Scoresby's "Arctic Regions" volume 1 page 122.) The case appears to me rather obscure: for that part of the mountain which is protected by a mantle of snow must be less subject to repeated and great changes of temperature than any other part.

I have sometimes thought that the earth and fragments of stone on the surface were perhaps less effectually removed by slowly percolating snow-water than by rain, and therefore that the appearance of a quicker disintegration of the solid rock under the snow was deceptive.

(15/2.

I have heard it remarked in Shropshire that the water, when the Severn is flooded from long-continued rain, is much more turbid than when it proceeds from the snow melting on the Welsh mountains.

D'Orbigny tome 1 page 184, in explaining the cause of the various colours of the rivers in South America, remarks that those with blue or clear water have their source in the Cordillera, where the snow melts.) Whatever the cause may be, the quantity of crumbling stone on the Cordillera is very great.