10 MR. A. W". WATERS ON THE MARINE 



from opposite sides may interdigitate, or in a few cases may be fused 

 together, thus forming a good protection. In some cases there is a second 

 spine on each side near the proximal end, but this is not common ; also in 

 one zocecium I have found a central proximal spine. The oral spines stand 

 erect, and therefore it is impossible to give an idea of them as seen from 

 above — consequently a few pressed down are drawn to show the shape. 



Cribrilina radiata (Moll) , Smitt. 



The specimens from the Cape Verde Islands are fairly typical C. radiata, 

 a species subject to much variation. They have about 11 ribs, and there are 

 5-7 pores in a ray. There is no suboral pore visible, but in dried specimens, 

 afterwards decalcified, a row of b' pore * disks is seen below the aperture with 

 a single pore in the middle, the same size as the others, but this semicircle of 

 pores must probably be considered as the upper series of the radial pores. 

 On each side of the central pore there is usually a small nodule. By the 

 side of the more or less oval zooecia there are elongate avicularia, as fre- 

 quently described ; as is generally the case in Cribrilina, there is no 

 avicularian cross-bar, though one occurs in C. monoceros, Busk, which, how- 

 ever, most workers would now remove from Cribrilina on other grounds. 

 The large bordering spines to which reference will be made are about 11, 

 the two upper ones, being much longer than the others, have been mistaken 

 for setae or vibracula. 



In a stained specimen, from Rapallo, the small frontal pores are each 

 in the centre of a much larger disk. 



The ovicells have nodules, and sometimes a keel dividing to each lower 

 corner. 



The opercula are separable and are of quite the same type as in C. punc- 

 tata, Hass., C. Gattya, Hincks, C. Balzaci, Waters, and I fail to see any 

 reason for separating a genus Puellina, Jullien, as maintained by Levinsen. 



The primary zooecium has 11 spines, while that of C. punctata, Hass., 

 has 12 (Harmer), that of C. projecta, "Waters, has 12, that of C. 

 monoceros, Busk, has 9 (Harmer), that of C. Gattya; Hincks, has 11, while 

 the closely allied C. Balzaci, "Waters, has a primary zocecium similar to the 

 ordinary zooecia, and this, according to Harmer, is also the case in C. 

 annulata, Fabr. The primary of Membraniporella nitida, Smitt, has 10 

 spines fairly regularly surrounding the zooecium, but it may be seen that 6 

 are oral spines and 4 zooecial. Hincks evidently made a mistake in describ- 

 ing 14 spines, for his figure in Brit. Mar. Poly. pi. 27. fig. b" is clearly 

 not a primary. In 21. nitida a number of zooecia, after the primary, have 

 a Membraniporidan character somewhat like M. lineata, Hincks, and taking a 

 fairly typical colony of M. nitida from Oban, Scotland, the second series have 



* In speaking of pores it is not meant that there is a direct opening to the interior. 



