Dr. F.^A. Bather — Notes on Yunnan Cystidea. 259 



specimen, however, lies rather in the traces of Adorals I, which, as 

 was surmised, form a narrow ring round the peristome. The sutures 

 hetween them cannot he made out, but the number of plates cannot 

 have been less than six or more than eight; six is the more 

 probable. 



Five facets are also said to occur in H. hacidus Miller (1879), but 

 the statement is not well substantiated by the figure, and is some- 

 what opposed by the description of the peristome as subquadrangular ; 

 Adorals I "are not clearly determinable"; if five facets were 

 present the arrangement was probably as in H. spharoidalis. The 

 figure of H. elegans Miller (1878) is like those of H. plenus and 

 JET. pustulosus, but the holotype has been attacked by a boring- 

 organism, so that its plates are anchylosed. In H. perlongus Miller 

 (1878) the five facets are not clear in the figure, and there is a lack 

 of detail. 



The general conclusion to be drawn from the figures and 

 descriptions of specimens with five facets is that they differ from 

 those with four facets mainly in the presence of a facet on the 

 anterior Adoral II and of a food-groove leading to it between the 

 two anterior Adorals I. 



Finally we come to the two species said to have only three arm- 

 facets. 



In " ffolocystites" ampins, Miller (1892) mentions three facets on 

 large bosses, and seems to imply that this is all by describing the 

 peristome as in the centre of the triangle formed by them ; but in 

 the absence of a figure and of other details it is impossible to 

 consider the case. Miller compares the species with H. ventricosus 

 (1879), which is equally ill-defined, and may well be a synonym of 

 If. dyeri (1879). . H. ampins may therefore provisionally be regarded 

 as based on either an imperfect specimen or an individual variation 

 of H. dyeri. 



The number three is also found in H. gyrinus Miller & Gurley (1 894). 

 The drawing of the adoral face (our fig. 30) shows apparently three 

 facets corresponding to the left anterior, left posterior, and the 

 united right-hand pair of the normal plan. The food-grooves 

 leading from the two left-hand facets meet at what would in a 

 normal form be regarded as the left end of the peristome. From 

 this point a straight groove connects them with the right-hand facet. 

 Thus the groove-system is roughly in the form of a >-. These 

 grooves are drawn of approximately equal width throughout, and, 

 for the most part, with a structure that can only be interpreted as 

 cover-plates. For these two reasons it is not possible to judge of 

 the precise position or extent of the peristome, but the arrangement 

 of the adoral plates suggests that it occupies about two-thirds of the 

 stem of the Y measured from the fork. The length of this tract is 

 about 9 mm., and the length of each food-groove is then also about 

 9 mm. The left posterior groove, however, is a little longer, 

 perhaps as much as 11 mm., and it widens at its distal end as 

 though it were forking. The left anterior groove has a slight curve, 

 with the convexity on the left; and the stem of the Y is also 

 slightly curved, with posterior convexity. The right-hand facet is 



