38 Mr. A. R. Wallace's Remarks on the Value of 



cation would surely oblige us to consider the three groups of 

 only equal rank," This extract^ I think, proves that I both fully 

 appreciated the uriity of this group and accurately defined its 

 limits some years before M. Blanchard's publication ; for though 

 it is (in its separate form) altogether without date, yet he quotes 

 works in 1857 — a year after the publication of my paper*. 



No one can be more convinced than myself of the utility of 

 osteology, and especially of the sternum, in the classification of 

 birds, and I sincerely trust this great work may be brought to a 

 conclusion. I cannot, however, allow that osteological characters 

 are an all-sufficing guide. Like every other character taken 

 singly, osteology is a very uncertain and irregular test of affinity, 

 and is, moreover, in almost every case accompanied by parallel 

 external characters. Sometimes one sometimes another part of 

 the bird's organization has varied more rapidly, so that one 

 group exhibits the most striking constancy of a part which 

 in another group is subject to extreme modifications. The 

 sternum is no exception to this rule, and by following it alone 

 we should make the greatest errors in classification. For 

 example, the sterna of the Finches and the Flycatchers are 

 scarcely distinguishable, notwithstanding the great dissimilarity 

 in almost every part of the structure of these birds — their bills, 

 their feet, their plumage, their habits, food, and digestive organs. 

 On the other hand, the sterna of the several genera of the 

 Caprimulgidce difier from each other more than do those of the 

 most distinct families of the restricted Passeres. The Bee-eaters, 

 the Barbets, and the Woodpeckers, again, are three very distinct 

 families, which, in a classification founded upon all parts of a 

 bird's organization, cannot be brought in close contact ; and yet 

 their sterna, according to M. Blanchard, much resemble each 

 other. It is evident, therefore, that the whole structure of a 

 bird and its corresponding habits may be profoundly modified, 

 and yet the sternum may retain a very close resemblance to a 

 common form ; and, on the other hand, the sternum may undergo 

 important changes, while the general organization and habits are 

 but little altered. 



* M. Blanchard's paper was published in the ' Annales d. So. Nat.' for 

 185:). Sec Ibis, 1860, p. 93.— Ed. 



