12 



BULLETIN 114, UNITED STATES NATIONAL MUSEUM. 



Fig. 3,— Lampropeltis getulus splendida (U.S.N.M. no. 



22374). 2X NAT. SIZE. SHOWING NORMAL ARRANGEMENT 

 OF HEAD PLATES FROM THE SIDE. 



In nearly all forms the influence of sex upon the scale formula is 

 evident. While the extremes for the form are generally attained by 

 both sexes, more females possess the higher formulae and more 

 males the lower, and in some forms the sex can be told in many 

 cases by the formula alone. Thus in multicincta and pyrrJiomelaena 

 the formula 23-21-19-17 appears to be possessed by males only, and 

 in some of the forms of triangulum the formula 21-19 is possessed 

 in the great majority of cases by females and 21-19-17 or 19-21- 

 19-17 by males. The ex- 

 act situation for each form 

 is included with its descrip- 

 tion. As a general rule it 

 is common for the male to 

 have one less row of scales 

 near the posterior end of 

 the body than the female. 

 This is undoubtedly corre- 

 lated with the fact that the 

 body of the female has a greater diameter in this region, due to the ne- 

 cessity at the breeding season for containing the eggs. This is another 

 illustration of the fact insisted upon by Ruthven that the number 

 of rows of scales is correlated with the size of the body. 



The rule derived by Ruth- 

 ven for the garter snakes 

 that decrease in number of 

 scale rows is d to a dwarf-, 

 ing of the body is, in a gen- 

 eral way, borne out by the 

 present study. Khombomac- 

 ulata and elapsoides are ex- 

 cellent examples. However, 

 it does not appear that de- 

 crease in size of body is neces- 

 sarily accompanied by decrease in number of scale rows, or that the 

 latter is evidence of a decrease in size of body. Thus Tiolbrookij nigerf 

 and getulus are undoubtedly derived from splendida and their scale 

 formulae are decidedly lower, but it is certainly yet to be proved 

 that they are smaller snakes ; furthermore, conjuncta is in all proba- 

 bility a derivative of hoylii, but while its scale formula is the same as 

 that for hoylii, it gives strong evidence of being a smaller snake. It 

 seems quite evident that these facts are to be explained by a geo- 

 graphic change in shape of the dorsal scales. In the western forms 

 of the getulus group the scales are more elongate, in the eastern forms 

 relatively shorter and broader. 



Fig. 4.— Lampropeltis getulus getulus (U.S.N.M. no. 

 14140). IJ X NAT. size. Showing extra supralabial 

 scute interpolated between normal second and 

 third. 



