433] COLORS OF TIGER BEETLES-SHELFORD 39 



1. The humeral lumile is usually without any enlargement at the 

 end suggesting au expansion in the place of spot Bj and j. 



2. The middle baud is withdrawn from the margin in all cases 

 and in only one case, figure 556 g, is there auy longitudinal extension. 



3. The angle of the middle band is always less acute and the for- 

 ward extension less pronounced. 



4. The longitudinal portion of the middle band is oblique to the 

 anal or inner margin (suture) of the elytron. 



5. The end of the middle baud is not hooked but rounded, and 

 rarely even parallel with the transverse bands. 



A close examination of the marking of the experimental individuals 

 show that there is correlation in all the respects in which the middle 

 baud is modified, in general the most oblique middle band is almost 

 withdrawn from the margin and shows least hook at the end. 



Figure 461 shows au unusual type of marking and of modification, 

 the most reduced marking in specimens of C. linibalis subjected to the 

 same experimental conditions as the tranquebarica shown above. The 

 usual type of modification which is quite general in experimental speci- 

 mens has the longitudinal portion of the middle band shortened. It is 

 also more oblique and thus less like the simple type. The middle bands 

 of these specimens approach those of the variety splcndida (Kansas). 

 They represent a more extreme modification of the simple type than 

 the experimental middle bands of specimens of C. tranquebarica. The 

 markings in two out of about twenty individuals (Fig. 461) surviving 

 the liigh temperature showed a sharp bend for\\ard. This is the reverse 

 of the usual tendency in the purpurea group but is a strong tendency 

 in some other species shown in plates XIII and XV. One indi^^dual 

 out of several hundred collected from the habitat in question, reared 

 as controls, and reared for ontogeny showed this character. Apparently 

 the tendency to respond by a sharp forward bend is little developed in 

 purpurea. 



Figure 463 a to d, 464 a to c, and 466 show the patterns resulting 

 from the high temperature experiment witli C. Iccontei while 466 a' to 

 c' and 467 u'', x', y', and z' show the conti'ol which survived and the 

 range of variation in a series of specimens collected from the same 

 area from which the larvae for the experiments were obtained. First 

 of all the high temperature experiments show patterns with reduced 

 markings. The markings shown in 463r; are joined in a way which rarely 

 or never occurs in the stock from which they were collected and which 

 is on the other hand characteristic of the varieties of this species which 

 occur on the Atlantic coast. Also 463rf shows a pattern which is smaller 

 in markings than auy that have ever been collected near Chicago, 



