girdle is not homologous throughout the members of the same species,
but that two-thirds of the morphological elements of the girdle are
capable of undergoing some process of absorption. The capability of
being absorbed possessed by the two outer elements of the girdle (./;
and z) will not, however, explain those instances where the girdle is
asymmetrically situated. Take for example No. 64 (Table I). On
one-side (fig. 5 (a) ) the girdle is pierced by the 36th and 37th nerves.
On the other side it is pierced only by the 37th (fig. 5 (l>) ). Now on the
side-fold excalation theory ^he 37th and 36th nerves should bear the
same relation to the three elements of the girdle. This can only be so
if we assume that the middle element y has been dropped out. Conse-
quently, in order to bring the theory into line with the different
.variations presented by the pelvic foramina and the nerves piercing
them, we must assume that any of the three postulated elements of the-
girdle are capable of undergoing absorption, and, as has already been
mentioned, we must further assume that a girdle which is pierced by
one nerve is not homologous with a girdle pierced by two nerves.
Also that a girdle pierced by one nerve is not necessarily homologous
with any other that is also pierced by a single nerve (ef. fig. 4 (")
and (c) ).
If, on the other hand, we seek to explain such variations on the migration theory we are not met with such difficulties. We may assume that the pelvic girdle in one specimen is strictly homologous with the girdle in any other specimen. We may suppose that the presence of one or of two nerve canals is determined by the position in which the " Anlage" of the girdle happens to be laid down. Thus in fig. 3, if we assume, for the sake of argument, that the most caudal nerve shown in each diagram is homologous, type C would result, if the " Anlage " of the girdle were laid down, in the most caudal of the three positions shown. If the " Anlage " be laid down a little more rostrally we should get type B ; still more rostrally, and type A would result. Though it is possible that there may be some ontogenetic migration,* yet the facts are capable of explanation on this theory without the assumption of such a process. The matter has been dis- cussed by Waite ((9), p. 86), who considers that in Necturus "the abnormal position of the girdle represents the development of a w.w girdle at a new point." The idea there discussed is similar to that given here with the exception of the new girdle. There seems to be no valid reason for supposing that the girdle in all cases is not a strictly homologous structure.
VI. Variations iti the Number of Fin liuys.
The number of fin rays is subject to some variation. In order to ascertain whether there exists any correlation between the fin rays
- Cf. Brans (11), p. 586.
