i2; 



The ovicell is not unlike that of a typical Crisia in its general proportions, and in the 

 fact that the narrow proximal part expands gradually into the inflated part. In length the ovicell 

 corresponds with 2 1 /,, to 3 zooecial series of one side. It usually occurs at about the middle of 

 the unbranched part of a branch, but it may be found just on the proximal side of a bifurcation. 

 Other ovicells may be developed on the distal side of the first one. Thus in iii.D. the fertile 

 part of the branch bifurcates in one instance, and of the two branches thus formed one has 

 an ovicell in the middle of its length, and the other has formed a new bifurcation, one of the 

 branches of which bears another ovicell. In all cases observed the ovicell is simple, not bifurcating 

 with the branch. 



Most of the ovicells are either incomplete or broken; and I have not been able to 

 distinguish the ooeciostome with certainty. In fig. 1 it is perhaps the second opening from the 

 middle of the series of four tubes which constitute the second zooecial series from the proximal 

 end, on the right side of the figure. 



Two specimens from Japan, in the Collection of the University Museum of Zoology at 

 Cambridge, show considerable resemblance to the present form in their zooecial characters. But 

 there are differences in the ovicells, particularly in the existence of a long ooeciostome projecting 

 laterally on one side of the ovicell and curving distally, which make it doubtful whether they 

 should be referred to the same species or not. 



M easu remen t s , in a: 



Width of branch, fig. 2, 300 ; 



Width of branch, just before bifurcation, fig. 3, 500 ; 



Length of ovicell, fig. 1, 1,300; 



Diameter of orifices, about 100. 



The specimens here considered are both from deep water (275 — 469 Metres), and may 

 represent an abyssal form of the species from shallower water which I have referred to T. 

 atlantica. The description, unaccompanied by a figure, which was given by Pourtales would 

 have been insufficiënt by itself for identification-, but the specimens figured by Smitt, who 

 regarded I. Jïexnosa as a form of /. atlantica, have a close resemblance to those dredged by 

 the 'Siboga'. Busk's var. tennis of the same species seems indistinguishable from Smitt's form. 

 In the biserial character of their zooecia the 'Siboga' specimens resemble I. gracillima Busk ') 

 (nee Ortmann). Busk states that the outer zooecia are the longer; but, from the examination 

 of the slide labelled "Adventure Bank, Porcupine, 92 fathoms", in the British Museum I do 

 not feel convinced that this is correct. The specimen consists of two minute fragments which 

 are probably more or less worn. The median zooecia are much more distant from the basal 

 surface than the outer zooecia, as in T. atlantica. It is to this assemblage of species, in which 

 the median zooecia are at the same time the most raised and the longest, that Busk's specimen 

 seems to me to belong-. 



From the locality in which they were found, the 'Siboga' specimens might be supposed 



1) Busk, G., 1875. p. 14. PI. VII, figs 5. 6. I. gracillima Busk is antedated by /. gracillima Retiss, 1S69, "Pal. Stud. iilt. 

 Tertiarsch. Alpen", II. Denkschr. Ak. Wiss. Wien, math.-naturw. Cl. XXIX, p. 2S2. 



128 



