OBSERVATIONS AND EXPERIMENTS 217 



dently some modifications, probably a chemical modi- 

 fication, which was transmitted to the butterflies and, 

 through their eggs, to the new generation of larva?. 



The larvae of Ocneria dispar feed habitually on oak 

 or birch leaves. Pictet fed them walnut leaves to 

 which they finally became accustomed although with 

 difficulty. The butterflies produced by those larva? 

 presented changes in the colouration and pattern of 

 the wings which became accentuated when several gen- 

 erations of larva? were kept on the same diet. In the 

 normal state the male is ash grey or brownish, with 

 four zigzag lines across the upper wings; the female 

 is lighter coloured, being greyish white or yellow, with 

 a less distinct pattern. By feeding the larva? walnut 

 leaves, one produces in the first generation smaller in- 

 dividuals of lighter hue, with fainter pattern; the 

 females become almost colourless. If the diet remains 

 unchanged these characters reappear in the second 

 generation. 



In another experiment (Pictet's sixth experiment), 

 the hereditary influence persisted in spite of a return 

 to the normal diet. The first generation was fed on 

 walnut leaves and presented the characters described 

 above ; the second and third were fed on oak leaves but 

 the characters due to feeding on walnut leaves per- 

 sisted. In the third generation the female showed 

 some signs of reversion to type such as slightly plainer 

 markings. 



