64 ILLINOIS BIOLOGICAL MONOGRAPHS [458 



meant by ortliogensis this group illustrates the orthogenetie priuciijle. 



The illustration above is concerned, however, with oul.y one of sev- 

 eral kinds of tendencies which appear in the group. Still another 

 Ijrinciple is suggested by the experiments. If extension of the unpig- 

 mented areas is indicated by the experiments with cold conditions dur- 

 ing ontogeny, which would be supported by geographic variation in 

 many species, one is forced to the conclusion that different kinds of 

 stimuli acting on the pattern mechanism produce different responses. 

 One type of response is the extension of the unpigmented ai"eas. From 

 an inspection of the figures it appears that this may take place on the 

 basis of a pattern in any stage of reduction. As a rule it occurs in cor- 

 relation with some marked change in the basal structures of the elytron 

 at least when the extensions violate the original plan of the pattern. 

 The mechanisms of pattern heredity and pattern development possess 

 the capacity both to respond to stimuli by changes in form and by the 

 extension of the unpigmented areas. This extension of the unpigmented 

 areas may take place in almost any form of pattern shown in the entire 

 series and may be concentric or in part linear. This is shown in plates 

 XII to XVIli and XXXVIII. The concentric extension at least would 

 seem to constitute a sort of reverse principle to that illustrated by 

 the changes in form resulting from my experimental conditions, such 

 as high temperature. In dealing with definite directions of response 

 which may be termed orthogenetie if desired, one must recognize pro- 

 gressive modification on the basis of a mechanism which may move in 

 any one of two or three or more directions under the stress of external 

 stimuli. Some evidence for a progressive series of modifications in 

 the same direction running through a series of species in the tiger 

 beetles is afforded by the experimental results. In general tlie pattern 

 of C. hirticollis is more angular and as a whole conforms to the original 

 groimd plan better than tliat of C tranquebarica. The modification of 

 the patterns (middle band) of C. hirticollis is in a direction toward 

 that of C. tranqueharica, but is not carried so far as are the modified 

 patterns of C. tranqueharica. C. limhalis is usually, in the less modi- 

 fied forms of middle band, about as far from the original angular type 

 as are the more modified forms of C. tranquebarica. Stimulation of 

 the mechanism of the middle band in limbalis at this stage usually 

 throws the band still further toward that of splendida or typical pur- 

 purea (Fig. .537, PI. XXVIII). Since the middle bands of the three 

 species differ normall.y only in the extent to which such oblique shift- 

 ing occurs, and each differs from the original plan to a greater degree 

 than the other, the peculiar character of the direction taken must re- 

 sult from a similarity of mechanism in tlie different species concerned. 

 Abundant evidence for stages in such shifting as fixed hereditary 



