when unopened anthers
were kept in water for 2 weeks there was a slight breaking of tissue at the
suture but no dehiscence. On the other hand, anthers which had once
dehisced and from which the pollen had been shed closed at once when
placed in water and remained closed during the 2 weeks of the test.
Opened anthers held for 4 days in a saturated atmosphere under a bell jar
did not absorb sufficient moisture to close them ; and the experiments of
Goflf (5) showed that plum anthers did not open in a saturated atmosphere
under a bell jar in 56 hours at a temperature of 65° to 70° F. Goff (5)
also showed that in a dry atmosphere low temperatures (about 51° F.)
112 Journal of Agricultural Research voi. xvii. no. 3
retarded but did not prevent anthers from opening. This shows clearly
the relation of dehiscence to water.
The fact that empty anthers close during a rain and open afterwards
probably has been the basis for the popular conception that rain washes
pollen away.
With this statement, then, of anther action in relation to water, the
question arises as to what extent rain removes pollen from anthers which
have just dehisced. In investigating this point a branch of flowers was
brought into the laboratory, and after the anthers opened it was stirred
vigorously for 8 minutes in a pail of water. All anthers closed com-
pletely during the time of stirring. The larger part of the pollen lost
occurred with the first impact with the water. After this treatment it was
estimated that those anthers which were open before being put into the
water still contained, when they opened again, from one-quarter to two-
thirds of their pollen. These results agree with observ^ations made in
the orchard both during and after a rain.
The effect of rain in washing pollen away, even in the quantity noted
above, is partly modified by the unevenness of anther opening, there
being in some cases as much as 3 days' difference between the first and
last opening of anthers. The unopened anthers have a light yellowish
color in contrast to the water-soaked appearance of those which have been
closed by rain.
These observations show that anther action is a reversible process and
is controlled by water. The presence of the anther sap until the maturity
of the pollen creates an internal condition unfavorable to dehiscence. If
dehiscense takes place only after sufficient drying, there must be an inter-
nal control of water as well as a means for external loss. These two con-
ditions are met by a breaking of the epidermis at the suture and by the
drying or death of the cells of the filament at the point of union with the
anther where there is a pronounced constriction of the filament. At this
point the cells typically turn brown before dehiscence, a condition which
suggests an early cutting off of water. The browning slowly extends
down the filament and at the time the petals fall, 3 to 4 days after bloom-
ing, the filament is dead for a distance of i to 2 mm.
Under some conditions pollen is shed more quickly than under others.
When anthers of Surprise were allowed to open in a dry, still room at
about 72° F., at the end of four days pollen had not been shed except in
very small amounts. This was due partly to the adhesive action of a
yellowish, oily substance about the pollen grains which is characteristic
of some varieties, and partly to the absence of w4nd. The persistence of
pollen is further shown by specimens of Surprise gro\Mi in the green-
house, which, at the time of abscission of the calyx tube, 10 days after
blooming, still had an abundance of pollen present. But in some
varieties with sticky pollen, under orchard conditions as much as one-
half may still be present at the time the petals drop. On the other hand,
Jane i6, 1919 Relation of Weather to Fruitfulness in PIu?n 113
in some varieties of P. americana, pollen may almost completely dis-
appear from the anther during a wind, undoubtedly due to drying and
shaking the first day, or even the first few hours after opening. Wind
pollination would be more effective in these varieties than in the others,
although it is probable that it would be insufficient because plum pollen
has no appendages as in Pinus spp. to give it greater carrying capacity.
The importance of the rapid closing of anthers upon coming in contact
with water, together with the fact that they remain closed as long as they
are wet, needs emphasis in this connection. It will be e\ddent that
pollination is impossible when the anthers are closed. Furthermore,
the conditions whiph close anthers in most cases also prevent insect
flight, but, even if insects were working, pollination could not take place
for the reason that pollen is not available. It appears, therefore, that
too much emphasis has been placed upon the action of rain in washing
pollen away because anthers close quickly enough largely to prevent it.
RAIN INJURY TO PLUM POLLEN
It has been shown above that anthers take up water in sufficient quan-
tities to close them before there is complete loss of pollen. Accompany-
ing the drying process which takes place in the anther and the disap-
pearance of the anther sap, there is a similar drying in the pollen. Before
dispersal, pollen changes from the typical spherical shape to one distinctly
oblong, and deep folds appear at the sutures. When subjected to drying
immediately after removal from the anther, this change in shape takes
place in 5 to 10 minutes and is quickly reversible in 3 to 5 minutes when
placed in water. With these changes in mind, the question arises as to
the effect of a prolonged rain upon pollen.
The rainy period at blooming time in 191 5 started with a trace on
April 24 and ended with rain all day on April 26 and 27. The hea\'iest
rain, accompanied by a moderate wind, fell on April 26. During the
period of the rain there was a relatively high temperature ranging from
58° to 62° F.
Following the usual cytological procedure, before drying, there were
fixed in Flemming's medium anthers from 48 hybrids and varieties after
the rain of April 26 and from 30 others after the rain of April 27. In
all, pollen was collected from 63 crosses and 13 varieties, representing 6
species, namely, Prunus americana, P. Besseyi, P. nigra, P. triflora, P.
pissardi, and P. cerasus.
It would appear that this material would furnish conclusive evidence as
to whether or not plum pollen is burst by rain, as is held by Hedrick (8)
and generally by fruit growers. A careful examination of mounted
sections from each lot fixed as mentioned above, showed (i) that the
pollen was not burst and had every appearance of being normal; (2)
that only an occasional anther was devoid of pollen, although most of the
sutures were broken; and (j) there was no apparent difference in the
pollen condition of the different species.
108123°— 19 3
114
Journal of Agricultural Research
Vol. XVII, No. 3
EFFECT OF WATER UPON THE VIABILITY OF PLUM POLLEN
The effect of water on the viability of plum pollen was tested in the
sand cherry (P. Besseyi). The results are presented in Table II. The
time of soaking, lo minutes, while relatively short, was decided upon
because it was thought that if water was injurious at all, it would be
desirable to test its effect at the shorter exposure. The time of soaking,
however, is much shorter than the actual time the pollen was subjected to
water, since it required some time to dry. Sixteen hours elapsed before
this pollen was applied to the stigma. It will be seen from these results
that soaking pollen of this species in water and drying before using has
no injurious effect.
Table II. — Viability test of Sand Cherry (P. Besseyi) pollen after being soaked lo minutes
in water and then allowed to dry for l6 hours, the pollen in one series having been taken
from unopened anthers and allowed to dry in the sun and in the other series from open
anthers and allowed to dry in the shade
Cross made and pollen treatment.
Treated :
Tree No.
Tree No.
Tree No.
Tree No.
Tree No.
Tree No.
Tree No.
Checks:
Tree No.
Tree No.
Tree No.
Tree No.
Tree No.
Tree No.
Tree No.
Tree No,
iX, pollen,
iX, pollen,
2X, pollen,
2X, pollen,
3X, pollen,
iX, pollen,
3X, pollen,
soaked
soaked
soaked
soaked
soaked
soaked
soaked
10 minutes.
10 minutes.
10 minutes.
10 minutes.
10 minutes.
10 minutes.
10 minutes.
iX, pollen,
2X, pollen,
1X2, pollen
2X1, pollen
3X4. pollen
3X5, pollen
4X3, pollen
5X3, pollen
not soaked . .
not soaked . .
not treated.
not treated .
not treated .
, not treated .
, not treated .
not treated .
Condition of
anthers.
Unopened
, . .do
..do
..do
..do
Opened. . .
..do
Num-
ber of
flowers
on
May 25.
9
26
31
28
Num-
ber
swell-
ing on
June IS
6
6
20
9
25
Num-
ber
set on
June 26.
4
6
19
9
4
9
7
II
13
II
3
I
In addition to this, germination tests were made with selected varieties
to determine the effect of the rain of April 26 and 27 upon the viability
of pollen. Pollen was taken from anthers which had been closed by the
rain and placed in a hanging drop of 20 per cent cane-sugar solution.
There was no germination even in the checks from tented trees or from
unopened anthers subjected to rain. The temperature, however, which
was very changeable, w^as quite low a good part of the time, especially
at night, and the negative results with the check make it impossible to
draw conclusions as to rain injury to pollen under orchard conditions.
It has been shown that on account of anther adjustment less pollen
is actually washed away than has been supposed. Also, considerable
quantities of pollen may be retained by anthers which have opened
juneie. 1919 Relation of Weather to Fruitjulness in Plum 115
immediately preceding a rain, owing to the rapidity with which they
close. Anthers open as a result of drying, a condition which is brought
about by cutting off the water supply at the constriction of the filament,
and by evaporation, particularly from the suture. Anthers which have
dehisced close quickly when brought in contact with water, and, like
those which have not dehisced, remain closed as long as wet. Con-
sequently, pollen is not available for dissemination during a rain. A
careful distinction must be made between the normal shedding of pollen,
which takes place for the most part the first day or even the first few
hours an anther is open, and the washing away of pollen by rain, for the
reason that empty anthers close when wet but open again after a rain
when dry. Insect visits are reduced to a minimum, if not prevented,
under the same conditions that impede pollen dispersal. The cytological
studies show that plum pollen does not burst when wet by rain and
crossing tests show that it is not killed by moderate exposures to water,
although the results of Sandsten {14) indicate that humidity decreases
its longevity. x\s far as the pollen is concerned, therefore, a prolonged
rain acts primarily to delay pollination until conditions are again restored
which are favorable to dehiscence and dissem.ination.
THE STIGMATIC SURFACE
As in the case of anther and pollen, a study has been made of the
changes of the pistil during the functional period, which may be regarded
as a critical stage viewed from the standpoint of the relation of adverse
weather to the setting of fruit.
Immediately before the receptive period the outer cells of the stigma
are turgid (PI. 14, C and D) and their papillate structure gives to the
surface a characteristic velvety appeal ance Vv^hich is readily distinguished
from the glossy, moist surface when receptive. Where the suture termi-
nates, the stigma has a distinct depression, and in the plum its surface
is more or less oblique to the axis of the style, with the higher margin
opposite the marginal suture fold.
The terminal cells are one layer thick, and in longitudinal sections are
clearly distinct from the cells below on account of their large size, scant
cytoplasm, and conspicuous vacuoles. There is a slight variation in
the length of these cells in different species. In some, as in Sapa (P.
BesseyiXP. triflora), they contain spherical bodies, which stain deeply
and vary much in size, the larger ones being somewhat greater in cross
section than the nucleus. The scant cytoplasm in the terminal cells is
mostly located at the extreme terminal end in the form of a crescent.
THE RECEPTIVE STIGMA
Decided changes are noticeable in the terminal cells after the stigma
has become receptive. In sections made from stigmas 48 hours after
first becoming receptive the papillate cells are very irregular in outline
ii6 Journal of Agricultural Research voi. xvii. No. 3
and typically are collapsed and shrunken. A few cell walls appear to
be broken. The cytoplasm is much contracted and drawn out into
irregular vacuolated strands. The nuclei are generally irregular in
outline and show evidence of disintegration. In many of the stigmas
the papillate cells are partly broken away from those beneath, and the
pollen grains are found among, or even beneath, the collapsed and partly
separated sheath composed of the tenninal cells.
Heideman states (9, p. 191) that the "actual time during which
fertilization may be effected scarcely exceeds two hours." Obser\'ations
here show that under normal conditions the plum stigma remains recep-
tive for a maximum period of about one week. At the end of three to
five days, however, the stigma begins to turn brown, and as it becomes
dead and dry at the end of the receptive period the color gradually
deepens to a dark brown and then black. The dark color slowly extends
down the style, which, as a rule, abscisses before turning brown more
than two-thirds of the way to the abscission layer. In this way the
possible time of pollen-tube growth on the stigma is limited to a rela-
tively short period. The significance of this will be emphasized in
cormection with the discussion on the rate of tube growth.
THE ACTION OF RAIN T-TPON THE STIGMA
The prevailing belief among fruit growers is that the chief injury of
rain to the stigma, aside from washing pollen from it, is the dilution of
the stigmatic fluid to such an extent that the growth of the pollen tube
is prevented. Immediately after a heavy rain during full bloom on
May 9, 1 91 8, a study of stigmas under orchard conditions showed that
even those which were past the receptive stage, dark brown in color and
partially dead, were distinctly moist and turgid.
Following these observations an investigation was made of the action
of water upon the stigma. When one which had been receptive for
about three days was dipped in water and carefully withdrawn, a small
droplet about the size of the stigma adhered. This droplet was absorbed
in approximately one minute. The dipping was repeated eight con-
secutive times in as many minutes, and in each case the droplet was as
quickly absorbed as in the first instance. As a result of the absorption
of water the papillate cells became distinctly turgid. A similar test was
made with an unreceptive stigma and also one which had passed the
receptive stage and of which the papillate cells had become dark brown
and partially dead. The same imbibition of water took place with
these two as with the receptive stigma.
It will be evident that absorption of water in such quantities acts to
dilute the cell sap of the papillate cells. This, however, would appear
to be of no immediate consequence, since pollen does not take up the
stigmatic fluid until it is secreted, and even if pollen in this way came
June i6. 1919 Relation of Weather to Fruit fulness in Plum 117
in contact with water before the stigmatic fluid, this, as has been shown,
would not be prohibitive of subsequent normal development. Further-
more, since tests show that germination takes place in a considerable
range of concentration in a sugar solution, a partial dilution of the stig-
matic fluid as a result of water absorption would probably not alone
prohibit tube growth. Under greenhouse conditions and in the orchard
under favorable conditions a stigma, like a leaf gland, has more than
one period of active secretion. If the first fluid to be secreted was com-
pletely removed by rain, it would be again renewed under favorable
conditions, so that a short rain alone would not necessarily be detri-
mental. Even if the secretion were considerably diluted following a
rain, evaporation from the surface would result in a gradual concentra-
tion. Furthermore, the influence of rain upon the stigmatic secretion
could be considered of more importance if the stigma had only a single,
short period of activity.
WASHING OF POLLEN FROM THE STIGMA
The adherence of pollen to the stigma was first noticed in pistils which
had gone through the washing and numerous changes of solution in the
preparation for sectioning by the usual cytological procedure. Stigmas
which had passed through a 2 -day rain, in addition to the cytological
process, still held as many as 40 to 50 pollen grains.
An examination under orchard conditions of stigmas which had been
subjected to a heavy rain of over 14 hours duration, showed that most
of the stigmas still retained a considerable quantity of pollen (PI. 14, B).
On one stigma 42 grains were counted. On another, which had passed
through a 2 -day rain while in bloom, there were 32 pollen grains, and 6
days afterward on still another there were 176. However, in the last
instance a part or all of the pollen could have reached the stigma after
the rain.
In order to determine how readily pollen can be washed away, an
abundance of pollen was placed on a stigma which was then immersed
in water, the results being observed with a binocular miscroscope. At
the first impact of the water a few of the outlying grains were washed
away, but at the end of 10 minutes of vigorous stirring and dipping in a
pail of water, 73 grains still adhered to the stigma. While the number of
grains at the start was not counted, it was estimated that less than one-
fourth were lost. The outstanding fact is that not all of the pollen
was removed by a washing action, certainly as vigorous if not as pro-
longed as a rain.
An explanation of the adhesion of pollen is found in the condition
of the respective stigma. In some of the fixed preparations there is a
slight staining area beyond the terminal cells of the stigma (PI. 14, A
and B), in depth about equal to the thickness of two or three pollen grains.
This undoubtedly represents the area in cross section of the stigmatic
II 8 Journal of AgricultMral Research voi. x\ai, no. 3
jfluid. Sections of stigmas show, as mentioned above, that during
the later stages of receptiveness pollen may be even partly sunken
in among the terminal cells. This, together with the gelatinous or
viscous nature of the stigmatic fluid, expecially some time after receptive-
ness, largely accounts for the difficulty in washing pollen from the stigma.
Also, the inward movement of water would partly counteract the washing
action, especially of light rains. In addition, during the early stages of
pollen germination the tubes tend to prevent pollen from being washed
away. At this time, however, the tube becomes the important considera-
tion instead of the pollen.
All pistils are not subjected alike to rain action. Those on the upper
side of limbs and in terminal positions receive the direct impact of rain,
while those in the more protected positions, as in the interior parts of the
tree and on the under side of clusters, are shielded from the direct force of
the rain.
It will appear from the foregoing that pollen is not so completelv washed
away by rain as has heretofore been supposed. This belief has become
general on account of the changes which take place in pollen when
it is placed upon a receptive stigma. Immediately upon coming in
contact with the stigmatic fluid, pollen becomes turgid and is more or less
immersed in it. Under these conditions its appearance closely resembles
that of the terminal cells of the stigma. If a dilution of the stigmatic
fluid and the washing away of pollen are the important inhibiting factors
in the setting of fruit, a short dashing rain at blooming time would,
at certain stages, do as much dam^age as a prolonged rain, because it
would be necessary for the pistil to again become receptive and pollination
to again take place. This, however, does not correspond with the general
observations of fruit growers nor with the conditions reported here.
LIMITATIONS UPON FERTILIZATION
If the statements regarding the effect of rain upon pollen and stigma
are correct, the failure of the plum to set fruit during unfavorable weather
conditions will have to be explained in another way. At the time the
pollen and pistil are maturing and functioning other factors are operating
which place certain definite limits upon the time fertilization is possible.
On account of self-steriUty, the relative time of dehiscence and
receptiveness within the variety is not an important factor in the plum.
However, because the pollen is mature before the stigma and virtually
in a "resting stage" protected by a thick covering in addition to the
anther wall, it is less susceptible to injury than the stigma, in which
growth changes are still taking place. This difference in the relative
maturity of the two structures may largely account for the greater
hardiness of pollen during frosts. Upon germination the pollen enters
a phase of less resistance, and it shares to a greater extent the lot of the
stigma and style, which constitute the substratum for the pollen tube.
junei6, I9I9 Relation of Weather to Fruit fulness in Plum 119
The factors, then, which place a time limit upon the mutual functional
period and which have a direct bearing upon the setting of fruit are (i)
the longevity of the pollen, (2) the length of the receptive period and life
of the stigma, (3) the abscission of the style, (4) the rate of the pollen-
tube growth, and (5) the influence of low temperature upon pollen
germination.
THE LONGEVITY OF PLUM POLLEN
The results of Sandsten (14) showed that plum pollen collected from
such widely separated sources as Washington, Missouri, Tennessee, and
Minnesota retained its germinating power for six months when subjected
to the normal humidity and temperature changes incident to the period
of the test. There was a gradual decline, however, in the percentage of
germination from an average of 54 per cent at the end of the first month
to about 8 per cent at the end of the sixth. Furthermore, relatively
adverse conditions do not affect the longevity of the pollen, since short
exposures to water do not kill it and freezing temperatures only retard
germination. Under favorable conditions, therefore plum pollen retains
its viability considerably longer than it is functional under orchard
conditions.
LENGTH OF RECEPTIVE PERIOD AND LIFE OF THE STiGMA
As has been noted, the plum stigm.a is receptive under orchard con-
ditions for a maximum of one week but begins to turn brown at the end
of approximately three to five days. Adverse weather conditions may,
however, extend the functional period somewhat, particularly when
accompanied by low temperatures. The delay in pollination up to a
certain point does not prevent tube growth. Crosses were successful in
the greenhouse on stigmas which were receptive four days previous to
the application of pollen. Under these conditions, however, drying and
browning does not take place as quickly as in the orchard where the active
period of secretion is over at the end of three to five days and is followed
by a period of partial inactivity of the stigma.
Furthermore, the stigma is more easily dried by the wind late in the
receptive stage than immediately after becoming receptive. Tube for-
mation would undoubtedly be more uncertain if pollination were delaved
until late in the receptive period, as would be the case during a prolonged
rain. Pollen germination, as well as considerable tube growth, must,
therefore, take place if fertilization is to be effected within a relatively
short time and before the conditions of the stigma prohibit tube growth
or before dying back in the style overtakes tubes which have been formed.
ABSCISSION OF THE STYLE
The styles do not begin to absciss until about two weeks after blooming
(PI. 14, E), although the abscission layer at the point of abscission near the
I20 Journal of Agricultural Research ' voi. xvii, No. 3
base becomes very distinct in some varieties, as Winnipeg (P. nigra), as
early as 8 days after blooming. In this variety at the 8-day period the
cells in the abscission layer had reached an advanced stage in their
disintegration, and while the style was still persistent, it was much
lighter in color above the point of abscission, a condition which suggests
the cutting off of food material. If the pollen tube had not passed the
abscission layer by this time, it is probable that it would not have done so,
since it would have had to grow through a region of partly disintegrated
cells. Consequently all tubes which had not passed the abscission layer
by the time of the abscission of the style (Pi. 14, F) would be definitely
eliminated as far as fertilization is concerned. Tube growth from the
stigma to the abscission layer, therefore, must take place between the
beginning of receptiveness and the shedding of the style.
If pollination occurs late in the receptive period, the condition of the
stigma begins to change so rapidly that only favorable growing conditions
for the tubes will enable them to pass the abscission layer before the
style drops. In this way the abscission of the style sets a definite time
limit to a certain minimum of tube growth which may be as short as 4
days and as long as 1 2 . It will be clear then that the later in the receptive
period pollination takes place and the more tube growth is retarded, the
more uncertain fertilization becomes.
RATE OB' POLLEN-TUBE GROWTH
It will be seen from the above that the rate of pollen-tube growth
becomes an important factor in fertilization, especially during unfavorable
weather accompanied by rain and low temperatures. In order to de-
termine the rate pollen tubes advance down the style, this point has been
studied in fixed preparations of pistils taken under orchard conditions
and also from controlled crosses in the greenhouse where the time of
pollination could be determined definitely. The greenhouse temperature
during this experiment was not recorded, but varied from 55° to 65° F.
Pistils from the orchard in all cases were collected after a period of vari-
able weather of rain and low temperatures. The results showing the
extent of tube growth under different conditions are presented in Table
III.
Sandsten (14), in determining the time required for the pollen tube to
reach the ovary, cut the pistils off controlled crosses at interv^als of 48
and 60 hours, respectively. From the data he presented it appears
that the plum is fertilized at the 60-hour period. It should be stated,
however, that this shows that tube growth had merely extended below
the point at which the style was cut in that time. The 7-day period,
at which time the final observations were made, is too soon to determine
certainly whether fertilization has taken place judging from size alone.
June i6, 1919
Relation of Weather to Fruitfulness in Plum
121
Table III. — Rate of pollen-tube growth found in the plutn in controlled crosses in the
greenhouse and under orchard conditions
UNDER GREENHOUSE CONDITIONS
Cross.
Minn. No. 10 "XMinn. No.
12."
Minn. No. 10 K<P. Besseyi .
Minn. No. 12 '■XMinn. No.
21.*
Do
Minn. No. 12 "XP. Besseyi .
Minn. No. 21 '•XMinn. No.
Minn. No. 21 ''XMinn. No.
10."
Minn. No. 6 " X Surprise . . , .
Time.
16 hours
do
9:30 a. m. to 3:30 p. m.
17.5 hours
ig hours
51 hours
69 hours
6 days
Tube growth in greenhouse.
3^^ of style length. Cross
sterile.
'2 of style length.
None.
Vio of style length.
>4 of style length.
Do.
Vio of style length.
Full style length. None
fertilized.
UNDER ORCHARD CONDITIONS.
Minn. No. 21 '•Xopen-pol
linated.
Minn. No. 35 ^Xopen-pol
linated.
Do
Minn. No. 12 ''Xopen-pol
linated.
Do
Minn. No. 12,^ selfed
Do."
Minn. No. 6, * selfed
Manitoba, selfed
P. Besseyi, selfed
Surprise, selfed
3 days after blooming .
do
6 days after blooming .
do
10 days after blooming
4^2 hours
24 hours
2 days
8 days
12 days
6 days
No tube growth. Rain
and frost.
Do.
% of style length.
Do.
Tube in embryo sac.
Tube just formed.
Vio of style length.
Do.
X oi style length.
5 3 of style length. Ovule
' aborted.
Vio of style length.
» A cross between Burbank and Wolf. *> A cross between Abundance and Wolf. <^ See PI. 14, A.
From Table III it appears that pollen-tube growth is relatively slow in
the plum and that the time required for the tubes to reach the ovary is
much longer than Sandsten estimated. Furthermore, it should be
emphasized that in the above table the maximum tube growth is gi\-en.
It will be seen in the case of Miimesota No. 21 and No. 35 that there
was no tube growth three days after blooming when open-pollinated under
orchard conditions. The weather conditions previous to the time stigmas
were collected from these two varieties will be of interest here. Both
came into bloom on May 20, 191 7, which was clear, with a maximum
temperature of 62° F., with a slight rain in the evening, and a medium
wind the latter part of the day. At night the temperature fell and
there was frost. May 2 1 was cloudy, with a heavy rain accompanied by
a strong wind lasting from early morning up to 2 p. m. May 22 was cool
and clear, and the stigmas of these two varieties were collected in the
early forenoon.
122 Journal of Agricultural Research voi. xvn, no. 3
The time of pollination is uncertain^ but bees were present in large
numbers on May 20. On a single stigma of Minnesota No. 35 there were
162 pollen grains, mostly embedded in the stigmatic fluid. There
were fewer grains on the stigmas of Minnesota No. 21. In the field
records, made at the time of fixing this material, it was stated that the
"stigmas were brown in all cases and dead in some." From this it will
be seen that the receptive period was much shorter than is common in
the plum. The condition, then, in these two varieties was as follows:
(i) Pollination had taken place, (2) on the third day after bloom no
tubes had formed in the stigmas examined, and (3) the end of the recep-
tive period had been reached.
On each variety the dying back in the styles averaged 5 mm. by
May 31, and by June 2, 13 days after bloom, the abscission layer was
fully formed and disintegration of the cells in it had started. On this
date additional pistils were collected and fixed, and in these pollen tubes
could not be found in the micropyle, nor had embryos formed in any
of the six which were sectioned. This is not surprising when it is noted
that under the favorable conditions of the greenhouse, Surprise pollen
tubes required six days to grow the full length of the style.
These trees under observation were" 8 years old from planting and were
under clean cultivation. On Minnesota No. 21, 25 per cent of the
buds were winterkilled and only 5 per cent of the flowers set fruit; on
Minnesota No. 35, 10 per cent were winterkilled and the percentage of
fruit to set was 10. On each there was a light crop of ripe fruit.
In the case of these two varieties, then, the small percentage of fruit
to set is not necessarily due to a lack of pollination, but apparently to
the delay in tube formation, during which the stigmas turned brown
and some died, conditions which either prevented or delayed tube
growth. According to this, in those fruits which set, tube growth had
either started on the 20th, before the rain, or was sufficiently rapid
after it to pass the abscission layer before the style fell. The weather
conditions for this season are analyzed in Plate 15.
From Table III it will be further seen that under the favorable grow-
ing conditions of the greenhouse, the rate of tube growth is so slow that
the abscission layer is passed dangerously near the time of dehiscence.
In the orchard, however, during the most suitable conditions, fully as
many fruits set as in the greenhouse, and it is very probable that the tube
extension is even more rapid.
The bearing of low temperatures upon the status of tube growth
noted above warrants further discussion. The lower temperature limit
of pollen germination in the plum was determined by Goff (5), as pre-
viously noted, to be approximately 40° F. At 70° F. there was an
abundance of tube growth, and at 51° F. the rate of growth was inter-
mediate between the two extremes. Entering the factor of humidity in
relation to temperature, his experiments further show that there was
junei6, J919 Relation of Weather to Fruitfulness in Plum 123
greater germination after five days, when pollen was kept in saturated
air in a refrigerator (the temperature is not given), than under the same
conditions at room temperature. This being the case, the cooler tem-
peratures usually accompanying prolonged rains would be more favor-
able to a higher percentage of germination than higher temperatures.
From Plate 15 it will be seen that each season the minimum tempera-
ture falls below the lower limit of tube growth a number of times and
occasionally the lower limit of tube growth is even approached by the
maximum temperature. It is probable that the temperature influence
on tube growth would be similar to that on tube formation.
The slow pollen-tube extension found under greenhouse conditions
serves as a basis for estimating what can be expected during periods of
low spring temperatures. That low temperatures have a much greater
influence some seasons than others is clearly shown by the extent the
minimum-temperature curve extends below the line of no tube growth
(5) drawn through each graph (Plate 15) at 40° F. The tem.perature
factor, therefore, has an important bearing upon the extent to which
fertilization fails to take place some seasons. While cool weather re-
tards tube growth, it does not appear to change rhaterially the time of
abscission of the style.
RELATION BETWEEN THE WEATHER AT BLOOMING AND THE
SETTING OF FRUIT
With the foregoing analysis of weather in mind, it now remains to be
seen whether there is any correlation between the v/eather conditions
prevalent at bloom and the setting of fruit. While an ample set of fruit
does not certainly insure a full crop, a full crop can not be obtained
unless there is a set up to a certain point. In this way the weather
determines the possibility of a crop.
During the years 1915, 191 6, and 191 7 there was a light set and a
light crop of plums at the Fruit-Breeding Farm. An inspection of
Plate 15 shows that different weather combinations occurred during
each of the three years. In 191 5, the outstanding features are the
frequent rains during bloom and the low-temperature period for one
week following. This single factor, according to the work of Goff (5)
on the temperature limits of tube growth, would make fertilization un-
certain, but it will be noted that following the cloudy, rainy weather of
the first four days of bloom there were two days of unusually windy
weather which interfered with bee flight at a critical time, and hence
rendered ample pollination uncertain. The following year, 191 6, bloom
was nearly a month later and was accompanied by a period of unusually
high temperature which extended to the period of tube growth. This
alone would have been very favorable to pollination, but during early
bloom there were two unusually heavy rains and five lighter ones.
Moreover, aside from actual injury to the bloom during such rains as
124 Journal of Agricultural Research voi. xvii, No. 3
occurred on May 21 and 25, as well as the interference with insect flight,
pollination would appear to be uncertain because pollen was not avail-
able for dissemination a large part of the time. This year, therefore, it
appears that pollination was uncertain instead of fertilization, as was
the case the year before. At any rate, during these two seasons the
temperature at bloom was very different. In 191 7 rain, high winds, low
temperatures, and even frost, were prevalent during bloom, and at the
close of bloom there were nearly 3 days of cool, rainy weather which
came at a critical time during tube growth. In addition to this, frequent
rains and a relatively low temperature at the latter part of the lo-day
period following bloom supplemented the retarding effect of the 3-day
rainy period. The wind on May 20, 21, 22, and 26 was strong enough to
interfere with the work of bees. Both pollination and fertilization were
uncertain this year.
In contrast to the slight set of these three seasons there was a good set
in 1912, 1913, and 1918, and a heavy set in 1914. It now remains to be
seen whether there were conditions at bloom these seasons which differ
markedly, as far as the influence on pollination and fertilization is
concerned, from the others. In 191 2 the temperature was relatively
high, except for three days, during the entire period. The rains were
slight at bloom. Also, in 1913 the temperature was within the range
of fast tube growth a good part of the time and rains were unusually
scant at bloom. The unusually high temperature in 19 14 is in marked
contrast to the low temperature the following year, and in the absence
of heavy rains there was the greatest setting of fruit as well as the heaviest
crop of all season included. The high temperature at the beginning of
bloom in 1918 gradually fell toward the end and there was a frost the
night of May 12. The rains were not prolonged during bloom, but the
heavy rain of May 9 delayed pollination in the later blooming varieties.
The warm period following bloom, however, counterbalances the cooler
4-day period at the end of bloom, so that the rate of tube growth was in
general increased. The setting of fruit was sufficient for a good crop this
season.
It will be seen from this brief analysis that there are conditions each
season v/hich can be correlated with the set of fruit. With a light set
it is impossible to get a heavy crop. As early as the 5- or 6-week period
the possibilities of a crop are determined.
SUMMARY
(i) Unfavorable weather at blooming time may completely prevent
the setting of fruit in the plum, even though there be a full bloom, A
study of the manner in which weather affects the processes at bloom
shows that rain and low temperatures are the most important factors,
although strong winds when prolonged are also important.
jimei6. I9I9 Relation of Weather to Fruilfulness in Plum 125
(2) Wind has its influence indirectly by interfering with insect action
and, hence, pollination at critical times. It is seldom strong enough to
cause much direct mechanical injury. The experiments of Waugh show
that wind pollination is insufficient, even under the most favorable
conditions. Frosts during bloom are only occasional and injure the
pistil more than pollen. The greatest damage from low temperatures is
in the retarding of pollen-tube growth. Other conditions being favorable,
cloudiness does not prevent the setting of fruit. Rain prevents pollen
dissemination by closing the anthers or by preventing them from opening,
but does not burst pollen nor kill it.
(3) On account of the adhesive action between stigma and pollen,
rain does not completely wash pollen from stigmas. The stigma is
receptive for 4 to 6 days, and following the active period of secretion the
stigmatic cells rapidly disintegrate. The style abscisses in 8 to 12 days
after bloom. Tube growth appears to be relatively slow in the plum
even under favorable greenhouse temperatures. As a result of the
rapid disintegration in the stigma and the abscission of the style, a delay
in pollination or slow tube growth when the temperature is low renders
fertilization uncertain.
(4) An analysis of the prevailing weather at bloom shows that each
season certain sets of conditions can be singled out as being largely
responsible for the status of the setting of fruit. In one season rain
during bloom may be the limiting factor and in another low temperature
during the period of tube growth. Unfortunately, practical remedies
under orchard conditions do not appear readily available. Late blooming
has not escaped unfavorable weather, and, since tube growth seems to
be the process most directly affected by low temperatures, remedial
measures can most effectively be sought in suitable pollinizers which
show the fastest tube growth.
LITERATURE CITED
(i) Backhouse, W. O.
1912. THE POLUNATioM OF FRUIT TREES. In Gaxd. Qiron., s. 3, V. 52, no.
1352, p. 381.
(2) Ballantyne, a. B.
i913. blooming periods and yields of fruit in relation to minimum
TEMPERATURES. Utah Agr. Exp. Sta. Bui. 128, p. 243-261.
(3) Beach, S. A., and Fairchild, D. G.
1893. THE EFFECT OF RAINFALL LTON POLLINATION. NOTE ON PRELIMINARY
EXPERIMENTS. In N. Y. State Agr. Exp. Sta. nth Ann. Rpt. 1892,
p. 607-611.
(4) GoFF, E. S.
1894. FLOWERING AND FERTILIZATION OF THE NATIVE PLUM. In Gard. and
Forest, v. 7, no. 332, p. 262-263.
(s)
1901. A STUDY OF CERTAIN CONDITIONS AFFECTING THE SETTING OF FRUITS.
In Wis. Agr. Exp. Sta. i8th Ann. Rpt. [i90o]'oi, p. 289-303, fig. 61-S0.
126 Journal of Agricultural Research voi. xvii. No. 3
(6) Halsted, Byron D.
1890. REPORT OF THE BOTANICAL DEPARTMENT. INFLUENCE OF RAINFALL AT
BLOOMING-TIME UPON SUBSEQUENT FRUITFULNESS. In N. J. Agr.
Exp. Sta. nth Ann. Rpt., 1890, p. 330-332.
(7) Hedrick, U. p.
1908. the relation of weather to the setting op fruit; with blooming
DATA FOR 866 VARIETIES OF FRUIT. N. Y. vState Agr. Exp. Sta. Bui.
299, p. 59-138.
(8)
1915. THE BLOOMING SEASON OF HARDY FRUITS. N. Y. State Agr. Exp. vSta.
Bul. 407, p- 365-391-
(9) Heideman, C. W. H.
1895. classification of the sexual affinities of prunus americana var.
In Ann. Rpt. Minn. State Horticultural Soc, v. 23 (Minn. Hort., v.
23, no. s), p. 187-195, illus.
10) Kenoyer, L. a.
I917. THE WEATHER and HONEY PRODUCTION. la. Agr, Exp. Sta. Bul. 169,
26 p.
11) Lord, O. M.
1S94. NATIVE PLUMS. In Ann. Rpt. Minn. State Hort. Soc, v. 22 (Minn.
Hort., V. 22, no. 2), p. 62-65.
12) Phillips, Everett Franklin.
1915. BEEKEEPING ... 457 p., front., illus., pi. New York, London.
13) Sandstsn, E. p.
1906. conditions WHICH EFFECT THE TIME OP THE ANNUAL FLOWERING OP
FRUIT TREES. Wis. Agr. Exp. Sta. Bul. 137, 21 p.
14)
1909. SOME CONDITIONS WHICH INFLUENCE THE GERMINATION AND FERTILITY
OF POLLEN. Wis. Agr. Agr. Exp. Sta. Research Bul. 4, p. 149-172, 5 fig.
15) Waite, Merton B.
1894. THE POLLINATION OP PEAR FLOWERS. U. S. Dept. Agr. Div. Veg. P?th.
Bul. 5, 86 p., 2 fig., 12 pi.
j6) Waugh, F. a.
1898. report op the horticulturist. problems in plum pollination.
In Vt. Agr. Exp. Sta. nth Ann. Rpt., 1897/98, p. 238-262, illus.
17)
REPORT OP THE HORTICULTURIST. THE POLLINATION OF PLUMS. In Vt.
Agr. Exp. Sta. 12th Ann. Rpt. 1898/99, p. 189-209, illus.
I901. REPORT OP THE HORTICULTURIST. FURTHER WORK, IN PLUM POLLINA-
TION. In Vt. Agr. Exp. Sta. 13th Ann. Rpt. 1899/1900, p. 355-362,
3 fig-
PLATE 13.
Plum tree and fruiting branch showing difterence between number of flowers borne
and quantity of fruit set:
A. — The appearance of a plum tree bearing a normal crop of bloom.
B. — ^A single fruiting branch 2 years old showing the contrast to A. Only 2 fruits
have set out of approximately 100 flowers borne by this branch. Note the stubs where
flowers have dehisced.
Relation of Weather to Fruitf ulne<'.s in Plum
Plate 13
Journal of Agricultural Research
Vol. XVII, No. 3
Relation of Weather to Fruitfulness in Plum
Plate 14
Journal of Agricultural Research
Vol. XVII, No. 3
PLATE 14.
A. — Stigma of Minnesota No. 21, a greenhouse tree, 24 hours after being selfed,
showing the condition of papillate cells in the stigma, pollen tubes, and also traces of
the stigmatic fluid.
B. — Stigma of Minnesota No. 35, open to cross pollination, showing the condition of
a stigma three days after bloom, having withstood a rain of 0.87 inch which fell in the
two days previous, lasting in all 18 hours. Note the slight staining area of the stigmatic
fluid in which two pollen grains are embedded.
C. — The turgid papillate cells in Sapa before receptiveness.
D. — Opata. Same as C. Pollination has not yet taken place.
E. — ^Abscission layer Minnesota No. 35, showing the cells of the layer 11 days after
bloom,
F. — The surface at the abscission layer of Assiniboin after the style has fallen, 12
days after bloom. There appears to be no marked disintegration of the cells imme-
diately below the abscission layer, which suggests that in cutting off the style by this
method the breaking down of the middle lamella is restricted to a few cell layers.
108123°— 19 4
PLATE IS
Graphic analysis of the weather from the standpoint of wind, sunshine, rain, and
temperature for seven years from 1912 to 1918. The maximum and minimum tem-
perature range is given for each day during bloom and for a period of 10 days
afterwards.
Relation of Weather to Fruitfulness In Plum PLATE 15
6 7 6 9 /O II /2 J3 14 15 16 i7 18 19 20 21 3Z 23 24 25
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
—
Ma
r
] ,
r-
s
\JUNE. 1 1
-
y
y
^
u.
/
A
\
^^Z>s^
19K
A^
"
\
^
-
<^
■
'^
U=:
'A
^>
1
^.csr:
T<^^Z>
r-—
^
J
/
V
^. .
• *~^
L 1
1
"- J
_ T_E
IB E
H
■
r 1 III ^H
■■
n ^HH
',1 1 1
1/<J
8
0
24 25 26 27 28 29 30 I 2 3 4 5 6 7
9 /O 1/ 12 13 14
f 4 5 6 7 8 9 10 II I a /3 14 15 16 17 IB 19 cO 21 22 23 24 <
Journal of Agricultural Research
Vol. XVII, No. 3
STRUCTURE OF THE MAIZE EAR AS INDICATED IN
ZEA-EUCHLAENA HYBRIDS
By G. N. Collins
Botanist, Office of Acclimaiization and Adaptation of Crop Plants, Bureau of Plant
Industry, United States Department of Agriculture
INTRODUCTION
In attempting to trace the origin of maize {Zea mays L.) the center of
interest is the evolution of the pecuHar form of inflorescence, especially
the pistillate inflorescence, or ear.
Since Euchlaena {Euchlaena mexicana Schrad.) or teosinte, the nearest
known relative of maize, has a very different type of pistillate inflorescence,
it may be instructive to compare the two genera and trace the successive
changes that would be required in passing from the Euchlaena form of
pistillate inflorescence to the maize ear.
Euchlaena and maize cross freely, resulting in intermediate hybrids
which in subsequent generations grade back to the parental forms (PI. i6).
It is therefore possible to present a complete series of intermediates, gradu-
ated to any desired degree of minuteness. It should be kept in mind that
although we may be able to arrange a continuous series of forms ranging
from Euchlaena to maize, these forms may not represent the course of
evolution. A study of these intermediate hybrids may be expected,
however, to throw light on the morphology of the ear and to explain its
evolution, at least in a mechanical sense.
DESCRIPTION OF MATERIAL
The fonns here described as intermediate between maize and Euchlaena
appeared for the most part among the descendants of a cross between
Florida teosinte and a diminutive variety of popcorn called "Tom
Thumb." Of this cross six first-generation plants were grown and from
the self -fed seed of one of these a second generation consisting of 127
plants was produced. Several hundred third-generation plants from open
pollinated seed were also examined.
Although in general appearance the pistillate inflorescences of maize
and Euchlaena are so unlike that comparisons are difficult, the structure
of the flowers is practically identical. The chief differences are therefore
to be sought in the structure of the inflorescence and the arrangement of
the spikelets.
Journal of Agricultural Research, (127) Vol. XVII, No. 3
Washington, D. C. June 16, 1919
rx Key No. 0-175.
128 J our7ial of Agricultural Research voi. xvii, no. 3
To avoid circumlocution it is necessary to consider as a morphological
unit the association represented by a sessile and pedicelled spikelet, as
they occur in the staminate inflorescence. It would be misleading to
refer to this unit as a pair of spikelets, because the same unit must also
be kept in mind in considering the pistillate inflorescence where one of
the spikelets may be suppressed. The two spikelets of a pair probably
arise from a single metamer, at least they seem never to become sepa-
rated. In the pistillate inflorescence, however, the individual metamers
can be distinguished with difficulty and the pairs of spikelets become so
profoundly and diversely modified that a general term is needed to
designate this structural unit in all its forms.
In the pistillate inflorescence the members of this morphological unit,
whether it is represented by one or two spikelets, occupy a single
alveolus, and the complex might be described as the contents of an
alveolus. In the staminate inflorescence, however, the depression in
which the spikelets are borne is usually too slight to be termed an alveolus.
It seems desirable, therefore, to derive the general term from some word
that carries the same implication as alveolus but which has not been used
in a specific morphological sense. The word alicole ^ is proposed and will
be used in the following description to designate the spikelet or spikelets,
whether staminate or pistillate, that are borne in a single alveolus or at a
single point on the rachis, considered as the axil or point of attachment
of a reduced branch.
The principal differences between the pistillate inflorescences of Zea
and Euchlaena may now be contrasted as follov.'s:
Euchlaena Zea
Single spikelets Paired spikelets
Two-ranked alicoles Many-ranked alicoles
Separate alicoles Yoked alicoles
SINGLE AND PAIRED SPIKELETS
The difference between single and paired spikelets will be best under-
stood by considering first the arrangement of the spikelets in the stami-
nate inflorescence of Euchlaena, which is identical with that of the
lateral branches of the staminate inflorescence of maize. Since Euch-
laena lacks the specialized central spike of the maize tassel it may be
taken to represent the primitive arrangement of the spikelets.
In these staminate inflorescences each alicole consists of two spikelets —
one sessile, the other pedicelled. The alicoles are disposed on the two
sides of the branch, leaving the lower, and, to a less extent, the upper side
of the branch, naked. The sessile spikelet is borne slightly below the
pedicelled, that is, toward the abaxial side of the branch. Thus when
viewed from the end of the branch the arrangement of the spikelets
would be such that instead of an alternation between pedicelled and
1 Ala, armpit -t-co/t), inhabit.
June i6, 1919
Structure of Maize Ear
129
sessile spikelets, the two sessile spikelets would stand next each other
as would the two pedicelled spikelets (see fig. i, A). This lack of
radial symmetry will be shown to be a very persistent and important
feature.
Fig. I. — Diagram showing arrangement of pedicelled and sessile spikelets in A, undifferentiated four-rowed
branch; E, eight-rowed ear, the result of the fasciation of two tmdifferentiated branches; C, eight-rowed
ear the result of twisting a single undifferentiated branch; D, i6-rowed ear, the result of fasciation;
E, i6-rowed ear, the result of a further twisting of "C."
Turning now to the pistillate inflorescences a striking contrast appears.
Both spikelets are sessile in Zea, and it is usually impossible to deter-
mine which of the pair is the homologue of the pedicelled spikelet. In
the pistillate inflorescence of Huchlaena, the spikelets are borne singly
instead of being paired. It is the pedicelled spikelet that is suppressed,
as is clearly shown in the hybrid plants where all stages of suppression
can be observed (PI. 17, A). I<'urthermore, in pure Euchlaena starainate
branches frequently have pistillate spikelets at the base. In such speci-
mens at the place v^^here the transition occurs, rudiments of a pedicelled
staminate spikelet can often be seen by the side of a sessile spikelet
bearing a well-developed seed.
130 Journal of Agricultural Research voi. xvri. No. 3
TWO-RANKED AND MANY-RANKED ALICOLES
The manner by which the number of rows has been increased iu the
pistillate inflorescence of maize has been the subject of much contro-
versy. Two ways of developing additional rows have been suggested —
by fasciation of long lateral branches of a compound inflorescence like
the tassel/ or by the reduction of branches until each branch is repre-
sented by a single pair of spikelets.
The fasciation theory v/ould explain the ear and the central spike
of the tassel in the same way, by assuming that a many-rowed spike has
resulted from the fusion of simple spikes or branches. In the terminal
inflorescence of pure Euchlaena there is no indication of a central spike,
all the branches being similar, except that the lower are again subdi-
vided. If two of the upper branches of such an inflorescence were to
coalesce, an eight-rowed spike would be formed, and if then the pedicelled
spikelets should become sessile and all the spikelets pistillate, an eight-
rowed ear would result.
According to the second or reduction hypothesis the development of
the ear and the central spike of the staminate inflorescence is supposed
to have been accomplished through a shortening of the branches in the
upper part of an inflorescence similar to the staminate inflorescence of
Euchlaena, the branches being reduced until each was represented by a
single pair of spikelets. In apparent conflict with this view is the abrupt
transition between the uppermost branch and the lowest spikelets of the
central spike, that characterizes all normal varieties of maize. But in
the mutation known as Zea ramosa the abrupt transition is lost, so that
the branches become gradually shorter and pass by imperceptible grada-
tion into simple pairs of spikelets like those of a normal tassel. Thus,
Z. ramosa may be looked upon as representing an intermediate stage
in the formation of a central spike, and as such constitutes the chief
support of the reduction theory.
The evidence derived from hybrids of maize and Euchlaena does not
support either of these theories. On the contrary, the hybrid plants pro-
vide an unbroken series of stages connecting the Euchlaena spike with the
maize ear that clearly indicates a third method of increasing the number
of rows and forming a central spike or ear. This is by shortening and
twisting the rachis of a single spike of Euchlaena, accompanied by an
increase in the number of alicoles. The stages in this process will be
discussed in more detail later.
SEPARATE AND YOKED ALICOLES
In the pistillate inflorescence of pure Euchlaena the joints of the
rachis, each of which bears a single alicole, stand almost directly above
one another, resembling a string of triangular beads. One of the most
' The earliest published statement of the fasciation theory that has thus far come to light is an anonymous
account (Sexual flowers in Indian corn), in Meehan's Monthly, v. 3, p. 105, 1S93.
June i6, 1919 Structure of Maize Ear 131
frequent and obvious indications of admixture with maize is a shortening
of the rachis. The reduction in length, however, is not uniform but is
more pronounced in alternate internodes, with the result that the alicoles
become associated and yoked in pairs, the members of which stand nearly
opposite to each other.
In the staminate inflorescence of either Euchla,ena or the common
varieties of maize there is little indication of this yoking of the alicoles.
The pairs of spikelets stand on opposite sides of the rachis, but usually
they are equally spaced with no indication of yoking, this tendency not
even appearing in the pistillate inflorescence of the first generation of
the hybrid between maize and Euchlaena. Yoking of the alicoles is,
however, a striking characteristic of the second generation and appears
in all the stages between the four-rowed spike and a well-formed ear.
With the increase in the number of ranks of alicoles this yoking of the
alicoles into pairs is obscured, but there are evidences that it still persists
even in the fully developed many-rowed ear.
In addition to the sharply contrasted characters discussed above, the
pistillate inflorescence of maize differs from that of Euchlaena in having
the alicoles much more numerous and more closely crowded.
EUCHLAENA X MAIZE HYBRIDS
Having outlined the nature of the differences between the pistillate
inflorescences of Zea and Euchlaena, the pistillate inflorescences of the
hybrid plants may now be examined. In the first generation the spike-
lets are paired, the alicoles separate, and two-ranked. In number of
alicoles and degree of crowding they are intermediate between the
parents. This mixture of characters derived from both parents creates
the general impression that the inflorescence is intermediate.
SECOND AND LATER GENERATIONS
Treating the three contrasted characters of maize and Euchlaena as
alternative, there are eight possible combinations: (i) Spikelets single,
alicoles separate and two-ranked; (2) spikelets single, alicoles separate
and many-ranked; (3) spikelets single, alicoles }^oked and two-ranked;
(4) spikelets single, alicoles yoked and many- ranked; (5) spikelets paired,
alicoles separate and two-ranked; (6) spikelets paired, alicoles separate
and many-ranked; (7) spikelets paired, alicoles yoked and two-ranked;
and (8) spikelets paired, alicoles yoked and many-ranked. With the ex-
ception of No. 6, all of these combinations have been found in second-
generation plants and most of them in the descendants of a single cross.
To class the individuals into the above eight combinations is, however, a
very inadequate expression of the diversity. The dominance shown in
the first generation was not followed by any clear-cut segregation in the
second. On the contrary, a complete series of intermediates connected
the parental forms with respect to each of the three contrasted pairs of
characters.
132 Journal of Agricultural Research voi. xvii. xo. 3
TRANSITION FROM A TWO-ROWED SPIKEJ TO A MANY-ROWED EAR
The pistillate inflorescence of Euchlaena may be looked upon as a two-
rowed ear. In hybrids between maize and Euchlaena the initial step from
such a two-rowed ear to one with four rows may be made in two quite
different ways. The more common method is for the pedicelled spike-
lets, which are suppressed in Euchlaena, to reappear. This converts
the flat two-rowed spike into a flat four-rowed spike, the condition that
obtains in the first generation of the hybrid (PI, 17, B).
In some instances, however, another method is followed. Alternate
intemodes of the spike become shortened until the alicoles, each with a
single spikelet, are yoked in pairs, the members of which stand opposite
or nearly so. The rachis then twists until each pair of alicoles, instead of
standing over the one below, stands at right angles with the pair immedi-
ately above and below (Pi. 16, D). This results in a square four-row ear.
The pairs of alicoles are crossed and fitted into each other in a way that
has suggested the name "saddleback" for this type of spike with four
rows of alicoles.
In some instances still another step is taken before the spikelets are
doubled in the alicole. The rachis is still further shortened and twisted,
resulting in a six-rowed ear. Six-rowx-d ears are sometimes found in
which both sessile and pedicelled spikelets are developed. In such cases it
appears that the definite relation which ordinarily exists between yoked
alicoles has been lost, and starting with the flat four-rowed ear every third
alicole has slipped around so that it occupies a plane between the other
two, which in turn are slightly displaced (PI. 17, C).
Returning now to the more common form of a four-rowed ear, it is to
be noted that the spike is four-rowed and the pedicels have been short-
ened, though the distinction between sessile and pedicelled spikelets can
still be made out with certainty. The rachis also has been shortened
and forced into a series of sharp angles and as a result of such crowding
it has now begun to twist (PI. 17, A).
The next clearly marked stage is the eight-rowed ear. The shortening
of the rachis has continued, with increased crowding and twisting of the
axil, forcing the alicoles, each bearing a pair of spikelets, to slip past one
another into the unoccupied spaces of what were the upper and lower
sides of the original horizontal branch. This is again a saddleback type,
with the alicoles associated as in the square four-rowed ear described
above, though each alicole contains two spikelets instead of one (Pi. 15, F).
Intermediate stages between the flat four-rowed ear and the eight-rowed
saddleback stage can sometimes be found where the twist is not quite a
quarter turn, but all such appear to be unstable. The saddleback, on
the contrary, is stable and will sometimes be shown consistently through-
out a plant of the second generation (PI. 18).
junei6, I9I9 Structure of Maize Ear 133
A further shortening of the rachis brings about the next stage, which is
that of a lo-rowed ear. Intermediate stages are more common during
the acquisition of this stage, and when they occur the seeds, as might be
expected, are not arranged in regular rows.
With these facts in mind, the spike can be understood as composed of
opposite or yoked ahcoles, each with a pair of spikelets. These yokes
are superposed, and as crowding increases there results a further twisting
and the formation of a more complicated spiral. With seeds of a uni-
form size a compact spiral would result in the formation of longitudinal
rows, though these might not run exactly parallel to the axis of the ear,
as, indeed, they seldom do even in ears of maize.
MORPHOLOGY OF THE MAIZE EAR
It has been shown that the intermediate forms that appear in hybrids
between maize and Euchlaena afford no support for the f asciation theory.
Evidence from the ear of pure maize may now be presented.
If a number of four-rowed branches were forced together and their
axes united, the conditions found in an ear of maize might result (see
fig. I, B). There is, however, evidence in the ear itself that it is not
constructed in this way.
It is not an uncommon occurrence for an ear to drop rows. For
example, there may be 12 rows at the base and only 10 rows at the tip.
A study of how this transition is made throws light on the morphology
of the ear. In the first place, the loss is almost invariably two rows,
and both are lost at the same distance from the butt of the ear. There
is no region with an odd number of rows. A normal ear is made up of
a series of paired rows and this is usually accepted as an adequate explana-
tion of the fact that the number of rows is always even. A pair of rows
is looked upon as the fundamental structural unit of the ear, a view in
accord with the theory of fasciation. Since two rows are dropped at
once, it might be expected that the interrupted rows would be adjacent.
This would follow from the suppression of a pair of rows representing
the sessile and pedicelled spikelets arising from a single series of alicoles.
There is, however, abundant evidence to show that rows are usually
interrupted by the abortion of pedicelled spikelets only. This can be
seen in abnormal maize tassels in which the base of the central spike is
pistillate, forming in reality a section of an ear. At the place where the
transition occurs it can be seen that the sessile spikelets are more per-
sistent and produce larger seeds. ^
' With the idea of determining to what extent differentiation between pedicelled and sessile spikelets
persists in the fully developed maize ear, the weight of each of the two seeds from individual alicoles was
compared. An ear of flint corn was chosen in which the alicoles were clearly marked and the individual
seeds were carefully weighed. There were 135 alicoles with two comparable seeds. The average weight
of the individual seeds for all the seeds was 430 mgm. The average difference between the seeds of an alicole
was 21.0 mgm. ±19.5.
It would appear, therefore, that if there was any consistent difference between the weight of the seeds
borne in pedicelled and sessile spikelets in this ear, the difference must have been something less than
S per cent of the weight of the seed.
134 Journal of Agricultural Research voi. xvii. no. 3
With the recognition of the fact that the interrupted rows represent
pedicelled spikelets instead of the pedicelled and sessile spikelets of a
row of alicoles the position of the interrupted rows with respect to one
another becomes of importance in studying the formation of the ear.
Following the fasciation theory, if both of the rows of pedicelled spike-
lets of a single branch aborted leaving the sessile, we should find the
two interrupted rows separated by two remaining rows. (This may
be illustrated by reference to fig. i , D. If the two rows of pedicelled
spikelets marked a were aborted the two missing rows would be sepa-
rated by two rows.) This is not what occurs. In the examination of
many ears in which rows were dropped no instance has been found where
the dropped rows were either adjacent or separated by two rows. In
cases where the location of the dropped rows can be determined with
reasonable certainty the dropped rows are on opposite sides of the ear.
Yet they are not exactly opposite, but missing, it by just two rows. This
is what should occur if the two pedicelled spikelets were dropped simul-
taneously from a pair of yoked alicoles. It will be recalled that the dorso-
ventral arrangement of the spikelets in the original four-rowed spike re-
sults in bringing the pedicelled spikelets not exactly opposite, but sepa-
rated by two more rows on one side than on the other. (See fig. i, E.
The pedicelled spikelets of a pair of yoked alicoles are marked a. It
will be seen that they are separated on one side by six rows and on the
other by eight.)
The persistence with which ears of maize maintain an even number of
rows is therefore more wonderful than has been supposed, for it can not
be fully accounted for by the fact that the spikelets are bom in pairs. It
must, in addition, be recognized that when a pedicelled spikelet of one
alicole is suppressed there is a simultaneous suppression of the pedicelled
spikelet in another alicole. The further evidence afforded by Kuchlaena
hybrids is that the two alicoles are the members of a yoked pair which
though standing on opposite sides of the ear, have not lost their identity
as a structural and developmental unit.
SUMMARY
Before the pistillate inflorescences of maize and Euchlaena could be
compared in detail it was found necessary to recognize as a morphological
unit the organs borne by a single metamer of the rachis. This unit,
whether staminate or pistillate, whether composed ot one or more
spikelets, has been called an alicole.
The stages between a Euchlaena spike and a maize ear as they appear
in hybrids between the two genera may be summarized as follows:
(i) The suppressed pedicelled spikelet in each alicole reappears.
(2) The alicoles become more crowded and their -number is increased.
(3) The alicoles associate themselves in pairs or yokes. (4) The axis
twists, increasing the rows of alicoles.
junei6, I9I9 Structure of Maize Ear 135
The order in which these changes occur is by no means fixed, but taken
together they comprise all the changes necessary in deriving the maize
ear from the Euchlaena spike.
In this series of intermediate stages nothing was observed that affords
support for either the fasciation or "reduced branch" theory of ear
formation. There is also evidence from the maize ear itself that the
association of alicoles into pairs is more fundamental than the linear
arrangement.
In all the hybrids between maize and Euchlaena that have been ob-
served there has appeared no suggestion of either pod com or Zea ramosa.
Since it can scarcely be doubted that the peculiar characteristics of both
of these mutations represent the reappearance of ancestral characters
common to the Andropogoneae, it would seem that in crossing maize
^nd Euchlaena, and thus calling forth a series of intermediate forms, we
are not returning to the point in the ancestry of maize where it became
differentiated from the Andropogoneae.
Furthermore, if the stages shown in the hybrid plants w^ere to be taken
as indicating the path of evolution of the ear, it would be necessary to
assume that the central spike of the staminate inflorescence or tassel had
evolved separately and along different lines. The close homology
between the ear and the central spike of the tassel makes such an assump-
tion unreasonable.
In the present article emphasis has been placed on the shortening and
twisting of the axis of a single spike as a possible method of deriving a
structure like the maize ear from the inflorescence of Euchlaena. This
has been done, not because the method is believed to represent the most
probable course of evolution, but because the present discussion has
been restricted to the evidence afforded by hybrids of maize and Eu-
chlaena, which seems to require such an interpretation.
Facts of other kinds are more easliy interpreted by the theories of
fasciation and reduction of branches, but there are also facts that do not
seem to accord with any of the theories yet proposed. Until the appar-
ently contradictory evidence can be reconciled, it seems best to keep
the several possibilities in mind and await additional evidence before
attempting a complete interpretation.
PLATE 1 6
Intermediate stages between a simple spike of the pistillate inflorescence of Euch-
laena and an ear of maize:
A. — Spike of pure Florida teosinte.
B. — Spike vv'ith slightly shortened axis.
C. — A* still more compact spike with an increased number of seeds. A-C have
single spikelets and separate two-ranked alicoles.
D. — Spike with single spikelets and yoked alicoles, irregularly four-rowed.
E. — Compact spike with two-ranked separate alicoles and single spikelets.
F. — Spike with paired spikelets and four ranks of yoked alicoles.
G. — Transition stage between four-rowed and eight-rowed ear.
H. — Ear of maize with eight rather poorly defined rows of seeds.
structure of the Maize Ear
Plate 16
Journal of Agricultural Research
Vol. XVII, No. 3
Structure of the Maize Ear
Plate 17
Journal of Agricultural Research
Vol. XVII. No. 3
PLATE 17
Pistillate inflorescences of hybrid between Euchlaena and maize:
A. — Showing pedicelled staminate spikelets with sessile pistillate spikelets.
B. — Closely compacted inflorescense with two rows of alicoles and four rows of seeds.
C-E. — Spirally twisted inflorescences, with three rows of alicoles.
PLATE i8
Pistillate inflorescences of hybrid between Huchlaena and maize, showing yoked
alicoles:
A-C— The alicoles are in four rows corresponding vo an eight-rowed ear.
D. — The alicoles are in five rows, corresponding to a ten-rowed ear.
structure of the Maize Ear
PLATE li
Journal of Agricultural Research
Vol. XVII, No. 3
Vol. XVII JULY 15, 1919 No. ^
JOURNAL OF
AGRICULTURAL
RESEARCH
coNXKisnrs
p«g«
Carbohydrate Metabolism in Green Sweet Com during
Storage at Different Temperatures _ _ _ _ 137
CHARLES O. APPLEMAN and JOHN M. ARTHUR
(Contiibution from Maryland Agricultural Experiment Station)
Certain Relationships Between the Flowers and Fruits of
the Lemon --------- 153
HOWARD S. REED
(Contribution from California Agricultural Experiment Station)
Ultra-Microscopic Examination of Disperse Colloids in
Bituminous Road Materials _--.-- 167
E. C. E. LORD
(Contribution from Bureau of Public Roads)
PUBUSHED BY AUTHORITY OF THE SECRETARY OF AGRICULTURE,
WITH THE COOPERATION OF THE ASSOCIATION OF AMERICAN
AGRICULTURAL COLLEGES AND EXPERIMENT STATIONS
WASHINGTON, D. C.
WASHINQTON : GOVERNMENT PRINTINO OFFICE : Itlt
EDITORIAL COMMITTEE OF THE
UNITED STATES DEPARTMENT OF AGRICULTURE AND
THE ASSOCIATION OF AMERICAN AGRICULTURAL
COLLEGES AND EXPERIMENT STATIONS
FOR THE DEPARTMENT
KARL F. KELLERMAN, Chairman
Physiologist and Associate Chief, Bureau
of Plant Industry
EDWIN W. ALLEN
Chief, Office of Experiment Stations
CHARLES L. MARLATT
Entomologist and Assistant Chief, Bureau
of Entomology
FOR THE ASSOCIATION
H. P. ARMSBY
Director, Institute of Animal Nutrition, The
Pennsylvania State College
J. G. LIPMAN
Director, New Jersey A gricultural Experitnent
Station, Rutgers College
W. A. RILEY
Entomologist and Chief, Division of Enio-
mology and Economic Zoology, Agricul-
tural Experiment Station of the University
of Minnesota
All correspondence regarding articles from the Department of Agriculture should be
addressed to Karl F. Kellerman, Journal of Agricultural Research, Washington, D. C.
All correspondence regarding articles from State Experiment Stations should be
addressed to H. P. Armsby, Institute of Animal Nutrition, State College, Pa.
JOINALOFAGRICDLTIMIESEARCH
Vol. XVII Washington, D. C, July 15, 1919 No. 4
CARBOHYDRATE METABOLISM IN GREEN SWEET CORN
DURING STORAGE AT DIFFERENT TEMPERATURES
By Charles O. Appleman, Plant Physiologist, and John M. Arthur, Assista7it
Plant Physiologist, Laboratory of Plant Physiology, Maryland Agricultural Experi-
ment Statio)i. ^
THE PROBLEM
The present paper deals with the character and kinetics of the processes
involved in the rapid depletion of sugar in green sweet com after it is
separated from the stalk and more particularly with the relative rates of
these processes at different storage temperatures, accurately controlled.
WORK OF PREVIOUvS INVESTIGATORS
In the course of an extensive sweet corn investigation, Straughn^
clearly shows that the loss of total sugars from green sweet corn i<=- very
rapid during the first 24 hours of storage at ordinary summer temper-
atures. Working with Sto well's Evergreen, he claims that about one-
third of the total sugars disappeared during the first 24 hours oi storage
at a room temperature of about 25° C. A further loss occurred during
the next 24 hours, but when the sugars reached 1.80 per cent no further
loss was noted. This rate of sugar loss for the first 24 hours of storage
at one uncontrolled temperature must be considered as merely an approx-
imation, since analyses of different ears before and after storage were
compared. The percentage of sugar in the different ears at the time of
picking showed considerable variation. In the same paper it is con-
cluded that there is no material advantage to be gained by storing the
corn in a refrigerator. It should be noted, however, that the refrigerator
showed a temperature of 23.5° C. during the first 24 hours and there-
after 17° C.
In a later paper by Straughn and Church ^ results are reported showing
the change in the sugar content of green corn after a period of 36 hours'
storage at room temperature. The data furnish very little additional
information on this problem, as the experimental corn was secured upon
the open market and the sugar loss in this com had nearly ceased before
1 The curves in figures i and 2 were drawn by John Paul Jones, of this laboratory.
2 Straughn', M. N. sweet corn investigations. Md. Agr. Exp. Sta. Bui. 120, p. 37-7S. 1907.
'Str.\ughn, M. N., and Church, C. G. the influence of environment on the composition op
SWEET corn, 1905-1908. U. S. Dept. of Agr. Bur. Chem. Bui. 127, 69 p, 11 fig. 1909.
Journal of Agricultural Research, Vol. XVII, No. 4
Washington. D. C. July 15, igig
ry Key No. Md. -i
(137)
138 Journal of Agricultural Research voi. xvii, No. 4
the experiment was begun. However, the data are interesting; they
show the usual low sugar content of green corn as it is now purchased
on the market. The percentage of sugar in this com ranged from 1.70
to 1.49.
EXPERIMENTAL METHODS
One of the first problems to solve was a method by which the rate of
the carbohydrate changes at different temperatures could be determined
without comparing analyses of different ears. The following method
was finally adopted: The ears for each experiment were brought to the
laboratory within 15 minutes after picking and numbered consecutively.
The first set of samples was taken from ears i and 2 , and all ears were then
placed immediately under the experimental conditions. At the end of
24 hours the second set of samples was taken from ears i and 2 and the
first set from ears 3 and 4. After 48 hours the second set of samples
was taken from ears 3 and 4 and the first set from ears 5 and 6. This
overlapping method of sampling was continued every 24 hours until the
experiment was completed. The change in chemical composition dur-
ing each consecutive 24-hour period of storage could then be determined
by comparing the analytical results of the first and second sets of samples
from the same ear.
Stowell's Evergreen corn was stored at seven different temperatures —
namely, 0°, 5°, 10°, 15°, 20°, 30°, and 4o°C. All the temperatures were
controlled within about i °. The 30° temperature was controlled within
0.1°. The com was stored with the husks on, and, in the case of the
higher temperatures, the ears were placed in large desiccators, with the
tubulure on the side left open to allow ventilation. Preliminary experi-
ments showed that, as far as the carbohydrate changes are concerned,
active aeration of the small number of ears used in each experiment was
not important during the short experimental period of four days.
Under the conditions of the experiments there was very little change
in the percentage of water in the corn at any temperature. However,
the analytical results from the second set of samples were all calculated
to the moisture of the first set in order to avoid false percentages due to
loss or gain in water content during storage. In a few cases at the higher
temperatures the percentage of water slightly increased on account of
the accumulation of respiratory water and possibly water set free by
condensation of polysaccharides.
ANALYTICAL METHODS
SAMPLING
Three rows of kernels were removed for each set of samples, care being
taken to remove the entire kernel. In order to take the first set of sam-
ples, the husks were split lengthwise with a sharp knife and then cut
July 15. 1919 Carbohydrate Metabolism in Green Sweet Corn 139
half way around at the base. After the kernels were removed the husks
were brought back to place and held by means of rubber bands. For
the second set of samples the husks were removed and three rows of
kernels taken from the opposite side of the ear.
The com was thoroughly ground to a mash in a small unglazed mortar
and sampled immediately. On account of the short time required to
sample the mash it was found unnecessary to surround the mortar with
cracked ice. Each set of samples furnished material for the following
determinations: Moisture, total sugars as invert sugar, sucrose, free-
reducing substances, and starch. The starch was determined as glucose
after hydrolysis with dilute acid.
Moisture. — Approximately 5 gm. of the mash were placed between
tared watch glasses ground tight and held together by means of a clamp.
After weighing, the cover glass was removed and the material covered
with I CO. of alcohol. The samples were then dried to constant weight
in a vacuum at 80° C. During the first drying a stream of warm, dry
air was passed through the chamber. The watch glasses were clamped
together before each weighing.
Sugars. — When all things are considered, the alcohol method for the
extraction of sugars from plant material in general seems preferable to
any other yet devised. Since the procedure by different authors varies
considerably, a large number of preliminary experiments were performed
to determine the best procedure for the alcoholic extraction of sugars
from the particular material at hand — namely, green sweet corn at
different stages of maturity. The chief problem was to obtain complete
extraction of the sugars and at the same time prevent any inversion of
cane sugar as well as diastase action.
These experiments show that there is no appreciable hydrolysis of
either sucrose or starch during boiling in 40 or 50 per cent neutral alcohol
as long as 60 minutes. However, complete extraction was obtained by
a much shorter period of boiling, and consequently the loss of alcohol
during extraction is very much reduced.
The procedure finally adopted was as follows: Samples of 16 gm. each
were weighed out into counterpoised 200 cc. Kohlrausch sugar flasks.
A small amount of calcium carbonate was added to neutraUze any acids
liberated in the mash. It was latp r found that in the case of sweet com
this is not as important as in the case of many other plant tissues. The
samples were covered immediately with 75 cc. of hot 95 per cent alcohol,
the alcohol being previously measured into small boiling flasks and
brought to boil on an electric hot plate. 'After the mixture began to
boil on the steam bath, 50 cc. of hot water were added. This brought
the extraction alcohol down to about 50 per cent. The water 'in the
sample was taken into consideration in making this calculation. The
foregoing method precluded any possible enzyme action in the weak
alcohol while heating up to the boiling point. Small funnels were placed
140 Journal of Agricultural Research voi. xvii, no. 4
in the necks of the flasks to condense the alcohol and the mixture was
allowed to boil 30 minutes. While still hot, the flasks were made up to
the mark with 95 percent alcohol and allowed to stand over night. They
were then shaken, again made up to the mark, tightly stoppered, and
stored. The final strength of the alcohol in which the samples were stored
was about 64 per cent.
When a large number of samples are taken during a comparatively
short time, as was the case in this work, it becomes necessary to store
most of the samples for some time. Since the storage problem is an
important one, a number of experiments were conducted to determine
the best treatment of the samples to prevent any carbohydrate changes
during long periods of storage. The final method, previously described,
was found to preserve the samples for at least 145 days without any
appreciable carbohydrate changes. After boiling, the samples may be
safely stored in 50 per cent alcohol. Cold treatment of the sam.ples
with 52 per cent alcohol inhibited invertase action, but there was con-
siderable starch hydrolysis after a long period of storage. If the number
of volumetric flasks is limited, a measured quantity of the filtered extract
can be stored. In this work 150 cc. were frequently stored.
The method employed for the determinations of the sugars in the
solutions was essentially the same as the one described by Bryan, Given,
and Straughn.^
Starch. — Ten gm. of the mash were weighed into counterpoised 200
cc. Erlenmeyer flasks and immediately covered with 50 cc. of 95 per cent
alcohol. About 0.05 gm. of calcium carbonate was added and after
thorough shaking the flasks were tightly stoppered and stored. The
strength of the cold alcohol in the mixture was approximately 80 per
cent. This was found sufficient to preserve the samples for several
weeks without any appreciable change in the carbohydrates present.
The method of weighing out the samples in small flasks, counterpoised
on torsion balances sensitive to one-fifteenth gm. was found to give
just as good duplicates as weighing the samples to the third place in
weighing bottles. By the former method, the samples could be placed
in alcohol in a very much shorter time.
The determinations were made according to the following procedure:
Decant the preserving alcohol on to a 9 cm. No. 1 Whatman filter paper;
add 75 cc. of 50 per cent alcohol and extract 24 hours at room tempera-
ture, shaking noon and evening; decant completely the 50 per cent
alcohol; add 50 cc. more of the 50 per cent alcohol and allow to stand
two hours, shaking three times; decant the alcohol and when all has
run through the filter transfer the mash to the filter; apply suction
and drain; add 50 cc. of 50 per cent alcohol to the flask to wash down
1 Bryan, A. Hugh, Given, A., and Straughn. M. N. extraction of grains and cattle foods
FOR THE determinations OF SUGARS; A COMPARISON OF THE ALCOHOL AND THE SODIUM CARBONATE
digestions. U. S. Dept. Agr. Bur. Chem. Circ. 71, 14 p. ign-
July 15. 1919 Carbohydrate Metabolism in Green Sweet Corn 141
the sides and transfer to the filter. With small portions of 50 per cent
alcohol, transfer to the filter any material still remaining in the flask;
after the alcohol has drained out of the filter, fill up once more with
50 per cent alcohol and drain. All sugars and any other reducing ma-
terials are now removed from the residue on the filter. The filter is
filled twice with 95 per cent alcohol and the residue allowed to dry on
the filter. The filter paper may be folded over the sample and placed
in small stoppered vials for another period of storage if necessary.
The filter paper containing the sample was placed in a Kjeldahl flask
and covered with 200 cc. of water; sufiacient hydrochloric acid was added
to give a final strength of acid in the mixture of 2.5 per cent. Hydrolysis
was effected by boiling under a reflux condenser for three hours.
A number of the filter papers used for the filtration were hydrolyzed in
the same strength of acid and for the same length of time as the samples.
Although the papers were claimed by the manufacturers to be starch
free, they were found to give a small amount of reducing material after
hydrolysis. However, the amount of this material was consistent in all
the boxes and in different parts of the box, so it was very easy to make the
necessary correction for the filter paper in the final results. The starch
was determined as glucose, but of course it includes any other poly-
saccharides which furnished reducing substances during the acid
hydrolvsis. ■
EXPERIMENTAL DATA
The work had not progressed far until it was evident that if the moisture
in the com at the time of picking had fallen below a certain percentage it
became a factor in controlling the rate of sugar loss. In order to eliminate
this variable factor, so that attention could be focused upon the tempera-
ture relation, the experimental ears were carefully selected to represent
a fairly definite stage of maturity — namely, the typical milk or best
eatable stage. Ears falHng below 80 per cent water were excluded from
the final calculations.
The work of the first year was repeated on another crop the succeeding
year. The results of the two years' work are averaged in Table I. In
the experiments of the first year, the carbohydrate changes for each
consecutive 24-hour period were not determined in duplicate ears as
described for the experiments of the second year. Each percentage in
the table, therefore, is the mean of three ears, except in a very few cases
where the results of one ear were excluded on account of the moisture
content's falling below the arbitrary standard. The results of the experi-
ments at 5° and 15° C. are not given, as they add nothing to the general
conclusions. The average percentage of sugars in the corn at the be-
ginning and end of each storage period is indicated by (a) and (6),
respectively.
142
Journal of Agricultural Research
Vol. XVII. No. 4
Table I. — Loss of sugar from green sweet corn during consecutive 24-hour periods of
storage at different temperatures
ALL SUGARS
Number
of hours
in
storage.
Ear
lot.
Storage temperatures.
o°C.
10° c.
20° C.
30° c.
40° C.
Total.'
Loss.i
TotaH
Loss. I
Total.i
Loss.i
Total .1
Loss.'
Total.'
Loss.'
0
24
24
48
48
72
72
. 96
la
16
2a
26
3a
4a
Ab
5-91
5-43
6. 70
5-96
6.63
6.36
6. 10
5- 75
0.48
•74
.27
•35
5^83
4.83
3^95
3^43
4. 61
3^92
3- 54
2-93
I. 00
•52
.69
.61
6.17
4^59
3.68
2. 69
3^o7
2. 52
2. 24
1.97
1.58
•99
•55
.27
5-34
2.65
3^ii
2.68
2. 10
2.03
1-59
1.49
2. 69
•43
.07
. 10
6.72
3-64
2.30
1. 69
2. 00
I. 90
3.08
.61
. 10
0
24
la
lb
3-87
3-73
0. 14
3-77
3.00
0.77
3.68
2.54
I. 14
3.68
1.50
2.18
4-50
2. 18
2.32
24
48
2a
2b
4. 06
3-77
•29
2-53
1.99
•54
1.84
I. 17
.67
1.52
1.24
.28
I. 18
.76
•32
48
72
3b
4.49
4-25
.24
2. 74
2.30
•44
1.38
I. 12
.26
I. 02
•97
•05
I. OS
.91
• 14
72
96
4a
Ab
3-84
3-56
.28
1.87
I. 41
.46
I. II
•85
.26
•71
.67
.04
FREE-REDUCING SUBSTANCES
0
24
la
lb
1.84
1.70
0. 14
I. 8s
1.68
0.17
2. 07
1.77
0.30
1.6s
I. 16
0.49
1.98
1.32
0.66
24
48
2a
26
1.66
1-55
. II
I. 29
1.28
. 01
1.68
I. 41
.27
1.56
1.42
.14
I. 06
.80
.26
48
72
3a
3b
I. 91
1.89
. 02
1-73
1.50
•23
1-55
1.28
.27
1.07
I. 04
•03
.90
•93
. CO
.72
96
4a
Ab
2. 05
2. 00
•05
1-57
1-45
. 12
I- 13
1.04
.09
.88
.8i
.07
' Total quantities of all sugars before and after storage and losses during storage are expnessed in per-
centages.
The data in Table I show that the loss of sugar from com during
storage is not uniform, but becomes slower and slower as a final equi-
librium is approached. The relative rates of processes of this kind at
different temperatures can be determined accurately only by comparing
July IS, 1919
Carbohydrate Metabolism in Green Sweet Corn 143
the times required to perform equal amounts of work at all temperatures,
and not by comparing the amounts of work performed in equal times_
In other words, we must compare the times required at the different
temperatures to bring the process to the same stage. We are thus com-
paring stages where the ratio between the reacting material and the
products is the same. Osterhout ^ has recently emphasized this point
in a "Note on measuring the relative rates of life processes."
In order to make it possible to determine, on this basis, the relative
rates of sugar loss at the different temperatures, the experimental results
in Table I can be easily interpolated by a simple graphic method, to
be described later, if they can be expressed in curves all starting from the
same point.
This could be decided only alter a careful consideratioti of all the
factors involved. If mass action alone were responsible for the gradual
decline in the rate of sugar loss, then, at a given temperature, the average
rate of change in any unit of time would be proportional to the sugar
concentration. For a range of original sugar from about 4.5 to 7 per
cent and of water from 78 to 80 per cent this was found to be the case
for the first 48 hours of storage even at 30° C. (Table II).
Table II. — Proportion of sugar lost during first 48 hours of storage at jo° C.
Reducing
sugar
before
storage.
hoss during storage.
Total
sugar
before
storage.
Loss during storage.
Sucrose
before
storage.
Loss during storage.
Ear
No.
Actual.
Propor-
tional.
Actual.
Propor-
tional.
1
I
2
3
4
5
6
Per cent.
1.70
2. 40
2. 60
.96
1.66
3.00
Per cent.
0.68
I. 02
1.05
. 22
.82
I- 31
Per cent.
37
42
40
43
49
44
Per cent.
6. 16
7.20
6.74
4-47
5-91
6-55
Per cent.
3-59
3-87
3-88
2-59
3-41
3-69
Per cent.
58
54
58
58
58
57
Per cent.
4.24
4.80
4. 14
3-33
4-25
3-55
Per cent.
2.77
2-95
2.89
2-33
2.68
2-43
Per cent.
65
61
69
69
63
68
The sugar loss ceases when an appreciable amount of sugar is still
present. Therefore, the speed of the counter process, that is, the for-
mation of sugar, becomes a factor to be reckoned with when the processes
have nearly reached an equilibrium. If at the beginning of storage the
percentage of sugar in ear i is considerably greater than in ear 2, the
latter would reach the equilibrium position sooner than ear i. At the
end of 72 hours of storage at 30° C. ear i might still have 2 per cent sugar
while in ear 2 the sugar content might be only i per cent. The sugar
loss in ear 2 being nearer the equilibrium point, the speed of the counter
process would be greater in this ear than in ear i. Therefore, during the
next 24 hours the proportionality between the sugar lost and the sugar
' Osterhout, W. J. V. note on measuring the relative rates of life processes. In Science.
n. s., V. 48, no. 1233, p. 172-174, 3 fig. 1918.
144 Journal of Agricultural Research voi. xvii, no. 4
present would not be the same in the two ears. This was proved experi-
mentally.
In considering the rate of the counter reaction in connection with the
problem at hand — namely, the possibility of expressing the experi-
mental results in curves all starting from the same point — it must be
borne in mind that it becomes appreciable only near the point of equi-
librium, and even then it would affect the proportionality between
the sugar present and the sugar lost in different ears at the same tem-
perature only when the percentage of sugar in the ears at the beginning
of storage varied considerably.
A decrease in the quantity of active enzymes present would produce a
steady fall in the values of the velocity constants; this would cause a
decreasing rate of actual sugar loss. There is no evidence that this
occurs up to 30° C.
In view of the foregoing facts, together with the fact that the ears
selected for the final calculations were all in practically the same stage
of maturity and therefore contained nearly the same percentage of
original sugar, the following procedure in preparing the data for con-
struction of curves all starting from the same point seemed justified.
The sugar lost during each 24-hour period of storage was calculated as
proportions of the sugar present in the ears at the beginning of each
period. The percentages of sugar found in the ears analyzed at the
beginning of each experiment were then averaged. Ten ears with not
less than 80 per cent water were included in the final average.
Taking this mean as the starting point for all temperatures and apply-
ing the proportions of sugar lost during each succeeding 24-hour period,
calculated from the experimental data, a new set of proportions was
obtained, based upon the same original sugar content in all cases. A
single concrete case may serve to clarify the foregoing procedure. The
total sugar in all the ears analyzed at the beginning of each experiment
averaged 5.766 per cent. During the first 24 hours of storage at 30° C.
the average loss of total sugar in three ears was 50.28 per cent of the
initial sugar present; that is, the total sugar in the com was 50.28
per cent less than at the beginning of storage. Applying this propor-
tion to an initial sugar content of 5.766 per cent, we obtain, after
the first 24-hour period of storage at 30°, a total sugar content of
2.867 per cent. Making use of all the experimental proportions in the
same manner, the percentage of total sugar present at the end of each
24-hour period of storage was calculated, assuming that the sugar con-
tent at the beginning of storage was 5.766 per cent. Each calculated
percentage was then substracted from 5.766, the initial sugar present.
The sugar loss, expressed as percentages of the initial sugar, could then
be calculated for the following storage periods: 24, 48, 72, and 96 hours.
The same procedure was followed for all the sugars at all the tempera-
tures, with the results shown in Table III.
July 15, 1919
Carbohydrate Metabolism in Green Sweet Corn 1 45
Table III. — Sugar loss from green sivect corn during different periods of storage at different
temperatures , expressed as percentages of the same initial sugar at all temperatures
TOTAL SUGARS
24.
48.
72.
96.
Number of hours in storage.
Per cent.
8. 12
14- 51
18.03
22. 00
Storage temperature.
3°C.
Per cent.
16.98
27-95
38.71
49.22
Per cent.
25. 61
45-73
55- 50
62. 10
30° C. 40° c
Per cent.
50. 28
57-09
59. 00
61. 84
Per cent.
45-79
60. 15
62. 16
24.
48.
72.
96.
3-
51
10.
39
15-
08
21.
25
20.
78
37-
49
47-
46
60.
54
31-05
56. 12
64. 2 2
70. 16
59-42
66.76
68.55
70. 19
51-03
64.68
69. 24
KREE-REDUCING SUBSTANCES AS INVERT SUGAR
24.
48.
72.
96.
7-58
13. 61
14. 62
16.97
9-
26
10.
52
16.
71
23-
Z3
14.
72
28.
84
40.
79
45-
05
29. 96
36. 19
39-74
43- 19
33.48
49.76
49.76
The data in Table III, showing the rate of actual loss for total sugars
and sucrose, were plotted as curves (fig. i and 2).
The curve for 0° C. shows a more rapid sugar loss than is typical for
this temperature. In the first place, it required some time for the corn
to cool down to this temperature. At the end of each 24-hour period a
pair of ears were removed from the cold chamber in order to take the first
set of samples. Although the sampling period was short, the temperature
of the corn would soon rise a few degrees above 0°. The loss of sugar
at the sampling temperature is accumulative in the curve.
The inversion of sucrose appears to be the controlling process in the
sugar loss, as the curves for the decrease of sucrose are very similar to
those for the loss of total sugar.
Temperature coefficient. — Since the curves in figures i and 2 all
start from the same point, by means of a simple graphic method the
relative rates of sugar loss at the different temperatures can now be
determined by comparing the times at different temperatures required
to do the same amount of work. As an illustration we will choose a
stage in the depletion of sugar- when 40 per cent of the total sugar is lost;
in other words, at this point the sugar in the corn is 40 per cent less
than at the beginning of storage. A horizontal line is drawn from
108124°— 19 2
146
Journal of Agricultural Research
Vol. XVII, No. 4
this point tlirough all of the curves. Vertical lines are now dropped
from the points of intersection to the base line. The times in hours
required at the different temperatures to bring the sugar loss to this point
are read off on the base line (see fig. i). The procedure was repeated
for all the percentages given on the ordinate.
The relative rates of sugar loss at the different temperatures are ex-
pressed in Table IV as the reciprocals of the times in hours required to
bring the process to five different stages. The temperature coefficients
were obtained from these reciprocals. The results at 40° C. were not
10
20
« 30
40
50
60
70
^
~
1 1 1
TOTAL SUGAHS
Y
.
"^
--^
o°c
\
\
\
^\
\
\
\
\.
\.
\
\
.
^^^
^^^10°.C
-^
30° C
IlP^
20° C
24
48
H 0 U R.S
72
96
Fig. I. — Depletion of total sugars in green sweet com during consecutive 24-hour periods of storage at
different temperatures. The ordinates are given by the numbers on the left of the figure and represent
the loss of sugar expressed as percentages of the initial sugar, which was 5.91 per cent, wet weight.
included in the foregoing calculation as there was evidently destruction
of the enzymes or other alteration in the system by the high temperature.
Some of the cur\^es for the sugar loss, expecially those for sucrose,
approach true logarithmic curves; and satisfactory constants were
obtained for most of the storage period by applying the simple uni-
molecular equation. During the latter part of the period there was a
falling off in the velocity constants, due no doubt to the counter reaction.
The simple uni-molecular equation assumes that the reaction proceeds
to completion or so near completion that the speed of the counter may
be ignored. However, as Osterliout has shown in the paper previously
July 15, 1919
Carbohydrate Metabolism in Green Sweet Corn 147
cited, it is not necessary to determine the true velocity constants of a
process under different conditions if only the relative rates are desired.
This may be accomplished in the manner indicated in Table IV by com-
paring the reciprocals of the times required to do the same amount of
work.
In general, it may be stated that up to 30° C. the rate of sugar loss in
green corn is doubled for every increase of 10°. This applies to both
total sugars and sucrose. It should be noted, however, that between
0° and 10° the temperature coefficient for sucrose is considerably greater
than 2.
10
20
30
40
50
60
70
^^
1
S 0 C H 0 S B
\
^^
-^
^^
o°c
A
\ '
\
\\
\^
-^
\
\
^
\10°0
\
*"v»,^^
ili:irr
^
30° C^"""""
.^zCc
24
48
H 0 0 B S
72
96
Fig. 2.— Depletion of sucrose in green Sweet com during consecutive 24-hour periods of storage, expressed
as percentages of the initial sucrose in the com, which was 3.87 per cent, wet weight.
Respiration. — In a former paper the writer ^ has shown that respira-
tion in green sweet corn after it is first pulled from the stalk is compara-
tively high. During the first 24 hours of storage at 30° C. the com
with the husks removed respired at an average rate of 50 mgm. of carbon
dioxid per kgm. per hour. This rate became slower and slower until it
reached, in eight days, a constant rate of about 18 mgm. of carbon
dioxid per kgm. per hour. Respiration of course consumes sugar and
therefore accounts for some of the depletion of sugar in sweet com
1 Appleman, Charles O. respiration and catalasE ACTrviTv in sweet corn. In Amer. Jour.
Bot., V. 5, p. 207-209, 1918.
148
Journal of Agricultural Research
Vol. XVII. No. 4
during storage. During each consecutive 24-hour period of storage the
percentage of sugar in the corn, however, is only slightly altered by res-
piration, as shown by the following illustration. Straughn, in the paper
previously cited, averaged the weight of kernels and cobs from 18 ears
and found that the kernels averaged approximately 50 per cent of the
total weight. If we assume that all of the carbon dioxid came from the
kernels, then 500 gm. of kernels would produce 1,200 mgm. of carbon
dioxid during the first 24 hours' storage at 30°. From the formula
C6Hi20e + 6O2 = 6CO2 + 6H2O
1,200 mgm. of carbon dioxid would correspond to the consumption of
818.61 mgm. of sugar. The consumption of this amount of sugar by
respiration would free in the system 491.343 mgm. of water.
TabIvE IV. — Reciprocals of the times, in hours, required at different temperatures to bring
the sugar depletion in sweet corn to five different stages. Also the temperature coefficients
obtained from these reciprocals
Percentage of initial sugar lost.
Storage
tempera-
ture.
Reciprocals of time
periods.
Total
sugars.
Sucrose.
Temperature
coefficients.
Total
sugars.
Sucrose.
30-
40.
SO-
60.
Average temperature coefficient.
o
10
20
30
10
20
30
10
20
30
10
20
30
10
20
30
10
20
30
o. 0303
.0666
. 1041
.2083
. 0122
. 0320
.0520
. nil
. 0184
• 0354
.0724
.0131
• 0252
• 055s
. OIOI
• 0173
. 0406
o. 02 13
• 0833
. 1282
.2777
. 0108
. 0416
.0666
.1388
.0268
.0427
.0925
. OI9I
.0326
. 0694
.0136
• 0256
•0555
. 0104
. OI7I
. 0416
2. 2
1.56
2. 00
2. 04
3-91
•1-53
2. 16
2. 62
I. 62
2.13
3-85
I. 60
2.08
1. 92
2. 04
1-59
2. 16
I. 92
2.30
1. 71
2. 12
I. 71
2.34
1.88
2. 17
I. 64
2-43
2. 14
For the sake of simplicity, we will confine the system to 100 gm. of
com and suppose that at the beginning of storage it contained 5 gm. or
5 per cent sugar and 80 gm. or 80 per cent water, a fair average for the
July 13, 1919 Carbohydrate Metabolism in Green Sweet Corn 149
com used in this work. According to the foregoing rate of respiration
this system would lose 163.72 mgm. of sugar during the first period of
storage of 24 hours. At the same time 98.269 mgm. of water would be
freed in the system. Our system would now contain 80.0983 gm. of
water and 4.8363 gm. of sugar. By correcting for the slight loss of dry
matter, the system would contain 80.1507 per cent water and 4.8395
per cent sugar. These percentages would be those found by actual
analysis of the 100 gm. of corn after 24 hours' storage, assuming that no
other changes occurred besides respiration.
If we now calculate the percentage of sugar on the basis of the original
water in the system, as was done in all cases in this work, the percentage
of sugar would be 4.8726, showing a loss by respiration of 0.1274 per cent.
It should be noted that the rate of respiration chosen for this illustra-
tion was the rate for the highest period at 30° C. It was also assumed
that all of the carbon dioxid came from the kernels. During the imma-
ture stages of the com it is very probable that some of the sugar in the
cob is consumed by respiration.
During the later periods at the high temperatures and for all periods
at the low tempeVatures, the change in the percentage of sugar by res-
piration during the short periods of 24 hours would be practically within
the experimental error for the sugar and moisture determinations.
One ton of husked green sweet com, during the first 24 hours of stor-
age at 30° C. would lose approximately 3.2 pounds of sugar on account
of respiration.
Under certain conditions, however, respiration may become an impor-
tant factor in accelerating the depletion of sugar from green sweet com.
One of the products of respiration is heat. This heat of respiration will
raise the temperature on the inside of large piles of green corn to a very
marked degree. The increased temperature accelerates not only the
respiratory process itself but also the other processes responsible for
most of the sugar loss. Aeration of green corn is therefore important in
order to dissipate the heat of respiration. In other words, green corn
should not be allowed to remain in large piles for even a short time,
especially during midsummer temperature.
Starch formation. — If the sugar is all converted into starch or other
polysaccharides, hydrolyzed by dilute acids, then the sum of the total
sugars and the polysaccharides as glucose should be the same before and
after storage. During the first period there is a slight deficit after
storage, especially in the more immature ears. A part of this deficit is
due to the high respiration of this period; but some of it is probably
accounted for, in the immature ears when the sugar is high, by the for-
mation of cellulose. During the later periods of storage many of the
ears, depending largely upon the stage of maturity, show a slight increase
in the sum of the total sugars and polysaccharides. This is true espe-
cially at the higher temperatures and is probably accounted for by the
I50
Journal of Agricultural Research
Vol. XVII. No. 4
sugar of the cob being drawn into the grain for starch formation as the
sugar in the grain is depleted. Analyses of cobs from immature ears
gave a total sugar content of about 7 per cent. The sugar in the cob
decreased slightly during storage, but there was no starch formation in
the cob.
After noting these exceptions, which alter the balance only slightly,
it may be stated in general that most of the sugar loss in green sweet
corn is balanced by the gain in polysaccharides, chiefly starch (Table V).
Table V. — Depletion of sugar in green sweet corn balanced chiefly by formation of
polysaccharides hydro ly zed by dilute acid
0° C.
Total sugars plus polysaccharides as glucose.
Ear No.
First period.
Second period.
Third period.
Fourth period.
0 hours.
24 hours.
24 hours.
48 hours.
48 hours.
72 hours.
72 hours.
96 hours.
I
Per cent.
II. 01
10. 72
10.86
Per cent.
II. 04
10. 67
10.86
Per cent.
12. 40
14.84
13.62
Per cent.
12-39
15-35
13-87
Per cent.
II. 07
t
Per cent.
II. 29
Per cent.
13-57
13-41
13-49
Per cent,
13-56
12.68
Average.
II. 07
II. 29
13. 12
10° c.
Average.
14.56
11.79
13.18
14-53
"•37
12.95
15. 26
14. 01
14.64
15. S9
13-50
14. 70
15.86
11-93
13.89
15-35
11.58
14. 29
10.81
13-47 12.55
15-05
10.49
12.77
20° C.
Average.
11.50
10. 82
16
II. 09
9-97
IO-53
15- 19
13.64
12. 13
13-65
15- 13
13-58
II. 20
13-30
11. 2»
10. 21
12. 67
11-35
10.56
10. 24
12. 20
11. 00
10.93
13-34
12. 14
10. 46
13-76
50° C.
Average.
14-75
12. 20
13.48
14. 10
10.53
12.31
13-39
16.77
13- 13
13-55
17. 24
12. 01
14. 43 j 14. 26
13-30
12. 16
10. 80
12.08
13-32
12. 66
12.36
12. 78
12.48
12. 17
ii-43
12.66
12. 61
12. 17
July 15. 1919 Carbohydrate Metabolism in Green Sweet Corn 151
SUMMARY
The data recorded in this paper apply to Stowell's Evergreen corn,
picked in the typical milk or best eatable stage and having a water
content of approximately 80 per cent.
A method was devised by which the rate of sugar loss from green
sweet com could be determined for consecutive 24-hour periods of storage
by comparing analyses of corn from the same ear.
The depletion of sugar in green sweet corn after it is separated from
the stalk does not proceed at a uniform rate but becomes slower and
slower until finally the loss of sugar ceases when the initial total sugar
has decreased about 62 per cent and the sucrose about 70 per cent.
Calculated on the basis of original moisture, the corn contained,
when the depletion of sugar ceased, approximately 1.5 per cent total
sugar as invert sugar, 0.7 per cent sucrose, and 0.8 per cent free-reducing
substances. The actual percentage of sugars would depend upon the
amount of water in the corn after storage. Under the experimental
conditions there was very little change in the percentage of water
in the com employed in this work.
Reversibility of the chief processes involved in the sugar depletion,
resulting in an equilibrium between the rate of sugar loss and the rate
of sugar formation, would account for the cessation of actual sugar loss.
During the early periods of storage, the falling off in the rate of
actual sugar loss is due to mass action. When the equilibrium is nearly
reached the counter reaction, that is the formation of sugar, also tends
to slow up the rate of sugar loss. Any destruction or decrease in the
quantity of enzymes present would produce a falling off in the value
of the velocity constant, with a consequent decrease in the rate of
actual sugar loss. There is no evidence that this occurs up to 30° C.
At 40° there is actual destruction of the enzymes or other altera-
tion on the system. The rate of actual sugar loss must not be con-
fused with the velocity constant.
Raising the temperature simply hastens the attainment of the equilib-
rium positions, which seem to be about the same for all temperatures.
At 30° C, 50 per cent or most of the total sugar loss occurs during the
first 24 hours of storage. At 20°, 25 per cent, and at 10°, or good
refrigerator temperature, only about 15 per cent is depleted during the
same period.
Relative rates at different temperatures, of processes that become
slower and slower until an equilibrium is reached, can be accurately
determined throughout this entire course only by comparing the times
required to bring the process to the same stage at all temperatures.
In order to make this comparison possible the experimental results
were interpolated by a simple ■ graphic method. The temperature
coefficient was then obtained by comparing the reciprocals of the times
152 Journal of Agricultural Research voi. xvii, no. 4
required to do the same amount of work at the different temperatures.
In this manner the temperature coefficients were determined for six
different stages. Up to 30° C. an average coefficient of 2.03 was obtained
for the loss of total sugars and 2.14 for sucrose. From 0° to 10° it
was greater than 2 in the case of sucrose.
In general, it may be stated that the rate of sugar loss, until it reaches
50 per cent of the initial total sugar and 60 per cent of the sucrose, is
doubled for every increase oi 10° up to 30° C.
Respiration in green corn is comparatively high when the corn is
first picked but falls off rapidly with storage. This process, however,
accounts for only a small part of the actual decrease in the percentage
of sugar in the corn during the consecutive 24-hour periods of storage
even at 30° C. One ton of husked green sweet com during the first
24 hours of storage at 30° would lose approximately 3.2 pounds of
sugar on account of respiration.
Respiration may become indirectly a more important factor in accel-
erating the depletion of sugar by raising the temperature on the inside
of large piles of green corn.
Most of the decrease in the percentage of sugar in green sweet corn
during storage is due to condensation of polysaccharides, chiefly starch.
CERTAIN RELATIONSHIPS BETWEEN THE FLOWERS
AND FRUITS OF THE LEMON'
By Howard .SI Reed
Professor of Plant Physiology, University of California
The physiological characteristics of the cultivated lemon make it an
interesting object for study, since its period of blossoming and fruiting
extends through much of the year. On most lemon trees it is possible to
find all stages of development between blossoms and mature fruit through-
out the year, though in varying amounts. There are distinct cycles in
both the vegetative and fruiting activities of this tree whose limits are
recognized by those engaged in its cultivation. One of the objects of the
present study was to obtain quantitative records of these cycles and
especially of the relations between flowers and fruit.
The present study attempts to discuss :
(a) The seasonal distribution of the fruit buds;
(6) The size and productiveness of the inflorescences ;
(c) The time required for the growth of fruit and the relation of this
time to the season at which the buds appear;
(d) The numerical ratio of flower buds to mature fruit.
The material studied consisted of a small group of Lisbon lemon trees
located on the Limoneira ranch near Santa Paula, Calif. The trees stand
in the midst of a large lemon grove and have received good orchard treat-
ment with respect to cultivation, irrigation, and so forth. No especial
attention in these particulars was given to the trees during the time
observations were being made. All were free from injurious insects and
fungous diseases. Each month for one year approximately 50 fruit
twigs bearing fruit buds ready to open were selected and marked with
identification tags. The twigs were chosen on seven adjacent trees, six
of which were full-bearing trees 22 years old. The seventh was 6 years
old, but was very fruitful. As soon as a twig was selected an entry was
made on a special blank on which full records could be subsequently kept
concerning leaves, buds, fruits, and new twigs. Once a month the twigs
were examined and the data recorded on special blanks.
When the first year ended 12 lots of twigs had been selected and marked,
and from that time the records on all twigs were continued for another
year. Thus the first twigs selected were under observation for two years
and the last for one year. A total of 610 twigs was selected and observed,
but there was some loss due to the removal of tags by winds, so that the
final number was somewhat less than 600. The partial records of twigs
» Paper No. 54, University of California, Graduate School of Tropical Agriculture and Citrus Experi-
ment Station, Riverside, CaUf.
Journal of Agricultural Research, (153) Vol. XVII, No. 4
Washington, D. C. july 15. 1919
" „ Key No. Calif.-20
100 1 24 — 19 3
154
Journal of Agricultural Research
Vol. XVII, No. 4
whose tags disappeared were discarded. The writer is indebted to the
management of the Limoneira Co. for their friendly cooperation in this
work, as well as to various members of the staff of the Citrus Experiment
Station for their assistance in the tedious work of obtaining and com-
piling data.
All biological work, especially work done in the field, is accompanied
by inevitable error. The present is no exception. In August, 191 6, much
of the small fruit on these trees was killed during their fumigation with
hydrocyanic-acid gas to kill insects. The following November the trees
blossomed profusely, perhaps due to the earlier loss of a portion of their
crop. In the early months of 191 7 some of the small fruit was killed by
freezing temperatures, in spite of the fact that oil heaters were used in
the grove and all vigilance was exercised to avoid losses. In June, 1917,
following a period of very hot weather, much of the more mature fruit
fell from the trees. It might be thought advisable to discard a portion
of the records which are known to be subject to these errors, but why
should one discard errors due to climatic conditions which he recognizes
while retaining other possibly greater errors which he does not recognize ?
SEASONAL DISTRIBUTION OF FRUIT BUDS
The lemon tree continually produces fruit buds, yet their distribution
through the year is not uniform. Information upon their seasonal dis-
tribution was obtained from the data for 4,545 "new buds"; that is, buds
which appeared on twigs subsequent to the selection of these twigs for
the purpose of study. These data were used to avoid the results of
conscious or unconscious selection by the person who chose and tagged
the original twigs. For example, because large clusters of buds are more
conspicuous, a larger percentage of the fruit buds on the tree may have
been chosen at one time than at another. The effect of this would have
been to give larger records at one season and lower at another. It
should be remarked, however, that since we were dealing with what is
recognized as "fruit wood" the average number of buds on the twigs
selected might be higher than for the average twig of the tree.
The figures given in Table I show the percentage of the new buds
which were produced in the different months of the year and are based
upon the observations of 4,545 buds during a period of two years.
Table I. — The distribution of lemon buds by months
[Average for 1916 and 1917]
Januar}- .
February
March . .
April . . . .
May ....
June. .. .
Buds (per-
centage of
total).
0.31
•37
29.74
36- 13
4-55
3.00
Mouth.
July
August. .. .
September
October. ..
November.
December.
Buds (per-
centage of
total).
4.81
2. 08
I. t8
1.83
13. II
2.56
July IS. 1919 Relation between the Flowers and Fruits of the Lemon 155
A survey of this table and of figure i shows that there are two periods
in the year at which fruit buds were principally produced. In round
numbers, about 66 per cent of the buds appeared in March and April,
about 13 per cent appeared in November, and 20 per cent between April
and November. There was, therefore, a very pronounced seasonal dis-
tribution of fruit buds on the trees observed.
MO
12.0
/
1
1 10
IQO
90
80
/
/
1
1
"iio
60
50
^0
30
20
10
1
i
/
\
/
\
\
y
\
/
\
1
\
^^
y
Jan. Feb. AAar. Aj^r. ^\o\( June July A^S- 5ept. Ott Nov/. Dec.^
Fig. I. — Average monthly production of lemon buds during the year.
The appearance of large numbers of fruit buds in March and April is
undoubtedly related to the greater activity of the tree, following its
slower winter growth. The secondary maximum in November, following
the last growth cycle of the tree for the season, is not so easy of explana-
tion. It might be assumed that the appearance of buds at this time was
a reaction to the large supply of elaborated food material in the tree
156
Journal of Agricultural Research
Vol. XVII. No. 4
and that the tree responded by putting forth fruit buds, while hindered
by cHmatic conditions from producing vegetative growth.
In view of the fact that there are two periods of the year in which
there is a maximum production of buds, it is logical to expect that the
coefficient of correlation between time and number of buds would be
negative, since the maximum production is in the early part of the year.
Table II shows the array for the figures representing the production
of new buds by months.
Table II. — Correlation between new buds and the month in which they appear
Months, be ginning with March
I
i
2 ! ^
4
5
6
7
8
9
lO
II
12
•|
I
4
6
7
8
lO
15
i6
42
95
IIS
1
I
I
2
I
R
I
I
i^
i
I
I
1
I
I
(.4
I
.D
I
1
T
fl
^
I
I
I
I
I
I
I
I
I
I
I I
i
12
r=— o.6so±o.ii2
The coefficient is strongly negative and is in harmony with the obser-
vations, showing that the numbers of new buds decrease after the spring
months, though not in a strictly linear regression.
A certain synchronism was frequently observed in the production of
new fruit buds. If a branch blossomed heavily in March, it would
blossom heavily again in July. A branch which blossomed heavily in
August was likely to blossom heavily in November.
THE SIZE AND PRODUCTIVENESS OF THE INFLORESCENCES
The lemon flowers occur singly or in clusters. During the rapid growth
of spring large inflorescences are common; at other seasons the inflores-
cences are smaller and many of them possess ohly one flower.
Statistical studies were made to ascertain the range of variability and
the productiveness of the inflorescences on these particular lemon trees.
Data on 1,363 inflorescences which appeared during the course of the
observations were examined. The number of flowers per inflorescence
ranged from i to 28. The relative frequency of the inflorescences in rela-
tion to the number of buds per inflorescence is shown in Table III. The
data show that the greatest frequency occurred in the class of inflores-
cences which bore a single bud and that the frequencies decreased quite
July 15. 1919 Relation between the Flowers and Fruits of the Lemon 157
uniformly as the number of buds per inflorescence increased in succeeding
classes. From these data the following constants were calculated:
Mean number of buds per inflorescence = 4.784 ±0.071.
Standard deviation= 3.916 ±0.050.
Coefficient of variability = 81. 86 ± 1.62.
An inspection of the figures shows several interesting relationships.
The number of buds per inflorescence shows no tendency whatever to
follow the normal curve of errors; therefore we may conclude that the
number is not determined by pure chance, but, on the contrary, is fixed
by some other influence. If the number of buds had been determined
by purely casual factors, such as position on the tree, age of wood, or
climatic conditions, we should be warranted in expecting* a purely chance
distribution. In a following paragraph it is shown that there is a cor-
relation indicating that larger inflorescences occur in the spring months,
but the coefficient expressing this correlation is not such that much
emphasis can be laid upon it.
Table III. — Frequency of inflorescences in relation to number of bud'; they produced
Number
of buds
per inflo-
rescence.
Number
of inflo-
rescences
observed.
Number
of buds
per inflo-
rescence.
Number
of inflo-
rescences
observed.
Number
of buds
per inflo-
rescence.
Number
of inflo-
rescences
observed.
Number
of buds
per inflo-
rescence.
Number
of inflo-
rescences
observed.
I
2
3
4
5
6
7
8
239
216
173
168
134
125
62
53 ,
9
10
II
12
13
14
15
16
51
37
18
20
18
7
7
II
17
18
19
20
21
22
23
24
7
4
2
0
0
3
2
0
25
26
27
28
2
I
I
2
^,3(>3
Since the distribution of the buds on the inflorescences departs so
widely from that to be expected upon the basis of pure chance, it seems
logical to assume that it is determined by factors which reside in the
tree and not by external factors. In other words, the Lisbon lemon
tree has an inherited tendency to produce few-flowered inflorescences
which outweighs the effect of external influences.
The writer has found very few recorded studies upon this question,
though it would seem worthy of study both from practical and theoretical
standpoints. The frequency of distribution of the number of seeds in
receptacles of the lotus (Nelumbium luteum) was found to agree very
closely with that of a chance distribution.^ It should be noted, however,
that the two cases differ in the morphology of the organs in question.
In lotus we are dealing with an organ developing from a compound
ovary — that is, with one flower; but in the lemon inflorescence we are
dealing with a short branch bearing flowers. It is possible that the
Pearl, Raymond. Variation in the number of seeds of the lotus. In Amer. Nat., v. 40, no. 479,
P- 757-768, 4 fig., 1906.
158
Journal of Agricultural Research
Vol. XVII, No. 4
number of seeds developed from a compound ovary is dependent upon
a set of external, casual factors, such as amount or variability of pollen,
conditions under which pollination occurs, and access of the mother plant
to suitable supplies of nutriment. On the other hand, the number of
flowers produced on an inflorescence may be more largely predetermined
in the mother plant by such internal factors as those which determine
the position and arrangement of leaves and others which act to produce
generic and specific characters.
The study of the inflorescence may be carried a step further by attempt-
ing to determine whether the larger inflorescences were more character-
istic of one season than of another. If so, it might show whether the
size of the inflorescence is in any way influenced by seasonal conditions.
Data for 403 inflorescences were available and represented a fair random
sample as far as seasonable distribution is concerned. Table IV shows
the correlation between the average number of new inflorescences on the
seven trees and the average number of buds per inflorescence. It seemed
more nearly correct to make this sort of correlation than one between
months of the year and number of buds per inflorescence, since it eliminates
irregular regression due to periodicity, leaving numbers of buds as the
two factors for correlation.
Table IV. — Correlation between monthly average size of inflorescence and numbers of
inflorescences produced
Average size of inflorescence (number of buds)
o 9
O 4<
I.I
1-3
1.6
1.8
1.9
2-3
3-3
4.6
4-7
4.9
5-0
5-4
2
I
9
12
18
ZZ
40
44
59
170
I
I
•
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I I
I
I
I
12
r=o.35i±o.i7i
The average number of new inflorescences per month on the seven
trees ranged from 2 to 170; the average size by months ranged from i.i
buds to 5.4 buds. The coefflcient of correlation between these factors
is 0.351 ±0.171. Since the coefficient is only twice its probable error
we must regard it as rather doubtful in indicating a correlation between
these factors. It may be taken, however, to indicate that larger inflores-
cences were more abundant in seasons in which the number of new buds
July IS, 1919 Relation between the Flowers and Fruits of the Lemon 1 59
was greatest — namely, in the spring months, or at other times at which
the activity of the tree is at its height. This conclusion is in agreement
with the repeated observation that thrifty trees most commonly bear
lemons in clusters.
The next question to be investigated was one of considerable physio-
logical interest: What is the correlation between the number of buds per
inflorescence and the numbers of fruits matured per inflorescence?
A positive correlation approaching i is to be expected in case buds
on all sizes of inflorescences have equal chances of development ; a value
much below this indicates that a bud on a larger inflorescence has a
poorer chance. It is certain from the nature of the case that there must
be some relationship between the two, since an inflorescence having only
I flower could not produce more than i fruit, but an inflorescence pos-
sessing 20 flowers may or may not mature a proportional number of
fruits. If we take the fruits whose history could be definitely ascertained
and arrange them with regard to the size of the inflorescence from which
they developed, we get the arrangement shown in Table V.
Table V. — Correlation between numbers of buds and fruit on inflorescences of various
sizes
Fruits matured per inflorescence
0
I
2
3
4
5
I
2
3
4
S
6
7
8
9
10
II
12
13
14
IS
16
17
18
19
22
23
25
26
27
28
217
186
144
133
107
96
50
44
38
31
12
14
14
5
5
5
6
3
2
3
2
2
I
I
22
29
27
27
22
23
4
7
8
4
5
4
3
I
2
239
216
173
168
134
125
62
53
51
37
18
20
18
7
7
II
7
4
2
3
2
2
I
I
2
I
2
7
4
6
5
I
5
2
I
2
I
I
2
4
I
I
I
I
I
I
I
I
I
I
I, 121
188
46
4
3
I
^^3^3
r=o.i78±o.oi7
i6o Journal of Agricultural Research voi. xvii, no. 4
As determined from these figures, the coefHcient of correlation is o. 178 ±
0.017, indicating a positive correlation between the size of the inflores-
cence and the number of fruits it brings to maturity. The small value
of the coefficient, however, is worthy of note. It was previously men-
tioned that we should expect upon a priori grounds a definite relationship
between flowers and fruits, especially in the case of the smaller inflores-
cences. It will therefore be in order to inquire whether the larger
inflorescences are relatively as fruitful as the smaller.
In order to secure a measure of the relative fruitfulness of the inflor-
escences, I have used a formula which Harris ^ published several years
ago. The correlation coefficient as here used is intended to measure the
correlation between the number of fruits produced per inflorescence and
the deviation of this number from its probable value, in case the number
of fruits per inflorescence is in the same proportion to the number of
buds per inflorescence as the total number of fruits to buds in the entire
population. It is computed from the formula,
^xy * xl ^ V
where x' = buds and >' = fruits per inflorescence, Vx and Vy are the coeffi-
cients of variability of the two characters, and z is to be read as "the
deviation of the number of fruits per inflorescence from its probable
value."
The value of this correlation as computed for the lemon inflorescence
is Yxz = o. 1 83 ± o.oi 8.
This negative value of r^z is interpreted to mean that there is a distinct
negative correlation between the size of the inflorescence and its power
to develop its buds into fruits. In other words, a bud on one of the
smaller inflorescences has a greater chance of becoming a mature fruit
than a bud on one of the larger inflorescences. The competition between
individual buds on larger inflorescences seems to be too severe to allow
all to survive. It is not, therefore, probable that the larger inflorescences
are able to mature proportionally larger numbers of buds.
Heinicke ^ has recently reported that the reverse relationship exists in
the case of the apple trees he studied. His figures indicated that a higher
percentage of flowers develop into fruits on spurs producing six flowers
each than on spurs producing four or five flowers. Further data on this
question are to be desired.
' Harris. J. Arthur. The correlation between a variable and the deviation of a dependent
VARLABLE from ITS PROBABLE v.\LUE. In Biomctrika, V. 6. pt. 4, p. 43S-443. 1909.
Correlation in the inflorescence of Celastrus scandens. In Mo. Bot. Card. 20th Ann.
Rpt. p. 116-122. 1909-
2 Heinicke, Arthur J. Factors influencing the abscission of flowers and partlally developed
fruits of the apple (Pyrus malus L.). N. Y. Cornell Agr. Exp. Sta. Bui. 393. p. 41-114. iHus. 191 7.
Bibliography, p. 112-114.
July IS. 1919 Relation between the F'lowers and Fruits of the Lemon 161
THE TIME REQUIRED FOR FRUIT TO DEVELOP FROM BLOSSOM TO
MATURITY
Lemon fruits grow slowly in comparison with the fruits of most decidu-
ous trees. About two months are usually required from the unfolding
blossom until a lemon fruit reaches a diameter of % inch. The
time required for the fruit to reach a size of 2X inches (a desirable
commercial size) varies according to conditions from 7 to 14 months.
From the standpoint of the producer it is desirable to have the lemons
reach mature size as soon as possible. A lemon which grows uniformly
and rapidly is usually of superior quality. It reaches the size required
for marketing without turning yellow to any appreciable extent; it de-
velops the desired flavor after being artificially cured and withstands
deteriorating influences during transportation and storage.
In the lemon the ovary begins to grow very soon after the perianth
withers and falls off. The style adheres for some time after this, but
eventually the stigma and a part of the style separate and fall away.
"Time of maturity" is regarded as the time at which a lemon is har-
vested. This time is usually determined by one of two things, either
the fruit has reached a diameter of 2^ inches or it has lost all green color
from its surface.
The records of 239 fruits were examined and the time at which they
were set was ascertained. The observations on the time required for
maturity are given in Tables VI and VII. It was found that the time
required to mature individual lemons ranged from 7 to 14 months. The
average time for all lemons in these records was 10.2 months. On
account of the small number of lemons set in the months of November,
December, January, and February, the figures for these months are
omitted from Table VI, because of the inevitably larger error involved
in averaging a few numbers. The reader will understand that fruit set
in a given month was a bud in the preceding month. If this is borne in
mind, there is no confusion in comparing Table VI with other tables in
this paper.
Table VI. — Average time required for growth of lemons to maturity according to the
m,onth in which fruit set
Month in which fruit
was set.
Number
of fruits
observed.
Mean time
required for
maturity. Num-
ber of months.
Month in which fruit
was set.
Number
of fruits,
observed.
Mean time
required for
maturity. Num-
ber of months.
March
8
43
70
20
10. I ±0.47
10. 3 ± . 16
9. 8± .10
9.8± -3°
Tulv
41
23
10
9
9. 3±o. 18
ri. 7± .23
II. 6± .38
II. 7± -39
April
August
May
September
October
June
The table shows that there was a variation of about 2% months in
the average time required to produce a lemon, depending upon the month
in which it set. Lemons which were set in May, June, and July came
l62
Journal of Agricultural Research
Vol. XVII, No. 4
to maturity in minimum time, and those set in August, September,
and October required the maximum time.
Further light on the relation of the time of maturity to the time of
setting was afforded by determining their correlation coefficient. Table
VII shows the data. February is denoted as the first month in the
subject column, since active growth begins in that month ; and January
is the last month.
Table VII. — Correlation between the rnonth in which letnons set and the time required
for maturity
7 S
!
9
10
II
12
13
M
Totals.
¥^
I
3
4
5
6
7
8
9
lO
II
12
I
I
"6"
I
3
I
4
5
8
43
70
20
41
23
10
9
6
I
3
•c
I
II
3
3
18
2
I
13
8
2
2
I
4
6
5
3
4
8
I
6
3
01
E
C
1
5
32
2
6
2
2
19
10
3
6
2
2
3
3
2
7
3
I
I
I
I
I
t
X
g
I
I
id
2
13 36
31
53
49
40
12
5
239
r=o.i38±o,043
The correlation coefficient denotes a positive relationship between the
two factors, though its magnitude is not sufficient to warrant much
emphasis. We can conclude that the season at which fruit is set in-
fluences, but does not absolutely determine, the length of time which
will be required for maturity. Thus, fruit set in May had a range in
time of maturing from 7 to 14 months, although about half the fruit
required 10 months.
THE R.\TIO BETWEEN BUDS AND FRUIT WHICH REACHED MATURITY
The lemon may develop without pollination of the flower, therefore
the proportion of fruit to buds may be expected to indicate the produc-
tiveness of the tree without entire dependence upon the chance of polli-
nation as already intimated; however, the productiveness of the tree is
greatly influenced by various environmental factors, especially by
meteorological factors. Soil environment, as influenced by the applica-
tion of fertilizers or water, affects fruit production; but its influence is
not so abrupt and does not make itself quite so conspicuous as the former
complex of factors.
An examination of these records may be of interest as an indication
of what happens under good commercial culture. These trees stood in
juh- 15. 1919 Relation between the Flowers and Fruits of the Lemon 1 63
a large plantation, receiving regular orchard treatment during the time
obser\^ations were being made, and exposed to the vicissitudes which
beset the commercial orchards. The effects of winter cold and of sum-
mer heat are plainl\' \'isible at places in the records, yet there is no reason
to expect that the average lemon tree may escape the vicissitudes which
befell these. The figures showing the proportion of buds which develop
into mature fruit were based only upon observations made early enough
to allow fruit to mature before the close of the observations.
Several definite stages in the development of fruit have been recog-
nized in making this study. They are as follows: (i) the plump bud
just ready to open; (2) the flower; (3) the first stage of the fruit at which
the corolla has fallen but the style is still attached, designated for con-
venience "style attached"; (4) the young fruit having a diameter of
% inch and having lost the apical portion of the style; (5) the fruit
having a diameter of 2% inches, ready to pick.
The individual histories of a random sample of lemon buds were
followed from stage to stage to see what proportion survived and to
locate, approximately, the time of heavy mortalities. A sample of
4,440 buds which appeared during the observation period was chosen.
Their developmental history is shown in Table VIII.
Tabi,E VIII. — Ratio of lemon buds to fruits -which reached various stages
Stage ol developiQent.
Number
observed.
Buds ready to open. . .
Styles attached
Fruit X inch diameter
4,440
2,308
964
Fruit mature 294
100. 00
51.98
21. 71
6.62
These figures sho^v that there is a large mortality between the young
buds and the mature fruit, and that the mortality seems to increase
with the age of the fruit. It should be stated, however, that losses
were comparatively small after the fruit had reached a diameter of
I inch. There is no reason, however, to regard the losses of fruit on
these trees as abnormally high, since the crops produced on these trees
were above the average for this district.
The effect of seasonal conditions upon the survival of young fruits
seemed worthy of study in determining the ratio between buds and
fruit. Since the time at which the fruit sets is the time at which it
begins to grow, calculations were begun with the stage designated
"style attached." Data were available for 2,453 fruits of this size
which had the chance of developing into mature fruit during the time
observations were made. Records were taken of the numbers of fruits
reaching this stage in each month and of the numbers which matured
from each of the several samples of "style attached" fruits.
164
Journal of Agricultural Research voi. xvii. no. 4
Table IX. — Relation of the survival of fruits to the months of the year in which they were set
Month.
January. . .
February . .
March
April
May
June
July
August
September.
October . . . .
November .
December .
Total
Number
of ' ' style-
attached '
fniits
observed.
262
"3
185
51
215
189
232
168
106
191
240
2,453
Mai ure fruits produced.
303
Number. Per cent.
1-5
1.8
^5 13-5
41. 2
106 21. 2
24 II. 2
57 30- 2
19 8. 2
25 15-0
1.9
16 8.4
It is apparent from these figures that there is considerable variability
in the chances of survival, depending upon the time of year at which
fruit is set. Fruit set in September or October is liable to be killed by
cold weather in January, or if set in May or June is liable to be killed
by hot weather in June. That which was set in the winter months was
repeatedly chilled by low temperatures at night and its vitality was
probably lowered. Fruit set in April and July appears to have the best
chance of survival, though these relations may vary from one year to
another.
It will perhaps be more nearly correct to group the records by seasons,
since conditions are not necessarily confined to months.
Table X. — Relation of survival to season at which fruit was set
Season.
Spring:
March
April
May
Summer:
June
July
August. . . .
Autumn:
September
October . . .
November
Winter:
December .
January. .
February .
July IS. 1919 Relation between the Flowers mid Fruits of the Lemon 1 65
These figures show that, upon the trees observed, a fruit set in one
of the spring months had the best chance of survival and of reaching
maturity. The chance of reaching maturity diminished as the seasons
advanced until the next spring. On the other trees or in other localities
the chances of survival might be quite different.
SUMMARY
(i) A small group of Lisbon lemon trees was studied for two years
to obtain data upon their fruiting habits. The trees stood in a large
commercial orchard and received no special treatment during the time
observations were being made.
(2) Approximately 66 per cent of the fruit buds appeared during
March and April, 13 per cent appeared in November, 17 per cent appeared
between April and November, and about 3 per cent appeared during
the winter months.
(3) The distribution of buds on an inflorescence showed no tendency
to follow the normal curve of errors. Few-flowered inflorescences pre-
dominated numerically over many-flowered inflorescences. A bud on a
small inflorescence had a greater chance of developing into a mature
fruit than one on a large inflorescence. The competition between indi-
vidual buds on larger inflorescences seems to be too severe to allow all
to survive.
(4) The time required for the fruit to reach maturity varied from 7
to 14 months, according to conditions. Fruit which was set in May,
June, and July came soonest to maturity. The season at which fruit
was set appeared to influence, but not wholly to determine, the time
which was required for maturity.
(5) The records for 4,440 buds showed that 51.98 per cent set fruit,
21.71 per cent reached a diameter of X inch, and 6.62 per cent reached
the stage of maturity.
(6) A fruit set in the spring months had the best chance of survival
and of reaching maturity. The chances of reaching maturity diminished
as the season advanced.
ULTRA-MICROSCOPIC EXAMINATION OF DISPERSE COL-
LOIDS PRESENT IN BITUMINOUS ROAD MATERIALS
By E. C. E. Lord
Petrographer, Bureau of Public Roads, United States Department of Agrictilture
INTRODUCTION
On a cursory examination of bituminous solutions by means of the
ultra-microscope, varying amounts of finely divided solid material held
permanently in suspension will invariably be found. In certain hard
native asphalts highly dispersed mineral matter is present in large quan-
tities, and the high adhesive properties of such asphalts have been at-
tributed largely to the selective absorption exerted by these colloids on
certain portions of the bitumen.^ Further investigations have led Rich-
ardson to conclude that some bitumens were absorbed in larger quantities,
and consequently had a greater colloid -carrying capacity than others,
and that this variation was apparently in accordance with their viscosity
and the general character of the particular bitumen.^
As a result of these investigations, it seemed desirable to develop a
reliable method of ultra-microscopic analysis whereby the number of dis-
perse colloidal particles could be determined accurately in any type of
bitumen, thereby furnishing a ready means for comparing their colloidal
capacities and at the same time establishing a possible method for esti-
mating the relative value of this property from a road-making standpoint.
The fact, however, should be emphasized that the present investigations
were undertaken essentially to develop a method for counting colloidal
particles in bituminous solutions, and that in drawing comparisons of the
relative supporting values from the results obtained, the original con-
sistency of the materials employed should receive due consideration.
METHODS OF ULTRA-MICROSCOPIC EXAMINATION
When examined under the ordinary microscope, the great bulk of the
colloidal material common to bituminous solutions is invisible. Early
investigations by Siedentopf and Zsigmondy ^ have shown that the re-
solving power of the microscope is very greatly increased when particles
are viewed in a powerful light against a dark background. This illumi-
nation was obtained originally by allowing a beam of light to enter the
cell through a narrow slit at right angles to the axis of the instrument,
' Richardson, Clifford, the theory of the perfect sheet asphalt surface. In Jour. Indus, and
Engin. Chem., v. 7, no. 6, p. 463-465. 1915.
* . importance of the REL.'^TIOX of solid surfaces and liquid films in SO.ME TYPES OP EN-
Glneering CONSTRUCTION. In Sci. Amer. Sup., v. 83, no. 2152, p. 198-199. 1917. Printed also in sepa-
rate {onn.
' Zsigmondy, Richard, erkenntnis der kolloide. 1S6 p. Jena, 1905.
Journal of Agricultural Research, Vol. XVII, No. 4,
Washington, D. C. July 15, 1919
sa Key No. D-16
167
1 68 Journal of Agricultural Research voi. x\ai, no. 4
where part of the rays were deflected from the surface of the suspended
particles into the microscope, thus rendering them self-luminous and
clearly visible while the remainder of the field remained dark. A similar
effect may be obtained by means of a substage parabaloid condenser,
with central stop, whereby the outer rays from the beam of light entering
the microscope from below are brought by a series of reflections to a
short focus within the cell and are totally reflected from the lower surface
of the cell cover, leaving the field dark as before. Particles whose indices
of refraction vary from those of the inclosing liquid intercepting these
oblique rays diffract a portion of the light into the microscope and be-
come luminously visible as in the former case while the remainder of the
field is perfectly dark.^
In order to avoid loss of light through refraction of the rays issuing
from the condenser, an immersion liquid, such as cedar oil or glycerin,
should be employed between it and the cell containing the liquid under
examination. This parabaloid illuminator is interchangeable with the
substage Abbe condensor of the ordinary microscope and was conse-
quently found most convenient for the present investigations. The micro-
scope selected was provided with an accurately calibrated micrometer
screw for vertical measurements and a mechanical stage for lateral
orientation. A diamond point object marker with circular movement
graduated to milHmeters and insertible in the revolving nosepiece of the
microscope will also prove a useful accessory. Light was furnished by
a special arc lamp run on either direct or alternating current and regu-
lated by a rheostat of 4.5 ampere capacity. Before entering the micro-
scope the light was passed through a cooling solution, acting as a ray
filter, of 10 mgm. diamine green dissolved in i liter of distilled water.
A photograph of the microscope with arc light and ray filter used is
shown in Plate 19, A.
The ordinary lens system of the microscope consisted of eyepieces
X 7.5 and X 12.5 and objective 3, 4, and 16 mm., giving linear magnifica-
tions of 50 to 740 diameters at a tube length of 160 mm., while the
best combination for counting was obtained with eyepiece X7.5 and
objective 4 mm., magnifying 320 diameters.
A counting device was inserted in the focal plane of this eyepiece,
consisting of a cross-line micrometer scale with ground glass border
divided into 25 square areas each side of which measured 1.25 mm.
and corresponded exactly to 0.05 mm. of a stage micrometer at a tube
length of 166 mm. With this micrometer the areal dimensions of any
liquid under examination could be accurately determined, while the
vertical element was obtained by means of the micrometer screw record-
ing an interval of 0.00254 mm.^
1 Burton, E. F. the physical properties of colloidal solutions, p. 46-47. London and New
York. 1 9 16.
* This micrometer was calibrated against that of a standard Fuess microscope registering a minimum
vertical interval of o.cxji mm.
July IS. 1919 Disperse Colloids in Bituminous Road Materials 1 69
PREPARATION OF THE ULTRA-MICROSCOPE CELL
In order to carry out a quantitative analysis of bituminous solutions
it was found necessary to employ a cell of minimum capacity that might
be readily cleaned and hermetically sealed to prevent the escape of the
volatile solvent. Efforts were made to utilize a container constructed
on the principle of the Zeiss haemocytometer, as employed by Burton
and Perrin in their examination of colloidal water solutions (hydrosols)/
but it was found that all types of cement used in constructing this cell
were attacked by the benzol solution and, furthermore, that the rulings
on the bottom of the cell when filled with the solution were almost
invisible under the microscope.^
In order to overcome the above-mentioned difficulties, it was found
necessary to excavate a suitable cavity in the object glass itself, thereby
doing away entirely with the superimposed glass plates of the haemo-
cytometer slide. The object glass selected was as free as possible from
air bubbles and other inclusions and had perfectly smooth plane sur-
faces and a thickness varying from 1.25 mm. to 1.75 mm. to assure
a proper focus within the cell of rays from the dark held illuminator,
he excavation was carried out by means of a stationary upright drill
provided with a pointed vulcanized liber cylinder having a flat grinding
surface about 2 mm. in diameter. The drill was run by an electric
motor at i,Soo revolutions per minute, using coarse emery mixed with
a little heavy lubricating oil as an abrasive. In operating the drill
great care was taken to apply a moderate uniform pressure, and the
glass plate was protected from sudden strain by a folded towel or
felt cushion placed beneath it. After grinding for one or two minutes
the drill was removed and the cavity examined. In general, the central
portion was found to be essentially flat and surrounded by deeper cir-
cular grooves, produced by the larger fragments of emery becoming
lodged in the drill during the process of grinding.
From this stage in the operation the grinding was carried on by
means of an electrically driven, flexible shaft drill constructed on the
principle of the dental drill and using volcanic ash or ground pumice
with water as an abrasive. This drill, operating at a speed of 1,540
revolutions per minute, was provided with a grinding point of vulcanized
rubber or belata gum which also proved very effective in polishing the
cell. The polishing was begun with diatomaceous earth and water and
1 Burton, E. F. op. cit., p. 118-120.
Perrin, jean, mouvement brownien et realite molecul-mre. In Ann. Chim. et Phys., s. 8, t.
18, p. 40-42. 1909.
2 The indices of refraction for ordinary light flint glass and benzol at 21.5° C. are 1.5710 (D) and 1.5304
(H), respectively. Smithsonian physicai, tables, ed. 6, p. 184, 192. Washington, D.C, 1914. (Smithsn.
Misc. Collect., v. 63, no. 6.) Hence light passing through glass and meeting etched lines on a cell bottom
mounted in benzol are but slightly diffracted and consequently appear indistinct under ultra-microscopic
illumination while plainly visible when viewed in water or air.
lyo
Journal of Agricultural Research
Vol. XVII, No. 4
was continued with a mixture of freshly precipitated calcium carbonate
and magnesia hydrate until a microscopically smooth and essentially
flat surface was obtained. Finally, in the center of the cell a circle i
mm. in diameter was inscribed with the diamond-point marker of the
microscope in order to limit the field of observation. A diagram of the
slide, with2>2 mm. cell (A) containing circular area (B) drawn to natural
scale, is shown in figure i , where the depth is indicated as lying between
0.044 3-11^ 0-143 mm. These values were determined as accurately as
r
-4^
^efiMofcell O oaa fo O /43 mm. t
/J /S saiJCer~5hape<f ce/f 2imm /o cf/ometer
O is Gtrcu/or fie Id I mm in diameter in center of eel/ floor.
Fig. I. — Glass slide with ultra-microscope cell drawn to natural scale.
possible for each cell by means of a strain dial recording intervals of
o.oooi inch and were checked with the microscope micrometer under a
magnification of 740 diameters. A blunted, highly polished needle point
inserted in the vertical arm of the dial enabled readings to be taken at
different points within the cell, and the average of these readings was
compared with that of an equal number taken around the cell from
without. The results of these measurements for a number of cells con-
structed in the manner outlined above, together with microscopic check
determinations, inclosed in parentheses, are given in the following table :
Table I. — Depth of ultra-microscope cells determined by strain dial
Slide number.
Depth in
mm.
Thickness
of slide in
mm.
Maximum
variation
in depth
in mm.
in TTITTI ^
Percentage
of variation
in thickness
of slide.
I
2
3
4
5
6
Average
a(
1209
1220)
1321
0434
0423)
1432
1450)
0795
0790)
0892
0890)
1014
I- 5316
I- 5382
I- 7374
I. 6103
I. 4880
I. 6812
I- 5978
o. 0058
0160
0036
0094
0071
0043
• 0077
o. 0061
. OIOI
. 0096
. 0046
■0053
. 0061
. 0070
12. 10
8.30
6. 50
9. 20
4.82
7-53
o. 40
66
55
28
36
36
43
a Determined by microscope micrometer.
July 15, 1919 Disperse Colloids in Bituminous Road Materials 171
In comparing these values it will be noted that the average depth
of all cells is but slightly in excess of o.io mm., while the maximum
variation in depth and in the thickness of slide is approximately the
same (0.007 rnii^-)' indicating a cell floor closely approaching a true
plane.
Dial measurements also were undertaken to determine the depth of
cell, including cover glass after mounting in the asphaltic oil solution
employed in counting (see below) and in air to form an estimate of the
relative thickness of the liquid film between cover glass and slide beyond
the cell area. In every case lesser values were obtained for cells mounted
in this solution than in air, indicating a more perfect contact through
the release of atmospheric pressure and the adhesive character of the
bitumen.
PREPARATION OF SOLUTIONS
Before describing the method of counting colloidal particles employed
in this investigation, it will be found desirable to outline briefly the
general character of bituminous solutions containing colloidal matter
and the manner in which these solutions have been prepared for micro-
scopic analysis.
When viewed under the ultra-microscope the colloidal portion of the
solution will appear as a mass of very finely divided and more or less
widely dispersed particles undergoing a constant and, under certain
conditions, perpetual movement (Brownian movement). This move-
ment has been ascribed to the molecular energy of the suspending liquid
and may be regarded as a function of the size of the particles and their
degree of dispersion which, in turn, is limited by the viscosity of the
solution.^ In order, therefore, to count successfully these suspended
particles it was found necessary either to retard their movement by
suitable concentration of solution or to increase it by dilution to such a
degree that they settled out within the cell inclosure.^ In the latter
case, however, it frequently happened that the particles were in part
resorbed on exposure to light, thus destroying the accuracy of the count.
To assure concordant results, therefore, the particles were always counted
in a somewhat viscous solution of colloid-free asphaltic oil to which a
definite amount of paraffin had been added. This solution was pre-
pared by fluxing 2.5 gm. Mexican oil asphalt (penetration 148) with
0.5 gm. crystalline paraffin and diluting to 100 cc. with benzol contain-
ing 10 per cent alcohol. This was then evaporated to constant weight
on the water bath, brought to original consistency with benzol and
passed through an alundum tube or filter tube clogged with macerated
filter paper until approximately all suspended matter had been removed.
• OsTWALD, Wolfgang, die welt der vernachlassigten dimensionen. p. 34-35. Dresden and
Leipzig, 1915.
2 The solvent used in these investigations was c. p. benzol, since carbon bisulphid was found to contain
an appreciable quantity of colloidal sulphur.
172 Journal of Agricultural Research voi. xvii. No. 4
When properly prepared, the diluting solution should contain not
more than 10 particles to a X i"™- square field in cell No. 3, at a mag-
nification of 320 diameters. A definite portion of the solution to be
examined was introduced into this standard dilutant after the coarser
mineral matter had been removed. That was accomplished by dis-
solving I gm. of the original sample in 50 cc. benzol in a stoppered cen-
trifuge tube, allowing the solution to stand overnight (17 hours) and
centrifuging for i hour at a speed of 800 revolutions per minute. A
small portion of this solution was then drawn off from the tube at a
depth of 10 mm., and i cmm. transferred to a glass-stoppered graduate
and brought up to 10 cc. with the paraffin oil dilutant. By this means
a dilution of i to 5,000 of the colloids present in the original sample was
obtained. This was found to be sufficient in most cases, but in certain
bitumens where the dispersed mineral matter was in a state of extreme
subdivision a further dilution of i to 50,000 was necessary before the
colloid particles could be conveniently counted.
METHOD OF COUNTING COLLOIDAL PARTICLES
The samples of bitumens selected for examination were obtained from
the commoner types of road material, ranging from hard native asphalt
to lighter oils and containing varying amounts of colloidal matter.
After having been subjected to the preliminary treatment mentioned
above, one or two drops of the properly diluted solution were rapidly
transferred from the lo-cc. graduate to the cell by means of a i-cc.
pipette and covered immediately by a i8-mm. cover glass, using a
bluntly pointed wooden rod to expel all excess liquid and assure a
perfect contact between slide and cover glass. After the excess solution
had hardened sufficiently by evaporation and the slide beyond the cell
limit appeared perfectly clear and colorless, the cover glass was sealed
with a 30 per cent solution of boiled Canada balsam in ether applied
with a hair-line paintbrush (No. o). A photomicrograph of a part of
the mounted cell with cross-line micrometer scale magnified 320 diame-
ters is shown in Plate 19, B.
In order to obtain consistent results, the cell and cover glass should
be microscopically clean before mounting and the dilutant examined
from time to time to allow for corrections in the final results.^ When
properly mounted the cell should be free from air bubbles and the
colloidal particles should appear under the microscope evenly dis-
tributed and in constant, though restricted, motion. In correct focus
these particles were clearly defined as brilliant points of light against
a dark background, but a change of focus resulted in the development
' The cleaning was accomplished by first boiling slide and cover glass in concentrated sulphuric acid,
then rinsing in water, alcohol, and benzol, drying with soft cotton or silk cloth, and rubbing with optical
tissue paper until thoroughly clean.
July IS, 1919 Disperse Colloids in Bituminous Road Materials 1 73
of concentric halos or diffraction rims around each particle that de-
tracted greatly from the definition of the images. In order to overcome
this so far as possible, it was found necessary to employ cells below
o.io mm. in depth, having a capacity less than o.io cmm. (Table I,
No. 3, 5, and 6). Counts were made of all particles in suspension as
well as those that might have settled out on the cell floor or become
attached to the cover glass during the process of counting. The area
examined was taken from within the central millimeter circle of the
cell and represented exactly one-fourth of i square millimeter (equiva-
lent to four fields of the cross-line eyepiece micrometer at a magnifi-
cation of 320 diameters), while the volume of liqtiid was obtained from
this area and the depth of cell employed (0.043-0.089 mm.). The
number of particles counted in each of the 25 square subdi\asions of
the micrometer through the entire depth of liquid was recorded, and
from the average of four such determinations the value for i cmm. of
solution and i gm. of bitumen was computed.
The results of the analyses were recorded on a special form which,
in addition to the data indicated above, contained information regarding
the physical properties of the bitumen, together with the relative size
and distribution of the colloidal particles. In general, it may be stated
that these particles varied in size from submicrons having a minimum
diameter of about 15/^^1 (0.000015 rnm) to particles within the visibility
of the ordinary microscope (above 0.25 ^ = 0.00025 mm.).^ These
dimensions may be determined by direct microscopic measurement or
they may be calculated by dividing the total volume of particles con-
tained in a definite quantity of solution by the number of particles
found where the volume represents the weight of the particles divided
by their specific gravity. The quotient thus obtained will equal the
volume of one particle (x). Assuming the particles to be spheres of
diameter a,
Then - t: a^ = x
3
3/ X
RESULTS OF THE ULTRA-MICROSCOPIC EXAMINATION OF
BITUMINOUS SOLUTIONS
In order to standardize the method of ultra-microscopic analysis
outlined above, a number of determinations were made of various
colloidal materials contained in different types of bitumens. The
results of these determinations are shown in Table II.
' ZsiGMONDY, Richard, op. cit., p. S8b-97.
174
Journal of Agricultural Research voi. xvii, no. 4
■<&,
•t?.
•J
0^*0-0000 0 0
5oOOf<OOQ«
<*! C^ fOO r- 1
O O '-' O O '
■^ ^ -^ ■* ^
r-O O O O •"*
+T *»-t *T-1 »m nH
■>f ■* Tt -^
o!2;
• ■* O 0> Ov •
^ ^ , , ,_„„' 5oo do CO
00000000000000
000»*^'TOC>0
d «; o
juiyis, I9I9 Disperse Colloids in Bituminous Road Materials 175
o
•<i.
"^
:i
S°-c§
« 0! •* C3
ag
.i' ^ V-i ^ -u
Pu 2
r^
6^
Bii 6
00000
00000
0000000000
OsOOCO 00f00\0-00 'OOO^OWOOO^O
OOQOCOO OOOnO^OcO f^c^MWC*c^M•HM
ooocxoo cooooo
I O O ^ O O 'O •■
<-. °
*J o
o-aa o 3 o o otsm
^ :S : : :fH
i
6
-4-1
u
c
"ci
ja
0
0
S
< 00 00 00 o o
o O o o o o o
•5 "O '5 "O "O "O "O
06 ■^00 Ti-oo -^
000000
o ■^o ■^<7»0'*0-<t
00 ^00 "l-OOCO -TOO ■^
000000000
00000000
fO fO fO fO
-)0 r^oo O O
■^ -^ ^ •* -^ rr Tj ■
1 -6 Journal of Agricultural Research \o\. xvii, .\o. 4
It will be noted that duplicate check counts were made generally in
cells of varying capacity, giving the maximum numerical and percentage
variation and indicating as well the supporting value or colloidal capac-
ity of each type of bitumen, based on that of refined Trinidad 'asphalt
considered as 100. All results were computed on a basis of i to 5,000
dilution of the original colloidal portion of the sample. (See p. 175.)
These bitmuens have been separated into groups containing clay as col-
loidal material (No. 1-19) and into others in which this mineral matter
was replaced by carbonates, sulphates and acetates of copper, iron,
zinc, and lead (No. 20-47). The samples included in the first class,
except untreated refined Trinidad asphalt (No. 1-5), were prepared by
incorporating 33 per cent sandy clay in each type of bitumen by Richard-
son's method of heating an aqueous emulsion of clay and bitumen until
all moisture and gas had been expelled.^ In the case of refined Trinidad
asphalt and clay (No. 6-10) all insoluble mineral and organic matter
originally present in the bitumen was removed before emulsifying by
dissolving in benzol and adding about 2 per cent shellac dissolved in
alcohol and evaporating to constant weight, redissolving in benzol and
filtering through an alundum tube until the solution was essentially
void of colloidal particles.
In the second group of bitumens (No. 20-47) the salts were intro-
duced in an anhydrous condition and the mixtures were heated to about
170° C. under constant stirring until all evolution of gas had ceased.
On examining solutions of this kind under the ultra-microscope it was
found that the copper carbonate salts had been largely reduced to red
cuprous oxid, accompanied by an enormous colloidal dispersion (No.
20-30), while with the remaining salts the reduction had been much
less complete (No. 31-39) or entirely lacking (No. 40-47), and the de-
velopment of colloids correspondingly less. It may be stated, there-
fore, that the colloidal capacity of the second group of materials taken
as a whole was dependent largely upon the degree of chemical reaction
between the bitumen and the salts employed, while in the first group
this supporting value was related more directly to the physical char-
acter of the bitumen.
A comparison of the duplicate counts recorded in columns 3 and 4 of
the table indicated that a maximum variation of less than 10 per cent
was attained in samples of the first group (No. 1-19), while in the second
group (No. 20-47) the results were, on the whole, less concordant, owing
largely to the greater dispersion of colloidal matter.
In conclusion, it may be stated that the accuracy of this method for
counting colloidal particles in bituminous solutions depends chiefly
upon accuracy in construction and calibration of the cell employed, as
well as upon the proper consistency and optical purity of the support-
ing liquid.
' Richardson, Clifford. 1917. op. cit.
PLATE 19
A. — Microscope with ray filter and arc lamp for dark field illumination.
B. — Photomicrograph of cross-line micrometer scale, showing colloids in dark field.
X320. Taken by E. A. Shuster, jr., Photographic Laboratory, United States Geo-
logical Survey.
Disperse Colloids in Bituminous Road Materials
Plate 19
Journal of Agricultural Research
Vol. XVII, No.4
Vol. XVII AUOUSX 15, 1919 No. 5
JOURNAL OF
AGRICUIvTURAL
RESEARCH
CONXKNXS
Page
Derds as an Insecticide ------- 177
N. E. McINDOO, A. F. SIEVERS, and W. S. ABBOTT
(Contribution from Bureau of Entomology)
Effects of Heat on Trichinae ------ 201
B. H. RANSOM and BENJAMIN SCHWARTZ
(Contribution from Bureau of Animal Industry)
Effect of Removing the Pulp from Camphor Seed on Germi-
nation and the Subsequent Growth of the Seedlings - 223
G. A. RUSSELL
(Contribution from Bureau of Plant Industry)
Bacterium abortus Infection of Bulls - - - - 239
J. M. BUCK, G. T. CREECH, and H. H. LADSON
(Contribution from Bureau of Animal Industry)
PUBUSHED BY AUTHORITY OF THE SECRETARY OF AGRICULTURE,
WITH THE COOPERATION OF THE ASSOCIATION OF AMERICAN
AGRICULTURAL COLLEGES AND EXPERIMENT STATIONS
WASHINOTON, D. C.
WASHINOTON : QOVERNMENT PRINTINQ OFFICE : III*
EDITORIAL COMMITTEE OF THE
UNITED STATES DEPARTMENT OF AGRICULTURE AND
THE ASSOCIATION OF AMERICAN AGRICULTURAL
COLLEGES AND EXPERIMENT STATIONS
FOR THE DEPARTMENT
KARL F. KELLERMAN, Chairman
Physiologist and Associate Chief, Bureau
of Plant Industry
EDWIN W. ALLEN
Chief, Office of Experiment Stations
CHARLES L. MARLATT
EntomoloQist and Assistant Chief, Bureau
of Entomology
FOR THE ASSOCIATION
H. P. ARMSBY
Director, Institute of Animal Nutrition, The
Pennsylvania State College
J. G. LIPMAN
Director, New Jersey A gricultural Experiment
Station, Rutgers College
W. A. RILEY
Entomologist and Chief, Division of Ento-
mology and Economic Zoology, Agricul-
tural Experiment Station of the University
of Minnesota
All correspondence regarding articles from the Department of Agriculture should be
addressed to Karl F. Kellerman, Journal of Agricultural Research, Washington, D. C.
All correspondence regarding articles from State Experiment Stations should be
addressed to H. P. Armsby, Institute of Animal Nutrition, State College, Pa.
JOIMAL OF AGEDLTiAL RESEARCH
DEPARTMENT OF AGRICULTURE
Vol. XVII Washington, D. C, August 15, 191 9 No. 5
DERRIS AS AN INSECTICIDE
By N. H. McIndoo, Insect Physiologist, Deciduous Fruit Insect Investigations, Bureau
of Entomology, A. F. SiEVERS, Chemical Biologist, Drug-Plant and Poisonous-
Plant Investigations, Bureau of Plant Industry, and W. S. Abbott,' Scientific
Assistant, Bureau of Entomology, United States Department of Agriculture
INTRODUCTION
The investigation of the possibilities of Derris as an insecticide is a con-
tinuation of the cooperative work inaugurated by the Bureaus of Ento-
mology and Plant Industry, and the most important results pertaining to
the study of Derris are discussed in this paper. There are now on the
market several standard insecticides : Arsenicals, used as stomach poisons •
nicotine solution, used as a contact insecticide; pyrethrura powder, em-
ployed as a dusting powder; and soaps, lime sulphur, oil sprays, etc. Not
one of these acts both as a stomach poison and a contact insecticide. The
following pages will show how well Derris acts in both of these ways.
In a search through the vegetable kingdom for plants possessing toxic
principles with a view toward utilizing them as insecticide material,
attention was directed to the large class of plants which are used exten-
sively in the tropics as fish poisons. There are many hundreds of these
plants, included in several families, and their habitat extends over prac-
tically the entire Tropics. That many of them belonging to particular
families and genera display a remarkably toxicity to fish has long been
known, and probably for ages the natives of the Tropics have used some
of these plants as a means of catching fish.
While a plant toxic to fish need not necessarily be poisonous to in-
sects, nevertheless, some of the fish poisons have already been recom-
mended and used in the Orient as insecticides. If the fish poisons prove
to be efficient insecticides, their practical utilization is at once suggested,
because many of them are known to be very abundant in the Tropics.
The present investigation deals with six or seven species, all belonging
to the same genus, which is widely known as Derris.^ Of these species
only Derris elliptica Benth. seems to have been used widely as a means
for catching fish; it is regarded as a powerful fish poison.
* A portion of the experimental part of this investigation was performed at the Insecticide Board's testing
laboratory, located at Vienna, Va., by W. S. Abbott and E. W. Scott, Entomologist, Enforcement Insec-
ticide Act, under the direction of the latter.
2 Although this genus has commonly been known as Derris, the rules of botanical priority require the
use of the name Dcguelia of Aublet. Of the six species mentioned, the following have received names
under DegueUa: Deguelia elliptica (Wall.) Taub. [Derris elliplica (Wall.) Benth ]; Deguelia robusta (Roxb.)
Taub. [Derris robusta (Roxb.) Benth.]; £)e(7Me//a iimorensis (DC.) Taub. [Derris scandens (Roxb.) Benth.]
Degtielia uliginosa (DC.) Baill.; [Oerri. tiligmosa (DC.) Benth.]. S. F. BlakE.
Journal of Agricultural Research. Vol. XVII, No. s
Washington, D. C. Aug. 15, 1919
sb 177 KeyNo.K-76
178 Journal of Agricultural Research voi. xvii. No. s
The material available for the present study was secured in most cases
from various agricultural and botanical agencies through the Office of
Foreign Seed and Plant Introduction, United States Department of Agri-
culture. The following is a list of the material used and the sources from
which and through which it was secured: Powdered roots of a Derris
species, most likely Derris ellipiica Benth., from the open market where
it is sold as an insecticide; roots of D. ellipiica, called "tuba" or
"toeba" in the Dutch East Indies, from the 's Lands Plantentuin,
Buitenzorg, Java; stems of D. uliginosa Benth., from Mr. C. H.
Knowles, Suava, Fiji Islands; stems of D. koolgibbcrah ^ Baill., and of
D. oligosperma,^ from the director of the Botanical Gardens at Brisbane,
Queensland, Australia; roots of D. scandens Benth.; and stems and
roots of D. robusta Benth., from the director of the Botanical Survey
of Sibpur, Calcutta, India.
HISTORICAL REVIEW
The genus Derris, belonging to the family Papilionaceae, tribe Dal-
bergieae and subtribe Lonchoecarpinae, is practically native throughout
the Tropics, but is far more abundant in the Old World than in tropical
America. Its members are climbing shrubs, having trunks 3 or 4 feet
in height and about 4 inches in diameter; the trunks send out numer-
ous long branches, which climb over the neighboring vegetation, and
the tips of which hang freely downward.
Watt (io,p.8oy describes Derris as a genus of arborescent climbers or
trees, and states that the roots of Derris elliptica furnish a useful insec-
ticide for gardening purposes. A number of other species are mentioned
in the literature as being used for fish poisons, and in some cases reference
is made also to their use as insecticides; but these cited cases seem
to be no better than mere reports. Correspondents in the Philippine
Islands and Java report that D. elliptica is probably the species most
commonly used as a fish poison. In all cases, so far as known, only
the roots are employed. It seems that the most widespread treatment
is one in which the roots are buried in mud, brackish mud preferred,
for a period of several weeks; then the roots are crushed and placed in
water inhabited by fish. The roots of Derris, in all probability D. ellip-
tica, are used as insecticides in the Dutch East Indies; and a correspond-
ent reports that Derris is commonly used by the Chinese gardeners in the
Malay Peninsula as an insecticide and that the parts of the plant used
are sold by Chinese storekeepers. However, it is said that the poison
loses its activity when the plant is dried.
Hooker (5, p. 43) reports on a specimen of Derris elliptica, obtained
from Singapore where it is known as "tubah" and where it is used as
an insecticide ; the roots are steeped in water and the resulting decoction
is said to be an efficient insecticide for garden purposes.
' No record of the publication of these specific names could be found.
^ Reference is made by number (italic) to "Literature cited," p. 200.
Aug. 15, 1919 Derris as an Insecticide 1 79
Probably the first investigator to report on a chemical examination
of Derris elliptica was Greshoff (j) in 1890. He found the most im-
portant constituent of the bark on the root to be a nitrogen-free, non-
glucocidal resin which he called "derrid." He describes this resin,
which he did not succeed in obtaining in crystalline form, as readily
soluble in alcohol, ether, chloroform, and am)^ alcohol, but soluble
with difificult}^ in water and potassium hydroxid. The yield obtained
from the whole root was 2.5 to 3 per cent. The resin was found to be
extremely toxic to fish.
Dymock, Warden, and Hooper {2, p. 421) record that in India Derris
uliginosa is used as an insecticide against larvae of insects.
In 1892 Wray (jj) worked on Derris elliptica and appears to have
been unaware of Greshoff's paper, because for the resinous principle
which he isolated from the root in an impure state and which he used
in his experiments on fish he proposed the name "tubain." This sub-
stance is without question the same as Greshoft*'s "derrid," judging
from its physical properties. The crushed roots when boiled in a retort
with water yielded an opalescent distillate, the odors of which strongly
resembled those from the roots. This distillate was found poisonous
to fish.
In 1899 van Sillevoldt (9), working on Derris elliptica, reported on
the extraction of Greshoff's "derrid." He used practically the same
method of extraction as did Greshoff and describes the "derrid" obtained
as a yellow, amorphous powder. In the impure "derrid" he found a
crystallizable substance which was very insoluble in ether, by which
means it could be separated from the soluble portion of the "derrid."
He found the melting point of "derrid" to be near 73° C. and he de-
scribes it as being readily soluble in alcohol, ether, benzol, aceton, glacial
acetic acid, acetic ether, carbon disulphid, and chloroform, and very
insoluble in petroleum, ether, and water. Van Sillevoldt assigned the
formula C33H30O10 to "derrid."
In 1902 Power {8) investigated the stems of Derris uliginosa. His results
led to the conclusion that the poisonous constituent of the plant is a
resin, thus concurring in the views of Greshoff and van Sillevoldt. He
noted further that this resin consists of two components, one being
soluble in chloroform and highly toxic to fish, and the other insoluble
in chloroform and inactive to fish.
In 191 1 van Hasselt {4) investigated the physiological action of
"derrid" on fish, frogs, mice, rabbits, and cats, and studied its effects
on the blood, respiration, circulation, intestinal tract, and nervous
system. From his experiments he concluded that "derrid" is a powerful
poison, causing characteristic symptoms in all the animals treated, and
that it kills by causing respiratory paralysis.
In 1 91 6 Campbell (/) investigated the poisonous actions of Derris
elliptica, and his work seems to be the most recent along this line. He
tested the water and saline extracts of the roots on fi.sh, mosquito larvae,
i8o Journal of Agricultural Research voi. xvii, no. s
tadpoles, toads, and monkeys. The following are the salient conclu-
sions of his investigation: (i) Boiling does not destroy the toxic action
of the sap ; (2) roots kept three months in a cupboard retain their strength ;
(3) milky extract introduced into a fish's stomach is rapidly fatal;
(4) tadpoles are fatally affected, but stronger extracts are required to
kill them than to kill fish ; (5) much stronger doses are required to kill
mosquito larvae than to kill either fish or tadpoles; (6) the extract from
1/50 gm.of the roots when injected subcutaneously is fatal to toads, and
the extract from i /i 2 gm. causes death when introduced into the stomach ;
and (7) when the extract from 2 gm. of the roots is injected subcuta-
neously or introduced into a monkey's stomach death results. Campbell
further states (p. 134-135) '■
From the results on different animals it is evident that the poison affects the more
highly developed members of the animal kingdom more readily than it does the
primitive members. This is only to be expected since its action concerns the
brain and one particular part of this, namely, the medulla oblongata.
It could be used to destroy mosquito larva:, but it should be used in solutions not
weaker than i in 1,000, that is just enough of the extract should be added to the pool
to make the water cloudy.
METHODS OF PREPARING AND TESTING EXTRACTS FROM DERRIS
MATERIALS
The many preliminary experiments performed indicate that Derris
(probably D. elliptica) is promising as a contact insecticide and as a
stomach poison but is of no practical use as a fumigant. The best
methods of applying it — whether in the form of powder, suspended in
water or in the form of extract mixed with water or with soap solu-
tion— now remain to be determined.
A vegetable insecticide is usually applied either in the form of fine
powder or as a spray mixture. This mixture may consist of any one
of the following four combinations: (i) Powder suspended in water; (2)
aqueous extract of the material diluted with Avater; (3) a solution con-
sisting of water and a small amount of a concentrated form of the active
constituent; and (4) a small quantity of a concentrated form of the
active principle suspended in water.
Since Derris material must be imported, only dried roots and stems
may be secured for insecticidal purposes. As already stated under the
historical review, the natives pound the roots of Derris into a pulp
which they then throw into the water inhabited by fish. This allows the
juices of the plant to mix freely with the water and is the simplest way
of obtaining a water extract, but will water remove the toxic principle
after the roots have become dry ? The chief object of the investigation
discussed under the preceding heading was to make a study of the
different methods of extracting Derris and to determine the value of
various solvents in order that a simple and economical method might
be devised for obtaining the active principle and applying the extracts.
Aug. IS, 1919
Derris as an hisecticide
181
quantitative; extractions of derris and preliminary tests of
EXTRACTS obtained
Five series of quantitative extractions were made as follows : In each
series 20 gm. of fine powder of Derris sp. (probably D. elliptica) were
exhausted successively with the following five solvents in the order
named: First series, petroleum ether, ether, chloroform, alcohol, and
water; second series, ether, chloroform, alcohol, water, and petroleum
ether; third series, chloroform, alcohol, water, petroleum ether, and ether;
fourth series, alcohol, water, petroleum ether, ether, and chloroform; and
fifth series, water, petroleum ether, ether, chloroform, and alcohol. No
heat was used in any of these extractions. Table I gives the percent-
ages of extracts thus obtained. The sequence is shown by the letters
a . . . a, b . . . b, c . . . c, etc., beginning with the first extraction
in each case.
Table I. — Successive quantitative extractions of Derris sp. with various solvents, start-
ing with a different solvent for each series
No. of extraction.
Solvents used.
^^ethe?™ Ether. Chloroform. Alcohol. Water.
First. ..
Second.
Third..
Fourth .
Fifth...
Per cent.
4.07 a
4-55 e
.55d
.50 c
. 10 b
Per cent.
7. 90 b
4. go a
2. 00 e
.49d
. 20 c
Per cent.
10. 60 c
•75 b
. 20 a
• 50 e
.2od
Per cent.
II. 25 d
3.60 c
3- 30 b
2. 59 a
I. 10 e
Per cent.
9-75e
8.45d
5. 00 c
5- 05 b
10. 80 a
Attention is called to the following points in the preceding table.
From the first extractions it will be seen that petroleum ether is a poor
solvent, while the other four may be called good ones; of these four, only
alcohol and water can be regarded as economic solvents. Other points in
this table will be referred to later. Since the amount of an extract need
not necessarily correspond to its toxicity, the following preliminary tests
were performed.
Experience has taught that the honeybee {Apis mellifica h.) is ex-
tremely sensitive to stomach poisons ; therefore this insect Was fed small
quantities of the foregoing extiacts in order to determine the degree of
toxicity of each one. It was furthermore considered desirable to know
the effect of heat on the extracts. Consequently five of these extracts
were obtained without the application of heat and the other five with
the use of it. The following method of procedure was employed: Since
all of these extracts, except those obtained with water, have a consist-
ency similar to that of thick paste and are not soluble in water, it was
necessary to dissolve a small quantity of each in alcohol; therefore 0.4
gm. of the petroleum-ether extract was dissolved in 10 cc. of 95 per
cent alcohol. The same method was employed for each one of the
other nine extracts, including the water extract, so that the effect of
the alcohol would be the same in all the tests; and then X cc. of one
of these solutions was mixed thoroughly with 5 cc. of honey in a small
1 82 Journal of Agricultural Research voi. xvir, No. s
feeder, which was so covered with wire that the bees could not waste
any of the food. After the lo feeders, containing supposedly poisoned
food, had been placed in as many wire-screen cases, 50 normal bees
were introduced into each case; the bees were thereafter observed care-
fully and the dead ones were counted at regular periods. As a control,
honey containing the same amount of alcohol as mixed with the other
food was used; and whenever the bees required more food, pure honey
was given to them. These experiments were repeated and were so
arranged that the probable errors were minimized. Reference to Table
II (extracts No. 246-254) shows that all of these extracts, except the
water extract, are almost equally toxic to the honeybee within 48
hours and that there is practically no difference in toxicity between the
extracts obtained with the use of heat and without it. The water
extract apparently had no effect on the bees tested. Similar results
were obtained by using the same extracts against aphids, fall webworms
(Hyphantria cunea Dru.), and tussock-moth caterpillars (Hemerocampa
leucostigma S. and A.) (see No. 246-249, 252-253, Table IV, and No.
253, Table V). The water extract from the powder of Derris sp. (filtered
mixtures) killed only a small percentage of the aphids sprayed (see
lower half of Table IV), while the nonfiltered spray mixtures, consisting
of powder and soap solution, were efhcient against aphids.
To determine whether the solvents had removed all of the toxic
principle from the powders extracted, these five powders (No. 240-244
in Table II) after having been thoroughly dried were fed to other honey-
bees in the same manner as already described. In these tests ys gm.
of powder was thoroughly mixed with 5 cc. of honey. Reference to
Table II shows that the powders exhausted with ether, chloroform,
and alcohol had very little effect on the bees tested, while the powder
exhausted with water killed 94 per cent of the bees within 48 hours.
The results pertaining to the powder exhausted with petroleum ether are
not reliable (see note at bottom of Table II).
To ascertain the effect of powder exhausted successively with i to 4
of the solvents and also the effects of the resulting extracts, other experi-
ments were performed. Reference to Table II (No. 260, 261, 264, and
266) shows that powder successively extracted is only slightly less
effective than powder extracted once, and that the third and fourth
successive extracts (No. 263 and 265) have no effect at all. These
results agree in only certain respects with the successive quantitative
extractions, expressed in Table I.
To determine whether any poisonous volatile substance can be
removed from Derris by steam distillation, 50 gm. of the powder were
so treated and the distillate was collected. Later some of this distillate
and a portion of the distilled powder, after it had been dried, were tested
on silkworms. The distillate had no effect whatever, but the powder
was as poisonous as ever.
Derris as an Insecticide
183
To facilitate the handling of a product which might be used as a
proprietary insecticide, an alcoholic extract was incorporated into a
soft linseed-oil soap at the rate of i gm. of extract to 4 gm of soap.
This product was later dissolved in water in the proportion of i to 2,400,
which would be equivalent to about i pound of powder to 200 gallons of
soap solution. All of the small fall webworms sprayed with this solution
died, but none of the controls sprayed with soap solution of the same
strength died.
Table II. — Effects on the honeybee of eating extracts and powders of Derris sp. {probably
D. elliptica)
No. of
extract
or pow-
der
used.
246
247
248
249
250
251
252
253
254
240
241
242
243
244
245
258
260
261
264
266
259
262
263
265
Extracts, powders, and controls.
Petroleum-ether extract (no heat used) . .
Petroleum-ether extract (heat used) . . .
Ether extract (no heat used)
Ether extract (heat used)
Chloroform extract (no heat used)
Chloroform extract (heat used)
Alcoholic extract (no heat used)
Alcoholic extract (heat used)
Water extract (heat used)
Control , alcohol in honey
Control, honey alone
Powder exhausted with petroleum ether
Powder exhausted with ether
Powder exhausted with chloroform ....
Powder exhausted with alcohol
Powder exhausted with water
Powder not exhausted with any solvent.
Control, wheat flour in honey
Powder exhausted with petroleum ether
Powder exhausted with above solvent
and ether
Powder exhausted with above solvents
and chloroform
Powder exhausted with above solvents
and alcohol
Powder exhausted with above solvents
and water
Control, wheat flour in honey
Control, honey alone
Petroleum-ether extract from original
powder
Ether extract from above powder (2d
extraction)
Alcoholic extract from above powder
(3d extraction)
Water extract from above powder (4th
extraction)
Control, alcohol in honey
Number
of bees
tested.
150
150
150
150
150
150
150
150
150
150
100
100
100
100
100
100
100
100
100
ICO
100
100
100
100
100
100
100
100
100
Percentage of bees dead within-
48 hours.
After eat-
ing ex-
tracts and
controls.
67
73
48 hours.
7 days.
After eating powders
from which one or
more extracts had
been removed, and
controls.
ICO
4
2
6
94
100
73
16
18
16
IS
a This powder emitted an odor resembling that from petroleum ether; the bees ate very little of the honey
containing it, and therefore most of them probably died for lack of suitable food.
1 84
Journal of Agricultural Research
Vol. Xvn, No. S
Tables I and II show that 95 per cent ethyl alcohol is the only good
economic solvent used and that heat has no effect on the extract obtained.
It was decided, therefore, to make quantitative extractions of several
species of Derris by using hot denatured alcohol, since this solvent is
comparatively cheap.
Table III. — Quantitative extractions of various species of Derris made with hot denatured
alcohol
Name of species.
Part of plant used.
Percentage
of extract
obtained.
Derris sp. (probably D. elliptica)
D. elliptica, called "tuba"
D. uliginosa
D. koolgihberah
D. scandens
D. oligosperma
D. rohusta
D. robusta
Roots.
Roots.
Stems
Stems
Roots .
Stems
Roots.
Stems
14.25
8.50
8.50
10.30
20. 30
22. 50
16. 70
15-70
The foregoing table shows that denatured alcohol is a good solvent and
that the percentages of extract obtained vary considerably ; this variation
is certainly due in part to the fact that the eight powders used varied con-
siderably in fineness. Results showing the effectiveness of these extracts
are discussed on page 188 and in Table V.
EXTRACTION OF" TOXIC PRINCIPLE; FROM DERRIS SP. BY TWO METHODS
As already stated, Greshoff (j), van Sillevoldt (9), and Powers {8) have
agreed that the toxic principle in Derris elliptica and. D uliginosa is a resin
and have called the active portion of it " derrid." In the present investi-
gation it was considered expedient to isolate a small quantity of the resin
and to test it on insects and on a few higher animals.
VAN SILLEVOLDT 'S METHOD
One kilo of the powdered root of Derris sp. was repeatedly extracted
wdth boiling water until the extract was only slightly colored. After the
powder had been filtered and thoroughly dried it was boiled under a reflux
condenser with successive portions of 95 per cent alcohol until exhausted.
The combined alcoholic extracts were mixed with one-fourth their volume,
of water, and the alcohol was distilled under reduced pressure. As the
alcohol was removed, the material in the flask became milky in appearance
and the resinous substance collected in a mass on the bottom of the flask
The last portion of the water was removed by transferring the material to
an open dish on a steam bath. The residue was a resinous, sticky mass
which weighed 1 10 gm., representing 1 1 per cent of the dry root. It was
Aug. IS, I9I9 D err is as an Insecticide 185
dissolved in boiling alcohol, and then the solution was heated with animal
charcoal and filtered. Upon evaporation, the resin closely resembled the
appearance it had before being treated with the charcoal and seemed to
consist of two forms, the greater portion being of a soft and pliable nature,
while the other portion was hard and brittle. The latter had a melting
point of 66°-68° C. Van Sillevoldt reports the melting point of "derrid"
as being about 73° C.
Two gm. of the soft portion were dissolved in 50 cc. of 95 per cent alcohol
by means of a low heat ; upon standing, a fine, yellowish-white powder
settled to the bottom of the flask; then this powder was separated by
means of a force filter, and after being washed with a small quantity of
alcohol and ether it was dried. This material appeared like an amorphous
powder, but under the microscope it was found to consist of small plate-
like crystals. The melting point of these crystals was 170° C.
A dilute alcoholic solution of the above crystals, as well as the alcoholic
solution of the resin from which the crystals had been separated, was
found to be very toxic to fish. A subcutaneous injection of 0.00066 gm.
of the crystals was fatal to a mouse in two hours.
The preceding method of extracting the resin is not very practicable
on a large scale. Several of the operations involved could possibly be
dispensed with.
power's method
One kilo of the powdered root of Derris sp. was repeatedly extracted
with boiling alcohol until exhausted. Upon removal of the alcohol by
distillation under reduced pressure, 173 gm. of a dark extract of a pillular
consistency were obtained; this amount is equivalent to 17.3 per cent of
the dry root. Then the extract was repeatedly extracted under a reflux
condenser with hot petroleum ether until the latter was no longer
colored. From the combined extracts the petroleum ether was removed
and a waxy, yellow residue weighing 16 gm. remained. This residue was
designated A .
The alcoholic extract after having been exhausted with petroleum ether
was heated on a steam bath with 95 per cent alcohol until it was brought
into solution, whereupon it was poured slowly into a large quantity of cold
water; a fine suspended precipitate resulted. The precipitated resin
was filtered by means of a force filter, then washed with water, dried by
means of an electric fan, and finally pulverized to a No. 20 powder which
was grayish in color and weighed 102 gm., being equivalent to 10.2 per
cent of the original material. This was designated B.
Sixty-five gm. of the resin B were extracted in a Soxhlet extractor with
chloroform until exhausted; 11.4 gm. or 17.5 per cent remained undis
solved. This portion was removed from the extractor, was dissolved in
alcohol, and then precipitated in cold water. After the precipitate had
1 86 Journal of Agricultural Research voi. xvii. no. s
been filtered and dried, a chocolate-brown powder resulted. This was
designated C.
The chloroform extract from B was placed on a steam bath to remove
the chloroform. The residue resulting was a dark, sticky material which
became hard and brittle when cooled below room temperature. It was
ground while hard and was designated D.
The three substances designated A, C, and D were tested in very dilute
form on small chinook salmon and were found to be exceedingly toxic.
The extract A appeared to be the most powerful, while the chloroform-
soluble resin D was much more toxic than was C. The effect of the
extract A on the fish might have been influenced to a considerable extent
by a trace of petroleum ether which seemed to remain in the extract and
imparted to it a distinct odor.
The three substances called A , C, and D were tested also on small tent
caterpillars by being sprayed on foliage. Within eight days .4 had killed
70 per cent, C 92.3 per cent, and D 54.4 per cent of the caterpillars tested;
but only 22.1 per cent of the control lar\^ae had died.
EXTRACTION OF DERRIS ELUPTICA AND TESTS OF EXTRACTS OBTAINED
The roots of "tuba" or "toeba" were ground as fine as their fibrous
nature would permit, and 200 gm. of this powder were macerated for
two days with a quantity of cold water. After the mixture had been
filtered, the water extract measured 600 cc, each cc. representing K gm- oi
the roots. Half of this cold water extract was tested on small tent cater-
pillars ; within eight days only 30.9 per cent of them had died. The other
half of this extract was evaporated to one-half its volume on a steam bath
and then again made up to its original volume with water. This portion
of the extract was later tested on small tent caterpillars; within eight
days only 14.3 per cent of them had died. This does not mean that the
application of heat affected the toxicity of the extract, for 22.1 per cent
of the control larvae died.
The marc from the preceding water extractions was dried by means of
a current of air and was macerated with several portions of cold petroleum
ether. The combined extracts were then divided into two equal portions.
While the petroleum ether evaporated spontaneously from one portion in
an open dish, it evaporated on a steam bath from the other portion.
The residue resulting was a waxy, yellow substance which represented
1.4 per cent of the original material. Spray solutions containing these
petroleum-ether extracts were tested on aphids ; there was practically no
difference in effectiveness between the extract obtained without the use
of heat and the one with it (see No. 288 and 289, Table IV).
The powder left after the preceding extractions was spread out, and
the residual petroleum ether was allowed to evaporate. It was then
A-ug. IS. I9I9 Derris as an hisecticide i^'J
divided into two equal parts; one part was macerated with successive
portions of cold 95 per cent alcohol until exhaustea, and the other part
was boiled on a steam bath with successive portions of 95 per cent alcohol
until exhausted. The combined extracts from the first part represented
4.17 per cent of the original powder and those from the second part 4.26
per cent. Spray solutions containing these alcoholic extracts were
tested on aphids, small fall webworms, and on large tussock-moth cater-
pillars. There was practically no difference in their effectiveness on these
insects (see No. 290 and 291, Table IV).
EXTRACTION OF DERRIS ULIGINOSA AND TESTS OF EXTRACTS OBTAINED
The stems were reduced to a coarse powder, and 100 gm. of this material
were repeatedly extracted on a steam bath with petroleum ether until
exhausted. Upon evaporation of the petroleum ether, there remained a
yellow, shiny, somewhat brittle substance which represented 1.02 per cent
of the original stems.
The marc from the above extraction was dried thoroughly and then
exhausted with 95 per cent alcohol on a steam bath. The residue left
upon the evaporation ot the alcohol represented 7.82 per cent of the
stems. The above petroleum-ether and alcoholic extracts were found
very effective against aphids (see No. 293 and 294, Table IV).
EXTRACTION OF VARIOUS SPECIES OF DERRIS WITH DENATURED ALCOHOL
AND TESTS OF EXTRACTS OBTAINED
Since the preceding results have shown that alcohol is the most suit-
able solvent for the toxic resins, the use of denatured alcohol as the
best economic solvent was at once suggested. By the use of suitable
apparatus this solvent can be recovered with very little loss and con-
sequently can be used repeatedly.
For the tests described below, 50 gm. of powdered material in each
instance were extracted with denatured alcohol on a steam bath, and the
extract was concentrated to 25 cc. so that i cc. was equivalent to 2 gm.
of material.
In the tests performed in the laboratory, the general plan for each
test was to spray or dust about 500 aphids or 100 caterpillers on foliage,
and then to place this foliage in a bottle of water inside a battery jar
which was covered with cheesecloth. A record of the dead insects was
taken at regular periods. The tests with aphids usually covered a
period of 24 hours, and those with caterpillars and potato beetles
{Leptinotarsa decemlineata Say) 10 or 12 days. The results of most of
these tests are given in Tables IV and V. Table V gives chiefly the
results obtained by using denatured alcoholic extracts and the powders
of various species of Derris, applied as dust. Attention is called to the
1 88 Journal of Agricultural Research voi. xvii, No. s
following points: (i) The alcoholic extracts of elliptica, uliginosa, and
koolgibberah (No. 296, 295, and 298) were generally efficient, while those
of oligosperma, scandens, and robusta (No. 299, 300, 400, and 401) were
only seldom efficient; (2) the powder of Derris sp. (No. no), mixed
with water or soap solution, was usually efficient, while the other pow-
ders (No. 402-406) tested by this method were found inefficient; and
(3) of the eight powders used as dusts, only those of Derris sp., elliptica,
and uliginosa (No. no, 408, and 407) were found efficient.
Aug. IS, 1919
Derris as an Insecticide
189
"^
•t^~
013
"« 2
•S a
Q
•^^
2 0
o lu
O O M ■*
O 00 O O
00 00 i-t r^ o^ •
o\ c^ ON ^ o\ ov OGO 00
■0,43
5 en 01 _
0,-g.H 2
a§5t<
ly-i On »^ r^
, !^ 6 f^ >>)
I On ON 0\ t^
lO O 00 NO (
(> t^ -* ^ <
•a i-a ••0 --d -"d • • ■
aT) St) s'd s'd aTSTSTa
«^^^ S!^ oi-M So Si «
S333Sa23g333
oa>O'l>04>0(l'0'l'4>1>
6 0) <s wi
NO t^oo ON <
) O •-< 00 0> CO -^
I ON ON 00 00 ON On
. . . t3 iJ T3 ii Ci ,„
: : ;i:sK<os« >.
: : <u^ CI J. >. rt
: : :(n a!53^ °
: : i^^SSBt; 1
.9 : .'O'u'O'O S.8 ■
•^ i; •,^> s p=3-3 t
s a O-S p'c.s.s.s.s.s.s -i
""^O^OOOOOO
p'.
Ph W <l -< Ph Pn-^KcnPOlWl/lOTCflOO fi<.
p- P CD K . c:
9 « <n S & 3
O o u
" o ga
o b
'o5
M
, M
O o "
S "> G & "
Sgca 00
'2'S-d-d^
" n f^
t^ <^ K. !^ 3
.9.9 ^^a
>i >■ .
4> 0)' , ., ^
« « 1*
.2.2 § g'i
^.2 3 3 5
rt 3 2'S'3
O « c e o
.2.2 ggm
CJ CJ p
3 3-
^.33 3
" ":'o'o
rssiiiig
88888=22
"o'o'o'o'o o ° °
•a "2 "d -a -o ^-S ■S
d g c c c^ - -
3 3 3 3 3
oSoco3oo
aaaaao aa
a a!
.9 -9
S 5 <u
■^ *^ -S!
rt CTl rt
^ &:?
o o o
*J ■♦-» ■*-'
o o o
rt ca g
u-.— .„.''v_v«°Ooo^_'-r
O O O^ o O " " O O "
.2.2 2 2 2.2 o. a.2.2-2
rt t3 rt ts tats Jl j; rt «^
c c e o s
aa
p p
190
Journal of Agricultural Research
Vol. XVII, No. 5
<a ISO
V.
V.
•a
.g
J3
5
0
0
~
• .
M
5
i
i
p
; :
w 0
"d
0
•s
3
■^
0, \ri
•4-J
a
0
•a
•So
:§
"
^"'~"
1
1
•5! 2
3
t
p
0 m
ON ON
S
0
"o
0
0
a
"
0 a
_aj 0
S5
•d oJ •
■KXI 0
8
0 0
6 d
j 0 0 0 0 0 0
• 666666
ill
tfl 0 ^.'
e.D 0
8
0
0000
o\ 0* 0 0
• 000000
• 6 i^ M 0 6 »^
■ 0 vo 00 <f.O «
II
8
0 0
88
• 000000
■ 6 6 10 d\ d> d
• 00 0\ 01 C3i «
•3 J?
llo
.&.a
8
0 0
■Si:
0
-It
a
% ^
C8
0
8
t^ H 00 0
00 ^ ^ w
•vO 0 0 t^ ■* «
'• 4 4o6 i^ t^ ■*
C3
4^00 01 0^
■ o> •o t- « t M
01
i s •~
■U oi .
0
^f 10
.4
10
•3
e
-6^
i
3\
•i
<u ,
"o
d
^s&
8
0
.0
S
8
a
—
;-•
3
°
^
pi
^«°
0
V w
,
fO W 0 »0 TtOO 0\ 0 0 M .
0 +j
■0 te C-a
■/10\OiOSO>t7ie3>000 •
+J ^-*
; ^"
u a
■ 0
k
nJ rt
; a
IH »-
22
:S
0 a
a 0
00
. 0
>,
^1
^' ^' 0 0 ^' 0
"a
rt rt 0 0 rt 0
0
!-• Cd
d
g-s
^ '^ 'A ^ f^ a
0- a>T3T3 OJfQ
.0
0 0 Q-' CJ 0 CJ
3
•S3
•02
.X.^ (_ ^,.r- t-
sisslr
0
•a
0
0
•a
a
0
<<oo<,o
•d
s
In
_to
•i.^,.-^^ '
.«
Sl^ :
• 0
° K&e :
•3
c
*. 'S, -Si's a s-::!
u
W
:§ =; Ci qQQQQ5
^
0
P. .
•§1
CO 3
^2
u S
g-o
*■'
K
W
.2.2
"a a
a a
to a
o o
'S'o
<S<a
a a
V V
•Ot3
a a
a a
.9 "3
a c
.2 2
0*0
a a
o o
. a'B
■..■,, a '^ rS
.2 o cs -^
■^•3_2 "!.3
O o M§ Oj
00 o2«
S iftt!^ a
0 o o a c.
"s — _ o 2
'S'2-a'2 8
S a 5 a ^
o o i o "
c. c. g, a vh
S2 P2 o
<J c o tj u
Aug. 15, 1919
Derris as an Insecticide
«
•n
s
^1
•^j
e
u
0
0
■s-.
q
?s
>^
0
S
a
0
"a
M
■tv.
^^
'T^
•d
3
c«
•d
a
a
ca
4J
a
V
>
1
3
t;
"B .
2 3
n! >
M
II
ai in
2
t
a
CO
12
0.
"o
4-)
z
0
I
0
•a
"o
0
In
potato
patch.
u
0
0
8
s
1
<
0
0
0
t
a
0
0
8
■0 A •
US
8
0
1
^|3
■^ 0 t-i
5|s
0
0
e
0
8
0 0 0 10 0 0
6\ 6 6 «t "o i>i
0
Jl
IP
3-SS
0
0
e
0 VO i^-o 0 t^ 0 0 00
0 o\co 0 a\ 0^ o\ 0
.a-.a
a a u
0
0
0
0
ea
0
0
0 0 r-- 0 000
06 0 Ov "j -JjO
'"t
5|2
0
0
e
Ov '• j j
a
a
1
'0
8
a
0
1 0 0000
1 6 >o 10 10 10
-
•Isb
fi;S2
e 1
0 0 00 f^0^*^*^0\^MrOtN
6 6 ^ voodoi-j^^'^^ro
7.
No.
of the
ex-
tracts
and
pow-
ders
used.
0 0 0 (N fO ^ ^<2 0 '■^ ^ -* ^^ 1^00
M M M 0 OOOOiHOOOOOOO
k
a a
fa
5S
•33
•oS2
3.S
3^
Powder soaked in water 'A
hour.
< Powder soaked in soap solu-
tion Y hour.
IPowder soaked in soap solu-
tion 2 days.
Powder soaked in soap solu-
tion Y hour.
fin
■b- o-c5
•o-cd
"S 0 0 0" c
3 ■ • ■
Q : : :
3 0 C
5 c
3T
■ i
3 -^
: j:
1
s
1
p
•0
1
a
=i a E
Q Si
3 . -S ^
i :
'Sea
IIJI
» -o c c
£).sp. (roots)
D. knolgibberah (stems) . .
D. oligospcrma (stems). .
D. scandcns (roots)
3|i
3 a
i S
h
" 1
• c
J.
• 1
•0
a
M
1
1
3
1
3o
Is.
3
3 a
•a
>» rn
t; 0
1;
•0
(1<
5
3
a
o o
'o'S
.ail
ta<a
-O'O
a a
3 3
•d-d
TJ'd
td ca
>, >.
•CO
01 rt
3 3 rt
o o 3
3 3ti
EA (0 O
a a J2
a = l
° ° Si
000
Ut t- tj
(U OJ 0)
•d'O'd
& &|
000
a aQ
i« Ml vt:f
000
•d'd'd
a a g
3 3 3
000
a a n
sea
OOP
« CVS !»
t- 1.4 iH
(U t< 01
192 Journal of Agricultural Research voi. xvii. No. s
EFFICIENCY OF DERRIS AS AN INSECTICIDE
The experiments relating to the efficiency of Derris as an insecticide
were performed at the Insecticide Board's Testing Laboratory, located
at Vienna, Va.. The Derris material used was purchased on the open
market and is distinguished from the other materials employed by being
called powder derived from Derris sp.
EFFICIENCY OF DERRIS AS A CONTACT INSECTICIDE
The commercial powder, when used as a contact insecticide, was
applied in two forms: (a) as a dry powder and (b) as a spray mixture
with or without soap.
DERRIS APPLIED AS A POWDER AGAINST VARIOUS INSECTS
Dog FEEAS.^Eight dogs badly infested with fleas (Ctenocephalus cams
Curt.) were dusted thoroughly. The material was applied with a shaker
and well rubbed into the hair with the hands. At the end of 48 hours
no living fleas were observed. Several dead ones were seen still clinging
to the hairs.
Chicken i^icE. — Twelve hens badly infested with several species of
lice (Mallophaga) were thoroughly treated with the powder, which was
well rubbed in through the feathers. When the hens were examined two
or three days later, they were free from lice.
Chicken mites. — When this powder was freely dusted over the
chicken mites (Dermanyssus gallinae Redi) , confined in jars, all were
killed within 24 hours, but when used under practical conditions in a
badly infested chicken house, all of the mites were not killed.
Bedbugs. — Derris was tested against bedbugs {Cimex lectularius L.)
by placing 20 bugs in a jar with a quantity of excelsior and then thor-
oughly dusting the contents of the jar. In nine tests under these very
severe conditions 24.4 per cent of the bugs were killed in 24 hours and
52.8 per cent in four days. This material would be of no practical value
against bedbugs.
Roaches. — Six small cages were thoroughly dusted and 20 roaches
(Blattella germanica L.) were placed in each cage. At the end of one
week an average of 57.5 per cent of the roaches were dead, which indi-
cates that this material would be of very little value under practical
conditions.
House flies. — In cage tests, where house flies {Musca domestica L.)
were dusted in ordinary flytraps about 10 inches high, all were dead or
inactive within 24 hours. In room tests, where the powder was freely
blown into the air and all parts of the room with a small hand dust gun,
all of the flies were dead at the end of 1 6 hours.
Aug. 15, 1919
Derris as an Insecticide
193
In one test several hundred flies were liberated in a room which had
been thoroughly dusted seven days before. Twenty-four hours later
very few active flies were to be seen, and on the second day only three
or four were living.
Plant insects. — Derris applied as a dust was of no value against the
mealy bug {Pseudococctis citri Risso), the Orthezia {Orthezia insignis
Doug.), red spiders (Tetranychus bimaculatus Harv.), and the crawling
young of the oyster-shell scale (Lepidosaphes ulnii L.) ; but it was eff"ective
against nasturtium aphids {Aphis rumicis L.) and the green apple aphis
{Aphis pomi De Geer).
DERRIS APPLIED AS A SPRAY MIXTURE
Derris applied as a spray mixture was tested against the green apple
aphis {Aphis pomi De Geer) under field conditions and was found to be
very effective. Young apple trees, about 10 feet high, were used. The
spray mixtures were applied with a knapsack sprayer, except in tests
No. 3 and 6, in which a barrel sprayer was used. When soap was used
it was employed at the rate of i pound to 25 gallons of water. One
dusting experiment was performed, the powder being applied with a
large hand duster.
Table VI. — Results of field tests, using Derris powder in spray mixtures and as a dust
against the green apple a phi (Aphis pomi De Geer)
No. of
test.
3
4
5
6
7
8
9
10
Ratio of powder to water or soap solution used, dusting test, and
controls.
I pound of powder to 25 gallons of water
I pound of powder to 50 gallons of water
do
I pound of powder to 50 gallons of soap solution .
I pound of powder to 100 gallons of water
do
I pound of powder to 100 gallons of soap solution
I pound of powder to 150 gallons of water
I pound of powder to 150 gallons of soap solution
I pound of powder to 200 gallons of water
I pound of powder to 200 gallons of soap solution
Dusting test
Control , soap solution only
Control, trees untreated
Num-
ber of
trees
used.
5
5
32
3
52
5
5
5
5
5
Dura-
tion of
tests.
Days.
3
3
4
3
3
4
3
3
3
4
3
Aphids
killed.
Per cent.
100
ICO
95-100
100
100
80-90
100
100
100
"98-100
"98- 1 GO
100
O
O
a Four trees entirely free of aphids.
Table VI shows that Derris, even at the rate of i pound to 200 gallons
of water, was very effective against the green apple aphis under field con-
ditions and that on apple foliage the addition of soap does not increase
its effectiveness. It also shows that this powder is effective as a dust.
Under greenhouse conditions, in tests against the nasturtium aphis,
this material was found to be effective when used at the rate of i pound
122501
194
Journal of Agricultural Research
Vol. XVII, No. s
of powder to 400 gallons of water, with soap at the rate of i pound to
100 gallons.
OysTER-shelIv scale. — At the rate of i pound of powder to 20 gallons
of water, either with or without soap in the proportion of i pound to 100
gallons, Derris was ineffective against the crawling young of the oyster-
shell scale (Lepidosaphes ulmi L.).
While taking records of numerous greenhouse tests with Derris against
aphids, it was noticed that all the aphids were not killed during the first
24 hours but continued to die for several days. Since a contact insecti-
cide which continued to kill for a period of five or six days seemed an
anomaly, the following experiments were made to determine definitely
if this were the case and over how long a period this killing would extend.
The aphids on small potted plants, were counted, and the plants were
then thoroughly dusted or sprayed. Paper disks were placed around the
plants to catch the aphids that fell. Careful counts were made ever}- day
until all of the aphids were gone. In these counts each aphid was
observed through a lens, and when necessary each one was touched wath
the point of a knife to determine whether it was still alive. A single
untreated plant was used w'ith each series as a control.
The aphids began falling from the plants wathin an hour, but for the
first 24 hours most of those on the paper disks were alive. After this the
aphids that fell were practically all dead. In the case of the dusted plants
a few dead aphids were found clinging to the leaves the third day, but
as a rule only the living ones remained on the plants.
These tests fully confirmed the earlier observations and, furthermore,
showed that some of the aphids did not die until five or six days after
the application of the insecticide. The results are presented in Tables
VII and VIII.
Table VII. — Results of tests against nasturtium aphids (Aphis rumicis L.), using Derris
powder in spray mixtures at the rate of i pound of powder to 100 gallons of water
Percentage of aphids living on plant at end of —
Number.
First
day.
Second
day.
Third
day.
Fifth
day.
Sixth
day.
Seventh
day.
Eighth
day.
Aphids treated :
182
52.2
25-3
19. I
33-7
24.7
22. 6
II. 0
20.3
10. 4
7-3
9.0
23. 2
4.4
2. 0
8. I
19.7
2. 2
2.0
5-7
19.7
0-5
.0
3-3
18.6
0-5
. 0
ICO
200
2.8
172
12.7
Average
35- 0
19. 6
12. 7
8.5
7-4
5-6
4. 0
Aphids untreated :
I CO
96. 2
105.6
104.4
137- I
144.6
169.7
235-2
Aug. 15, 1919
Derris as an Insecticide
195
Table VIII. — Results of tests against aphids (Myzus persicae Sulz.) on cabbage plants,
using Derris powder as a dust
Percentage of aphids liring on plant at end of —
Number.
First
day.
Second
day.
Third
day.
Fourth
day.
Sixth
day.
Eighth
day.
Tenth
day.
Thir-
teenth
day.
Fif-
teenth
day.
Aphids treated :
q6
43-7
38.4
49-5
47.6
36.4
21. 0
29.7
31.0
31.2
6.3
II. 7
16.5
16.6
2-5
9.0
15-3
3- I
.6
4-5
9-7
0.0
. 0
1.8
2. I
0. 0
. 0
•9
1. 2
0. 0
. 0
•9
•4
0. 0
157
Ill
. 0
2^C
Average
44.8
29-5
16. 4
10.8
4-5
I. 0
0-5
0-3
0. 0
Aphids untreated:
180
103-3
128.3
146. I
170. 4 231. 6
315-5
315-5+
315-5+
315- 5+
Reference to Tables VII and VIII shows that the percentage of un-
treated aphids gradually increased from the first day of the tests onward ;
This increase was due to the birth of aphids on the untreated plants,
aphids were born likewise on the treated plants from the time the insecti-
cide was applied until all the reproducing females had died. Since
practically all of the aphids on the treated plants were dead at the close
of the tests, the newly born young ones must have been killed by coming
in close proximity to the particles of powder still remaining on the plants.
EFFICIENCY OF DERRIS AS A STOMACH POISON AGAINST VARIOUS INSECTS
Potato-beetle larv^. — Derris powder as a stomach poison was
tested on a small scale against potato-beetle larvae {Leptinotarsa decem-
lineata Say) at several strengths, ranging from i pound of powder to 16
gallons of water up to i pound to 128 gallons and was found to be very
effective. Practically all of the larvae were killed within 48 hours and
the plants were little eaten.
Since these spray mixtures might have acted as contact poisons,
because the larvae were already on the plants when the latter were sprayed,
a second series of tests was arranged to eliminate this factor. The same
plants were used and from 20 to 40 larvae were placed on them one or
two days after they had been sprayed. The results obtained were prac-
tically the same as in the first series of tests. Very few living larvae
were found three days later and the plants were little eaten.
When applied as a dust, Derris was equally efficient against potato-
beetle larvae.
Tent caterpillars. — Derris was tested against young tent cater-
pillars {Malacosoma americana Fab.) in a series of strengths ranging
from I pound of powder to 8 gallons of water to i pound to 200 gallons.
All the mixtures were found to be effective.
196 Journal of Agricultural Research voi. xvii, no. s
Apple tree branches were thoroughly sprayed, and after the foliage
had dried from 20 to 40 newly hatched larvae were placed on each branch.
The caterpillars began to show signs of discomfort within 48 hours and
were practically all dead in from 5 to 10 days. In no case was any mate-
rial amount of feeding observed.
In a second series of tests the larvae were placed on the branches and
sprayed after they had begun to form their tents. Under these condi-
tions sprays containing i pound of powder to 50 gallons of water and i
pound to 100 gallons killed all of the larvae within 24 hours. When i
pound to 200 gallons and i pound to 400 gallons were used all thelar\^ae
were not killed within 11 days, but the few which remained alive were
very small and inactive.
Used as a dust, this material killed all of the treated larvae within one
week.
Fall WEbworms. — These caterpillars (Hyphantria cunea Dru.), about
one-third grown, were killed within a week by a spray containing i
pound of powder to 5 gallons of water. Mixtures ranging from i pound
to 50 gallons to i pound to 200 gallons were not satisfactorily effective,
since nearly all of the sprayed foliage was eaten and not all of the cater-
pillars were killed.
Oak worms. — ^Two small oak trees, on which about 300 caterpillars
{Anisota senatoria S. and A.) were feeding, were sprayed thoroughly
with Derris at the rate of i pound of powder to 25 gallons of water;
soap was added at the rate of i pound to 50 gallons, and a knapsack
sprayer was used. Within 24 hours the larvae became inactive and
ceased to feed, and at the end of 6 days no living ones could be found.
As a check on this test, powdered arsenate of lead was applied at the
rate of i pound to 50 gallons of water, and almost identical results were
obtained.
A second test was made in which a small tree was sprayed, and 24
hours later about 50 larvae were placed on it. The caterpillars ate very
little and gradually disappeared, evidently leaving the tree, since no dead
ones were observed; and at the end of 5 days they were nearly all gone.
Datana larv^. — Two apple trees, on which large colonies of nearly
full grown apple datanas {Datana ministra Dru.) were feeding, were
sprayed with Derris at the rate of i pound of powder to 50 gallons of
water. Twenty-four hours later one living lar\^a was found on one
tree and two on the other. The ground under the trees was thickly
sprinkled with dead larvae and many had lodged in the trees.
Cabbage worms. — In two cage tests against cabbage loopers (Auto-
grapha hrassicae Riley), Derris, applied at the rate of i pound to 25 gallons
of water, killed all of the larvae within 24 hours.
Aug. IS, 1919 Derris as an Insecticide 197
PHARMACOLOGICAL EFFECTS OF TOXIC PRINCIPLE
The preceding experiments show that the toxic principle contained in
Derris kills insects both as a contact insecticide and as a stomach poison.
It now remains to be shown how this poison kills insects. This phase of
the work involves a careful study of the physiological effects of the toxic
principle on insects and of how it reaches the internal tissues.
PHYSIOLOGICAL LFFECTS
In the foregoing experiments it was observed that the various spray
mixtures and powders were effective only when they came in actual
contact with the insects tested. The following experiments were per-
formed to determine whether they would kill insects without coming in
actual contact with them. In these experiments only the powder from
Derris sp. was used.
Ten small fall webworms, confineo in an observation wire-screen case,
were placed }i inch above the surface of a strong mixture of Derris
powder and water so that the exhalation and vapors from the mixture
could pass freely through the wire screen. No effects on the insects were
observed which could be attributed to the presence of the insecticide.
Fall webworms, ants {Monomorium pharaonis L.), various species of
aphids, roaches, and the larvae of Prodenia ornithogalli Guenee were con-
fined in large, air-tight glass tubes with Derris powder so that they could
not touch it. As a rule, the exhalation from the powder had little effect
upon the confined insects. None of the webworms or larvae of Prodenia
died, and only a small percentage of the ants and aphids and only the
recently hatched roaches succumbed.
Most of the aphids dusted with Derris powder fell within a few hours
in a paralyzed condition from the plants bearing them, and then they lay
more or less helpless for a few hours before they died. Aphids sprayed
with Derris mixtures and extracts behaved almost normally and showed no
symptoms of paralysis; in short, they died very slowly and their behavior
was similar to that of those sprayed with quassia extracts, described by
Mclndoo and Sievers (7, p. 523). Honeybees fed extracts of Derris
seemed to die of motor paralysis; and their behavior was similar to that
of those fed nicotine, described by Mclndoo (<5, p. pj); but it was some-
what different from the behavior of those fed arsenic.
HISTOLOGICAL METHODS OF TRACING DERRIS POWDER AND SPRAY MIX-
TURES IN INSECTS
Small individuals of fall webworms, caterpillars of Datana, silkworms,
and cockroaches, confined in wire-screen observation cases, were dusted
with Derris powder (No. 200) . Three hours later all of them were " stupid,"
and after being removed from the cases they were put in vials containing
thick celloidin. After remaining in the celloidin an hour they we're put in
other vials containing chloroform. Then an hour later they were cut into
198 Journal of Agricultural Research voi. xvii.no. s
small pieces and were fixed in a liquid containing equal parts of absolute
alcohol and chloroform with corrosive sublimate to excess. The thick
celloidin completely covered the integuments of the dusted insects and
held the particles of powder where they were already adhering to the
hairs and integuments. It did not pass into the mouth, anus, or spiracles
but ran into all of the crevices and surrounded the hairs. The
chloroform soon made the celloidin hard, thereby forming a hard layer
around the insect, and thus holding the powder in position. Sections
made from this material were stained with eosin in equal parts of absolute
alcohol and chloroform. This method kept the celloidin hard and thus
firmly held the particles of powder in position.
A study of the sections described above showed the following : A thick
layer of celloidin, dotted with particles of powder, completely surrounded
the integument, and processes from it ran into all of the crevices or inden-
tations of the integument. The heat in the paraffin bath caused the
celloidin to shrink, thereby drawing it away from the integument at
places ; but at other places it remained in contact with the integument.
Most of the powder in the layer of celloidin lay against the integument
and none could be seen inside the insect, except particles here and there
which seemed to have been dragged inside by the microtome knife; none
was seen in the tracheae and only occasionally was a small amount
observed in the spiracles, but never enough to clog them.
To be able to trace the powder better and distinguish it irom the par-
ticles of food in the intestine, the following experiments were performed:
Eight fall webworms were dusted with a mixture of Derris powder and
lamp-black, and eight more with a mixture of Derris powder and carmine;
the lamp-black and carmine were finely pulverized and were mixed thor-
oughly and in equal proportions with the Derris powder. The first four
of each set were three hours later fixed intact in the modified Carnoy's
fluid (equal parts of absolute alcohol, chloroform, glacial acetic acid, and
corrosive sublimate to excess) ; and the second four of each set were
treated by the celloidin process, described above. Many sections were
made from the material of each set; one-half of those from the material
dusted with the Derris powder and lamp-black mixture were stained in
the mixture of absolute alcohol, chloroform, and eosin; and the other
half were left in the paraffin-ribbon stage on the slides and not stained.
The sections from the material dusted with the Derris powder and carmine
mixture were likewise treated, one-half being stained with methylin blue
in 95 per cent alcohol and the other half being left unstained in the par-
afRn-ribbon stage.
A study of these sections showed the following: The black and red
powders were easily traced around the outside of the integuments but
never in the tracheae, and only occasionally did a small amount lie in
the mouth of a spiracle. In many of the sections, small masses of the
Aug. 15. 1919 Derris as an Insecticide 199
colored particles lay inside the integument; but most of them, if not all,
seemed to have been dragged there by the microtome knife, or washed there
by the staining liquid and xylol. However, a careful study of the par-
affin-ribbon sections, from the material dusted with the Derris powder
and carmine mixture, showed red powder only on the outside of the integu-
ment and none inside, except a small amount here and there in the
intestine.
Mclndoo (6, p. 103) has shown tnat nicotme spray solutions not con-
taining soap do not pass into the tracheae of certain aphids and cater-
pillars, and the same is true for quassia-spray solutions not containing
soap. Quassia-spray solutions containing soap, however, pass freely into
the tracheae and finally reach the various tissues (7, p. 525). In view of
these results it was not considered necessary to trace Derris extracts con-
tained in water and in soap solution.
The preceding histological study seems to show the followmg: Derris
powder dusted upon insects does not pass into the tracheae, but a limited
amount of it may lodge in the spiracles, though never sufficiently to inter-
fere with breathing. In order that the vapors and exhalation from a
. nicotine-spray solution be efifective, it is necessary for the insects sprayed
to carry some of this solution on their bodies; likewise it is necessary for
the insects dusted with Derris powder to carry some of this powder on
their bodies in order that its exhalation may pass into the spiracles in as
undiluted a condition as possible. After being dusted the insects seem
to swallow some of the powder, which later may act as a stomach poison.
Soap solutions containing Derris extracts pass freely into the spiracles
and finally reach the various tissues, but probably the extracts kill by
first affecting the ner\-e tissue.
SUMMARY
Derris, known widely as a pow^erful East Indies fish poison, was
found to fulfill several of the requirements of a general insecticide; it
acts both as a contact insecticide and as a stomach poison, but is of no
practical value as a fumigant. Six species of Derris were tested, but
only two of them {elliptica and uliginosa) were found to be satisfactory
for insecticidal purposes.
According to the views of various authors, the toxic principle in Derris
is a resin, which affects the various classes of animals according to the
development of their nervous systems. It kills some insects easily and
others wath difficulty, but it usually acts slowly and seems to kill by
motor paralysis.
Denatured alcohol was found to be a good economic solvent for ex-
tracting the toxic principle, which when applied in spray mixtures proved
to be efficient against certain aphids, potato-beetle larvae, and small
fall webworms. For proprietary insecticides it is possible to incorporate
the extracts from Derris into soft soaps which when greatly diluted
with water are readv for use.
200 Journal of Agricultural Research voi. xvii. no. 5
Derris powder, used as a dust under practical conditions, was found
to be efficient against dog fleas, chicken lice, house flies, three species
of aphids (Aphis rumicis L., Aphis pomi De Geer, and Myzus persicae
Sulz.), potato-beetle larvae, and small fall webworms, but of no practical
value against bedbugs, roaches, chicken mites, mealybugs, Orthczia
insignis, red spiders, or against the crawling young of the oyster-shell
scale. Used as powder in water with or without soap under practical
conditions, it proved to be efficient against most of the aphids sprayed
and also against cabbage worms (Autographa brassicae Riley), the larvae
of apple datanas (Datana ministra Dru.), oak worms (Anisota sanatoria
S. and A.), small tent caterpillars, and potato-beetle Isltvse:.
LITERATURE CITED
(i) Campbell, J. Argyll.
1916. AN EXPERIMENTAL INVESTIGATfON CONCERNING THE EFFECTS OF " TUBA "
(derris elliptica) fish-poison. In Jour. Straits Branch Roy.
Asiatic Soc, no. 73, p. 129-137.
(2) Dymock, William, Warden, C. J. H., and Hooper, David.
1890. pharmacographia indica. v. I. London.
(3) GrEshoff, M.
1890. mittheilungen aus dem chemisch-pharmakologischen labora-
TORIUM DES BOTANISCHEN GARTENS ZU BUITENZORG (jaVA). In
Ber. Deut. Chem. Gesell., Jahrg. 23, p. 3537-355°-
(4) Hasselt, E. H. van.
191 1. uEBER DIE piiysiologische wirkung von derrid, pachyrhizid vnd
NEKOE. In Arch. Intemat. Pharmacod. et Ther.,v. 21, fasc. 3/4, p.
243-279. 6 fig.
(5) Hooker, J. D.
1878. report on the progress and condition op the royal gardens
AT KEw, 1877. 53 p., I fig.
(6) McIndoo, N. E.
1916. EFFECTS OF NICOTINE AS AN INSECTICIDE. In Jour. Agr. Research, v.
7, no. 3, p. 89-122, pi. 1-3.
(7) and SiEVERS, A. F.
1917. QUASSIA EXTRACT AS A CONTACT INSECTICIDE. In Jour. Agr. Research,
V. 10, no. 10, p. 497-531, 3 fig. Literature cited, p. 528-531.
(8) Power, Frederick B.
1902. the CHEMISTRY of THE STEM OF DERRIS ULIGINOSA BENTH. AN EASTERN
FISH POISON. In Pharm. Arch., v. 5, no. 9, p. 145-160, v. 6, no. i,
p. 1-14.
(9) SiLLEVOLDT, H. E. Th. van.
1899. UEBER DAS DERRID UND PACHYRHIZID, EIN BEITRAG ZUR KENNTNIS DER
INDISCHEN FISCHGIFTE. In Arch. Pharm., Bd. 237, p. 595-616.
(10) Watt, George.
1890. DICTIONARY OP THE ECONOMIC PRODUCTS OF INDIA. V. 3. London,
Calcutta.
(11) Wray, Leonard, Jr.
1892. ON THE MALAYAN FISH POISON C.\LLED AKER TUBA, DERRIS ELLIPTICA.
In Pharm. Jour, and Trans., v. 52 (s. 3, v. 23), p. 61-62.
EFFECTS OF HEAT ON TRICHINA
By B. H. Ransom, Chief of the Zoological Division, and Benjamin Schwartz, Junior
Zoologist, Bureau of Animal Industry, United States Department of Agriculture
INTRODUCTION
It is a well-known fact that the larvae of Trichinella spiralis, which
are of rather common occurrence in pork, may be killed by thorough
cooking and the meat thereby rendered safe for food so far as concerns
the danger of trichinosis. As to the actual temperature required to kill
the parasites, however, various writers give very different figures, so
that the question of the thermal death point has been rather uncertain.
The thermal death point of trichinae is a matter of great practical
importance in connection with the control of cooking processes employed
by meat-packing establishments in the preparation of cooked products
containing pork. The simple rule of cooking pork until it is well done,
which can be applied satisfactorily by a careful cook in the household
kitchen, is not suited to conditions in meat-packing establishments.
Instead of such a rule a more exact statement of requirements is desir-
able. In fact, the Bureau of Animal Industry, which is charged with
the enforcement of the federal meat-inspection law, requires that pork
or products containing pork cooked in establishments operating under
Federal inspection shall be heated sufficiently to insure a temperature
throughout all portions of the meat that will destroy the vitality of any
trichinae which may be present, specifically a temperature of 137° F.
(58° -^ C). This temperature is several degrees higher than the tempera-
ture that has been accepted by the bureau as representing the thermal
death point of encysted trichinae, but the difference between the two
represents no more than a reasonable allowance as a margin of safety.
Before a decision could be reached as to the degree of heat required
to destroy the vitality of encysted trichina, it was found necessary to
supplement the investigations on this question which are recorded in the
literature with further experimental work; and it is the purpose of this
paper to set forth the results obtained. This work was begun by the
senior writer in 191 3, continued in 191 4 and 191 5, and in the latter part
of 1 91 5 taken up by the junior writer.
REVIEW OF LITERATURE
Haubner, Kiichenmeister, and Leisering (5) * state that trichinae are
killed by prolonged salting, followed by 24 hours of smoking, but do not
give data as to the temperature of smoking.
1 Reference is made by number (italic) to "Literature cited," pp. 220-221 .
Journal of Agricultural Research. Vol. XVII. No. 5
WASmNGTON, D. C. Aug. 15. 1919
sc Key No. A-48
(201)
202 Journal of Agricultural Research voi. xvii. no. s
Fiedler (i, p. 26-29) found that if small particles of trichinous meat
were heated to 35° R. (43.75° C.) in water the heating had no other effect
than to render the parasites more active when viewed at the same
temperature under the microscope. Similar results were obtained by
heating to a temperature of 40° R. (50° C). The trichinae in finely
chopped meat held at a temperature of 50° R. (62.5° C.) for 15 minutes
and then cooled were found to show movement when gently warmed,
but reexamination of the meat 24 hours later failed to show any trichinae
that would move when warmed. This experiment was frequently
repeated with similar results, and similar results were obtained with a
temperature of 52° R. (65° C). Temperatures of 58° R. (72.5° C.)
and upward,, allowed to act for a period of 10 minutes in all cases, affected
the parasites so that no movement occurred afterward when gentle heat
was applied. Three rabbits and a cat were fed trichinous meat after it
had been heated 10 minutes at a temperature of 50° R. (62.5° C), and
none became infected. Trichinous meat heated 10 minutes at a temper-
ature of 40° to 42° R. (50° to 52.5° C.) infected a rabbit. In another
experiment meat heated at 40° R. (50° C.) for 10 minutes failed to infect
a young cat. Trichinous meat heated at 60° R. (75° C.) for 10 minutes
failed to infect two rabbits.
In another paper Fiedler (2, p. 467-468) reported an experiment in
which he fed two rabbits with minced trichinous meat that had been
heated in water for 10 minutes at a temperature of 50° R. (62.5° to
65° C). No infection resulted. He also reported an experiment in
which two rabbits were fed with trichinous meat that had been heated in
water for 10 minutes at a temperature of 45° to 46° R. (56.25° to 57.5° C).
No infection resulted.
Haubner (4) states that the smoking of pork at a temperature which
reaches and exceeds 52° R. (65° C.) kills the trichinae or brings about
their early death.
Rodet (12) states that trichinae do not die at a temperature of 55°
to 60° C. He also asserts that they survive even a temperature of 70° to
80° C. and succumb with certainty only to a temperature of 100° C. In
support of his views Rodet presents very imperfect experimental evidence.
He states that he placed pieces of trichinous muscle in water at a temper-
ature of 70° to 80° C. and allowed them to remain there for some time.
Upon being taken out of the water the trichinae in the meat were still
lively. When plunged into water at 100° C. they were killed and became
completely uncoiled.
Fjord and Krabbe (j) concluded that encysted trichinae die at 52.5° C.
after a 30 minutes' exposure. At 54° C. they sur\dved 10 minutes and
at 55° to 56° C. they died in 5 minutes. Their method of procedure
consisted in cutting up trichinous meat and heating it in a vessel contain-
ing warm water while agitating the contents with a thermometer. To
Aug. IS, I9I9 Effects of Heat on Trichince 203
determine the effects of the heating upon the vitality of the parasites
they fed the meat to rabbits, which were examined for trichinae 15 to 30
days after feeding.
Perroncito (7) records observations on the behavior of the larvae under
the influence of high temperatures and draws the conclusion that a tem-
perature of 48° to 50° C. is sufficient to kill the parasites. He placed
decapsuled larvae as well as encysted larvae in salt solution and examined
them on a warm stage. He observ^ed that as the temperature increased
the larvae became more active, but that at 45° C. their activities ceased.
If the temperature was lowered they resumed their activities. If the
temperature was raised to 48° or 50° C. they became completely inactive
and remained so even when the temperature was lowered.
Vallin (ij) records a series of experiments on the effects of heat on
trichinae. He heated small pieces of trichinous meat in tubes containing
water, placed the tubes on a sand bath, and read the temperatures on a
thermometer with which each tube was provided. He found that a
20-minutes' exposure to a temperature of 60° C. resulted in a complete
destruction of the vitality of the larvae. He fed the heated meat to two
rabbits and four guinea pigs and failed to infect them. Vallin states
that temperatures below 60° C. are uncertain in their effects, since after
heating meat to 56° C. he succeeded in infecting with it one guinea pig,
although two rabbits to which the meat was fed escaped infection.
He tried temperatures lower than 56° C. and found them ineffective.
Leuckart (6) states that Trichinella spiralis does not perish until it is
acted on by a temperature ranging betw^een 62° and 69" C.
Piana {8) concluded as a result of certain experiments that a temper-
ature of 56° C. is fatal to the larvae of Trichinella spiralis.
Ransom (jo, p. 159) states:
With reference to the effects of high temperatures upon the vitality of trichinae,
various statements are found in the literature which seem to have for the most part
rather imperfect experimental evidence as a basis. From a rather small series of
experiments conducted within the last two years, I have found that encysted trichinae
regularly die when exposed for a short time to a temperature somewhere between
53° and 55° C.
The earlier of these experiments supplied the data upon which w^as
based the following statement (9): "The results already obtained in the
investigations . . . show that the parasites die after a brief exposure
to a temperature between 53° and 55° C."
Winn {14) records a series of experiments in which trichinous meat
was heated to certain temperatures, maintained at those temperatures
for 15 minutes, and then fed to experimental animals. The effect of
the heat was judged by the degree of infection as compared with that of
animals fed on similar quantities of meat w^hich were unheated. Winn
found that temperatures below 53° C. produce no apparent effect upon
204 Journal of Agricultural Research voi. xvii, no. s
the vitality of the worms. At 53° C. he found the vitality of the worms
slightly reduced, but the results were variable. At 54° C. there was a
further reduction in vitality, but meat which was heated to 55° C. and
maintained at that temperature for 15 minutes was not capable of
producing an infection.
EXPERIMENTAL WORK
Experiments by the present writers on the effects of heat on the larvae
of Trichinella spiralis have been made with meat containing encysted
larvae as well as with larvae freed from their capsules by artificial diges-
tion. In the former case there is more or less difficulty in obtaining accu-
rate data, since the temperature in the interior of the meat does not neces-
sarily correspond to the temperature of the medium in vvhich it is heated.
This difficulty may be overcome, however, if small pieces of muscle tissue
are used and if the temperature is raised gradually. In experiments on
larvae freed from their cysts by artificial digestion more accurate deter-
minations can be made, since the temperature of the medium is an excel-
lent index to the temperature of the parasites themselves. From a com-
parison of the results obtained by the two methods definite conclusions
regarding the thermal death point of the larvae may be drawn.
OBSERVATIONS ON THE SURVIVAL OF DECAPSULED LARV^ IN
VARIOUS MEDIA
In comiection with experiments on the effects of heat upon decapsuled
larvae, the question of their sur\nval in various media following artificial
digestion is important, since such experiments are complicated by the
factor of abnormal environment, and results obtained might not corre-
spond with those obtained in experiments in which the parasites are sub-
jected to heat while still inclosed in their capsules in pieces of meat. En-
cysted trichinae may be kept alive for many months and may still be
viable in meat that has become badly decomposed. Although decap-
suled larvae are unlikely to survive as long as encysted larvae, they can be
kept alive for considerable periods of time. In a paper by the senior writer
(Ransom, 11), it has been shown that decapsuled larvae may retain their
normal activity and appearance when kept in tap water or 0.6 per cent
salt solution at a temperature of about 20° C. for a period of from 10 days
to two weeks or more, and that they have been kept alive and very active
for as long as 1 1 days in 2 per cent salt solution. On the other hand, at a
temperature of 38° decapsuled larvae kept in tap water became inactive
within a few hours, whereas when kept in 0.6 per cent salt solution at the
same temperature for the same length of time they suffered no apparent
injury.
Further observations have been maae by the junior writer which show
quite clearly that the longevity of the larvae after artificial digestion de-
pends upon both the medium in which they are kept and the temperature
Aug. IS. 1919 Effects of Heat on Trichince 205
to which they are subjected. Pure water as compared to physiological
salt solutions was found to be distinctly injurious, the injurious action va-
rying directly with the temperature. Larvae kept in distilled water at a
temperature of 39°-4o° C. were all dead at the end of 22 hours, while in 0.7
per cent solution of sodium chlorid or in Ringer's solution they lived longer,
although they all died within 48 hours. In distilled water at a tempera-
ture of 32°-33° decapsuled larvae were nearly all uncoiled at the end
of 48 hours, while in 0.7 per cent sodium-chlorid solution or in Ringer's
solution some were still alive at the end of 5 days. Similar differences
were observed in the case of lower temperatures. In distilled water at
25°-26° larvae remained alive for 4 days; in physiological salt solutions
at 25 °-2 7° some were still alive at the end of 13 days; in distilled water
kept at a temperature of about 8° only a few larvae were still alive at
the end of 12 days; while in physiological salt solutions at the same tem-
perature some larvae were still alive at the end of 50 days.
From these observations and our general knowledge of the phenomena
of osmosis it would appear that the loss of salts from the tissues of the
worms into the water and the penetration of the water into the tissues of
the worms are important factors in bringing about the death of the worms
when kept in hypotonic media, such as distilled water. This belief is borne
out also by the fact, noted in a former paper (11, p. 849) and repeatedly
observed since that paper was written, that larvae kept in a hypotonic
solution until they have begun to show distinct evidence of its effects, such
as loosening of their coils and paling of their protoplasm, if transferred
to a physiological salt solution before the injurious action of the hypo-
tonic medium has gone too far, will usually resume a normal state of con-
traction and a normal or almost normal brown color. Another indication
that the death of decapsuled larvae kept in hypotonic solutions may be
dependent upon osmotic processes is that they die more quickly at high
than at low temperatures, which is in harmony with the fact that osmosis
is hastened by raising the temperature.
Another factor or factors, however, enter into the matter, inasmuch as
in isotonic solutions as well as in hypotonic solutions the larvae do not
survive so long at high temperatures as at low temperatures. It may be
supposed that at the higher temperatures death of the lar\^ae kept in iso-
tonic and comparatively inert solutions is brought about by exhaustion
resulting from the greater activity of the worms and consequently more
rapid oxidation of their tissues than at lower temperatures. Such an
explanation is complicated by the fact that larval trichinae encysted in
the muscles of a living animal may live for many years, although con-
stantly subjected to a temperature at which they live only two or three
days when removed from their cysts and kept in salt solutions. Possibly
in the living animal they are kept in a relatively inactive condition
through the operation of factors no longer present when they are removed
2o6 Journal of Agricultural Research voi. xvii, No. s
from their normal environment, and it is possible also that they may be
able to replace waste through the absorption of nutritive materials from
their host.
A natural corollary to experiments on the effects of hypotonic solutions
are experiments on the effects of hypertonic solutions. A typical
example of such an experiment is one in which decapsuled larvae were
kept for 22 hours in a molar solution of dextrose. At the end of this
time they were found to be partially uncoiled; their protoplasm was dull
in appearance; the cuticle was wrinkled, particularly in the posterior
portion of the body; the body wall was wrinkled; and the cells of the
esophagus were indistinct. After having been transferred to and kept
in 0.7 per cent salt solution overnight, they were found to be tightly coiled
and normal in appearance. Similar results were obtained in a repetition
of this experiment.
So far as concerns the purposes of the present paper, the foregoing
observations are of interest because they show that trichinae freed from
their cysts by artificial digestion may be kept alive for a long time in
physiological salt solutions, in water, and in certain hypertonic solutions,
and that, although within a temperature range the upper limit of which
does not exceed 40° C. their longevity decreases as the temperature at
which they are kept is raised, they do not in any case die quickly.
EXPERIMENTS WITH DECAPSULED LARVAE
Inasmuch as trichina larvae that have been freed from their cysts by
digestion of finely chopped trichinous meat in artificial gastric juice ^ at
a temperature of 38° to 40° C. for a period of about 20 hours can be kept
alive for long periods of time, they can be conveniently used in experi-
ments on the effects of heat. In a medium such as a 0.6 per cent or 0.7
per cent solution of sodium chlorid, but also in plain water if not kept
too long, they display more or less activity even at ordinary room tem-
peratures but commonly assume a posture in which they are tightly
coiled spirally; and their movements are often limited to a tightening or
loosening of the coil. Their protoplasm, when unaffected by heat or
other injurious agents, exhibits a certain brilliancy in appearance; and
pigment in the cells of the alimentary tract, especially of the esophagus,
gives them a distinct brownish color. After a little experience, depar-
tures from the normal both as to their behavior and appearance of their
protoplasm can easily be detected by microscopic examination. As a
rule, in experiments in heating decapsuled larvae, the larvae were placed
in a beaker or test tube containing sometimes water but usually a phys-
' The following fluid has yielded satisfactory results:
Scale pepsin (U. S. P.) 2. s gm.
Sodium chlorid 2 gm.
Hydrochloric acid (sp. g. 1.19) 10 cc.
Water. 1,000 cc.
Aug. IS, 1919 Effects of Heat on Trichince 207
iological salt solution or Ringer's fluid; and this was heated to the
desired temperature on a water bath over an open flame, or in an incubator.
After being cooled, individuals were removed with a pipette to hollow
ground slides, or in some cases transferred to a Petri dish or shallow
stender dish and allowed to cool. They were then examined directly on a
warm stage, either on slides or in the dishes, in order to determine the
results of the experiment.
BEHAVIOR OF DECAPSULED LARV^ WHEN HEATED
When trichina larvae are heated on a warm stage their reactions may
be directly observed with the microscope. As the temperature rises
they begin to uncoil and become very active, their activity gradually
increasing. When the temperature has reached the neighborhood of
50° C. spasmodic contractions are commonly observed, and the larvae
twist themselves into various shapes. With a further rise of temperature
they grow sluggish and may become either uncoiled and inactive or else
tightly coiled and quiescent. After passing into this sluggish condition
they may again become lively if the temperature is lowered, but if sub-
jected to a sufhciently high temperature for a sufficient length of time
they do not recover when removed to a cool place.
Decapsuled trichinae killed by heat usually become uncoiled and assume
a characteristic shape resembling the figure 6. If allowed to stand for
some time the protoplasm becomes dull, certain granulations appear,
and often the cell partitions in the gonads can no longer be distinguished.
Larvae in this condition are readily recognizable as dead. Sometimes,
however, larvae that have been subjected to heat may remain loosely
coiled and the protoplasm may not undergo any conspicuous changes.
From experience it has been learned that larvae in this condition are
usually dead. A generally satisfactory test of life is heat stimulation;
if still viable the larvae will usually uncoil and move. Even individuals
with a minimum amount of vitality will move the anterior or posterior
end very sluggishly. However, the most reliable test of life, or at least
of their viability from a practical standpoint, is feeding them to experi-
mental animals and thus determining their ability to reproduce; and
this has been done in some instances but not so regularly as in experi-
ments on encysted trichinae.
DETAILS OF EXPERIMENTS
Some experiments on the effects of heat on decapsuled trichinae were
made by the senior writer in 1913, 1914, and 1915, after which the work
was taken up by the junior writer and continued along the same general
lines.
Experiment i (April 5 and 7, 191 3). — A decapsuled larva was sealed
under a cover glass in salt solution on a slide and heated to 54° C. on a
2o8 Journal of Agricultural Research voi. xvii. no. 5
warm stage. The temperature was held at 54° for a few moments. The
worm was inactive at this temperature but resumed its movements when
the sHde was cooled. The same worm was reheated to 55° and became
entirely motionless at this temperature. The temperature was raised
to 55.5° and the slide then cooled. The worm became active again on
cooling.
Another decapsuled larva was heated in the same manner. It became
sluggish in its movements and coiled up at a temperature of 48° C. The
temperature was raised slowly to 56°, and the slide was allowed to cool as
soon as this temperature was attained. The worm resumed its active
movements when cooled. In order to check the correctness of the tem-
perature indicated by the thermometer in this experiment, some crystals
of diphenylamin having a melting point of 54° were placed on a slide
under a cover glass and heated on the stage. They melted when-the
thermometer registered 54°. A second trial gave the same result.
On April 7, a decapsuled larva was heated as described above. The
temperature was raised slowly to 56° C. and then held for five minutes at
56° to 56.5°. When cooled the worm did not resume its movements, its
internal structure showed slight disorganization, and it was undoubtedly
dead.
Experiment 2 (March 28, 1914). — Decapsuled trichinae, isolated by
artificial digestion from a mixture of meat from three trichinous rats,
were heated in a beaker of constantly stirred water over a hot water
bath to a maximum of 53.6° C, 10 minutes being required for the tem-
perature to rise to this point from 30°. The temperature dropped to
46.2° in another 10 minutes, after which 233 larvae were examined at
room temperature. All were inactive. Unheated larvae from this lot
when examined at room temperature were active. Another lot of larvae
from the same source was heated in the same maimer, the temperature
rising from 20° to 51° in 21 minutes, and then dropping in 6 minutes to
45.8°. One hundred and thirty-nine larvae were then examined at room
temperature, and 65 of them were found to be inactive. Of the 74 active
larvae, all but 2 were sluggish. A third lot of larvae from the same source
was heated in the same manner from 24° to 50° in 12 minutes, and then
cooled to 46° in 6 minutes. Out of 159 examined, 18 were inactive.
Some of the 141 active larvae were sluggish.
Experiment 3 (May 16, 1914). — Decapsuled larvae, isolated by artifi-
cial digestion from a mixture of meat from two trichinous rats, were
heated in a beaker of constantly stirred water over a water bath. The
temperature was raised from 23° to 48.4° C. in 8 minutes and held at
48.4° I minute. The beaker was then allowed to cool. One hundred and
ten larvae were examined on a warm stage. Thirty-five were inactive
and 75 active, mostly very lively. Another lot of larvae from the same
source was heated in the same manner from 22° to 51° in 10 minutes.
Aug. 15. I9I5, Effects of Heat on Trichince 209
Examination of 213 larvae on a warm stage showed 179 inactive and 34
active, most of them very lively. Another lot was heated from 30° to 5 1 .9°
in 10 minutes. Ninety-nine were examined, and of these 93 were inactive
and 6 active. Another lot was heated from 30° to 53° in 4 minutes.
One hundred and eighteen were examined, and of these 72 were inactive
and 46 active, sluggish. Another lot was heated from 22° to 53° in 12
minutes. One hundred and forty-seven were examined, and of these
109 were inactive and 38 active, sluggish. As a control upon the re-
sults of this experiment 158 unheated larvae from the same source as those
subjected to heat were examined on a warm stage. Of these 22 were
inactive and 136 active.
Experiment 4 (November 17, 1914). — Decapsuled larvae, isolated by
artificial digestion from the meat of a trichinous hog, were heated in a
beaker of water over a hot water bath for a period of 10 minutes, during
which time the temperature gradually increased from 23° to 53.4° C.
The beaker was then cooled. Seventeen of the larvae were examined on
a warm stage and one was observed to move slightly. Fifteen minutes
later the lar\^ae remaining in the beaker were reheated to a temperature of
53.6° C, seven minutes being required to raise the temperature to this
point from 38°. Twenty-four larvae were examined after this reheating ; one
exhibited definite movements on a warm stage. The others were more
or less tightly coiled and presumably still alive. Thirteen minutes later
the larvae remaining in the beaker were heated a third time, the tem-
perature being raised rapidly (in 3 minutes) from 43° to 55°. Thirty-
nine larvae were examined; all were motionless and failed to react to heat,
evidently dead.
Experiment 5 (November 17, 1914). — Decapsuled larvae from the same
source as those used in Experiment 4 were heated in the same manner
from 1 6° to 54° C, 7^2 minutes being required for raising the temperature.
Twenty-three lar\^ae were examined after heating and all were found to
remain inactive on a warm stage. The remainder of the larvae in the
beaker were left on the laboratory table until the following day when 42
of them were examined on a warm stage heated to 45°. Most of these
were inactive but more or less tightly coiled. Thirty-five others were
placed on a warm stage heated to 61 °. Six of these exhibited convulsive
movements before they succumbed to the heat, the others showing no
response to stimulation.
Experiment 6 (November 17, 191 4). — Decapsuled trichinae, isolated
by artificial digestion from a mixture of meat from six trichinous hogs,
were heated in a beaker of water over a hot water bath to a temperature of
53.4° C. Some of them showed signs of life when examined on a warm stage.
The beaker was reheated to 55°. Fifty larva were then examined on a
warm stage and all were found to be dead.
122501°— 19 3
2IO Journal of Agricultural Research voi. xvii, no. s
Experiment 7 (December 19, 191 4). — Decapsuled larvae, isolated by
artificial digestion from meat of a trichinous hog, were heated in 0.6 per
cent salt solution in a corked bottle over a water bath. The temperature,
determined by a thermometer inserted through the cork, rose from 24.4°
to 56.7° C. in 44 minutes and remained at this maximum for 30 seconds,
after which the bottle was allowed to cool, the temperature dropping to
34.4° in 38 minutes. Three hundred and sixty-five of the larvse were
then examined on a warm stage and all were found to be inactive. As
a control on the results of this experiment 22 unheated larvae from the
same source were examined on a warm stage; 4 were inactive, 18 active.
Experiment 8 (April 6, 191 5). — Decapsuled trichinae, isolated by
artificial digestion from a mixture of meat from six hogs, were kept 7
days in 0.6 per cent salt solution at ordinary room temperature. Some
were then heated in a beaker of the salt solution, constantly stirred, over
a water bath. The temperature rose from 20° to 54° C. in 7 minutes, and
remained at this maximum for 30 seconds, after which the beaker was
allowed to cool. Examination of some of the larvae from the beaker
showed that most of them were more or less uncoiled, but some were
tightly coiled and practically normal in appearance. The beaker was
kept until the following day at ordinary room temperature and the con-
tents again examined. The great majority of the worms were still alive,
but most of them were not tightly coiled.
Another lot of lar\^ae from the same source was heated in a similar man-
ner but more slowly, the temperature rising from 23° to 54.8° C. in 56
minutes, remaining at 54.8° for i minute, after which the beaker was
allowed to cool. Four hundred and seventy larvae were examined; all
were uncoiled, and their protoplasm was rather dull in appearance. The
beaker was kept at room temperature until the following day, when
examination of 200 larvae showed that all were dead.
Subsequent experiments on the effects of heat on decapsuled larvae
were performed by the junior writer.
Experiment 9. — Decapsuled trichinae in a physiological salt solution
were placed in a test tube and a thermometer immersed in the solution.
The test tube was placed in a beaker of water, which was heated rapidly
until the thermometer registered 55° C. This temperature was attained
in four minutes. The contents of the test tube were immediately trans-
ferred to a stender dish and allowed to cool. The larvae were then exam-
ined. Nearly all were unaffected. A few days later this experiment
was repeated, increasing the time of heating to about eight minutes.
Similar results were obtained.
Aug. 15, 1919
Effects of Heat on TrichincE
211
The results of other experiments with various lots of decapsuled lar\^ae
are shown in the followinsf table :
Table I. — Effect of various temperatures on decapsuled larvcp
Maximum tempera-
ture.
Time required to reach maxi-
mum temperature.
Results.
''C.
Not recorded
Some alive.
XIT,
t;o minutes
Nearly all alive.
Sf:)me alive.
Not recorded
54
:;4.6
. . .do
Do.
42 minutes
All dead.
54 minutes
Do.
14.8
Not recorded
A few showing sluggish movements.
None active.
cc
do
^c
do
Do.
cc
77 minutes
All dead.
cc
60 minutes
All expanded.
Do
cc
65 minutes ....
c:r
•^7 minutes . . .
All dead.
c6
Not recorded
Do
=6
52 minutes. .. .
Do
s6
83 minutes. .
Do
Experiment 10. — A 0.6 per cent salt solution was heated to 56° C.
At this point some decapsuled larvae were spurted into the solution from
a capillary pipette. The temperature dropped from 56° to 55° in 75
seconds, and the contents of the vessel were then emptied into a shallow
dish and examined. Of 25 larvae, 14 were uncoiled and 1 1 tightly coiled.
The same experiment was repeated. Of 21 larvae, only 3 were com-
pletely uncoiled. In another test the larvae were spurted into the solu-
tion at 55° after which the temperature was allowed to drop to 54°,
which required 85 seconds. On examination following transfer to a shal-
low dish, only 3 out of 18 larvae were found to be completely uncoiled.
In order to control the results of direct examination of decapsuled
larvae after heating, the junior writer in two instances fed some of the
larvae to rats. Thus larvae heated rapidly to 55° C. in Experiment 9
were fed to two rats, which when killed at the end of a month were found
to be moderately infected, a result in agreement with the results of direct
examination of the larvae. In another case — one of the experiments
summarized in Table I — larvae heated gradually for 60 minutes to 55°
were fed to two rats, which were found free from trichinae a month later.
Another rat fed unheated decapsuled larvae from the same source became
infected.
From the foregoing experiments it is evident that decapsuled trichina
lar\^ae are killed by a temperature of 55° C, provided this temperature is
gradually attained. Many may be killed by lower temperatures, but the
results of heating to temperatures lower than 55° are uncertain. It is
also apparent that a momentary exposure to a temperature of 55° is not
sufficient to destroy the vitahty of decapsuled larvae, as is shown by the
results of Experiments 1, 9, and 10.
212 Journal of Agricultural Research voi. xvii, no. 5
EXPERIMENTS WITH ENCYSTED LARV^
The experiments on decapsuled larvae were supplemented by experi-
ments on encysted larvae in their natural location in pieces of infested
muscle, the earlier of these .experiments being made by the senior writer,
the later, as noted, by the junior writer.
Experiment ii (March 31, 191 3). — Small pieces of meat from a tri-
chinous rat were placed in a beaker of water (about 500 cc.) in a constant-
temperature oven. The temperature of the water increased from an
initial temperature of 18.4° to 48.4° C. in i hour and 10 minutes, at
which time a piece of the meat was removed. Ten minutes later, when
another piece was removed, the temperature had reached 51°. Eleven
minutes after this at a temperature of 52.8° another piece was removed.
After another period of 15 minutes, when the temperature had reached
55°, another piece was removed. Thirty-seven minutes later, when the
thermometer registered 59.8°, another piece of meat was removed. A
few larv^ae were isolated by dissection from these various pieces of meat
and examined under the microscope. The larvae from the pieces heated
to 48.4° and 51° were alive and active. One out of four larvae from the
piece heated to 52.8° showed slight movements; the others were inactive.
Those from the pieces heated to 55° and 59.8° were inactive when exam-
ined. The results of direct examination of the larvae were checked by
feeding the various pieces of meat to guinea pigs. The guinea pigs fed
with the meat which had been heated to 48.4° and 51° became heavily
infected; those fed the pieces heated to 52.8°, 55°, and 59.8^ remained
free from trichinas.
Experiment 12 (April i, 191 3). — Several small pieces of rat muscle
were placed in a vessel containing 500 cc. of water and heated in an
oven from an initial temperature of 16° to a temperature that reached
55° C. at the end of two hours. Pieces of meat were removed at tem-
peratures of 51.2°, 52.2°, 53°, and 55°. A few larvae from each piece
of meat thus removed were isolated and examined directly on a warm
stage. Samples from these pieces of meat were also fed to guinea pigs,
which were killed about a month after feeding. The direct examination
of the larvae on a warm stage showed that, with the exception of those
from the meat heated to 55°, the majority were alive and responded to
thermal stimulation. Those heated to 55° were loosely coiled and did
not become active on the warm stage.
The results of the feeding experiments were as follows : The guinea pig
that was fed meat heated to 51.2^0. was killed seven da3^s after feeding
because it became sick. The muscles were negative, but one pregnant
female trichina was found in the intestine. The guinea pig that was fed
meat heated to 52.2° was killed about five weeks after feeding, and only
one encysted larva was found in the diaphragm. No parasites were
found in the intercostal muscles. The guinea pig that was fed meat
Aug. 15, I9I9 Effects of Heat on TrichincE 213
heated to 53° was killed five weeks after feeding and was free from
parasites. The meat heated to 55° also failed to infect two guinea pigs.
Experiment 13 (April i, 191 3). — Small pieces of meat from a trichi-
nous rat were heated as in the previous experiment; but an open flame
was used instead of an oven and the temperature was allowed to go up
very rapidly, the water in the beaker meanwhile being stirred con-
stantly. Meat heated from 27.8° to 53° C. in 3J^ minutes was fed to a
guinea pig and resulted in a mild infection. Meat heated from 27.8° to
52° in 3 minutes and from 20° to 49.2° in 6 minutes when fed to guinea
pigs produced h£avy infections.
Experiment 14 (April 3 and 4, 1913). — A small piece of meat from a
trichinous rat was heated in a beaker of water which was constantly
stirred. The temperature rose from 17° to 53° C. in 13 minutes and
remained between 53° and 53.6° for 2 minutes. One larv^a afterwards
isolated by dissection was inactive except at the anterior end which
moved slightly; another was active, though the appearance of its proto-
plasm was somewhat altered.
Another piece of meat was similarly heated from about 20° to 54° C. in
about 10 minutes. Larx^ae isolated by dissection were alive and active.
Another piece was similarly heated from 28° to 53° in 11 minutes and
remained in the water another minute, during which time the tempera-
ture rose to a maximum of 53.8°. Larv^as isolated by dissection v^'ere
alive and active. Two pieces were heated from 28° to 55° in 13 minutes.
One piece was held at a temperature of 55° for i minute, the other
piece at the same temperature for 2 minutes. Trichinae isolated by dis-
section from these pieces were inactive. Another piece of meat from the
same rat was heated from 30° to 54° in 5 minutes and held at a tem-
perature of 54° to 54.8° for I minute. Larv^ae isolated by dissection were
found to be inactive.
Experiment 15 (April 9, 191 3). — Small pieces of meat from two
trichinous rats were tied in a cloth around the bulb of a thermometer,
which was immersed in a beaker of water and heated. The temperature
was held at 54.6° to 54.8° C. for five minutes. Ten larvae were after-
wards isolated by dissection. All were inactive except one, which showed
a very slight movement of its anterior end.
Experiment 16 (May 16 and 19, 1914). — Portions of the diaphragm
of a trichinous rat were heated in a beaker of water stirred constantly
over a water bath. Trichinae were dissected out cf the meat after heating
and examined under the microscope at room temperature. A portion was
heated from 24° to 54° C. in four minutes. Four larv'se examined; i inac-
tive; 3 active, sluggish. Another portion was heated from 24° to 53° in
6 minutes. Ten lar\'£e examined; all active. Another portion was
heated from 23° to 54° in 5 minutes. Twelve larvae examined ; 3 inactive ;
9 active but very sluggish; appearance of protoplasm abnormal.
214 Journal of Agricultural Research voi. xvii, no. ?
In the following tests portions of the diaphragm of another rat were
heated. A portion was heated from 24° to 54° C. in 5/^ minutes. Ten
larvae examined; 9 inactive; i active, very sluggish. A portion was
heated from 24° to 52° in ^}4 minutes. Ten larvae examined; all active,
lively. A portion was heated from 24° to 58° in 3X minutes. Ten
larvae examined; all inactive. A portion was heated from 26° to 53°
in 3>2 minutes. Five larvae examined; all active but not very lively.
A portion was heated from 26° to 55° in 4 minutes. Twenty-three
larv^ae examined; 21 inactive; 2 active, very sluggish. A portion was
heated from 24° to 52.6° in 9 minutes. Twelve larvae examined; 2 in-
active; 10 active, but very sluggish ; appearance of protoplasm abnormal.
A portion was heated from 23° to 52.9° in 2% minutes. Eight larvae
examined; all lively. A portion was heated from 22° to 52° in 3^4 min-
utes. Twenty-four larv^ae examined; all lively.
Experiment 17 (May 20, 191 4). — Portions of the diaphragm of a third
rat were heated as in Experiment 16, but more gradually. Examination
was made as in Experiment 16. A portion was heated from 26° to 53° C.
in 12K minutes and cooled to 48.8° in 5 minutes. Sixteen larv^ae exam-
ined; all active, but sluggish; appearance of protoplasm duller than nor-
mal. A portion was heated from 23.2° to 52° in 14 minutes and cooled
to 46° in 7 minutes. Thirteen larvae examined; all active, fairly lively
but not as vigorous as unheated larvae ; no conspicuous change in appear-
ance of protoplasm ; larvae not coiled as tightly as normal larvae. A por-
tion was heated from 23° to 55° in 16 minutes and cooled to 50° in 5
minutes. Fifteen larvae examined; all inactive; protoplasm dull and
dead in appearance. A portion was heated from 37° to 54° in 9 minutes
and cooled to 49.4° in 6 minutes. Twenty-three larvae examined; all
active but very sluggish; protoplasm dull and dead in appearance. A
portion was heated from 27° to 54° in ii^ minutes and cooled to 49°
in 5 minutes. Twenty-four larvae examined; 16 inactive; 8 active but
very sluggish; protoplasm dull and dead in appearance.
Experiments on encysted trichinae were made by the junior writer as
follows :
Experiment 18. — Small pieces of meat from a rat killed one month
after infection with trichinae were heated in a physiological salt solution to
52°, 53°, 54°, and 55° C, respectively, and then allowed to stand in a
refrigerator for two days. The larvae were then freed from their capsules
by teasing out the meat, and examined directly. Those heated to 52°
were still tightly coiled, although a number of loosely coiled larvae were
also seen. Most of the larvae heated to 53° were uncoiled, but a few
were coiled normally. Those heated to 54° and 55° were entirely
uncoiled, dull in appearance, and failed to become active when warmed.
Experiment 19. — Larger pieces of meat from a trichinous hog were
heated as in the experiment just described, kept in a refrigerator for
two days, and then fed to mice. The post-mortem examinations yielded
nesrative results in all cases.
Aug. IS. 1919 Effects of Heat on TrichitKE 215
The results obtained from the experiments in which pieces of trichin-
ous meat were heated agree with the results of those in which the larv^ae
were first freed from their cysts by artificial digestion and then heated in
water or physiological salt solution. The larvae are killed if the meat is
gradually heated to a temperature of 55° C, though some may escape if
the temperature rises rapidly to 55° and soon falls again. They may
survive a temperature of 54°; but meat which has been exposed to a
temperature of about 53°, gradually attained, is likely to be non-
infective.
It may be concluded that meat which has been heated so that the tem-
perature throughout reaches 55° C. (131° F.) will be innocuous so far as
concerns the possibility that persons eating such meat will become
infected with trichinae, inasmuch as under ordinary conditions of cooking
the rise of temperature will be gradual enough to insure the destruction
of the parasites if the temperature of the meat actually reaches 55° C.
or higher. Under the regulations of the Bureau of Animal Industry the
minimum temperature that must be attained throughout all portions
of pieces of pork or products containing pork that are cooked in estab-
lishments operating under federal meat inspection has been fixed
somewhat higher than 55° C, namely 137° F. (58.33° C), which allows
a margin of safety of several degrees above the temperature that has
been shown by our investigations to be fatal to trichinae.
THE EFFECTS UPON TRICHINA OF CONTINUED EXPOSURE TO HEAT
AT TEMPERATURES BELOW THE THERMAE DEATH POINT
It has been shown that trichina larvae are killed by brief exposure to a
temperature of 55° C, gradually attained; and since they will not after-
wards resume their activity when thus heated, this temperature may be
considered the thermal death point. The vitality of the larvae may be
destroyed also by exposure to lower temperatures, provided the appli-
cation of heat is long enough continued. In the former case it may be
assumed that death results from irreversible coagulations of the proto-
plasm, in the latter case either as the result of coagulation changes which
become irreversible if the heat acts for a sufficient period, or as the result
of exhaustion following excessive activity to which the larvae are stimu-
lated by heat. We may, therefore, distinguish three ranges of lethal
temperatures: The highest, in which death comes quickly from rapid
and irreversible coagulations of the protoplasm; an intermediate range,
in which death results probably from somewhat similar coagulation
changes, changes, however, from which the parasites may more or less
completely recover if the temperature is lowered before death occurs;
and the lowest range, in which death is apparently brought about by
exhaustion from increased activity.
2i6 Journal of Agricultural Research voi. xvii, No. s
The following experiments to determine the effects of the continued
exposure of decapsuled larvae to temperatures below 53° C. were carried
out by the junior writer. The larvae in 0.7 per cent salt solution or in
Ringer's solution were first heated to a given temperature and then
placed in an incubator at the same temperature for a given period.
When taken out of the incubator the larvae were kept at room tempera-
ture at least an hour before they were examined.
Experiment 20. — In one test the larvje were all dead after exposure
for three hours to a temperature of 48° C, but generally an exposure to
a temperature of 48° for less then four hours failed to destroy their
vitality. In every case, however, after they were heated four hours at a
temperature of 48° they were all uncoiled, having assumed the shape of
the figure 6; and they failed to react to heat stimulation.
Experiment 21. — When exposed to a temperature of 49° C. nearly
one-half the larvae in one lot were still alive at the end of two hours.
Another lot from a different host animal succumbed to a similar treat-
ment, but in no case did a briefer exposure to 49° prove effective. When
subjected to 49° for 3X hours all the larvae became completely uncoiled,
rigid, and insensitive to thermal stimuli.
Experiment 22. — At a temperature of 50° to 50.6° C, the vitality of
the larvae was completely destroyed after an exposure of i hour and 20
minutes. At a constant temperature of 50° an exposure of i K hours
proved fatal.
Experiment 23. — An exposure of one hour to a temperature of 52° C.
was sufficient to destroy the vitality of decapsuled larvae.
From the foregoing experiments it is evident that decapsuled trichina
larvae die in a comparatively short time when exposed to temperatures
in the neighborhood of 50° C. and that the time required for their de-
struction increases as the temperature is lowered. If the results of
these experiments are considered in connection with the question of the
length of time that decapsuled larvae survive at temperatures ranging
below 40°, already discussed in this article, it may be concluded that
between limits at which the larvae become altogether quiescent because
of the effects of heat on the one hand and of cold on the other their
longevity varies inversely with the temperature. It would, however,
not be safe to conclude from the experiments just described that exposure
of trichina larvae to the temperatures given for the stated periods of
time would be sufficient in all cases to destroy the vitality of the para-
sites. It is not improbable that in these experiments the larvae had
already become som.ewhat exhausted as a result of abnormal activity
during the process of artificial digestion, and furthermore it is possible
that different lots of trichinae vary considerably with respect to their
store of vitality. The following experiments by the senior writer show
that the vitality of encysted trichinae as well as that of decapsuled
Aug. IS. 1919 Effects of Heat on TrichincE 217
trichinae may be destroyed by continued heating at temperatures lower
than that which kills on brief exposure. Like the experiments with
the decapsuled larvae, however, they are not sufificiently extensive to
allow definite conclusions to be drawn as to the periods of time necessary
to insure the destruction of trichinae exposed to temperatures lower than
the thermal death point.
Experiment 24 (April 7, 191 3). —A small piece of the diaphragm of
the same rat which supplied the meat used in Experiment 14 was tied
in a cloth around the bulb of a thermometer, which was immersed in a
beaker of water heated to about 50° C. and the entire apparatus placed
in a constant-temperature oven. The temperature, as indicated by the
thermometer, varied from 50.2° to 51.6° during the two hours of heating
the meat. Larvae isolated from the meat by dissection were dead.
Experiment 25 (April 9, 191 3). — Two small pieces of meat from the
same rat used in Experiments 14 and 24 were tied in cloths around the
bulbs of two thermometers and heated in a beaker of water as in Experi-
ment 24. During the experiment the temperature, as indicated by the
thermometers, varied between 49.6° and 50° C. One piece was removed
after an hour's exposure. Tvv^o larvae isolated from the meat by dissec-
tion were alive, but rather sluggish. The other piece was removed after
an exposure of i^ hours. Two larvae were examined, one of which
was dead, the other alive, but rather sluggish. Two guinea pigs were
fed with the meat, but neither became infected. Another piece of meat
from the same rat was similarly heated for one hour at a temperature of
50.1° to 50.4° C. A larva isolated from the meat after heating was alive
and active. Another piece was similarly heated for i}^ hours at 50°.
Five larvae were isolated from the meat and examined. Four were cer-
tainly dead, the other inactive, but with protoplasm less changed than
that of the others.
Experiment 26 (August 31, 191 4). — Finely chopped meat from a
trichinous rat was placed in water in a flask, which was kept 2 1 hours in
an oven maintained at a temperature of 49° to 52° C. The temperature
of the water during this time varied from 48.8° to 51.4°. Four larvae
dissected out of the meat after heating were dead. The meat was fed
to two rats, both of which remained free from trichinae. Some finely
chopped meat from the same rat was heated 2 1 hours in a covered Petri
dish in the same oven at a temperature of 49° to 52°. Five larvae dis-
sected out of the meat after heating were dead. Two rats to which the
meat was fed remained free from infection.
Experiment 27 (September 3, 191 4). — Finely chopped meat from a
trichinous hog was heated in a closed jar in a constant -temperature oven
for 19 hours. The temperature of the meat during this time varied
between 47.8° and 48.4° C. Twenty-five trichinae were dissected out of
the meat after heating and all found to be dead.
2i8 Journal o[ Agricultural Research voi. xvii. no. s
Experiment 28 (September 8. 1914). — The eviscerated carcass of a
trichinous rat was heated 17 hours in an oven at a temperature of 48°
to 50° C. On removal from the oven the carcass had a bad odor; the
upper surface was dried, the lower still moist. Twenty trichinae were
dissected out of the meat after heating and all found to be dead. Meat
from the carcass was then fed to two rats, one of which remained free
from trichinae, while the other was found moderately infected when
killed three months after feeding.
Kxl'ERiMENT 29 (September 19, 191 4). — Finely chopped meat from a
trichinous rat was heated 5 hours in an oven at a temperature of 48° to
49° C. A few trichinae afterward dissected out of the meat were shrunken,
but their protoplasm was bright in appearance. After being soaked in
water for 30 minutes some of the lar\-ae became lively, and 2 days later
the remainder of the isolated larvae kept in water at room temperature
had also become actixe and normal in appearance. Some of the same
meat was left in the oven until September 21, and thus exposed for 48
hours to a temperature of 48° to 49° C. It was hard and dry. Trichinae
isolated from the meat by dissection after it had been softened by soaking
were very clear, pale, motionless, and apparently dead.
Additional data regarding the effects of the continued action of tem-
peratures below the thermal death point were obtained by the junior
writer. In these experiments, which are summarized in tabular form
(Table II), the method of procedure was as follows: Meat from trichinous
hogs was finely chopped by passing it through a meat chopper several
times. A bottle with a capacity of about 200 cc. was half filled with
the meat. Through a perforation in the cork a thermometer was inserted
into the bottle and the top of the cork then paraffined. The bottle of
meat was placed in a constant-temperature oven and the temperatures
read on the thermometer in the bottle.
Inasmuch as the meat before being placed in the oven was kept in a
refrigerator at a temperature of S° to 10° C, a considerable period was
required to bring its temperature near that of the oven. In nearly all
the experiments shown in Table II the meat was in the oven about 2
hours before the first reading of the thermometer, given in the table as
the minimum temperature, was made. Between the first and the final
reading there was a slight fluctuation of the temperature but nearly
always between the limits recorded in the table.
At the end of each experiment a portion of the meat was artificiallv
digested in the usual way and the condition of the larvae noted. As a
control on the microscopic findings in each experiment two rats were
fed portions of the meat, being given an average of about 10 gm. each.
Unless they died earlier the test animals were killed about a month after
feeding. The following table gives the record of i o experiments :
Aug. 15. 1919
Effects of Heat on Trichince
219
Tabi.Iv II. — Effects of continued action of temperatures below thermal death point <
encysted trichince
Temperature.
Appearance of larvae after artificial digestion.
Ap]>arently dead
Profoundly disorganized
Showing evidence of having been partially
digested :
Uncoiled ; evidently dead
Apparently dead
Probably dead
Uncoiled and pale
....do
....do
Results of feeding
experiments.
Coiled Negativ
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
From a practical standpoint the results of the experiments on the
effects of continuous heating at temperatures below the thermal death
point of trichinae are of comparatively^ little importance so far as con-
cerns the destruction of the vitality of trichinae in fresh pork by cooking.
Obviously, as compared to cooking at a higher temperature for a short
time, there would be no advantage in subjecting meat to a lower tem-
perature, which would require a very great lengthening of 'the period of
heating. If for no other reason, the probable spoiling of the meat would
preclude the use of such a method of destroying the vitality of the
parasites. In connection with the preparation of certain kinds of cured
pork products, however, the fact that heating at low temperatures for
considerable periods of time is destructive to the vitality of trichinae has
been put to practical use. In this case there is also another factor
which comes into play — namely, the destructive action of salt in hyper-
tonic percentages, which increases greatly as the temperature increases.
The question of the destruction of trichinae in cured pork by heating at
low temperatures will be discussed in another paper.
CONCLUSIONS
The vitality of the larvae of Trichindla- spiralis is quickly destroyed
by exposure of the parasites to a temperature of 55° C, gradually
attained, the result apparently of irreversible coagulation changes in the
protoplasm. This temperature may be considered the thermal death
point.
Trichina larvae exposed to temperatures slightly below 55° C. for short
periods of time may recover from this exposure; but they die if exposed
for longer periods, recovery or death depending apparently upon whether
or not beginning coagulation of the protoplasm has proceeded bevond
a stage from v/hich a return to normal mav occur.
220 Journal of Agricultural Research voi. xvii, no. 5
Exposed to temperatures in the neighborhood of 50° C, trichina larvae
die if the application of heat is sufficiently long continued, apparently
as a result of exhaustion following excessive activity to which they are
stimulated by the heat.
The longevity of trichina larvae freed from their cysts by artificial
digestion and kept at temperatures ranging between limits at which they
become quiescent from the effects of heat and cold, respectively, varies
inversely with the temperature.
Methods of destroying trichinae b}^ heating at temperatures below the
thermal death point, which may be utilized in connection with the
preparation of certain kinds of cured pork products, appear not to be
applicable in the case of fresh pork.
Upon the basis of the results of experiments recorded in this paper the
Bureau of Animal Industry has selected a temperature of 137° F.
(58-33° C.) as the minimum temperature to which pork and products
containing pork are required to be heated when cooked in establishments
operating under federal meat inspection.^ This temperature is several
degrees above the thermal death point of trichina larvae, thus providing
a certain margin of safety.
LITERATURE CITED
(i) Fiedler, A.
1864. BEITRAGE ZUR ENTWICKLUNGSGESCHICHTE DER TRICHINEN, NEBST
EINIGEN MITTHEn.UNGEN UBER DIE EINWIRKUNG EINZELNER MEDI-
CAMENTE u. ANDERER AGENTIEN AUF DIESELBEX. In Arch. Heilk.,
Jahrg. 5, p. 1-29.
(2)
1864. WEITERE MITTHEILUNGEN UBER TRICHINEN. Ill Arch. Heilk., Jahrg. 5
p. 466-472, 511-520.
(3) Fjord, N. J., and Krabbe, H.
1868. EFFECTS OP HIGH TEMPERATURE ON TRICHINA SPIRALIS. (Translntion.)
In Veterinarian, London, v. 41 (s. 4, v. 14), p. 323-328.
(4) Haubner, Karl.
1864. UEBER die TRICHINEN, MIT BESONDERER BERUCKSICHTIGUNG DER SCHUTZ-
MITTEL GEGEN DIE TRICHINENKRAXKHEIT BEIM MENSCHEN. In Mag.
Gesam. Thierheilk., Jahrg. 30, Stuck 2, p. 129-176, pi. 2.
(5) KucHENMEiSTER, and Leisering.
1863. VERSUCHE MIT TRICHINEN. In Ber. Veterinarw. Konigr. Si^chs., 1862,
p. 114-120.
(6) Leuckart, Rudolf.
1886. THE PAR.\SITES of man, AND THE DISEASES WHICH PROCEED FROM THEM.
Translated from the German, with the cooperation of the author, by
William E. Hoyle. xxvi, 771 p., 404 fig. Edinburgh.
(7) Perroncito, E.
1877. DIE TRICHINA SPIR.'^LIS IN it ALIEN. MITTHEH.UNG UBER EINEN FALL VON
TRICHINOSE IM INTERMUSKULOSEN BI.NDEGEWEBE EINES JAGDHUNDES.
(Translation.) In Ztschr. Veterinarwiss., Jahrg. 5, Heft4/5, P- 200-203.
' This requirement has reference to the temperature actually reached in the interior of the meat and
not merely to that of the water or oven in which it is cooked. It should also be understood that when
meat is cooked for purposes of sterilization because of conditions other than trichinosis a higher tempera-
ture is necessary than that sufficient to destroy trichina:.
Aug. 15, 1919 Effects of Heat on Trichincr 221
(8) PlANA, G. P.
1887. STUDIO SULLA TRICHINA SPIRALE R SULLA TRICHiXOSI. In Clin. \'et.,
ann. 10, no. i, p. 17-28, fig. 1-24; no. 2, p. O9-72; no. 3/4, p. 108-117,
fig- 25-37; no. s/6, p. 197-200; no. 7, p. 304-312, fig. 38-42; no. 8/9,
P- 383-390; ^'3- 10. P- 438-442; no. II, p. 502-505.
(9) Ransom, B. H.
1914. TRICHINOSIS. In U. S. Dept. Agr. Ann. Rpts., 1913, p. 101-102.
(10)
1915. TRICHINOSIS. In Rpt. i8th Ann. Meeting, U. S. Live Stock Sanit.
Assoc, p. 147-165.
(11)
1916. RFFECTS OF REFRIGERATION UPON THE L.A.RV^ OF TRICHINELLA SPIRALIS.
hi Jour. Agr. Research, v. 5, no. 18, p. 819-854.
(12) RoDET, Henr>^
1866. DE LA TRiCHiNE ET DE LA TRiCHiNOSE- ed. 2, 50 p., I pi. Paris.
(13) Vallin.
1881. de la rjgsistance des trichines a la chaleur et de la temperature
CENTRALS DES viANDES PREPAREES. In Bul. Acad. Med., Paris, ann.
45 (s. 2, t. 10), no. -8, p. 264-265.
(14) Winn, Henry Newton.
1915. EFFECT OP HEAT AND COLD UPON THE LARV.E OF TRICHINELLA SPIRALIS.
In Wis. Med. Jour., v. 14, no. 2, p. 59-60.
EFFECT OF REMOVING THE PULP FROM CAMPHOR
SEED ON GERMINATION AND THE SUBSE-
QUENT GROWTH OF THE SEEDLINGS
By G. A. Russell
Expert, Office of Drug, Poisonous, and Oil Plant Investigations, Bureau of Plant In-
dustry, United States Department of Agriculture
INTRODUCTION
Heretofore but slight attention has been paid to the germination of
camphor seed. The few statements on this subject which occur in the
literature refer only to the percentage of seeds germinating under condi-
tions existing at the place of experimentation, and all the recorded
results indicate a uniformly low germination. Likewise in Florida, pre-
vious to the experiments recorded in this article, the germination of
camphor seed has been extremely low.
In commercial plantings in Florida, in which unpulped seeds have been
planted with a modified cotton-dropping machine, the average number
of seedlings brought to transplanting age on i acre of seed bed has
been approximately 20,000. To plant an acre of seed bed requires 3
bushels of camphor seed, or approximately 200,000 seeds. The germina-
tion on a commercial scale, therefore, has averaged only about 10 per
cent, which corresponds closely with the results obtained in various
foreign countries. As a consequence of this low germination there has
been no considerable extension of large plantings because of the limited
number of seedlings available each year.
EXPERIMENTS IN 1916-17
In the fall of 19 16 it was decided to make germination tests of camphor
seed to determine if possible the cause or causes of the low germination
obtained both experimentally and commercially. Accordingly seed was
gathered from six individual trees growing in the vicinity of Orlando, Fla.
Seed from one parent tree. A, was selected from a row grown for shade
and ornamental purposes. This tree was 20 years old and a typical
representative of the camphor trees in Florida from which seed is gathered
for commercial planting. The conditions under w^hich the various lots
of seed were collected and the treatment of each before planting are
shown in Table I.
Journal of Agricultural Research, Vol. XVII, No. s
Washington, T). C. -A-"g- ^S. i9i9
sjj Key No. G-176
(223)
224 Journal of Agricultural Research voi. xvii. No. s
Table I. — Condition and treatment of camphor seed selected for germination tests con-
ducted in igi6-ij
PARENT TREE A
Experimental row No.
Treatmeat of seed just previous to planting.
6
As they came from the parent tree.
Pulp removed.
Pulp removed. Soaked in water at 2 5°C. for K hour.
Pulp removed. Soaked in water at 5o°C. for % hour.
Picked up from the ground. Pulp removed.
Picked up from the ground as they fell from the parent tree.
As they came from the parent tree. Gathered after a severe
freeze.
Pulp removed. Gathered after a severe freeze.
As they came from the parent tree. Gathered after a severe
freeze and soaked in water for i8 hours.
Pulp removed. Gathered after a severe freeze and soaked in
water for i8 hours.
7
JNorthhalf
^°\vSouth half
T •;
14. -
I C
i6
North half
171
South half
The first experiments were conducted in the winter of 191 6-17. From
some previous experience it was found that by removing the pulp from
around the seed, germination was hastened if not materially increased.
It was decided, therefore, to give special attention to the effect of removal
of the pulp from the seed, since if it proved to be a decided aid to germi-
nation, the adoption of this method of treatment by commercial growers
would be of distinct advantage. The remainder of the seed from the
selected trees was pulped and planted. The percentage of germination
was high, but the results are not recorded here since no data were
secured on unpulped seed from the same trees.
The seed bed had been well prepared one week previous to the plant-
ing of the first seed, and a quantity of dry velvet-bean vines had been
turned under. Drills from i ^ to 2 inches deep were opened with a hoe
and the seed carefully hand-planted at intervals of 2 inches. The soil
was placed back in the drill and very firmly packed. At the time of
planting the soil was moist and in good condition, but later in the spring
after the seedlings were several inches high it became necessary to
water the bed three times in order to maintain the moisture content.
On May 7 and August i, 191 7, the seed bed was fertilized with goat
manure analyzing: Moisture 20 per cent, ammonia 1.5 per cent, and
potash (as KjO) 2.5 per cent. One hundred pounds were used at each
application, which was at the rate of i ton per acre. The seedlings were
well cared for by cultivating and hoeing. Table II gives the date of
planting, rate of germination, and total percentage of germination in
the 1 91 6-1 7 trial of seed from parent tree A.
Aug. IS, 1919 Effect of Removing Pulp from Camphor Seed
225
0\
N
8
*>.
&;
0
z
w
«
«
s
<
S
04
<^
0^
>.
lg
5 e O^ O O M
Q, I V3 O vO
i^j Lo -^ rj CO t>
. M 00 ^01
O w LO 10 rJ-OO
ro >o Tj-o6 CO CI
w 00 M 00
Q;a
O vO 10 M O 01
■^ 10 "^ d rj 10
On cs <N O
lO w M
M >0 >JO cs COMO
CO lO Tj- O H 1^
M CO W
to cooO
Ci
00
0\ PI \0 \o
0)
a!
goo
■* 0 10
10 10 Tj-
■* 10 M
M M
CO
0
CO
i^ CO t^ CO
00 ^ 0) M
10 M M
0\ C» O 00 o
M CO H
^.
^
"
0
^
C<
VO
VO
M
0 VO
D
Q,
5;
00
10 M t^
10 10 ■^
10 10
0
ro
VO
CO
M VO M 00
00 VO CO w
VO M H
t^ VO C* vO COvO
Tj- VO o r^
M CO >-*
a,s
.00 " 00 vO
a, t
0. 5 fe O 3
g 11 a
oi
--3
c
•2
-6
^•s
Qrfs « -e
•S c c c
,?' as cfl oj
T' 1^ u, vi Vh Vh ;-!
O O
•* o
O O VO VO
M
M
0 0
c
6
-a
VO
>
0
6 c
-co
4
fe
fe
esa
BBB
3 3 3
§§§
3 S K
01 CO 01
aea
.S'H.S
.2 a--
"3— 3
9 2-9
d S S
^ S-3
fl o S
V r^ N
N « O
.5 tni-J2
I'S fe fc'3
B .0 u -a
122501 — IC
226 Journal of Agricultural Research voi.xvn.No.s
No germination was secured when the seeds were artificially dried at
a temperature of approximately 55° C. Seeds that were allowed to air-
dry for several weeks in an attic likewise failed to germinate. Soaking
in water at a temperature as high as 50° neither hastened nor impaired
germination. Pulped seed treated with sulphuric acid of 5 per cent
concentration by weight failed to germinate.
At the commercial planting beds so much seed is received in a ferment-
ing condition that it was deemed advisable to ferment one lot of seed
during an extended period in order to ascertain the effect on their germi-
nating power. One thousand seeds fermented for 35 days in a closed
jar failed to show a single case of germination. At the end of this period
the pulp surrounding the seeds had almost entirely decomposed and the
resulting liquid was sufficient to cover practically all the seeds. Ship-
ments of seed for commercial use, however, are seldom enroute longer
than from 8 to 10 days and do not reach such an advanced stage of decom-
position. No marked ill effects due to fermentation have been noted in
the commercial seed beds, which is attributed to the fact that the seed
pulp has not entirely decomposed and that the liquid is constantly
leaching from the barrels and boxes in which the seed is shipped, thus
eliminating any chance for the seed to soak. Moreover, as soon as the
seeds reach the camphor plantation they are spread out to cool and dry
and fermentation ceases.
The results obtained from seed picked up from the ground are of special
interest, such seed being often used in commercial work. Camphor seeds
even when quite ripe do not drop readily from the tree ; and a large per-
centage of the seeds which fall early in the season are defective, since the
fallen unpulped seed showed a germination of only 5.9 per cent as com-
pared with 9.4 per cent germination of seed picked from the tree. How-
ever, these defective seeds when pulped showed a germination of 15.6
per cent as compared with 60. i per cent of pulped seed picked from the
same tree. The seeds picked up from the ground were planted one
month later than those picked from the tree, but they had fallen during
the interim.
The idea that frozen camphor seed will not germinate is widely dis-
seminated throughout Florida. A special experiment was made with
seeds obtained after a relatively hard freeze during which the tempera-
ture fell to 26° F. The results obtained prove beyond doubt that
camphor seed subjected to a freeze will germinate (fig. i). This fact is of
special value since freezing weather is liable to occur at any time during
the late fall months in the camphor-seed producing areas, especially in
those farthest north. The total gennination, however, is decreased,
being approximately 50 per cent of that obtained with unfrosted seeds.
A greatly increased gennination of the seed is secured by removing the
pulp before planting. This increase was found to amount to 539 per cent.
A graphic representation of the increased germination is presented in
Aug. 15. 1919 Effect of Removing Pulp from Camphor Seed
227
figure I , which shows not only
the increased germination
when the pulp is removed
but the variation in germi-
nation of seed secured under
varying conditions. The
seed planted on December i,
1 91 6, which was picked from
parent tree A, was first-class
in every respect. On Janu-
ary 6, i9i7,aquantity of seed
was planted that was picked
up from the ground under the
same tree, and on February
6 frozen seed from this tree
was secured and planted. In
every instance the removal
of the pulp before planting
greatly increased the germi-
nation.
The percentage of germina-
tion of the seed picked up
from the ground is much less
than that of seed picked from
the tree (fig. i ) . This differ-
ence in germination was an-
ticipated and was due in
great part at least to defec-
tive seeds which fell from the
tree, in other words, those
which are considered as
" drops. ' ' However, even the
germination of these "drops"
increased by 16.4 per cent
when the pulp was removed.
Seed picked from the tree af-
ter a severe freeze germinated
remarkably well, especially
when the pulp was removed.
By soaking these frozen seeds
in water at a temperature of
approximately 25° C. for 18
hours a rather remarkable
result was obtained. The
e-^
GO
x5-s5-
so
^s
t
«?o
/•s
/o
^\
^ s.
0.0
^
-c«^
■iJ 1,
?l
^o.ze>r>
I
l^^
r
0 c «
"b
%
\
0
^
%
K
^ ^
^ ^
& <
?" V
>•
^ 1
"^
1
1
d)
1
<b
«j ■
(.
:^ ■
■VJ
9 1
\\
^^
I
o;l
*^l
•J ~ <!>■
4;
<!,
^5 -■
Hj
1
0-?^
"I;^
^
^
0
4.
IT)
.^
0
^
:B
^
1
^
(r
'^ i
to
05 h
Oote_of_ P/anti'n^r
,5l
Fig. I. — Diagram showing percentage of geniiination of
camphor seed secured from parent tree A imder varying
conditions.
228
Journal of Agricultural Research
Vol. XVII, No 0
percentage of germination of unpulped seed was reduced, whereas that of
pulped seed was increased 23 per cent over the germination of the pulped
seed not soaked. The reduction in germination when the unpulped
seeds were soaked can not be attributed to the direct action of the water
alone. Since soaking favors a more rapid decomposition of the pulp
when the seed is placed in the ground, the degree of fermentation reached
may have been sufficient to destroy the vitality of some of the seeds.
In addition to the marked effect on the rapidity of germination which
was noted when the pulp was removed from the seed, it was also noted
that as the planting season advanced germination was more rapid and
much less time \\'as required to reach the point of maximum germination.
Table III. — Rapidity of germination of camphor seed secured from parent tree A under
various conditions
Experimental row
No.
Treatment of seed.
Date of
planting.
Days
required
to reach
maximum
germina-
tion.
6
As they came from the tree
1916.
Dec. I
Dec. 2
...do
161
"J
Pulp removed
120
[North half. . .
Pulp removed. Soaked in water at 25° C. for
'i^ hour
109
109
84
[South half. .
Pulp removed. Soaked in water at 50° C. for
' -> hour
.do.. .
i^
Picked up from the ground. Pulp removed. .
Picked up from the ground. As they fell from
the tree
1917.
Jan. 6
...do
14
98
1;
As they came from the tree. Gathered after
hard freeze
Feb. 10
...do
89
16
Pulp removed. Gathered after hard freeze. . .
49
[Nortli half. . .
[South half. ..
As they came from the tree. Gathered after
severe freeze and soaked in water 18 hours.
Pulp removed. Gathered after severe freeze
and soaked in water 18 hours
...do
...do
89
64
Three special points of interest are brought out in Table III: First,
the time for camphor seed to reach maximum germination; second, the
shortening of this time by removing the pulp before planting; and third,
the decrease in time required for the seed to germinate as the season
advances. These points are more fully illustrated in the graph showing
the time required for camphor seed to reach maximum germination
(fig. 2). In every trial the pulped seed germinated much more quickly
than the unpulped seed, irrespective of the condition at the time of
gathering. As the season advanced and the soil warmed up, germina-
tion naturally took place in a shorter time. But what is of more interest
from the commercial point of view is the fact that seed gathered and
planted early in the fall will remain in the ground in good condition
Aug. 15, 1919 Effect of Removing Pulp from Camphor Seed
229
I I I I I . I l,.t I J I .1 ,1 I I L_I I I I I I I I I I I I I I I I I I I I I
D£C.- :/S/e ^AN. /&//' r£B./3/F
Date of p/onting
Fig. 2. — Graph showing time required lor pulped and unpulptd camphor seed to reach maximum germi-
nation. The seeds were planted at intervals of approximately one month during the winter of 1916-17.
230
Journal of Agricultural Research
Vol. XVir, No. s
Dec. /. /9/e
UAN 6} /9/P'
r£B. /0./3/7
GO
>SS
■SO
-*5J-
o
30
/•s
/o
j
^^
\
1
\
]
1
/
i
/
1
1
•
1
/
1
1
1
•
f
•
/ 1.
b^
I'
^^
»^^ ^
1
-""^
y
,yj^
,--"
»«^*^*«l
^\
o
V
^
0
"^N
^ N
:^N
^ N
ffi^
p
?^
t^
H
k
%
\
^
P€:r/oo
'^or
GA?OWTH
Fig. 3. — Graph showing time required for camphor seed secured from parent tree A at various times and
under various conditions to reach maximum germination. The percentage of germination is also shown.
Aug. 15, 1919 Effect of Removing Pulp from Camphor Seed 231
until favorable germinating weather is reached. Figure 3 shows graphic-
ally the time required to reach maximum germination and the per-
centage of germination of seed secured from parent tree A at various
times and under various conditions. The maximum germination of all
pulped seed was reached by April i, 191 7, although the time of planting
extended over the period from December 2, 191 6, to February 10, 191 7;
whereas the maximum germination of the unpulped seed occurred about
May 10, 1 91 7, the planting period extending over the same period of
time as that of the pulped seed.
The results of the germination experiments of 191 6-1 7 were so pro-
nounced in favor of removal of the pulp from the camphor seed that the
work for 191 7-1 8 was planned to include a more extended comparison
between pulped and unpulped seed.
EXPERIMENTS IN 1917-18
During the first week of November, 191 7, a seed bed was prepared on
soil practically identical with that employed in the experiments of 1916-
17. A quantity of fertilizer made by composting rose-geranium leaves
and stalks — after distilling the volatile oil — was turned under at the
time the seed bed was plowed. This fertilizer material consisted only
of leaves and stalks and, being somewhat intact at the time of applica-
tion, had a tendency to keep the soil from packing, at the same time
supplying some plant food as it decomposed. At the time of planting
the soil was moist, and no subsequent watering of the bed was necessary
throughout the time of the experiments. No fertilizer was applied
during the growing season. The plants were given the usual cultivation
and hoeing. Commercial conditions, with the exception of the applica-
tion of fertilizer, were approximated as closely as possible.
Seed was selected from a row of ornamental camphor trees, and trees
were chosen which bore an abundance of fruit. Camphor seed which
ripens on the tree falls readily into the hand when picked. All the seeds
used in these experiments were fully ripened and easily secured by picking,
care being taken to secure seed from all sides of the tree. Each sample
therefore was representative of the entire yield of the individual tree.
All the seeds were gathered on November 27, 191 7, and planted Novem-
ber 28, 1917. A severe freeze occurred February 2, 1918, but as none
of the seedlings had appeared above ground no damage was done. In
Table lY are given data in reference to the treatment of the seed, rapidity
of germination, and percentage of total germination.
232
Journal of Agricultural Research
Vol. XVII. No. s
Table IV. — Rate and percentage of germination of camphor seed in the experiments of
IQIJ-18
Parent
tree.
A.
B
C.
D
E.
G
H
K
O
P,
No. of row.
, /North half..
ISouth half. .
/North half .
nSouth half. .
/North half .
nSouthhalf..
/North half
"^l South half..
/North half.,
nsouthhalf..
jNorthhaU..
*! South half..
/North half.
nSouthhalf..
JNorth half .
'^l South half..
/North half .
5'1 South half. .
JNorth half .
'°\ South half..
Treatment of
seeds planted.
None
Pulp removed...
Pulp removed...
None
None
Pulp removed . . .
Pulp removed...
None
None
Pulp removed...
Pulp removed...
None
None
Pulp removed...
Pulp removed . . .
None
None
Pulp removed...
Pulp removed...
None
Rapidity and percentage of germination.
191a
Feb. II. Feb. 23. Mar. 18. Apr. 37- July 6. Dec. 30,
No.
16.0
36.5
9.8
36.2
13.0
54-6
28.4
31-4
22.8
30.2
No.
199
3"
15
II
345
362
15
3
218
372'
15
3
301
315
Per
ct.
6
66.
82
No.
223
325
67
22
375
367
40
9
314
38S
42
4
360
336
397 :
409!
31
No.
232
317
117
55
31SJ63
367 73
80 16.
80 16,
326,65
No.
18
232
317
10 1
55
259
377
69
63
310
442
63
33
373
349
31
37
382
418
36
iPer
No.l ct.
18 3.
232I77
3i7i84
loi
55
M.—
per
ct.
o 77
2|23
O II
8,75'
4 75
8 16.
6' 16
o 65.
489
2 16
6 7
674
873
2: 8.
4 8.
480.
695
In this trial the seeds from a total of lo individual trees were tested.
The results obtained by merely pulping the seed before planting were
so favorable that commercial planters adopted the pulping plan when
its merits were brought to their attention. In commercial work the
pulp is removed by rubbing the seeds through a wire screen of the proper
mesh. Many of the pulps are left behind and are swept off the screen.
Those that fall through with the seed cause no inconvenience in plant-
ing, for the seed is spread to dry for about 24 to 48 hours; and during
this time the pulps dry and shrink to such an extent that they readily
pass through the plates of the corn planter which is now used to plant
the pulped seed. Unless the seed is dried before planting the plates
of the planter become clogged, causing an uneven distribution of the
seeds in the row.
TOTAL GERMINATION OF CAMPHOR SEED
The greatly increased germination obtained when the seed is pulped
is remarkable. Figure 4 shows graphically the total germination of
both the pulped and unpulped seed from 10 parent trees. The increased
germination of the pulped over the unpulped seed ranged from 270 per
cent for tree B to 2,466 per cent for tree A, the average increase for the
entire lot of 10 trees being approximately 525 per cent.
Germination was found to be uneven with seed from various parent
trees. Moreover, the ratio between the percentage of germination
of the unpulped and the pulped seed was by no means constant; and no
correlation can be established between the percentage of germination
Aug. 15, 1919 Effect of Removing Pulp from Camphor Seed
233
when the seed is pulped and when it is not. The variation in germina-
tion of 10 individual lots of seeds from as many parent trees is shown
graphically in figure 5. The upper line in each case indicates the germina-
tion of pulped seed, the lower line the germination of unpulped seed.
Under period of growth, a indicates the date of planting, November 28,
1 91 7. The percentage of seeds germinating was determined by counting
the number of seedlings in the beds at stated times, indicated in the
figure as follows: 6 = February 11, 191 8; (- = February 23, 191 8; d =
£0
/o
1
p
r
p
Fig. .
r/?£€ NUMBER
-Diagram showing percentage of total germination of pulped and unpulped camphor seed from
10 parent trees. Black bars represent pulped seed; white bars, unpulped seed.
March 18, 1918; e = April 27, 1918; / = July 6, 1918; ^ = December i,
1918, on which date the seedlings were transplanted. It will be noted
that the pulped seed germinated in much shorter time than the un-
pulped seed and that after the appearance of the first seedlings the
major portion of the gennination took place in a relatively short time.
The apparent falling off in the germination of the pulped seed, as indi-
cated in the graphs (fig. 5), is due to the effect of the hot sun on the
tender seedlings. A large number of the seedlings were burned off at
the ground level soon after they pushed up through the hot sand, and
234
Journal of Agricultural Research
Vol. XVH, No. s
as a result many of those represented in the count of one day had dis-
appeared by the time of the next count. Likewise some of the seed
that germinated never entered into the calculations, the seedlings being
lost to observation between counts.
-I
\
?
\
\
1 ^
\
Q.
\
^
1
Ch
\
/
\
's
I
\
\.
\
K
\
^
-^
0
-I
— 1
--
■V
f
/
\
/
>
/
\
/
\
\
\
\
'
-1
~~~"
— -
^
<
-1
^
■^
/
\
S
\,
1
\
\ _
V
\
\
■^
-1
--
^
_J
"^ -SI
0 d:
V
\
\
\
^
^^
■ — ,
'
1 '
1 <
■^ V
: 1
■^ ■'
1
N
\
/
\
\.
\
s.
»i
<% >^ " o O) <D N
Two of the parent trees, A and D, show no apparent falling off in the
cur\'-e of germination for pulped seed; and parent trees A and C show
likewise no apparent falling off for unpulped seed. Careful obser\^ation
Aug. IS, 1919 Effect of Removing Pulp from Camphor Seed 235
has shown that the burning off of the young seedUngs occurred to a
greater or less degree throughout all the late winter and early spring
and that the burning off was by no means uniform in all the experimental
rows. It appears, therefore, that the curves given in figure 5 are not the
true curves of germination, but rather the curv^es of count. However,
the true curve of germination follows closely the curve of count with
some striking exceptions. As the season advanced a large number of
seeds germinated within a short period of time, and as the heat of the sun
became more intense a large number of seedlings were burned off. In
the case of eight of the trees the burning off of the seedlings between
counts reached a point where it included practically all the younger
seedlings just pushing up through the ground as well as some of the
more tender seedlings of the count previously recorded. Consequently
a drop occurred in the curv^e, which shows as an apparent decrease in the
percentage of germination. This explanation is further supported by
the fact that somewhat later in the spring, during more favorable weather
conditions, germination of more of the seed took place, as shown in Table
IV, which caused a rise in the curve of count, especially noticeable in the
curves for trees G, K, and P (fig. 5).
In the case of parent tree A, no apparent falling off in the germination
of camphor seed is recorded in either the pulped or unpulped seeds; in
tree C no apparent falling off is recorded for the unpulped seed; and in
tree D no apparent falling off is recorded for the pulped seed. In the
case of these trees. A, C, and D, the burning off of the young seedlings
which occurred between observations never reached a stage where the
total number of seedlings burned ofi' was large enough to cause a decrease
to show in the count. For this reason the curve of count for trees A,
C, and D, as indicated in figure 5, probably closely coincides with the
true curve of germination for the seed from these trees.
CAMPHOR SEEDLINGS BROUGHT TO TRANSPLANTING SIZE
Of more economic importance than the number of seed that germinate
is the number of seedling camphor trees which can be brought to trans-
planting size. Out of 4,800 seeds planted as they came from the tree
only 508 seedlings reached a sufficient size for transplanting, whereas
from 4,675 seed planted after pulping 3,499 such seedlings were secured.
The increase therefore in the percentage of seedlings of transplanting
size from the pulped seed over those from the unpulped seed amounted
to approximately 600 per cent.
The loss of seedlings due to the burning off by the hot sun is relatively
large and has a marked influence on the percentage of seedlings secured.
In this experiment 14.5 per cent of the total number of seedlings obtained
from the unpulped seed and 5.5 per cent of the seedlings from the pulped
seed were burned off. However, the total loss of seedlings from pulped
236
Journal of Agricultural Research
Vol. XVII, No. s
seed is much less than from unpulped seed, and this lower percentage
is due to the advanced growth obtained by the seedlings before the
extreme hot weather commenced.
In Plate 20 is shown a nursery bed of camphor seedlings just previous
to transplanting on December i, 191 8. In this bed the pulped and the
unpulped seed were planted in alternate rows. The small seedlings from
the unpulped seed are almost obscured by the alternate rows of large
seedlings which were obtained from the pulped seed.
At the time of transplanting, the seedlings had reached the compara-
tive size shown in Plate 21 A. In all instances the pulped seed had pro-
duced hardier and more vigorous seedlings, which, when trimmed and
cut back as shown in Plate 21 B, were in a better condition to withstand
the shock of transplanting. The superior growth of the seedlings from
pulped seed was so marked that measurements were taken of 258 seed-
lings of this lot and compared with measurements taken of the same
number of seedlings that came from unpulped seed. The results are
given in Table V.
Table V. — Average groivth attained by camphor seedlings during the growing period
between germination of the seed and transplanting of the seedlings
Treatment of seed.
Number of
seedlings
measured.
Average growth.
Length of
stem.
Length of
taproot.
Diameter
of crown.
Pulp removed . . . .
Pulp not removed .
2=;8
258
Inches.
Inches.
17-7
^5-0
Inches.
o- 473
.320
If the growth of seedlings from seed on which the pulp remained is
considered as 100 per cent, then the increased growth in the seedlings
from the pulped seed is: For length of stem, 18.1 per cent; for length of
taproot, 15.6 per cent; and for diameter of crown, 47.8 per cent. The
latter vigorous growth is of special interest from the commercial point of
view, since the loss from transplanting is much less with roots of large
diameter than it is with small roots, which are more easily dried out
during the period that elapses between the removal of the seedlings from
the seed bed and the date of the beginning of growth the following
growing season. The increased growth of the seedlings is brought about
directly by pulping the seed, since it insures a more rapid germination and
gives the plant an early start in the spring and consequently a much
longer growing season. The increased growth and vigor reduce to a
very low figure the loss through transplanting.
Aug. IS. 1919 Effect of Removing Pulp from Camphor Seed 237
SUMMARY AND CONCLUSIONS
In the season of 191 6-1 7 camphor seeds were planted under various
conditions. The experiment was repeated in the season of 1917-18, and
commercial conditions were closely approximated.
Removing the pulp from the seed was found to hasten germination by
an average of two weeks; it also gave an increase in germination of
approximately 525 per cent over that of unpulped seed.
Drying the seed with artificial heat at 55° C. destroyed all vitality.
Soaking the seed in water apparently did not hasten germination;
neither did it increase the percentage of seed that germinated.
Soaking the seed in sulphuric acid of 5 per cent by weight destroyed
all vitality.
Allowing the seed to ferment and the pulp to decompose in a closed
vessel destroyed all vitality.
Seed picked up from the ground showed less vitality than those
picked from the tree, but removal of the pulp increased and hastened
germination.
A freeze on three successive nights, during which the temperature
fell to 26° F., did not destroy the vitality of all the seed; it did, however,
impair the vitality and reduce the number of seeds that germinated by
approximately 50 per cent.
Seeds planted early in the winter required a longer time to germinate
than those planted in midwinter. The former averaged more sturdy
trees.
When the pulp was removed and the germination of the seed thus
hastened, a larger and more sturdy seedling tree was obtained for trans-
planting than when the pulp was not removed. The increase in the
number of seedlings of transplanting size secured by pulping the seed
amounted approximately to 600 per cent.
From a commercial point of view, removal of the pulp is desirable
even though the labor must all be done by hand. The increased germi-
nation and the well -developed trees that result will repay many times
the cost of the labor involved.
It is believed that in commercial plantings the removal of the pulp
from the seeds will increase the percentage of germination by at least
200 per cent, thus producing 40,000 more seedling trees to each acre of
seed bed. This estimate is believed to be very conservative, and even a
much greater increase may be expected.
PLATE 20
A camphor seed bed, showing the growth of seedlings from pulped and unpulped
camphor seed planted in alternate rows. The seedlings growing from pulped seed
have been cut away on one side in order to expose the seedlings growing from un-
pulped seed, which are otherwise almost completely covered by the luxuriant growth
of the former.
(238)
Effect of Removing Pulp from Camphor Seed
Plate 20
Journal of Agricultural Research
Vol. XVII, No. 5
Effect of Removing Pulp from Camphor Seed
Plate 21
Journal of Agricultural Research
Vol. XVII, No. 5
PLATE 21
A. — Camphor seedlings at the time of transplanting. The tree on the left is a
representative produced from pulped camphor seed; the one on the right is a repre-
sentative produced from unpulped camphor seed. Both seedlings are from seed of
the same parent tree and both are of the same age from planting of the seed.
B. — Camphor seedlings cut back and trimmed ready for transplanting. These
seedlings are the same as those shown in A. The one on the right is from pulped seed.
BACTERIUM ABORTUS INFECTION OF BULLS
[PRELIMINARY REPORT)
By J. M. Buck, G. T. Creech, and H. H. Ladson, Pathological Division, Bureau of
Animal Industry, United States Department of Agriculture
Numerous investigators have called attention to the fact that Bacterium
abortus agglutinins and complement-fixing bodies can frequently be dem-
onstrated in the blood serum of bulls from abortion-infected herds. Such
animals in consequence have frequently been referred to as being sys-
temically infected. While the presence of these bodies constitutes strong
evidence that abortion infection exists, or has been present, success has
been reported in associating positive reactions with the causative infec-
tion in so few instances as to have resulted in a certain amoimt of specu-
lation regarding the significance of these reactions in male animals.
Literature, it is true, records no great amount of investigative work
in connection with bulls suspected of being infected with abortion disease
where the object has been the isolation of the causative microorganism
from the organs or tissues of the animals or the demonstration of lesions
associated therewith.
Schroeder and Cotton ^ in investigating this problem describe two cases
that came under their observation. They state that one of the bulls at
the time of autopsy showed the presence of an abscess involving the
epididymis of one testicle from which Bad. abortus was isolated. The
other animal was permitted to serve a cow that was considered to be
free from abortion disease. Seminal fluid which was recovered from the
vagina immediately following the service and injected into numerous
guinea pigs produced Bad. abortus lesions in one of the experimental
animals.
Rettger and White ^ describe endeavors to associate the presence of
the infection with positive serum reactions in three cases which they
studied. In two of the animals neither abortus infection nor pathological
changes could be demonstrated. In the third they call attention to the
finding of two abscesses or cysts in the region of the groin, near the
point of attachment of the scrotum; but from these abscesses they were
unable to isolate the abortion organism, thus failing to obtain bacterio-
logical evidence of the infection.
In view of the positive bacteriological findings of Schroeder and Cotton
the present writers were prompted to undertake further investigations,
1 Schroeder,
E.
C, and Cotton, W.
E. SOME
FACTS ABOUT ABORTION DISEASE. In JoUT
• Act.
Re-
search, v. 9, no.
I, P
. 9-16. 1917.
2 Rettger, L.
F.
, and White, G. C.
INFECTIOUS ABORTION
l.\' TATTLE.
Conn.
Storrs Agr.
Exp.
Sta.
Bui. 93, p. 199-2
49.
1918. References, p. 246.
Journal of Agricultural Research, Vol. XVII, No. s
Washington, D. C. (239) Aug. 15, 1919
sf Key No. A-49
122501°— 19 5
240 Journal of Agricultural Research voi. xvii. no. $
involving a considerable number of animals, in an endeavor to ascertain
with what frequency abortus infection could be demonstrated in the
generative organs of bulls giving positive or suspicious reactions to the
agglutination test for this disease, and to determine whether or not
pathological changes are commonly associated with such infection.
EXPERIMENTAL PROCEDURE
The procedure employed by the writers consisted in securing blood
samples from the animals as they arrived at one of the abattoirs in close
proximity to Washington, D. C, for slaughter. No information was
available regarding the original source of the bulls or the exposure sus-
tained. Each sample was given a number corresponding to the serial
number of a tag that was attached to the animal's ear at the time of
bleeding. The blood samples were then taken to the laboratory for the
application of the agglutination test. At the time of slaughter, which
was usually the following day, those animals giving positive or suspicious
reactions were autopsied as carefully as abattoir conditions permitted
and the organs of the genital system were secured for further study. Cul-
tural work was depended upon as a means of detecting infection, the
medium employed consisting of 3 per cent glycerin infusion agar to which
approximately 5 per cent sterile blood serum was added. To reduce
the oxygen tension the inoculated tubes were subjected to incubator
temperature in closed jars in the presence of fresh cultures of Bacillus
subtilis.
During the period from December 9, 1916, to July 7, 1918, the agglu-
tination test for abortion disease was applied to 325 mature bulls. Of
this number 288 gave negative results to the test. The manner in which
the remaining 37 reacted is of considerable interest, inasmuch as the
intensity of the reactions appeared to bear some relation to the cultural
results.
The manner of applying the test consisted in the making of a i to 10
basic dilution of the blood serum. To the four tubes utilized for each
case were added 0.4, 0.2, o.i and 0.05 cc. of this basic dilution. The
amount of test fluid added to each tube was i cc.
The vesiculae seminales, vasa deferentia, testes, and epididymides
were secured from the 37 bulls whose blood serum showed the pres-
ence of Bad. abortus agglutinins. From 15 to 20 tubes of medium
were utilized for culturing the various organs from each bull. These
investigations resulted in the demonstration of the presence of Bad.
abortus infection in four animals — No. 88, 98, 136, and 409 — and in the
detection of marked lesions in bulls 98 and 409.
A brief description of the work performed and the findings in these
cases follow.
Aug. IS, 1919
Bacterium abortus Infection of Bulls
241
The agglutination reactions of the animals appear in the following
table.
Table I. — Results of agglutination tests
Animal No.
86.
88.
89.
98.
103
105
109
^33
134
136
137
143
146
147
150
154
165
Suspected serum
0.04 cc.
0.02 cc.
o.oi cc.
0.005 cc.
SI
SI
_
_
+
SI
—
—
SI
SI
—
—
+
+
+
SI
SI
SI
—
—
+
+
+
SI
+
+
+
+
+
SI
—
—
+
SI
—
—
+
+
+
+
SI
SI
SI
—
+
+
+
SI
+
+
SI
SI
SI
—
—
—
+
SI
~
—
SI
—
—
—
+
+
SI
—
SI
SI
—
—
SI
SI
Animal No.
177,
178,
179,
189,
198,
265
271
280
301
326
338
348
409
451
453
454
Suspected serixm.
0.04 cc. 0.02 cc. O.OI cc. 0.005 cc.
+
SI
+
SI
4-
+
+
+
SI
+
+
SI
SI
+
+
+
+
SI
SI
SI
SI
+
SI
+
SI
SI
SI
+
SI
+
+
SI
+
SI
SI
SI
SI
+
+
SI
SI
+
+= Complete agglutination.
— =No agglutination.
Sl= Partial clumping of bacteria.
EXAMINATION AND FINDINGS OF BUI.Iv 88
February 12, 1918. Agglutination test: 0.04 cc.4-, 0.02 cc. + , o.oi
cc. + , 0.005 cc. SI.
February 14, 1918. Slaughtered.
Macroscopic examination. — Fluid of left seminal vesicle turbid in
appearance and slightly more excessive in amount than that contained
by other organ. No indication of abnormal conditions noted elsewhere.
Bacteriological findings. — Of the 16 tubes of medium inoculated
from the various organs enumerated, after four days' incubation three
tubes developed from 40 to 60 colonies of an organism suggestive of
Bad. abortus and subsequently identified as such. These inoculations
were from the left seminal vesicle.
EXAMINATION AND FINDINGS OF BULL 98
March 8, 191 8. Agglutination test: 0.04 cc. + , 0.02 cc. + , o.oi cc.4-,
0.005 cc. SI.
March 9, 191 8. Slaughtered.
Macroscopic examination. — Marked pathological changes involved
the left seminal vesicle. The organ was increased from 8 to lo times
242 Journal of Agricultural Research voi. xvii, No. 5
its normal size. On cross section of the vesicle numerous hemorrhagic
areas were observed, as well as a number of necrotic centers, the latter
being confined chiefly to the more central portions of the organ. So
softened were some of these foci that the necrotic material assumed
a semifluid character. The capsule of the organ showed considerable
thickening. (See PI. 22.)
HisToi^OGiCAL EXAMINATION. — Sections from the left seminal vesicle
showed varying stages of the diseased process, ranging from exfoliation
of the epithelial lining of a few of the acini to complete obliteration
of the normal glandular structure. There was marked proliferation of
the interstitial tissue with round cell infiltration, which was more pro-
nounced immediately surrounding the acini and just beneath the
epithelial lining of the acini. In those areas exhibiting the more pro-
nounced pathological changes many of the acini were filled with detached
epithelial cells and cell debris. In other areas where the mere outline
of the acini could be traced, a homogeneous substance was present,
together with more or less granular detritus. There were hemorrhages
into and between the acini. Occasionally large areas of degeneration
and necrosis were observed. As a result of the inflammatory changes
little normal glandular structure was recognized in many of the sections
examined (PI. 24, A and B). Plate 23 A, representing a normal seminal
vesicle, is inserted for comparison with Plate 23 B, and Plate 24, A
and B.
BacterioIvOgical findings. — Eighteen tubes of medium were utilized
for the culturing of the different organs. The six tubes from the left semi-
nal vesicle after three days' incubation developed from 75 to 1 50 colonies
of an organism that appeared typical of Bad. abortus. All tubes inocu-
lated from other sources remained sterile, although incubated for several
additional days. Subsequent work with the organism isolated estab-
lished its identity as Bact. abortus and indicated that pure cultures of the
organism were isolated in all instances.
EXAMINATION AND FINDINGS OF BULL 1 36
May 15, 1918. Agglutination test: 0.04 cc.-H, 0.02 cc.-|-, o.oi cc.^-,
0.005 cc. SI.
May 16, 1918. Slaughtered.
Macroscopic examination. — The right seminal vesicle showed slight
enlargement. The fluid contained by this organ presented a turbid
appearance. No lesions were elsewhere detected.
BacteriologicaIv findings. — Five of the i8 tubes of medium inocu-
lated from the different organs developed from 40 to 60 colonies of an
organism that was subsequently identified as Bad. abortus. These inoc-
ulations were from the right seminal vesicle.
Aug. 15, 1919 Bacterium abortus Infection of Bulls 243
EXAMINATION AND FINDINGS OF BULL 409
August 15, f9i8. Agglutination test : 0.04 cc.+ , 0.02 cc. + , o. 01 cc. + ,
0.005 cc. + .
August 16, 1918. Slaughtered.
Macroscopic examination. — Left seminal vesicle showed evidence of
disease. This organ was approximately twice the size of the right and
was incised with considerable difficulty on account of fibrous tissue pro-
liferation. The fluid contained was decidedly turbid. Other organs
presented a normal appearance.
Microscopic examination. — Sections from the left seminal vesicle
showed marked proliferation of the interstitial tissue with areas of round-
cell infiltration. Degeneration and exfoliation of the epithelial cells lining
the acini were observed. A few of the acini contained cells and cell
detritus ; others had been completely obliterated as a result of the inflam-
matory process. (See PI. 23 B.)
Bacteriological findings. — Two of the 20 tubes of medium that were
inoculated from the various organs developed colonies typical of Bad.
abortus. The colonies were few in number and appeared on but 2 of 5
tubes that were sown with material from near the same point. These
tubes were from the left seminal vesicle. All tubes inoculated from
other sources remained sterile. The infection was subsequently estab-
lished as Bad. abortus.
It has been previously suggested that the intensity of the serum reac-
tions appeared to bear some relation to the cultural results. Of the 2>7
bulls exhibiting agglutinating properties for a Bact. abortus suspension,
the blood serum of but 7 caused perfect agglutination of a suspension
with o.oi cc. of the serum. It may be observed that 4 of these 7 animals
yielded positive cultural results and that in no instance was the presence
of the infection demonstrated in animals when their blood serum failed
to cause perfect agglutination with such an amount of serum.
examination and findings of guernsey bull
Since the isolation of Bad. abortus infection from the cases previously
described, the writers have had an opportunity to demonstrate the
presence of the infection and observe lesions in a fifth bull where the
isolation of abortion bacteria was carried out under different conditions
and where it was possible to obtain a somewhat more complete history
in regard to the development of the pathological changes that were
associated with the infection.
This pure-bred Guernsey, 8 years of age, was acquired by the present
owner in June, 191 8, and appeared at the time to be in perfect physical
condition.
The writers were informed that during the following January an asym-
metrical enlargement of the scrotum was noted. Mechanical injury was
244 Journal of Agricultural Research voi. xvii, No. s
suspected which had prompted the application of fomentations and
counterirritants. When the condition failed to respond toJ:his treatment
and an area of softening that appeared to involve the left testicle was later
detected, a canula had been introduced through which had been evacuated
a considerable quantity of a semifluid material. It was furthermore
stated that the animal had at times discharged through the urethra a
substance bearing some resemblance to that removed by the surgical
procedure.
On April 22, 191 9, about three months after the swelling was first
observed, a sample of blood was secured for the application of the aggluti-
nation test for abortion disease. The specimen caused clumping of a
Bad. abortus suspension with o.oooi cc. of the serum.
When the animal was examined on the following day with the object
of obtaining material for bacteriological work, the enlargement involving
the left testicle was found to be four or five times the size of the normal
organ. When a needle was passed into its lateral wall, little resistance
was encountered after the instrument had been inserted for about i}4
inches. Through the needle were aspirated from 400 to 500 cc. of a gray-
colored substance of the consistence of heavy cream. At the same time
from 20 to 30 cc. of a turbid fluid were obtained from the urethral opening.
This material was secured by exerting pressure on the urethra and by
massaging the seminal vesicles through the walls of the rectum. During
this procedure it was detected that the seminal vesicles differed markedly
in size, enlargement of the right organ being pronounced.
CuivTURAiv RESULTS. — Eight tubes of serum agar were inoculated with
the semifluid substance aspirated from the interior of the enlargement in-
volving the diseased testicle. Numerous dilutions were made of the fluid
recovered from the urethra with physiological salt solution, and serum-
agar tubes were sown with these dilutions. When the tubes were exam-
ined after six days' incubation one colony of abortuslik^ appearance was
observed on one of the tubes from the substance obtained by aspiration.
The infection was later established as Bact. abortus. Further inoculations
of medium with like material resulted in the isolation of additional
abortus colonies, although fewer in number than were anticipated from
the extenj: of the lesions. No Bact. abortus was isolated by cultural
methods from the material secured from the urethra, but excessive
contamination made these results inconclusive.
On May 9, or about two weeks after the condition was diagnosed as
abortus infection, an opportunity was afforded for the making of a more
thorough examination of the diseased process involving the external
genitals and for further bacteriological work, for the affected testicle with
its coverings were at this time removed and forwarded to the Patholog-
ical Division.
Aug. IS, 1919 Bacterium abortus Infection of Bulls 245
The weight of the mass of tissue was 53^ pounds. On section it was
found to consist of an outer wall or capsule of from iX to 2 inches in
thickness. This abnormal structure had evidently resulted from pro-
liferative changes involving mainly the connective tissue coverings of
the testicle. Firmly embedded in this external layer could be distin-
guished areas of tissue that upon microscopic examination were identified
as epididymis that had undergone severe inflammatory changes. The
cavity formed by this dense fibrous wall contained a considerable quantity
of a grayish -colored, semifluid material identical with the substance pre-
viously obtained by aspiration. Floating free in the cavity was also a
mass of tissue that was recognized as the remains of the testicle, it having
the same general form although somewhat reduced in size. Blood vessels
no longer communicated with the organ, and the serous membranes
which normally envelop it had seemingly been entirely obliterated. The
close resemblance existing between the semifluid substance and softened
portions of the testicle strongly indicated that the organ was undergoing
liquefaction necrosis.
Microscopic examination. — The thick wall surrounding the testicle
consisted largely of dense fibrous tissue with a certain amount of round-
cell infiltration. Different portions of the epididymis which were em-
bedded in this mass showed extensive interstitial proliferation, which had
resulted in a wide separation of the tubules. Chronic inflammatory
changes were noted in sections from the testicle proper. Many tubules
were surrounded by zones of round-cell infiltration. There was exfolia-
tion and more or less disintegration of the epithelium lining the tubules,
causing the latter to be largely occupied by cell debris. Advanced
degenerative changes, verging on necrosis, were observed in all the sec-
tions examined, the peripheral portion of the organ exhibiting little more
than a mere outline of the testicle structure.
BacterioIvOGICAl findings. — Tubes of serum agar that were inoculated
with the exudate at this time developed numerous colonies of an organism
that was identified as Bad. abortus.
It has been suggested by writers on abortion disease that Bad. abortus
infection when acquired by bulls remains active for a comparatively
brief period, the resistance offered being sufficient for its destruction.
The encountering of a considerable number of animals giving slight
agglutination reactions and the isolation of abortus infection from only
a small percentage of the bulls cultured would tend to strengthen the
theory that the infection may commonly terminate in this manner. On
the other hand the extensive pathological changes and the chronic
character of the lesions exhibited by three of the five bulls where abortus
infection was demonstrated suggest that it may be unwise to assume that
long-standing cases of infection never exist.
246 Journal of Agricultural Research voi. xvii, no. s
CONCLUSIONS
Bad. abortus infection may involve organs of the generative apparatus
of bulls, producing chronic inflammatory changes.
Of the generative organs, the seminal vesicles appear to furnish the
most favorable site for the lodgment and propagation of abortion
infection.
The presence of Bad. abortus infection in bulls appears to be more
strongly indicated by relatively marked than by slight reactions to the
agglutination test for this disease.
PLATE 22
Photograph of normal and diseased seminal vesicles of bull 98, showing the
marked increase in size and the gross pathological changes of one of the organs.
Bacterium abortus Infection of Bulls
Plate 22
■■««k'.«r
I
'i-^ \
•!<-*
«r'%
'/f"-
U
14 -t^^
l^
r
;/
Journal of Agricultural Researcli
Vol. XVII, No. 5
Bacterium abortus Infection of Bulls
Plate 23
Journal of Agricultural Research
Vol. XVII, No. 5
PLATE 23
A. — Photomicrograph of a section from a normal seminal vesicle of bull. X 92.
B. — Photomicrograph of section from seminal vesicle of bull 409, showing inflam-
matory changes. X 92.
PLATE 24
A. — Photomicrograph of section from seminal vesicle of bull 98, showing tissue
proliferation and exfoliation of epithelium lining acini. X 92.
B. — Photomicrograph of section from seminal vesicle of bull 98, showing ad-
vanced pathological changes with cell degeneration and necrosis. X 92.
Bacterium abortus Infection of Bulls
Plate 24
Journal of Agricultural Research
Vol. XVII, No. 5
Vol. xvn skf»xje:mbe:r 15, 1919 No. 6
JOURNAL OP
AGRICULTURAL
RESEARCH
CONTKNXS
Page
Investigations on the Mosaic Disease of the Irish Potato - 247
E. S. SCHULTZ, DONALD FOLSOM, F. MERRILL
HILDEBRANDT, and LON A. HAWKINS
(Contribution from Bureau of Plant Industry )
Temperature in Relation to Quality of Sweetcorn •- - 275
NEIL E. STEVENS and C. H. HIGGINS
( Contril>ution from Bureau of Plant Industry)
Variation of Ayrshire Cows in the Quantity and Fat Con-
tent of Their Milk _-_-_- - 285
RAYMOND PEARL and JOHN RICE MINER
( Contribution from Maine Agricultural Experiment Station )
Index and Contents of Volume XVII - _ - - 323
PUBLISHED BY AUTHORITY OF THE SECRETARY OF AGRICULTURE,
WITH THE COOPERATION OF THE ASSOCIATION OF AMERICAN
AGRICULTURAL COLLEGES AND EXPERIMENT STATIONS
V^ASHINOTTON, r>. C.
WAtHINQTON ! OOVEKNMENT PRINTINO OFTICe t l»l|
EDITORIAL COMMITTEE OF THE
UNITED STATES DEPARTMENT OF AGRICULTURE AND
THE ASSOCIATION OF AMERICAN AGRICULTURAL
COLLEGES AND EXPERIMENT STATIONS
FOR THE DEPARTMENT
FOR THE ASSOCIATION
KARL F. KELLERMAN, Chairman H. P. ARMSBY
Physiologist and Associate Chief, Bureau
oj Plant indusirv
EDWIN W. ALLEN
Chief, Office of Experiment Stations
CHARLES L. MARLATT
Entomologist and Assistant Chief, Bureau
of Entomology
Director, Institute of Animal Nutrition, The
Pennsylvania State College
J. G. LIPMAN
Director, New Jersey A gricultural Experiment
Station, Rutgers College
W. A. RILEY
Entomologist and Chief, Division of Ento-
mology and Economic Zoology, Agricul-
tural Experiment Station of the University
of Minnesota
All correspondence regarding articles from the Department of Agriculture should be
addressed to Karl F. Kellerman, Journal of Agricultiural Research, Washington, D. C.
All correspondence regarding articles from State Experiment Stations should be
addressed to H. P. Armsby, Institute of Animal Nutrition, State College, Pa.
JOmALOFAGRIdllMLESEARCH
Vol. XVII Washington, D. C, September 15, 1919 No. 6
INVESTIGATIONS ON THE MOSAIC DISEASE OF THE
IRISH POTATO^
[PRELIMINARY PAPER]
By E. S. ScHWtz,^ Pathologist, Cotton, Truck, and Forage Crop Disease Investigations,
Bureau of Plant Industry, United States Department of Agriculture, Donald Folsom,
Assistant Plant Pathologist, Maine Agricultural Experiment Station, F. MERRILL
HiLDEBRANDT, funior Chemist, and LoN' A. Hawkins, Plant Physiologist, Plant
Physiological and Fermentation Investigations, Bureau of Plant Industry, United States
Department of Agriculture
INTRODUCTION
The economic importance and wide distribution of the mosaic or
"calico" disease of tobacco (Nicotiana tahacujti L.), as well as its dis-
tinguishing characteristics, have been a matter of common knowledge
among pathologists and practical growers for many 3'ears. The fact
that mosaic occurs also on certain others of the Solanaceae is well
recognized, but it has been known for only a comparatively short time
that the Irish potato (Solanum tuberosum L.) is subject to a similar
malady.
As will be shown, potato mosaic, although more common and appar-
ently more destructive in certain sections of the United States than in
others, is widely distributed in North America. While the data regarding
it which have so far accumulated are necessarily limited, there is a
tendency among those pathologists who have given the subject special
study to regard it as a disease of great economic importance. The
results of the studies described in this paper, chiefly those which throw
light on the means of transmission of the disease, are made more sig-
nificant by the fact that they were obtained in four different laboratories,
partly through collaboration and partly as the result of independent
work.
' This paper was read at the conference of potato pathologists on Long Island, June 26, 1919. An abstract
was published in Phytopathology.
The investigations were conducted as a cooperative project between the Office of Cotton, Truck, and
Forage Crop Disease Investigations of the Bureau of Plant Industry and the Department of Plant
Pathology of the Maine Agricultural Experiment Station.
2 The authors wish to acknowledge their indebtedness to Dr. H. A. Edson and Dr. W. J. Morse for helpful
suggestions and criticism of the manuscript and to Dr. Joseph Rosenbaum, Mr. M. Shapovalov, and
Mr. G. B. Ramsey for assistance in furnishing material and collecting data.
Journal of Agricultural Research, Vol. XVII, No. 6
Washington, D. C. Sept. 15. 1919
sg Key No. G-177
(247)
248 Journal of Agricultural Research voi. xvii, No. «
GEOGRAPHICAL DISTRIBUTION OF POTATO MOSAIC
Orton (9, p. 42)^ in 191 1 first observed potato mosaic in a field at
Giessen, Germany, where it was very common on some varieties. The fol-
lowing year it was found to be prevalent in the potato fields in northern
Maine but was not found in Wisconsin, Minnesota, Colorado, and other
western states during an extended survey made in 1912 and 1913. In 1913
Melchers (<5, p. 15s) observed symptoms of this disease in the greenhouse
on potato plants from tubers from New York. More recently Worfley
{12) reported it as very prevalent on the Bliss Triumph variety in
Bermuda and on Long Island, and Murphy (8) said that the disease
occurred to a considerable extent in New Brunswick and to a less extent
in western Canada. In 1917 and 1918, collaborators for the Plant
Disease Surv^ey reported it from the following states: Alabama, Arkansas,
Connecticut, Delaware, Florida, Georgia, Kentucky, Louisiana, Maine,
Massachusetts, Michigan, Minnesota, New Hampshire, New York,
North Dakota, Ohio, Oregon, Texas, Vermont, Virginia, and Wisconsin.
From these reports it is apparent that potato mosaic occurs rather
generally throughout the United States.
Although potato mosaic, named as such, has been reported for the
first time within the last decade, the following statement made by
Johnson (4) before the middle of the nineteenth century is of interest.
In a description of a potato disease which seems to have somewhat
resembled mosaic he says :
The stem is tinbranched, brownish green or mottled, and here and there sprinkled
with rusty spots, which penetrate to the pith, so that it is not white but rust colored
or sometimes black. The upper surface of the leaves is not as smooth as is usual in
the case with potato leaves but rough, ivrinkled, or curled. The leaves are far more
sessile than usual, and are not of a uniform brownish or dark green color, but spotted?
Johnson further says that this trouble can be produced by repeated
removal of the sprouts before planting.
EFFECTS UPON YIELD
The yield from affected plants is less than that from healthy vines of
the same variety. Orton {9, p. 42) as the result of an experiment with
Green Mountain potatoes in northern Maine reports a difiference in yield
of 22 per cent between 80 healthy and 80 diseased hills. Wortley {12)
states that 200 healthy Bliss Triumph hills yielded more than twice as
much as 200 diseased hills of the same variety and that mosaic of potatoes
in Bermuda frequently causes a reduction in yield of from 10 to 75 per
cent. Murphy {8) compared 682 diseased Green Mountain hills, scat-
tered over II plots, with the same number of healthy hills growing
adjacent to the diseased hills. He found the yield of the former to be
' Reference is made by ntunber (italic) to " Literature cited," p. 272-273.
^ Italics in this quotation are supplied by the writers.
Sept. IS, 1919 hwestigations on Mosaic Disease of the Irish Potato 249
but 58 per cent of the latter and concluded that in New Brunswick,
Canada, the yield is reduced about i| bushels for every i per cent of
mosaic present. Reduction in yield reported by collaborators of the
Plant Disease Survey {11) ranges from 5 to 30 per cent. Comparative
results secured by the writers in northern Maine will be discussed later
in connection with the questions of hill selection and roguing.
The preceding statements refer to the comparative yields of healthy
and entirely diseased lots and so may seem to be somewhat inapplicable
to conditions where a large number of the plants are not diseased and
where these may possibly be able to make up for the deficiency of
affected plants by making better growth at their expense. However, the
writers have found that often, in the absence of any control measures, a
healthy lot of a susceptible variety will show symptoms of the disease
in some hills the next year after being grown near to diseased stock
and will thereafter from year to year have a larger percentage of hills
affected.
APPKARANCE OF THE DISEASED PLANTS
Some of the symptoms will be described here, although the subject
has already received considerable attention (<5, 7, 8, 10). On Green
Mountain or Bliss Triumph potatoes, the leaves of affected plants are
characterized by mottling (PI. A; B; 25), which is produced by the
presence of light green areas on the foliage. These areas may occur on
any part of the leaf; they may include or adjoin sections of the larger
veins or not come in contact with them. The light green patches vary
greatly in shape, being punctate, elongate, circular, angular, and irregular.
Considerable variation in the degree of paleness may be seen even in the
same small discolored patch, from a barely discernible fading of the green
to an almost pure yellow. The abnormal spots differ in distinctness of
outhne, usually in proportion to the degree of discoloration. Their
dimensions seldom exceed a few millimeters. Their frequency varies,
usually becoming greater as the disease progresses and thus giving to the
general appearance of the leaves a much lighter color than that of healthy
foliage. In the more severely affected plants the foliage may become
spotted with brown flecks of dead tissue. Furthermore, in the more
advanced stages the foliage presents a characteristic crinkled or corru-
gated appearance. In these stages the diseased plants are frequently
dwarfed because the stems, the leaf petioles, and leaf blades are con-
siderably shortened or reduced in size.
The symptoms as described above are not so marked in certain other
varieties — for example, in Blue Victor, Early Rose, Irish Cobbler, Pearl,
White Bliss, Carmen, Early Dix, Netted Gem, Peach Blow, Portuguese
Purple, and Spaulding Rose. In the first five named, decided rugosity
is a characteristic of the disease.
250 Journal of Agricultural Research Voi. xvu, no. 6
So far no symptoms have been discovered by which mosaic can be
recognized in the dormant tubers ; nor has any effect upon the percentage
of germination or the time of blossoming been observed, although pre-
mature death may occur.
The presence of mottling on the leaves is apparently modified by cli-
matic conditions. It was found by planting a part of the same affected
stock and strain in northern Maine and Colorado, that, whereas distinct
mottling occurred in northern Maine, none whatever could be detected
on the stock in Colorado during the same season. Similar tests were made
at Washington, D. C, and in northern Maine; and although some mottling
was noted at Washington there were a number of doubtful cases, while
the same stock in northern Maine showed very distinct mottling. Melhus
(7) found that progeny of plants which were mottled in northern Maine
did not show such s)anptoms in Iowa but showed symptoms of "curly
dwarf."
For three successive seasons a number of lots of mosaic and healthy
seed potatoes have been divided and planted at the two experimental
farms of the Maine Agricultural Experiment Station. One of these
farms is located in the northeastern and the other in the southwestern
part of the state. Usually the part of a lot grown in southwestern Maine
showed considerably less mottling than the part grown in northeastern
Maine, while the reverse has never been noted. In two out of the three
seasons these differences have been very marked. On the other hand,
when the same lots which showed practically entire absence of mosaic
mottling in one location — in southwestern Maine — one season were re-
moved to the other and planted the following season, the mottling again
appeared in marked degree.
TRANSMISSION STUDIES
TRANSMISSION BY TUBERS
Orton (9) cites a preliminary experiment and suggests the probability
of tuber transmission. Wortley {12) found that all tubers from affected
plants produced foliage with mottled leaves. Stewart {10) says that
mosaic is transmitted through the tubers. As pointed out before, Melhus
(7) showed that, under Iowa conditions, plants from diseased tubers
might not exhibit the mottling of the leaves but might show a dwarfing
and curling of the foliage similar to "curly dwarf." Murphy {8) says,
"Mosaic is perpetuated by planting the tubers from diseased hills."
These conclusions are confirmed by evidence which has been secured
by the writers and which will be presented later in connection with the
questions of hill selection and roguing.
Sept. IS, 1919 Investigations on Mosaic Disease of the Irish Potato 251
TRANSMISSION BY GRAFTING
Experiments were carried out in the winter of 191 6-1 7 at Washington,
D. C, to see if it were possible to transmit the disease by grafting. In
these experiments two methods of grafting were followed, the cleft-graft
and the inarch. According to the first method the top of a young,
apparently healthy, potato plant was removed, the base sliced down to a
thin wedge and grafted in the place of the top of a diseased plant. The
scion was held in place by winding with adhesive tape. Of six plants
grafted in this way that grew well, all the scions showed evidence of
the disease (PI. 26, A). Four of the plants from which the scions were
taken remained apparently healthy. The other two showed evidences of
the disease. Grafts were made according to the inarch method by placing
a healthy and a diseased plant side by side, slicing away a thin layer of
the outer tissue of the stem, bringing the cut surfaces in close contact, and
fixing them by wrappings of adhesive tape. After the plants had re-
mained in contact for several days the stem of the healthy plant was cut
below the point of attachment and the top of the diseased plant removed.
In three grafts made in this way the scion of one became diseased while
the parent plant remained healthy. The other two were doubtful.
This last-mentioned method of grafting seemed not to be adapted to
potato plants because, unless maintained in a very humid atmosphere
the scions wilted. However, Giissow (j) in 191 8 by inarching a mosaic
shoot on a healthy one found that no mosaic symptoms formed on the
foliage of the sound plant but that tubers from it produced mosaic
plants.
The results obtained in these preliminary experiments were corrobo-
rated by a number of experiments in the field in 191 7, the results of which
are shown in Table I. In this series no attempt was made to control
aphids, noi were any observations made, after grafting, on the plants
which supplied the scions. However, these plants were from 4 to 6
inches high and free from mottling at the time of grafting.
Table I. — Grafts of potato vines, Presque Isle, Me., igi'j
Date.
Variety.
Graft.
Num-
ber
grafts.
Num-
ber
mot-
tled.
Num-
ber
non-
mot-
tled.
Num-
ber
doubt-
ful.
Per-
cent-
age
mot-
tled.
August
Do. .
Bliss Triumph . . .
Green Mountain . .
Healthy scion on af-
affected stock.
do
17
17
II
10
2
5
4
2
64.71
58.82
During the summer of 191 8 grafting experiments were continued in
northern Maine. Although more than 100 grafts were made, relatively
few of these made sufficient growth, 4 to 12 inches, to show distinct
252
Journal of Agricultural Research VoI.xvu.No.
mottling. In order to study the behavior of the plants from which the
scions were taken, these plants as well as the scions and stocks were
labeled. Their performance is indicated in Table II under the heading
of "Condition of parent vine." The grafts were made when the plants
were from 4 to lo inches in height. In the majority of cases the cleft-
graft method was used. After the insertion of the scion the contact
between scion and stalk was effected by wrapping tightly with adhesive
tape. The performance of these grafts is recorded in Table II.
Table II. — Grafts of potato vines, Presque Isle, Me., igi8
Date.
Variety.
Graft.
Condition
of parent
vine.
Num-
ber of
success-
ful
grafts.
Num-
ber of
grafts
mot-
tled.
Num-
ber of
grafts
non-
jnot-
tled.
July 6 to Aug. 17.
Do
Bliss Triumph ....
Green Mountain
Healthy scion
on diseased
stock.
. ..do... .
Healthy
to end of
season
do. .
14
6
5
3
14
19
2
3
Do
Bliss Triumph . . .
Green Mountain. .
Healthy scion
on healthy
stock.
do
...do
. ..do. ..
6
Do
S
Do
...do
Affected scion
on healthy
stock.
. . . do . .
Do
Bliss Triumph
These results indicate plainly that distinct mottling of the healthy
scions grafted upon diseased stocks had developed by the end of four or
five weeks, whereas no mottling developed on either the parent plants
or the healthy scions grafted upon healthy stocks. (See PI, 27, A, B.)
A few new shoots from stocks supporting affected scions showed mottling,
but since only a small number of these grafts were made the results are
inconclusive.
In the winter of 191 8-1 9, 61 Green Mountain grafts were made at
Orono, Me., by means of the cleft-graft method already described. Of
the 50 which survived, 14 consisted of healthy scions on healthy stocks
and remained entirely healthy for from 43 to 82 days, 9 making new growth
from the stock and i from the scion; 15 consisted of healthy scions on
mosaic stocks; and 7 of these, or 41 per cent, developed mosaic on the
scion in from 21 to 44 days, although the plants from which the scions
came remained healthy. In the 7 mosaic scions there was usually a con-
tinuation of leaf expansion, and the mosaic symptoms developed in the
youngest leaves. The scions which remained healthy usually showed no
good growth. Of 21 grafts consisting of a mosaic scion and a healthy
stock, the one whose stock produced the most new growth showed
Sept. IS. I9I9 Investigations 07i Mosaic Disease of the Irish Potato 253
much wrinkling and some mottling on this new growth; 3 other stocks
showed wrinkling only; and the rest remained healthy, even in the
rather poorly developed new shoots.
TRANSMISSION BY PLANT JUICES
Attempts were made to inoculate tubers of the Green Mountain and
Bliss Triumph varieties with juice from diseased plants. In these
inoculation experiments the method followed was to divide the potato
in half longitudinally, make a cavity in one piece, fill this cavity with
the juice from the crushed stems and leaves of the diseased plants, and
then plant this treated piece. The other half of the potato was planted
in a separate pot as a control. In the first experiment, with four Bliss
Triumph and four Green Mountain tubers, all the stalks from four of the
inoculated portions of these tubers were typical mosaic plants. One of
the control plants, corresponding to one of the inoculated portions
which developed mosaic, also showed the disease. These experiments
were repeated a number of times with larger numbers of tubers, but only
occasionally did the inoculation appear to be successful. In all of the
transmission experiments it has been difficult to secure seed-tuber lots
which were absolutely free from mosaic contamination, so it was to be
expected that occasionally both the inoculated and uninoculated parts
of the same tuber would produce diseased plants. On the other hand,
the uninoculated controls remained healthy in some experiments where
the inoculated seed piece produced a mosaic plant, while the converse
did not occur. Hence the evidence secured is presumptive that the
disease can be transmitted by inoculating seed tubers with juices of
affected plants.
In northern Maine during the season of 191 8, 50 hills of apparently
healthy potato plants of the Green Mountain variety were treated with
the filtered and unfiltered extracts from diseased tubers and leaves.
These juices were applied by means of painting upon rubbed, bruised,
or slashed leaves, and by hypodermic injection into the petioles. The
plants at the time of this treatment, July 9 and 10, were about 12 inches
tall and in actively growing condition. Obser\-ations on July 20 and on
August 17 indicated that no treated plants had developed mottling but
appeared like the controls, which were treated with water. In order to
note whether this treatment of the foliage had any effect upon the tubers,
progeny of these hills was reserved for study in 191 9.
On November 23, 1918, in a preliminary experiment in the green-
house at Washington, D. C, juices extracted from potato vines were
transferred to foliage of the Bliss Triumph variety. This operation was
performed several times in the course of a month, the first inoculation
being made when the plants were 3 to 6 inches high. By December 20,
1 91 8, fully 30 per cent of the inoculated plants showed mottling on the
254 Journal of Agricultural Research voi. xvn.No. 6
youngest leaves. It was noted also that this mottling occurred only in
connection with two very similar treatments. In view of these sug-
gestive results, a similar experiment with the more promising of the
methods employed in November, 191 8, was begun February 22, 191 9,
In this experiment healthy plants from 17 different tubers of the
Green Mountain variety were inoculated according to the methods indi-
cated at the foot of Table III. At the time of planting, each of the 17
tubers was halved lengthwise, so that for each treated plant an untreated
control plant of the same tuber was obtained. The halves of each tuber
were designated respectively x and y and with the same number. Each
half tuber was planted in an 8-inch pot.
At the time of the first inoculation, the height of the plants varied
from 2 to 6 inches, and the number of shoots to each half tuber varied
from two to seven. As shown in Table III, plants from the tuber halves
472X, 483X, 473y, 484X, 471 y, and 485y were treated with juices from
healthy plants according to the methods indicated and served as control
to the plants treated similarly but with juices from mosaic potato vines.
The remaining 1 1 plants, from as many tuber halves, were treated with
juices from mosaic foliage. All juices were taken from vines of the
Green Mountain variety.
The performance of the treated and untreated plants is noted in Table
III. Number 472y represents the untreated plant and 472X the treated
plant developed from the same tuber. At this time observations on
foliage of the plants treated according to method 3 with juices from
mosaic plants indicated that no mottling had developed. This method
failed to produce mottling in the November experiment also. How-
ever, with method 5 and with method 7 seven different plants had
developed new leaves since March 22; and five of them, or 71 per cent,
showed distinct mosaic mottling on the younger leaves formed after
the time of the last inoculation on March 22 (Pi. 28, A). The first mot-
tling on any of these plants was noted on March 25. On examination
March 25 and April 3, 191 9, no mosaic mottling was found on either
the old or newly formed leaves in any of the controls, treated or
untreated (PI. 27, A, B). All the plants in this experiment were free
from aphids.
Sept. IS, 1919 Investigations on Mosaic Disease of the Irish Potato 255
2 o
^^^
000000 0005
M rt o
^ -^^
■w 13 -d o
.§tII§
^ 01 o o r
S.ii a 2 S
, O rt O rt
« " «;: o
■^ S.g g ^
■S.s^.g^.a
iillii.g"
§.a?.sg^§
.a
u
JS
a
Ui
Tl
a
0
r^
«
a
0
J3
g
a
is
UJ^
>
a>
■ V
c<
. M
is
t-
*-i
13 -d G'5^2
.2 -a ca o
^0 0 iz:
g « 3 fl ^'o ^
SS o o-" " ^ 2
g o ^c cS J'-
« g o o M^ S n
■g « o5 d & 2 ti
ja S K 3 d
■Bsi^:3
5 000000
o ._, ._, .— o
mil 00000 int3 000
ii o-axs-djaTa « n-d-dja'
'^ :^
fc/j (U
o-g
•35
•dg
OJ bfl
^ o
.. o
•d^J
So
.S"B
5<
4JX|
32
35
■d j3
0
«
a
A
«
c
t
ffl
rO
Q.
•rt
bO
0
d
«-s
256 Journal of Agricultural Research voi. xvii, no. e
TRANSMISSION BY APHIDS
The fact that plant diseases are frequently carried by insects is well
recognized. In this connection the work of Allard (z, p. 626) on the
mosaic of tobacco is of special interest. This writer showed that the
virus of tobacco mosaic is readily carried by the common green peach
aphis, or spinach aphis {Myzus persicae Sulz.). More recently McClintock
and Smith (5) have demonstrated that the spinach-blight, which ap-
parently is a virus disease, also is transmitted by plant lice, the pink
and green potato aphis (Macrosiphum solanijolii Ashmead) and the
spinach aphis both acting as carriers. Doolittle's work (2) with cucum-
ber mosaic is also worthy of mention in this connection.
FIELD EXPERIMENTS WITH INSECT CAGES
From findings of these writers it has seemed possible that the mosaic
of potato might be spread by some insect. To study this question an
attempt was made to grow plants in the field under cages that were
supposedly insect-proof. These were 22 by 30 by 36 inches, covered
with cheesecloth, one side being arranged so that it could be opened — a.
type that was used also by McClintock and Smith (5, PL 5 and 6).
Potatoes were planted about 14 inches apart, so arranged that one
cage covered two hills. During the season of 1917 at Presque Isle, Me.,
potato plants were grown throughout the season under these cages and
observations made from time to time on their condition as regards
mosaic. It was found that the percentage of mosaic in the cages was
practically the same as that to be found in the same stock planted in the
adjoining plots. However, since the disease may be acquired in one
season without showing the symptoms until the tubers develop their
shoots the following season, it was necessary to continue the comparison
through 1 91 8. It was then found that not more than 5 per cent of the
tubers from healthy plants caged in 191 7 were mosaic, the lowest season-
to-season percentage of increase on record for lots grown on the experi-
mental plots.
Tubers from plants grown under the cages in 191 7 and not showing
the characteristic mottling of the disease during the season were selected
for planting in the cages in 191 8. On account of the poor quality of the
cheesecloth obtainable in the second season the cages were not insect-
proof; and within them there were found, at the end of the season,
considerable numbers of aphids as well as some insects of other kinds.
However, since the dispersal of aphids probably was checked more or
less by the cages, tubers were reserved for the 191 9 season for comparison
with uncaged lots.
Sept. 15, 1919 Investigations on Mosaic Disease of the Irish Potato 257
GREENHOUSE EXPERIMENTS WITHOUT INSECT CAGES
Since greenhouse conditions are more favorable to the control of
aphids, experiments with the pink and green potato aphis were conducted
in the greenhouse at Washington, D. C, during the winter of 1917-18.
The insects were allowed to develop on stock of the Bliss Triumph
variety which during the preceding summer had been rogued in the field
in northern Maine — that is, had the plants showing mottling eliminated
from the stock. However, as Tables IV and V show, about 22 per cent
of the plants developed mottling on January 28, 191 8, when they were
from 2 to 6 inches tall. From these afifected plants the aphids were
permitted to disperse to the neighboring, apparently healthy plants;
and in addition on March 5 artificial transfers of aphids from diseased
to nonmottled plants were made on fully a dozen different plants. By
March 19, 191 8, it was noted that many of the plants infested with aphids
had developed a crinkling and mottling, on the newly formed leaves
only, very similar to mosaic mottling (Pi. 29, A). The number of such
mottled plants increased so that by April 6, 191 8, 50 per cent of the plants
showed mottling. On the other hand, only 15 per cent of the remainder
of this 1917-grown stock were diseased when grown at Presque Isle in
the season of 191 8. This 15 per cent, as well as the 22 per cent which
first showed mosaic in the greenhouse experiment described above,
undoubtedly were progeny of hills that had become diseased in 191 7
in spite of the roguing. The increase to 50 per cent seems to be explained
best by the dispersal of the aphids from the diseased plants. Moreover,
the percentage of plants to which the aphids transmitted the disease in
this experiment was really 100, inasmuch as all plants, whether or not
eventually becoming mottled in 191 7-1 8, produced progeny which was
decidedly mottled in the winter of 191 8-1 9. (See Table IV, "Perform-
ance of second generation.-') That is, all the tubers which were saved
from nonmottled plants, as well as all tubers from the mottled plants,
produced mosaic vines in the following winter when planted in the same
greenhouse with no aphids present.
258
Journal of Agricultural Research voi.xvn.No.e
Table IV. — Effect ofaphids on mosaic of potato at greenhouse, Washington, D. C, 1918-1Q
[Bliss Triumph variety]
Pot No.
i04{b
{i
105
io6|^
'{?
Condition of plants on "
Jan. 28,
1918.
108^
\
"°{b
fa
112-
"3{b
"4b
"sjb
ii8J
i
A
"4b
4
tb
"9{b
^"4b
"3{b
"4b
"4b
I26'
H".
H. .
H..
H..
M..
M. .
H..
H. .
M..
M..
H..
H..
H..
H. .
H.
H.
M.
M.
H.
H.
H.
H.
H..
H.
M.
M.
H.
H.
H.
H.
H.
H.
M.
M.
H.
H.
H.
H.
M.
M.
H.
H.
H.
H.
H.
H.
H.
H.
H.
H.
H.
H,
Mar. I,
1918.
H,A. .
H,A. .
H,A..
H,A. ,
M, A. .
M,A..
H,A..
H,A. .
M,A. .
M,A. .
H,A. .
H,A. .
H,A..
H,A. .
M,A..
M, A. .
M, A.
M,A..
H,A..
H,A. .
M,A..
H,A. .
H,A..
H,A..
M,A. .
M,A. .
H,A..
H,A..
H,A.,
H.A..
H,A..
H,A..
M,A.
M.A.
H,A..
H,A.
M,A.
M,A.
M,A .
H, A.
H,A.
H,A.
H,A.
H,A.
H,A.
H,A.
H,A.
H,A.
H,A.
H,A.
H.A.
H,A
Mar. 19,
1918.
Apr. 6, 1918.
H,A..
H,A..
M,A..
M, A.
M, A.
M,A..
M,A..
M,A..
M,A..
M,A. .
H,A. ..
H,A. ..
H,A...
H,A...
M,A...
M,A...
M,A...
M,A...
M,A...
M,A...
Dead. .
H,A. ..
M,A...
M,A...
M,A...
M,A...
H,A..
H,A..
H,A..
H,A. .
M,A. .
M, A. .
M,A..,
M,A..
M,A..
M,A..
M,A. .
M,A..
M,A..
M,A..
H,A..
H,A..
H,A..
H,A. .
H,A. .
H,A..
H,A. .
H,A..
H.A..
H,A. .
H.A. .
H,A..
A.,
A.
A.
A.
A.
A.
A.
A.
A.
A.
A.
A.
A.
A.
A.
A.
A.
A.
A.
A.
Perform-
ance of
second
genera-
tion, in-
spected
Jan. 4,
1919.
H, A.
Dead.
Dead.
Dead.
H,A
H,A
Rapidly maturing .
Mm.
Mm.
Mm.
Mm.
Mm.
Mm.
Mm.
Mm.
Mm.
Mm.
M, A
Dead
M, A
M, A
M, A
M, A
M, A
Dead
H, A
H,A
H, A
M,A
H,A
H,A
H,A
Dead
Young leaves dead .
M.A
M.A
Young leaves dead .
Mm.
Mm.
Mm.
Mm.
Mm.
Mm.
Mm.
Mm.
Mm.
Mm.
Mm.
Mm.
» All potatoes were planted Dec. 19, 1917- . . . , ^. . • r » j „ui, „„u;^.,
•> H=liealthy. M=mosaic. Mm=having a medium mosiaic mfecUon. A=mJested with apluds.
Sept. 15, 1919 Investigations on Mosaic Disease of the Irish Potato 259
Table IV. — Effect of aphids on mosaic of potato at greenhoitse, Washington, D. C,
igiS-iQ—ContmueiA
Condition of plants on —
Perform-
Pot No.
Jan. 28,
1918.
Mar. I,
1918.
Mar. 19,
1918.
Apr. 6, 1918.
second
genera-
tion, in-
spected
Jan. 4,
1919.
"M
H
H
H
H
M
M
H
H
H
H
H, A .
H,A...
H,A...
H,A...
H,A...
M,A..
M,A...
H,A. ..
H,A...
H,A. ..
H, A. ..
HA
H
H
H
M
M
H
H
H
H
H
H
M
H
M
M
M
H
H
H
M
M
H
H
H
H
H
H
H
H
M
H
H
H
H
H
H
H
H
H
M
M
H
H
H
H
H
H
H
H
A...
A...
A. .
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A..
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A...
A.v.
A...
A...
A...
A...
A...
A...
A
H. A
Al
H. A
Mm
H. A
"'{b
M
Mm
Dead
Mm
Ab
H,A
Dead
Ml
Too mature
Ml
H, A. .
Mm
H,A ..
Mm.
i33{b
M,A...
H,A..
M,A. ..
M,A. ..
M,A...
H,A. ..
H,A. ..
H,A. ..
M,A...
M,A...
H,A...
H,A...
H,A...
H,A...
H,A...
H,A. .
Dead
Mm
H
Mm.
^34{b
M, A
Dead
i35{b
M, A
Mm.
H, A
Mm.
Ml
Too mature to observe
A
Ml
M
Dead
Ml
M
Mm.
M
Mm.
^39{b
H, A
H, A
Ml
Too mature to observe
^^Hb
H,A. ..
H,A...
M,A...
H,A...
H,A...
H,A...
H,A. ..
H,A. ..
H, A. .
A
Mm.
A
Mm.
Ml
M,A
Mm.
Dead
Mm.
Ml
M, A
Dead
/a
M, A
M, A
Ml
H,A..
Ml
H,A...
H,A...
M,A...
M,A...
H,A ..
H,A...
H,A...
H,A...
H,A...
H,A ...
H,A...
M,A...
M,A
Mm,
M, A
Mm.
Ml
M, A
M, A
Ml
H, A
H,A
Ml
H,A
H,A
Ml
H,A
Mm.
H,A
Mm.
Ml
Dead
Mm.
Mm.
26o
Journal of Agricultural Research voi.xvn,No. 6
Table V. — Swmmary of Table IV
Condition of plants on
Jan. 28, 1918.
Condition of plants on
Mar. I, 1918.
Condition of plants on
Mar. 19, 1918.
Condition of plants on
Apr. 6, 1918.
Total
number
of
plants
loia to
132a.
Num-
ber of
mosaic
plants.
Per-
cent-
age of
mosaic.
Total
number
of
plants
loia to
151b.
Num-
ber of
mosaic
plants.
Per-
cent-
age of
mosaic.
Total
number
of
plants
loia to
151b.
Num-
ber of
mosaic
plants.
Per-
cent-
age of
mosaic.
Total
number
of
plants
loia to
151b.
Num-
ber of
mosaic
plants.
Per-
cent-
age of
mosaic.
62
14
22.5
102
28
27
102
40
39
102
SI
50
Total number of plants grown from above progeny in second generation is 44.
Number of second generation plants showing mottling is 44.
Number of plants without mottling in first generation but mottled in second is 21.
Percentage of plants mottled in second generation but not in first is 48.
Similar experiments were performed at Washington in the winter of
1 91 8-1 9. Bliss Triumph potatoes, from stock that had been rogued
during the preceding season in northern Maine, were planted in two
lots. One lot was kept free from aphids by fumigation while the other
was subject to a heavy infestation. In the former, ii per cent — the
progeny of 2 out of 18 halved tubers — became mottled as soon as the
first leaves appeared, evidently as a result of field infection. In the
latter, 67 per cent, or 31 out of 46 plants — progeny of 23 halved tubers —
developed mottling. The difference between 11 per cent and 67 per
cent evidently was the result of aphid dispersal from neighboring mosaic
plants of the same variety. The aphid-free lot was planted December
17, 1918, and was fully matured by March 22, 1919. The infested lot
was planted in 8-inch pots on February i, 191 9, in a separate greenhouse
but with growing conditions practically the same as those of the other.
Hundreds of aphids were present upon the plants by the time they had
developed to a height of 6 to 8 inches. The plants were arranged in
five rows, the plants in row i being in contact with the originally aphid-
infested plants and the other rows following in numerical order at re-
spectively greater distances from them and therefore being less infested
by the dispersing aphids. As shown in Table VI, all the plants in rows
I and 2 showed mottling by April 4, while at that time some but not all
of the plants in the other three rows were mottled.-
Sept. IS. 1919 Investigations on Mosaic Disease of the Irish Potato 261
Table; VI. — Relation of aphids to mosaic of potato: Continuation of experiments at
greenhouse, Washington, D. C, iQl8-ig
[Planted Feb. i.
1918; observed Apr.
4, 1918]
Seed piece No.
Condition of foliage.
Num-
ber of
8-inch
pots re-
moved
from
aphid-
in-
fested
plants.
Remarks.
X.
y-
Mottled .
Mottled .
...do
I
I
I
2
2
2
I
...do
Mottled fromi beginning.
...do
Mottled.
...do
Mottled
...do.....
...do
...do
...do
...do
...do
...do
...do
2
2
3
3
3
4
4
4
5
5
3
3
4
4
5
5
S
Mottled from beginning.
...do
...do
...do....
...do
Healthy
Healthy.
Mottled
Healthy.
Healthy.
Mottled.
Mottled .
Healthy.
Healthy
...do
Mottled .
...do
...do
...do
Healthy
Mottled .
. do....
Healthy.
...do
...do
...do
Mottled .
...do
Mottled .
Mottled from beginning.
...do
...do
Do.
Total number of plants is 46.
Number of plants showing mottling Apr. 4, 1919, is 31.
Percentage of plants showing mottling Apr. 4, 1919, is 67.
Somewhat similar evidence was secured during the same winter at
Orono, Me. Some Green Mountain potatoes were used that had been
grown in a rogued plot in northern Maine during the season of 191 7 and
had been kept for about a year in cold storage. One lot of 10 tubers was
planted immediately and 2 of them, or 20 per cent, produced plants that
were mottled when very young, evidently through field infection. An-
other lot of 30 tubers was stored in a cellar for a few weeks and then was
found to have produced sprouts that had become lightly infested with
green peach or spinach aphids. These aphids apparently had dispersed
from a neighboring heavily infested lot of sprouted tubers that had
come from a purely mosaic stock and that later produced mosaic plants.
The number of insects on a tuber varied from o to 30, and there were
few skins and but little honey-dew deposit present. The infested lot
was fumigated and planted. Five tubers, or 17 per cent, produced
plants that became mottled when very small, in 25 to 30 days after
planting, evidently the result of field infection. In addition to these,
6 other tubers, or 20 per cent, produced both mottled and healthy shoots.
This increase can be explained only by the infestation of the sprouts by
the aphids from the diseased tubers. This explanation receives support
from the observation that the mottled shoots of the 6 partly diseased
262 Journal of Agricultural Research voi. xvu.no. 6
tubers showed the symptoms later, averaging 44 days after planting,
and that they usually came from eyes of the bud end and therefore were
probably the first to become exposed to aphid attack.
• GREENHOUSE EXPERIMENTS WITH INSECT CAGES
As has been indicated already, plants that appear healthy may pro-
duce tubers that develop mottled plants. In studies with potato mosaic,
therefore, it is ver>' desirable to grow a second generation if the effects
of a given treatment are to be fully disclosed. Under greenhouse con-
ditions, especially in Maine, it is necessary to furnish treated plants
with as much light as possible if a satisfactory crop of tubers is to be
secured. This makes it appear better, in experiments involving the
artificial introduction of aphids, to remove as soon as possible any cages
that were used. This can be done without compromising the results of
the experiments if frequent inspection and fumigation are employed to
keep insects reduced to negligible numbers.
During the winter of 191 8-19 an experiment was performed in the
greenhouse at Orono, Me., with Green Mountain potatoes that had been
grown in a rogued plot in northern Maine during the season of 191 7 and had
been kept for about a year in cold storage. Fifteen tubers were planted,
of which 3, or 20 per cent, produced plants which showed mosaic symp-
toms when only a few inches tall. The same rogued stock when planted
in the field in 191 8 had shown mottling in 11 per cent of the hills. The
other 12 tubers, each being divided into 2, 4, or 5 sets, furnished 53
plants. Twenty-one plants, i or 2 from each tuber, were kept as untreated
controls throughout the experiment and remained healthy. Eighteen
plants, I or 2 from each tuber, were fed upon by aphids introduced from
mosaic potato plants; and 13 of them, or 72 per cent, eventually devel-
oped typical mosaic symptoms. Five plants, from 5 tubers, were fed
upon by aphids introduced from a healthy potato plant; 8 plants, from
8 tubers, were infested by aphids from radish plants; but all of these
remained healthy.
In this experiment spinach aphids ^ were used and were never found,
during frequent inspections, to be parasitized by other insects or by
fungi or to be mixed with predatory enemies or with individuals of another
aphid species. They were secured from two colonies, one on a mosaic-
diseased potato plant and the other on an apparently mosaic-free one.
Stock from the former was kept on mosaic-diseased potato plants and
that from the latter on healthy ones or on radish plants until ready for
use. The aphids were transferred to the plants of the experiment by
methods that seemed favorable to the transmitting of mosaic: (i) By
laying one or two leaves, bearing feeding aphids, upon the plant so that
1 Determinations were made by Dr. Editli M. Patch, Entomologist of the Maine Agricultural Experi-
ment Station, who informs one of the writers that this species frequently is abundant upon potato
plants in Aroostook County and other parts of Maine.
Sept. IS. 1919 Investigations on Mosaic Disease of the Irish Potato 263
the insects could crawl most easily to the new host; (2) by introducing
aphids when the new host was young, 3 to 13 inches tall; and (3) by
introducing a rather large number, 40 to 220 by estimate. Cylindrical
cages consisting of coarse wire screening covered with fine cloth gauze
(5, PI. 6 B) were used to confine the aphids to the individual treated
plants. These effectually served their purpose. For the three treat-
ments with aphids — from mosaic potato, healthy potato, and radish —
the height of the tallest shoot of the new host when the aphids were
introduced was on the average 6, 5, and 7 inches, respectively. The
number of aphvds introduced was on the average respectively 130, 80,
and 120, while the average number of days the insects remained was
respectively 7, 14, and 9. After feeding on the new host for a week or
longer, the ap'iids were killed by nicotine fumigation intense enough to
cause the marjins and tips of some leaflets to become yellow and later
to die. This yellowing occurred on both the aphid-infested and aphid-
free plants and was in no way similar to mosaic mottling. It did not
occur on the leaves which were the first to show mosaic symptoms.
Frequent cyanid and nicotine fumigation of the uncaged plants was
practiced. No white flies (Aleyrodes vaporariorum Westw.) and very
few dispersed aphids were found at any time in the room occupied by
the plants included in this experiment. No other species of aphids was
found in the greenhouse. Thrips fed somewhat upon all the plants,
both those within cages and those uninclosed.
The aphids were introduced in December and January. Symptoms
of mosaic were first seen in 1 8 to 3 1 days and then consisted of the mot-
tling characteristic of "slight" mosaic, but the mottling soon became
more pronounced and sometimes was accompanied by considerable
wrinkling. The average number of days that elapsed between the
introduction of the aphids and the time when mosaic symptoms were
first ascertained was 26. It might have been shorter if the plants had
been examined daily instead of semiweekly. The average height of
the tallest shoot at the time when the symptoms were first ascertained
was 20 inches. The symptoms appeared first in the one, two, or three
topmost leaves of an affected shoot, which, if already formed, w-ere still
very small at the time the aphids were feeding on the plant.
The fact that a large percentage of the plants treated with aphids
from mosaic potatoes showed mosaic while the others, either untreated
or treated with nonvirulent aphids, all remained healthy, can be attri-
buted only to aphid transmission. A.s pointed out before, the group of
plants that showed mosaic came from the same tubers as the healthy
controls. Moreover, special precautions were followed because of the
previous tendency to regard mosaic as a physiological disease and there-
fore to neglect some operations normally followed in pathological work.
All the plants were grown in the same greenhouse room and were
arranged so that those with each type of treatment were distributed
122502°— 19 2
264 Journal of Agricultural Research voi. xvii, No. 6
over the bench, all four treatment groups thus being mixed and appar-
ently exposed to similar conditions of light, temperature, and humidity.
Each plant had enough space so that it was not in contact with any
other. Soil fertilization and watering were similar for all plants. There
was as much variation in the amount and type of soil used for the plants
that showed mosaic symptoms as for the others. The untreated controls
came from neither eye-end sets nor stem-end sets. Each tuber was
cut with a flamed knife, and the seed pieces were planted in steam-
sterilized soil. Finally, the objection that the method used for intro-
ducing aphids brought in the factor of contact with diseased leaves, is
met by the results of 14 checks in another room of the greenhouse.
These 14 tubers from the same stock produced tuber hills, each of which
remained entirely health}^ for 38 days after a mosaic leaf or shoot had
been placed upon it when it was 8 inches high.
In connection with the experiment reported in Table VI another
aphid experiment was conducted at Washington, D. C, in the winter
of 1918-19, but withaGreen Mountain lot and with insect cages emplo3'ed.
From this stock for the last three seasons the mosaic plants had been
eliminated, so that but 13 per cent of the plants developed mosaic as
soon as new leaves were formed. In this experiment, plants from 5
different tubers were used. Each of these tubers was halved, making
10 sets. Plants from 5 different sets, designated 474X, 48 ix, 486X,
470X, and 478X, were kept in the greenhouse without a cage; and plants
from the corresponding 5 different sets, designated as 474y, 481 y, 486y,
47oy, and 4783^, were placed in two cages which were kept in the same
greenhouse with the uncaged plants. Three of these plants, 474y, 48 ly,
and 486y, when from 3 to 6 inches tall were placed in one cage, while
the two remaining plants, 47oy and 4783', were placed in another cage.
On February 26, 191 9, a few hundred aphids taken from health3' Green
Mountain plants were transferred to each of plants designated 47oy
and 478y, and similar transfers were made on March i and on March 15.
The aphids were brushed upon cardboard with a camel's-hair brush
and then transferred to the plants. Before the transfers were made
on March 15 the plants were fumigated. Upon 4743% 4813% and 486y
aphids from mosaic plants were transferred in a similar manner, but
with a different brush and cardboard. Here also three distinct trans-
fers were made on February 26, March i, and March 15. The plants
were fumigated on March 15 and the third transfer made. At this
time a few hundred aphids were placed upon each of the plants. This
last set of aphids was allowed to feed on the plants until March 22, when
the lower half of each stalk had become defoliated. Then another
tobacco fumigation was applied, and the cages were removed from the
plants. At this time a few of the newly formed leaves showed distinct
mottling. On April 2 newly formed leaves on all 7 stalks representing
the 3 different plants were distinctly mottled (PI. 30, A). At this time
Sept. IS, 1919 Investigations on Mosaic Disease of the Irish Potato 265
the 3 plants which had developed from the other half tubers and had
been kept free from aphids were free from mottling. On examination
of the control plants 4707 and 4787 as well as 470X and 478X on April 4,
no mottling whatever was found (PI. 30, B).
In January, 191 9, some Green Mountain potatoes were secured at
Orono, Me., supposedly from a field that had been found free from mosaic
the previous season. Seventeen tubers were divided each into 6 seed
pieces. Eight tuber groups of 6 plants each developed mottling when
very small, and the other 9 did not. The 6 plants from each tuber
were subjected to 6 different treatments: One plant was kept as an
uncaged control ; another was a control , caged until the plant was over 2
feet tall; the third was grown intertwined with a mosaic potato plant
from a separate pot; the fourth was fed upon for a week by wingless
green peach aphids from a mosaic potato plant, an average number of
about 130 being introduced on a piece of gauze when the plant was 3
inches high; the fifth received the same treatment as the fourth except
that the average number of insects was about 170 and that they were
introduced on leaves which were impaled upon a sterile stick thrust
into the soil in such a way that there was no contact between the intro-
duced leaves and the plant or soil (see PI. 26, B) ; on the sixth plant when
I inch high there were placed 20 winged aphids secured from a mosaic
plant with a camel's-hair brush and introduced within a small open
bottle.
All of the 18 controls remained healthy. Of the 9 plants with aphids
introduced on leaves, as described above, 8, or 89 per cent, became mot-
tled in 20 to 31 days — averaging 25 days — or when the plants had be-
come 14 to 29 inches high — averaging 25 inches. One of these plants,
together with an untreated plant from the same tuber, is shown in Plate
29, B, and corresponding leaves from these two plants are shown in
Plate 27, C. Of the 9 plants with aphids introduced on gauze, 2, or 22
per cent, became mottled in 20 to 26 days, when the plants were 22 to 29
inches high. Of the 9 plants with winged aphids introduced, i, or 11
per cent, showed signs of mosaic in 27 days, when 17 inches high. Of
the 9 plants kept in contact with mosaic plants, all remained healthy
but I. This I was brought into contact with the diseased plant on
March 7, was found to be free from aphids on March 31, was fumigated
on April 7 because of the presence of several aphids, and showed signs of
mosaic on April 17, when 35 inches high. This plant became diseased
apparently either because of transmission by very few aphids after
March 31 or because of contact. The latter cause seems more prob-
able, but would make this the only case of contact transmission known
at present to the writers.
This experiment seems to have demonstrated that aphids can transmit
mosaic, even better than the first experiment conducted in this green-
house (p. 262-264). The same precautions were used in this experiment,
266
Journal of Agricultural Research voi.xvn.No. 6
and in addition each tuber was split lengthwise so that each seed piece
included eyes from the bud-end and the stem-end. Also, no mosaic
leaves were put in contact with the plants when aphids were introduced,
and all cages that had been used previously were steam sterilized.
PHYSIOLOGICAL STUDIES
Some work on the chemical differences between the healthy and dis-
eased potato plants was carried out in connection with this investiga-
tion. These experiments included a determination of the reducing and
total sugars and starch content in the healthy and diseased leaves of
potato plants grown under the same environmental conditions. In this
work the potato leaves were picked off the stems and weighed. They
were dried in a hot oven to constant weight and extracted with alcohol
in a Soxhlet extractor. The sugars were determined in the extract, and
the starch in the residue. The results of these determinations are given
in Table VII. Data are given in this table as to the location of the
plants from which the samples were taken, whether caged or in the open,
the time of sampling, and whether the day was bright or cloudy. In
the other columns of the table are given the results of the sugar and
starch determinations. The determinations compared in the table are
from plants grown under as nearly the same conditions as possible.
TablS VII. — Sugars and starch in healthy and mosaic Green Mountain potato foliage
Percentage of sugars as dextrose on basis of dry weight.
Percentage of
starch on basis of
dry weight.
Time, treatment of plants,
and weather conditions
when sampled.
Reducing sugar.
Nonreducing sugar. 1 Total sugar.
Healthy.
Mosaic.
Healthy.
Mosaic.
Healthy.
Mosaic.
Healthy.
Mosaic.
11 a. m. Plants caged.
1.6
2-3
2-5
I. 0
2-3
2. S
2.4
3-0
1.6
4.4
3-S
5-5
7- 0
I. I
4. 2
3-7
5-5
5- I
3-3
4-3
5- I
7.8
95
2. I
6.5
6-S
7-9
8.1
4.9
8.7
2S
28
21
18
23
17
13
17
11.30 a. m. Plants in
open. Bright day. . . . .
9.30 a. m. Plants in
open. Bright day
2.10 p. m. Plants caged.
19
16
19
IS
4.1S p. m. Plants in
open. Rainy day
IS
IS
1.94
2.84
4. j6
4-38
6.20
7.22
20. 7
16.6
From the results shown in Table VII it appears that mosaic plants
have a higher sugar content than the healthy plants grown under the same
conditions. This is true of both reducing and nonreducing sugars,
though the differences in the latter are not so marked. There is an
average of about i per cent more total sugar in the mosaic plants than
in the healthy. With starch this relation is reversed, healthy plants
having an average of about 4 per cent more starch than those affected
with mosaic. It is, of course, to be remembered that the investigations
Sept.is.i9i9 Investigations on Mosaic Disease of the Irish Potato 267
in this paper are preliminary in character. It is hardly possible to draw
conclusions from so limited an amount of data, though the facts seem
worth recording.
METHODS OF CHECKING NATURAL TRANSMISSION
The experimental results previously described in this paper suggest at
least one way in which transmission of potato mosaic may occur in the
field — namely, by aphids. Both species of aphids that were experimented
with are commonly found on potatoes, including those in Aroostook
County, Me. In 191 8, a year in which aphids were unusually abundant
upon potatoes in northern Maine, they began to appear upon the plants
about the middle of July. Since in the experiments mottling did not
appear after the plants had finished elongating and had produced blos-
soms, it is quite probable that aphid transmission in the field occurs too
late for the effects to be shown during the same season. The possibility
of this was demonstrated in one experiment (p. 262-264) in which
after aphid transmission some plants remained unmottled but produced
progeny that showed disease the next season. Before evidence had
accumulated regarding insect transmission, control of the disease was
attempted by means of hill selection and roguing. The results of such
attempts, together with notes made at the same time on yields, will
now be discussed.
HILL SELECTION
A number of hill selections were made in 191 6 and 191 7 in northern
Maine in order to ascertain more especially the progress of mosaic from
one season to another upon the same strain and stock. Plants in three
different stages of the disease as well as healthy checks were included in
these selections. The term "slight stage" was used when the plants
had just begun to show a few mottled spots on the leaves though the
foliage otherwise appeared like that of healthy plants. "Medium
stage" was used when the leaves had apparently just begun to become
slightly corrugated, had six or more mottled areas, and had begun to
show slight dwarfing. "Bad stage" indicated that the leaves were
mottled, corrugated, and decidedly dwarfed. The results of the obser-
vations on the behavior of the foliage are presented in Table VIII.
268
Journal of Agricultural Research voi. xvii, no. 6
Table VIII. — Hill selection: Effect of mosaic of potato on vines in laboratory plots,
Presque Isle, Me., IQ18
Variety.
Planted for—
.g
6
1
a
6
•6
CI
a
■5.
tu
.a
"o
d
■ft
•d
6
•a m
2o
6
SI
on
bj
a .
Sc
a ts
ma
6
Z
0 ]S
tx.2
c c,
1-2
•a"
£ c
%l
£, 0
3-d
*^ c
d^
Z"
'3
0
a
d
.s
"o
d
No. of plants
in each stage
of mosaic.
a
.5
■3
liH
1^
CI]
Bliss Triumph
Control to mosaic
do
8
6
II
5
3
II
9
2
4
5
8
4
9
4
2
5
6
5
S
3
10
3
2
4
2
3
8
10
10
II
2
3
11
4
21
3
31
30
39
14
8
122
119
284
82
I
4
3
0
0
S
6
0
0
3
0
0
0
0
0
0
7
2
3
I
0
3
3
0
I
0
0
0
0
0
0
0
0
s
4
3
3
0
7
4
4
3
I
I
0
I
4
0
0
2
0
53
24
147
70
69
95
137
12
3
I
40
24
146
70
44
10
Do
Do
Do
Irish Cobbler
Do
Green Mountain
Do
do
.. .do::
do
do
31 ;ii6
34
22
7
74
6
70
82
95
0
do
Slight
.do
2
IS
12
24
14
20
8
S
II
14
10
10
6
20
6
4
8
4
6
16
20
20
22
4
6
17
7
47
7
7
74
84
lil
82
79
32
30
63
75
39
40
36
80
24
IS
3-
16
24
32
39
40
76
12
24
32
42
202
42
7^r-
Do
I
78
87
66
12
76
27
8
12
54
6
52
16
20
72
39
28
36
16
8
4
28
Do
do
Do
Shght
.. ..do
79
32
20
63
7S
39
40
4
S
6
0
0
8
Do
... .do
Do
do
8
Do
do
3
Do
do
Do
. do
Do
. . .do
Bliss Triumph
Bad
6^
Green Mountain
do
t6
Do
... .do
Do
do
33
16
24
32
39
40
76
12
24
32
42
Do
do
Tfi
Do
.do
16
8
Do
. . do
33
White Bliss
39
32
76
12
24
32
42
14
1;
Bliss Triumph. . . .
do
P
Irish Cobbler
Do
do
Green Mountain
Do
do
Do
Slight
Irish Cobbler
Control to streak and
mosaic.
. do ...
s
2
it8
Do
7
■
From these data it is apparent that progeny from plants seeming to
be healthy in one season may develop both healthy and diseased stock
the following season. It will be noted further that the mottling on this
control stock the following season may develop to such a degree that it
falls under all three stages, slight, medium and bad, and does not neces-
sarily begin with a slight stage as one might expect. Furthermore, it is
shown that mottled and nonmottled plants may develop from the same
hill and even from a single tuber, similar observations having been
recorded by F. C. Stewart (/o).
The observations regarding degree of mottling and dwarfing in the
control stock also obtained in the slight, medium and bad stages. In
none of these stages did the stock necessarily run true to the stage for
which it was selected. It will be noted also that mottled foliage devel-
oped wherever the progeny came from plants showing the slight, medium,
or bad stage the previous season. In but two strains where the foliage
Sept. IS, 1919 Investigations on Mosaic Disease of the Irish Potato 269
appeared to be very slightly mottled but questionable for mosaic in 191 7
were mottled and nonmottled plants noted in 191 8. Observations upon
the foliage symptoms were made at three different times during July
and August, the first observations being made when the vines were from
2 to 6 inches tall and the last just before the vines began to die.
In connection with the observations on the behavior of the vines of
hill-selected stock, studies were made also on the effect of mosaic on
yield. The hills were selected from some of the same stock on which
notes upon the performance of the foliage were taken, and hence the
stages of the disease indicated here answer the same description as those
presented in connection with the notes upon the behavior of the foliage
as indicated in Table VIII. Table IX gives the effect on yield of these
hill selections.
Tablu IX.
-Hill selections: Effect of mosaic of potato on yield in laboratory plots,
Presqtie hie, Me., igiy-iS
Plot
No.
4
4
4
4
4
5
5
5
5
5
5
16
16
16
25
25
25
25
25
36
36
36
37
37
37
37
37
37
37
37
37
37
37
Condition of stock in 1917
and 1918.
Healthy
....do.
....do..
. . . .do. .
....do..
....do..
....do..
....do..
....do..
....do..
....do..
Slight...
....do..
....do..
Bad . . .
....do..
...do..
...do..
....do..
Medium
...do..
...do..
....do..
....do..
....do..
....do..
....do..
....do..
....do..
....do..
....do..
....do..
....do..
Hill
unit.
4A
4C
4D
4K
4F
5A
5B
5C
5D
5E
5F
I6R
I6H
I6I
2SA
2sE
25F
25H
2SJ
36A
36B
36C
37A
37B
37C
37D
37E
37F
37G
37H
37I
37J
37K
Num-
ber of
hills.
Total
yaeld.
15
II
6
3
3
5
7
14
Average
yield
per hill.
Lb. Oz.
I 15
I 14
1 14
2 5
1 14
I 12
I 15
I 12
1 14
2 A
2 \\
I 10
I 4
I 3
4
8
5
7
5
I 13
2 y2
3
13
6
Yield of
parent
hill in
1917.
Oz.
5
13
9
9
14
9
6
6
4
14
12
Yield per hill in
1918 compared
with that of 191 7.
In-
crease.
De-
crease.
Oz.
10
I
5
12
Oz.
3
9
6
10
2
5
10
4
6
I
10
II
7
2
3
I
2
12
8
6
3
3
10
3
3
4
3
270 Journal of Agricultural Research voi. x\tj, no. s
In this table the yield per hill in the season of 1918 is compared with
that of 1 91 7. From control plants and those slightly mottled there was
a slight increase in yield, whereas from plants showing medium and bad
stages of mottling there was in some an increase but in a larger number
a decrease. Although it will be necessary to study the performance of
such stock for a number of seasons before final conclusions upon the effect
of mosaic on the yield can be submitted, nevertheless it is clear that mosaic
hills can be depended upon to produce diseased progeny, while apparently
healthy hills can not be depended upon to produce healthy progeny.
Consequently hill selection is an unsatisfactory method of control, at
least when practiced in a field that contains a considerable percentage of
affected hills.
ROGUING
Additional observations on the effect of mosaic of potato on yield were
made in connection with the experiments on roguing. In these experi-
ments the stock was not hill-selected but was harvested in bulk after the
affected hills were removed from the plots during the growing season.
The results of these observations as well as of those on the behavior of
the vines are indicated in Table X.
From these data a reduction in yield of from 23 to 30 per cent is apparent
where progeny from wholly diseased lots is compared with the progeny
from lots of the same strain and variety but with a low percentage of
mottled plants. Furthermore, the mottled plants vv^ere reduced from 45
per cent in 191 7 to 13 per cent in 191 8 as the result of but one thorough
roguing in 191 7, when the plants were about 12 inches tall. However, in
order to note how much the percentage of mottled plants can be reduced
by roguing it will be necessary to study the effect of this procedure on
the same stock and strain for several seasons and under as nearly uniform
conditions as possible.
By reducing the number of diseased plants in the seed stock the effect
of the aphids in spreading the disease is apparently considerably reduced.
It is quite evident that such roguing must be carried on with the greatest
care and by persons who are thoroughly acquainted with the symptoms
of the disease. Even though practically all diseased plants can be elimi-
nated with a single roguing in one season, the work can be done more
efficiently with two or three roguings, beginning when the plants are
from 6 to 10 inches tall. With this method it is advisable to begin with
a stock which runs relatively low in the number of affected plants.
Whether it is possible entii-ely to eliminate mosaic by roguing has not
been proved. From the results of the study of aphid transmission here
reported it is evident that attempts to eliminate mosaic by roguing should
be made on an isolated seed plot removed from aphid-infested fields. In
addition, insects of all kinds should be kept off the seed plot by adequate
spraying. Naturally the same precautions should be taken if one wishes
to prevent transmission of the disease to seed plots or fields planted with
mosaic-free seed tubers.
Sept. IS, 1919 Investigations on Mosaic Disease of the Irish Potato 271
t^
'^
"^
X
n
«
«
3
1
k-
xn
.S
.3
.9
B
M
CU
rt
Vh
i^
k
rt
>;
"
rJ
■s
"^ d
=i
(A
(U
;g
0)
0
Si
;S 0
•d C
1
a
i
0.
E
■5,
0 *->
.a t
^? 8
r'
8
8
OC
Ci
MD
^" i
"
8
S
i
•^
2
^§
^
•d
d
d
a
a
X
J3
>.
*j
ai
8
s
'11
^
^
^
sa
ai
%
^
Si
►^
■3.
s
"n
Q
^
Oi
• in
p^
„
0
c
00
•aS
^" a
«:
to
c
"S^a
S "
H
H
tH
">;s
<i
a
d.
sS
.aS Yr
C
t VC
^
{:
c
' Z
"5 '^
w cv t?5
^
r* ^
-r
^
f- I
&
w CC 0
r-
00
K
t^
■^R
0
2g
;3 ■^ •-<
H
„
"«
0
t
5 2
0
0
°
<
' ".
a
o„ .
J. 0 ^
- Tf •*
~
f ,
1- N
%\ ".
0
o>
0
*2
^c
t^
o'5
S fo "
^■?.
^
0 vO
0
00
00
«
o\
tS
^ 00 " CC
^ 4
8
8
g
O-OO 0
0
d
C
OC
<
10 '^
V4
0» vO
10 N
T
c
•*
%S
? :^
s
0
00
§3
N N
«
c
Z 0
T
T
I.*
SI
•z
w
p
t^
"c
t
3
^
1
S.
E
d
•d
■^
b
) CA
^
SJ
•c
s
•
■ 0
3 fl
1
w
C/"
.D
a
a;
1 =
j
=
«+
'■d ^
i
«
d M
«i ° b
'+ J
03
•d -d-w -c
E II E
'"2
u b
d £
E 2
0 0
n c
0
a 0 «
c
0 c
w 0
P^ rt
p:
Pi
rt
12
12
z
.g
5
c
1,
.s"
'3
1
§
E
§
0
? =
8 "d
C
0
d
4*
c
c
c
•0
•c
•t:
•d 1
«
a
0
5
0
•a
;
S I
"3 c
"c
.0
"3
til
*" 'c
i
"J
"
4i 0
IJ
i
<-
il
1
c
"d
jj
M OC
>o
,
*■ I
r) vO
0 d
Ez
1
272 Journal of Agricultural Research voi. xvn. No. e
SUMMARY
(i) Mosaic of the Irish potato has become well distributed over the
United States.
(2) It has a decidedly detrimental effect upon yield.
(3) It produces characteristic symptoms upon the aerial parts of the
plant, especially on the foliage. These symptoms may be modified or
obscured by differences in environment or variety.
(4) Tubers of diseased plants carry the disease.
(5) Grafting a healthy scion upon a diseased stock, or a diseased scion
upon a healthy stock, may result in the development of the disease by
the originally healthy scion or stock.
(6) Mosaic may be transmitted by transferring juice from a diseased
plant to a healthy plant.
(7) At least two species of aphids can transmit potato mosaic, whether
the aphids are transferred artificially or disperse naturally,
(8) Mosaic apparently tends to increase the sugar content of the leaves
and to reduce their starch content.
(9) Hill selection has not proved successful for maintaining healthy
stock when practiced in fields having a considerable number of mosaic
plants.
(10) Roguing or eliminating mosaic plants before aphids become abun-
dant is indicated indirectly by certain experimental evidence here presented
as being helpful and also has been found actually efficient for checking the
spread of the disease. It appears also that isolation of the rogued seed
plot is very desirable.
LITERATURE CITED
(1) AiLARD, H. A.
191 7. FURTHER STUDIES OP THE MOSAIC DISEASE OP TOBACCO. In JoUT. AgT.
Research, v. 10, no. 12, p. 615-632, pi. 63.
(2) D00UTT1.E, S. P.
1916. A NEW iNTfECTious MOSAIC DISEASE OP CUCUMBER. In Phytopathology,
V. 6, no. 2, p. 145-147.
(3) Giissow, H. T.
1918. OBSERVATIONS OxX OBSCURE POTATO TROUBLES. In Phytopathology, v.
8, no. 9, p. 491-495' ^g- 2-5-
(4) Johnson, G. W.
1847. THE POTATO. ITS CULTURE, USES, and history. i8ip.,front. London.
Reprinted from Gard. Monthly, v. i.
(5) McClintock, J. A., and Smith, L. B.
1918. TRUE nature op SPINACH-BLIGHT AND THE RELATION OP INSECTS TO ITS
TRANSMISSION. In JouT. Agr. Research, v. 14, no. i, p. 1-60, i fig.,
pi. A, i-ii.
(6) Melchers, L. E:
1913. the mosaic disease of the tomato and related plants. in ohio
Nat., V. 13, p. 149-173. I fig-, pl- 7-8. Bibliography, p. 169-173.
Reprinted as Contrib. Bot. Lab. Ohio Univ., no. 74.
Sept. 15, I9I9 Investigations o?i Mosaic Disease of the Irish Potato 273
(7) Melhus, I. E.
1917. NOTES ON MOSAIC SYMPTOMS OF IRISH POTATOES. (Abstract.) In
Phytopathology, v. 7, no. i, p. 71.
(8) MuitPHY, P. A.
1917. THE MOSAIC DISEASE OF" POTATOES. In Agr. Gaz. Canada, v. 4, p. 345-
349, illus.
(9) OrTon, W. a.
1914. POTATO WII.T, LEAF-ROLL, AND RELATED DISEASES. U. S. Dept. Agr. Bul.
64, 48 p., 16 pi. Bibliography, p. 44-48.
(10) Stewart, F. C.
i916. observations on some degenerate strains of potatoes. n. y.
State Agr. Exp. Sta, Bul. 422, p. 319-357, 12 pi.
(11) U. S. Department op Agriculture. Bureau of Plant Industry. Plant
Disease Survey.
1917-18. plant disease bulletin, v. [iJ-2. 1917-1918.
(12) Wortley, E. J.
1915. THE transmission of potato mosaic through the tuber. In vScience,
n. s. V. 42, no. 1043, P- 460-461.
PLATE A
Foliage of Irish potato, Green Mountain variety. Note distinct mottling and
slightly lighter color of diseased leaves on plant at left. Single dark green leaflet
from healthy plant at right. Pres'-ue Isle, Me., 1915.
(274J
Investigations on Mosaic Disease of the Irish Potato
Plate A
Journal of Agricultural Research
Vol. XVII, No. 6
Investigations on Mosaic Disease of the Irish Potato
Plate B
V'l -
Journal of Agricultural Research
Vol. XVII, No. 5
PLATE B
Foliage of potato, Bliss Triumph variety. Note decided crinkling of leaf paren-
chyma on diseased leaf at left. More severely affected than diseased leaf on Plate A.
Healthy leaf of same variety at right. Greenhouse, Washington, D. C, 1919.
FLATE 25
Leaf of Irish potato, Green Mountain variety, infected with mosaic. Medium stage
of disease. Note mottling and crinkling of laminar parenchyma. Specimen taken
from field, Caribou, Me., 1914.
Investigations on Mosaic Disease of the Irlsii Potato
Plate 25
Journal of Agriculturai Research
Vol. XVII, No. 5
Investigations on Mosaic Disease of the Irish Potato
PLATE 26
Journal of Agricultural Research
Vol. XVII, No. 6
PLATE 26
A. — Healthy scion grafted upon diseased stock. Yotmger leaves on scion show
typical mottling. Greenhouse, Washington, D. C, 1916.
B. — An illustration of a method used for introducing aphids. This method resulted
in 89 per cent of infection before the plants matiu'ed. In practice the insect cage
was left in place while the aphid-bearing leaves on the stick were introduced. Green-
house, Orono, Me., April, 19 19.
PLATE 27
A. — Leaves from graft shown in B, of this plate: At right, from healthy parent of
scion; at left, from mosaic stock; in center, from mosaic scion.
B. — At left, healthy scion grafted to diseased stock, Green Mountain variety; at
right, two mosaic shoots of stock. Grafted July 6, 1918. Scion decidedly mottled
August 17, 1918. In field, Presque Isle, Me., 1918.
C. — Leaves from corresponding parts of the plants shown in Plate 29, B. These
leaves were near the tops of the plants and matured long after all aphids had been
removed. The characteristic mottling was obscured by the use of reflected light,
but the contrast in the evenness of the leaf surfaces is evident. Greenhouse,
Orono, Me., 1919.
Investigations on Mosaic Disease of tlie Irish Potato
Plate 27
«i .•,
■^W*
t".^.
■'^.*t"
N.
'X
^^j^
Journal of Agricultural Researcii
Vol. XVII, No. 6
Investigations on Mosaic Disease of the Irish Potato
Plate 28
CD
Journal of Agricultural Research
Vol. XVII, No. 5
PLATE 28
A. — 491X, inoculated artificially with unfiltered juices from mosaic plant February
22 to March 22, 1919. Note mottled leaves on the two new shoots at apex of plant.
Green Mountain variety. 49iy, control, untreated plant from half of same tuber as
491X. Greenhouse, Washington, D. C, 1919.
B. — 473y, inoculated in same way as 49 ix, but with juices from healthy plant.
48sy, also inoculated with juices from healthy plant. Greenhouse, Washington, D. C.
122502°— 19 3
PLATE 29
A. — Mosaic of potato transmitted by aphids. 142a, infected plant, Green Mountain
variety. Plant developed beside a badly mosaic plant, thus allowing free infection
by the aphids. Upper leaves distinctly mottled and crinkled. Lower leaves without
mottling. 142b, healthy plant from same tuber as 142a. Greenhouse, Washington,
D. C, March 14, 1918.
B. — Two plants from the same tuber treated alike, except that about 200 aphids
were introduced upon one when it was 2 inches high. Photographed 46 days after
the introduction of aphids and 26 days after the first signs of mosaic were shown.
Greenhouse, Orono, Me., April, 1919.
Investigations on {Mosaic Disease of the Irish Potato
Plate 29
Journal of Agricultural Research
Vol. XVII. No.6
Investigations on Mosaic Disease of the Irish Potato
Plate 30
■^•>*v
Journal of Agricultural Research
Vol. XVII, No. 5
PLATE 30
A. — Inoculated by means of artificial transfers of aphids from diseased plants-
Green Mountain variety. Transfers made February 26, March i, and March 15, 1919.
Distinct mottling and crinkling of younger leaves noted April i, 1919. Green-
house, Washington, D. C.
B. — Plants inoculated in same way as those in A of this plate, but with aphids
taken from healthy plants. No mosaic April i, 1918. Greenhouse, Washington,
D. C.
TEMPERATURE IN RELATION TO QUALITY OF SWEET-
CORN
By Neil E. Stevens, Pathologist, Fruit Disease Investigations, Bureau of Plant
Industry, United States Department of Agriculture, and C. H. HiGGiNS, Instructor in
Chemistry, Bates College, Lewiston,Me}
INTRODUCTION
The temperature at which green sweetcorn {Zca mays) is held after
picking has an important relation to its quality. Certain features of
this relation are discussed in the present paper. That sweetcorn canned
near the northern limit of its cultivation is sweeter and its general
quality superior to that canned farther south seems to be generally
accepted (9, p. 24^; 10, p. 36)? The correctness of this belief is attested
by the fact that it has been customary for many corn growers in Mary-
land, for example, to purchase northern-grown seed in the belief that a
sweeter corn would thus be obtained {10, p. 31), and by the reputation
of "Maine sweetcorn."
That any difference in the quality of the canned corn is not due to a
difference in the sugar content of the corn when it is picked seems fully
proved by the investigations of Straughn and Church {14). These
investigators determined the sugar content of freshly picked corn of the
same variety at a series of stations located in Florida, South Carolina,
New Jersey, Connecticut, and Maine during the four years from 1905 to
1908. In contrast to the condition found in sugar beets, this work
failed to show any direct relation between the latitude in which the
com was grown and the sugar content. Corn grow^n in South Carolina
showed the highest percentage of sugar, that grown in Connecticut the
lowest, that from Maryland and Maine (Crosby variety) intermediate
and about equal {14, p. 62)}
The writers believe that the advantage of northern-packed corn lies,
at least in part, in the lower temperatures at which it is handled, and the
present paper aims to present the follov>'ing salient points in this con-
nection: (i) That sweetcorn deteriorates very rapidly after it is picked,
(2) that the rate of this deterioration depends upon temperature, and
(3) that the differences in climatic temperatures, and consequently in the
' The work on which the present paper is based was done while the writers were investigating the
diseases of sweetcorn in Maine, through the courtesy and at the expense of the Office of Cereal Investiga-
tions. Bureau of Plant Industry, United States Department of Agriculture.
2 Reference is made by number (italic) to "Literature cited," p. 2S3-284.
3 The cur\-es published by Straughn and Church (p. 59-60) are somewhat misleading, since, as ex-
plained in the text, results from analyses of both Crosby and Stowell varieties are included for Maryland
and only the Crosby variety, having a higher sugar content than the Stowell, was grown in Maine.
Journal of Agricultural Research, Vol. X\^I, No. 6
Washington, D. C. Sept. 15, 1919
si KeyNo. G-178
(275)
276
Journal of Agricultural Research
Vol. XVII, No. 6
temperatures at which the corn remains after picking, are sufficient to
cause marked differences in the rate of deterioration. "Whether the
main thesis be accepted or not, the data presented are sufficient to
indicate that a close relation exists between the quality of sweetcom
and the temperature at which it is handled.
LOSS IN SUGAR AFTER PICKING
That green corn deteriorates rapidly after picking is a matter of
common observation, and that an important factor in this deterioration
is the loss in sugar has been pointed out by Straughn, Appleman, and
others. Straughn, working with Stowell's Evergreen in Maryland, re-
ports (/J, p. 6g) that in freshly pulled samples 4.59 to 4.74 per cent
total sugars were found. On standing 24 hours at room temperature,
unhusked, about one-third of the sugars disappeared; after this the loss
continued until the sugars reached 1.80 per cent. More recently, Apple-
man and Arthur (2, Table HI), working with the same variety stored
at accurately controlled temperatures, report that at 20° C. more than
25 per cent of the total sugar was lost during the first 24 hours after
picking, and that at 30° C. more than 50 per cent of the total sugar was
lost in the same period. Analyses of Golden Bantam com made at
Lewiston, Me., during 191 8 showed rapid loss in sugars in stored corn.
The ears were split lengthwise and a sample from one half analyzed
immediately, while the other half was stored.
TablS I. — Total sugars in Golden Bantam corn in edible condition, calculated as per-
centage of invert sugar on original wet weight, Lewiston, Me}
Ear No.
Percentage
Percentage
of sugars
of sugars
in fresh
in stored
half.
half.
5.36
2-39
4.40
2.79
5-84
2.72
5-94
2. 46
Approxi-
mate
number
of hours
stored at
20° C.
6.
9-
10
24
20
48
48
I The method used was essentially that of Bryan, Given, and Straughn as modified by Hassclbring
and Hawkins. The total sugars were calculated as invert sugar by the methods of Mimson and Walker.
Bryan, A. H., GrvEN, A., and Straughn, M. N. extraction of grains and cattle foods for the
DETERMINATION OF SUGARS ... U. S. Dept. Agr. Bur. Chem. Circ. 71, 14 p., 1911; Hasselbring, Heinrich,
and Hawkins, Lon A. physiolooicai. changes in sweet potatoes during storage. In Jour. Agr.
Research, v. 3, no. 4, p. 335, 1915; Wiley, H. W., ed. offici.vl and provisional methods of analysis ...
U. S. Dept. Agr. Bur. Chem. Bui. 107 (rev.), p. 241, 1908.
RELATION OF TEMPERATURE TO RATE OF SUGAR LOSS
The recently published careful researches of Appleman and Arthur (2)
explain the earlier and somewhat conflicting statements of Straughn,
Church, and Wiley (ij, 14) and leave no doubt that the rate of loss of
sugar in stored sweetcorn is directly dependent on temperature. Apple-
Sept. 15, 1919 Temperature in Relation to Quality of Sweetcorn
277
man and Arthur summarize two years' work on Stowell's Evergreen
stored at seven carefully controlled temperatures, namely, 0°, 5°, 10°,
15°, 20°, 30°, and 40° C, as follows:
In general, It may be stated that up to 30° C. the rate of sugar loss in green com
is doubled for every increase of 10°. This applies to both total sugars and' sucrose.
It should be noted, however, that between 0° and 10° the temperature coefficient
for sucrose is considerably greater than 2.
Before the conclusions of Appleman and Arthur were available, the
writers made a few tests to determine whether temperature influenced the
rate of sugar loss in sweetcorn. Their results agree closely with his, but
since the work was done on another variety grown at a considerable dis-
tance the data secured may still be of sufficient interest to warrant pub-
lication. Freshly picked ears of Early Bantam corn in edible condition
were split lengthwise. One half was placed in a small refrigerator which
maintained a temperature of approximately 10° C. and the other half
placed in a box at room temperature, about 20°. Determinations made
at the end of 26 to 30 hours showed uniformly a lower sugar content in
the half kept at the higher temperature. In most cases the ears were
too small to make three satisfactory samples, so no data are available for
the original sugar content of the ears used. The freshly picked ears listed
in Table I were, however, of the same variety, grown in the same plot,
and picked at the same stage of maturity as those shown in Table II;
and if the ears used in the keeping test had about the same average sugar
content as those in Table I (5.38 per cent of wet weight) then the halves
kept at 20° lost, during the first day after picking, more than twice as
much sugar {2,.2)(> per cent) as the halves kept at 10° (1.41 per cent).
Table II. — Total sugars in Golden Bantam corn in edible condition, calculated as per-
centage of invert sugar on original -wet weight, Lewiston, Me., September, igi8
Ear No.
Percentage of
sugar remain-
ing in half
stored at 20° C.
Percentage of
sugar raniain-
ingin half
stored at io° C.
II . .
2.43
I. 90
2.28
I. 90
1.86
1.78
4. 06
12
3-14
5-54
3-71
3.18
4.21
I^
14.
16
17 . ,
Average
2. 02
3-97
Further evidence that the rate of vital activities of green sweetcorn
varies with temperature is afforded by tests of the rate of respiration.
The curv^es of respiratory intensity of sweetcorn during storage published
by Appleman (j, p. 20 j) show that the rate of respiration is very high
during the first day after the corn is pulled from the stalk but falls off
rapidly with storage. They clearly show also that throughout a storage
278
Journal of Agricultural Research
Vol. XVII, No. 6
period of nine days respiration continued more rapidly at 30° C. than at
25°. In storage tests made by the present writers during 191 8, 48 ears of
freshly picked green corn were placed in an air-tight can which had a
capacity of 46.6 liters. The corn displaced somewhat more than 20
liters, leaving about 26 liters of air. The can was then sealed, and loo-cc.
samples of the air were withdrawn at intervals through a stopcock and
analyzed by means of a commercial Orsat apparatus. In the results
thus obtained (see Table III) some error was caused by replacing with
fresh air that withdrawn for analysis; and the temperature of the cool
samples rose slowly, while that of the warmer sample fell somewhat
during the test. The differences, however, are far too great to leave any
doubt as to the facts. With corn at a temperature of 25° (picked near
noon on a warm day) there was over 19 per cent carbon dioxid at the
end of 4 hours. With corn at a temperature of 15° (picked in the morn-
ing) 8 hours were required to reach practically the same point, while
with still cooler corn the point was not passed in 10 hours.*
Table; III. — Oxygen and carbon-dioxid content of air in which green sweetcorn had been
stored in a sealed container
Tem-
pera-
ture
of
corn.
Content of air.
Number of hours after com was sealed.
I
2
3
4
5
6
7
8
9
10
°C.
6
10
3-6
15-4
0. 2
10
19.8
fOxvEfen
16. 0
3-2
\2. 2
7.0
12.8
5- 2
8.0
12. 0
9-5
8.3
3-2
16. 0
7.8
10. 4
I. 0
19.8
6.0
13.0
. 2
22.7
2. I
17.8
0.4
19. 6
0. 0
22.8
15
\Carhon dioxid
2S
l_Carbon dioxid
TEMPERATURE OF GREEN CORN IN RELATION TO AIR TEMPERATURE
In view of the rapidity with which green corn loses its sugar and the
relation of this loss to temperature, it is apparent that if the temperature
of the corn itself is near that of the air there must be a difference in the
extent of deterioration which would occur during a given interval in
different localities and that this difference must correspond to the dif-
ferences in climatic temperatures. Observations in Maryland and in
Maine indicate that the temperature of green corn on the stalk in the
shade is usually near that of the air while in the sun it is often well above
that of the air (see Table IV),
It will be noted that in the cases cited in the table, which are typi-
cal of several others, the corn was 10° or more than 10° C, warmer in
the afternoon than in the morning.^
1 It may be of interest to note that the corn which had been kept for some time in an atmosphere defi-
cient in oxygen was of extremely poor quality.
2 Compare in this connection the condition reported in small fruits (72).
Sept. 15. 1919 Temperature in Relation to Quality of Sweetcorn 279
Table IV. — Temperature (°C) of ears of green sweetcorn on the stalk on clear days
Glencoe, Md.,* Aug. 6,
1918.
Glencoe, Md., Aug. 9,
Iyi8.
Lewiston, Me., Aug. 13,
1918.
Air.
Com.
Air.
Corn.
Air.
Com.
a. m.
6
23. 6
24. 6
26. 0
29- 5
32.0
ZZ- 7
34-5
24.7
24.7
25.0
28.7
31.6
32.8
33- 5
19.9
21-5
21.3
28. I
30. I
31-5
32.5
33-1
33-7
33-8
33-5
27. 8
19. 6
22. 0
26. 7
31. 2
35- 0
36.6
36.7
35-3
35- 6
34-2
33- 5
30-3
16. 0
17.0
18.5
18.5
19. 0
20. 0
21. 0
22. 5
23.0
22. 5
24. 0
24. 0
23-5
22. 0
19. 0
15-3
7
17. 0
8
18.5
22. 0
22. 5
II
23- 5
24. 0
p. m.
25.0
2
34-5
35- 0
35-3
34-5
34-3
34-5
34-7
35- 0
35-6
35-2
26. 0
26.0
26.0
c
27.0
6
23- 5
23. 6
24. 2
22. 0
8
19. 0
1 The temperatures in Maryland were taken by Mr. William E. Seifriz.
TEMPERATURE AT CORN-PICKING TIME IN MARYLAND AND IN MAINE
In attempting to study the temperature of different regions in their
relation to plant growth the investigator must still depend chiefly on
meteorological data taken in cities. Thus Cox (4, p. 10), working on so
highly specialized a crop as the cranberry, in order to compare the differ-
ent regions was forced to use temperature readings observed in shelters
over hard land, even though his own work had shown the great difference
between air temperature over the marshes and air temperature over hard
land.
In comparing the temperature of the corn-canning districts of Mary-
land and Maine, use will be made of the data from the observation
stations of the Weather Bureau at Baltimore and Portland. The sweet-
corn canning district of Maryland extends from Dorchester County
north to Harford County and west to Frederick County. The most
important localities lie north and west of Baltimore. In Maine, corn is
canned commercially from northern York County to southern Penobscot,
the most important localities lying north and west of Portland. It is
probable then that the observations at Baltimore and at Portland furnish
a fairly reliable index of the difference in temperature between the
sweetcorn producing districts of Maryland and Maine. Maryland and
Maine were chosen for comparison because they are the most southerly
and the most northerly of the important corn-canning districts on the
Atlantic seaboard and were among those included in the work of Straughn
28o
Journal of A gricultural Research voi. xvii, no. e
and Church. The exact date on which sweetcorn reaches edible condi-
tion naturally varies somewhat with different seasons;' but corn-canning
time in Maryland almost always falls during August, and in Maine during
September. In order, then, to give some idea of the temperature condi-
tions under which sweetcorn is handled in the two States, it will be neces-
sary to compare the temperature of Baltimore in August with that of
Portland in September.
Table V.
-Daily normal temperatures and corresponding indices for Baltimore, Md.,
August 2 to ji, and Portland, Me., September I to JO
Daily normal tempera-
tures.
Remainder indices
Exponential indices.
Physiological indices.
Baltimore.
Portland.
Baltimore.
Portland.
Baltimore.
Portland.
Baltimore.
Portland.
" F.
" F.
° F.
" F.
76
64
37
25
4. 0000
2. 5198
82.333
30. 000
76
63
37
24
4. 0000
2.4245
82. 333
27. Ill
76
63
37
24
4. 0000
2. 4245
82. 333
27. Ill
76
63
37
24
4. 0000
2.4245
82.333
27. Ill
76
63
37
24
4. 0000
2- 4245
83- 333
27. Ill
76
62
37
23
4. 0000
2-3331
82.333
24. 333
76
62
37
23
4. 0000
2-3331
82. 333
24. 333
76
62
37
23
4. 0000
2-3331
82. 333
24- 333
76
62
37
23
4. 0000
2.3331
83- 333
24. 333
76
61
37
22
4. 0000
2.2451
82.333
22. 000
75
61
36
22
3. 8480
2.2451
78. Ill
22. 000
75
61
36
22
^. 8480
2. 2451
78. Ill
22. 000
75
60
36
21
3. 8480
2. 1603
78. Ill
19. 883
75
60
36
21
3. 8480
2. 1603
78. Ill
19- 883
75
60
36
21
3. 8480
2. 1603
78. Ill
19. 883
75
60
36
21
3. 8480
2. 1603
78. Ill
19. 883
75
59
36
20
3. 8480
2. 0786
78. Ill
17.778
74
59
35
20
3- 7024
2. 0786
73- 667
17.778
74
59
35
20
3. 7024
2. 0786
73- 667
17. 118
74
58
35
19
3- 7024
2. 0000
73. 667
16. Ill
74
58
35
19
3. 7024
2. 0000
73. 667
16. Ill
74
58
35
19
3. 7024
2. 0000
73- 667
16. Ill
74*
57
35
18
3. 7024
I. 9240
73- 667
14.444
73
57
34
18
3- 5629
I. 9240
69. 000
14.444
73
57
34
18
3- 5629
I. 9240
69. 000
14. 444
73
56
34
17
3- 5629
1.8512
69. 000
12.778
73
56
34
17
3- 5629
I. 8512
69. 000
12.778
73
56
34
17
3- 5629
1.8512
69. 000
12.778
72
55
33
16
3- 4283
1.7815
65- 333
II. 667
72
55
33
16
3-4283
1.7815
65-333
II. 667
Av. 74. 6
59-5
35-6
20. 5
3- 7940
2. 13504
76. 2591
19. 5992
Table V gives the daily normal mean temperatures of Baltimore, Md.,
from August 2 to 31 and of Portland, Me., from September i to 30, with
three sets of corresponding temperature efficiency indices. The normal
mean temperatures are those calculated by Bigelow (j) from obser\^ed
temperatures. That mean temperatures furnish only a very unsatis-
1 The harvest dates given by Straughn and Church (14) are: for Maryland, first week in August, 1905:
about the first of August, 1906; about Aug. 15, 1907; and Aug. 23, 1908; for Maine, about Sept. 15, 1505;
Sept. 25, 1906; frost before crop matured, 1907; and Sept. 19, 1908.
Sept. IS, 1919 Temperature in Relation to Quality of Sweetcorn 281
factory basis for estimating the temperature value for physiological
processes of a given climate has long been recognized, and the three sets
of indices represent three suggested methods of deriving from mean
temperatures some index which would more nearly represent temperature
efficiency.
Remainder indices, derived by subtracting a constant quantity (in
this case 39) from each daily mean temperature, have been in use for
a considerable time. The other methods were suggested recently by
Livingston (7, 8), and all three are fully discussed by him in the papers
cited.*
The exponential system is based on the supposition that plant growth
rates follow the chemical principle of van't Hoff and Arrhenius, which
states that the velocities of chemical reactions about double with each
increase in temperature of 10° C. The physiological indices were cal-
culated from actual temperature values for the growth of corn (maize)
seedlings from 10 to 12 mm. high, as worked out by Lehenbauer (6). In
view of Appleman and Arthur's conclusion (2) that the average tempera-
ture coefficient of sugar depletion in sweetcorn is about 2, Livingston's
"Exponential Indices" based on a coefficient of 2 are of special interest.
The degree of accuracy with which any index derived from mean
daily temperatures — }4 (maximum -{-minimum) — expresses the tempera-
ture of the day must depend somewhat on the daily temperature range
and on the shape of the curve of hourly temperatures. In figure i are
plotted the curves of normal hourly temperatures for August at Balti-
more, as published by Fassig (5, p. 61), and the mean hourly tempera-
ture at Portland for September, 1918.' It will be obseived that the
curves are of the same general shape and that the daily ranges of tempera-
ture are similar.
The curves of mean hourly temperatures shown in figure i furnish a
striking evidence of the difference in the temperatures of the contrasted
regions during the corn-packing season. The highest mean temperature
at Portland, 62.6° F., is 6° below the lowest mean temperature for
Baltimore, 68.6°.
From the purposes of the present paper, however, it is unimportant to
determine which method most nearly represents the actual rate of loss,
since on whatever basis the comparison is made it is evident that the
average day during the corn-packing season in Maryland is much warmer
and therefore much more severe in its effect on sweetcorn than the
average day of the corresponding season in Maine. Deterioration of
corn after picking during a given period would then ordinarily be much
' These three kinds of temperature efficiency indices have been compared by one of the writers in con-
nection with studies of the growth of fungi in relation to temperature (i^).
2 Thenormalhourly temperature for Portland has not been computed. The curve for Portland was pre-
pared from data for the month of September, 1918, kindly furnished the writer by Jlr. Edward P. Jones,
Meteorologist, in charge of the Portland. Me.. Station. This, according to advice from Dr. P. C. Day, Chief
of Climatological Divisiou, of the U. S. Weather Bureau, gives a fairly representative curve.
282
Journal of Agricultural Research voi. xvn, No. e
/^o" ^ ^ & 7 <ff o /e' /yAtxn^/ ^3-:r^sp'<ff&/o ///y//^T
^3.
/
/
/"
\
\
V
/
/
7Z
■2/P
7?^
V
'sy
?
\
\
\
\
V
\
i
/
/
s
\
\
V.
V
/
GO
/
/'
y^
y
«->
\
\
/
/
f
7Z
•
PZ
7/^
f7
\
^.
\
V,
\
^
N
^
^
,_,^
/
/
r
>
s
Fig. I. — Meanhourly temperatures for August at Baltimore, Md., and for September,
1918, at Portland, Me.
Sept. 15. 1919 Temperature in Relation to Quality of Sweetcorn 283
greater in Maryland than in Maine. The original quality and the
methods of handling being equal, corn handled at a mean tempera-
ture of 59.5° F., the mean temperature at Portland in September, must
inevitably be superior to corn handled at 74.6°, the mean tempera-
ture at Baltimore in August.'
SUMMARY
The rate at which sugar is lost increases with rise of temperature at
least up to 20° C.
The rate of respiration also varies with temperature, being greater at
higher temperatures, at least up to 30° C.
Observations in Maryland and in Maine indicate that the temperature
of green corn on the stalk while in the shade is usually near that of the
air, while in the sun it often is above that of the air.
The corn-picking season in Maryland (August) has a much higher
average temperature than the corresponding season (September) in Maine.
The difference is sufficient to cause considerably greater deterioration in
picked corn during a given period.
LITERATURE CITED
(i) Appleman, Charles O.
T918. RESPIRATION AND CATALASE ACTIVITY IN SWEET CORN. In Amer. JoUf.
Bot., V. 5, no. 4, p. 207-209.
(2) ■ and Arthur, John M.
19 19. CARBOHYDRATE METABOLISM IN GREEN SWEET CORN DURING STORAGE AT
DIFFERENT TEMPERATURES. In Jour. Agr. Research, v., 17, no. 4, p.
137-152.
(3) BiGELOW, F. H.
1908. THE DAILY NORMAL TEMPERATURE AND DAILY NORMAL PRECIPITATION OP
THE UNITED STATES. U. S. Dept. Agr. Weather Bur. Bui. R, 186 p.
(4) Cox, Henry.
I918. FROST AND TEMPERATURE CONDITIONS IN THE CRANBERRY MARSHES OF
WISCONSIN. U. S. Dept. Agr. Weather Bur. Bui. T, 121 p., illus.
maps.
(5) Fassig, Oliver Lanard.
1907. REPORT ON THE CLIMATE AND WEATHER OF BALTIMORE AND VICINITY.
In Marjdand Weather Service, v. 2, p. 27-514, 170 fig., 24 pi.
(6) Lehenbauer, Philip Augustus.
I914. GROWTH OF MAIZE SEEDLINGS IN RELATION TO TEMPERATURE. In Physiol.
Researches, v. i, no. 5, p. 247-288, 4 fig. Literature cited, p. 287-288.
(7) Livingston, Burton Edward.
I916. PHYSIOLOGICAL TEMPERATURE INDICES FOR THE STUDY OF PLANT GROWTH
IN RELATION TO CLIMATIC CONDITIONS. In Physiol. Researches, v. i,
no. 8, p. 399-420, 4 fig. Literature cited, p. 420.
(8) and Livingston, Grace Johnson.
I913. TEMPERATURE COEFFICIENTS IN PLANT GEOGR/\PHY AND CLIM.\TOLOGY.
In Bot. Gaz., v. 56, no. 5, p. 349-375. 3 fig-
1 The practical application of the facts here presented in such matters as home canning and handling are
too obvious to need comment. As indicated by Table V, com picked early in the morning is much cooler
and can be handled with much less loss of sugar than that picked later in the day.
284 Journal of Agricultural Research voi. xvii, no. e
(9) Pearl, Raymond, and Surface, Frank M.
1910. EXPERIMENTS IN BREEDING SWEET CORN. Maine Agr. Exp. Sta. Bui.
183, p. 249-316, fig. 221-234.
(10) Stabler, Augustus.
1904. SWEET CORN. BREEDING, GROWING, AND CURING FOR SEED. Md. Agr.
Exp. Sta. Bui. 96, p. 31-43.
(11) Stevens, Neil E.
1917. THE INFLUENCE OP TEMPERATURE ON THE GROWTH OF ENDOTHIA PARA-
SITICA. In Amer. Jour. Bot., v. 4, no. 2, p. 112-118, i fig. Literature
cited, p. 118.
(12) and Wilcox, R. B.
1918. TEMPERATURE OP SMALL FRUITS WHEN PICKED. In Plant World, V. 21,
no. 7, p. 176-183. Literature, cited, p. 183.
(13) Straughn, M. N.
1907. SWEET CORN INVESTIGATIONS. Md. Agr. Exp. Sta. Bui. 120, p. 37-78.
(14) and Church, C. G.
1909. THE influence OF ENVIRONMENT ON THE COMPOSITION OF SWEET CORN
1905-1908. U. S. Dept. Agr. Bur. Chem. Bui. 127, 69 p., 11 fig.
VARIATION OF AYRSHIRE COWvS IN THE QUANTITY
AND FAT CONTENT OF THEIR MILK^
By Raymond Pearl and John Rice Miner
The present paper has for its purpose a biometrical analj'sis of the
normal individual variation in the milk flow and the fat content of the
milk in Ayrshire cattle.
This work has been undertaken because of a strong conviction on the
part of the authors that a fairly comprehensive knowledge of the normal
variation of a character which is to be made the basis of genetic study is
essential if such study is to be critical. This viewpoint is entirely inde-
pendent of any position which one may hold regarding the genetic signifi-
cance of different kinds of variation. As a matter of biological fact one
never deals actually with one sort of variation absolutely free from the
influence or effect of all others. For, even though we may be studying
a discontinuous variation of strictly germinal origin and control, there
will be, in the actual somatic expression of this variation, a superimposed
fluctuating variation of nongerminal origin. The student of genetics
ordinarily, and quite rightly, neglects these superimposed fluctuations
and confines his attention to the underlying germinal variation.
This is logically a perfectly justifiable procedure, but an essential to its
successful operation is that one shall have such an intimate and thorough
knowledge of the normal variability of the character in question that he
can make his rejections of the unimportant with substantial correctness
and hence safety.
These considerations become particularly significant when the char-
acter dealt with is one especially subject to environmental influences, in
consequence of which the fluctuations assume highly significant propor-
tions in relation to the underlying germinal differences. Such characters
are, for example, fecundity, fertility, and, to a marked degree, milk pro-
duction in cattle. Any milk or fat record represents the result of the
action of a complex of factors, of which those classed broadly as environ-
mental certainly play a very important part. To arrive at any sound
conclusions regarding the inheritance of these characters it will be essen-
tial to form some sort of judgment as to the proportionate parts which
genetic and environmental factors play in the production of particular,
individual records. It seems perfectly clear that a prerequisite to
' Papers from the Biological Laboratory of the Maine Agricultural Experiment Station, No. 125.
This work was begun while the authors were actively connected with the Maine Agricultural Experiment
station. It was interrupted by the entry of the United States into the war and has been completed in the
Laboratory of Biometry and Vital Statistics of the School of Hygiene and Public Health of Johns Hopkins
University.
Journal of Agricultural Research, Vol. XVII, No. 6
Washington, D. C. Sept. 15, 1919
sh Key No. Me. -15
(285)
286 Journal of Agricultural Research voi.xvn.No. 6
anything approaching a sound basis for such a judgment is a thorough
analytical study, with the best of biometric tools, of the normal varia-
bility of milk and fat production.
MATERIAL FOR INVESTIGATION
The present study is based on the records of Ayshire cattle published
in the Reports of the Ayrshire Cattle Milk Records Committee of Scotland,
compiled by Speir {26) ^ and Howie ^ (6), Portions of the very valuable
records gathered by this committee have been used by other students of
the problems of milk production, notably Wilson {30), Pearson {23), and
most recently Vigor {28). Wilson made use of the 1908 records, and
Vigor those of 1909 for the Fenwick district only.
The reports under consideration include, so far as it is possible to get
the information, the following items :
1. Total milk produced (in gallons).
2. Average percentage of fat, determined from periodic tests.
3. Total milk calculated to a 3 per cent fat basis.
4. Weeks in milk.
5. Age of cow.
6. Date of last calving.
7. Miscellaneous information about the cow, particularly of abnormal
circumstances of any sort during the test.
In many cases information is lacking on some one or more of these
points, so that, while altogether 8,132 cows were tested in 1908 and 9,202
in 1909, nothing like these numbers are available for analytical study.
Another difficulty arises in the fact that there is, of course, much over-
lapping of calendar years by the lactation periods. Again there is in
some districts frequent failure to state the age of the cow.
In the present study all available records from the 1908 and 1909
reports have been used, if they came within the following regulations
which we established in order to secure critical material for variation
study:
(a) The recoid must be complete in all particulars — that is, cover
items I to 6 in the list above.
(6) The record must be based on 32 or more weeks in milk.
(c) There must be nothing of an abnormal or unusual nature about
the cow or the lactation, so far as discoverable from the records.
The first of these restrictions requires no comment.
Regarding the second it may be said that the reason for imposing
this restriction was that, for present purposes, we desired to use long
term averages, rather than to consider lactations of all durations. There
' Reference is made by ntunber (italic) to "Literature cited," p. 320-322.
' It is a great pleasure to acknowledge, with grateful thanks, the kindness of Mr. John Howie, of Ayr,
Scotland, the secretary of the Milk Records Committee, in furnishing a set of the committee's reports for
this investigation.
Sept. IS, I9I9 Variation in Milk of Ayrshire Cows 287
seemed good reason for the belief that one was likely to get better — that
is, more nearly physiologically normal — values for the two characters
here studied — mean fat percentage and mean weekly yield — if one con-
sidered only lactations eight months or more long. Furthermore it is
clear that no error of any consequence can be introduced by leaving out
of account short lactations, since Vigor {28) has shown that there is no
significant net correlation between duration of lactation and either
percentage of fat or average weekly yield of milk, the two characters
studied in this investigation.
The third restriction obviously needs no argument in its justification.
Under it were excluded cases of abortion, "off -food" at particular tests,
diseases, and accidents of various sorts. Undoubtedly some records
were excluded which might fairly have been regarded as normal; but it
was thought best, where one was working entirely from records and could
not see the cow itself, to err, if at all, on the side of too great rather than
too little strictness.
The two characters dealt with in this paper are (a) average milk yield
per week in gallons, and (6) average fat percentage. The values for the
former were obtained by dividing the total yields as given in the reports
by the weeks in milk. The fat percentage figures were taken directly
from the reports. The ages were taken as centering at the mid-point of
each year. For example, all cows recorded as 3 years or more in age but
less than 4 years were put in the 3-year class in the tables of the present
paper. Hence a 3-year-old is to be taken as including anything between
3 and 4 years.
The biometric methods used were the ordinary ones. All of the dis-
tributions containing enough individuals to make the results significant
were fitted with Pearson's {18, 20) skew frequency curves, following in
the computations some simplifications of method.
FREQUENCY DISTRIBUTIONS
The frequency distributions, showing the variation in the two char-
acters studied, are exhibited in Tables I and II, in both absolute and per-
centage figures.
122502°— 19 4
288
Journal of Agricultural Research voi. x\ai. no. e
Table I. — Frequency distributions for variations in average weekly milk yield of Ayrshire
cows of different ages
2-year-old cows.
3-year-old cows.
Yield (in
gallons).
1908.
1909.
Combined
years.
1908.
1909.
Combined
years.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
2
3
5
5
3
18
19
21
36
47
36
66
64
70
60
73
51
61
44
41
24
23
13
17
3
0. 24
•36
.61
.61
•36
2.18
2.30
2-55
4-37
S-70
4-37
8.00
7-76
8.48
7-27
8.85
6.18
7-39
S-33
A- 91
2.91
2.79
.85
1.58
2. 06
•85
•36
.48
2
6
7
7
5
24
28
35
56
68
70
107
118
124
119
133
87
102
78
76
43
43
28
20
22
14
5
6
3
2
0. 14
.41
.48
.48
•35
1.67
1-94
2-43
3-89
4.72
4.86
I
2.86
I
1. 16
3
2
2
2
6
9
14
20
21
34
41
54
54
59
60
36
41
34
35
19
20
21
7
5
7
2
2
3
I
I
0.49
•32
•32
•32
.98
1.46
2.27
3-25
3-41
S-52
6.66
8.77
8.77
9-s8
9-74
5-84
6.66
5-52
5.68
3- 08
3-2S
3-41
1. 14
.81
I- 14
•32
•32
•49
.16
.16
8.00
8.50
I
2.86
2
2
2
2
2
1
7
3
I
4
5
2
I
5
2
3
3
I
3-92
3-92
3-92
3-92
3-92
I. 96
13-73
S-88
I. 96
7-85
9.81
3-92
1.96
9.81
3-92
5-88
5-88
1.96
3-92
3
2
2
3
2
2
7
7
3
10
9
5
2
6
S
S
5
2
3
3-49
2-33
2-33
3-49
2-3i
2-33
8.13
8.13
3-49
11-63
10.47
5-8x
2-33
6.98
5-81
S-8i
5- 81
2- a
3-49
I
2.86
I
2.86
4
2
6
4
3
I
I
3
2
2
I
I
11-43
5-71
17-13
11-43
8.57
2.86
2.86
8-57
5- 71
5-71
2.86
2.86
7-43
8.19
8.61
8.26
9-23
6.04
7.08
5-41
2.98
2.98
17.50
18.00
18.S0
I
1.96
1
I
I. 16
I. 16
I
2.86
:::::::i::::::;
21.00
. 12
I
.16
X
i . . . .
1
Total
35
51 1
86 j
616
82s
1,441
Sept. 15, 1919
Variation in Milk of Ayrshire Cows
289
TablB I. — Frequency distributions for variation in average weekly milk yield of Ayrshire
cows of different ages — Continued
Yield (in
gallons).
S-oo. .
SSo- •
6.00. .
7.00. .
7-SO- •
8.00. .
8.S0. .
9.00. .
9SO-
10.00.
10.50.
11.00.
11.50.
12.00.
12.50-
1300.
13.50.
14.00 .
14-50.
15-00.
IS-SO.
16.00.
16.50.
17.00.
17-50-
18.00.
18.50.
19.00.
19. so.
20.00.
20.50.
21.00.
21.50.
22.00.
22.50.
23.00.
23 SO.
24.00.
2450.
25.00.
25-50.
37.00.
39.00.
Total.
4-year-old cows.
1908.
Fre-
quen-
cy.
526
Per-
cent-
age.
•38
.76
•38
1. 14
I. 14
1. 90
3-04
s- 70
4-37
5- 13
4-75
5-51
7.04
7.80
8-75
5- 70
5-9°
6-60
4- 18
5-71
3- 80
2.28
2.47
1-33
• 19
Fre-
quen-
cy.
Per-
cent-
age.
■17
• 17
5- 40
7. 60
8. II
9. 12
6- 59
7-43
S-07
7.09
5-4°
3-38
2.87
2-53
2-03
I. 18
2-53
•51
1-52
Combined
years.
Fre-
quen-
cy.
1,118
Per-
cent-
age.
• 27
•71
•45
3- 76
4- 74
5-28
6.26
6.89
8. 14
7. 16
8-05
5-37
6-53
5-99
3-76
5-year-old cows.
1908.
Fre-
quen-
cy.
Per-
cent-
age.
I- 19
•71
1. 90
2. 14
4.04
4.28
5-23
7-36
7-84
8.79
8-08
8.31
4.51
3- 80
7- 13
5-46
3-80
4.28
2.38
1.42
I- 19
I- 19
•47
• 71
•47
•47
•47
Fre-
quen-
cy.
Per-
cent-
age.
.41
. 20
•41
1-85
1.44
2.0s
2. 67
3-(>9
3-48
5-94
9-63
5-94
8.20
9. 02
6-35
7.17
6.15
4-30
4.92
2. 46
3- 69
2- 46
2. 26
1. 64
•41
.62
.82
■41
Combined
years.
Fre-
quen-
cy.
Per-
cent-
age.
•33
•33
•44
1-54
I. 10
1.98
2.43
3^8s
3^85
5-61
8.59
6.83
8.44
8.59
7.27
$•94
5- 06
561
517
3- 08
3- 96
2.43
1.87
1-43
•77
■SS
•77
•44
■33
.44
290
Journal of Agricultural Research voi. xvn. no.6
Table I. — Frequency distributions for variation in average weekly milk yield of Ayrshire
cows of different ages — Continued
Yield (in
gallons).
8.00. .
8.50. .
9.00. .
9-5° ■
10.00.
10. so.
11.00.
11.50.
la.oo.
13.50.
13-00.
13-50.
14.00.
14-50.
15.00.
»SSO.
16.00.
16.50.
17.00.
17-50.
18.00.
18.50.
19.00.
19-50.
30.00.
30.50.
31.00.
21.50.
32. 00.
32.50.
33.00.
23-50.
24.00.
34-50.
35.00.
35-50-
36.00.
36.50.
37.00.
39.00.
30.50-
6-year-old cows.
1908.
Fre-
quen-
cy.
per-
cent-
age.
o- 61
•31
•31
•31
.61
•31
6. 12
9- 17
7- 6s
4.89
5. 20
6. 12
7-95
7-34
3-36
3-98
2.44
1-83
2.44
•31
1-83
.92
•31
.61
•31
•31
•31
Total.
Fre-
quen-
cy.
Per-
cent-
age.
4- 19
S-03
6.29
5-66
S-45
9.64
5-03
6.92
7-34
S-24
4.61
4.40
4.40
4- 61
3-97
I- 26
2-52
I- 68
1-05
Combined
years.
Fre-
quen-
cy.
804
Per
cent-
age.
•50
•2S
I. 00
1.99
1.99
3- 61
4-23
S-72
S-3S
4.98
8.21
6. 72
7.21
6-34
5.22
5-32
S-8S
5.60
4. 10
3-98
1.74
2.24
1.99
•75
•87
.62
•50
•75
•37
•^S
7-year-old cows.
1908.
Fre-
quen-
cy.
316
Per-
cent-
age.
I. 26
2-53
2- 22
3-16
4-43
4- 75
6-65
7- 28
7^59
3- S3
8.54
4^43
6-33
S-70
S-38
4-75
2-85
3-48
4. II
.63
I-S8
1-58
1.58
•95
Fre-
quen-
cy.
396
Per-
cent-
age.
0. 25
•25
•25
.76
•2S
1. 01
1.76
2. 27
5- 30
5-56
6-82
S-8i
7-07
9.60
7-83
S-S6
6^57
6.31
4.80
3-54
4- 55
2- 02
1- SI
2- 27
I- 26
I- 52
■• 25
.76
Combined
years.
Fre-
quen-
cy.
Per-
cent-
age.
.14
• 14
.38
.84
.28
1.54
1.54
239
2. 12
4-35
5- 06
5-92
6. 18
7- 16
8.71
$•48
6.88
5- 6a
6.33
5. 20
4-3S
4-63
2-39
2-39
3- 09
.70
I. 26
.70
1-40
X- 36
.14
.98
Sept. IS, 1919
Variation in Milk of Ayrshire Cows
291
TablS I. — Frequency distributions for variation in average weekly milk yield of Ayrshire
cows of different ages — Continued
Yield (in
gallons).
8.S0. .
9SO. .
11.00.
11.50.
12.00.
12.50.
1300.
1350.
14.00.
14.50.
15.00.
15-50.
16.00.
16.50.
17.00.
17-50.
18.00.
18.50.
19.00.
19-50-
20.00.
ao.50.
21.00.
21.50.
22.00.
32.50.
23.00.
23-So.
34.00.
34-50-
35.00.
35-50.
36.00.
36.50.
37.00.
27.50.
28.50.
39.00.
8-year-old cows.
Total.
1908.
Fre-
quen-
cy-
Per.
cent-
age.
0.37
•74
•37
•37
4-45
i-8s
4-45
3-33
8- IS
5-56
8-52
4-81
8. IS
7.41
5- 19
8. IS
7-04
2- 22
2. 22
2. 22
2. 22
I. II
•37
I. II
Fre-
quen-
cy.
367
Per-
cent-
age.
0. 27
•55
2. 18
•27
•27
1. 09
4-09
2- 73
3-82
6.81
4.90
5-45
7.90
7-63
6.81
7.08
5-45
4.09
4-63
5-99
4-63
2- 18
1- 64
2- 73
I. 64
1-36
•27
Combined
years.
Fre-
quen-
cy
637
9-year-old cows.
1908.
Per-
cent-
age.
0. 16
•31
• 16
•47
1. 41
•78
I. 10
1. 26
4- 24
2- 35
4-08
5-34
6.28
5-49
8. 16
6-44
7-38
7-22
5-34
5-81
5- 6s
4.40
3- 61
2. 20
1.88
2. 04
I- 57
I- 57
•63
I. 10
. 16
Fre-
quen-
cy.
Per-
cent-
age.
0.47
•95
•47
•47
1.89
2.83
2- 36
3-30
3- 30
4- 72
6.60
5-66
II. 79
5-66
7-55
6-60
2-83
6.13
6.60
3- 30
3-30
4-25
1.89
2-36
■95
•47
•95
•47
•47
•47
•47
Fre-
quen-
cy.
Per-
cent-
age.
o. 40
.40
2.84
2.84
2-43
5-67
4-05
5-67
5- 26
9-31
6.48
6.48
6. 07
4.86
6.48
2.84
4-05
6-48
4-45
2.84
Combined
years.
Fre-
quen-
cy.
Per-
cent-
age.
0-44
•6s
•44
•6s
•44
2. 40
2.83
2. 40
4- 5a
3- 70
5-23
5-88
7-63
8-93
6. 10
6. 75
5-66
4-79
4-36
5-33
S-oi
3-93
2. 61
1.96
2- 6t
1-74
•44
•6s
87
• 33
• 33
292
Journal of Agricultural Research voi. xvii. No. 6
TablB I. — Frequency distributions for variation in average weekly milk yield of Ayrshire
cowsof different ages — Continued
Yield (in
gallons).
7.00. .
ISO- .
8.00. .
9.00. .
10.00.
11.00.
11.50.
12.00.
12.50.
13.00.
13.50.
14.00.
14.50.
15.00.
15.50.
16.00.
16.50.
17.00.
17.50.
18.00.
18.50.
19.00.
19.50.
20.00.
20.50.
21.00.
21.50.
22.00.
22.50.
23.00.
23.50.
24.00.
24.50.
25.00.
2550.
26.00.
26.50.
37.00.
38.00.
39.00.
Total.
10-year-old cows.
1908.
Fre-
quen-
cy.
Per-
cent-
age.
2.38
1-59
3-97
3- 17
• 79
3-97
9- 53
7-94
7-94
5-56
3-97
3-97
8.73
3- 17
4- 77
5-56
5-56
3-97
3- 17
2.38
• 79
•79
Fre-
quen-
cy.
Per-
cent-
age.
4. 12
3- 09
5-67
8.2s
4.64
6. 19
9- 79
3-61
6. 19
4.64
5-15
6. 19
4.64
4.64
3-09
4. 12
2. 06
I. 03
•52
•52
1.03
I- 03
Combined
years.
Fre-
quen-
cy.
Per-
cent-
age.
0.31
■31
.94
• 31
• 31
•94
2.81
1.88
2.50
3-75
2. 19
5- 00
8.75
5-94
6.88
8. 12
3-75
5-31
6. 25
4-38
5.62
5.00
5.00
3-44
3- 75
2. 19
•94
.62
•31
n-year-old cows.
1908.
Fre-
quen-
cy.
Per-
cent-
age.
3- 19
2. 13
2- 13
4. 26
4. 26
7-45
4. 26
6.39
3- 19
8.51
8.51
5-32
6.39
6-39
I. 06
I. 06
3- 19
2- 13
1. 06
2. 13
I. 06
Fre-
quen-
cy.
Per-
cent-
age.
•93
•93
3- 70
1.85
1.85
9. 26
5-55
3- 70
8-33
7.41
8-33
3- 70
5-55
7.41
6.48
3- 70
4-63
2. 78
2.78
I. 8;
"l.'Ss
I. 8s
•93
•93
•93
Combined
years.
Fre-
quen-
cy.
Per-
cent-
age.
.50
• 99
•99
• 50
•99
•99
■50
3-46
2.47
1.98
5-94
4-95
3- 96
7.92
5-94
7-43
3-46
6.93
7.92
5-94
4-95
5-44
1.98
1.98
2.47
•99
1.48
1.98
•99
1.48
• SO
•50
Sept. IS, 1919
Variation in Milk of Ayrshire Cows
293
Table I . — Frequency distributions for variation in average weekly milk yield of Ayrshire
cows of different ages — Continued
12-year-old cows.
13-year-old cows.
Yield (in
gallons).
1908.
1909.
Combined
years.
1908.
1909.
Combined
years.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
I
I
1.52
1-52
I
I
0.86
.86
I
4-76
1. 61
I
2-44
1. 61
I
2.00
I
I
2
2
2
3
6
8
4
6
3
4
I
I
2
3
S
2
3
1-52
1.52
3- 03
3-03
3- 03
4-54
9.09
12. 12
6.06
9.09
4-54
6.06
1.52
I- 52
3-03
4-54
7-s6
3-03
4-54
2
I
2
4
6
6
10
10
8
II
4
5
4
7
4
S
7
3
3
1.72
.86
1.72
3-45
S-I7
S-I7
8.62
8.62
6.90
9-49
3-45
4-31
3-45
6. 04
3-45
4-31
6.04
2-59
2.59
I
4-76
I- 61
2
3
4.88
4.88
3-23
3-23
2
4
3
4
2
4
S
I
I
3
6
2
2
2
I
4.00
8.00
6.00
8.00
4.00
8.00
ID. 00
2.00
2. 00
6. 00
12. 00
4.00
4. 00
4. 00
2.00
3
3
I
3
2
4
2
3
I
3
2
3
I
I
2
7-32
7-32
2.44
7-32
4.88
9-75
4.88
7-32
2-44
7-32
4.88
7-32
2-44
2.44
4.88
6
3
3
2
5
3
I
2
I
I
4.84
8.06
2
9- 52
1. 61
I
4.76
6- 45
3- 23
11.30
6.45
9-68
4.84
4.84
3- 23
8.06
18.00
3
2
3
2
14.29
9-52
14.29
9-52
18.50
2
2
9-52
9-52
4.84
I. 61
3-23
i-6i
3
6.00
2
I
3- 03
1-52
S
I
I
2
2
4-31
.86
.86
1.72
1.72
I
4.76
I
2.44
I -61
I
I
I
2. CO
2.00
2.00
I
I
1-52
1-52
I
4-76
I
I
I. 61
I
2.44
I. 61
38.00
I
2.00
.86
Total
SO
66
116
21
41
62
14-year-old cows.
iS-year-old cows.
Yield (in
gallons).
1908.
1909.
Combined
years.
1908.
1909.
Combined
years.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
I
5-26
I
4-76
I
I
4-55
4-SS
I
I
2.50
2.50
I
S-26
I
4-76
I
4- 55
2
II. II
3
7- so
I
S-26
I
4-76
•'•5
2
I
I
I
2
II. II
5- 56
S-56
S-56
II. II
2
2
I
2
3
3
3
3
3
I
2
a
3
5- 00
5.00
2.50
5- 00
7- 50
7- SO
7- SO
7-50
7- so
2.50
S-00
S-oo
7- SO
I
4- 55
:::::' :::::'■
I
I
3
I
2
3
I
I
I
2
4-55
4-55
13-63
4-55
9.09
13-63
4-5S
4-SS
4-SS
9.09
I
2
I
2
s-26
10.53
s-26
10-53
4-76
9-53
4- 76
14-29
2
t
H. II
5-56
I
50
18.00
18.30
3
I
I
2
I
I
I
IS- 79
s-26
5-26
10.53
S.26
s-26
S-26
14.29
4-76
4-76
9-53
4-76
4- 76
4- 76
I
X
I
5-s6
S-S6
S-S6
I
I
5-s6
S-S6
I
I
I
I
I
2.50
2.50
2.50
2.50
2.50
I
I
4-SS
4-SS
I
so
4-76
I
5-s6
Total
22
18
40
2
19
21
294
Journal of Agricultural Research voi. xvn.No.e
Tables I. — Frequency distributions for variation in average weekly milk yield of Ayrshire
cows of different ages — Continued
i6-year-old cows.
Yield (in gallons).
1908.
1909.
Combined years.
Frequency.
Percentage.
Frequency.
Percentage.
Frequency.
Percentage.
I
I
16.7
16.7
^
12-5
12. 5
I
50.0
12. 5
I
I
I
16.7
16.7
16.7
12.5
12. S
12-5
I
50.0
12. S
I
16.7
12. S
2
6
8
Table II.
-Frequency distributions for variation in percentage of fat in the milk of
Ayrshire cows
-year-old cows
3-year-old cows.
Fat
percentage.
1908.
1909.
Combined
years.
1908.
1909.
Combined
years.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
2
2
5
12
17
19
45
52
57
6S
68
71
43
44
38
19
17
7
16
8
4
I
I
I
I
I
0.33
■33
.81
1-95
2.76
3.08
7-31
8-44
9-25
10.55
11.04
11-53
6.98
7.14
6.17
3.08
2.76
1. 14
2. 60
1-30
•6s
.16
.16
.16
.16
.16
3
10
28
34
57
98
124
143
157
1 58
178
107
109
80
46
37
17
28
10
8
2
2
I
I
I
0. 14
I
=;
16
17
38
53
72
86
92
90
107
64
65
42
27
20
10
12
2
4
I
z
0. 12
.61
1.94
2. 06
4. 61
6. 42
8.73
10.42
II-I5
10.90
12.97
7.76
7.88
5- 09
3- -27
2-43
I. 21
1.46
.24
•49
. 12
. 12
. 21
.69
I
I
6
4
4
6
4
6
7
2
S
3
I
I
1.96
1.96
11-77
7.84
7.84
11-77
7.84
11.77
13-73
3-92
9.80
S-S8
1.96
1.96
I
I
10
5
5
8
12
9-
10
8
8
I
I
1. 16
1. 16
11-63
5- 82
5- 82
9-3°
13-95
10.47
11.63
9-30
9-3°
4-65
3-49
I. 16
I. 16
1.94
2.36
4
I
I
2
8
3
3
6
3
I
2
11-43
2.86
2.86
5-71
22.86
8.57
8.57
17.14
8.57
2.86
5-71
3-96
6.80
8.61
3.6s
g. 92
10. 90
:,.8c:
10. 96
12- 3S
7-43
7-56
5-SS
3-19
2-57
I
2.86
1. 18
A f>Z
1-94
1
.69
!
.56
t
.14
.14
.07
S-4S
.07
0.3s
6.6s
.07
Total
35
SI
86
616
825
I>44I 1
Sept. 15, 1919
Variation in Milk of Ayrshire Cows
295
TablB II. — Frequency distributions for variation in percentage of fat in the milk of Ayr-
shire cows — Continued
Fat
percentage.
2-95-
3-oS.
3-I5.
3-25-
3-35-
3-45-
3-55 ■
3-65.
3.7S-
3-85.
3-95 ■
4-05.
4.15.
4.25-
4-35 •
4-45 .
4-S5-
4.6s-
4-75-
4-95-
5-I5-
6.25.
Total.
4-year-old cows.
Fre-
quen-
cy.
526
Per-
cent-
age.
•38
1. 14
2.85
4-56
8. 75
9-13
II. 22
10. 6s
11. 22
11.41
7.41
6.6s
4. 18
3- 80
1. 14
•95
I. 90
•76
•38
Fre-
quen-
cy.
Per-
cent-
age.
.63
I. iS
1-35
5-41
5-74
8-9S
10.98
12.84
13-34
10.47
9.46
4-73
5-41
4- OS
1.86
I-S2
I- 01
•17
•17
Combined
years.
Fre-
quen-
cy.
S6
So
lOI
124
132
138
122
9S
63
54
44
Per-
cent-
age.
•53
I. 16
2.06
5.01
7-16
9- 03
II. 09
II. 81
12.34
10. 91
8. SO
5^64
4-83
3^94
1-52
1-25
1^43
•45
S-year-old cows.
1908.
Fre-
quen-
cy.
Per-
cent-
age.
3- 09
4.99
4.28
7-36
9- 03
12-35
9. 26
9^74
5^46
5^46
Fre-
quen-
cy.
per-
cent-
age.
1-43
2. 2S
3-69
4.92
9^63
9.02
13-93
12. 70
lo. 66
9^43
8.40
4.92
2.87
1^43
1^43
Combined
years.
Fre-
quen-
cy.
82
120
lol
102
87
74
47
37
23
Per-
cent-
age.
0.44
•55
•99
2. 64
4. 29
4. 62
8.s8
9. 02
13- 21
II- II
II- 22
9- 57
8-14
5-17
4.07
2-53
I- 21
.66
.88
•77
•ii
Fat
percentage.
6-year-old cows.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Combined
years.
Fre-
quen-
cy.
Per-
cent-
age.
7-year-old cows.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
Per-
cent-
age.
Combined
years.
Fre-
quen-
cy.
Per-
cent-
age.
0.31
.61
.61
.92
4.89
5.20
7^65
11-31
13-46
12-84
9.48
9.48
II. 62
4.28
3^37
1^83
.61
.92
0-63
.42
.84
2.09
3^98
7^34
9-43
13-00
II-9S
11-74
10-48
9-43
7-34
3^77
2-94
1. 26
1.47
1.68
. 21
Total.
99
lOI
804
0- so
•50
-75
1- 62
4-35
6-47
8- 70
12-31
12-56
12.18
10-07
9-45
9.08
3^98
3-"
0-3 2
2-53
4- 1 1
6. 33
10- 76
10- 76
12-34
11-39
12-03
9-49
6.6<;
5^70
2-8?
316
396
4
l-OI
6
1-52
23
5-80
20
5- OS
34
8-59
40
10- 10
50
12-63
49
12-37
46
II. 61
45
11-36
27
6-82
24
6.06
36
o- 70
1-97
S-OS
S6a
9-SS
10- 3»
12. SO
11-94
11.80
10- S3
6-74
S-90
2-8x
2- 25
• 84
•99
.14
296
Journal of Agricultural Research voi. xvn. no. 6
Tabl^ II. — Frequency distributions for variation in percentage of fat in the milk of Ayr-
shire cows — Continued
8-year-old cows.
9
-year-old cows
Fat
percentage.
1908.
1909.
Combined
years.
1908.
1909.
Combined
years.
Fre-
quen-
cy.
Per-
cent-
age-
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
I
0.47
I
0. 22
I
2
3
9
8
17
25
35
36
32
32
25
15
16
5
6
I
2
0-37
•74
I. II
2-96
6.30
9. 26
12.97
13-34
11-85
11.8s
9. 26
5-56
5- 92
1-85
2.22
■37
•74
I
0.27
2
2
5
17
26
49
67
74
83
71
68
61
34
36
16
15
4
3
I
2
I
0.31
-31
.78
2.67
4.08
7.69
10. =;2
II. 61
13.09
11. 14
10.67
9-57
5-34
S-65
2.51
2-35
•63
•47
•IS
•31
•IS
2 Kz
I
s
7
8
18
26
30
23
22
20
11
19
12
6
3
•47
2-36
3-30
3-77
8-49
12.26
14- IS
10.8s
10.38
9-44
5-19
8.96
5-66
2-83
1.42
I
0.40
2
5
13
22
31
Si
68
54
59
41
27
35
26
13
5
3
•44
1.09
8
18
32
42
39
47
39
36
36
19
20
II
9
3
I
I
2
I
2.18
4.90
8-72
11-44
10.63
12.80
10- 63
9.81
9.81
S-18
S-4S
3.00
2-45
.82
•27
•27
-55
•27
6
14
13
27
38
31
37
21
16
16
14
7
2
3
2.43
S.67
5. 26
10.93
15-39
12-55
14.97
8-50
6.48
6.48
5-67
2.84
.81
1.22
2.8?
4-79
6- 75
II- SS
14.81
11-77
3.65
12.86
8-91
3 85
5.88
7-63
5-66
2.83
1.09
•6s
A 61:
I
.40
I
.22
270
Total ....
367
637
212
247
459
lo-year-old cows.
11 -year-old cows.
Fat
percentage.
1908.
1909.
Combined
years.
1908.
1909.
Combined
years.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
2 85
I
4
2
9
12
12
20
10
16
17
6
8
5
I
2
0-79
3-17
1-59
7.14
9-53
9-53
15-87
7-94
12. 70
13-49
4-77
6-35
3-96
•79
I- 59
I
2
3
15
17
20
29
27
23
23
13
9
4
6
I
0.52
1.03
I -.54
7-73
8.77
10.31
14-95
13-92
11-85
11-85
6. 70
4.64
2.06
3- 09
•52
2
6
5
24
29
32
49
37
39
40
19
17
9
7
3
0.62
1.87
1.58
7-50
9.06
10.00
15-31
11-56
12. 19
12. SO
5-94
5-31
2.81
2. 19
•94
I
2
3
8
13
17
6
12
7
7
6
4
S
2
1.06
2-13
3-19
8-51
13-83
18.08
6.39
12.76
7-45
7-45
6-39
4-25
5-32
2-13
4
2
9
8
12
13
11
13
12
9
7
3
3
1
I
3-70
I.8S
8.33
7.41
11. II
12.04
10.18
12-04
II. II
8-33
6.48
2.78
2.78
•93
•93
5
4
12
16
25
30
17
25
19
16
13
7
8
3
I
2-47
1.98
S-94
7.92
12.38
14.85
8.41
3.6s
12.38
9.41
3.8s
7.92
6-44
3-46
3- 96
1-48
• SO
I
•52
J
•31
I
1.06
I
■5°
.* 6c
I
•79
I
•31
Total . . .
126
194
320
94
108
202
Sept. 15, 1919
Variation in Milk of Ayrshire Cows
297
Table II. — Frequency distributions for variation in percentage of fat in ike milk of Ayr-
shire cows — Continued
Fat
percentage.
2.85.
2-95 •
30s.
3-iS-
3-35-
3-3S-
3-4S-
3-SS.
3-6s.
3-75 •
3-8s.
3-95 •
4-05.
4-iS.
4-35.
4-45 ■
Total .
12-year-old cows.
1908.
Fre-
quen-
cy.
Per-
cent-
age.
4.00
8.00
8.00
8.00
10.00
14.00
14.00
14.00
8.00
2.00
2.00
Fre-
quen-
cy.
Per-
cent-
age.
Combined
years.
Fre-
quen-
cy.
Per-
cent-
age.
3-45
.86
2-59
6. 04
8.62
II. 20
12-93
12.07
13-79
8.62
6.04
4-31
3-45
3-45
.86
1.72
1 3-y ear-old cows.
1908.
Fre-
quen-
cy.
Per
cent-
age.
4.76
4.76
9.52
33-34
14.29
9-52
9-52
4.76
4.76
4-76
Fre-
quen-
cy.
Per
cent-
age.
14.63
4.88
19-51
21-95
9-75
4.88
7-32
2.44
4.88
Combined
years.
Fre-
quen-
cy.
Per-
cent-
age.
1. 61
6-45
I. 61
9.68
3-23
16. 13
25-80
11.30
6-45
8.06
3-23
4.84
l.6l
14-year-old cows.
15-year-oId cows.
Fat
percentage.
1908.
1909.
Combined
years.
1908.
1909.
q
Combined
years.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
quen-
cy.
Per-
cent-
age.
Fre-
uen-
cy.
Per-
cent-
age.
3-05
3
2
II. II
II. II
5-56
II. 11
5-56
16.66
5-56
16.66
II. II
2
4
3
6
4
6
3
5
5
I
I
5-00
10.00
7-50
15-00
10.00
15.00
7-50
12. 50
12. so
2.50
2.50
3.15
2
2
4
3
3
2
2
3
I
9-09
9.09
18.18
13.64
13.64
9-09
9-09
13-64
4-54
3
I
4
4
S
15-79
5-26
21.06
21.06
26.31
3
14.29
4-76
23.81
19. OS
28.57
I
50.00
3.4s
I
50.00
3.6s
3.7s
I
S-26
4.76
3-95
4.IS
X
5.26
4-76
4.4s
5-56
Total
22
18
40
2
19
21
16-year-old cows.
Fat
percentage.
1908.
1909.
Combined years.
Frequency.
Percentage.
Frequency.
Percentage.
Frequency.
Percentage.
3.3s
16.67
16.67
16.67
33-33
16.67
I
I
2
2
I
I
3.4s
3.55
I
50.00
3.6s
3.75
3.9s
I
50.00
I2- 50
ToUl
J
6
298
Journal of Agricultural Research voi. xvn.No. e
VARIATION CONSTANTS
Before undertaking any discussion of the distributions given in Tables
I and II it is desirable to have at hand the simple physical constants,
means, standard deviations, and coefficients of variation deduced from
them. These constants are accordingly given in Tables III and IV. In
the calculation of the standard deviations vSheppard's correction of the
second moment was used in all cases.
Table III. — Constants for variation in weekly milk yield
Age of cow.
6 years.
8 years.
16 years.
Weighted means (total)
Year.
1908
1909
Combined.
1908
1909
Combined,
1908
1909
Combined.
1908
1909
Combined.
1908
1909
Combined.
1908
1909
Combined
1908
1909
Combined.
1908
1909
Combined
1908
1909
Combined
1908
1909
Combined
1908
1909
Combined
1908
1909
Combined
1908
1909
Combined
1908
1909
Combined
1908
1909
Combined
Mean weekly
yield (in
gallons).
3. 907±a 274
3. 407± -243
3. 6io± . 183
4. 029±
3- 701 ±
3- 84i±
S- 289 ±
5-i77±
S- 230±
6. S58±
6.382±
6.463±
7-698±
7-3l4±
7- 470±
8. 278±
7. 866±
8. 049 ±
8. I09±
8.37I±
8. 26o±
8.698±
8. 434±
8. S56±
8.683±
8. 773 ±
8. 738±
8. 367±
7- 889 ±
8. iii±
8. 790 ±
8. 205±
8.4S7±
9. 036±
8. 6i6±
8. 750±
7-932±
7. 972±
7- 9SO±
I. 500±i
7. 776±
8. I3i±
8. 500±i
7. 500±
7.875±
16.489
Standard de-
viation (in
gallons).
2. 4oo±o,
2. s69±
2. 5i4±
2. 392±
2. 480 ±
2. 449±
2- 734±
2. 7Soi:
2- 743±
2. 798 ±
2. 7I5±
2. 7SS±
2. 933±
2. 972±
2. 962±
3-079±
2. 877±
2. 97o±
2. 984±
2. 9S2±
2. 969 ±
2. 9II±
2. 901 ±
2. 909±
5. 049±
3- 030±
3'037±
3- 593 ±
3059±
3-330±
3- 143 ±
3-025±
3 o87±
2-99S±
2. 968±
2.986±
3-S09±
3- 795 ±
3. 640 ±
3- 747±I
3. 046±
3.3o6±
2. 24S±
2. 389±
2. 364±
Coefficient of
variation.
7. 26o±i. 432
9- l65±i.326
8.47i± .980
7-053±
8. io3±
7- 690±
7- 88s±
8. ii8±
8. 0I2±
337
310
229
384
367
266
:6. 899± . 572
6. S7S± -368
16. 736± . 272
6. 575± .449
7. I63± .385
i6.95S± -294
[6. S46± . 464
6. 103 ± .397
6. 4s6± .322
6. 477± .491
6. 07I±; .410
6. 2s8± .315
5. 566± . S23
5. 739± .489
5-675± -358
6.3i8±
6. I38±
6. 2io:i:
9- 56o±
7. IOI±
8. 384±
712
566
444
638
6. 726±I. 160
6. 6i7±i.oo2
6. 723 ± . 761
5. 773 ±1-682
S.943±I-2l8
S.924± -989
19. s66±2. 06s
21. ii6±2. 477
20. 28o±i. S91
17- 429 ±6. 053
17. I38±i.929
18. 234±i. 958
12. I37±4-IS3
13- 653 ±2. 707
13- 22S±2. 269
17. 081
Sept. 15, 1919
Variation in Milk of Ayrshire Cows
299
Table IV. — Constants for variation in percentage of fat in the milk of Ayrshire cows.
Age of cow.
Mean fat per-
centage.
Standard de-
viation (in
percentage).
Coeflficient of
variation.
3 years.
6 years.
8 years .
16 years.
Weighted means (total).
1908
1909
Combined.
jgo3.
1909
Combined.
190S
1909
Combined.
1908
1909
Combined.
1908
1909
Combined .
1909
Combined .
1908
1909
Combined.
1908
1909
Combined.
1909
Combined.
1908
1909
Combined.
1908
1909
Combined.
1908
1909
Combined.
1908
1909
Combined.
1909
Combined.
1908
1909
Combined.
891 ±0.
82s ±
852 ±
924±
9IS±
89s i:
903±
770±
78s ±
780 i:
77S±
741 ±
694 ±
7x6i:
705 ±
672±
68s ±
694 ±
b6g±
69I±
658 ±
668±
664±
6i8±
652±
636±
599±
601 ±
6oo±
6S3±
608 ±
629±
6io±
591 ±
599 ±
7oo±
559±
6o6±
595 ±
589i:
592 ±
500±
Soo±
500 i:
8oo±
6i7±
662 ±
O. 292±0. 024
.3I3± .021
.3lii: .016
1 .4I2± .008
6 .384^ .007
.344± .006
.36l± .005
.359± .008
c.342± .007
<i .326± .006
.342± .005
.3S8± .008
•335± -oo?
.346± .006
.3I4± -008
.322± .007
.3I9± .005
.3I2± .008
.305± .008
• 3o8db .006
.3o6± .009
.323± .008
• 3l6± .006
.3I3± -oio
.294± .009
•303± -007
.3o8± .013
• 278± .009
.29I± .008
.312± .015
• 3o6± .014
• 3io± .oio
.286± .019
.338± .020
.326i .014
.288± .030
.273± .020
.285± .017
.27I± .028
■iSi± -040
.3I0± .023
.096± .032
.225± .023
.2i6± .022
.igSi .067
.I3i± .026
.'!■^o±. .029
7.496±0. 611
". i8o± .552
.o8oi: .415
da.
3-738
SOO± . 202
8i6± .191
820± . 148
26oi: . 118
5i3± .200
034± . 179
626± . 171
055± .131
579± -223
074± . 198
320± . 149
481 ± .225
772i: ■ 193
664± . 147
448± .228
259± . 200
334± • ISO
358± .244
801 ± .221
6i7± . 164
64o± . 28s
047 ± .245
336± .187
569± .367
73 2 ± .266
072± .217
5S2± .424
48oifc .392
S36± .289
936± .538
4I2± .557
o66± .405
785± .815
684± .576
942± .487
7. S27± .770
9. 83s ±1.116
&64ii: .656
2. 736± .923
6. 436± .707
6. i8o± .646
$. 2o8±i. 761
3.62S± . 707
4-65o± .786
8.827
"I Including the two very high testing cows.
* Without the two very high testing cows.
<^ Including the one very high testing cow.
<* Without the one very high testing cow.
From these tables a number of points are to be noted,
I. It is evident that the mean weekly yield and the fat percentage
change with the age of the cow. The nature of these changes will not,
however, be discussed here but will be analyzed in detail in a later sec-
tion of the paper.
300
Journal of Agricultural Research
Vol. XVII. No. 6
2. From comparison of the results here given with those of Vigor {28)
it is seen that while, the general, the agreement is fairly close, there are
some rather striking differences. Taking weighted means from Tables
III and IV, we see that the mean weekly yield is slightly lower, and the
mean fat percentage slightly higher, in the whole group than in the
Fenwick district data alone. The differences in the means, however,
are small and probably of no significance. When we turn to variation
as measured by standard deviations, there is a striking difference in
weekly yield. The weighted mean standard deviation for the whole
group is 2.806 gallons, while ^^igor finds for the Fenwick district alone
4.0704 gallons. This is a large and statistically significant difference.
In fat percentage the standard deviations are practically alike for the
two sets of data, our weighted mean value being 0.330 and Vigor's 0.3229
3. The explanation for the difference in variability in weekly yield
between our figures and Vigor's is not far to seek. It lies mainly in the
fact that Vigor has dealt with cows of all ages lumped together, while
in the present paper each year of age is dealt with separately. Natur-
ally when dealing with a character which changes with age so extensively
as does milk yield, as has recently been discussed by Pearl (12), the
variation exhibited will be markedly increased if animals of all ages are
lumped together. In order to determine how much of the difference in
variation was due to this cause and how much to other factors Table V
has been prepared. This table gives the distribution for weekly yield
obtained by adding together all of the "combined" distributions for the
several years, as set forth in Table I.
Table V. — Distribution for weekly yield combi?ted for all ages and for the whole area to
compare with Vigor's data for the Fenwick district alone
^
1
10
I
T
12
I
■K
I
J
I
■;
I
5
I
7
iS
I
3
18
33
81
146
275
442
592
752
819
850
769
636
19
20
21
22
23
24
25
26
27
28
29
30
Total.
587
2 •so
i^i
lO^
61 -^o
iR
8
3
4
I
6.935
Mean = i5.99i±o.o27. Standard deviation=3.329±o.oi4. Coefficient of variation= 20.81 6 ±0.088.
4. The difference between this value of the standard deviation and
Vigor's, while reduced, is still sensible. It amounts to about 0.742 ±
0.082. This remains to be accounted for. We find it difficult to suppose
that the selection of relatively long lactations in the present data can be
the cause, since Vigor himself has shown that there is no sensible cor-
relation between either mean weekly yield or fat content and duration
of lactation. We are much m.ore inclined to the view, especially in the
light of unpublished results on milk production in other breeds of cattle,
that the Fenwick district returns give somewhat abnormal values in the
Sept. 15, 1919 Variation in Milk of Ayrshire Cows 301
direction of heightened variability and also in certain other respects which
need not be gone into here.
5. Comparison of the present results with those of Gavin (2, 3) leads
to the same conclusions as those reached in the preceding paragraphs.
In Gavin's first paper (2), where 1,233 normal lactations of cows of all
ages lumped together are discussed, coefficients of variation are given as
follows: For total lactation yield 25.72; for average daily yield 25.78;
for maximum daily yield 24.68; and for revised maximum daily yield
24.77. These values are of the same order as those from Vigor's data
(coefficient of variation about 24.2) and from the total combined dis-
tributions (Table V) of the present paper. In a later paper Gavin (j)
deals with each of the first five lactation periods separately for a group
of about 375 cows. From his data we find the weighted mean coefficient
of variation for these five lactations to be 17,998, a value sufficiently
close to our weighted mean value for single years of age.
6. Turning to the fat percentage, we see that Vigor's values of 3.681
for the mean and 0.323 for the standard deviation are substantially
identical with our weighted mean values of 3.738 and 0.330. Pearson
(2j) has also given some reductions for variation in Ayrshire fat per-
centages, and the present values are again in close accord with his.
7. It may be concluded that the values of the means and variabilities
here given represent essentially normal values for Ayrshire cattle.
These constants will be of considerable usefulness as time goes on, for
purposes of comparison with other breeds and in the study of special
problems.
THE COMPARATIVE VARIABILITY OF MILK PRODUCTION
Milk production is essentially a physiological character. It is a matter
of some interest and significance to examine the variability of the
character in comparison with other physiological characters and also
with some that are more strictly morphological, as, for example, bone
measurements. Such comparisons may be made through the coefficients
of variation. It must, however, always be kept clearly in mind just w^hat
a coefficient of variation is ; and care must be taken to avoid drawing too
sweeping or even entirely unjustified conclusions from comparison of
these constants. What the coefficient of variation measures is the per-
centage which the "scatter", or variation exhibited by a distribution as
measured by the standard deviation, is of the mean of the character vary-
ing. For some purposes this percentage is meaningless. It is therefore
idle to try to force its use for those purposes. It will undoubtedly be
presently supplemented by some other constant for the measurement of
other aspects of comparative variability. It has, however, a perfectly
definite, if limited, meaning. It is a unique constant of any distribution,
expressed in abstract units. As such it may be used for purposes of
comparison, always remembering that one must be cautious as to the
maimer in wliich conclusions drawn from such comparison are stated.
302
Journal of Agricultural Research Voi. xvii, No. 6
In Table VI are given coefficients of variation for a number of char-
acters for purposes of comparison with milk yield. The coefficients are
arranged in order of descending magnitude.
Table VI. — Coefficients of variation for various characters
Characters.
Coefficient
of varia-
tion.
Authority.
Number of children per family (New South Wales)
Area of comb (domestic fowl)
Weight of spleen (English males)
Size of litter (mouse)
Lambs per birth (sheep)
Dermal sensitivity (English males)
Annual egg production (domestic fowl)
Size of litter (Poland-China swine)
Size of litter (Duroc-Jersey swine)
Milk yield (total lactation)
Milk yield (daily average)
Fecundity o (horse)
Heart weijiht (English males)
Weight of kidneys (English males)
Weight of liver (English males)
Swiftness of flow (English males)
Body weight (English males)
Revised maximum daily milk yield (for given age)
Weekly milk yield (Ayrshire cattle)
Breathing capacity (English males)
Strength of pull (English males)
Weight of shell of egg (domestic fowl)
Body weight (domestic fowl)
Weight of albumen of egg (domestic fowl)
Length of red blood corpuscles (Bufo tadpoles)
Weight of yolk of egg (domestic fowl)
Amount of fat in mixed milk (daily fluctuations) . .
Yield of mixed milk (daily fluctuations)
Weight of egg (domestic fowl)
Brain weight (Bavarian males)
Length of forearm (English males)
Length of femur ( French males)
Length of egg (domestic fowl)
Stature ( English males)
Horizontal circumference of skull (English males) . .
Specific gravity of egg (domestic fowl)
78
Powys (25).
Pearl and Pearl (14).
Greenwood (4).
Weldon (2q).
Pearl (n).
Pearson (19).
Pearl and Surface (i;).
Surface (27).
Do.
Gavin (2).
Do.
Calculated from data in Pearson(27).
Greenwood and Brown (5).
Do.
Do.
Pearson (lo)-
Greenwood and Brown (5).
Gavin (j).
This paper.
Pearson (19).
Do.
Curtis (r).
Do.
Do.
Pearson (22).
Curtis (j).
Pearl (70).
Unpublished data in this labora-
tory.
Pearl and Surface (16).
Pearl (0).
Pearson and Lee (,24).
Pearson (iq).
Pearl and Surface (.16).
Pearson and Lee (.24).
Macdonell (7).
Pearl and Surface (j6).
a Fecundity here means the fraction which the actual number of offspring arising from a given number
of coverings is of the possible number of offspring under the circumstances.
This table brings out the well-known fact, which has been discussed in
some detail by Pearl (9), Gavin (2), and others, that, in general, physio-
logical characters exhibit high coefficients of variation as compared with
strictly morphological characters. Characters which are intermediate in
their quantitative determination, as, for example, the length of the egg in
the domestic fowl, give coefficients of variation intermediate in value.
Purely physical characteristics which are usually regarded by physicists
and chemists as "constants," such as the specific gravity of eggs, show
very low coefficients of variation.
It is of interest to compare the coefficients of variation for total yield
and absolute amount of fat in the mixed milk of a large herd with
those for milk yield as discussed in the present paper. It is seen that
the former are about 9, whereas the coefficients for milk yield give values
of about 17 to 25, depending upon whether cows of all ages or of a single
age are considered.
In secular variation in the amount or quality of the mixed milk of a
large herd, individuality of the animal as a source of variation is entirely
Sept. IS, I9I9 Variation in Milk of Ayrshire Cows 303
eliminated. The observed variation must, therefore, be due to the com-
bined action of all the external environmental influences which affect in
greater or less degree the milk yield of every cow.
On the other hand, the constants of variation for milk yield determined
in this paper are based upon the diversity or variation in weekly yield
exhibited among a large number of different cows. Here one primary fac-
tor in the causation of the observed variation must be the individuality of
the animal with respect to milking ability. By individuality in this sense
is meant the genotype of the individual with regard to the character named.
But in the causation of the variation in milk yield as here discussed there
must be involved the combined influence of the individuality of the
animal plus that of all the environmental factors which act in producing
variation in the mixed milk of the herd, since each of these causes influ-
ences every individual animal while it is making its individual record.
It is therefore possible to make comparison here between observed
variations (as measured by the coefficient), due, on the one hand, to en-
vironmental influences alone and, on the other hand, to genotypic differ-
ences plus environmental influences. The difference should represent
in a general way that part of the observed variation due to genotypic
differences.
The figures as they stand suggest that roughly about one-half of the
variation (measured by the coefficients of variation) in milk production
results from the varying genotypic individuality of the animals with re-
spect to this character, and the other half results from the varying exter-
nal circumstances to which cows are subjected during lactation and
which have an effect upon the flow of milk. Or, to put the matter in
another way, if the conclusion just stated were true it would mean that
if a large number of cows were placed in environmental circumstances
which were at once ideal and uniform we should expect the variation
exhibited in milk production to be roughly about one-half of that which
we actually find when we measure this variation under ordinary cir-
cumstances.
Another point of interest in connection with Table VI is the com-
parison of the coefficients of variation for milk yield with that for the
weight of the albumen of the &%g of the domestic fowl. Both of these
are secreted products. The weight of the shell of the &gg is another
character falling in the same category. The figures here given indicate
that the variation in these characters, taken in relation to their respec-
tive means, is greater for milk secretion than for albumen or shell secre-
tion. Or, put in another way, the oviduct as a secretory organ appears
to work truer to type than does the udder of the cow. This result is
what would be expected from all that is known of the physiology of the
two organs. The secretory activity of the cow's udder is apparently
very much more easily influenced by external circumstances and by
nervous impulses than is the oviduct of the fowl.
122502°— 19 5
304
Journal of Agricultural Research
Vol. XVU, No. 6
ANALYTICAL DISCUSSION OF VARIATION IN MILK PRODUCTION
Turning next to the analysis of the variation in mean weekly yield
and in fat percentage by fitting skew curves to the observed frequency
distributions, we have the results set forth in Tables VII and VIII.
Table VII gives the anal3^tical constants for mean weekly yield and
Table VIII those for fat percentage. In fitting the curves the combined
distributions for the two years 1908 and 1909 have been used throughout
in the case of weekly yield. In the case of fat percentage the combined
distributions have been used for all ages except 3 and 4 years. Since
there was some doubt as to whether, at these ages, the fat distributions
for the two years were significantly different from each other, it was
thought best to fit the 190S and the 1909 fat distributions separately
for both of the ages mentioned.
TablB VII. — Analytical constants for variation in mean weekly yield
6 years.
V/Si.
Skewness
d (gallons)
a (gallons) ,
Mean (sallons). . .
Mode (frallons) . .
Range (gallons.).
+end of range. . .
—end of range. . .
Yoper cent
P. E. V^
-p.-E.Pi
P. E. skewness. .
1,441
23.9798
17-7255
1,830.3431
.0228
.1509
3- 1830
•1830
.2977
.0578
.0688
.1686
2.4485
13-8413
13.6727
1,118
30. 1002
21.6903
986. 7220
•0173
•13 13
3-2965
-2965
•5413
.0240
.0560
•1537
2-7432
15-2299
15-0762
8.28
± -0435
± .0S70
± .0217
7-50
± -0494
± .09S8
± .0247
909
30-3681
61.4079
3,604.8731
.1346
.3669
3.9089
.90S9
1-4138
.0746
.1302
•3587
2-7554
16.4634
16. 1047
35-0959
55-2151
4, 201.6107
.0705
-2656
3.4112
.4112
.6107
.0883
• liiS
• 3304
2.9621
17.4701
17-1397
7-79
± .0548
± . 1096
± -0274
6.99
± .0583
± -1165
± .0291
712
35-2890
80. 0331
.307.9103
•1458
• 3818
3-4593
-4593
.4813
•2357
.1661
•4934
2.9702
18.0492
I7^555S
6-97
± .0619
± .1238
± -0309
Constant.
/3i...
/32...
Skewness
d (gallons)
a (gallons)
Mean (gallons) . .
Mode (gallons) .
Range (gallons).
+end of range. .
—end of range. . .
Yo per cent
P. E. -n/^i
P.E.Oj
P. E. skewness.
8 years.
637
35-2522
59-3923
4,334-0982
.0805
• 2838
3-4876
.4876
• 7336
.0842
.1157
•3435
2.9687
18.2602
17.9167
7- 03
± • 065s
± • 1309
±•0327
459
33.8404
48-3269
3,464.0612
.0603
-2455
3.0249
.0249
— . 1309
— -3505
• 1283
•3732
2.9086
18.5561
18. 1829
64.4991
67-2343
2-7352
6.85
±•0771
±.1542
±.0386
320
36-9035
44-3365
4,387-5182
•0391
.1978
3.2217
.2217
.3260
.0909
.0895
.2718
3-0374
18.7375
18.4657
6-57
±•0924
±.1847
± . 0462
202
44. 3448
—22.3850
7,459.8357
.0057
-0758
3-793 5
•7935
I. 5698
.0009
— .0259
— .0863
1-3296
18. 1114
18-1977
6.46
±.1163
±-2325
±.0581
4,642
116
38. I07S
4164
5S35
2491
4991
1970
1970
3535
1203
2817
8695
0866
4569
5874
0305
4776
4471
52
1534
3068
0767
Sept. IS, 1919
Variation in Milk of Ayrshire Cows
305
In the graduating of the observational data Pearson's (18, 20) skew
frequency curves and his method of moments have been used through-
out. In the tables the moment coefficients are given in terms of units of
grouping.
Only the distributions for the ages 3 to 12 years, inclusive, have been
subjected to analytical treatment. Outside of these limits the numbers
involved become so small as to make the discussion of them from the
point of view of mean weekly yield and fat percentage not worth the
labor involved.
Tabi,E VIII. — Constants for variation in fat percentage
ft. ..
VS.
fi2...
Skewness
d (percentage)
a- (percentage)
Mean (percentage).
Mode (percentage) .
Range (percentage .
+ End of range. . . .
— End of range. . . .
Yo percent
P.E. VS
P. E.02
P. E. Skewness. . , .
614
14- 7683
21. I714
734-3389
• 1392
• 3730
3-3669
• 3669
.3164
■ 3416
.1697
.0652
• 3843
3-9151
3. 8499
10.6700
± .0667
± -1334
± • 0333
82s
II- 7975
II. 600s
434- 5862
.0820
.2863
3- J225
. 1225
. 0009
- 6.8108
•1432
.0492
•3435
3- 8947
3-8455
1.4607
11.6700
± -0575
± .1150
± .028S
S26
. 12.8584
19.9750
582-6387
.1887
•4332
3-5239
•5239
.4848
•3044
.1886
.0676
.3586
3- 7696
3- 7020
11.0700
± .0721
± .1441
± . 0360
591
10. 6362
5-8796
342. 3461
• 0287
.1695
3. 2036
.0262
•0338
.6413
.0857
.02S0
.3261
3- 7804
3-7524
IS- 7482
16. 5764
• 8283
12.2400
.0680
•1359
.0340
909
11-9935
10. 5807
460. 8242
.0649
.2548
3- 2036
.2036
. 2126
•2327
. 1192
.0413
•3463
3- 7160
3- 6747
± .0548
± . 1096
± -0274
6 years.
10. 1970
7-6514
330- 2351
•0553
-2350
3- 1760
. 1760
• 1863
•2254.
.1108
-0354
-3193
3- 6852
3-6498
. 7000
-0583
- ii6s
■ 0291
N
/*2
W
M<
ft^
V/3i
ft
|82-»
Kl
K2
Skewness
d (percentage) ....
<r (percentage). . . .
Mean (percentage)
Mode (percentage)
Range
+ End of range . .
— End of range. . ,
Yo per cent
P.E. -VS
P.E- ft
P. E. Skewness. .
)- 4598
J-OI93
I- 7592
.0298
•1725
!. 7400
- 2600
.6093
•0370
. IO7I
•0330
.3076
.6906
•6576
.6461
-2643
.6182
- 2600
.0619
•1237
•0309
8 years.
637
9. 9688
9. 8014
309.5716
.0970
■3114
3-1151
-1151
. 0607
1 . 2 2 70
.1589
.0502
-3157
3-6637
3-6136
13-3421
15-2865
1.9445
12. 7600
•065s
• 1309
.0327
459
9- 1889
4-8556
247. 6547
•0304
•1743
2- 9330
.0670
.2251
.1021
-3031
3- 6362
3-6076
4- 57-0
6-6211
2.0441
13.0100
.0771
.1542
.0386
320
8.4448
7-5919
228. 2603
.0957
•3094
3. 2008
. 2008
.1144
.6425
.1492
-0434
. 2906
3.6003
3- 5569
13-9500
± .0924
± . 1847
± . 0462
202
9- 5987
9- 1894
250. 6424
.0955
•3090
2. 7204
. 2796
- .8457
. 0872
•2194
.0680
-3098
3-6292
3-5612
2- 7333
4- 9959
2. 2625
12.3700
•1163
•2325
.05S1
116
10. 6494
10.0965
361. 60H
.0844
-2905
3- 1885
.1885
-1237
-5225
-1397
-0456
-3263
3S99I
3-5535
12. 4200
± • 1534
± .3068
± -0767
3o6
Journal of Agricultural Research Voi. xvii. No. 6
The histograms and their fitted curves are shown in figures i to 4.
^ 15
A.
5 10
10
^
K
3-YEAR-O^D COWS
J
/
/
b.
\1
/
c:
^
6
V
4-YEAR-Ol.D COWS
/
7^ ■
\
/
r
\
?
[\
^
5- YE
AR-OI,
p COWS
/
\
^
y
/
/'
\
^
r
H
6-YE
AR-OI
.D CO
WS
/
A
K
-,
r?
/
A
^_^
/
X
7-YE
AR-OI
.D CO
WS
r
[K
"vl
V.
^
/]
^
-M.
7 9 11 13 15 17 19 21 23 25 27 29
MEAN WEEKLY YIELD - GALLONS
Fig. I. — Histograms and fitted curves for variation in mean weekly milk yield of Ayrshire cows of ages
3 to 7 years. The ordinates are plotted on a percentage basis, and since the base unit (i gallon) is the
same for all diagrams the areas of all are equal.
Sept. IS, 1919 Variation in Milk of Ayrshire Cows
307
11 13 15 17 19 21 23 25 27 29
MEAN WEEKLY YIELD - GALLONS
FlC. a. — Histograms and fitted curves for variation in mean weekly milk yield of Ayrshire cows of ages
8 to 12 years.
3o8
Journal of Agricultural Research voi. xvn, no.6
Fig. 3. — Histograms and fitted curves for variation in fat percentage of milk of Ayrshire cows of ages
3 to 7 years. For purposes of illustration the 1909 curves are used in the 3- and 4-year classes.
Sept. IS. 1919 Variation in Milk of Ayrshire Cows
309
27 2.9 3.1 3.3 3.5 3.7 3.9 4.1 4.3 4.5 4.7 4.9 5J
FAT PERCENTAGE
Fig. 4.— Histograms and fitted curves for variatioa in fat percentage of milk of Ayrshire cows of ages
8 to 12 years.
2IO Journal of Agricultural Research voi.xvn, no. e
The equations to the various curves are as follows
Mean weekly yield
-22.7692 10.7802 tan"'
/ 2 \ —22.7692 10.7002 tan '
3 years. •J'=33 -7848(1+^^1^) ' '""" ^^^^'^'^ ^^"^^ ^'^'
4 years. •>'=52.7759(i+^7;^)
5 years. .>'=47-4o6(^i+^^^_^^^ J ^ -Uo.
26.698s
J.2 \— 136324 3-9661 tan 1
2 N -6.8866 3.3430 tan-l
2 \ —12.4973 7-1576 tan 1-
/ , x2 \-"-4973 7-IS76 tan '^-^^^
6years. .j'=2o.457(^i+^^^-y^j e -Uo.
/ , x^ \-xs-4198 16.0162 tan-1^^3^^
7years..j'=o.9i98(^i+^^^;8i86; ^ ^'
^, s -10.8431 5.9709 tan-l j^;^
8years..y=i9.87i(i+6^j;^765; ^ Do.
. .21.0838 , y. V 66.9477
9 years. .^=3 1.42 78(1 +3-^) (^-^71^) ^^^^^"'^ '^^^P^ ^-
10 years. .^=21.0148 e ~ -a-^i^ Normal curve.
. 2 \ "6.2805
11 years. .y=x3.o538(i+^^,) ^^^arson's Type II.
N 5-6158, N 25.4498
12 years. .>'=7.5643(x+^^) {^-^^^ ^^ '^ '^P^ ^-
Fai percentage
,.2 \ —22.1236 30.4278 tan-i-
/ ,.2 \ -22.1230 30-4270 •.ail - , >,, TTT
3 years, ,,08. .,.„.^,,(^+^^ e "-" Pearson^s Type IV.
— 2.0483I , >. 48.8478
3 years, 1909. .^=96.^95^ e C^+ip^e) ^^=^'^"'' '^^'P" '"•
(o \ — ic.Assg 19.1242 tan ' t, 3
T I ^ ^ 5 4559 9 'SSooS ,^ jy_
^+249.6647/ '^ ^
. s65.3S87, ,^ .286.7682
4 years, 1909. .)'=7-36o5(x+,-^) (^"x-^S^^) ^^^''^"'^ '^^P^ '•
9 X — 11.1847 75.2496 tan ' 7
/,_^i_V / ''■'" Pearson's Type IV.
5 years. .>'=o.o2 2 1(^1+-^^^;^^ e
X
(o N —It;. 1461 ?6. 8364 tan '■ ■
,_^l_^ ^^-'^^^ ^3-os5i Pearson's Type IV.
"^531-5363/
. X 5-8303/ „ N9-OI23
Q a,./tJ- ^ 1 fi- - — ) Pearson's Type I.
7years..>=87.2637(^i + — 3^^^^ l^i 16.0669/
Sept. IS, I9I9 Variation in Milk of Ayrshire Cows 311
24.7018 , V 172.7511
'+16:6^; (^-^6:7;^; Pearson's Type I.
(^ .16.6417/- ^ X32-07SS
—34/ 3;2 \ -36-1987 147.9872 tan-'^^—
10 years. .>'=9.6682 X 10 ( i4" ) e ' Pearson's Type IV
11 years. .7=24.9882(^1-^3^-^^ J Pearson's Type II.
~^°/ X- \— 32.0148 106.7238 tan-i^y^g—
12 years. .)'=2. 5479 X 10 ( i-j ^-7 — j g " Pearson 'sTypelV.
From Tables VII and VIII and the accompanying curves the follow-
ing points are to be noted :
1. It is apparent that the fitted curves give very good graduations
of the data throughout. Pearson's generalized probability curve has
been shown by experience to be applicable in one or another of its types
to so wide a range of cases that a new application calls for no special
mention. However the continued addition of new classes of data
easily and perfectly graduated by these curves is the best refutation
of the criticisms which were formerly made against them.
2. The general tendency of these milk production and fat percentage
variation curves is plainly toward positive skewness. All of these
curves show a positive skewness, with the single exception of the mean
weekly yield curve for cows 1 1 years old. There the skewness is minus
but in comparison with its positive error (on the basis of the normal
curve) is insignificant. In other words, this curve for 11 -year-old cows
is, within the limits of error of random sampling, a symmetrical distribu-
tion. All the others are skew in the positive direction, or, in other
words, the mean is greater than the mode.
3. Considering the probable error of this skewness on the basis of a
normal curve, it is seen that in 7 out of the 10 curves for the mean weekly
yield the skewness is three or more times its probable error. In 2 cases
it is somewhat less than three times its probable error, while in i case
the skewness is certainly insignificant — that for 11 years, as already
noted. In the case of fat content 9 out of the 11 curves show a skew-
ness three or more times the probable error. In 2 of the remaining
cases the skewness is nearly three times the probable error, while in i
case — that of the 12-year-old cows where the number of individuals
concerned is small — the skewness is distinctly less than three times its
probable error. From these figures it is plain that in general these
Ayrshire milk variation cur\^es show a significant tendency toward
an asymmetry indicated by a positive skewness.
312 Journal of Agricultural Research voi. xvii, no. 6
4. It is of some interest to examine the weighted mean value of the
skewness for all the curves, the weighting being in proportion to the
number of individuals involved, in comparison with the skewness ex-
hibited in the variation curves of other characters. AVe have for the
weighted mean value of the skewness for mean weekly yield, the 11-
year curve being omitted, a value of +0.1047. ^o'' the variation curves
for fat content the weighted mean value of the skewness is +0.1338.
It was shown by Pearl and Surface (75) that in variation in annual
egg production in Barred Plymouth Rocks the skewness is always
negative and usually significa.nt. This difference in skewness between
the two characters milk production and egg production is striking.
Curv'es of variation in egg production tail off more on the side toward
low egg production, whereas the curves of variation in milk production
tail off more on the side toward high production. The weighted mean
values of the skewness for annual egg production in three successive
years were found to be —0.280, —0.122, and —0.108. In other words,
the values in general were of an order of magnitude not far from that
here found for the skewness of curves of variation in milk production.
5. It might at first thought be supposed that the direction of the
skewness in milk productive curves was due to selection — that is, to the
continued culling out of the poor producers. Since, however, the same
factor of selection in the direction of the high producers was operative
to as great or even a greater extent in the making up of the flocks from
which the annual egg production variation curves were obtained, it seems
perfectly clear that selection can have had very little to do with bringing
about the difference in direction of skewness exhibited by egg and
milk production curves respectively. The inference would then seem
strongly justified that selection had nothing to do with the production
of the asymmetry of the variation curves in either case considered by
itself.
6. Additional interest is given the matter when an examination is
made of the facts regarding the direction of the skewness in the varia-
tion of the hen's egg in size characteristics. vSuch data have been fur-
nished by Pearl and Surface (/6). They show (/?. 184) that in the
variation of egg length, egg bulk, egg weight, and egg breadth the skew-
ness is positive and significant in all cases except that of breadth. We
see here again emphasized a point which comes out frequently in bio-
metrical work — namely, that there is frequently between characters
a parallelism in variation corresponding to a parallelism in the
underlying physiological bases of the characters. This relation
is clearly apparent in the present instance. The size of the egg is
primarily determined by the secretory activities of the oviduct. It is
a character which is physiologically much more directly comparable
to milk production than is total annual production. Primarily the
latter depends physiologically upon quite another thing — ^namely, the
Sept. IS. I9I9 Variation in Milk of Ayrshire Cows 313
inherited genes for fecundity which determine the frequency and regu-
larity of ovulation. Corresponding to the physiological parallelism
in egg size and milk production is found corresponding asymmetry of
the variation curves, as well as a closer relationship between other of
the variation constants in the two cases than is found when milk produc-
tion is compared with egg yield.
7. Considering the types of the curves, we find that 7 out of 10 curves
for mean weekly yield give upon analysis unlimited range curves —
in 6 cases the skew Type IV and in one case the symmetrical normal
curve. Something approaching the reverse condition is found with
respect to variation in fat percentage. Five out of the 1 2 distributions
for this character lead upon analysis to curves with the range limited
at both ends (Type I) and one to a cur\^e of Type III, which is limited
at the lower range end. The remainder of the curves are of Type IV,
but near the border line of passage over to the limited range types. It
would then appear that the physiological fact that variation in percentage
of fat content will necessarily tend to be confined within relatively nar-
rower limits than variation in total flow of milk is reflected in the dis-
tribution of the several curves in respect to type.
8. The estimation of the range ends in the case of the limited range
curves is on the whole fairly good, leading in no case to absolutely impos-
sible values regarding the probable errors involved. The determina-
tion of the range ends in the Type I curv-^e is subject to rather considerable
probable errors. The most extreme range end estimation in mean
weekly yield is that given by the curves for 9-year-old cows. This
gives for the upper range end 67.2 gallons a week. This of course would
be an extraordinarily high average weekly yield, yet it probably can not
be regarded as physiologically impossible. It certainly would not be
for a single week. Indeed such a record is rather frequently exceeded
by Holstein-Friesian cows which on official tests may occasionally go
to a production over 100 gallons per week. In fat percentage the most
extreme range estimation is for the 1909 curve for 4- year-old cows,
which gives for the upper end of the range 16.6 per cent fat. Again
this figure, while of course extraordinary for an average test, probably
indicates no physiological impossibility for brief periods of time. That
such is the fact is indicated by some of the short period tests of Jersey
cows.
9. In all of the curves for mean weekly yield the kurtosis is positive.
In other words, these curves show a tendency of greater or less degree
toward the leptokurtic condition. They are more peaked than would
be noi-mal curves of corresponding standard deviations. The value of
the kurtosis is probably significant in all the mean weekly yield curves,
with the exception of those for 3-year-old, 9-year-old, 10- year-old, and
12-year-old animals. In curves of variation in fat percentage there is
no such unifonn tendency in regard to the value of the kurtosis. The
314
Journal of Agricultural Research
Vol. XVU. No. 6
1908 3-year-old curve is probably significantly leptokurtic. The 1909
3-year-old curve does not appear to differ significantly from the normal
in this respect. The same relations hold in regard to the 1908 and 1909
4- year-old curves. The 5 -year-old distribution is distinctly leptokurtic.
The 6-year-old distribution is probably mesokurtic. The 7-year-old
distribution is probably platykurtic. This is the first of the fat curves
to give a negative value for the kurtosis. The remainder of the curves
are significantly mesokurtic.
CAN THE VARIATION IN MEAN WEEKLY YIELD BE BETTER REP-
RESENTED BY THE SUM OF TWO NORMAL CURVES OR BY A
UNIMODAL SKEW FREQUENCY CURVE?
An examination of certain of the raw distributions for variation in
milk yield suggested that possibly we were dealing here with bimodal
distributions. Such a possibility is well worth testing thoroughly on
theoretical grounds, since if it were found that milk production curves
were bimodal this fact might be used as a first point of departure in the
determination of the number and characteristics of the (presumably
multiple) genes concerned in the inheritance of this character. We
have consequently subjected certain of the distributions to the method
of analytical dissection discovered by Pearson (77).
The distribution chosen for dissection were those for 5- and 6-year-old
cows, the combined distribution for the two years (1908 and 1909)
being used in both instances. (Compare Table I.)
It will not be necessary here to go over all the details of the laborious
arithmetic involved in this work. It will suffice to show, as is done in
Table IX, the best solutions when these two distributions are regarded
as the sum of two normal curves in each case.
TablS IX. — Constants of the component normal curves in the variation in m,ean weekly
yield
S-year-old cows.
First
cojnponent.
Area 716. goo
Mean (gallons) 16. 115
S. D. (gallons)
Modal ordinate
62. 220
Second
component.
192. 100
17. 764
3-752
10. 210
6-year-old cows.
First
cojnponent.
496. 600
16. S90
2. 418
40. 970
Second
component.
307. 400
18. 408
3-476
17. 640
From this table it is seen that the dissection gives in both cases a
lower component curve of large area and small standard deviation and
an upper component of smaller area and much larger standard deviation.
This is exactly the sort of result which might well be expected if milk
yield depended upon two hereditary factors, the higher one of which
was linked with sex or some other factor.
Sept. 15, I9I9 Variatio7i in Milk of Ayrshire Cows 315
The graduation obtained by the summing of two curves was, in
general, a good one. But before drawing any conclusions regarding
genetic factors from these successful resolutions of the variation curves
into two components it will be well to determine quantitatively, by
means of Pearson's test for goodness of fit, whether the two component
curves or the unimodal skew curves give the better graduations. Carry-
ing out this test the following values are found:
2-cojnpoiient
curve.
5-year-oId COWS.
6-year-old cows.
P=o.774
P=o.624
P=o.7i7
P=o.599
It thus appears that while both the skew curves and the 2-com-
ponent curves graduate this material rather well, there is a distinct,
if not large, advantage with the skew curve in each case.
To sum the whole matter up it may be said that, while it is possible
to graduate milk production variation distributions as the sum of two
normal curves, the resulting fit is not so good as that obtained with the
appropriate skew frequency curves. There is no evidence fiom the
analysis of the variation curves to indicate either that milk production
distributions are bimodal or that this character depends upon two
rather than some other number of genetic factors.
THE RELATION OF MILK AND FAT PRODUCTION TO AGE
With the analyzed variation data in hand it is possible now to con-
sider the problem of the changes in milk production per unit of time and
in mean fat percentage, with advancing age of the cow. The great
importance of a thorough and comprehensive knowledge of these rela-
tionships, if one is to make any adequate investigation of the inheritance
of milk and fat production, is sufficiently obvious. It is a perfectly
well-known fact, incorporated in all rules for advanced registry of
dairy cattle, that milk production does change with age, and to a marked
degree. Until investigations on this subject were undertaken in the
Biological Laboratory of the Maine Station some years ago it has always
been assumed by those (such as advanced registry officials) who have
had to deal with the problems that the changes of milk production with
age were linear up to "mature" age, usually taken as 5 years, and that
after that time there was no further change with advancing age. How
far wrong such an assumption is will be shown graphically below. It
was pointed out two years ago by Pearl {12) in a preliminary paper
based on calculations then completed that the fundamental law of
change with milk flow with age is logarithmic.
Let us now examine the facts for Ayrshires, considering first mean
weekly yield. The necessary data are given in Table III. The mean
3i6
Journal of Agricultural Research
Vol. XVU. No. 6
weekly yields in gallons for the combined distributions from age 2
to age 16, inclusive, are exhibited graphically in figure 5. The zigzag
line gives the observed production as ordinate against age as abscissa.
The smooth curve is a logarithmic curve of which the equation is
y= 12.4766 +0.61 46x — o.o3663;2 + 3.6641 log re,
where y denotes mean weekly yield in gallons and x age in years, taking
origin from i year. This curve was fitted by the method of moments
(compare Miner 8).
20.000
J 9, 000
/a 000
17.000
1 6,000
16,000
/ 4,0 00
/3,000
12.000
^
\
y
7^
\
>
SSs
/
//
^
^
/
/ 1
1
h
1
0 1 2 3 4 S e 7 a 9 JO // 12 13 /4 /£ 16 /7
AGE IN YEARS
Fig. s. — Showing the change in jnean weekly yield of milk in Ayrshire cows. The smooth curve is
of the form v=a+hx^cx^-\-d log x.
The actual figures, observed and calculated, are given in Table X.
It is evident from Table X and the diagram that the change here is
logarithmic. No better agreement between obser\^ation and theory than
that here shown could be expected. The law of change may be stated in
words in the following way : In these Ayrshire cattle the absolute amount
of milk produced per unit of time increases with the age of the cow until
a maximum is reached, but the rate of increase diminishes with advanc-
ing age until the absolute maximum of production is reached. After the
time of maximum productivity the absolute production per unit of time
decreases with advancing age, and at a continually increasing rate. This
conclusion agrees with that of Pearl and Patterson (ij) for Jerseys.
Sept. IS, 1919
Variation in Milk of Ayrshire Cows
317
Table X.— Comparison of observed mean weekly yields at different ages ivilh those cal-
culated on the assumption that the change is logarithm,ic
Age (in years).
Mean weekly yield (in
gallons).
Age (in years).
Mean weekly yield (in
gallons).
Observed.
Calculated.
Observed.
Calculated.
2
13. 610
13. 841
15-230
16. 463
17.470
18. 049
18. 260
18. 556
13- 055
14. 656
15- 730
16. 544
17. 183
17.684
18. 067
18. 344
10
18. 738
18. Ill
18. 457
18. 750
17- 950
18. 131
17-875
18. 524
18 610
•2
II
4
12
18. 608
e
J7
18. 519
18. 346
18. 091
17- 754
6
14.
7
le
8
16
0
With the equation relating to mean weekly yield and age in hand we
may consider the important problem of the age at which milk produc-
tion is at a maximum in these cows. To get an answer to this question
we have obviously only to equate -j- to zero and solve for x.
We have
^=0. 6146-0. 0732x + -^-^.
dy
When -3^=0, we have
dx
3^=10.4720.
Or, we may say that in the large group of cows here dealt with the
maximum rate of milk production per unit of time is reached only when
the cow is io>2 years old.
Turning next to the relation of fat percentage to age, we have the es-
sential data exhibited in Table XI, the values being taken from Table IV.
Table XI. — Mean fat percentage at different ages
Age (in years).
Mean fat percentage.
Age (in years).
Mean fat percentage.
Observed.
Calculated.
Observed.
Calculated.
2
3-852
3-903
3-775
3-716
3-685
3-691
3.664
3-636
3.862
3.827
3-793
3-759
3-725
3.690
3- 656
3.622
10
3.600
3.629
3-599
3. 606
3-592
3-500
3.662
3.607
3.604
3.601
3-598
3-595
3-593
3-590
T,
II
4
12
C
I?
6
14
I q
7
8
16
9- •"
From an examination of the observed figures it appears that in general
the fat percentage tends to decline with advancing age until the tenth
3i8
Journal of Agricultural Research voi. x\ai. no. 6
year is reached. From that point on, allowing for chance fluctuations
and the fact that the numbers dealt with get progressively smaller, the
fat percentage appears to remain about constant for the rest of the cow's
milking life. Consequently, it has seemed best to break the curve at the
lo-year point and fit the two parts separately, each with a straight line.
The resulting figures are given in the "calculated" column of Table XI,
and are shown graphically in figure 6. The equations to the two lines
4,000'
3.900
FiQ. 6.— Showing the observed (zigzag line) and calculated (straight line) changes ia the mean fat per-
centage of the milk of Ayrshire cows with advancing age.
are as follows, the fitting having been done by the method of least
squares.
From 2 to lo years of age:
;^' = 3.896-o.o343x.
From ID to i6 years of age:
y= 3.610— 0.002 8x.
SUMMARY
This paper presents the results of a biometrical analysis of variation in
the quantity per unit of time, and the quality, as indicated by fat per-
centage, of the milk of Ayrshire cows. Its purpose is to establish normal
constants for interindividual variation in these characters, to serve as a
base of reference in future genetic studies on milk production.
Sept. IS, 1919 Variation in Milk of Ayrshire Cows 319
The chief results of this first part of the investigation may be sum-
marized as follows:
(i) The mean weekly yield and fat percentage of the milk change in a
considerable degree and definite manner with increasing age of the cow.
(2) The weighted mean standard deviation and coefficient of varia-
bility for mean weekly yield of cows of any given age are 2.806 gallons
and 17.081 per cent respectively. Reasons are given tending to show
that these may be taken as very close approximations to true normal
values. For cows of all ages lumped together the corresponding values
are 3.329 gallons and 20.816 per cent.
(3) For fat percentage the weighted mean values for cows of any given
age are as follows: Mean = 3.738, standard deviation = 0.330, and coeffi-
cient of variation =8.827.
(4) A table is presented (p. 18) showing the relative variability of milk
production as compared with other physiological characters. The udder
as a secreting organ is compared with the oviduct of the hen; and it is
shown that the oviduct considered as a mechanism operates with some-
what less variability than does the udder, having regard to the absolute
weight of the product in the two cases.
(5) Evidence is presented which indicates that about one-half of the
observed variation in milk production results from the varying genotypic
individuality of the animals with respect to this character and that the
other half results from varying environmental influences.
(6) Milk production curves, analytically considered, tend definitely
toward positive skewness. This is true in respect to yield and to
quality. The weighted mean value of the skewness for mean weekly
yield is found to be 4-0.1047, and that for fat percentage -{-0.1338.
(7) Evidence is presented which indicates that selection can have had
little if anything to do with determining the direction or the amount of
skewness showTi by milk production curves.
(8) The curves for milk yield tend on the whole to fall more frequently
in unlimited range types, while those for fat percentage tend more to
limited range types. The estimation of range ends given by the theoreti-
cal curves are, on the whole, good.
(9) In general the tendency of milk yield curves is toward the
leptokurtic condition — that is, they are more peaked than the corre-
sponding normal curves would be. Fat percentage curves do not show
any definite tendency with respect to kurkosis.
(10) Certain of the milk yield curves were dissected into two normal
curves by Pearson's method. The resulting graduation was not so good
as that given by the appropriate unimodal skew frequency curv^e. There
is no evidence that variation curves for milk production curves are
biomodal.
(11) The change in mean weekly yield of milk with advancing age is
found to be represented by a logarithmic curve, and to be in accordance
122502° — 19 6
320 Journal of Agricultural Research voi. xvii. no. 6
with a law which may be stated in this way: The absolute amount of
milk produced per unit of time increases with the age of the cow until the
maximum is reached, but the rate of increase diminishes with advancing
age until the absolute maximum of production is reached. After the
time of maximum productivity, the absolute production per unit of time
decreases with advancing age at a continually increasing rate.
(12) The mean fat percentage of the milk was found to decline with
advancing age until the tenth year of the cow's life is reached. From
that point on, the fat percentage remains about constant through the
remainder of the milking life of the cow.
LITERATURE CITED
(i) Curtis, Maynie R.
1914. A BIOMETRICAL STT.TDY OF EGG PRODUCTION IN THE DOMESTIC FOWL.
IV. FACTORS INFLUENCING THE SIZE, SHAPE, AND PHYSICAL CONDI-
TION OF EGGS. In Arch. Entwicklungs-mech. Organ., Bd. 39, Heft 2/3,
p. 217-327, 18 fig. pi. 6-10.
(2) Gavin, William. ^
1912. THE INTERPRETATION OF MILK RECORDS. In Jour. Roy. AgT. Soc,
V. 73. P- 153-174.
(3)
1913. STUDIES IN MILK records: ON THE ACCURACY OP ESTIMATING A COW'S
MILKING CAPABILITY BY HER FIRST LACTATION YIELD. In Jour. Agr.
Sci., V. 5, pt. 4, p. 377-390-
(4) Greenwood, M., Jr.
1904. A FIRST STUDY OF THE WEIGHT, VARIABILITY, AND CORRELATION OF THE
HUMAN VISCERA, WITH SPECIAL REFERENCE TO THE HEALTHY AND
DISEASED HEART. In Biomctrika, v. 3, pt. i, p. 63-83.
(5) Greenwood, M., Jr., and Brown, J. W.
1913. A SECOND STUDY OF THE WEIGHT, VARIABILITY, AND CORRELATION OF
THE HUMAN viscER.\. In Biometrika, v. 9, pt. 3/4, p. 473-485.
(6) HowiE, John.
I91O. REPORT ON MILK RECORDS FOR SEASON I909. RECORD OP 9,202 COWS.
[Ayrshire Cattle Milk Records Committee.] 313 p. Kilmarnock.
(7) MacdonELL, W. R.
1904. A STUDY OP THE VARIATION AND CORRELATION OF THE HUMAN SKULL,
WITH SPECIAL REFERENCE TO ENGLISH CRANIA. In Biometrika, V. 3,
pt. 3, p. 191-244, illus., 50 pi., 8 fold. tab.
(8) Miner, John Rice.
1915. fitting LOGARITHMIC CURVES BY THE METHOD OF MOMENTS. In JoUf.
Agr. Research, v. 3, no. 5, p. 411-423. Literature cited, p. 422-423.
(9) Pearl, Raymond.
1905. BIOMETRICAL STUDIES ON MAN. I. VARIATION AND CORRELATION IN
BRAIN-WEIGHT. In Biometrika, v. 4, pt. i, p. 13-104.
(10)
19 13. CONSTANTS FOR NORMAL VARIATION IN THE PAT CONTENT OP MIXED MILK.
In Maine Agr. Exp. Sta. Bui. 221, p. 299-305.
1913. NOTE REGARDING THE RELATION OF AGE TO FECUNDITY, /n Science, n. s.
V. 37, no. 945, p. 226-228.
(12)— —
19 14. ON THE LAW RELATING MILK FLOW TO AGE IN DAIRY CATTLE. In PrOC.
Soc. Exp. Biol, and Med., v. 12, no. i, p. 18-19.
Sept. 15. I9I9 Variation in Milk of Ayrshire Cows 321
(13) Pearl, Raymond, and Patterson, S. W.
1917. THE CHANGE OF MLK FLOW WITH AGE, AS DETERMINED FROM SEVEN DAY
RECORDS OF JERSEY COWS. Maine Agr. Exp. Sta. Bui. 262, p. 145-162.
(14) , and Pearl, M. D.
1909. DATA ON VARIATION IN THE COMB OF THE DOMESTIC FOWL. 7w Biometrika,
V. 6, pt. 4, p. 420-432, 3 pi.
(15) , and Surface, Frank M.
1909. a BIOMETRICAL STUDY OF EGG PRODUCTION IN THE DOMESTIC FOWL. I.
VARIATION IN ANNUAL EGG PRODUCTION. U. S. Dept. Agr. Bur.
Anim. Indus. Bui. no, pt. i, p. 1-80, 17 fig.
(16) , and Surface, Frank M.
I914. A BIOMETRICAL STUDY OF EGG PRODUCTION IN THE DOMESTIC FOWL. III.
VARIATION AND CORRELATION IN THE PHYSICAL CHARACTERS OP THE
EGG. U. S. Dept. Agr. Bur. Anim. Indus. Bui, no, pt. 3, p. 171-241,
14 fig.
(17) Pearson, Karl.
1894. contributions to the mathematical theory of evolution. In Phil.
Trans. Roy. Soc. London, A, v. 185, pt. i, p. 71-110, pi. 1-5.
(18)
(19)
(20)
(21)
(22)
(23)
1895. contributions to the mathematical theory of evolution. II.
SKEW variation in HOMOGENEOUS MATERIAL. In Phil. Trans. Roy.
Soc. London, A, v. 186, pt. i, p. 343-414, pi. 7-16.
1897. THE chances OF DEATH. ... v. I. London.
1901. MATHEMATICAL CONTRIBUTIONS TO THE THEORY OF EVOLUTION. X. SUP-
PLEMENT TO A MEMOIR ON SKEW VARIATION. In Phil. Trans. Roy.
Soc. London, A, v. 197, p. 443-459.
1902. ON THE systematic FITTING OF CURVES TO OBSERVATIONS AND MEASURE-
MENTS. In Biometrika, v. i, pt, 3, p. 265-303; v. 2, pt. i, p. 1-23,
illus.
1909. A BIOMETRIC STUDY OF THE RED BLOOD CORPUSCLES OF THE COMMON
TADPOLE (rANA TEMPORARIA), FROM THE MEASUREMENTS OP ERNEST
WARREN, D. sc. In Biometrika, v. 6. p. 402-419.
19 10. NOTE ON THE SEPARATE INHERITANCE OF QUANTITY AND QUALITY IN COW'S
MILK, /n Biometrika, v. 7, pt. 4, p. 548-550.
(24) , and Lee, Alice.
1903. ON THE LAWS OP INHERITANCE IN MAN. I. INHERITANCE OF PHYSICAL
CHARACTERS. In Biometrika, v. 2, pt. 4, p. 357-462, illus., 7.8 tab.
(25) POWYS, A. O.
1905. DATA FOR THE PROBLEM OF EVOLUTION IN MAN. ON FERTILITY, DURATION
OP LIFE AND REPRODUCTIVE SELECTION. In Biometrika, v. 4, pt. 3,
p. 233-285, 17 fig.
(26) Speir, John.
1909. REPORT ON MILK RECORDS FOR SEASON I908. RECORD OP 8,132 COWS.
[Ayrshire Cattle Milk Records Committee], p. 1-268. Kilmarnock.
(27) Surface, F, M.
1909. fecundity OP SWINE. In Biometrika, v. 6, pt. 4, p. 433-436.
(28) Vigor, H. D.
1913. THE CORRELATION BETWEEN THE PERCENTAGE OP MILK FAT AND THE
QUANTITY OP MILK PRODUCED BY AYRSHIRE COWS. Sup. Jour. Bd.
Agr. [London], no. 11, 28 p.
322 Journal of Agricultural Research voi. xvn. no.6
(29) Weldon, W. F. R.
1907. on heredity in mice from the records op the late w. f. r. weldon.
i. on the inheritance of the sex-ratio and op the size of litter.
In Biometrika, v. 5, pt, 4, p. 436-449.
(30) Wilson, James.
I9IO. THE SEPARATE INHERITANCE OF QUANTITY AND QUALITY IN COWS' MILK.
In Sci. Proc. Roy. Dublin Soc, v. 12, no. 35, p. 470-479.
INDEX
Page
Abbott, W. S., et al. (paper): Derris as an In-
secticide 177-200
Abortus infection of bulls 239-246
Acetic-acid test, used in analyses of meat ex-
tracts 13-14
Aceton, "derrid" soluble in 179
Acid —
acetic, used in analyses of meat extracts . . . 13-14
phosphoric, in meat extracts 4, 7-12, 14, 16
sulphuric, effect on camphor seed 226, 237
Acidity, soil, influence on rootrot of tobacco. 53-60
Agglutinins, test for in Bacterium abortus. . 239-246
Alcohol, "derrid ' ' soluble in 179
Alkalinity, soil, influence on rootrot of to-
bacco 53-60
Amsotasenatoria, toyiity of Derris to 196-200
Ants. See Monomorium pharaonis.
Aphids —
toxity of Derris to 181-200
transmission of mosaic by 256-266
Aphis —
false cabbage. See Rhopalosiphum pseudo-
brassicae.
false turnip. See Rhopalosiphum pseudo-
brassicae.
green apple. See Aphis pomi.
green peach. See Myzus persicae.
green potato. See Macrosiphum solanifolii.
melon. See Aphis gassy pit.
pink potato. See Macrosiphum solanifolii.
spinach. See Myzus persicae.
sunflower. See Aphis heliantki.
tulip-tree. See Macrosiphum liriodendri.^
Aphis —
gossypii, toxity of Derris to 190-200
helianthi, toxity of Derris to 190-200
mellifica, toxity of Derris to 181-183
pomi, toxity of Derris to 193-200
Tumicis, toxity of Derris to 193-200
spiraecola, toxity of Derris to 190-200
Appleman, Charles O., and Arthur, John M.
(paper): Carbohydrate Metabolism in
Green Sweet Com during Storage at Differ-
ent Temperatures 137-152
Arthur, John M., and Appleman, Charles O.
(paper): Carbohydrate Metabolism in
Green Sweet Corn during Storage at Differ-
ent Temperatures 137-152
Ascochyta sp., on seeds, use of formaldehyde
vapor against 36
Ash in meat extracts 3, 7-8, 14, 16
Asphalt —
disperse colloids in 167-176
ultra-microscopic examination of 167-176
Autographa brassicae, toxity of Derris to. . . 196-200
Ayrshire cows, variation in quantity and fat
content of milk 285-321
(3
Page
Bacillus caratcnorus, on seeds, use of formalde-
hyde vapor against 36
Bacterium abortus Infection of Bulls
(paper) 239-246
Bedbugs. See Cim.ex Icctularius.
Beetles, potato. See Leptinotarsa decemli-
neata.
Benzol, "derrid" soluble in 179
Bitumen, examined by ultra-microscope. . . 167-176
Blattella germanica, toxity of Derris to 192-200
Buck, J. M., Creech, G. T., and Ladson, H. H.
(paper): Bacterium abortus Infection of
Bulls 239-246
Bulls, abortus infection of 239-246
Cabbage —
aphis, false. See Rhopalosiphum. pseudo-
hrassicae.
worms. See Autographa brassicae.
Calcium —
carbonate —
effect on plant growth and composition . 90-100
in Oregon soils 89
used in grinding cell for ultra-microscopic
examination 169-171
oxidin Oregon soils 89
sulphate, effect on plant growth and com-
position 90-100
Camphor seed , effect of removing the pulp
from 223-237
Carbohydrate >Ietabolism in Green Sweet
Corn during Storage at Different Temper-
atures (paper) 137-152
Carbohydrates in meat extracts 16-17
Carbon —
dioxid, relation to storage of sweetcom 278
disulphid, "derrid" soluble in 179
Caterpillars, tussock-moth. See Hemero-
campa leucostigma.
Certain Relationships between the Flowers
and Fruits of the Lemon (paper) 153-165
Chicken —
lice. See Mallophaga.
mites. See Derma nyssus gallinae .
Chlorid, sodium, in meat extracts 3,7-8,14,16
Chloroform, " derrid " soluble in 179
Cimex leciularius , toxity of Derris to 192-200
Colletotrichum gloeosporioidcs, on seeds, use of
formaldehyde vapor against 36
Collins, G. N. (paper): Structure of the Maize
Ear as Indicated in Zea-Euchlaena Hy-
brids 127-135
Colloids-
method of counting 172-173
ultra-microscopic examination of 167-176
Complement-fixing bodies in Bacterium
abortus 239-246
Copper test of meat extracts 15-16
23)
324
Journal of Agricultural Research
Vol. XVII
Page
Com, sweet. See Sweetcom.
Cows, Ayrshire, variation in quantity and fat
content of milk 285-321
Creatin in meat extracts 6-16
Creatinin in meat extracts 6-16
Creech, G. T., et al. (paper): Bacterium abor-
tus Infection of Bulls 239-246
Clcnocephalus cants, toxity of Derris to . . . 192-200
Dalana niinistra larvse, toxity of Derris to. . 196-200
Datanas. See Datana minisiro.
Deguclia spp. Syn. Derris spp.
Dehiscence of anthers, effect of rain i lo-i 13
Dermanyssus gallinae, toxity of Derris to. . . 192-200
'"Derrid," substance derived from Derris
elliptica 179
Derris as an Insecticide (paper) 177—200
Derris —
elliptica —
fish poison 177-178
insecticide 177-200
Tobusla, insecticide 177-200
scandens, insecticide 177-200
uliginosa, insecticide 177-200
spp., extracts from, methods of preparing 180-191
Derris —
contact insecticide 189-200
histological methods of tracing poison of. 197-199
pharmacological effects of 197-199
stomach poison 182-200
Disinfection of seed by formaldehydevapor. . . 33-39
Disperse colloids, ultra-microscopic exam-
ination of 167-176
Dog fleas. See Ctenocephalus canis.
Dorsey, M. J. (paper): Relation of Weather
to Fruitfulness in the Plum 103-126
Effect of Removing the Pulp from Camphor
Seed on Germination and the Subsequent
Growth of the Seedlings (paper) 223-237
Effects of Heat on Trichinae (pajjer) 201-221
Emery, James A., and Henley, Robert R.
(paper): Meat Extracts, Their Composition
and Identification 1-17
Errata and authors' emendations iii
Euchlaena —
alicoles —
separate 128-133
two-ranked 128-130
spikelets single 128-133
Euchlaena inexicana —
structure of ear of Zea mays compared to. 127-135
pistillate inflorescence of 127-135
Ether, "derrid" insoluble in 179
Ewes' milk, quantity and composition of . . . . 19-32
Extracts, meat. See Meat extracts.
False cabbage aphis. See Rkvpalosiphum
pseudobrassicae.
False turnip aphis. Rhopalosiphum pseudcf-
brassicae.
Fertilization, in plum, limiting factors 1 18-124
Foliage inoculations with mosaic 253-255
Folsom, Donald, et al. (paper): Investiga-
tions on the Mosaic Disease of the Irish
Potato 247-274
Formaldehyde vapor, seed disinfection by . . . 33-39
Free-reducing substances, loss from sweetcorn
during storage 142-152
Page
Frequency distributions for variation in milk
yield 287-321
Fruit buds of lemon, seasonal distribution of 154-156
Fruit-
lemon —
relation to flowers 153-165
time required to mature 161-163
setting of, in plum 105-123
Fusarium vasinfeclum, on seeds, use of formal-
dehyde vapor against 36
Germination, camphor seed, effect of —
drying 226
fermentation 226. 237
freezing 224-237
heat 226, 237
picking from ground 224-237
removing the pulp 223-237
soaking in sulphuric acid 226, 237
soaking in water 224. 237
time of planting 225. 237
Glacial acetic acid, "derrid" soluble in 179
Grafting, transmission of mosaic by 251-253
Green apple aphis. See Aphis pomi.
Green peach aphis. See Myzus persicae.
Green potato aphis. See Macrosiphum solani-
folii.
Hartman, R. E., and Johnson, James (paper):
Influence of Soil Environment on the Root-
rot of Tobacco 4 1-86
Hawkins, Lon A., et al. (paper): Investiga-
tions on the Mosaic Disease of the Irish
Potato 247-274
Heat-
effect on —
toxity of Derris extract 182-184
trichinae 201-2? i
See also Temperature.
Hemerocampa leucostigma, toxity of Derris
to 182-200
Henley, Robert R., and Emery, James A.
(paper): Meat Extracts, Their Composition
and Identification 1-17
Higgins, C. H., and Stevens, Neil E. (paper):
Temperature in Relation to Quality of
Sweetcom 275-2S4
Hildebrandt, F. Merrill, et al. (paper):
Investigations on the Mosaic Disease of the
Irish Potato 247-.; 74
Hill selection, effect upon mosaic of Solanum
tuberosum 267-370
Histological examination for Bacterium
abortus 242-246
Histological methods of tracing Derris
poison 197-199
Honeybee. See A phis mcllifica.
House flies. See Musca domestica.
Hybrids, Zea-Euchlaena, structure of ear. . 127-135
Hyphanlria cunea, toxity of Derris to 182-200
Iddings, E. J., and Neidig, Ray E. (paper):
Quantity and Composition of Ewes' Milk:
Its ReUtion to the Growth of Lambs i9-3»
Illinoia, sp., toxity of Derris to 190-200
Inflorescence —
lemon, size and productiveness of 156-160
maize ear, structure of 127-13S
Apr. 15-Sept. IS, I9i<)
Index
325
Page
Influence of Soil Environment on tlie Rootrot
of Tobacco (paper) 41-86
Inoculation with mosaic through plant
juices 253-255
Insecticide, Derris 177-200
Investigations on the Mosaic Disease of the
Irish Potato (paper) 147-274
Irish potato. SceSolanum tuberosuTn.
Johnson, James, and Hartman, R. E. (paper):
Influence of Soil Environment on the Root-
rot of Tobacco 41-86
Ladson, H. H., et al. (paper): Bacterium
abortus Infection of Bulls 239-246
Lambs, growth of, relation to ewes' milk .... 19-32
Larvae —
potato beetle, toxity of Derris to 195-200
tent caterpillar, toxity of Derris to 195-200
trichinae —
decapsuled, effects of heat on 204-212
encysted, effects of heat on 212-221
Lemon —
fruit buds, seasonaldistributionof 154-156
fruit, time required to mature 161-162
inflorescences, size and productiveness of. 156-160
relation between flowers and fruits of 153-165
Lepidosaphes tdmi, toxity oiDerristo 193-200
Leptinoiarsa decemlincata, toxity of Derris to,
187-200
Lord, E. C. E. (paper): Ultra-Microscopic
Examination of Disperse Colloids Present in
Bituminous Road Materials 167-176
Mclndoo, N. E., Sievers, A. F., and Abbott,
"W. S. (paper) : Derris as an Insecticide .... 1 77-200
Macrosiphum —
liriodendri, toxity of Derris to 189-200
solanifolH, carrier of mosaic of Solanum
tuberosum 256-266
Magnesia hydrate, used in grinding cell for
ultra-microscopic examination 169-171
Maize. See Zea mays.
Malacosoma americana, larvae, toxity of Der-
ris to 195-200
Mallophaga, toxity of Derris to 192-200
Mealybug. See Pseudococcus citri.
Meat extracts —
acetic-acid test of 13-14
ash in 3, 7-8, 14, 16
carbohydrates in 16-17
clarification of 2
copper test of; 16-17
creatin in 6-16
creatinin in 6-i5
Molisch test of 15-16
nitrates in 6-7
nitrogen in 4-17
normitrogenous organic matter in 5-10
phosphoric acid in 4, 7-12, 14, 16
phosphorus in 4-16
physical characteristics of 12-13
portions cf carcass used in preparation of . . i
preparation of —
commercial method i
laboratory method 2
purins in 5-8
qualitative investigation of 13-1 7
quantitative investigation cf 3-12
sodium chlorid in 3, 7-8, 14, 16
water in 3,7
Page
Meat Extracts, Their Composition and Iden-
tification (paper) 1-17
Melon aphis. See Aphis gossypii.
Metabolism, carbohydrate, in sweetcom
during storage 137-152
Milk-
cows' —
effect of age on frequency distributions for
variation in 285-321
variation of Ayrshire cows in fat content
of 285-321
variation of Ayrshire cows in quantity
of 285-321
ewes' , quantity and composition of 19-32
Miller, H. G. (paper): Relation of Sulphates
to Plant Growth and Composition 87-102
Miner, John Rice, and Pearl, Raymond
(paper): Variation of Ayrshire Cows in the
Quantity and Fat Content of Their Milk. . 2S5-321
Moisture content of soil, relation to rootrot of
tobacco 49-53
Molisch test, used in analyses of meat extracts. 15-16
Mcnilia fruciigena, on seeds, use of formalde-
hyde vapor against 36
MonoTnorium pliaraonis, toxity of Derris to. 197-200
Mosaic disease of Solanum tuberosum —
effect of —
hill selection 267-270
roguing 270-271
effect on —
starch content of plant 266-267
sugar content of plant 266-267
yield 24S-249, 269
geographical distribution 248
symptoms 249-250
transmission by —
aphids 256-266
grafting 251-253
plant juices 253-255
tubers 250, 253-254, 261-262
Mw^coifomei/ica, toxityofDerristo 192-200
Myzus persicae—
carrier of mosaic of Solanum tuberosum . . 256-266
toxity of Derris to 1S9-200
Nasturtium aphids. See Aphis rumicis.
Neidig, Ray E., and Iddings, E. J. (paper):
Quantity and Composition of Ewes' Milk:
Its Relation to the Growth of Lambs 19-33
Nicotianatabacuin, rootrot of^
effect of transplanting diseased seedlings Si-83
influence of^
clay and sand content of soil 76-78
organic matter in soil 73-76
soil acidity 53-60
soil alkalinity 53-60
soil compactness 80-Si
soil fertility 78-So
soil moisture 49"53
soil temperature 60-73
Nitrates in meat extracts 6-7
Nitrogen in meat extracts 4-17
Nonnitrogcnous organic matter in meat ex-
tracts 5-10
Oak worms. SecAnisotasenaloria.
Orthezia insisnis, toxity of Derris to 193-200
Oxygen, relation to storage of sweetcorn 278
326
Journal of Agricultural Research
Page
Oyster-shell scale. See Lepidosaphesulmi.
Peach aphis, green. See Myzus persicae.
Pearl, Raymond, and Miner, John Rice (pa-
per): Variation of Ayrshire Cows in the
Quantity and FatContent of Their Milk.. 285-321
Petroleum ether, "derrid " insoluble in 179
Phosphoric acid. See Acid, phosphoric.
Phosphorus in meat ejctracts 4-16
Phosphorus pentoxid in Oregon soils 89
Pink potato aphis. See Macrosiphum sola-
nifolii.
Plants, growth and composition of, relation
of sulphates to 87-102
Plum-
relation of weather to fruitfulness of 103-126
setting of fruit, effect of —
rain 110-118
stmshine no
temperature 108-110
wind 107-10S
Pollen of plum —
injury by rain 113-115
longevity 119
Pollen-tube growth, rate of in plum 120-123
Polysaccharides, increase in sweetcom during
storage 149-152
Pork, temperature required tokilltrichinse. 201-221
Potassium oxid in Oregon soils 89
Potato aphis —
green. See Macrosiphum solanifolii.
pink. See Macrosiphum solanifolii.
Potato beetles. See Laptinotarsa decetnlineaia.
Potato, Irish. See Solanuin tuberosum.
Prodenia ornitkogalli, larvse, toxity of Derris
to 197
Pseudococcus citri, toxity of Derris to 193-200
Pulp, effect of removing from camphor seed. 223-237
Purins in meat extracts 5-8
Quantity and Composition of Ewes' Milk:
Its Relation to the Growth of Lambs
(paper) 19-32
Ransom, B. H., and Schwartz, Benjamin
(paper): Effects of Heat on Trichinae 201-221
Red spiders. See Tetranychus bimaculaius.
Reed, Howard S. (paper): Certain Relation-
ships between Flowers and Fruits of the
Lemon 153-165
Relation of Sulphates to Plant Growth and
Composition (paper) 87-1C2
Relation of Weather to Fruitfulness in the
Plum (paper) 103-126
Rhopulosiphum psetidobrassicae, toxity of
Derris to 190-200
Roaches. See BlaUclla germanica.
Road materials, ultra-microscopic examina-
tion of disperse colloids in 167-176
Roguing, effect on mosaic of Solanum tuber-
osum 270-271
Rootrot of tobacco. See Nicoiiana lobacum,
rootrot of.
Russell, G. A. (paper): Effect of Removing
the Pulp from Camphor Seed on Germina-
tion and the Subsequent Growth of the
Seedlings 223-237
Page
Schultz, E. S., Folsom, Donald, Hildebrandt,
F. Merrill, and Hawkins, Lon A. (paper):
Investigations on the Mosaic Disease of the
Irish Potato 247-274
Schwartz, Benjamin, and Ransom, B. H.
(paper): Effects of Heat on Trichinae. . . . 201-221
Season, realtion to fruit buds of lemon 154-156
Seed, camphor, effect of removing pulp. . . . 223-237
Seed Disinfection by Formaldehyde Vapor
(paper) 33-39
Serum, blood, test of in Bacterium abortus. 239-246
Sheep. See Lambs.
Sievers, A. F., et al. (paper): Derris as an
Insecticide 177-200
Sodium —
chlorid in meat extracts 3, 7-8, 14, 16
nitrate, effect on plant growth and com-
position 90-100
sulphate, effect on plant growth and com-
position 90-100
Soil-
acidity, influence on rootrot of tobacco 53-60
alkalinity, influence on rootrot of tobacco. . 53-60
clay and sand content, influence on rootrot
of tobacco 76-78
compactness, influence on rootrot of to-
bacco 80-81
fertility, influence on rootrot of tobacco. . . . 78-80
moisture content, influence on rootrot cf to-
bacco 49-53
organic matter, influence on rootrot of to-
bacco 73-76
temperature, influence on rootrot of to-
bacco 60-73
Solanum. tuberosum., mosaic disease of 247-274
effect of hill selection 267-270
effect of roguing 270-271
effect on —
starch content of plant 266-267
sugar content of plant 266-267
yield - 248-249, 269
geographical distribution 248
symptoms 249-250
transmission by —
aphids 256-266
grafting 251-^53
plant juices 253-255
tubers 25o,.2S3-254, 261-262
Spinach aphis . See Myzus persicae.
Starch content of foliage, effect of mosaic
upon 266-267
Starch, formation in sweetcorn during
storage 149-152
Stevens, Neil E.,andHiggins,C.H. (paper):
Temperature in Relation to Quality of
Sweetcom 2 75-284
Stigma of plum —
effect of rain on 1 15-1 18
length of receptive period 119
Storage , effect on carbohydrate metabolism
in sweetcorn 137-152
Structure of the Maize Ear as Indicated in
Zea-Euchlaena Hybrids (paper) 127-135
Style, abscission of in plum 119-120
Sucrose, lossfrom sweetcom during storage. 142-152
Apr. 15-Sept. 15, 1919
Index
327
Page
Sugar content of foliage, effect of mosaic
upon 266-267
Sugar —
loss from sweetcorn after picking 276-278
loss from sweetcorn during storage 142-152
Sulphates, relation to plant growth and
composition 87-102
Sulphur in Oregon soils 89
Sulphuric acid. See Acid, sulphuric.
Sunflower aphis. See Aphis helianlhi.
Sweetcorn —
loss of sugar after picking 276-278
storage of —
carbohydrate metabolism during 137-152
formation of starch during 149-152
loss offree-reducing substances during. 142-152
loss of sucrose during 142-152
loss of sugar during 142-152
relation of—
carbon dioxid to 278
oxygen to 278
temperature to loss of sugar 276-283
Temperature —
effect on —
carbohydrate metabolism in sweetcorn
during storage 137-152
germination of camphor seed 232-237
setting of fruit in plimi 108-110
toxity of Derris extract 182-184
trichinae 201-221
relation to loss of sugar from sweetcorn . . 276-283
soil, influence on rootrot of tobacco 60-73
Temperature in Relation to Quality of
Sweetcorn (paper ) 2 75-284
Teosinte. See Euchlaena tncxicana.
Tdranychus bimacvlatus , toxity of Derris to 193-200
Thermal death point of trichinae 201-221
Thielavia basicola, cause of rootrot of tobacco . 41-86
Page
Thomas, Cecil C. (paper): Seed Disinfection
by Formaldehyde Vapor 33*39
Tobacco. See Nicoliana tahacum.
"Toeba." SeeDcrriVspp.
Trichinae. See Trichinella spiralis.
Trichiitella spiralis —
effect of heat on 201-221
larvae —
decapsuled, effects of heat on 204-212
encysted, effects of heat on 212-221
"Tuba." See £>err!.f spp.
"Tubain, " substance derived from Derris
elliptica 179
Tubers, transmission of mosaic by
250, 253-254, 261-262
Tulip-tree aphis. See Macrosiphutn lirio-
dendri.
Turnip aphis, false. See Rhopalosipkum
pseudobrassicae.
Tussock-moth caterpillars. See Hemerocam-
paleuco stigma.
Ultra-Microscopic Examination of Disperse
Colloids Present in Bituminous Road
Material(paper) 167-176
Variation of Ayrshire Cows in the Quantity
and Fat Content of Their Milk (paper) . . 285-321
Water in meat extracts 3)7
Weather . relat ion to f ruitf ulness in the plum 103-1 26
Web worms, fall. See Hyphantria cunea.
Zea- Euchlaena hybrids, structure of ear. . . 127-135
Zea —
mays —
alicoles, many-ranked 128-134
alicoles, yoked 128-134
spikelets, paired 128-134
raviosa, intermediate between Euchlaena
and Zea mays 130
New York Botanical Garden Library
3 5185 00263 3921
:-;j:3;>iJi:j)
iiiiiiiliiliiiiiiiiiiiii
Slllf
■::iiii;i?i?i?5?:^-
iiliiiiiiii
iijlliiii
iiliiiilii
.:: lilfc^^^^^^
iiliiii.
iiijipiipiji lijiiipj
j:";;:::;;'
liiilBil