476 MR. A. W. WATERS ON 



vibracular seta lies. Exactly the same thing, though not so 

 marked, is seen in Scrtdpocellaria, for in several species the 

 A'ibraculum extends beyond the median line, as in S. macandrei 

 Busk, S. incu7^vata Waters, etc. 



As I told Dr. Levinsen that I did not find two chambers, 

 he kindly sent me some vibracula, skilfully separated, which, 

 however, only confirmed what I had seen in my own specimens. 



In the Scrupocellaridse, so far as I have seen, the ovaria a,re 

 large, situated near the distal end, and contain several ova which 

 are developed into large ova before they pass into the ovicell. 

 On the other hand, in Bugula, Bicellaria, etc., the ovaria are, at 

 the proximal end, lasually very near to the base of the Cfecum ; 

 they are very small with usually two small ova, and when still 

 extremely small, an ovum passes into the ovicell. It is very 

 interesting to find these generic differences in the ovaria, and 

 undoubtedly the form, size, and position of the ovaria will be 

 found to furnish useful characters in many species of Bryozoa. 



The direction of evolution of the Scrupocellaiidse seems to be 

 indicated in the articulation, and a comparison of the changes in 

 this family may help us to understand the Catenicellidse better. 



SCRUPOCELLARIA FEROX Busk. (PI. LXYIII. figS. 11-15; 



PI. LXIX. figs. 7, 20.) 



Scrupocellaria ferox Busk, B. M. Cat. Mar. Polyzoa, p. 25, 

 pi. xxii. figs. 1, 2, & 5. 



The avicularia vary considerably in size, being largest just 

 below a bifurcation, and smallest or wanting in the younger 

 zooecia. The avicularian chamber has the lateral projection to 

 which I referred and figured in S. mansueta Waters*, from the 

 Red Sea, and the long dorsal opening of the vibracular chamber 

 in the older zooBcia has a calcareous band across dividing it in 

 two (^see fig. 14). There is one vibraculum to a bifurcation some- 

 what directed towards the front, as in >S'. cyclostoma Busk, and 

 the radicle, which is hooked at the end, is not ringed as in 

 S. cyclostoma^ but there are only a few complete radicles in the 

 specimens. No ovicells occur on the Zanzibar specimens. 



There are about 24 tentacles. 



The rosette-plate into the vibracular chamber is at the base of 

 the chamber and has many pores ; as this rosette-plate is not 

 always veiy distinctly marked ofi^ these pores might be looked 

 upon as several plates. Each zooecium has its own lateral wall, 

 so that when prepared in Eau de Javelle they may separate. 

 Stained preparations show a band near the distal end bent 

 back upon itself (fig. 12 6). The contents are granular, with 

 hollow places at intervals, and proba.bly the function is the same 

 as in the bodiesf I mentioned in Bugula bicornis BuskJ, and they 



* .Tourn. Linn. Soc, Zool. vol. xxxi. p. 134, pi. x. fig. 15 (1909). 

 t Those on page 474 are compared with somewhat similai' structures in 

 M.flaqellifera Ix, Flustra ahijssicola Sars, and Crihrilinajiqularis Johnst. 

 X Result, du Voyage S.Y. Belgica, " Bryozoa," p. 21, pi. i. fig. 4 (1904). 



