184 BULLETIN 114:, UNITED STATES NATIONAL MUSEUM. 



Arkansas and Oklahoma with an average of 188, and southern Mis- 

 souri with 190 show a condition closely like amaura, the average for 

 which throughout its range is 193. In east central Missouri, and in 

 southern Indiana and the adjacent portion of Illinois the averages are 

 200 and 201 respectively. This is closely like the averages for triangu- 

 lum in Indiana and Ohio, and in Michigan, which are respectively 201.6 

 and 202. The figures for syspila in Iowa and eastern Kansas are 208 

 in each case. These are high like triangulum in northern Illinois and 

 Wisconsin (202), and gentilis in Kansas (202). The point to be noted 

 is that on the three sides of its range where syspila meets a closely 

 allied form it shows in the number of its ventral plates a definite 

 approach to these respective forms. 



Since the dorsal scale rows vary somewhat with sex, and the 

 specimens available are very limited in number, geographic variation 

 in this character cannot be conclusively shown. The totals for the 

 whole range, however, are of interest for comparison with alhed forms. 

 The table shows that the general expectation of a greater proportion 

 of high formulae in females holds here. While in amaura about half 

 of each sex has the formula 21-19, in syspila only about one-seventh 

 of the males have this number; and, whereas, in the former only a 

 tenth of the females and half of the males have 17 rows at the end, 

 in this form about one-third of the females and three-fourths of the 

 males have this number at the end. This indicates a distinct dif- 

 ference between amaura and syspila and a decided approach to 

 triangulum. 



Differences in other matters of scalation between this form and 

 its near relatives are too slight to demonstrate change with locahty. 



In its color pattern, as in the number of its ventral plates, syspila 

 proves its close relationship to its three nearest neighbors. The 

 line of separation on the south is arbitrary at best. Here it is the 

 change from a ringed pattern to a saddled pattern that marks the 

 change from amaura to syspila, while on the north and east it is the 

 change from a saddled to a blotched pattern that indicates the change 

 from syspila to triangulum. Whatever may be our conception of 

 the origin of these forms, we must agree that syspila gives every 

 evidence of being an intermediate or connecting link between amaura 

 and triangulum. On the south, near the range of the former, it is of 

 smaller size, tends to partake of the ringed pattern of amaura, has 

 more black on the head, and only scant indications of the complicated 

 head pattern of triangulum. In Missouri the body is larger, the 

 saddles of red well-defined, long, and broad, with usually a ventro- 

 lateral series of smaller spots alternating with them, the black is 

 restricted to a black band across the posterior portion of the parietal 

 plates, forming an anterior border to the first whitish cross band, 



