Page:Proceedings of the Royal Society of London Vol 69.djvu/31

Rh is 29. This he takes as evidence that the girdle migrates caudally between the stages of 32 '0 and 40 '0 mm. and then rostrally again between the stage of 40 -0 mm. and the adult ( (11), pp. 592, 593). To base such an important deduction on four specimens, whilst com- pletely ignoring the possibility of his 400 mm. stage being merely a variation, seems greatly to weaken the case for ontogenetic migra- tion of the pelvic fin in Spinax niger. Braus' argument rests on the assumption that no variations occur in the position of the girdle-piercing nerve in this species. Though I have never had an opportunity of examining Spinax, I have been able to make observations on a number of specimens of each of the following species *Mustelus vulgaris, M. Icfvis, Acanthias vulgaris, Galeus canis, Carcharias glaucus, Scyllium cani- cula, and Scyllium catulut*. In all these species there occurred variations of several metameres in the position of the girdle. Until, therefore, it has been definitely shown that Spinax is constant with regard to the position of its pelvic girdle, any deductions based on the assumption that no variations occur in this species are to be accepted with the greatest caution.!

As has already been stated, the idea with which this investigation was commenced lay in ascertaining the amount of variation shown on certain points in a given species from a given locality, and to discover how far these were explicable on either of the two rival theories the side-fold theory and the migration theory of the origin of limbs.J

It has been seen that certain of the facts are not in accord with the side-fold theory, without considerable strain on the imagination, though all are explicable on the theory of migration. Consequently the general tenour of the paper is on the side of limb migration.

with regard to the girdle-piercing nerve, post-girdle nerves, and whole vertebra: amounting to four metameres. This species is the common form in the Channel. I have never been able to procure M, Icevis, and according to Day ((2), vol. 3, p. 295) it does not occur here. At Naples, M. Icevis is the common form, M. vulgaris being much scarcer. I have previously given evidence to show that M, vulgaris is a more stable form than M. Icevis in the Mediterranean ((12), p. 342). We may look upon this as evidence for the greater variability of the more abundant form, or possibly we might regard the M. vulgaris of the Mediterranean as having acquired greater stability by the splitting off of a variable factor as M. Icevis. This, of course, IB mere speculation.
 * Twenty-three specimens obtained from Plymouth show a range of variation

t Since the above was written I have had the opportunity of examining a number of specimens of Spinax niger on this point, and find that here, as in all other Elasmobranchs examined, there exist variations in the position of the pelvic girdle amounting to several metameres.

t This paper is not concerned with the more precise origin of the free limb, whether it has been derived from gill-arch rays, according to Gegenbaur, or from external gills, as recently suggested by Graham Kerr (' Proc. Cambridge Phil. Soc.,' 1899).