140 BULLETIN 15 4, UNITED STATES NATIONAL MUSEUM 



and sciitellational identity of the two forms, and the fact that 

 gulans is central in position both in relation to the presumably 

 simpler southern stock of Cnemidofhorus and to the northern sexlin- 

 entiis and ferplexus^ the latter is thought to be a direct derivative 

 of gularis. Further support of this view will be presented in the 

 summary of the group (pp. 144-146). 



As pointed out under the discussion of the affinities of sexlineatus^ 

 specimens of that species approach jjerplexus in the panhandle dis- 

 trict of Texas and Oklahoma and the two forms probably intergrade 

 along a common boundary here or in eastern New Mexico. Further 

 collecting is necessary before this point can be definitely settled. 

 The differences between sexlineatus and 'perplexus are greater than 

 those between gularis and 'perplexus and involve scutellational as 

 well as colorational considerations, as shown in the diagnosis of />e?'- 

 plexus. Therefore, the Sonoran race runner seems closer to gularis 

 than to sexlineatus and apparently intergrades with the latter only 

 secondarily. The geographical position of the three subspecies and 



lalaialis 



perp\ex-us •* giglgiris . 



"h-u perxjthrus 



FiGUKE 24. — Diagram of the relationship of C sexlineatus perplexus 



the fact that both perplexus and sexlineatus intergrade extensively 

 w4th the central gularis, Avhich lies nearer to the presumed center 

 of origin of the genus, but only slightly with each other, tend to in- 

 dicate that the former subspecies, perplexus and sexlineatus, are not 

 direct derivatives of one another. Thus, again it is concluded that 

 they are derivatives from the central gularis. 



The relationship Avith deppii is not direct, as suggested by Gadow 

 (1906), but indirect and through gularis, so a consideration of this 

 point is not given here. 



As shown in the discussion of the affinities of lahialis, perplexus 

 approaches that species very closely at times and was apparently 

 derived from the same ancestral stock. 



An elaboration of the relationship existing between perplexus 

 and liyperythrus is reserved for presentation in the genej'al sum- 

 mary at the end of this work (pp. 251-260), since this is in reality 

 a consideration of the origin of the hyperytJirus group from the 

 sexlineatus group. 



A diagram of the relationships of perplexus is given below. 



