404 C. H. DANFORTH 



the same that the true value (to be obtained from an infinite 

 number of chicks produced under identical circumstances) 

 would fall above 43. This may be taken to mean that the 

 chances against the observed discrepancy being simply a chance 

 occurrence are more than 16 : 1. In 1-C on the other hand 

 the correspondence between observed and expected results 

 is extremely close. 



Polydactyly and color are treated in the same manner as 

 brachydactyly, it being assumed on the basis of the controls 

 that the normal incidence of Polydactyly is 36 per cent, and on 

 the basis of genetic literature that the incidence of white color 

 should be 50 per cent. 



DISCUSSION 



In the foregoing sections the purpose and conditions of the 

 experiments have been set forth and the data that they yielded 

 have been presented. We may now examine these data and 

 attempt to interpret their significance. The three characteris- 

 tics especially investigated were brachydactyly, Polydactyly, 

 and color. These will be discussed first. 



Brachydactyly. Table 4 brings out the fact that in each 

 instance the percentage of brachydactyly in A is considerably 

 in excess of that in C, the differences in the four experiments 

 being 7.5, 8.2, 10.9, and 17.9, respectively. The results of 

 either of these experiments taken separately would point strong- 

 ly to the conclusion that alcohol is capable of influencing the 

 percentage of brachydactyly, and the fact that all four of them 

 show such close agreement gives strong assurance that such 

 is the case. If alcohol were without effect, the departures 

 from the normal distribution would not be all in the same 

 direction and the sum of the data from the four A's should 

 show a percentage approximating that from the four Cs. Such 

 does not prove to be the case. In the C experiments there 

 were 464 individuals of which 39 per cent were brachydactyl, 

 while in the A experiments there were 369 individuals of which 

 48.2 per cent were brachydactyl. These figures are sufficiently 



