Massachusetts Agricultural Experiment Station.

Bulletin - Massachusetts Agricultural Experiment Station (Volume no.379-398) online

. (page 19 of 77)
Online LibraryMassachusetts Agricultural Experiment StationBulletin - Massachusetts Agricultural Experiment Station (Volume no.379-398) → online text (page 19 of 77)
Font size
QR-code for this ebook

and dysentery bacilli from a contaminated water. So, instead of using
a direct method for the isolation of disease-producing bacteria from water,
the bacteriologist makes use of an indirect method.

There normally is present in the intestines of man and animals an or-
ganism known as "B. coli." This strain of bacteria is given off in the waste
materials in large numbers. The organism is harmless, and resembles
the typhoid and dysentery bacteria in that it is rod-shaped. It can be
rather easily identified by somewhat simple cultivation methods. The
presence of this organism in a water supply indicates that the supply
either is or has been receiving waste material of animal or human origin.
The numbers present will indicate to some degree the extent of contam-
ination. The argument presented when this test was devised was, of
course, that if "B. coli" could gain access to a water supply then typhoid or
dysentery bacteria might also gain access to the supply. Thus in the
•laboratory test this "B. coli" organism is used as an "indicator" of contam-
ination in water. This is what was meant when it was said that the
laboratory test was an indirect test rather than a direct test.

So far it has been impossible to distinguish in the laboratory between
the "B. coW of human origin and that of animal origin, and there is little
reason to believe that such a distinction would have any practical value
in routine water testing. If either type is present in a water it means that
faulty construction or poor location is responsible. The mere presence
of "B. coli" in a water supply, be it of human or animal origin, indicates
that the water is unsatisfactory for drinking purposes. And, further,
drinking water containing waste material from any source is potentially


It is only fair to call the reader's attention to the fact that this bac-
teriological test is not infallible. Water should not be judged wholly on
such a test. A common-sense sanitary survey of the supply and its im-
mediate surroundings should supplement the laboratory analysis. In most
cases this survey will indicate the source of contamination and the kind —
whether it is of human or animal origin. Further, there may be occasions
when the laboratory test fails to reveal any contamination, but the sanitary
survey may show certain potential sources of contamination. Such a
condition should arouse questions about the future purity of the supply
even though the present laboratory tests failed to reveal any impurities.
This survey will reveal also faulty construction in the supply, which is an
important item particularly in shallow dug wells. If persons building
new homes in rural areas would only remember to locate their water
supply first, and then so construct their homes that the disposal of waste
materials can not contaminate the supply, little trouble would ever be
experienced with impure drinking water.

The results of a single analysis are not always too dependable. A
drought or a rain storm just previous to sampling can influence the results
considerably. As Dr. Feemster points out in his letter to the writer:

No doubt in your bulletin you will emphasize the fact that bac-
terial examination of water should not be depended upon solely
to give an idea of the safety of a particular supply since a sample
taken during a wet period or after a heavy rainfall might show
evidences of pollution of a well which under more favorable cir-
cumstances might appear by bacteriological examination to be
perfectly safe.

Careless sampling methods, and careless handling of the sample between
sampling and testing, can also influence the results considerably.

Before concluding this section some mention should be made of chemical
tests of drinking water. In the laboratory at the Massachusetts State
College chemical analyses of drinking water are not made. One reason
is the belief that the bacteriological test is a much more dependable and
sensitive test than the sanitary chemical test. The bacteriological test
can detect minute amounts of contamination which might be missed in the
chemical tests. Further, the chemical analysis is of little value when
applied to rural private water supplies unless several such tests can be
made over a considerable period of time. Individual waters vary con-
siderably in their chemical composition, and unless normals have been
established by means of frequent tests it is dii^cult to justify an opinion
of the quality of the water upon one test. In the case of town and city
water supplies the State Board of Health makes frequent routine chemical
analyses of the supplies. This has been done for many years, and in this
way it has been possible for them to establish normals for the several
chemical components of the water. If at any testing period results show
that any one or several of these chemical components varies from the
normal, check tests, as well as an immediate sanitary survey, are made
to locate the source of the change. Incidentally, the bacteriological
analysis is one of the most important check tests.

In conclusion there is another type of chemical analysis for which
frequent requests are made. So far as we know there is no water in the
State of Massachusetts that might contain any one, or several, chemical


compounds which might have a medicinal "cure-all" value. Frequently
persons send us samples of water which they believe to contain some
unusual chemical constituent with remarkable curative properties for prac-
tically all known human ailments. Our answer is that there is no such
water and such analyses can not be made.

All analyses of this type, as well as analyses to determine what type
of piping or hot water tank to install, should be referred to a commercial
laboratory. The laboratory at the Massachusetts State College deals
only in health applications of any tests made.

Results of 1,016 Analyses

A very brief survey of 1,016 analyses made in the Service Laboratory
during the past ten years is presented in the final section. The very fact
that these samples were sent in to us for analysis suggests that in the
majority of cases some doubt existed in the minds of the owners as to
the purity of their supplies. Therefore a rather high percentage of con-
taminated supplies is to be expected. The reader must not interpret the
final figures given in tables 1 and 2 to imply that this is a general picture
of the condition of the rural private water supplies in the State of Massa-
chusetts. These figures mean' just what they represent: that 44 percent
of all the supplies suspected of being contaminated, and scut to the laboratory
to he tested, were contaminated.

Table 1.— Total number and type of each sltpply tested, and the
number and percentage of each showing contamination.

Type of Number Number Number not Percent

Supply Tested Contaminated Contaminated Contaminated

Spring 193 77 116 39

Dug Well 452 250 202 54

Driven Well 332 119 213 35

Totals 977 446 531 45

Table 1 presents the total number of samples tested from each type
of supply together with the number and percentage of each that showed
contamination. These results show, as might be expected, that the high-
est percentage of contaminated supplies occurred in the group classified
as dug wells. Being for the most part rather shallow (the average depth
of those tested was only 15 feet), dug wells are the most difficult type of
supply to protect from contamination. The majority of those tested were
of rather ancient vintage; that is, from 30 to 150 years old. At the time
of their construction a general knowledge of water sanitation was lacking.
Field stone was used as a curbing, and the cover, if any, was of loosely
laid planks. Further, for convenience the well was located in or near
the house. So also were the toilet facilities. The result is obvious. The
loosely laid field stone curbing and plank cover allowed surface-water
contamination, while the nearness to the sewage disposal system resulted


in sewage contamination. Modern knowledge of water sanitation, requir-
ing a tight tile curbing and a tight inetal or stone cover, in addition to
proper location of the supply, results in a fully protected drinking water

As stated previously, spring water as it emerges from the ground is
generally of high quality. Contamination of the supply occurs only after
the water has emerged from the ground, and it is for the most part of
the surface-drainage type rather than the sewage type. Proper excavation
back to the actual point where the water emerges and a well-constructed
spring house will prevent this subsequent contamination.

The records show that in practically every case where contamination
occurred in a driven well it could be traced to a poor location. It is true
that soil will act as a natural filter for the water percolating through it.
However, when it becomes saturated with impurities, it tlien ceases to
act as a filter. Further, even though the source of the impurities was
removed, it still would take the soil a considerable length of time to
purify itself. The answer to this problem is to so locate the well in respect
to the waste disposal systems from the house that no contamination from
these sources can gain access to the well, or to the soil immediately sur-
rounding the well.

Table 2. — Location of supplies tested, by counties, and the number



Number Number Number not Percent

County Tested Contaminated Contaminated Contaminated



Plymouth 9 9

Bristol 17 7 10 41

Barnstable 118 29 89 24

Norfolk 3 1 2 33

Suffolk 8 7 1 87

Middlesex 21 6 15 28

Essex 17 11 6 64

Worcester 223 121 102 54

Hampden 104 41 63 40

Hampshire 235 118 117 50

Franklin 184 84 100 45

Berkshire 11 29 48 2>1

Totals 1,016* 454* 562* 44"

* Includes drilled wells.


In table 2 the supplies tested are classified by counties. Here again
tlie results given are in no way a reflection on the general condition of
tlie private supplies located in the respective counties.

As migiit be expected, the majority of the supplies tested were located
in counties near to the laboratory. The 118 samples received from Barn-
stable County were an exception to this. It is interesting that the lowest
percentage of contaminated supplies also occurred in the case of the
supplies tested from this County. This directly reflects the fact tliat
people are becoming more alert to make sure their water supjdies are
satisfactory, for most of these supplies were of new construction. This
is brouglit out more clearly wiien the results of Barnstable County are
compared with those of Worcester, Hampden, or Hampshire Counties.
In tlie case of these three latter counties the majority of supplies tested
were not newlv constructed.

It has been the intention in preparing this bulletin to present certain
information, little known or little understood by the public, regarding
the contamination in, and the bacteriological testing of, rural private water
supplies. The containination of such supplies with typhoid and dysentery
bacteria has ceased to be a public health problem in Massachusetts. The
State Board of Health Reports for the past few years fail to reveal any
case of typhoid fever directly traceable to the drinking of contaminated
water from a rural private supply.

Water can and does, however, become contaminated witli the waste
materials from man and animals. The routine bacteriological test i>
designed to isolate and identify an organism known as "B. coli," which is
present in large numbers in the feces of man and warm-blooded animals.
This organism is used as an indicator of tlie presence of fecal material
in water. The test is both sensitive and dependable, for the presence of
such inaterial in a drinking water supply is at least objectionable, and
should always be considered as potentially dangerous. No attempt is
made in (his test to directly isolate disease-producing bacteria from the

Publication op this Document Ahhroved ay Commission on Administration anu hiN,-
4i]i-4-4 1 — ."i7Vl2

agricultural experiment station

Bulletin No. 384 July 1941

The Importance of

Length of Incubation Period

in Rhode Island Reds

By F. A. Hays

This represents an attempt to determine whether length of incubation period
may serve as a criterion of the future performance of chicks.



By F. A. Hays, Research Professor of Poultry Husbandry


A limited amount of preliminary unpublished data on Barred Plymouth Rocks,
collected by Dove (1935) at the Maine Station, suggested that chicks emerging
early from the shell were more likely to be females than males and that such
chicks were probably more viable than chicks emerging later. It is also com-
monly observed, in chicks from the same setting of eggs, that those emerging late
are less likely to survive and often grow more slowly than those emerging early.
Such observations suggested the possibility that the length of the incubation
period might be a useful criterion of the future outcome of chicks. A study of
the length of incubation period for Rhode Island Reds in its possible relation to
sex, viability, and characters afifecting fecundity was therefore undertaken m
the spring of 1937 and continued through three hatching seasons.

Data Available

All chicks were pedigreed Rhode Island Reds hatched in the same forced-
draft electric incubator in 1937, 1938, and 1939. Observations covered si.\
hatches obtained at weekly intervals between March 7 and April 23. Seven
emergent periods, each covermg eight hours, were included in the study. The
first chicks appeared during the last third of the 20th day and the last chicks
appeared during the last third of the 22d day under the methods of incubation
used. Data were secured on 4730 chicks in the three-year period. Records were
taken on sex, mortality at different ages, body weight at about six months of
age for both males and females and at sexual maturity and at the end of the first
laying year for females, age at sexual maturity in females, winter and annual egg
production, and the emergent period of chicks from sires and dams classified
according to their individual emergent periods.

In order to reduce to a minimum the temperature shock to the newly hatched
chicks, it was necessary to keep the chicks out of the incubator for the shortest
possible time during each eight-hour observation. For purposes of identification
a series of water colors was used for marking. It was found that the color mark
was most satisfactory when placed beneath the wing with a small brush. The
pedigree baskets were opened at each eight-hour observation. All chicks that
were out of the shell at the first observation were marked; and at the second and
later observations, all unmarked chicks were given the respective color mark
for the period. When the hatch was taken out of the incubator at the end of
the 22nd day, the emergent period of each chick was recorded. It was found
that the chicks could be gone over and marked in a short time by this method.

♦Special credit is due J. W. Locke, Plant Foreman, for assisting in the observations on time of


Relation of Length of Incubation Period to Weight of Eggs

Huggins and Huggins (1941) have recently reviewed the literature on varia-
tion in the length of the incubation period in wild birds. These workers are of
the opinion that fresh egg weight is an important factor in observed variability.

Byerly (1934) and McNally and Byerly (1936) found egg weight to be an
important factor in variation in length of the incubation period. Heavier eggs
generally required a longer incubation period than lighter eggs.

In the studies here reported, the mean weight of the first ten eggs laid during
the hatching season was used as a measure of egg weight. By setting this mean
against the emergent period of each chick from the respective dams, the correla-
tion between egg weight and length of incubation period was approximated.

There were 430 dams that laid at least ten eggs and the mean egg weight of
these dams was tabulated against the emergent period of their 4730 chicks in a
correlation table. The constants arrived at were as follows:

Number of dams 430

Number of chicks 4730

Mean egg weight of dams, grams 62.1

Mean emergent period of chicks 4.2 (last third of 21st day)

Coefficient of correlation —.0981 +.0097

Regression was found to be strictly linear, so that the coefficient of correlation
may be used to measure association. Using either 400 or 500 degrees of freedom,
the magnitude of the coefficient of correlation is insufficient to be definitely sig-
nificant. It seems apparent, therefore, that this small negative value of the
coefficient of correlation indicates no relation between weight of eggs and length
of the incubation period in Rhode Island Reds.

Relation of Hatching Date to Length of Incubation Period

The range in hatching dates was limited, extending from March 7 to April 23.
All the data are thrown together in table 1 to indicate the percentage of chicks
emerging in the different periods as the hatching season advanced.

The data indicate that the majority' of chicks hatched in periods 3, 4, and 5
which represent the second third of the 21st day, the last third of the 21st day,
and the first third of the 22d day. The number hatched on the last third of the
20th day was extremely small, and the number hatching at the end of the 22d
day was also small.

In the hatches produced at the end of March and during the first half of April,
there were fewer early emerging chicks and a significant increase in the number
of chicks emerging on the 22d day of incubation. The first two hatches, produced
before March 15, agreed closely in the percentage of chicks emerging during the
different periods. A slight tendency was observed in the third hatch, produced
after the middle of March, for the chicks to emerge somewhat earlier than was
noted in the first two hatches. Since there was a tendency for the chicks in the
earlier hatches to emerge somewhat earlier than in the later hatches, this may in
part account for a more rapid growth in chicks from the earlier hatches during
incubation, to two weeks, and to four weeks, as reported by Hays and Sanborn


Table 1. — Percentage of Chicks Emerging in Different Periods, by
Hatches, with Summary of Time of Emergence of All Chicks


Percentages of Chicks Erne


Summary of
Chicks Emerging



































































Mean Length of the Incubation Period in the Normal Hatching Season

As previously indicated, the chicks were hatched in March and April, and
this may be considered the normal hatching season for this locality. All of the
hatches during the three-year period have been thrown together in the "Sum-
mary" in table 1 to get a general picture of the proportions of chicks emerging
during each of the seven 8-hour periods under the uniform conditions of m-
cubation used.

About 42 percent of the chicks emerged during the last third of the 21st day
of incubation. At the end of the 21st day only 66.26 percent of the chicks had
emerged. The remaining one-third hatched during the 22d day of incubation.
Precautions were taken to have the incubator heated to about 100° F. for several
hours before receiving the eggs, and the temperature was maintained within a
half degree of the manufacturer's recommendations. The data in general indicate
that there was a wide fluctuation in the length of the incubation period, which
may or may not be a normal condition.

Table 2. — Percentage of M.\les from
Different Emergent Periods




of Males


















Relation of the Length of the Incubation Period to Sex of Chicks

It is desirable to know whether the length of the incubation period is in any way
associated with the sex of chicks. For purposes of study, table 2 was constructed
to show the percentage of males obtained in the seven different emergent periods.

There is considerable evidence to indicate that the sex ratios were lower in
the chicks hatched through the second third of the 21st day of incubation. Chicks
emerging on the 22d day — periods 5, 6, and 7 — showed a high sex ratio. The
mean emergent period for all males was found to be 4.24, and for all females 4.09.
The difference in these values is small and does not appear to be of any great

The data tend, however, to substantiate the postulate that females predominate
in the chicks emerging early and that males predominate in the chicks emerging

Relation of Length of Incubation Period to Viability

Mortality rates for the first six months of life may be taken as a measure of
viability. These rates have been calculated for the first week, for one to four
weeks, between four and eight weeks, from eight to twelve weeks, and from twelve
weeks to housing time at about six months of age. Some chicks died on an
unknown date during the first eight-week period without the sex being known,
and are included as a separate group in the table. Total mortality for six months
is also included. Table 3 gives the summarized results.

Mortality for the first four weeks showed essentially no relationship to the
time of emergence from the shell. Between the ages of four and eight weeks,
however, there appeared to be an important relationship between the length
of the incubation period and mortality rate. During this period the chicks in
emergent groups 1 and 2 showed no mortality; the chicks that hatched during
the last two periods of the 21st day (groups 3 and 4) showed considerable mortality;
and the chicks emerging in the last three periods, that is on the 22d day, showed
excessive mortality.

There were 68 chicks missing at the end of the first eight weeks so that the
age at death as well as the sex was unknown. These chicks were placed in their
respective emergent period and the mortality recorded in the fifth line of the
table. No relationship between emergent period and mortality rate was ob-
served in this small group, except that the very late emerging chicks had a mortali-
ty rate of about 4.5 percent compared with about 1.5 percent for the chicks
from the other emergent periods.

Mortality rates are recorded separately for the sexes after eight weeks. No
consistent relationship between time of emergence and mortality rate appeared
during the age period from eight to twelve weeks in either sex. Between twelve
weeks and six months of age, males had a significantly higher mortality rate than
females, but the relationship between time of emergence and mortality rate is not
conspicuous if it exists at all. The data do seem to indicate that both males and
females from the last emergent period were low in viability. The data in table 3
appear to indicate that chicks emerging late are very likely to exhibit low viability
between four and eight weeks of age and not at other ages during the first six

The last line of the table records the total mortality to the age of six months,
obtained by adding together the preceding mortality rates. The data indicate a
consistent increase in mortality rate with each eight-hour increase in length of


the incubation period. This fact furnishes rather definite evidence that the
early emerging chicks are the more viable.

Table 3. — Number of Chicks and Percentage Mortality by Emergent


Emergent Period

Online LibraryMassachusetts Agricultural Experiment StationBulletin - Massachusetts Agricultural Experiment Station (Volume no.379-398) → online text (page 19 of 77)