312 MAINE AGRICULTURAL EXPEjRIMLNT STATION. I912. 



recorded numbers would have fallen very far below those 

 which they have actually, and most fortunately for the good of 

 biology, been able to obtain. 



The foregoing remarks are not in any sense intended as an 

 apologia for the statistical portion of this paper, because in the 

 opinion of the writer, who is thoroughly acquainted with the 

 practical difficulties whicli beset the study of inheritance of fe- 

 cundity, no apology is needed. The data here presented are 

 about as extensive as it is practically possible to obtain in an 

 interval of time and with an experimental equipment equal to 

 what has been available in the present investigation. It is hoped, 

 however, that what has been said may help the reader, who may 

 not be practically familiar with the rearing and trap-nesting of 

 large numbers of fowls, to understand the reason why more 

 extensive data are not forthcoming in this paper. In every case 

 where the number of birds to a family was too small to warrant 

 any conclusion this fact is particularly noted. The data for 

 these small families are not suppressed, however, but are in 

 most instances separately tabulated. 



One convention which is used throughout in the tabulation o£ 

 the material should be explained. In case a bird has a winter 

 &gg record of exactly 30 eggs, she evidently falls on the boun- 

 dary line between the two fecundity classes already discussed 

 and defined (p. 302). The number of such cases is not large, 

 but in order to be perfectly impartial in their treatment it was 

 decided to split such a bird in two, in a metaphorical sense, and 

 credit one-half of her to the 'Over 30' winter fecundity class, 

 and the other half to the 'Under 30' class. This explains the 

 fractional records which occasionally appear among the fre- 

 quencies in what follows, and which might otherwise puzzle one 

 used to thinking of a hen as an individual unit, at least during 

 the fecund portion of her existence. In calculating the mean 

 winter production (in eggs) of the several classes these few 

 birds with records of exactly 30 eggs have been omitted alto- 

 gether. There are obviously two equally fair ways of dealing 

 with them in getting these averages. One is to include each 

 one in both 'over' and 'under' classes ; the other is to include 

 each one in neither class. The latter alternative is adopted be- 

 cause simpler. 



