Vol. 68.] THE GEOLOGY OP MYNTDD-T-GADEE. 359 



Nodular rhyolites. — The great development of these rocks 

 (PI. XXXIX, fig. 1) on parts of the southern slopes of Mynydd-y- 

 Gader forms one of the most remarkable features of the geology 

 of the district. The nodular structure does not occur in a patchy 

 manner, but is limited to one or two well-marked bands running- 

 parallel to the general strike of the rocks. These bands lie 

 within the Ashy iSeries. 



Two kinds of nodules may be distinguished : — 



(a) Regularly spherical or spheroidal nodules, rarely more than an inch 

 in diameter. They are often hollow, and show a marked concentric 

 structure when weathered. They ocour chiefly in the western part of 

 the outcrop. (See PI. XXXIX, fig._2.) 



(h) Larger ovoid lumps, sometimes reaching a length of from 4 to 5 inches. 

 These are not hollow, and do not show a concentric structure. They 

 come from the eastern part of the outcrop. (See PL XXXIX, fig. 4.) 



The nodules of the first of these types stand out on the weathered 

 surface, and cause the rock superficially to resemble a conglomerate. 

 Sometimes a number of nodules of approximately equal size are 

 compacted together, but generally the spaces between the larger 

 nodules are occupied by smaller. 



In many places, instead of nodules, the weathered rock -surface 

 shows a network of sinuous ridges of flinty material (PI. XXXVII, 

 fig. 1) standing out boldly. The spaces between the ribs of the 

 network vary from a quarter of an inch to over 2 inches. Patches 

 of nodular rock may be mingled with patches without nodules, 

 or the nodules may be enclosed in a network of flinty ridges. 

 It is possible that the flinty network may be in part the result of 

 "flow-brecciation. 



Sometimes the nodule is evidently composed of three or four con- 

 centric layers (PI. XXXIX, fig. 2), in other instances it is uniform. 

 Frequently the nodules are hollow, and sometimes, in addition to the 

 central hole, a space occurs between the concentric layers, or, instead 

 of there being one central hole, several irregular spaces may occur. 

 There seems every reason to suppose that all these cavities are of 

 secondary origin, are due to weathering, and that all the nodules 

 were originally spherulites. This is the view adopted by Prof. Cole^ 

 and Mr. Harker ^ in the case of the Carnarvonshire rocks, and by 

 Mr. Thomas ^ for those of Skomer. No examples of nodules were 

 found in which the cavities were filled by or lined with quartz. 

 None of the solid unweathered nodules show a radial arrangement 

 -when broken across ; but this structure is sometimes strongly 

 marked in certain of the nodules, in which weathering has etched 

 out the several concentric layers. 



Phenocrysts, whether of quartz or felspar, are very much less 

 •common in the nodular rhyolites than in either the compact or 

 the banded varieties. 



In thin sections the nodules are sometimes seen to pass gradually 



1 Q. J. G. S. vol. xli (1885) pp. 162-68; & ibid. vol. xlii (1886) pp. 183-90. 

 '^ ' The Bala Volcanic Series of Caernarvonshire ' 1889, pp. 28-40. 

 ^ Q. J. G-. S. vol. Ixvii (1911) pp. 185-90. 



^Q.J.G.S. No. 271. 2d 



