50 
cular investment of the tube, which in this region is thicker, al- 
though of the same character as in the anterior part. 
The secretion of this part of the apparatus takes the form of , 
round transparent spheres (figs. 54 and 57). They are formed 
inside the cells and are again enclosed in vacuoles which some- 
times exceed the size of the cell in which they originated. Some- 
times they have the appearance of lying loose between the cells (fig. 
54) and are provided with a proper membrane. In other sections 
I found them yet unmistakably enclosed within the cell and in 
others on their way to be expelled into the central cavity. The 
spherule is then no more enclosed in its sac, measures about 0.018 
mm. and often appears to contain a few diminutive granules (fig. 57) 
In the sections further backwards through RR’ (figs. 32 and 55) 
the ciliated epithelium is no longer distinet, the cells being alto- 
gether filled with similar spherical globules but of much smaller 
dimensions. The contents of the subjacent cells nn’ are in these 
sections also partly transformed into the same tiny globules (fig. 55) 
which was not yet the case in the region where the section of fig. 54 was 
taken. These smaller globules, stained yellow after treatment 
with pierocarmine and secreted into the central cavity, now 
and then project like beadstrings out of the superficial layer of 
cells. The process of their formation appears to take place on 
the same plan as that of the larger corpuscles above described. 
This change in the character of the epithelium is wholly analo- 
gous with what v. GRAFF and Hansen notice in the corresponding 
tubes of Chaetoderma (branchial sacs, v. GRAFF; mucus glands, Han- 
SEN) which they characterise as a diffuseness of the protoplasm and 
as a loss of the cell-boundaries. At the same time it appears to 
indicate that the secretion is stronger, the more we approach the 
external opening. | 
Nothing of importance remains to be mentioned but that the two 
lateral tubes coalesce (fig. 32), the longitudinal folds persisting, that the 
muscular investment increases in thickness, whereas the lumen nar- 
rows rather considerably (fig. 34) and finally communicates with 
the anal cavity just below the opening of the rectum. There are 
