6i 



in this position of the operculum there is an epidermic recess {op. rcc.) which opens into the 

 pharynx by the mouth («/.). I am thus inclined to believe that the mouth in fig. 93 of 

 Masterman's later memoir (98, 2) is the aperture immediately internal to the "ventral grooves". 



The epidermic recess which I describe is not in any way continuous with the ventral 

 groove of the pharynx, but it comes to an end, as it began, on the outer side of the mouth, 

 ceasing to be visible in the sections immediately after the mouth is passed. This indeed might 

 be anticipated from the sagittal sections shewn in Plate IV. 



The extra-oral "ventral grooves" are seen, in C. gracilis^ in the sections shewn in 

 figs. 66, 6; (PI. \T). In the case of this individual I have made a plasticine reconstruction from 

 the actual sections ; and there can be no question, from the examination of this reconstruction, 

 that the grooves are merely the junction of the operculum with the rest of. the body, and that 

 they have no relation to any grooves of the pharjnx itself. 



In fig. 3 of Masterman's memoir a depression on each side of the ventral wall of the 

 pharynx, immediately adjacent to the mouth, is identified as the beginning of a gill-slit. I do 

 not understand how the gill-slit is supposed to occur in this position. An inspection of figs. 

 152 — 157 (PI. XII) will shew that the first indication of the gill-slit appears in connexion with 

 the dorsal side of the pharynx, as indeed Masterman has himself pointed out, in insisting on 

 the close connexion between pleurochords and gill-slits. In fig. 153 the dorsal ends of the 

 pleurochords are seen as vacuolated tracts of the lateral pharyngeal epithelium, and distinctly 

 on its dorsal or posterior side. The exact point at which these will open into the diverticulum 

 is indicated in this section by a pair of minute depressions of the epithelium which are somewhat 

 nearer the posterior than the anterior wall of the cavity. In fig. 154 the pleurochordal grooves 

 are well established, while on the right side of fig. 155 the anterior end of the groove is 

 recognisable as the dorsal wall of the gill-slit. In fig. 156 both gill-slits are completely formed, 

 while in fig. 157 the vacuolated tissue of the pleurochords extends along the sides of the 

 pharynx, beyond the points where the gill-slits have been given off. The tissue gradually fades 

 away in passing along the sides of the pharynx, and it disappears completely some time before 

 the oesophagus is reached. 



With the exception of the point I have indicated with regard to the origin of the gill-slits, 

 all this is completely in accordance with Masterman's accounts of the pleurochords (97, 2, 98, 2). 



The close relation between the gill-slits and the dorso-lateral or pleurochordal grooves 

 of the pharynx is shewn in the reconstruction of C. levmseiii drawn in fig. 24 (PI. III). The 

 cavity of the pharynx is exposed by the removal of its anterior or ventral half, the whole of 

 the upper lip and of the pharyngeal diverticulum being left. In the lower part of the figure is 

 seen the posterior or dorsal wall of the pharynx, with a strongly convex surface, on either side 

 of which are the two pleurochordal grooves. These become much deeper in passing forwards 

 and they finally pass into the sides of the pharyngeal diverticulum, from which tlie tubes leading 

 to the gill-pores open. In the dorsal view of the same specimen (fig. 23) the wall of the left 

 gill-slit is seen curving outwards from the region of the diverticulum to open by its external 

 aperture. The same relations are illustrated by figs. i2i — 123 (PI. X). In fig. ^23, the dorsal 

 wall of the pharynx is convex towards the lumen of that organ, and on either side of it is one 



