78 FALCONID^. 



by their rufous tail-bands. On the other hand, Mexican examples have grey tail- 

 bands, and appear at first sight to be distinct ; hence Prof. Ridgway has separated the 

 northern birds as Bupornis griseicauda, and those from Cozumel, which also have 

 grey tails, as R. gracilis. 



We cannot admit the specific distinctness of B. ruficauda and B. griseicauda, for in 

 Costa Eica, Nicaragua, and Guatemala we find a perfect intergradation between these 

 two forms, some specimens having rufous tails suffused with grey, while others have grey 

 tails suffused with rufous. B. gracilis is a slightly smaller race of the Mexican B. grisei- 

 cauda, but Cozumel birds are matched by others from Yucatan and Southern Mexico. 



The measurements taken from our large series prove that no reliance can be placed 

 on differences of dimension. As a rule, the sexes attain an equal length of wing. The 

 males of B. ruficauda from various parts of Central America vary in this respect from 

 8"1 to 9"4 inches, while the females vary from 8"6 to 9'6 inches. The wing of the 

 grey-tailed Mexican race measures from 8'8 to 9*8 inches in the male, and from 8'9 to 

 9'9 in the female ; that of the Cozumel bird from 7'9 to 9*2 inches, while Bonacca 

 specimens reach to 95 inches. Thus it will be seen that, although the smallest male 

 bird we possess is from Cozumel, there are many others from the same island which 

 equal the dimensions of those from the mainland, and as our series shows every possible 

 gradation in size, we have not felt justified in separating them specifically. 



Young birds are undoubtedly darker than the adults, and show less rufous on the 

 wings. The tail-bands in immature specimens are six in number, whereas old birds 

 have only four or five light ones. 



As will be seen by the list of localities given above, there is scarcely a part of Central 

 America in which B. ruficauda does not occur, and from the dates attached to our 

 series of specimens it is apparently a resident throughout our region. In Mexico 

 Sumichrast ^^ states that it is found on both coasts, up to an elevation of 1200 metres, 

 where it is abundant. Mr. Witmer Stone says that in Yucatan it was the commonest 

 Hawk in the neighbourhood of Izamal ^4. 



In Guatemala it abounds in the hot coast-region, and is probably the most abundant 

 of the Accipitres. We obtained specimens from the plains of Zacapa, the Pacific 

 coast-region, the valley of the River Polochic, Choctum, in the forest-region of Vera 

 Paz, and the district of Peten i«. In Nicaragua B. ruficauda is abundant, according to 

 Mr. Nutting 28, who also found examples in the Gulf of Nicoya, in Costa Rica, where 

 it is the commonest Hawk of the country ^7. 



Although widely distributed, this species is not found in the high mountain districts, 

 but it frequents the lowland plains up to an elevation of about 2500 feet. Its food in 

 Guatemala consisted chiefly of snakes and lizards, though we have also found locusts 

 and centipedes in its stomach ^ i«. In Costa Rica, Mr. Nutting says that it prefers open 

 woods, although it also occurs in the thickest forests ^s 29. 



The nest and eggs have not been discovered. 



