104 



branching. It was, however, noticed before the specimen was damaged that branches were given 

 off from the 2" d or 3 rd zooecium of one side, or higher in the internode; and that two branches 

 might be given off by the same internode, on the same or opposite sides. The basis ra mi 

 is short, usually not reaching the zooecium on the proximal side of the one with which it is 

 connected. The internode is narrow, the two zooecia of a pair being in contact with one another 

 medianly along the whole length of their "immersed" part. The peristome is of some length, 

 and is directed somewhat obliquely, so that the outline of the internode appears strongly serrate. 



The proximal part of the colony in the Japanese specimen (fig. 14) has so close a 

 resemblance to the 'Siboga' specimen that there is good reason for referring it to the same 

 species. The zooecia are, however, larger, and the basis rami is long enough to reach the 

 preceding zooecium. The specimen shown in this figure is an incomplete internode, which bears 

 two branches on the left side, foliowed by a third branch on the right side. An ovicell is 

 commencing to develop at the distal end of the principal internode. The remainder of the 

 Japanese specimens consist of isolated internodes, some of which are fertile. In one case three 

 branches arise from one side of an internode ; and some of the internodes cannot have had less than 

 30 zooecia. When seen in side view (fig. 1 6) the zooecia are observed to overlap one another 

 to a considerable extent, and in a way very different from that characteristic of C. elongata. 

 In the latter a zooecium does not overlap its predecessor on its basal side, but it extends for 

 a considerable distance on its inner side down the middle of the branch. The internode is thus 

 broad as well as being flattened. In C. cuneata the branch is of considerable fronto-basal depth, 

 and it appears narrow when seen in frontal view, in consequence of the way in which the 

 zooecia overlap. In some of the internodes of the Japanese material, obviously belonging to 

 the peripheral parts of a colony, the peristomes are much elongated. 



C. cylindrica was originally described by Busk *) from Tristan da Cunha. The present 

 species agrees with it in its narrow branches and in the median contact of the zooecia of the 

 same pair when seen in frontal view. The elongated internodes and the mode of branching 

 seem also to be very similar in the specimens from the three localities under consideration. It 

 thus apjiears possible that the 'Siboga' specimen and that from Japan should be referred to 

 C. cylindrica. But in the absence of ovicells in Busk's fio-ures, and taking- into consideration 

 the distance which separates Tristan da Cunha from the localities considered in this Report, I 

 think it would be unsafe to assume that the specimens here described belong to Busk's species. 

 I have therefore referred them to C. cuneata, in which the ovicell has been described, even 

 though imperfectly, and which comes from a nearer locality. The Japanese form referred by 

 Ortmann to C. cylindrica is insufficiently described and figured ; but in view of the practical 

 identity of locality in the two cases, the Japanese specimens in the Cambridge Collection may 

 fairly be considered to belong to the same species as Ortmann's. Waters ") has recorded C. 

 cylindrica, though with some hesitation, from the Sudanese Red Sea; and I should have been 

 inclined to regard this as belonging to the same species as the Japanese form if it had not 



1 i Bi >K, (1., 1886, p. 7, PI. II, figs 2, 2«, 4, 4<7. 



j Waters, A. W., 1910, p. 235, PI. XXV, figs 14, 15. 



104 



