110 DK W. A. HERDMAN ON 



in which the intestine lias been turned ventrally so as to expose the whole 

 alimentary system) to about the middle of the intestine. At this point the 

 duct divides, and its two branches run over the wall of the intestine, subdivid- 

 ing as they go. The twigs branch freely and sometimes anastomose (PI. XX. 

 fig. 3). Many of them end in short caecal projections, and in some cases these 

 are enlarged to form terminal knobs (see PI. XX. fig. 3, c), which may contain 

 irregularly rounded bright bodies (concretions ?) similar to those described by 

 Chandelon in Perophora. 



The wall of the main duct is lined by regularly arranged fusiform cells 

 placed with their long axes parallel to the length of the duct (PI. XX. fig. 2). 

 The tubules on the intestine are lined by flattened epithelium bulging into the 

 lumen where the nuclei occur, and enlarged into cubical cells in the terminal 

 knobs. 



The apertures of the reproductive organs lie at the posterior end of the 

 body behind the alimentary canal, and usually in the 6th intermuscular space. 

 All the " Triton " specimens of Doliolum denticulatum examined belong to the 

 sexual generation, Keferstein and Ehlers' "generation A," and have both 

 male and female organs well developed. 



The ovary is an ovate mass placed usually in front of the 7th muscle band 

 (PI. XX. fig. 7, ov), but occasionally behind it (PI. XX. fig. 6, ov). Ova of 

 different sizes were almost always distinctly visible in it (PI. XX. fig. 1, g, and 

 fig. 13, ov). It opens on its dorsal edge into the atrial cavity. 



The testis, as Huxley * first correctly described, is in the form of a greatly 

 elongated tube, usually nearly as long as the body, terminating posteriorly on 

 the anterior face of the ovary, and extending forwards for a variable distance 

 with rather an irregular course (PI. XX. figs. 6 , 7, &c, and PI. XVIII. figs. 1-4). 

 At its posterior end, where it abuts against the ovary, it turns dorsally, 

 forming a tube which may be called the vas deferens, and opens into the atrial 

 cavity (PI. XX. figs. 13 and U, v.d.). 



The anterior end of the testis is very variable. Keferstein and Ehlers 

 state that it may terminate any place between the 3rd and the 1st inter- 

 muscular space, and they figure it at the posterior end of the 3rd in one case 

 and the anterior end of the 4th in another. Grobben states that it extends 

 forward to the 4th muscle band, while Huxley figures it as reaching nearly to 

 the 1st. In most of the specimens which I have examined the anterior end is 

 placed close to the 2nd muscle band, as shown in Plate XX. figs. 6 and 9. No 

 previous investigators, so far as I am aware, either describe or figure the extra- 

 ordinary variability in form of this anterior end of the testis. A glance at 

 figs. 6, 7, 8, 9, 10, and 11 on Plate XX. shows the extent of this variability. 

 In fig. the tube becomes rapidly smaller opposite the 3rd muscle band, 



* Phil. Trans., 1851, part ii. \>. 602. 



