106 BULLETIlSr 154, UNITED STATES NATIONAL MUSEUM 



tliis second variety is oljvioiisly the yoiiiiji- oi' (he lii'sl \ariety, 

 further comment seems unnecessary. 



C . inarkiimm Avas described from the Tres Marias Ishmds, off the 

 west coast of Xayarit, by Giinther in 1885. A dia<2;nosis was not 

 given in the ori<>-inal description, hut it was said tliat the caudals 

 are weakly keeled and that the keels run nearly ])arallel to the lateral 

 margins of the scale. A comparison of a large series of specimens 

 from the Tres Marias Islands and the mainland of Mexico indicates 

 that this character is valueless. The keels of all gidaris are much 

 weaker than those of fe^.'fellatus. Although the keels run " nearly 

 parallel to the lateral margins of the scales," they are nevertheless 

 (>blique, and in some cases decidedly so. It was this, perhaps, that 

 lead Van Denburgh (1897«, p. -103) to write that " A large number 

 of lizards from Mazatlan, San Bias, and Tepic seem to be identical 

 with Giinther's specimens from the Tres Marias Islands. Whether 

 they are also identical with Cope's communis and Peter's 7/iexicanus, 

 I have been unable to decide." Upon admitting mariarum to the 

 nuiinland. Van Denburgh specified tliat " the upper lateral light lines 

 are much farther apart than in (fulat-'ts^ causing the specimens to bear 

 resemblance to sexlineatusP Since he was familiar chiefly with the 

 gularis {=2}erpJ€xus) of the Sonoran region, it was only logical for 

 him to draw this distinction, which is, of course, useless here since 

 it does not separate gularis, as here defined, from maricurum. 

 Stejneger (1809) identified the mainland specimens as G. gularis 

 mexicanus, thus confining mariaruw. to the Tres Marias Islands again, 

 stating tluit " This is the only species of reptile hitherto collected 

 which is peculiar to these islands." He also called attention to the 

 " different keeling of the caudal scales " as the chief characteristic 

 of the form. The spots in these specimens are always more or less 

 ))oorly defined the yellowish middorsal streak is often unusually wide, 

 the upper lateral stripes are sometimes obsolete, and the tails of the 

 young are usually decidedly pink. l)ut all of these characters are 

 frequently seen in mainland specimens of gularis, and even of 

 pcrplexus as well. They are also significantly varied in marimmin 

 itself. Therefore, because of the slight degree of difference between 

 mariaru/i) and the bulk of guJa/'i.s, and because of its identity with 

 many mainland specimens, mariaruin is here reduced to the synonymy 

 of gularis, leaving Tanfilla nclsoni. recently described by Slevin 

 (1926, p. 200). as tlie one apparently distinct reptile from the region. 



Cncinidophorus hocourfii was described by Boulenger (1885) 

 from an indefinite locality. "California; Mexico." In tracing its 

 affinity, it seems important that hocourtii was mentioned in the 

 original description as sometimes having " traces of six longitudinal 

 lines on the back."' This and reference to an enlarged series of 



