No. 433.] SNAKES OF SOUTHERN MICHIGAN. 



helped me to imagine the steps by which crythrogaster might 

 have developed from sipedon, although it by no means follows 

 that such were the steps. The following ten are the stages 

 that I have selected, but they are of course arbitrary, and I could 

 easily have subdivided the sixty-three sipedons into a dozen color 

 varieties, would such a division have been of any service. 



VI 



Some brown and much black 



Brown and black 

 Brown and black 



Little brown ; much blackish slate 



Much slate on anterior half of gastrostege> 

 Much slate on anterior half of gastrost. 



Deep rufous on mid-ventral surface ; very 



Although these are such hypothetical stages, more than 

 half of them occur among the ninety snakes I have examined. 

 The following diagram will show at a glance their relative 

 abundance, but it will of course be borne in mind that the first 

 four stages, which include all of the sipedons, might have been 

 divided up into a much larger number of color varieties had it 

 been desirable. This would not, however, have affected in any 

 way whatever the great gap between stages 4 and 8. The only 

 purpose of this diagram is to show plainly that gap. 



Although, of course, this diagram is not really comparable 

 with that showing the number of gastrosteges, since we are 

 dealing here with a purely artificial arrangement and not with 



