4 BULLETIN 15 4, UNITED STATES NATIONAL MUSEUM 



lliiit arc based only on color or color pattern. A number of island 

 sLibsiiecies are recognized here, but none of these vary greatly from 

 their mainland i)r()totypes in sciitellation or in bodily proportions. 

 In these cases an apparently constant ditferenee in coloration is usu- 

 ally maintained by each form. Thus, two or more geographical 

 I'aces, which are obviously closely related, may show a really funda- 

 mental color divergence or evolution in spite of a continued uniform- 

 ity in other physical characteristics. 



The amount of evolution undergone is to be judged largely by 

 quantitative methods. If the color or pattern clilferences are great 

 and all color stages in the life history of each supposed form are 

 strikingl}^ distinct, the divergence is held to have full specific value, 

 but if these differences are small and show no great break from the 

 parent stock, it is evident that a full specific designation for them 

 would hide the true state of their evolutionary divergence. Popu- 

 lations belonging to the latter type are in reality "mere color 

 varieties," as they have frequently been styled, and in order that 

 their true value may not be overemphasized and their relationships 

 thus obscured, they have been given subspecific rank in the following 

 pages even though their free interbreeding be now prevented by one 

 or more water barriers. All of the island subspecies recognized in 

 this paper have apparently formed a separate population for only 

 a short time geologically, and the full species have apparently formed 

 an isolated colony for a much greater length of time. If the iso- 

 lation of these subspecies continues, many of them will presumably 

 become full species with the passing of another geological epoch, or 

 less. To give them full specific rank now, however, Avould obviously 

 be a purely theoretical anticipation, rather than a logical interpre- 

 tation, of nature as it exists to-day. 



At times overlapping color phases between insular forms appear 

 on intermediate islands. In such cases it is usually necessary to 

 abandon the name for the central group, retaining the two sufficiently 

 differentiated colonies on the neighboring islands as subspecies. The 

 placing of the synon3'm under the name of the nearest geographical 

 form has almost always been followed, although the actual basis 

 for this act is found, of course, only by a careful comparison of the 

 characteristics of the biological units concerned. 



In Cneinidopliorus^ an admittedly plastic genus, it is apparent 

 that colonies of the same general stock are much more homogeneous 

 on small islands than on larger ones, and that insular forms as a 

 whole show a smaller amplitude of variation than do mainland 

 types. The explanation of this may be reached, perhaps, by a brief 

 consideration of the environment. 



From this study it seems logical to assume that a modern wide 

 ranging species has met with and tolerated a large variety of habitat 



