TllK KITKCl' OF COLD ON DKVKLOl'.MKNT. 1385 



tliov iijivc been made in a variety of instances, the elirysalis which would 

 nonnaliy j)ro(hice tlie smnuiier type liaving l)een made in many instanees 

 (thougli neitiier uniformly nor always to a like degree) to yield the spring 

 type. The most extensive, complete and varied of these experiments have 

 been made in this country by ATr. ^V. II. Edwards, paiticularly (with 

 success) u[)on three s|)ccics, Pi>lygonia interrogationis, Piiyciodes tharos 

 and Iphit'lides ajax. These researches show that experiment cannot com- 

 pete with nature in the completeness and invariability of its results, but 

 can only etfect changes in the same direction and with some uncertainty. 

 Thus in the first and last of the species mentioned ai)ove, there is a difl'er- 

 ence between the two broods, not only in the coloring but in tlic shape of 

 the wings. The artificial ap[)lication of cold to the chrysalis has so far 

 proved unable to cause the anticipated change in the shape, but only in 

 the coloring of the wings ; the resultant is the summer type in shape, the 

 spring type in the more conspicuous though less deep seated characteristic 

 of coloring. The efi^ect in Polvgonia interrog-ationis has been to melanize, 

 in Iphiclides ajax to ali)inize the general effect of the coloring : in Phyciodes 

 tharos, where suffusion is not unknown in nature, the effect, when the change 

 of coloration is not complete, has been to imitate to a greater or less 

 degree such suffusion, and this has been most frequently the result when 

 the exposure to severe cold has begun before the chrysalis has hardened. 

 This would lead to the presumption that such instances in nature occur 

 when the change to puj)a takes place (as it ordinarily does, not) at the 

 close of the day, followed Ijy a colder night than usual. 



The actual temperature of the chrysalis indeed depends very much upon 

 the temperature of the circumambient air or of the object on which it may 

 rest, ne^-er varying from it many degrees when it is quiet (as chns'salids 

 usually are) nor indeed when it is active. Once when the air was 64°. 5 F. 

 I placed a delicate thermometer against the body of the chrysalis of Vanessa 

 cardui and found the chrysalis one degree warmer than the air. I annoyed 

 it, causing it to keep up a vibratory motion implying much muscular 

 activity for about a minute, and then the thermometer placed against it 

 rose another degree, to 60°. 5. On another occasion at 10 A. M., the 

 thermometer at 71°, I placed it against the hanging chrysalis of the same 

 buttei-fiy and it rose to 73° ; then I placed it against tiie chrysalis of Thorybes 

 pylades, which had been lying upon a cool piece of marble throughout the 

 night, and it fell to 70°. 5. 



There is a fiirther difference between the generations, extending in a 

 limited way to the different summer broods, in the period requii-ed for the 

 complete transformations, affecting all the preparatory stages. Cold in- 

 duces inactivity and all stages are retarded during cold or even cool 

 weather. There is much individuality in this respect and the differences 

 that appear in successive seasons are often marked, as every field entomolo- 



»74 



