No. 511] SELECTION INDEX NUMBERS 



2. A study of the detailed figures of the table makes it 

 apparent that the index is a perfectly impartial and ac- 

 curate measure of a bird's breeding performance. Thus 

 to take an example, bird no. 402 has a low selection index 

 (2.9), while bird no. 168 has an index more than ten times 

 larger (31.6). Do these figures represent fairly the dif- 

 ference between these two birds in respect to breeding 

 performance! Let us examine the detailed figures. But 

 14 per cent, of the fertile eggs from bird no. 402 hatched 

 as against 88 per cent, of the fertile eggs of no. 168. In 

 regard to number of eggs laid during the breeding sea- 

 son the two birds are about equal, with what advantage 

 there is in favor of no. 168. But on turning to the ques- 

 tion of fertility of eggs we see that 30 per cent, of the 

 eggs of no. 402 were infertile as against only 4.7 per cent, 

 in the case of no. 168. The same condition obtains in 

 regard to the vitality of the chicks. Fifty per cent, of the 

 chicks from no. 402 died before attaining the age of three 

 weeks, whereas but 13.89 per cent, of no. 168 's chicks died 

 within this time. There is no doubt of the great superior- 

 ity of 168 over 402 in breeding performance. The index 

 gives an exact measure of its degree or amount. 



3. The detailed figures bring out clearly the further 

 point that the value taken by the index is not unduly in-* 

 fluenced by any one factor. Low values for one variable 

 may be offset by high ones in another. In general, the 

 variables which form the numerator of the index fraction 

 are seen to increase as the index increases. The reverse 

 holds in general for the variables in the denominator of 

 the index. The facts regarding the two variables, a 

 (percentage of fertile eggs hatched) and c (percentage of 

 infertile eggs) are shown graphically in Fig. 1. 



The diagram shows that the percentage of fertile eggs 

 hatched in general increases as the selection index in- 

 creases in value, but with much greater fluctuations from 

 bird to bird. On the other hand the percentage of in- 

 fertile eggs in general decreases as the value of the index 

 increases, but again with much greater fluctuations. The 



