R. C. PUNNETT 
319 
error of the difference, and is therefore doubtless significant. A comparison of 
these figures for the embryos reveals a similar condition of things with this small 
difference, viz. that the means of the two sexes are not quite so widely separated 
as in the adults. See Table 1 bis, p. 320. With regard to half vertebrae the 
embryos tend to show a rather larger number than the $ s. The difference however 
is not so marked as in the features just considered, and falls within the limit of 
the probable error. Consequently, whilst it would be hazardous to attach much 
weight to this difference, yet the fact that it is found equally well marked in 
both embryos and adults lends some support to the view that it may not be 
altogether without significance. 
In the cases of the 1st g. p. nerve and the post-girdle nerves, the differences in 
the means for the two sexes are so great that there can be no doubt but that we 
are here dealing with characters exhibiting a marked sexual dimorphism. The 
difference in the number of collector nerves is also nearly double the probable 
error of the difference and is therefore possibly significant. 
The fact that for both cases the mean of the embryos is in all cases less than 
that of the adults, coupled with the fact that these differences are probably in 
every case significant, would seem to denote one of two things. Either selection 
has been at work during post-embryonic existence, and in such a way that the 
factors concerned have operated in the same direction for both sexes, or else an 
increase in the number of segments takes place during this period. But these 
are points to which reference will be made later. 
The figures given in Table 1 show that a well-marked sexual dimorphism 
occurs, and that it cannot be due to selection, since it is almost equally apparent 
among the unborn young. Such sexual differences have been shown to exist also 
among other Elasmobranchs where considerable numbers have been examined, viz. 
in Mustelus laevis ('OO, p. 339), and in Acanthias vulgaris ('Ol, p. 24). From 
which it follows that in attempting to trace ontogenetic changes from a numerical 
standpoint, this factor must be taken into account as well as the variability of 
these structures ; and any attempt that neglects these factors must necessarily 
be of comparatively little value. It seems a plausible view that the sexual 
dimorphism which occurs in the Elasmobranchs is due primarily to the presence 
of the claspers which are found in the ^. The enlargement of the distal part of 
the pelvic fin necessitated by the development of these structures has led to a 
more rostral position of the pelvic girdle, which in its turn involves correlated 
changes throughout the meristic series. If there be any truth in the view that 
the claspers are the determining cause of the meristic differences in the sexes, one 
would be led to expect no such difference among the Teleostei, where such organs 
are absent. The only member of that group which, so far as I am aware, has 
been examined in this connection is Glwpea harengus. In his elaborate study of 
that fish, Heincke ('98, p. 95) has shown that there are no sexual differences 
with regard to the vertebrae or to the position of the fins, paired and unpaired. 
