298 PKOE. T. G. BONNET ON THE [May I9OO, 



Sandstone. One represents the grits between the two conglomerates, 

 and was taken from a fairly hard band from 200 to 300 feet below 

 the second one. A long description is needless. The grains are 

 commonly a little smaller than in the last specimen, and the slice 

 contains a large proportion of ' volcanics' as already described, but 

 perhaps fewer are blackened. Another specimen represents the 

 lowest grit (under No. 1 conglomerate). The grains are rather 

 smaller, and some of them have a slight green tinge, but otherwise 

 it closely resembles the last described. The fifth specimen is from 

 a fine-grained grit, associated with argillites, near Castle Pulver- 

 batch. The fragments are yet smaller, and more distinctly angular, 

 but it also is crowded with similar volcanic materials. 



The foregoing samples, I think, suffice to give a fair idea of the 

 composition of the Longmynd Torridonian. Some of their materials 

 have been undoubtedly derived from coarse granitoid rocks, and it is 

 also true that fragments of ' rhyolite ' may be occasionally detected 

 in the Scottish Torridonian. But they are not common ; I am 

 certain of their presence in two slices only, out of ten in my collec- 

 tion, and I do not detect them in any of six slices from Staffordshire 

 pebbles, though they are present in one from the Bunter near 

 Bun corn. 1 Thus the Longmynd Torridonian s are quartz- rhyolite 

 rather than quartz-felspar grits, and on lithological grounds alone 

 cannot have been important contributors to the Staffordshire Bunter. 



(3) Cornwall and Devon. — Whether Cornwall and Devon, 

 even supposing a large extension southward and westward, could 

 supply the sand is doubtful, if what remains is a sample of what 

 has gone. 



The Triassic pebbles, as exhibited at Budleigh Salterton, were 

 described in 1895, 2 but I may add that the red quartz -felspar 

 grit is practically identical, microscopically as well as to the 

 unaided eye, with the Torridon Sandstone of Scotland, and that 

 further study of the tourmaline-bearing pebbles from the Midlands 

 reveals ra' her more resemblance to some rocks from Devon and 

 Cornwall 3 than I had at first anticipated. But even now there is 

 an overlapping rather than a general identity. The same variety of 

 quartzite, as I then pointed out, occurs in both localities, but 

 abounds in the former and is rare in the latter. 



The quartzite dominant at Budleigh Salterton does not exhibit 

 under the microscope any distinctive characteristic ; though the grains 

 are rather angular, more oblong, and smaller, perhaps also with less 

 secondary quartz than in the Midland and Scottish specimens. The 

 difference, however, in form is very marked, and I have never seen 

 in Staffordshire one of the flat ellipsoidal pebbles characteristic of 

 Devon. Besides this, whatever evidence we possess is hostile to the 



1 I have to thank Mr. Bolton, of Owens College, Manchester, for this and 

 other pebbles from that locality. 



2 Geol. Mag. 1895, p. 75. 



3 I am indebted to Mr. J. J. H. Teall, F.E.S., for the opportunity of studying 

 a number of specimens from this region. 



