J60 CRIMSON NUT- CRACKER. 



less of the conic form than any of those now repre- 

 sented ; its shape, in fact, is intermediate between 

 that o&Dertroides and Coccothraustes to which it 

 leads. Coccothraustes, again, fig. 3, has a remark- 

 ably strong upper, hut comparatively a weak under 

 mandible, and the commissure is regularly and gra- 

 dually arched from the base. Some of the species, 

 however, of Coccoborus have the under mandible 

 rather the smallest, so that by these birds f a passage 

 is formed from one to the other. Coccoborus is 

 nevertheless very different ; for in the greater part 

 of the species the two mandibles are perfectly equal, 

 fig. 2, and the margins, instead of being curved, are 

 sinuated ; the upper one, moreover, has a very small 

 but obvious tooth, which none of the other genera 

 possess. Now, to unite this genus (which is strictly 

 confined to the warm parts of America) with that 

 of Pyrenestes, there should be a species with an un- 

 usually large bill, yet furnished with a notch ; such 

 a bird would obviously unite in itself the characters 

 of both genera, and such a bird we accordingly have 

 in the Loxia Angolensis of Linnaeus, erroneously 

 supposed to inhabit Africa, but which we ourselves 

 shot in the forests of America. We have chosen to 

 illustrate this progression by the form of the bill 

 only, because this organ is the most obvious to the 

 generality of students, and will admit of more ac- 

 curate delineation; but this chain of connexion is 

 equally apparent in the variation of the wings and 

 feet. We have thus presumptive evidence of a 

 circular group. Let us now see whether this group 



