Page:Transactions of the Geological Society, 1st series, vol. 1.djvu/265

 grotesque cliffs, remarkable both for their variety of colour and the multiplicity of short pyramids implanted one upon the other.

This loose sand is also the matter which fills up for the most part the inclined semi-trough, which from the shallow inner harbour of Poole extends to the northern acclivities of the chalk hills by Corfe Castle. It is there of a whitish-grey colour, covered with heath reduced in some places to the state of turf. To this formation of sand belong the following mineral substances:

(a) Coarse hard ferruginous sandstone, passing to a conglomerate, composed of rolled quartzose pebbles of different sizes, united by a ferruginous cement which does not effervesce with acids. This rock strikes fire, and has at first the appearance of slag, being rough, hollowed out, and covered externally with a crust of an earthy brown colour, arising from the oxidation of the iron. It has no action whatsoever on the magnet. It occurs in the form of flat scattered masses at the depth of a few feet from the surface of the sandy soil; the upper strata of loose sand being sometimes washed away, these pieces project, making a sort of cornices, and protecting the inferior strata of sand. I suspect this conglomerate to be of a very late origin, and indeed, daily forming by a process somewhat like that of cementation, viz. the percolation of water strongly impregnated with iron coming from the upper strata, and thus agglutinating the loose sand.

This rock is to be seen plentifully on the road from Shorwell to Chale in the Isle of Wight, and in. the sandy trough of Poole.

(b) Potter's clay.

It alternates with the loose sand in the trough of Poole, where it is found in beds of various thickness at different depths. It does not effervesce with acids, and from a cursory chemical examination which Dr. Marcet had the kindness to make at my request, we