838 MR JAMES RITCHIE ON 



Budding is not common. Buds arise from the lower half of the stem dilatations as 

 mounds of ectoderm which push outwards between the muscle-fibres. They resemble 

 closely the stolon buds of Pedicellina, consisting of a thick layer of ectoderm, 

 protected on the outside by a thin cuticle, and including a gelatinous matrix with 

 many connective-tissue cells. The mature bud is always separated from the parent 

 stalk by a single chitinous septum similar to those which form the stem nodes. 

 In Barentsia ramosa (Robertson), not only do the buds arise from beneath the 

 muscular dilatation, but two septa intervene between each mature bud and the original 

 stem, enclosing between them a short segment apparently homologous with the non- 

 polypiferous segment of the stolon.* In one case only was the occurrence of a pair of 

 close-set septa observed in our specimens, and there they separated the basal dilatation 

 from the succeeding segment. The budding remains simple, and complicated branching 

 where secondary branches are frequent, such as that of Barentsia ramosa, is 

 altogether unknown. 



Calyx. 



The general structure of the calyx and the arrangement of the organs it contains 

 are similar to those of typical Pedicellinidse. The calyx is asymmetrical in position, 

 overbalanced towards the anal side, and the lophophore lies almost at right angles to 

 the long axis of the stem, at least in expanded specimens, for some obliquity is notice- 

 able when the tentacles are retracted. In expansion, the lophophore is circular and 

 broad, with a distinct, transparent, membranaceous ' ' collerette. " The tentacles vary 

 in number from ten to fourteen, but by far the most common number is twelve, which 

 occurred in 70 per cent, of the animals examined. 



Apart from the alimentary canal, the gonads, and other organs, the calyx is filled 

 with a hyaline, perfectly transparent, gelatinous matrix, throughout which are scattered 

 cells similar to those in the inner portions of the stalk. 



Muscle-fibres appear in small groups, most characteristic being the atrial sphincter 

 band running around the margin of the lophophore and containing few strands, which 

 probably vary in number from about eight to ten. An intestinal sphincter, consisting 

 of one or two fibres, lies in the wall of the intestine almost at the point where it and 

 the rectum join (fig. 1, sph. int.). A similarly placed sphincter has only been described 

 by Davenport in the fresh-water Urnatella gracilis,]' but similar fibres, one or two in 

 number, occur in my specimens of Barentsia gracilis and of B. benedeni from Ostend. 

 There are also present somatic strands, very few in number, which are attached at one 

 end to the anterior basal portion of the calyx, and at the other to the infero-lateral 

 portions of the oesophagus. These correspond to the " Seitenwandmuskel " described 

 in Barentsia macropus by Ehlers, } although the latter appear to be shorter, for 

 they are attached half way up the calyx. They may also correspond to part of the 



* Robertson, Alice, op. cit., p. 340. 



t Davenport, C. B., "On Urnatella gracilis," Bull. Mus. Comp. Zool., vol. xxiv., 1893, p. 11, pi. iii. fig. 18. 



X Ehi.eks, E., op. cit., p. 64. 



