DEPARTMENT OF EXPERIMENTAL EVOLUTION. 117 



Another immediate effect of alcohol is upon the fecundity. In one 

 set of experiments a male and a female rat that had been alcoholized 

 were mated and the number of their progeny was compared with that 

 of unalcoholized controls. The results are given in table 2. Since 30 

 pairs of alcohoUzed rats produced 108 young, while 29 pairs of un- 

 alcoholized rats produced 300 young, it can not be doubted that the 

 imbibition of alcohol causes, either directly or indirectly, a reduction 

 in the number of offspring born. 



A test was made also upon the learning capacity of the alcoholized 

 rats as compared with that of their normal brothers and sisters. The 

 training was given each morning before the alcohol was administered. 

 No summaries have been made from these data. 



A study of the second generation, which constitutes the main objec- 

 tive of the experiment, has been pushed rapidly ahead. A large number 

 of rats (see table 3) has been taken through a 3 months' schedule, but 

 no summary of the results can be offered at present. 



Table 3. — Number of rats trained November 1916 to August 1917. 



Total. 



Parents 12 



First generation offspring (alcoholized, 62; normal, 57) 119 



Second generation offspring (from alcoholized parents, 49; from control parents, 66) 115 



Second generation offspring (one parent alcoholized and one normal, 6; from control 



parents, 7) 13 



Total 289 



EFFECT OF STARVATION ON THE GERM-PLASM OF BEANS. 



Dr. Harris has continued his studies on the influence (or rather the 

 absence of influence) of the depauperization through starvation of the 

 earlier generations upon later generations of beans. This work is 

 nearly completed. 



ALTERATION OF THE QUALITY OF THE GERM-PLASM OF A POPULATION 



BY SOMATIC SELECTION. 



Experiments with Drosophila. — The question of the degree to which 

 the quality of the germ-plasm of a population may be altered by 

 somatic selection remains much mooted, despite Castle's prolonged 

 investigations of the subject. During the year under review Dr. 

 MacDowell has contributed to this topic in his paper, ''Bristle Inheri- 

 tance in Drosophila: II. Selection." Like Castle in his studies of 

 rats, MacDowell with Drosophila has selected for more and for less, 

 only MacDowell, unlike Castle, has dealt with a numerical trait — the 

 number of bristles on the back of this rapidly breeding fly. Castle 

 has selected his rats for only 16 or 17 generations; MacDowell selected 

 his flies for 49 generations. Castle believes his experiments prove that 

 by selecting parents with the greatest amount of pigmentation the 

 germ-plasm in the later generation tends to produce a more extensive 

 pigmentation than in the earlier generations and that this change may 

 go on indefinitely or until the rat is entirely covered with pigment. 



