Pectoral Fin o/'Eusthenopteron. 473 



Finally, tlic fourth piece of the axis is somewhat con- 

 stricted in the middle, and. quite distinctly bifurcated poste- 

 riorly (a feature not marked in the figure of A. S. Woodward, 

 18f 8). When looked at with a maguifying-glass, these two 

 posterior branches seem to continue in two separate ossifica- 

 tions, so that the composition of this fourth axonost of two 

 separate parts is not improbable, although not to be affirmed 

 with certainty, the separating line between the two being 

 perhaps due to a crack. One sees also with the magnifying- 

 g'ass the clear attachment of a dermal ray to the left of these 

 two bifurcations, while a fragment of somewhat crushed bony 

 matter attached to the right bifurcation also probably represents 

 dermal rays. 



There are three preaxial radials in our specimen. The 

 uppermost radial is attached to one of the two articulating 

 surfaces of the basal axonost ; it is bent inwards in the 

 middle and constricted posteriorly. The new preparation 

 shows the attachment of the dermal rays to this radial very 

 clearly. The second radial, attached to the smaller of the 

 two articulating surfaces of the second axonost, is also con- 

 stricted posteriorly, but not sufficiently preserved in its poste- 

 rior part. The third radial, better preserved than the second, 

 is constricted in the middle, but the limit of its posterior part 

 is indeterminable. It is attached to the smaller of the two 

 articulating bifurcations of the third axonost. 



There are only two postaxial processes in our specimen, and 

 no postaxial radials at all. The first process is a large pro- 

 longation of the basal axonost (this prolongation is not well 

 visible in the figure of A. S. Woodward, 1898), and the 

 second a prolongation of the third axonost, while the second 

 and the fourth axonosts are devoid of similar processes (on 

 the left side of the second axonost some bony matter is visible 

 in our specimen, but it is evidently a crushed scale). 



Having finished the description of the fin in question, I 

 will add some remarks concerning the problem of the origin 

 of the tetrapod limb. The resemblance of the internal skeleton 

 of the pectoral (and also of the pelvic) tin in Eusthenopteron 

 to the internal skeleton in the tetrapod limb has been empha- 

 sized by several authors (by Patten, Watson, Broom, Gregory), 

 and Watson especially has tried to point out in detail the 

 homologies of both (comp. Watson, 1913, p. 25 scq. and 

 figs. 1 & 2). But his restoration of the pectoral tin of 

 KuslhenopUrou (/. c. fig. 2) is wrong, inasmuch as he takes 

 no account of the posterior bifurcation' of the fourth axonost 

 (in this respect the restoration of Broom, 1913, p. 460, fig. 1, 

 is more accurate) and represents the postaxial process of the 



