SCHMIDT, THE HERPETOLOGY OF PORTO RICO 187 



Number of scales between occipital and semicircles 1 2 3 4 



Number of specimens, Mona Island 14 18 8 



Number of specimens, Vieques 5 16 12 7 



The vertical rows of loreals in the same series are as follows : 



Loreals , 5 6 7 



Mona Island 27 11 2 



Vieques 12 22 S 



In the present series of Porto Rican specimens, adult males which 

 wholly lack the tail "fin" are frequent, and such specimens are even more 

 frequent in Vieques. Thus, of thirty-nine males collected in Vieques, 

 none have a high continuous "fin" like those of Culebra or Mona, twenty- 

 seven have a low serrated crest, about one-third as high as the diameter 

 of the tail, and twelve lack the crest entirely, having merely a compressed 

 tail with a denticulate row of dorsal scales. This is evidently the con- 

 dition referred to by Bernhardt and Luekten (1863, Vidensk. Med. 

 naturh. For. Kj0benhavn, p. 249), whose comment was inexplicable to 

 Stejneger (1904, p. 640) because he lacked a sufficient series from 

 Vieques. In going from Vieques to Culebra, the difference between the 

 tail crests of the males is very striking, and if they were not linked by 

 Porto Bican specimens they would certainly be regarded as distinct forms. 

 Thus, out of twenty Culebra males, only four have a crest as low as the 

 highest found in the Vieques specimens, and in the remaining sixteen 

 it varies from a height equal to the vertical diameter of the tail to twice 

 the diameter. Evidently we have an excellent example of the beginning 

 of the process of differentiation through isolation on islands of this 

 plastic species. The specimens from Desecheo present no peculiarities. 



Anolis cristatellus reaches an altitude of at least two thousand five 

 hundred feet, being associated with Anolis pulchellus on the deforested 

 hills near Maricao. It is evident that the differences in the distribution 

 of this species and A. gundlachi are due not to altitude, but to habitat 

 conditions, of which light seems to be one of the determining factors, 

 A. cristatellus being the species of open fields and roadsides, A. gundlachi 

 of the thickly planted coffee plantations and of the forests. 



The examination of one hundred stomachs yields the following in- 

 formation as to food habits: Empty, 22; unidentified insect remains, 15; 

 beetle remains, 20 (larva and adults; a species of Diaprepes very abun- 

 dant) ; Orthoptera, 16 (cockroaches, grasshoppers, and a single cricket 

 and mantis) ; ants, 10; caterpillars, 9; bugs, 5 (mostly heteroptera, one 

 large cicada); flies, 3; spiders, 3; vegetable matter, 9 (mostly brightly 

 colored seeds) ; vertebrates, 2 (Anolis sp.). 



