155 
which we assume will not be doubted, then we must con- 
clude that we have, in this clearly monophyletic group, a 
genuine case of fin migration, since the pelvic fin of 
Teleosts may be either abdominal, at the side of the anus, 
thoracic, just behind the pectoral fin, or jugular, in front 
of the pectoral. Further the fin must have moved from 
behind forwards, since it may be clearly deduced from the 
fossil forms that the abdominal position is unquestionably 
more primitive than the jugular. Now in Elasmobranch 
fishes, according to Gegenbaur’s hypothesis, the pelvic fin, 
having been formed from the branchial skeleton, must 
have migrated from before backwards, unless the addi- 
tional hypothesis of the extension of the branchial region 
further back than we have any knowledge, is evoked.t It 
must be noted at once, first, that all paleontological evi- 
dence is against such an assumption, and second, that this 
migration, if it occurred at all, must have taken place at 
a remote geological period. On the other hand, the 
migration of the pelvic fin of the Teleost is not only some- 
thing more than an hypothesis, but it must also have 
occurred within comparatively recent times. Now several 
attempts have been made to deduce the migration of the 
Klasmobraneh fin from its innervation, and so far the 
Teleosts have been ignored, in spite of the fact that here 
we might with much more reason expect to encounter 
such evidence. Unfortunately, however, in the latter 
group, the jugular pelvic, as in the case of the Plaice, is 
supplied by the anterior spinal nerves of its region, and 
the abdominal pelvic, as in the several fishes investigated 
by Stannius, is supplied by the posterior nerves of its 
} It may, of course, be maintained, as it is in the case of some Elasmo- 
branchs, that the pelvic fin of Teleosts migrated jirst backwards and then 
forwards. We have ignored the former possibility (which after all would 
be purely hypothetical), in order to concentrate attention on the latter, 
where the migration theory may be the more satisfactorily tested. 
