STUDIES OF THE MACEOCHIRES. 



379 



being satisfied with that (taken in connection witli a few other 

 salient characters), very often the rest of the bird's economy 

 has not been examined nor even taken into consideration at all. 

 Why the pelvis has not proved an equally valuable character 

 in the list of classificatory characters, is simply because the 

 systematist cannot so readily say pelvis 2-notched, pelvis un- 

 notched, and so on. Yet the pelves o£ birds, when carefully 

 compared, offer fully as good distinctive characters for taxo- 

 nomic purposes as the sternum. I have already pointed out the 

 fact thcbt the pelvis of a Trochilus is as different from the pelvis 

 of a Cypselus as any two birds' pelves can well be. Further, their 

 Sterna, when we really take all their characters into considera- 

 tion, apart from the fact that both happen to be unnotched, are 

 very differently fashioned bones. Both are unnotched, to be 

 sure, — but so are the sterna of some Petrels ! Were the fact that 

 the sterna of both Cypseli and Trochili are unnotched of any signi- 

 ficance, so far as affinity is concerned, then surely the remainder of 

 the organization in these birds would be more or less in harmony, 

 and not at the widest variance, as is the case ! AVhat I mean by this 

 is easily shown in the shoulder-girdles of the two types in question : 

 thus, the coracoid of a TrocMliis is a very uniquely-formed bone 

 (P. Z. S. 1885, pi. Ix. fig. 5), and very difi'erent from the great 

 majority of birds. In the Swifts the coracoid is like that of the 

 Swallow^s. Again, the scapula in Trochilus is unlike the corre- 

 sponding bone in a Swift : consequently, this being the case, I 

 attach little or no importance, so far as affinity is concerned, 

 to the fact that their furcula3 happen to possess some marked 

 resemblance. For we well know that this latter component of 

 the girdle is that which becomes modified in accordance with 

 the flight of its owner, while the coracoid can be far better 

 relied upon for any affinity it may show as a character amongst 

 forms more or less related. Swifts are birds of long-sustained 

 flight, Humming-birds are great fliers, and so are Albatrosses; and 

 were w^e to increase in size the os furcula of a Swift and a Hum- 

 ming-bird to the size of tlie bone in an Albatross, we should be 

 surprised to find how much they resemble each other. 



Seeing now how very difi'erent the thoracic and pelvic, or 

 really the trunk-skeletons of Swifts and Humming-birds actually 

 are, let us next examine into some of the organs and viscera 

 which they enclose. 



LINN. JOURN. — ZOOLOGY, VOL. XX. 29 



