46 CALIFORNIA ACADEMY OF SCIENCES. 
The structure of the nephridium is considerably complicated. The nephridio- 
stome, found in front of the ventral sete, is unusually minute, and leads to a com- 
partively short duet, which connects with the body of the nepridium proper, at the 
point where the single-tubed outlet leaves the folds. This tube is not conyoluted (fig. 
59 G@)), but almost straight, very hard and solid, widening toward the base. When 
it reaches the folds it does not at once enter a tube, but forms a long, cylindrical, 
spongy mass (fig. 63), which extends the whole side or one-third of the length of the 
nephridium, before it assumes the proper shape of a clear canal (fig. 59, @) to @), 
and fig 64). In the beginning at (@), this mass shows no regular lumen, but a num- 
ber of irregular pores and tubes, which might best be compared to the inner canals 
of a common washing sponge (fig. 64, @),@,@)). In the center of this mass are 
imbedded the two parallel folds of the main nephridial canal (fig. 64, a@ and 6). At 
first the spongy tube is located principally above the two canals, but soon the mass is 
shifted and the canals become imbedded in the center of the mass, or very nearly so 
(fig. 64, G)toG)). The small connecting tubules at (2), which were at first so ir- 
regular, soon assume a greater degree of regularity at (8), while at the same time two 
longitudinal lumens are formed—one on the upper (@) and one on the under side 
(@),, fig. 64) of the two central canals (a and }). At first these canals (@) and @) 
are indistinct and irregular, but soon they assume the character of regular longitudinal 
canals (from () toG@), @ and @). From @) to @) these canals (@) and G) are 
connected by the transverse tubules, which completely surround the two central 
canals (a) and @). At G) the lower canal G@) becomes narrower, and the trans- 
verse ducts drop into the main canal (©), while the former lower canal @) assumes 
the character of a rather thick, epithelial lining. The general effect of this arrange- 
ment is, that, seen with a lower power, the spur between (¢) and () appears to con- 
sist of four distinct parallel canals, while from («) to () the fold contains only three 
parallel canals. At (¢) these canals become very crowded, the whole fold being nar- 
rower. The length of this narrower part varies, but generally already at (), or 
shortly before entering on the bent plexus at (®), the fold has regained its original 
width. At @) the canals turn; that is, @) connects with @) and the central canal 
folds upon itself, and for a short distance we have four almost parallel canals. At 
(3) the formerly central canal (from (3) to @), 2) and @)) leaves the lobe and crosses 
over to the other fold, and at () becomes the original canal @), which runs its 
course all through the two folds, at @) turning downward, becoming the lower canal, 
which again at @) becomes canal ©), from there on pressing backward and again 
forward from (@) to @), at which point it separates itself from the fold, and, running 
along the inner body-wall, forms the outlet duct opening at the exterior pore at (°). 
The connecting bridge between the two main folds, or rather between one fold and 
the free lobe of the other, between (@) and (®), is very narrow at (1), suddenly 
increasing at (4), gradually tapering toward (5). As regards the ciliation of the 
canal, it may be stated that it does not extend all through the tube. It appears that the 
canal is ciliated at places where the passage of the excretions is difficult. The narrow 
duct from the funnel to () is ciliated as well as the funnel itself. Again, the ciliation 
