408 c. H. DANFORTH 
There are several considerations that may serve to explain 
this- discrepancy. Pearl studied the effects of alcohol acting 
over long periods, while in the present work it was desired to 
have the treatments extend over the shortest possible periods 
consistent with securing a statistically sufficient number of 
eggs. Since Arlitt and Wells ('17) have shown that in the 
rat testis cells in different stages of spermatogenesis are affected 
differentially by alcoholic poisoning, it may be surmised that 
injury done to cells in early stages of gametogenesis may not 
show any effects in breeding tests for relatively long periods. 
It is possible on this assumption that Pearl's data were taken 
mostly after the full effects of the alcohol had been established, 
while most of my experiments were stopped too soon to get the 
later results. The data from experiment 3 are in harmony 
with this supposition, for in C, which followed immediately 
after A, the percentage of infertile eggs was actually higher 
than it was during the time when alcohol was being administered. 
Another possibility that may be mentioned is differential 
susceptibility to alcohol. It might be inferred from Pearl's 
paper that one hour is practically the maximum time that a 
fowl can live in an atmosphere saturated with alcohol vapor. 
The males used in experiments 1, 2, and 3 were able to endure 
much more than that, but of these three males no. 28 showed 
unmistakable signs of being most severely affected. It may be, 
therefore, that although stronger dosages were administered, the 
actual physiological reactions were less except in experiment 3. 
That male no. 28 was not naturally infertile is shown by his 
subsequent history. During July and till the 10th of August 
he was mated to no. 21, and this hen gave no infertile egg till 
August 26, sixteen days after having been separated from the 
male. Aug. 10, he was mated to eight black Minorcas, the 
remnants of experiments 2 and 3. Eggs from this mating 
were saved beginning Aug. 19. Despite the fact that during 
much of the time the weather was unfavorable and the male 
was moulting, out of the 101 eggs laid only three were infertile, 
in marked contrast to the more than 27 per cent of non-fertile 
eggs during the experiment. Since the number of hens in the 
