150 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 9i 



indicated by variants of the several species. Only by sheer guess- 

 work can the process of divergence of these two types from some 

 common prototype be imagined. 



Evolution within each group is relatively clear, and follows amaz- 

 ingly parallel trends. 



In the hiscutatus group are two closely related sections, of which 

 guadruplex is the most primitive of one, paucimaculatus of the other. 

 Of these two species, the latter exemplifies a more primitive pattern, 

 but both have large blotches and identical ventral counts, and they 

 differ from each other only in subcaudal counts and in extent of 

 subdivision of the blotches. In paucimaculatus the spots are very 

 broad but are divided only across the middle by a light streak or 

 spot; in guadruplex they are also divided medially, but the light 

 streak has completely split each blotch, and each of the resulting 

 spots is again split medially, so that superficially guadruplex very 

 strongly appears to have double the normal complement of blotches 

 of the group. 



Modification of the pattern of guadruplex resulted in the develop- 

 ment of hiscutatus. This form differs from guadruplex only in its 

 pattern, which appears to have been produced by suppression of the 

 alternate blotches of guadruplex. That this was the procedure is 

 indicated by the fact that (1) the primary blotches in the northern 

 form are widely separated and number about half as many as in 

 guadruplex; (2) the spaces between the blotches in hiscutatus are 

 frequently occupied by narrow, interrupted dark bands, which oc- 

 casionally are of the same shape as the primary blotches (more or 

 less H -shaped, light-centered) ; and (3) these "secondary" bands (sup- 

 pressed primary blotches), if enlarged to the size and character of 

 the primary bands, would reproduce the pattern of guadruplex. 



The same process apparently has been followed in the section in- 

 cluding paucimaculatus^ with the production of lyrophanes^ lambda^ 

 and vandenhurghi. The most primitive pattern type among the de- 

 rivatives of paucimaculatus is, curiously enough, that of vanden- 

 hurghi (structurally the most highly modified species of the genus) , 

 which represents a phase intermediate between paucimaculatus and 

 lyrophanes. To explain, the first step beyond the pattern type of 

 the former is the production of quadruple blotches, or, in other words, 

 double the usual number of primary blotches (as in guadruplex). 

 The next step is suppression of alternate blotches; in vandenhurghi 

 about half have been suppressed (and accordingly the number of 

 blotches is distinctly higher than in paucimaculatus). In lyrophanes 

 nearly all alternating blotches have been suppressed, and secondary 

 bands are made evident betwen the primary blotches; sometimes one 

 or two of the alternate blotches are not completely suppressed but 



