CANONS AND DOMES. 



69 



filled with ice, as was first pointed out by Mr. Clarence King. This has been 

 questioned; but 18 months since I found a small amount of perfectly pre- 

 served glacial polish on the north wall. Ice, however, cannot have excavated 

 this valley, for the moraines are of trifling extent, and facts enough have 

 been adduced above to show that the glaciers of the Sierra have made no 

 such canons. If the fault system studied in this paper extends to the Yo- 

 semite, as the fissure system certainly does, the valley cannot be due to a 

 local subsidence. On the other hand I can see no objection to the hypothesis 

 that the shattered zone of rock was disintegrated and eroded in preglacial 

 times, the process being completed by a glacier which left a small moraine 

 near the entrance and thus converted the valley into a lake. 



Formation of Domes. — Decomposition converts polyhedral fragments of 

 granite into approximately spherical nodules found the world over ; and I 

 have on previous occasions examined the cause of the process, which, shortly 

 put, is simply that the rate of decomposition varies directly as the surface 

 exposed per unit volume. Similarly a prism of rock projecting from a sur- 

 face is reduced to a more or less raammillary or dome-shaped mass. The 

 following figure shows such a mass about five feet high near Dumont meadows, 

 at a locality where the prismatic fractures are very pronounced. 



In these instances examination makes it perfectly clear that the modelling 

 is directly dependent upon the fissure systems from which decomposition has 



FiGUEE 8— Weathered end of Prism. 



proceeded, and, as has been shown in the earlier part of this paper, the fissure 

 systems are those to be expected in a mass which on a large scale is to be 

 regarded as homogeneous. Thus these rounded forms are not due to any- 

 thing like ball structure or flow structure. 



From such little masses as shown in figure 8 to immense masses simihvi-ly 

 modelled there is every possible transition, and the great domes of the Yo- 



