MORPHOLOGY OF THE GALLINACE^. 225 



p. 702, namely, " that the furcula presents a median prolongation is undeniable ; but 

 the following statements on two points will show reasons for supposing it to be an out- 

 growth derived from the clavicles only at a late date." For these " reasons," I must refer 

 the reader to the paper itself, and also to my old figures, but more especially to the new 

 ones, representing the extremely lacertian condition of these parts in Opisthocomus, in 

 which a long styloid Inter clavicle lies in front, between the thin lower ends of the clavicles, 

 and behind is bound closely to the flat lower surface of the sternum, exactly as in Lizards 

 (see op. cit. plates 9-11). 



One word more on this subject : the figures given by Miss Lindsay on p. 701 (fig. II., 1-1 

 not 5) represent the parts of the shoulder-girdle in such a non-differentiated state, that 

 any deduction from them is mere loss of time and labour. Those very diagrammatic 

 figures must of necessity represent, not formed parts, but masses of embryonic cells, which 

 will, in time, become, muscle, fibrous tissue, vessels, nerves, cartilage, and bone. 



X. — The Wing of Chicle and Fowl. 



In a recent paper on the Wing of the Fowl (Phil. Trans. 1888, B, plates 62-65), I 

 have described the structural changes that take place after the three normal digits of the 

 wing are formed. In plate 62, figs. 1 and 2 of that paper, and in this paper (Plate XXII.), 

 it is shown what a remarkable three-fingered hand the wing is essentially. Now the develop- 

 mental completion of this part of the skeleton takes place by the fusion of a consider- 

 able number of the elements that still remain after the suppression of the 1th and 5th 

 digits. That was the first step, and the most archaic form of bird known — namely Archceo- 

 pteryx — and some of the extinct Reptiles that come nearest that incipient-flying fowl, have 

 only three digits in the fore foot. This practical amputation of the hind part of the fore 

 paw is correlated with a very minute development, as we have just seen, of the shoulder- 

 girdle, which is the " root and foundation" of the limb ; part of it indeed — its proximal end ! 



In the 1st stage (Plate XXII.), the 1st and 2nd digits are perfect, each with its proper 

 number of phalanges, namely, two and three (or three and four, if we count the meta- 

 carpals, the proximal elongated elements of the manus). The normal number of carpals 

 for a three-fingered hand would be present if there were three in the proximal series ; 

 but the inner or radiale is seldom in two pieces ; it is in some types, and then we have 

 an intermedium also. 



In the 1st stage (Plate XXII.) the 1st of the distal carpals is ventrally displaced, and 

 lies on the inner face of the head of the 2nd metacarpal ; it is a small knob of cartilage 

 when just clearly seen, which is not the case until the 3rd stage (op. cit. plate 62. 

 figs. 3, 1, d.c. 1 ) *. 



After the primary parts are formed, there is a considerable amount of what appears 

 to me to be atavism. In the 3rd stage (op. cit. fig. 3, mc. 3 ') the 3rd metacarpal 

 has attached to it near its hind margin, on the flexor side, a small rod of cartilage ; 

 this is evidently the remnant of the 1th metacarpal. Also between the 2nd and 3rd 



* In figs. 1 and 2 of that plate, the soft cartilage inside the carpals is marked off as a 1st distal carpal (d.c. 1 ) ; 

 that is not correct, it is not defined until we get to the 3rd stage (figs. 3, 4), after 10 days of incubation. 



