Dr. F. A. Bather — Eocystis. 51 



4, Undetermined jjlates. E 7609-10 ; G. F. M. k 



There seem no d priori grounds for supposing that all these 

 specimens belong to the same species, or even genus. Nor is there 

 any external evidence that they do not. They lie close together on 

 the same surfaces of shale ; at the same time they are never in any 

 natural juxtaposition, but always occur as isolated plates or ossicles. 



These four sets will therefore be described separately. 



1. The Stellate Plates are probably to be regarded as thecal plates 

 of a Pelmatozoan, and since a parallel is to be found only among 

 Cystidea we may regard the Pelmatozoan as a Cystid. 



Each plate has a small, but usually distinct, central umbo, from 

 which folds radiate to the periphery. The ridges due to these 

 folds on the outer surface of the plate are more marked than the 

 corresponding furrows on the inner side of the plate (Figs. 7, 10, 11, 

 12). The number of folds varies from five to perhaps as many as 

 ■seventeen (Fig. 6). In many instances three folds, arranged in 

 Y fashion, are more prominent than the others, and merge into the 

 umbo. Two folds nearly as pronounced are often seen to approach, 

 or perhaps join, the umbo, one on each side of the Y and at about 

 right angles to its stem. Thus arise five main folds, which, it will 

 be noted, are not of necessity pentamerously symmetrical. The 

 remaining folds do not reach the umbo, but while a few come rather 

 near to it the rest are mere golferings of the margin of the plate. 

 Thus the total number of folds tends to increase with the size of 

 the plate. 



These folds present two points of difference from the folds and 

 ridges so common on the thecal plates of many Cystidea and 

 Crinoidea.' First, the rare specimens in which the margin of the 

 plate is indicated (G. F. M. a, b) appear to show that the main folds 

 were directed towards the angles of the plate, whereas the axial 

 folds of Crinoids and the rhomb-folds of Cystids normally pass across 

 the sides of the plates, the more prominent among them bisecting 

 the sides at right angles. Secondly, whereas the minor ridges of 

 Cystidea Rhombifera and of all Crinoidea except Porocriniis are 

 parallel to the major axial folds, thus giving rise to the characteristic 

 rhomb-structure, the folds of Eocystis primceva radiate from the 

 umbo like the spokes of a wheel. It follows from this that in two 

 adjacent plates of E. primrBva the folds cannot liave passed in 

 a straight line from one plate to another. It also seems probable 

 that the plates cannot have fitted very accurately together, and that, 

 if in any case two ridges of adjoining stellate plates did lie in the 

 same straight line, then there must have been smaller irregular plates 

 filling up the interspaces. A similar, though more definite and 

 regular arrangement of folds is seQixin Amygdalocystisjlorealis; but 

 the plates of Macrocystella marim, which, as they lie in their matrix 

 of Shineton Shale, are strongly reminiscent of Eocystis primceva., may 

 be distinguished by the axial nature of the folding. 



Since not a single stellate plate in Dr. Matthew's collection 

 preserves a complete outline the exact measurements cannot be 

 given. The long diameter varies from 25 mm. to 6 mm., but 

 perfect specimens may have exceeded this. In the majority of 



