42 Effects of Alcohol o?i Selenia bilunaria 



and the other to be enclosed in exactly the same kind of cage to serve 

 as a control and therefore minus the alcohol. The value of the controls 

 thus used was greatly increased by the fact that they were part of the 

 same family as the treated lot. 



The treatment was commenced soon after the ova were deposited, 

 those to be alcoholised being shut up in airtight glass-topped tin boxes 

 with a soaked sponge, and the others put into a similar box without 

 alcohol. 



The sponge was charged twice per day, once in the early morning 

 and once at 8 p.m. Even after hatching, during their first instar, the 

 larvae were kept in the same tins, but as they grew larger they were 

 removed to bigger cages, and the dose of alcohol was simultaneously 

 increased. 



After pupation the treatment was still continued ; be it ever so 

 small, respiration takes place in the pupae, gametogenesis is being 

 completed, and therefore the germ cells are capable of being influenced 

 at a most critical time. 



Throughout their early stages, both lots were provided with exactly 

 the same food, hawthorn (Crataegus oxyacantha), and at the same time. 

 Further, since the mortality (as will be seen below) was greater amongst 

 the " alcoholists," whenever the cages were cleaned out the number of 

 individuals in the control cage was reduced so as to correspond with 

 those still alive in the other ; this step was, of course, taken to eliminate 

 possible errors arising from differences in the number of cubic inches of 

 cage room allowed to each insect. 



III. The Course of the Experiment. 



Although both sections of ova produced their full complement of 

 larvae about May 4th, the effects of the alcohol were very soon visible 

 when once the larvae commenced to feed, for almost immediately the 

 mortality rate amongst the treated individuals became very great, n» 

 less than fourteen ( = 31*l7o) dying in the first instar. On the other 

 hand, only two ( = 4*4°/^) succumbed in the untreated controls — a 

 number showing no sensible divergence from one's usual losses with 

 first stage larvae. At this point, as shown in Table I, the controls were 

 reduced to 31 so that once again the cages contained equal numbers. 

 With the weeding out of these presumably very susceptible specimens 

 the death rate in the intervals between the changing of the food and 

 the cleansing of the cage slowed up, 137o. ^^'^"L ^'^d l7-37o being the 



