40 ILLINOIS BIOLOGICAL MONOGRAPHS [434 



467 y representing the smallest, which makes the marking of 463 un- 

 doubtedly reduced by experimental condition. 



Figures 465 a to & show experiments in which the larvae pupae 

 were iced from the beginning of the pupal stage ; all either by remark- 

 able accident or through the effects of the experimental conditions show 

 the widest type of markings ; a third specimen was only slightly modi- 

 fied. In 465& the form of the end of the elytron is rounded in an 

 unusual way and the surface appearance of the entire body and the 

 elytron are different from the normal types. 



Figure 468a and w' show the type of modification occurring in 

 experiments on C. Jiirticollis. The middle band is modified as follows: 

 the hooks and angles are rounded, the transverse part which usually 

 turns forward and has a sharp angle as in 468 w' is oblique in the 

 opposite direction. These modified patterns are identical with those in 

 southern and western localities. This modification is of the same kind 

 as that in C. tranquebarica and C. 'purpurea. 



Thus it is evident that C. tranquebarica, Jiirticollis, and lecontei 

 may be modified in structure and pattern by high temperature during 

 the pupal and prepupal stages. Experiments performed on C. repanda, 

 lepida, and punctulata show no such modification, or pattern modifica- 

 tion of any other type so far as has been noted. Specimens stimulated 

 by a' temperature of 37 °C. in the fall and forced through the winter 

 were modified only in case of the specimens which emerged early, 

 January 1. Specimens which emerged in the spring earlier than the 

 normal were not modified. One specimen of C. Jiirticollis (Fig. 566, 

 PL XXXI) coming through without any winter was very much smaller 

 than the normal. A specimen of C. lecontei shown in color plate XXIX, 

 figure 556, was different in form, the abdomen being broadest at a 

 point not usual for lecontei. 



One of the patterns of tranquebarica produced in this way (Fig. 

 459) was one of the most striking modifications obtained. 



Thus so far as the species which show modification are concerned 

 the modification appears to be in definite directions and the modifications 

 of C. tranquebarica, C. Jiirticollis, and C. linibalis are in the general 

 direction in which the modification of the pattern plan has proceeded 

 in many patterns which have deviated from it in course of their evolu- 

 tion. The experimental results further show a basis for the interpre- 

 tation of the geographic variation of the group which is our next topic 

 for consideration. 



GEOGRAPHIC VARIATION OF PATTERNS 



C. tranquebarica, very widely distributed in North America, 

 (PI. XXII) shows great variation in color and markings, but the 



