380 



REV. A. IRVING ON THE PHYSICAL HISTORY OP 



while even the darkest green earths (e. g. at St. Anne's Hill) may 

 he observed coloured in places of a bright orange-red by slow 

 oxidation. It is trivial to argue from the small amount of colour- 

 change observed in some instances (e. g. at Goldsworthy, Q,. J. G. S. 

 vol. xlii. p. 415) to a negation of such change during the ages for 

 which some of these beds may have been exposed near the surface. 

 Moreover, it requires a comparatively dry exposure to allow of the 

 penetration of the beds by oxygenated rain-water ; and cases may 

 occur in which the beds are kept so saturated with water from the in- 

 terior of the beds holding carbonaceous matter in solution, that the 

 work of oxidation may be arrested altogether. 



7. Microscopic Structure. — A detailed examination of the speci- 

 mens from the Wellington-College Well-section, subdividing the beds 

 for this purpose as they are given in vol. iv. of Mem. Geol. Survey, 

 p. 425, shows sizes of the quartz-grains ranging from 0-005 to 

 0-030 of an inch. A pure clay is seldom met with. Even the so- 

 called clays of the Middle group contain very often some 50 per 

 cent, of fine clear quartz-grains, cohering^in masses by the cementing 

 action of the argillaceous material. In the Upper group the 

 investment is generally a mere clay ; in the Middle and Lower there 

 is generally more or less of amorphous organic matter, either of 

 itself constituting the investment or mingled with clay*. The 

 quartz-grains are for the most part subangular, and therefore water- 

 worn. In one bed (no. 7) the majority of the grains were found to 

 be well rounded ; and in beds no. 8, part of 9, and 11, there was a 

 noticeable proportion of rounded grains. In some of the beds a 

 considerable quantity of glassy silica, in minute clear flakes, is seen. 

 This glassy silica may consist partly of fragments of frustules of 

 Diatoms ; but hardly so the larger flakes. These have the appear- 

 ance under the microscope of flakes of silica artificially precipitated 

 by a process well known in chemical laboratories. If so, the solvent 

 action of the humus-acids upon silicates may have contributed to 

 the result. 



8. Distribution of the Sarsens. — A careful investigation of the 

 matter which I have recently made has tended strongly to confirm 

 the statements of Profs. Prestwich t and Rupert Jones J, as to the 

 occurrence of these blocks and slabs of ' saccharoid sandstone ' in 

 the Upper Sands. The probable genesis of them has been discussed 

 by me elsewhere §. Their extensive distribution far beyond the 

 present range of the Bagshot strata is well known. The larger 

 blocks are generally very angular, sometimes water- worn on the 

 upper surface, but rough and uneven beneath. As a student of 

 glaciation, there seem to me insuperable physical difficulties in the 

 way of attributing to them a morainic origin. For these reasons 

 it seems that their wide distribution may be regarded as indicating 



* Geol. Mag., September 1883. 



t See Quart. Jouru. Geol. Soc. vol. iii. p. 384. 



\ Proc. Geol. Assoc. vol. vi. p. 436. 



§ Report of Brit. Assoc. (Soutbport Meeting), 1883 ; alsa Proc. Geol. Assoc. 

 vol. viii. pp. 156-160. 



