584 GEOLOGY. 



assume that stress- differences would be distributed throughout the 

 mass, and bring into play a large portion of its stress- accumulating 

 competency. When the mass yielded, it would not be by crushing, 

 but by ^'flowage,'' which would be more or less general throughout the 

 mass. It might, however, be partially concentrated, as, for example, 

 ,on the borders of sectors of different specific gravity. 



Stress-differences may arise from physical changes within the rock 

 itself. Whenever there is a re- aggregation of matter, or a change of 

 any kind which involves change of volume, a change of stress is liable 

 to be involved. It may be of the nature of relief or of intensification. 

 In an earth built up by the haphazard infall of matter, a very hetero- 

 geneous mass must result, and the subsequent changes may be sup- 

 posed to be intimately distributed through the mass, being slight at any 

 point, but present at innumerable points. An immeasurable number 

 of small stress-differences may, therefore, be developed throughout 

 the mass. Until these overmatch the effective strength of the mass, 

 they may continue to accumulate. These are not necessarily con- 

 nected with stresses that arise from sphericity, and may work more or 

 less independently of them. It is not improbable that the great stress- 

 accumulating power of the globe finds an essential part of its explana- 

 tion in supplemental considerations of this kind, and not wholly in its 

 spheroidal form. 



The actual configuration of the surface. — The foregoing computations 

 relative to the power of shells of the earth to sustain pressures are based 

 on ideal forms and structures that are not realized in fact. IIow far 

 the earth fails to conform to these conditions must now be considered. 

 When compared with the earth as a whole, the inequalities of its surface 

 are trivial. If the great dynamic forces acted through the whole or 

 the larger part of the body of the earth, the configuration of the surface 

 can be supposed to have done little more than mfluence the location 

 of the surface deformations and their special phases. But if the forces 

 were limited to a crust of moderate thickness, the configuration of the 

 surface is a matter of radical importance. 



Concave tracts. — There is need, therefore, to inquire if any consider- 

 able breadth of the crust is outwardly plane or concave, for the principle 

 of the dome is obviously not applicable to a plane or concave surface. 

 To be a source of fatal weakness, the concavity must be broad enough 

 to cause the planes of equal cooling, the isogeotherms, to be concave 



