1868.] nARBAGE PARALLEL ROADS.. 275 



fragments of stone, until it mig-lit at last even sweep down in little 

 miniature avalanches, and thus cover the edges of the frozen lake 

 with a multitude of loose stones which had previously rested upon 

 the inclined surface of the mountain, and even with some of those 

 imbedded in its soil. 



Thus a mound of snow containing the debris of the mountain it- 

 self would be formed round the margin of the lake during the 

 ^vinter. On the return of summer, this margin of snow as well as 

 the icy covering of the lake would slowly melt, and the accumulated 

 debris of the mountain above it would thus be (juietly deposited in 

 the still water on the margin of the lake. 



An analogous case, arising in its commencement from the same 

 causes, although differing in its ultimate results, occurs when the 

 snow falling upon the higher elevations of great mountain-ranges 

 accumulates, and from time to time rolls down their sides in enor- 

 mous masses, carrying with it the fragments of rocks which were 

 loose upon their surface, and tearing up others more or less firmly 

 attached. The valleys thus filled up are called '' glaciers ;" and, the 

 same causes continuing to act, there are often formed upon their 

 surface one or more lines of these broken fragments of rocks near 

 the edge of the glacier, which are then called '' moraines." 



The parallel roads of Glen Iloy are, in my opinion, the results of 

 ancient moraines, generated by rain and snow, upon a very contracted 

 scale. 



It has been observed that in the highest portions of glaciers the 

 substance of which they consist is opaque ; it is, in fact, condensed 

 snow; whilst at the lower ends of the glaciers which terminate in 

 alpine valleys, the opaque structure of the condensed snow has been 

 changed into clear transparent hard ice. 



This transformation, which I have explained more fully elsewhere*, 

 arises from the circumstance that the masses of frozen snow in their 

 transit down valleys many miles in length, are continually cracked 

 and split, generally in vertical planes. 



The friction of the fractured sides against each other generates 

 heat sufiicient to melt a small portion of the indurated snow. The 

 water thus produced is retained within the narrow fissures, or is 

 spread over the surface of the opaque ice and freezes very slowly. It 

 therefore becomes a thin layer of transparent ice. 



The ice thus formed is much harder than the opaque snow-made 

 ice, which is full of bubbles of air ; consequently, when the next 

 move down the valley produces a new fracture, it will not occur in 

 the transparent ice, but in the opaque consolidated snow. 



Thus in their long passage down the glacier the blocks of con- 

 densed snow are gradually converted into pui^e transparent ice. 

 At the lower end of one of the glaciers, I was enabled for a short 

 time to get close to some of the terminal blocks. I could distinctly 

 trace in these the wave-like structure of the transparent ice. The 

 preceding explanation rests upon the passage of matter not in a state 

 of fluidity, nor yet altogether solid,'over substances of a harder nature. 



* Passages from the Life of a Philosopher, chap, xxxiv. p. 464 : 8to, 1864. 



