40 EFFECT OF CRYSTALLISING FOECES 



do this it must either widen or lengthen the crevice which contains it. 

 When the next thaw comes, the widened or deepened crevice again fills 

 with water, and the next frost repeats the action described. This process, 

 going on at hundreds of places on the surface of a rock, soon breaks it 

 up into smaller fragments, and these in turn are subjected to the same 

 action. The process is necessarily confined to the outermost layer and 

 is only of much importance when frost and thaw alternate rapidly. 

 Long-continued frost appears to protect rocks from weathering, the ice 

 formed cementing the whole surface together. 



The above is the generally accepted explanation of the disintegrat- 

 ing effect of the alternate freezing and thawing of wet rocks and soils. 

 But the writer, from observations he has made in hilly districts and 

 from facts that have come under his notice, is of opinion that there is 

 another important factor to be. taken into account, viz., the forces 

 exerted by the crystals of water, during freezing, in striving to main- 

 tain their original direction of growth. The writer has noticed that 

 sandy surfaces near cliffs in mountainous districts are sometimes 

 covered, after a frosty night, following wet weather, with a moss-like 

 growth or efflorescence, consisting of fine filaments of ice, in apparently 

 prismatic crystals. In some cases, this growth attains a height of 

 several inches, and pebbles and fragments of gravel are lifted by it, two 

 or three inches above the level of the ground. 



Moreover, it has been shown l that a disintegrating effect perfectly 

 similar to that produced by the alternate freezing and thawing of 

 water on porous earthenware can be produced by replacing the water 

 by melted, crystallised sodium thiosulphate, Na.,S 2 3 .5H 2 O, which is 

 fluid at about 50 C. and which is denser in the solid than in the liquid 

 state, and which therefore, unlike water, contracts very slightly in 

 solidifying. The disintegrating effect, in this case, can only be 

 ascribed to the forces brought into play by crystallisation. A some- 

 what similar effect is sometimes to be noticed with a photographic dry 

 plate. If a gelatine plate be removed from the fixing bath and, with- 

 out washing, set aside to dry, the residual sodium thiosulphate crystal- 

 lises out on drying and often tears holes in the gelatine film. The 

 expansive effect of solidification of water is probably the more potent 

 factor in bringing about the disruption of large fragments of rock, but 

 with the smaller fragments and with clods of clayey soil, the effect of 

 the crystallising forces may be the more important. 



(iii) As glaciers. Glaciers have played a very important part, both 

 in grinding and wearing down rocks into the finest powder and also in 

 transporting the materials, in some cases for hundreds of miles. The 

 water which issues from the snout of a glacier is always heavily laden 

 with the finest mud, and huge heaps of debris, known as moraines, 

 mark the successive positions of the termination of the glacier. The 

 fine portions of the glacial deposits possess a composition similar to 

 that of the rock from which they were formed, comparatively little 

 chemical change having taken place in their production. 



(iv) Chemically. As already stated, many minerals, e.g., felspar, 

 exposed to the action of water, undergo chemical changes leading to 



1 Cobb, Jour. Soc. Chem. Ind., 1907, 390. 



