THE " TRITON " TUNICATA. 105 



A glance at Plate IX. of Keferstein and Ehlers' work suggests that the 

 specimens there figured may have been young, and the number of stigmata 

 shown (thirteen to fifteen in the dorsal row) is just about the number present 

 in the smallest " Triton " specimens (2 mm. long). Perhaps this may also 

 account for the great anterior extension of the dorsal rows of stigmata which 

 are represented as reaching in front of the 2nd muscle band, while in the 

 " Triton" specimens they were never seen in front of the 3rd (see PI. XVIII. 

 figs. 8 and 11). The ventral band, containing fifteen stigmata, is shown by 

 Keferstein and Ehlers extending to the front of the 3rd intermuscular 

 space, while in all the specimens which I have examined, it has terminated 

 some place in the 4th intermuscular space. Grobben"* speaks of forty-two as 

 the largest number of stigmata upon each side which he observed, Keferstein 

 and Ehlers t say that the number may vary from twenty-six to forty-three, 

 while the usual number in the " Triton " specimens was about seventy ! 

 Grobben also describes and figures J the series of stigmata as extending exactly 

 one intermuscular space further anteriorly than I found to be the case. As 

 they appear always to terminate posteriorly in the neighbourhood of the 6th 

 muscle band, it is obvious that there must be a greater number of stigmata in 

 each intermuscular space in the " Triton " specimens than in those from the 

 Mediterranean, and a comparison of my figures on theone hand, with those of 

 Grobben and of Keferstein and Ehlers on the other, shows that this is the 

 case. 



The bars separating the stigmata are covered in the usual manner with 

 ciliated cells placed in such a position that the cilia project across the 

 stigmata. These cells are not placed in a single row, as a surface view of the 

 branchial sac such as that shown in fig. 2, Plate XIX. might lead one to 

 imagine, but are placed in groups of four or five elongated cells placed closely 

 side by side § (see PI. XIX. fig. 3). This arrangement can only be made out 

 by viewing the bar upon which the cells are placed from the interior of the 

 stigma. An osmic acid preparation showed with a Zeiss x^-in. oil immersion 

 objective that these cells were nucleated and nucleolated, and had a striated 

 band upon the free edge, from which the cilia project (PI. XIX. fig. 4). At the 

 rounded ends of the stigmata the ciliated cells are very numerous, forming 

 many rows. They also change their character (see PI. XIX. fig. 2), and become 

 cubical, spherical, or polygonal in shape. 



The endostyle is always a well-marked feature in the ventral middle line 

 of the branchial sac. It extends from midway between the 2nd and 3rd 

 muscle bands anteriorly (PI. XVIII. figs. 7 and 11, en) to somewhere in the 



* Loc. cit, p. 16. f Loc. cit, p. 57. J Loc. cit., p. 16, and pi. i. fig. 1. 



§ Grobben has figured a similar arrangement in the case of the asexual forms of the same species 

 (Loc. cit., pi. v. figs. 34, &c). 



