BUDDING PROCESSES OF CEPHALODISCUS DODECALOPHUS. 523 



Such a condition is found in the stage with one pair of plumes, and the series of sections 

 here (figs. 78' to 84) described, lie between figs. 40 and 43. They are drawn by 

 camera under a Zeiss microscope with obj. F. In fig. 78 the mesodermic cells forming 

 a thickening of the anterior wall of the sinus are cut through, whilst the next section 

 (79) passes through the sub-neural sinus itself. In fig. 80 a few mesodermic cells are 

 cut through, apparently loose in the blood-sinus, but in reality they are of the same 

 nature as the cells seen in the next section, but cut obliquely. In fig. 81 is seen the 

 pre-oral sac invaginated from the pre-oral ccelome, and it still opens freely into the 

 latter. It appears to be filled internally with a homogeneous darkly-staining material, 

 which, in later development, thins out to form a uniform layer around the walls of the 

 pre-oral sac as in the adult, and then is indistinguishable from the chondroid tissue. 



Attached to the mesodermic wall of the sac are a number of cells (t.c.) which have 

 a characteristic form, being somewhat fusiform with the nucleus near the centre, one end 

 free and the other attached. They radiate in all directions except posteriorly, where the 

 head of the sub-neural gland lies. A few can be seen in fig. 82 attached to the posterior 

 edge of the pre-oral sac, which lies slightly over the sub-neural gland. In fig. 8 1 the 

 dorsal wall of the sinus is seen to consist of ectoderm alone, and the sub-neural sinus 

 can be traced without break into the dorsal sinus in figs. 83 and 84. In fig. 82 it is seen 

 to be limited on the left by the pre-oral ccelomic wall, and, owing to a slight obliquity, 

 on the right by the wall of the collar-ccelome ; in the following sections the collar- 

 ccelome is alone seen. In the latter figure may be noticed the sub-neural gland, free 

 in the blood-sinus, which surrounds it on all sides. In the sinus are a few cells, 

 either free blood-corpuscles or cells detached from the radiate cells. Such is 

 the peculiar morphology of these parts, and their functions are not easy to conjec- 

 ture. 



A stage differing little from this is found in fig. 60, in nearly median longitudinal 

 section. 



Subsequent development consists in an expansion of the pre-oral sac, which becomes 

 of somewhat irregular outline, in a loss of regular arrangement by the radiate cells, and 

 in a partial closure of the aperture of the pre-oral sac. The posterior wall of the 

 sub-neural sinus grows backwards across the aperture, and comes in contact with the 

 pre-oral ccelome immediately covering the head of the sub-neural gland. Thus, in the 

 adult, a median longitudinal section through this region gives one the impression that 

 the pre-oral sac is closed ventrally and has no communication with the pre-oral ccelome, 

 thus seeming to bear out the statement of Dr Harmer (2). The aperture, however, 

 ■secondarily closed on the middle line, is still open on each side. Thus, in the adult, the 

 pre-oral sac still communicates with the pre-oral ccelome by a pair of short channels, 

 which are easily demonstrable by longitudinal sections a little to either side of the 

 median plane. 



On the other hand, I have entirely failed to find any trace of a communication 

 between the pre-oral sac and the dorsal blood-sinus, such as should exist if Dr Harmer's 



VOL. XXXIX. PART III. (NO. 17). 4 I 



