420 EVOLUTION AND GENETICS 



ment of the tail. At the end of the experiment he produced mice 

 of the same kind as the original generation. This experiment and 

 other evidence furnish abundant proof that mutilations are not 

 inherited. 



Parallel Induction in Insects. A number of experim.ents with 

 insects have shown that conditions of excessively low or high 

 temperature or relative humidity produce melanism. The off- 

 spring of insects in which melanism has thus been produced have 

 in some cases also shown melanic color, and these have been 

 interpreted as due to parallel effects upon the germ plasm of the 

 melanic parents. The cases are open to a variety of interpreta- 

 tions, however, none of which is conclusive. In any case, if 

 parallel induction could be proved it would have little significance 

 since the effect on the germ plasm would be direct and the charac- 

 ters therefore not of the type significant in evolution. If parallel 

 induction were well established and frequent in occurrence its 

 standing would be different, for then we might conclude that the 

 soma and germ plasm were normally susceptible to the same 

 stimuli. 



Schroder's Insects. Schroder experimented with the caterpil- 

 lars of a willow moth which usually form a case by rolling the tip 

 of a willow leaf. By cutting off the tips of the leaves on which they 

 were allowed to feed, he forced a number to roll the sides of the 

 leaves in order to conceal themselves. These conditions were 

 repeated with a second generation, and of the third generation 

 of caterpillars four out of nineteen continued the acquired habit 

 even when given entire leaves. 



In another experiment he dealt with beetles which live on a 

 smooth-leaved species of willow. Their eggs were removed to 

 pubescent leaves for three generations and at the end of that 

 time he reported that the pubescent leaves were preferred even 

 when free choice of the two varieties was allowed. 



Detlefsen objects to Schroder's conclusions on the ground that 

 they are based on too few cases and that the normal range of varia- 

 tion is not taken into consideration. These objections seem sound, 

 but another question which he raises, viz., ''whether the modifica- 

 tions would persist or disappear gradually if a reasonable number 

 of additional generations were followed under normal conditions," 

 seems pointless. Either habit may be looked upon as positive. 

 The fact that one prevails under normal conditions may well be 



