REVISION OF THE KING SNAKES. 



105 



Since the forms are thus linearly arranged and related, evolution 

 may have proceeded from either extreme toward the opposite one, 

 or from some intermediate point in two diverging directions. Flori- 

 dana, hroolcsi, and conjuncta must be excluded from further con- 

 sideration as ancestral forms, for reasons already discussed, chief of 

 which are that they belong to regions that have received their faunas 

 from the mainland to which the peninsulas are attached and which 

 have thus not been centers of dispersal, and that they are very 

 clearly derivatives of the adjacent forms, getulus and hoylii, respec- 

 tively. Also, californiae and nitida are too obviously derived from 

 hoylii to receive consideration in this connection. There are left 

 therefore six forms to which to look for an approach to the ancestral 

 condition of the group. 



A comparison of the characteristics of these various forms may 

 give some clue as to whether evolution has proceeded from one coast 

 to the other or from some intermediate point in two directions. 

 Certain structural features will be considered first. 



Scale formulae of the forms of the getulus group. 



Form. 



Maximum. 



conjuncta j 23-25-23-21-19 



hoylii 23-25-23-21-19 



yumensis ' 23-25-23-21-19 



splendida ; 23-25-23-21-19 



holbrooki 21-23-21-19 



niger 21-23-21-19 



getulus i 23-21 



floridana i 23-21 



Muiunum. 



21-23-21-19 

 21-23-21-19 

 21-23-21-19 

 21-23-21-19 



19-17 

 19-21-19-17 

 19-21-19-17 



21-19 



Average. 



23-21-19 



23-21-19 or 21-23-21-19 



23-21-19 



23-21-19 or 21-23-21-19 



21-19 or 19-21-19 



19-21-19 



21-19 or 21-23-21-19 



21-23-21-19 



The table giving maximum, minimum, and average scale formulae 

 shows that the forms from splendida westward are identical in the 

 matter of scale rows, and that those east of splendida have a decidedly 

 lower range and average, with the exception of Jloridana. As ex- 

 plained in the discussion of getulus, the drop to 17 rows posteriorly 

 is rather common from Virginia to New Jersey, and since this form 

 must have reached this portion of its range from the south where 

 17 rows at the end is almost unknown, this must be considered a 

 secondary and not a primitive character. Seventeen rows occur most 

 frequently in holbrooki in the vicinity of New Orleans, and in niger in 

 the northern portion of its range, both places that would be more 

 recently populated whether the origin be placed in the southeast 

 or in the southwest. This indicates that in these places a reduction 

 in number of scale rows is taking place, and strongly suggests that 

 the formula 19-21-19 is also a result of reduction from a higher 

 formula. That evolution has taken place in this group in Jloridana, 



