318 Alexander G. Ruthven, 



from the land they tlirow tliemselves lieadlong into the water, often 

 from heig'lits of from 25 to 40 feet. When they strike the water 

 they sink like a plummet, and are not seen again. In places where 

 trees overhang the water in numbers, particularly along; the San 

 Juan River, we would often cause a rain of iguanas by Walking* 

 through tlie jung'le along- the bank. Quite frequently they were 

 observed on the ground, but when surprised in this Situation they 

 at once took to the trees before throwing themselves into the water. 

 When surprised from the water we found that they would not always 

 throw themselves into the water but often rush back into the veg-e- 

 tation on the bank. 



The stomachs examined contain only vegetable natter, princip- 

 ally leaves. 



Ctenosatira acanthura eomjßleta (Bocourt). 



Not uncommon on the Hacienda de Cuatotolapam, but not ob- 

 served in numbers. 



The specimens obtained exhibit few diiferences among- themselves, 

 and do not agree with the descriptions and the specimens that I 

 have before me of typical acanthura. In the Cuatotolapam specimens 

 the dorsal scales are larger, and the spines on the caudal verticles 

 decidedly stronger than in typical acanthura, so that they agree with 

 the C. cycloides and C. completa of Bocouet and Boulenger's variety c. 

 The uniformity of our material and the fact that the specimens re- 

 fered to completa by authors have mostly come from southern Mexico 

 and Central America, while the typical acanthura material is mostly 

 from more northern localities, seems to indicate that the former is 

 to be recognized as a geographic variety. The data at hand at 

 least justify the tentative recognition of the southern Mexican 

 material as a subspecies. 



The habits of the individuals observed by us are so different 

 from the description given by Ditmaes^) as to suggest either that 

 there are racial diiferences or that individuals vary in habits in 

 different environments. Our observations agree with those of Sümi- 

 CHEAST -) in that we found the form to be decidedly arboreal, mostly 

 residing in the large trees. The latter frequently have one or more 

 large cavities and into these the lizards retreat when alarmed. Only 



1) Eeptiles of the World, p. 141 — 142, New York. 



2) in: Ann. Mag. nat. Hist. (3), Vol. 13, p. 502—503. 



