438 BOULENGER. 



concluded, from the material studied, that the average length of the 

 head increases from north to south, which does not answer to Cope's 

 definitions, the specimens from Texas and some parts of Mexico being 

 referred by him to the brachycephala form in which the head is stated 

 to be 3| times in the length from snout to vent. 



The shape of the snout varies much individually, exactly as in 

 R. esculenta. The most acutely pointed snouts are in specimens 

 from Florida (var. sphenocephala) Mississippi, Texas, and Minnesota, 

 and broadly rounded snouts are to be found among specimens from 

 Lucknow, Ont., Cuttyhunk, Mass., Ithaca, N. Y., Texas, Presidio, near 

 Mazatlan, Jalapa, Nogales, Omilteme, La Cumbre de los Arrestados, 

 Guatemala, and Costa Rica. 



The development of external vocal vesicles does not seem to be 

 correlated with the distribution, although the fullest development is 

 shown in specimens from Long Island, Florida (var. sphenocephala) 

 and Texas; they are large in males from Lucknow, Texas, and 

 Nogales near Orizaba, small or absent in the others. 



I cannot reconcile the presence or absence of cross-bars on the tibia 

 with the habitat; these bars are present in two of the specimens from 

 Lucknow and absent in the two others, and they are present or absent 

 in the specimens from the northwestern parts of America which 

 should, according to Cope, be referred to the subspecies brachycephala. 

 The dark longitudinal band on the front of the thigh is absent in most 

 of the specimens from Lucknow which, for the length of the head 

 and the large external vocal sacs, answer to Cope's subspecies mres- 

 cens; 1 find it more or less well marked only in single specimens from 

 Lucknow, Cambridge, New Jersey, Eau Gallic, and Pensacola, Florida, 

 Brownsville and Duval Co., Texas, Bloomington, Ind., Wet Mountain 

 Valley, Colorado (8000 feet), and various localities in Mexico and 

 Costa Rica. 



In his description of the subspecies sphenocephala, Cope ascribes 

 to it 3 phalanges of fourth toe free, and I find it so in specimens from 

 Eau Gallic and Moon ; but in the two male specimens from Pensacola, 

 and in a female from Canaveral, which are surely referable to this form, 

 only 2 phalanges are free, as in most specimens from other localities. 



Miss Dickerson, who allows specific rank to R. sphenocephala, 

 mentions that the dark spot on the upper surface of the snout is- 

 constantly absent whilst it is usually present in the true R. halecina 

 {pipiens) ; but I note its very frequent absence in specimens from all 

 parts of the habitat. 



As a rule the yellow or pale green border to the dark spots on the 



