318 Dr. F. A. Bather — Notes on Yunnan Cystidea. 



the higher beds of Limestone dip north-eastward at 45° beneath the 

 alluvium of tbe lliver Alwen. We may, therefore, be justified in 

 assuming on stratigraphical grounds, and on the evidence of the 

 gravels themselves, that the coal in the latter may have been 

 derived from some portion of an outcrop that was demolished by 

 glacial action. The erosion may have removed the stratum entirely 

 if it were quite a small wedge between the limestone and a branch 

 of the main Bala Fault concealed by the alluvium on the northern 

 side of the outcrop. Our experience shows that such little wedges 

 of rock, clipped in at the junction of important faults like that 

 which traverses the Dee valley from Corwen to Llangollen and the 

 Bala Fault, are the rule rather than the exception. 



The shales and sandstones in such an outlier might have yielded 

 some of the material of the finer silts and loams and sandy patches 

 in the late-glacial beds. Whether the sandstones were purple 

 might depend upon whether they were overlain by any Triassic 

 Sand like that which succeeds the purple sandstone in the Vale of 

 Clwyd. No boulders of purple sandstone have been found (it is 

 very friable material), yet the patches of brick-red Triassic-like sand, 

 mentioned above as being found in the drift, are distinctly 

 suggestive. 



How far the Trias overspread the Carboniferous rocks in this 

 area is not known, but it is fairly certain on other evidence that the 

 Bala and other large structural faults were in action both before 

 and after the deposition of the Trias. 



IV. — Notes on Yunnan Cist idea. III. Sinocfstis compared with 



similar Genera. 



By F. A. Bather, D.Sc, F.B.S. 



(Published by permission of the Trustees of the British Museum.) 



B. — Comparison with Megacystis {continued). 



4. The Thecal Openings. 



b. The Periproct. 



IN Megacystis the periproct — which S. A. Miller always called the 

 mouth — lies between adoral circlets II and III, being bounded 

 by the two posterior facet-bearing plates and the posterior interradial, 

 i.e. 3 of Adorals II, also by 2 or 3 of Adorals III, making 5 or 6 

 plates in all. The number 5 is the more usual ; it occurs in 

 M. aspera, faberi, gorbyi, indianensis, ornatissima, ornata, parvula, 

 parva, perlonga, ple?ia, scitulus, and spangleri, also in British Museum 

 specimens E 7631, —33, —35, —37, —38, ?— 39, —40, —42, —44, 

 _45 f _74, _76, —77, E 16167, and E 16168 (see our figs. 24, 25, 

 antea). The number 6, due to an additional Adoral III, is found in 

 M. bacillus, commoda, splendens, and perhaps gyrinus, also in E 7630, 

 —34, —36, ?— 75, and E 16171 (see our figs. 22, 23, 28, 30, antea). 

 This does not seem to be a difference of such constancy as to be 

 relied on for the discrimination of species. It is the sole character 

 distinguishing Miller's M. commoda from his M. gorbyi, not to 

 mention M. scitulus and M. parva, which may be younger stages of 



