DISINTEGRATING WORK OF HEAT. 435 



longer than the rock below. When the surface is shorter than the main 

 mass of the rock below, if the elastic limit be surpassed tensile fracture 

 will take place nearly at right angles to the surface. When during the 

 hot part of the day the surface is longer than the average of the rock 

 below, powerful shearing stresses are set up parallel to the surface. If 

 these surpass the elastic limit of the rock, shearing rupture will take 

 place roughly parallel to the surface. When once a rupture is started, 

 it will extend until the rupture feathers out at the surface or until the flake 

 or spall has become long enough to permit the stresses to be relieved by 

 buckling. 



The surface scaling of rocks, and parting parallel to the surface 

 exfoliation, are attributed to differential expansion and contraction caused 

 by change of temperature, because the fractures are peripheral. In massive 

 rocks, whatever the topographic forms and shapes, the scaling produced by 

 temperature changes should be roughly parallel to the surface. For 

 instance, if there be a great dome of granite, the fractures which form 

 should circumscribe it (see PI. I) as they would circumscribe a sphere, 

 being at any given place parallel to it. In the United States such fracturing 

 is well exhibited by the granite of Dunn Mountain, in North Carolina; by 

 the granite domes of central Missouri; and above all by the great granite 

 domes of the West, such as those in the Sierra Nevada. The scales are 

 ordinarily but a few centimeters in thickness, and are rarely more than 60 

 cm. thick. Of course, in none of these cases does every fracture occur 

 precisely parallel to the surface; rather, they are inclined to it at small 

 angles in various ways, thus frequently intersecting. However, the corre- 

 spondence is, in a general way, as given, and therefore the positions of the 

 fractures furnish evidence of their production in consequence of shearing 

 stresses due to changing temperature. 



Another reason for believing that the peripheral or zonal scaling and 

 parting are due to changing temperature is the fact that the ruptures occur 

 nearly parallel to the surface without reference to previous structures. 

 Thus it often happens that great slabs, 2 cm. to 5 cm. thick, of schistose or 

 gneissic rocks form, which are diagonal or even at right angles to the 

 structures. 



