48 JAMES Francis ABBoT, 
of its opacity, but in ©. willeyi they are very evident and the 
circulation may easily be watched with a low power. It may then 
be seen that the colorless elements of the circulating fluid are driven 
through the canals with considerable rapidity. The speed of the 
circulating current varies with the movements of the animal and, 
in the peripheral portions at least, is apparently controlled solely 
by them. The disposition of the canals in the center of the body 
is very difficult to make out in the living animal but toward the 
periphery they branch much more profusely and it is seen that they 
form an anastomosing network, ending blindly in finger-like pockets 
and not fusing to form a circum-peripheral canal. 
In connection with the canal system, dorsal respiratory 
tentacles are frequently to be observed, especially when the animal 
is floating ventral face upward, on the surface of the water (Fig. 2). 
These organs are superficially much like the cirri of Thysanozoon, 
and consist simply of hollow finger-shaped projections from the dorsal 
surface of the body, in direct communication with certain of the 
digestive canals. In C. willeyi these tentacles are cylindrical or 
slightly club-shaped (Fig. 1), whereas in ©. mitsukurii they end in 
digitate processes which give them a fringed appearance and pro- 
bably assist the animal in imitating the vegetable growth or the 
debris, in the midst of which it lives. The body tissue of C. willeya 
is so loose and the animal itself so ‘amoeboid’ in its movements, 
that it is impossible to tell, in the living specimen, just what 
arrangement these dorsal tentacles have, and it is only by the aid 
of a series of sections that their relations may be worked out. But 
in C. mitsukurii the body is firmer in texture and its shape more 
consistent, so that the general arrangement may readily be observed. 
It is then seen that the tentacles are arranged about the sense 
organ as a center and radiate from it in four directions. ‘There 
are 6 to 8 tentacles in each row, and the distal ends of the rows 
curve in toward the tentacular planes so that the four rows together 
assume somewhat the appearance of a figure 8 with the upper and 
lower curves broken. It is easily determined that these four rows 
follow four main channels of the digestive canal system. With a 
low power it may readily be seen that the fluid contents of these 
tentacles is in direct communication with that of the gastric canal 
immediately underneath, and the solid formed elements of the circu- 
lating fluid are caught and whirled about in eddies within the 
tentacles or at their bases. The very evident function of these 
