56 EDWIN 8S. GOODRICH. 
fatty nature. Others are formed of more opaque and finely 
granular protoplasm, prolonged into fine branching processes 
of great length (fig. 28,.): corpuscles intermediate in character 
between these two forms also occur. 
The second variety is very characteristic of the Enchy- 
treide in general, and has been described by almost every 
writer on this group of the Oligocheta. Much larger than 
the amceboid form (nearly twice the size), these corpuscles 
are of regular and constant oval shape, somewhat flattened, 
thicker in the middle than at the periphery, and frequently 
bluntly pointed at one or both ends. The body of the cell is 
made up of large spherical, clear, and refringent granules, 
embedded in a delicate meshwork of protoplasm (fig. 15, a). 
The nucleus is round or slightly oval, in which case its long 
axis is invariably situated at right angles to that of the cor- 
puscle ; staining reveals a dark nucleolus lying in the less 
deeply staining nucleus, n (fig. 12, a). These corpuscles are 
found either freely floating in the ccelom, or attached to the 
body-wall or septa by means of a short stalk, apparently of 
cuticular nature. In the latter case the corpuscle is always 
fixed by its more pointed extremity. The remains of the stalk 
of attachment may generally be detected on those corpuscles 
which float freely (st., fig. 15, a). Although my observations 
as to the origin of these corpuscles are unfortunately neither 
complete nor conclusive, yet I am strongly inclined to believe 
that they arise from the coelomic epithelium lining the body- 
wall and septa. Certain it is that smaller, flatter, and younger- 
looking corpuscles of similar structure may sometimes be seen 
attached to the epithelium. The fact that similar apparently 
young corpuscles may occasionally be seen floating freely in 
the fluid does not seem to me fatal to the view that normally 
they remain attached to their place of origin until they have 
reached their complete development. Never have I found 
intermediate stages between these oval corpuscles and the 
ameeboid variety, or the granular cells of the modified epithe- 
lium covering the gut, from which they differ markedly, as 
already noticed by Vejdovsky (13). In the youngest worms I 
