BRYOZOA OF THE CAPE VERDE ISLANDS. 11 



6 oral spines and two or three delicate zooecial ones. The following series 

 have 4 distal spines and 2 oral. None of the subsequent zooecia have lateral 

 spines, while in the next series (circle) there are two zooecia with two distal 

 spines and one oral on each side ; whereas two other zooecia in the same 

 series have the flat spines characteristic of nitida and meeting in the median 

 line. Thus the early zooecia are distinctly Membraniporidan, in each row 

 approximating gradually to the mature nitida. 



There are several cases in the Cellularida?, and other families, in which 

 the zooecia immediately following the primary are intermediate between it 

 and the following zooecia. 



Hindis, Norman, and Levinsen have placed Membraniporella under 

 Cribrilinidse, and Norman* has written a lengthy description of the bars or 

 spines, comparing them with other Cribrilinidfe, but what he describes as 

 the loop is onlv the base of the spine arising from the lateral wall. His 

 figures (pi. 8. figs. 8, 9) are difficult to understand without examination of 

 specimens and are far from satisfactory. The bars are simply spines such 

 as we see in many Membraniporce, and they do not touch their neighbours 

 at the side continuously for the whole length. These spines are an arch 

 over the frontal membranous wall with which they are not in contact, and 

 the operculum, which is of a Membraniporidan type, is not separable, and is 

 in the membrane in no way connected with the spines. We may ask if 

 there is any reason for generically separating it from Membranipora, and 

 must answer, No — as Smitt has already done. 



The ridges of Cribrilina have been supposed to show that it was very 

 closely allied to Membraniporella nitida, but the difference is much greater 

 than first appearances suggest. 



In 1879 t I mentioned minute spines round the area of C. Gattyce, Hincks, 

 and since then Harmer % has described spines surrounding C. radiata §. 

 I have made decalcified and stained preparations, and even decalcified 

 preparations of dried specimens of various species, for they often show im- 

 portant structures, and these preparations led me to doubt the theory of the 

 frontal wall being formed of confluent spines. My decalcified preparations 

 of C. radiata from several places show a row of bluntly pointed, erect spines 

 round the border of the zocecium, with the two distal ones much longer and 

 narrower than the others, and acute. Of course, these free erect spines have 



* " Natural History of East Finmark," Ann. Mag. Nat. Hist. ser. 7, vol. xii. (1903) pi. 8. 

 tig. 13, pi. 9. figs. 4, 6. 



t "Bry. of the Bay of Naples," Ann. Mag. Nat. Hist, ser. 5, vol. iii. (1879) p. 30, pi. 9. 

 fig. 6 a. 



X "On the Morphology of the Cheilostomata," Q. Journ. Micr. Sci. vol. xlvi. p. 326. 



§ Probably these must be compared with the large stout spines which I figured round 

 Eippothoa Brongniartii, d'Orb. in Ann. Mag. Nat. Hist. ser. 7, vol. xv. (1905) p. 10, pi. 1. 

 figs. 2-4 ; and also with the spines of Lepralia 1'oissonii, Hincks. 



