iv CONNECTING LINKS 43 



little digit is the thumb, and that the other two are Nos. 

 2 and 3 respectively, is unsupported by evidence, he 

 seems to me to be stating what is undeniable. When 

 he goes further and argues that they are Nos. 3, 4, 5, 

 he is flying in the face of facts. In the embryos of 

 the Swift and Tern several good observers have seen 

 a fourth unmistakable metacarpal on the ulnar side 

 {i.e. the side on which in our hand the little finger is) 

 and in the embryo of that extraordinary South 

 American bird, the Hoatzin, there is a remnant on 

 the same side of a fourth finger though the meta- 

 carpal has disappeared. There are, then, only two 

 alternatives : the surviving digits are either I, 2, 3, or 

 2, 3, 4. Now the Emeu has only one, the central one 

 of the three, and all analogy would lead us to believe 

 that this is the third of the original five and not the 

 second, since, when reduction proceeds very far with 

 the digits of birds' feet, or with those of the fore or 

 hind feet of mammals, they are lost, as far as can be, 

 symmetrically, not in lopsided fashion. And since it 

 is the ulnar side of the wing on which mainly the 

 strain falls in flight, it is not likely that all the weak- 

 ening would go on on this side and all the strength- 

 ening on the other. Moreover, in the embryo Hoatzin 

 there has been found beyond the so-called " thumb," 

 besides vaguely suggestive cartilage, a bone, small yet 

 solid and well defined, that may be a trace of the true 

 thumb that has disappeared. 1 In any case, the bird's 



1 See Leighton, Tuffs College Studies TIL, on " The Develop- 

 ment of the Wing of -Sterna Wilsonii," and W. K. Parker, 

 Trans. Zool. Soc, Part 2, April, 1891, on "The Morphology of 

 Opisthocomus Cristatus " 



