GENERAL DISCUSSION. 95 



The term "primitive" is a relative one and, in the working out of the 

 evolution of the color-pattern, much more value lies in the ascertainment 

 of the ancestor of the genus or subgenus than of a more remote ancestor. 

 There are not enough data at present and there probably never will be to 

 make the attempt to study the evolution of the color-pattern from genus to 

 genus profitable, and this is not attempted in this paper. The study of 

 evolution is served better by concentration upon the evolution of the pat- 

 tern within the genus, subgenus, or species, and this has been my task. 

 The distribution and number of spots differ in the different genera. These 

 differences I believe have arisen by the different ways in which the vittse 

 have given way to spots in the phylogeny of the genus. 



The beetles of each genus, with a few exceptions, where the unit is a 

 subgenus, are referable to one set of spots. They constitute a pattern- 

 unit. Within this pattern-unit the work of evolution is principally the loss 

 or confluence of spots. Only rarely are new spots added. Change or. 

 shape of elytron sometimes shifts the position. Otherwise the position is 

 very conservative. 



The reasons for believing the spotted pattern to be primitive for Hippo- 

 damia at least are (1) the wide distribution of the spotted species, and (2) 

 the narrow distribution and varietal nature of so many of the forms which 

 have deviated from the spotted condition. But so difficult is the change 

 of number and position of spots between the several genera that it is easier 

 to conceive of them arising by independent origins from some primitive 

 vittate condition. 



EVOLUTION. 



We have seen, in the preliminary discussions, that natural selection must 

 be very feeble in the evolution of the color-pattern of coccinellid beetles. 

 Since the pattern is for the purpose of association with the bad taste, if it 

 has a purpose, its highest utility would lie in constancy and idiosyncrasy. 

 The spotted pattern being in these beetles the commonest and character- 

 istic one, natural selection, in so far as it is operative, should favor this 

 pattern. If, then, we find any evolution away from this pattern, it must 

 have taken place either without the aid of natural selection or in opposi- 

 tion to it. There is occasion for some doubt as to the primitive pattern of 

 some of the genera and species. Let us consider, then, a case where the 

 evidence that the spotted pattern is ancestral is conclusive, viz, Hippo- 

 damia convergens and its varieties. Here we have a species with the same 

 spotted pattern that is possessed by several other species of the genus and 

 which is widely distributed. In this beetle the spotted condition gives way 

 to marked deviations from its original spotted condition in several ways in 

 different varieties (fig. 11). 



The question is, What has caused this evolution ? Recourse to natural 

 selection is debarred. The inheritance of somatogenic characters is open 

 to such grave questions about the conceivable mechanism as to be very 



