1894.] Development of the Wing of Sterna wilsonii. 767 
latively of thesame size. Later (fig. 6) it becomes united with 
carpale II--III, the whole forming a single piece equivalent to 
the separate os magnum and unciforme of some birds. 
Metacarpats. The only metacarpal which requires notice 
is V (IV of many authors). This has been more or less per- 
fectly described by several students since its first discovery by 
Rosenberg (73). "This author describes it in the chick as a 
distal process of à common mass of cartilage which clearly 
contains two carpal elements, IV 4- V, since to it is also joined 
metacarpal IV. In the case of his figures there can be no 
doubt that this distal prolongation is a true digital element, as 
it is clearly homonomous with the other metacarpals. It is 
to be noted that according to Rosenberg this new metacarpal 
lies at a lower level than the others, being flexed towards the 
palmar surface. Zehntner (90) finds the same element in 
Cypselus melba, but existing there, as in Sterna, as a piece dis- 
tinet from the basal (carpal) element with which it is at first 
joined in thechick. According to Zehntner after 9 or 10 days, 
this metacarpal "geht... bei Cypselus einen vollstündigen 
Atrophie.” This is certainly not the case in Sterna, nor is it 
in those forms studied by Parker. Here it retains its discrete 
nature for sometime and in the fowl, toucan and cariama it 
even becomes ossified before its final union with the basal end 
of metacarpal IV. 
That this is a true metacarpal is, I think beyond question. 
Owing to the method of study adopted by Parker he failed to 
recognize its earlier conditions, and his observations, unsup- 
ported by other evidence might be interpreted, as has been 
done by several, in another way. However, the evidence ad- 
dueed by Rosenberg, Zehntner and myself, clearly removes 
this from the category of tendinous ossifications, the pisiforme 
and the like. 
Naturally the structures which I have described should be 
compared with those of the reptiles, but this to be at all ade- 
quate would require a detailed knowledge far greater than I 
possess. It isto be noted, however, that if, as contended in 
the next section, the avian “pollex” is not the first digit of 
the pentadactyle hand, a portion of the reasons adduced for 
