DIFFERENTIAL PRESSURE 309 



regarded, I think, as itself normal. If this view is correct, some other 

 cause must be sought for a special stress component parallel to the rock 

 face. 



I think such a cause exists in differential friction. The friction per 

 unit area of the glacier on its rock bed at any point is the measure of the 

 force there applied by the glacier in a direction parallel to the local rock 

 surface. It varies with the material of the two bodies in contact and is 

 directly proportional to the force, normal to the contact surface, by which 

 they are pressed together. Therefore during the period in which the 

 hypothetic boulder communicates an excess of pressure to a small area of 

 the rock bed, the same area experiences a proportionate excess of sliding 

 friction, and is consequently subject to a proportionate excess of force in a 

 direction lying in the plane of contact. The composition of this force 

 with the differential force normal to the plane of contact gives a resultant 

 parallel to the general system of oblique stresses in the surrounding ice. 

 This reasoning appears to warrant the statement that the differential 

 pressure occasioned by the approach of the boulder to the prominence of 

 the rock bed is oblique to the local rock surface and is directed forward. 



Deformation and Eupture 



To obtain an idea of the nature of the deformation resulting from the 

 differential pressure just mentioned I have tried a few simple experi- 



Figurb 7. — Ideal oblique View of originally plane Rock Surface. 



Showing, with vertical exaggeration, its theoretic deformation under the external 

 stress causing a crescentic gouge. The arrow shows the direction of the stress. The 

 direction of ice motion is from right to left. The position of the conoid fracture is indi- 

 cated by a broken line. Compare figure 8. The lines of the drawing are parallel equi- 

 distant profiles of the deformed surface. 



ments. If a liquid jelly be allowed to cool in a large bowl it assumes 

 the condition of an elastic solid with a level and smooth upper surface. 

 Pressed by the ball of the finger its surface is deformed, the hollow under 

 the finger being surrounded by a low circling ridge, the slope of which is 

 relatively steep toward the finger but very gentle in the opposite direc- 

 tion. If the pressure of the finger be made oblique the ridge becomes 

 steeper and higher on one side of the hollow, and is correspondingly re- 



