PAIRED FINS 75 



vated from the dorsal roots (p. 3 and Fig. 1). In the head region 

 although the visceral arches may be supplied with epi- and hypo- 

 branchial muscles derived from myotomes, yet the great bulk of 

 the musculature of these arches belongs to the lateral -plate 

 mesoblast. It is true that the trapezius muscle, attached to the 

 scapula, is of similar origin and is supplied from the vagus, but it 

 does not penetrate into the fin ; and, at all events, in the pelvic 

 ergion there is no trace of other than segmental muscles. Still 

 more inexplicable on Gegenbaur's theory is the position of the 

 girdles with regard to the nerves, blood-vessels, coelom, etc. For, 

 whereas these lie outside the gill-arches, in the case of the girdles 

 we find not only that the relative position of the structures is 

 reversed, but that the nerves often actually pierce the girdle to 

 reach the fin. In fact, the girdles lie in the outer body-wall, while 

 the visceral arches lie in the wall of the alimentary canal. Finally, 

 this theory offers no explanation whatever of the striking 

 resemblance borne by the paired fins to the median fins in every 

 detail of structure and development — a resemblance so close that 

 it can only be supposed that they are organs of essentially the same 

 nature. Especially remarkable is the identity in structure of the 

 dermal fin rays (pp. 122, 212). 



Turning now to the rival fin-fold theory, it is found that if 

 difficulties in its application do occur, yet the greater part of the 

 evidence of embryology and of comparative anatomy is distinctly 

 in its favour. 



The paired limbs, especially the fins of fish, as already mentioned, 

 always appear as longitudinal folds. The folds may be very short. 

 The continuity of the folds from the pectoral to the pelvic regions 

 is not really an essential point. Possibly from the first the paired 

 fins, and indeed the median fins also, were discontinuous. Never- 

 theless, such facts as the great extension of the muscle-buds and of 

 the 'nerve-plexus' both before and behind the fins [46], the con- 

 centration of the fins, the frequent presence of a greater number 

 of buds in earlier than in later stages, the presence in some fish 

 (such as Pristiurus and Scyllmm: Dohrn [118], Braus [47, 50]) 

 of such buds in all the trunk segments, may be considered as 

 evidence supporting Balfour's view of the original continuity of the 

 folds. 



It has been convincingly shown (Thaeher [434], Mivart [300]) 

 that the various types of the endoskeleton of the median fins of 

 fishes, with more or less well-developed basal plates, or rays branching 

 from an axis (Figs. 48-50, 52), have been formed from a series of 

 primitively discrete segmental radial s (somactidia) by a concrescence 

 or fusion of their bases, often accompanied by concentration or 

 gathering together at the narrow base of the fin (p. 106). That the 

 similar manifold types of the endoskeleton of the paired fins have 



