BARBOUR AND NOBLE: LIZARDS OF THE GENUS AMEIVA. 421 



America was under water probably while this ancestral stock was 

 migrating, but a short period of emergence or the presence of an ancient 

 land mass joining north Central and northwest South America, but 

 lying to the westward of the present Middle America, would have made 

 possible the migration. This land mass, the existence of which we 

 have suggested, has been postulated by various other writers on other 

 grounds. The sinking of this area would then have left the A. u?idu- 

 lata allies free to distribute themselves in Central America as it as- 

 sumed its present form and also to reach the Greater Antilles while 

 they were in connection with Central America. The temporarily 

 isolated South American stock then spread widely through the conti- 

 nent and passed into Antillea extending to Haiti. Finally with the 

 completion of the appearance of lower Central America in its present 

 form, we find it invaded by the Ameiva ameiva types in the form 

 of ^. praesignis, while western South America received some immigrant 

 representatives of A. undulata, which on reaching this region so pecu- 

 liarly favorable for speciation in reptiles became transformed into the 

 curious and hardly recognizable A. edracantha and A. bridgesii. The 

 latter of these reached Gorgona Island off the Colombian coast. 



An alternative would have been to conclude that possibly the 

 genus arose in Antillea and spread to Central and South America, but 

 this seems hardly likely in view of the definite grouping of the species 

 about the two prominent mainland types. 



Two other stocks remain to be mentioned, which show a somewhat 

 anomalous condition. The maynardi-wetmorci-polops group does not 

 seem to show any very close relationship with the other species, and 

 we can only conclude that these three very distinct species all represent 

 chance survivors from some stock which once had a wider distribu- 

 tion, but which has completely disappeared. The other anomaly 

 is afforded by Ameiva hifrontata and its subspecies divisa. These are 

 not very dissimilar to Ameiva ameiva, but yet occur side by side with 

 other races which are probably more closely related to Ameiva avieiva 

 than either of them are. Whether these represent the survivors of a 

 primary unsuccessful elaboration of Avieiva avieiva itself or are the 

 remnants of some other stock, which in the same environment has 

 come to look much like Ameiva avieiva, it is impossible to say. One 

 gropes in the dark in treating all of this problem. It is even far from 

 easy to surmise which are the more primitive types, while, of course, 

 we know but little of skeletal variation within the group and there is 

 no particular object to seek it out when it cannot be applied to palae- 

 ontology. How sadly different are the opportunities for the mammal- 

 ogist and the herpetologist in essaying studies of this sort. 



