R. C. PUNNETT 
317 
to that vertebra. iSoinetiiues it articulates with the interdorsal of vertebra x 
and the basidorsal of x + \ (Plate I, Fig. 1 c). In such cases it has been reckoned 
as belonging to segment x. Lastly, it may be carried mainly by a basidorsal, 
though to a slight extent by the interdorsals in front and behind (Plate I, Fig. 1 a). 
Here it has been regarded as pertaining to the segment to which the basidorsal 
belongs. 
5. Posterioi' spine. A uniform system of nomenclature has been adopted 
throughout, and what applies to the anterior applies also to the posterior spine. 
6. 1st girdle-pierci72g nerve (1st g. p. n.). The pelvic girdle is pierced by one 
or two foramina transmitting certain of the ventral divisions of the fin-nerves. 
When only one foramen is present it transmits the ventral division of the nervus 
collector. When, as is less frequently the case, two foramina are found, the 
anterior of them transmits the collector. The most caudal of the nerves which 
go to form the collector therefore passes down almost directly to the foramen, and 
it is this, the 1st girdle-piercing nerve, that I have used to estimate the position 
of the pelvic girdle with reference to the axial skeleton. 
7. The nervus collector contains in Spinax several branches lying along the 
lateral vein (cf. Braus, '98, Fig. 5, Plate XI). By injecting this vessel with 
osmic acid, in the way that I have previously described for Mustelus ('OO, p. 332), 
the smallest branches can be readily made out. 
8. As post-girdle nerves (p. g. ri.), I designate all such as run to the fin on the 
caudal side of the 1st g. p. n. Thus, if there are two foramina in the pelvic girdle, 
the nerve passing through the hinder one will be the 1st post-girdle nerve. It 
not infrequently happens that the fin br .nch of the most caudal fin nerve joins 
with the fin branch of the most caudal but one before entering the fin, thus 
forming a "posterior collector"; more rarely the "posterior collector" may be 
composed of branches from three nerves. Such a condition is brought about by 
persistence of the early embryonic state of things in which a posterior collector is 
normally present, as Braus has already shown ('02, p. 56G). 
9. The Jin rays were counted for the pelvic fins of a number of $s. A 
slight difficulty was introduced here by the fact that the last ray may be more 
or less fused with the distal part of the metapterygium. Whether it was more 
or less fused, or quite free, the last ray was always reckoned independently. 
10. The length of the whole and half vertebral series was determined in a 
number of c/s. The measurements were made from centrum to centrum. 
The observations made upon the various above-mentioned characters were 
tabulated separately for adults and embryos, and again for each sex apart. For 
each of the series so obtained the mean and standard deviation were calculated 
together with the probable error in each instance. The results are given in full 
in the Appendix, Tables 12 — 19. 
