Amphibiaus and Reptiles of Mexico. 299 



dwelt lipon. Biit so little work has been done lipon tlie liabitat 

 relations of the Middle American foi-ms that a siimmary of tliose 

 of the species stiidied will not be out of place. 



From the standpoint of the amphibians and reptiles seven major 

 habitats may be recognized in the region studied. (The moimtain 

 fbrests Fig. 1, were not studied.) 



Lowland forests. — The dense jungle that now occupies the 

 low ground. The trees are large and form a dense shade (Figs. 5, 9). 



Lowland forest Clearings. — These are grass-grown or thicket- 

 covered areas that have resulted from the Clearing of the lowland 

 forests. The cane fields are included here (Figs. 6, 10). 



Lowland forest ponds. — The pools that occiir throughout the 

 jungles during the rainy season. 



Rivers and lakes. — All the streams — San Juan, Hueyapam, 

 La Laja and Arroyo Negra — are included here with Lake Catemaco. 

 The animals considered characteristic of these bodies of water are 

 both aquatic and semi-aquatic forms (Figs. 8, 9). 



Savannah forests and thickets. — The areas of woodland on 

 the higher parts of the piain. These may consist of rather exten- 

 sive forests or be limited to a few trees or to larger or smaller 

 areas covered with low bushes (Figs. 3, 7). 



Savannah grassland. — The grass covered areas on the higher 

 parts of the piain and surrounding the patches of savannah forest. 

 Where studied the grass was kept short by grazing (Figs. 3, 7). 



Savannah ponds. — Nuraerous shallow ponds dot the savannah, 

 and these all contain water during the rainy season. The largest 

 poud investigated was the Laguna de Chacalapa, which is roughly 

 a mile long by one half mile wide and less than a meter in depth 

 except where a ditch has been dug at one end. 



An analysis of our material indicates that the liabitat distri- 

 biition of the amphibians and reptiles obtained is as shown in the 

 table on page 300—301. 



It is hardly necessary to point out that further work will modify 

 the liabitat distribution of the species as expressed by this table. 

 It is highly improbable that we found each species in all of the 

 liabitats in which it occurs, and there are, with little doubt, many 

 other species that frequent the different sets of conditions studied. 

 On the other band, it is believed that the general results will not 

 be vitiated, as particular attention was given to habitat study and 

 an effort was made to collect specimens of all species in each habitat. 



