Dv. F. A. Bather — Notes on Yunnan Cystidea. 321 



second circlet, and so induced shiftings or numerical changes in 

 the opening hetween them. In Sinoeystis the number of the 

 periproctals was more fixed, being six in S. loczyi and 

 £. yunnanensis, five in S. mansuyi (antea, 1918, p. 511). 

 A rough correspondence with the bounding plates is shown in 

 Heed's pi. i, fig. 3, but the available evidence is not enough to 

 warrant any general statement. The greater irregularity and 

 variability of the thecal plates in Sinoeystis, as well as the absence 

 of any connexion between the periproct and the relatively fixed 

 Adorals II, suggest that the constitution of the theca had here little 

 influence on the periproctals. The preservation of the periproctals 

 in the fossils of Sinoeystis indicates that they were more solid, or 

 more firmly united to the frame than in Megacystis. Therefore, in 

 Sinoeystis the number of the periproctals dominates the bounding 

 plates, whereas in Megacystis it is more affected by them. To some 

 extent the periproctals and the bounding plates form two systems, 

 each subject to its own hereditary and environmental influences, 

 and therefore liable to be brought into conflict. 



c. The Hydropore and Gonopore. 



When the various openings in Megacystis were discussed by 

 P. H. Carpenter in 1891 (J. Linn. Soc, Zool., xxiv, pp. 48-50), 

 on the basis of Miller's figures, he recognized all four openings in 

 M. commoda, though in general he regarded the openings as confined 

 to peristome, periproct, and " nephridial opening" [ = hydropore[|, 

 and in some cases (e.g. M. elegans) to the two former alone. His 

 view was that the periproct served as an osculum, embracing the 

 gonopore as a rule and in some cases the hydropore as well. 



Jaekel (1899) did not specifically mention or figure the hydropore 

 and gonopore in his Trematocystis, but, since he regarded the 

 presence of both those openings as characteristic of the Aristocystidse, 

 he must at any rate have assumed their presence in Trematocystis. 



Miller's figures and descriptions are rather insecure evidence, and 

 there are reasons for believing that the opening observed in 

 a number of species cited by Carpenter, and regarded by him as 

 "excretory" or "nephridial" and "equivalent to the fourth 

 opening of Aristocystis and Glyptosphcera", i.e. what we now take 

 as the hydropore, is really the gonopore, and that the hydropore, 

 though present, was not observed. The reasons for this belief are 

 first a morphological argument, secondly comparison with other 

 genera, thirdly actual observation of the British Museum material. 



The morphological argument depends on a structural feature to 

 which attention has not, it seems, previously been directed. This 

 is the tendency in so many of these primitive echinoderms for the 

 hydropore to be a slit crossing the suture between two plates. So 

 it is in E&rioaster (Bather, Studies in Edrioasteroidea, Geol. Mag. 

 1914, p. 168, pi. xi, fig. 2), Aristocystis (Barrande, 1887, Syst. 

 Silur., vol. vii, pi. ix, figs. 2, 3, 6, 13 ; Bather, 1906, Palseont. 

 Indica, II, 3, p. 9.), Sinoeystis {antea, 1918, p. 535), Pleurocystis 

 (Jaekel, 1899, Stammesges. d. Pelmatozoen, pp. 101, 138), Sehizocystis 



DECADE VI. — VOL. VI. — NO. VII. 21 



