228 COI.ORATION IN I^EPTINOTARSA. 



more or less stable condition which exists in the germ plasm for longer or 

 shorter periods of time, through whose influence certain characters are taken 

 from the general phyletic backgound and speciaHzed in various ways to build 

 up the color pattern of the race or species. 



(6) The color phenomena of insects show, first, a general phyletic color 

 pattern which consists of the metamerically repeated homodynamous centers 

 of coloration and the wing stripes and bands ; and, second, ordinal, generic, 

 and racial types of color pattern, which consist of modifications and specializa- 

 tions within groups, superimposed upon the phyletic type. The phyletic type 

 of coloration belongs with other characters of the phylum, and its explanation 

 can be undertaken only when we know better the origin of phyla. The ordi- 

 nal, generic, and racial types are those with which we are concerned in present 

 day studies of evolution, and they are open to investigation by experimental 

 and other methods of research. 



In this discussion of coloration data and observations from several lines of 

 research have been brought together and used in building up the final con- 

 clusions and conceptions given in the last section of this paper. Coloration 

 is here shown to be a fundamental character of animals, and not an accident. 

 Moreover, it is found to be governed by laws as strict as those which control 

 all morphological and physiological characters of animals in general. It is 

 therefore a trustworthy character for use in experimental evolution, being as 

 capable as is any other character of giving accurate data as to the laws of 

 evolution. Finally, the conclusions reached herein may be used for the pur- 

 pose of more fully verifying and comprehending the facts and tentative con- 

 clusions set forth in the discussions of variation and distribution. 



Although in this study of coloration important and fundamental conclu- 

 sions are reached as to the laws governing its evolution, we in nowise derive 

 any valid basis for speculation as to the causes of evolution beyond that 

 already brought forward in previous chapters. The experiments in somatic 

 color modification do remove, to a large extent, if not absolutely, from the 

 range of possibility as causes of color evolution, the neo-Lamarckian factors — 

 the inheritance of somatic variations acquired through the action of external 

 stimuli upon the soma during ontogeny. Hence it is to the germ plasm in 

 the period before cleavage and differentiation begins that we must turn our 

 attention in the study of evolution. 



If somatic variation is eliminated as a cause of color evolution, the only 

 other possible explanation of the phenomenon that presents itself is that varia- 

 tions of permanence and utility in evolution arise in the germ plasm; but 

 whether these variations are due to selection, mutation, or to the direct influ- 

 ence of environment, is a m.atter for observation and experiment, and not for 

 idle speculation. 



