/. HopMnson — Graptolites in Arenig Rocks. 467 



fronds witli the stem, but they are the best evidence for this that we 

 can expect in Fossil Botany short of actual organic union. Now the 

 fruit of Pecopteris arhorescens is so near to that of Cyathea that I can 

 find no characters whereby they can be separated. Our classifica- 

 tion based on the stems must of course yield to that derived from 

 the organs of fructification, and our group of Ferns, instead of being 

 made into a new order, as would be the case by some who publish on 

 Fossil Botany, must be grouped with a tribe of recent Polypodiacece. 

 It may seem that this is a forced and arbitrary grouping together 

 of plants that in some important characters so remarkably differ ; 

 and so it is undoubtedly to those who with rash confidence generalize 

 on the systematic position of plants from stem structure alone. But 

 what can such objectors say to the practice of placing in close prox- 

 imity plants that are beyond question nearly related to each other in 

 all essential characters, though some have caudices while others 

 possess rhizomes ? Yet these two forms of stems are more widely 

 separated from each other than the extinct Paleeozoic group is from 

 the recent forms. 



III. — On the Occubkence op a Eemaekable Group of Graptolitbs 



IN THE Arenig Eocks of St. David's, South. Wales. 



By John Hopeinson, F.G.S., F.R.M.S. 



IN a series of black, iron-stained shales, about 1000 feet in thick- 

 ness, which form the lowest beds of the Silurian rocks in the 

 immediate vicinity of St. David's, the author noticed the occurrence of 

 about twenty species of Graptolites, which, he considered, furnished 

 conclusive evidence of the equivalency of these beds with the Quebee 

 group of Canada, the Skiddaw slates of Cumberland, and the Arenig 

 group of Shelve. From their stratigraphical position, and from the 

 evidence afforded by the fossils they had previously yielded, these 

 rocks had already been inferred to. be of Arenig age. 



The Graptolites were collected in the lower beds of the series at 

 Eamsey Island and Whitesand Bay, by Messrs. Hicks, Homfray, 

 Lightbody, Kirshaw, and the author, in the course of a few days 

 they spent together at St. David's in July. 



Of the true Graptolites, or Bhabdophora, the only genera of un- 

 doubted occurrence are Dddymograptus, Tetragraptus, and Phyllo- 

 graptus. Didymograptus is represented by five species, three of 

 which — D. extensus, Hall ; D. patidus, Hall ; and D. pennatulus, Hall 

 — are characteristic of the Quebec group and the Skiddaw slates, D. 

 patulus also occurring in the Arenig rocks at Shelve; the other 

 species are new. Of Tetragraptus but one species, T. serra, Brong., 

 a Quebec and Skiddaw form, has been found. Phyllograptus also is 

 only represented by a single species, which is. new. There is also 

 another new species — a very peculiar branehing form referred pro- 

 visionally to Loganograptus. The absence of any specimens un- 

 doubtedly referable to Dicliograptus is i-emarkable, as this is a 

 common Quebec genus. Diplograptus and Climacograptus, genera of 

 very rare occurrence in the Quebec group, have not as yet been 

 found here. 



