1883,] NATURAL SCIENCES OF PHILADELPHIA. 253 



anchored a quarter of a mile from the face of the Muir glacier 

 the portion to the southeast for a distance of perhaps a thousand 

 feet, as examined by the field-glass, was of a different character to 

 the rest of the face in having a milky white, marble-like look. 

 The line of demarkation between this opaque and the transparent 

 ice was exactly defined. It was not possible to get nearer for a 

 more satisfactory examination, but the conclusion of all was that 

 this portion was compressed snow. At this point the ice-sea had 

 to draw in, through passing an intruding bluff of rocks, and the 

 lateral pressure must have been enormous between the bluff and 

 the solid ice. It would be the best possible opportunity for a 

 mass of snow, carried down from the mountain side, and floated 

 along on the margin of a wide glacier, to become ice if pressure 

 would ever do it. It cannot, of course, be positively stated that 

 this opaque section was compressed snow, in the absence of 

 actual handling, but there is little room for doubt that it was. It 

 was, at any rate, an opaque section, and whollj' different from the 

 glacier-ice as generally seen. Again, from the amount of air- 

 cavities in snow, and the resistance these must oflTer to the self- 

 pressure of snow, and also from actual experience of deep snow- 

 drifts in ordinary mountain ranges, there is nothing to warrant a 

 belief, outside of an actual demonstration, that the pressure of any 

 depth of snow is of itself sufficient to turn it into glacier-ice. 



If now we admit that above the glacial snow-line and under the 

 great snow-cap there may not be solid ice formed by compression, 

 but there may be a huge lake of water held back by the icy 

 breast-work at the snow's edge, we may conceive of a method of 

 forming the glacial sea quite different from any alread}'^ proposed. 

 The water must and will flow out from the edge of the snow-line 

 when the temperature is far below freezing-point, and form a 

 fringe of ice all along the line. How this is done can be readily 

 seen passing under the snow-sheds of a mountain railroad. 



On the Denver and Rio Grande Railroad, passing over Marshall's 

 Pass, 14,000 feet altitude, as the speaker did in May of the present 

 year, the melted snow passed as water through the mass to the 

 bottom^then passed down the mountain-side under the snow to the 

 snow-shed, where it formed real glaciers down the railroad — cutting 

 under the sheds to the railway track. The law must of necessity 

 be the same on a mountain-top in Alaska as on a mountain-top in 

 the Rocky Mountain region. Snow occurring after this ic}'^ deposit 

 was Tormed, would extend down the mountain over the ice, and 

 new layers of ice would be continually forming over the old layers, 

 or on their edges with the occasional retrocession of the snow. 

 A portion of the water at the snow-head will naturally course 

 under the ice, and form a channel beneath. This will increase in 

 width and depth wdth time. In the torrent which sprung out 

 from above the mouth of the Muir glacier myriads of stones, 

 some of them of many cubic feet in size, were borne along by the 

 muddy waters. The force of the water, as well as the added 



