58 EDWIN S. GOODRICH. 
Like the ordinary oval cells, the thread-corpuscles are found 
both floating and attached by a stalk to the body-wall or septa. 
In the latter case the thread-coil is invariably situated on 
that side of the nucleus which is far from the point of attach- 
ment. Looking in the adult worm for possible stages in the 
development of thread-corpuscles, smaller cells can be found 
in which the coil is smaller relatively to the granular cell 
body (figs. 10, 11,d, e, f, and 12,6). Such small and appa- 
rently young cells are scarce, and generally attached. Turning 
now to very young worms, we find in them the thread- 
corpuscles scarcely less abundant than in the adult relatively 
to the whole number of ccelomic corpuscles. They are, how- 
ever, as a rule, flatter and more oval in shape (fig. 9), with 
more loosely packed granules and smaller thread-coils. In- 
deed, these refringent bodies differ from those of the adult, 
not so much in the circumference of the disc as in its thickness 
(cp. figs. 9, d, and 138, a, both side views). Similar shallow 
discs occasionally occur in the adult (fig. 10, a). It will be 
seen, moreover, that in what I have taken to be the younger 
stage of development the disc presents its edge to the nucleus 
(figs. 9 and 10) ; whilst in the later stages the nucleus is lodged 
on its flat surface (figs. 12, c, and 13). Whether the disc has 
during growth turned round or merely changed shape is a 
doubtful point. These facts lead to the conclusion that the 
thread-corpuscles originate from the coelomic epithelium lining 
the body-wall or septa, passing through stages essentially 
similar to those of the young corpuscles of the ordinary oval 
type, and are not formed by the development of a thread 
within full-grown oval corpuscles. The thread itself would 
appear to be formed at the expense of the granules, cp. (figs. 
9,10, and 13); but this appearance may be deceptive, and 
merely due to the increase in size of the thread alone. The 
chemical nature of the substance composing the thread is 
discussed in detail below; it differs widely in its chemical 
properties from that of which the granules are formed, and 
from chitin. 
On the function or fate of the thread in the ccelomic cor- 
