56 W. G. RIDEWOOD. 
pinnules except the few youngest at the base of the plume-axis are slightly 
swollen at their extremities. The most distal pair of pinnules arise as a rule 
immediately to the proximal side of the end-bulb, and are comparatively short 
(figs. 383-35, plate 5); exceptionally they arise from the end-bulb itself (fig. 36). The 
longest pinnules vary from *5 to *7 mm in length in different individuals. 
The bulb at the extremity of the plume-axis of C. dodecalophus is a hollow 
enlargement, of the shape of two-thirds of a sphere, and with a wall composed of 
taller epithelial cells than those of the plume-axis. The cavity is continuous 
with that of the plume-axis, and is traversed by a few coelomic trabeculae. 
These end-bulbs have been described in detail by Masterman (22, p. 344, 
and 24, pp. 516 and 521) and Cole (2). The former author regarded the 
terminal swelling of the plume as a rudimentary monostichous compound eye, 
bearing a remarkable resemblance, both in appearance and _ structure, to the 
“branchial organs” found in the sedentary Annelids, such as Potamilla and 
Sabella, but he states in a later paper (28, p. 725) that he had already 
abandoned this view before the publication of the observations of Cole, who 
considers the terminal bulb to be a “rhabdite-battery,” composed of rhabdite 
“cells” resembling those of Turbellaria and Nemerteans. 
Harmer (10, p. 38) states that end-bulbs with highly refringent vesicles, 
similar to those of C. dodecalophus, occur in the buds of C. gracilis on the 
first pair of plumes, sometimes also on the second and third (p. 20 and p. 94), 
and that these vesicles probably disappear as the adult stage is reached. No 
end-bulbs or refringent vesicles occur in C. levinseni, nor in neuter individuals 
of C. sibogae, but the males of C. sibogae have two long arms, representing 
the first pair of plume-axes without pinnules, and these have closely-set refringent 
vesicles covering the greater portion of their length. On page 90 he suggests, 
with some diffidence, that possibly the vesicles “have the nature of reserve 
supplies of nutritive material, developed precociously in the young bud for the 
nutrition of the future testes.” 
The end-bulbs of C. hodgsoni are less truly spherical in their curvature than 
are those of C. dodecalophus, since the tall epithelium extends farther along the 
neural than the grooved face of the plume-axis, and thus present a lop-sided 
appearance when examined from the side (figs. 33 and 35, and text-fig. 1). The 
clear refractive beads of previous authors occur among the tall cells of the end-bulb, 
and are present in greater profusion in immature than in full-grown polypides. 
A minute examination of thin sections and of teased preparations suggests the 
possibility of the refringent vesicles of Harmer and the rhabdite “cells” of Cole 
being the globules of the tubarium of the colony in process of secretion, like the 
globules of mucus in the goblet-cells of a mucous membrane. When the secretion 
is forming, it is of course, surrounded by protoplasm, and if the bead of secreted 
material is large, it will bulge upon the surface, being covered by a thin pellicle 
