TRANSACTIONS OF SECTION C. 767 



appear on the Survey map as greenstone, between St. Isbmael's on the E., Dale 

 on the S., Wooltack Park on the W., and Musclewick Bay on the N., including 

 also Midland Island and part of Skomer. The rock seen in a small quarry at Crah- 

 hall was described by Mr. Teall as a somewhat basic porphyrite. What appears 

 to be the same rock is found exposed in several places on the opposite side of the 

 Mullock stream. At Marloes Sands the continuation of the Crabhall rock appears 

 in the cliff as a black dense rock much traversed by quartz and epidote. Micro- 

 scopically it is a perfect dolerite, generally ophitic in structure with plagio- 

 clase in augite plates, but sometimes granular ; there seems to be some hypersthene, 

 and serpentine is present, probably after the same mineral. From Marloes 

 Beacon this rock seems to continue, until it appears in contact with Llandeilo rocks 

 in Musclewick Bay. The rock found here, however, is shown by the microscope to 

 be a felsite (probably a soda felsite), and not a variety of the dolerite. 



At Martins Haven the same ophitic hypersthene dolerite is found which occurs 

 at Marloes Beacon and in Marloes Bay. 



IV. Age of the Igneous Rocks. — The age of these rocks appears to have been 

 regarded by the Survey as post-Carboniferous, while some of them at any rate have 

 been claimed by Dr. Hicks as pre-Cambrian. The felsites in the Benton area are 

 almost entirely associated with beds of Old Bed Sandstone, and there does not seem 

 to be any satisfactory evidence of intrusion. Such a continuous mass of quartz 

 felsite, with well-marked spherulitic and fluxion structure, seems to suggest 

 rather a flow than an intruded mass. On this supposition the beds must be at 

 least post-Silurian. 



The rocks from Llangwm to Talbenny are in almost every instance associated 

 with Llandovery beds on the one hand and Carboniferous strata on the other. 

 The Carboniferous strata are reversed in dip, and the line of junction is in our 

 opinion a line of thrust. The evidence seems to indicate that the rocks are not 

 post-Carboniferous, the Culm measures being apparently unaltered near the junc- 

 tion, and that, judging by lithological character, they did not occupy their present 

 position as a ridge in Llandovery, Old Bed, or Carboniferous times. With regard to 

 the southern area we could not find clear evidence of intrusion, all the chief 

 junctions appearing to be faulted ones ; still there seems to be little doubt that 

 rocks microscopically similar rest on diilerent measures of Silurian age. 



Grassholme, a small island about seven miles from the mainland, but rarely 

 visited, appears to be the continuation of the ridge from Wooltack Park and 

 Skomer towards the S.E. corner of Ireland. 



No clastic rocks were found, the main rocks being ophitic dolerite, with 

 corroded augite and some bands of secondary epidote and quartz. 



7. Notes on a Hornblende Pikrite frovi Greijstones, Co. Wicldow. 

 By W. W. Watts, M.A., F.G.8. 



[Communicated by permission of the Director-General of the Geological Survey.] 



In this paper the author gave a description of a rock which forms a dyke in the 

 Cambrian slates and grits of Greystones, in Co. W^icklow. It is a dark, dense, 

 coarsely crystalline rock, showing large crystals of hornblende with lustre-mottling, 

 owing to the weathering-out of olivine crystals. It becomes finer-grained at the 

 margins. An analysis by Dr. Sullivan was added. 



The hornblende is of the usual gi-een type, and occurs in large crystals 

 enclosing pseudomorphs of olivine, now made up of magnetite and probably a 

 colourless amphibole. A colourless hornblende also occurs either as cores or 

 borders to the green crystals. A third type of hornblende present shows few 

 cleavage cracks and much magnetite dust. Apatite is a constituent, but there is 

 no felspar in the rock. The margin of the dyke is much sheared and phacoidal 

 in structure. 



8. Hejjort on the Registration of Type Specimens of Fossils. 

 See Reports, p. 482. 



