30Q DR. H. GADOW OX EVOLUTIOX [Mar. 20, 



A Key, or (jeneral definition of the main (jfoups of Mexican 

 and North- American Cnemidophoi'i. 



4 supraoculars. Collar compo.seii of large scales. 

 Postei'ioi- surface of foi'earin covered with 

 granules. U.S.A. into Northern Mexico. 



C. sexlineatns*, p. 302. 



4 supraoculars. Oollai' composed of large scales. 

 Foi-e;irm normally with scutes, or enlarged 

 polygones Central or gularis-yron}), p. 327. 



4 supi'aocixlars. Collar composed of small scales, 

 especially the rows which form the posterior 

 edge. Foi-eai-ni posteriorly covered with 

 granules North-Western or tessellatus-yroup'f, p. 367. 



3 supraoculars. Collar composed of large scales. 



Foreai'm granular Southern or dejij^ei-grouj'), p. 308. 



For comparison I give a few data of th.e South-American forms 

 of Gnemidophoras. The fii'st 5, C ocellifer to C. longicauda, are 

 closely allied to each other, and are all from Southern South 

 America, i. e. extratropical. They are distinguished by the very 

 small number of femoral pores ; the position of the nosti-il is in 

 the first nasal, ?". e. " anterior to the ;iasal suture," as in the 

 Mexicans (in the other South Americans the nostril lies between 

 the two nasal plates) ; when there are only 3 supraoculars, this is 

 due to reduction of the anterior scute, instead of the posterior as in 

 the Mexicans. In this respect, therefore, the mere number is no 

 indication of relationship with the f?e^;j9ev group ; it is rather the 

 reverse. 



We observe the same differences in the shape of the collar, 

 some of the tessellatus, others of the gidaris type : with or Avithout 

 granules at the edge, and, e. g. G. leachi, with intermediate 

 conditions. The humerus is covered either with many small 

 rows, or a few large i-ow^s, followed by smaller rows. But all the 

 South- American species agree with the tessellatus and dejypei groups 

 in the entirely granular covering of the posterior side of the 

 foi'earm ; there being not one specimen known with scutes or 

 even enlarged polygones. 



Concerning the colour-pattern there seems to be the same 

 amplitude of variation from adult species with many sharj^ly 

 defined stripes to those in which the stiipes fade away, or are 

 dissolved into round spots, mostly also with field-spots, e. g. in 

 G. lemniscatvs ; lastly there are some which attain a monochrome 

 condition. 



* Hereto possibly also C. hi/j^erythrns, distinguished by the fusion of the two 

 frontoparietals into one plate, p. 307. 



t Hereto possibly C. Jahialis, distinguished by the nasal opening being in contact 

 with the second upper labial, p. 374. 



