218 THE THEORIES OF EVOLUTION 



Pictet observed certain cases in which the insects had 

 become so completely accustomed to walnut leaves that 

 they finally reverted to their original type. This 

 would indicate that a defective organic condition pre- 

 vailed, resulting from an imperfect adaptation, and 

 that the characters observed were mere symptoms of 

 that weakened condition. The fact that walnut leaves 

 produce a lighter colouring and obliterate the pattern 

 of the wings seems to point to the same conclusion. 

 Other experiments, however, show that a change in 

 diet does not always produce that result, and may on 

 the contrary deepen the colouring and make the mark- 

 ings more distinct. This occurred when Pictet fed 

 his larvae esparcet and dandelion. He also combined 

 the various abnormal diets by first feeding the larvae 

 walnut leaves and then keeping the next two genera- 

 tions on esparcet. The characters due to walnut 

 leaves persisted in spite of the change to esparcet. 



We can therefore suppose that there was in this case 

 a mere weakening of the organism. While there is no 

 positive proof of this, a doubt can be entertained as to 

 whether Pictet's experiments prove the transmissibility 

 of acquired characters. 



Fisher's experiments, also on butterflies, are less 

 open to criticism. He concentrated his attention upon 

 the influence of temperature. Pupae of Arctia caja 

 were exposed to a very low temperature, 8° centi- 

 grade; the butterflies showed certain abnormal modi- 



