A TELEOSTEAN FISH. 255 
trichia (/p.) have also been formed at the margin of these blocks 
of cartilage. 
Stage itl (1:05 em.) (fig. 3).—Darkly staining skleroblastic 
tissue “(sk' .) has now made its appearance above the notochord im- 
mediately dorsal to what will later become the ‘ventral hypural” 
(a.). It is significant to note that just at this time the notochord 
is becoming straight again. This secondary dorsal flexure is 
started, I suggest, by some stimulus due to the appearance of this 
new mass of skleroblastic tissue. More cartilage has now been 
formed ventrally, in front of the two pieces already described 
(Stage IT), in the anterior skleroblastic mass, which two pieces 
have probably fused to form a, fig. 3. This new cartilage consists 
of a pair of proximal pieces and a median distal piece. The 
former, 6, represent a hemal arch and the latter is a radial. 
That the proximal pair of cartilages represent, or are serially 
homologous with, the arches which appear later and more ante- 
viorly seems probable, because they correspond well enough in 
size and in the manner of their appearance. I must confess that 
1 am not quite satisfied that the proximal and distal parts are 
actually separated, as the appearance of the line of separation 
may be due to optical section. In any case, the fusion between 
arch and radial in the next anterior segment is quite clearly seen 
(figs. 5 & 6), so that the evidence for the fusion of the above- 
mentioned elements is presumptive. 
One may regard these cartilages, perhaps, as being in a con- 
dition intermediate between that of the hypural behind, which 
apparently is formed from the skleroblastic tissue as fused arch 
and radial, and the condition of the arch in front, where fusion 
of the originally separate elements can be seen as development 
goeson. I may say here that the term “hypural” is limited to 
those pieces of cartilage (which may later become ossified) which 
represent or are actually made up of fused arches and radials. 
The hypural (a, fig. 3), behind the one whose development I have 
just traced, is a good deal larger than this latter, and is somewhat 
rounded when seen laterally. It is beginning to form an anterior 
peg which later on passes through the arch formed by the 
proximal end of the hypural next in front, and lies just dorsal 
to the caudal artery and vein which run in the arch. The 
caudal fin is now beginning to be constricted off ventrally from 
the median fin-fold. 
Stage IV (fig. 4).—Examination of specimens 1:3 cm. long 
shows that the upturning of the notochord has gone on further, 
and produced from the straight condition seen in the specimens 
1-05 em. long one in which there is a distinct dorsal flexure. The 
epaxial elements (ep.), formed by the skleroblastic tissue mentioned 
in the last stage, lie in the concavity thus formed. The posterior 
hypural cartilaginous block (phy.) can now be seen to be made 
up of two partially-fused blocks, or, rather, the original mass of 
skleroblasts has secreted two masses of matrix, which are not 
completely divided the one from the other. Similarly, the hypural 
