UNIVERSITY OF MASSACHUSETTS
LIBRARY
S
73
E5
no. 4-15
1921-28
massachusetts
Agricultural Experiment Station
TECHNICAL BULLETIN No. 4 NOVEMBER, 1921
Development and Pathogenesis
OF THE
ONION SMUT FUNGUS
By P. J. ANDERSON
Onion smut is the most destructive of all onion diseases in
New England. In the Connecticut Valley it is probably re-
sponsible for more loss to the growers than all the other diseases
of this crop combined. This paper embodies the results of funda-
mental work on a project having for its chief aim the control
of onion smut. Beginning with the germination of the spores,
the development of the fungus is followed through the saprophytic
stage, infection of the host, distributive stage within the host,
and final sporogenesis.
Requests for Bulletins should be addressed to the
AGRICULTURAL EXPERIMENT STATION
AMHERST, MASS.
Publication of this Document
approved by the
Supervisor of Administration.
CONTENTS.
II.
III.
IV.
V.
VI.
VII.
VIII.
I
PAOB
Introduction, .......... 99
Germination of the spores, ........ 100
Review of the literature on essential conditions for smut sporo
germination, ......... 100
Essential conditions for germination of Urocystis crpulac spores, . 104
The process of germination, ....... 108
Comparison with the germination process in other species of Uro-
cystis, . . . . . . . . . . 110
Comparison with the process as described by Thaxter, . . Ill
Saprophytism, . . . . . . . . . .112
Isolation, .......... 112
Cultural characters, . . . . . . . .112
Effect of freezing the cultures, . . .115
Microscopic characters of the mycelium in culture, . . .115
Fate and function of the detached hyphal cells, . . .116
Life in the soil, . . . . . . . .117
Summary of the saprophytic stage m the natural life history, . 117
Infection, ........... 118
Development of the onion seedling, . . . . .118
Period of susceptibility, ....... 120
Point of infection, ........ 122
Character of the inoculum, ....... 123
Method of entrance, ........ 123
Passage through the epidermal cell.s, ..... 124
Multiple infection, ........ 125
Incubation period, ......... 125
Young hyphse in the intercellular spaces, .... 125
Haustoria, .......... 126
Absorptive hyphal expansions, ...... 127
Progressive infection of new leaves, ..... 128
Sporogenesis, .......... 128
Summary, . . . . . . . . . . .131
Litarature cited, .......... 132
TECHISriCAL BULLETIN" J^o. 4.
DEPARTMENT OF BOTANY.
DEVELOPMENT AND PATHOGENESIS OF THE
ONION SMUT FUNGUS.^
BY P. J. ANDERSON.
I. Introduction.
Onion smut is the most destructive of all onion diseases in New Eng-
land. In the Connecticut Valley it is probably responsible for more loss
to the growers than all the other diseases of this crop combined. Despite
the fact that a method of control by the use of formaldehyde has been
developed, manj'- fields are now planted to other less profitable crops on
account of the ravages of smut; every year sees fields plowed up because
smut has so reduced the stand that it is not worth while to tend them;
more important in the aggregate, perhaps, is the smaller toll which the
disease exacts from each onion grower throughout considerable sections
of the valley.
Investigation of the disease yviih the primary object of finding better
methods of control was begun by the Department of Botanj'- of the Massa-
chusetts Agricultural Experunent Station in 1918, and has been continued
to date. Since control measures are necessarily conditioned by the nor-
mal life history of the pathogene, and since a review of the published re-
search of other investigators showed that the development of the fungus
had been inadequately^ studied, this phase of the problem has been made
the subject of no inconsiderable part of the writer's stud3^ Results of the
work which deals directly with control are to be presented in another
pubUcation. The present paper concerns certain phases of the life cycle
of the causal organism {Urocystis cefXilae Frost) in which it seemed to the
writer that further investigation was desirable. Beginning mth germina-
tion of the spores, the development of the fungus will be followed through
its saprophytic stage, infection of the host, distributive stage within the
host and final sporogenesis.
• This paper embodies the results of preliminary and fundamental work on a project having
for ita chief aim the control of onion smut. A report on the more practical phases of this project
is to be published shortly.
100 TECHNICAL BULLETIN 4.
II. Germination of the Spores.
The spore of Urocystis cepulae is compound, having one large central
fertile cell to the surface of which are attached 15 to 40 smaller hemi-
spherical sterile cells. There are said sometimes to be two fertile cells at
the center, but in a thiee-year study of the fungus the writer does not
remember ever having seen a spore ^ith more than one. To conform to
the nomenclature of certain other genera of smuts, the entire structure is
usually called a spore ball, the peripheral cells being termed pseudospores.
Since we have here only one cell capable of germination, it is perhaps
better to term the whole structure a spore and then distinguish between
fertile and sterile cells. The hemispherical cells are attached to the fertUe
cell by their flat surfaces, but do not cover it entirely. They stand apart
as indicated in Fig. 1 (page 109). The sterile cells are tinted brown, while
the central cell is a more solid opaque brown. Sterile cells average 5/z in
diameter by 4.25/i in height. The fertile cell is usually spherical, but
frequently oval or ovate, averaging about 12/i in diameter. The entire
spore averages about 19/^ in diameter.
For the germination of most fungous spores it is only necessarj'' to place
them when mature in a drop of water, and, after a few hours, or, at most,
a few days, the whole process may be watched under the microscope.
But for Urocystis cepulae the case is not so simple. Germination tests,
conducted in the same way in which the writer had brought to germina-
tion the spores of many species of fungi, were entirely vrithout result for
the onion smut fungus. Apparently there are other essential conditions
which had not been obtained in these trials. This preliminary failure led
to a thorough search through the literature to find what conditions were
essential for the germination of spores of other species of Qstilaginales.
It seemed probable that the same conditions which brought about germi-
nation in other smuts might also be successfully applied to Urocystis cepulae.
A condensed summary of the Uterature of this phase is given below, fol-
lowed by a description of the experiments with the spores of Urocystis
cepulae.
Revieiv of the Literature on Essential Conditions for Styntt Spore Germination.
The Water Require^nent. — No spores uill germinate without water in
some form, sometimes, to be sure, merely as vapor in a saturated air. In
the simplest cases, and, in fact, for the majoritj^ of the smut fungi, it is
only necessary to immerse the spores, as soon as mature, in a drop of water
on a shde, or in a hanging drop. Enough air to satisfy all requirements
seems to be present dissolved in the water, or else the spores remain on the
surface of the drop. Brefeld (3), in his experiments, germinated the spores
in a film of water which adhered to the inside of the walls of flat glass
chambers after the bulk of the liquid had drained out. This probably in-
sured greater access to air than where the hanging drop or drop on slide has
been used, and this fact should be kept in mind in interpreting his results.
ONION SMUT FUNGUS. 101
The following species can be germinated in water as soon as mature:
Cintradia deiisa McAlp. (11); C. Sorghi vulgaris (Till.) Clint., 12 hours, ^
(11); Entiilomn canesccns (14); Schizonella mdanocjrnmma D. C. (4);
Sorosporium Rcilianttm (Kiihn) McAlp., tap water, 17 hours (11); Tilletia
zonata Bref. (4); Urocystis occulta Wallr. (11); U. 'priviulicola Magn., 10
hours (14); U. Violae Sow. 5 days (4); Ustilago Avenae (Pers.) Jens., 6 to
S hours (8), (11) and others; U. Boutelonae humulis Bref. (4); U. Carbo
Tul., 6 to 10 hours (5); U. jloscvhnim, 5 to 6 hours (7) (he finds that
fresh spores germinate most quickly); IJ. grandis Fr., 24 hours (3); U.
longissima Sow., 3 to 4 hours (3), (7); U. major 24 hours (14); U. Panici
glaiici Wallr., 8 daj^s (3); U. Readeri Syd. (11); U. segetum, 6 to 8 hours,
"fresh spores germinate better" (14); Ust. violacea Pers. (3) and many
others. In the most favorable cases germination begins within two to
three hours, while at the other extreme McAlpine (11) mentions species
the spores of which did not begin to germinate until they had been in water
for several weeks, ^\^lere such extreme lengths of time are required, the
question arises as to whether this is not really the time required for the
weathering process such as takes place when they are kept in damp soil,
as in Brefeld's experiments.
Air. — Some spores require only a moist air for germination, and will
not germinate at all or only abnormally when immersed in water. Thus
Fischer von Waldheim (7) writes : —
For the normal germination of the different species of Ustilago, a certain quantity
of water or moisture is usually necessary. For this purpose, the spores need only
be placed in a drop of water, or upon moistened earth, or even merely in an atmos-
phere kept moist; for instance, under a glass globe placed over a dish of water.
But Tilletia and Urocystis germinate only in damp air (for instance, under the
glass globe mentioned) , and their germinating spores, coming in contact with water,
only show abnormal appearances.
In Brefeld's germinating apparatus the spores v/ere never entirely im-
mersed in water, but in the thin film chnging to the chamber walls must
have always had a sufficient quantity of air. This probably contributed
to his remarkable success in germinating the spores of a very large number
of species. McAlpine also found that he was able to secure germination
in many cases only by floating the spores in a watch glass over water.
Both Brefeld (3) and Fischer von Waldheim (7) mention the fact that the
spores of Tilletia caries germinate in damp air. Plowright (14) had a
similar experience vidth Tubercinia trientalis. McAlpine (11) was able to
germinate the spores of Tilletia Tritici (Bjerk) Wint. best by keeping them
on moist filter paper or blocks of plaster of Paris kept moist by capillary
water from a dish in which the blocks were partially immersed. He (11)
makes the following interesting observation on the necessity of air for
germination of spores of Ustilago Readeri Syd.: —
1 Figures after the species and not in parentheses indicate the time required for germination to
begin after the spores were placed in v/ater. Omission of them indicates that the investigator
gave no data as to time required. Numbers in parentheses refer to bibliography on pp. 132 and
133.
102 TECHNICAL BULLETIN 4.
Immersed in the liquid they do not germinate as readily as when floating on the
surface. Thus, after eighteen hours on one occasion, the spores in the water had
failed to germinate, while by simply altering the focus and examining the spores
on the surface they were all found, with very few exceptions, to have germinated.
In the descriptions of germination given by the majority of writers
there is no way of determining just how much influence the presence of
air had.
It seems probable that, in general, the presence of air is essential to
the germination of smut spores, but that different species vary in respect
to the amount required; some need scarcely any, others must have verj^
free access to air, and there are probably all gradations between these two
extremes.
Nutrient Solutions. — Very early in the investigation of smut spore
gennination it became apparent that the spores of some species could not
be germinated merely by placing them in water when mature. Conse-
quently solutions of various substances supposed to have nutritive quali-
ties have been tested for their ability to induce germination. Hallier (6),
in 1868, was apparently the first to use such solutions. He used a great
many substances such as albumin, starch, milk, sugar solution, etc. Others,
since then, have used almost every kind of a salt, acid, or other substance
for which one could imagine any germinative influence. One should con-
sult Osner's (13) bulletin on "Leaf Smut of Timothy" to gain some idea of
the number of substances that can be used for that purpose. IMcAIpine (11)
seems to have had most success with a hay infusion, although he also used
various other solutions. Sugar solutions and decoctions of the host plant
have proved fairly successful.
The nutritive solution which has been used most extensively and proba-
bly most successfully is the "nahrlosung, " a sterilized aqueous decoction
of horse dung which was employed first by Brefeld (3, 4). In this "nahr-
losung" he was able to bring to germination the spores of many species
which showed no sign of germination in water, e.g., Cintractia spinificis
(Ludw.) McAlp. (McAlpine (11) also confirmed Brefeld's results), Doas-
sansia Limosellae Kunze, Ustilago Andropogonis tubercidati Bref., Ust.
Arundinellae Bref., Ust. Coicis Bref., Ust. Cynadontis Hem., Ust. Ischdemi
Fckl., Ust. major Schroet., Ust. Panici leucophaei Bref., and Ust. T^dasnei
Klihn. Other species, e.g., Ust. Maydis, which gave scanty or only occa-
sional germination in water, germinated to almost 100 per cent in this
"nahrlosung." In almost every case the growth and size of the germ tube
(promycelium) was increased; and frequently sporidia were produced in
this nutritive solution where none at all were developed in water. On the
whole, however, it should be kept in mind that in by far the majority of
cases the function of the nutritive solution was to bring the germling to
complete development after it had started, rather than to cause it to start
in the first place. Only in the case of the comparatively few species men-
tioned above did he fail to get some germination in water also, and very
commonly the percentage of germination was as high in water as in "nahr-
ONION SMUT FUNGUS. 103
losunc;." On the other hand, he found that TiUclia Trilici would not
gonnuiate at all in nutritive solution, but could be germinated easih' in
water.
His experinieiits with nutritive solutions led Brefeld to believe that smut
spores in the soil are brought to germination and further development
through the influence of manure which has been used to fertilize the soil.
On this theory he explains the common observation of German farmers
that cereal smuts are more destructive on freshly manured fields.
Host Stimulus. — One might expect that some stimulus from the host
plant would be necessary for germination, and consequently that a decoc-
tion from the host, or the presence of bits of it in the germinative medium,
would be necessary for starting germination. Although such host decoc-
tions have been successfully used, we find in the literature no instance in
which the}'^ furnished the only conditions under which the spores would
germinate. There seems, then, to be no evidence to indicate that a smut
spore must be in close proximity to, or in actual contact with, its host
before it T\ill germinate.
Period of Rest. — But, even with the aid of nutritive solutions, and all
other conditions which have been tried, there is a considerable number of
species, the spores of which cannot be brought to germination immediately
after maturity. For these species, a period of "rest'" is necessary during
which the}'' must be exposed to certain natural conditions which operate
in some way to bring them into the proper condition for germination. For
our knowledge of this phase of the problem we are indebted, above all,
to Brefeld, and we cannot present it better than by quoting from his sum-
mary of it ( (4), page 128) : ^ -
Only a part of these forms germinate at once even in nutrient solution, more
rarely in water; many will not germinate at all, but must be made capable of
germination by special methods. . . . The spores of many species are so adapted
in their time of germination that they do not proceed at once, but only after passing
through a shorter or longer resting period. In cases of this kind one has only to
wait until after the expiration of the resting period in order to bring them to ger-
mination. But one would often wait long and in vain, if he only kept the spores
dry in the house. Under these circumstances, the external influences are not
brought to bear, which operate in nature during the period of rest, and which must
operate in order to bring about those changes on which the initiation of germination
depends. For the most part, when simply kept dry the spores die without ger-
minating, except in a few cases, as, for example, the corn smut, . . . but even
here germination is always incomplete. It is necessary to obtain the conditions
which in nature operate on the spores and influence them to germinate, if one
wishes to succeed in observing germination. The simplest method would be to
expose the spores in nature or leave them in their natural habitat and observe
from time to time whether germination has begun. But in most cases it is entirely
impossible in this way to get and keep the material pure.
He then describes in detail his method of keeping the material in steri-
lized damp sand in pots in a cool cellar. Then he continues: —
' Translated by P. J. Anderson.
104 TECHNICAL BULLETIN 4.
By this method it has been possible to bring to germination most spores which
otherwise would not germinate. The length of time required to bring about ger-
mination varies greatly. The spores of some species usually germinate after a
few months, others after a half or an entire year, others require several years be-
fore germination, some even five years. ... In this methodical way, which is, to
be sure, nothing but an imitation of what takes place in nature, ultimately all
spores can be induced to germinate. Therefore it can be scientifically proved that
the earlier or later germination is only an adaptation, a resting period, which under
the natural conditions must be passed through, if the inner and apparently chemical
changes are to operate, through which the germination of the spores is slowly pre-
pared and finally made possible.
In this way Brefeld was able to germinate the spores of the following
species none of which would germinate when flist mature (length of time
in moist earth given after each): Anthracoidea (Ustilago) Caryces Bref.,
over winter; Anthracoidea subindvsa Bref., 1 year; Doassansia Alismatis
Nees, 1 year; D. LimosellaeK.unze, 1 year; D. punctiformis Niesse, more
than a year; D. Sagdlariae Fckl., over winter; Melanotaernnim cingeus
Bref., 4 years; Neovossia Barclay aria Bref., 2 j^ars; Sphacelotheca Hydro-
piperis Schurn., 6 months; Tilletia controversa Kiihn, 2 years; Tilletia
decipiens Pers., 3 years; Tolypospornmi btdlatiim Schroet., 9 months;
Tol. Jtmci Schroet., 6 months; Tol. Penicillariae Bref., 1 year; Urocystis
Anemones Pers., 6 months; Ur. Filipendidae Tul., 1 year; Ustilago Adoxae
Bref., 1 year; U. anomola Kunze, over winter; U. Bistortanim D. C, 1
year; C/". Coias Bref., 2 years; ?7. dowes^ica Bref., 6 months; U. Holostei
D., 3 years; U. utriculosa Nees. Other "^Titers also have found that for
various species, a weathering under natural conditions was necessary in
order to secure germination.
Stibstitution of Nutritive Solntion for Weathering Period. — In the case
of some species Brefeld believes that the same changes which are ordinaril}"-
induced by storage in damp soil for a long period may be induced at once
by the use of his "nahrlosung." For example, he finds that the corn
smut spores when first mature will not germinate in water, but if kept until
the following spring they germinate in water. If, however, the freshly
matured spores are put in nutritive solution, they germinate overnight
almost without exception. He concludes, therefore, that the changes
induced by one are the same as those induced by the other, or, in other
words, that each may be substituted for the other.
Freezing. — Wliether or not freezing has any influence on germination
seems never to have been determined. Brefeld makes no mention of freez-
ing, and one infers from his publications that his buried spores were never
frozen. Since the spores of practically all species of smuts have been suc-
cessfully germinated without freezing, it may be safely said that freezing
is not a necessary condition of the process.
Essential Conditions for Germination of Urocystis cepulae Spores.
Search through all available literature on the subject revealed only one
reference to previous attempts at germination of the spores. Thaxter (18)
was unable to genninate fresh spores either in water or in moist air. When,
ONION SMUT FUNGUS. 105
however, the smutted onions were stored until January, then mixed with
wet earth and frozen for a week or more, the spores germinated when
kept moist in a warm room. They also germinated in an onion decoction.
He also made pure cultures in onion decoction horn fresh spores and from
sporiferoils hypha?, but does not mention germination in this respect.
Such, in full, is the extent of our present knowledge of the necessary condi-
tions. The purpose of the wTiter's experiments was twofold: (1) to dupli-
cate Thaxter's work and (2) to extend the inquiry in order to determine
more exactly many points which Thaxter either did not touch or treated
insufficiently. The experiments are summarized below.
Fresh Spores in Water. — Spores from a fresh but mature lesion were
scattered in a drop of w^ater on a slide kept in a Petri dish with water in
the bottom of the dish to prevent evaporation of the drop on the shde.
This common and familiar method was used in all the experiments where
water or a water solution was tested. Both distilled water and tap water
were tried. The spores were examined daily for aver two weeks, but no
indication of germination was observed. The experiment was repeated
many times, and the temperature and light relations were A'^aried in differ-
ent sets, but always without result. Spores taken from lesions which had
been kept dry for a. year in the laborator}'- gave no better results.
Fresh Spores in Soil Water. — A soil extract was made by filling a
beaker wdth good onion soil (taken from a field where smut was abundant) ,
adding water until the soil was saturated and the water was 1 cm. deep on
top of it, stirring thoroughly several times and filtering off after several
days. Results were the same as with tap and distilled water.
Influence of the Germinating Onion Seed. — These tests were in every
way like those described above with water, except that a few germinating
onion seeds were placed in each drop in addition to the spores. With one
exception, in these tests the spores failed to germinate. On one slide a
very few spores germinated in close proximity to the young cotjdedon.
Fresh Spores in Soil Decoction.- — A mixture of soil and water w^as
cooked for one hour on two successive days in the autoclave at 14 pounds'
pressure, filtered, tubed and sterilized. It was hoped that in this way
more of the soil substances would be brought into solution, and that they
might bring about germination. But, just as in the case of the soil extract,
so with this more concentrated soil decoction, there was no germination.
Fresh Spores in Dung Decoction. — This decoction was prepared just as
Brefeld prepared his "nahrlosung" which he used so successfully on the
spores of a large number of species. Fresh spores failed to germinate in
it. In these experiments the solution was concentrated. It is possible
that if it had been more diluted the results might have been different.
Fresh Spores in Onion Decoction. — This decoction was prepared by
boiling a shced onion in a pint of water for one hour. It was then filtered,
tubed and sterilized one-half hour at 15 pounds' pressure. This appears
to furnish an excellent medium for the growth of bacteria and fungi, and
in working with it every possible precaution must be used to prevent con-
tamination. These organisms grow so fast that they soon obUterate the
106 TECHNICAL BULLETIN 4.
more slowly germinating smut spores. It was found necessary not only
to sterilize, bj^ boiling, the slides, Petri dishes and all instruments used,
but also to wash the seedlings from which the spores were taken, first, in
mercuric chloride, 1 to 1,000, and then in sterile water, before the lesions
were opened. In drops of this decoction some of the spores began to
germinate within three days at laboratory temperature. The percentage
of germination, however, was always very low. In dozens of slide tests
made in this way, not over 25 per cent germination has ever been observed ;
and in most cases it is lower, averaging 5 to 10 per cent. It is apparent
from these tests that there is some substance in the onion which is capable
of inducing germination of fresh spores. In the light of other tests de-
scribed below, however, one would not be justified in concluding that this
substance is peculiar to the onion alone.
Fresh Spores in Sugar Solutions. — Sterile solutions of |, 1, 2, 3, .5, 7
and 10 per cent cane sugar were used just as the onion decoction mentioned
above. There was some germination in all of them, but very little in the
^ per cent and the 10 per cent. The highest percentage of germination
was in the 2 per cent solution, where 50 per cent of the spores which were
on the surface of the drop germinated. When spores are mixed with a water
solution of any kind, some of them remain on the surface while others
sink to the bottom. Only a very small percentage of those which were
immersed germinated. Since the spores on the surface are better located
for obtaining air, it is apparent that air is an important factor in germina-
tion. It is also apparent that sugar is at least one of the substances which
may induce germination. Since onions contain a high percentage of cane
sugar, it seems probable that this is also the effective element in the onion
decoction which induces germination.
Fresh Spores on Onion Decoction Agar. — Onion decoction agar was pre-
pared by adding 2 per cent of agar to the onion decoction. Sterile plates
were poured and permitted to become hard. Spores were mixed with
onion decoction or water and floated over the surface of the hard agar.
After permitting the spores to settle to the bottom the liquid was poured
away and the spores were left distributed over the agar. This insured a
sufficient quantity of air, and at the same time access to nutrient substances
in the agar. The percentage of germination varied ^vith different experi-
ments, but always it was as high as 10 per cent; sometimes 50 per cent.
This was found to be the most rehable of all the methods and was largelj'
used. Here also it was noticed for the first time that the spores did not
all germinate on the same day, but that there was a progressive germina-
tion, new ones starting each day for as long as three weeks, after which
the plates had dried too much, or possibly the supply of food had become
exhausted.
Fresh Spores on Czapek's Agar, Sugar Potato Agar, etc. — The Czapek's
agar contains 3 per cent of cane sugar. Several other agars containing
sugar were tried and always with a small percentage of germination, but
none higher than on onion decoction agar.
ONION SMUT FUNGUS. 107
Fresh Spores in Soil Decoction Agar. — After the role playerl by air was
determined, it seemed that the writer's previous failure to induce germina-
tion in soil decoction might have been due to exclusion of air. Therefore a
medium was prepared by adding 2 per cent of agar to the soil decoction.
Tests were made as with the onion agar, using soil decoction, however,
for floating the spores over the surface. After five days, germination of
1 to 2 per cent was observed. With each day, however, more of them
germinated, and this continued for several weeks until the plates became
too dry or were exhausted. We may conclude from these experiments
that (1) the soil contains all the essential sthnulating elements for ger-
mination, and (2) not all the spores germinate at once, but there is a pro-
gressive preparation.
Fresh Spores on Dung Decoction Agar. — This medium was prepared by
adding 2 per cent of agar to the dung decoction mentioned above. Since
the soil used in making the soil decoction had been heavily manured dur-
ing the previous season, it was thought that some element in the manure
might furnish the stimulus and a higher percentage of germination would
be secured. The percentage of germination in this medium, however, was
scarcely as high as for the soil decoction. Here is proof, however, that
stable manure contains some substance which is capable of inducing
germination.
Effect of Freezing the Spores. — ■ It has been previously mentioned that
Thaxter froze smutted mature onions in the soil and then found the spores
capable of germination. This experiment was duplicated as nearly as
possible by the writer, but he was entirely unable to get the spores free
from bacteria and other fungi, and abandoned the method rather than
work ^\'ith contaminated cultures.
An attempt was next made to freeze the spores under sterile conditions.
Smutted seedlings were sterihzed with mercuric chloride 1 to 1,000, washed
in sterile water, and sealed in sterile test tubes with a drop of water in the
bottom of each tube. After being exposed for nine days during December,
during which there were some light freezes, they were tested in onion
decoction. There was a germination of about 2 per cent. In similar tests
during January, in which they were frozen soUd for ten days or more,
buried under the snow in zero weather, the spores were apparently killed.
No germination at all was observed, although tried on or in the various
media described above. In \dew of the fact that the mycelium in culture
is not killed by freezing, these results are difficult to explain. In a later
series of tests smutted seedUngs, sterihzed on the surface, were buried in
sterile soil in test tubes and then frozen out of doors for eight weeks. On
onion decoction agar plates, varying percentages of germination were then
secured, but it was never as high as for spores which had been kept in damp
soil during the same length of time, but not frozen. The conclusion seems
warranted that freezing does not kill spores in the soil, but it does not
render them more capable of germination, and is not necessary.
Effect of a Period of Rest in Damp Earth. — Seedlings with unopened
lesions were sterilized and buried in sterile soil in test tubes. The tubes
108 TECHNICAL BULLETIN 4.
were then sealed and kept in the laboratory. After two weeks, germina-
tion was found to be somewhat higher than in the case of spores from fresh
lesions. Tests at the end of four weeks gave 50 per cent germination. At
the end of three months the average percentage was not higher, though in
indixddual slides it mounted to about 65 per cent. A higher percentage
of germination has not been seen in any test. In removing these seedlings
from the damp earth it was constantly noticed that the soil remained cling-
ing to the lesions and could be washed off with difficulty, while it was very
easily removed from other parts of the plant. Microscopic examination
showed that the soil particles were attached by numerous fungous hyphae.
When these hyphse were transferred to sterile agar tubes they gave pure
cultures of Urocystis. It was not possible to determine whether these
hyphse arose from germination of spores in the sori, from vegetative hyphse
in the seedling or from both. If the spores germinate while still inside the
lesion, this may explain why not all the spores taken from the weathered
sori germinate; they may have already germinated. This experiment
demonstrates clearly one way, at least, in which the smut myceUum gets
back into the soil from the diseased planto.
Natural Conditions of Germination. — It is a common impression among
laymen that the spore remains dormant in the ground, lying in wait until
an onion starts to grow near it, upon which it germinates and infects the
onion. Such, however, is apparently not the case. The seedling does not
seem to furnish any stimulus which causes the spore to start. Whatever
substances are necessarj^ for starting the process are in the soil itself. As
soon as the spores are released into the soil — if not before — a few of
them germinate; the others become capable of germination gradualh^
and it seems likely that all of them finally come to germination, but that
the period of preparation differs in length for different spores, so that the
germination extends over many months and possibly years. This period
of preparation may be shortened artificially by the use of certain stmiulat-
ing substances, such as cane sugar.
The Process of Germination.
Germination begins in three to six days after the spores are placed in the
water solutions or on agar plates as previously described. The time varies
somewhat with the medium used, and also apparently ^^^th other factors
which have not been explained. Three days are usually sufficient in
onion decoction or onion agar, while on soil decoction agar six days were
found necessary.
The first indication of germination is the appearance of a hyaUne hemi-
spherical vesicle (Fig. 1, B) on one side of the spore. This is apparently
an extrusion from the central fertile cell, but whether it comes out by a
rupture of the spore wall or by a regular pore could not be determined.
The covering of sterile ceUs renders exact observation of this point difficult.
Tliis vesicle when first observed is of about the same size as one of the
sterile cells, and can at first be distinguished from the latter only by the
ONION SMUT FUNGUS.
109
fact that it is hyaline while the sterile cells are brown. During the suc-
ceeding stages, however, it increases rapidly in size until it may be almost
as large as the spore itself (Fig. 1, K). This hemispherical or subglobose
Fig. 1. — Gormiiuition of spores.
vesicle corresponds to the promycelium or hemibasidium of other smuts
and rusts, and will be so designated. A stout tube grows out from the
surface of the promycelium (Fig. 1, C-G) and is quickly followed by
others in succession until a w^horl of diverging branches is produced (Fig.
1, H-N). The number of branches in the whorl is not constant, neither
110 TECHNICAL BULLETIN 4.
do they arise simultaneously but usually in succession. None has been
observed, however, which showed more than eight branches on a promy-
celium (Fig. 1, S). These primary branches are 2 to 3// in diameter, usu-
ally somewhat undulating, with broadly rounded tips. They soon become
septate, and almost invariably a lateral secondary branch grows out from
the top of the cell just below each septum (Fig. 1, 0-S"). The angle of
divergence between the primary and secondary branch is very broad,
often approaching a right angle. This manner of branching is character-
istic of onion smut mycelium wherever it is found, and is a good diagnostic
character. By continued branching, a dense mass of mycelium is developed
about the spore, and it becomes increasingly diflacult to follow the course
of single hyphse. Fig. 1, T, represents the latest stage in which the sep-
arate branches could be followed. On agar plates the older cells lose their
denoO protoplasmic content, and only the more distant tip cells appear
to be alive. In onion decoction, as the hyphse become older they become
more constricted at the septa and the cells rounded in the middle until
they appear almost separated from each other, the hypha having somewhat
the appearance of a string of beads t,Fig. 1, U). Frequently in the older
mats of mycelium from the germinated spore it has been observed that
some of the hyphal tips are recurved in the form of croziers. They have,
however, never been seen to develop further, and it is impossible to say
whether this development has any relation -to development of spores,
which are never produced except inside the tissue of the host. In hun-
dreds of germination tests which have been made during three years in a
large number of media, no conidia have ever been observed on the pro-
mycelium or its branches or anywhere else throughout the development
of the organism. Sometimes the short lateral branches appear like conidia,
but continued observation soon convinces one that they are merely vege-
tative branches which will elongate apically like other branches unless the
supply of nutriment is exhausted.
Comparison with the Germination Process in Other Species of Urocystis.
Let us now compare this process with the process of germination which
other investigators have described for other species of Urocystis.
Urocystis occulta Wallr., causing the flag smut of rye, was apparently
the first species of this genus which was studied with respect to germina-
tion of spores, that process having been first observed and described by
Kuhn in 1858. It was later studied by Wolff (19), Brefeld (4), McAlpine
(11) and others. According to Brefeld a promyceUal tube of varying
length is first produced, and at its apex it branches verticillately into a
whorl of four to six branches. These branches increase in length by apical
growth, and they, as well as the promycelial tube, become progressively
septate, while the protoplasmic content of the older cells constantly dis-
appears and the only living cells are those at, and just back of, the growing
tips. The verticillate branches never produce conidia, but form mycelium
by continued growth. McAlpine considers the verticillate branches them-
ONION SMUT FUNGUS. Ill
selves as conidia, but does not state that he ever found them detached.
This resembles the process in U. cepulac in (1) the ])roduction of the whorl
of branches, (2) the complete absence of sporidia and (3) the progressive
empt5ang of the cells. The main point of difference is in the elongated,
ultimately septate, promj'cclium in U. occulta which replaces the globose
vesicle of ['. cepulac.
In U. Tritici Koern. the process is almost identical with that of U.
occulta according to McAlpine (11), but the promj^celium is at times uni-
cellular, a condition which suggests that of U. cepulae.
The germination process in U. Anemones (Pers.) Wint. has been studied
by Fischer von "Wnldheim (7), Plowright (14) and Brefeld (4). As de-
scribed and figured by Brefeld it is almost identical with the process which
the writer observed in U. cepulae except that the promycelium is not so
large. The whorl of 2 to 4 branches arises very close to the surface of the
spore on a very much reduced promycelium, and they remain permanently
sterile.
In U. Filip&uhdae Tul. (Brefeld (4) ) the whole process is identical with
that of U. Anemones.
Germination of the spores of U. Violae Sow. has been studied by Prillieux,
Dangeard, Brefeld (4) and others, being a favorite subject for study because
of the ease with which germination can be brought about in water. Each
fertile cell of the spore ball produces an elongated promycelium which
becomes septate just as in U. occulta. A whorl of three to eight diverging
branches is produced at the apex. Each verticillate branch grows out at
the distal end into a slender sterigma on which is borne a long cylindrical
conidium. In nutrient solution these primary conidia may produce sec-
ondarj^ or tertiary conidia. The process in this species differs from that
of U. cep^dae (1) in the length of the promycelium, and more especially (2)
in the development of conidia.
In general, then, we may conclude that U. cepidae differs in its germina-
tion from the other species of Urocystis (except U. Violae) only in theshape
of the promycelium which is here reduced to a nonseptate hemispherical
vesicle. All other details of development appear to be identical.
Comparison with the Process as described by Thaxter.
As described and figured by Thaxter the spores germinate by a single
long irregularly branched tube on the tips and lateral branches of which
are borne small ellipsoidal to long ovoidal conidia. He does not mention
a globose promj^celium or whorl of branches such as the writer has always
observed. The marked differences in the process as observed by the writer
and as described by Thaxter are difficult to explain, unless they are due to
contamination in the cultures used by the latter. He states that he was
unable to obtain the material pure, and that "all the cultures swarmed
with bacteria." The presence of these same bacteria might produce a
difference in the development of the germination process. The writer in
attempting to aecure germination by Thaxter's method also failed to keep
the spores free from bacteria and therefore changed to a different method.
112 TECHNICAL BULLETIN 4.
III. Sapeophytism.
The early botanists and mycologists believed that smut fungi were obli-
gate parasites, i.e., they developed only when in parasitic relation with host
plants from the living cells of which they must take their nourishment.
We now know, however, that at least most smut fungi have in their life
cycle a saprophytic period during which they may develop extensively
and propagate for a long time, deriving nourishment only from dead
organic material in the soil or other substrata. Also most of them may
be propagated indefinitely in artificial culture media of various composi-
tions. Our knowledge of this stage began ^vith the extensive investigations
of Brefeld (3), and has been increased later bj'' numerous smaller contribu-
tions from a large number of workers. Urocystis ceptdae is no exception to
the rule, and is very readily isolated and grown in a large number of cul-
ture media and on soil. It is probably able to exist and grow in the soil for
years in entire absence of onions.
Isolation.
Two methods of isolation have been used by the writer. By the first
method a germinating spore on an agar plate is located under the micro-
scope by a ring of India ink, care being taken that this spore is far enough
removed from all others to prevent confusion. WTien themycelium from the
germinating spore has increased to such an extent that it is visible to the
naked eye as a tiny white speck it is transferred to an agar slant where it
gradually spreads to the agar of the tube and can be grown for a long
period. This method was used especially in the original isolations when
it was necessary to know for certain that the resulting fungus originated
from a single spore of Urocystis cepulae. In later work a more rapid method
was used. A part of a cotyledon or young leaf containing a lesion which
had not yet broken open was washed for a few minutes in mercuric chloride
1 to 1,000 and then in sterile water. The lesion was then cut into as many
pieces as desirable and the pieces transferred to agar slants. One hundred
per cent of pure cultures could be obtained in this way. Lesions of any
age could be used, but the youngest were found to be most satisfactory.
Cultural Characters.
The range of media on which the fungus will develop is almost unlimited.
Those which the WTiter has used are listed below along with a brief state-
ment of the pecuharities exhibited by the organism on that particular
medium.
Potato Agar.
The ordinary potato agar containing a boiled decoction from a large potato and
17 grams of agar to a liter of water. No sugar was added and the acidity was not
determined. Growth very slow, reaching a diameter of 1 cm. in about ten days;
very dense and compact like fine felt, snow white, dry, fiat, but with considerable
ONION SMUT FUNGUS. 113
aerial mycelium; margin very definite and oven. After about ten days the mycelium
shows more and more of a tendency to grow beneath the surface of the agar, and
the edge has the appearance of gradually fading away into the surrounding agar.
Growth may progress for several weeks, but is gradually checked by the drying
out of the agar. Some of the cultures show indistinct zonation. With age the
surface of thp felt may become rugose.
Oat Agar.
Growth more luxuriant than on potato agar, showing denser zones of white
mycelium. No change of color in mycelium or in the medium. Growth not suf-
ficiently different from that on potato agar to have any diagnostic value.
Nutrient Beef Broth Agar.
The standard agar of bacteriological work. Growth scanty, much less than on
potato agar, slimy, and taking on the color of the medium; never dry, very little
aerial mycelium. A very poor medium for growing the organism.
Czapek's Agar. ^
This was found to be a very favorable medium, the giowth being more rapid
and with a greater abundance of white, cottony aerial mycelium than on potato
agar. After about two weeks the agar below the growth, especially in the upper
part of the tube, turns maize yellow, ^ due to the suffusion of a pigment. After
about four weeks the color becomes more intense — aniline yellow or citrine yel-
low. With age this darkens to orange citrine or to various shades of olive. Also
the mycelium as seen from above loses its white color after three or four weeks,
showing various shades of greenish yellow — citrine drab, olive lake, etc. These
color changes on Czapek's agar offer one good diagnostic character.
Onion Decoction.
Prepared by boiling a sliced onion in a liter of distilled water and sterilizing the
filtered product for one hour at 15 pounds' steam pressure. Growth very slow,
resulting in development of little compact balls of mycelium; brown when in the
bottom of the tubes or white when on the surface of the liquid. Growth continues
for months very slowly, but the little balls of mycelium do not attain a diameter
of over 1 cm.
Onion Agar.
Prepared exactly like potato agar, but the onion decoction as described above
is used instead of potato juice. This was found to be not only the best medium for
culturing Urocystis, but also very much better than potato agar for growing many
other fungi which the writer had occasion to try on it. It is very easily prepared,
has a minimum of sediment even when not filtered, and altogether forms a very
superior general purpose agar. Its only objectionable qualities are the obnoxious
odor in the laboratory during preparation, and the fact that the growth of certain
fungi is too luxuriant for some purposes. The growth starts with a dense white
felt much like that on potato agar, but more rank. After about a week wrinkles
begin to appear near the center, and these spread and become sharper and the
irregular ridges more elevated with age, also at the same time the crests of the
ridges become hygrophanous and gray. This appearance spreads until it involves
' For method of preparation see Soil Science, 2:113.
' All colors according to Ridgway's Color Standards.
114 TECHNICAL BULLETIN 4.
the entire center or wrinkled part of the growth. The convoluted gray growth on
onion agar is perhaps the best diagnostic cultural character of the species. It has
been very constant in the many series of cultures which the writer has made with
this agar. After a few weeks the color in reverse becomes darker, reaching cinna-
mon brown in about five weeks.
Sugar Potato Agar.
Prepared as potato agar with the addition of 3 per cent of saccharose. Growth
is coarser in texture, more luxuriant and spreads more rapidly than on potato
agar. The aerial mycelium is not snow white, but early assumes a cream color
changing to cartridge buff after a few weeks.
Effect of Concentration of Sugar on Growth of the Mycelium in Culture.. —
In the series of cultures on different media it was observed that the best
growth occurred on media containing considerable sugar, viz., Czapek's,
sugar potato and onion agar. This led the writer to suspect that sugar is
the essential element of nutrition both in culture media and on the host
itself, since the onion contains a high percentage of saccharose. In order
to determine the effect of sugar on the development of the organism,
Czapek's synthetic agar was prepared first without any sugar and next
with .5, 1, 2, 3, 5, 7 and 10 per cent of cane sugar. Five tubes of each
were inoculated at the same time and accurate notes taken each day. No
growth whatever occurred where no sugar was included. At the end of
three weeks there was very little difference in the diameter of the growths
on all the othei concentrations, but those on the higher concentrates were
a little more dense. The most apparent difference was in the color which
was imparted to the agar. In the . 5 per cent the culture was pure white
in reverse, while in the 10 per cent it was bright yellow. The other con-
centrates formed a perfectlj^ graded series between the two. The only
other difference noticed was a wiinkling of the surface of the growth in
some of the higher concentrates, and its entire absence from the cultures
of low sugar content. Certain conclusions seem warranted from this
experiment: (1) agar and inorganic salts alone do not furnish food for
growth; (2) the yellow color in the agar is due to some reaction with the
sugar; (3) the amount of groAvth (at least for three weeks) does not depend
on the amount of sugar present. Any one of the concentrates apparently
contained more than the maximum amount which the organism could
utilize.
Substitution of Starch for Sugar. — In order to see whether the fungus
can utilize starch as a source of carbon, agar tubes were prepared identical
with Czapek's except for the substitution of soluble starch for saccharose.
A scanty growth occurred, but even after four weeks it had not attained a
diameter of 1 cm, and was very thin. Apparently, then, Urocystis can
utilize starch, but it is a very poor source of carbon.
Soil Decoction Agar.
Prepared by adding 2 per cent of agar to the soil decoction described above.
Growth was much less vigorous than on potato agar, and thin, but, on the other
ONION SMUT FUNGUS. 115
hand, spread filmost as rapidly over the surface for the first few weeks. There can
be no question whatever but that soluble elements in the soil furnish sufficient
food for the development of the mycelium.
Dung Decoction Agar.
Prepared by adding 2 per cent agar to the dung decoction previously mentioned.
Growth much thicker than on the soil decoction agar, but not as heavy as on
Czapek's, sugar potato, etc. Dense white aerial mycelium. The conclusion seems
warranted that horse manure furnishes all the elements necessary for the growth
of the fungus, and is more favorable medium than a good soil. Apparently a heavily
manured soil would be more favorable for the propagation of smut than one which
was not manured.
Tolerance of Acid. — Four series of cultures were made on onion agar, —
the first series without lactic acid; second, with 1 drop of lactic acid per
tube; third, with 2 drops per tube; fourth, with 3 drops. All were inocu-
lated at the same time. Growth was rank and normal in the series in which
no lactic acid was added; no growth whatever in the series in which 3
drops were added ; a very sUght growth where 2 drops were added; growth
much retarded in the 1-drop series. This series was begun with the pur-
pose of finding a method of excluding bacteria from cultures of the smut
fungus, but the latter was apparently checked by acid just as much as the
bacteria.
Effect of Freezing the Cultures.
Cultures on potato agar and on onion agar were kept out of doors for
two months during the most severe winter weather of 1919-20. Transfers
were then made to fresh agar tubes, and the mycelium grew luxuriantly
and rapidly on the surface of the slants. In fact, the growth at first seemed
to be even better than when transfers were made from cultures which had
not been frozen. Accurate measurements on a second series showed a
slight difference in favor of the transfers from frozen mycelium during
the first few days, but it was not permanent. We may conclude, then, that
freezing not only does not injure the mycelium, but possibly stimulates
it to even better growth.
Microscopic Characters of the Mycelium in Culture.
The characters of the mycelium differ somewhat with the age of the
culture. Microscopic examination of a culture a week old shows slender
hyaline hyphse of rather uniform diameter, about 2/<, with rather indistinct
septa and homogeneous contents. Branches arise almost exclusively from
the upper ends of the cells and diverge at a wide angle. The characters
have not changed from the condition previously described under germina-
tion of the spores. Not all of the cells of the mycehum appear to be alive;
some of them are empty and apparently dead; others are full of homoge-
neous protoplasm with no vacuoles. Under the oil immersion lens one
notices certain very refractive granules scattered throughout the dense
protoplasm (Fig. 2, A). The cells are easily broken apart, and when a
116
TECHNICAL BULLETIN 4.
mount is made the hyphae appear in segments as represented in the figure.
At this early stage they show no constrictions at the septa. No conidia
can be found. Clamp connections have not been observed.
If, however, cultures several weeks old are examined microscopically it
will be observed that certain changes have taken place. The aerial myce-
lium may remain about the same as described, except that the cells appear
vacuolated, but there will now be found a different kind of mycelium be-
neath the agar surface. These hyphce are stouter, averaging S.2fi in diam-
FiG. 2. — Details of hyphae in culture. Detached hyphal cells at C and further development
of same at D.
eter, the ceUs are much shorter, the septa very distinct, and the hj^phae
decidedly constricted at the septa, so much so that the hyphse appear almost
like strings of separate cells. A large proportion of the cells become shaped
like dumb beUs. When disturbed, as in mounting, the cells of the thread
break apart very readily so that when one makes a mount of an old luxuri-
ant culture, such as on onion agar, he hardly finds mycelium at all, but
only these irregular separate units. Most of them are branched at the
tip. A strand of this mycelium is represented in Fig. 2, B, with a young
ordinary hj^^ha for comparison. The appearance of the separate ceUs
from an onion agar culture as seen floating about in the microscopic prep-
aration is represented in Fig. 2, C.
Fate and Function of the Detached Hyphal Cells.
Since these large detached cells appear so early in the development of a
culture and in such large numbers, it does not seem probable that they
represent merely a stage in the degeneration or breaking down of the
mycelium. Apparently they have some role in the life history of the
organism. In order to determine whether they are capable of further de-
velopment, a culture was thoroughly shaken in water and the detached cells
floated out on sterile agar plates as desciibed previously for germination of
ONION SMUT FUNGUS. 117
the spores. Within twenty-four hours slender tubes of about half the diam-
eter of the original cells could be observed growing out from them. These
tubes originate from one or from both ends of the cell, quickly become
septate and branched, and within three days each is the center of a white
mycelium which can be seen with the naked eye. The centrifugal empty-
ing of the cells, the branching, and all other characters are the same as
those of the growths from the chlamydospores. Practically 100 per cent
germinated. No conidia could be found on them at anj'' stage. The de-
velopment of these cells is represented by Fig. 2, D.
Taken in connection with the fact that no true conidia have appeared
in any of the cultures, the conclusion seems warranted that these cells
detached by division of the vegetative h3rphae are analogous to and serve
the same purpose as the sporidia (conidia) of other smut fungi in propa-
gation and dissemination. In fact, almost any cell of the mj'^celium which
retains its protoplasm is a potential spore, and may serve all the functions
of the same. Since the cells are so easily detached and germinate so quickly
and universally, their importance in the distribution of the disease can
hardly be overestimated.
Lije in the Soil.
There are at least two ways in which the organism may pass from tlie
host into the soil; (1) when the spores are mature and the sorus is exposed
by rupture of the enclosing host tissue, the spores fall out or are blown or
shaken out by various agencies and fall to the ground; (2) as previously
described, mycelium from any buried lesion may grow from the disinte-
grating tissues directly into the surrounding soil. It has also been indicated
in cultures on soil extract naedia that the soil contains all the elements
necessary to induce germination of the spores and to nourish the mycelium
into further growth. In order to study further this period of development
of the organism, pure cultures on soil were made by inoculating Ehrlen-
meyer flasks of steriUzed soil, some by placing a small portion of diseased
cotyledon on the center of the surface of the soil, others by placing bits
of mycelium from agar tubes in the same position. Within a few days
the mycelium could be seen plainly with the naked eye passing from both
into the soil and spreading over its surface. After four weeks it was iso-
lated from all points of the soil surface. After more than a year it could
still be isolated in pure culture. Microscopic examination of mycelium
from the soil showed the same characters that are previously described
for cultures and the same detached cells.
Summary of the Saprophytic Stage in the Natural Life History.
From all that has preceded concerning this stage we may draw some
conclusions.
1. The fungus lives naturally in the soil, especially where there is an
abundance of organic material.
2. It derives sufficient nutrient materials from the soil to grow and
spread extensively during this stage.
118 TECHNICAL BULLETIN 4.
3. It enters the soil either as spores or as mycelium from the buried
parts of diseased onions.
4. No typical conidia (sporidia) are produced but it can be widely dis-
seminated by the detached mycelial cells which may be carried about by
water, wind, rain, tools, animals, workmen, etc.
5. It probably lives in the soil in this state for years without the pres-
ence of onions.
6. As will be shown later, infection may take place directly from this
mycelium, and the presence of spores is not necessary.
7. The number of years which must elapse before onions can be grown
safely on an infested piece of land is not necessarily decided by the lon-
gevity of the chlamydospores, but in all probability by the length of time
during which the mycelium can continue to live and develop saprophyti-
cally without having to pass again through a parasitic stage.
IV. Infection.
Very little has been published concerning infection except the bare fact
that it occurs at an earh' period in the growth of the plant. Concerning
the method and point of entrance, character of inoculum, etc., nothing
has been previously ascertained.
Development of the Onion Seedling.
In order to understand the description of infection given below it is
necessary that the reader should know something of the stages through
which an onion seedling passes during the process of germination. The
resting seed consists of a hard, black outer seed coat, a nutritive endosperm,
and an embryo. The embryo is coiled like a snail within the endosperm
(Fig. 3, A). The larger part of the coil represents the cotyledon; only a
short portion of the free end is the radicle. In the lower part of the coty-
ledon, just above where it joins the radicle, there is, even at this early
stage, a small cavity. A minute bud, lb, arises from the base of and pro-
jects into the cavity. This bud is the primordium of the first leaf, and the
cavity in this and later stages is called the cotyledonary cavity, cc. Sev-
eral layers of elongated cells throughout the length of the center ef the
embryo indicate the position which the fibrovascular bundle of the seedling
will occupy. Germination begins ^\^t]l rapid elongation of the embrj'o, the
radicle and lower part of the cotyledon being thus pushed through the
micropyle, a small opening in the seed coat. This elongation is effected
both by longitudinal stretching of the cells of the embrj^o and by cell
division. Food and water for this acti\ity are absorbed by the upper end
of the cotyledon which remains attached in the endosperm. On the third
day after planting, the projecting radicle is about 3 to 4 mm. long. The
root usually points upward as it emerges, but geotropism soon causes it
to turn downward and the cotyledon describes a sharp curve as indicated
in Fig. 3, B and C. It will be noticed that the tip of the leaf bud now
ONION SMUT FUNGUS.
119
A. Section through a resting
seed.
B. Longisection of a seedling
four days after planting.
C. Successive stages in the de-
velopment from the third
day to the twenty-fifth
day.
D. Diagrammatic longisection
through the growing zone
at the end of two weeks.
Symbols for parts are the
same in all:
sc, seed coat.
en, endosperm.
em, embryo.
lb, leaf bud or primordium of
first leaf.
re, root cap.
r, radicle or first root.
en, cotyledon.
rj, root joint.
kn, knee.
al, ascending leg.
dl, descending leg.
CO, exterior opening of coty-
ledonary cavity.
rS and r3, first and second
secondary roots.
cc, cotyledonary cavity.
gp, growing zone, region of
origin of all leaves and
roots.
Ib2, primordium of second
leaf.
Fig. 3. — Development of an onion seedling.
120 TECHNICAL BULLETIN 4.
points upward. At this early date the point of division between radicle
and cotyledon is indicated by a slight swelling, the root joint, rj. As all
the parts continue to elongate rapidly the curve in the cotyledon becomes
a sharp knee, kn , the part between the knee and seed is the ascending leg,
al, while that between the knee and the root joint is the descending leg,
dl. The primary root grows down very rapidly and is soon several times
as long as the cotyledon. From about the fifth day it will be noted that
the descending leg elongates more rapidly than the ascending leg. The
first part to appear above ground (seventh to tenth day) is the tip of the
knee, and each part becomes green as soon as it has reached the light.
The seed may still remain in the ground for a week or more after the knee
has appeared, but since it is firmly attached, and since the descending leg
continues to elongate more rapidly than the ascending leg, the seed is
finally carried into the air (Fig. 3, C). The knee then has a tendency to
straighten out, but its position is indicated as long as the cotyledon lives
by a sharp kink. On about the ninth or tenth day the first secondary root,
r2, may be seen pushing out from the swollen root joint, and this is fol-
lowed later by others in rapid succession, rS. Meanwhile the first leaf
bud has been elongating rapidly. The cotjdedonary cavity elongates also
in proportion (Fig. 3, D). It should not be understood that this cavity
is absolutely included, without any opening to the outside; on the con-
trary, its upper narrowed apex communicates with the outside air through
a small longitudinal slit in the side of the cotyledon (CO in Fig. 3, C and
D). As the leaf bud pushes its way upward the sides of the cavity are
distended, and finally from about the seventeenth to twenty-fifth day the
tip passes through the slit and appears on the outside as the first leaf
{lb in Fig. 3, C). But before this time the primordium of the second leaf,
lbs, has appeared in a depression at the base of the first, and successive
leaves follow rapidly, each starting from the base of the next preceding
at a very early stage. The successive secondary roots also start from the
same region. This very active meristematic region, the growing point, gp,
is very restricted, and remains stationary in the onion until after the bulb
is formed. The limited size and stationary position of the growing point
from which all new organs, roots or leaves, originate are characters of
prime importance in the spread of the smut fungus within the host plant.
Period of Susceptibility.
It is a well-known fact that onions are susceptible onlj^ in the seedling-
stage, and are immune after a certain stage of maturity is reached. But
we have no exact knowledge of the duration of this period of susceptibility,
the exact stage or time at which infection first occurs, or the stage or
time at which it ceaocs. The establishment of two points is thus nec-
essary: (1) the first day on which infection takes place, and (2) the last
day during which the plant can be infected. The latter of these two points
was established by the following experiment. Seed was planted in a flat
of sterilized soil. Beginning with the third day, when the radicle on the
ONION SMUT FUNGUS.
121
most advanced was less than J cm. long, and had not even started in many
of them, oO plants were transferred each day to soil which was badly
infested and which could be depended on to produce almost 100 per cent
of infection. Notes were made on the stage of development of the seed-
lings each day, and a careful record was kept of all the plants which be-
came smutted. After six weeks, when the plants were mostly in the
fourth leaf (after which infection never starts), all of them were pulled,
and the following table compiled to show the complete results of the
experiment : —
Days between Planting and Transplanting.
Percentage of
Infection.
3
4
5
6
7
8
IC
11
12
13
14
17
18
19
Check (left in original sterile sand)
100
100
95
100
100
98
87
87
70
59
15
6
The follo\\'ing conclusion may be drawn from this experiment: Under
greenhouse conditions the greater part of the infection occurs within two
weeks after planting, and the plants are no longer susceptible after the
seventeenth day. Since it seems probable that the period of suscepti-
bility is not limited by the number of days during which the seeds have
been in the soil, but by the length of time required for the seedling to pass
through certain stages of development, we may express this first conclu-
sion by stating that susceptibiUty begins to diminish from the time that
the knees emerge from the ground, and that little if any infection occurs
after the first leaf has emerged from the side of the cotyledon. In a large
number of experiments in the greenhouse at all times of the year it has
been found that the knees begin to appear above ground in seven to twelve
days. In one experiment, where the house was very cool, it required over
two weeks, and in this case the percentage of infection was 100, and the
individual plants were more thoroughly smutted than in any other experi-
ment tried. Since, then, the period of susceptibility might be increased
122 TECHNICAL BULLETIN 4.
by the length of time required for the seedlings to reach a certain stage,
it is well to inquire how the rate of growth in the greenhouse compares
with that in the field. During the spring of 1920, when the spring was
late and cold, onions in the field did not come up for over two weeks in
most cases, but growers have frequently told the writer that they have
had fields which came up within eight days. Apparently weather and soil
conditions may materially affect the length of this period. Depth of
planting might also influence slightly the length of the period and also the
chances of infection. The experiment reported above, however, gives us no
information as to the date when infection begins, but only indicates that
it ends -with about the seventeenth day.
In order to determine the stage at which the earliest infection starts, —
and at the same time to work out other points in the early life history, — ■
another bed of onions was started in the greenhouse with soil knowai to
give 100 per cent of smut infection. Beginning with the third day, a cer-
tain number of plants was dug up each day, fixed in Flemming's weaker
solution, run up into paraffin, sectioned, mounted serially and stained
with triple stain. No mycelium was found in the tissues of those which
were fixed on the third and fourth days. The first infection was found
in a plant which was dug up on the fifth day after planting, and was
apparently a very young infection because it had at no point penetrated
more than to the fifth layer of cells below the epidermis, and at its furthest
point was not more than 150/; from the point of infection. Fifteen other
plants dug at the same time were carefully searched under high power
through every section of 92 slides, but no other trace of mycelium was
found. It is probable, therefore, that only rarely, if ever, has the my-
celium entered the tissues of the plant on the fifth day after planting
(second day after germination has started). Since cultural experiments
with the smut fungus have shown it to be of very slow growth, at least in
the saprophytic condition, it seems hardly possible that it could have
succeeded in entering the tissues before the second day after germination
of the seed starts.
It may be concluded from everything which has been learned up to the
present in regard to the period of susceptibility that infection may take
place at any time between about the second day after the seed starts to germinate
until the seedling is in the first leaf (a period of about twelve days in the
greenhouse) .
Point of Infection.
In the studj^ of the plants fixed and stained as mentioned above, many
very young infections were found where it was possible to determine the
point of entrance for the mycelium. Infections were found at the knee
above, at the root joint below, and at various points between, also at least
one through the interior wall of the cotjdedonary cavity. The conclusion
is, therefore, that all points of the epidermis at least between the root
joint and the knee are susceptible to penetration by the smut tubes. In-
fection was never found taking place in the roots proper or between the
ONION SMUT FUNGUS. 123
seed aiul knee. From observation of mature sori, however, it seems jirob-
able that infection sometimes occurs above tlie knee. Mycehum in various
quantities has been found in the cotyledonary cavity of many plants, even
in the j'oungest stages, and by tracing it to the opening of this cavity it
can be se^n that it comes in from the outside through the natural opening,
but in most cases it has been impossible to trace a direct connection be-
tween this mycelium and any hyphte inside the tissues between the cells.
This mycelium has the size and all the other distinctive characters of
smut mycelium, but it is not possible to prove that it is such. It was
thought at first that this was the usual infection court, but after it was
demonstrated be\"ond anj' question that in a large number of cases young
infections could have no connection whatever with this cavity, the con-
clusion was reached that onlj^ a small part of the infection could be ac-
counted for in this way. It is still doubtful whether the mycelium which
was found in the canity was always that of [Jrocystis, or whether it may
have been that of another soil fungus.
It is probable that all infection takes place through the cotyledon. A case
was never noted where the leaf became smutted while the cotyledon
remained healthy. ]\Iore careful experiments on this point, however,
might show that the leaf does sometimes become infected fir.'t. It is
probable that all infection takes place beneath the surface of the ground.
Character of the Inoculum.
In all literature on onion smut it has been assumed that the spores of
the organism must be present in close proximity to the seedling in order
that infection may occur. The possibility that the mj'-celium might be
present and growing saprophytically and indefinitely in the soil, and might
infect without the immediate presence of spores, has been left out of con-
sideration. In order to determine the abihty of saprophytic mycehum
to produce infection, onion seeds were germinated beneath the surface of
agar cultures in test tubes in such a way that the developing seedling as
it elongated must pass through the mat of mycelium. Over 50 per cent
of the seedlings became infected, although no smut spores could have
been present. In the stained sections which were studied, in a few cases
mycelium was found outside the walls of the epidermal cells where infec-
tion has occurred. Only in one case were spores found in these sections,
and at that time there was no infection beneath them. It is probable,
however, that spores would usually be removed by the washing process,
and this could hardly be adduced as conclusive evidence against the neces-
sit}'' of spores for infection. It is probable that either spores or saprophytic
mycelium in the soil can serve as the inoculum.
Method of Entrance.
The infecting hj'pha enters the epidermal cell by boring directly through
tlie outer wall. Since in the younger infections the stomates are not yei
open, and mechanical wounds have not been found, there is no other route
124
TECHNICAL BULLETIN 4.
by which it could make its way into the interior tissues of the plant. A
stage of infection has not yet been found so young that the tube has just
entered the epidermal cell and has not progressed further.
Passage through the Epidermal Cells.
In the youngest infections observed, the mycelium had already grown
through the epidermal cells, and its tips could be found in the intercellular
spaces at a depth of two or more layers below. In some cases a piece of
the infecting hypha still remained on the exterior of the cuticle, but was
always devoid of contents and consisted only of somewhat crumpled walls
Fig. 4. — Infection through the epidermal cells; A, B, C from outside the cotyledon, D from
the cotyledonary cavity.
(Fig. 4), A broad clear canal passes inward from the outer wall usually
directed toward the cell nucleus. The wall of the canal appears to be
continuous with the cell wall as if merely an inward extension of the same.
Commonly it is much thicker at the point of entrance, and resembles a
slender funnel or trumpet in shape. It was not found possible to deter-
mine whether part of the wall of the canal is an inward growing sheath of
the same substance as the cell wall, or whether it is merely a thickened
wall of the hypha. In all the cases observed, the canal was empty at the
point of entrance. The host nucleus appears to exert an attractive influ-
ence. When the tube has reached the depth of the nucleus, it branches
to form a tangle of stout, swollen, gnarled, hyphse which maj^ be confined
to the region immediately about the nucleus, or may reach to all parts of
the lumen of the host cell (Fig. 4). They may be entirely devoid of con-
ONION SMUT FUNGUS. 125
tents or — depending on the stage at which one finds them — may con-
tain protophisni and bright red nuclei scattered singly or in pairs. The
hj'phal tangle may be confined to the lower (inner) part of the cell, and
is always more dense there (Fig. 4, C). Its windings are difficult to follow.
These intracellular windings stain red with the triple stain. There is a
marked contrast between the large, swollen winding intracellular hypha;
and the trim, slender, straight intercellular hypha; between the cells be-
low, which stain violet and are of only about one-half the diameter of the
former. LTsually the tangle is confined to one epidermal cell, but some-
times the adjacent cells may be invaded (Fig. 4, B). The attacked epi-
dermal cells do not collapse, and, in fact, appear practically normal.
Hypha? pass down from the tangle through the inner wall of the epidermis
into the intercellular spaces immediately beneath.
Multiple Infection.
The same plant may suffer from a number of infections. In one plant
fixed eight days after planting, the mycelium was found passing in through
the epidermis at six points on a piece of the cotyledon less than a centi-
meter in length. In young stages it is not difficult to trace each mycelium
to its limits between the cells, and in this case no one of the six had come
into contact with another. It is not unusual to find seedlings which show
five or six sori on the same cotyledon. Microscopic examination indicates
that these are not the results of a single infection, but that for each sorus
there is at least one infection thread which penetrated the epidermis from
the outside. This statement, however, does not apply to the sori which
appear later on the true leaves.
V. Incubation Period.
The incubation period is the time which elapses between infection and
the first externally \'isible sjTnptom of disease. Since the first external
sjTnptoms appear at approximately the same time that the spores are
forming, we may say that the incubation period is that segment of the life
cycle between infection and sporogenesis. In the greenhouse the fu'st
symptom, a sHght curving and thickening of the cotyledon, has been ob-
served here on the tenth day. Since, as previously stated, infection may
take place as early as the fifth day, we may consider that this period occu-
pies a space of about five days under favorable conditions in the green-
house. It may be longer outside, but, at most, is a comparatively short
period. During this period the parasite grows rapidly, spreads inside the
host and prepares to form spores.
Young Hypha: in the Intercellular Spaces.
After passing through the epidermis the hypha? are intercellular during
the remainder of their development. Just below the inner epidermal wall
they spread in all directions. They are long, slender, and, as they pass
126 TECHNICAL BULLETIN 4.
along the longitudinal walls, appear very straight. They appear to pro-
gress somewhat more rapidly up and down the cotyledon than in a radial
direction inward. In the young stages the}^ do not occur in strands or
bunches between the cells, but one finds them running singly (Figs. 4, A
and 5, D). They do not appear to be going toward any definite point,
but are spreading more or less in all directions. They are undoubtedly
septate, but the septa in the very young hypha? are difficult to distinguish.
The protoplasm passes to the growing tips, and leaves empty the cells
behind it. These tip cells stain deep violet with the triple stain, while
those cells behind them take less and less stain until only the thin line of
the walls can be seen. The nuclei stain bright red and are very prominent,
especially back of the deep violet tip cells. These nuclei may occur singly
or in pairs distributed along the hyphae. At this stage it is not always
possible to tell whether the two nuclei of a pair are in the same or different
cells, but by a comparison with what is found in hyphse somewhat older,
it is probable that here also the cells may be either uninucleate or bi-
nucleate. The contents of the hyphal cells appear homogeneous, and at
this stage there are no vacuoles or oil drops. The hyphse seem to be
mostly tightly pressed against the walls of the cells, but at places can be
seen passing from the wall of one cell to that of another across the open
spaces. The cells are long and the branching not close as in the later
stages. The branches always arise monopodiall}'^ from just below the
septum, as previously described.
Haustoria.
These absorbing organs are not numerous, but are not uncommon. In
some infections none could be found, while in others they are fairly com-
mon. They are of various sizes and of ver.y irregular shape (Fig. 5, A-E).
They are not much different from the haustoria of other smuts as described
by various ^vriters. The}^ are always very much branched, but the
branches may be reduced to mere knobs or short stubs which are fre-
quently bifid at the apices (Fig. 5, A) . In the larger haustoria, however,
the branches are longer and more lax, and may go to all parts of the cell
(Fig. 5, B and C) . The branches of these larger haustoria are usually — ■
but not always ■ — imbedded in the protoplasm about the nucleus. la
some cases they seem to be tightly gripping the nucleus, and the latter
appears indented by the pressure. Their shape and size can be best
understood by reference to the figures. In many of them an appressorium-
like expansion of the hypha can be seen flattened against the outside of
the cell wall, and from the lower side of this expansion a narrow neck
passes through the wall (Fig. 5, A, E). It is not certain, however, that
this appressorium is always present. In the larger haustoria, red nuclei
can be distinguished in varying numbers, but in smaller ones, and, in fact,
in many of the larger ones, no nuclei can be seen. In some, the position
of the nucleus in the stalk of the haustorium is evident (Fig. 5, C) but
ONION SMUT FUNGUS.
127
apparently there is \w uniformity either in the position or nunii:)er of
nuclei. The haustoria usually stain yellowish brown with the orange G
of the triple stain.
Fig. 5. — Haustoria (A-E) and absorptive hyphal expansions (F, G).
Absorptive Hyphal Expansions.
Frequently during the incubation period one finds the tips, especially
of short lateral branches, flattened 'out like spatulas against the cells of
the host. In some sections, just before sporogenesis, these structures may
be found in great numbers. Csually they are terminal (Fig. 5, G), but
not infrequently they may be found intercalary within the ordinary
course of a hj^jha which, beyond the expansion, continues in its normal
size and shape (Fig. 5, F). They resemble the appressoria previously
mentioned as the bases from which the haustoria arise, but their number
is out of all proportion to the number of haustoria which one finds in the
same sections. No description of these organs has been given elsewhere, and
their function or meaning is not clear. One can only conjecture that their
purpose is to present a broad absorbing surface for securing more nourish-
ment from the host cells. It seems doubtful whether haustoria are really
necessary in this connection, because many infections have been studied
under the microscope in which no haustoria could be found.
128 TECHNICAL BULLETIN 4.
Progressive Infection of New Leaves.
It is a common belief, supported by statements in the literature of the
disease, that when a seedUng once becomes infected it never recovers.
Such, however, is not the case. The waiter has watched the development
of many seedlings which had infected cotyledons, but which developed
into healthy onions. On the other hand, he has not seen an onion, in which
the first leaf was affected, which produced a healthy bulb. Usually each
successive leaf will show smut sori, and they are not always in any appar-
ent relation to the sori on older leaves. As previously stated, all infec-
tions come through the cotyledon, but the fate of the plant depends on
the point in the cotyledon at which infection takes place. If it occurs
only high up toward the knee, or above it, there is a pretty good chance
that the host tissue will have become mature or dead and no longer suit-
able for spread of the mycelium before the latter has reached the gromng
zone, and the bulb will develop normall3^ But if infection occurs at or
very near the root joint, the mycelium quickly penetrates to the gro^^dng
zone from which all future leaves arise. This meristematic tissue furnishes
the ideal condition for continuous vegetation of the pathogene, and as
each new leaf pushes out from this restricted stationary zone it contains
filaments from which the new sori of the successive leaves develop. When
the parasite is once established in this growing point, the host seems never
to be able to shake off its grip, and is doomed. It is not quite so clear why
the mycelium does not enter the tissues of the developing roots in the
same wa}', but the writer has never been able to find it in these organs,
VI. Sporogenesis.
The approach of spore formation is first indicated by massing of the
mycelium between the cells. Up to this time only long straight slender
hyphse are found spreading singlj^, or at most not more than two or three
together, between the cells. The period during which the pathogene ap-
pears to be spreading as widely and rapidly as possible between the cells
has just been described as the incubation stage. The distributive hyphse
now begin to branch profusely, and the branches are not straight and
parallel to the main hyphse, but become twisted and interwoven into
dense tangles which push the cells apart and increase the area of inter-
cellular spaces within which the spores are to be formed. The hyphse
now become highly vacuolated, and the protoplasm between the colorless
vacuoles stains densely blue with the triple stain, while the old cells from
which the protoplasm has passed take the orange stain. The beaded
appearance of the alternating vacuoles and densely staining cytoplasm is
the surest indication of approaching sporogenesis.
These spore nests or sori always occur between the cells of the mesophyll
anywhere between the epidermis and the bundles, but have not been found
inside the bundles. They are extended in the direction of the length of
the leaf or cotyledon.
ONION SMUT FUNGUS.
129
Observation of the exact course of events in the fonnatioii of a spore is
rendered difficult by the denseness of the mass of developing spores, and
by the fact that in the j'oung stages all the developing parts stain so
deeply on account of their very active protoplasm that the nuclei and
septa can hardly be made out. In all cases which have been observed,
the spore begins as a lateral or terminal branch which curves back on itself
in the form of a crozier (Fig. 6, A-I). These hook-like croziers may be
seen in enormous numbers in the mycelial tangle at the initiation of sporo-
FiG. 6. — Stagesof sporogenesis. A-P, development of the crozier and origin of theenvelop-
ing hyph.-r; Q, section through young spore which is shown in surface view at R; S, sec-
tion through mature spore.
genesis. Even after the spores at the center of a sorus are fully formed,
one may still find various stages of development extending as far back as
the crozier, as he passes from the center toward the periphery of the tangle.
The croziers remind one of those from which the asci of the Ascomycetes
are developed. They stain very deeply, and apparently the protoplasm
from the other cells of the hyphse passes into them. The various shapes
which they may assume are best understood by consulting Fig. 6. By
growth from the apex of the crozier a complete circle is soon formed and
then a spiral if further terminal elongation occurs (Fig. 6, F, L, N, P).
At about this time the crozier or spiral begins to appear angular and
130 TECHNICAL BULLETIN 4.
irregular (Fig. 6, M), due to protuberances whicli mark the origin of short
lateral outgrowths which soon curve inward along the surface of the devel-
oping ball (Fig. 6, P). The whole structure becomes so complicated at
this time that it is not always possible to make certain of the exact course
of events. The surface view now shows a dense ball of interwoven hyphae
(Fig. 6, R) . A cross section (Fig. 6, Q) shows that at the center there is a
larger cell which represents what will later be the fertile cell of the spore.
This cell appears to be the enlarged terminal cell of the crozier, though it is
not certain that this is always its origin. Also it is not entirelj^ certain that
all the branches which form the outside of the tangled mass arise directly
from the surface of the crozier. In some cases one gets the impression
that other hyphee may be involved, or that branches arise from below the
crozier on the same hypha. The transformation from the stage represented
in Fig. 6, Q, R, to the mature spore is very rapid. The central cell enlarges
while the cells of the surrounding hyphse become pressed tightly against
and united with it. The union between the central cell and the cells of
the enclosing hyphse appears to be stronger than that between the cells
of a single hypha of the latter; at any rate, the hyphse now break up and
their elements no longer appear as cells of individual hyphse, but as scat-
tered conical cells whose flattened bases are firmly attached to the surface
of the central cell (Fig. 6, I). This involves a decided change in shape as
well as orientation. Nothing has been seen in this process which could
be called a gelatinization of cells, such as has been described so often as
occurring during sporogenesis in the Ustilaginales.
Approximately at the center of the fertile cell of each fully developed
spore there is a nucleus which stains very prominently at this stage of
development (Fig. 6, S) . In thousands of beautifully stained spores exam-
ined by the writer, more than a single nucleus has never been found. It
is 3 to 4fi in diameter, with a prominent very red single nucleolus of about
.Qfi diameter, usually in contact with the nuclear membrane. The mem-
brane is very plain, but the nuclear content, ■with the exception of the
nucleolus, appears only as a few fine granules of cromatin aggregated about
the nucleolus or around the inside of the membrane. In each accessory
cell there is a single small nucleus of about the diameter of the nucleolus
of the fertile cell. In Urocystis Violae, Dangeard reported that there were
no nuclei in the accessory cells. With the staining methods used it was
impossible to determine whether the nucleus of the mature spore results
from the fusion of two nuclei. In U. Anemones (Pers.) Wint., Lutman
found that the cells of the vegetative hyphse are binucleate and remain
so until after the formation of the spore ball, and that the large nucleus
of the mature fertile cell results from fusion of the two nuclei. Such might
well be the case here, because in the vegetative hyphse, as previously
mentioned, about half of the cells are binucleate, while in the mature
spores all cells are vminucleate.
With the full development of the sorus, the host tissue above it dries
out and may split open and permit the escape of the dry powdery mass of
ONION SMUT FUNGUS. 131
spores. In tlie largor loaA'Os the oi)ening of the sorus may first occur on
the interior of the hollow leaves, thider moist conditions other funpi,
such as P\isarium, may cause the tissue to decay more rapidly, and thus
aid in the liberation of the spores.
The first outward indication of disease in a j'oung seedling is a slight
curvature of the cotyledon accompanied by some enlargement of the
affected part. In the greenhouse I have found these s\^llptoms as early
as the tenth day after planting. Within another day or two, when an
affected seedling is held so that the light will shine through it, the lesions
may be located by the darker appearance. As soon as the spores are
mature the dark sorus can be seen through the tissue without holding it
up to the light. The length of time which elapses before it splits open
and permits the escape of spores varies greatly with the weather, age of
leaf, and other factors.
VII. Summary.
1. Spores as soon as mature germinate in the laboratory in onion
decoction, sugar solutions, onion decoction agar, soil agar, manure decoc-
tion agar and various agars containing sugar.
2. The}' do not germinate in tap water, distilled water or soil water.
3. The presence of the onion or any substance from the onion is not
necessar3^
4. Freezing does not increase or hasten germination, but when spores
are frozen in the ground they are not killed.
5. Free access to air increases the percentage of germination.
6. A period of rest in damp soil increases the percentage of germination,
but is not necessary.
7. In the soil the spores do not all germinate at once, but become pro-
gressively prepared for germination. They do not wait until a host plant
starts to grow near them.
8. Germination begins in three to six days after the spores are brought
under favorable conditions.
9. A short hemispherical promycelium is first developed, and from this
a whorl of branches grows out.
10. The branches grow as mycelium indefinitely without producing con-
idia (sporidia) . The older cells become devoid of their protoplasm progres-
sively.
11. The germination process is very similar to the same process in
other species of Crocj^stis, being almost identical with that of Urocystis
Anemones. Of the investigated species of this genus, only U. FioZae pro-
duces sporidia.
12. Urocystis cepidae lives and grows as a saprophyte indefinitely in the
soil, its gro-vvi;h being favored by manure.
13. It may be grown in pure culture on a wide range of culture media,
and shows cultural peculiarities by which it may be distinguished from
other fungi.
132 TECHNICAL BULLETIN 4.
14. Sugar in the media greatly increases the growth. The same sub-
stance probably accounts for its rapid growi^h in the host.
15. Starch furnishes a very poor source of carbon.
16. Decoctions from soil or manure furnish all the essentials for growth.
17. A small amount of acid checks its growth.
18. Freezing does not kill the mycelium.
19. No sporidia (conidia) have been found by the WTiter in pure cul-
tures or in soil.
20. The mycelium at an early stage breaks up into short ])lump cells
which have all the functions of sporidia and are probably of great im-
portance in dissemination.
21. The organism gets into the soil either by means of spores when the
sorus is broken up, or as mycehum which grows from the lesions when in
contact with moist soil.
22. Infection occurs during the time from the second day after the
seed germinates until about the time that the first leaf appears on the side
of the cotyledon, after which the plant is immune.
23. Infection occurs only through the cotyledon, and any part of its
epidermis may serve as the point of infection.
24. The infecting hypha bores directly through the outer wall of the
epidermal cell, forms a hyphal gnarl inside the cell, and then passes
through the inner wall into the intercellular spaces where it grows during
the rest of its development.
25. Many infections may occur on the same cotyledon.
26. The incubation period is less than a week.
27. Large complicated haustoria are formed within the host cells.
28. An infected plant recovers if the fungus fails to reach the growing
zone; but if it once becomes established in this zone, the plant never
recovers, and most if not all the leaves will contain lesions.
29. At the close of the incubation period the mycelium is in dense masses
between the cells, and from this the spores develop in sori.
30. The spore begins as a recurved lateral or terminal branch, forming a
crozier, circle or short spiral.
31. Branches arising from the circle (crozier) form a close covering
about the terminal (fertile) cell.
32. By adhesion of the cells of the covering hyphse and rapid expansion
of the fertile cell the enclosing hyphce are separated into the scattered
elements which appear as the sterile ceUs of the mature spore.
33. The fertile cell contains a single, large nucleus, and each sterile cell
a single small nucleus. Probably the large nucleus is a result of fusion.
VIII. Literature Cited.
(1) Baiy, A. de. Untersuchungen ilber die Brandpilze. Berlin. 1853.
(2) Bary, A. de. Comp. Morph. and Biol., etc. 1887.
(3) Brefeld, O. Untersuch. a. d. Gesam. Myk. 5.
(4) Ibid. Vol. 12.
ONION SMUT FUNGUS. 133
(5) Duggar, B. M. Physiological Studies with Reference to the Germination of
Certain Fungous Spores. Bot. Gaz., 31:38. 1901.
(6) Hallier, E. Phytopathologie. Leipzig. 1868.
(7) Fischer von Waldheim. Contribution to the Biology and History of the De-
velopment of the Ustilagineae. Trans. N. Y. Agr. Soc, 1870:280.
(S) Kcllerman, W. A., and Swingle, W. T. Loose Smut of Cereals. Ann. Rpt.
Kans. Sta., 2:213. 1890.
(9) Ki'ihn, J. Uber die Entwicklungsformen des Getreidebrandes, Naturf. Ges.
Halle. 1874.
(10) Kiilm, J. Krankbeiten der Kulturgewachse. Berlin. 1858.
(11) McAlpine, D. Smuts of Australia. Melbourne. 1910.
(12) Norton, J. B. S. A Study of the Kansas Ustilagineae, especially with Regard
to their Germination. Trans. St. Louis Acad. Sci., 7:229. 1896.
(13) Osner, G. A. Leaf Smut of Timothy. N. Y. (Cornell) Bui. 281. 1916.
(14) Plowright, C. B. A Monograph of the British Uredineae and Ustilagineae.
London. 1889.
(15) Prevost, I. B. Memoire sur la Cause immediate de la Carie oux Charbon
des bles. Paris. 1807.
(16) Prilleaux, E. E. Sur la Formation et la Germination des Spores des Uro-
cystis. Bui. Soc. Bot. Fr., 27:204. 1880.
(17) Schmaler, F. Uber den Einfluss verschiedener Temperaturen auf die Keim-
fahigkeit der Stein Brandsporen. Forsch. Geb. Agr. Physic, 3:288. 1880.
(18) Thaxter, R. The Smut of Onions {Urocystis cepulne, Frost). Ann. Rpt. Ct.
Exp. Sta., 1889:129.
(19) Wolff, R. Beitrage zurKenntniss der Ustilagineon. Bot. Zeitung, 31:657. 1873.
(20) Woronin, M. Beitrag zur Kenntniss der Ustilagineen. Abh. Senck. Nat.
Ges., 12:559. 1882.
Massacl. '■icnltural Collcuc,
AMHERST, MASS.
massachusetts
Agricultural Experiment Station
TECHNICAL BULLETIN No. 5 AUGUST, 1922
Concerning the Diagnosis of Bac-
terium PuLLORUM Infection in
THE Domestic Fowl
By GEORGE EDWARD GAGE
For several years the Experiment Station has been con-
ducting studies of Bacterium pullorum, the object being to
determine factors which aid in accuracy of diagnosis. The
data obtained are recorded in this bulletin, and indicate that
there are two forms of Bacterium pullorum, both of which are
distinct from Bacterium sanguinarium and can be distinguished
from it by certain biochemical tests; that Bacterium san-
guinarium is not widely distributed in Massachusetts; that
neither Bacterium pullorum nor Bacterium sanguinarium is the
cause of the so-called "paralysis" common in Massachusetts
at certain seasons of the year; and that the agglutination
test, when carefully controlled through epidemiological work,
is the best method now available for locating Bacterium pul-
lorum infection and furnishing to poultr\men a starting point
for its elimination.
Requests for Bulletins should be addressed to the
AGRICULTURAL EXPERIMENT STATION
AMHERST, MASS.
Publication of this Document
approved by the
Supervisor of Administration.
CONTENTS.
PAGE
Historical ............ 61
Experimental ........... 64
Change of reaction in carbohydrate media bj' the 112 strains of Bac-
terium pullorum .......... 69
Conclusions from the fermentation tests ...... 70
Fermentation tests with Bacterium sanguinarium . . . .71
Distribution of fowl typhoid in Massachusetts . . . . .79
Does either Bacterium jmUorum, or Bacterium sanguinarium play any
part in the so-called "paralysis" so widely distributed in Massa-
chusetts? 79
Influence of infection upon the hatching quality of eggs and upon the
viability of young stock ........ 80
The present status of the specificity of the agglutination test as a means
of control of Bacterium pullorum infection in young chicks . . S3
Summary ............ 86
Literature cited ........... 88
TECHE"ICAL BULLETIN No. 5.
DEPARTMENT OF VETERINARY SCIENCE.
CONCERNING THE DIAGNOSIS OF BACTERIUM
PULLORUM INFECTION IN THE DOMESTIC
FOWL.
BY GEORGE EDWARD GAGE.
During the years 1916, 1917, 1919 and 1920 special studies have been
conducted in this department concerning the diagnosis of Bad. pidlorum
infection in chicks and adult birds. The object in view has been to de-
termine factors which aid in accuracy of diagnosis. Therefore the plan
here is to set forth the data obtained which may be of some value in sub-
stantiating the work of others, and to add any data from experimental
studies and routine which may assist those who have to do with the
puUorum problem.
Among the points to be considered by the laboratory and field worker
in the Bad. pulloruni problem, the following are of interest: —
1. Are there a Bacterium pullorum A and a Bacterium pullorum B?
2. Can infections with Bacterium pullorum and Bacterium, sanguinarium be
differentiated?
3. Is Bacterium sanguinarium (fowl typhoid) widely distributed in Massachu-
setts?
4. Is it necessary to submit suspicious Bacterium pullorum cultures to biochemical
tests before a diagnosis is justified?
5. Is either Bacterium pullorum or Bacterium sanguinarium related to the so-
called "paralysis" so widely distributed at certain periods of the year in Massa-
chusetts?
6. Is Bacterium sanguinarium of any significance as the cause of epidemic disease
in very young chicks?
7. What is the present status of the specificity of the agglutination test as a
means of control of Bacterium pullorum infection in young chicks?
Historical.
The presence of cholera-like or typhoid-like epidemics in domestic birds
dates back manj'- years, but careful study extends only from the last
quarter of a century. For a most excellent historical r6sum6 of these
studies from 1789 to 1913, the reader is referred to Hadley (1).
Since 1913 several investigators have added much to our knowledge
concerning the biology of Bad. pullorum. Smith and Ten Broeck (2),
62 TECHNICAL BULLETIN 5.
carr5dng out five sets of experiments in which serum of rabbits immunized
with heated cultures of human typhoid, fowl typhoid and Bad. pullorum,
considered that the agglutination tests were sufficiently definite to enable
them to group the fowl typhoid and pullorum types together, both demon-
strating the same intimate relation to tj^Dhoid bacilli. Again, in another
paper (3), these writers demonstrated that fowl typhoid has many diag-
nostic features in common with the human t^^phoid bacillus, namely, the
behavior toward carbohydrates and the agglutination reactions.
Rettger and Koser (4) carried out agglutination tests using reacting
sera from rabbits immunized by subcutaneous injections, first of killed
suspensions and later of living suspensions of Bad. pullorum and Bad.
sanguinorium. Five days after the injections of heated vaccine, the rabbits
were bled and the agglutinative power of the sera tested against definite
suspensions of both Bad. pullorum and Bad. sanguinarium. No difference
in agglutination properties was manifested. Attempts were made to in-
crease the agglutination titre by the injection of living organisms. The
titre remained the same and no change in the agglutinative ability of the
two sera was manifested. Although these organisms have several char-
acters in common, and particularly the serological reactions, they consti-
tute two separate and distinct types, each bearing a specific relationship
to the disease with which it has been associated, namely, either bacillary
white diarrhcea or fowl typhoid. Taylor (5) concludes from his studies
on fowl typhoid that the lesions produced in fowls which are infected with
Bad. sanguinarium resemble in many respects those produced by Bad.
pullorum, but, although there is a still closer resemblance in the biological
characters of the two organisms, there is enough difference to warrant the
conclusion that they are distinctly different diseases. Ward and Gallagher
(6), studying forty-seven birds for comparison of agglutination and intra-
dermal tests on naturally infected birds, report the absolute failure of each
test as judged by the other test and by an autopsy, findings being similar
in amount. Field tests on two hundred and thirty-one birds made simul-
taneously with the agglutination test at thirty-eight hours failed to detect
one case reported positive to the other test.
Pfeiler and Rehse (7) present the clearest description of an epidemic
in fowls due to the fowl typhoid bacillus. The feraientative reaction
showed the organism to be similar to the human typhoid bacillus. Accord-
ing to Goldberg (8) the principal differences between the strains of Bad.
pullorum and Bad. sanguinarium studied lie in the fact that Bad. pullorum
produces gas in various carbohydrates while Bad, sanguinarium lacks
this power in any of the carbohydrates he used, which included sugar-free
media containing dextrose, lactose, saccharose, mannite, dextrine, inuline,
galactose, levulose, raffinose, amygdalin, arabinose, adonite, dulcite, xylose,
salicin, isodulcite, mannose, starch, glycerine, erythrol. The difference in
gas production, as well as in their actions on milk, maltose, dulcite, dex-
trine, and isodulcite seems to indicate that these two organisms are distinct
species of bacteria.
BACTERIUM PULLORUM INFECTION IN FOWL. 63
Hadley (1) concludes from his studies on the colon-typhoid intermediates
that in carbohydrate media used kno^^^l types of Bad. pullorum, B.
gallinarum, B. avisepticus, B. parotypJwsns A and B, manifest definite
fermentative differences which justify regarding them as distinct species.
Since paratyphoid A does not ferment xylose, a close relationship is shown
between tlie types from poultry (pullorum and gallinarum) and para-
tj-phoid B. The data presented indicate that pullorum is much less active
than gallinarum on xylose. Aside from gas production there is a closer
fermentative relation between B. gallinarum and the paratyphoids than
between Bad. pullorum and the paratyphoids; this is due to the fact that
pullorum is maltose-dextrine-dulcite negative. Hadley also finds that all
the maltose-dextrine-dulcite negative strains isolated from chicks have
been aerogenic, while all of the maltose-dextrine-dulcite negative strains
isolated from adult birds were anaerogenic. The author has been able to
isolate from the eggs of fowls experiencing infections with the maltose-
dextrine-dulcite negative anaerogenic strains both aerogenic and anaero-
genic forms. The gas production may vary quantitatively within wide
limits. The writer has found that no one of the many original aerogenic
pullorum strains, cultivated for years in artificial media, has lost its
aerogenic power when placed under favorable conditions for growth ; and
none (either pullorum or gallinarum) that originally lacked this power
ever attained it. According to these data one may conclude that if a
strain, possessing otherwise the characteristics of pullorum or of galli-
narum, is aerogenic it is not B. gallinarum, while if it is anaerogenic it
may be either Bad. pullorum B or 5. gallinarum. This indicates that it
is necessary to make use of the maltose-dextrine-dulcite fermentation
tests only when the strain in question is anaerogenic. In another paper
(9) this same author concludes from his data that gas production by
Bad. pullorum may depend upon whether the cultures are grown in glucose
extract or glucose infusion broth. Propagating cultures for many years
on artificial media does not cause them to lose their gas-producing
ability. Bad. pullorum isolated from epidemics of bacillary white diar-
rhoea in young chicks or from infected eggs is aerogenic; there exist also
anaerogenic strains which, in all the cases in which they have been ob-
served, have been isolated from adult fowls experiencing acute or subacute
infections simulating fowl typhoid in both clinical symptoms and patho-
logical alterations of tissues. Therefore the writer proposes tentatively
to postulate for Bad. pullorum: (1) Bad. pullorum A, aerogenic; and
Bad. pullorum B, anaerogenic, pathogenic for adult stock only.
Hadley (10) suggests that Bad. pullorum appears to stand as a border-
line group in the colon-typhoid intermediates, separating the actual para-
typhoids from the actual paracolons; and further suggests that, in order
to facilitate bringing about some degree of order in the group of colon-
typhoid intermediates, gas-forming strains be referred to the paracolon
group which should be revived; and that anaerogenic forms only be re-
ferred to the paratyphoid group, in which B. gallinarum (Klein) might
stand as the type species.
64 TECHNICAL BULLETIN 5.
Rettger and Koser (4) present data which indicate that dextrine, maltose
and dulcite are attacked by Bad. sanguinarium with the production of
acid but no gas. Bad. pullorum produces, on the other hand, no visible
change of media containing these agents except slight alkali production.
Bad. pullonmi acts upon dextrose and mannite with evolution of appre-
ciable amounts of gas, while Bad. sanguinarium, whether recently isolated
or artificially cultivated for many years, does not produce gas in any of
the carbohydrate media. Prolonged cultivation of Bad. pullorum in
the laboratory does not cause this organism to lose its power of producing
gas in dextrose and mannite broth. These authors conclude that Bad.
pullorum manifests itself only as the cause of natural epidemic infection
in young chicks. They further maintain that Bad. sanguinarium. attacks
fowls of different ages, and is of relatively little, if indeed any, significance
as the cause of epidemic disease in very young chicks.
Mulsow (11) concludes from his studies that B. avisepticus may generally
be distinguished from Bad. sanguinarium by its action in milk, indol
production, fermentation of carbohydrates, agglutination reaction and
pathogenesis. Bad. pullorum and Bad. sanguinarium do not produce
indol, generally form hydrogen sulphid in lead acetate medium, and pro-
duce a temporary acidity in milk, but later alkalinity. As regards fer-
mentation, Bad. pullorum produces acid and generally gas in the same
carbohydrates, and in addition produces acid in dulcite and maltose.
According to this author. Bad. pullorum may be distinguished from Bad.
sanguinarium by the inability of the former to ferment dulcite, while the
latter ferments this carbohydrate. Bad. sanguinarium generally produces
acid promptly in maltose, and does not produce gas in any of the carbo-
hydrates. Rhamnose is fermented promptly by Bad. pullorum, while
Bad. sanguinarium produces acid only after forty-eight hours' incubation.
It appears that there are sufficient differences, reported in this paper by
Mulsow, between Bad. sanguinarium and Bad. pullorum to regard these
as separate types.
Krumwiede and Kohn (12) report results which indicate that the
essential characteristic of the paratyphoid-enteritidis group is the ability
of its members to produce acid from rhamnose, differentiating both the
aerogenic and anaerogenic members from B. typhosus. They point out
that, without due regard to low and latent avidity for carbohydrates in
relation to variability and practical differentiation, erroneous differential
significance might easily be given to variation even among members of the
fixed groups.
Experimental.
In the experiments presented, a study has been made of 112 different
strains of Bad. pullorum isolated from diseased materials from poultry
plants in various parts of Massachusetts, to determine, if possible, bio-
chemical and cultural details which are constant enough to warrant their
recommendation as a part of the procedure in diagnosis. The following
organisms, listed in Table 1, have been isolated from cases of chick disease.
BACTERIUM PULLORUM INFECTION IN FOWL.
65
clinically white diarrhoea, and these conformed mor])hologically, bio-
chemically and serologically to this group of organisms. It was further
decided to study the uniformity of these 112 cultures biochemically and
serological!}', and to determine how many of them gave reactions which
were similar to the reactions of its close relative, the fowl typhoid organism
{Bad. sanguinarium) . The cultures of Bad. sangidnarium were isolated
from birds sent here for diagnosis, and the Smith, Cornell and Gage
strains. There were five strains in this list. The two other than the
three mentioned appeared typical of sanguinarium, were isolated during
the early part of 1920, and designated the Humphrey and Massachusetts
strains, respectively.
The follo-\\-ing table lists the cultures of Bad. pullorum isolated and
studied during the course of this work: —
Table 1. — Strains of Baderium Pullorwn studied in this Investigation.
Bacterium Pullorum.
Source of Culture.
When Isolated
and Studied.
Strain No. 1
Strain No. 2
Strain No. 3
Strain No. 4
Strain No. 5
Strain No. 6
Strain No. 7 ,
Strain No. 8 .
Strain No. 9 ,
Strain No. 10
Strain No. 11 .
Strain No. 12
Strain No. 13
Strain No. 14
Strain No. 15
Strain No. 16
Strain No. 17
Strain No. 18
Strain No. 19
M. A. C. Amherst, Jlass. Isolated March, 1914,
from M. A. C. chick. Used in summer of 1914 as
Strain A.
Experimental material from this laboratory. From
unabsorbed yolk of chick inoculated summer of
1913 with Sa ^S.s from Cutler egg). Used in
summer of 1914 as Strain B.
Isolated from material sent to laboratory. Used as
Strain C in summer of 1914.
Bridgewater, Mass. Isolated from Cutler chick.
Used as So in 1913. Used as Strain D in 1914.
Maryland. Used at Maryland Experiment Station
in 1911.
Sterling, Mass. Isolated 1914. Trask Strain. Used
as Strain F in summer of 1914.
Holliston, Mass. Isolated from chicks sent by C. E.
Cristman, Silvervvood Farm, Holliston, Mass.
These chicks were bought of A. B. H. Arnold,
Holliston, Mass.
M. A. C. Amherst, Maes. No. 231 (2703) from un-
absorbed yolk (chick).
Holliston, Mass. Isolated from unabsorbed yolk of
chick. Isolated from liver of chick.
Northborough, Mass. Isolated from liver of chick .
Franklin, Mass. 11-1 isolated from unabsorbed
yolk of chick No. 2; 11-2 isolated from liver of
chick No. 5.
North Hadley, Mass. 12-1 from unabsorbed yolk
of chick No. 1 ; 12-2 from unabsorbed yolk of
chick No. 4; 12-3 from unabsorbed yolk of chick
No. 9.
Kingston, Mass. Isolated from unabsorbed yolk
of chick No. 2.
Center Marshfield, Mass. Isolated from unabsorbed
yolk of chick No. 4.
Brookline, Mass. Isolated from unabsorbed yolk
of chick No. 1.
Amherst, Mass. Isolated from liver of chick No. 1;
16-2 isolated from unabsorbed yolk of chick No. 1 ;
16-3 isolated from liver of chick No. 2.
Southborough, Mass. 17-1 isolated from liver of
chick No. 1; 17-2 isolated from heart of chick
No. 2; 17-3 isolated from heart of chick No. 3;
17-4 isolated from unabsorbed yolk of chick No.
4; 17-5 isolated from unabsorbed .volk of chick
No. 5; 17-6 isolated from unabsorbed yolk of
chick No. 6.
Littleton, Mass. lS-1 isolated from heart of chick
No. 1; lS-2 isolated from liver of chick No. 1.
Andover, Mass. Isolated from unabsorbed yolk of
chick No. 1.
March, 1914
May 1, 1914
Feb. 20, 1915
Mar. 31. 1915
Apr. 1, 1915
Apr. 5, 1915
Apr. 5, 1915
Apr. 6, 1915
Apr. 7, 1915
Apr. 12, 1915
Apr. 16, 1915
Apr. 17, 1915
Apr. 22, 1915
66
TECHNICAL BULLETIN 5.
Table 1 . — Strams of Bacterium Pullorum shidied in this Investigation ■
Continued.
Bacterium Pullorum
Source of Culture.
When Isolated
and Studied.
Strain No. 20 .
Strain No. 21 .
Strain No. 22 .
Strain No. 23 .
Strain No. 24 .
Strain No. 25 .
Strain No. 26 .
Strain No. 27 .
Strain No. 28 .
Strain No. 29 .
Strain No. 30 .
Strain No. 31
Strain No. 32
Strain No. 33
Strain No. 34
Strain No. 35
Strain No. 36
Strain No. 37
Strain No. 38
Strain No. 39
Strain No. 40
Strain No. 41
Strain No. 42
Strain No. 43
Strain No. 44
Strain No. 45
Strain No. 46
Westborough, Mass. Isolated from unabsorbed yolk
of chick No. 2.
Amherst, Mass. Chicks hatched from eggs bought
at Hickory Farm, Ludlow, Ma.ss. 21-1 isolated
from heart of chick; 21-2 isolated from liver of
chick.
Shrewsbury, Mass. Isolated from unabsorbed yolk
of chick No. 1.
Natick, Mass. Isolated from liver of chick No. 1 .
Lowell, Mass. 24-1 isolated from unabsorbed yolk
of chick No. 1 ; 24-2 isolated from unabsorbed
yolk of chick No. 2.
South Hadley, Ma.ss. 25-1 isolated from liver of
chick No. 1; 25-2 isolated from unabsorbed yolk
of chick No. 2.
Amherst, Mass. 26-1 isolated from liver of chick
No. 1 ; 26-2 Lsolated from liver of chick No. 2.
Dedham, Ma.ss. 27-1 isolated from liver of chick
No. 1; 27-2 isolated from liver of chick No. 2.
Belchertown, Mass. Isolated from liver and un-
absoibed yolk of chick.
Nobscot, Mass. 29-1 isolated from liver and un-
absorbed yolk of chick; 29-2 isolated from liver
and unabsorbed yolk of chick; 29-3 isolated from
liver and unabsorbed yolk of chick; 29-4 isolated
from liver and unabsorbed yolk of chick.
Concord, Mass. 30-1 isolated from liver and un-
absorbed yolk of chick; 30-2 isolated from liver
and unabsorbed yolk of chick; 30-3 isolated from
liver and unabsorbed yolk of chick; 3U-4 isolated
from liver and unabsorbed yolk of chick; 30-5
isolated from liver and unabsorbed yolk of chick;
30-6 isolated from liver and unabsorbed yolk of
chick.
Holliston, Mass. 31-1 isolated from unabsorbed
yolk of chick; 31-2 isolated from liver of chick;
31-3 isolated from unabsorbed yolk of chick.
Shiewsbury, Mass. Isolated from unabsorbed yolk
of chick.
Morrisville, N. Y. 33-1 isolated from unabsorbed
yolk of chick; 33-2 isolated from unabsorbed yolk
of chick.
Egypt, Ma.ss. Isolated from unabsorbed yolk of
chick.
Plainville, Mass. Isolated from unabsorbed yolk
of chick.
Fitchburg, Mass. 36-1 isolated from liver of chick;
36-2 isolated from liver of chick.
Lunenburg, Mass. Isolated from liver of chick;
37-2 isolated from liver of chick.
Sutton, Mass. 38-1 isolated from unabsorbed yolk
of chick; 38-2 isolated from liver of chick.
Southborough, Mass. Isolated from hver of chick
Cohasset, Mass. Isolated from unabsorbed yolk
of chick.
Amherst, Mass.
yolk of chick;
yolk of chick;
yolk of chick;
yolk of chick.
Shirley, Mass. 42-1 isolated from unabsorbed yolk
of chick; 48-2 isolated from unabsorbed yolk of
chick.
Middleton, Mass. 43-1 isolated from ovary of chick ;
43-2 isolated from ovary of chick.
Spencer, Mass. Isolated from liver of chick .
Greenfield, Mass. 45-1 isolated from liver of chick;
45-2 isolated from liver of chick.
Winchendon, Mass. 46-1 isolated from liver of
chick; 46-2 isolated from liver of chick.
41-1 isolated from unabsorbed
41-2 isolated from unabsorbed
41-3 isolated from unabsorbed
41-4 isolated from unabsorbed
Apr. 23. 1915
May 15, 1915
May 13, 1915
May 14, 1915
May 15, 1915
June 2, 1915
June 2, 1915
June 2, 1915
May 2, 1916
July 28, 1916
Mar. 24, 1916
May 2, 1917
Feb. 28, 1917
Mar. 28, 1917
Mar. 16, 1917
Apr. 15, 1917
Apr. 13. 1917
Apr. 13, 1917
Apr. 13, 1917
Apr. 16. 1917
Apr. 16, 1917
Apr. 15, 1917
Apr. 18, 1917
Apr. 21, 1917
May 2, 1917
May 3, 1917
May 8, 1917
BACTERIUM PULLORUM INFECTION IN FOWL.
67
Table 1 . — Strains of Bacterium. Pullorum studied in this Investigation —
ContiiuuHl.
Bacterium Pctllouum.
Source of Culture.
When Isloated
and Studied.
Strain No. 47 .
Strain No. 48 .
Strain No. 49 .
Strain No. 50
Strain No. 51
Strain No. 52
Strain No. S3
Strain No. 54 ,
Strain No. 55
Strain No. 56 ,
Strain No. 57 ,
Strain No. 58 ,
Strain No. 59 ,
Strain No. 60
Strain No. 61 ,
Strain No. 62 ,
Strain No. 63 .
Strain No. 64 ,
Strain No. 65 ,
Strain No. 66 ,
Strain No. 67 ,
Strain No. 68 .
Strain No. 69 .
Strain No. 70 ,
Strain No. 71 ,
Strain No. 72 ,
Strain No. 73 .
Strain No. 74 ,
Strain No. 75 ,
Strain No. 76 ,
Strain No. 77 ,
Strain No. 78
Strain No. 79
Strain No. 80
Strain No. 81
Pittsfield, Mass. Isolated from liver of chick
8-1 isolated from unabsorbed
18-2 isolated from unabsorbed
Pe.ibody, Mass.
yolk of chick;
yolk of chick.
Weymouth, Mass. 49-1 isolated from unabsorbed
yolk of chick; 49-2 isolated from unabsorbed
yolk of chick; 49-3 isolated from unabsorbed
yolk of chick; 49-4 isolated from unabsorbed
yolk of chick.
Westfaeld, Mass. Isolated from unabsorbed yolk of
chick.
Methuen, Mass. Isolated from liver of chick .
Methuen, Mass. Isolated from unabsorbed yolk of
chick.
Methuen, Mass. Isolated from unabsorbed yolk of
chick.
Methuen, Mass. Isolated from heart of chick .
Webster, Mass. Isolated from unabsorbed yolk of
chick.
Webster, Mass. Isolated from heart of chick .
Webster, Mass. Isolated from unabsorbed yolk of
chick.
Andover, Mass. Isolated from unabsorbed yolk of
chick.
Andover, Mass. Isolated from liver of chick .
Natick, Mass. Isolated from unabsorbed yolk of
chick.
Natick, Mass. Isolated from unabsorbed yolk of
chick.
Natick, Mass. Isolated from heart of chick .
Natick, Mass. Isolated from unabsorbed yolk of
chick.
Hubbardston, Mass. Isolated from liver of chick .
Hubbardston, Mass. Isolated from liver of chick .
Hubbardston, Mass. Isolated from unabsorbed
yolk of chick.
Hubbardston, Mass. Isolated from liver of chick
Lexington, Mass. Isolated from heart of chick
Lexington, Mass. Isolated from liver of chick
Lexington, Mass. Isolated from liver of chick
Lexington, Mass. Isolated from heart of chick
Longmeadow, Mass. Isolated from liver of chick
Plymouth, Mass. Isolated from liver of chick;
Essex, Mass. Isolated from heart of chick
Worcester, Mass. Isolated from unabsorbed yolk
of chick.
Worcester, Mass. Isolated from unabsorbed yolk
of chick.
Belchertown, Mass. Isolated from unabsorbed yolk
of chick.
Bridgewater, Mass. Isolated from liver of chick
Bridgewater, Mass. Isolated from unabsorbed yolk
of chick.
Wellesley, Mass. Isolated from unabsorbed yolk
of chick.
East Braintree, Mass. Isolated from liver of chick
May 7, 1917
May 24, 1917
Apr. 10, 1917
May 24
Mar. 7
Mar. 7
Mar. 7
Mar. 7
Mar. 15
Mar. 15
Mar. 15
Mar. 19
Mar. 19
Mar. 19
Mar. 19
Mar. 19
Mar. 19
Mar. 23
Mar. 23
Mar. 23
Mar. 23
Apr. 8
Apr. 8
Apr. 8
Apr. 8
Apr. 3
Apr. 3
Apr. 9
Apr. 9
Apr. 9
Apr. 9
Apr. 12
Apr. 12
Apr. 14
Apr. 14
, 1917
. 1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
,1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
, 1920
68
TECHNICAL BULLETIN 5.
Table 1. — Strains of Bacterium Pullorum studied in this Investigation ■
Concluded.
BACTERItnU PULLOHUM,
Source of Culture.
When Isolated
and Studied.
Strain No. 82 .
Strain No. 83 .
Strain No. 84 .
Strain No. 85 .
Strain No. 86 .
Strain No. 87 .
Strain No. 88 .
Strain No. 89 .
Strain No. 90 .
Strain No. 91 .
Strain No. 92 .
Strain No. 93 .
Strain No. 94 .
Strain No. 95 .
Strain No. 96 .
Strain No. 97 .
Strain No. 98 .
Strain No. 99 .
Strain No. 100 .
Strain No. 101 .
Strain No. 102 .
Strain No. 103 .
Strain No. 104 .
Strain No. 105 .
Strain No. 106 .
Strain No. 107 .
Strain No. 108 .
Strain No. 109 .
Strain No. 110 .
Strain No. Ill .
Strain No. 112 .
M. A. C. Amherst, Mass. Isolated from liver of
chick.
M. A. C. Amherst, Mass. Isolated from unabsorbed
yolk of chick.
M. A. C. Amherst, Mass. Isolated from unabsorbed
yolk of chick.
Chester, Mass. Isolated from unabsorbed yolk of
chick.
Chester, Mass. Isolated from liver of chick .
Chester, Mass. Isolated from liver of chick .
Boston, Mass. Isolated from liver of chick
Leominster, Mass. Isolated from liver of chick
Medway, Mass. Isolated from liver of chick .
Medway, Mass. Isolated from liver of chick .
Wakefield, Mass. Isolated from liver of chick .
Wakefield, Mass. Isolated from liver of chick
M. A. C. Amherst, Mass. Isolated from unabsorbed
yolk of chick.
M. A. C. Amherst, Mass. Isolated from liver of
chick.
Littleton, Mass. Isolated from heart of chick .
Bedford, Mass. Isolated from liver of chick .
Bedford, Mass. Isolated from liver of chick .
Worcester, Mass. Isolated from liver of chick .
Worcester, Mass. Isolated from liver of chick .
West Acton, Mass. Isolated from liver of chick
West Acton, Mass. Isolated from liver of chick
Woonsocket, R. I. Isolated from liver of chick
Woonsocket, R. I. Isolated from liver of chick
Woonsocket, R. I. Isolated from liver of chick
Belchertown, Mass. Isolated from unabsorbed
yolk of chick.
Segreganset, Mass. Isolated from liver of chick
Waltham, Mass. Isolated from liver of chick .
Charlemont, Mass. Isolated from unabsorbed yolk
of chick.
Hampton Falls, N. H. Isolated from liver of chick
Southwick, Mass. Isolated from liver of chick
Hudson, Mass. Isolated from unabsorbed yolk of
chick.
Apr. 20, 1920
Apr. 20, 1920
Apr. 20, 1920
Apr. 21, 1920
Apr. 21, 1920
Apr. 21, 1920
Apr. 21, 1920
Apr. 21, 1920
Apr. 27, 1920
Apr. 27, 1920
Apr. 27, 1920
Apr. 27, 1920
Apr. 27, 1920
Apr. 27, 1920
Apr. 30, 1920
Apr. 30, 1920
Apr. 30, 1920
May 4, 1920
May 4, 1920
May 7, 1920
May 7, 1920
May 11, 1920
May 11, 1920
May 11, 1920
May 14, 1920
May 18, 1920
May 21, 1920
May 28, 1920
May 29, 1920
May 19, 1920
June 3, 1920
BACTERIUM PULLORUM INFECTION IN FOWL. 69
Change of Reaction in Carbohydrate Media by the 112 Strains of Bacterium
Pnllorum.
Tlic cultures of Bact. pvUorum were gro\\Ti in test tubes of uniform
length and caliber and in standard beef extract bouillon containing 1 per
cent of the carbohj^drate. These results were somewhat lower than those
obtained by Goldberg (8), who found bj^ using infusion broth that the
percentage was higher. According to Hadley (10), on an average 0.7 per
cent more acid is produced in sugar-infusion broth than in sugar-extract
broth. Two drops of a bouillon suspension of each strain were used as the
inoculum for a test, tripHcate titrations made, and the average percentage
acidity noted at the end of the fifth daj'. It appeared from our work in
relation to time of acid production that the maximum occurred between
the fifth and tenth day. Therefore the tables and curves represent the
amount of acid at the end of a five-day period, at 37.5° C, expressed in
percentage normal acid. All titrations were made in the cold, using
N N
— NaOH and — HCl and phenolphthalein as the indicator. Gas produc-
20 20
tion was determined in dextrose, galactose, mannite, levulose, arabinose,
salicin, mannose, xylose, adonite, erythrol, saccharose, dulcite, dextrine,
lactose, raffinose, inuhn, maltose and glycerine. Durham double-barreled
fermentation tubes were employed, and the percentage of gas in the inner
tube read off on the Frost gasometer chart at the end of five days' incuba-
tion at 37.5° C.
Dextrose. — This sugar was fermented by all the 112 strains. The lowest amount
of acidity was 0.6 per cent and the highest 1.8 per cent, the mean of 108 determina-
tions being 1.4 per cent acid. Gas was produced in this carbohydrate by all strains,
ranging in quantity from a bubble to 55 per cent, the average for all the 112 strains
being 20 per cent.
Mannite. — The acid production in mannite was greater than in dextrose and
much more variable. After five days' growth the 112 strains had produced an
average of 1.0 per cent acidity. The exceptions to this average were strains 23,
46 and 72 which produced 2.0 per cent, 2.2 per cent, and 1.7 per cent, respectively.
Gas was produced by all strains, ranging in quantity from 20 to 50 per cent, with
an average for the 112 strains of 30 per cent.
Galactose. — This sugar was fermented by all strains, being very much like
mannite and dextrose. The acidity ranged from 0.1 to 2.1 per cent, the average
for all cultures being 0.9 per cent. There were four exceptions which make a wide
variation in the curve, — strains 29, 33, 42 and 49, which produced 0.1, 1.9, 2.0
and 2.1 per cent, respectively.
Levvlose. — This sugar was fermented easily by all strains of Bact. pullorum,
and the changes in reaction here correspond with those in dextrose, mannite and
galactose, the acidity ranging from 0.2 to 2.0 per cent, the average for the 112
strains being 0.9 per cent. The exceptions were strains 63, 72 and 73, which pro-
duced 2.0, 1.9 and 1.5 per cent acidity, respectively.
Arabinose. — All strains fermented this carbohydrate, the acidity ranging from
0.5 to 1.0 per cent, with an average for the 112 strains of 0.7 per cent. This carbo-
hydrate was fermented in a very variable manner.
Salicin. — None of the 112 strains fermented salicin. On the fifth day there was
marked alkaline reaction in some strains. The average acidity for the 112 strains
was 0.1 per cent.
70 TECHNICAL BULLETIN 5.
Mannose. — This sugar was fermented by all the strains. The minimum acidity
by any strain was 0.6 and the maximum 1.3 per cent. The average for the 112
strains was 0.9 per cent acid.
Xylose. — This sugar was fermented by all the strains, but none produced marked
quantities of acid. The minimum produced by any strain was 0.1 and the maximum
0.4 per cent, with a mean of 0.25 per cent for the 112 strains. Therefore it may be
said that these pullorum strains are not strongly xylose positive.
Adonite. — For the most part the initial acidity was not greatly changed. The
minimum figure observed was an alkalinity of 0.1 per cent and the maximum an
acidity of 0.1 per cent. As a group these strains were adonite-negative, the curve
of results from the 112 strains running close to the line of initial aciditJ^
Erythrol. — This carbohydrate was not fermented significantly by any of the
cultures of Bad. pullorum studied. All strains gave a reduction of the initial acidity.
The acidity ranged from a minimum of — 0.4 per cent to a figure which represented
no change from original acidity. Therefore these 112 strains are erythrol negative.
Saccharose. — There was no appreciable amount of acid produced in this carbo-
hydrate. The minimum reading was — 0.2 per cent and a few readings showed no
change from the initial acidity. The average acidity determination for the 112
strains was — 0.2 per cent. There were two exceptions, strains 67 and 84, which
showed a determination of — 0.4 and — 0.5 per cent for acidity. Therefore in
saccharose there is no acid formed by Bad. pullorum.
Dulcite. — All the 112 strains of Bad. pullorum showed a marked reduction of
acidity. A few strains did not change the initial acidity, the range being between
no change of acidity and — 0.4 per cent. There were three exceptions, however,
cultures 32, 46 and 47, which produced the following results: — 0.6, — 0.5 and
— 0.5 per cent, respectively. Therefore it may be said that the results from these
determinations indicate that Bad. pullorum is dulcite negative.
Dextrine. — The initial acidity was readily reduced by all strains studied. The
readings ranged from no change in acidity to — 0.3 per cent. There were no ex-
ceptions, all cultures demonstrating this reduction.
Lactose. — The initial acidity was reduced by all strains. The readings ranged
from no change in acidity to — 0.4 per cent, the mean reading being — 0.12 per
cent. Bad. pullorum may be considered, consequently, lactose-negative as regards
acid production. Two strains, 93 and 109, were unusually prompt in this particular.
Both strains gave a reading of — 0.4.
Raffinose. — The acidity was reduced by all the pullorum strains. The average
reading for the 112 cultures was — 0.2 per cent. Strain 48 was capable of greater
alkaline production than the others, giving a result of — 0.5 per cent.
Inulin. — All strains of Bact. pullorum were negative in this carbohydrate, the
mean reading being — 0.19 per cent. There was a prompt reduction in initial
acidity, only one culture of the 112 showing no change in the initial acidity.
Maltose. — None of the 112 strains produced any acid. The change was usually
marked in all tubes on the fifth day. There was an average reduction of acidity of
— 0.18 per cent.
Glycerine. — None of the 112 strains produced any acid in glycerine. The
determination on the fifth day showed a reduction in the final acidity, averaging
— 0.1 per cent.
Conclusions from the Fermentation Tests.
From the tests reported concerning the fermentation of the 112 strains
of Bact. pullorum, it appears that this organism is positive in dextrose,
galactose, mannose, mannite, levulose, xylose and arabinose; and negative
in glycerine, maltose, adonite, dulcite, lactose, dextrine, saccharose, inulin,
erythrol and raffinose. In salicin there is a slight indication of fermenta-
tion, at least a shght acidity in a large percentage of the strains. All strains
of this organism studied showed a marked tendency to produce gas in
70
Mam.
by any
strains ■^
Xylos
quantiti
0.4 per '
said tha
Adon
minimu
acidity
of resul
Eryth
cultures
The aci
no char
Saccf
hydratt
change
strains
showed
sacchar
Dulc
acidity
no cha:
culture
— 0.5p
determ
Dext
reading
ceptior
Lad
from n
cent,
acid pr
Both s
Raff
readinj
alkalin
Inui
mean
aciditj
Mat
marke
—0.18
Glyc
detern
—0.1
Frc
of Ba
galaci
ingly
eryth
tion, ;
of th.
BACTERIUM PULLORUM INFECTION IN FOWL. 71
dextrose. This aerogenic property of the pullorum strains is persistent.
Cultures of ])ulloruni carried for fourteen months in France during the
war, and kept under adverse conditions, when planted again under favor-
able conditions regained their aerogenic properties, and the activities in
this direction were as marked as in the original cultures. The 112 strains
of Bad. pullorum studied, even after being transferred eighteen times, still
retain active gas production in dextrose and mannite. In one exception,
culture No. -±4, there has never developed more than a bubble of gas in
the dextrose. This is recorded in the table in the dextrose column as B,
meaning bubble. All strains are methyl red negative. Therefore from
previous morphological and cultural tests, linked with these biochemical
findings, it may be concluded that the organism classed to-day as Bad.
■ptdlorum A should be a slender, non-motile, non-liquefying, gram-negative
bacillus. It does not coagulate or peptonize milk. It produces gas in
dextrose and mannite, forms HoS in lead acetate medium, does not produce
indol, and does not reduce nitrates.
Fermentation Tests with Baderiitm Sanguinarium.
Dextrose. — This sugar was fermented by all the five strains, 0.8 per cent being
the highest amount and 0.7 per cent the lowest, the mean being 0.7 per cent.
Mannite. — All cultures of Bact. sanguinarium produced about the same quantity
of acidity, 0.8 per cent.
Galactose. — Fermented by Bact. sanguinarium, the percentage acidity being
0.7, 0.7, 0.6, 0.8 and 0.7 per cent, respectively.
Levulose. — Fermented more variably than galactose, 0.6 per cent being the
lowest figure, and 0.9 per cent the highest.
Arabinose. — All strains fermented this carbohydrate, the readings being be-
tween 0.6 and 0.8 per cent acid.
Salicin. — Xot fermented by the five strains.
Mannose. — This carbohydrate was fermented by Bact. sanguinarium about the
same as mannite.
Xylose. — Fermented less actively in this carbohydrate, the readings being
0.5, 0.3, 0.2, 0.5 and 0.4 per cent acidity, respectively.
Adonite. — Not appreciably fermented by Bact. sanguinarium. The maximum
figure obtained was 0.1 per cent acidity.
Erythrol. — Not fermented significantly by any of the five strains of Bac. san-
guinarium.
Saccharose. — Not fermented by Bact. sanguinarium. There was increased
alkalinity.
Dulcite. — In this carbohydrate the initial acidity was increased, 0.4 per cent
being the maximum amount determined in any of the five cultures.
Dextrine. — There was a marked increase in acidity, four of the five strains of
Bact. sanguinarium showing 0.6 per cent.
Lactose. — There was no increase in acidity by Bact. sanguinarium. There was
a marked production of alkalinity.
Raffinose. — There was no increase in acidity in this carbohydrate; the initial
acidity was markedly reduced.
Inulin. — There was no increase in acidity in this carbohydrate; the initial
acidity was markedly reduced.
Maltose. — Large increase in acid was noted by all strains of Bact. sanguinarium
in this carbohydrate.
Glycerine. — None of the strains of Bact. sanguinarium produced any acid in
glycerine. The determination on the fifth day showed a reduction in initial acidity.
72
TECHNICAL BULLETIN 5.
l.O
OS
0.9
V
Q&
/
\
^
^
08
on
/
\
^
y
0.7
■v
s
^
-^
0.6
o.e
0.5
05
HO.4
r>^
+ 0.3
M,
kd
re
o
1],
^0.3
Dt.Xti^Qbtl
0.Z
OZ
01
0.1
O
o
O.I
O.l
CULT
St
«
1
;
b
^
c
5
2
3
4-
-5
l.O
0.9
Q9
08
^
^
s
V,
Q6
1 —
07
a?
/
■v.
^
OH
0.6
N
^
/
0-5
05
r,o.4.
§0.4
fQ3
<0.3
oz
M;
\/^
/I
GL
^f
OE
3;
iL.
ac
i3
L
0.1
0.1
n
o
O
01
0.1
1UM
TJKL
MR.
^
?
-
^
z
L
c
3
^JLT
UHt
Btfc
1
c
?
i
^
-
c
3
90.Z
M
|o[nl
Tfl
90.2
LApTQS
L
1
<o.i
^0.1
..JT
o
^
o
^1^
Ol
0.1
\
MOM
BtR.
2
3
?.
^
- ^
5
X*.""
6tt
1
^
^
^
^
-
3
qQ5
^0.?
o.l
GL
fCLi
LL
^C
__,
o
O.I
"^v^
run
'^."r
"c
p
">
^
.
c
5
O0.3
^0^
O.l
o
/I
u
-L
"t
Ol
^
^
^
^
•v
xj
«««
ste,l
2
^
3
I
.
1
c
?
Fig. 2. — Curves showing change of reaction in carbohydrate media by cultures of Bac-
terium sanguinarium. Percentage of acid produced at end of five-day period. Titra-
tion of 5 c.c. samples in the cold, using _ NaOHand— HCl.
20 20
BACTERIUM PULLORUM INFECTION IN FOWL.
73
0.9
1
J
08
1
M
0.7
/
0.6
'
/
Q5
1
OOA
<0.3
A
\f\Jh
OpEL
o.z
O.I
1
o
1
O.I
r~
UlTl
z :
h A-
5
09
0£
on
.^
X
as
^
V,
^
•^
Q5
y 04
1
•<
+ Q3
ARA6l/fcli>L
Q2
O.I
o
O.l
MUM
z
3
4-
5
OS
03
o*
07
OB
01
0.6
o.a
05
/
OS
^0.4
/
1
^OA
^03
<ai
oc
^E^'XT
PLI.
riL
oz
MJ
\n
n\
rt>
0.1
o.l
o
o
Ol
o.\
2
?
b
A
-
c
>
AUW
i^
r
C
■>
;
^
^
■
c
3
9Q2
N
5ACC
m;
\^
^Et
dOZ
N
lU
c
A
,
Wo.l
i 1
koA
1
o
\1
o
\
1
QI
1-
o.l
\
-<^^ 1
RtR.
i 1
2
1 [>
3 4-
s
L '
->
OX-TUW.
2
? "Z
i 1 1 1
aOZ
i
oO.-h
kaz
'jaz
ol
Lr
yi
b
i^bil
o.\
lA
rf
n
sd
k
o
' ' 1 i
o
o.l
^
^
O.I
^
^
V
v
>
nun
Sltl
1 1
2 ' 3 ' ^
-
c
>
i
c
I
b
^
.
c
>
Fig. 2. — Curves showing change of reaction in carbohydrate media byjcultures of Bac-
terium sanguinarium — Continued.
74
TECHNICAL BULLETIN 5.
Fig. 2. -Curves showing change of reaction in carbohydrate media by cultures of Bac-
terium sanguinarium — Concluded.
Table 2. -Gas Production of the 112 Strains of Bacterium pullorum in
Carbohydrate Broth.
[Percentage of gas in closed arm of fermentation tube.]
rcuL-
6
i ^ s
6
Cul-
ture
i
d
6
o5
o
s
13
a
6
s
6
i
Iture
;num-
1 BEB.
o
X
o
Q
g
o
%
a
<
g
a
"3
Q
Num-
ber.
1
Q
d
3
■5
1
X
Q
1 .
17
20
0
0
0
0
26
0
0
57 .
20 45
0
0
0
0
25
0
0
2 .
33
50
0
0
0
0
30
0
0
58 .
15
30
0
0
0
0
30
0
0
3
37
35
0
B
4
43
30
5
25
59 .
18
25
0
0
0
0
20
0
0
4 .
43
30
0
0
0
25
0
0
0
60 .
17
40
0
0
0
0
20
0
0
5 .
25
25
0
B
0
0
20
0
0
61 .
10
25
0
0
0
0
25
0
u
6 .
30
30
5
0
0
0
20
0
0
62 .
25
45
5
B
0
0
30
u
u
7 .
55
25
0
0
0
0
30
0
0
63 .
18
45
0
0
0
0
22
0
0
8 .
12
20
0
0
0
0
20
0
0
64 .
22
40
0
0
0
0
20
0
0
9
15
30
0
0
0
0
30
0
0
65 .
10
35
0
0
0
0
25
0
0
10 .
10
25
0
0
0
0
20
0
0
66 .
12
30
0
0
0
0
30
0
0
11
16
30
0
0
0
0
B
0
0
67 .
19
25
0
0
0
0
30
0
0
12 .
16
25
0
0
0
0
B
0
0
68 .
22
25
0
0
0
0
B
0
u
13 .
10
25
0
B
0
0
B
0
0
69 .
23
45
0
0
0
0
20
0
0
14 .
22
25
0
0
0
0
B
0
0
70 .
16
35
0
0
0
0
25
0
u
15
10
25
0
0
0
0
25
0
0
71 .
22
30
0
0
0
0
20
u
0
16 .
17
30
0
0
0
0
0
0
0
72 .
17
30
0
0
0
0
30
0
0
17 .
13
30
0
0
0
0
12
0
0
73 .
17
30
0
0
0
0
20
0
0
18 .
14
30
0
0
0
0
15
0
0
74 .
20
20
0
0
0
0
20
0
0
19 .
20
30
0
0
0
0
25
0
0
75 .
17
30
0
0
0
0
20
0
0
20 .
10
25
0
0
0
0
25
0
0
76 .
28
30
0
0
0
0
25
0
0
21
20
40
0
0
0
0
22
0
0
77 .
17
30
0
0
0
0
26
u
u
22 .
13
30
0
0
0
0
18
0
0
78 .
18
30
0
0
0
0
20
0
0
23 .
13
30
0
0
0
0
25
0
»
79 .
20
25
1 0
1
0
0
0
25
u
u
■n 1
3„KV
In
0-l>
Jo ga
IS.
Adonite,erythrol.saccharose:dex-trine,lactose.raffinose,inulin. maltose andglycerine produced
no gas with any of the cultures.
BACTERIUM PULLORUM INFECTION IN FOWL.
75
Table 2. — Gas Prodtiction of the 112 Strains of Bacterium PuUonmi in
Carbohydrate Broth — Concluded.
Cul-
ture
Num-
ber.
1
Q
a
6
s
1
<s
O
i
h3
6
s
d
'3
'■i
1
u
"3
Q
Cul-
ture
Num-
ber.
6
1
o
'3
a
5)
1
"a
o
s
"3
>
a
1
.5
"3
I
a
X
6
■3
Q
■2i .
17
20
0
0
0
0
5
0
0
80 .
27
40
0
0
0
0
20
0
0
25 .
15
30
10
0
0
0
35
0
0
81 .
20
30
0
0
0
0
25
0
0
26 .
28
30
0
0
0
0
B
0
0
82 .
13
30
0
0
0
0
20
0
0
27 .
23
30
0
0
0
0
25
0
0
83 .
15
25
0
0
0
0
20
0
0
28 .
18
25
0
0
0
0
15
0
0
84 .
27
25
0
0
0
0
25
0
0
29 .
20
45
0
0
0
0
B
0
0
85 .
20
25
0
0
0
0
20
0
0
30 .
10
20
0
0
0
0
B
0
0
86 .
13
25
0
0
0
0
20
0
0
31 ,
20
45
0
0
0
0
32
0
0
87 .
15
30
0
0
0
0
25
0
0
32 .
13
40
0
0
0
0
35
0
0
88 .
25
25
0
0
0
0
20
0
0
33 .
30
35
0
0
0
0
15
0
0
89 .
22
30
0
0
0
0
20
0
0
34 .
27
30
0
0
0
0
35
0
0
90 .
20
20
0
0
0
0
25
0
0
35 .
25
25
0
B
0
0
25
0
0
91 .
23
20
0
0
0
0
30
0
0
36 .
27
30
0
0
0
0
25
0
0
92 .
47
45
5
B
0
0
20
0
0
37 .
25
45
0
0
0
0
35
0
0
93 .
10
35
0
0
0
0
25
0
0
38 .
28
35
0
0
0
0
35
0
0
94 .
10
30
0
0
0
0
20
0
0
39 .
25
40
0
0
0
0
0
0
0
95 .
20
25
0
0
0
0
30
0
0
40 .
25
30
0
0
0
0
30
0
0
96 .
25
30
0
0
0
0
0
0
0
41
29
35
0
0
0
0
18
0
0
97 .
10
25
0
0
0
0
0
0
0
42
45
45
10
10
0
0
5
0
0
98 .
23
20
0
0
0
0
20
0
0
43 .
8
30
0
0
0
0
15
0
0
99 .
27
50
0
0
0
0
25
0
0
44
B
20
0
0
0
0
28
0
0
100 .
17
35
0
0
0
0
15
0
0
45 .
20
30
0
0
0
0
22
0
0
101 .
13
30
0
0
0
0
10
0
0
46 .
48
25
15
0
0
0
35
0
0
102 .
17
25
0
0
0
0
25
0
0
47 .
5
25
0
0
0
0
0
0
0
103 .
40
30
10
B
0
0
30
0
0
48 .
27
30
0
0
0
0
0
0
0
104 .
33
25
0
0
0
0
20
0
0
49 .
10
0
0
0
0
0
0
0
0
105 .
30
20
5
B
0
0
20
0
0
50 .
20
50
0
0
0
0
30
0
0
106 .
28
30
0
0
0
0
25
0
0
51 .
17
30
0
0
0
0
B
0
0
107 .
25
25
0
0
0
0
25
0
0
52 .
30
25
0
0
0
0
B
0
0
108 .
17
30
0
0
0
0
30
0
0
53 .
12
25
0
0
0
0
25
0
0
109 .
20
25
0
0
0
0
20
0
0
54 .
32
30
10
0
0
0
B
0
0
110 .
28
20
0
0
0
0
20
0
0
55 .
17
20
0
0
0
0
40
0
0
Ill .
10
25
0
0
0
0
25
0
0
56
22
30
0
0
0
0
25
0
0
112 .
22
25
0
0
0
0
25
0
0
B=Bubble. 0=Nogas.
Adonite, erythrol, saccharose, dextrine, lactose, raffinose, inulin, maltose and glycerine produced
no gas with any of the cultures.
76 TECHNICAL BULLETIN 5.
Table Z.-~ Summary of Biochemical Data as Regards Fermentation
of the 112 Strains of Bacterium Pullorum.
[Acidi and gas- production.]
1 4- = acid production.
— =alkali production.
0= neutral.
! -|- = gas produced.
— =no gas produced.
B = bubble (not enough to measure).
BACTERIUM PULLORUM INFECTION IN FOWL.
//
Table 3. — Swnmnry of Biochemical Data as Regards Fcrmcnlation
of the 112 Strains of Bacterium PiiUorxnn — Concluclcd.
[Acid' and gas- production.]
57
58
59
©O
61
62
63
64-
65
66
67
66
<o9
lO
7 1
72
73
74-
75
l(b
77
78
79
SO
61
82
83
e,^
as
86
67
83
89
90
91
92
93
94
95
96
97
98
99
100
lOl
102
103
loa.
105
106
I07
106
I09
no
in
112
1—
X
o
§
—1
y
i
K
<:
>-
1
I-
•<
1—
y
0
X
0
2
<
§
a-)
^9
1^
r,
-2
9lo
^1^
9'^
^,3
9lin
<
0
elo
9'o
u <
0
<
gS
y 0
0
in
<
19
4-
1-
4i-h
^-
—
4
-
4I-
4'-
4:-t-
4-
_
0
-
oi-
-
_
0 -
_
_
_
_
_
_
1-
t-
4
4-
4-
-
4
—
+r
4
-
^\ +
f
-
0
-
—
—
-
-1 —
—
0
-
-
-
-
-
-
-
—
-
1-
t-
4
4-
4
—
4
-
+i-
4
-
t
4
4
-
1-
-
—
-
0
-
-1 -
-
0
-
-
-
-
-
—
—
—
-
i-
1-
•h
4
4
—
4
—
4 -
+
—
4
4
4-
—
0
-
-
-
-
-
-
-
-
0
—
-
-
-
-
-
-
—
-
1-
♦-
4
4
4
—
4
—
4-
t
-
4-
4-
4
—
0
-
-
-
—
-
-
—
-
-
-
-
-
-
—
0
—
-
-
■1-
\-
4
4
4-
-1-
4
S
4I-
4-t-
■t
4
t
-
4-
-
-
-
-
-
-
-
—
-
_
-
_
-
_
0
-
-
_
+ I+-
4
1-
4
—
4
—
t-
H
4-
4-
4
-
0
-
-
-
—
-
-
-
-
-
-
-
-
-
-
-
-
-
-
+ 1*-
+
4
4
—
4
_
4;-
4'-
4
4
4
-
4
-
-
—
-
-
-
-
—
—
-
—
-
-
-
-
-
—
-
-
-t-l-
4
4
4-
—
4
-
4 -
4 -
4
4
4
-
4
-
-
-
-
-
-
-
-
-
-
-
-
--
-
-
-
-
¥ ^-
4
4
4
—
4
-
+ !-
4 -
4
4
4
-
0
-
-
—
-
-
-
—
-
-
-
-
-
_
-
-
-
-
-
•1-
(-
4
4-
4
-
41-
4|-
+ '-
4-
4
4-
-
0
-
-
-
-
-
-
-
-
-
-
— 1-
-
-
-
-
-
+-
f-
+ !-!-
4
-
4
—
4|-
4;-
-t-
4-
-
4-
-
-
-
_
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-(-
+-
4
4-
4
-
4
—
+ 1
4 -
4-
4
4
—
4-
-
-
-
-
-
-
-
-
-
0
-
-
-
-
-
-
-
—
-
-(-
1-
+
4-
■1-
-
-1-
-
4|-
i- -
4-
4-
4
-
4-
-
-
-
0
-
-
-
-
0
-
-
-
-
-
-
-
-
-
-t-
1-
4
4-
4-
-
4
—
4|-
4-; —
4-
t
4-
-
0
-
-
-
-
-
-
-
-
0
-
-
-
-
-
-
-
-
-
1-
-(-
4
4
4
-
+
-
t
-
4,-
4
1-
4-
-
4
_
_
_
-
_
-
_
0
_
_
-
_
—
-
-
-
-
-
-
-(-
(-
4
4-
4
-
4-
-
4-
-
+1-
4
4
4
—
0
_
-
-
_
_
-
—
-
0
-
—
-
-
—
-
-
—
+
4-
4
4
4
-
t
—
+!~
4-' —
4
4
4
-
4
-
-
-
-
—
—
-
—
-
0
-
-
-
-
-
-
-
-
-
+
4-
t
4-
+
-
4-
—
4!-
4 -
4-
4-
4
-
0
—
-
-
—
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
+
4-
4
4-
4
-
4-
-
+
-
4- -
-1-
+
4-
-
0
—
-
—
-
-
-
-
—
—
—
—
-
-
-
-
-
-
-
+
+-
4-
4
+ h
1-
-
4
—
4- -
+
4
4
-
0
-
-
-
—
-
-
—
-
-
—
-
-
—
—
—
—
—
-
-
+■+-
4
4
4!-
1-
—
4
0 _
4-4-
4-
-
^'\~
-
—
-
-
-
-
-
-
-
-
-
-
-
-
-
-
4-
4
4
4
4-
-
4
—
4
-
+ -
f!4
4
-
°1-
—
-
—
-
0
-
-
-
-
-
-
-
-
-
-
-
-
-
-t-
4-
4
4
4
-
4
-
41-
4- -
41 +
+
-
"^r
-
-
—
-
0
-
-
-
-
-
-
-
-
-
- —
-
-
4-
4-
4
-4
4
-
4
-
ti-
-1
4i+-
4
-
0 -
-
-
—
-
-
—
-
-
-
-
-
-
-
-
-
-
-
-1-
4
4
4
4-
-
4
-
+r
4 -
4
4-
+
-
4
-
-
-
0
-
—
-
-
-
-
-
0
-
-
-
-
-
-
-
+
t-
4-1
4
-1-1-
-4
-
4;-
t -
4
i-
4-
-
0
—
-
-
0
-
—
-
0
-
-
-
0
-
-
-
-
-
-
-
-t-
4
-t
+
4
-
4
-
4
—
4' —
4
4-
-1-
-
0
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
—
-
-t-
4
-M4
4
-
t
-
4
—
41-
t-
4
+-
-
t-
-
-
-
-
-
0
-
-
—
-
-
-
-
-
-
-
-
—
-
^^
4 4
-1-
—
4
-
+
—
4;-
-1-
4
-f
-
0
-
-
-
-
-
-
-
-
-
-
-
—
-
-
-
-
-
-
-
-t-
4
4-i4
4-[-
4
—
4
—
-^-
4-
4
4
-
0
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-l-
4
4'4
4 -
4<
—
+■
-
4 -
4-
-1-
4
—
0
—
-
—
-
—
—
—
-
—
-
-
-
-
-
-
-,-
-
-
+
4
4 '4-
4- —
-h
-
4-
4 -
4|-t-
4
-
0
-
-
-
0
—
-
-
-
-
-
-
—
-
-
-
—
-
-
-
4
4
4;-4
4
—
4
—
1-
-
4!-
4-1 4
4
-
0
-
-
-
0
-
-
-
-
-
-
-
-
-
-
-
-
_
-
—
+
4
4;4-
4-
-
4
—
4|-
4-; —
+-i +
4
-
0
—
-
-
0
-
-
-
-
-
—
-
-
-
-
-
—
-
-
4
4
4-^
4-
'r
4fi
4I-
4-i —
4!4
4
-
0
-
-
-
0
_
-
-
-
-
—
-
-
-
-
-
-
-
-
-
-1-
4
4 4
^\-
-t-
—
+ 1-
^-1-
4!4
4
_
4
_
_
_
_
_
0
_
—
_
_
_
_
_
_
—
_
-
-
-
4
4
-t 4-
-^i-
4
-
4
4
—
4
4
4-
—
0
_
-
_
-
_
-
-
-
-
-
-
-
-
-
-
-
—
-
4
4-
4i4
^-l —
-4
—
4
-
4
—
4
4
+
—
4
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
+
4
-•-4
-4-1 —
4
-
4-
—
4
—
4
_
4
-
4-
-
_
-
-
_
-
-
-
-
-
-
-
-
-
-
-
-
-
-
4^
4
-+!-+
4 -
4
-
4
—
4
-
4-
_
4
_
0
_
-
_
-
-
-
-
-
-
_
-
-
-
-
-
-
-
-
-
4'-l-
4 4
4
—
+
—
4
—
0
4
-t-
4
-
4-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
—
-
-
+ i-t-
4(4
4-
—
-t-
—
-f-
~
0
—
4-
4
4
-
•t-
-
-
-
-
-
-
-
-
—
-
-
-
-
-
-
—
-
-
-f
4
+14
4
-
4^
-
4
4
—
4-
4-
4
-
4
-
-
-
-
-
-
—
-
-
-
-
-
-
~
r
-1 -
—
-
4
4
4-4
4
—
4-
-
4
-
1-
-
4
4-
4
—
4
-
-
-
—
-
—
-
-1-
-
-
-
—
~i~
-
-
+
4-
-f-4
+ ' —
4-
—
-*-
4
-
+
4-
-t
—
0
—
-
-
-
-
-
—
-' -
-
-
-
-
-
-
-
-
-
4- 4
-I|-^
-^4
+ |£
4-
—
4 -
t
-t
4-
—
4
-
-
-
0
-
-
-
-: —
-
-
-
-
-
-
-
-
-
-
414
hU
4
—
4 -
4-
—
+1-
4-
-t-
4-
—
4
-
-
—
-
—
—
-
-
-
-
-
—
-
-
-
-
-
-
-
-t-
4
^h
4
4
4
B
4-
—
■^|-
4
4
4
-
4
_
_
_
-
_
-
_
_
-
_
-
-
-
-
_
-
_
-
_
4
-1-
^v
4-
—
4
-
4
—
-1- -
-1-
t
4-
-
0
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
^1+
+'4
-t-
—
4-1
4-
—
4-' -
4-
4-
4
-
4
-
-
-
0
-
-
—
-
-
-
-
-
-
-
-
-
-
-
-
-+|4
4|4
-^
—
4
—
4
-
4
—
4-
4
4-
-
4
-
-
-
0
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-H4-
-^'4
-h
—
4-
—
4-
-
4
—
4
4
4
-
0
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-4
4
-^ 4
4- -
4 -
-4
—
T
-
4-
-t-
4-
-
0
—
-
—
-
-
-
-
-
-
-
-
-
-
-
-
-
-
—
-
4
4
-t H-
4-;-
4|-
4
-
4-
-
4-
-4-
4-
-
0
-
-
-
0
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-^^1
41-4
4|-
+ |-
4
-
•+
-
-4-
4-
4-
-
0
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
1 4-= acid production.
— =alkali production.
0=neutral.
' 4- = gas produced.
— =no gas produced.
B = bubble (not enough to measure).
78 TECHNICAL BULLETIN 5.
A comparison of the tables which have to do with Bad. pullorum with
those which have to do with Bad. sanguinariuni shows that Bad. pullorum
is maltose-dextrine-dulcite negative, while Bad. sangumarium is maltose-
dextrine-dulcite positive. All freshly isolated strains of Bad. pullorum
(139 strains) have produced gas, while the five strains of Bad. sangui-
nariuni have never produced gas. The 112 strains of Bad. pullorum
studied have been maltose-dextrine-dulcite negative. This agrees very
well with the work of Hadley. Thus far we have isolated but one organism
from chicks, showing typical sjanptoms of white diarrhoea, which did not
produce gas in dextrose. This particular strain was maltose-dextrine-
dulcite negative, and therefore would correspond to Bad. pidlorum B or
the anaerogenic pullorum form. During the current year, 1920-21, several
anaerogenic forms have been isolated from adult hens, and they were
maltose-dextrine-dulcite negative, which in a way helps to substantiate
Hadley's claim that the Bad. pullorum infecting adult hens is maltose-
dextrine-dulcite negative, but anaerogenic. The number of cases thus
studied is meager, and future studies with more cases ought to give suffi-
cient data to establish this point. Since Hadley has been able to isolate
both aerogenic and anaerogenic forms of Bad. pullorum from the eggs of
fowls experiencing infections with the maltose-dextrine-dulcite anaerogenic
strains, and since the maltose-dextrine-dulcite negative strains isolated
by him from chicks have been aerogenic, while all the maltose-dextrine-
dulcite negative strains isolated from infections in adult birds have been
anaerogenic, the duality of the Bad. pullorum type appears to be justified.
The work presented in this paper substantiates Hadley's results. Besides,
the gas production is of great value as a differential characteristic. There-
fore it is essential in making a differential bacterial diagnosis for Bad.
pullorum to note its special morphological characteristics; to ascertain
its fermentation activities in maltose, dextrine and dulcite, and its aero-
genicitJ^ Doubtful cultures of Bad. pullorum should be submitted to the
above biochemical tests before a differential diagnosis is justified. As a
routine in this department, all doubtful cultures are tested for aerogenicity
in dextrose, and for acidity in maltose; methyl red being used as an
indicator for the increased acid production. The data at hand indicate
that there are maltose-dextrine-dulcite negative strains which do not pro-
duce gas in dextrose, and these, whether found only in adult birds or not,
should be classed as the Bad. pullorum B, different from the one so gen-
erally isolated from chicks, which is maltose-dextrine-dulcite negative,
but produces gas in dextrose.
The fowl typhoid {Bad. sanguinarium) is characterized, aside from its
specific morphology, as an anaerogenic non-motile bacillus. It does not
form indol, nor reduce nitrates. It forms H2S in lead acetate media. It
is a maltose-dextrine-dulcite positive organism.
BACTERIUM PULLORUM INFECTION IN FOWL. 79
Distribution of Fowl Typhoid in Maftsachuscttst.
During the seasons of 1919-20 and 1920-21, ol)servations were made on
all specimens sent to the laboratory for diagnosis, especially to note the
presence of Bact. sanguinarium. During that time more than 600 different
specimens were examined, and this anaerogenie, non-motile bacillus which
was maltose-dextrine-dulcite positive was isolated but six times, — three
times in the season of 1919-20 and three times in the season of 1920-21.
These cases exhibited all the post-mortem findings peculiar to this disease.
Especially noticeable were the enlarged spleen and the marked leukemic
condition. There were, however, several maltose-dextrine-dulcite negative
forms isolated which were anaerogenie, these classifying as Bact. pullonmi
B. During this same period 289 chicks, sent here with a history of bacillary
white diarrhoea, were examined, and the true Bact. pullorum was isolated
from all but one. This one strain was anaerogenie, and persistently gave
a faint acid reaction in maltose when methyl red was used as an indicator.
From this it would appear that in this one chick we were dealing with an
organism which came close to the Bact. sanguinariuvi type. From these
findings the writer is led to believe that the fowl typhoid infection in
Massachusetts is infrequent, and that the Bact. pullorum B type is far
from common. In our work of the last few years we have never isolated
from eggs a Bad. pullorum form which was anaerogenie. All cultures have
been aerogenic and have produced little or no acid in maltose, dextrine
or dulcite.
Although this represents but two years' observations, there appears to
be sufficient e\adence to indicate that fowl typhoid is not widelj' distributed
in Massachusetts; that it is not transmitted by the egg; and that Bact.
pullorum of the B iypQ is found frequently in adult stock.
Does either Bact. Pullorum or Bact. Sanguinarium play Any Part in the
so-called "Paralysis" so widely distributed in Massachusetts?
During the course of the studies concerning the diagnosis of Bact.
pullorum, there were brought to the laboratory manj^ birds suffering with
the so-called "paralj'sis," which even now is assuming a vast economic
importance in the poultry industry in Massachusetts. The weakness of
the legs and the listlessness of these birds were not essentially different
from conditions produced in rabbits when inoculated with pure cultures
of Bact. ptdlorum. With this in mind, all specimens exhibiting the paralytic
symptoms were examined bacteriologically,.with special reference to Bact.
pullorum and Bact. sanguinarium. There were 83 paralytic specimens
examined, and from 5 of them only was isolated Bact. ptdlorum of the
aerogenic type. None of the 83 specimens exhibited the marked enlarged
spleen and leukemic conditions found in fowl tyjjhoid, as laiown to us in
this laboratory. The anaerogenie maltose-dextrine-dulcite positive
organism of fowl typhoid was not isolated from any of the 83 specimens.
Cultural examinations were made of liver tissue, spleen, intestinal mucosa.
80
TECHNICAL BULLETIN 5.
ovarian tissues, and lumbar region of the spinal cord. In this so-called
"paralysis" all birds during life showed a rather bright red comb, the
paleness being evident only a short time before death. There was never
found at autopsy a marked leukemia. In fowl typhoid this leukemic
condition is highly prominent, and for this reason Moore has called this
paratyphoid type of infection ' ' infectious leukemia. ' ' Hadley has observed
a similar epidemic in fowls showing pronounced leukemic symptoms
associated with Bad. pullorum. The writer has never observed this con-
dition in relation to Bact. pullorum infections in adult birds.
From these observations on the 83 paralytic birds, with only 5 showing
the presence of the Bad. pullorum infection, — these five probably having
carried the infection since chickhood, — the evidence does not indicate
that the paralytic disease so widely distributed at certain periods of the
year in Massachusetts is due to the presence of either the pullorum or
sanguinarium type.
Influence of Infection upon the Hatching Quality of Eggs and upon the
Viability of Young Stock.
In 1917 and 1918 several sets of experiments were carried out under
the best known conditions for poultry husbandry. Eggs from 60 hens
known to have reacted positively to the agglutination test were set in an
electrobator. When tested at the end of the first seven days of incubation,
30 were found to be infertile and 2 were found dead in the shell. Of the
28 left, 10 were hatched; 3 chicks died at the end of the first day and
Bad. pullorum (aerogenic type) was isolated from the unabsorbed yolk.
All eggs containing fully developed chicks were examined especially for
Bad. pullorum, with the following results. The egg number in each case
represents the number of the hen laying the egg.
Table 4. — Results of Tests for Bacterimn P^dlorum in Dead Chicks
from Eggs laid by Positively Reacting Birds.
Egg Number.
Bact. pullorum.
Egg Number.
Bact. pullorum.
8001
+
7925
—
8384
+
7998
-
8388
—
8430
+
8002
•
8430 . . .
-
8002
—
8565
+
8430
+
8388
+
7925
-
7998
+
8565
—
8430
—
8001
+
S.'lSl
1
+ = present.
= not present.
BACTERIUM PULLORUM INFECTION IN FOWL,
81
From this table it will be seen that with the methods used it was not
possible to detect Bad. pullorum in all the dead chicks, although adult
hens were all positively reacting to the agglutination test. From 8, Bad.
pullorum was isolated without difficulty; from the other 10, the cultures
were negative.
After three months, following out three sets of incubation, the author
was able to obtain from the three sets of eggs set, 60 in each lot, all from
l^ositively reacting hens, 7 livable chicks on the first set, 9 on the second
set, and 9 on the third set, and these chicks were all given the numbers
of the parent stock from which they came: 7811, 7895, 7925, 7997, 7998,
8001, 8002, 8020, 8082, 8084, 8094, 8139, 8171, 8180, 8202, 8204, 8294,
8384, 8388, 8389, 8430, 8431, 8544, 8565, 8810. These 25 birds, all reared
from positively agglutinating hens, were yarded together and blood taken
at various times to determine whether their blood would show any signs
of agglutinative powers.
When the chicks had grown to a w^eight of at least 400 grams, they were
put together in the yard on Aug. 10, 1917. The following table Show^s
the weight of each bird at that time: — ■
Table 5. — Weight of Chicks on Aug. 10, 1.917.
Chick Number.
Weight (Grams).
Chick Number.
Weight (Grams.)
7811
870
8180
680
7895
1,200
8204
450
7925
1,240
8202
580
7997
860
8294
780
7998
1,249
8384
620
8001
1,160
8388
530
8002 . . . .
1,1.30
8389 .
540
8020
680
8430
540
8082
950
8431
380
8084
1,490
8544
510
8094
730
8565
530
8139
1,050
8810
670
8171
!
780
Agglutination tests were run on these birds, the first being on July 17,
1917. The following table shows the reactions for this and subsequent
tests : —
82
TECHNICAL BULLETIN 5.
Table 6. — Records of Agglutination Tests on Chicks hatched from Eggs
laid by Positively Reacting Hens. ^
J
18
17 AXD
1917.
July
21,
1917.
Aug
3,
1917.
Aug. 26,
1917.
Nov
7. 1917.
DILUTION OP
DILUTION OP
DILUTION OP
DILUTION OF
DILUTION OP
SERUM.
SERUM.
SERUM.
SERUM.
SERUM.
O
7
5
o
I
1
o
I
1
i
1
i
1
V
1
1
1
i
1
o
7
o
1
o
CD
1
1
o
o
1
i
o
o
o
o
1
1
rM
7811
9
?
1
1
?
c
7
7
?
7
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
7895
•>
?
1
1
9
0
0
0
0
0
c
c
c
0
0
?
?
?
?
7
7925
0
0
0
0
0
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
7997
c
c
0
0
0
0
0
0
0
0
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
7998
0
0
0
0
0
0
0
0
0
0
c
c
0
0
0
7
7
7
7
7
7
7
7
7
7
8001
0
0
0
0
0
7
7
7
7
7
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
8002
0
0
0
0
0
c
0
0
0
0
c
c
0
0
0
7
7
7
?
7
0
0
0
0
0
8020
0
0
0
0
0
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
8082
c
?
1
?
?
0
0
0
0
0
c
c
0
0
0
c
c
0
0
0
8084
1
?
1
?
?
0
0
0
0
0
c
c
c
c
c
c
c
c
c
c
c
c
0
0
0
8094
0
0
0
0
0
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
8139
?
7
?
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
8171
1
1
1
?
7
7
?
7
?
7
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
8180
c
c
c
c
c
c
c
c
c
c
c
0
0
0
0
c
c
c
0
0
c
c
c
c
c
8202
0
0
0
0
0
0
0
0
0
0
c
c
c
c
c
c
c
c
c
c
8204
?
1
1
0
0
c
c
c
c
7
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
8294
0
0
0
0
0
c
c
c
c
c
c
c
0
0
0
c
c
c
c
c
8384
c
c
c
c
c
0
0
0
0
0
c
0
0
0
0
c
c
c
0
0
c
c
c
c
c
8388
?
?
?
?
?
c
c
c
c
c
7
7
7
?
7
c
c
0
0
0
c
c
0
0
0
8430
?
?
?
0
0
c
c
c
c
c
c
c
c
c
c
c
c
c
c
.c
c
c
c
c
c
8565
?
?
?
?
7
c
c
c
c
c
c
c
c
c
0
c
c
c
c
c
c
c
c
c
c
8810
?
0
0
0
0
0
0
0
0
0
0
0
0
0
0
7
7
?
7
7
n
-, , ., _ ■ ■ - ■.:■--■ — 1
' The symbols indicating various degrees of agglutination have been taken from Hadley,
Journal of Immunology, Vol. 2, p. 46.3, 1917, as follows: C=complete agglutination; ?=doubtfuI
agglutination; 0 = no agglutination.
These experiments indicate that in chicks liatched from eggs laid by
positively reacting hens, at least six months' time should elapse before
the normal agglutination power of such sera would be sufficiently definite
to furnish indication of past or present infection. The birds reared from
hens 8001, 8139 and 8810 never showed any agglutinative power to their
blood sera. The length of time a serum maintains its agglutination power
.has not as yet been detennined.
BACTERIUM PULLORUM INFECTION IN FOWL. 83
The Present Status of the Specificity of the Agglutinatio7i Test as a Means
of Control of Bacterium Pulloruin Infection in Yoking Chicks.
During the last few years the agglutination test has become a popular
means of recognition in the domestic fowl of those individuals which have
contracted Bad. pulhrum infections in chickhood; and conseciuently, as
adult productive fowls, may have become, through infections in their
ovaries, carriers of infection to the offspring. Several writers have demon-
strated that there are certain factors which have influenced the test and
which suggest the need of modification of the method in the direction of
securing a higher degree of specificity. Hadley suggests that we stand in
need of a means of diagnosis which shall distinguish between a latent
(presumably ovarian) and a past infection. The data presented up to
date indicate that not all adult hens with Bad. pullorum have infections
localized in the ovaries; and also that not all infection has its origin in
an attack of bacillary white diarrhoea experienced in the chick stage.
This point, as Hadlej- suggests, is of less significance in its bearing upon
the validity of the results of agglutination tests for Bad. pullorum in-
fection than is the question of the specificity of the test. This author as
well as others has demonstrated the interagglutinability of Bad. pullorum,
fowl typhoid and other antigens in both Bad. pullorum and fowl typhoid
serum. Fowls which have been experimentally immunized against dif-
ferent t\-pes of fowl typhoid possess serum which agglutinates Bad.
pullorum antigens quite as well as it agglutinates its homologous antigen.
According to these data the agglutination test for the recognition of Bad.
pullorwn infection appears to lose some of its claim to specificity; and to
this extent, without carefully going over the results as obtained in field
and laboratory co-operating, it may be open to criticism.
If all operations both in field and laboratory are considered, however,
the reader will be convinced that the test yields valuable results. From
our work, already reported, during the seasons of 1919-20 and 1920-21,
there were only six cases where the anaerogenic type of organism was
isolated and the post-mortem examinations revealed the enlarged spleen
associated with leukemic conditions. This indicates that, at least so far
as this laboratory has been called upon to make examinations, fowl typhoid
infections are infrequent. That all infections are localized in the ovary is
yet to be proven. It can be said, however, that the ovarian infections
are not rare, and when they are present they persist". During the course
of the examination of hundreds of eggs for Bad. pullorum infection, only
the true aerogenic fonn of Bad. pullorum was isolated. Strains of these
cultures, even after four years, maintained this aerogenic property and
were maltose-dextrine-dulcite negative. Therefore these studies indicate
that fowl typhoid is not transmitted to the egg. In all of our work in the
bacteriological examination of young chicks, in all cases showing large
unabsorbed yolks, we have been able to isolate only the aerogenic type of
organism, and this in hundreds of cases. This shows an apparent lack of
84 TECHNICAL BULLETIN 5.
susceptibility of young stock to the Bad. sanguinarium type of infection,
and appears to substantiate the work of Dr. Hadlej', who states that he
has examined large numbers of cultures derived from j'oung stock, but
has not encountered among them the Bad. sanguinarium type.
In this laboratory hundreds of agglutination tests have been made to
demonstrate the interagglutinability of Bad. pullormn with Bad. sangui-
narium, B. typhosus, B. paratyphosus A, and B. paratyphosus B. The
results obtained here agree with those from other laboratories: i.e., that
the agglutinative tests are sufficiently definite for grouping the fowl typhoid
and pullorum types together, both demonstrating the same intimate re-
lation to tj^phoid bacilli. In every test made, the Bad. pullorum immune
serum agglutinates typhoid antigen better than typhoid serum agglutinates
pullormn antigen. Bad. sanguinarium immune serum agglutinates Bad.
pullorum much better than it does typhoid. There has never been demon-
strated any indication of an affinity of interagglutinability between B.
avisepticus (fowl cholera) and the pullorum and sanguinarium types. While
it is true that by our present methods it is difficult to differentiate san-
guinarium and pullorum by agglutination, this does not mean that appli-
cation of the test will not yield valuable results. Already, from the work
of three years, the typical maltose-dextrine-dulcite positive anaerogenic
fowl typhoid organism has been isolated six times, and in this studj^ more
than 600 specimens were examined. This indicates that fowl typhoid is
not widespread, at least in Massachusetts.
From the preceding biochemical data the establishment of Bad. pullorum
and Bad. sanguinarium as separate tyj^es is justifiable. Therefore if it
can be proven that breeding birds showing a positive agglutination reaction
may lay eggs, from which are hatched chicks developing white diarrhoea
sjmaptoms, and at death the internal organs yield cultures which demon-
strate morphologically an organism which is slender, non-motile, gram-
negative, gelatine non-liquefying, and is aerogenic, demonstrating no
acidity in maltose, dextrine and dulcite, the agglutination test would not
be invalid as an economic measure in the identification of this infection.
With this in mind, an experiment was carried out to this end.
Twenty breeding flocks were selected, all showing positively reacting
birds, and the following spring all the dead chicks from these places were
examined bacteriologicallj^ with special reference to identifying the small
gram-negative, maltose-dextrine-dulcite negative organism which was
aerogenic. The following table shows the details of the tests: —
BACTERIUM PULLORUM INFECTION IN FOWL.
85
Table 7. — Rcsitlif< on Identification of CtiUurcs iKoIated frovi Dead TA/VA-.s
which had been hatched front Eggs laid by Positively Reacting Breeding
Birdf^.
(Matoriiils for stiul.\- obtained from 20 difforonf parts of >rassachusotts.)
bueedixo
Birds.
n
2.
(Fekmentation)
Acid in —
o
si
3
.2.S5
■3Q
O
.2 a
o >,
§1
a
11
■-3 S
<
a
K
a
tj
O
.J
o
iZ
c
1
E
3
■z
(£■1
S-- B
a.t; w
s
"3
6
a
Q
6
"3
D
c
o
g
c
1
51
16
2 Y
3 H
3 Y
-
-
-
+
+
+
+C (1-400)
+C (1-400)
+C (1-400)
Bad. pullorum A
2
219
26
1 L
2 Y
3 Y
4 H
-
-
-
+
+
+
+
+C (1-400)
+C (1-400)
+ C (1-400)
+C (1-400)
Bad. pullorum A
3
216
45
29 Y
-
-
-
+
-1-C (1-400)
Bad. ■pullorum A
4
51
20
22 L
-
-
-
+
+C (1-400)
Bad. pullorum A
5
36
3
24 Y
25 Y
—
—
—
+
+
-1-C (1-400)
-l-C (1-400)
Bad. pullorum A
6
1,194
244
29 Y
-
-
-
+
+C (1-400)
Bad. pullorum A
7
784
14
31 Y
32 Y
33 Y
-
—
-
+
-h
+
-1-C (1-200)
f C (1-200)
-1-C (1-200)
Bad. pullorum A
8
250
51
39 L
40 L
—
—
—
+
+
-1-C (1-200)
fC (1-200)
Bad. pullorum A
9
89
13
45 H
-
-
-
+
+C (1-200)
Bad. pullorum A
10
393
29
52 L
53 L
54 L
—
+
+
-t-
-1-C (1-200)
+C (1-200)
-1-C (1-200)
Bad. pullorum A
11
138
21
60 L
-
-
-
+
-1-C (1-200)
Bad. pullorum A
12
76
6
61 Y
-
-
-
+
-1-C a-200)
Bad. pullorum A
13
882
129
1 L
2 Y
3 Y
4 H
-
-
-
+
+
+
+
-1-C (1-200)
-t-C (1-200)
-1-C (1-200)
+C (1-200)
Bad. pullorum A
14
116
33
2 Y
3 H
3 Y
—
—
+
+
+
-1-C (1-200)
-f-C (1-200)
f C (1-200)
Bad. pullorum A
15
264
71
1 Y
-
-
-
-
+C(l-200)
Bad. pullorum ?
16
110
46
2 Y
-
-
-
+
+C (1-200)
Bad. pullorum A
17
239
33
1 L
-
-
-
+
-t-C (1-200)
Bad. pullorum A
18
66
10
1 Y
-
-
-
+
+C (1-200)
Bad. pullorum X
19
38
11
2 Y
3 H
3 Y
—
—
-
+
+
+
-1-C (1-200)
-t-C (1-200)
-t-C (1-200)
Bad. pullorum A
20
407
103
1 L
2 Y
3 Y
4 H
-
-
-
+
+
+
+
+C (1-200)
-1-C (1-200)
+C (1-200)
+C (1-200)
Bad. pullorum A
Y = unabsorbed yolk; H = heart blood; L = liver.
86 TECHNICAL BULLETIN o.
The results presented in this table need no comment. It can readily be
seen that, with the exception of one culture obtained from flock No. 15,
all cultures obtained from dead chicks which had been hatched from
positive-reacting birds were maltose-dextrine-dulcite negative, and pro-
duced gas in dextrose. This is significant in that these flocks were widely
distributed, and the only exception to this rule was the one noted above.
This culture was maltose-dextrine-dulcite negative and was anaerogenic.
At any rate, it gave none of the reactions for Bad. sanguinarhim. On
this experiment were 5,619 breeding hens and 924 were positive reactors,
giving a positive agglutination up to dilutions of 1,000 and over. It is
reasonable to believe that these results would be substantiated by a
repetition of the experiment. While there are, as already noted, certain
factors which have influenced the test and which may suggest need of
modifications, — such as the validity of the agglutination tests, based on
interagglutinability of Bad. p^dlorum, Bad. sanguinarium and other
antigens in both Bad. pullorum and Bad. sanguinarium serum, — yet the
fact remains that in the twenty flocks mentioned the agglutination test
definitely located infection in 924 birds in a total number of 5,619. The
differential characteristics of the cultures isolated from dead chicks which
had been hatched from the eggs laid by these positive-reacting birds proved
to be typical Bad. p^dlorum, conforming morphologically and biochemically
to the standard set as a result of fermentative, serological and morpho-
logical studies completed.
After all is said about chances of error with the test, data are constantly
being accumulated which indicate that the agglutination when carefully
controlled through epidemiological work is at present the best method we
have of locating Bad. pullorum infection and furnishing poultrjanen a
starting point for its elimination.
Summary.
From the foregoing data the following conclusions appear justified con-
cerning the diagnosis of Bad. pullorum infection in the domestic fowl: —
1. From the fermentation studies conducted over a period of three
3'ears, it was found that Bad. pullorum is maltose-dextrine-dulcite negative
and aerogenic, while all cultures of Bad. sanguinarium studied have been
maltose-dextrine-dulcite positive and anaerogenic. These characteristics
are constant. Whenever there has been question as to cultural and
morphological differentiations, these investigations have shown that the
biochemical tests have aided in making a final decision. Variations in
morphology of the pullorum strains are frequent; therefore doubtful
cultures should be submitted to the maltose-dextrine-dulcite test and
checked by gas production in dextrose. Experience has shown that this
procedure should be followed as a routine in all laboratories having to do
with the pullorum problem.
2. From the examination of 600 avian specimens for the anaerogenic,
non-motile, maltose-dextrine-dulcite positive form which produced en-
BACTERIUM PULLURUM INFECTION IN FOWL. 87
laro;o(I s])loons associated with marked leukemic conditions, it was of
sonu' significance that the true sanguinarium culture was identified only
six tunes. Chick examinations conducted during this same period, repre-
senting several hundred examinations, all yielded tj^Dical pullorum cultures.
There was but one exception, and this culture was probably an atypical
])ull()rum form which had become anaerogenic. In the examination of the
adult avian specimens, the maltose-dextrine-dulcite negative forms isolated
from several dead hens indicate that Hadley is correct in his contention
that Bad. pullornm may assume a dual role: Bad. pullorum A being
maltose-dextrine-dulcite negative and aerogenic, infecting young chicks;
and Bad. pullorum B being maltose-dextrine-dulcite negative and an-
aerogenic, infecting adult hens. Cultures from eggs have always been
aerogenic. If knowledge of Bad. sanguinariuni is based upon the anaero-
genicity of cultures, the absence of this property in cultures isolated from
adult hens, chicks and eggs sent from all parts of the State would appear
to indicate that fowl tj^phoid is not widely distributed in Massachusetts.
3. From pathological and bacteriological examination of 83 birds
suffering with the so-called "paralysis," the evidence at hand does not
indicate that the disease, so widely distributed at certain periods of the
A'ear, is due to the presence of the pullorum or sanguinarium type of
organism.
4. The agglutination test has become a popular means of recognition in
the domestic fowl of those individuals which have contracted infections
in chickhood, and consequentl}^ as adult productive fowls, may have
become, through infections in their ovaries, carriers of infection to the
offspring. During this investigation hundreds of agglutination tests have
been made, demonstrating that there is an interagglutinability of Bad.
pullorum Avith Bad. sanguinarium, B. typJiosus, B. paratyphosus A and
B. paratyphosus B antigens, with a consequent tendency to make the test
lose in terms of specificity. The fact remains, however, as a result of
experiments in this department, that in twenty flocks studied, representing
5,619 breeding birds, the test located infection in 924. Furthermore, the
differential characteristics of the cultures isolated from dead chicks which
had been hatched from eggs laid by these positively reacting birds proved
them to be tjT^ical Bad. pullorum, conforming morphologically and bio-
chemically to the standard set for this organism. Therefore, from these
data, the conclusion seems justified that the agglutination test, when
carefull}' controlled through epidemiological work, is at present the best
method we have for locating Bad. pullorum infection and furnishing to
poultrymen a starting point for its elimination.
88 TECHNICAL BULLETIN 5.
Literature Cited.
(1) Hadley, Philip. 1918. The Colon-typhoid Intermediates as Causative Agents
of Disease in Birds. 1. The Paratyphoid Bacteria. R. I. Agr. Expt.
Sta. Bui. No. 174.
(2) Smith, Th., and Ten Broeck, C. 1915. A Note on the Relation between
Bacterium pullorum (Rettger) and the Fowl Typhoid Bacillus (Moore).
In Jour. Med. Research, Vol. XXXI, pp. 547-555.
(3) Smith, Th., and Ten Broeck, C. 1915. Agglutination Affinities of a Patho-
genic Bacillus from Fowls (Fowl Typhoid) (Bacterium sa^iguinarium ,
Moore), with the Typhoid Bacillus of Man. In Jour. Med. Research,
Vol. XXXI, pp. 503-521.
(4) Rettger, Leo F., and Koser, Stewart G. 1917. A comparative study of Bac-
terium pullonim (Rettger) and Bacterium sanguinarium (Moore). In Jour.
Med. Research, Vol. XXXV, No. 3, 1916, pp. 443-458.
(5) Taylor, Walter J. 1916. An Outbreak of Fowl Typhoid. In Jour. Amer.
Vet. Med. Assoc, Vol. 49, pp. 35-49.
(6) Ward, Archibald R., and Gallagher, Bernard A. 1917. An Intradermal
Test for Bacterium pullorum Infection in Fowls. U. S. Dept. Agr. Bui.
No. 517.
(7) Pfeiler, W., and Rehse, A. 1913. Bacillus typhi galHimrum alkalifaciens and
the Disease which it causes in Fowls. Mitteilungen d. Kaiser-Wilhelms
Institut f. Landwirtschaft in Bromberg, Vol. 5, pp. 306-321.
(8) Goldberg, S. A. 1917. A Study of the Fermenting Properties of Bacterium
pullorum (Rettger) and Bacterium sanguinarium (Moore). In Jour.
Amer. Vet. Med. Assoc, Vol. 51, pp. 203-210.
(9) Hadley, Philip, Caldwell, Dorothy W., and Heath, Bertha M. 1919. Bac-
teriological Notes. In Jour. Bacteriology, Vol. IV, No. 1, p. 65.
(10) Hadley, Philip. 1917. Infections caused by Bacterium pullorum in Adult
Fowls. R. I. Agr. Expt. Sta. Bui. No. 172.
(11) Mulsow, F. W. 1919. The Differentiation and Distribution of the Para-
typhoid-enteritidis Group. VI. Avian Typhoid Bacilli: a Comparative
Study of Bacterium pullorum and Bacterium sangxdnarium. In Jour.
Infect. Diseases, Vol. 25, pp. 135-162.
(12) Krumwiede, Chas., Jr., and Kohn, Laurence. 1917. Studies on the Para-
typhoid-enteritidis Group. IV. The Differentiation of the Members of
the Paratyphoid-enteritidis Group from B. typhosus, with Special Reference
to Anaerogenic Strains and Observations on the Fermentative Character-
istics of the Avian Types. In Jour. Med. Research, Vol. 36, p. 509.
A^AHERST. ^AA-^
MASSACHUSETTS
Agricultural Experiment Station
TECHNICAL BULLETIN No. 6 JANUARY, 1924
THE INHERITANCE OF FERTILITY AND
HATCHABILITY IN POULTRY
By F. A. HAYS and RUBY SANBORN
DeteiTnination of fact as to inheritance of characters is essential to suc-
cessful poultry breeding. This work is peculiarly within the province of
tlie Agricultural Experiment Station, for records must be made on large
numbers of individual birds, the work must extend over a period of years,
a wearisome amount of data must be preserved. The data recorded in
this bulletin are the result of eleven years' work. Individual records were
made on 886 birds. Resulting data are now analyzed statistically in the
iiuht of all that genetic science has to offer. It is through work such as
this that a basis of sound fact, in poultry breeding work, will ultimately
replace one based largely on opinion and tradition.
Requests for Bulletins should be addressed to the
AGRICULTURAL EXPERIMENT STATION
AMHERST, MASS.
Publication of this Document approved by the Commission on Administration and Finance
20
THE INHERITANCE OF FERTILITY AND HATCH-
ABILITY IN POULTRY.
Bt F. a. Hays and Ruby Sanborn.
Introduction'.
The importance of a thorough understanding of the mode of inheritance of
factors affecting fertility of hens' eggs needs no stressing. Neither does the value
of a complete understanding of the way hatching power of eggs is inherited require
emphasis, for the proper functioning of the factors for high fertility and high hatch-
ability is of fundamental and vital importance to ever}"" poultry breeder.
The purpose of this report is to consider only the question of the inheritance of
fertility^ and hatchability- from as many angles as our data will permit. The
inheritance of these two characteristics is discussed first from the standpoint of the
dams and then from the standpoint of the sires. The fact should be recognized
at the outset that numerous variable environmental factors such as weather con-
ditions, health of birds, exposure of eggs, variation \\ithin the same and different
incubators, etc., are in constant operation. The combined action of these con-
stantl}' varAang environmental factors may largelj^ obscure the inherent capacity
of the bird to produce fertile eggs that are largely hatchable. A further lack of
knowledge of the fundamental factors concerned in breeding for high fertiUty and
high hatchabiUt}^, as pointed out by Dunn ('23), makes proper matings impossible.
Data Available.'
The data used in this bulletin have been collected each hatching season from 1913
to 1923. All records kept represent the pullet year or cockerel year unless otherwise
stated. All records were made by pedigreed Rhode Island Red birds. The atten-
tion of the reader is called to the fact that stud matings have been used almost
exclusively and this will account for a lower degree of fertility than might be
obtained from pen matings. Uniform methods of incubation have been used and
care has been taken to maintain a definite system of management throughout the
eleven-year period. Only females whose daughters were trap-nested are included
in this report.
PART I.
The Female's Role in the Inheritance of Fertility and Hatchability.
Fortunately a measure of individual fertiUty and hatchability is possible in the
female. The accuracy of such a measure depends very largely upon the number
of eggs laid by the pullets in question during the hatching season. Some pullets
will lay fifty eggs during a two months' incubation season, while others may lay
as few as five or ten eggs. Fertility and hatchabiUty records on the first type would
certainly be much more significant than those on the second type. The major
portion of the records here reported upon were made between the hatching dates
of March 25 and May 15 of the respective .years. In some cases chicks were hatched
beyond the above dates, but not as a rule. Since the flock was being bred for egg
production, considerable care was exercised to use pullet breeders that would lay
a goodly number of eggs during the hatching season.
Section L Correlation between Fertilihj and Hatchability.
A hen to be able to produce a large number of chicks must lay highly fertile
eggs. Furthermore, her eggs must hatch well. In ordinary usage, good hatching
hens are those from which almost all eggs laid give rise to vigorous chicks. Fer-
tility and hatchabihty are bound together in the sense that there can be no hatch-
1 The term fertility as used here refers to the percentage of eggs that are fertile; the test being made on the fifth
day of incubation. ,
• The trriii hatchability as used here refers to the percentage of fertile eggs hatched.
3 The data used in this report icere collected by Dr. H. D. Goodale until 19S1; for the year 1922, by Professor
William Sanctuary and the junior author.
21
ability without fertility; but there may be one hundred per cent fertility and zero
hatchability, or there maj' be only five per cent fertility and one hundred per cent
hatchability.
The above facts show that the coefficient of correlation between fertility and
hatchability could neither be zero nor nep;ative. Pearl ('09) found a correlation of
— .127±.071 between the percentage of infertile eggs and the percentage of fertile
eggs hatched from pullets. Such a factor, in view of the large probable error, in-
dicates no sensible correlation between the degree of fertility and the percentage
of fertile eggs hatched.
In table 1 presented below, the percentage of fertile eggs from 758 pullets is
correlated with the percentage of fertile eggs liatched. These percentages represent
each pullet's average fertility record and her average hatching record for the season.
The records were obtained in eleven breeding seasons. The table includes all
pullets used as breeders during the period covered, except those showing zero
fertiUty. The zero-fertility class had to be omitted because zero fertility always
means zero hatchability, and if the fifty-three pullets that laid no fertile eggs were
included, a spurious correlation would arise and not the true correlation coefficient.
Table 1
.
Correlation Between
Fer
tility and Hatchability.
Pullets' Hatchability, Per Cent.
1
o
C5
o
1
o
o
7 7
o
T
lO
1
1
U5
O
CD
at
1
s
1
i
1
o
1
I
o
o
1
OS
f.
1-4
5-9
4
2
6
10-14
2
1
1
1
5
15-19
2
1
1
1
1
6
20-24
5
1
2
1
1
1
1
1
1
14
25-29
1
1
2
1
2
2
1
2
12
^
30-34
3
1
1
1
1
3
10
0)
U
35-39
1
1
1
1
1
5
1
1
2
14
40-44
3
1
1
1
2
4
1
1
14
^
45-49
3
1
2
1
2
3
2
14
50-54
3
1
1
1
2
2
1
1
4
1
1
5
23
"to
55-59
1
1
1
1
1
1
1
1
2
1
1
13
_o
60-64
1
2
2
1
3
2
2
6
1
2
1
24
PL,
65-69
6
1
2
2
2
1
1
2
2
1
7
1
3
4
3
1
40
70-74
5
1
3
1
2
1
2
3
4
3
2
3
6
2
2
41
75-79
4
3
1
1
2
1
1
3
1
4
3
1
3
5
1
3
2
2
42
80-84
1
2
2
4
2
1
2
7
2
6
6
3
4
4
4
3
53
85-89
5
1
3
3
1
3
7
4
7
3
5
7
9
9
4
2
2
1
1
77
90-94
10
3
1
2
3
4
4
3
2
3
3
11
8
5
8
9
8
4
91
95-100
28
5
4
10
5
7
10
29
9
28
9
24
7
24
13
38
12
35
20
74
22
55
15
50
22
70
17
47
18
38
20
33
6
35
259
f.
r
87
13
16
18
21
23
758
Constants calculated from Table 1
Mean fertility ....
Fertility standard deviation .
Mean hatchabifity
HatchabiUty standard deviation
Coefficient of correlation
.688272 ±.005466
.2231 ±.003865
.637875±. 007119
.2906±. 005034
.0672±. 024390
Table 1 gives a positive correlation coefficient of .0672±. 02439 which must be
interpreted in the light of a probable error of more than one-third as signifying
22
almost complete independence between degree of fertility and hatchability.
From the genetic standpoint, the results in table 1 are significant. The table
shows that a flock of pullets may carry the factors that are conducive to high fer-l
tility and vet lack the ability to be good hatchers. Stated simply, these results
mean that "the degree of fertility in a hen's eggs is an entity independent from the
hatchal)ilitv of her eggs.
The mean fertility shown in table 1 is .6883, while the mean hatchability is .6379.
Of the total eggs hiid by these pullets during the hatching season, 68.83 per cent
were fertile, and 63.79 per cent of these fertile eggs hatched. Two possible avenues
are open for increasing the number of chicks per pullet. First, Increase the per-
centage of total eggs that are fertile. Second, Increase the percentage of fertile
eggs that hatch. Selection for high fertility and high hafcchabihty is possible only
where hens are used as breeders. Hens have been used to only a very minor extent
in this flock. Hence there has not been much progress in fertility and only moder-
ate pi-ogress in hatchabihty, as will be shown in section 17 of this bulletin. The
general deduction must be made, therefore, from the study of table 1, that fer-
tility and hatchability are independent of each other. The stability of each char-
acteristic may next be considered.
Section 2. The Constancy of Fertility in Hens.
In order to test the constancy of fertility in hens, the records of 253
female breeders that were used first as pullets and again as yearhngs have been
placed in table 2. In practicality all cases a different male was mated to these
females the second year. If there is a sensible correlation in fertiUty between the
pullet-year record and the yearUng record from the same hens, the natural assump-
tion must be that degree of fertility is more or less constant in the female, regard-
less of the male to which she is mated.
Table 2. — Correlation Between First and Second-Year Fertility.
Yearling Hens' Fertility,
Per Cent.
o
o T
in 2
cr.
T
1
O
Ci
CM
1
o
CO
J.
o
1
1
1
02
1
CO
1
o
OS
1
o
00
5
00
1
o
g
T
f.
0-4
3
1
5
9
5-9
1
1
2
10-14
1
1
1
3
15-19
1
2
3
20-24
1
1
2
1
5
25-29
1
1
1
3
c
30-34
1
1
U
35-39
1
1
1
1
9
6
(m
&
40-44
1
2
3
.§
45-49
1
1
1
1
1
2
7
50-54
1
1
3
2
7
55-59
60-64
1
3
2
1
3
6
f^
65-69
70-74
75-79
80-84
85-89
90-94
95-100
1
2
2
1
2
1
1
1
1
2
2
1
1
1
1
1
1
1
1
1
2
1
1
2
1
1
1
3
1
3
2
1
1
1
1
1
1
1
2
3
7
2
1
2
2
1
7
3
24
3
3
2
2
2
5
8
32
4
3
3
6
5
19
50
15
12
9
20
20
42
77
f.
15
1
3
3
6
1
2
3
3
2
3
4
3
11
7
4
17
109
253
23
Constants ralcnlatcd from Tabic 2.
Pullets' mean fertility 7r)S<)± .011288
Pullets' standard deviation 2C.()2± ,(H)79S2
Yearlin-i; hens' mean fertility 7825^.012111
Yearling hens' standard deviation 28r)()± .(K)8r)()4
Coefficient of eorrelat ion 27;«i: .():«)238
The mean fertility of the birds used in table 2 was slightly fjireatcr for the year-
Kng than for the pullet-year. The ditference, .02oG=b.()l()r)79, is not great enough
to be significant. The range of variability measured l)y the standard deviation is
slightly wider as yearlings than as pullets, but the closeness of agreement in the
two years signifies a degree of fixedness. From tiie breeding standpoint, the chief
deduction that may be made from a study of table 2 is that the percentage of fer-
tility for a pullet is a good guide as to her probable fertility as a y(>arling.
A positive coefficient of correlation, .2733 ±".039238, between the first and second
year fertility supports the \\e\\' that fertility is a trait that is fairly constant for the
individual hen. Lamson and Card ('20) have pointed out this fact in I/'ghorns.
Pearl ('09) found a negative correlation of .lir2±.092 between infertility the first
year and the second year in Barred Plymouth Rocks. Our data, however, indicate
tliat a bird with good fertility as a pullet will probably show good fertility as a
j yearling.
Section 3. The Constancy of Hatching Power in Hens.
The group of 253 birds studied in table 2 are correlated for hatchability in table
3 to discover if there is a relationship between the percentage of fertile eggs hatched
as pullets and as yearlings. In other words, does hatchabihty approach any degree
of constancy in the same individual in two successive years? Does a good hatching
record as a pullet mean a good hatching record as a yearling?
Table 3. — Correlation Between First and Second-Year natchabilit}j.
Yearling Hens' Hatchability, Per Cent.
1
o
OS
1
s
1
1
o
^ 7
3
T
o
T
1
o
1
1
1
oo
OS
1
U5
1
O
o
o
'k
f.
0-4
10
2
1
2
1
1
2
I
20
5-9
2
1
1
1
5
10-14
3
1
1
5
15-19
3
1
I
1
1
7
20-24
3
3
25-29
2
2
1
1
1
1
8
c
30-34
1
1
1
1
1
5
(5
35-39
3
1
2
1
1
8
^
40-44
2
1
1
1
1
2
2
10
>;
45-49
1
3
1
1
2
1
9
J3
03
50-54
5
2
2
1
1
2
1
2
1
1
1
1
20
"S
55-59
1
2
1
1
3
2
3
1
2
1
17
♦J
S
"5
60-64
65-69
1
2
1
1
1
3
3
2
1
1
3
1
3
2
2
1
2
2
1
1
2
1
2
20
19
Ph
70-74
75-79
80-84
8.5-89
90-94
2
4
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
3
2
1
2
1
2
2
3
3
1
1
1
2
1
2
5
3
4
1
1
2
1
1
3
1
2
2
4
3
1
1
1
1
22
24
8
18
14
95-100
2
10
10
1
17
19
4
17
12
18
2
15
1
16
7
1
7
11
c- —
f.
"
5
7
4 9
6
12
5
11
253
24
Constants calculated from Table 3.
Pullets' mean hatchability .
Pullets ' standard deviation .
Yearling hens ' mean hatchability
Yearling hens ' standard deviation
Coefficient of correlation
.5678±.011313
.2668 ±.008333
.4791 ±.012963
.3057±. 009166
.4346±. 034409
The mean hatchability for pullets is .5678±.011313. The mean hatchabiUty
for the same birds as yearlings is .4791±.0r2963. There is a difference of .0887±
.0172 in favor of using pullet breeders. This difference is significant in the Hght of
its probable error. Stewart and Atwood's ('09) records with White Leghorns do
not agree with these results. They found both the mean fertihty and mean hatch-
abiUty to be higher in hens than in pullets. Their records are scarcely comparable
with ours because they did not compare the same birds. Furthermore, in a yearling
or two-year-old flock, most of the poor hatchers will have been discarded if they
were tested as pullets. Pearl ('09) obtained a slightly higher mean fertility in the
pullet year and an insignificant difference in hatchability between pullets and
yearlings, using the same flock of Barred Pl^-mouth Rocks.
The range of variabihty measured by the standard deviation is significantly
greater in the yearling hens. This difference may possibly be ascribed to vari-
ability in physical condition in the older birds. Hatchabihty, however, seems to be
a trait that behaves mth a good deal of constancy in hens. This fact makes the
indi\adual hatching record valuable, at least in making use of a hen for several
years to increase flock numbers. The ability of the hen to transmit this hatching
power to her daughters will be considered in section 5.
The coefficient of correlation calculated from table 3 is .4346±. 034409. Hatch-.
abiUty is therefore more constant than fertility, for the coefficient for fertility in
the same flock was only .2733. In breeding for high hatchability there is ample
justification for discarding the poor hatchers the first year and retaining the good
hatchers to perpetuate the flock.
Section 4. Correlation in Fertility between Mothers and Daughters.
In order to discover if there is any relationship between mothers and daughters
in degree of fertihty, the average fertility of pullet breeders has been correlated
with each of their daughters that were used for breeding as pullets. In case only
one daughter was used, there was but one insertion in the table. If a pullet dam
had more than one daughter used as a breeder she is paired with each of these
daughters and an insertion made in the table.
Tarle 4. — Correlation in FcrtilHii Bdwcni Mother find Daughter.
25
Dauohteus' Fertihtv, 1'er Cent.
■*
Oi
T
o>
■^
o>
^
7
T
S
o
s
r
o>
^
g
S
g
f.
o
.i
o
o
U5
o
CO
o
-r
o
U5
S
g
CO
g
K
s
g
o>
0-4
:.-9
lO-U
15-19
1
1
2
20-24
25-29
1
1
80-34
35-39
1
1
1
1
2
1
1
1
1
5
5
(£
40-44
1
1
1
1
2
2
1
4
13
>i
45-49
1
2
2
3
8
t2
50-54
55-59
1
1
1
1
1
2
1
3
2
4
8
9
c
Q
60-64
2
1
1
1
1
2
1
1
7
17
65-69
3
2
2
1
2
1
2
1
1
2
3
3
8
31
70-74
2
1
1
1
1
1
2
1
5
2
3
4
13
37
75-79
1
1
1
4
2
2
3
2
3
5
24
80-84
5
3
1
1
2
1
1
4
2
1
2
7
17
47
S5-S9
5
1
1
2
2
3
1
1
2
2
1
2
4
3
7
10
17
64
90-94
12
1
1
1
1
3
1
1
3
4
5
3
7
6
4
5
10
11
45
124
95-100
23
4
2
4
6
5
5
7
6
6
8
9
15
20
21
23
31
44
45
132
416
f.
54
6
5
7
14
n"
10
13
14
14
23
12
24
40
41
42
54
77
91
259
811
Constants calculated from Table 4.
Dams' mean fertility 8765 ±.003503
Dams ' standard deviation 1479zb. 002477
Daughters ' mean fertility 7378±. 006831
Daughters ' standard deviation 2884it .004830
Coefficient of correlation 0147±. 023679
The standard deviation in dams in fertihty is .1479, while the standard deviation
of their daughters is twice as great or .2884. There is a positive correlation co-
efficient in fertility of .0147±.023679 between the dams and the 811 daughters
that were used as breeders. Since this coefficient is less than its probable error,
it can have no significance. This table must therefore indicate that a pullet with
low fertility is as likely to give daughters high in fertihty as is a breeding pullet
that shows high fertihty herself. These observations are essentially in agreement
with Pearl ('09), for he found a negative correlation of .035db.072 in infertility
between mother and daughter. The conclusion seems justified, therefore, that the
fertihty of the dam 's eggs is no indication as to the probable fertihty of her daughter's
eggs. In section 2, the fertility record of a pullet was shown to be a guide as to her
second-year fertility. Since the dam's fertility record is not a dependable index of
her ability to breed true for fertihty, the only satisfactory test is the progeny test,
for fertihty seems to depend upon many as yet unrecognized factors, or else'is not
an inherited characteristic.
Section 5. Correlation in Hatchability between Mothers and Daughters.
The identical group of dams and daughters used in table 4 has again been cor-
related in table 5, using percentage of fertile eggs hatched.
26
Table 5. — Correlation in Hatchahility Between Mother and Daughter.
Daughters
' Hatch
ABILITY,
Per Cent.
o
z
o
3
S
g
r
1
OS
1
1
1
1
3
"3
o
CO
03
to
1
1
o
o»
1
{2
1
Oa
oo
§
T
03
f.
0-4
5-9
1
1
10-14
1
1
15-19
1
1
20-24
1
1
2
25-29
3
2
5
c
30-34
1
1
1
1
1
2
1
1
9
35-39
4
2
1
5
3
1
2
1
19
(S
40-44
6
2
2
3
3
1
1
1
1
1
3
24
^
45-49
7
3
2
1
2
1
3
1
2
3
2
27
1
w
Q
50-54
55-59
7
10
2
1
1
3
1
1
4
1
4
1
3
2
3
4
1
1
1
2
1
3
3
9
1
3
1
5
7
5
1
2
1
1
1
3
2
1
36
67
60-64
65-69
13
18
1
2
3
4
3
2
4
4
2
3
4
6
2
3
5
2
4
8
11
4
6
7
4
9
7
4
5
3
2
3
4
5
3
84
86
70-74
8
2
1
2
2
2
4
3
1
1
9
3
6
2
3
8
5
4
3
1
70
75-79
28
4
4
3
4
3
2
4
6
2
7
12
5
8
6
9
1
7
115
80-84
11
2
1
6
3
2
2
3
4
8
5
6
6
4
10
5
4
82
85-89
11
3
3
1
1
3
2
2
1
1
4
3
8
6
9
6
6
4
5
79
90-94
6
2
1
1
2
1
3
1
2
3
2
7
6
7
7
5
6
5
67
95-100
2
2
2
2
2
3
2
3
4
6
2
3
3
36
f.
137
14
16
17
22
23
28
27
23
24
38
34
75
53
50
70
48
40
33
39
811
Constants calculated from Table 5.
Dams' mean hatchability 7064±. 003891
Dams' standard deviation 1643±. 002752
Daughters' mean hatchability 509 l=b .007340
Daughters' standard deviation 3099 ±.005 190
Coefficient of correlation 1960it .022805
Table 5 undoubtedly shows that hatching power is transmitted from mother to
daughter, yet while the dam's mean hatchabihty is .7064, her daughter's mean was
only .5091. The standard deviation of dams was .1643 and their daughters'
standard deviation was .3099. Thus the range of variation in daughters as meas-
ured by the magnitude of their standard deviation is almost double that of their
dams. Such would be the case if a dominant factor is present for high hatcha-
bihty. This relative variability is in exact agreement with the same observation
on fertility" as pointed out in section 4.
There is a positive correlation coefficient of .1960±. 022805 between dams and
daughters in hatchability. During the progress of the experiment, the pullet
breeders used on successive years came from pullet mothers that showed a good
hatching percentage. In other words, the pullets that were used as breeders in any
one year came from pullet dams that had laid eggs of good hatching power.
According to Pearson ('03) rigid selection in parents may reduce the correlation
between parent and offspring for the character in question. Since we have no fer-
tility and hatchabihty records for the flock as a whole, it is impossible to mathe-
matically measure the effect of such selection on our flock.
27
Pearl ('09) reix)rts a correlation coefficient of only .031 ±.072 between mothers
and daughters in hatchahility, but only 87 indi\iduals were studied. Dunn ('23)
states that he was unable to separate high and low hatching lines by two genera-
tions of selection. He did find, however, that families tend to become different
in hatching power and to retain this ditTerence.
Table 5 clearly indicates that hatching power is transmitted from mother to
daughter, even though rigitl control of the many environmental factors that modify
the liatching power is very difficult. These varying conditions often obscure the
tiiie hatching ability of the pullet as an individual. The use of t)re('ding females
of liigh hatching power is the first stej) toward improving the flock in this par-
ticular characteristic. We have shown in section 3 that the hatching power
of a pullet is sensibly correlated with her later hatching power. Follow this by
using breeding hens that transmit high hatchability to all of their daughters. The
male 's part in heredity of hatchability will next be considered.
PART II.
The Male's Role in Inheritance op Fertility and Hatchability.
Section 6. The Constancy of Fertility in Males.
In studying the question of the inheritance of fertility and hatchability, much
importance should be attached to the male side of the flock, for the male is more
than half the flock from a genetic standpoint because each male furnishes half the
inheritance to the progeny of several hens.
The measure of the male's fertilizing ability is the mean degree of fertility from
his different matings. The accuracy of such a measure will of course depend upon
whether or not high fertility is governed in inheritance by dominant or recessive
factors, or whether it is independent of Mendelian factors. If high fertility depends
upon recessive factors, we should expect less variation in the daughters from a hen
that carries these factors pure, so that she herself is genetically highly fertile, than
would be the case if high fertiUty is dependent on dominant factors and these were
not in homozygous condition. The fact that manifestation of fertility in the eggs is
probably dependent on both miile and female makes the classification of either
males or females with regard to this characteristic a hazardous undertaking. A
careful analysis of the results from mating specific males to a number of females
in successive years with conditions kept uniform would help much to explain this
confusing problem.
The problem of the constancy of a male's ability to transmit a certain degree of
fertility to his daughters may be elucidated by correlating the fertility of his
daughters sired during his first breeding year with that of his daughters sired during
the second breeding year, using pullet records in all cases. In other words, if males
transmit a certain degree of fertility to their daughters in successive years, a posi-
tive correlation will exist. Such a tabulation is made from data available in table
6. Unfortunate!}'', records on only 51 pairs of daughters are obtainable for study.
The number is small because few males are used as breeders after their cockerel
year.
28
Table 6.
Correlation in Fertility between Males' First and Second-Year Daughters.
Fertility op Males' Second Year Dauqhters, Per Cent. |
a>
1
1
1
OS
7
CO
1
-:)<
-<1<
OS
T
Oi
CO
to
ts
OS
s
OS
?
s
1
o
1
U5
o
W5
o
"5
o
CO
O
S
iS
o
s
s
s
1
o
s
§
1
0-4
2
1
1
1
5-9
1
1
10-14
15-19
1
«
u
20-24
1
1
Lh
(S
25-29
1
30-34
-iS
M
35-39
tS
p
40-44
1
1
p>
i
45-49
1
r^
-fi
50-54
I-
s
55-59
1
'n
s
"cS
60-64
1
s
"o
65-69
1
1
^
70-74
1
2
®
75-79
1
1
1
2
tN
80-84
85-89
90-94
95-100
1
1
1
1
1
1
1
1
1
1
3
4
2
1
5
f.
2
0
1
0
1
0
0
2
0
1
4
1
1
1
3
2
3
2
7
20
10
Constants calculated from Table 6.
First-year daughters' mean fertility .
First-year daughters' standard deviation
Second-year daughters' mean fertility
Second-year daughters' standard deviation
Coefficient of correlation
. 6651 db. 031064
.3289±. 021966
.7700zb. 025001
.2647±. 017678
.2151 ±.090076
In table 6 the mean fertility of the first-year daughters was .6651 while the mean
for the second-year daughters was .77. There is a difference of .1049±.0399,
which, judged by the magnitude of its probable error, is of doubtful significance.
There is also no sensible difference in the standard deviation of first-and second-
year daughters. A sensible degree of correlation between first-and second-year
daughters is questionable because r = .2151 zb .090076. The probable error is almost
half as great as the coefficient itself. The only logical interpretation that can be
placed on the Hmited data in table 6 is that mean fertility in the daughters of the
same group of males in successive years is strikingly constant, and in the second
place that a positive correlation coefficient of questionable magnitude exists be-
tween first-and second-year daughters in fertility. More data of a similar nature
are required to clear up this question.
Section 7. The Constancy of Hatchability in Males.
The male's ability to transmit fertility is still questionable, as has been pointed
out in section 6. In the present section the subject of the constancy of hatchabihty
in the male, as measured through his daughters, will be considered. The same
difficulties are encountered in studying this question that have already been men-
29
ioned for fertility. Possibly environmental factors are of less importance in hatch-
ibility than in fertility. Pearl ('00) believes that hatchinp; cjuality is more of an
innate eonstitntional character than is fertility. If hatchinjj; (|uality is (lej^-ndent
upon Mendelian factors in inheritance, the dejjree of correlation between hatch-
:i!)ility of the egss of first-year daufj;hters and th(» e^S^ of second-year daut^hters
wonld vary with the number of factors concerned, and with the dcfirce of liomo-
/viiosity in tlie males for tliese factors. Should there be a sensil)le ))ositive correla-
tion, it would indicate that the male as well as the female transmits hatching power
to the offspring.
In tal)le 7, the group of 51 pairs of daughters studied in secticm 0 is talnilated for
hatchability.
T.uiLE 7. — Correlation in Hatchability between Males' Fird and Second-Year
Daughters.
Hatchabilitt of Males' Second Yeah Dauohteks, Peu Cent.
-»< CS
Tf
05
•<»<
OJ
■»■
01
•*
o>
■»)<
OS
■*
o>
S
03
■^
0
0
f.
CM
CM
.«<
to
to
1^
0
o>
T T
1
1
1
1
1
1
1
1
1
1
1
1
0 "5
0
"5
0
W5
0
10
0
it3
0
0
>o
0
»o
0
»c
0
CM
CM
■^
to
to
1^
00
00
0
0-4
6
1
2
1
1
2
2
3
1
1
20
5-9
10-14
1
"S
15-19
1
u
u
20-24
1
25-29
1
1
1
S
30-34
1
1
1
1
1
53
35-39
1
Q
1«
40-44
1
1
2
45-49
1
1
2
1
50-54
1
I
1
1
4
£
s
55-59
2
1
3
■5
60-64
1
1
2
"0
65-69
1
I
2
'"
70-74
oj
0
75-79
1
1
2
oj
K
80-84
85-89
90-94
95-100
1
1
2
f.
13
1
1
2
2
0
2
1
2
0
2
3
3
0
8
2
3
3
2
1
51
Constants calculated from Table 7.
First-year daughters' mean hatchability 2965=b .025445
First-year daughters' standard deviation . . . .2694± .017992
Second-year daughters' mean hatchability . . . .4484± .031130
Second-year daughters' standard deviation . . . .3296± .022013
Coeffi
cien
it 0
f CO
rrel
atic
n
2c
)96:
t.C
)85C
72
Referring to table 7, the mean hatchability of first-year daughters is. 2965, while
the second-year daughters of the same male have a mean of .4484. The difference
is .1519=b.0336, which is a significant difference. The second-year daughters
appear to be superior to the first-year daughters in hatching power. To draw any
conclusion, however, on such meager data would be more than hazardous. The
standard deviation does not differ significantly in the two groups of daughters.
30
A sensible positive correlation of .2996 ±.085972 appears between first-year
pullet daughters and second-year pullet daughters in hatchability. Table 7 thus
furnishes a very small amount of evidence that hatching power is transmitted
through the male, and that it is a more constant character than would be possible
were it independent of heredity.
Section 8. Relation betiveen the Fertility of the Sirens Dam and His Phenotypical
Fertilizing Ability.
As there is no direct measure of a sire's phenotypical fertilizing power, it is
necessary to resort to the indirect, which is the average fertility of his mates. The
degree of fertility in the sire's dam may be something of a guide to his inheritance.
The pertinent question at this point is: Is the degree of fertility of a cockerel's
mother a guide to his abiht}^ to fertiUze the eggs of his mates? If such be the case,
there should be a sensible positive correlation between sire's dam 's fertility and his
mates' fertility. In table 8 the dams of cockerels used throughout the eleven-year
period have been tabulated with the mates of these cockerels. The record of any
particular dam was used against each of the mates of her son. The total number
of mates was 647.
Table
8.
— Correla
ion
in
Fertility Between
Sires'
Dn
ns and Sires'
Mates.
Sires' Mates' Fbrtilitt, Per Cent. 1
1
o
1
o
T
1
o
C5
1
o
■*
i
■*
05
T
in
r
o
1
■<r
OS
CD
1
o
OS
1
1
o
OO
1
oa
o
o
o
05
f.
0-4
5-9
1
2
1
4
10-14
15-19
20-24
1
'
1
1
3
3
2
11
25-29
<u
30-34
1
35-39
1
1
1
1
9
13
^•
40-44
45-49
50-54
oi
55-59
1
1
1
2
1
12
18
C
60-64
1
2
1
1
1
2
1
3
7
5
4
15
43
65-69
70-74
75-79
1
1
1
1
1
1
2
1
1
1
7
2
2
4
2
4
2
1
4
3
2
4
2
7
14
2
15
21
6
49
51
15
80-84
2
2
3
3
2
1
5
5
5
10
6
9
11
23
48
135
85-89
1
1
1
2
3
18
26
90-94
1
2
1
1
4
2
1
3
2
5
10
15
46
93
95-100
1
1
2
3
3
1
1
6
5
4
9
14
14
26
99
189
I.
2
2
3
3
6
9
9
9
7
8
17
23
30
27
43
57
100
292
647
Constants calculated from Table 8.
Sires' dams' mean fertility 8157±. 004492
Sires' dams' standard deviation 1694±. 003176
Sires' mates' mean fertility 8531 ±.004587
Sires ' mates ' standard deviation 1730±. 003244
Coefficient of correlation - .1890± .025363
31
The mean fertility of the sires' mates is .0374±. 00642 greater than the mean
of the sires ' dams. This is a small but significant difference and indicates that more
attention was given to fertility from the female standpoint than from the male stand-
point. The standard deviation is almost identical for both groujis of females. A
negative coefficient of correlation of .1890 ±.0253(53 appears rather difficult to
explain. Jt certainly does indicate that the degree of fertility shown by sire's
mother is not an index to the degree of fertility that such a sire will exhibit in his
mates — his phenotypical fertilizing ability. This negative correlation may be
due to selection of females to be used as breeders with more regard to high fertility
in ancestry than is practised in selecting male breeders; or possibly males from the
very fertile ancestrj'' were mated to pullets that were lacking in fertility but other-
wise desirable.
Section 9. Relation between the Hatchability of the Sire's Dam and His Phenotypical
Hatching Ability.
The question of hatchability may be considered by the same methods used in
section 8 in studying fertility. The identical group of birds is again tabulated for
hatchabihty in table 9.
Table 9.-
- Correlation in Hatchability between Sires' Dam
9 and Sires
' Mates.
Sires' Mates' Hatchability,
Per Cent.
"*
05
7
■*
05
■*
en
Tf<
Ol
to
g
jj<
O!
•^
OS
•*
§
f.
o
o
o
ta
o
o
o
«
o
o
■o
lO
§
ia
o
K
o
00
S
o
Oi
s
0-4
1
1
1
1
4
5-9
10-14
15-19
1
1
2
1
5
20-24
■*^
25-29
1
1
1
2
1
2
0
10
o
30-34
1
1
1
4
3
4
2
4
3
1
1
1
1
27
^
35-39
,
1
1
1
1
1
5
>,
40-44
1
4
1
1
2
1
2
2
14
■g
45-49
1
1
1
3
3
2
3
2
4
2
4
3
2
1
3
35
'a
50-54
2
1
4
1
3
4
4
4
4
2
2
1
2
34
i
a
55-59
60-64
1
1
1
3
3
2
1
3
1
4
5
6
2
3
7
8
5
4
5
3
10
3
5
3
2
3
2
2
41
59
2
65-69
1
2
2
1
2
2
2 '
1
2
2
7
6
3
2
35
cc
70-74
1
1
1
1
2
4
6
4
3
4
7
6
7
8
9
7
6
2
8
87
75-79
1
2
1
2
1
1
3
4
2
10
9
7
11
3
4
2
3
66
80-84
2
1
2
1
1
1
1
8
7
5
6
6
6
7
10
9
1
74
85-89
1
1
2
1
3
3
3
3
5
7
7
6
2
3
1
1
49
90-94
1
1
1
3
2
6
8
4
5
5
1
37
95-100
1
2
2
2
1
2
1
6
10
4
8
8
9
6
3
65
f.
1
4
2
7
9
14
21
20
27
30
46
38
67
67
61
77
48
49
34
25
647
Constants calculated from Table 9.
Sires' dams' mean hatchabihty 6977±. 005115
Sires' dams' standard deviation 1929=t .003617
Sires' mates' mean hatchability 6488±. 005229
Sires' mates' standard deviation 1972±. 003698
Coefficient of correlation ....... .1579± .025856
32
The average hatching abilitv of the sires' dams is .6977rt.005115, while that of
the sires' mates is .6488±. 005229. There is a difference of .0489 ±.0073 14, which
means that the males used as breeders came from dams of higher hatching power
than was inherent in the pullets to which they were mated. The almost identical
standard deviation for the two groups points to a similar variability in hatching
power for the two.
The coefficient of correlation between the sires' mothers and their mates is .1579
rt .025856, a small but sensible correlation. Possibly this can be interpreted as
meaning that males tend to show a phenotypical hatching power comparable with
that of their dams. In selecting cockerels for breeders, hatching power of their
dams is something of a guide to their ability to contribute hatching power to the
eggs they fertilize. There is considerable probability that the male does influence
the hatching power of his mates' eggs.
Section 10. Relation of Sire's Average to his Daughters' Individual Fertility.
In considering the fertilizing and hatching power of males, it is necessary to use
some measure of their phenot3^pical character. This fact has been pointed out by
Pearl ('09) and, as he states, the average fertility and hatching power of hens
mated to a male may be used as his index. In table 10 the average fertility of
each sire from his different mates is tabulated against the fertility of each of his
daughters. This average figure for each sire is thus inserted a number of times to
correspond with the number of his daughters that were used as breeders.
Table 10.-
— Correlation
in Fertility Between Sires' Mates and Sires' Daughters.
Sires' Daughters' Fertility, Per Cent.
j
-•1
p
o
"5
o
1
1
S i
04 CO
OS
7
n
o
OS
T
1
o
to
J*
1
o
05
1
■<»i
1
o
1
1
§
g
f.
0-4
6-9
10-14
15-19
20-24
25-29
O
t-i
30-34
85-39
Ph
40-44
45-49
1
1
50-54
1
55-59
1
1
2
1
2
2
4
13
S
60-64
1
1
00
65-69
70-74
1
1
1
1
1
1
1
2
1
4
1
6
1
3
11
17
19
75-79
1
1
1
1
1
2
4
1
5
2
14
33
80-84
1
1
1
2
3
4
3
5
4
11
11
46
85-89
1
2
1
1
5
2
8
4
8
22
54
90-94
1
1
3
1
1
5
4
3
3
10
12
34
78
95-100
1
1
1
2
2
1
1
2
4
13
8
5
13
17
21
58
150
f.
2
2
3
1
6
6
5
6
5 13
19
25
17
35
46
64
157
412
33
Constants calculated from Table 10.
Sires' fertility mean ....
Sires' fertility standard deviation
Sires' daughters' mean fertility .
Sires' daughters' standard deviation
Coefficient of correlation
.8761 ±.008522
.1060±. 002491
.84 16 ±.005599
.1685 ±.003959
.0244±. 033211
A difference, amounting to .0345 ±.00661 4, will be observed between the sires'
mean fertility and their daughters' mean fertility. This significant difference is
easily explained if the same factors are operating to aft'ect fertility of males and
females. A wider range of variability in the daughters as compared with their
sires, measured by the standard deviation, seems to indicate that there is little or
no constancy in fertilit.y between father and daughter.
No sensible correlation in fertility exists between sire and daughters as table
10 show^s. In the face of this fact, there is no e\adence that factors for fertifity
are transmitted from sire to daughter. In other words, fertility does not seem
to be an inherited trait that is transmitted from parent to offspring, as has already
been shown in both tables 4 and 10.
Section 11. Relation of Sire's Average to Daughters' Individvxil Hatchability.
The same group of birds used in table 10 is correlated in table 11 to study the
relationship between sire and daughters in hatching power.
Table 11. — Correlation in Hatchability between Sires' Mates and Sires' Daughters.
Sires' Daughters' Hatchability, Per Cent.
o
1
O
ea
1
O "5
o
CO
1
S3
T
o
s
05
U5
■*
1
O
to
at
to
1
s
•*
1
o
oo
oo
1
5
o
OS
o
o
1
f.
0-4
5-9
10-14
15-19
20-24
c
25-29
1
2
3
6
30-34
S,
35-39
>>
40-44
1
1
1
2
5
!3
J3
"cS
45-49
50-54
1
4
4
2
1
3
2
1
5
4
'8
2
2
4
1
1
2
2
2
2
12
41
K
55-59
1
1
1
1
6
10
1
5
, 3
3
4
3
1
40
a
60-64
65-69
2
2
3
1
2
3
1
2
3
1
2
4
2
5
2
1
3
6
5
6
2
8
5
6
9
4
4
5
3
2
1
3
2
62
48
M
70-74
1
2
3
1
2
2
5
5
10
10
8
10
3
8
8
4
82
75-79
1
1
2
1
2
2
1
2
5
4
2
3
26
80-84
1
1
6
1
7
10
4
11
8
4
4
3
60
85-89
1
2
2
3
2
4
1
15
90-94
1
2
5
3
2
2
2
1
18
95-100
4
4
6
10
14
12
16
25
23
41
57
33
34
29
17
f.
c — ■ ..
1
1
46
39
412
34
Constants calculated from Table 11.
Sires' hatchability mean
Sires' hatchability standard deviation
Sires' daughters' mean hatchability .
Sires' daughters' standard deviation
Coefficient of correlation
.6824±. 004084
.1229 ±.002888
,6753±. 006217
.1868±. 004396
.2268zt .031523
The mean hatchabihty of the sires is almost identical with that of the daughters.
This is in striking contrast to the mean of dams and daughters given in table 5
where the figures are .7064±. 003891 and .5091 ±.003740, respectively. Such
evidence might be interpreted as showing that a closer relationship exists between
sires and daughters than between dams and daughters in hatching power. Such a
relationship is probably due entirely to the somewhat dissimilar methods for meas-
uring hatching power in sire and dam. The range of variabihty is greater in daugh-
ters than in sires evidently because of the variable nature of the males mated to
these daughters.
The coefficient of correlation between sires and daughters is .2268±. 031523.
Comparing this factor with the factor calculated from table 5 where mothers and
daughters are concerned, the two are found to be of almost identical magnitude
when their probable errors are considered. Table 11 furnishes convincing evi-
dence of the heritability of hatching power. In this instance, hatching power of
sires is carried on in their daughters. Table 11 further points to the necessity of
using tested males in developing a flock carrying uniformly high hatching power.
Section 12. Relation of Sire's Dam to his Daughters' Fertility.
In section 8 the relation between sire's dam and his phenotypical fertiUzing
ability has been considered. A negative relationship was found to exist in that
case. The present section is an attempt to discover if the sire transmits to his
daughters a degree of fertility similar to that of his dam, so that when these
daughters are mated with other males their probable fertility may be forecasted.
In table 12, 748 pullet fertility records are tabulated with the fertiUty records
of their sire's mother as a pullet.
Table
12.-
—Correlation in Fertility between Sires' Darna
a7id Sires
35
' Daughters.
SiREs' Daughters' Fertility, Pek Cent.
1
o
T
o
U5
CM
1
o
1
o
1
f
o
05
T
3S
1
o
J.
3
to
<o
■J
o
1
1
o
at
U5
1
o
§
f.
0-4
5-9
1
1
1
1
4
10-14
15-19
20-24
1
1
1
1
1
2
1
3
11
a
e
u
25-29
30-34
35-39
1
2
1
3
7
40-44
't,
45-49
?
50-54
C3
55-59
1
2
1
1
1
2
4
5
17
60-64
2
1
4
1
1
1
1
1
2
5
4
4
2
1
3
17
50
65-69
1
1
2
3
2
2
2
1
4
4
6
5
6
28
67
70-74
5
1
2
4
2
2
2
2
1
7
4
2
7
3
7
21
72
75-79
1
1
1
1
1
3
1
9
18
80-84
15
2
1
2
1
3
1
1
3
2
4
6
6
9
5
9
27
52
149
85-89
1
1
1
1
1
1
1
1
3
3
4
4
7
29
90-94
14
1
1
1
3
2
3
6
3
4
5
7
5
10
16
14
29
124
95-100
12
1
3
5
1
3
4
6
5
4
4
9
7
12
8
13
27
19
57
200
f.
52
5
3
6
14
8
8
14
13
14
22
12
23
33
40
39
50
73
88
231
748
Constants calculated from Table 12.
Sires' Dams' Mean Fertility
Sires' Dams' Standard De\'iation
Sires' Daughters' Mean Fertility
Sires' Daughters' Standard Deviation
Coefficient of Correlation .
.8183±. 003009
.1585±. 002764
.7364zb. 007108
.2882 ±.005026
.0501 ±024599
The mean fertility of the dams of the males used in this study is .0819±.008112
greater than the mean for the daughters of these males. The males used, there-
fore, came from dams of high fertility but the daughters of these males failed to
measure up to such a standard. The standard deviation of the daughters is almost
twice as great as for the sires' dams, showing that the daughters are a highly vari-
able lot. The coefficient of correlation is negative but insignificant because of the
magnitude of its probable error. The conclusion seems justified, therefore, that
the degree of fertility of a sire's dam is no index to the degree of fertility that his
daughters will exhibit.
Section 13. Relation of Sire's Dam to his Daughters' Hatchability.
If the hatching power of a sire's dam is something of an index to his probable
inheritance of factors affecting hatchability, such relationship will appear when the
hatchability records of the daughters are tabulated with the records from the
sires' dams. Table 13 is thus made up of the same birds used in table 12.
36
Table 13. — Correlation in Hatchability hetiveen Sires' Dams and Sires' Daughters.
Sires'
Daughters
Hatchability, Per Cent.
•*
OS
^r
OJ
■^
OJ
T
T
■^
7
OS
CO
■<»<
en
S
OS
S
o
o
f.
1
o
■o
o
»o
o
>o
o
Ui
o
o
>o
g
§
o
J2
o
o
o
o
0-4
5-9
10-14
15-19
1
1
1
2
1
1
7
20-24
a
25-29
2
1
1
1
1
1
1
1
2
11
6
30-34
3
3
1
2
2
2
2
1
2
1
4
1
1
3
1
1
1
31
(S
35-39
1
1
1
1
1
5
>,
40^4
6
1
1
2
1
1
2
2
1
2
2
3
3
2
29
1'
1
45-49
50-54
16
2
2
4
2
1
2
2
1
1
3
1
1
4
2
2
2
4
1
3
1
1
1
1
I
2
3
2
2
1
2
43
30
ffi
§
55-59
6
1
2
1
2
1
1
3
1
1
8
2
3
2
2
3
3
42
Q
60-64
12
1
1
3
1
1
3
3
3
10
3
6
11
6
5
3
72
£
65-69
1
1
1
2
1
4
10
c»
70-74
13
2
1
1
4
3
5
6
7
2
4
8
10
12
7
11
4
12
8
2
122
75-79
12
1
3
2
3
1
3
1
4
3
6
6
7
9
8
2
4
3
78
80-84
18
2
2
2
2
5
5
3
5
3
5
14
8
9
7
7
3
6
9
lis
85-89
7
1
2
1
1
3
4
4
2
1
1
2
29
90-94
2
3
1
3
2
1
1
1
2
3
2
4
4
4
1
4
38
95-100
26
1
1
4
2
4
1
1
4
3
5
5
2
5
9
3
1
7
2
86
f.
126
12
12
16
19
21
28
27
23
23
32
35
66
52
45
66
44
36
32
33
748
Constants calculated from Table 13.
Sires' Dams' Mean Hatchability
Sires' Dams' Standard Deviation
Sires' Daughters' Mean Hatchability
Sires' Daughters' Standard Deviation
Coefficient of Correlation .
.70 19 ±,004664
.1891 ±.003298
.5096 ±.007588
.3077 ±005366
.0588 ±.024576
The mean hatching power of the hens whose sons were used for breeding was
.7019. The daughters of this group of males averaged only .5096 of fertile eggs
hatched. This difference in the means amounts to .1923 ±.008906 and is a much
more striking difference than was observed between the same group of females in
fertility. The standard deviation of the two groups agrees with that found for
fertility in table 12. Again the daughters of the males show almost double the range
in variability of their sires' dams.
The coefficient of correlation is here positive, but of no significance since it is
a httle more than twice its probable error. The lack of correlation between sire's
dam and sire's daughters in hatchability can scarcely be interpreted to show that
hatchability is not governed by factors transmitted from sire to daughter. The
hatching power of a cockerel's dam is only the phenotypical manifestation of her
ability and may be affected by her mate as well as by numerous environmental
factors. She furnishes, moreover, but a part of the heritage of her son. If several
factors governing hatchability are transmitted equally by males and females and
if both parents have an influence on the hatching power of eggs laid and fertilized,
respectively, this apparent independence of hatching power in inheritance will be
explained.
37
It fertility be governed by genes transmitted in Mendelian fashion and without
-x-linkage, this fact should be brought out by correlating the sire's record with his
Ill's record. The only measure is the fertility record of the eggs laid by females
Kited to siu'h males. If it were possible to compare males by a system of mating
1 the same group of females, the variable factors could be refiuced to the male
lie alone., 8uch a system seems impossible to attain because of nimierovis factors
lo well understood to require mention.
Section 14- Relation of Sire and Son in Fertility.
Table 14. — Correlation in Fertility between Sires and Sons.
Sons'
FERTir.iTY, Tkr Cknt.
1
O
Ca
1
1
o
1
o
1
•>»<
7
T
1
o
s
1
to
1
1
o
s
1
00
1
g
T
f.
0-4
5-9
10-14
15-19
20-24
25-29
30-34
•«J
<s
35-39
1
2
3
40-44
1
1
45-49
50-54
1
2
3
^
55-59
1
1
2
HI
60-64
65-69
70-74
75-79
80-84
85-89
90-94
1
1
1
1
1
1
1
1
1
1
2
1
1
2
2
1
1
1
4
3
2
2
1
1
3
2
3
3
2
3
2
3
1
3
4
4
6
5
9
1
6
14
19
15
25
95-100
1
3
2
1
2
3
3
3
6
6
12
32
74
f.
1
1
4
4
3
1
1
4
7
9
8
18
17
25
67
170
Constants calculated from Table 14-
Sires' Mean Fertilitv . . . . ' 8682 ±.007041
Sires' Standard Deviation 136 1 ±.004979
Sons' Mean Fertilitv 8441 ±.008660
Sons' Standard Deviation 1674±. 006124
Coefficient of Correlation 0685±. 051486
In table 14 each pullet mate of a sire is paired with a pullet mate of his son. The
number of pairs concerned is 170 and the number of sires included is about the same
as the number of sons included. The mean fertility of the sires and their sons is
not significantly different, and the range in variability of sires and sons, as measured
by the standard de\aation, is about the same. The coefficient of correlation is very
small and its probable error renders it negligible. The only conclusion that may be
drawn from this small amount of data is that either the fertility record of a male 's
mates is not a reliable index to his inherent fertilizing ability, or else degree of fer-
tility is not transmitted from sire to son.
In the next section the relation of hatchability of sire and son will be considered
for the same birds that were used in studying fertility.
38
Section 15. Relation of Sire and Son in Hatchability.
Table 15. — Correlation in Hatchability Between Sires and Sons.
Sons' Hatchability, Per Cent.
■w
o=
T}<
■?
05
7
a>
"T
s
Q>
■*
05
to
S
K
^
OS
00
s
o
o
f.
o
o
U5
o
"5
IM
S
J?
o
^is
o
CD
S
R
K
§
00
§■
0-4
5-9
10-14
15-19
1
1
20-24
1
1
25-29
1
1
1
1
1
5
30-34
1
1
2
U
35-39
1
1
1
1
1
1
1
7
P^
40-44
1
3
1
1
2
1
1
10
iS
45-49
2
1
2
1
2
1
1
10
03
50-64
1
1
1
2
1
1
1
1
9
55-59
1
1
1
1
1
4
4
1
2
16
2
60-64
1
3
1
1
3
2
1
1
14
i»
65-69
70-74
75-79
80-84
85-89
90-94
1
1
1
1
1
1
2
1
1
1
1
2
1
1
2
1
1
2
1
1
3
3
3
1
1
3
1
2
2
1
1
3
2
4
1
4
1
3
2
1
2
1
1
1
1
1
1
1
1
13
10
18
13
14
12
95-100
2
1
1
1
1
1
2
2
1
1
1
15
f.
4
3
2
5
6
5
9
15
8
22
19
19
21
8
11
5
8
170
Constants calculated from Table 15.
Sires' Mean Hatchability
Sires' Standard Deviation
.6738±. 010274
.1986 ±.007265
Sons' Mean Hatchability
Sons' Standard Deviation
.64 18 ±.009720
.1879 ±.006873
Coeffic
ien
tof
Co
rre]
atic
n
.07
55 zl
= .0
514
40
Reference to table 15 shows that the mean degree of hatchability is almost the
same in sires and sons. The two groups are also closely similar in standard devia-
tion. There is no sensible correlation between father and son in hatchability. The
degree of correlation here is practically the same as that observed for fertility in
table 14. If we are using the correct measure for a male's hatchability, there is
no evidence in these data to show that hatching power is transmitted from sire to
son.
Section 16. Mendelian Interpretation of the Inheritance of Fertility and Hatchability.
Before entering upon a discussion of the "possibihties of Mendelian inheritance of
factors governing fertility and hatchability, it would seem desirable to present
the mean records in the flock from year to year. These means are given below in
table 16 along with the number of birds tested each year.
39
T VHLE 16. -
— Mean
Fertility and Hatchahility Records from the Mass
%chusetts Agri-
II cultural Experiment Station Flock.
Yeau.
Average Average
Number
Fertility. Hatchahility.
of Birds.
1913
.7562±. 016855
.5910±. 016578
73
1914 .
.8300 ±.01 5294
.5793±. 016514
67
1915 .
.8308±, 012692
.56 13 ±0130 15
118
1916 .
.8834±, 010973
.6469 ±.01 5942
62
1917 .
.9158±. 009776
.6217±. 014709
78
1918 .
.8821 ±.009917
.6502 ±.013599
89
1919 .
.8882±. 014611
.6941±. 014602
56
1920 .
.8647±. 014243
.6861 ±.01 7473
51
1921 .
.9107±. 012241
.748;3±. 014129
59
1922 .
.8746±. 010910
.7449±.011125
89
1923 .
.7749±.011944
.7051±. 011399
144
The fertiUty mean has fluctuated appreciably from year to year and has not
increased during the past six j^ears. The low fertilitj^ of 1923 can be attributed to
no other cause than adverse weather conditions throughout the \\'inter and spring
months. The majority of the males seem to have suffered from more or less frosting
of combs and wattles during the winter of 1922-23. The basis of selecting breed-
ing males for 1923 was not voluntarily changed from that of previous years. The
general deduction must therefore be made, as Pearl ('09) has done, that fertilit}'
is dependent largely upon environmental factors and that it is not an inherent char-
acteristic that is transmitted in inheritance.
Table 16 indicates an increase of .1141 ±.0206 in mean hatchability from 1913 to
1923. This increase is mathematically significant. There has been a gradual
upward trend in mean hatching power since 1915. This increase has accompanied
the use of breeding pullets and breeding cockerels from mothers showing good
hatching power. The .04 drop in hatchability in 1923 is within the range of prob-
abilit}" and need not be considered.
Relatiox of Male to the Hatching Power of his Mates' Eggs.
Unmistakable evidence is available to show that the male contributes to the
hatching power of his mates ' eggs. For want of any more suitable term we have
used "male's phenotypical hatching power" to express the male's part. In table
9 a positive correlation coefficient of .1579±. 025856 was observed between the
sire's dam, and his phenotypical hatching power. A sensible correlation could not
exist unless the male contributes to the hatching power of his mates' eggs.
The most conclusive evidence that the male influences the hatching power of his
mates ' eggs lies in the fact that the same hen shows different hatching power when
mated to different males in successive years or even in the same year. Such data
should be placed beside data showing the degree of constancy of hens in hatchability
when mated to the same male on successive years. No data are available on the
last-named question from our flock, although table 3 brings out a degree of cor-
relation between first and second year hatchability in hens, amounting to .4346±
.034409. The correlation should be much greater if the male did not play a part.
In section 5 a sensible correlation between mothers and daughters was discovered.
Reference to the constants calculated from table 5 shows that the hatching power
of a hen is an uncertain guide to the probable hatching power of her daughters.
The relative magnitude of the standard deviation of dams and daughters indicates
that the phenotypical hatching power of a hen is an uncertain index of her true
genetic constitution. This fact would seem to indicate that the male obscures the
true genotype of the hen.
Data from the flock of the Massachusetts Agricultural Experiment Station on
the constancy of hatching power in males is very limited. In table 17 a comparison
is made between the first-j^ear hatching power and second-year hatching power of
40
15 males. The figure used for each male represents the mean for all of his mates.
These males were used on the following years: — 2 in 1913 and 1914, 4 in 1914 and
1915, 2 in 1915 and 1916, 2 in 1916 and 1917, 1 in 1917 and 1918, 2 in 1919 and
1920, 2 in 1922 and 1923.
Table 17
. — Mean Hatchabiiity of Males.
Male No
-
First Year.
Second Year.
A323 . .
57.00
55.80
A324 .
59.19
57.93
68. .
38.67
52.17
228.
59.50
67.75
619 .
59.00
49.75
A271 .
70.71
67.40
A274 .
50.23
63.50
3617 .
53.93
64.40
5470 .
62.00
70.75
5581 .
59.29
65.00
8528 .
71.83
72.62
B2776 . .
67.00
75.00
B2828 .
64.13
85.50
C901 .
76.20
65.00
C938 . .
70.57
74.44
Mean first year, .6128±.016043; Mean second year, .6580±.015825; Difference in
means, .0452±.0225.
Although the data are meager in table 17, we can give it no other interpretation
than as indicating that the male does partly control the hatching power of his mate
through dominant factors.
The mean hatchabiiity for the fifteen males during the first year is .6128±
.016043, for the second year .6580±. 015825. There is a difference of .0452±.0225.
This difference is just double its probable error and can therefore be of no conse-
quence. The point we wish to emphasize in table 17 is the striking constancy in
phenotj'pical hatching power of the same male, even when mated to different hens
on two successive years. Such a degree of constancy was not found to exist in
hens, as table 3 shows. The mean pullet-year hatching power of the hens was
.5678±.011313. The mean second-year hatching record of the same hens was
.4791 ±.012963. The standard deviation is nearly three times as great for the hens
as for the males. The difference in the mean hatching power for the same hens on
two successive years is .0S87±.0172, which is significant. The genetic interpreta-
tion given below will serve to elucidate several apparent compUcations.
Genetic Factors Concerned ^
One dominant gene seems to be concerned in the production of high hatchabiiity.
We use the symbol H to designate this gene. There is no sex linkage and all results
obtained are to be expected in a simple mono-hybrid ratio. With this hypothesis,
three possible genotypes of males and females exist, namely, HH, Hh, and hh in-
dividuals. The genotj^pe is obscured in most cases for both males and females.
Such being the case, only the breeding test can be used as a guide for matings.
Hatching records on 886 females studied in this report show that these birds
divide themselves into three general classes or phenotypes: — (1) Those showing
hatchabiiity of 85 per cent or above, we call high. (2) Those with a hatchabiiity
of 55 to 84 per cent, we call medium. (3) Those below 55 per cent, we call low.
Since factor H has a cumulative effect, the range for the medium class is twice as
great as for the high class. The minimum for the low class has not yet been de-
terrnined. Below are summarized the males' pheno typical and genotypical classes:
> A detailed report on the genetics of hatchabiiity will appear in another pubhcation.
41
Males' Phenotypiml Chnrncl-cr.
HH males on IIII hens give all high hatchahility.
HH males on Hh hens give all mediimi hatchability.
HH males on hh hens give all medium hat('hal)ility.
Hh males on HH hens give all high hatchahility.
Hh males on Hh hens give all medium hatchability.
Hh males on hh hens give all low hatchability.
hh males on HH hens give all medium hatchability.
hh males on Hh hens give all low hatchability.
hh males on hh hens give all low hatchability.
Males' Oenotypical Character.
HH males on HH hens give all HH daughters.
HH males on Hh hens give 50% HH and 50% Hh daughters.
HH males on hh hens give all Hh daughters.
Hh males on HH hens give 50% HH and 50% Hh daughters.
Hh males on Hh hens give 25% HH, 50% Hli, and 25% hh daughters.
Hh males on hh hens give 50% Hh and 50% hh daughters,
hh males on HH hens give all Hh daughters.
|j hh males on Hh hens give 50% Hh and 50% hh daughters,
hh males on hh hens give all hh daughters.
Both parents must carry the H factor in order to be phenotypically good hatchers.
Hens cannot rank in the first class unless they carry the gene H in homoz^ygous
cnndition and are mated to H-bearing males. These observations indicate a
cumulative value for the factor H and show wh}- the male by failure to contribute
at least one-half H-bearing sperm ranks a genotypically high hen as a medium
hatcher. Furthermore, both HH and Hh males probably give about the same
hatching record from HH hens. The progeny test alone can give a clue to the
genetic composition of males if pullets of unknown formuliB are used as breeders.
Selection for high and low hatchability did not give results in two generations
according to Dunn ('2.3). The probable explanation is that he used in his low line
genotypicalh'- high (HH) hens that gave medium hatching records because they
were mated to hh males. If such were the case, no appreciable separation could
take place in but two generations. There may also have been a lack of HH or Hh
males in his high line. Selection for high hatchability with the female as a guide
and using cockerels from hens that hatched well has been a slow but progressive
process in our flock, as already shown in table 17. In table 9, the mean hatcha-
bility of the dams of the males used for breeders is about 70 per cent. This would
indicate thti^, on the average, the breeding males came from Hh hens. Thus, only
in the later years of the period could any considerable percentage of males have
been of the formula Hh. A stud}' of earlier records shows that practically all the
males must have been of hh composition, because they came from medium or low-
hatching dams.
SUMMARY.
1. No correlation was found between fertility and hatchability in 758 pullets.
2. Fertility in the hen behaves as an individual characteristic with a fair degree
of constancj'^ from j^ear to year.
3;' Fertility does not appear to be transmitted from mother to daughter.
4, Hatching power is more constant from year to year in the same hen than is
fertility.
5. Hatching power gives evidence of being transmitted from mother to daughter.
&r Fertility in the male behaves as an individual characteristic and probably
with some constancy in the same individual from year to year.
7. The fertility record of a hen is no index to the fertilizing abilit}' of her sons."
8. Fertihty does not appear to be transmitted from sire to daughter-;
9. Hatchability is more constant from year to year in the same male than is
fertility.
10. Fertility does not appear to be transmitted from sire to son.-
11. The hatching power of a male cannot be judged by his dam's hatching
record.
42
12. Hatching power gives evidence of being transmitted from sire to daughter.
13. Insufficient data are available on the transmission of hatching power from
sire to son. ^|
14. Fertility is evidently not an inherited characteristic. ^0
15. Hatchability is evidently an inherited trait. High hatchability is dependent
in inheritance upon one dominant gene. Both male and female parent govern the
hatching record, thus obscuring the true genetic composition of either parent.
16. Genetically pure hens for high hatchability maj^ be discovered through their
own hatching record. Genetically pure males for high hatchability can be dis-
tinguished from males heterozygous for the factor only by the progeny test com-
bined with mating tests. Both the mating and the progeny test should be used
for choosing males to improve the flock in hatchability.
REFERENCES.
1. Atwood, H., West Virginia Experiment Station Bulletin 124, 1909.
2. Dunn, L. C., Address before the Annual Meeting of Instructors and Invest:
gators in Poultry Husbandrj', 1923.
3. Lamson, G. H. and Card, L. E., Connecticut (Storrs) Experiment Station Bui
letin 105, 1920.
4. Pearl, Raymond, Maine Experiment Station Bulletin 168, 1909.
5. Pearson, K., On the Laws of Inheritance in Man. I. Inheritance of Physical
Characters V. II, 1903.
i
LIMAffY OF THE
lassactinsptts A^rlCQimnl rollege,
AMHEHST, i)^^&^.
Massachusetts
Agricultural Experiment Station
TECHNICAL BULLETIN No. 7 MARCH, 1926
BROODINESS
IN RELATION TO FECUNDITY
IN THE DOMESTIC FOWL
By F. A. HAYS and RUBY SANBORN
This bulletin is the seventh in the series of bulletins reporting the in-
vestigations of the Massachusetts Agricultural Experiment Station on
heredity in the Rhode Island Red breed of poultry; and the second giving
report of the study on broodiness in the same breed. In addition there
have been published at various times scientific papers presenting the results
of certain more or less minor phases of this study.
Expressed in terms of change in the character of the breeding flock,
the data show that the percentage of broody birds has decreased from 90
in the foundation flock of 1912 to 27 in 1923, the last year reported in
this publication. Associated with this decrease in broodiness, the average
annual egg production has increased from 114 to 200 eggs. The data
sho^v, however, that decrease in broodiness is but one of many factors
which have contributed to increased production.
Requests for bulletins should be addressed to the
AGRICULTURAL EXPERIMENT STATION
AMHERST, MASS.
BKOODIXESS IN RELATION TO FK( INDITV IN THE
DOMESTIC FOWL
Bv F. A. IIAVS and RUDY SANBORN
XATrKK UK CiiAiiA(.Ti:u Bi;iMi Sti i)n;i)
Broodiiics.s is tlic tendency of female birds tu incubate or atteni])t to incii-
liate eggs. Tlie broody hen stays on the nest, clucivs, rulHes featliers wl»en
disturbed, etc. It is a recurring cyclical trait in birds and should l)e consid-
ered as a normal phase of their reproductive process. It has no iiomologue
in mammals since they reproduce viviparously (developed young). In rep-
tiles, which are closely related to birds, we have oviparous reproduction, but
the eggs are hatched without the attention of the mother.
.VU breeds of domestic chickens exhibit broodiness to some extent. The
Asiatic or meat breeds are ail intensely broody; tlie American l)reeds all
exhibit the trait to a considerable extent; and the Mediterranean lireeds,
although said to be non-broody, always give some broody females.
There thus appear to be widely different degrees of broodiness. There have
iieen birds in the Massachusetts Station flock that first showed broodiness in
November of their pullet year and continued to exhilut its cyclical recurrence
to the extent of ten or twelve times during the first laying year. Contrasted
with tins is hen C 960 — non-broody during pullet year, twice broody the second
year, and non-broody the third year. Also hen C 1347 — non-broody as a
pullet, broody once for 23 days her second year, and non-broody her third
year. Hen C 476.3 was brood}' once as a pullet for 17 days and non-broody
her second year. On the other hand, we now have two hens (B C-ITC i.nd
B8797) that have completed four annual records without going broody. In
general, three measures of broodiness may be used: nan)ely, (1) the number
of broody periods per year, (2) mean length of each broody period, and (3)
total days of non-productiveness associated with broody periods. In all cases
the length of a broody period has been taken as the period between last egg
previous to going broody and first egg following "recovery."
Effect of Metholl of Hcmdliufi
With the domestic fowl efforts are made to check the manifestation of
Viroodiness so that the hen may begin laying again. Modern j)rai'ticc is to
coop such hens in slat-bottom coops, making nesting almost impossiiile. After
four to six days of such confinement, the bird may ordinarily be returned to
the flock without resuming nesting. Such hens .show wide diversity in length
of time before resuming laying.
Trapne.sting and regular removal of all eggs from the nests seem to "dis-
■ durage" the on.sct of broodiness. Punnett reports two cases of hens from ;.
liroody-frce race that were themselves non-broody for two years, later actually
incubating and hatching eggs. This particular phase of the prol)lem needs
further elucidation.
Broodiness thus apjjcars to be a normal ]ihase of the re])rod!icfion of
domestic chickens. Its occurrence seems to depend upon environmental and
])hysiological stimuli as will l)e pointed out later.
56 TECHNICAL BULLETIN 7
Work Already Doxe.
By Other Investufntors.
Bateson (1902) and Hurst (1905) both present data (in crosses between
broody and non-broody races, indicating tbat l)roodiness is a dominant cbar-
HCter. No furtber information was obtained at tliat time.
Punnett and Bailey (1920) report some results using Black Langslians,
Brown Legborns and Gold-pencilled Haiiilnirgs. Results:
Langsban $ x Leghorn $ gave all F,* pullets broody. Of the F.* genera-
tion 16 pullets were retained, 8 of which went liroody as pullets. Punnett
states tluit if the Langshans were of composition A ABB and Leghorns aabb,
F, should give 9 broody to 7 non-broody, a close approximation to actual ratio.
The reciprocal cross, Legliorn $ x Langsban $ gave all broody in F„ l)ut in
F. there were but 19 broody to 47 non-broody. Most of these birds were re-
tained but one year. A few that were kept the second year added more
broodiness so that the ratio is not 9 to 7, probably because of delayed appear-
ance of broodiness.
In the Hamburg-Langshan cross, the Fi hens were either non-l)roody or
showed very little broodiness. Of 38 Fo pullets, 4 were broody, 34 non-broody.
Tliese results suggest a third jfactor, N, which inhibits. Fj birds would l)e
NnAaCc, but factor N did not inhibit in all cases. The Fo ratio gave 4
broody to 34 non-broody. Punnett states that his results are far from con-
clusive as to the true nature of the l)roody trait.
Pearl (1914) found much less broodiness in Barred Plymouth Rocks tiian
exists in Reds. His method of measuring the intensity of broodiness was l>y
the length of non-productive period. Otlier known factors, sucli as winter
pause and molt make such a measure subject to error.
Work Done by the Massachusetts Station.
Goodale began the study of this trait in 1912. From that time up to 1921,
when he severed his connections with the Station, very satisfactory ])rogress
was made in eliminating the tendency from the egg-laying strain of Rhode
Island Rcds.t
Recent M'ork at the Massachusetts Agricultural Experiment Station.
In the fall of 1922 the writer took up this project using the same general
plan with some modifications. The non-broody strain has been carried along
witii the intense broody strain and not as a part of tiie general flock in so
far as the matings are concerned. The non-broody birds are now being carried
along through the second and third laying years to definitely test their l)e-
havior with regard to broodiness. Similarly, the breeding males are being
carried over and tested for genetic composition. In a paper entitled, "In-
breeding the Rhode Island Red Fowl with Si)ecial Reference to Winter Egg
I'roduction," the broody trait has been shown to confirm Goodale's AC theory
which suggests that broodiness is due to the presence of two dominant, auto-
somal, complementary genes, A and C. Both must be jiresent to i)roduce
broodiness, but either may be carried alone by non-broody birds.
* Fi and F2 refer to generations one and two.
t See Jla-ss. Agri. Expt. Sta. Bull^. 199 and 'ill.
r>K()()l)lNi:SS AM) F1XIN1>1 1 ■^ IN* FOWL 57
l'i)(Kiui:ss TO Datk.
O'eiierol Proj/resx from Year to Year.
Taiii.k 1. Mean Degree of IJrooclincss by Years.
Average
Year
Birds liroody.
number of
Total nunil)cr
Annual
Hatclied
j)er lent
broody periods
per broody hen
of birds
available*
Production
llU-i
89.60
4.4
125
114
1913
91.03
5.4
78
124
1911.
85.95
4.3
121
103
1915
89.25
4.3
428
122
1916
86.31
3.5
431
134
1917
48.84
2.7
432
166
1918
61.40
2.9
215
169
1919
No annual records
1920
46.03
2.9
126
200
1921
44.56
2.7
285
200
1922
28.91
1.9
399
200
1923
27.35
1.9
340
189
* This coluinn iiuliules all Rhode Island Reds except intense bi-oodies and inl)reds.
It will be observed that the percentage of broody birds has been reduced
from 90 in 1912 to 27 in 1923. Great significance should also be attached to
the fact that the mean annual egg yield has increased from 114 to 200 in the
same period. In the 1912 flock each broody hen lost 75 days in broodiness her
first year, while in the 1922 flock each broody hen lost but 29 days. The
assumption seems justified, therefore, that progress in eliminating broodiness
has been two-fold: namely, reduction in the percentage of broody birds, and
reduction in the mean degree of broodiness.
The average number of days spent in broodiness for the 112 broody birds
in the 1912 flock is 74.8. For the 71 broody birds in the 1913 flock the figure
is 78 8 days. In the 1922 flock, made \\\) of all birds except those bred for
intense broodiness, there were 112 birds that were broody, with a mean of
28.71 days .sjient in broodiness. In the 1922 flock there were 33 birds l)red
for intense broodiness. These Itirds averaged 42.94 days l)roody for the ]iullet
year.
Speci/ic liesulttt. *
The non-broody strain has been strengthened during the past two years
by the retention of non-broody hens up to five years old. Such hens have
been used as breeders each season so that their genetic character for broodi-
ness may be confirmed by the progeny test. Aged breeding males have also
lieen retained for similar purposes.
An intense broody strain has been carried on from year to year. Females
selected to peqietuate this strain iiave been selected with a view of combining
the maxinutm number of broody ])eriods with desirable traits from the stand-
])oint of annual fecundity. This intense broody strain will eventually diflfer
from the non-liroody strain only in possessing the broody trait. It is pw.ssible
.38 TECHNICAL Bl'LLETIX 7
in this manner to measure directly the effect of hroodiness on fecundity. This
intense broody strain differs from the foundation birds more in the distribu-
tion of broody periods throughout the laying year tlian in the ninnl)cr of
broody periods.
Complete records of hroodiness are also maintained on every female of tlie
experimental flock to augment data collected in the broody experiment.
ExD TO Be Attained
A flock of poultry breeding true for hroodiness and non-broodiness.
Scope of This Report.
In this bulletin consideration is given to the actual relationship lietwecn
pullet-year egg production and the broody trait as manifested during tiie fir.st
laying year. Coefficients of correlation have been calculated as follows:
Between hroodiness and rate.
December rate — Sections 1, 2, 3, 4, 5, Iti, 17.
Winter rate— Sections 6, 7, 8, 9, 10, 18, 19.
Annual rate— Sections 11, 12, 13, 14, 1.5, 20, 21.
Between times broody and length of broody periods.
Section 22.
Between winter rate and annual rate.
Section 23.
Between winter rate and annual eyy yield.
Section 27.
Between annual rate and ^annual eya yield.
Section 28.
Between hroodiness and eyy yield.
Winter production — Sections 24, 25, 26.
Annual production— Sections 29, 30, 31, 32, 33, 34, 3.5.
COEFFICIEXT OF CoRREI.ATIOX.
The coefficient of correlation furnishes a concrete measure of the tendency
of two characteristics to nio\e together, to move in opposite directions, or to
behave independently. In tliis particular study the characteristics studied
both belong to the same individual fowl. Either a significant positive or
negative correlation coefficient is useful to the breeder as a guide, and the
itiagnitude of the coefficient shows him the relative amount of dependence
l)etween the traits or characters considered. The value of a coefficient of
correlation from the biological standpoint depends upon its ai)Solute magni-
tude and upon its relation to its probable error. A coefficient at least tliree
times as great as its probable error is generally considered significant, e\en
though its absolute magnitude is small. The deductions reported in this
bulletin are based on the al)Ove conception. King (1923), however, states
that the correlation coefficient should lie more tiian six times its proliable
error. He further states that a correlation coefficient of less than .30 indicates
a lack of marked correlation, that over .50 shows decided correlation. Further-
more, the correlation coefficient with its regression coefficients may be used
for jnirposes of prediction. The value of a knowledge of the degree of correla-
HUOOniXKSS AND FECUNDITY IX I'OWI. 59
tion lies mainly in its use for selecting a group of lueeders and nol in the
selection of individual breeders.
Tiie true eoellkicnt of correlation may only be calculated for a race pure
with regard to the characteristics being studied, as Harris (1915) points out.
False correlations result when two or more genetically different races are
concerned in any calculation Broody birds have been shown to i)e genetically
different (Hays, 19'JI-) from non-broody birds. In studying the relation of
broodiness to fecundity, it lias been deemed advisable to make three general
grou])ings: namely, (1) total jjopulation of broody and non-broody combined,
(2) only birds that went liroody during the pullet year, and (3) broody or
non-broody without regard to the degree of broodiness. The first series of
calculations was made for two purposes: first, to confirm that broody and
non-broody races are genetically different; second, to furnish evidence on the
intensity characteristics in relation to the_broody trait even in a mixed popu-
lation of broodies and non-broodies. The third series of calculations was
made by Yule's formula for presence and absence of a character, as given by
Davenport (1907). All other calculations were made by the ordinary method
for calcidating the correlation coefficient for fluctuating variables.
The regression coefficient is readily calculated after the correlation coeffi-
cient is determined. It is useful to the breeder for selection purposes. If a
group of liens, each five times broody, were selected, the regression coefficient
might be used to estimate its probable average egg production. If the degree
of correlation between days broody and annual production is known, it is a
simple matter to calculate the probable annual egg record of hens broody for
25 days or for any other period of days. Thus the regression coefficient merely
represents the amount of change in one character with respect to a unit change
in another. For example, the regression coefficient of days broody on annual
production is — .1171, and the regression of annual production on days broody
is — .3295. "What should be the average annual egg yield of hens broody for
thirty davs?
42 87 average days broody of all hens
30.00
— 12.87 days broody below the average
— 12.87 X— .3295 — 4.2407 + 16 1..8S5 (average production of all)=r
169.1257, probable record of hens broody for 30 days
The correlation ratio is comi)arable to the correlation coefficient and has a
similar use. The former is made use of where the correlation coefficient
would be false. As a measure of association in mixed races the correlation
ratio is reasonably accurate, but it is of less value than the correlation coeffi-
cient for prediction purposes. Since a con.stant is calculated for each of the
two variables in correlation ratio, a difference in magnitude of these two con-
stants sometimes occurs, probably due to genetic impurity. Correlation ratio
has not been used extensively in these studies because the correlation coeffi-
cient has been calculated on the three classes of hens with respect to broodi-
ness: namely, broody and non-broody, different degrees of broody, and broody
or non-broodv, so that regressions closely approach linearity.
60 TECHNICAL BULLETIN 7
Character of Birds Used.
Beginning in tlie spring of 1916 the plan of breeding Rhode Ishind Reds
for high egg production was somewhat modified. On that year matings were
planned to consider early sexual maturity, no winter pause, intensity, i)er-
sistency, and especially non-broodiness. Particular attention was given to tiie
elimination of the broody tendency by using females non-liroody during the
pullet year and males from non-broody mothers for breeding purposes. The
original foundation stock was all standard-bred Rhode Island Red. No new
blood has been introduced into the flock since the plan of mating for the five
characteristics above referred to was inaugurated in 1916. Inbreeding has
not been practiced to any considerable extent, but the line of ancestry has
been markedly reduced so that the present flock traces to but a small number
of the best foundation birds.
Records Kept
Records used in the study of broodiness include complete pedigree of all
birds used; complete trapnest records of every female as long as retained;
date hatched; date of first egg; age at first egg; weight at first egg; nesting
records; date of appearance of broodiness; date of placing into broody cooj);
date of return to laying house; hatching record of females used as breeders;
complete family record of the progeny from each mating; and daily, winter
and annual records on all surviving females.
Intensity.
Intensity and rate are terms used interchangeably in this report. They
refer to the number of eggs laid in a specific interval of time on a percenFage
basis of the maximum possible number of eggs in the time considered.
December Rate, as used here, is a figure obtained by dividing the number of
eggs laid by 31 if the hen began laying on or before December first. For
birds that lard their first egg later than December first, tiie rate was calcu-
lated by di\iding the number of December eggs by the number of days from
first eijij to the end of December. As a short-time measure of intensity this
may be considered more accurate than the actual number of eggs laid during
December, for obvious reasons. Winter Rate is calculated by dividing the
total number of eggs from first egg to March first by the number of days
from first egg to March first, less all pauses of four or more days in duration
from November first to March first. Annual Rate is calculated by dividing
the total eggs from first egg to 364 days thereafter, for all birds that showed
no 30-day pause after March first, by the number of days from first to last
egg. When a bird stopped laying for thirty or more days after March first,
her laying year is assumed to terminate at the beginning of this pause, and
her annual rate is calculated by dividing the number of eggs laid by the
nvuTiber of laying days before the pause.
Broodiness.
Broodiness has already been defined as the tendency of the female fowl to
incubate or attempt to incubate eggs. The intensity of liroodiness may l)e
BUOOniN'KSS AMI KiXlNDll^ IN l-OWl. (Jl
lutMsmvd l>y tiu- luimluT of lnHn)d\ periods and l)y llic mean Iciiglli uf hroody
periods. Botli I'earl (1!>U) and Goodule (1920) have measured the length
of eaeli broody period by the cessation of egg })roduction associated tlierewith.
Goodale (loc. cit.), however, stresses the fact that winter pause and fall molt
may prolong the non-productive period for a consideraitle time interval beyond
the normal broody period.
In the present studies, tlu- observation has been made that tlicre is a remark-
aiile degree of uniformity in length of broody periods in the same individual.
In the occasional bird that goes broody during the fall or winter of her pullet
year, the winter pause may greatly lengthen the period of non-production.
In such cases we have allowed four days for the l)ird to begin laying after
removal from tiie broody coop to the laying house. Such birds are removed
from the broody coop only when they no longer show signs of broodiness. In
such cases any pause up to March first, of greater duration than four days
following removal of hen from broody coop to laying house, is not considered
a broody pause.
Very frequently the laying year terminates with a broody period and no
more eggs are laid for two or three months. In such cases the length of the
last broody period is calculated in the same manner as outlined above for the
winter season. This long period of non-production is without question due
largely to the onset of complete molt and not to broodiness. The fact that
non-broody birds exhibit this long period of non-production during molt is
very convincing evidence on the point in question.
Rki.atiox of Bkoodixess to Fecuxdity.
In studying the relation of broodiness to fecundity, it has ijeen necessary
to study the degree of correlation between broodiness and rate of laying,
times broody and mean length of broody periods, winter rate and annual
rate, winter rate and annual egg yield, annual rate and annual egg yield, and
broodiness and annual egg yield.
Unpublished data at this Station indicate that rate of laying or intensity
is the most important single characteristic affecting egg yield. For this rea-
son, the relation between broodiness and rate is of extreme importance.
Either a positive or negative correlation between broodiness and rate would
be far more significant genetically than would the absolute correlation between
broodiness and egg production; for egg production has already been shown by
Goodale and Sanborn (1922) to depend upon at least five characteristics and
one of these characteristics is rate. In the present study of the relation of
broodiness to fecundity these facts are fully considered.
'. Correlation Between Times Broody and December Rate — Pullet Year.
In this study pullets are included that were hatched on the following years:
1916, 1917, 1918, 1920, 1921, 1922 and 1923. The flock hatched in 1919 is not
included because no annual records are available for that year on account of
a disease epidemic. All Rhode Island Red pullets with normal records are
included. In addition to the major portion of each flock that was bred for
egg: i)roduction, there are included a small number of inbred birds, a small
number bred for intense broodiness, a small number bred for color, and a
small nimiber used in studying the inheritance of hatchability. Inasmuch as
this report is a study of the relationship between broodiness and fecundity,
there is no conceivable reason why a rather heterogeneous flock should not
62 TECHNICAL BULLETIN' 7
be as valuable for study as one of marked uniforiiiit\" for all cliariicteristics.
Some short-time record of production is often made use of by commercial
poultrymen in predicting the laying ability of a puliet for the year. Winter
pause is likely to appear in many birds during December and is very ]iro-
nounced in earlier birds. Other birds beginning tl\eir laying year early and
continuing to lay regularly through December, as well as those starting their
laying year in December, will as a rule have higli December rate. Possibly
November records would be freer from the winter pause, but such records
would be less valuable than December records for predicting either winter or
annual egg records, as Harris and Goodale (1922) have shown. It therefore
seems advisable to use December rate in studying the relation of liroodiness
to rate.
A total of 194.5 birds consisting of both broody and non-broody are included
in the study. The range in times broody is from 0 to 12, divided into 1.3
classes. The range in December rate is from 1 to lOii, di\ided into five-unit
classes. Constants calculated from this study follow:
Number of birds ....... 194.5
Mean times broody 1.41
Times broody standard deviation . . . ±1.98
Mean December rate ...... .59.60
December rate standard deviation . . . ±20.40
Coefficient of correlation -J-.0639 ± .01.52
The constants given above impress the reader with the marked variability
in the birds studied, lioth witii regard to times broody the first year and
December rate of production. The apparently abnormal standard deviation
in times broody is due to the large percentage of non-broody birds (51.23
per cent). In other words, an impure population is concerned.
The magnitude of the standard deviation in December rate signifies \ery
marked variation in rate of laying for December. Even such a short-time
measure of fecundity is subject to excessive variability.
The coefficient of correlation between times broody and December rate,
although more than three times as great as its probable error, is of question-
able magnitude and is a false correlation as Section 2 shows.
2. Correlation Between Times Broody mid December Rate for Broody Birds
Alone — Pullet Year.
In order to measure the relation of degree of broodiness, as indicated l)y
the number of periods, to December intensity, only birds actually going broody
have been used in the calculation of the correlation coefficient. Of the group
of 1945 individuals studied in section 1, 949 birds actually went broody dur-
ing the pullet year. This number has been used to study the relation of
degree of broodiness to December rate. Constants arrived at follow:
Number of birds 949
Mean times broody 2.89
Times broody standard deviation . . . ±1.95
Mean December rate 61.24
December rate standard deviation . . . ±20.11
Coefficient of correlation -^.0145 ± .0219
Regression broodiness on rate .... -f.0014
Regression rate on broodiness .... -f-.1498
HKOODIXKSS AM) KKC rxniT^ 1\ |-( )\\ 1 . (53
Tlu" very liiryv standard doviatioii in tinies ln-oody .su{,''<>:csts a most pro-
nouiu'cd varial)ility in mimluM- of l)roody periods. The actual range is from
1 to 1'2. Since tlie modal class is broody l)ut once, there can he l)ut little
furtlier progress in reducing tiic mean numl)cr of liroody jicriods witlijn tlie
broody race.
'Hie mean December rate i.s slightly higher than that for both broodies and
non-broodies combined, in section 1. The standard deviation in rate is of the
same magnitude as that in section 1.
The coefficient of correlation between degree of broodiness and December
rate is actually less than its probable error, and since it is of verv small
magnitude, the interpretation seems justified that December rate is indepen-
dent of degree of broodiness, and that the correlation in section 1 is false.
• '. Correlation bet'ui'een the Presence of Broodiness and December Utile above
the Mean of Broodies and Xon-broodies Combined — Pidlet Year.
The actual correlation between the presence of broodiness and higii rate
is of much importance to the breeder. Such a constant was calculated for
tlie 191.5 broody and non-broody birds by the method of Yule (loc. cit.).
Decemher Rate | Broodv \ Xon-Broodv
1 ■ 1
Xuniber a')o\e population mean
632 1 595
1
Number below population mean 317 | 401
1
Totals 949 996
1 1
Coefficient of correlation -I-.1466 rt .0150
Although the degree of correlation between the presence of broodiness and
high December rate is not large, there can be no justification for any other
deduction than that the presence of broodiness is i>artially linked with high
December intensity. The elimination of the broody trait .should result in
something of a reduction in December rate for the flock as a whole.
A furtlier consideration of this relationship in a flock higli in liroodiness and
in a flock low in broodiness .seems advisable. The 1916 flock showed 86.31 per
cent broody and is unimproved, at least for broodiness. The 1923 flock siiowed
27.35 per cent broody and may be classified as an improved flock.
4- Correlation Between the Presence of Broodiness and December Rate above
the Mean of Broodies and Non-broodies Combined — Pullet Year (Unimproved
Flock 1916).
In the total of 253 birds the followiixg results were obtained: —
64
TECHNICAL BULLETIN 7
December Rate
Broody
Non-Broody
Number above population mean
138
14
Number below population mean
85
16
Totals
223
. 30
Coefficient of correlation +.2996 ±; .0386
Tlie above constant suggests tbat in the 1916 flock there was a rather dir-
tinct tendency for broody birds to lay at a higher rate during December than
non-broody birds. The constant given in section 3 for the entire period
reported upon is -j- .11-66 ±: .01.50. A comparison of the two constants assigns
them a similar value in comparison with their probable error, as each is about
eight times its probable error. There is the possibility that December rate
is higher in the early flock because they were slow to reach sexual maturity, so
that winter pause was less pronounced in December than in later flocks.
5. Correlation Bettaeen the Presence of Broodiness and December Rate above
the Mean of Broodies and Non-Broodies Combined — Pullet Year (Improved
Flock 1923).
A total of 404 birds is studied in the 1923 flock, distributed as below: —
December Rate
Broody
Non-Broody
Number above population mean 78
157
Number below population mean 51
118
Totals 129
275
Coefficient of correlation 4-.0695
.0334
The degree of correlation amounts to insignificance compared with its prob-
able error. It indicates no dependence between the presence of broodiness
and December rate above the mean. It is conceivable that early maturity
may affect December rate, and winter pause is more pronounced in tlie flocks
since the age at maturity lias been reduced.
6. Correlation Between Times Broody and Winter Rate — Pullet Year.
This study included 2221 pullets hatched the same seven years as those
studied for December rate. Winter rate is calculated on the period from
first egg to March first, as already explained. Unpublished data at this Sta-
tion indicate a rather intimate correlation between winter rate and annual
production. Winter rate was calculated on a greater number of pullets than
were studied for December rate, because the latter could only be calculated
on individuals hiying one or more eggs in December. The same classes were
used in tabulating times broody and winter rate as were used in studying
December rate. Constants calculated Are as follows: —
J
lUiOODlNKSS AM) KKCrXDITV 1\ rowi, (I.,
Xiiiiilicr ol' liirds . . . .
MiMii times l)n)(Kly
J'iiiies l)r()t)dy staiulaiil tU'\ iatioii
Mean winter rate . . . .
A\'iiitVr rate standard deviation
Cuettic'ient of correlation
2221
l.VA
rtl.f)!)
±9M7
+.0706 ± .0142
Tlu' al)o\e I'onstants sliow tlie mean winter rate to be greater tlian tiie mean
December rate ])reviously calculated. The above winter rate really signifies
that, on the average, the birds laid 66.45 per cent of the maximum possii)le
number of eggs when they were laying, since all pauses of four or more days
have been deducted in calculating winter rate. The standard deviation in
winter rate is only ± 9.37 compared with a figure of ± 20.40 for I)eceml)er
rate. The winter pause and the fact that many of the birds actually lay their
first egg during December account for the wider variability in December rate.
The coefficient of correlation between times broody and winter rate is almost
identical with that between times broody and December rate. This is a con-
stant of small magnitude, and is a fal.se correlation because the population is
made up of both broody and non-broody birds.
7. Correlation Beticeen Times Broody and Winter Rate for Broodt/ Birda
.Hone — Pullet Year.
In order to ascertain any possible relationship between winter rate and
degree of broodiness, the correlation between times broody and winter rate
has been calculated for broody birds alone. The constants obtained are as
follows: —
Xum!)er of birds 1098
Mean times broody 2.89
Times Itroody standard deviation .... ±1.93
Mean winter rate ....... 67.57
Winter rate standard deviation .... ±9.63
Coefficient of correlation —.0314 ± .0203
Regression l)roodiness on rate .... — .0063
Regression rate on broodiness .... — .1564
The mean winter rate in those iiirds that actually went l)roody during their
pullet year is 67.57 compared with 66.45 for broodies and non-broodies com-
liined. Such a difference is of no significance.
The coefficient of correlation is negative. Its small magnitude, together
with the size of its probable error, leads to the assumption that there is ab.so-
lute independence between winter rate and degree of broodiness as measured
liy times broody.
.S'. Correlation Between the Presence of Broodiness and Winter Rate ahove
the Mean of Broodies and Non-Broodies Combined — Pullet Year.
Tlie absolute correlation between the presence of broodiness and higli rate
is of importance to the breeder. Such a constant will indicate whether or not
the broody trait carries with it higher winter intensity than does the non-
l)roody trait. The coefficient of correlation is calculated below according to
Yule.
66
TECHNICAL BULLETIX 7
Winter Rate Broody Non-Broody
Nu!.ii)er above i)opulation mean 67i
558
Number below population mean-
422
565
Totals
1096
1123
Coefficient of correlation
.2358 z= .0135
The magnitude of the above constant points to a linkage between broodiness
and high winter intensity. Herein lies a probable explanation why the heavier
l)reeds, all of which carry the broody trait, are in general superior winter
layers to the non-broody lighter breeds. In the liistory of the flock under
consideration, the highest average winter records, 67.65 and 74.5 eggs, were
made by the 1920 and 1921 flocks with a percentage of broodiness amounting
to 46.03 and 44.56 respectively of birds included. The 1923 flock, for example,
showed 27.35 per cent broody and a mean winter egg record of but 51.04.
Probably broody birds carrying early sexual maturity and no winter pause are
superior as winter layers to non-broody birds possessing the same two traits,
because of some linkage between broodiness and high intensity. Further
consideration is given to this important question in sections 9 and 10.
9. Correlation Between the Presence of Broodiness and Winter Rate above
the Mean of Broodies and Non-broodies Combined — Pullet Year (Unimproved
Flock 1916).
Winter rate and broody records are complete for 332 birds in the 1916 flock.
These have been correlated below:
Vx'inter Rate
Broody
Xon
-Broody
Xumber above population mean
174
15
Nimiber below population mean
115
28
Totals
289
1
43
Coefficient of correlation
.4770 it .0286
This is a rather pronounced correlation and siiows winter intensity was
associated with broodiness in an early flock.
10. Correlation Between the Presence of Broodiness and Winter Rate above
the Mean of Broodies and Xon-Broodies Combined — Pullet Year (Improved
Flock 1923).
Winter rate and broody records for 430 birds hatched in 1923 are tabulated
below :
liROOniNKSS AND KKCUNDIT^ JN FOWL
07
\\inli-r K.itr
N'iiiiil)c'r al)()vr jxipuljition mc;in
Broody
Non-Broody
ST
HI
N'umbor' holow ]»opiil;ition moan
■51
lol
Totals
138
292
Coefficient of correlation -\- .2925
.0297
A significant coefficient of correlation between l)roodiness and liigli winter
rate sugijests tliat there is linkage between broodiness and high winter in-
tensity. Fiirtiier evidence lias already been presented in sections H and 9.
Herein lies the probalile snperiority of broody breeds over non-broody breeds
in winter intensity-
JJ. Correlation Between Times Broody and Annual Rate — Pullet Year.
Tiie nictiiod used in calcvdating annual rate does not allow for winter pause
or for time lost while broody. It is simply a figure intended to measure the
actual rate of laying between the time of laying the first pullet egg, and time
of laying the last egg before the complete molt. Winter pause birds and
broody birds are actually penalized in calculating annual rate. If there is
absolute independence between broodiness and winter pause, the only normal
liandicap that the broody bird carries over the non-broody is the production
loss during broody periods. Inasnmch as the magnitude of the annual rate
depends most largely upon yearly egg production, this method of measuring
rate should Vie most significant in breeding for fecundity. It is believed that
this is a true measure of actual rate of laying during the year. Constants
calcidated from the 2245 individuals studied follow: —
Xuiniier of l)irds
Mean times liroody
Times liroody standard deviation
Moan annual rate .
Annual rate standard deviation
Coefficient of correlation
221.5
\A\-
±1.98
56.48
±9.85
—.2620 ± .ni;J3
The above constants show that tiie 2245 birds actually laid on 56.48 ]ier
cent of tlie possible days between their first egg and the time they ended
their year with the complete molt. Tlie standard deviation agrees closely with
that for the winter rate. A mean rate of such a magnitude inuuediately
suggests high annual production.
The coefficient of correlation between times broody and annual rate is nega-
tive; and its magnitude, together with its small probable error, suggests that
liroodiness and low rate tend to move together.
15v tlie use of the regression coefficient we find that those l)irds with a
mean rate of 60.48 will be less broody than the mean of all birds stirtlied
n 44 .21 = 1.23). The fact is very evident, therefore, that broodiness tends
to lower annual rate of laying. The coefficient as determined, however, does
not represent the true correlation, since the flock of 2245 liirds is made up of
both broodv and non-broody races.
68
TECHNICAL BULLETIN 7
JL Correlation Between Times Broody and Annual Rate for Broodi/ Birds
Alone — Pullet Year.
A pure race in so far as the broody trait is concerned is to l)e found in tlie
birds actually going broody during tiieir first laying year. The total number
of birds in this class for the seven years is 1122. By tabulating the annual
rate of eacli individual against her number of broody periods a measure of
the degree of correlation between degree of broodiness and annual rate is
obtained. Constants calculated on this group follow: —
Number of i)irds
Mean times liroody
Times broody standard deviation
Mean annual rate
Annual rate standard deviation
Coefficient of correlation
Regression broodiness on rate
Resrression rate on broodiness
112-2
2.89
±1.91
.54.93
±9.24.
—.3232
—.0669
—1.5610
.0180
The mean annual rate for the broody birds is slightly lower than was found
for the total population in section 11 (56.48). No significant change is observ-
able in standard deviation.
Tlie coefficient of correlation is slightly larger than that obtained for tlie
total number of birds, and represents a rather intimate negative correlation
between times broody and annual rate. Degree of broodiness as measured
by number of periods is therefore very inimical to high annual rate.
13. Correlation Between the Presence of Broodiness and Annual Rate above
the Mean of Broodies and Non-broodies Combined — Pullet Year (Flocks 1910-
1923).
The true relation or correlation between the presence of broodiness and
annual rate above the mean is of interest and value to poultrymen. Such a
determination has been made for the 2245 birds being studied, by Yule's
metliod.
Annual Rate
Broody
j Non-Broodv
1
Numlier above population mean
513
1 675
1
Number below population mean
609
448
1
Totals
1122
I 1128
1
Coefficient of correlation
.2828
.0131
Tlie above constant does not differ significantly from tJiat representing the
whole population. In this particular case the mingling of a broody and a
non-broody race in the same correlation table did not result in skew correla-
tion. The constant — .2828 ± .0131 is known to represent a true value for
tlie flock in question, and emphasizes the importance of breeding for non-
broodiness to secure maximum annual records.
1U{()(>I)I\KSS AM) Fl',( INDir^' IN I'OWI.
()0
'riie lu'xt two sections arc devoted to tlic correlation between the iiresenee
of liroodiness and annual rate above the mean of broodies and non-l)roodies
combined, using the high broody flock of 191(i and the low iiroody tlocl< of 192;}.
Such a study siiows tlie rehitive imi)ortanee of broodiness in determiniiifi-
annual rate in a flock of low and iiieh fecundilv.
1.}. Correlation Between the Presence of Broodiness and Annual Rate above
the Mean of Broodies and Non-broodies Combined — Pullet Year (Unimproved
Flock 1916).
Annual Rate
Broodv
Non-Broody
N'umber above population mean
165
31
Number lielow population mean
Totals
159
18
324.
49
Coeflicient of correlation — .24.80 ± .0328
This constant agrees well with that for the whole eight-year period. It is
significant and illustrates the negative relation between broodiness and high
■ innual rate in an unimp^o^■ed flock.
IJ. Correlation Betxceen the Presence of Broodiness and .hniudl Rate above
the Mean of Broodies and Non-broodies Combined — Pullet Year (Improved
Flock 1923).
Annual Rate | Broody Non-Broody
N'ltmber aliove iiojiulation mean 60 164
Number below population mean
76 1 129
Totals
136 1 293
1
Coefficient of correlation — .2338
.0308
This coctficient does not difler significantly from the coefficient obtained on
the 1916 flock in section 14. or from the constant on all flocks in section 13.
Evidently the relation of broodiness to annual rate has not changed with the
improvement in fecundity.
In the previous sections, the relation between times broody and rate or
intensity of production has been considered. A considerable body of evidence
has been presented to indicate first, that hens with the broody trait do tend
to lay more intensely during the winter season than non-broody hens; second,
that broodiness is a considerable handicap to annual production in that it
lowers the annual rate. The next consideration is the relation of total days
broody during the pullet year to December rate, winter rate and annual rate.
70
TECHNICAL BULLETIN 7
16. Correlation Between Total Days Broody and December Rate — Pullet
Year.
In this study the same group of 194.5 birds botii broody and non-broody that
were studied in relation of times broody to December rate (section 1) is con-
sidered. It is important to know which is the more important from the stand-
point of rate, the number of broody periods or the actual number of days
spent in l)roodiness calculated so as to avoid winter pause and fall molt.
Constants calculated on this group of birds follow: —
Number of birds ....
Mean total days broody
Total days broody standard deviation
Mean December rate
December rate standard deviation
Coefficient of correlation
191.5
23.20
±27.07
59.60
It 20.40
+.0.529
.0153
The standard deviation in total days broody exceeds the mean total days
broody because of the large percentage of non-broody birds in the grouj)
studied.
The coefficient of correlation agrees rather closely with the figure given in
section 1 where times broody and December rate are considered. Evidently
broodiness may be measured either by periods or iiy total days. The degree
of correlation is slight, and it is really a false correlation because based upon
a mixed population — broody and non-broody.
17. Correlation Between Total Days Broody and December Rate for Broody
Birds Alone — Pidlet Year.
The relation between degree of broodiness, as measured liy total days of
non-production associated with broodiness during the pullet year, and Decem-
ber rate may be determined by using only the birds that went broody the first
year. Such a determination was made for the same group of 94.9 birds that
was considered in section 2. The following are the constants: —
949
42.84
±27.38
61.24
±20.11
—.0002
—.0003
—.0001
Number of liirds ....
Mean total days broody
Total days broody standard de\'iation
Mean December rate
December rate standard deviation
Coefficient of correlation —.0002 ± .0219
Regression l^roodiness on rate
Regression rate on broodiness
That degree of broodiness and December rate are entirely indeiiendent is
shown by the above coefficient of correlation which is practically zero. This
is rather conclusive evidence that December intensity bears no relation to tlie
])resence or absence of the liroody trait.
18. Correlation Between Total Days Broody and Winter Rate — Pullet Year.
A total of 2221 birds studied in section 6 are included in this study to dis-
cover the degree of dependence or independence between total days l)roody
and winter rate. The constants calculated follow: —
I>K()()ni\i:SS AMI FF.C rNDll'i IN I'OWl. 71
XuiiilnM- (if birds ....... 2"J'_M
.Moan total day.s l)roody ..... 2'.i.'Hf
'!\)tal days broody standard deviation . ±27.01
Mean winter rate 6(>.|..5
Winter rate standard deviation .... ±9.37
Coeflieient of eorrelation +.0178 ± .0142
Tlie coeflieient of eorrelation is praetieally the same fi}>iire as was ol)iained
between times broody and winter rate. This is also a false eorrelation be-
eau.se l>roodies and non-broodies each represent a genetic type. Since winter
production and annual production are so intimately correlated (Hervey 1923;
Hays, Sanborn and James 1924), high winter record is of very great import-
ance in lireedlng for fecundity.
Blakeman's test for linearity of regression has lieeh api)lied in this studv
with the following results:
Correlation ratio for days broody . . . -)-.1134
Correlation ratio for winter rate . . . -(-.1695
(Cor. Ratio)2— (Cor. Coeff.)2=i . . . .007.5 ± .0024
(Cor. Ratio)2— (Cor. Coeff.)2 = . . .0233 ± .00^3
The difference between the correlation ratio for winter rate squared and
the correlation coefficient squared is .0233 ± .0043, a difference more than five
times as great as its probable error. This fact indicates that the coefficient
of correlation is false, as might be anticipated from the fact that two genetic
races are concerned.
10. Correlation Between To^nl Dai/s llroo'l;/ and Winter Rale for Broo'l;/
Birrh Alone — Ptillet Year.
Winter rate records are available on 1098 birds tiiat were broody the first
year. In this study days broody is tabulated against winter rate to further
disco\er the correlation between degree of broodiness and winter rate. Con-
stants are as follows: —
Xuniber of liirds 1098
Mean total days broody 42.8.5
Total days broody standard deviation . . . ±27.14
Mean winter rate ....... 67.57
Winter rate standard deviation .... ±9.63
Coefficient of correlation —.0241 ± .0203
Regression broodiness on rate .... ^.0679
Regression rate on broodiness .... — .008.5
The coefficient of correlation as siiown above signifies independence between
degree of broodiness and winter rate. The intensity of tlie broody trait is
therefore of no concern in affecting winter intensity.
20. Correlation Beticeen Total ]>ai/s Hroodii and Annual Rate — Pullet Year.
The total days broody for each bird are tal)ulated against iier aiuiual rate.
The lowest rate class is 16-20; the higiiest rate cla.ss is 86-90 Tlie lowest
broody class is 0-9; tlie highest broody class is 150-159 days. This study on
the 2245 birds used in section 11 will show if broodiness is an advantage or
disadvantage from the standpoint of annual rate. Are broodiness and high
intensity linked toeether? Constants calculated are:—
72
TECHNICAL BULLETIN 7
Number of birds ....
Mean total days broody
Total days broody standard deviation
Mean annual rate ....
Annual rate standard deviation
Coefficient of correlation
2245
23.68
±26.98
56.48
±9.85
—.2720
.0132
The coefficient of correlation between days broody and annual rate is nega-
tive and of such magnitude as to be of considerable significance, were it not
for the fact that the two races of birds give a false correlation.
21. Correlation Between Total Days Broody and Annual Rate for Broody
Birds Alone — Pullet Year.
The same group of birds considered in section 12 is used in this study.
The coefficient of correlation is here used to measure the degree of association
between degree of broodiness and annual intensity. Constants obtained are
the following: —
PrPCENT BPOOPy
\
_ M5^N WINTER R/^TE
1 -
^ 60
^ 50
«Xi
X
K
^ 40
\ 30
IS/6
/9/7
/9JS /9J9 /G20 192/ /92£
/923
/E^f? h/^TC/^EP
CiiAitT 1. Relation of ])ercentage of birds broody to mean winter rate i^y yeans.
HHOODIXKSS AM) I'l-XIXDIT^ 1\ I'OWI.
7:3
XuiiiIht of liirds ....
.Mean total days broody
Total days broody standard (U'\ialion
.Mean annual rate ....
Annilal rate standard deviation
Coetticient of correlation
Regression broodlness on rate
Resrression rate on broodiness
1122
1-2.87
=:2().8|.
54..}».'J
:^i).21.
-..•J<i22
—1.0526
—.1246
.0175
A ratiier marked degree of negative correlation exists between days broody
and annual rate. The degree of broodiness influenees annual rate l)ecause of
the loss of time while broody. This constant agrees substantially with the con-
stant for times broody and annual rate (—.3232 ± .0180).
Relatlitu of Broodiness lo ]\'iiifer Ihiie (dkI Annual Rote.
In charts 1 and 2 the mean percentage of birds liroody on the different
.ears is illustrated gra])hica]ly by a solid line. The mean winter rate and
fc
$
I
I
■90
PERCENT BPOOPy
ME^N A^A/A/L/^L P^T5
1916 1917 1918 19/9 /920 192/ /922 1923
Yf^R H//TCHEP
CiiAur 2 Relation of percentage of birds broody to mean amiual rate by years.
74 TECHNICAL BULLETIN 7
mean annual rate are represented liy In-oken lines. Tlie groups of birds used
in making the two charts are not identical because winter rate records are
availal)le on a considerable number of individuals that did not survive to com-
plete annual records. However, the two groups are so nearly identical that
the mean percentage of broody birds closely agrees in the two charts.
The increase in mean winter rate from 1916 to 1923 is 4.06, while the increase
for annual rate in the same period is 6.00. This fact indicates that annual
rate has increased more rapidly than winter rate as the percentage of broody
birds has been reduced from year to year. The greater degree of parallelism
in the two graphs on chart 1 suggests tliat a change in percentage of liroody
birds is usually accompanied by a change in winter rate. Chart 2 shows a
lesser relationsliip between percentage broody and annual rate.
In general the two charts furnish evidence that lioth mean winter rate and
mean annual rate may be increased while the percentage of broody birds is
being reduced. The lowering of the percentage of broody birds to at least 30
per cent, as has been accomplished in the flock studied, appears to be advan-
tageous from the standpoint of annual jiroduction.
Tiie next section is devoted to a study of the relation between the number
of broody periods and the mean length of broody periods. It seems desirable
to ascertain if the average length of broody period is affected by the numl)er
of periods. Does the frequency of onset of broodiness tend to shorten or
lengthen the period? The coefficient of correlation is again made use of and
the number of broody periods is tabulated against the mean length of period,
using 1135 birds that were broody in the pullet year.
22. Correlation Beticeen Times Broothj and Mean Lenath of Brooihi
Periods — Pullet Year.
Any attempt to decrease the intensity of liroodiness must be accomplished
either by reducing the number of periods or by reducing the length of these
periods. The coefficient of correlation is here calculated to discover a possible
relationship between number and length of broody periods. Constants calcu-
lated are as follows:
Number of birds ....... 113.5
Mean times broody 2.S9
Times broody standard deviation .... ±3.67
Mean length of periods 15.10
Length of periods standard deviation . . . ±3.78
Coefficient of correlation —.2338 ± .0189
Regression times broody on length . . . — .4620
Regression length on times broody . . . — .1183
The total number* of birds showing one or more broody periods is slightly
greater than the number in sections 12, 21, 30 and 32, broody records being-
available on a few birds on which annual rate records are lacking. The stand-
"The total number of hirds goin^ bi-oody was 1135. Of this group, 1017 individuals
were first broody after March first so that the actual length of the period of non-
production attributable to broodiness could be definitely recorded. There were 118
birds broody before March first. The mean lensjth of broody period for the 1017
birds is 1.5.95 days, while that for the grou]) of 1135 birds is 15.10 days. This slight
difference in mean length of period is not significant and may be attributed to our
inability to separate broody pause from winter pause in those 118 birds going broody
before March first. The method of allowing a bird but four days to begin laying after
her return to the laying house following broodiness during the winter season is faulty
in that it actually assigns a shorter broody period during winter than the mean of
summer broody periods.
HKOODINI'.SS AND FI'.ClN Dll ^ 1\ l-()\\l. 75
;ird tk'vi;ilimi in times lirootly is jiivalrr tli.ui liic iik'.iii lu'taiisr !).')1 hirds (H;J
JUT cent) fell into classes 1-4, Icavinir only 17 jxt cent in classes H-lli. Tiic
iiuulal class is 'J.
The mean lenglli of Ijroody jjcriods is 15.10 days witli a standard deviation
of ± 3.78. There is, therefore, much {greater uniformity in len^lli of period
than is observed for number of periods. Evidently the numiu-r of i)eriods
otVcrs a more fertile held for imi)rovement than is offered by the lenfrth of
period.
The negative coetticient of correlation indicates tiiat an increase in numi)er
of broody periods is accompanied by a decrease in tiieir average length. A
reduction in number of periods would therefore be accompanied by an increase
in their length. That this relationship is far from absolute is shown by the
magnitude of the correlation coetticient. Certainly the time lost in non-pro-
duction has been very significantly reduced by decreasing the innnber of
broody periods, as table 1 shows.
.'.). Correlation Between Winter Rate and Annual Rate — Pullet )'ear.
The records for 2212 individuals both broody and non-broody are available
for study. This relationship is important liecause both rates bear ;i rather
intimate relation to egg production. The fact has previously been pointed
out that broody birds tend to be more intense winter layers than are non-
broody birds, but that the former are likely to carry a lower annual rate. An
intimate correlation between winter rate and annual rate would suggest that
rate of laying for the year may be predicted from the winter rate. Constants
calculated are as follows: —
Number of birds ....... 224.2
Mean winter rate ....... 66.41
Winter rate standard deviation .... ±.9.38
Mean annual rate ....... 56.46
.Vnnual rate standard deviation .... ±9.85
Coefficient of correlation +.4900 ± .0108
The above con.stants indicate a slightly greater relaLive standard deviation
in annual rate than exists for winter rate. Such a condition might he sur-
mised from the fact that broodiness and complete molt may both affect anmial
rate but for the most i)art are not concerned in winter rale.
A rather intimate correlation exists between winter and annual rate. Evi-
dently those birds above the average in winter rate would be exi)ected to be
above the average in annual rate. The practice of selecting for high winter
rate is without doubt sound from the standpoint of securing high annual rate.
J4. Correlation lietween the Presence of Broodiness and Winter Production
above the Mean of Broodies and Non-ljroodies Comfrined — Pullet Year (Flocks
1916-1923).
The absolute correlation !)et\veen the presence of l)r()odine.ss and winter ))ro-
duction above the mean of all birds is of much concern to poultrymen striving
for high winter records. Such information will show whether or not itroody
birds tend to lay more eggs before March first than do broody-free birds. In
section 8 some evidence is presented to indicate that broody birds do actually
lav at a .slightly higher rate than non-broodies when they are laying; but late
J...... .pi pinturitv, winter, pause and the occasional winter l)roody ])eri(td may
76 TECHNICAL BULLETIN T
possibly be more pronounced in the l)roody population. The actual correla-
tion between the presence of broodiness and winter production above the mean
is siiown by the following table: —
Winter Production I Broody Non-Broodv
Number above population mean
1
561
561
Number below population mean
334
563
1
Totals
1
1 1095
i 1124
1
Coefficient of correlation — .02^4 = .0143
The above coefficient is so small as to iie of no significance and it is less
than three times the magnitude of its probable error. The deduction nuist be
made from this study that broodiness and winter egg production are entirely
independent e\en though broody birds do lay at a sligiitly higher rate in
winter when they are laying.
25. Correlation Between the Presence of Broodiness and Winter Production
above the Mean of Broodies and yon-hroodies combined — Pullet Year (Un-
improved Flock 1916).
Winter Production
\ Broo'dy
Non-Broody
Nunil)er above population mean
1
14
Number below population mean
140
29
Totals
289
43
Coefficient of correlation — .3759 it .0318
The above constant shows that broodiness l)ears a rather intimate correhi-
tion to high winter production in the 1916 flock. Such an assumption is based
on the conclusion that the individuals laying more eggs in winter than the
average of the flock (46.87 eggs) are high producers. Even though a small per-
centage of the 1916 flock reduced this winter record by being broody before
March first, broody birds appeared to carry intensity to a sufficient extent to
enable them to lay more eggs for the period than did the non-broody l>irds.
It is rather striking that the total population (section 24) should not exhibit
a con-stant similar to that for the 1916 flock. No doubt changes in early ma-
turity and winter pause have operated to modify winter production to a
greater extent than any possible lowering of intensity by the elimination of
broodiness has been responsible for.
nR()()i>iNKss AND i'K( rxnnv i\ fowi.
77
.'(.'. Vorreltidon Hcticeeii tin Presence of Broodiness and ]\'inler Production
(il'ove the Mean of Brood'ux and Xon-broodies Combined— Pullet )'fiir ( I m-
/.rni-ed Flock IV2S).
Winter I'rodiu'lion Hioocly
Non-Hroody
Xumber above population mean 77
1
140
1
Xumher lieUnv jxiinil.ition mean | 61
1
162
Totals 1 138
1
292
Coefficient of correlation + .1563 ±: .0317
This constant is of questionable magnitude and signilies tliat winter produc-
tion of the 1923 flock above the mean of 53.62 eggs is but little dependent
upon the presence of the broody trait. The fact should be recalled, however,
that the maxinnmi winter production (74.5 eggs) was made by the 1921 flock
with 44.56 per cent of the birds broody during the pullet year.
The later sections of this report are devoted to a consideration of the cor-
relation between rate and egg yield and broodiness and egg yield.
J7. Correlation Betxceen Winter Rate and Annual Eyy Production — Pullet
Year.
In commercial poultry breeding for fecundity, a short-time measure of
probable annual production is of vast importance. If the winter rate could
be used as a l)asis for selecting breeding females as efficiently as the yearly
record, it would be of vast economic importance. By making use of the co-
efficient of correlation, a measure of the probable worth of tlie winter rate in
selecting for large yearly records is obtained. The constants arrived at in
this study are given below: —
Xiuiiber of birds ....
.Mean winter rate
Winter rate standard deviation .
Mean annual production
.\nnual production standard deviation
Coefficient of correlation
2242
66.41
±9.38
174.37
±44.59
+.4561 ± .0113
The mean annual egg production of the 2242 birds used in section 27 was
174.37, with a standard deviation of 44.59, or a coefficient of variation of
about 25 per cent. The class range in egg production is from 21 to 300 with
class intervals of 10. This wide range in production is due to the hetero-
ireneitv of the flock and to the number of characteristics that affect production.
The magnitude of the correlation coefficient, together witli the small ])roi)-
ahle error, suggests that winter rate is rather intimately correlated with
annual esift production.
,'cS. Correlation Beticeen Annual Rate and Annual E</(j Yield.
Annual rate as calculated for this flock is a rather concise measure of in-
78
TECHNICAL BULLETIN 7
tensity for the entire pullet laying year. It should furnish a reasonalily true
measure of the bird's ability to lay throughout the year. Since the relation
of broodiness to annual rate has already been considered, it seems advisable
to correlate annual rate witli annual yield. The calculations gave the follow-
ing constants:
Number of birds ....
Mean annual rate ....
Anniuil rate standard deviation .
Mean annual egg yield .
Annual egg yield standard deviation
Coefficient of correlation
2289
.56.38
±9.86
172.21
±46.61
-|-.6ri7 ± .0077
A very sensible positive correlation was found l)etween annual rate and
annual egg yield. Annual rate is thus a very dependal)le measure of a bird's
ability to lay during her pullet year.
39. Correlation Befiveen Times Brooihj and Annual Production — Pullet Year.
Tlie records of 2215 birds broody and non-broody are tabulated and the
coefficient of correlation calculated between times broody and annual produc-
tion. Constants arrived at follow: —
Number of birds ....
Mean times broody
Times broody standard deviation .
Mean annual production
Annual production standard deviation
Coefficient of correlation
224.5
1.44
±1.98
173.06
±46.40
—.2126
.0136
This constant is false because the table is made up of two genetically dis-
tinct races, namely, broody and non-broody.
Table 2. — Relation of Broodiness to Egg Record.
Times Broody Number of Birds Egg Record
0
1
2
3
■ 4
5
6
7
8
9
10
11
12
13
1121
312
259
220
149
72
47
28
17
9
5
2
1
1
181.31
178.32
156.62
156.50
158.65
162.58
1.53.59
140.14
1.55.50
147.72
145..50
160.50
1.55.50
145.50
Reference to table 2 above shows that the 1121 non-broody birds averaged
181 31 eggs per year. Close to tiiis group in production is the class of 312
birds with but one broody period, averaging 178.32 eggs. A somewhat gradu-
al but not regular decline begins with the group broody twice. No further
HHOOniXKSS AM) FKcrNDM^ IN I'OWI. 7V)
decline is observed until tlie group with six liroociy jhtIixIs is rc.iclicd, nt'U-r
wliidi tiie mean egg yield fnlls signilicantly. The jji-ohable error inereases so
rajudly due to small numiiers when the elass with eight broody jjeriods is
reaehed tiiat very little signifieanee can be attaehed to the mean in this and
later elasses. On the whole, this table suggests in a general wav iliat increased
liroodiness does lower the annur.l record.
■yO. Correlation Jieficecu Times Jiroodi/ (ind .linntal E(iii Yield for liroodtj
Birds Alone — Pullet Year.
A true measure of the correlation between times broody and animal pro-
duction can only lie found within the liroody population as ))reviouslv staled.
In the eight-year period IVl'l broody birds are concerned. This groujt has
been tabulated and constants calculated.
Xiunber of birds ....... 1122
.Mean times broody ...... 2 89
Times broody standard de\ iation .... ±1.91
Mean annual production ..... 161-.89
Aiuiual production standard de\iation . . . ±45.03
Coefficient of correlation ..... —.1791 ± .019.5
Regression broodiness on production . . . — .007()
Regression ]iroduction on broodiness . . . — 4.2167
A negative coefficient of correlation of .1791 ± .0195 indicates that the
correlation between times broody and annual ])roduetion may not be consid-
ered intimate. Such a constant leads to the assumption that broodiness as
measured by periods has played some part in limiting annual production for
the eight-year period studied.
31. Correlation Betzceen Total I>ai/s Brood tj and Annual Production — Pullet
Year.
Broodiness may next be measured in total days for the year. Tliis method
of measuring has been made use of between total days l)roody, and annual
egg production. The birds u.sed in this tabulation are 2245 in number, both
broody and non-l)roody. Constants calculated follow: —
Xinnber of birds ....... 2245
Mean total days broody ..... 28.68
Total days broody standard deviation . . . ±26.98
Mean anniuil ]>roduction ..... 173.06
Annual production standard de\iatiou . . ±46.40
Coefficient of correlation —.2200 ± .0135
An interesting and important fact is brought out by the aliove constants
in that the correlation between total days broody and annual egg record is
negative and of almost the same magnitude as that obtained between times
broody and annual egg record (section 29). The deduction that broodiness
lowers annual record again seems warranted, but the coefficient obtained in
this table is false because of the presence of the genetically diflFerent i>roodies
and non-b.roodies.
80 'lECHXICAI. BULLETIN 7
32. Correlation Jietween Total Days Broodi/ and Annual Egg Yield for
Broody Birds Alone — Piillet Year.
A tabulation made of the 1122 broody birds gives the true correlation be-
tween degree of i)roodiness and pullet-year production. Constants obtained
are the following: —
Number of birds ....... 1122
Mean days brood}- 42.87
Days broody standard deviation .... ±; 26.84
Mean annual production ..... 164.89
Annual production standard deviation . . ±45.03
Coefficient of correlation — .1964 ±: .0194
Regression days broody on production . . . — .1171
Regression production on days broody . . . — .3295
Tlie degree of correlation between total days broody and annual egg record
is not at all intimate. It is in very close agreement with the constant for
times broody and annual egg yield, in section 30. On the whole, liroodiness
has been shown to be negatively correlated with annual production to a rather
moderate degree over the eight-year })eriod covered in this report.
■3-3. Correlation Betzveen the Presence of Broodinei^s and Annual Production
above the Mean of Broodies and Non-broodies Combined — Pullet Year (Flocks
lOin-1923).
Tlie true correlation between the presence of broodiness and high annua!
egg ])rodurtion is determined below.
Annual Production
Broody
Non- Broody
Number abo\e population mean
514
1 665
1
Number below po))ulation mean
608
458
Totals
1122
1123
1
Coefficient of correlation — .2640 ± .0132
The above constant is statistically significant and is of sufficient magnitude
to warrant the assumption that broodiness is negatively correlated with high
annual production. The fact that annual egg record depends upon a vast
array of genetic and non-genetic factors should not be overlooked. During
the eight-year period being considered there has been constant progress in
eliminating broodiness, yet the mean annual egg records of the flocks have
been stable since 1920. Very likely broodiness has played a greater part in
affecting production in some years than on others. The two following sec-
tions show that the correlation between broodiness and annual record has not
been intimate either in 1916 or 1923.
BHOODIXKSS AND !'K( IN DI r> IN lOWI.
81
. ;. C'orrelofion lietzceen the Presence of lirnodiufxn diid Aimititl I'lodiirlion
iihofe (he Menu of Broodies and Xnn-hroodies Ct)>til>lned~ I'lillel )'e(ir (Un-
improved Flock 1976).
Anmi.il Production
liroody
Non-Uroody
Numl)er above population mean
1G5
27
Number below population mean
159
22
Totals
324
•!•!)
Coefficient of correlation
.0837 ± .034.7
The correlation coefficient is negative and less tlian tiirec times its proljaijle
error. The conclusion seems to be warranted that in tlie 1916 flock there is
no significant correlation between broodiness and annual production.
35. Correlation Between the Presence of Broodiness and Annual Production
above the Mean of Broodies and Non-broodies Combined — Pullet Year (Im-
proved Flock 1923).
Annual Production
Broody
Non-Broody
Number above population mean
67
164
Niunber lielow population mean
69
129
Totals
136
293
Coefficient of correlation — .1339 =t .0320
This coefficient is about four times its probable error and is prol)al)ly of
some significance. The degree of correlation between broodiness and annua!
production is slight on the two years studied; a fact that probably indicates
more correlation on the intervening years.
82
TECHNICAL BULLETIN 7
General Sum:marv and Deductions
The means for the broodies ;ind non-broodies for the three rates and for
winter and annual production for the ciaht-year period are as follows: —
Broody
Non-broody
Number
of Birds
!Mean
Number
of Birds
Mean
December Rate
949
61.24
996
58.05
Winter Rate
1098
67.57
1123
65.36
Annual Rate
•1122
51.93
1122
38.04
Winter Production
1094
60 46
1123
58.79
Annual Production
1122
161.89
1121
181.31
1. Birds carrying the broody trait lay at a slightly higher rate when they
are laying than do non-broody birds. This characteristic is oliservable in both
December records and in winter records.
2. A much more intimate correlation between tlie presence of broodiness
and high December intensity was observed in the 1916 flock than in tlie 1923
flock. This difference may prol)ably be attributed to changes in sexual ma-
turity and to the time of onset and termination of winter pause.
3. Intensity for the winter period is rather intimately correlated with the
presence of broodiness. Such a relationship is observed in the total po]mla-
tion for the eight years, in the 1916 flock and in the 1923 flock.
4. Degree of broodiness is not correlated with either December rate or
winter rate.
5. The degree of correlation between broodiness and high annual rate is
constant but negative and significant.
6. Degree of broodiness may be measured with equal accuracy by number
of broody periods or by total days broody during the pullet year.
7. "Winter rate and annual rate are distinctly positively correlated.
8. Tlie duration of the ))roody period is somewhat lessened as the number
of periods increases.
9. Correlation between the presence of Ijroodiness and winter production
above the average is negligible when all the birds are considered ()\er the
eight-year period. There is a significant positive correlation for the 1916 flock
and probably a slight correlation within the 1923 flock.
10. The mean winter egg record of broodies is not significantly greater
than that of non-broodies.
11. Annual egg production is significantly negatively correlated with
l)roodiness in the total population studied; to a very minor degree in tiie 1916
flock; and to a rather moderate degree in the 1923 flock.
12. The elimination of broodiness has had but little significance in breed-
ing for high winter production but a pronounced significance in breeding f"i-
hisi'li annual records.
BROODINESS VXD FKCrNDnV 1\ FOWL gjj
Referexcks
B.iteson, W. 1!)()2. Kxperiments with iionltiv. lijits. to Evolution Com.,
Roy; Soc. (London] 1:87-121-
Bljikeman, ,1. 1!)().5. On tests for linearit\ of regression in fre(|nene\ dis-
tribution. Hionietrika 41:;j;j2-o'5().
!).i\ eniiort, K. 1907. !'rineii)le.-; of i)reedinsi\ p. 1-71. Boston, Xew York,
ete.: Ginn <.\: ("o.
Goodale, H. D., Ruby Sanborn and Donald Wiiite. 1920. Broodiness in
doniestie fowl. Mass. Agr. Expt. Sta. Bull. 199.
Goodale, H. D. and Ruby Sanborn. 1922. Changes in egg production in
the station flock. Mass. Agr. Expt. Sta. Bull. 211.
Harris, J. A. 191.'). IMiysical conformation of cows and milk yield. Jour.
Hered. 6:;M.8-.-^50
Harris, J. A. and 11. I). Goodale. 1922. The correlation between the egg
production of the various periods of the year in the Rhode Island Red
breed of domestic fowl. Genetics 7:44G-i65.
Hays, F. A. 1924. Inbreeding the Rhode Island Red fowl with special
reference to winter egg production. Amer. Xat. 58:1.3-.59.
Ha.\s, F. A., Ruby Sanborn and L. L. James. 1924. Correlation studies
on winter fecundity. Mass. Agr. Expt. Sta. Bull. 220.
Hervey, G. W. 1923. The prediction of egg records. N". J. Agr. Expt.
Sta. Bull. 389.
Hurst, C. C. 190.5. Experiments with poultry. Rpts. to Evolution Com.,
Roy. Soc. i London] 2:131-1-54.
King, V\'. I. 1923. Tiie elements of statistical method, p. 215. X'ew York:
Macmillan Co.
Pearl, Ii. 1914. Studies on the physiology of reproduction in the domestic
fowl. VII. Data regarding the brooding instinct 'Ji relation to egg
production. Jour. Anim. Behavior 4:2Gfi-288.
Punnett, R. C. and P. G. Bailey. 1920. Genetic studies in poultry. II.
Inheritance of egg color and broodiness. Jour. Genet. 10:277-292.
Massachusetts
Agricultural Experiment Station
TECHNICAL BULLETIN No. 8 JULY, 1926
WINTER CYCLE AND WINTER PAUSE
IN RELATION TO
WINTER AND ANNUAL EGG PRODUCTION
By F. A. HAYS and RUBY SANBORN
In this bulletin are reported the results of a statistical study of winter
cycle and winter pause records taken over a period of nine years, with as
many separate flocks. The records show that winter cycle furnishes a sig-
nificant short-time measure of probable annual egg production in the flock.
Winter pause is shown to be a potent cause of lowered annual egg produc-
tion; £ind any increase in length of pause is but partially compensated for
by later increased activity in egg production.
Requests for bulletins should he addressed to the
AGRICULTURAL EXPERIMENT STATION
AMHERST, MASS.
WINTER CYCLE AND WINTER PAUSE
IN RELATION TO WINTER AND ANNUAL
EGG PRODUCTION.
By F. A. HAYS and RUBY SANBORN
IXTRODUCTION
Winter cycle is represented by a period of continuous egg laying during the
winter season. February 28 or 29th is arbitrarily chosen as the closing date
of the winter period. A distinct cessation of laying before the end of Febru-
ary may be assumed to mark the end of the winter laying cycle. Just what
length of pause should be chosen to mark the end of the winter laying cycle
is purely arbitrary. Goodale (1918) suggested that a ten-day cessation of
laying is sufficiently small to mark the end of the winter cycle. In the studies
reported here, a four-day interval is considered as a winter pause because, in
the flock studied, such a pause generally means the omission of one clutch*
of eggs. The omission of one clutch of eggs really marks a distinct break
in the functioning of the reproductive system and probably represents the
termination of a laying cycle. Pauses due to broodiness or to injury or dis-
ease are not considered, and only pauses occurring between November 1 and
February 28 or 29 are classed as winter pauses or as marking the end of the
winter cycle.
Winter cycle may further be defined as total days from dkte of first egg
to a pause of four or more days, the pause being considered only between
the dates November first and March first. Winter cycle length can therefore
be determined only for the pause class of birds, as Goodale (1918) pointed
out. Winter cycle is probably inherited as a recessive, according to Goodale
(loc. cit.) ; but he failed to discover a relationship between number of eggs
laid in the winter cycle and duration of winter pause.
Winter pause may be defined as the period when egg production ceases,
following the termination of the winter cycle and previous to the initiation
of the spring laying cycle. Pearl (1912) and Goodale (1918) make reference
to winter pause in relation to fecundity. Pearl (loc. cit.) found that the
winter cycle in Barred I'lymouth Rocks was characteristically terminated l)y
a cessation of production. Goodale (loc. cit.), on the other hand, observed a
cessation of production previous to March first in but a part of a Rhode
Island Red flock. Goodale (1922) states that winter pause is due in part
to inherited characteristics and in part to environmental conditions. He dis-
covered a rather intimate relation between the time of beginning to lay in
the fall and the appearance and duration of the winter pause.
Winter pause is usually associated with the shedding of some feathers, or
partial molt, as Hays (1924) suggested. Furthermore, a cessation of pro-
duction during broodiness or at any other time during the hiying year is
* The term "clutch" refers to the number of eggs laid on successive days, which
js more or less characteristic of the individual hen.
WINTER CYCLF, AND WINTKH I'M'Sl'.
!()'
generally eliaraeterized hy partial or eomplete iiiiilt. Apparently nioilinf;-,
wliieh jueeedes the development, of a new growlli of feathers, is a phenome-
non initiated by a eessation of aetive funetioning of the rejjroduetive system.
A number of environmental intluenees as well as inherent feeimdity tluis
appear to' be eoneerned in the dilVerent pauses of the pullet laying year. The
sunj total of the winter })ause may or may not be represented by a continu
ous period of non-production. In other words, there may be several pauses
of four or more days with some production intervening. In these studies such
pauses have been added together and such individuals ])laeed in the san)e
class with birds whose pause is unbroken.
The duration* of winter pause is recognized to be de])endent upon environ-
mental conditions such as hatching date, feeding and housing, weather con-
ditions, and all other influences that may affect the physical condition and
state of metabolism of the pullet. Since age at which .sexual maturity is
attained is modified by such controllable conditions as date of hatching, as
Hays and Sanborn (192+) point out, and since age when sexually mature
largely governs the time of beginning to lay, a complex relationship must
exist between age at first egg and winter pause.
The presence or absence of winter pause depends upon inheritance, as
Hays (1924') points out. Pullets that exhibit a winter pau.se of a week or
more before March first are known to carry a dominant factor M, while
non-pause pullets are recessive and lack factor M. This factor is transmitted
equally by both males and females. In this connection, breeders should bear
in mind that genetically non-pause pullets may exhibit a winter pause
brought on by disease, abrupt changes of feed, moving to new quarters, and
other environmental influences largely within control of the poultryman. In
tlie breeding flock of the Massachu.setts Agricultural Experiment Station
extreme care has been exercised to keep environmental conditions constant
from year to year in order that inherited traits affecting fecundity may be
studied.
Character of Birds Used
This study includes all Rhode Island Red females, hatched from 1916 to
1021, on which pullet-year trapnest records are available. The flocks each
year are made up of all the daughters from each individual female whose
progeny was retained. Culling within the family has not been practiced.
The major portion of birds in each flock belong to the fecundity experiment.
There are, however, a limited number of birds bred for non-broodiness, some
lor intense broodiness, some for luitchai)ility, some for color, and a few
inbreds that arc included, l^ullet-year records alone are used in this study.
The heterogeneity of the flock can scarcely be considered as a factor a fleet-
ing the constants presented in this report.
Two possible methods are open for attacking these questions in a large
population made up from nine years' breeding for fecundity. The first
method is the use of the coefficient of correlation. The second po.ssible mode
of attack is through the presentation of actual data by families tracing to
both foundation males and females through nine successive years. The possi-
bilities of the two methods may be briefly considered.
* In a small percentage of the flock the pause begun Inte in February and sucn
birds did not resume laying until some time in March or later. In these cases, dura-
tion of pause is calculated when production is actually resumed.
168
TECHNICAL BULLETIN 8
The Coefficient of Correlation.
A general survey of a series of individual egg records myy lead to certain
general deductions some of which may actually be true and others false. The
breeder needs to know just how nnich stress to lay upon difl'erent character-
istics associated with the traits concerned in high fecundity. Tiie simple
correlation coefficient affords a concise measure of the degree of association
between specific traits and higii fecundity as well as a measure of the rela-
tionship between the presence of particular traits and high fecundity. The
correlation coefficient may thus l)e made use of by the breeder in two ways:
first, for prediction purposes; and second, in the selection of breeding stock
to obtain the most valuable combination of traits. Stated concisely, the cor-
relation coefficient is the only direct and specific measure for degree of
association of characteristics where large numbers of individuals are con-
cerned. Fecundity records may be modified by a vast number of environ-
mental conditions as well as by tiie live traits pointed out by Goodale and
Sanborn (1922) which are shown to be inherited. Hays (1924.). Only by
the use of large numbers of records made over a period of years under uni-
form conditions of management and in a ilock bred for uniformity can a
true value of the relative importance of characteristics concerned with fecun-
dity be approached. The coelT'cient of correlation thus becomes an invalu-
able tool in breeding for fecundity.
Presentation of Data by Families.
A study undertaken to consider the winter cycle and winter pause by
separate families would necessitate the presentation of page after page of
abstract data. Such data should be accompanied by detailed and complete
discussions and such general deductions as would seem justified. No definite
constants could be determined on numbers so small as the individual family.
Possibly all the descendants of particular individuals could be considered
as units, but from the genetic standpoint such a consideration should be
classed as questionable. A general tabulation of the whole population,
giving such information as mean hatching date, mean age at first egg, mean
weight at first egg, percentage of birds pausing, mean length of pause, mean
winter production, mean annual production, etc., by years could be made.
Such a tabulation would again be open to the criticism of not furnisliing
specific information. Only general deductions could be made and no evi-
dence would be furnished as to relative values. In view of the above facts,
this method of handling the data is not considered feasible.
Winter Cycle
Winter cycle may be considered in three general categories: namely, (a)
in its relation to environmental conditions, (b) in its relation to heritable
characteristics concerned in fecundity, and (c) in its absolute relation to egg
production.
(a) Relation of Enviroiimental Conditions to Winter Cycle.
Hatching date belongs to the definitely controllable class of conditions in
tiiat it may be varied at will of the investigator. Date of first egg depends
WINTER CYCLE AND AVINTKH I'Al'Si:
KIJ)
both upon en\ ironiiu'iit aiul inlKTilaiuT. I'lic tiiiii- when a f:riiii|i of pullets
will begin to lay depends in jnirt ujxm liateliin}!: date, method of feeding an«l
management, and upon weatiur eondilions — all of whieii may l)e elassed as
rn\ ironmental. The dependence of date of first egg uj)on age when hegin-
lung to lay, however, is a relation to a heritable trait, since Hays (1!»'24-) has
shown age at first egg to be inherited.
/. Correlation Between Jlatchiiuj Date and Lenyth of Winter Cycle.
lime of hatching is generally believed to hold an important relation to the
time of appearance of winter pause. Since the appearance of winter pause
marks the termination of the winter cycle, the possibility exists of a relation-
siiij) between hatching date and length of winter cycle. The table presented
1 flow tends to substantiate a relation l)etween date of hatching and length
(I winter cycle for the total population of birds actually manifesting a
winter cycle terminated by a pause:
Hatches Number of Birds Mean Length of Winter Cycle
1 32i> 68.33 Days
2 267 62.54 "
3 286 59.56 "
4. 281 51.91 "
5 258 47.05 "
6 237 42.00 "
7 225 36.21 "
8 195 38.78 "
Grand Average 52.26 "
The mean length of winter cycle is shown to consistently decrease as the
hatching date advances, with but a single exception in the last hatch. There
are eight hatches each year at one week intervals from March 25 to May 15.
The total difference in age between the first and last hatches is 49 days,
while the difference in mean winter cycle length is 30 days. The ability of
the later hatched pullets to reach sexual maturity at a slightly earlier age
than do early hatched pullets (Hays, Sanborn, and James, J924) [irobably
accounts for the minor inconsistencies in the above table. The means of the
eight different hatches for the nine-year period covered by the table indicate
a rather important relationship between date of hatching and length of
winter laying cycle, which is determined by the onset of winter pause. In
this connection, the reader should bear in mind that only winter pause birds
are included in the tabulation because no winter cycle can be ascertained
in non-pause birds.
The absolute relation between hatching date and length of winter cycle
may be discovered by means of the coefficient of correlation. Available for
study are 2078 birds. Class intervals of ten days are u.sed for winter cycle
in calculating the following con.stants:
Number of birds 2078
Mean hatching date (Apr. 17) 4.18
Hatching date standard deviation .... ±2.26
Mean length of winter cycle ..... 52.26
Winter cycle standard deviation ..... ±34.23
CoeflScient of correlation ...... — .3174 ± .0133
Regression of hatching date on winter cycle length . — .021
Regression of winter cycle length on hatching date . — 4.811
170
TECHNICAL BULLETIN 8
A significant negative coefficient of correlation informs that, in general,
early-hatched pullets have a longer winter laying cycle than late-hatched
pullets of the same flock. The magnitude of the constant does not estahlish
an intimate relationship, however, and for this reason the influence of other
forces is evident. An increase in length of laying cycle is important from
the hreeding standpoint, because it signifies a greater mean winter record,
and winter production is intimately correlated with annual production (Hays,
Sanborn and James, 1924).
2. Correlation Bcitceen Date of First Ef/ff and Length of Winter Cycle.
Date of first egg is very important economically. Its significance biolog-
ically depends upon the influence of weather conditions on egg production.
Specific data concerning the influence of weather on fecundity are not avail-
able, however. There is a considerable body of evidence pointing toward a
seasonal periodicity of production which has led a number of workers to
consider winter, spring, summer and autumn cycles of laying.
In this experiment 2078 pullets with definite winter cycles are available for
study. Fifteen-day class intervals are used in grouping data of first egg,
and the range in dates is August 24 to February 20. Below are the con-
stants calculated:
Number of birds 2078
Mean date of first egg (Oct. 29) . . . . 5.93
Date of first egg standard deviation .... ±2.09
Mean length of winter cycle 52.26
Winter cycle standard deviation ..... ±34.23
Coefficient of correlation —.5307 ± .0106
Regression of date of first egg on winter cycle . — .032
Regression of winter cycle on date of first egg . — 8.689
The date of first egg fluctuates widely in the population studied. Tl»e
mean date of first egg for the 2078 birds studied is October 29. In breed-
ing for fecundity this variability in time of beginning to lay may be reduced
genetically and also by providing a more uniform environment.
A negative coefficient of correlation of substantial magnitude demonstrates
that early laying makes for a long winter cycle. The relation that winter
cycle length holds to egg production remains to be considered in sections
4 and 5 of this report.
(l)) Relation of Heritable Traits to Winter Ci/cle.
Age at first egg is a definitely heritable trait (IIay.s, 1924). It has been
shown by a number of workers to be intimately correlated with both winter
and annual fecundity. This study shows how age at first egg is related to
length of the winter laying cycle. Sexual maturity is the only heritable
characteristic reported on in relation to winter cycle.
3. Correlation Between Age at First Egg and Length of Winter Cycle.
The same group of 2078 pullets has been studied to ascertain the correla-
WINTER CYCLK AND WINTER PAUSE
171
tion between age at first egg and k-ngtli of the winter laying cycle. Class
intervals of ten days have been used for age, and the respective ages of the
individuals tabulated against their winter cycle length. The constants de-
termiiuHl ave as follows:
Number of birds ....
Mean age at first egg .
Age at first egg standard deviation
Mean length of winter cycle
Winter cycle standard deviation .
Coefficient of correlation
Regression of age on winter cycle
Regression of winter cycle on age
2078
203.66
±25.92
52.26
±34.23
— .4529± .0118
—.343
—.598
Age at first egg varies within moderate limits. Since genetically-early
and genetically-late birds are concerned, and because environment probably
modifies age and sexual maturity, tiie standard deviation of age is not ex-
cessive.
A significant negative correlation is shown, as might be anticipated from
the constants obtained in section 2. Age of sexual maturity may be classi-
fied as a characteristic influencing the length of winter cycle as determined
by the onset of winter pause. Here is an example of a heritable trait being
negatively correlated with duration of winter c\cle.
(c) Reldllon of Winter Cycle to E(fy Production.
A knowledge of the relation of winter cycle length to winter fecundity
and annual fecundity is of value for prediction purposes. If any short-time
measure of fecundity that is reasonably accurate in predicting winter and
annual production is discovered, it will be of much economic import. Proper
culling enables the poultryman to raise mean flock production by disposing
of mediocre layers. If a relatively short season of trapnesting gives a clue
to probalde jirodiiction for tiie year, such information will greatly assist
poultrymen. This section considers the correlation between length of winter
cycle and winter prtxluction and length of winter cycle and annual produc-
tion. Since the winter cycle length for each bird is tabulated against her
egg record, a true measure of degree of correlation is arrived at.
ft
4- Correlation Between Length of Winter Cycle and Winter Ei/g Record.
Both length of winter cycle and winter egg record are placed in class
intervals of ten for the 2078 indivi(hial pullets studied. The following con-
stants were determined:
Number of birds
Mean length of winter cycle
Winter cycle standard deviation
Mean winter production
Winter production standard deviation
Coefficient of correlation
Regression of winter cycle on production
Regression of production on winter cycle
2078
52.26
±34.23
56.99
±23.40
+.6538 ±.0085
-f.956
+.4-47
172
TECHNICAL BULLETIN 8
Mean length of winter cycle is 52.26 days while mean winter production is
5ti.99 eggs. Winter production exceeds the producdon of the winter laying
cycle because most of the pullets resume laying previous to March first
following a pause. An arbitrary termination of the winter season at the
close of February in all cases does not give a true measure of winter pro-
duction and no definite calendar date will suffice.
The standard deviations of both winter cycle and winter production are
excessive. This fact further establishes the variability as due to inheritance
and environment.
The above constant discloses a very intimate positive correlation be-
tween length of winter cycle and winter production. Here is concrete evi-
dence establishing an important relation between long winter cycle and liigh
winter egg record.
The importance of optimum hatching date, age at first egg, and date of
first egg in relation to the length of the winter laying cycle has been pre-
sented in sections 1, 2, and 3. Possibilities of shortening the pause period
by breeding methods are to be handled in another publication. Probably the
most important consideration is the correlation between length of winter
cycle and annual egg production, studied in t4ie next section.
d. Correlation Between Length of Winter Cycle and Annual Prodnction.
Annual egg record depends upon a vast array of environmental forces
and upon a series of Mendelian factors. Specific information concerning
many of these influences has never been presented. This section attempts to
present in concrete form the relation of length of winter cycle to annual
production over a period of years. On 131t pullets the following constants
appear:
Number of birds 1314.
Mean length of winter cycle ..... 53.52
Winter cycle standard deviation .... ±34.87
Mean annual production ...... 172.53
Annual production standard deviation . . . ±41.13
Coefficient of correlation . . . . . . ±.4027 ± .0156
Regression of winter cycle on production . . . -f" 341
Regression of production on winter cycle . . . 4.475
This group of birds averaged slightly under 173 eggs during their pullet
laying year of 365 days beginning with their first egg. The standard devia-
tion in production illustrates a wide range of fluctuation. The actual range
in annual egg production was from 35 to 275.
A positive correlation of substantial magnitude exists between winter
cycle length and annual egg record. Length of vnnter cycle, therefore, fur-
nishes a rather dependable short time measure of probable annual produc-
tion for a population. Winter cycle length is discernable only for pullets
exhibiting winter pause. By trapntsting during the first part of the laying
year, it is possible to discover tlie length of the winter cycle and conse-
quently the time of appearance of its complement, the winter pause.
WIN TKR CYCLE AND WINTKU PAUSE 3^73
\Vint?:k Pause
In the following ta!)le are given tiie number of pause and non-pause birds
by years, togetiier with tlie percentage of birds in tlie pause class.
Year Non-Pause Pause Total % Witli Pause
Birds Birds
1916 120 15«) 279 56.99
1917 153 239 392 60.97
1918 115 248 363 68.32
1919 59 109 168 64.88
1920 48 133 181 73.48
1921 276 175 451 38.80
1922 201 376 577 65.16
1923 109 353 462 76.41
1924 160 340 500 68.00
Tiie average length of j)ause for those birds pausing is as follows:
1916 flock 29.46 days
1921 flock 19.23 days
1924 flock 32.97 days
6. Correlation Between Length of Winter Cycle and Length of Winter Pause.
The range in cycle length was found to be from 5 days to 175 days. The
class interval used was 10 days. Since winter cycle and winter pause are
complementary to each other, it is important to discover their possible rela-
tion to each other. The coefficient of correlation will illustrate any tendency
for length of winter cycle and length of winter pause to move in the same or
in opposite directions. Such information will make clearer their physiological
relationships and possible genetic linkage.
Xuiiibcr of birds ......
Mean length of winter cycle
Length of winter cycle standard deviation .
Mean length of winter pause
Length of winter pause standard deviation
Coefficient of correlation ....
Regression winter cycle on pause
Regression pause on winter cycle
2078
52.26
±3t.23
31.91
±21.68
—.1385 ± .014.5
—.219
—.088
The ler>gth of winter cycle as measured in the.se studies is sul)ject to wide
fluctuation as indicated by the relative magnitude of the mean and its stand-
ard deviation. Such fluctuations are to be anticipated in a population highly
variable for the seven pairs of inherited factors concerned in winter produc-
tion, subjected to uncontrollable variation in environmental influences. The
fact should be observed, however, that the mean length of winter cycle is
about 63 per cent greater than the mean length of pause.
A small negative correlation coefficient indicates a very slight tendency for
long-cycle birds to pause for a shorter period than do short-cycle birds. This
7 74 TECHNICAL BULLETIN 8
constant is subject to error, however, in that any increase in length of winter
cycle reduces the possible pause interval before March first. The coetticient
of correlation as calculated is of value in that it gives the actual relationship
between length of winter cycle and duration of pauses within the winter
season. These data furnish very good evidence that the length of winter
pause does not depend upon the length of the winter cycle of laying.
In this report consideration is given to the winter pause from three gen-
eral standpoints, namely, (a) Environmental factors affecting duration of
the pause; (b) Inherited characteristics concerned with fecundity in relation^
to winter pause, and (c) The absolute relationship of winter pause to egg
production.
(a) Environmental Factors Alfectiny Duration of Pause.
Much concern should be given to the relation of environmental factors
affecting the duration of the winter pause since these conditions are more or
less under the control of the poultryman. In the group of environmental
factors the following have been placed: hatching date and time of beginning
to lay in the fall. The time of year when pullets begin to lay is clearly de-
pendent both upon management and inheritance. Management is a factor
when the hatching date remains constant because housing, range, and feeding
may either retard or accelerate sexual maturity. Just how significant these
environmental influences are on time of beginning to lay in comparison with
inherited early or late sexual maturity remains to be determined. At any
rate, hatching date can be definitely controlled and time of beginning to lay
may be considered as partially controllable.
7. Correlation Between Hatching Date and Length of ]Vint(r Pause.
A very common observation among poultrymen is that early-hatclied pulletg
arc more likely to exhibit winter pause than are late-hatched birds of the same
dock. In other words, the belief is prevalent that the earlier the hatching-
date, the longer the winter pause. Such observations have naturally led to
the assumption that the pullet possesses capacity to lay a certain number of
eggs in the fall and winter and if this number is laid early there will be a
cessation of laying until the spring season. That hatching date is only one
of several conditions operating to aifect the onset and duration of winter
pause has been shown by Hays (1924) and Hays, Sanborn, and James (1924).
Age at sexual maturity has been pointed out as an inherited characteristic,
and as a characteristic having greater effect than hatching date upon winter
egg production. Furthermore, winter pause of seven or more days' dura-
tion is an inherited characteristic. The importance of knowing just how
intimate a relationship exists between date of hatching and duration of
winter pause becomes apparent and may be discovered by means of the
coefficient of correlation.
A total of 2134 birds exhibited a pause of four or more days and are in-
cluded in these calculations. The winter pause class interval is ten days in
all cases. Constants obtained from this study follow:
WINTER CYCLE AND WINI'ER I'AUSE
175
Number of liirds ....
Mean liatehing date (Apr. 17)
Hatdiinjr date standard deviation
Mean length of winter pause
Winter pause standard deviation
Coeftieient of correlation
Regression of hatching date on winter pause
Regression of winter pause on hatching date
2iai
i.l5
It -2.26
32.26
±21.92
—.2480 ± .0137
r pause
—.026
ng date .
—2.404
The mean length of the winter pause over the entire nine-year period
amounts to 32.26 days for the pause birds, but a striking fluctuation in the
duration of the pause is revealed by its standard deviation. Environmental
influences may be considered as largely responsible for the fluctuations. Any
jiossible changed environment to shorten the pause would be advantageous
economically.
The coefficient of correlation measuring the degree of association between
time of hatching and duration of winter pause is negative, of moderate
magnitude, and certainly significant. Clearly, a reduction in the range of
hatching dates would tend to reduce the length of the period of non-produc-
tion during the winter season. The mean length of winter pause of the eight
different hatches studied follows:
Hatches
I
2
3
4
5
6
7
8
Grand Average
No. of Birds
353
271
293
289
261
245
227
195
Mean days Paused
39.41
37.35
36.46
33.84
28.22
25.99
25.85
24.32
32.26
S. Correlation ]Jeticeen Hatching Date Earlier Than The Mean and the
Presence of Winter Pause For Entire Population.
Yule's short method as cited by Davenport (1907) is used in this study.
This tabulation includes the total population, 337.5 Rhode Island Reds classi-
fied as pause and non-pause individuals.
Hatching Date
Pause
Non-Pause
Earlier than Population mean
1206
555
Later than Population mean
928
686
Totals
2131
1241
Coefficient of correlation
-f.2326± .0110
176
TECHNICAL BULLETIN 8
This positive coefficient of correlation is of sufficient magnitude to establish
a definite relationship between early hatching and the appearance of winter
pause. This being the case, the assumption must be made that inheritance is
not the sole controlling force concerned in tlie manifestation of winter pause.
0. Correlation Bet^ceen Dote of First Eyg and Length of Winter Pau^e.
I'he date on which a pullet lajs her first egg is dependent upon many
factors. Among the most important of these are hatching date and age at
first egg. Environmental influences such as character of ration, amount of
free range, and weather conditions may, to some extent, hasten or retard
the date of first egg. Date of first egg is important economically if not
biologically.
The relation between date of first egg and duration of winter pause has
been determined by means of the coefficient of correlation on the same group
of 2134 birds studied in section 7. The birds were again classified as to
winter pause into class intervals of ten days. The class interval used for
date of first c^g was fifteen days. Constants calculated are as follows:
213t
5.88 •
±2.13
32.26
±21.92
—.320.5 ±
.0131
—.031
—3.297
Number of birds
Mean date of beginning to lay (Oct. 29)
Date of beginning to lay standard deviation
Mean length of winter pause ....
Winter pause standard deviation
Coefficient of correlation
Regression of date of first egg on winter pause .
Regression of winter pause on date of first egg .
The date on which tlie birds began to lay ranged from August 16 to March
29 making 15 class intervals. Its standard deviation may be expected to be
of considerable magnitude in relation to the mean as is shown above.
A negative coefficient of correlation of .3205 ±.0131 between time of be-
ginning to lay and pause duration stresses an important relation between
the two. Early laying, on the average, tends to increase the duration of tlic
pause.
10. Correlation Betiveen Time of Beginning to Lay Earlier Than the Mean
and the Presence of Winter Pause for Total Population.
Time of Beginning to Lay in tiie Fall Pause I Non-Pause
Earlier than Population mean ; 1210 I 456
Later than Population mean i 924 I 785
Totals 2134 1241
Coefficient of correlation -t-.3854 ± .0099
WINTER CYCLE AND VVINIEU 1>ALISE 1 77
A definite and sipnifieant correlalion exists i)ctween early laying and tlie
presence of winter i)ause. Ihis tact snjrgests tiie importance of breeding for
a specific age at first egg, and liatcliing on some special dale to meet condi-
tions of environment.
(b) Inherited I'haracterislks Concerned With Fecundili/ In
Relation to Winti'r Pause.
In tiie category of inherited fecundity traits tiiat may be considered in their
relation to winter pause, tb.e following may be grouped: age at first egg,
weight at first egg, winter rate or intensity, length of winter cycle, size of
winter clutch, annual rate or intensity and annual persistency. A study of
the relative degree of correlation l)etween these inherited characteristics and
duration of winter pause as well as its presence or absence furnishes con-
structive information in breeding for high egg yield. Such analyses bring
out important relationships as well as pointing out possible cases of genetic
linkage.
]1. Correlation Between Aye at First Egg and Length of Winter Pause.
Age at first egg marks sexual maturity in the pullet. Age at first egg is
inherited in Mendelian fashion according to Hays (loc. cit.). The importance
of early laying to high winter and annual egg yield has been stressed in our
publications as well as in those of other workers. The significance of know-
ing if there is a correlation between age at first egg and duration of winter
pause is therefore very evident, since both are inherited traits and both are
concerned in winter and annual egg yield. A study was therefore made on
the 2134 pause birds already considered in sections 7 and 9. Age at first
egg class intervals of ten days are used and the same class interval used for
length of winter pause. The following constants were calculated:
Number of birds 2134
Mean age at first egg 203.26
Age at first egg standard deviation .... ±26.28
Mean length of winter pause ..... 32.26
AVinter pause standard deviation .... ±21.92
Coefficient of correlation — .2329 ± .0138
Regression of age at first egg on winter pause . . — .279
Regression of winter pause on age at first egg . . — .194
The above coefficient of correlation is almost identical with that between
hatching date and winter pause duration given in section 7. The range in
hatching date covers 49 days, while the range in age at first egg covers 180
days. The fact therefore becomes evident that a slight change in hatching
date would cause a greater change in winter pause duration than would the
same change in age at first egg, as brought out by their respective regres-
sion coefficients. Herein lies the reason for emphasizing hatching date as of
greater significance in relation to winter pause than age at first egg when
they exhibit identical coefficients of correlation to winter pause durntion.
178
TECHNICAL BULLETIN 8
J2. Correlation Between Age at First Eyg Beloiv the Mean and the Presence i
of Winter Pause for the Total Population.
Age at first egg
Pause
1
1 Non-Pause
1
Earlier than Population mean
1337
549
Later than Population mean
797
692
Totals
2134
1241
Coefficient of correlation
.3578 ± .0101
Attention should be called to the fact that time of beginning to lay and
age at first egg each show almost identical correlation coefficients to the pres-
ence of winter pause. The inter})retation is that age at first egg is the chief
determinant of time of beginning to lay when the hatching dates are constant
from year to year.
13. Correlation Between Weight at First Egg and Length of Winter Pause.
Available for this study are the records of 2106 birds, classed as pause
birds, on which the body weight on tiie day of laying their first egg was
secured. Thus a very small number of the 2134 birds previously considered
is omitted from this study. The class interval used for body weight was the
half pound and the ten-day class interval was again used for winter pause.
The following constants were determined:
Number of birds
Mean weight at first egg
Weight at first egg standard deviation
Mean length of winter pause
Winter pause standard deviation
CoefRcient of correlation
Regression of weight on winter pause
Regression of winter pause on weight
2106
5.55
±.72
32.32
±22.01
-(-.0161 ±.0147
+.0005
-f.4908
A range in body weight from 3.25 to 8.25 lbs. occurs in the population
studied. The magnitude of the standard deviation in weight indicates, how-
ever, that extremely small or extremely large birds are the exception, since
the coefficient of variability for body weight is only about 13 per cent.
The coefficient of correlation between body weight and winter pau.se dura-
tion is mathematically insignificant. This furnishes rather concrete evidence
that a pullet's body weight when she lays her first egg bears no relation to
the length of her winter pause.
WINTER CYCI.K AND WINTKK i'AL'SK
170
14. Vorrelat'uin Betxceen Body WeUjht at First E<j<j Loic'er Than tin- Mean
and the Preaence of Winter Panne for Entire Population.
Coefficient of t-orrelation
Wfiiilit at First Egg
Pause
Non- Pause
Below i'opulation Mean
1139
655
Above Population Mean
967
544
Totals
1
1 2106
1199
.0110 zii .0117
The complete independence between weight at first egg and tiie presence
(»f winter pause is shown by the above correlation coelticient. Evidently
body weight is not a factor in either the manifestation of winter pause or
its duration.
JS. Correlation Between Winter Rate and Length of Winter Pavne.
The group of 2131- birds exhibiting winter j)ause is used in these calcula-
tions. Winter rate or intensity was calculated for each individual bird in
the following manner: —
The total number of eggs from first egg to March first was divided by the
number of days from first egg to March first, less all pauses of four or more
days in duration from November first to March first. By this method of
calculation the actual net rate of laying is arrived at if the assumption is
correct that a cessation of laying for four or more days during winter actu-
ally constitutes a winter pause. A four-day cessation of laying may gener-
ally be assumed to necessitate the omission of one clutch of eggs for the
average bird and such omissions suggest the manifestation of winter pause.
The following constants were calculated:
Number of birds ....
Mean winter rate ....
W^inter rate standard deviation .
Mean length of winter pause
Winter pause standard deviation
Coefficient of correlation
Regression of rate on winter pause
Regression of winter pause on rate
2134
65.69
±8.74
32.26
±21.92
—.1023 ± .0144
—.041
—.257
The above mean winter rate expresses the net rate of laying of all birds
exhibiting winter pause. This rate of laying is compared in section 16 with
that of the total population and that of the non-pause group above. The
standard deviation for rate is of moderate magnitude compared with the
standard deviation of many other fecundity characteristics.
A small but significant negative correlation suggests a very moderate tend-
ency for high-rate birds to pause for a shorter period than do low-rate birds.
Such a relationship is important from the breeding standpoint in that it
180
TECHNICAL BULLETIN 8
hints at some linkage relation between the dominant genes for iiigii winter
intensity and the recessive gene for non-pause.
16. Correlation Between Winter Rate Beloic the Mean and the Presence of
Winter Pause for the Total Population.
Winter Rate
1 Pause
1 Non- Pause
Below Population Mean
1 1688
1 784
Above Pcjpulation Mean
1
446
1 454
Tetals
1
j 2134
1 1238
Coefficient of correlation
+.3734 lii .0100
Winter rate as used in all the calculations is tiie net rate of laying with
all pauses of four or more days deducted. The above table shows the rela-
tion of net rate of laying to the presence of winter pause. This table dis-
plays a moderately intimate relation between low net rate and the presence
of winter pause.
17. Correlation Between Size of Winter Clutch and Lenf/th of ]Vinter Pause.
Size of clutch represents the number of eggs laid on successive days. In
very extreme cases a pullet may lay as many as hfty eggs in succession pre-
vious to March first and the same bird may exhibit a few clutches of one.
In order to arrive at a constant to represent the clutch size of an individual
bird, it has been necessary to calculate mean clutch size during the winter.
Such calculations have been made on ail pause birds. The range in mean
clutch size of individuals was found to be from 1 to 11.9. The class interval
used was 1. Only one bird was omitted from this study because its class
range fell between 1-5 and 15.9. CluLch size is really a measure of intensity
of laying. Its relation to winter pause duration is of marked significance in
breeding for fecundity.
Constants obtained in this correlation studv ure as follows:
Number of birds ....
Mean winter clutch size
Winter clutch standard deviation
Mean length of winter pause
Winter pause standard deviation
Coefficient of correlation
Regression of winter clutcli on winter ])ause
Regression of winter pause on winter clutch
2133
2.41
±1.11
32.27
±21.92
—.0674 ± .0145
—.003
—1.325
On the average, winter clutch size closely approaches 2.5 but the magni-
tude of its standard deviation indicates considerable variability in clutch size.
A small negative correlation was discovered between clutch size and winter
pause duration. While this correlation is significant as judged by its prob-
WINTER CYCLE AND WINTER TAl^SE
181
able error, it is of such small ina<;nitude .as tt> indicate praclical iiKicpciid-
cncc between the characteristics beinK considered.
/<"i. Correlation Between Winter Clutch Size Beloic the Mean and the Pres-
ence of Winter Pause for Entire Population.
Winter Clutch
Pause
Non-Pause
Below Population Mean
1425
616
Above Population Mean
709
1
1 624
Totals
1
1 2134
1 1240
1
Coefficient of correlation
+.3412 ± .oias
A significant positive correlation between small winter clutch and the pres-
ence of winter pause appears above. In general, there is a greater tendency
for birds that lay in small clutches to pause than for birds laying in larger
clutches. The rate of functioning of the reproductive system" must therefore
bear a relation to winter pause.
19. Correlation Betzceen Annual Rale or Inlensity and Length of Winter
Pause.
Annual rate represents or approximates the intensity of each individual
bird for the pullet laying year. Inasmuch as this constant has been discussed
in Technical Bulletin No. 7 of this station, space will not be occupied here
by further discussion. Since winter pause represents a period of non-pro-
duction, there must of necessity exist a negative correlation lietwcen annual
rate and length of winter pause unless pause birds lay at a higher net rate
than non-pau.se liirds. This la.'^t point is, in part, discussed in section 16 of
this report, where the net winter rate of the total population in relation to
the pause and non-pause groups is considered. The important positive rela-
tion between annual rate and annual egg record makes the correlation be-
tween annual rate and length of winter pause of importance. Included in
this study are the 1348 birds exhibiting winter pause and having complete
annual records. The following constants were determined:
Number of birds
Mean annual rate ......
Annual rate standard deviation .
Mean length of winter pause . .
Winter pause standard deviation
Coefficient of correlation ....
Regression of annual rate on winter pause
Regression of winter pause on annual rate
1348
53.79
±9.07
32.29
= 21.77
—.4091 ±: .0153
—.170
—.982
182
TECHNICAL BULLETIN 8
The mean annual rate of laying is lower than the mean winter rate of
laying, wliich is 65.69. This difference may be attributed largely to the fact
that in calculating annual rate no account is taken of winter pause or of
broody pauses. In the winter rate calculations, winter jjause days are not
included and very few birds become broody before the end of the winter
season. The standard deviation in annual rate is relatively small and suggests
uniformity in annual rate of laying.
The coefficient of correlation is negative and of sucli magnitude as to indi-
cate a significant relation between rate and length of pause. In other words,
low annual rate and long winter pause tend to move together. In breeding
for high annual intensity, winter pause nnist certainly be reduced in duration.
20. Correlation Betxaeen Annual Kate or Intensity Below the Mean and the
Presence of Winter Pause for the Total Population.
Annual Rate
1 Pause
1
1
1 Non-Pause
1
Below Population Mean
i
858
1
I 209
i
Above Population Mean
490
1
1 593
Totals
1348
1 802
Coefficient of correlation -)-.6649 ±: .0081
The sul^stantial magnitude of the above coefficient of correlation points to
a, pronounced tendency for low annual rate to occur with winter pause. The
table above also sliows that 80 per cent of the low-rate birds are pause birds
while only 45 per cent of the high-rate birds are in the pause group. The
conclusion, therefore, seems justified that winter pause operates very sig-
nificantly to lower tlie annual rate of laying.
21. Correlation Betwetn Annual Persistencif and Len(jth of ]\'inter Pause.
Annual persistency represents the nimiber of days of laying from tlie first
egg to a pause of thirty or more days after March first. If no thirty-day
pause occurs between March first and the date 364 days after the first egg,
the bird is given a persistency of 365 days on ordinary years and 366 days
on leap years. A cessation of laying for a period of thirty days or more
during summer is a rather dependable indication of the onset of complete
molt, which always signifies the conclusion of the biological laying year.
Persistency as indicated by time of molting has long been recognized as
affecting egg yield, and poultry investigators have recommended the use of
late molting birds for breeding purposes. Hurst (1921) was the first to
offer a definite hypotliesis concerning its mode of inheritance. He believes
high persistency is transmitted as a single factor recessive. If a rest period
in winter enables the bird to lay later in the fall than does the bird without
the rest periwl, then persistency must depend in part upon the previous
physiological activity of the reproductive organs, or possibly there is linkage
between winter pause and high persistency. The same group of 1348 birds
used in the two previous sections is studied lielow. Persistency range lies
between 67 and 366 days with class intervals of 15 days. Following are the
constants:
WINTER CYCLE AND WINIKU TAUSK
I8:i
Nuiiil)or of birds ......
Me.in .iniui.il ))ersisti'iu-y . . . .
Annual persistency standard de\ iation
Mean len<rth of winter pause
Winter pause standard deviation
Coefficient of correlation ....
Regression of persistency on winter pause
Regression of winter pause on persistency
1348
3<)!).0.'J
±. 54.-89
32.39
±21.77
-i-.l()17
+.256
-I-.040
.0182
Mean annual persistency closely approaches ten months, hut the range of
\ariahility is rather wide as shown hy its standard deviation. This vari-
ability is no doubt due in part to many environn>entaI influences as well as
to dit^'erences in the inherited capacities of the birds. Only about five per
cent of the population fall below 200 days in persistency so that the range
200 to 3G6 is a close approximation of the actual range. A study of fre-
quency distribution for persistency does not reveal a bimodal curve as might
be expected for a population made up of genetically early and late molting
birds. Such information suggests two possibilities, namely, Ihat environ-
mental influences completely obscure the genetic phenotj'pes, or else that
high persistency is not inherited in simple Mendelian fashion. The mode of
inheritance of persistency is out of the scope of this report.
A small but significant positive correlation coeflicient exists between per-
sistency and winter pause duration. Thus tiiere is a very slight tendency for
birds with long winter pau.se to lay later in the fall than do short pause
birds. Relatively little significance should be attached to a constant of such
small maenitude, however.
32. Correlation Between Annual Perttislevcy Greater Than the Mean and
the Presence of Winter Pause for the Total Population.
Annual Persistency
I Pause I Non- Pause
I I
Above Population Mean
85.5 1
423
Below Population Mean
493 1
1
378
Totals
i 1
1 1348 1
801
Coefficient of correlation
-f .21.56 It .0139
A moderate degree of correlation is shown between the presence of winter
pause and high persistency. There is thus a slight tendency for pau.se birds
to lay later in the fall than do non-pause birds Po.ssibly the functional abil-
ity of the reproductive organs is somewhat extended by a period of non-
production in winter. The relation does not appear to he pronounced, how-
ever.
Ij^^ TECHNICAL BULLETIN 8
(c) The Absolute Relationship of Winter Pause to Egp Production.
The duration of winter pause may he considered a factor affecting the
number of eggs laid before March first as well as for the entire year. Since
winter fecundity alone depends upon the inheritance of seven pairs of
Mendelian factors (Hays, 1921.), it is desirable and necessary to know some-
thing of the relation of winter pause to winter and annual egg record. Al-
though fecundity is very complex in its mode of inheritance, its manifesta-
tion depends in part upon environmental conditions as division (a) of this
report shows. The correlation between size of winter clutch and winter egg
yield is first considered, then the correlation between winter pause and winter
egg record, and finally the correlation between winter pause and annual pro-
duction is studied.
2S. Correlation Bet-iveen Size of Winter Clutch and Winter Egg Yield.
The relation of winter clutch size to winter pause has already been con-
sidered in sections 17 and 18. In this section the relation of winter clutch
size and winter egg production are studied. Since size of winter clutch is so
often used as a criterion for selection by poultrymen, knowledge of its rela-
tion to winter fecundity is important. Records are available on 3376 birds
upon which the following constants were ascertained:
Number of birds 3376
Mean size of winter clutch ...... 2.51
Winter clutch standard deviation .... ±1.23
Mean winter production ...... 61.08
Winter production standard deviation . . . ±2.5.79
Coefficient of correlation ...... -)-. 472.5 ± .0090
Regression of winter clutch in production . . -}-.023
Regression of production on winter dutch . . -)-9.884
The fact will be observed that tlie mean winter production above is greater
than the mean length of winter cjcle given in section 6. The mean length of
winter cycle is less than mean winter production because the end of the
winter cycle is determined by a four-day pause before Marcii first while
winter egg record does not cease imtil February 28 or 29. Winter egg rec-
ord is highly variable on account of the complexity of its inlieritance.
The magnitude of the above correlation coefficient emphasizes an important
tendency for clutch size and winter production to move together. As a cri-
terion of winter fecundity large clutch size is very important.
2^. Correlation Betxveen Length of Winter Pause and Winter Egg Record.
Winter pause represents a definite period of non-production, but the tend-
ency of winter pause and winter production to move in o})posite directions
<an only be measured by means of the coefficient of correlation. Tlie group
of 2134' pause birds has been tabulated to give this relationship. The follow-
ing are tlie constants obtained:
\i
WIXTKH CYCLE AND WlNl'I'.K
Number of birds
Mean lengtb of winter pause
Winter pause standard deviation
Mean winter prodiiotion
Winter production standard deviation
CoelRcient of correlation
Regression of winter pause on production
Regression of production on winter pause
iwrsi',
18
2134
32 26
±21.92
56.8T
±23.51
—.2873
± .0134
-.268
—.308
Mean winter productioi\ is lower in tiic above group of pause birds tban
for tbe total population given in section 23 because non-pause birds tend to
have higher winter records than do pause birds. About the same degree of
variation in winter records occurs in l)oth ca.ses.
A rather significant negative correlation coeflicient shows tluit in general
an increase in length of pause is associated with a decrease in number of
winter eggs. A coefiicient of much greater magnitude would appear if the
time element were the only consideration. There is the possibility that pause
birds tend to possess desirable fecundity traits that are lacking in non-
pause birds.
25. Correlation llelween W^intcr Production Belou' the Mean and Presence
of Winter Pause for Total Population.
Winter Production
1
1 Pause
1
1
1 Non-Pause
1
Below Population Mean
1
1 1273
1 463
1
Above Population Mean
1
861
1 777
1
Totals
1
2134
1
1240
Coefficient of correlation
-f .4.255 ± .0095
The above table illustrates a rather pronounced correlation between low
winter egg production and the presence of winter pause. Winter pause lias,
therefore, proven to be a trait inimical to high winter egg record throughout
the nine-year period of tlie experiment here re])orled.
26". Correlation Between Leni/th of Winter Pause and Annn(d Production.
There are available for study 1348 pause birds with annual egg records.
Tabulations have been made to discover how the length of winter pause
affects annual egg production. Following are the constants:
186
TECHNICAL BULLETIN 8
Number of birds
Mean lengtli of winter pause
Winter pause standard deviation
Mean annual production . . . .
Annual production standard deviation
Coefficient of correlation ....
Regression of winter pause on production .
Regression of production on winter pause
1348
32.39
±2L77
172.51
±4.1.07
—.2107
—.112
—.398
.0176
Tbe mean annual record of the pause birds tiiroughout the period is about
17B eggs. Tlie range of variation in annual egg yield is wide, as is shown bj"
its standard deviation. Greater honuigeneity in iieritable factors concerned
in fecundity should reduce such varialiility.
The magnitude of the coefficient of correlation is sufficient to indicate that,
in general, an mcrease in length of winter pause is accompanied by a de-
crease in annual egg production. The time lost in winter pause is not com-
pensated for by heavier production either before or after the pause in any
class of pause birds.
m. Correlation Befzveev ^nwni] Production BeloTi' the Mean and the Pres-
ence of Winter Pause for the Tolid Population.
Annual Production | Pause | Non- Pause
1 1
Below Population Mean 715 | 301
1
Above Population Mean 633 501
Totals 1 1348 i 802
1 1
Coefficient of correlation
+.3056 ± .0132
Low annual production is significantly correlated with the presence of
winter pause as shown in the above table. Even thougli such a short period
as a four-d;\y pause is considered, this correlation coefficient is of appreci-
able magnitude. Winter pause nuist, therefore, be classed as inimical to
highest annual egg yield, for the pause birds averaged but 173 eggs while
the non-pause group averaged 189 eggs.
Generaf, Discrssiox axu Su.mmary.
The lenglh of the winter laying cycle is unquestionably modified by a
series of environmental influences. Some of tliese influences are within
while others are beyond control of the poultry breeder. Winter pause is the
complement of the winter laying cycle and is important in that it vitally
alTects total fecundity.
Two distinct classes of pullets appear in the flock studied, namely, pause
and non-pause. A group of pause birds studied i)eside a group of non-
pause birds, both groups hatched on the same date and both groups starting
WINTER CVri.F, AMI WINTER I'AUSE
18'
til hiy at tlic same age, yet the first slidwin/i a distiiu'l winter iiause, jilai'cs
lilt' clifTereiiee in tlie groups as iiiliereiit. Siuli a study wltliiu tlie i'ainily
unes (ielinite ratios of pau.'^c and non-]>ause ]Mdlels as Mays (lf>21')
l>i>ints out.
riu- line of deinarkation between genetically non-pause birds thai exhiliit
winter pause due to environmental influences and birds earryinji' the dominant
Tutor (M) for pause cannot be drawn. The i)re.sent paper is devoted to a
( unsideration of the non-heritable and some heritable factors that may or
may not affect winter cycle and winter pause. fJenetic factors concerned
with the inheritance of winter cycle length and winter pause duration have
not been dealt with.
The major teachings of this study may be summarized:
1. In general, early-hatched pullets have a longer laying cycle than late-
hatched pullets of the same flock.
:.'. Date of first egg exhibits a rather intimate negative correlation to
length of winter cycle.
;J. Age at first egg shows an appreciable negative correlation to length of
winter cycle.
I. The winter egg record is intimately positively correlated witii length of
winter cycle.
5. Annual egg production is significantly correlated with length of winter
cycle though less intimately than is winter record.
6. A minor though significant degree of negative correlation appears be-
tween length of winter cycle and length of winter pause.
7. Hatching date bears a significant but not intimate negative correla-
tion to length of winter pause in the pause population.
8. Early hatching is jiositively correlated with tiie presence of winter
pause in the total population of pause and non-pause birds.
n. Time of beginning to lay is significantly negatively correlated with
length of winter pause in the pause population.
10. Time of beginning to lay is appreciably positively correlated with
early hatching in the total population.
II. Age at first egg shows the identical degree of negative correlation to
length of winter pause that it shows to lcn(jih of winter cycle.
V2. Early sexual maturity is positively correlated with the presence of
winter pause in the total population.
13. Weight at first egg is independent of length of winter pause.
14. Light weight at first egg is not correlated with the presence of winter
pause in the total population.
1.5. Tlie net winter rate of laying holds a very slight negative correlation
to length of winter pause in the pause population.
16. Slow rate of winter layina: is rather intimately jxisitivelv correlated
with the pre.'^enee of winter jtause in the total population.
17. The average size of winter clutch is iuit \ery slightly correlated with
length of winter pause in the jiause pi>|uilali(in.
18. Small size of winter clutch is moderately positively correlated with
the presence of winter pause in the total population.
188
TECHNICAL BULLETIN 8
19. Annual intensity shows a considerable degree of negative correlation
to length of winter pause.
20. A very intimate positive cori'elation exists between low annual in-
tensity and the presence of winter pause.
2L Annual persistency is hut slightly positively correlated with length
of winter pause.
22. Birds that pause during winter slu)w a tendency to lay later in the
fall than non-pause birds.
23. The mean size of winter clutches is rather intimately positively cor-
related with winter fecimdity. Clutch size is a very good measure of in-
tensity.
21'. Length of winter pause is negatively correlated with winter pro-
duction.
25. Low winter production exhibits a considerable degree of correlation
to the presence of winter pause in the total population.
26. Length of winter pause is negatively correlated with annual egg
record in the pause population.
27. Annual production heh'W the mean is substantially correlated with
the presence of winter pause in the total population.
28. Although winter cycle and winter pause are complements of each
other, they are practically independent in duraiion in the pause group.
29. Winter pause is definitely shown to lie a characteristic detrimental to
both winter and annual fecundity, and should tiierefore be eliminated from
flocks bred for epg production.
References
Davenport, E. 1907. Principles of Breeding, p. 471. Boston, New York,
etc.: Ginn & Co.
Goodale, H. D. 1918. Winter cycle of egg production in the Rhode
Island Red breed of the domestic fowl. Jour. Agr. Res. XII.: 547-574.
Goodale, H. D. 1918. Internal factors influencing egg production in the
Rhode Island Red breed of the domestic fowl. Amer. Nat. 52:65-94,
209-232, and 301-321.
Goodale, H. D., and Ruby Sanborn. 1922. Changes in egg production in
the station flock. Mass. Agr. Exp. Sta. Bull. 211.
Hays, F. A. 1924. Inbreeding the Rhode Island Red fowl with special
reference to winter egg production. Amer. Nat. 58:43-59.
Hays, F. A., Ruby Sanborn and L. L. James. 1924. Correlation studies
on winter fecundity. Mass. Agr. Exp. Sta. Bull. 220.
Hays, F. A., and Ruby Sanborn. 1926. Broodiness in relation to fecun-
dity in the domestic fowl. Mass. Agr. Exp. Sta. Tech. Bull. 7.
Hurst, C. C. 1925. Experiments in Genetics. Cambridge: Cambridge
University Press.
Pearl, Raymond. 1912. Tlie mode of inheritance of fecundity in the
domestic fowl. Papers from Maine Biol. Lab. 37.
Massachusetts
Agricultural Experiment Station
TECHNICAL BULLETIN No. 9 OCTOBER, 1926
ANNUAL PERSISTENCY
IN RELATION TO
WINTER AND ANNUAL EGG PRODUCTION
By F. A. HAYS and RUBY SANBORN
This bulletin is the third in a series dealing with the five inherited traits
in relation to fecundity. Those already published shew the relation of
broodiness and of winter pause to egg production; while a later publica-
tion will consider intensity in relation to fecundity.
The records show that high persistency is a trait much to be desired
from the standpoint of production, and that there is no reason why it
may not be combined with other desirable traits in the same individual.
Requests for bulletins sliould he addressed to the
AGRICULTURAL EXI'KIIIMKNT STATION
AMHERST, MASS.
ANNUAL PERSISTENCY IN RELATION TO
WINTER AND ANNUAL EGG PRODUCTION
By F. A. HAYS and RUBY SANBORN
Annual persistency, as terminated by the onset of complete molt, has been
emphasized for more than the past three decades as of marked significance in
the selection of breeding females for egg production. The cessation of egg
production in summer or fall is generally accompanied liy a complete change
of plumage and this period of non-production may continue for 30 to 120
days. The exceptional hen may lay a considerable number of eggs while
molting, but such individuals are of infrequent occurrence.
Hurst (1925) classifies laying hens into complete and partial-molt classes
and states that there is complete cessation of laying in the first class while
the second class sheds its feathers gradually and continues to lay for 13 or 14
months after the first pullet egg. According to Hurst, complete early molt
depends upon the inheritance of a dominant Mendelian gene.
Goodale and Sanborn (1922) note that cessation of production in the summer
or fall at the end of the pullet laying year has a genetic foundation as indi-
cated by the beliavior of families in this respect. Data collected on the
Massachusetts Agricultviral Experiment Station flock of Rhode Island Reds
show that the biological laying year may extend to 14 or 1.5 months as a
maximum with 6 or 7 months as the minimum for normal birds. A study of
all factors affecting the duration of the pullet laying year in the flock in
question has not yet been completed.
A flock bred for egg production should theoretically consist of two general
classes of birds with respect to persistency, namely, a high persistent class
and a low persistent class. In reality these two classes do not stand out
distinctly to form a bimodal curve when all the birds with annual records for
the nine-year period are tabulated in persistency classes using 1.5-day class
intervals. (See chart 1.) The proliability exists, however, that environmental
forces largely obscure these expected classes. A tabulation of the 2179 birds
with annual persistency records does give a frequency distribution that is
indistinctly bimodal and furnishes the basis for classification of those birds
laying for a shorter period than 315 days as low in persistency and those
laying for 315 days or longer as high in persistency. Such a classification is
largely arbitrary', however, and is used in these studies only as a working
basis until tlie true genetic point of division may be discovered.
Scope of This Rkpokt
This study was undertaken for a three-fold purpose, namely, to show (a)
tiie relation between controllable environmental conditions and persistency,
(I)) tlie relation between inherited characteristics concerned with fecundity
and persistency, and (c) the relation between persistency and fecundity.
From the practical breeding standpoint these considerations are of great im-
ANNUAL PERSISTENCY AND EGG rRODUCTION
191
r^ lo > «^v ^
^ c\j c^ t\^ W
"^M <^i csj
be
S
J
a*
<
X
o
Sc7<!//ff JO d3dN/]N
-.go TECHNICAL BULLETIN 9
portance. A knowledge of the part played by environment as well as the part
played by inheritance in limiting the manifestation of a desirable character-
istic is well worth considering when breeding for egg production. Analytical
studies on the complex nature of fecundity should further disclose valual)le
information that might otherwise be obscured.
Character of Birds UsEn
The birds used in this study are identical with those used in the two previ-
ous reports. Included are the records of all Rhode Island Red females
hatched from 1916 to 1924 in the experimental flock upon which pullet-year
trapnest records are available. The flock each year is made up of all the
daughters of each hen whose progeny was retained. The major portion of
the birds belong to the fecundity experhnent. There are, however, a limited
number of birds bred for non-broodiness, some for intense broodiness, some
for hatchability, some for color, and a few inbreds that are included. Pullet-
year records are used exclusively in this study.
The CoEFnciEXT ot Correlation
The simple coefficient of correlation is subject to certain limitations in bio-
logical data yet it afl'ords a basis upon which to select groups of breeders and
also a basis for predicting future possibilities. Tlie fact is self-evident that
the simple coefficient of correlation is not an absolute measure of the degree
of association between the variables being studied because each variable may
be partially dependent upon other variables. For example, section 3 shows a
negative correlation of .6146 ±: .0090 between age at first egg and annual
persistency. It is a known fact that both the dependent variable, age at first
egg, and the independent variable, annual persistency, are dependent upon
hatching date and environmental conditions, and that persistency is also de-
])endent upon weight at first egg and possibly upon length of winter pause.
The true relation between age at first egg and annual persistency could only
be arrived at by making the hatching date and environmental conditions con-
stant, as well as by making weight at first egg and winter pause duration
constant. Such procedure necessitates the use of partial coefficients of cor-
relation which require the use of the simple coefficient in their calculation.
Both the partial coefficient and the multiple coefficient will be employed in a
concluding bulletin of the series. Simple correlations are of very significant
practical value to the poultryman, however, in that they show him the relative
importance of different environmental influences and inherited character-
istics in relation to fecundity, and enable him to formulate his breeding pro-
gram accordingly.
(A) Relation Between Environmental Conditions axd Persistency
The only controllable environmental condition that will be considered in
relation to persistency is hatching date. Extreme care has been exercised
throughout the experiment to employ the same methods of feeding and brood-
ing. Hatching dates have been kept constant each year, but there have been
eight hatches each year at weekly intervals between Marcii 2.5 and May 15.
The range in hatching date thus amounts to 49 days. Hatching date, how-
ever, may be controlled at will by the poultryman.
ANMAI. I'KKSISri'.NC'V AM) V.GC. I'UODUCTION
193
1. Corrchition I^etxce"n Hdtrhhxi half aud Annual Perslstencii.
ilic class interval for liatcliiiisi date is seven days, and the class interval for
annual persistency is 15 days with a ranfie of from (J7 to 366 days. The fol-
liiwin": constants were calculated on 217J) hirds:
Number of liirds . . . . .
-Mean hatching date (April 18)
Hatching date standard deviation .
Mean annual persistency . . . .
Persistency standard deviation
Coefficient of correlation ....
Regression hatching date on jiersistency
Regression persistency on iiatching (hile
217.9
4.28
±2.21.
300.i7
±62.64.
—.2208 ±.01 37
—.008
—6.187
The .ihove constants indicate that, on tlic average, the hirds hiid for 300
days l)efore the onset of complete molt. This figure is somewhat lower than
it would he if maximum persistency had not been ])laced at 366 days, because
some of tiie birds laid for a greater time interval. The standard deviation
in persistency amounts to almost 63 days and furnishes statistical evidence
of very marked variability in persistency.
The coefficient of correlation between iiatching date and annual persistency
is negative and statistically significant. While this is not an intimate correla-
tion, it does demonstrate a tendency for early-hatched birds to lay longer
than late-hatched birds. The fact should be kept in mind, however, that the
earliest hatch was taken off" each year about March 2.') and that this date
should not be considered very early in this latitude.
2. Correlotiun lietxceen Hcdchinti l)<tfe E-irller than the Men?i nnd Hi()h
Persistencfi.
As previously stated, the birds have been cli\ idcd into two classes with
regard to persistency, namely, high and low. All birds are classed as hUjh in
persistency when they lay for 315 days or more before molting. Birds laying
for a shorter period than 315 days arc classed as low. By dividing the popu-
lation of 2179 birds into these two classes for persistency, and by again classi-
fying these as hatched earlier or later than the population mean, an absolute
measure of the correlation lietween early hatching and high persistency is
obtained. The results of this classification follow:
Hatching Date
I High Persistency | Low Persistency
I ' I
I'.arlier tlian population mean
720
458
Later than popidation mean
Totals
416
1136
585
1043
Coefficient of correlation
+.3771 ±.0124
194
TECHNICAL BULLETIN 9
The division of tlie population into high and low persistency groups in the
above table rests on a possible genetic foundation as already stated. The
mean persistency of the entire 2179 birds was found to be about 300 days as
section 1 shows. AVhen the point of division between high and low persist-
ency birds is taken between 314 and 315 days, there are 1136 individuals
classifying as high and 1043 as low in persistency. The low persistency class
ranges from 67 to 314 days while the high persistency class ranges from 315
to 366 days. The wide range for the low class enables them to bring the mean
persistency of the population down to 300 days even tiiough there are more
high-persistency birds than low-persistency birds in the above classification.
A positive coefficient of correlation .3771 ±.0124 signifies that early hatch-
ing is associated with high persistency. Possibly early hatching better equips
the pullet for a long laying year because she begins to lay earlier in the fall
and is also able to finish her laying year under more favorable weather con-
ditions than is her late-hatched sister. These data signifiy, therefore, that
hatching before the middle of April tends to increase persistency for the
pullet year.
(B) Relation Between Inherited Characteristics Concerned
WITH Fecundity and Annual Persistency.
In the class of inherited characteristics concerned with fecundity the fol-
lowing will be considered in relation to annual persistency: Age at first egg,
weight at first egg, winter rate, length of winter pause, and total days broody,
all records being based on the pullet year.
3. Correlation Between Age at First Egg and Annual Persistency.
Both age at first egg and annual persistency have been found by Goodale
and Sanborn (loc. cit.) and by the writers to be of appreciable significance
in breeding for egg production. Both early maturity and high persistency
are essential in the high producer and for this reason their relation to each
other should be known. The identical group of birds studied in section 1 is
used to determine the following constants:
Number of birds
Mean age at first egg
Age standard deviation .
Mean annual persistency
Persistency standard deviation
Coefficient of correlation
Regression age on persistency
Regression persistency on age
Mean age at first egg is about 209 days, which is a figure falling within the
limits of genetic early maturity. Age at first egg is a characteristic that
fluctuates widely, and in this particular population the extremes are 140 and
379 days, respectively. Class intervals of ten days for age have been used in
these correlation studies.
The mean annual persistency of the population is about 300 days. The
2179
208.56
±31.28
300.47
±62.64
— .6146±.0090
—.307
—1.231
ANNTAl. rKIJSISll-,N( ^ AND KCICI I'HODUCTIOxX
10.5
(xt nines iiro (>7 ;ind 366 cl;iys, rtsprctivrly. The sliindiird deviation for per-
sistenev is very large and indicates tliat a nuniher of faetors is concerned.
'riie negative ciKnicient of correlation is of siicii niiiiiiiitudc as to suggest an
intimate relation between age at first egg and annual persistency. Those
pullets that lay at an early age ajipear to he nuuh more persistent layers
than those maturing later. Herein lies a partial explanation of the significant
iclation between early maturing and high annual production. These studies
point to age at first egg as a criterion of importance for i)redicting per-
sistency.
4. Correldtiun Belxaeen Aye al First E(j(j Deloxv the Mean and High Per-
sistency.
The population has again been di\ ided into the two possible genetically dif-
ferent classes for persistency as in section 2. These classes have been tabu-
lated against age below the mean and age above the mean as follows:
Age at First Egg
I
High Persistency | Low Persistency
Below ))ojnilatii;n mean
860
387
Above population mean
276
656
Totals
1136
1043
Coefficient of correlation
.6816±.0077
A very intimate correlation is shown by the above coefficient between early
sexual maturity and high persistency. This relationship is very significant to
the breeder, disclosing possible genetic linkage between two desirable inherit-
ed traits that may later be cleared up on a factorial basis.
5. Correlation Betzceen Weight at First Egg and Annual Persistency.
Body weight is a convenient standard to use for selection purposes. Weight
in poultry is inherited on a multiple factor basis according to Punnett and
Bailey (1914). If weight should prove a criterion of persistency, its value
for culling purposes soon after pullets begin to lay is very evident. Weight
records are available on 2125 of the birds being studied, and when correlated
with persistency give the following constants:
Xuniber of birds
Mean weight at first egg .
Weight standard deviation
Mean annual persistency .
Persistency standard deviation
Coefficient of correlation .
Regression weight on persistency
Regression persistency on weight
2125
5.58
±.75
302.64
±58.00
— .3225±.0131
—.004
—25.002
196
TECHNICAL BULLETIN 9
Tills group of birds averajred about five and one-half pounds at first egg
and the extremes are 3 and 9.5 pounds, respectively. Class intervals of .5
pound were used in making these studies. Weight shows a coefficient of vari-
ability of about 13 per cent.
The coefficient of correlation exhibits something of a tendency for ligiit
weight and high persistency to move together. Such a coefficient might have
been anticipated from the fact that weight and age at first egg are positively
correlated (Hays, Sanborn and James, 1924), and because hatching date and
weight at first egg are negatively correlated (Hays, Sanborn, and James, loc.
cit.). In view of these* facts, it is doubtful if weight at first egg is a true
criterion of persistency.
(). Correlation Beticeen Bocli/ Weit/bt at First Eijcj lieloxv the Mean and Hi<ih
Persistency.
Weight at First Egg | High Persistency Low Persistency
1 ^ 1
Below population mean | 714 | 468
Above population mean 417 526
1 1
Totals 1 1131 1 9.94
1 1
Coefficient of correlation ..... 4~-3161it.01.'32
The above table presents the absolute correlation between weight at first
egg below the population mean and high persistency. Those birds weighing
less at first egg than the mean of the whole population may be considered
small while the high persistency class includes only those individuals laying
for 315 days or more before molting.
The coefficient of correlation is positive and of statistical significance. There
is a tendency for the persistent class to weigh less at first egg than does the
low ])ersistcncy class. Although the correlaticn is significant, it is not pro-
nounced and probably does not im])ly tiiat factors for rapid growth are in-
imical to high persistency.
7. Correldliori BetTveen Net Winter Rate and Annua} Persistencij.
In order to discover if there is any association between the net rate of lay-
ing throughout the winter season and persistency of laying the following fall,
a correlation table was made between winter rate and persistency, using the
2147 birds with records for lioth characteristics. The constants are as follows:
Number of birds ....
Mean winter rate ....
Winter rate standard deviation
Mean annual persistency
Persistency standard deviation
Coefficient of correlation
Regression winter rate on persistency
Regression persistency on winter rate
2147
()7.41
±8.87
302.98
±59.n3
-f-.1835±.0141
-f.028
+1.222
ANNIAI, I'KUSISTKNCV A XI) KCG l'l{()!)l'( TION
197
A slifi'nt luit siiiiiilKMiit <'i»rrfl.ili(ni is fdiind lo fxist luMweeti winter rate of
layiiiir and persisteiu\ . This correlation indicates lliat, in jjeneral, there is
some tendency for tlie more intense winter layers to jiersi.st in layiii'; later in
the fall than do les> intense layers.
,s. ('(irrel'ilii>ii lutxceeii ]\'iiiltr Rate (/ri'dtii llniii the Mtaii (iiitl Hii/li
f^ersLtfenci/.
Ry classifying all liirils with higher winter rates than the mean of the
whole |)oi)idation as hiuh for rate, and by classinju' as highly persistent all in-
dixiduals laying for ;JI5 days or more, the following table gives the correla-
tion between high winter rate and the presence of ]iossible genetically high
persistency:
Winter Rate
;-
.gh
Persistency
Low Persistency
Above i)oi)ulation mean
651
4.66
Below ])oi)iilation mean
1
1
181
516
Totals
1
1135
■
1012
Coeflicient of ciirrelation +.2236±.0138
The above tabulation presents a moderate degree of positive correlation
between two inherited characteristics concerned in high fecimdity. The very
significant fact is brought to light that high winter rate and high persistency
are ]iartially conn)lementary, and there is no evidence of antagonism between
t1ie two.
f). Corrc'iillnii Bi'xiueii LeiKjth of Winter Pause (iti'l JiDinal Pemintenci/.
The j)resence or absence of winter ]iause has been shown by Hays (1921)
to dej)cnd u])oii genetic factors. '1 he duration of the pause, however, may
depend upon en\ironment as well as inheritance. Most environmental forces
alTecting the duration of pause are ])roi)ai)ly beyond control of the breeder
;;nd may not properly be considered in this rei)ort. This section is devoted
to a study of the correlation between length of pause and persistency as has
already been done by Hays and Sanborn (I{)26b). In the ])()pulation being
studied there were 131S birds with winter pause records which were divided
into ten-day class inters als and the following constants arrived at:
Number of birds 1348
Mean length of winter pause 32.39
Pause standard deviation ..... ±21.77
Mean annual persistency 309.03
Persistency standard deviation .... ±54.89
Coefticient of correlation -f .1017±.0182
Regression persistency on pause .... -|— 256
Regression pause on persistency .... -(—040
198
TECHNICAL BULLETIN 9
Winter pause duration is subject to extreme fluctuations. Its range extends
from 4 to 130 days. The magnitude of its standard deviation indicates that
its duration is affected by a considerable number of variables.
The above coefficient of correlation gives a statistically significant yet far
from pronouaced correlation between length of winter pause and annual per-
sistency. There is but a very slight tendency for long-pause birds to persist
longer than do short-pause birds.
10. Correlation Between Annual Pers'istencij Above the Mean ami the Pres-
ence of Winter Pau^e.
This section is devoted to a consideration of the presence of winter pause
and annual persistency above the population mean of 303.20 days. Such a
correlation will bring out any possible association between the heritable char-
acteristic, winter pause, and high persistency which, in this instance, means
persistency greater than the mean of the population studied. The classifica-
tion follows:
Annual Persistency
1 Pause
1
Non-Pause
Above population mean
1
1 855
1
423
Below population mean
1
493
378
i
Totals
1348
801
Coefficient of correlation
-.2156±.0139
The correlation coefficient is significant though of only moderate magnitude.
Possibly winter pause birds tend to lay for a slightly longer period than do
non-pause birds because the former are more likely to be early-hatched (Hays
and Sanborn 1926b), and early-hatched birds tend to be more persistent than
late-hatched birds. The exact relation between pause and persistency can
only be discovered through the partial coefficient of correlation and will be
reported in a later publication.
11. Correlation Between Total Days Broody and Annual Persistency.
The heritable trait, broodiness, will next be considered in so far as its in-
tensity as measured by total days broody is correlated with persistency. Only
the pullets that exhibited broodiness during their first laying year are used
to obtain the constants below:
ANNUAL PKUSISI'KNC ^' AND KC.C. I'liODlC'TION
11)1)
Nmiil>or of birds ....
Mean total days broody .
Broody standard deviation
Mean annual persisteney .
Persistency standard deviation
Coefficient of correlation .
Repression days broody on persistency
Regression persistency on days broody
1037
i2.()9
±27.33
294.05
±64.-82
+.0532:
+.022
+.126
.020.0
Intensity of broodiness, as measured In' tiie total days spent in broodiness
dtiring the pullet year is subject to wide fluctuations. Its standard deviation
.shows marked variability in the population. In view of this fact, it seems
probable that intensity of broodiness depends on a number of variables.
The .small and .statistically insignificant coefficient of correlation indicates
practical independence between degree of broodiness and annual persistency.
12. Correlation Between Annual Persistency Above the Mean and the Pres-
ence of Broodiness.
This section deals with the absolute relation between tlie presence of tiie
inherited characteristic, broodiness, and persistency greater than the mean of
the population studied. Herein lies a definite basis for selection which may
or may not be useful in breeding for high persistency. The following results
appear:
Annual Persistency
I
1 Broody
1
1 Non-broody
Above population mean
1 566
1 715
Below population mean
1 471
390
Totals
1 1037
1
1 1105
1
Coefficient of correlation
—.2081 ±.0139
The above negative correlation coefficient is statistically significant though
it does not reveal an intimate correlation. Eliminating the broody character-
istic should in some measure increa.se annual ]»ersislency.
(C) The Relation Bktwekn Pkuststexcy axd FECuxDiri'
Since high annual persistency appears to be a desirable cliaracteristic to
develop from several standpoints, it is higlily desirable tiiat its relation to
both winter and annual egg record be ascertained. Both relations may be
considered first from the quantitative standpoint and then from tiie qualita-
tive standpoint by use of long and short correlation tables, respectively.
200
TECHNICAL BULLETIN f)
13. Correlation Beticeen Winter I'rodnctioii oikI .liinnni J'ersi.s-tcnci/.
Winter production diiriiifi' tlie jiullet year is represented by tlie nuiuher of
eggs laid from first egg to the end of Fchruary. It lias already lieen pointed
out by many workers as a valuable criterion of annual production. Class
intervals of ten lia\e been used to make tlie correlation taiile for the 21.51
birds with winter records. Constants computed follow:
Number of birds ....... 21.51
Mean winter production ...... 62.19
Winter production standard deviation . . ±2.5.M
Mean annual persistency ...... 302.82
Persistency standard deviation .... ±59.32
Coefficient of correlation . .... -{-.1.5.51iii.011.5
Regression production on persistency . . . 4-. 19.5
Regression persistency on ])roduction . . . -|-1.0()1
The degree of correlation between winter production and annual jiersistency
is positive and of appreciable magnitude. Selection for jiersistency based on
winter records should be of consideraiile value. Such a condition might be
anticipated in view of the high correlation between early maturity and winter
production and between early maturity and jH-rsiskncy.
14. Correlation Between Winter Production (Irealer Tlxtn the jMean and
h i(/ h Pe rsis t en cij .
In the tabulation below the population is classified into four qualitative
groups, namely, high winter produ.cers, low winter producers, possible genet-
ically highly persistent, and possible aeiielically low persi.stent. The correla-
tion is then determined between production above the mean and high jier-
sistencv.
Winter Production
1 1
1 High Persistency (
1 ' 1
Low Persistency
Above population mean
1 1
1 ^1- 1
1 1
345
1 i
Below population mean | +24 | 670
Totals ! 1136 I 101.5
Coefficient of correlation +.5306±.01O4
This coefficient of correlation demonstrates a jjositive relation between high
winter egg record and high persistency. In other words, selection based on
winter records greater than the average should increase the percentage of
late-molting or persistent birds.
ANNUAL I'EKSISTKNCV AND KCtl I'HODrCTION
15. Corn Uittdii Between Ainntal I'rodiictioii mnl .tinniiil Pernisteiiri/.
201
Other coiulitions being etpiiii, ;miv iiuTease in persisteney should l)e .iccoin-
panied Ity an inerease in annual ejrfr yield. These are i)urely quantitative rela-
tions and in this manner some information concerning the value of high per-
sistency from the fecundity standpoint may be ascertained. The same popu-
lation of 217!) individuals is tabulated, using ten-day classes for production,
to obtain the following constants:
Number of birds ....... 2179
Mean annual production ...... 177.16
Production standard deviation .... ±44.73
Mean annual persistency ..... 300.47
Persistency standard deviation .... ±62.64
Coefhcient of correlation +.7082±.0072
Regression production on jicrsistency . . . -{-.506
Regression jiersistency on {production . . . -(-.992
The ai)ove coelticient rc\eals an intimate correlation between annual egg
yield and annual persistency or the length of the laying year. These data
furnish definite evidence to conunend the practice of emphasizing late molt-
ing in breeding for high fecundity. On the average, any increase in persist-
ency within the limits of the pullet laying year is advantageous.
16. Correlation Between Annual Production Above the Mean and Hii/h
Persistenci/.
By classifying all birds as high producers if they laid more eggs than the
population mean of 177.46, and as high in persistency those birds that laid
for not less than 315 days before molting, a definite relation between high
production and high persistency may be established.
Annual Production | High Persistency | Low Persistency
Above population mean i 872 | 280
. \ , \
Below popidation mean j 264 | 763
L__ \
I I
Totals I 11.36 I 1043
Coefficient of correlation -f .8000±.0052
The above coefficient of correlation establishes a very intimate relation be-
tween the presence of possible genetic high persistency and annual egg yield
above the average of the total population. This fact points to high ])ersist-
ency as being closely associated with high annual fecundity. High ])ersist-
ency must, therefore, be classed as a trait of vital importance in breeding for
fecundity and one that .should be stressed greatly by the breeder. Siiould
high persistency breed as a true recessive, it would be a com])aratively simple
matter to establish the characteristic in the laying flock.
202
TECHNICAL BULLETIN 9
Discussion axd Summary.
Annual persistency is a characteristic bearing a vital relationship to fecund-
ity. Its duration is affected bj' environmental influences and by inherited
traits concerned in fecundity. No conclusive evidence is presented in this
report to indicate that high persistency behaves as a simple recessive in in-
heritance as has been suggested by Hurst (loc. cit.). In this climate persist-
ency may be increased to some extent by hatching before April 15, with such
birds as are studied here. Early sexual maturity, npn-broodiness and high
winter rate probably show some linkage with high persistency. At any rate,
there is no evidence of antagonism in attempting to combine these desirable
traits in the same individual. Valuable information for selection purposes
has been disclosed by these studies. Partial correlation coefficients will be
used in a later publication to remove some complications.
The general relation of persistency to winter and annual production for
the whole population studied is shown in the following table:
Character of Birds Winter Production Annual Production
Persistency above population mean 69.84 198..59
Persistency below population mean 51.57 145.67
Persistency of 3l5 days or more 71.13 201.98
(Mean 347 days)
Persistency below 315 days 52.83 150.75
(Mean 249 days)
Year
1916
1917
1918
1920
1921
1922
1923
1924
Total and average
The chief findings of this report may be snmmed up as follows:
1. Early hatching is moderately correlated with high annual persistency.
2. Age at first egg is very intimately negatively correlated with high per-
sistencj''.
3. Weight at first egg shows significant negative correlation to persistency.
4. Winter rate of laying is moderately correlated with persistency. The
two traits appear to be partially complementary.
5. Length of winter pause is l)ut .slightly positively correlated with per-
sistency.
6. Total days broody is not significantly correlated with persistency.
7. The presence of broodiness shows a fair negative correlation to high
persistency.
in Persistency
'>y
Years
Number of
Birds
A\
erage Persistency
278
247.53
347
280.74
194
285.49
125
325.29
314
301.00
379
329.67
317
316.33
225
320.47
2179
300.19
ANNUAL PERSISTENCY AND EGG PRODUCTION
203
8. M'inter production and persistency are rather signilicantly positively
correlated.
9. Annual production is pronouncedly correlated with persistency.
10. Persistency behaves as a trait much to be desired from the production
standpoint.
References
Cloodale, H. D. and Ruby Sanborn. 1922. Cluuiges in egg production in tiie
station flock. Mass. Agr. Expt. Sta. Bull. 211.
Hays, F. A. 1924. Inbreeding the Rhode Island Red fowl with special refer-
ence to winter egg production. Amer. Nat. 58:43-59.
Hays, F. A., Ruby Sanborn and L. L. James. 1924. Correlation studies on
winter fecundity. Mass. Agr. Expt. Sta. Bull. 220.
Hays, F. A. and Ruby Sanborn. 1926a. Broodiness in relation to fecundity
in the domestic fowl. Mass. Agr. Expt. Sta. Tech. Bull. 7.
Hays, F. A. and Ruby Sanborn. 1926b. Winter cycle and winter pause in re-
lation to winter and annual egg production. Mass. Agr. Expt. Sta. Tech.
Bull. 8.
Hurst, C. C. 1923. Experiments in genetics, p. 490. Cambridge: Cambridge
University press.
Punnett, R. C. 1923. Heredity in poultry. New York, London, etc.: Macmil-
lan and Co.
Punnett, R. C. and P. G. Bailey. 1914. On inheritance of weight in poultry.
Jour. Genet. 4:23-39.
f
K^
i' < .-'^■jidi'MiKai College
massachusetts'^*^*'*'' **ajis, -
Agricultural Experiment Station
TECHNICAL BULLETIN No. 10 NOVEMBER, 1926
THE THERAPEUTIC EFFICIENCY OF
AVIAN DIPHTHERL\, ROUP, AND BIRD POX
VACCINES AND BACTERINS
By Norman J. Pyle
Avian diptheria, roup and bird pox cause serious loss to Massachusetts
poultrymen by decreasing egg production during the season when eggs are
bringing the highest prices. In this bulletin the Department of Veterinary
Science and Animal Pathology reports results of their study of the prob-
lem. A filtrable virus was found to be the cause of all three types of the
disease. None of the commercial vaccines produced immunity, neither did
they effect a cure when the disease was present, although they caused a
slight improvement in the general condition of the birds. Autogenous
bacterins, when used in the early stages of the disease, caused an improve-
ment in the general health of the birds, but were not of sufficient value
to make their use economically profitable.
Requests for bulletins should be addressed to the
AGRICULTURAL EXPERIMENT STATION
AMHERST, MASS.
THE THERAPEUTIC EFFICIENCY OF AVIAN DIPHTHERIA,
ROUP, AND BIRD POX VACCINES AND BACTERINS
Bv NORMAN J. PYLE
Introijuctiox
Avian diphtheria, roup, and bird pox have caused serious financial losses
to the poultry industry of Massachusetts during the fall and winter months
of past years. Diphtheritic roup has been the predominating form of the dis-
ease. It has not been attended with great mortality, but has become of grave
economic importance because it has caused a decrease in production, occurring
at a time of the year when poultry products bring maximum prices.
Two biological products, a powdered pox virus vaccine and an avian mixed
infection bacterin, have been used extensively in an attempt to control the
disease. The results obtained following the use of the preparations have been
confusing. Some reports claim the vaccine to be 100 per cent efficient, while
others claim it to be an absolute failure.
Many factors contribute towards the efficiency of the vaccine and bacterin.
It is of primary importance to ascertain the nature of the causative micro-
organism or virus and whether it is incorporated in either of the preparations.
When this is accomplished, it is assured that either the vaccine or bacterin is
the logical product to develop specific antibodies against the disease.
It is also necessary to determine whether avian diphtheria, roup, and bird
pox are separate etiological entities or various manifestations of a common
cause. On the answer to this problem depends the need for one common
vaccine or bacterin or for separate ones for each entity.
Historical.
Moore (1), a pioneer American worker on avian diphtheria and roup, iso-
lated a non-motile, pathogenic bacillus from lesions of the disease. He claimed
that this organism was "apparently the etiological factor". He was unable,
however, to determine its specificity for the affection. Harrison and Streit
(2) demonstrated that Bacillus pijocyanens would produce typical lesions of
avian diphtheria and roup. These authors also found a second virulent bac-
terium associated with the diseases, which they called the roup bacillus or
B. cacosmos. Hausser (3), Bordet and Fally (4), Beach, Lothe and Halpin
(5), and Crofton (6) have all added specific organisms to the long list of
causative factors.
Bird pox or contagious epithelioma has not been studied from the stand-
point of its etiology to the extent that has avian diphtheria. Marx and Sticker
(7) reported investigations wherein they found ai filtrable virus to be the
cause of bird pox. Schmid (8) and Sigwart (9) confirmed this work.
V. Betegh (10), De Blieck and V. Heelsbergen (11), and others advanced the
theory that all forms of the disease are caused by one and the same virus.
Several references in the early literature maintain that the various poxes,
skin eruptions, and variola affecting animal life are all caused by a common
virus, which adapts itself over a period of successive generations to a specific
host. If this were true, vaccinia or cowpox would have an etiological rela-
tionship to bird pox.
BIRD VOX VACCINES AND RACTERINS
207
Lowentli.tl, Kadowaki, and Kondo (12) wore al)Ie to transmit vaccinia to
tlic f o\s 1 tliroiijxii five successive {generations, hut tiie affections hecaine less
and less pronounced and finally died out. Fowls recoverinj; from vaccinia
were imnume to vaccinia, and tliose recovering from bird pox were immune to
bird pox. They were unable to produce a neutral or combination innnunity
and concluded that the causes of vaccinia and bird pox were very different.
Exi'KRIMEXTAL DaTA ON THE El lOI.OOV OF AviAN DlI'HTUElUA
Roup, ano Biud Pox.
Bacteriological examinations of diphtheritic patches were made and many
organisms were isolated, the majority of which were contaminating invaders.
In order to avoid this the i)atches were aseptically removed and the bacteri-
ological examination made directly from tiie underlying, denuded surface. The
same technic was employed in the ]>ox form of the disease; that is, bacterio-
logical cultures were made from the pitted areas after removal of the pox
scabs.
Pseitdomoiw.i itenufinosa (Bdcilius pi/ocyaneus) was found associated with
pox and diphtheritic lesions. This organism has been previously observed in
diphtheritic roup by Harrison and Streit (2), Hausser (3), Jackley (13),
Kaupp (14), and others. Various other pyogenic bacteria were isolated,
namely, Staphi^lococrus aureus, Oajfkifd {Staphylococcus) tetragena, and
Staphiflococcus alhus. A Fasteurelki avlcUla-like organism was isolated from
infected birds suffering with avian diphtheria, also one similar to the roup
bacillus or Bacillus cacosmus of Harrison and Streit (2).
These organisms are at least prominent secondary invaders, but their ability
to cause diplitberitic roup is in doubt. Psewlomonas aeruginosa, when found
in an infected flock, was readily isolated from the heart blood, liver, and spleen
of those birds dead of the disease. The organism was injected into the wing
veins of several healthy birds and death ensued in from fifty-six to eighty-
four hours. The germ was recovered from the dead fowls, especially from
exudates in the nasal passages, indicating that the organism was associated
with roup. Other exj^eriments with the organism, such as injection beneath
the skin and application to scarified wounds of the comb, wattles, and mem-
branes of the mouth, failed to produce any type of the disease.
Fresh pox scabs obtained from a Massachusetts infected flock were dried,
passed through a coffee mill, and finally pulverized in a ball mill. One gram
of this powdered virus was macerated for twelve hours and afterwards tritur-
ated in 100 cc. of physiological salt solution. It was then passed through a
controlled Berkefeld filter of medium porosity. The filtrate was vigorously
rubbed into the scarified comb and wattles of healthy birds and failed to
reproduce the disease in forty-three days. These birds were susceptible to
avian diphtheria and bird pox for they later succumbed to inoculation with the
imfiltered virus. The experiment was repeated, using scab virus from two
•other States, and again using a filter of medium porosity. The results were
the same. The ex])eriments were controlled by the respective unfiltered scab
viruses which produced t\"]iical pox lesions in the usual incubation period.
It is known that the filtrable virus of smallpox will not pa.ss through a
filter of fine porosity, but will pass through one of coarser porosity. Accord-
ingly, Berkefeld filters No. V (coarse) were next used and the results recorded
in the following table. The filtrate proved "sterile" upon cultural examination.
208
TECHNICAL BULLETIN 10
«
U
o
CO
1-5
■a
o
■3
O
0/
o
o'
0,
•3
o
13
a>
o
a>
c
00
^5
o ,
o
o .
o "'
o .
■-Si
o ""
o
c
c .
"C a;
o
o .
'Z ©
d
o
11
>>
1
.2 2
S £
111
00 *; V © ■
® c i c «
P- .
c.r
4)
o ^
o
o ^
Pn
IH
a
.4
o S o
« C8 >
o
a
o
|2!
o3
S 2 « ■* rt
£ S
BS
•Z a)
d >
^ 0)
o ^•
1"
CM
u
o
a
«
u
u
"2 6 .
a) o °
o
p.
o
12;
y.
o
a
o
X
o
p.
o
12;
11
CO •
§2S
o !; o
p ><
1l
O X
CM
U
a
a>
o
o
P.
o
o
o
1^
o
p.
o
!Z5
o
p.
O
> a
A° >
PM a)
x'3
k<
a)
.Q
S
a)
o
«
■«2
iJd si.
" J, o
^ u o
» si
^ tH O
Ills
U M 13 O
^ t. o
Ills
^ w =e o
J 1^
T3 o
03 aj-Tj-d —
^.•jj a) 0- g
nils
CO
£ . M
a! »T3'd-5
^tfi ® © ?
o i^ a^
O wtrt 03 o
6 "5
c . ®
O O j-
o 5
C • a;
BIRD POX VACCINES A'SD BACTKRINS
209
Interpretation
1. The filtered virus produced pox, roup, and avian di])litiieria, indicating
thai one and the same virus is capable of causing all forms of the disease.
2. Bird No. 2 had a simple catarrh when inoculated. This evidently low-
ered the resistance of mucous membranee surfaces and avian diphtheria
followed. ■ *
3. The incubation period of the filtered virus was from twelve to eifthteen
days, while in the case of the unfiltered virus it ranged from seven to nine
days. The liltered virus also i)roduced a less pronounced form of tiie infec-
tion than did the unhltered virus. These latter two points confirm the work
of Schinid (8) in 1909.
A bacteriological examination of the unfiltered powdered virus revealed
several secondary invaders, such as P.seudomonas aeruginoxa and various
Staphylococci. These organisms imdoubtedly assisted the unfiltered virus in
causing a shorter period" of incubation and a more pronounced form of the
di.'^ease.
The Uniformity of the Viuulexce of Commercial Viruses.
Before studying the efficiency of the powdered pox virus vaccines, it was
desirable to ascertain the strength of tlie viruses which make up these com-
mercial products.
Four groups of birds were inoculated on October 5th with four different
strains of powdered pox virus. The course of the disease subsequent to the
inoculation is represented by the respective lines A, B, C, and D in the fol-
lowing graph. Virus A was obtained on October 1 from a natural infection
in Massachusetts and viruses B, C, and D were of commercial origin.
Chart I. Variance in Strength of Powdered Pox Viruses.
-|- Period of incubation.
-|-.|- Appearance of a few or several well formed pox nodules.
-|--|.-|- Appearance of many pox nodules of mature development.
-|--|--j--|- Maximum development of pox nodules.
Downward curves — Periods of recovery.
^'irus A siiowed the greatest potency. The period of incubation was seven
days, the disease reaching its maxinuun development three days later, and
death following within twenty-three days with no appreciable evidence of
recovery.
Virus B showed the greatest potency of the three commercial stock viruses.
The period of incubation was eight days, maximum development four days
later, and com])lete recovery in twenty-two days more.
210
TECHNICAL BULLETIN 10
Virus C presented an incubation period of nine days, niaxinunn develop-
ment fifteen days later, and complete recovery within an additional six days.
Virus D was very weak. The period of incubation was twelve days. There
was jiractically no further develoi)ment of pox and recovery soon took place.
• Interpretation.
The degree of efficiency of the powdered pox virus vaccine depends upon
the potency and antigenicity of the virus of whicli it is composed. The fol-
lowing conclusions are then evident: —
1. The viruses, being non-uniform in potency, would produce vaccines of
varying efficiency.
2. A method of standardizing the virus and vaccine, which is lacking at
the present time, would be essential to the efficiency of the vaccine.
3. An autogenous virus would produce a vaccine of greater value than one
composed of a stock virus.
The Efficiency of Powdered Pox ^'II{us Vaccines.
The powdered pox virus vaccine was first used by Manteufel (15) and
by Hadley and Beach (16). The vaccine as conunercially distributed to-day
is a development of the original methods of these workers.
Scabs collected from pox nodules are the source of the virus. In order to
produce large quantities of scabs it is necessary to maintain a flock of young
cockerels, preferably white leghorns. The combs and wattles are scarified
and the powdered scab virus after being "emulsified" in physiological salt
solution is vigorously rubbed into the woimded areas. Typical pox scabs will
develop and mature on susceptible bird-s in from seven to twelve days. The
scabs are then collected, thoroughly dried, passed through a coftee mill, and
finally pulverized in a ball mill. The product is stored away as the stock
virus.
The vaccine is made by takin."' I aram of the powdered virus and thoroughly
triturating it in 100 cc. of physiological salt solution. It is then attenuated
at 55° C. for one hour in a water bath. Finally it is filtered through sterile
cheesecloth into vaccine bottles, and after cooling is ready for use. The entire
procedure should be handled in as sterile a manner as possible. The vaccine
should be used within ten to fifteen days after it is manufactured because it
deteriorates rapidly.
In the following experiments having to do with the efficiency of the powdered
pox virus vaccines each bird was housed in a separate compartment. The
final conclusions are based on a repetition of experiments and the average
reaction of a group of birds. The vaccine used was manufactured as described
by J. R. Beach (17), a brief description of which is given above.
Experiment 1.
Part A.
A freshly made vaccine, composed of virus B, was administered subcutane-
ously to a grouj) of six liealthy birds, 1 cc. being given to each liird beneath
the skin of the breast under the right thigh. The group was divided into three
lots of two birds each.
Lot 1. Fourteen davs after vaccination both liirds were inoculated on comb
BlUn POX VACCIXKS AXI) llAriKUINS
•ill
and wattles witli virus IJ. I'ox nodules developed eijilit days later and reaelied
a inaxinuuii jirowth in an additional ten days.
Lot -. Twenty-six days after vaeeination both birds were inoeulated on
eonil) and wattles with virus B. A mild ]U)x developed eif^ht days later, hut
soon disappeared without further develoi>nient.
Lot o>. Forty-two days after vaeeination i)oth birds were inoculated on eonib
and watHes with virus B. I'ox was pronounced eijrht days later, one bird
showinti' dijihtheriiie patches in mouth as well i'.s pox.
Control: two non-vaccinated liirds inoculated \\i(li \irus B. Incubation
period of eight days, maximum (k'vcloiunent four (l,i\,-, later.
Part B.
Two injections of a virus B vaccine were given a second grouj) of six birds
in the same maimer. The second injection was gi\en six days after the
first. Ihe grou]) was likewise divided into three lots of two birds each, and
inoculated with \irus B tiftcen, thirty, and forty-two days resj)ectively after
the second injection.
Lot L Incuiiation i)eriod of ten davs, pox becoming pronounced five days
latei-.
Lot 2. Slight pox de\elo]>ed in eight days in t)nly one bird, clearing up
within the next seven days. Second bird showed no evidence of the disease.
Lot o. Pox developed in ten days, persisting for three weeks in a mild form.
Control: two non-vaccinated, healthy birds inoculated with virus B. Pox
developed in eight days, reaching a maxinmm development four days later.
Part C.
Three injections of a virus B vaccine were given a third group of six birds
at intervals of six days. The group was again divided into three lots of two
birds each, and inoculated with virus B sixteen, thirty-one, and forty-two days
respectively after the third injection.
Lot L Pox developed in eleven days and persisted in mild form.
Lot 2. Pox developed in eiglit days and became pronounced in another
week.
Lot o. Pox developed in twelve days and persisted in mild form for three
weeks.
Control: two non-vaccinated, healthy birds inoculated with virus B. Pox
deve]oi)ed in eight days, reaching a maxinnun development three days later.
Result.
One, two, and three injections of the vaccine failed to produce an ab.solute
protection against artiHcial infection with homologous virus B.
Experhnent 2.
Pari A.
This experiment was similar to Experiment 1, excei)t that the vaccine was
nuide of virus C and the check inocidations were made with virus B. The
first group of six birds was given a 1 ec. injection of the vaccine, divided into
tliree lots of two birds each, and inoculated with virus B seventeen, twenty-
six, and forty days respectively after the vaccine injection.
Lot 1. Pox developed nine days later, grew worse, and death followed in
one bird.
Lot 2. Pox develoj)ed within eight days, but in weak form, and cleared up
in two weeks.
Lot 3. Pox de\eloped within eight days in one bird and diiihtheritic roup
within ten days in the other.
Control: two non-vaccinated, healthy birds inoculated with virus B. Pox
developed in nine days, reaching a maximum develo])ment four days later.
212
Part B.
TECHNICAL BULLETIN 10
A .second group of six lieallhy birds was given two injections of a virus C
vaccine of 1 cc. each at five day intervals. The group was divided into lots
1, 2, and 3 and inoculated with virus B sixteen, thirty, and forty days re-
spectively after the second vaccine injection.
Lot 1. Pox developed in ten days and persisted in a mild form.
Lot 2. Pox developed in eight days in both birds and avian diphtheria in
one bird of the lot.
Lot 3. Pox developed in eigiit days, becoming pronounced, and complicated
with roup.
Control: two non-vaccinated, healthv' birds inoculated with virus B. Pox
developed in eight days and reached a maxinuun development tliree days later.
Part C.
A third group of six healthy birds was given three injections of a virus C
vaccine of 1 cc. each at five day intervals. The group was divided into lots
1, 2, and 3 and inoculated with virus B fifteen, thirty, and forty-one days
respectively after the third vaccine injection.
Lot 1. Pox developed in ten days and persisted in severe form for tiiree
weeks.
Lot 2. A slight pox developed in eight days, persisting in a mild form.
Lot 3. Pox developed in eight days, becoming severe and persisting as such.
Control: two non-vaccinated, healthy birds inoculated with virus B. Pox
developed in eight days and readied a maxinuim development four days later.
Result.
One, two, and three injections of the vaccine failed to produce an absolute
protection against artificial infection witli iieterologous virus B.
Other vaccine and virus combinations were used, such as a vaccine made of
virus B, and its immunizing ability checked with virus C. The results were
comparable to experiments 1 and 2.
Infection by Contact.
A healthy, young cockerel was added to each lot of the foregoing experi-
ments after the disease developed in the supposedly immune birds. This addi-
tion of a strange bird to each lot of birds instigated fights, and minor wounds
of the comb followed. This allowed a point of entrance for the virus which
contaminated the food, water, and litter. Pox developed in about 50 per cent
of those birds in contact with the diseased ones. The infection persisted in a
mild form, never reaching the severity evidenced in those birds with which
tliey were in contact.
Experiment 3.
An ett'ort was made to determine the curative value of the vaccine. A
group of vweive birds was inoculated witli virus C. Pox nodules apjieared in
nine days and a moderate degree of development, wiiicii proved to be the maxi-
mum, followed in seven days. Tlie group was liien divided into two lots of
six birds eacli and placed in separate pens. A virus C vaccine, in a 1 cc. dose
was administered to each bird of one lot, tiie other lot being used as the con-
trol. No apparent decrease in number and severity of the pox nodules fol-
lowed the injection of the vaccine. The injected lot, however, appeared bright-
er and more active, and lo.ss of flesh was arrested after seven days following
the injection. The non-injected lot steadily lost flesii for two weeks, but from
then on gained in general ajipearance and physical conditions.
BIRD rOX VACCINES AM) HACTKHIXS
Results.
213
[lif use of the vaccine as a curative measure resulted in a slifzlit iiiii)r<ive-
lUfut in the general condition of the treated liirds, luif did not cause any
diminution in number or extent of the lesions.
OxE Attack of Bir» Pox Confeus ax Ijimuxity.
All hirds recovering from the infection during tiie exjieriments were held
over for future use. Approxinialcly fifty days following complete recovery
from both types of the disease, a group of such birds was inoculated wtih
viruses B and C. Lesions of the disease failed to develop, indicating that one
attack of the disease, whether of avian pox or diphtheritic type, confers an
imnumity of at least fifty days' duration. Four healthy birds which served
as controls developed pox in eight days with virus B, and in nine days with
virus C.
This actively acquired immunity is undoubtedly of greater duration than
that demonstrated by the abo%e experimental data. Evidence indicates that it
lasts from two months to two years, depeneling upon the virulence of the infec-
tion among the birds which acquire this protection.
The Efficiexcv of Bacterixs.
Several infected flocks were available during the fall and winter of the past
year for treatment with bacterins. Autogenous bacterins were resorted to for
the control of the outbreaks. Eleven different organisms, aside from
the common Siibtilis group, etc. contaminators, were isolated from diseased
birds obtained from five outbreaks of bird pox and avian diphtheria. These
organisms were not constantly present in all cases of the disease, and as has
been previously stated, they are secondary invaders only. It appears, there-
fore, that an autogenous bacterin is indicated in preference to a stock bacterin.
Also, such a preparation is limited to the control of secondary comjjlications
of the disease.
Commercial avian mixed infection bacterins were not used. Their bacterial
content does not correspond to the specific bacteria isolated from lesions of
birds affected with the disease as it exists in Massachusetts. McNutt (18) in
referring to experimental data on the use of such a biologic concludes, "In
every case the death loss among the treated equaled or exceeded the loss among
the untreated. Usually the loss was greater among the treated."
Flock 1.
A pen, consisting of 112 birds affected witli iiotii pox and avian diphtheria,
the latter predominating, was treated with an autogenous bacterin. Several
of the worst cases of both fomis of the disea.se were examined bacteriologic-
ally and the following organisms isolated: Stnphi/lococci aureus and nU>vs,
Gaffkija (Slaphj/lococcus) tetr(uje}in, and an unknown, gram negative, short,
rod-shaped organism of the colon group. The bacterin, composed of these
organisms, was standardized so that one dose of 1 cc. contained 2,000,0(30,000
organisms.
214
TECHNICAL BULLETIN 10
An initial injection of 1 cc. was given to 80 birds of the pen, selected prom-
iscuously, and 32 birds remained uninjected as controls. All birds were laying
well prior to the outbreak of the infection. Both the injected and control
groups averaged 43 per cent production at the time the first symptoms of
the disease were noticed. Three days prior to the first injection the egg pro-
duction of both groups dropped to 18 per cent. Fifty per cent of the total
number of birds showed symptoms of one or more forms of the disease. There
was no appreciable decrease in number or extent of lesions or increase in
egg production of both groups during the next few days. The infection aj)-
peared to be arrested, however. Nine days following the first injection, a
second one of the same dose was given. Three days afterward the injected
group of layers improved in general condition and the egg production began
to increase gradually. The condition and production of tiie control group
remained at a standstill. These results were evident in the same proportions
for the following two weeks, at the end of which time the last reading was
taken. The injected group had reached 41 per cent egg production and the
control group averaged 3.5 per cent. Lesions of the disease persisted, however,
in all birds, but were somewhat less extensive in type.
In estimating the percentage of egg production care was taken to consider
factors other than disease, which would tend to influence it.
Results.
The administration of the autogenous bacterin was followed by an improve-
ment in the general condition and production of the injected group. No
diminution in number or extent of the lesions was noted. Local treatment of
the lesions would probably have served the purpose.
Flock 2.
An autogenous l^acterin was administered to a second flock affected with
avian diphtheria. The following organisms, which were used to make the
bacterin, were isolated from typical cases of the affection: Staph i/lococci
aureus and albvs, Gaffkya {Staphylococcus) tetrayena, a gram negative, short
rod, bi-polar staining bacillus, and an organism of the Escherichia grouj),
typical of Escherichia schaeferi.
A severe infection of a similar nature had existed in these same pens dur-
ing the previous season. At the time of the injection a moderate degree of
the infection was present in the birds of houses 1, 2, and 3. One injection of
1 cc. of the bacterin, having a concentration of 2,000,000,000 organisms per cc,
was given each bird. Previous to the treatment the egg production had
dropped to 40 per cent. From four to six weeks later when final readings
v.ere made the production had increased to 66 per cent. House 1 contained
2.H per cent injected birds showing mild symptoms of the disease as opposed
to 11 per cent of the non-injected birds or controls in the same condition.
House 2 showed 12,3 per cent infection in injected birds and 30 per cent in-
fection in the controls. House 3 showed .5.4 per cent infection in injected
biros and 41.7 per cent infection in the controls. No attempt was made to
treat the symptoms of the disease.
Results.
One injection of the autogenous liacterin arrested the course of the infection
and brought about an increase in egg production.
Flock S.
*
A third flock of 2,000 birds was injected, each bird receiving 1 cc. of an
autogenous bacterin consisting of Staphylococci aureus and albus, and Pseu-
domonas aeruginosa. The iiacterial concentration in this instance was but
500,000 organisms per cc. Complete data on the results of the treatment were
BIRD POX VACCINES AND BACTERINS
215
not oht.iiimhie. An early report from tlie owner sliowed an improvement in
cjij: i>ro(liKtion and Init a few mild cases of tiie disease amonfj the treated
I'irds. It was cpiestionahle, however, whether the increase in production was
due to recovery from the disease or from an existing neck moult. No data
were availahle concerninu- the controls. No conclusion can he drawn from the
use of tije bacterin in this instance.
Fluch 4.
A fourth flock of 300 cockerels was injected with an autogenous hacterin
composed of Siaphylococcii.t aureus, a Pasteurella oricida-like organism, and
a bacillus similar to the rouj) bacillus or Ji. caco.stntis of Harrison and Streit
(2). The infection had practically run its course at the time of the treatment.
Two injections were given, the first of 0.5 cc. containing 500,000 organisms,
and a second six days later of 1 cc. containing 1,000,000 organisms. The dis-
ease entirely cleared u]> during the following tliree weeks. No difference was
noted between the injected and control groups.
Summary.
1. Several organisms were isolated from the lesions of avian diphtheria,
diphtheritic roup, and pox. They proved to be of no causative significance,
but were prominent secondary invaders. A filtrable virus was found to be
the common cause of all types of the disease.
2. Commercial stock powdered pox viruses varied markedly in ability to
produce the disease. The need of a method of standardizing the virus and
vaccine was indicated.
3. One, two, and three injections of the powdered pox virus vaccines failed
to produce an absolute protection against artificial infection with homologous
and heterologous viruses.
4. Infection by contact occurred in 50 per cent of all cases.
5. The powdered pox virus vaccine caused a slight improvement in the
general condition of diseased birds when administered as a means of bringing
about recovery from the infection.
6. One attack of either or both types of the disease conferred an immunity
of at least fifty days' duration against both types.
7. Autogenous bacterins, when administered in the early stages of the dis-
ease, caused an improvement in the general health of the birds. As avian diph-
theria and pox advance in severity the egg production of hens decreases.
"With the injections of these bacterins, data at hand indicate that the egg
production is increased. While all these observations are interesting and
point to a certain degree of therapeutic efficiency; the time consumed in the
manufacture, standardization, and administration of these bacterins would
w'ork against their use as an economic practice.
Bibliography.
(1) Smith, Theobald, and Veraxus A. Moore: Investigations concerning
infectious diseases among poultry. U. S. Dept. Agric, Bur. Anim.
Indus., Bull. 8, p. 56, 1895.
(2) Harrisox, F. C, axd H. Streit: Roup: an experimental study. Onta-
rio Agric. College Bull. 132, 18-22, 1903.
(3) Hai'sser, Albert: Bacteriologische Untersuchungen iiber Gefliigeldiph-
therie. Centralbl. f. Bakteriol., Parasitenk. u. Infekt. — Krankh., 1908,
1 Abt. Orig., Bd. XLVIII, S 535.
216
TECHNICAL BULLETIN 10
(4) BoRDET, J., ET V. Fally: Le Microbe de la Diphtheric des Poules.
Ann. de L'Inst. Pasteur, 1910, T. 24, p.. 563.
(5) Beach, B. A., H. Lothe, and J. G. Halpin: An outbreak of roup and
chicken-pox in which the high mortality was apparently caused by a
secondary invader. Jour. Infect. Diseases, 1915, 3, XVII, 554-58.
(6) Crofton, W. M.: Diphtheria of fowls: Its cause, prevention, and cure.
Jour. Path, and Bacteriol., 1924, 4, XXVII, 456-58.
(7) Marx, E., uxd A. Sticker: Untersuchungen iiber das Epithelioma con-
tagiosum des Gefliigels. Deut. med. Wochcnschr., 1902, S 893. Cited
in Kansas Agric. Exp. Sta. Tech. Bull. 4, p. 7, 1917.
(8) ScHariD, G.: Untersuchungen iiber der Beziehungen zwischen Gefliigel-
diphtherie und Epithelioma contagiosum. Centralbl. f. Bakteriol.,
Parasitenk. u. Infekt.— Krankh., 1909, 1 Abt. Orig., Bd. LII, S 200.
(9) SiGWART, A.; Experimentelle Beitrage zur Frage der Identitat von
Gefliigeldiphtherie und Gufliigelpocken. Centralbl. f. Bakteriol., Par-
asitenk. u. Infekt.— Krankh., 1910, 1 Abt. Orig., Bd LVI, S 428.
(10) Betegh, L. von: Uber die Beziehungen zwischen Gefliigeldiphtherie
und Gefliigelpocken. Centralbl. f. Bakteriol., Parasitenk. u. Infekt. —
Krankh., 1912, 1 Abt. Grig., Bd., LXVII, S 43.
(11) DE Blieck, L. tJND T. VAN Heelsbergen : Impfung gegen Diphtheric
und Gefliigelpocken bei Huiincrn. Deut. Tierarztliche Wochcnschr.,
1923 Feb., 8, Bd. XXXL
(12) LowENTHAL, W., Y. Kadowaki UND S. KoNDo: Untcrsuchuugcn iiber
das Verhaltnis der Gefliigelpocken zur Vakzine. Centralbl. f. Bak-
teriol., Parasitenk. u. Infekt.— Krankh., 1925., Feb. 18, 1 Abt. Grig.,
Bd. XCIV, S 185.
(13) Jackley, J. G.: A study of the etiology of roup in birds. Kansas
Agric. Exp. Sta. Tech. Bull. 4, p. 12, 1917.
(14) Kaupp, B. F.: A chromogenic bacillus from a case of roup. Jour.
Infect. Diseases, 1918, 568-571.
(15) Manteufel: Beitrage zur Kenntnis der Immunitatserscheinungcn bei
den sogennannten Gefliigelpocken. Arb. a. d. k. Gsndhtsamte., 1909-10,
Bd. XXXIII, S 305. Cited in Jour. Amer. Assoc. Instructors and
Investigators in Poult. Husb., 1920, 1, VII, p. 3.
(16) Hadley, F. B., and B. A. Beach: Controlling chicken-pox, sore-head,
or contagious epithelioma by vaccination. Proc. Amer. Vet. Med.
Assoc, 1913, p. 704. Cited in Jour. Amer. Assoc. Instructors and
Investigators in Poult. Husb., 1920, 1, VII, p. 3.
(17) Beach, J. R.: The Treatment and prevention of chicken-pox (con-
tagious epithelioma) of fowls. Jour. Amer. Assoc. Instructors and
Investigators in Poult. Husb., 1920, 1, VII, p. 3.
(18) McNuTT, S. H.: Vaccination of poultry. Jour. Amer. Vet. Med.
Assoc, July 1926, 4, XXII, 472-77.
Publication of this Document
Approved by the Commission on Administration and Finance.
Massachusetts
Agricultural Experiment Station
TECHNICAL BULLETIN No. 11 NOVEMBER, 1927
INTENSITY OR RATE OF LAYING
IN RELATION TO FECUNDITY
By F. A. HAYS and RUBY SANBORN
This bulletin continues the series dealing with inherited traits in relation
to fecundity in the Rhode Island Red breed of domestic fowl. Intensity is
an inherited trait which vitally affects fecundity. In this study four meas-
ures of intensity have been used: first sixty-day egg record, mean size of
winter clutch, net winter rate of laying, and cuinual rate of laying. From
the standpoint of the breeder, mecin size of winter clutch is the most
satisfactory criterion of intensity because it can be accurately determined
and because it is inherited.
Keqiiests for l)ulletin.s sliould be addressed to the
AGRICULTURAL EXPERIMENT STATION
AMHERST, MASS.
INTENSITY OR RATE OF LAYING
IN RELATION TO FECUNDITY
By F. A. Hays and Ruby Sanborn
INTRODUCTION
Intensity or rate of laying was first cited by Goodale and Sanborn (1922)
as of vital importance in relation to total number of eggs laid. The sig-
nificance of intensity in relation to winter fecundity' has been further
stressed bj' Hays (1924) and a theory concerning inheritance of winter
intensity proposed. The probabilities are that higli winter rate behaves
as a dominant in inheritance and that two Mendelian factors are concerned.
Possible measures of intensity are numerous and diverse. A number of
.such measures will first be considered.
Meaisures of Intensity.
Since rate can only be obtained on a time basis, it becomes necessary to
<nake use of some definite time interval. Furthermore, the time element
must be based either on calendar months or upon a specific period in the
laying year of the individual bird. From the biological standpoint the
second method of assigning time in calculating rate might be preferable,
yet diverse weather conditions throughout the year may partially nullify
effects of biological differences.
Goodale and Sanborn (loc. cit.) suggest as time units the month, the
initial cycle, the inter-broody periods, the summer period and the spring
period. These workers also mention length of clutch as a possible meas-
ure and make the statement that the calendar montli may be employed for
flock comparisons though unsuitable for individual comparisons.
A rather common measure of intensity is the greatest number of eggs
laid by a bird during any calendar montli of the pullet year. Such a
criterion of intensity is certainly very crude, as Harris and Goodale (1922)
have shown for Rhode Island Reds. Probably a long time interval, gov-
erned in part by the individual date of first pullet egg and terminated at
a definite calendar date, offers the most reliable period for calculating
intensity.
In these studies four measures of intensity liavc l)een used:
i. First si.vti/ daiftt' prodiirtion.
The number of eggs laid from the first pullet egg for a period of sixty
days was tabulated. This figure represents a definite interval in the lay-
ing year and furnishes a clue as to the rate at winch the birds begin lay-
ing. In this instance the actual calendar date varies both with hatching
date and with age at which sexual maturity is attained. In other words,
the production of the first sixty days is not comparable with production
tor two specific calendar months.
KMK OK l.A'^INC AND I'.CC I'UODIH IION IM
•_'. Menu iciiittr rlittch nizf.
Tlie term clutch repres«^iits tlic mmilicr nf cfifi" l;ii<l on Miccc.><si\ c (biys.
I'lu- size of clutcli vjiries widely, witli a mean of from 1 to 21 ep:gs in the
population .studies. The total number of clutches from first egg to March
tirst was tabulated for each pullet. The number of eggs up to March first
was then divided by the number of clutches to obtain the mean winter
clutch size. In this manner ;i miMsurc of intensity was jirrivcd ;it without
employing the time interval. Moreover, it offers an index to the cajiacity
of the rei)r(>ductive organs for elaiiorating eggs on successive days. An
cb.jectionablc feature is apparent in that no evidence is ol)tained relative
to the ability of the birds to continue producing a definite clutch size.
'A. Xet xc'iiitf-r rate.
Net winter rate is determineil by dividing the total number of eggs
from first pullet egg to March tirst by the number of days from first egg to
.March first less all pauses of four or more days from Xoveml)er first to
.March first. This method expresses the rate of laying excluding winter
pause and an occasional liroody pause before March first. Compared with
mean clutch length, net winter rate is probably a sujierior measure of in-
tensity because it covers a definite time limit.
\. .linuKil rate.
Annual rate is indicative of the gross rate of laying from first pullet
egg to the onset of complete molt. This rate is obtained by dividing the
total eggs laid during the period by the number of days from first egg to
the beginning of complete molt in the late summer or fall if within 364
(lays from the date of first pullet egg. In cases where molting does not
begin until after the 365-day period, time is figured at 365 days. The rate
thus obtained does not make allowance for time lost in winter pause or for
iionproduction during broody ])eriods. Such a rate is, therefore, but a
crude ai)i)ro\irii.itioii of annual intensity.
Character of Birds Used.
The birds used in these studies are identical with those reported (ui in
.NFassachusetts .Agricultural Experiment Station Technical Bulletins 7, 8
and 9, except that data on the flock hatched in 1925 are added. All records
were made during the pullet year and all birds are pure bred Rhode Island
Reds. The flocks are somewhat heterogeneous in character including, in
addition to the major portion bred for high fecundity, a smaller propor-
tion bred each year for non-br()odiness, intense broodiness, high batcli-
..bility, good color, and inbreeding studies.
Scope of This Report.
-Vs previously stated, four measures of intensity are made use of here.
.Vttention is given to some prominent environmental influences and to in-
lierited traits that may affect intensity. Major consideration is given to
tiie relation of intensity to winter and annual jiroduction. The report is
divided into sections .V. B, (" and 1) oti the basis of the four criteria of
intensity emjiloyed.
182 TECHN'ICAI. lU'I.I.Kl IX II
A. FIRST SIXTY DAYS' PRODUCTION.
In dealing' with a cniiiiilfx liiolojiical and ])liysii)|ofiical prohlf in like
t'cH'iindity, strict attcntioji mnst hf siixen Imtli to inht-rited and to environ
mental factors. Strictly speaking' the lirst sixtx days' egg' record of two
or more iiullets woidd only be coin])aralilc if tlic i)irds were hatched tlic
same day, hefiJUi layinji' the same day and at all times were fed and man-
aged identically. I'nder such restrictions nund)ers wonid be so limited
as to be of (juestionable wortii. In studying the relaticm of the lirst sixty
days' ]irodnction to fecundit>, tiie same ))ro(hicti\e period is used for eacli
bird even though the actual calendar time \aries with liatching date and
age of sexual maturity. For these reasons, this should be a sui)erior mea-
ure to that employing one or two calendar months for all birds.
1. Cnn-tliiri(ni lietxcccii First Si.rli/ ptii/x' I' mil iirtKin mid Siilixciiiifiil \]'iiilir
rroihn-tinn.
\ population of ;i.5 1-'2 birds was used to discover the corrclaticm between
the first sixt>' days' record and jiroduction for the rest of tlie wintc
l)eriod. In this study only birds that had laid for sixty days by Marcli
first could be included. The correlation coetfi<-ient shows any tendency
of ])rodiicti(m during the lirst sixty days and sul)se(pient production u]i tn
March first to m()\e together. The following constants were obtained:
Nund)er of birds .'15 42
Mean sixty-day record 37.83
.Sixty-day record standard deviation ±9.95
Mean subsequent winter record 3n.(i.'5
Subse(]uent winter record standard deviation ±17.96
CoetHcient of correlation -I-.3-1-J-.5 zn. 01(10
The fact should be observed that the average length of laying jierio 1
up to March first for this population was 123 days. The first sixty-da>
record therefore covers about half of the winter laying jieriod. It is inter-
esting tb note that the mean numl)er of eggs laid during the first sixt\
days is 37.83 while the mean number laid during the next sixty-three days
was 30.65. The variability in production during the second half of the
winter period is also much greater than during the first half. Since winter
pause is more frequent in January and February than during previous
months, lower production during these months might be anticijiated.
The coefKcient of correlation between lirst sixty-day record and subse-
quent winter record is significant, and indicates that the ])nllets which lay
tne most during the first sixty days of their year tend to continue at a
higher rate than do those of less intensity during the first sixty days. The
absolute magnitude of the coefficient is, however, scarcely adeciuate for
selection purposes when maximum M'inter records are desired.
2. CorreJidioit lieticeen First !<ixtii l>iii/s' I'md net ion miil Sul'siiiiwii! Aiiii:i<il
Production.
Accurate methods for predicting ])robable annual etig yield are wanting.
The discovery of reliable criteria in advance would mark an importani:
step in progress and would be of outstanding value in selecting pullets for
H A ri-: ()i- I \^ INC \N!) i:(i(; ruonrc rioN ih'a
cfifi-liiyiiij:' (■(intfst.s ;i>- \m-I1 .i^ in m.-ikiiij: iiji lirccdiiifi Hock.s iiiul (•(iinriicr-
(•i;il prodnrtidii fl()ck>. Tlu' v.iliic of cnlx scmi.iI iiiiitiirity in flock seloc-
titiii ha.s ;ilTr,i(l\ Ix't-ii ]M)int»'d out (ll;iys and IW-iiiu'tt, Ift'iH), yet scxnal
inatnrit> is inadf(|uate as tiie soli- ;iiii(i<- of tlic l)ii'cdt"r in ins selections.
If siinic sliort-tinie test can be diseox trt-d, its luactical worth is self-evident.
A poimlation of 'Jiifjo birds is studied ovef the i)eiiod from Ifllfi t(>
I !>•_'."). The si\t\-ii,i> eiifi record of each individual is tabulated ajtainst her
jiroduetion for the remainder of the scar, a time interval of MD.t days.
Constants obtained follow:
Number of birds 2o<j(>
Mean sixty-day record 37. 8()
.Sixty-day record standard deviation it9.94
.Mean subsequent annual record lti.27
Subseciuent annual production standard deviation — ■tO.O+
Coefficient of correlation -f-.3082 ±:.0121
Comparinii the variability in sixty-day record with the variability in
rubseijuent annual record, the coefficients of variability are 26 per cent
;Mid 28 ])er cent, respectively. Theoretically, tbe standard deviation in
rmnual record should be five times as great as tbe standard deviation in
sixty-day record because the time interval of the former is ien months, and
of the latter two montlis. In reality, the two constants are about the
same, showiiifi that egg production fluctuates most during the early months
of the pullet laying year.
A positive coefficient of correlation of ..■J082 ±.0121 indicates a signif-
icant tendency for heavy sixty-day production to be associated with heavy
jiroduction for the remainder of the year. The degree of correlation is
somewhat less than the — .1380 ±.0131 reported by Hays and Bennett
Moc. cit.) between age at first egg and annual egg yield. The probabilities
are that selection upon first sixty days' record as a partial measure of
intensity, and upon age at first egg as a?iother valuable criterion, will in-
crease fecundity.
B. MEAN WINTER CLUTCH SIZE.
An expression for the clutch si/e of a pullet furnishes information rela-
tive to iier ability to elaborate few or many eggs on successive days. In
other words, it is an index to functional capacity. Mean clutch size
tliroughout the winter season otters ])ossibilities as a measure of intensity
over a considerable period of time. Moreover, clutch size can also be
(Icfinilely measured for each individual and measurable characteristics are
most useful in biological studies.
Behavior of Clutch Size in Inheritance.
.V frecjuency distributi<m of any of the flocks included in tin's report with
regard to clutch size clearly presents a bimodal aspect. A more exact
classification of each individual bird for clutch size places the modes at a
clutch size of 2 and 2.2, respectively. There is a very pronounced depres-
sion in the frequency .sraph at the clutch size of 2.1. The freciuency dis-
tributicm for clutch sizes from 1 to 2 rather closely approaches a straigrht
line with a positive slope not far from 1. On the other hand, frequency
(iistribuf ion for clutches greater than 2 is less regular and if fitted to a
184. TECHNICAL BULI.F/JIN 11
straight line gives a less abrupt negative slope. An examination of tliese
frequency distributions has brought to light \aluabl(' information concern-
ing the inheritance of winter clutch size.
Goodale (1918) recognized the fact that characteristic types of rhythm
in laying exist in Rhode Island Reds. He applied the time element and
classed hens as one-half, two-thirds, three-fourths, etc. with respect to
rhythm. He fully recognized the importance of rhythm in egg laying.
Riddle (1925) presents data to show that the common pigeon lays the
greatest percentage of single eggs during January and February and the
smallest percentage during July and August. Since the characteristic
clutch of the pigeon consists of two eggs, it seems probable that adver.se
weather conditions tend to reduce the rate of laying and in consequence
the mean clutch size.
Daily egg records of the wild ancestors of the domestic fowl are not
available for study in comparison with the records of improved flocks.
Information on the question of clutch size must for this reason be obtained
on flocks of domestic fowl largely unimproved in fecundity and from some
information on wild species of birds.
The mean winter clutch size of the foundation birds hatched in 1912,
from which the flocks reported upon are descended, is 1.9 for the 119 birds.
The mean for the 276 birds hatched in 1916 is 2.5, and for the 541 birds
hatched in 1925, 3.1. The normal clutch size for the common pigeon is 2.
but this normal may be modified by weather conditions as Riddle (loc.cit.)
shows. Data on the Massachusetts flocks of Rhode Island Reds hatched
from 1916 to 1925 indicate the existence of two modal classes for winter
clutch size and probably a third modal class higher than these two. The
first mode occurs at a clutch size of 2 and is very distinctly separated
from the second mode, which is about 2.2. The third mode probably is
at the 2.6 class. What then is the behavior of clutch size in inheritance?
Proposed Theory.
Extensive studies of available data on Rhode Island Reds indicate that
the normal clutch size for the domestic fowl is two and that adverse
weather conditions may operate to reduce some of the clutches to one.
Further bearing on this point is the fact that the hen not infrequently
liberates two ova almost simultaneously making a double-yolk egg, but
that the occurrence of more than two ova in the same shell is an extremely
rare phenomenon. It seems probable, therefore, that the hen ordinarily
ovulates twice either on the same day or on successive days, and that a
greater length of clutch than two represents a modification of the normal.
The first modal class appears to consist of the normal individuals that
ovulate twice in favorable environmental conditions and thus have a char-
acteristic clutch size of two or less than two. The second modal class
In the flocks studied is made up of birds modified for clutch size, so that
their mean is greater than two. The third modal class occurs at a clutch
size of 2.6.
On a Mendelian factor basis the following seems warranted after very
extensive studies of clutch size in families of sisters: That the normal un-
improved hen is a recessive for clutch size. That a gene I added to normal
gives a clutch size greater than 2. That a second gene I' makes a clutch
size of 2.6 or more possible. That genes and I and I' together give the
RATK OK LAYING AND ¥XiG I'RODl'CTION 185
greatest cliitoli size — more than 2.6. Both genes are autosomal and no
linkage has been observed to date. The four general classes of hens with
regard to clutch size are: i i i' i' individuals with a clutch size of 2 or less;
I I i' i' individuals with a clutch size of 2.1 to 2.5; i i I' I' individuals with a
clutch size of 2.6 or more; and I I I' I' individuals with a clutch size
greater 'than 2.6, and possibly as great as 21 or more for the winter
season.
3. Correldtioii Betxceen Hidcliniii lUtte mid Mcmi Cliilch Size.
Ilatcliiiig date is a controllable environmental condition. It may be
varied at the will of the breeder. In the series of years covered by these
studies the hatching dates have been kept on the same calendar dates. The
first hatcli came off each year on March 2.5 and there was one hatch each
week thereafter until May 15, or a total of eight hatches per year over a
period of 49 days. If time of hatching is associated with size of winter
clutch, it may be discovered by means of the coefficient of correlation.
The population consists of 3867 birds upon which constants were calculated
as follows:
Number of bird.s 3867
Mean hatching date (Apr. 19) 4.35
Hatching date standard deviation ±2.28
Mean size of winter clutch 2.64
Winter clutch standard deviation ±1.29
Coefficient of correlation — .0167 ±.0108
Clutch size exhibits a variability of 49 per cent as shown by dividing
its standard deviation by the mean clutch size. This striking lack of uni-
formity in clutch size is in no small measure responsible for great varia-
bility in winter egg records of these flocks. No correlation is shown between
hatching date and clutch size.
4. Correlation Between Age at First Egg and Mean Clutch Size.
In this section the relation between two inherited traits is brought to
light. Both are of significant importance in breeding for fecundity and
any linkage relation should be understood. The constants calculated for
the population follow:
Number of birds 3867
Mean age at first egg 206.18
Age standard deviation ±29.52
Mean size of winter clutch 2.64
Winter clutch standard deviation ±1.29
Coefficient of correlation — .2273 ±.0109
Age at first egg shows a moderate negative correlation to winter clutch
size. In other words, there is something of a tendency for early-maturing
pullets to lay larger clutches than do later-maturing birds. Here then
is one of the reasons for the pronounced negative correlation between age
at first egg and winter egg record (Hays and Bennett, 1923). There
appears to be a significant linkage between early sexual maturity and
]86 TECHNICAL BULLETIN 11
large winter clutch size tliat may be employed advantageously in breeding
for egg production.
5. Corrchition Heticeeii M'eufhf tif Firnt J'^;/;! <ni<l Menu Clutch Size.
Because of the importance both of body weight and of clutch size in
breeding for fecundity, it is very desirable that their relation to each
other be ascertained. The correlation between weight at first egg and
mean clutch size was calculated with the following constants;
Number of birds 3797
Mean weight 5.53
Weight standard deviation ±.72
Mean size of winter clutch 2.65
Winter clutch standard deviation ±1.29
Coefficient of correlation — .1714 i!=.O106
The coefficient of correlation is found to be iiegiitive and its absolute
magnitude is not very great. Statistically, however, the correlation is
significant and suggests something of a tendency for smaller birds to ex-
hibit larger clutches. For purposes of prediction or selection the relation
is not sufficiently pronounced to be of value.
(j. Correhifi'Hi lieticeeii ]\'uitcr l'n><h(rt!i»i hikI Menu Clufcli Xize.
Heavy winter egg production is important both genetically and economic-
ally. Genetically, winter egg record depends on seven pairs of Mendelian
factors as has been shown by Hays (1924), and winter production is also
known to be intimately correlated with annual production (Hays, Sanborn
and James, 1924). Economically, the number of winter eggs is of no small
value in affecting the net income per bird. In breeding for high fecundity
it is necessary to recognize both environmental influences and hereditary
factors that are concerned in winter egg yield. By tabulation of the entire
population for clutch size and winter record the constants below were
secured:
Number of birds ;^867
Mean winter production 62.92
Winter production standard de\ iation ±25.95
Mean size of winter clutch 2.64
Winter clutch standard deviation ±1.29
Coefficient of correlation +.4727 ±.0084
The above coefficient of correlation suggests a rather intimate positive
relation between clutch size and winter record. This constant may be
considered a true measure of correlation and indicates an important rela-
tion between clutch size and total egg yield for the winter. In view of
this fact large clutch size should be considered as a vital factor in breed-
ing for maximum winter egg records.
7. Correlation Beticeen Annual Record and Mean Clutch Size.
Clutch size appears to behave in Mendelian fashion in inheritance. It
KATE 01' LAYlNc; AND KCit! I'RODICTION 187
l;;is been slunvii to be rather significantly related to winter fecundity.
I'robably a more important consideration is the relation of clutch size to
annual production. The entire population with annual records has been
tabulated in a correlation table and the constants obtained follow:
Xuniber of bird.s 2;j.'}2
Mean annual production 182. 80
Anniial production standard deviation ±42.91
Mean size of winter clutch 2.7(t
Winter clutch standard deviation ±1.29
Coefficient of correlation +.3.544 ±.0117
On the studies reported herein egg production over a period of 36.5
days from first egg is taken as the standard for measuring fecundity. The
coefficient of correlation is positive and certainly significant so that winter
clutch size may be employed as a valuable criterion in the selection of
prospective heavy annual egg producers. Since winter clutch size appears
to be an inherited trait, there is opportunity for increasing fecundity by
breeding for greater mean winter clutch size.
C. NET WINTER RATE
Net winter rate is an expression for the rate of laying throughout the
winter season after deducting time spent in winter pause. It represents
the rate of laying for an average time interval of about 120 days in the
population studied. Net winter rate is considered first in relation to date
of hatching.
S. Correlutiiin Betic^en Hatcliiixi Utile <(n<l Net }\'inter Rale.
The population for the ten-year period (1916-192.5) is considered, and
tlie following constants are derived:
N'uini)er of birds 3863
Mean hatching date (Apr. 18) 4.3-5
Hatciiing date standard deviation ±2.28
Mean winter rate 67.79
Winter rate standard deviation ±8.86
Coefficient of correlation -f.0'100 ±.0109
.ludgeil l)y the magnitude of the coefficient of correlation, there is no
relation between hatching date and net winter rate of laying. This fully
agrees with the findings set forth in section 3 where hatching date and
winter clutch size are found to be independent,
f). Correlation Between Age at First Egg and Net Winter Rate.
Age at first egg is an inherited characteristic which has been shown by
the writer as well as by many other workers to be intimately correlated
with winter and annual production. In this report an attempt is made to
discover the relation of intensity to fecundity as well as to other character-
istics concerned in fecundity. Age at first egg has therefore been tabulated
against winter rate to derive the constants below:
188 TECHNICAL BULLETIN II
NuMber of birds 3863
Mean age at first egg 206.14
Age at first egg standard deviation ±29.47
Mean net winter rate 67.79
Winter rate standard deviation ±8.86
Coel/icient of correlation ■ — .2274 ±.0103
A moderate degree of negative correlation is found between age at first
egg and winter rate of laying. Evidently there is some tendency for the
early maturing pullets to lay somewhat more intensely than do the late
maturing individuals. This difference is no doubt due to larger clutch size
in the earlj- maturing birds as has been pointed out in section 4. The fact
should be noted here that the degree of correlation is identical between
age and clutch size and between age and net winter rate.
10. Correlation Between WeUjlit at First Ef/fi and Net Winter Rate.
Mature body weight in poultry is inherited on a Mendelian basis accord-
ing to Punnett (1923). Weight at first egg has been shown by Hays,
Sanborn and James (1924) to depend both upon hatching date and upon
age at first egg. If there is any relation between weight at first egg and
winter rate of laying, body weight might be used as a partial criterion in
selection. To gain this information the constants below were determined:
Number of birds 3794
Mean weight at first egg 5.53
Weight standard deviation ±.72
Mean net winter rate 67.81
Winter rate standard deviation ±8.82
Coefficient of correlation — .1756 ±.0106
The above coefficient of correlation is of the same magnitude as was
the coefficient of correlation reported in section 5 between weight at first
egg and winter clutch size. Again the correlation coefficient is statistically
significant but of little practical value for prediction and selection pur-
poses.
11. Correlation Between Winter Egg Production and Net Winter Rate.
To discover the relation between net winter rate and winter egg record,
the population of 3863 birds is studied. The following constants appear:
Number of birds 3863
Mean winter production 62.95
Winter production standard deviation ±25.92
Mean net winter rate 67.79
Winter rate standard deviation i ±8.86
Coefficient of correlation +.5444 ±.0076
The above constants indicate an intimate positive correlation between
winter rate and winter production. This correlation points rather con-
clusively to the importance of winter rate as a factor in winter egg pro-
duction. Since winter rate is inherited (Hays, 1924), it becomes evident
that one important means of securing high winter production lies in the
development of a high winter intensity strain.
HATK OK LAY INC AM) V.GG I'HODUCTION 189
12. (.'orri'ldfiim lUtxcicii .liiininl l''.<l<i Ixtcoril (iitd A'<7 Winter Rule.
Net wiiittM- rate lias been considered one measure of intensity or rati"
of layinii. There yet remains to be studied the relation between winter
rate and annual egfT yield. A correlation table has been constructed for
the entire population and tlie followinp,- constants have been arrived at:
Xiunhrr of birds 2528
-Mean annual egg" record 182. 9()
Annual egg record standard deviation ±-t2.83
Mean net winter rate 68.27
Winter rate standard deviation ±8.81
Coefficient of correlation +.4769 ±.0104
A rather intimate correlation exists between net winter rate and annual
production. As a criterion for predicting annual egg production, net
winter rate is superior to any other thus far considered. This fact would
seem to indicate that the rate of laying up to March first is an index to
the probable rate throughout the year.
D. ANNUAL RATE OF LAYING
Annual rate is taken as a long-time measure of intensity. The time
interval employed in calculating the rate for each individual is the persist-
ency interval or period from first pullet egg to complete molt. The mean
time interval for the 1916 flock is 248 days, and the maximum time interval
was in the 1922 flock with 330 days. For mean persistency of other flocks,
see Massachusetts Agricultural Experiment Station Technical Bui. 9. An-
nual rate is a gross representation of the percentage of time in the produc-
tion year that each bird actually laid. The relation of annual rate to hatch-
ing date, age at first egg and weight at first egg is first considered, and
finally the relation of annual rate to annual production.
13. Correlation JietiCeeii Hatrhini/ hate and Annmd Rate.
Date of hatching is known to l)e of considerable practical importance.
In previous publications it has been shown to influence age at sexual
maturity, winter pause, and annual persistency, as well as rate of growth.
It is therefore desirable to know if hatching date shows any influence upon
annual rate. The population over the ten-year period was tabulated in a
correlation table which gave the constants below:
Number of birds 2560
Mean hatching date (Apr. 18) 4.29
Hatching date standard deviation ±2.26
Mean annual rate 57.57
Annual rate standard deviation ±9.57
Coefficient of correlation +.0318 ±.0133
The mean annual rate above shows that the birds averaged to lay 57.57
per cent of the time between first egg and the onset of complete molt.
This population of 2560 individuals had a mean annual egg production of
181.59. Simple calculation of the time interval shows it to be 315 days.
\9() TEC'HN'ICA[. I'.l'I.I.KTIX 11
!l i.s possible that aniuuil ratf mi.iilit lie iiUTt-ased hv proper inctlidds of
breeding.
The eoerticieiit of corri'lalion Ix'twcen liatchiiiii (late and annual rate is
of very small magnitude and not statistieaJly signitieant. it can on]\ 1) ■
interpreted as meaning tliat annual rate is indei)endeiit of hatching date.
1-t. CorrehitUnt Heticeeii .li/i "/ l-lr.s-/ l'-'j;i "ii</ .liiininl liith-.
Age at first egg was sliown in section •") to be signiru-antiy correlated
with winter ehitch si/e, and in section 9 to be correlated to a similar degree
with net winter rate. Of im])ortance now is a consideration of tlie relation
between age at first egg and annual rate. .Vll of the birds with annual
records are included in the correhition taliie to derive the <-onstants:
Number of l)irds 25(i0
Mean age at first egg 205.37
Age standard deviation ±30.56
Mean annua] rate 57.57
Annual rate standard deviation ±9.57
Coefticient of correlation — .Ofi57 ±.0133
The magnitude of the coefficient of correlation given abo\e is insufficient
to establisii any relation between age at first e^^ and annual rate of lay-
ing. The assumption therefore seems warranted that age at first egg and
annual intensity are independent of each other.
15. Correliilion rxliocn W'c'nilil n! First />_'/// "//'/ .liniiiiil Rale.
Body weight at the beginning of the laying year might be thought of
as a crude measure of capacity for food consum])tion and as such body
weight might be correlated with annual intensity. The entire population
was assembled in a correlation table to discover jiossible relations. The
constants calculated are as follow ■-:
Nund)er of birds 250 1
Mean weight at first eg.a 5.5 t
Weight standard deviation ±.73
Mean annual rate 57.62
Annual rate standard deviation ±9.57
Coefficient of correlation — .llTi ±.0133
Tiie degree of correlation shown between body weight at first egg and
annual intensity is small but .statistically significant. There is but a slight
tendency for smaller birds to exhil)it higher annual rates. This correlation
may be attributed to the somewhat greater intensity of early maturing
pullets and such pullets would normally show lower body weights than
later maturing pullets becau.se of the time element. The degree of corre-
lation formed above is of no practical importance, as it is too slight to be
used for prediction or selection purposes.
i<\ri-. oi- i.v^iMi AM) i-,c.(i I'liODi ( rioN nn
111. ('iin-> III' iiiit liil\ciiii .liininil I'.ilil I'liniil niiil .liiininl h'uli.
Ill tlif coiicliKliiii; si'ction of' tlu->i- >lii(lirs, tlic (•(•rrrlal inn lu'twt't'ii ;mi-
!iii;il r.itf ;nul ;utmi;il protliict idii is ctuisidrrcd. Siicli studies will briiijf to
lirht soim-tliinj: of tin' iniixn-tanct- of iiitciisitv nuMsurfd over a long period
(tf tiiiii' as a t',uti)r in aiimial ejiji yield. In makiiin' ii)) the correlation
lal)li' tlie aniuiai recDrd of eacli individual \n as tal)nlatcd against her in-
tciisitv i-fc'Ofd. riif constants a])pear lit'low :
NnniinT of liirds 2o(iO
Mean annual produftiint 181.59
Annual ])rodiiction standai-d deviation iJ'-i.SO
.Mean annual rate 57.57
Annual rate standard (h-viatinn ±9.57
C'oeftieient correlation +.710() ±.00fi()
The con-elatiiui l)et«een annual i-ate and annual i)roduction is po.sitivc
and dccidedl\ intimate. .\s ,i criterion of annual ])rodiiet ion, annual rate
is of outstandin.i: \alue. Only out other criterion of annual i)roduction,
namely, annu.il ])ersistency (llays and .Sanliorn. 192(i b), shows a .similar
dcfiree of correlation to yearly iiroductiiui. .\nnual intensity, therefore,
shoidd he ade<|u itel_\ stressed in a iiro^irani of hreedinp.' for fecundity.
The Measures of Intensity Compared.
In these studies four nieistM'cs of intensity have been considered in re-
lation to winter and annual eji'.ir ])roihiction. The measures of intensity
<-ini)loyed are: (a) first sixty-day record: (h) mean size of winter clutch;
(c) net winter rate of laying': and (d) annual rate of laying'. Ba.sed on the
intimacy of con-elation with \\inter )n-oductioii three criteria rank as fol-
lows: net winter rate, mean winter clutch size, and first sixty-day jirodnc-
tion. On t!ie intiniaty of cnrreiation with yearly ])roduction the rank is
annual rate, net w intei- rate, mean winter clutch si/.e, and first sixty-day
record.
.Vs sho)-t-time measures ot intensity net winter rate and mean winter
clutch size are siijyerior to first sixty-day ejifi' record, .ludfied by the intimacy
of correlation with both winter and annual eji'K records, net winter rate is
somewhat siijjerior to mean clutch si/.e. Both show the same relation to
liatching date, a.ae at first egfi' and weijiht at first egg. Mean clutch
^izc is more definite than is net winter rate because short ])ause intervals
atfect net winter rate and apparently do not aflt'ect clutch size. Mean
( Intch size has been shown to behave as an inherited trait and to depend
upon two autosomal genes. From the breeding standjioint, therefore,
winter clutch size may be considered the best measure of intensity thus
far de\elo)-)ed. since annual rate is valueless for jirediction ])url10';^s and is
known onl\ after the first laying year closes.
192 TECHNICAL BULLETIN' 11
RELATION OF AGE AT FIRST EGG, BROODINESS, AND INTENSITY
Age at first egg has previously been shown to be an inherited trait
(Hays, 1924), and to be intimately correlated with fecundity (Hays and
Bennett, 1923) in the Rhode Island Reds being studied. It is desirable
to discover if age at first egg and degree of broodiness are dependent or
independent.
17. Correhtfioti Between Acje at Firtit Eijg and Total Days Broody.
The broody population for the ten-year period has been tabulated in
a correlation table for age at first egg and total days broody during the
pullet year. Constants obtained are as follows:
Number of birds 1207
Mean age at first egg 207.76
Age standard deviation ±31.52
Mean total days broody 42.81
Days broody standard deviation ±27.41
Coefficient of correlation +.0062 ±.0194
The magnitude of the coefficient correlation between age at first egg
and total days broody is insufficient to indicate any correlation between
age at first egg and degree of broodiness. Because of this fact it is very
probable that these two inherited characteristics are in no way linked in
inheritance.
Age at first egg has been shown in sections 4 and 6 of this report to be
rather intimately correlated with winter intensity. Apparently there is
linkage between the genes E and E' for early maturity and genes I and I'
for large clutch size. Tlie degree of linkage has not been determined as
yet.
18. Correlatidii Betiveeu Total Dayts Broody and Mean Winter Clutch Size.
Hays and Sanborn (1926a) report no significant correlation between
degree of broodiness in the broody population and net winter rate of lay-
ing. Since mean winter clutch size is a useful measure of intensity, it is
desirable to ascertain if the degree of broodiness is correlated with mean
winter clutch size. The following constants liave been calculated for the
broody population over the ten-year period:
Number of birds 1188
Mean total days broody 42.89
Days broody standard deviation ±27.54
Mean winter clutch size 2.96
Clutch size standard deviation ±1.38
CoeQicient of correlation +.2205 ±.0186
In section 3 the mean winter clutch size of both broody and non-broody
birds is 2.64. The mean winter clutch size of the broody birds alone is
found to be 2.96. This fact suggests that, on the average, broody birds
tind to lay in larger clutches during the winter than do non-broody birds.
The coefficient of correlation between total days broody and mean win-
ter clutcii size is statistically significant, and suggests that degree of
HA1"K OF LAYING AXD KGCI PRODUCTION 193
broodiness is in a incasuro positively forrelated witli winter clutch size.
From the standpoint of annual fecundity, however, the broody trait should
he eliminated as Hays and Sanborn (loc. cit.) show.
GENERAL SUMMARY.
Four measures of intensity have been considered in relation to fecund-
ity; namely, first sixty-day egg record, mean size of winter clutch, net
winter rate, and annual rate. The data have been secured over a ten-year
l)eriod on succeeding flocks from the same foundation with reasonably
constant environmental conditions. From the standpoint of tlio breeder,
mean size of winter clutch is the ujost satisfactory criterion of intensity
because it can be accurately determined and because it is inherited in
Mendelian fashion.
In studying the correlation between these four measures of intensity
;ind different environmental conditions and inherited traits affecting fecund-
itj- as well as their correlation to winter and annual egg production, the
following facts appear:
1. There is a positive correlation between first sixty-day egg record
and subsequent winter record of .3445 ±.0100. Such a correlation shows
something of a tendency of production for the first sixty days to be asso-
ciated with a somewhat similar production for the next two months. But
production for both the first and last half of the winter period is often
reduced by the onset of inherited winter pause, making the absolute num-
ber of eggs laid during any part of the winter season an unreliable cri-
terion of intensity.
2. The number of eggs that a pullet lays during the first sixty days
of laying is correlated with the number she lays for the remainder of the
year; yet the degree of correlation is less than with some other measures
of intensity.
3. Size of winter clutch is inherited on a two-factor Mendelian basis.
4. Size of winter clutch is not affected by hatching date.
5. Factors for early sexual maturity indicated by age at first egg arc
linked with factors for large clutch size.
6. Body weight at first egg and mean winter clutch size are negatively
correlated to a moderate degree.
7. Winter egg production is intimately correlated with mean winter
clutch size making clutch size a valuable criterion of winter intensity.
8. The correlation between annual egg record and mean winter clutch
size is positive and of sufficient magnitude to establish winter clutch size
as a good measure of intensity.
9. Net winter rate is very similar to mean winter clutch size in rela-
tion to hatching date, age at first egg, and weight at first egg.
10. Net winter rate is somewhat more intimately correlated with win-
ter egg yield than is mean winter clutch size, but the former is a less
specific measure of intensity alone because of winter pause.
11. Net winter rate shows a coefficient of correlation of +.4769 ±.0104
with annual egg record compared with a coefficient of -f-.3544 ±.0117 be-
tween mean winter clutch size and annual record; but the former is not
as valuable a criterion of intensity as the latter because of winter pause
disturbance^.
l!)i TKCHNICAI, HUI.LKTIN 11
12. Annual rate of laying is not affected by date of hatching.
13. Annual rate is but very sliglitly correlated with age at first egg
in a negative direction.
l-i. Annual rate .shows but a small negative correlation with bod\"
weight at first egg.
15. Annual rate is most intimately correlated with annual egg record,
the constant being +.7106 ±:.0066 for the population studied.
16. Annual rate is valueless for prediction purposes during the pullet
year and is l)ut a gross approximation of rate for the entire pullet laying
year.
17. Age at first egg and degree of broodiness are independent.
IH. A positive correlation between total days broody and mean winter
clutch size suggests that birds carrying factors for broodiness show a
tendency to lay in larger clutches during winter than do non-broody birds
The degree of correlation, however, is not sufficient to indicate that non-
broody birds may not carry high intensity.
1J». In general, intensity or rate has been shown to vitally aft'ect fecund-
ity; and as a short-time measure of intensity, mean winter clutch size ha.s
been suggested because it is definitely measurable and because it is in-
herited.
REFERENCES
Goodale, H. D. 1918. Internal factors influencing egg production in th''
Rhode Island Red breed of domestic fowl. Amer. Nat. 52:65-321.
Goodale, H. D., and Ruby Sanborn. 1922. Changes in egg production in
the Station flock. Mass. Agr. Expt. Sta. Bnl. 211.
Harris, J. Arthur, and H. D. Goodale. 1922. The correlation between tiie
egg production of the various periods of the year in the Rhode Island
Red breed of domestic fowl. Genetics 7:446-465.
Hays, F. A., and J. S. Bennett. 1923. Correlation of sexual maturity to
annual egg record. Poultry Science II. No. 6:205-206.
Hays, F. A. 1924. Inbreeding the Rhode Island Red fowl with special refer-
ence to winter egg production. Amer. Nat. 58:43-59.
Hays, F. A., Ruby Sanborn and L. L. James. 1924. Correlation studies
on winter fecundity. Mass. Agr. Expt. Sta. Bui. 220.
Hays, F. A., and Ruby Sanborn. 1926 a. Broodiness in relation to fecund-
ity in the domestic fowl. Mass. Agr. Expt. Sta. Tech. Bui. 7.
Hays, F. A., and Ruby Sanborn. 1926 b. Annual persistency in relation
to winter and annual egg production. Mass. Agr. Expt. Sta. Tech.
Bui. 9.
Pearl, Raynmnd, and Frank M. Surface. 1909. Poultry Notes. Maine Agr.
Expt. Sta. Bui. 165.
Punnett, R. C. 1923. Heredity in Poultry. New York, London, etc.: Mac-
millan and Co.
Riddle, Oscar. 1925. Studies on tlie plnsiology of reproduction in birds.
Amer. Jour. Phisiol. 73:15-16.
Massachusetts
Agricultural Experiment Station
Technical Bulletin No. 12 December, 1927
NET CORRELATIONS OF CHARACTERS
CONCERNED IN FECUNDITY
By F. A. HAYS and RUBY SANBORN
This bulletin completes the series dealing with the five inherited traits
concerned with fecundity in the Rhode Island Red breed of domestic fowl.
Net correlations are presented, which more adequately portray the relative
importance of the several characters than do the simple correlations pre-
viously used. Annual egg production is shown to be entirely independent
of apfe at first essi to be dependent to an important and substantially equal
degree upon length of winter pause, intensity as measured by winter
clutch size, and degree of broodiness; but to be most intimately affected
by annual persistency. The multiple correlation of -|-.8642 shows that
the five characters here considered largely control the annual egg yield.
Requests for bulletins should be addressed to the
AGRICULTURAL EXPERIMENT STATION
AMHERST, MASS.
NET CORRELATIONS OF CHARACTERS CONCERNED
IN FECUNDITY
By F. A. Hays and Ruby Sanborn
Annual egg })roduction has been found by several workers to depend upon
the characteristics early sexual maturity, non-pause, high intensity, non-broodi-
ness, and high persistency. All five of these traits have been shown to be
inherited in Mendelian fashion (Hays, 1924 and 1927). In order to obtain
niaxiinuin annual egg yield in the domestic hen, a combination of the above
five characteristics in the same individual is necessary, as well as very favor-
able environmental conditions. Familiarity with the mode of inheritance of
any one of the above five traits clearly indicates the difficulties in combining
all in the same individual and serves to explain the marked variability occur-
ring in the egg production of individual hens in the same flock. Even when
hens that carry all five of the desirable traits are used as brseders, there will
not be marked uniformity in production of the daughters of such matings
unless the breeding hens are genetically homozygous for the fecundity char-
acteristics and are mated to males of like genetic composition. Too much
stress cannot be laid upon the importance of selecting as breeders both males
and females that are homozygous for the characteristics sought.
In previous reports (Technical Bulletins 7, 8, 9, and 11 of the Massachusetts
Agricultural Experiment Station), the relation of the five fecundity traits to
production has been fully considered by means of the simple coefficient of
correlation and the relative importance of each trait as a selection unit has
been studied. There yet remains to be considered in this concluding report
the net relation of each of the five traits to each other as well as to annual pro-
duction when all characteristics except the two under consideration are made
constant. These results are obtained by use of partial correlation coefficients
as calculated by standard formulae from the simple correlation coefficients.
For illustration: in calculating the partial correlation between age at first egg
and annual production; winter pause, intensity, broodiness, and persistency
are made constant. The use of partial correlation coefficients here accom-
plishes a two-fold purpose. First, it brings out any possible linkage between
traits — an invaluable item of information to the breeder. Second, it shows
clearly the relative importance of the five characteristics to fecundity so that
each may be properly stressed in the breeding program.
The concluding section makes use of multiple correlation to discover the
exact degree in which annual production depends on the combined influence
of the five traits under consideration, and also whether other factors are con-
cerned in fecundity.
Birds Used
All of these studies are based on pullet-year records on Rhode Island Reds
bred by the Massachusetts Agricultural Experiment Station from 1916 to 1925.
The major portion of the birds was bred for fecundity, but each year there
has been included a limited number bred for broodiness and in inbreeding
studies.
198 TECHNICAL BULLETIN 12
Relation of Characteristics Concerned in Fecundity to Each Other
1. Age at First Eyg and Length of Winter I'amte.
Age at first egg may be definitely recorded and lias been used extensively
as a criterion of future productive ability. The simple correlation between
age at first egg and annual egg yield in these flocks is — .4380±.0134 (Hays
and Bennett 1923), liut this apparent correlation may be due to linkage re-
lations between early sexual maturity and other high fecundity traits. By
making intensity, days broody, and annual persistency constant, the net or
partial correlation between age at first egg and length of winter pause is
— .2236±.0139. Here is a significant correlation which suggests some link-
age between early sexual maturity and long winter pause. The relationship
is far from intimate, however, and probably does not signify that genetically
early-maturing birds need be handicapped by winter pause.
2. Age at First Egg and Mean Size of Winter Clutch.
The mean size of winter clutch is obtained by dividing the total eggs laid
from first egg to March first by the number of clutches involved. Clutch size
has been shown by Hays and Sanborn, (1927) to be a good measure of in-
tensity and behaves in inheritance on a two-factor basis. The simple correla-
tion between mean winter clutch size and annual egg yield is -}-.3.54.4±.0117
(Hays and Sanborn, loc. cit.). If early sexual maturity is linked with large
winter clutch size, there should exist a significant net correlation between the
two. The net correlation between age at first egg and mean winter clutch
size, after making the length of winter pause, total days broody, and annual
persistency constant, is - — .1879±.O105. This is a statistically significant con-
stant, but its magnitude does not suggest that early sexual maturity is in-
timately linked with high intensity.
3. Age at First Egg and Total Days Broody.
Total days broody during the pullet year is a good measure of degree of
broodiness (Hays and Sanborn 1926a). These workers also show in the same
report that degree of broodiness as well as the presence or absence of broodi-
ness during the pullet year must be considered in the breeding program. The
simple correlation between total days broody and age at first egg is-(-.0062
±.0194, a constant of magnitude insuificient to indicate any relation between
early sexual maturity and days broody. By calculating the partial correlation
between age at first egg and total days broody where winter pause, intensity
and persistency remain constant, the true relation between sexual maturity
and degree of broodiness is arrived at. The partial correlation coeflFicient is
found to be -]-.0473±.0194, which establishes independence between age at
sexual maturity and degree of broodiness.
4. Age at First Egg and Annual Persistency.
Persistency or long laying period at the close of the pullet laying year is of
great significance as affecting annual egg production. Hays and Sanborn
(1926c) report the simple correlation between annual persistency and annual
CHARACTERS CONCERNED IN FECUNDITY 199
production as -{-.7082 ±.0072. Tlu- siiiiplc correlation between age at first
esrg and annual persistency is rei)orted as — .6146±.0090. There appears to
be an important relation between age at first egg and persistency, and this
may be accurately determined by the partial correlation coefficient where
winter 'pau.se, intensity and days broody are made constant. The calculations
give — .5956±.0093 as the partial correlation coefficient. This constant in-
dicates an important linkage between iieritable factors for early maturity
and for higli persistency as has been previously pointed out by Hays (1927).
Therefore, by tiie use of breeding females that carry the early maturity
factor tliat is linked with the persistency factor, it is entirely possible to com-
bine the two desirable traits in the same individual bird.
J. Length of Winter Pause and Mean Winter Chitch Size.
The simple correlation between lengtii of winter pause and winter clutch
size is — .0674±.0145 (Hays and Sanborn, 1926b). This constant does not
suggest a significant relationship. By applying the method of partial correla-
tion where age at first egg, total days broody, and annual persistency are
made constant, the coefficient of correlation between length of winter pause
and winter clutch size is — .0874±. 014.5. This constant is of very small magni-
tude and probably indicates no significant linkage between length of pause
and winter clutch size.
6. Length of Winter Pause and Total Days Broody.
The simple correlation between length of winter pause and total days
broody is — .1832±.0243 (Hays and Sanborn, unpublished data). Such a
constant would indicate a tendency for intensely broody birds to exhibit
shorter winter pause than do less intensely broody individuals. Possibly a
short winter pause is compensated for the following summer either by longer
liroody periods or by a greater number of broody periods. By means of the
partial correlation coefficient the correlation between the two characteristics
may be calculated when age at first egg, winter clutch size, and annual per-
sistency are made constant. The partial or net correlation between length
of pause and total days broody is — .1609±.0245. This constant is statistically
significant and indicates a slight tendency for intensely broody birds to pause
for short periods in winter. There is apparently no linkage relation betw^een
the dominant genes for winter pause duration and the genes that intensify
broodiness, but rather a tendency for short winter pause to be associated with
a long period of broodiness.
7. Length of Winter Pause and Annual Persistency.
The simple correlation between lengtii of winter pause and annual per-
sistency is +•1^1'''— -0182. The magnitude of this constant does not warrant
the assumption of an important relation between length of pause and annual
persistency.
The partial or net correlation of winter pause and annual persistency, when
age at first egg, winter clutch size, and days broody are made constant, gives
the true relation of winter pause to persistency. The partial correlation
200 TECHNICAL BULLETIN 12
coefficient is — .0393±.0183 and indicates complete independence between dura-
tion of winter pause and annual persistency.
8. Mean Winter Clutch Size and Total Days Brood;/.
Broody birds exhibit some tendency to lay in longer clutches during winter
than do non-broody birds. Stated diflFerently, there is a somewhat higlier
winter intensity in broody than in non-broody individuals (Hays and Sanborn
1927).
The net correlation between winter clutch size and total days broody for
the pullet year is -j-.2079±.0187. There are, however, many non-broody in-
dividuals showing large clutch size and these birds should be used as breeders
in the production flock.
9. Mean Winter Clutch Size and Persistency.
Both large clutch size and high persistency are desirable from the stand-
point of fecundity. They should be combined in the same individual to secure
maximum egg production. The net correlation between winter clutch size
and persistency when age at first egg, winter pause duration and total days
broody are made constant is -j— Ol^^— •*51*^- Thus complete independence is
established between intensity and persistency.
10. Total Days Broody and Annual Persistency.
Degree of broodiness may be measured by tlie total days spent in broody
behavior during the pullet laying year. The loss of production during broody
periods has a pronounced eflFect in lowering annual egg records. On the other
hand, high persistency is associated with large annual records. The net cor-
relation between total days broody and annual persistency where age at first
egg, intensity and winter pause are made constant is -f-.0579±:.0209. Thus
there is shown to be no significant linkage between degree of broodiness and
annual persistency.
Relation of Characteristics Concerned in Fecundity to Annual Production
In this final study of the relation of characteristics concerned in fecundity
to annual production the relative net correlation of each of the five traits with
annual egg record is calculated. In this manner the true value of each trait
as a selection unit may be discovered, as the method of partial correlation
eliminates any possible effects from interrelation of characteristics concerned.
11. Age at First Egg and Annual Production
Age at first egg is a good measure of early or late sexual maturity in the
pullet and has been used rather extensively as a selection unit in making up
both laying and breeding flocks. The reason why early sexual maturity is
desirable cannot be discovered without a knowledge of the relation of sexual
maturity to the other fecundity traits as presented in sections 1, 2, 3, and 4
of this report. The possibility also exists that pullets that begin laying at an
early age are able to complete their annual record under more favorable
weather conditions than are later maturing pullets. Since the last two or
CIIAHACTKUS I ONcKHXI'D IX FKClXDri^' 201
tlircc nu)ntlis of tlic i>ulk't liiyiiiu- \c;\v mark lier irrcalcst susfrptiliility to
id\crse en\ ironnioiital infiuoiu'os, it is oiitin-ly probable tliat early inatiirily
may enable tlie bird to persist late, as pointed out in seition i.
Hy ap)ilyiiiji' tlie iiietbod of partial correlation, winler pause duration, in-
tensity, degree of broodiness and persisteney are made eonstant and tlie net
eorrelation between age at tirst egg and annual egg i)roduetion arrived at.
The net eorrelation between age at first egg and annual production is — .0238
:t.0177. This insignificant t'onstant clearly discloses that age in itself is not
associated with annual ]n"oductioii. Sections 2 ;uid !• make clear the fact
that early maturing pullets may carry slightly greater intensity and that they
beyond (piestion tend to exhibit greater persistency than late maturing pullets.
Thus the high yearly producer must be early maturing not because early
maturity itself is of importance but because early maturity has some linkage
with high intensity and very intimate linkage with persistency.
/,'. Letififh of ]V inter Pans-/' anil ^Inumd Pniduction.
The probabilities are that the domestic hen possesses functional capacity
to lay a rather definite number of eggs previous to the time when she must
cease laying and renew her reserve of materials necessary in the ctunplex
phj'siological processes of egg production. The mean length of the winter
cycle for the flocks studied is 52.26 days (Hays and Sanborn, 1926b) and the
mean length of winter pause is 32.26 daj's. The standard deviation in winter
cycle is ±34.23 days, showing great variability in the length of time the birds
may lay before the onset of winter pause. "With these facts in mind, the
negative correlation between age at first egg and length of pause observed in
section 1 may be understood. Furthermore, there exists an appreciable
negative correlation between early hatching and duration of pause as might
be anticipated. The tendency to pause in itself is governed by a dominant
inherited factor (Hays, 192 1-).
The net correlation of length of winter pause with annual egg record is im-
portant and is determined by making age at first egg, clutch size, total days
broody and annual persistency constant. This constant Is — .5487±.0128.
Here is shown a rather intimate negative association between length of winter
pause and annual egg yield. In the breeding program hereditary pause should
be eliminated by the constant use of tested breeding males and females, and
environmental pause should be controlled by time of hatching and methods
of management.
/->. Average Size of Winter Clutch (ind Annmd Production.
The mean size of winter clutch has already been shown to be specific and a
workable measure of intensity. In order to discover the association between
mean clutch size and annual production the method of partial correlation is
applied. In this instance age at first egg, length of pause, days broody and
annual persistency are made constant. The net correlation between winter
clutch size and annual production is -f-.t94 lit .0101. This constant is of sulR-
cient magnitude to demonstrate that intensity as measured by winter clutch size
is a significant characteristic in relation to annual production and of aliout the
same importance as duration of winter pause.
202 TECHNICAL BULLETIN 12
14- Total Days Broody and .liinual J-'rodiicfiuii.
Degree of broodiness may be measured by tlie total broody days during the
pullet laying year. The degree of broodiness is affected by inherited factors
so that it may be reduced. The presence of broodiness in any degree has been
shown to be inimical to annual egg production (Hays and Sanborn, 1926a)
and tlie broody trait itself has been shown to be inherited (Hays, 1924).
To discover the true relation of degree of broodiness to annual production,
the method of partial correlation is applied with age at first egg, winter pause,
winter clutch size and persistency made constant. The net correlation between
total days broody and annual egg production is — .5630±.0097. This is an
important relation and makes clear that degree of broodiness is a vital factor
in annual egg yield. Degree of broodiness may be placed on a par with winter
pause and intensity as a characteristic affecting annual production (see sections
12 and 13).
The simple correlation between degree of broodiness and annual production
is — .1964±.0194 and the simple correlation between broodiness and higher than
mean annual production is — .2640±.0132 (Hays and Sanborn, 1926a). This
latter constant does not show the presence of broodiness to be as inimical to
annual egg yield as is the presence of winter pause. From these observations
it appears tliat a very important step has been the reduction in degree of
broodiness and that the increased annual egg record from non-broody birds
is only significantly greater than that from birds broody but once in the pullet
year.
15. Annual Persistency and Annual Eyy Production.
Annual persistency in laying (long laying period) at the end of the pullet
year has been shown to be very important in relation to annual egg record.
The true relation of persistency to egg record may best be arrived at by using
the method of partial correlation where age at first egg, length of winter pause,
winter clutch size and total days broody are made constant. The net correla-
tion between persistency and annual egg yield is -|-.7501±.0063. This is a
very intimate correlation and places annual persistency as the greatest single
characteristic affecting annual production.
A breeding program should therefore lay special stress on the high persis-
tency characteristic wliich is intimately linked with early sexual maturity
(Hays, 1927).
16. Midtifle Correlation Betzveen Five Fecundity Traits and Annual Eyy
Production. '
Theoretically, if all influences affecting annual egg yield were brought to-
gether and correlated with egg yield, the correlation should be perfect. In
this report five of the most important inherited traits are considered. No
account is taken of various environmental influences that operate to affect
fecundity, because such influences are not breeding problems but rather prob-
lems of management. If by this method a high total correlation is discovered,
it will be an indication that the breeding program has been directed along
constructive lines.
CIIAHAC'J'F.HS r()\('KI{\Kl) 1\ I'lHlNDIlV 'JnrJ
Uy means i>f imiltiplo i-orrt-latioii tlu- cliarac-torislii-s ajic at lirst rii^x, K-njrll)
lit" winter pause, winter eluteli si/e, total (lays i)r(><)(ly and annual perslstem y
have eolleetively been correlated with annual esrg production. The constant
obtained is R = -(-.861-'2. This constant sliows tliat the live traits considered
show ajiifrh decree of correlation with annual jiroduction and that they are
\ cry larjrely responsible for annua! produclion.
Summary
Relation to Each Otlier of Characters Concerned in l^'ecundity
Simple Net
Characters Compared Correlation Correlation
Aire at lirst eirg Length of winter ]>ause — .2329±.0138 — .2236±.0139
Age at (irst egg Winter clutch size — .2273±:.0103 — .1879±.0105
Age at first egg Total days broody +.0062±.0194 +.0473±.0194
Age at first egg Annual persistency — .6U6±.0090 — .5956±.0093
Length of winter pause Winter clutch size — .0671.±.0145 — .0874.±.0U5
Length of winter pause Total days broody — .1832±.024.3 — .1609±.0245
Length of winter pause Annual persistency +.1017±.0182 — .0393±.0183
Winter clutch size Total days broody +.2205±.0186 +.2079±.O187
Winter clutch size Annual persistency +.1692±.0130 4-.0190±.0134
Total days broody Annual persistency 4-.0532±.O2O9 +.0579±.0209
Relation to Annual Production of Characters Concerned in Fecundity
Age at (irst egg Annual production — .4380iii.0134 — .0238±.0177
Length of winter pause Annual production — .2107±.0176 — .5487±.0128
Winter clutch size Annual production +.3544±.0117 +.4944±.0101
Fotal days broody Annual production — .1964±.0194 — .5630±.0097
Annual persistency Annual production +.7082±.0072 -f. 7501 ±.0063
The partial or net correlations tabulated above indicate the following rela-
tionships:
1. Age at first egg and winter })ause arc sigiiiCicantiy but not intimately
negatively correlated.
2. Age at first egg and mean winter clutcli size are negatively correlated to
a significant but not intimate degree,
3. Age at first egg and total days broody are independent.
I. Age at first egg and anniud persistency are intimately correlated and
early sexual maturity is linked with high persistency.
5. Length of winter pause and winter clutch size are independent.
6. Length of winter pause and total days broody are negatively correlated
in a minor degree.
7. Length of winter pause and annual ))ersistency are independent.
8. Mean winter clutch size and total days broody arc positively correlated
to a moderate degree.
!). Winter clutch size and annual persistency are independent.
10. Total days broody and annual persistency are independent.
204 TECHNICAL BULLETIN 12
11. Age at first egg and annual production are independent.
12. Length of winter pause and annual production are negatively correlated
in an important degree.
13. "Winter clutch size and annual production are positively correlated in
about the same degree as days broody and annual production are negative-
ly correlated.
14. Total days broody and annual production are negatively correlated and
days broody is of about the same importance as length of winter pause
and clutch size in relation to production.
13. Annual persistency and annual production are verj- intimately correlated
with each other. The degree of correlation places persistency as the most
important cliaracteristic affecting fecundity in the flocks studied.
16. The multiple correlation of age at first egg, length of pause, winter clutcli
size, days broody, and persistency witli annual egg record is -\-.86-i2.
Literature Cited
Hay.s, F. A. and J. S. Bennett. 1923. Correlation of sexual maturity to annual
egg record. Poultry Science II, No. 6:205 — 206.
Hays, F. A. 1924. Inbreeding the Rhode Island Red fowl with special re-
ference to winter egg production. Amer. Nat. 58: 43 — 59.
Hays, F. A. and Ruby Sanborn. 1926a. Broodiness in relation to fecundity
in the domestic fowl. Mass. Agr. Expt. Sta. Tech. Bui. 7.
Hays, F. A. and Ruby Sanborn. 19261). AVinter cycle and winter pause in
relation to winter and annual egg production. Mass. Agr. Expt. Sta.
Tech. Bui. 8.
Hays, F. A. and Ruby Sanborn. 1926c. Annual persistency in relation to
winter and annual egg production. Mass. Agr. Expt. Sta. Tech. Bui. 9.
Hays, F. A. and Ruby Sanborn. 1927. Intensity or rate of laying in relation
to fecundity. Mass. Agr. Expt. Sta. Tech. Bui. 11.
Hays, F. A. 1927. Inheritance of persistency and its relation to fecundity.
Proc. World's Poultry Congress.
3500 l-'28. No. lOi;
Massachusetts
Agricultural Experimental Station
Technical Bulletin No. 13 May, 1928
Washing Powders For Dairy
Use
By A. W. Phillips, M. J. Mack and J. H. Frandsen
The importance of cleanliness in the production and handling of dairy
products is very generally recognized, and as a consequence there is much
interest in the use and manufacture of suitable cleaning compounds. It
is recognized that under our present system of advertising the use of a
product may not necessarily be proportional to its merit. It was in an
attempt to determine what constitutes merit in a cleaning powder that
this study of composition and properties was undertaken.
Recniests for bulletins should be addressed to tlie
AGRICULTURAL EXPERIMENT STATION
AMHERST, MASS.
WASHING POWDERS FOR DAIRY USE
By A. W. Phillips, Assistant Research Professor of Dairying, M. J. Mack,
Assistant Professor of Dairying, and J. H. Frandsen, Professor of Animal
and Dairy Husbandry'
Introduction
Tlie most iiii]vortaiit factor I'onnected with any food supply is cleanliness.
This is particularly the case in the handling of dairy products because, owing
to their greasiness and solid particle content, these products adhere very tena-
ciously to all utensils with which they come in contact, and, being very good
food upon which many micro-organisms can live, offer excellent breeding
grounds for bacteria. Therefore, it is extremely important that all milk
utensils should receive a thorough cleaning after each use.
This paper reports a study of the cleansing of dairy utensils and how it
may he accomplished satisfactorily. There are two phases of the problem due
to the complexity of the milk products themselves: first, the cleaning of the
utensils from adhering particles of the dairy product; and second, the steril-
ization of the cleaned surface. The cleaning operation is by far the more
important step, for without thorough cleaning sterilization would be extremely
difficult if not impossible. Also, sterilization is accomplished to a certain de-
gree during the cleaning process.
Chemical Changes in Cleaning
Milk and the products manufactured from milk are of a very complex
physico-chemical nature. In the handling of dairy products we are dealing
with a very delicately balanced system of emulsions, colloids, and solutions
which, even when unaltered by any treatment, possess great adhesive
properties for surfaces with which they come in contact. When the products
are altered by intrinsic agencies such as souring or by outside agencies such
as heat, then the original systems of enuilsions and colloids are changed and
the solid and liquid materials thus thrown out are found to adhere even more
tenaciously, and the difficulties of the cleaning operation are therefore greatly
increased. The problem in cleaning is, then, to re-einulsify these deposits or
bring them back into the colloidal state.
The fat or oily ingredients in dairy products can be separated from the
surfaces to which they adhere by the action of some emulsifying agent pro-
vided they are not bound there by some other ingredient. This enuilsifying
action is not, as is popularly l)clieved, due to neutralization of fatty acids
and saponification. If the fatty substance is not rancid there is scarcely any
free acid present and, moreover, saponification does not take place under con-
ditions existing in the washing process.
An alkali is capable of assisting in the formation of a surface layer by
reacting with the free fatty acid in the grease to be removed. The "surface
activity" of the detergent is increased by adding alkali (1). The alkali lowers
the interfacial tension between oil and water but does not saponify the oil to
any extent. In order to maintain a constant alkali strength buffer salts are
needed in the detergent.
1 Acknowledgement is made to Professor H. F. .ludkiiis. former head of the Dairy
Department, who originally proposed this problem.
(1) Shorter: Proc. Roy. Soc. London A92:'_';!l (1916).
180 TECHNICAL BULLETIN 13
The cleaning of dairy utensils is not merely "cutting the grease" withi
alkali. Milk proteins aresoluble in alkali. But once coagulated or dried, so-
lution is very slow even with a strong concentration of alkali. Therefore,
something besides alkali nuist be used in cleaning. The cleaning solution must
possess deflocculating or emulsifying power so that coalescence of the oily
substances into droplets is impossible. This is prevented by the formation of
a surface layer of the deflocculant around tiie fat globules. Surface activity is
therefore a vital function in the cleaning solution. The detergent must go into
true solution, not colloidal or crystalloidal, before it can become oriented and
adsorbed upon the surface of the particle. Deflocculation occurs when a cer-
tain minimum concentration of "absorbed oriented molecules" of the detergent
upon the dirt particle surface has taken place (2).
Methods of Comparing Washing Powders
The lirst step in this investigation was logically the cliemical analysis of
the various washing powders found on the market. However, the evaluation j
of complex and variable materials such as most cleansers are cannot be based
upon chemical analysis alone. A properly designed test for performance
often affords data of greater practical usefulness.
No test yet proposed for determining the cleansing efficiency of detergents
has received general acceptance, and chemists still depend almost entirelj-
upon data of composition for evaluating the respective merits of competitive
samples. Such dependence is justified, once the effect of each ingredient has
been studied in relation to its use alone and with other ingredients in vary-
ing proportions.
In making laboratory tests, no objections can be raised to tests which close-
ly parallel service conditions provided they are not impractical. Such tests
may actually prove capable of distinguishing between various samples with
sufficient precision for ordinary comparisons. The tests devised for this work,
which are described in detail below, may be considered practical in every
respect.
Tests 'based upon the determination of surface tension or interfacial tension
appear not entirely satisfactory (3). Such tests show wetting power but this
is not necessarily vitally connected with detergent action. The cleaning com-
pound must wet the dirt surface, but it must do more than this. It must
have the j^owers of emulsification and deflocculation.
The test of greatest importance is of course the washing power of the
powder. Yet this was the hardest test to conduct. Other investigators (4)
have found that when using a specific detergent results varied on the average
by ten per cent when all conditions insofar as possible were kept constant.
With milk in particular the cleaning action is quite varied, depending upon
the depth, hardness, et cetera, of the dried or gummed deposit.
Chemical Analyses of Powders
In order to ascertain the tj'pes and differences in washing powders on the
market, thirty-six brands of powders recommended for dairy use and on sale
in Massachusetts were collected. These powders were subjected to chemical
analysis to determine the kind and amount of ingredients present. Then the
(2) Chapin: Ind. and Eng. Chem. 17:1187 (1925).
(3) Chapin: Ind. and Eng. Chem. 17:461 (1925).
(4) Luksch: Seifenseeder. Ztg. 40:413.
WASHING POWDERS FOR DAIRY USE 181
Ii(>\\ders were subjected to laboratory- and i)}ant tests for efficiency of per-
l<irniance in the various cleaning operations.
riie results of the chemical analyses are listed below. If the figures given
tif not total 100 per cent tlie reader must bear in mind that the water content
ni.iy be verj'" high in some cases (15 per cent or over), and many ingredients
net determined but listed later may have been present. Chapin in some of
his work, (3) leaves undeteniiined quantities as high as 26 per cent.
Analyses of Washing Powders.
unple
Total Al-
Sodium car-
Sodium
Tri-sodium
Soaps as
Remarks
kali as
bonate
hydrox-
phosphate
fatty
^^aOH
ide
12H.,0
acids
<■;,
%
'/c
%
%
1
57.8
61.5
0.0
37.8
0.0
Chlorine
2
31.0
0.0
2.0
98.0
0.0
Trace insol.
3
57.2
61.5
0.0
38.1
Little
4
55.7
63.7
7.7
0.0
Little
5
47.9
45.5
8.8
14.8
27.7
6
15.5
4.5
12.1
0.0
Little
Much grit
7
89.:;
60.5
40.8
0.0
0.0
Trace insol
8
60.7
8 5., 5
0.0
0.0
0.0
Trace insol
9
71.8
52.3
32.4
0.0
8.0
10
39.6
38.6
10.5
0.0
19.8
Sulfates
11
34.0
3.6
3.0
95.0
0.0
12
60.8
88.6
0.0
0.0
0.0
1:3
53.5
63.7
5.5
0.0
17.9
14
88.3
55.5
46.4
0.0
0.0
15
58.5
66.0
0.0
32.7
0.0
16
58.5
55.5
0.0
45.0
0.0
17
59.7
61.5
7.3
18.7
Little
18
61.3
68.2
0.6
29.1
0.0
19
38.6
18.2
0.0
80.0
0.0
20
54.7
66.8
4.2
0.0
12.6
21
64.5
76.0
7.2
0.0
0.0
22
33.6
1.0
5.0
94.0
0.0
23
25.0
0.0
0.0
89.0
0.0
Trace insol.
24
34.6
7.7
0.0
93.0
0.0
25
63.8
66.0
14.0
0.0
Little
26
53.2
56.0
10.9
0.0
24.1
27
33.6
0.9
5.1
94.2
0.0
28
27.2
0.0
0.0
100.3
0.0
Trace insol.
29
62.0
94.6
0.0
0.0
0.0
30
93.0
46.8
54.6
0.0
0.0
31
C0.7
95.5
0.0
0.0
0.0
32
75.0
56.8
32.2
0.0
0.0
33
76.8
34.0
51.2
0.0
0.0
34
58.6
65.0
9.6
0.0
13.6
35
64.0
72.8
6.8
7.0
Little
Grit
36
62.7
95.5
0.0
0.0
0.0
A
100.0
0.0
100.0
0.0
0.0
Pure sodium
hydroxide
B
S1.2
100.0
0.0
0.0
0.0
Pure sodium
carbonate
C
26.8
0.0
0.0
100.2
0.0
Pure tri-
sodium
phosphate
D
14.3
0.0
14.3
0.0
85.1
Pure castile
soap
It may be observed from this table that many of the powders were of very
nearly identical composition, yet great variations are also shown. The
Ipovvders may be classified into four main groups: those containing pure car-
bonate, those containing tri-sodium phosphate, those containing free caustic
182 TECHNICAL BULLETIN 13
and those containing soap. Ihe roles played by the various ingredients will
be discussed later after the practical tests have been described.
The water content of the powders varied considerably. No determinations
of moisture content are given because this factor would tend to vary.
Pcvwders high in sodium hydroxide would tend to absorb water more rapidly
than others. All the powders when weighed were of a dry, somewhat dusty
nature, and not in the least gummy or sticky.
The ease of solution of the powders followed a general trend depending
upon composition. Those containing tri-sodium phosphate were aH very slow
to dissolve. Also, the soapy powders were somewhat slower to enter solution
and often lumped. The carbonate and hydroxide powders dissolved quite
readily although there was a tendency for the hydroxide powders to lump.
Those high in sodium hydroxide tended to heat considerably upon dissolving
and those high in tri-sodium phosphate tended to cool.
Practical Tests
In this part of the investigation it was aimed to duplicate actual working
conditions in the laboratory and then to conduct practical tests in the plant.
The tests used and the method of procedure for each are given below.
The strength of solutions employed in these tests was standard 0.6 per cent
based on the dry powder. Ihis was found to be the average concentration
recommended by the various manufacturers. A few reconuuend a weaker
or stronger solution, but in order to study the efficiency of the powders a
standard had to be adopted. Ihis concentration corresponds to five pounds
of powder to one hundred gallons of water.
The water softening power of the powders was determined by treating
lOOcc portions of the samples with 50cc quantities of standard hard water
and then adding standard soap solution until permanent bubbles appeared.
The washing power of the powders was deteniiined by tests on unifoniily
dirty bottles. Five cubic centimeter c]uantities of milk were run into bottles
and allowed to dry. While drying, the bottles were occasionally tilted and
rotated so as to wet the sides. The action of lOOcc amounts of the cleaning
solutions upon these dirty bottles was then observed.
The emulsifying power was tested by shaking a lOOcc portion of the powder
solution with l.Occ of butterfat, the whole contained in a tall cylinder. The
degree of emulsification was determined bj' the whiteness of the emulsion and
by the length of time the enmlsion persisted.
The ease of rinsing of each solution was tested by moistening the fingers
with the solution and then counting the numiber of half seconds the slippery
feel lasted while the fingers were held under a flow of water and gently
rubbed twice per second. The flow was from a faucet and was regulated to
give a stream one-fourth inch in diameter. This test was found to be satis-
factory and gave consistent and reproduceahle results. The average time of
three experiments was taken in each case.
The action of each washing powder solution upon the metals aluminum,
copper, nickel, tin and zinc was also studied. Strips of these metals were
cleaned and allowed to stand in solutions of the various powders. These
tests were run for a considerable length of time in order to determine the
effect of continued use of the powders.
All results are tabulated below. In case tlie figures given seem to show
some discrepancy the reader is reminded that the tests are not absolute, and
the previous workers on similar tests for washing efficiency obtained values
varying as much as 10 per cent.
\V ASHING I'OWDKllS FOR DAIRY USE IS.i
Results of Practical Tests on Washing Powders.
Water
ample
Softening
Washing
Kmulsification
Rinsing
Power*
Test
Test
Testf
1 '
200
Good
Good
9
2 ^
216
Good
Poor
12
3
28
Good
Pair
7
4
237
Good
Very good
15
5
194
Pair
Very good
17
6
28
Poor
Very good
2
7
181
Fair
Good
15
8
157
Poor
Good
5
9
231
Good
Very good
27
10
241
Good
Very good
25
11
75
Fair
Poor
15
12
237
Poor
Poor
1
13
231
Good
Good
7
14
178
Fair
Good
48
15
72
Poor
Poor
8
16
71
Poor
Poor
9
17
84
Fair
Very good
23
18
124
Good
Good
18
19
183
Good
Good
25
20
202
Poor
Very good
11
21
150
Poor
Good
7
22
27
Good
Poor
12
23
31
Fair
Good
13
24
52
Poor
Poor
12
25
250
Poor
Very good
10
26
276
Good
Good
10
27
143
Poor
Poor
14
28
112
Pair
Good
14
29
237
Pair
Very good
3
30
135
Pair
Fair
17
31
178
Pair
Fair
4
32
162
Pair
Good
40
33
94
Fair
Fair
15
34
243
Poor
Good
10
35
235
Fair
Good
4
36
234
Good
Poor
1
A
31
Very good
Pair
28
B
175
Very good
Good
9
C
153
Poor
Good
10
D
250 plus
Pair
Very good
8
* The figures in this column give the cubic centimeters of standard hard water (con-
taining the equivalent of 0.2 grams of calcium carbonate per liter) softened by 100 cc
of the standard solutions of the powders.
t The figures in this column give the number of half seconds the slippery feel lasted
when tested as outlined above. These figures are all comparable, the low ones indi-
cating rapid rinsing.
Further tests were conducted, which concerned the action of the typical
powders numbered 1, 2, 3, 6, 7, 8, 26, 23, 28, 31, A, B, and C, upon the metals
aluminum, copper, nickel, tin and zinc. The action of the powders upon th-^
metals was carefully watched during the first day, and then obsened daily
thereafter. Strong alkali powders reacted immediately upon aluminum.
Other combinations reacted less readily and in reverse proportion to their
caustic strengths. Powders iiigh in soap content tended to l)]acken and tarnish
the metals. The tests were all allowed to stand for fifteen days. A summary
of the results is listed below:
184 TECHNICAL BULLETIN 13
The Action of Washing Powders on Metals.
Sample
Aluminum
Copper
Nickel
Tin
Zinc
1
P
G
E
F
E
2
P
E
E
B
E
3
B
F
E
F
E
6
F
F
E
G
F
7
P
B
E
B
F
8
P
B
E
E
E
20
P
F
E
F
F
23
B
E
E
P
E
28
P
E
E
P
£
31
P
B
E
G
E
A
B
B
E
P
E
B
B
P
E
F
E
C
F
E
E
F
E
Key
E — Excellent condition
G — Good condition
F — Fair condition
P — Poor condition
B — Bad condition
Tests of the germicidal action of solutions of the above samples showed
that in solutions of 0.6 per cent strength, dirty, milky water was sterilized
after standing for thirty minutes in all tests at room temperature.
Plant tests were conducted upon representative powders by using them in
the daily bottle and can washing operations. The results in these cases sub-
stantiated those obtained from tlie laboratory tests.
Summarizing, we may place the ingredients in the order of effectiveness as
follows:
Water Softening Powers: Carbonate, phosphate, soap, hydroxide
Washing Powers: Carbonate, hydroxide, soap, phosphate
Emulsifying Powers: Soap, phosphate, carbonate, hydroxide
Ease of Rinsing: Carbonate, phosphate, soap, hydroxide
Action on Metals: Hydroxide attacks aluminum, copper and tin
Carbonate attacks aluminum, copper and tin
Phosphate attacks aluminum
Disinfecting Values: All powders in strength ordinarily employed
act as disinfectants to such a degree as to
make th,e washing solution sterile.
The Role of Constituents Used in Cleaning
Reviewing and correlating the data observed in the above experiments, it
is possible to draw somewhat trustworthy conclusions as to the roles played
by each jiarticular ingredient in the powder and tlieir action when combina-
tions are used.
Sodmm Carbonate has very good softening power, greatly aids washing mech-
anism, is a poor enuilsifying agent, rinses very easily, has slow action on
tin and very mild action on hands, and neutralizes odors.
Sodium Hydroxide is a very poor water softener, with good washing power
and poor emulsifying effect. It is very difficult to rinse, acts on metals,
has a severe action on hands, is a strong caustic, and will pit cans.
Tri-Sodium Phosphate is an excellent water softener, with poor washing power
and excellent emulsifying properties. It is easy to rinse, has very mild
action on metals and hands, is a good solvent for casein, and gives buffer
effect which keeps caustic strength uniform.
Soap has poor softening power, poor washing power and excellent emulsifying
properties. It is hard to rinse, has mild action on metals and hands, and
is apt to leave scum in bottles.
I
W ASlUN'c; I'OW HDERS FOR DAIUV USE 185
Olher hujredieiitu are i)ften added for specilic i)uri)()se.s. JIi/ptx-hlDriles arc
added in some instances because of their germicidal action. Sulfates are
occasionally added, l)iit have no desiral)le qualities. Siliciite.i of scjme
niet^^ls are added to render protection to the machinery parts in contact
with the solution. SodiuDi ziiirate and nlumiuate are sometimes added
for' protection of zinc and aluminum jiarts respectively. Rentiinles and
r/HHf.s- are also occasionally found in the powders and are usually added to
]>rotcct cojtper fittings. Grit is added to some to clean metafs. liornx,
although not often found, has certain advantages of emulsification and
disinfecting powers.
Thus it may be seen that no one ingredient is a perfect wasliiiig <'om]»oiiiKl
of itself, and for efficient washing- a mixture should be used.
General Considerations
Soap, owing to the difficulty of rinsing which is due in large measure to
suds formation, and owing to the dangers from odor, is not considered a wise,
nor is it an essential ingredient for washing powders for dairy use. Also it
is more expensive and will deteriorate upon long standing. Powders contain-
ing soap are light in weight and thus often deceive the purchaser as to the
quantity he is obtaining.
For hand washing any free caustic should be avoided because such solu-
lions readily attack the skin of the operator. It also attacks metals and in
general its use is not justified, for other ingredients of less drastic action are
e\en more efficient in the actual cleaning operation. Caustic lacks many of
the essential qualities of a good cleaning agent. It is very difficult to rinse off
and thus will cause pitting and darkening of .the cans.
For machine washing strong caustic may be used and is efficient in removing
grease. Although it acts upon metals and glass the action is not drastic and
is noticed only after continued use. The mechanical washer is perhaps the
most efficient because of the possibility of using stronger solutions and hotter
water. A 3 per cent caustic at 140° F. can be used with the mechanical
washer.
Tri-sodium phosphate is a particularly valuable constituent because of its
remarkable emulsifying and softening power. In softening water containing
calcium sulfate, calcium, magnesium and iron silicates and calcium carbonate
the action is to throw all these metals out of solution as phosphates while the
acid radicals will become united with sodium and remain in solution where
they will cause no harm. The phosphate also has a strong buffer effect which
keeps the caustic strength of the cleaning solution uniform by supplying more
caustic as soon as some is neutralized. The caustic strength of the phosphate
is only about one-tenth that of an equal amount of hydroxide and therefore
no great caustic concentration is necessary at the start of operations in order
to have the solution strength maintained during the washing when the phos-
I^hate is used. In this respect, using phosphate, added in quantities at the
start, gives the same effect as adding a small amount of hydroxide every
little while.
Casein is soluble in alkali in ])roportion to the caustic strength. However,
nmch of the precipitated or baked-on casein is in the form of calcium casein-
ate which is insoluble in alkali. Here tri-sodium phosphate acts to a much
greater advantage than the other components. Calcium in the water supply
is more efficiently removed by sodium phosphate. Casein is soluble in borax
although this substance is scarcely ever used in dairy w^ashing powders. It
would aid the disinfecting properties but it has only a mild cleansing action.
P>()rax is easily rinsed and fornss a good emulsion with milk fats.
186 TECHNICAL BULLETIN 13
The action of the various powders upon glass is ratlier severe over long
periods of time. With bottles, the length of life is so short that the corrosive
effect does not need to be considered. However, with tiie new glass-lined
vats the problem might become important. In the order of their severity of
action lipon glass the several ingredients may be listed: hydroxide, phosphate,
carbonate and soap.
Alkalies of any kind attack paint and so no strong cleaning solution should
be allowed to remain on painted surfaces for any length of time.
Cleaners containing abrasives should be used sparingly, and only when
absolutely necessary. Any abrasive will wear away the surface over which
it is rubbed. If the surface is plated, the under metal will soon be exposed
and form an electrolytic cell with the metal used in plating. This causes
increased corrosive action. Also the grit present is very hard to rinse off.
On the average, five to six pounds of powder to a hundred gallons of water
may be regarded as the proper strength solution for ordinary hand washing.
This of course will vary with the particular powder used and the hardness of
the water. Smaller amounts of any powder are not to be recommended even
though the sales agent may speak very highly of his product and what it will
do. The majority of the powders on the market may be classified under three
heads having nearly the same composition, and a high price does not neces-
sarily infer any greater efficiency. Prices quoted different consumers vary
greatly even for the same product from the same manufacturer and, of
course, vary with the size of the order. The current prices for the basic in-
gredients are as follows:
Sodium cart)onate $0.90-$1.30 cwt. depending upon purity.
The unrefined material is satisfactory. There are several grades
depending upon the water content: as soda ash, NajCOs; crystal
carbonate, NajCOs.HjO; and soda crystals or washing soda,
NaoCOs-lOHjO. Soda crystals contain 63 per cent water. One ton
of crystal carbonate contains as much alkali as 48 cwts. of soda
crystals and occupies only half as much space. The soda ash is
highly concentrated and is liable to cake. It is higher priced and
has no advantages.
Sodium bicarbonate $2.00 cwt.
Sodium hydroxide 3.00 cwt.
Tri-sodium phosphate 4.00 cwt.
The 60 per cent carbonate — 40 per cent phosphate (12 HjO) reconuuended
in the following pages would cost $2.20 per cwt., or a little over 2 cents per
pound for raw products. The cost of mixing would be negligible.
Soap prices are omitted because they would vary greatly depending upon
the purity of the product, and because soap is not considered advisable for
use.
Thus it is evident that users of large amounts of wasiiing powder would
find it profitable to buy their own stock and mix their own powders. Smaller
dealers might not be sufficiently reimbursed for the trouble. Great care is
necessary to protect the eyes from alkali dusts. Also the dust is irritating to
the throat and nose. Therefore the mixing should be done in an enclosed
room, and the man doing the mixing should be protected by a sponge through
whicii to ])reathe, and his eyes protected by goggles.
The temperature of the washing solution should not much exceed 140° F.
or 60° C. because, if allowed to do so, the precii)itated casein and gununy
WASHING I'OWHKRS FOR DAlin USE 187
deposits will bake on harder. For hand wasliiiifj tliis tcinperaturo is too
severe and a lower temperature must be resorted to. However, up to al)out
1W° V. tlie liotter the better as re{!;ard.s the eleaning aetion. The baeterieidai
action is markedly reduced l)elow .35° C. or 95° F.
From a survey of tlie powders used by a number of dairies in Massachu-
setts the outstandiufj; feature is that the majority of chanfres from powders
previously used has been for the adoption of powders containing tri-sodium
phosphate. The tri-sodiuni phosphate powders are held in high favor by
tliose using them and these powders are proving very desirable. There is a
tendency to avoid the use of strong caustic powders except for very sj)ecial
uses.
A Suggested Composition for Washing Powders
It may be safe to say that the best powder for dairy cleansing shouin
have no soap and no free caustic for hand washing, and little free caustic for
machine washing. A satisfactory powder should analyze apj)roximately as
follows:
Sodium carbonate (anhydrous) 60%
Tri-sodium phospliate (I2H2O) 40%
Total alkali as NaOH about 58%
Sodium hydroxide None (except
what may occur free in the carbonate)
Soap None
This combination has been found to give good results in all laboratory
tests as well as in all plant tests. It was found to be efficient in its cleans-
ing action and to jjossess all the desirable qualities of a good jiowder. Tliis
percentage coni])osition could be obtained from various mixtures of the com-
mercial chemical products which should be analyzed and mixed accordingly.
Conuiiercial tri-sodium j)hosphate ordinarily contains 12 molecules of water.
For machine washing it may be deemed advisable to increase the total
alkali content by the addition of a little sodium hydroxide.
Summary and Conclusions
1. Analyses of many washing powders on the market show four general
classes, containing carbonate, caustic, phosphate and soap respectively.
2. There is very slight variation in those powders which come in the same
class.
0. Laboratory and plant tests on these powders, on other mixtures, and
on the pure ingredients have demonstrated the specific roles played by each
ingredient.
4. A desirable composition for general dairy use has been indicated to be
60 per cent sodium carbonate and 40 per cent tri-sodium phosphate.
5. By buying the commercial chemicals the price per jiound of cleaner
may be reduced to close to 2 cents as compared with from 8 to 16 cents now
paid for a similar grade of product.
6. The wa.shing efficiency of the powders increased up to about 140° F.
Below 95° F. the bactericidal action is greatly reduced.
7. All powders showed disinfecting powers in 0.6 per cent solution, l)y
rendering the wash water sterile.
3 M-5-'28. No. 2366
Massachusetts
Agricultural Experiment Station
Technical Bulletin No. 14 June, 1928
Cutaneous Immunity in Relation
to Contagious Epithelioma
By Norman J. Pyle
The investigation of contagious epithelioma at this Station has for its
purpose the determination of an efficient preventive and curative treat-
ment for the disease. This is of vital importance to the Massachusetts"
poultry industry because the disturbance causes serious loss by decreasing
egg production during the season when eggs are highest priced. A sero-
logical study of immune birds was made for the purpose of standardizing
the vaccine already extensively used. As this work progressed it became
evident that immunity against the disease was not of a general nature,
and therefore a study of cutaneous immunity was undertaken in its stead.
Results of this latter investigation are reported in this bulletin.
Requests for Bulletins should be addressed to tlie
AGRICULTURAL EXPERLMENT STATION
AMHERST, MASS.
CUTANEOUS IMMUNITY IN RELATION TO CONTAGIOUS
EPITHELIOMA.
By Norman J. Pyle, Assistant Research Professor of Avian Pathology.
Contagious epithelioma, a disease of poultry coninionly known as chicken
pox, has caused serious loss to Massachusetts poultrymen by decreasing egg
production during the season when eggs are bringing the highest prices. The
treatment for the disease has not been satisfactory. For this purpose i
powdered pox virus vaccine, as recommended by Beach (1), was used quite
extensively up vmtil the winter of 1926-1927. In November, 1926, the writer
(2) published a report on the experimental use of this and other vaccines,
wherein it was shown that the vaccine did produce a slight degree of immun-
ity, but not enough to protect the bird fully against the disease under all
conditions. It was thought at the time that if a method of standardizing the
vaccine could be determined, its tiierapeutic efficiency would be considerably
eniianced. Accordingly, furtlier investigations were undertaken with tliat
objective in mind.
The logical method of attacking sucii a problem was to study immune birds
serologicallj' and attempt to produce a unit value of antiserum against which
the virus or vaccine might be standardized. It was naturally assumed that
immunity to contagious epithelioma was of a general nature; that is, due to
a sj)ecific antibody concentration in the blood serum. As the work progressed
antibodies were found, but they were present in the sera of immune birds in
such a low concentration that it was evident that a protective force other
than a general immunity was operating against the disease. Because of tills
conclusion and due to the fact that varying complement-fixation reactions
prevented the determination of a unit value of antiserum, the entire aspect
of the problem was changed.
The data herein reported include:
1. Serological studies on normal and inunune birds which demonstrate
that a specific antibody concentration in the blood serum of immune fowls is
not the sole protective force against contagious ejjithelioma.
2. Imnumity experiments which demonstrate that a cutaneous immunity is
the chief protective force against tiie disease.
3. Experiments with local or cutaneous vaccines and tlielr standardization.
SEROLOGICAL STUDIES
A review of the literature on the serological study of contagious epithelioma
of the domestic fowl reveals very few references on the subject. Beach (1)
mentions that the vaccine cannot be standardized by complement-fixation
methods, but he does not describe the technic used in the investigation. Sweet
(3) has published an interesting complement-fixation study of the serum of
fowls affected with the disease. He found that hemolysis was com})iete in
serum from normal birds, while in serum from birds in which the disease was
operating he obtained fixation in varying degrees, or to use his own words,
"There was considerable evidence of a specific antibody".
CONTAClIorS Kl'lTHKI.IOMA 203
Source of Experimental Birds
M'liitc I.cjiliorii c-ockerels aiul Hliode Island Red jiullets and cockerels were
used in the following experiments and were obtained from the ])oiiItry iilant
of the Massachusetts Agricultural College. They were progeny of liirds
which were free from Salmonella pullorum infection, as determined i)y the
agglutination test. The birds were reared on a clean grass range which had
not been used for this purpose for three years. At no tune were they siil)-
jei'ted to any infectious disease.
Technic of Bleeding Fowls.
The pin feathers are plucked from the inner surface of the wing over the
triceps region and the area is cleansed with a suitable antiseptic, i)referal)ly
alcohol. The brachial vein is selected for l)leeding. It is found sulicutane
ously in a nuiscular groove foriued by the bodies of the two major tricep:;
nuiscles. A Luer syringe of 10 cc. cai)acity with a 20 gauge needle, li
inches long, is used for aspiration. Aseptic precautions are used during the
bleeding process. From 20 to 25 cc. of blood can be taken from a mature
liird in good physical condition without any apparent ill results. A])])roxi-
mately 1.50 birds were bled during the course of this investigation, and an
average of 20 cc. of blood was taken from each individual without a single
loss. Frequently desirable birds were bled a second and third time at weekly
intervals. When more than 10 cc. of blood are required, as was the case in
this investigation, a second syringe and the brachial vein of the other wing
are used. After bleeding, the needle is removed from the syringe and the
blood gently forced into a sterile test tube (6/8 by 6 inches). The tube is
.slanted until the l)lood clots, and it is then placed in the ice box over night,
during which time the serum separates.
Source of "Immune" Seruni.
The first serum subjected to serological study was obtained from \\'hiti'
Leghorn cockerels, four to five months of age, which had been used for virus
production and had fully recovered from the operation. Later, serum was
ol)tained from yearlings that had recovered from either avian diphtheria or
bird pox, or both, during their pullet and cockerel year. Serum was also
taken from birds while they were in the active stages of contagious epithel-
ioma. There was no doubt w-hatever that the first two groups of birds were
immune to the disease. One might question the immunity of the yearling
group, but this was determined by check inoculation of coml)s with active
virus. Lesions of the disease did not appear. If specific antibodies were
associated with immunity, they should be found in one or more of these tlirec
groups of sera.
' In addition to tlie above, attemjits were made to produce sera containing
S])ecific antibodies by using various vaccine combinations of the active virus.
Healthv young White Leghorn cockerels were used in the experiments. It
was determined that the .siibcutaneous administration of 30 milligrams of
\irus per dose, suspended in a 1-0 per cent glycerol-physiological saline solu-
tion, fully ])rotected the birds, in the majority of cases, when they were check
inoculated on the comb with virus for imuumity determination. However,
scablike lesions, which were found to contain pox virus, invariably appeared
at the point of inoculation, and in .several instances pox nodules also aj)-
204 TECHNICAL BULLETIN 14
peared on the comb and eyelids. Those birds, in particular, which de\eloped
local and generalized reactions after the vaccine injection, demonstrated a
complete immunity after check inoculation on the comb with the virus.
This vaccine was labeled the triple streiigtli vaccine because it contained
three times as much virus per dose as that recommended by Beach (1). tfn-
attenuated vaccines of 10, 15, 20, and 25 milligrams of virus per dose were
also used. They failed to produce the same degree of immunity as did tho
triple strength vaccine.
-Daily rectal temperatures and weiglit readings of the birds in the triple
strength vaccine experiment were taken. It was found that a prevaccinating
average normal temperature of 106.6° F. to 107° F. was increased, beginning
late on tiie day of vaccination and reaching a crisis of 108.6° F. to 109.5° F.
three days later, then decreasing until it iiad regained normal temperature
on the ninth day after the injection. A noticeable effect on the weiglit of the
birds occurred. The cockerels were from five to six months old and gaining
in weight on the average of from 10 to 20 grams per day prior to the vaccina-
tion. After the vaccination this gain in weight was either retarded for eleven
to twelve days or decreased from an average of 1480 grams to 1420 grams
during the same period. Thereafter the birds regained the average of 1480
grams and steadily increased in weiglit
These reactions on the part of the l)od\ to tlie vaccine injection indicated
that the body was developing its protective forces against tlie disease and
from all probabilities these forces were specific antibodies.
Complement-fixation Test.
Anti(ien.
Sweet (3) concluded, in his work with tlie complement-fixation test in con-
tagious epithelioma, that the antigen which he was using lacked marked anti-
genic properties. He did not attempt to impro\e his antigen in this respect.
In this investigation a polyvalent, active virus was always used. An antigen
composed of 1 gram of the virus, thoroughly triturated in 100 cc. of pliysio-
logical saline solution and passed first through infusorial earth and next
through filter paper, was but slightly antigenic. Alcoholic extracts were pre-
pared from pulverized pox scabs and diphtheritic membranes, from pulverized
liver from pox infected birds, and from pulverized normal chicken hearts.
They were found to be without value as antigens.
The antigen finally selected as most suitable and perhaps the only possible
reagent for this purpose, was 1.5 grams of jjowdered pox virus, thorougiily
triturated in 100 cc. of 0.5 per cent phenolized physiological saline solution
and passed through ordinary filter paper. Its hj^drogen ion concentration
was varied in an attempt to improve the antigenic properties. The original
concentration of approximately 5.8 to 6.0 was the most satisfactory. Prior to
its titration for antigenic and anticomplementary properties, it was heated
at 60° C. for forty-five minutes.
A known four-plus positive serum was, of course, not available for use in
titrating the antigen for its antigenic properties. Varying doses of serum
and dilutions of antigen, and finally undiluted antigen, were used and a
degree of fixation obtained. It was necessary to use a two-plus serum in
the titration as a substitute for a four-plus serum. The reason for this will
be apparent when the results of the complement-fixation test are discussed.
In the antigenic titration of the antigen, 0.1 cc. of the undiluted antigen
was found to be the titre or unit dose. Four imits of the imdiluted antigen
were used as a working dose in the test.
CONTAGIOUS El'ITHKLlOMA 205
( "inplement.
XoniiMl guinea piii' senuu was used as coiui)lcmeMt. It was freslilv collected
iMuc a week by bleeding directly from tlu^ iieart of the g-uiiiea j)igs. The
iMuliluted coniplenient was jireservcd with a 12 per cent solution of sodium
.irclatc in the proportion of 1 part sodium acetate solution to 2 parts undi-
hitrd eoiuplement. A 1 to I dilution of the ])reserved C()iui)lenient was used
U<r the preliminary eomjilement titration. One and a half times the coniple-
iiMiit titre was used as the working dose.
//( iiioli/tic amboceptor.
Shee]> liemolytic amboceptor was used. In the routine titration of the
aml)!)ceptor it was determined that the titre was .05 cc. of a 1-600 dilution.
Five units or .25 cc. were used as the working dose.
Sheep cells.
Freshly drawn sheep blood was agitated in a 2 per cent solution of sodium
citrate to prevent clotting. For this purpose from 5 to 10 cc. of the citrate
were used to about 80 cc. of sheep blood. The red blood corpuscles were
washed three times and a 2 per cent suspension of cells used in the test, 0.5 cc.
being added to each tube.
Suspected or "immune'' serum.
Difficulties were encountered in obtaining senuii separation from the avian
Iilood. In the first place the amount of serum obtained from the blood was
small, considerably less than the usual 40 per cent. This is characteristic of
cockerels' blood. It oftentimes "jellied" and in many cases was anticomple-
ir.entary. It appeared that by exercising aseptic precautions in handling the
serum, a great deal of this anticomplementary action was eliminated. The
.■-erum in many instances would also "precipitate", or present a flocculent
appearance during or immediately after inactivation, esjiecially if the tem-
perature of the inactivating bath rose above 56.5° C. The suspected .serum
w^as tested in doses of 0.1. cc, 0.3 cc, 0.2 cc, 0.1 cc, and 0.05 cc.
Res-ults.
With normal fowl serum, aside from the usual number of "Jellied" and
anticomplementary samples, hemolysis was complete. Tabulated residts of
the complement-fixation studies on the "immune" sera are too lengthy to be
rejiorted herein. It is sufficient to say that about one hundred and fifty such
examinations have been made and in no instance was it possible to obtain
complete fixation. Fifty per cent of the sera presented on the average a two-
plus fixation of complement with 0.4 cc. of serum. An occasional three-plus
fixation was noted in the same amount of serum. The remainder demon-
strated a one-plus fixation, a mere inhibition of hemolysis, or Mere entirely
negative (complete hemolysis).
Precipitin Test.
The antigens used in this test contained varying amounts of powflered pox
virus, ranging from 0.5 gram to 1.5 grams, suspended in 50 cc. of 0.5 per cent
phenolized physiological saline solution, and passed through a Seitz filter.
Sera from a group of fifteen birds were subjected to the test. These birds
had recovered from a severe experimental contagious epithelioma infection
and had demonstrated a complete immunity after reinoculation with the virus.
s s
2 t«
LO
lO Ik
■ ■* -i -H
Tf<
■ Tjl ^ r-i
CO
(M lO »0
!N
CO lO o
■ ^ r-i r-i
6
:z;
w
a
D
H
■ T}l rt -H
CC.
units
units
cc.
Serum
Antigen
Complement
Saline
o
CO
Da
f^i^
N
«
7. <
rj "O
^ U3
C^l O
O O O
o -p -g o
c a —
u; < O w
C-l O O (O o
< M 0-
208
TECHNICAL BULLETIN 14.
Seven of the birds were given a dose of the triple strength vaccine in order
to make the immunity more intense, if such a thing were possible. The sera
were tested both undiluted and diluted. Normal fowl serum was used as a
control. All tubes were placed in a water bath at 37..5° C. for one hour and
tiiereafter placed in the incubator over night.
Table 2. Protocol of precipitin test.
Rack No. 1
Tube 1. Anti-pox serum
Tubo 2. Normal serum
Tubo 3. Anti-pox serum
Tubo 4. Saline
0.5 cc. + 1 cc. virus filtrate.
0.5 cc. + 1 cc. virus filtrate (control).
0.5 cc. -|- 1 cc. saline (control).
0.5 cc. -j- 1 cc. virus filtrate (control).
Rack No. 2
Tube 1. 0.2 cc. serum (undiluted)
Tube 2. 0.2 cc. serum (1-5 dilution)
Tube 3. 0.2 cc. serum (1-10 dilution)
lube 4. 0.2 cc. serum (normal)
Tube 5. 0.2 cc. serum undiluted
Tube 6. 0.2 cc. saline
+ 0.2 cc. virus filtrate.
+ 0.2 cc. virus filtrate.
+ 0.2 cc. virus filtrate.
+ 0.2 cc. virus filtrate (control).
+ 0.2 cc. saline (control).
+ 0.2 cc. virus filtrate (control).
Results:
In no case was there any evidence of a })recipitin reactit)n.
Passive Immunity.
Eight Rhode Island Red cockerels were divided into two groups of four
birds each. Each bird received three injections of "immune" serum in the
manner siiown below in Table 3. The serum was obtained from immune birds
tliat had demonstrated a three-plus complement-fixation reaction of their sera
at various times. The sera were collected and pooled for injection on the
same day that they were administered.
TABLE 3 — Degree of protection afforded by the intravenous and subcutaneous injection
of "immune" serum
Amount of
Results of Check In-
"Immune'" Serum
Check
oculation for Immu-
Group
Administered
Inoculation
nity Determination-
Method
OF
Administration
Bird
Numbers
Dec. 13
Dec. 19
Jan. 6
Jan. 13
Jan. 17
Jan. 20
Jan. 28
cc.
cc.
cc.
Group 1,
(103 to 106)
Intravenously
2
2.5
2.5
Comb and
wattles
scarified
+
(*104-)
+ -f-
(«104-)
(«104-)
Group 2.
(107 to 110)
Subcutaneously
3.5
3.5
5
and virus
suspension
vigorously
rubbed into
-1-
-|--h
-I--I--1-
(Controls)
(111 & 112)
the areas.
+
^*
+ + + +
-t- First evidence of pox nodules, immature and sparse.
-|- -t- Appearance of a few or several well formed pox nodules.
+ + -1- .\ppearance of many pox nodules of mature development.
4--|--f- + Maximum development of pox nodules.
- Negative.
COXTACIOrS F.l'l'niKI.IOM A 209
fiilcrprrtnfiiui.
riie above results iiiclioatcd tliat the intravenous and sul)Oiitaneoiis admin-
istration i>f "immune" serum, for the j)ur]>ose of produeinj>- a j^assive inuiniii-
ity, failed to protect tlie l)irds auainst artirK'ial inoculation with the \ii-us of
contaji'inus ei)ithelioma.
Conclusions.
1. Tlie varyina; and incomplete conn)lement-(ixation reactions of sera froia
birds imnnme to contagious epithelioma indicated that the formulation of a
unit value of protective antiserum, against which the vaccine or its virus
might be standardized, was impossible.
2. The low degree of specific antibody concentration in the sera of liirds
immune to contagious epithelioma and the failure to produce a passive im-
munity indicated that a general immunity was not the sole protective force
against the disease.
3. The development of lesions containing the virus of the disease at the
point of inoculation and on the comb, after the subcutaneous administration
of the triple strength vaccine, which was followed by the production of a
complete immunity, indicated that the skin jjrobably plays an important part
in the immimity against the disease.
CUTANEOUS IMMUNITY.
It appears that there is another inununity factor operating in contagious
epithelioma and it is evident that the subcutaneous injection of virus does
not consistently produce this factor.
De Blieck and Van Heelsbergen (4) have apparently solved the problem
of immunization against contagious epithelioma in European countries by
producing a local or cutaneous immunity. They use a vaccine, known as
"Antidiphtherin", which is "a thoroughly living vaccination material, the
vitality of which has not been decreased, either physically or chemically,
which always gives rise to a local pox-eruption, which never generalizes, which
immunizes against the experimental as well as against the spontaneous infec-
tion, and which is constant for all these properties during all seasons". The
vaccine is applied to an area of denuded feather follicles on the leg by means
of a vaccinating instrument or small trocar. A swelling (pox eruption) of
the follicles results and the investigators claim that birds showing such local
or cutaneous reactions are immune to both the experimental and spontaneous
infections. Hoi (5) claims to have had marked success with "Antidiphtherin"
in Holland. Doyle (6), in testing out samples of the vaccine procured on
the open market, found them to vary in degree of "attenuation", and to
cause a generalized infection in several instances. According to De Blieck
(7) these faults of the vaccine have since been overcome.
De Blieck and Van Heelsbergen do not mention their method of ])roducing
"Antidiphtherin". Their various publications refer but briefly to a descrip-
tion of the general properties of the vaccine and its method of administration.
Experimental Comb Vaccination.
The early attempts to produce a local or cutaneous iiiummity (July, 192fi)
210 TECHNICAL BULLETIN U
were modifications of the metliod of Panisset and Verge (8). Approximately
1 sq. c-:.i. t)f c'oiiil) area was curetted or scraped until Ipnpli was drawn.
Care was exercised not to produce a bleeding surface. Virus suspensions,
unattenuated and attenuated by moist heat at 55° C. for one hour, were
rubbed into the areas. Well formed epitheliomata always developed after
the usual incubation period of from four to seven days, but in at least .50 per
cent of all cases the lesions spread to contiguous surfaces of the comb, not
remaining localized. The virus content of the suspensions was decreased
until 200 milligrams of virus in 50 cc. of physiological saline solution were
used. The results were about the same. The suspensions were then injected
intracutaneously in 0.1 cc. doses, into a barb of the comb. This method of
administering the vaccine greatly decreased the number of cases that pre-
sented lesions on the comb after the vaccination other than at the point of
inoculation.
After the local lesions had fully cleared up in the l)irds used in these experi-
ments, the combs were lightly scarified and unmodified virus rubbed into the
areas to test the degree of immunity production. In all cases where local
lesions had developed following vaccination, the immunity was complete or
nearly so.
Experimental Skin Vaccination.
At the same time tliat tlie above experiments were in progress, various
efiforts to induce a local or cutaneous imnumity by feather follicle vaccination
were being made. Young and healthy White Leghorn cockerels, four to eight
months old, were used in tiie experiments.
In the following experiments the place selected for vaccination was on the
outside of the right leg just above the tibio-femoral joint. An area of ap-
proximately 1 square inch was denuded of feathers, cleansed with sterile
physiological saline solution, dried with sterile cotton, and the vaccine ap-
plied by rubbing into the follicles with a cotton swab attached to a wooden
applicator. The left leg was similarly treated for control purposes, saline
solution being used instead of vaccine. The experiments described below
differ only in the vaccine used and the method of preparing tJie feather fol-
licle area prior to the administration of the vaccine.
Experiment 1.
Twelve birds were used. The vaccine M^as an unattenuated suspension of
0.5 grams of pox virus in 50 cc. of physiological saline solution. The bared
feather follicles and interfollicular spaces were lightly scarified and the vac-
cine applied by rubbing into the area with a cotton swab attached to a
wooden applicator. Five days later a definite swelling of the follicles had
developed and scab formation over the orifices of the follicles was in progress.
On the tenth day after vaccination, the scab formation was fully developed.
The infection became generalized, pox lesions aj)pearing on the comb and
eyelids, in four of the birds. A systemic reaction, as evidenced by droopi-
ness, lethargy, etc., also occurred in these four birds.
While scab formation was taking place, a local pyogenic inflammation
develo])ed, which was undoul)tcdly due to the Staphylococcus aureus and
Pseudomonas aermjinotia content of the fresh virus (2) used in the vaccine.
It was evident that scarification of tlic skin and feather follicles induced this
local inflammation. Scarification was, therefore, contra-indicated.
Tiie fejither follicle scabs were removed, dried, pulverized in a mortar,
susjtended in saline solution and ap))lied to tiie scarified comli areas of normal
CON TACIlorS Kl'irHKl.IO.MA 211
lurds. l.osions of coiUjijiioiis (■|)itlu-li()iii;i (1<'\ cloix-d, iiulicitiiiir tli.i! the scnlis
contaiiu'd tlu' pox virus.
'I'lu' control leiis of the liirds sliowed no rrat'tioii.
After complete recovery from the gener.-d elTects of the viiccination (aliont
22 days), virus inoculation of tiic comb proved tiiat tiiere was an aj)i)arent
imiminity. evidently of a cutaneous nature. Three normal birds were inocu-
lated on the comb as controls for the virulence of the \irus used to check the
immunity. They all developed a heavy infection.
K.vfierlnn-iii J.
Twelve iiirds were used. In this experiment the \accine I'ontained the same
amount of \ irus. It was also unatteiuiated, but passed through several layers
of sterile gau/e and cotton for the purpose of freeing it of epithelial debri.s,
etc. It was injected into each denuded feather follicle, about 1 droj) to each
follicle. The })urpose of the injection was to penetrate the dermal ])apilla,
not dee])ly, init very su|)erficia!ly.
The resulting reactions were not conlined to the follicle alone, l)ut sjjread
fo the adjacent tissue as well. The systemic reaction was not so pronounced
as in the preceding experiment, nor did pox nodides appear on tiie comb and
eyelids. Scal)s did apjiear on the orifices of the vaccinated follicles.
.VU controls were satisfactory and the degree of immunity de\elo])ment, as
determined by check virus inoculation, was pronounced.
J-^.cpi ritiient J.
Fourteen birds \\ ere used. The vaccine in tiiis case contained but 200
milligrams of virus. It was suspended in .50 cc. of a 40 ])er .cent glycerol-
lihysiological saline solution, (llycerol was used for the same purpose as in
the triple strength vaccine; that is, it was an excellent medium for suspend-
ing the virus in solution and it was thought that it might have some attenu-
ating action. The vaccine was applied directly to the follicles by rubbing it
in well with a cotton swab attached to a wooden ai)pIicator. The area was
not previously cleansed nor scarified. The birds reacted to the vaccine in a
uniform manner and with but slight variation, as follows: —
(I. W'itiiin 1 to 8 days aftei' \acciiiation, a gradual swelling of follicles
occurred.
/'. "Within S to IS days after vaccination, a gradual a})pea ranee of follicu-
lar scabs occurred.
r. Within 18 to 31 days after vaccination a gradual disai)i)earance of
follicular scabs and swelling occurred, the latter being the last to disai)pear.
The reactions remained localized. Pox nodides did not appear on the comb.
The birds were check inoculated on tiic comb on the 31st day after vaccina-
tion with a su.spension of virus. They all showed a complete immunity, while
the usual number of controls on the virulence of the virus all presented a
heavy comb infection.
Johnson (9) has reported the experhnental use of such a vacciiu% but
found that it caused unfavorable reactions in laying birds and its use was
oftentimes attended with mortality. The virus content of his vaccine was
slightly higher than that used in Experiment 3. This investigator used a
pared down camel's hair brush in applying the vaccine to the feather fol-
licles. This method of application has since been u.sed in duplicating ex])eri-
ments and found to be both efficient and practicable. Gidlow (10) u.sed the
vaccine as recommended by John.son in .several flock trials and rej)orts that
212
TECHNICAL BULLETIN U
the vaccinated birds did not come down with pox during tlie subsequent
months.
Experiment Jf.
Tliis experiment was planned to determine if birds after recovery from
natural and exjjerimental infection would react to the local or cutaneous
vaccination of feather follicles on the leg. A grou}) of 28 White Leghorn
cockerels that had recovered from contagious epithelioma was used for this
purpose. The control birds of the previous experiments were also used, they
having recovered from the experimental infection. Tiie vaccine containing
200 milligrams of virus in .50 cc. of a 40 per cent glycerol-piiysiological saline
solution was applied to the feather follicles in the manner described.
The results v/ere clear cut. Not one bird developed a local or cutaneous
reaction (pox eruption) at the point of vaccination.
Experiment 5.
An effort was made to determine what effect the age of the cutaneous
vaccine of Experiments 3 and 4 had on its ability to produce a complete
immunity against contagious epithelioma.
TAB E 4 — Results of ageing of the cutaneous vaccine on the subsequent production of
complete immunity
No.
OF Days After
Virus Inoculation of
Comb to
Skin Vaccination
Determine Degree of
Immunity
Age
OF
Vaccine
Maximum
Maximum
E isappear-
6th day
10th day
20;'i day
follicular
swelling
scab
formation
ance of
scab
after
inocu'ation
arte-
inoculatiDn
after
inocu'ation
Bird No.
Days
V91
6
5
none
none
regati\-e
negative
negative
V92
11
6
IS
31
negati .-e
negative
negative
V 93
16
8
23
31
negative
negati -e
negative
V94
19
9
21
35
r.eg- tl e
negative
negative
V95
25
10
1.5
27
regati e
negative
negative
V96
.31
10
none
none
'ight
i iection
■ light
infection
light
infection
V97
40
10
none
none
regative
light
infectian
light
infection
V9^
46
11
none
none
'.ight
i ifection
light
infection
lig'it
i.ifection
62
(controls)
(c nt o'. )
(controls)
(cont.ols)
64
not
'ig. t
i iection
pronounced
infection
heavy
infection
67
vaccinated
Iv-terpretation. The above data indicate that when the vaccine is 2.5 days,
or less, in age at the time it is administered, it j)roduces a complete inmmn-
ity. When older it does not confer absolute protection against the experi-
mental infection. It cannot be said at the present time wliether this attenua-
tion was due to actual ageing of the ^irus in tiie vaccine, the action of the.
glycerol upon it, or some other imsusjiected factor. Furtiier experiments arc
planned for determining this question. The data also show that apparently
coxTAC.iors 1 rriiiKi.ioM.v 2i;j
the sral) form.itioii at tlu' j>oint of iiioiulation is essential to tlu' jirodiiction
of a coiiipletf iinmuiiity.
In c'onjuiu'tion w itii Exjieriiiieiit .5 a f:rou)i of birds was \ acciiiatod cutaiic-
ously witli tlif \at'cine used in that experiment, ilie birds were check inocu-
lated with virus for ininiunity determination at various jieriods after the day
of vaccination. As near as could be determined from the exyierimental evi-
dence in this aroup of birds and the birds of Experiment 5 as well, imnuuiity
began to develop on about tlie 20th day after vaccination, when the follicular
.scab had reached its maxinuan development, continuing until a (•omi)lete im-
nnmity had been produced on or altout the 2f)th to 31st day after vaccination,
at which time the follicular scab and swelling had practically disappeared.
Experiment 6.
A pen of 100 trapnested, ))edigreed Rhode Island Red pullets at the college
poultry plant was used in this exjierunent. The cutaneous vaccine (200 milli-
grams virus in 50 cc. of -l-O per cent glycerol-saline solution) was administered
to 70 of the birds, the remaining 30 being left as controls. The chief purpose
of this experiment was to determine the eflect that the vaccine had on egg
production.
Eighteen days after the vaccination, every one of the 70 vaccinated birds
demonstrated a well formed feather follicle reaction with scab formation.
Three of them showed one or two pox lesions on the comb. It was question-
able whether the lesions on the comb were due to generalization of the virus
in the vaccine or the result of contact infection. Presumably, they were due
to generalization of tlie virus for such lesions were not found in any birds of
the control group.
A close comparison of the trap nest egg records of the control and vacci-
nated groups for one month prior to vaccination and thereafter for an addi-
tional ten wrecks, indicated that the percentage of egg production in the
vaccinated group was materially decreased. This drop in egg production
liegan 8 days after vaccination, and on the 21st day after vaccination it had
reached its lowest point. From then on the production gradually, but slowly,
increased to normal.
Xo opportunity was given to check the immimity production by virus
inoculation of the comb.
This experiment demonstrated two important points. First, that the cuta-
neous vaccine, without exception, always produced a local pox-eruption of
the nature of a follicular .swelling and scab formation, which apparently is
essential to the development of a complete immunity. Second, that its admin-
istration was followed by a material decrea.se of egg production.
Complement-Fixation in Relation to Cutaneous Immunity.
All birds u.sed in Experiments 3, 4 and 5 were bled and the seru obtained
for complement-fixation tests prior to the check inoculation of the birds for
immunity determination. Prior to subjecting the.se sera to complement-fixa-
tion tests, known three-plus sera were u.sed to determine tlie antigenic prop-
erties of the antigen to be used. In the actual test of the sera from the
cutaneously vaccinated birds, positive and negative sera were used as controls.
The u.sual reaction was complete hemolysis. In two cases a slight anti-
complementary action of the sera occurred.
The results indicated that complement-fixing antibodies were not concerned
in the immunity acquired following cutaneous vaccination.
214 TECHNICAL BULLETIN U
Standardization of The Cutaneous Vaccine.
Tlie powdered pox virus used in tlie al)ove ex})erinieiits, wlien Mpjjlied to
the scarified areas of the comb and wattles, always produced pronounced
lesions of contagious epithelioma in susceptil)le l)irds, within four to seven
days after the inoculation. The virus, tiierefore, had an inculcation period
of from four to seven days. This variance of three days depended on tiie age
of the bird inoculated, or in otiier words on tiie individual resistance of the
birds to the virus, and on the age of the virus at tlie time it was used, ^'irus
over one year of age was eliminated.
By varying the virus content of the cutaneous vaccine it was determined
that less than 200 milligrams per 50 cc. of 40 per cent glycerol-physiological
saline solution failed to produce local vaccine reactions in all trials, while
more than this quantity of virus caused a more pronounced generalized reac-
tion. It was demonstrated in Experiment 5 that the vaccine became attenu-
ated in some manner to such an extent that it failed to produce a complete
immunity when more than 25 days old.
Based on this information, a tentative standard has been adopted for the
vaccine, until such time as it may be improved upon. The virus should be
less than one year old and produce definite lesions of contagious epithelioma
in from four to seven days after inoculation. The cutaneous vaccine should
contain 200 milligrams of such a virus suspended in 50 cc. of a tO per cent
glycerol-physiological saline solution. It should not be attenuated by heating.
It should be used within 25 days after manufacture, preferably within 10 to
15 days. When continuous ice-box storage is not available, 0.5 per cent phenol
may be added to the suspension.
Discussion.
It is apparent, from the residts of the fi^regoing experiments, that local or
cutaneous vaccination of an area of bared feather follicles, with the proper
virus suspension or vaccine, results in the production of a local pox-eruption
and the development of a complete immunity.
The theory of cutaneous immunity in relation to contagious epithelioma is
comparatively new. ^'erge (11) refers to it as a "cuti-immunity" and states
that it can be obtained by cuti-vaccination. The experimental evidence in
this paper confirms this statement. Verge used 1/10 to 1/20 cc. of his cuta-
neous vaccine injected intracutaneously into the wattle. He furtlier states
that "the general inuinmity is in reality only an innnunity of the ectoderm",
and claims "that the protection in the cutaneoiis and nuicous membranes
isolates the structures remaining sensil)le and thus creates a refractory state
that extends to the whole body".
It will be recalled that while producing "imnume" serum for serological
study the subcutaneous administration of a triple strength vaccine produced
a complete imnumity in all cases where scablike lesions containing pox virus
develoj)ed at the point of inoculation. In many of these cases pox nodules
also appeared on the comb. It was csideiit that the skin was actively con-
cerned in the immunitj' jiroduction. Beach (12) has made similar observa-
tions. He used subcutaneous injections of vaccines containing lesion tissues
froni fowls a fleeted with the disease and noted that the percentage of birds
immunized thereby was higher among those that developed scab lesions at
the point of inoculation. He, likewise, found these scab lesions to ccmtain
pox virus. The evidence is convincing. In order to [produce a uniform, con-
(•()\r Aciors i;i'ri'iii:i.i()M A 215
sistont, and finu|ilct«' iriiniuiiity tin- .skin must Ix- \ acciiiatcd and a local ])()x
reartioM |>r(>diK'ed at the jjoiiit of inoculation.
Hcsredka's (i;j) ex])Ianatioii of cutaneous ininuinity, as it is produced in
several luainnialian diseases, hinges on two propositions, wliich iti the writer's
mind are directly applicable to eontajrious epithelioma:
1. "'I'he susceptibility of the animal is limited principally, if not exclusively,
to the cells of the skin.
2. "The immunity of the animal is due to the vaccination of the reee])tive
cells."
The (piestion naturally arises, how docs this imnmnity s])reatl from a local-
ized area on tlie skin to the entire cutaneous surface? Additional investiga-
tion is necessary before the question can be answered. Besredka is inclined
to believe that "the immunity is local, but its effect reacts upon the rest of
the cutaneous surface, because of the large net of lymphatic vessels interested
in the jirocess". AVhatever the explanation may be, it has been demonstrated
that the entire cutaneous surface is immune to the virus.
SUMMARY.
A triple strength vaccine, containing 30 milligrams of active pox virus per
dose, when administered to birds subcutaneously, produced a high degree of
imminiity to contagious epithelioma. However, scablike lesions and pox
nodvdes developed at the point of inoculation and on the comb and eyelids,
respectively, following its administration. These lesions were found to con-
tain pox virus. Those birds showing such skin lesions were completely im-
nume to the disease, as deteriuined 1)V check inoculation of the comb and
wattles with the virus.
Blood was taken from these and other iiunuine birds for serological study.
It was found that from 20 cc. to 2.5 cc. of blood could be safely drawn from
the brachial vein of a mature bird in good physical condition without any
apparent ill results.
Complement-fixation reactions varied and on the average were but a two-
plus reading. Occasionally, a three-plus serum was found. Precipitin studies
were negative. The attemjit to ])roduce a ])assi\e inmiunity was a failure.
It was, therefore, concluded that, because of the relatively low concentration
of specific antibodies in the sera of immune birds, and the failure to produce
a passive nnnnmity, a general immunity was not the sole protective force
against contagious epitheliojua.
The development of lesions containing pox virus at the point of inoculation
and on the comb and eyelids after the subcutaneous administration of the
triple strength vaccine, which was followed by a complete iiunumity, indi-
cated that the skin was actively concerned in the immunity against the disea.se.
Various vaccines were applied to scarified com!) areas and injected into the
barbs of the coiub; also, to scarified areas of denuded feather follicles and
interfollicular skin surface on the leg just above the tibio-femoral joint.
Unfavorable reactions followed. The vaccines were then applied to the de-
nuded feather follicles, without previous scarification or cleansing, by rubbing
them directly into the follicles v\ith a cotton swab attaclied to a wooden aji-
plicator or a pared down camel's hair l>rush.
A cutaneous vaccine containing 200 milligrams of virus susj)ended in .50 cc.
of a 40 per cent glycerol-physiological saline solution (2 parts glycerol and 3
pai-ts saline) always caused a swelling of the feather follicles followed by
the development of scabs over the orifices of the follicles. This vaccine
216 TECHNICAL BULLETIN U
always produced a complete iiiiiimnity of a cutaneous nature after the de-
velopment of the local pox eruption.
Additional experuncnts showed that this cutaneous vaccine always produced
the local pox eruption, which was essential to the development of a complete
immunity, but its administration was followed by a decrease in egg produc-
tion. It was also determined that com})lement-fixing antibodies were not
produced during the development of cutaneous immunity against contagious
epithelioma.
A tentative standard for the cutaneous vaccine was adopted. The virus to
be used should have an incubation period of from four to seven days and,
therefore, must be less than one year old. The vaccine should contain 200
milligrams of such a virus suspended in 50 cc. of a 40 per cent glycerol-
physiological saline solution. The product shovdd not be attenuated by heat.
It should be used within 25 days after its manufacture, preferably within 10
to 15 days, because it does not always produce a complete immunity when
older. If continuoTi.'^ ice-box storage is not available, 0.5 per cent phenol
siiould lie added as a preservative.
BIBLIOGRAPHY.
(1) Beach, J. R. : The treatment and prevention of chicken-pox (con-
tagious epithelioma) of fowls. Jour. Amer. Assoc. Instr. and Invest.
Poultry Husb., 1920, vii. No. 1.
(2) Pyle, Normax J.: The therapeutic efficiency of avian diphtheria, rouj),
and bird pox vaccines and bacterins. Mass. Agric. Exp. Sta. Tech.
Bull. 10, 1926.
(3) Sweet, Clifford D.: A study of epithelioma contagiosum of the com-
mon fowl. Calif. Univ. Pubs. Zool., 1913, ii, 29-51.
(4) VAX Heelsbergex, T.: Vaccination against diphtheria and fowl po.\
with Antidiphtherin. Vet. Rec, 1925, v. No. 24.
(5) HoL, G. H. G. : Antidiphtherin vaccine for diphtheria and pox in poul-
try. North Amer. Vet., 1927, viii, 44-46.
(6) Doyle, T. M.: Tests of the de Blieck-van Heelsbergen method of im-
munization against fowl pox. Vet. Rec. 1926, vi, 741-743.
(7) DE Blieck, L.: Diphtheria and pox in chickens, and the combating of
this disease. Proc. World's Poultry Congres.s, Ottawa, Canada, 1927,
290-294.
(8j Paxisset, L. and "S'erge, J.: Immunity in a\ian diphtheria and con-
tagious epithelioma (trans, title). Compt. Rend. Acad. Sci. (Paris),
1923, clxxviii, 345-347. Cited in E. S. R., Aug. 1924, li, 184.
(9) JoHxsox, W. T.: Fowl pox prevention by immunization. Jour. Amer.
Vet. Med. Assoc, 1927, Ixxi, N. S. xxiv, 750-763.
(10) GiLDows E. M.: Fowl pox. New Eng. Homestead, Jan. 21, 1928, xcvi,
14-15.
(11) Verge, Jeax: Reclierches Experimentales sur I'Affection Diphtero-
Variolique des Oiseaux. 1926. 230 p. Toulouse, France, J. Bonnet.
(12) Beach, J. R.: The immunization of fowls against chicken-pox (Epith-
elioma contagiosimi) by subcutaneous injection of virus. Hilgardia
(Calif. Agric. Exp. Sta.), 1927, iii, 41-97.
(13) Besredka, a.: Local immunization. Edited and translated by Harry
Plotz, 1927. 181p. Baltimore, U. S. A., AVilliams and Wilkins.
Massachusetts
Agricultural Experiment Station
Technical Bulletin No. 15 June, 1928
The Extraction of Apple Juices
in the Manufacture of Jelly
By Carl R. Fellers
The many different practices used in jelly making at present would in
themselves indicate that the choice of a method has only experience as a
basis for its selection. This investigation is an attempt to establish cer-
tain principles which may lead to greater economy in production, and
provide a sounder basis for the housewife and commercial preserver for
the manufacture of a uniformly high quality product.
Requests for bulletins should be addressed to tlie
AGRICULTURAL EXPERIMENT STATION
AMHERST, MASS.
THE EXTRACTION OF APPLE JUICES IN THE
MANUFACTURE OF JELLY
By Carl R. Fellers
Research Professor of Horticultural Manufactures
INTRODUCTION
In the manufacture of apple jellies, pectin concentrate, extract, or syrup,
it is customary to extract the juice from the fruit with water by the use of
heat. Whether the operation be carried on in the home or in the commercial
manufacturing plant, the principles involved remain the same. In the past
very little attention has been given to the study of extraction methods not
only of apples, but of all the common juice or jelly-yielding fruits.
It is the purpose of this bulletin to present data bearing upon this problem.
It is not proposed to discuss cold pressed juices such as cider, because the
cold press is a distinctly different problem. In the latter case, very little
pectin is obtained in the juice, whereas in the heat extracted juice, consider-
able pectin is found. True, cider jelly without the addition of pectin may
be manufactured from cold pressed apple juice, but only by concentration
to one-sixth to one-eighth of its original volume. This is necessary liecause
of the small amount of jellifying pectin and sugar which is present in cold
pressed apple juice. Heat disintegrates the pectin-ricJi cell walls of fruits
and thus releases the pectin into solution. To a lesser degree freezing (5)
accomplishes the same purpose.
Directions for juice extraction are far from standardized in that various
ratios of fruit to water, and diverse periods of extraction at various temper-
atures are reconunended. In some cases the fruit is sliced, in others pulped
and yet again the use of added acid to aid in the extraction is advocated.
Often the directions call for one extraction only, sometimes two, and occa-
sionally three. How is the commercial plant or the farm factory operator
or even the housewife to know which methods are best? The literature
bearing on this subject is appalling in its diversity of methods and lack of
orderly scientific approach.
PLAN OF STUDY
In order to determine the yield of juice as well as the relative amounts of
the three recognized jelly essentials, namely sugar, pectin, and acid, whicii
were extracted from apples by the use of various methods of heat extrac-
tion, a series of laboratory tests were conducted during 1926 and 1927 on
Baldwin, Red Astrachan, Rhode Island Greening, Winesap, Mcintosh,
Wealthy, King David and Red Siberian Crab varieties. Under standardized
conditions, given weights of apples were successively extracted three times at
<each of the following temperatures, viz. 8S^C. (190°F.), 100°C. (212° F.),
and 109 °C. (228° F.), respectively for 15, 30 and 60 minutes. A 15-minute
extraction followed by a standing period of 10 minutes, was also employed.
Sliced apples were compared with chopped or pulped fruit. Similarly, the
effect upon the extractives of various concentrations of acid added to the
apples was considered. The ratios of apple to extraction water, by weight,
were varied in this study from 3:2 to 3:4.
APPLE JUICE EXTRACTION FOR JELLY 219
Equipment
Willi tlie exception of Red Astraciiaii, Rod Siberian Crab and Wealtliy
varieties, uniform Grade C apples kept in cold storage for from one to tliree
and one-half months were used. Just enough of the fruit was lirougiit to
tlie laboratory each morning for one day's run. Only sound, firm fruit was
used. A hand slicing machine adjusted to give slices one-eighth inch in
thickness, and a large sized food chopper equipped with medium knives were
used to slice or chop the fruit as desired. Pieces not over one-fourth inch
in diameter were obtained by the use of the food chopper. Tiie ciiopped
apples were somewhat finer in texture than the press stock usually obtained
liy the average shredder or mill in a cider factory.
Unciilorinated Amherst tap water (pH 6.9) was used for the extractions.
Aluminum stew pans with close-fitting covers were found to be convenient
utensils in which to cook the apples with water. Circular gas burners were
used as the source of heat except where temperatures al)ove 100°C. were
desired. In tiie latter case steam-heated retorts or autoclaves were used.
Extraction Methods
Three pounds (L36 kilograms) of either sliced or ciiopped fruit were
placed in the covered aluminum pans, together with the required weight of
cold water. The gas flame was turned on fully until the desired temj)erature
was reached. It was then adjusted so as to maintain this temperature for
the period of the extraction. Where the ratio of fruit to water was high,
as in the 3:2 ratio, the pan contents were stirred with an aluminum spoon
when necessary to prevent scorching; otherwise no stirring was done.
Upon completion of the extraction period, the contents were poured into
a moist cheese cloth (2 thicknesses), allowed to drain one minute, then well
squeezed by wringing both ends of the cloth in the hands for another minute,
after which the pulp was returned to the original stew pan, the desired
quantity of cold water added, and again extracted over the gas flame.
Similarly, a third extraction of the pulp was obtained. Thus for each apple
sample there was secured and kept separately, first, second and third ex-
tracts, as well as the pulp remaining after separation from the third ex-
tract. These various extracts and pulps were at once carefully weighed,
placed in glass fruit jars and pasteurized for 30 minutes at 71°C. (160°F.)
in case they could not be examined unmediately. Both chemical and organ-
oleptic examinations were made. Wiicn tartaric acid was used to acidify
the fruit, the desired amount in solution was added in the extraction water.
None was added except in the first extraction. For the sake of uniformity
the same quantity of water was added to the pulps remaining after the first
or second extractions, as was originally added to the apples. The several
juice extracts as well as the pulp were reserved for chemical examination and
for use in preparing the jelly samples.
Method of Preparing Jelly
For the sake of uniformity a given weight of juice, 511 grams (18 ounces),
was taken from each extract for conversion into jelly. When it was desired
to make a jelly representing the combined first and second extracts, amounts
proportional to the yield of each were taken so that the total weight was 511
grams. To this was added enough sugar to total 341 grams (12 ounces),
220 TECHNICAL BULLETIN 15
allowance being made for that present in the juice. The juice was then con-
centrated to a sheeting, spoon jelly test (104° to I05°C. or 219° to 221°F.).
The finishing point was checked by determining the weight of the pan
and contents from time to time. Inasmuch as jelly formation usually occurred
when the sugar concentration reached 67 to 69 per cent by weight, this fur-
nished a simple and accurate check upon the finishing point. The jelly was
weighed in the pan at once and poured through one thickness of medium
mesh cheesecloth into straight sided, 2-ounce jelly glasses. After skimming,
these were covered with melted paraffin and capped. Jelly strength, chem-
ical, and organoleptic determinations were made after 1 to 3 months storage
at 21° to 23° C.
At first, jellies were made from all three extracts individually, together
with combinations of the first and second, as well as all three extracts com-
bined. Third extractions produced juice which seldom gave satisfactory
jellies unless acid was added. Moreover the pectin content was usually too
low to give a firm jelly. Naturally a jelly made from the third extract was
of poor color and flavor. Although combinations of all three extracts gave
fair to good jellies, it was decided that a combination of the first and second
corresponded more closely to home or factory practice and this was the pro-
cedure finally adopted.
Chemical Methods
Each sample of juice, i. e., first, second, and third extracts, was examined
for total titratable acidity, hydrogen ion concentration, soluble solids by
Abbe refractometer, Brix and specific gravity hydrometers, pectin by the
alcohol precipitate and centrifugal methods, pectic acid by the A. O. A. C.
(2) and centrifugal methods, sterility, and such physical characters as taste,
color, turbidity and sediment. Determinations of insoluble solids, soluble
solids, alcohol precipitate, pectic acid, total aciditj', pH value, and sterility
were also performed on the pulp. Determinations of total sugars, reducing
sugars and sucrose were made in some case?. The A. O. A. C. methods were
used wherever possible.
The hydrogen ion concentration was determined in most cases colorimet-
rically, though several check determinations with the potentionmeter were
made. Notwithstanding the statement of Myers and Baker (8) that the
colorimetric method was of no value in the study of fruit juices, it was found
to l)e very useful and economical of time. Occasional checks by the electro-
metric method showed good agreement in pH values.
In order to prepare samples for pectic acid, acidity, sugar determinations,
etc., from the pulp or fresh fruit, 300 grams of the well pulped and mixed
samples in a 2-liter beaker with 800* c. c. of water were boiled one hour,
the volume being kept constant by the addition of hot water at intervals.
The contents were transferred to a 2-liter flask, cooled, diluted to volume,
and filtered.
It was found that, in drying the residue for the determination of total
solids, losses occurred where a temperature above 60° C. at 25 inches of
mercury in the vacuum oven was employed. This was due probably to
levulose decomposition. Similarly in drying alcoholic precipitate or pectic
acid it was found that decomposition occurred at 100° C, hence a tempera-
ture of 90 to 95 °C. was used. The absolute necessity of washing these
precipitates free from hydrochloric acid was also observed, as charring and
loss invariably occurred when this precaution was not fully carried out.
APPLE JUICE EXTRACTION FOR .lEI.I.Y •221
Color, taste, turbidity, and sediment were {riven ratings on a basis of organ-
oleptic tests only. Four classes were made in each case — for example, the
color or taste was considered excellent, good, fair, or poor. Likewise tiirhid-
ity and sediment were classified as much, moderate, slight, or none.
Sugar in Jellies was determined by the Abl>e refractometer. In tlie case
of juices a'comjiarison was made of tiie results obtained by the Brix hydro-
meter, refractometer and chemical determination of soluble solids. These
results are discussed elsewhere in this bulletin.
Centrifugal Method for Pectic Acid
A centrifugal method was developed during the course of the work and
will be described more fully in a separate publication. In brief, it consisted
of measuring either 5 or 10 c. c. of juice, according to concentration, into
15 c. c. tapered, graduated, glass centrifuge tubes. When 5 c. c. were em-
ployed, the juice was always diluted to the 10 c. c. mark before the addition
of alkali. One c. c. of a 10 per cent sodium hydroxide solution was added
to each tube, the contents mixed by shaking and allowed to stand for 15
minutes. Two c. c. of a 10 per cent hj'drochloric acid solution were intro-
duced and thoroughly mixed. The tube was then placed in boiling water for
5 to 8 minutes (or until the gelatinous precipitate was entirely flocculated
and freed from air bubbles), removed, cooled to below 25°C. and whirled
in a centrifuge 15 minutes at 2600 revolutions per minute on a 14-inch
head. It was found within certain limits that the vohmie of the precipitate
could be correlated with the chemical determination of pectic acid. Tiie
precipitate was also centrifuged without previous heating, but the readings
were much higher than with the heated precipitate and less consistent. It
was found that the centrifugal method did not yield reliable results when
applied to the alcohol precipitate of juices or pectin extracts.
For practical purposes, it was believed that the chemical determination
of pectic acid yielded more reliable results than the alcohol precipitate
(pectin). It has i)een proved recently by Nelson (11) that the pectic acid
obtained in this manner consisted of a definite stable substance, namely
digalacturonic acid. The composition and physical properties of the alcohol
precipitate w-ere foimd to be variable and it was difficult to obtain satis-
factory checks by using the method. The ratio of pectin (alcohol precipitate)
to pectic acid, A. O. A. C. method, was variable but averaged approximately
1.7:1 for Baldwin apples. Wichmann (19) estimated that this ratio for most
varieties of apples was about 2:1.
Jelly Strength Determination
Obviously the jelly strength test commonly employed, i. e., the resistance
of the jelly to the fingers, is inaccurate. Paine (12), Sucharipa (16), and
Tarr (18) have described the advantages of a suitable jelly-strength tester.
Sucharipa (16) devised a tester which would break a laj'er of jelly of definite
thickness by means of air pressure, the latter being registered on a mano-
meter. Later Tarr (18) and Baker (3) modified and simplified Sucharipa's
apparatus by using water pressure in place of air pressure.
By substituting a light paper cup in place of the heavier metallic one
furnished with the Bloom gelometer (13), a standard instrument used in
gelatin and glue testing, the writer found this instrument gave very satis-
factory results in determining the jelly strength of fruit or pectin jellies,
jams or sauces.
222
TECHNICAL BUI>LETIN 1^
rigure 1. Bloom Gelometer Used to Make Jelly Strength Tests.
The instrument is electrically manipulated and allows fine shot to flow
into the pan above the plunger until the latter has penetrated into the
jelly a definite depth, usually 5 nini., when the circuit is closed and the
flow stopped. The jelly strength is merely the weight in grams of the shot
on the pan. Straight sided jelly tumblers containing .56 grams (2 ounces)
of jelly, were used in the tests, all of which were conducted at room tem-
perature, 20° to 23°C. (68° to 74° F.). The flow of shot was always regu-
lated so that a definite weight flowed through the outlet in a given time — in
other words, the orifice was opened exactly the same distance (regulated
on the instrument) for each test. For details regarding the use of the
Bloom gelometer for jelly strength determinations in fruit and pectin jellies
the paper by Fellers and Griffiths (6) should be consulted.
CHEMICAL COMPOSITION OF THE APPLE VARIETIES
USED FOR JUICE EXTRACTION
Before extracting the juice by heat, representative samples of each variety
were subjected to chemical examination. In most cases several analyses of
APPLE JUICE EXTRACTION FOR .JEI.I.Y 2J3
a variety were made and the average taken. Baldwin ai)i)k's were analyzed
at approximately the same degree of maturity for three successive years.
All these data are presented in Table 1.*
The composition of the Baldwin apple varied somewhat from year to year
though these differences were not striking. In 1925 and 1927 there was a
small amount of residual starch even in the mature fruit. It is possible tiiat
■this may be the cause of the cloudiness which is often associated with Bald-
win apple jelly. The pectin is reported as pectio acid, but this figure may
be converted readily to pectin (alcohol precipitate) by nmltiplying by a
ifactor found to vary from 1.5 to 2.3. Inasnmch as the factor is not constant
and depends upon the amount of hydrolysis or deesterification that the pectin
has undergone as well as upon impurities, the pectic acid values are prefer-
table to the alcohol precipitate.
The ash of the several varieties showed but little variation. Pectin varied
Ifrom 0.29 per cent in the 1925 crop of Baldwins to 0.62 per cent in Red
Siberian Crab. Starch was usually present only as a trace in mature fruits.
Total sugars made up approximately 81 per cent of the soluble solids pres-
ent. The ratio of sucrose to reducing sugar proved extremely variable even
in the same varietj'. The insoluble solids averaged 2.5 per cent in Baldwins
and 2.7 per cent in other varieties. The mean aciditj- for Baldwins was 0.53
per cent as malic acid while other varieties varied considerably. The mean
pH of Baldwins was 3.46, the range for other varieties being from 3.18 for
Red Astrachan to 3.6 for Mcintosh. Shaw (14), Bigelow, Gore and How-
ard (4) and Alwood, Davidson and Moncure (1) reported analyses of a
large number of varieties of apples grown in different localities, at several
stages of maturity and during storage. For a more complete discussion of
the chemical composition of the apple, reference should be made to those
reports.
STUDIES OF THE EXTRACTED JUICE
Yield and Composition of Extracted Juice per Unit
Weight of Apples
Although analytical data were obtained on three successive extracts and
the residual pulp as well, only two extracts are usually considered here.
Ordinarily in jelly making, only two extractions are made. The third ex-
traction usually yields a juice which is too dilute to concentrate and use
economically for jelly. Fruit flavors and colors as w^ell as pectins were
injured readily or even destroyed completely by prolonged heating, and for
this reason it was considered inadvisable to mix the third extract with either
fthe first or second or both.
Table 2 was summarized from 126 tests including Baldwin, Red Astra-
•chan, Red Siberian Crab, Wealthy, Mcintosh, King David, Winesap and
Rhode Island Greening varieties. Obviously the ratio of fruit to extraction
water greatly influenced the amount of juice obtained from a given extrac-
tion period. For example the mean yield increase of a 3:3 and 3.4 ratio
over a 3:2 was 33 and 91 per cent respectively. Hence the juice from the
i3:2 ratio of fruit to water was much more concentrated than the others and
required less evaporation to convert it into jelly. Since the pectin and acid
content of such a concentrated juice is relatively high, more sugar may be
utilized thus greatly increasing the jelly yield. In general, where the juice
was concentrated as in the 3:2 extract, the total amount of pectin obtained
'The tables are presented in the appendix at the end of the bulletin.
224 TECHNICAL BULLETIN 15 t
from a given weight of apple was somewhat less than where greater dilu-
tions were used. On the other hand since longer heating was required for
the conversion of a dilute juice into jelly, there was a more or less serious
loss in the jellifying power of the pectin. In other words the total amount
of pectin maj' be greater in the more dilute extracted juices, yet because of
deesterification (10) of pectin by heat and acid, its jellifying qualities may
be impaired. This hydrolysis of the pectin may affect seriously the yield and
the quality of the resultant jelly if the extraction period is too extended.
This readily became apparent when a 30-minute extraction period was used,
and caused irreparable loss in jelly yield and in quality if extended to 60
minutes. (See Table 14).
Due to evaporation during extraction, the yields of juice from the longer
extraction periods were less in general than those from shorter periods,
though the former contained more solids including pectin. As already stated,
this may mean little, because the pectin may be partly demethoxylated and
of poorer quality (7, 9, 10). The results here presented indicate clearly that
a loss occurred. The least amount of pectin was extracted in the 15-minute
extraction period, the most in the 60-minute period. There was little differ-
ence in either juice yields or composition in the 15-minute extraction period
as compared with a 15-minute extraction period followed by 10 minutes
standing removed from the source of heat. An additional column in Table
2 was inserted to show the actual amounts of pectin obtained in the com-
bined first and second extracts from one kilogram of apples. The amount
increased with both the length of the extraction period and the widening of
the apple-water ratio. The same observation applied though in a different
degree in the case of soluble solids and acidity.
Composition of Heat Extracted Apple Juice (8 varieties)
(Extraction Period, 15 Minutes at 100°C.)
Table 3 is largely self-explanatory and was prepared to show the general
composition of the juice obtained from the first, second, and third extrac-
tions and of the residual marc or pulp. Since different temperatures or
periods of extraction showed only minor variations in the composition of the
juices, they have not been included. The chopped fruit yielded slightly more
concentrated juices than the sliced; similarly tlie longer the extraction, the
more concentrated became the juice, but for practical purposes these dif-
ferences were insignificant.
The averages reported in Table 3 were found to be very similar to others
compiled using different methods of extraction and even distinct apple va-
rieties. Some varieties were better for jelly making than others (see Tables
10 and 11), but as previously shown in Table 1, there were no striking varia-
tions in solids among the eight varieties analyzed. The differences in pectin
content were marked and merited some attention. The large amount of pectin
remaining in the residual pulp was striking. The A. O. A. C. methods (2)
of analysis, no doubt, were responsible for a part of this high percentage.
The method required boiling 300 grams of the pulp with 800 c. c. of water
for 60 minutes, replacing from time to time the water lost by evaporation.
Thus, much additional pectin was brought into solution. Moreover, by re-
peating the operation still more pectin was obtained. The method yielded/
arbitrary results only, and did not indicate the absolute amount of pectin
present in the pulp. Nevertheless it proved of value in comparative work
such as this. In general, a slightly higher percentage of pectin was found
APPLE JUICE EXTRACTION FOR JELLY 225
(see Table 3) in the residual pulp than was obtained in the first extract
from the fruit. However, since the weiglit of juice obtained in tiie several
extractions was nuich greater than that of the pulp, the actual amount of
pectin remaining in tiie pulp was much less. This is shown clearly in Tables
10 and 11. An average of 31 tests, where the ratio of fruit to extraction
water was 3:2 and the tune 15 minutes at 100°C., showed that 25.5 per cent
of the pectin remained in the pulp. AVliere the ratios were 3:3 and 3:4, the
percentages of pectin in the pulps were 20.4 and 19.1 respectively.
In Table 4 are found the mean actual amounts of soluble solids, pectin and
acid calculated to malic, in the combined first, second and third extracts plus
that found in the residual pulp. These figures approximated the composition
of the fruit itself. Altliough some differences existed among the several
varieties, there were only slight variations between tests on the same variety.
For example, in 91 different tests on Baldwin apples, the mean value for
soluble solids was 134.7 grams per kilogram of fruit with an average devia-
tion from the mean of 9.55 grams. Similarly the values for pectin were 4.84zt
0.50 and for acid calculated as malic, 4.01 ±0.57. Red Siberian Crab, Wine-
sap, and King David gave the highest yields of soluble solids; Red Astrachan
and Wealthy were among the lowest. The Winesap and Baldwin varieties
carried the most pectin while Mcintosh, Red Astrachan and King David
contained the least. Red Siberian Crab, Red Astrachan and King David
produced the juice of highest acidity for jelly, while Mcintosh and Baldwin
yielded juice of low acidity.
Comparison of Brix Hydrometer and Abbe Refractometer
for Solids and Sugar D eterndnations in Apple Juice.
Check determinations of soluble solids (mainly sugar) were made by both
the Brix hydrometer and the Abbe refractometer, to ascertain which gave
the more accurate results. Percentages of sugar from a direct reading suc-
rose scale on the refractometer also were compared with the refractive index
values as computed from the Tables by Schonrock (2) and with the total
sugars as determined gravimetrically. Temperature corrections were made
according to Stanek (2), though every effort was used to make readings at
20° C. so that such corrections usually were unnecessary. The A. O. A. C.
table (2) for correcting saccharonieters for temperature variations was used.
The results obtained from the chemical determinations of solids and sugar
made according to the A. O. A. C. methods are given for comparison in
Table 5.
The sugar constituted an average of only 76.5 per cent of the soluble solids
present in the juice. For this reason care must be exercised to report Brix
or refractometer readings in terms of solids rather than sugar. The Brix
reading was too high consistently as were the values olitained from the
direct reading sucrose scale on the refractometer. The latter registered a
mean difference of 1.45 per cent over the total sugars actually present. The
computation of solids from refractive index gave good results, though they
were usually slightly lower than those obtained by use of the gravimetric
method. As a result of 263 comparisons between Brix and refractometer
readings on total solids in heat extracted Baldwin apple juices, the Brix
method gave a mean increase in solids of 0.308 per cent. In 101 tests on
other apple varieties the increase was only 0.22 per cent. These results
corroborated certain data recently reported by Slierwood (15) on the re-
fractometer analysis of sugar beet juice.
226 TECHNICAL BULLETIN 15
Likewise comparisons were made on jellies between refractive index and
direct sucrose readings from the refractometer scale. Here the results
cliecked very well. For example in 66 determinations the sucrose readings
averaged only 0.053 per cent more than the solids calculated from the re-
fractive index. The direct reading sucrose scale on the refractometer should
not be relied upon in testing apple juices for solids or sugar. The refracto-
meter gave more reliable results than the Brix hydrometer.
Hydrogen Ion Concentration and Titrutable Acidity
of Extracted Apple Juices and Jellies.
Data collected relative to the titratable acidity and pH of apples are pre-
sented in Table 1, while the resulting juices and jellies are considered in
Table 6. In general, the hydrogen ion concentration of the juice was slightly
lower than that of the fruit itself. The second and third extracts were lower
than the first. The resulting jellies made from the combined first and second
extracts corresponded closely in pH to the juices. Tarr (17) stated that pH
3.46 was the minimum at which jelly formation occurred with a relatively
pure source of pectin. From Table 6 it is evident that some of the jellies
made from apple juice exceeded this value. Apparently natural fruit juices
did not behave like pure pectin in this regard.
The relation existing between total titratable acidity and hydrogen ion
concentration was reasonably constant, i. e., the higher the hydrogen ion
concentration the higher the acidity and vice versa. Some varieties appeared
to contain more buffer substances than others. The total acidity, calculated
to malic, of the finished jelly varied from 0.22 per cent in Mcintosh to 0.66
per cent in Red Siberian Crab, while the pH varied from 3.2 in Red Astra-
chan to 3.62 in Baldwin, 1925 crop.
Recovery of Soluble Solids, Pectin and Acid in Successive Extractions
Made under Various Conditions
Data were compiled bearing upon the influence of the following factors
upon the percentage composition of the extracts from eight varieties of
apples.
1. Successive extractions
2. Sliced and chopped apples
3. Ratio of fruit to water during extraction
4. Time of extraction
5. Temperature of extraction
6. Added acids
7. Yearly variations of Baldwin apples
Tables 7 — 12 and Charts 1 — 5 contain condensed analytical data showing
the mean percentage of soluble solids and pectin successively extracted from
eight varieties of apples by the use of various methods. Weighted averages
were used throughout. Due to space limitations, physical and organoleptic
observations were omitted. In spite of the number of tests some few incon-
sistencies occurred, yet on the whole the results were fairly uniform and
showed definite trends. Because of the large nimiber of tests with Baldwin
apples under controlled conditions, it is believed that considerable significance
may be attached to them.
APPLE JUICE EXTRACTION FOR .IKI.I.V 227
[iifltteiice of Surcetisive Extract ions
Three successive extracts, togetlier Mith the pulp ()l)(aiuc(l iu the luiiiuier
already described, froui 91 series cotuprisinji in all 361. sauii)Ies, were exam-
ined nu^iutitatively for soluble solids (chiefly sugar), pectin and acid. Know-
ing the amount of juice recovered at each extraction and keeping the amount
of apples used a constant, viz. 3 pounds (1.36 kilograms), it was possible,
with the aid of these analytical data, to calculate the percentages of soluble
solids, pectin and acid recovered in the several successive extracts and resi-
dual pulp.
Method of Calculating Resulls — To illustrate this method of calculation
assume that the first, second, and third extractions and the remaining inilp
weighed .50, 40, 32, and 16 ounces respectively. The soluble solids were de-
termined to be 10, 6, 3, and 3 per cent while the pectin was 0.20, 0.09, 0.0 1 and
0.12 per cent respectively, and the acid as malic 0.20, 0.12, 0.06, and 0.07.
The total weight of soluble solids in 3 pounds of apples becomes (50X-10) +
(40X-06) + (32X-03) + (16X-03)=8.84 ounces. The total amount of pectin and
malic acid extracted in each separation was calculated in a similar manner.
Knowing the amount and composition of each extract, the exact percentage
of the soluble solids, pectin or acid obtained in any of the three extracts or
in the remaining pulp was readily calculated. For example, the total solids
retained by the pulp after three successive extractions was found to be
(16X-03)-^-8.84=5.43 per cent.
Discussion of Results. Three successive IS-minute extractions of sliced
Baldwin apples with water removed from 80 to 94.4. per cent of the soluble
solids and from 63.5 to 88.3 per cent of the pectin, whereas the chopped
fruit yielded soluble solids ranging from 85.9 to 95.4 per cent, and pectin
from 63.1 to 83. In only two cases did the sliced fruit yield more soluble
solids than the chopped. In each of these, the extracting temperature was
109°C. (228° F.). Though more solids were obtained in the juice by chopping
the apple, still the actual gain was slight. It was realized that the amounts
of solids or pectin obtained from fruit under different methods of extraction
tended to become equalized when the totals of the three extracts were con-
sidered. Hence the percentages of solids and pectin recovered by one extrac-
tion and by two successive extractions .showed wider differences. For this
reason these results were included in the tables. Three successive extrac-
tions with equal weights of water removed from the Baldwin apple all except
about 10 per cent of the soluble solids, which remained in the residual pulp.
The maxinunn amount remaining in the pulp was 22 per cent in the case
where the ratio of apple to waiter was 3:2 and the extraction was carried
on at temperatures of either 88° or 100°C. The minimum amount of ex-
tractable solids remaining in the pulp was only 4.6 per cent. This occurred
where the ratio of fruit to water was low, viz. 3:4, and the extraction tem-
perature was 100°C.
In tables 7 to 12 and graphically in Chart 1 data are as.sembied .showing
the relative percentage recoveries of soluble solids, pectin and acid which
may be exi)ected from one, two, or three successive extractions at varying
periods and temperatures. Likewise the influence of acid added to the
apples, as well as varying ratios of apple to water, are shown. Charts 1 to
5 graphically portray the effect of a variety of conditions on tiie chenucal
composition and percentage yield of the extracted juice.
228
TECHNICAL BULLETIN 15
Chart 1. Relative Amounts of Total Soluble Solids, Pectin and Acid Recovered in
each of Three Sucessive Extracts and in the Residual Pulp. Random Tests.
Temperature of Extraction 100° C. Ratio of Sliced Apple to Water 3:2.
No Added Acid.
Black — first extract
Doable crosslines — second extract
Single crosslines — third extract
White — residual pulp
SOLUBLE SOLIDS PECTIN MALIC ACID
*Boiled 15 minutes and allowed to stand 10 minutes.
Thus Chart 1, constructed from individual tests, pictures the relative per-
centages of soluble solids, pectin and malic acid obtained in successive ex-
tracts under a constant set of conditions, i. e., temperature of extraction,
ratio of apple to extraction water and acidity, while only the extraction
period was varied. In this manner it was possible to determine the influence
of these several factors upon yield and composition of the juice obtained
from each successive extraction.
The limits in percentage for soluble solids recovered in the first, second,
third extracts and remaining in the pulp were respectively 33.0 — 66.8, 17.0 —
31.6, 4.4—19.7, and 3.3—22.9. For pectin, these limits were 21.5—64.0, 10.8—
33.2, 6.0—22.2, and 7.9 — 43.9; and for acid calculated to malic they were
34.9—82.3, 6.1—43.1, 4.5—26.2, and 2.0—31.2.
The relative lack of variation of solids or pectin in the second and third
extracts regardless of the treatment is particularly emphasized in Charts
1-5. In fact one of the outstanding results of this investigation was that
juice could be extracted successfully from apples by almost any method,
though with unlike results. It is fortunate indeed tiiat this is true, because
every conceivable method of juice extraction is practiced in the home and
factory. The wide limits within which a degree of success may be attained in
preparing apple jellies, for example, makes total failure difficult, though
improvement in yield and quality always can be secured by the use of
correct methods.
ATPLE JUICE EXTRACTION FOR JELLY
221)
Chart 2. Effect of Extraction Temperature upon Yield of Soluble Solids and Pectin
in First and Second Extracts of Sliced and Chopped Apples. Ratio of
Apple to Water 3:3.
100
90
70
60
W Mo
Fin^T EXTRACTION
^ecoMD EXTRACTION
Sliced Chopped Sliced Chopped
SOLUBLE SOLIDS PECTIN
What has Just been said relative to Baldwin apples normally held true
for Red Astraohan, Red Siberian Crab, Wealthy, Mcintosh, King David,
W^inesap, and Rhode Island Greening. Of course there were differences in
ease of extraction. For example, Red Astrachan, King David, and Wealthy
gave up their soluble solids, including pectin, more readily than most of the
other varieties. A relatively high acidity undoubtedly aided in breaking up
the cell walls and liberating the cell contents of the fruit. Referring to
Tables 7 and 8, it is evident that the cellular structure of certain varieties
was more easily broken down by heat extraction with water than that of
others. The Greening, Winesap, and Baldwin were among the varieties most
difficult to extract with water, while Mcintosh and Red Siberian Crab occu-
pied an intermediate position in this respect. At any rate it appeared that,
for all the varieties tested, two successive 1.5-minute extractions of the sliced
apples at the boiling point removed from 58 to 74 per cent of the pectin and
73 to 93 per cent of the soluble solids. For chopped fruit, the recovery was
slightly increased, but at the expense of clearness of the juice. The per-
centage of soluble solids remaining in the pulp after three extractions varied
from 2.2 in Wealthy to 14-.3 in Mcintosh, with a general average of 6.0 for
all varieties.
Comparison of Sliced and Chopped Apples for Juice and Jelly.
Influence of Temperature and Ratio of Fruit to Extraction Water. Com-
parative data are presented in Tables 7 and 8 and in Charts 2 and 3. In
general the chopped apples gave a slight increase in yield of soluble solids,
pectin, and malic acid over sliced fruit. On the other hand the expressed
juice of chopped apples was always more cloudy and turbid than that from
sliced fruit. This was true in the case of the first extract alone or of 2 or 3
successive extracts combined, though this increased recovery seldom amounted
230
TECHNICAL BULLETIN 15
Chart 3. Influence of Ratio of Apple to Water on Recovery (Yield) of Soluble Solids
and Pectin in Sliced and Chopped Baldwin Apples. Temperature of Extrac-
tion 100° C.
90
EZ2 TmsT EXraACTloK
CZl .SECOttD EXTRACTION
Sliced Chopped
SOLUBLE SOLIDS
* Ratio of apple to water by weight.
Sliced Chopped
PECTIN
to as much as 10 per cent of the total and averaged about 5 per cent. It
was greatest where the extraction temperature was low, i.e., 88° C. (190° F.) ;
it was negligible at the boiling temperature and finally became negative at
109°C. (228° F.). The probable cause of the decreased recovery of sugar,
acid and pectin at 109 °C. was the poor heat conductivity of the chopped
apple mass. This resulted in lower temperatures in the mass itself t!ian was
indicated by the retort thermometer. This was experimentally proved. It
took more than 30 minutes for a retort temperature of 109°C. to bring the
whole of the apple pulp to that temperature. For this reason, in all pressure
cooker extractions under 30 minutes, the temperature of the chopped pulp
mass was below 109°C. This readily explains the poor recoveries obtained
at this seemingly high extraction temperature. On the other hand in the
pressure cooker, sliced apples allowed ready penetration of heat, largely
due to unimpeded convection currents and increased conductivity in the
ai)ple — water medium. Where the amount of water used for extraction
was large, as in the 3:4 ratio, there was an increase in the time necessary
to bring the mass to the desired temperature. After reaching this point,
however, the temperature of the mass was relatively constant, due. to the
large volume of liquid present.
Where the ratio of fruit to water was 3:2 the chopping of the fruit had
a slight inimical effect upon yield of soluble solids, acid, and pectin. Where
the ratio was 3:3 or 3:4 the chopped fruit usually yielded considerably more
of these substances than the sliced. Thus it appeared that where the apples
were ground up or chopped, tlie ratio of fruit to extraction water siiould be
widened over that necessary where the fruit was sliced. For sliced fruit, a
Al'l'I.K .lUICK KXTHACTION FOR .IKI.I.V J.n
r;iti() of ;3:2 was foiiiui to l)c vt-ry suilal>lc. Ilioii.-li for cliopix-il fruit tlic
ratio which gave most satisfactory results was 3:3. The ratio 3:i was un-
economical because of tlie huge fuel consumption necessary to evaporate the
large volume of juice. This objection was not recompensed by a significantly
greater yield of solids or pectin in the juice. Furthermore, the juice was
inferior in quality for jelly making or pectin manufacturing purposes.
In Table 10 where a 15-minute extraction period at 100° C. was u.sed on
eight common ai)iile varieties, the gain in soluble solids of a ratio of fruit
to extraction water of 3:3 over 3:2 was 5 per cent. For pectin the differ-
ence was still less. Furthermore if two or even three successive extractions
were made using ratios of 3:2, 3:3 and 3:4, the yields of total soluble solids,
pectin, and acid tended to become equalized, though the greater the amount
cif water used, the higher the recovery of extractives became. In a 30-
iiiinute extraction period (Table 11) the difference in recovery of extractives
between these ratios was still further minhuized. All eight varieties reacted
similarly in this respect.
Comparing the several varieties it is seen that some yielded up their
soluble solids including pectin much more readily than others. Besides
Tables 7 and 8, additional data bearing upon this point may be found in
Tables 10 and 11. For example. Red Astrachan and Wealthy both yielded
nearly 70 per cent soluble solids and over 50 per cent pectin in a single 15-
minute extraction. King David, Mcintosh and Winesap varieties held these
substances more tenaciously in the fruit tissues. In most cases the sliced
fruit when suV)jected to 15 minutes at 100°C. liberated slightly less soluble
solids including pectin and malic acid than the chopped fruit. The Red
Astrachan variety proved to be an exception, while in Greening and AVine-
sap, but minor differences were noted between the two methods. The ex-
tracted juice from all varieties gave high grade jelh^, though in every case
that obtained from the chopped fruit was less clear and therefore of poorer
quality than that made from sliced fruit. Red Siberian Crab, King David,
Red Astrachan, Winesap and Mcintosh were all considered to be first class
jelly varieties. Baldwin and Wealthy are inferior to these while Greening
yielded an unattractive light colored jelly. For this very reason the latter
served as an excellent base for mint jelly.
Other things being equal, the optimum ratio of fruit to water is the one
that yields the highest concentration of soluble solids, including acid and
jellifying pectin, per unit volume of liquid. However, it should be noted
at this time that enough extraction water nmst be used to prevent scorching,
also that if too little liquid is present, especially where chopped fruit is used,
satisfactory filtration or separation of the juice from the pulp, is difficult.
The resulting juice may also be more cloudy and thus lower the jelly (juality.
Efect of Temperature Upon the Recovery of Soluble Solids,
Pectin and Acid in Heat Extracted Apple Juice.
In general there was a consistent increase in percentage recovery of soluble
solids, pectin and acid with rise in temperature from 88° to 109°C., though
the increase was more marked in passing from 88° to 100°C. than from the
latter temperature to 109°C. (See Tables 7, 8 and 9). Results were much
more consistent with sliced than with chopped apples, especially at the maxi-
mum extraction temperature of 109 °C. In the first extract using sliced
fruit, the average increase in yield of both soluble solids and pectin at 100°C.
varied from 4 to 13 per cent over the amounts extracted at 88° C. Smaller
232
TECHNICAL BULLETIN 15
increases were noted where chopped fruit was used. Likewise greater yields
were obtained at 100° than at 88° C. if the sum of either two or three suc-
cessive extracts is considered. For two extracts combined an average maxi-
mum increase of 15 per cent was noted, whereas for tliree extracts, the maxi-
mum was only 7 per cent. Thougii some beneficial effects were gained by
extracting at 109°C. as compared with 100°, still these were too slight to be
of importance when the difficulties involved in cooking the apple — water mix
under steam pressure, are considered. In fact for chopped apples, unless
the time of extraction was at least 30 minutes, there was usually no gain in
recovery of solids or pectin over the extraction carried on at 100° C.
The evidence indicates that any temperature above 88° C. is efficacious in
bringing into solution the sugars, acid, and pectin of apples, though (jn
account of the cost and difficulty of cooking under pressure it is probably
not desirable to exceed the boiling temperature. For ordinary purposes
100° C. appears to be the most satisfactory extraction temperature though
somewhat lower temperatures do not greatly influence the yield of soluble
solids, acid or pectin in Baldwin apples. Attention is again called to the
greater cloudiness of aU juice samples produced from chopped fruit. This
makes for a poorer quality jelly or pectin extract and is distinctly objec-
tionable.
Influence of Extraction Period upon the Composition of Apple Juice.
Baldwin Apples. Data bearing upon this question have been assembled
in Tables 9, 10 and 11 and Charts 1 and 4.
Chart 4 gives a clear average picture of the influence of period of extrac-
tion upon soluble solids and pectin recovered in the juice. For the removal
of soluble solids and pectin in one or more extractions the 15-minute cook
was practically as efficient as where, in addition to the regular extraction
time, a standing period of 10 minutes removed from the source of heat, was
Chart 4. Influence of the Length of the Extraction Period uison Recovery of Soluble
Solids and Pectin in Sliced Baldwin Apples at 100° and 109° C.
>-70
pi
cc
%^
o
'^ 50
Soluble Solids Pectin
100°C.
Soluble Solids Pectin
109°C.
*Boiled 15 minutes and allowed to stand 10 minutes.
Al'PLK JUICE EXTRACTION FOR JEI.LY 233
allowed. A sliylit gain in aiiiouiit of pectin extracted was found, but tliis
was of little significance particularly wlien two or more extractions were
considered.
It is interesting to note tliat one 30-niinute extraction was not ecjuivaleiit
by any Ineans to two IS-niinute extractions. At all tiiree temperatures used,
there was an average increase of less than 10 per cent in sohil)le solids,
pectin, or acid in the 30-minute as compared to the single lo-minute extrac-
tions. Two successive 15-niinute extractions yielded from 50 to 90 per cent
more soluble solids, pectin and malic acid than a single 30-minute extraction.
Of course other considerations enter here such as a greater volume of liquid
to evaporate when two 15-niinute extractions are made instead of one 30-
minute period, doubled labor in handling the two extracts and a small loss
of time. The cost of fuel is practically the same and, all things considered,
the data indicate that two short extractions were much more efficient and
economical than a single long one.
Although a slight gain in both total solids and pectin was observed in the
60-minute extract, it was far too long to be practical. Furthermore the in-
creased amounts of solids obtained did not repay for the huge fuel consump-
tion, time, and loss in jelly quality. This latter point is very important and
in almost every case where apples were extracted for as long as 60 miimtes,
a serious loss in jelly quality was observed. The long-continued boiling
hydrolyzed the pectin and reduced its activity. Usually apples could be
boiled 15 minutes without great apparent injury to the pectin or jelly qual-
ity though after 30 minutes boiling this was very apparent. A decrease in
both jelly yield and jelly strength in juices boiled for 30 minutes showed
that even in that short heating period, some pectin decomposition occurred.
Other Varieties. As regards apple varieties other than Baldwins, the same
conclusions hold. A comparison of the data in Tables 10 and 11 (15-and 30-
minute extractions respectively) shows very little difference in the amounts of
soluble solids obtained. Pectin recovery was slightly greater in the 30-min-
ute extraction period, though the increase was not significant. Since con-
siderably more evaporation loss occurred during a 30-minute extraction period
than during a 15-minute period, the volume in the former case was consid-
erably less. It was also more difficult to separate the liquid from the pulp
where the amount of liquid was small. There was a tendency toward mushi-
ness as the period of extraction increased and this physical condition of the
fruit pulp interfered somewhat with the removal of juice from it, that is the
pulp retained a higher percentage of the extractives. The data indicate quite
clearly that under the conditions of the experiment two 15-minute extrac-
tion periods with the ratio of fruit to water 3:2 to 3:3 were best for juice
extraction from the common varieties of the apple. It is possible that a
shorter period than this is desirable, but none was studied in this experiment.
By this procedure from 65 to 85 per cent of the total soluble solids, accord-
ing to the ratio of apple to water used and the temperature, and from 50 to
70 per cent of the pectin present in the apple can l)e utilized. Under the
average conditions of jelly manufacture a third extraction would be un-
economical.
Influence of Added Acid upon the Composition of the Juice.
Tartaric acid was added directly to the sliced or chopped api)les in quan-
tities sufficient to produce theoretically acidities of 0.2, 0.4, and 0.8 per cent
tartaric acid in the apple-water mixture. This was in addition to the natural
234.
TECHNICAL BULLETIN 15
malic acid constantlj- present in the fruit. No ;icid was added after the first
extraction had been made.
Table 12 and Chart 5 show the effect upon soluble solids and pectin yield
of extracting in the presence of increased acidity. In general a slight but
uniform increase in the amounts of soluble solids, including jjectin and malic
acid, was obtained wtih increased acidity. This increase was more marked
in the case of sliced than of chopped fruit. An acidity of 0.4 per cent tar-
taric acid was practically as efficient as double that amount. In other words
beyond a certain point little or no advantage accrued from increasing the
Chart 5. Effect of Added Acidity upon the Kecovery of Soluble Solids and Pectin in
First and Second Extracts of Sliced and Chopped Apples. Ratio of Apple
to Water 3:3.
a 50
140 -
I 50
20
10
0
FIRST EXTBACTIOK
SE.C0HO E,XTHACT10K
Sliced Chopped
SOLUBLE SOLIDS
Sliced
Chopped
PECTIN
*Per cent added aciidity.
acidity. There was but little effect upon the second or third extracts; the
increased yields of solids and pectin being much more pronounced in the first
extract. In l)ut few cases did the addition of tartaric acid in order to in-
crease the natural acidity of the fruit, produce more than a 10 per cent in-
crease in yield of solids or pectin even in the first extract. Pectin extraction
was aided somewhat more by the acid than was the extraction of soluble
solids; in neither case was the increase appreciable, especially if two extrac-
tions were considered.
It is questionable whether the addition of organic acids to apples yielded
a sufficiently richer juice to pay for the trouble. The resulting jelly was
improved in quality, yield, color, and taste by the acid, though of course if
sold, a declaration of added acid would be required by State and Federal
food laws. However, addition of acid may be advantageous on other grounds
inasmuch as Tarr (17) has demonstrated that the hydrogen ion concentration
must be at least 3.46 for jelly formation using pure pectin, with the optimum
from 2.9 to 3.0. It is likewise true that added acid in many cases appreci-
ably increased jelly yields and improved its physical properties and flavor.
For some non-acid varieties of apple a distinct improvement in jelly flavor
was effected 1)V the addition of a small amount of acid.
Arri.E .TUICE EXTRACTION' FOR .li:i.l.^- -.'.jj
Seasonal Variation of Baldwin Apples
Data were collected for the three years, 1925, li»'2(), and li)27, on llu- ilu-in-
ic-al composition of the fruit as well as the heat-extracted juiie. Tahle 1
shows no striking changes in composition from year to year. Similarly Tahic
13 presents only minor differences in the heat-extracted juices from ajipies
harvested during three successive years. Despite varying ratios of fruit to
extraction water, tiie range of soluble solids obtained by a single 15-minute
extraction was narrow, 47.9 to 59.6, and for the second extraction 22.2 to 27.7.
Similarly, the amounts of extracted pectin and also malic acid varied l)ut
little. In general, both jelly yields and jelly strengths were higher in the 3:2
ratio than in the 3:3 or 3:4 ratios.
STUDIES OF JELLY
Jelly Yields per Unit Weight of Apples
In Table 14 is summarized the average jelly yields from two extractions at
100°C. for the 8 varieties already mentioned. Additional data also are found
in Tables 10, 11 and 13. In interpreting these data, the method of making
the jelly must be kept in mind. That is, to 18 ounces (511 grams) of juice
composed of proportionate parts of the first and second extracts, was added
sufficient sugar in addition to that already present in the juice, to make 12
ounces (341 grams). The juice was then rapidly concentrated until the jelly
sheeted or, in the absence of a satisfactory jelly test, until the sugar content
reached 69 to 70 per cent. In general jellies containing over 70 or under 64
per cent of sugar were abnormally soft in consistency or were otherwise of
poor qualitj'.
Had the sugar been proportioned exactly to the pectin actually present,
the increase in jelly yields in columns 3, 8 and 13, Table 14, would not be as
marked, though the general trend would have been the same. This may be
.shown by a simple calculation using the pectin content of the juice or jelly
as tabulated respectively in Tables 2 and 3. The pectin content of jellies
made from a 3:2 ratio of fruit to extraction water was higher than where
the ratio was 3:3 or 3:4.
The effect of pectin content of the juice upon jelly strength is clearly pre-
sented in Table 14. Other things being equal, for short extractions the jelly
strength increased as the pectin content of the jelly increased. Most of the
jellies made from the 3:2 ratio of fruit to extraction water were tough, i. e.,
over 100 Bloom grams, while those made from a 3:3 ratio were of medium
firnmess, and those from the 3:4 ratio were considerably softer in consist-
ency. The grade of jelly depended upon several factors such as flavor, color,
texture, consistency, stickiness, syneresis and ability to withstand storage.
Two 15-minute extraction periods using equal parts of fruit and water, gave
high yields of juice containing sugar, acid and jellifying-pectin in suitable
amounts to produce in turn high yields of well flavored, higli grade jelly.
Normally, this procedure gave best results. In some cases it was possibly
more economical to use a ratio of apple to extraction water of 3:2 but the
danger of scorching, difficulty of separating the juice from the pulp and the
lower yields of solids, pectin and jelly usually outweighed the advantage
gained by having only a small amount of relatively concentrated juice. If
the sugar was properly proportioned to the amount of jellifying i)ectin
236 TECHNICAL BULLETIN 13
present in the juice and with due regard to tlie hydrogen ion concentration
of the juice, optimum yield and quality of the jelly resulted and there was
no great variation in either jelly yield or quality. The 3:4 ratio of fruit to
extraction water, though giving maximum yields of juice, solids, and jelly
was found to be impracticable, because of the dilution of the resultant juice,
the greater fuel consumption in concentration, and the loss of jelly quality.
Acidity of Apple Jellies
The total titratable acidity as well as hydrogen ion concentration was de-
termined on all jellies. These data are summarized in Table 6. The total
acidity calculated as malic acid varied from 0.22 per cent in Mcintosh to
0.66 per cent in Red Siberian Crab. The pH value varied between 3.62 for
the 1925 crop of Baldwins and 3.20 for Red Astrachans. Inasmuch as jelly
yields were dependent to some degree on the pH of the juice, the lower this
value, the better was the yield of jelly. Most people prefer an acid or sub-
acid jelly, hence those varieties possessing high total acidity and hydrogen
ion concentration are to be preferred for use in jelly making.
Sugar Content of Jellies.
The sugar content of all jellies was determined by the Abbe refractometer.
On 101 samples of Baldwin jellies the sugar ranged from 61 to 72 per cent
with an average of 65 per cent for jellies classed as marketaljle, i. e.. Grade
1 or 2. The soft jellies very often contained over 70 per cent of sugar and
crystallized badly during storage. The sugar content of high grade jellies
made from Red Astrachan, Crab Apple, Rhode Island Greening, Winesap and
King David averaged 67.5 per cent. Mcintosh and Wealthy gave jellies
with a sugar content of approximately 65.5 per cent.
Most directions or recipes for making jellies proportion the amount of
added sugar to the juice. This is satisfactory only when extractions are made
in exactly the same way, i. e., time and temperature of extraction, number
of successive extractions, ratio of fruit to water and still other factors must
remain nearly constant. Othewise there is danger of using too little or too
much sugar for maximum yields of good quality jelly.
It is far better to proportion the added sugar to the weight of fruit origi-
nally taken, viz., pound for pound or some other ratio as experience may
indicate. This basis is much more scientific and makes for a better product.
Each apple variety differs from the other in suitability for jelly manufacture
(Tables 10 and 11.) For example, Red Astrachan, King David and Red
Siberian Crab produce highly colored and well flavored jellies. Other va-
rieties such as Greening give very light colored jellies; still others are low
in acidity, or give jellies which are cloudy or of poor consistency. The va-
riety is important if a product possessing certain known qualities is desired.
On the other hand nearly all varieties will yield a fair amount of good edible
jelly.
Maturity is important, because in apples past their prime, the pectin un-
dergoes deterioration and produces not only a low yield but a poor quality
jelly as well. Seasonal differences in varieties (Table 13) do not appear to
be particularly significant.
By consulting Table 14, it is possible to estimate the yield of apple jelly
that may be obtained from a unit weight of apples providing the same manu-
facturing procedure is followed. Since approximately 67.5 per cent of the
I
APPLE JUICE EXTRACTION FOR .IKI.LY 2:U
weight of a finished apple jolly is sugar, and the weight of roncentraled juice
and the percentage of solul)le solids in it are known, the finishing point of
the jelly is readily determined by the simple ex])edient of weighing the sugar-
juice mixture from time to tune until the calculated yield of jelly is reached.
For <;xample in Table 14, it is shown that an average of 20 ounces of jelly
per pound of apples may be expected where the ratio of fruit to extraction
water is 3 to 3 by weight and the combined first and second I5-niinute ex-
tractions are utilized. Then if 3 pounds of fruit are taken, the yield of
sugar in 90 ounces of unconcentrated juice containing 5 per cent soluble solids
(Table 2) is approximately 90X-05 or 4.5 ounces of sugar in the juice. This
figure is not absolutely correct because only about 76.5 per cent of the solids
in extracted apple juice is sugar (Table 5), but this correction is unnecessary
for ordinary work. Since 3 pounds of apples are used, 3X20 or 60 ounces of
jelly should result. This jelly nmst contain 60X0.675 or 40.5 ounces of sugar.
Therefore 40.5 — 4.5 or 36 ounces of sugar must be added to the juice which
is then concentrated until the weight of jelly in the pan is exactly 60 ounces.
This is the finishing point of the jelly. The writer has used this method of
preparing samples of experimental jellies and jams in numerous cases. In
general, the results obtained by the procedure just outlined, usually approxi-
mate those obtained by the sheeting test as made by an experienced jelly
maker.
The temperature test was useful but could not be relied upon in all cases
if used by itself. The refractometer test was found to be of particular value
in following the evaporation of water in a jelly mixture. The results checked
those obtained by occasional weighings.
For a discussion of the role of sugar, acids, and salts in pectin jellies ref-
erence shoidd be made to the series of bulletins issued by Tarr, Myers and
Baker of the Delaware Agricultural Experiment Station. In many cases
their results are applicable as well, to fruit jellies. These fundamental con-
tributions on the chemical and physical factors influencing jelly formation
are very important and, when properly utilized by jelly manufacturers, should
enable them to produce economically standardized products of high quality.
ACKNOWLEDGMENT
The writer is deeply indebted to I'rofessor AV. W. Chenoweth, who original-
ly proposed this investigation, for his sustained interest in the project and
for many timely suggestions freely given during its entire progress, and to
Mr. Francis P. GritSths who assisted in the final phases of the laboratory
work.
SUMMARY AND CONCLUSIONS
Baldwin apples showed but slight yearly variation in composition of fruit,
juice or yield of jelly. Red Siberian Crab, King David, Red Astrachan,
Winesap and Mcintosh were the most suitable varieties for jelly manufacture.
A study of juice extraction by heat showed:
1. Two short (15 minute) successive extractions were usually desirable to
obtain an optimum yield of juice containing sufficient pectin and acid to
give satisfactory yields of high quality jelly.
2. When only one extraction of the fruit was made, there was a serious
loss in jelly yield. This was found to be due largely to the difficulty of ex-
tracting pectin.
238 TECHNICAL BULLETIN 15
3. Long extraction periods were unsatisfactory because of destruction of
the jellifying power of tlie pectin. Jelly yields and quality were injured
materially. For example, two 15-minute extractions removed from 50 to 80
per cent more soluble solids, pectin and acid than a single 30-minute ex-
traction.
4. Although fair yields of solids, pectin and acid were obtained at an
extraction temperature of 88° C. (190° F.), the optimum was found to be
100°C. (212°F.). Retorts or pressure cookers at 109°C. (228°F.) gave only
slightly higher yields of solids and pectin in the juice than were obtained at
the boiling point, and their use is not recommended for juice extraction.
5. The best ratio of fruit to water was 3:2 in the case of sliced apples, or
3:3 where chopped or grated apples were used. The yield of jelly per pound
of fruit was greater where the ratio was 3:i and least when it was 3:2.
6. Tartaric acid added to the apple-water mixture in concentrations of
0.2 to 0.4 per cent slightly increased the yield of solids and pectin in the
extract as well as the total jelly yield. Added acid always improved the
color and often the flavor and consistency of apple jellies.
7. Finely chopped apples gave a slightly more concentrated juice and
greater jelly yield than sliced apples, but because of the difficulty of filtering
and the cloudiness of the finished jellies, chopping is not recommended.
8. In short extraction periods the jelly strength of the jellies increased
with the percentage of pectin present. In longer extractions where the
jellifying power of the pectin had been injured by prolonged heat, this rela-
tion did not hold.
Apples suitable for jelly should yield approximately 20 ounces of jelly per
pound of fruit.
Jelly yields depended primarily upon the amount of jellifying pectin pres-
ent in the juice, and the amount of sugar used. Less sugar should be used
where pectin is present only in limited amounts or where its jellifying power
has been injured by hydrolysis.
In general those jellies containing the most pectin were very firm and
tough in consistency unless additional sugar was used. The hydrogen ion
concentration of the varieties tested was suitable for the formation of good
yields of well flavored jellies.
There was a fair degree of correlation between the total titratable acidity
and hydrogen ion concentration in apples, apple juice and jelly. That is, high
total acidity and high hydrogen ion concentration were present in the same
samples of apples, juices or jellies.
The sugar content of apple jellies ranged from 65 to 70 per cent with an
average of 67.5 per cent. Jellies containing less than 65 per cent sugar were
often tough while those with over 70 per cent were uniformly soft or syrupy.
In making jellies from fruit, it was found much more desirable to proportion
the sugar to the original weight of fruit than to an vmcertain yield of juice of
questionable composition.
The determination of jelly strength by means of the Bloom gelometer gave
very concordant results.
Brix hydrometer readings on extracted apple juice gave an average of 0.18
per cent higher than by the Abbe refractometer and 0.15 per cent higher
than by the gravimetric determination of solids in solution. The mean sugar
content of the soluble solids in heat extracted apple juice was 76.5 per cent.
I
Al'l'I.I-: JUICE KXTUALTIOX I'OU .IKIJ.V i>:}!>
LITERATURE CITED
1. Alwood, W. B., Davidson, R. J., and Moncure, W. A. 1'. l!)()l-. Tlie
clieniical ..'oniposition of apples and cider. U. S. Dept. Agr. Bur. of
Cheni. Bui. 88. 46 p.
2. Association of OtHcial Agricultural Chemists. 192.5. Methods of Anal-
ysis. "Washington, D. C.
3. ]5aker, G. L. 192G. Jelly strength of pectin jells. Jour. Ind. Eng.
Chein. 18, 89-93.
i. Bigelow, W. D., Gore, H. C and Howard, B. J. 1905. Studies on apples.
U. S. Dept. Agr. Bur. of Cheni. Bui. 94, 100 p., 5 pi.
5. Fellers, C. R. and Griffiths, F. P. 1927. Unpublished data.
6. Fellers, C. R. and Griffiths, F. P. 1928. Jelly strength measurements of
fruit jellies by the Bloom gelometer. Jour. Ind. Eng. Chem. 20: — .
7. Mehlitz, A. 1926. The alteration of pectin during the cooking process.
Cheniie Zelle und Gewebe 12, 3.53-361.
8. Myers, P. B. and Baker, G. L. 192.5. Fruit jellies. Del. Agr. Expt.
Sta. Bui. 141 (Annual Report of the Director), 14-19.
9. Myers, P. B. and Baker, G. L. 1927. Fruit jellies. V, The role of pectin.
1. The viscosity and jellifying properties of pectin solutions. Del.
Agr. Expt. Sta. Bui. 149, 46 p.
10. Nanji, D. R. and Norman, A. G. 1926. Pectin: Micro-method for the
determination of methanol and its application to the study of the con-
ditions governing the de-esterification of pectinogen. Jour. Soc. Chem.
Ind. 4.5, 337-40.
11. Nelson, E. K. 1926. Pectic acids. Jour. Amer. Chem. Soc. 48, 2412-2414.
12. Paine, H. S. 1922. Use of pectin in jams and jellies. Amer. Food Jour.
17, No. 3, 11-13.
13. Richardson, W. D. 1923. A new instrument for testing glue and pectin
jellies. Chem. & Met. Eng. 28, No. 12, 3-4.
14. Shaw, J. K., 1911. Climatic adaptations of apple varieties. Mass. Agr.
Expt. Sta. Ann. Rep. 23, Part 1, 177-245.
1.5. Sherwood, S. F. 1928. Use of the refractometer in the analysis of in-
dividual sugar beets. Jour. Agr. Res. 36, 41-52.
16. Sucharipa, R. 1925. Die Pektinstoffe. 188 p. Serger & Hempel. Braun-
schweig, Germany.
17. Tarr, L. W. 1923. Fruit jellies. I, The role of acids. Del. Agr. Expt.
Sta. Bui. 134. 38 p.
18. Tarr, L. W. 1926. Fruit jellies. Ill, Jelly strength measurements. Del
Agr. Expt. Sta. Bui. 142. 33 p.
19. Wlchmann, H. J. 1926. Personal communication.
<
a
Z o
W c
t 2
o
a
s:
r^
Cl
ro
lO rt cq 00 CD
00 CD
"^
:^
C-)
CO
CO CO CO (M C>)
CO CO
<
■Ci.
d
d
d
d d d d d
d d
>'■-
^
•- "5
^
o
^H
CO
^ rt t^ 00 t^
-. o
2 E
'5'
d
d
d
-H no ■* CO ■<»<
-h" d d d d
OS 00
d d
< S
■^
o
C^l
t^
00 t^ "O o o
lO o
"S,
L*^
^
■^
-; CO 'i; CD -^^
CO CO
n
M
CO
CO co" co' co' co"
CO co"
K !2
rt *u
c
c "
<w
o
N
t^
X Tj< rt t» o
t^ CO
"5 '^
■^
c^
CO
CO
N CO CO IM CO
•«< CO
o "5
X.
d
d
d
d d d d d
d d
■<A.
K
00 o
s
^
o
k.
*-^
'1 ^. i: o o
ij ^
w
•^
d
d
o o
d
^
K
o
"^
o
CO
-H o O CO o
CO 00
3
■^^
lO
r-;
Tf
M (N •* 05 t^
o t^
i..
•*'
m
TO
CO CO CO ci -H
r~i r^
w
^
be
K
IN
M
00
CJ! O CO CO O
00 t~
3 bO
u
CI
•C
Ol 00 O — 03
CD •«<
t
b-'
;c
t^
lo t^ d d 00
00 00
ci
■c.
s
C^
M
^H
O O ri 02 o
O O
u
t-;
05
q
(N q >0 p CD
CD CO
£ =
h.
«-4
d
^
d d CO c^i d
d d
H to
3 -^
K
u
c
(M
^ -H 05 00 CO
OO CO
IC
^^
o
CD ro t~ CO t~
cq r^
fe
IC
Tt"
CO
co' IM' lO CO 'h"
CO co'
w ""
■C
5 •§
£:
-X
o
o
-H O -^ O CO
00 r~
^j
o
t~
o
CO r-; CD r>; 00
03 rt
(21
j__
00
d
lO
lO "5 00 "5 CO
Tf to
■^
*tS 0)
° -S
. p.
M
(N
M
rH CO CO CO rt
r-1 CO
S E
Z 1
>-!
o
CD
^^
CD CD CD CO CD
CO CO
rt
C-1
?)
N
CO C) C^l CO CO
CO CO
o
o:
05
C: C3 O 03 c:3
03 en
/3
>.
a!
C bQ
u
>
c
'%
< ^. ^ 1 5
11
"1
2
2
2
n
B5
rt
OQ
« « ^ s ^
i4 «
APPLE JUICE EXTRACTION FOR JELLY
241
TABLE 2 — Yield and Composition of Juice per Unit Weight of Apples.
Combined First and Second Extracts at 100° C, 8 varieties.
Number
of
YIELD OF JUICE
COMPOSITION OF JUICE
Extraction
period
Per
Per
Soluble
Pectin
Pectin
Acidity
tests
lb. of
kg. of
solids
as pectic
per kg.
as
fruit
fruit
acid
of fruit
malic
Ratio of fruit to water 3 : 2
min.
OS.
grams
per cent
per cent
grams
per cent
15
20
21.6
1353
7.43
0.187
2.53
0.290
15 + 10*
5
21.1
1322
7.71
0.193
2.55
0.345
30
12
19.1
1196
8.26
0.220
2.63
0.368
60
5
18.0
1127
8.60
0.227
2.55
0.421
Ratio of fruit to water 3 : 3
15
20
30.7
1922
5.70
0.144
2.77
0.212
15 + 10*
5
31.8
1992
5.67
0.143
2.85
0.235
30
12
27.2
1704
6.08
0.175
2.98
0.250
60
5
28.1
1760
6.55
0.177
3.12
0.300
Ratio of fruit to water 3 : 4
15
20
41.7
2611
4.49
0.129
3.37
0.155
15 + 10*
5
40.2
2519
4.67
0.131
3.30
0.187
30
12
35.4
2218
5.05
0.152
3.37
0.175
60
5
35.5
2550
5.27
0.163
4.15
0.220
* 10 minute standing period.
245
TECHNICAL BULLETIN 15
TABLE 3 — Composition of Successive Extracts of Apple Juice.
Extraction Period, 15 minutes at 100° C, 8 varieties.
3 :
2*
3
: 3*
3
: 4*
Number
of
Extract
tests
Soluble
Pectin
Soluble
Pectin
Soluble
Pectin
solids
as pectic
acid
solids
as pectic
acid
solids
as pectic
acid
per cent
per cent
per cent
per cent
per cent
per cent
First extract
27
8.2
0.226
6.7
0.161
5.7
0.150
Second extract
27
5.3
0.170
3.7
0.122
2.9
0.111
Third extract
27
3.2
0.109
2.0
0.084
1.5
0.065
Residual pulp
24
3.1
0.240
1.9
0.195
1.7
0.170
* Ratio of fruit to water.
TABLE 4— Yield of Soluble Solids, Pectin and Acid per Unit Weight of
Apples. Combined First, Second and Third Extracts plus the
Residual Pulp. '
Number
of
tests
SOLUBLE SOLIDS
PECTIN AS
PECTIC ACID
ACID AS MALIC
Variety
Per
Per
Per
Per
Per
Per
lb. of
kg. of
lb. of
kg. of
lb. of
kg. of
fruit
fruit
fruit
fruit
fruit
fruit
o:.
grams
02.
grams
02.
grams
Baldwin
91
2.15
134.7*
0.077
4.84*
0.064
4.01*
Mcintosh
7
2.10
131.2
0.065
4.06
0.034
2.12
King David
7
2.36
148.0
0.064
4.06
0.106
6.64
Red Astrachan
7
1.41
87.9
0.065
4.08
0.115
7.20
Red Siberian Crab
S
2.42
151.6
0.070
4.3S
0.143
8.94
Wealthy
7
1.47
92.0
0.057
3.59
0.078
4.87
Rhode Island Greening
7
1.91
119.7
0.070
4.38
0.087
5.42
VVinesap
7
2.41
148.5
0.083
5.18
0.072
4.52
* Soluble solids 134.7 ± 9.65; pectin 4.84 ± 0.50; acid 4.01 ± 0.57.
APPLE JUICE EXTRACTION FOR JELLY
248
TABLE 5 — Relation of Soluble Solids and Sugar in Heat Extracted Apple
Juice as Determined by Abbe Refractoineter, Brix Hydrometer, and
Gravimetrically.
Brix
Computed
Direct reading
of sucrose —
Soluble
solids —
Total sugar as
invert —
degrees
from R. I.
refractometer
gravimetric
determination
gravimetric
method
6.2
.5.67
6.30
5.71
4.10
3.7
3.48
4.10
3.39
2.49
l.S
1.44
2.00
1.51
0.98
6.7
6.45
6.90
6.65
4.96
3.3
3.40
3.75
3.42
2.01
l.S
1.38
2.00
1.41
0.97
5.1
5.05
.5.20
5.09
4.05
7.9
7.90
S.OO
8.10
6.76
.5.0
5.00
4.61
4.71
3.81
10.8
10.65
10.90
10.71
8.93
Mean difference between Brix and refractometer determinations +0.176
Mean difference between Brix and gravimetric determinations +0.146
Mean difference between refractometer and gravimetric determinations — 0.030
Mean difference between sucrose scale reading on refractometer and total sugars (chemical) + 1.470
Mean sugar content of total solids in heat extracted apple juice 76.5%
TABLE 6 — Titratable Acidity and Hydrogen Ion Concentration of
Extracted Apple Juice and Jelly. Extraction Period 15 minutes at 100° C,
with Ratio of Apple to Water 3 : 3.
Variety
Season
First
extract
Second
extract
Third
extract
Jelly from com-
bined first and
second extract
pH
Acidity
pH
Acidity
pH
Acidity
pH
Acidity
as malic
as malic
as malic
as malic
Baldwin
1925
3.55
0.198
3.65
0.100
3.75
0.051
3.62
0.307
Baldwin
1926
3.47
0.201
3.60
0.110
3.67
0.047
3.57
0.310
Baldwin
1927
3.49
0.201
3.57
0.111
3.63
0.045
3.53
0.321
Mcintosh
1926
3.64
0.117
3.71
0.068
3.82
0.039
3.55
0.220
Winesap
1926
3.54
0.265
3.57
0.134
3.59
0.560
3.51
0.295
King David
1926
3.33
0.436
3.36
0.238
3.45
0.105
3.40
0.542
R. I. Greening
1926
3.38
0.312
3.40
0.183
3.55
0.077
3.39
0.383
Red Astrachan
1926
3.15
0.430
3.30
0.240
3.47
0.139
3.20
0.640
Wealthy
1926
3.47
0.181
3.52
0.093
3.62
0.400
3.50
0.280
Red Siberian Crab
1926
3.21
0.450
3.25
0.247
3.40
0.810
3.26
0.660
a
a
<
a
a
o
Xi
u
•o
4-1
lU
3
u
C
C3
■w
s
t«!
iri
W
i-M
a|T3
u
O
3
h
•-S
O
C
Oh
s
v>
o
•o
^
u
o
«
cc
u
a;
X
^
lis]
S
o
^
(M
o
c
o
^
3
X!
•T3
i;
"a
a
««ir-'>«<ooa3(Noct^ooooaioooiO(N
■*.
o
J3
^ CO •*■ CO ■* to n; lo t>; 00 1-; cj -^ to 00 -^
t>^
o
^lMOC^(NM(N<NC-)MlMiN<M(Ni-ilN
IM
(4
V4
O
«.
X
01
c
o
o
.^
s^
S t^ (35 X 05 O; 00 O X 05 ■>) 1>; C<l O t-. 00_
O
_o
'^ lO M im' t^' C^ ■*■ o •*' O O "3 W O ® "i
d
W
^(NININ(N(N(NINCqiNCSlC'IiMMrH(N
CO
-o
0)
a;
a
■i. C5 -; c-1 T)> 00 —1 o to ^J lo t^ ^ c-i o C5
»\
o
^ CO t^ to '* «' 05 O L-t CO O O -t -^ CO rt'
t^
^
V lO i-O O O O O •* •* Tt> to to t~ uo to lO
i<
rt
O
■o.
X
.4J
*J
u
E
"C
■g
(U
») CO O to CO to lO C) i-O CO t>. X C5 to •# 00
.^ to' CO d d 05 X t^ d so d d d co oi c^'
■*
r-'
W
^ Tt< lO CO i-O "O O CO ".O O to O to Tf to »o
■*
•a
u
a
■u t^ OS O X ■*_ to O CO T-i CO —_ o -; c^_ x_
Ol
■ij
o
U
^ d c>i lo rt lo ~r t>; t^ d m" r-' d d "O x'
d
^XOJXOiOiC. XXX005C:C;C:X
■o,
X
'S X
u "O
u o
3 C
o
•a
11
§ q ^. CD q (M_ -^ q -^ in x q x t^ -# cm_
"^ lo d <N d -I ■* d t^' d -^ d t-^ "-0 -t< «'
CO
<J
ci
w
^XX05X0i05XX05 05O0:X0505
05
)-•
-s^
m
■*-*
f^-^^tOTtotO-^OlNOllMINe^iM
(N
H 0
3
-M
Z
1)
u
3
G
1^
,o * * *
V4
■aXOOiXOCJXOOOOOOOO
O
C
JJXOOXOOXOOOOOOOO
o
(3^ I-H.-1 i-ti-H ^rH^i-.i-.rt^i-l
0^
■o
H
'o 2
OJ
C0COC0Tj<'#-f<INC^C^COCOCOCOC0^D
CO
O 4j
rt
S
cocococococococococococococoffo
CO
bo
a
^
"c
>
J2
c«
O
•a
cd *n Ji '>
(U
O
5 ^ s s •? 'm ? < w 5 2 Q g
eac«cdcac3cacadc«4;4j.r±:'-tr
^
M CO m
(sipqp3mDacQP:!ii:;:?2^4:s
oi
a
a
<
(U
a
a
o
u
s
"O
Si
M >n h, in o oo oi 00 c>i oi in oo —_ » ci >•■: o)
a
o
^ h^ CO O -- -r ^ 00 Cl -i "5 C^ p — ' 1< P "O
^ — O M C^ (N (M — (M C^) M M ?i ft — fl iM
1
U
«.
K
c
o
u
di
T3
S.^!OONPt^OC5«C<3CJ>.tt»e>J<3C^
_y
^ 05 ■*■ p' <N lo o — -^ -^ -r x" -^ d ci p' «'
i^
^ •-H (N IN IM 04 — W (N C^ -> C-l C-) 04 C-l CI <N
•c
« C! Ol in O p Tf .-< •!)< p (N t^ IN O) p M P
c.
'^ IN p" "ri p' '»• p "O oo' irf (N oo" in ec oo' r>^ -f
^
^•^'^mn-vnoommio-^'^m-fnm
^
u
■c
U.
X
dj
4->
k.
E
-a
<v
5 p r~ to « P t>; p 00 M o) p p_ CO p_ p_ -)<
^ c<5 t^' p' -^ cc X d p Tt! p' r-' ci p' -c' >n c-i
m
•ft.
•o
c.
S X a: P X P t~ -^^ 'n -; h- -!; T)< X P M P
■*S
c.
o
"^ 'N p' f' cc" c^' X cc n t^ p cc 'T ro -f" -r p'
o
i^t^t-t-t^Xt^Pt-t^XXXt-t^t^t^
> i
u
■Q,
to u
<J OJ
3 _C
m "2
TO r;
S IN o <N •* r>. ri in p r-; c-j IN p p p P ^_
O
y
"^ ■*' x' p x ^^ x' « f' d p' ■*' i>' p' •*' p' x'
55
^;opr-PXX'xt^t-xxxxt^pi-
U
?<
4-)
t^ -f -< P -f — O -r O M C-l C) C-) C-1 !N CI
3
z
V
u
3
G
rt
|lll?88S3?gS§8§8S8§
1
M ^r- .-(^ „„r-i — i-.-i^^,-i
^
Z
■a
H
0 ^
V^
c«:ccM-»'fT».MMCi^:rt:occ;r;TCM
"^ S
« i
c<5c<5mc»5ccM«Mccccrccc~;rc?;M
bc
c
>
3
CO ^
>
.S .£ £ .£ .5 .= .s ■= ■= M £ ^ = a t?
§ 1 II 1 1 1 1 1 ^ ^ > S 12 > 5
246
TECHNICAL BULLETIN 15
TABLE 9 — Influence of Period of Extraction and Temperature upon the
Percentage Recovery of Soluble Solids and Pectin. Ratio of Baldwin
Apple to Water 3 : 3.
Temperature
Number
of
tests
Extraction
period
3 successive
combined extracts
First extract
Second extract .
Soluble
Pectin
Soluble
Pectin
Soluble
Pectin
solids
solids
solids
degrees C.
minutes
per cent
per cent
per cent
per cent
per cent
per cent
88
10
15
87.8
69.3
48.4
36.7
26.0
19.8
88
3
15 + 10*
87.7
60.6
44.3
24.2
26.7
18.6
88
3
30
8.5.4
62.1
42.6
28.7
27.2
19.8
88
3
60
87.5
79.0
.50.7
36.8
24.8
25.9
100
24
15
90.3
77.3
54.0
39.6
24.6
23.5
100
6
15 + 10*
89.8
77.9
52.6
37.3
25.0
24.2
100
6
30
87.4
79.7
53.9
41.4
23.6
25.6
100
6
60
92.8
84.8
59.2
47.3
24.1
24.8
109
10
15
91.1
75.4
52.9
35.2
26.4
'25.5
109
4
15 + 10*
93.0
77.7
58.7
43.2
24.1
21.6
109
5
30
93.9
83.0
61.5
48.2
22.9
25.2
109
4
60
93.1
87.5
63.8
.56.4
19.4
20.0
* 10 minute standing period.
j:
— M
5 feS"^
S5§
°2S
10 wt
"5 ox
0-" t-
523
1 C
gOO »t^
cor- CO
r~!Oco
INX"1«
*
>.
V
1j
1
IN ?)C^
IN — --
C-l N —
— -< _«
CO N — .
CI — CO
01 N —
01 — —
—1
O
2.-S-3
.C^OC^
"5 X O
t^-^CJ
COMt>.
Ot>;l>;
Mh-t-
OiOlCO
fflOlX
o '* 9 3:
■*" N ffl
dsJO:
■* d ci
CO d CO*
■d d -r
id d —
I* d 'd
i-HCV)(M
«N01
•-NC^
— IN'N
--N
— 1 — N
— — IN
— OJOl
>> ,,
^
S 05 O M
^ (N m" (N
5000!
q o -i;
COOllN
too OS
>ONX
h-Oco
X q OI
•*'-': ro
-r -I'-!'
id id CO
CO dh-^
— dcd
eoh-'f'
•rt-^-)<
u ;=
■o.
c.
< -
"a
_c
5=5 t^M
^ 3" ?' =='
mxco
MX —
-i'_--eo
C0t«<O
(^•frjl
lOOOJ
oia> —
3
u
o TC — ■
in >c 1^
id id d
-»>'dx'
■*idd
t^ — ■*■
id — CO
ai
•o.
CO -'IN
M-^C-I
NC^ —
CO CON
OJO)-
a;
^
s;
?i
vCif^X
05 ^) r^_
CO ox
r-'N 03
qcoo
--jiO
— X-
•^t-co
^2z:^
t>;>OT|<
d -i> IN
•4 in" d
ci -t — *
— id id
id X* d
-r d d
o 2
w ■"
«,
>.
^
.5 .y
Srooot^
C5l-Ot~
cox-^
(N -<o
INOX
-J3OC0
t-iiCCO
05 — —
3 rtrt
^ oi d 3C
'■£t>--3
-i d d
CO d id
d COCO
— di^
— X tvT
— "dr-^
i! E
03
<
Ci,
c
s:
X
— t^?D
yr'.-f-i
SOCVi
INfflCO
-|)<X —
INf-CO
— XII
aj
^d^'|2
x' ■>] c:
— " X ?!
dido
oi -4 CO*
— d— ■
Tji'cdd
i<* r-.' IN
[E
■c
— -,
-.-,—
rt«
— — -.
— — —
— — —
— — —
IS
3J
^
3-d
3 —
^aot^X
■O C5 t^
IN uO m
N r- in
O-i-*
T^coo
co'-ooj
'l<^_01
^^■-v
dt^x
co' d »n
r-iii Tt<
d IN -H*
WIN —
idiNd
-i-'idcd
"o S
— — cq
w "
•Q.
>,
^
.t: .-i
■a .-< •*_ C>J
t^XN
t>;C<3C<:
T^qr~;
coo^
XOSO
qqx
x_ q o)
3 S'5
^ CO -^ rt'
•^co'i^
id c^j co'
diNffl
d>CR-4
dod-H
d-**co"
o'ldd
^ S
JjNNIM
IN — — ■
C^lC^l c^
— iC^ «
l-Hl^CJ
M — N
— IN —
INOJ —
0^
■O.
_c
t r-; o: O
-•con
COC^_-H
■*inx
t '^ t
xqr-
cocoas
OSOJ-t
'J
■^ ci -r 1^
CO* *f *^
dx'-<'
oi-Jd
d d d
oi N* M
d d x
d X oi
5
o
OJ
V
a,
^ rj c^) 'M
r-^J-N
IN N M
NN-
OICIOI
C-lOqiN
OIOIO)
c^
— CO
ts
4;sir-.c<:
OiN-f-
OI-^M
oi M q
NOt~
Xt-M
-JiX-
Oi-OOJ
3-^
"^ ^6 CO -^'
— or-'
id id oi
c^cdd
r-' d X
Tf'd —
— idd
d d CO*
"o °
^ M ■M M
C1M«
(NC^ IN
INCI'N
■NCO-H
(N — IN
c^ci —
oicooj
t« "
•Q.
- .a
5t^XX
X t^t-
-3 M 3!
'-;q-<
q q 'O
— r~co
'^P^
lO-fiO
iiSa
^ «■ in t^
MC0 50
x"d>d
oilNC^'
d d -t
d id t^*
id d -)<
^•i*
o c
^ lo ^n m
tor^o
Tj<;0 35
i.oor-
i.O f i.O
lO -.o -^
lO lO t-
1-0 mo
o
< "
■o.
rt
-^
'J
C O ?3 O
'" rt M CO
c^ OS 'd
® -O I')
co' r>.' ci
CO c-i c^'
— coco
CO dd
— wx
0005
X* id t>.*
lO^ —
dojol*
2
CL,
-r 'O o
coco-f
i.O lO 1.0
-tcoco
-l'C0 1<
MCOOI
CO CO in
^
t -r r: — ;
■£>_ t~_ 33
1.T X m
IN 3-. -.O
CO q q
t~^x
iNxq
q^tjco
= :=
'' OS C'J 30
dcdd
-■' d d
id d -t<
d CO* x'
— M— '
ad NCI
■nTfiS
o 2
^-r-oin
;0S3O
o -o -^
o-jr^
-T-V-r
■CfflS)
1<iOO
■Jn "
•Q.
n^ S
xox
M -N C-)
IN CO IN
ci M ri
— C>1 M
M M N
01 COM
01 01 01
Eos
— !>!«
3 -^
z.
.2
u.
OJCO •»•
■M?0 -t
N CO -I-
-MM I-
■Nco-r
N co-r
OJCOf
OJCO'*
'w
- s
(0
CO COM
:c rocc
C0CO?3
CO CO CO
CO CO CO
CO CO CO
coco w
CO CO CO
>.
C
•r
^
c
'i-
1^
>
pa
, 5
■a ^
as
1^
13
0
0)
Q
a
as
'•^
^
i2
:?
si
!/5
a
Xi
3
O
cc
O
O
>>o
bO
«^
1
a>
CC
OJ
'4-1
v
3
Xi
C
■w
C
s
o
o
u
re
01
•o
(Tt
C
^
OCU
■4J
o
3
'M
C>
b
C8
o
h
'<->
«
o
:z]
'M
C3
Ctf
•o
c
C3
>.
■w
a>
•pM
u
«
>
^S
goc^io
to -J to
1000
IN CO to
lOTf ■#
-*NO
r^cox
oxr>.
■j; c
or^oo
N COt^
t^ — t^
-*t^CO
CM*-<
^ i-
CM -H--I
CM-l -H
*
>.
o
M -hC^
N -Hrt
COM-H
rt NCO
CO CM CM
CM -^CO
CON .-1
COCO— 1
^^
O
?-^"l
.c^oo
•*OrH
lOMt^
OX!N
iCXO
t^OJCO
toco«o
to*0
gcodro
.^^CO
icjioo
'J'CJN
cooco
-*o-*
lOOlCM
*02 O
" — CSM
-HININ
-H WC<I
-H^Csl
rt rt CM
rtrtCM
.-(-HCM
rtrtCM
.$ .y
§>O00M
-Hcoo
coto t^
-H^O
COC5-<
rtCM'C
coxx
XCO-*
2 rt rt
^ m' x >o
^JiOLO
Otoco
■* coc^
-Hcoto
t^t^*
t^t^iO
COOS*
y E
a
< ^
■cs.
3
a
c
OJ
wt^POC^
MOCO
(N05tO
lOM'*
tOOCM
OJCMX
CM CO 0
UJO*
3
^r-^dtd
O-hO
t^ N OO
f~co "*
CO coo
10 CO *
tct^x
£
IN M rt
CM CM CM
CM CM CM
CMCM-H
CMCMrH
1)
X!-0
t^OcD^O
O-Ht^
to CO lo
COIN IN
•ococo
t^toos
t~*CO
*OX
"^ r^ oi »o
tot^-*
» >OM
C^NIN
r^a>^
coo>-*
tooo
CO CM to
■3 g
w "
•0,
d 0
gin o) to
oot^os
©■ot^
OOt^OJ
LOXCM
tOt^CM
CM05»0
X CO >o
3 S rt
"" c^i 00 to
COO
Tt<0"5
©■^IN
t^^d
-HX05
— I05X
^ S
4->
■e.
u
c
X
toiNcc
005CO
CONt).
,-irt to
CM-So-Oi
-**f-
toco*
il5*»0
0
^ d IN IN
e»iNN
IN-H-H
oooso
CO CM CM
CM*t^
* too
■E
(5!
■la.
— Irt — 1
H
^
3S
Socio
03-HO
t^ooco
coiOi-i
CM CM'*
■*OJtO
-*r^ro
o*co
^cod-H
oot^i>
10 »-< t>-
Tjtioeo
CMXCM
OJiOCM
COCMO
*CM'0
0 °
fc'-i '-
1^ t— (
f— 1 i-H T-4
Cfi "
■ft.
5- 0
trtOCiO
OONO
ocoto
N Csl —1
CO-*-*
lOXCO
CMt~(^
— OX
S «"«
"■^t^ iCli
'*I(N'0
t^iNN
TtlCOCO
CSiOOi
o*x
02Cir»
OOX
^ ^
^•ININ 0<
N<NN
IN<NCS
cscq r-<
CM C<|rt
CMCM-H
t-H »— 1 »-H
■C
_c
£c^ to c^
Ol tool
NOiN
(NC3C0
CM CM CM
CM i-lC^l
10 10 CO
000
0
^diod
oocc
IN "OiO
i-HNOO
CMiC-H
* -H t~
t^-^ t^
CMCStC
-^NCM
8
(1,
■c
w
QJ
J2-0
^00 toes
t^MOO
to to 00
to 00 00
05-*tO
to to*
COO-H
t0 05l>
3 =
tocor^
00 1-H c^
IO05 05
-^tOX
C0Of~
XOCJ
-iClCM
0 g
*;(N(N!N
N(NCS
C^CO CM
CM rt rH
CMCq — 1
CM CM —1
rtCM-H
CM — CM
C« "
■ft,
>>
CO 00 IN
t^to-*
CJOr^
05 Oi CO
XCOO
CO too
co*co
."H rt rt
^ t^t^-*'
-H-HOO
00000
r-<O05
©■OC^I
*tot^
CMCOt>.
* —X
^ s
^Tf "OCO
to CD 10
■* toto
totot^
* to to
TJIlOtO
lOtOtO
•OtOtO
OJ
■ft.
^
X
toiN M<CO
comiN
oox
(NX t^
0*^
ococo
t^tox
<j
^d-^io
N lOlN
t^O-<)<
COTft^
— NiO
tOCO'O
CO to to
CM CO "5
E
Oh
^TtH T|< ■*
■ft.
-tmin
CO'^TJI
10 »c 0
CO**
-* * *
— m
^
<tiNC5 C^l
■^om
r-,m~^
lO 10 05
•*rt t^
eoo-H
tOOCM
Ot^CM
3;z:
•*tot>-
cocoio
rtt^O
— inio
"o 2
^■<J. lO to
to to to
T}<into
tot^t^
10 010
* 10 to
lOiOtO
000
72 "^
■ft.
■-
•^^ "
M INM
NCOC-)
CMr>cM
CM NCM
C-JCMCM
CMMN
C^ CMCM
E 0 £
3 ■"
Z
o
MTO-*
INCOTfH
NCO-»<
CM COtJI
c^ CO-*
CM CO*
CM CO*
NCO *
"o52t5
cfl
a;
i: 5:
CO CO CO
CO CO CO
CO CO CO
CO CO CO
CO CO CO
CO CO CO
CO CO CO
c
>.
nJ
•0
c
c
1)
1
'$
•a
"(3
m
1 01
(U ■"
Pi
LO ttf
0
c
0
2
p
B
5
a
CO
1
0
I
a
a
<
G
2
OQ
<*i
O
o
A
G
o
•a
<
o
V
u
c
3
S
■T3
CJ
a
a
1
jr
feMIN(NIN-HMMN
U
■cx
•o
e
CJ
^•^OOOMrO^OOO
if5-HC0^C>1<ro-i
a:
5o
feMc^e^c^-MNiN
(-
«.
X
a
-J
z
13
.^
o
a
o
V X ^_ -fo t^ o m to
111
•a
. i2 t-' >o' r>: CO «D «■ to
C/3
^
feC^INC^INMMCNIN
1
U
•e>.
0)
3
s
;i
OJ
lO ■<(<■-}; N i6 00 W TO
^sj
Sq
•o
K
Si
o
^SDoqtDO-jNKIOO
55 — 1 «" N -j! •*■ lo lo
hTOrJ<T!<'fl<CO-^TOC>.
c
U
«. •
(J
Oi
.^
•o
c _
H
^OtC-^XOO-fcOO
U
u
. TO d C^i rf tg TO to' to'
<
S
5;ro'<»<M<-*<NTO'*T}H
a
*,
H
X
T3
e
tt
G
G
O
wajioioiot^xusoo
" ■^' to' u5 ci 6 IN d lo
b
T3
^U5iOiOU5iCiO-<><iO
1
U
■<»■
CJ
S
n
•o
s — „
O
u
^jINoqOMOO^DOJM
, d ■^" 00 OS m' «c t>: ci
W
,u
5q
•o.
XI
CJ
s
G
^J-JpO'OOrtj^OO
G
O
c
U
<i.
CJ
C/l
Ch
.^
H
•a
s
U
CJ
«- !C !N C^ t>. O TO M TO
'=' 5
_o
^ M' ■*■ c4 OO" !£> '^ o' t^
2: H
w
S;t^t^oot^;ct^t^r~
(j] Id _
«.
^ i
CJ
s:
c «
,
G
jjC^t^mOMC^OM
«
^ m' — c-i ri 00 -h' ci TO
o
•—
^OOOCJCCCJMO
u
■©
CJ —
U
«.
D
5
X!
o
tt^-irJciOr^MOOO
7J
.^
. ^ 00 — ■ d "O 00 (N oi
t^oocjowoOoo
7;
•c.
i->
•2 >_ "
r< « CI o c) cc r> t^
£ "o 2
2
.y °
Tartar
acid
added
applci
5 M I- oc -r T"
^ c d d d d d d d
(i
U
|g88Sgg2gg§
r
3
a>
a
a
<
G
if
TS
«
u
pa
o
fi
e
TS
y
r/1
<
3;
■M
.-
3
«
^B
•o
»«
fi
cs-o
c
0
u
Oi
(J
Pu
PU
C
^
0
<»
•o
u
^M
fT!
o
^
rr
■w
a>
td
X!
3
O
cc
"3 5
rt o S ii rt
s: O CO iM
iM 1-1 — ,_ rt ,-1
M O t-j O tN. t>.
CO OS ic CO 00 ■*'
:0 0 0;
-H 0 0
CO 00 05
Tt< >0 rt
C>1 (N O)
2SS
'f OJ t^
<N O) C^
C^CO-* (MCO-* OICO-I'
cococo cococo cococo
APPLE JUICE EXTRACTION' FOR JEM,Y
251
TABLE 14— Jelly Yields per Unit Weight of Apples.
Combined First and Second Extracts at 100° C, 8 varieties.
Extraction
period
Number
of
tests
YIELD OF JELLY
Per
pound of
fruit
Per
kilogram
of fruit
Jelly
strength
Grade*
Pectin
as pectic
acid
Ratio of fruit to water 3 : 2
minutes
ounces
grams
grams
per cent
15
16
14.26
893
118
1-2
0.283
15 + 10**
5
14.95
937
123
1-2
0.272
30
12
13.20
827
143
2
0.318
CO
5
12.90
808
78
3
0.317
Ratio of fruit to water 3 : 3
15
Ifi
20.02
1251
88
1
0.221
15 4- 10**
5
19.90
12.39
90
1
0.229
30
12
19.90
1239
95
1-2
0.239
60
5
18.33
1141
57
2-3
0.271
Ratio of fruit to water 3 : 4
15
16
24.60
1432
67
1
0.219
15 + 10**
5
22.40
1395
70
1-2
0.235
30
12
23.55
1471
76
2
0.229
60
5
19.40
1496
; 52
3
0.298
* Grade 1 — good; grade 2 — fair; grade 3 — poor, unmarketable.
** 10 minute standing period.
3M. 7-'28 No. 2857