23 



oped, more chitinised projecting marginal portion, »the pore-rings the outer open- 

 ing of which in the most developed rosette-plates is smaller than the size of 

 the plate a little in from it. Within the pore-ring we can distinguish between 

 two portions, differing in thickness, a thicker outer area and an inner surrounded 

 by the other, very much thinner and very small pore-area, which is pierced by 

 an extremely fine pore, and distinguished by a strong bluish lustre, which at 

 first glance makes it seem thickened. Such one-pored rosette-plates may appear 

 singly (the distal wall of Flustra securifrons, Fl. papyrea etc., the distal wall of 

 most Reteporidae), in a more or less numerous (of 2 — 12 plates) series (all walls 

 in many species of Smittina, in Adeonidae, in most Flustridae), or in groups some- 

 times consisting of more scattered, sometimes of more closely placed plates (e. g. 

 in Catenariidae, the distal wall in Scrupocellariidae and Thalamoporellidae). In 

 cases where the single rosette-plates are close together, they have a square or 

 hexagonal shape, and meet in a network of elevated ridges, which must be re- 

 garded as the pore-rings for the single rosette-plates. 



It is quite impossible to draw a sharp limit between a group of one-pored 

 rosette-plates and a multiporous rosette-plate, as the onlj' character, which can 

 be regarded as peculiar for the multiporous compound rosette-plate, namely, a 

 common pore-ring which surrounds all the single small plates, can be developed 

 to very different degrees, and does not always appear to be constant even in the 

 same colony or in the same zocecium. This is the case with for instance Arthro- 

 poma Cecili, in which species the distal wall as well as the distal half of each side 

 wall is furnished with an elongated or oval group of numerous uniporous rosette- 

 plates. A pore-ring may be lacking on some walls and appear on others, not 

 only in the same colony but also in the same zocecium, and wherever it appears, 

 it may either be exceedingly well developed, or only slightly indicated. Within 

 the pore-ling (PI. XVII, figs. 10 a, 10 b) which has a similar structure as in the 

 uniporous rosette-plate, we have an area, the large area, which may be filled by 

 the small plates to a very different degree, and while these for instance in Arthropoma 

 Cecili often form only a longitudinal belt along the middle part, in ^Lepralia* 

 Pallasiana (PI. XVII. fig. 10 a) they fill the whole or at any rate the largest part 

 of the plate. The rosette-plates may show many different degrees of calcification, 

 to some extent according to the degree of calcification of the species concerned. 

 In most members of the family Bicellariidae, in species of the genus Onychocella 

 and Selenaria we thus find rosette-plates, which with exception of the pore-ring 

 are quite uncalcified; on the other hand we find e. g. in Flustra foliacea and 

 Flustra carbacea, that the large area is calcified and the single small plates are 

 uncalcified. The pore area is always uncalcified, while on the contrary the outer 



