TEIID LIZARDS OF THE GENUS CNEMIDOPHORUS 229 



The second distinction given by Stejneger (1894) "vvas based on 

 the supraocular granules, thus: " I find, moreover, that in 37 of the 

 40 specimens of V. hyperytlira from Cape St. Lucas the frontal 

 shield is in contact witli the second supraocular and often with the 

 Ihii'd as well, while in T'. heidingl the frontal is separated from all 

 of the supraoculars by a complete series of granules." This second 

 discrimination was considered the weaker of the two by Stejneger, 

 because he found the three exceptions in his series of V. hyperythra 

 from La Paz, southern Lower California (U.S.N.M. No. 12613). 

 Van Denburgh (1895), on the other hand, emphasizing the impor- 

 tance of the second distinction, gave heldingi recognition on this 

 single feature, and presented a table in support of his stand. Data 

 secured from 13G specimens Mere summarized in tlie table, and it 

 was shown that in these morphological intergradation existed. Cal- 

 culation indicates that it was only 8 per cent. Because of this "in- 

 tergradation," Van Denburgh reduced F. heldingi from specific to 

 subspecific rank, and it has remained there ever since. 



Later the same author (Van Denburgh, 1922, p. 559) presented 

 a more significant table, which was based on data taken from 332 

 specimens, 266 of F. hyperythra and 66 of F- heldingi. Of these, 

 65 specimens of hyperytlira are found to present the morphological 

 condition supposedly characteristic of heldingi., and 5 specimens of 

 heldingi show the condition attributed to hyperythra. Calculation 

 from Van Denburgh 's figures shows that wath the increased numbers 

 of specimens examined the known percentage of variational overlap 

 was advanced to over 21 per cent. Thus, Stejneger's second 

 character distinction was shown to be even more unreliable than 

 was at first supposed. The examination of numerous specimens 

 shows that the supraocular granulation is relatively constant in 

 some widely separated localities but highl}^ variable in others. The 

 collection of more of these small lizards may change the known per- 

 centage of morphological overlap between the northern and southern 

 forms, but it can scarcely be expected to do away with it entirely. 

 Therefore, since inconstancy in the variation of the distinctive 

 characters is not confined to a point or line of contact between the 

 two diverging populations concerned, it seems impractical to retain 

 them longer as distinct units. Additional data bearing upon the 

 above subject may be found below in the discussion of variation. 



While examining type specimens of Cnemidophoms at the United 

 States National Museum, the writer found one. No. 64252, which 

 Miss Mary C. Dickerson intended to make the type of a new sub- 

 species of C. hypei^jtht^s. Her manuscript has never l)een pub- 

 lished and doubtless never will be. The specimen was collected by 

 C. H. Townsend at Castro Rancho in the region of Concepcion Bay 

 on the east coast of the mainland of Lower California on April 5, 

 1911. It was sent to the American Museum with other specimens 



