J. W. H. Harrison 51 



alcoholic selection than those of the male — which is entirely in accord 

 with one's expectations after giving due attention to the protection 

 afforded to the gametes of the female for such a prolonged period of 

 their existence. 



That the whole of the zygotes yielded by the mating of an 

 alcoholised male and a control female were substantially alike, and that 

 those from the reverse pairing were not so, was actually capable of 

 proof. When exactly one-half of the first-named lot had pupated they 

 were weighed, as were also the representatives of the other lot then in 

 pupa. A similar course was then adopted when the other half had 

 changed. It was then discovered that, whereas the two halves of the 

 former agreed in the mean weights of the male and female pupae, the 

 second (as well as the third) portion of the other cross was clearly 

 inferior to the first as will be readily perceived from the following table : 



TABLE VIL 



Alcoholised Males x Control Females Control Males x Alcoholised Females 



Average Average Average Average 



Male Weight Female Weight Male Weight Female Weight 



First Portion, August 15th -164 grm -IGl grm -163 grm -ISe grm 



Second Portion, August 23rd -I'JS grm -198 grra -141 grm -177 grm 



Third Portion, ) All emerged ) All emerged ) .ioa~t« -mi m-m 



September 6th / •• by Aug. 23rd) by Aug. 23rd)' ^^*» 8^™ lagrm 



If this hypothesis is in very truth the scheme that forms the ground- 

 work of the observed facts then my insects, when the requisite allow- 

 ance is made for the additional phenomenon of parental selection, fall in 

 line with Pearl's poultry and not with Stockard's guinea-pigs, for in the 

 latter the destruction and injury of the germ cells of treated parents 

 had proceeded much further than in the germ cells of my Selenia 

 hilunaria males. This, too, serves to demonstrate why neither Pearl 

 nor I reared malformed progeny while Stockard did so. 



In conclusion, only one more point needs explaining, and that is the 

 great feebleness of the controls. This is quite consistent with one's 

 expectations. The genus Selenia, like many (but not all) lepidopterous 

 genera, is extremely impatient of inbreeding and soon dwindles in size, 

 productivity, and vigour under its influence. That parental and gametic 

 selection was able to counteract this only emphasises further the lessons 

 of the experiment, for it must not be forgotten that controls and 

 " alcoholists " alike had the same great-grandparents and grandparents, 

 and that, necessarily, their immediate parents were brothers and sisters. 



4—2 



