254. MR. A, K. TOTTON ON 
I will now return to my description of the caudal skeleton of 
Pleuragramma untarcticum. 
One of the chief characteristics of this skeleton is the large 
size of the cartilages supporting the procurrent rays. No doubt, 
this 1s connected with the weakness of general ossification. The 
caudal skeleton of an adult specimen of Zrematomus newnesit, 
though generally similar to that of Plewragramma, differs in 
several points. As mentioned above, the general ossification is 
much stronger. Correlated with this probably is the smaller 
relative size of the cartilages bearing the procurrent rays. The 
epiural apophyses are three in number. The two terminal 
hypurals still retain something of their compound nature; finally, 
the penultimate centrum bears a well-defined double neural arch, 
but a single unexpanded hemal arch. The free end of the noto- 
chord extends further, and that of the nerve-cord not so far as 
in Plewragramma. 
In specimens of Pleuragramma antarcticwm of about 8 mm. 
(Pl. I. fig. 1), the tip of the tail and the notochord are bent 
slightly ventralwards, the angle thus formed being filled up by 
skleroblastic tissue (sk). - This latter consists of mesenchyme- 
cells, in which the cytoplasm has become clearer and the nucleus 
more distinct (7.e. takes up stain more readily) than in the cells 
of the surrounding tissue. The continuous fin-fold round the 
tip of the notochord is expanded and supported by actinotrichia. 
Whether this bending down of the tip of the notochord is appa- 
rent in the living animals or not I do not know. It certainly 
appears in large numbers of fixed specimens. One is reminded, 
in connection with this, of the prolongation of the vertebral 
column into the lower lobe of the vertical caudal fin of the 
Ichthyosauria. I think the condition I have described can 
searcely be due to fixation, because, since the skleroblastic tissue 
is denser than the tissue dorsal to it, the tendency of the fixative 
would be to shrink the upper tissue more than the lower, and so 
turn the tail upwards instead of downwards. I suggest that the 
presence of the skleroblastic mass under the notochord gives rise 
to some sort of stimulus causing this flexure. 1 should like to 
call attention to the fact that Ryder (U.S. Comm. of Fish & 
Fisheries, 1884, p. 1057), in describing the development of Alosa, 
says :—‘ Here the development of the hypurals is accompanied 
by a pressing inward of the ventral wall of the chorda.” He 
gives a figure (fig. 2, pl. i1.) to illustrate his point. I think that 
the illustration would be more correctly interpreted as showing 
a ventral flexure of the chorda, since the dorsal wall of it is also 
involved, 
Jn the next stage (-85 em.) (fig. 2), the skleroblastic cells in the 
concavity of the notochordal flexure have become divided into 
an anterior and a posterior mass, and a hyaline matrix has been 
secreted around and between certain of them, and in this way 
three blocks of cartilage are formed, two in the anterior mass 
of skleroblastic tissue and one in the posterior mass. Lepido- 
