Geol.— Vol. I.] TURNER— ORIGIN OF YOSEMITE VALLEY. 29 1 



The glacier ordinarily affords protection against the temperature changes 

 which fracture the exposed rock of the bordering mountain slopes above its 

 surface, yet the glacier keeps at a low temperature the rock surface it covers. 

 Therefore, if the glacier be broken across, so that communication may be 

 had with its bed by the outer air, the line of rock-bed thus exposed to excep- 

 tional contrasts of temperature maybe disrupted in consequence. The trans- 

 verse crevasse, carried forward at the top, will presumably have its foot at 

 the base of the bluff. In summer only will this crevasse as a rule remain 

 open. And in summer there may be set up, at the base of the buried cliff, 

 through the communicating air of the open crevasse, alternations across the 

 freezing point between night and day. The result must in such case be sap- 

 ping of the cliff. Alternations of temperature on the bare slopes without are 

 not diurnal but annual. This subglacial process of cliff weathering, with con- 

 centration of weathering at its foot, will be relatively rapid. The tendency of 

 course will be downward, as well as horizontally backward. But downward 

 fracturing will be defeated by the difficulty of removal of products. Hence 

 the floor will be maintained approximately flat, following with its extension 

 in the trail of the receding cliff. It will not, however, be absolutely flat; fur- 

 thermore, outward a little from the lee of the cliff, abrasion will resume its 

 work of smoothing and polishing, interrupted at the upper lip of the cliff, and 

 these irregularities of the approximately level floor, revealed by stream pond- 

 ing after glacial withdrawal, will constitute the shallow rock basins so much 

 discussed. 



Given the process of undercutting at the cliff base, the height attained by 

 the cliff, in its recession into a rapidly rising grade, will not affect the effi- 

 ciency of the process. No more work of sapping is required to undermine a 

 high than a low cliff, given complete removal of the debris. The height, 

 however, may be limited by the depth to which open crevassing will pene- 

 trate. In fact, there seems to be such a limit. Yet terrace steps of several 

 hundred feet, nearly sheer fall, occur. Above the limiting height, a second 

 step, of more vigorous recession, will be initiated. Terracing thus results. 

 As it is immaterial, in the progress of undercutting and undermining, what 

 height the cliff may eventually come to have, so it is immaterial how high 

 the side walls may become in this low-grade eating backward into the high- 

 grade slope. Hence, under favorable conditions— which must be of rare 

 occurrence — a Yosemite and a Hetch Hetchy Valley is a possibility. 



It is in the summit region of a glaciated range that work of this character 

 becomes most marked. Here, at the head of the individual ice-stream, the 

 cirque, or glacial amphitheater, unlike the head region of the stream of water, 

 nourishes the glacier full-born at once and exhibiting its maximum of destruc- 

 tive efficiency. The amphitheater wall is but the last in the series of cliffs, 

 and though the stream does not become broken to its foot by overriding it, 

 it nevertheless has its crevasse — the 'bergschrund ' — following the semicircle 

 a little out upon the body of the ice ; and the encircling amphitheater wall is 

 also a clitr of recession. It is exceptionally high, not because the glacier con- 

 tinues on a level for a larger space at its head here, but because, with the 

 glacier following persistently the same habit of horizontal burrowing, it works 

 into steeper slopes. Hence the amphitheater is the most notable of glacial 

 forms. It indicates the heaviest degradation. But since its wall is produced 



