390 



DE. E. W. SHrPELDT'S MOEPHOLOGICAL 



And this is the way, I suspect, that certain forms which we 

 now see in our modern Swifts were differentiated from the early 

 Hirundine stock. That this occurred early in the chapter of 

 avian life-history, for the world is old, may be conjectured 

 from the fact that Cypseli;are now quite cosmopolitan birds, 

 and, moreover, have many representatives among them which 

 present highly specialized organization. Even at the present 

 time, however, we yet have forms that structurally are nearer 

 the Swallows than others of the same group. To instance this, 

 we have but to glance at two such birds as Micropus mela- 

 noleucKs and Chcetura pelagica, in the first of which we still find 

 the general Swallow-like form of the body, the average depth of 

 the carina of the sternum, the non-pneumatic humerus, and other 

 points, all of which are far more Cypseline in character in the 

 latter bird. Chcetura, too, agreeing with other spine-tail Swifts, 

 shows its greater fixedness of characters in the very structures 

 which gives it its name, for the spines which terminate its rec- 

 trices are useful to the bird, yet can only have been developed 

 through ages of time. When we come to examine the still more 

 Swallow-like Swifts, Heoniprogne for example, and its allies, I am 

 sure we sliall meet with other points in their anatomy which will 

 lend support to this view of the origin of these types. 



In the present memoir I have, by extensiv^e and careful com- 

 parative investigations into structure, attempted to point out 

 how entirely diff'erent these Swifts are from the Humming- 

 birds, a group with which they have long been associated, to 

 my mind upon very meagre claims. During the course of my 

 present researches I have shown that Cypseli differ from 

 Trochili, (1) in their habits ; (2) in their nidification ; (8) in 

 the method of securing their food ; (4) in all their external 

 characters, and markedly in their external form ; (5) in their 

 pterylosis ; (6) fundamentally in their skeletons ; (7) every struc- 

 ture in their heads is as widely at variance as any two forms 

 of birds in the Class; (8) in their wing-structures; (9) in their 

 pelvic limbs ; (10) in their respiratory apparatus ; (11) in their 

 visceral anatomy ; and (12) in their digestive system. These 

 two groups have been associated together upon an entirely false 

 system of classification, which assumed first, that they are alike 

 in their wing-structure — a resemblance which I have shown to 

 be purely superficial ; secondly, that they both have an unnotched 

 sternum, although physiological law demands it, and when asso- 

 ciated with an entire organization that widely differs from that 



