216 THE THEORIES OF EVOLUTION 



worms whose butterflies only attained the size of mi- 

 crolepidopters. Scanty nutrition or change to a less 

 favourable kind of food also retarded the various 

 physiological processes, such as the metamorphosis, 

 and lowered the reproductive powers. Those charac- 

 ters were transmitted to three generations after which 

 the observations were interrupted by the death of all 

 the butterflies. 



A hereditary influence is apparent in this case which, 

 however, could be rejected as evidence, for the modi- 

 fications transmitted are too general and amount really 

 to a weakening of the organism. It is not so much 

 that characters due to the influence of the environment 

 have been transmitted but that weak parents have bred 

 weak offspring whose weakness could not manifest 

 itself except through those very characters. 



While Pictet's experiments are similar, more accu- 

 rate conclusions can be drawn from them. They do 

 not bear merely upon a general condition of health but 

 upon one single physiological character. The larvae 

 of every species of butterfly feed on certain leaves ex- 

 clusively and cannot easily accustom themselves to any 

 others. To some larvae which feed usually on oak 

 leaves, Pictet gave lettuce or walnut leaves; they 

 adapted themselves to the new food with difficulty, but 

 after the adaptation had become perfect, they pro- 

 duced butterflies whose larvae-progeny accepted the 

 new food without repugnance. There occurred evi- 



