On the Longicorn Coleoptera of Tropical America. 21 



Falcons from Sardinia which can hardly be any thing but a 

 new species; for they differ from every other Peregrine which 

 I have ever seen from Europe, and more closely approach 

 the southern forms F. meJanogenys and F. nigriceps. The 

 Sardinian birds, however, differ from these as well as the 

 common Peregrine in the very strongly marked oval or tear- 

 shaped spots on the chest, and the very broad and closely 

 marked bars on the breast. They approach F, melanogenys 

 in having a greater extent of black on the ear-coverts, which 

 nearly meets the cheek-stripe along the whole of its length. 

 Both specimens are fully adult females, and agree entirely ; 

 they were shot by Mr. Brooke in April 1869 and April 1871 

 respectively. I have no doubt that the characters above men- 

 tioned will be found to be constant, and therefore propose to 

 describe the Sardinian bird as 



Fal(^ Brookeij sp. n, 



F. similis F. peregrino, sed statura paullo minore, facie laterali tota 

 nigricante, at pectore latissime nigro transfasciato distingiiendus. 



Hub. Sardinia {A. B. Brooke). 



Mr. Brooke has very kindly presented one of the typical 

 specimens to the national collection ; so that the species can 

 be examined by any one visiting the British Museum. The 

 measm-ements of F. Brookei (in skin) as compared with F. 

 peregrinus are as follows : — 



Long. tot. 



F. perec/rinns, 5 ad 19 -0 



F. Brookei, $ ad 17-0 



V. — Notes on the Longicorn Coleoptera of Tropical America. 

 By H. W. Bates, Esq., F.L.S. 



Subfamily Rsinotbaoin^. 



The '■'■groupes^'' corresponding to our subfamilies, under 

 which Lacordaire classed the genera of Longicorns, and of 

 which he established about eighty in the family Cerambycida3 

 alone, arc seldom distinguished by definite group-characters. 

 The rule seems to be that in each "groupe" modifications of 

 form appear which do not occur in the same conjunction in 

 any other ; but every single modification is liable to disappear 

 in some members of the "groupe." Thus there is a looseness 

 and uncertainty of definition in the classification of this family 

 which cannot be agreeable to rigid systcmatists ; but they are 



