20 



J. A. PHILLIPS ON THE CONSTITUTION AND 



all subaqueous deposits ; the majority are in the state of minute 

 well-rounded pebbles. As in the case of water-borne sands, the 

 effects of attrition are more conspicuous in the larger fragments 

 than in the smaller ones ; but even the most minute particles are in 

 these sands much rounded. Exceptionally the quartz encloses 

 fragments of a felspathic material ; and fluid- cavities with bubbles 

 are not entirely absent in African specimens. 



A bed of sandstone, said to occur in a salt-producing district 

 sixty miles south-east of Tebessa, of which some years ago I brought 

 a specimen from Tunis to London, is entirely made up of rounded 

 grains. In that respect this rock resembles the millet-seed sand- 

 stones of Lancashire and Cheshire : but it is unlike them, inasmuch 

 as no crystals of quartz, or of any other mineral, have been deposited 

 upon the surfaces of the rounded granules. 



Modern blown sands, of which we have numerous examples in 

 this country, differ from desert sands and from those of certain 

 sandstones only in being usually somewhat less completely rounded. 

 Among the sands of this description which have been examined are 

 specimens from the dunes at Rhyl, Flintshire, Colwyn Bay, Den- 

 bighshire, Lytham, Lancashire, and from Perranzabuloe and Lelant, 

 in Cornwall. Speaking generally, the sands from the northern loca- 

 lities have been more completely rounded than those from Cornwall, 

 and consist of a mixture of worn quartz and various slaty and 

 other rocks, with a little felspar and a few fragments of shells. 

 Fluid- cavities with bubbles are rare in the quartz of these sands. 

 In addition to rounded grains of various slaty and other rocks, 

 quartz, felspar, and fragments of shells are present ; among these 

 quartz largely predominates. The Cornish sands contain a few 

 partially rounded prisms of tourmaline. 



Chemical Composition of Sandstones &fc. — -A microscopic examina- 

 tion of a large number of sections of grits and sandstones having 

 led to the conclusion that many of the published analyses of such 

 rocks must be of a very imperfect character, five different specimens 

 were selected for analysis. 



In making these analyses I have received the valuable assistance of 

 Mr. E. W. Yoelcker, A.R.S.M. ; and in each case fusion with alkaline 

 carbonates was adopted. The estimation of alkalies was made after 

 an attack by hydrofluoric acid, and was checked by a fusion with 

 carbonate of calcium and chloride of ammonium. 



