OF THE CHORDA DORSALIS. 13 
and between the cells a small triangular interstice is seen, which 
is filled by a transparent fluid (not by air), or at least by a sub- 
stance which refracts the light in a different manner from the 
cell-walls, just as it is represented in plate I, fig. 1c, from the 
onion. 
Young cells, which float free, form within the cells of the 
chorda dorsalis, as in plants. They are, however, in the larve 
of the frog so transparent, that very favorable light and good 
instruments are required to see them. The number of cells, 
also, in which new ones are formed in the larvee is not great, at 
least in such as are to be had in the latter part of autumn. In 
the above-mentioned species of Cyprinus, and also in other 
fishes, they are, however, easy to be seen, and in greater number. 
Vesicles of very various sizes may be perceived in the cavities 
of many of these cells, and also in those of the larvee of the 
frog, though they are more difficult of observation in the latter ; 
a single one of these vesicles sometimes fills the greater part 
of the cavity; and occasionally several lie in one cell. (PI. I, 
fig. 4, 6, 6, c.) They are commonly quite round; but not 
unfrequently two are in contact, and flattened against each 
other. That they le free in the cell, follows from the fact, 
that they may be isolated without rupture. If, for instance, 
a small portion of the chorda dorsalis be torn into minute 
pieces, and a thin plate of glass with water be placed upon 
them, by moving this lightly backwards and forwards a few 
times, some such isolated vesicles may often be brought into 
the field of vision. They may then be made to roll about, and 
thus demonstrate their globular form. I have taken great 
pains to discover a nucleus in their walls, but without suc- 
cess. The young cells of the chorda dorsalis, also, in the larvee 
so often mentioned, have often the appearance, so long as 
they are not isolated, of possessing a nucleus: but one may 
readily be deceived here, since such a nucleus may belong to 
a cell lymg above or below them. Caution must also be 
used, not to confound a globular epithelial cell, which may have 
shpped into the chorda dorsalis in making the transverse sec- 
tion, with the true cells of that structure. I have not as yet 
been able, with certainty, to observe any nucleus, at least not 
of the characteristic form, in isolated young cells of the chorda 
dorsalis. In rare instances, a very small corpuscle, (d, d, of 
