CONDUCTING BREEDING INVESTIGATIONS 593, 
Since so much has been said about the Maine Station investigations 
of fecundity in fowls, perhaps it would be of some interest to know how 
records are obtained there. The type of trap-nest in use is shown in 
Fig. 229. Details of construction need not be taken up here, except to 
remark that durability of materials is a prime requisite for continuous 
service. The absence of any springs or other involved contrivances has 
made it possible to use this type of trap-nest in extensive breeding in- 
vestigations involving a flock of about 2000 hens. Ten such nests are 
used in a pen of fifty birds, and an attendant visits the pens at intervals 
of one hour or more, depending upon the rate of egg laying. Obviously 
a method such as this is expensive even when reduced to the simplest 
terms, and it is, therefore, applicable only to the selection and production 
of breeding stock. It is difficult, however, to conceive of any other 
accurate criterion which might be adopted. 
It should be noted that statistical requirements do not demand that 
complete records be obtained, for the existence of modifiability and other 
kinds of individual variability make it impossible in any event to get 
anything but an approximate record. Accordingly in recording the data 
of production of dairy cows, for example, it is not necessary to weigh and 
test the milk every day for the whole period of lactation, but two or 
three 7-day periods at stated times with respect to the beginning 
of lactation will give a sufficiently accurate estimate for all practical 
purposes. Similarly in poultry breeding, Pearl has found that produc- 
tion during the winter period is a sufficiently accurate and distinctive 
index of the egg-laying capacity of a hen. 
Further the danger from unjust comparisons should always be empha- 
sized. A comparison between egg production of hens in the second 
laying season and pullets would favor the pullets, for pullets ordinarily 
lay more eggs during the first season than they do as hens in the second 
season. Moreover different parts of a given season are not equivalent. 
A pullet lays more eggs in a given length of time during the spring cycle 
beginning about March 1, than she does during the winter cycle. A cow, 
likewise, produces more milk during the early part of her lactation period 
than she does later on, and she reaches her maximum capacity at 5 
or 6 years of age. With respect to these points we have reproduced 
in Table LX VI, the comparative indices which Pearl has calculated and 
which provide a method of comparing the productions of cows of different 
ages at different stages in the lactation period. Asan additional variable 
in this case we should include the time at which a cow freshens, whether 
in spring, summer, or fall, as having a definite influence on herd produc- 
tion of milk and butter fat. We could go on recounting without end 
such factors which must be considered in making accurate comparisons. 
The point, however, is sufficiently obvious, namely, that even objective 
38 
Digitized by Microsoft® 
