CONOID FRACTURES 307 



closely the conoid of percussion. This feature occurs in a glacial trough 

 where there are many crescentic gouges, but it is not connected by grada- 

 tion with the ordinary gouges. 



Differential Pressure 



As a glacier moves forward its under surface is continuously adjusted 

 to the irregular shapes of its bed. The greater inequalities of the chan- 

 nel find expression on the upper surface of the glacier, but the minor 

 inequalities do not affect the upper part of the ice stream. The diagram 

 (figure 5) represents in profile a projection of the bed, of moderate mag- 

 nitude in relation to the total thickness of the glacier. The adjustment 

 of the glacier to this obstruction affects the flow lines of only the low3r 

 strata of ice, leaving unaffected all above some limiting plane, A B. 

 Below that limit the lines of motion first ascend and then descend in 

 passing the obstruction. Tf we think of the flow lines of the diagram 

 as separating layers of ice, then each layer becomes thinner in ascending 

 and gains thickness in descending.* 



A large boulder embedded in the glacier close to its base is not reduced 

 in thickness along with the enclosing ice layers, and its resistance to com- 

 pression develops differential stresses both above and below it. These 

 pressures tend to force it into the overlying ice, and at the same time to 

 force it into the rock bed. As the rock bed effectually opposes the down- 

 ward motion the boulder is actually forced into the ice body above it 

 (figure 6). A large share of the pressure thus brought to bear on the 

 upper side of the boulder is transmitted by the boulder to the rock bed at 

 their point of contact or approximation, and there is thus a concentration 

 of pressure on a small area of the rock bed. This concentration con- 

 tinues as long as the ice about the boulder is undergoing vertical com- 

 pression in passing the projection, and ceases when that compression 



*The statements of this paragraph are, I think, indisputable, but they are quali- 

 tative only. In making a diagram to illustrate them I could not avoid presenting a 

 more definite, and in some sense quantitative, conception of the forms of the flow lines, 

 but of the accuracy of this conception I am by no means confident. Not only is there a 

 third dimension of which account should be taken, but there are complicated inter- 

 actions connected with differential velocity. In order to pass through the diminished 

 cross-section above the projection, some of the lower layers of ice must take on tempo- 

 rarily a higher velocity, and that acceleration involves both internal resistance and basal 

 friction. The conditions affecting these resistances are notably different on the up- 

 stream and downstream sides of the projection, and it is therefore improbable that the 

 curves of the flow lines are symmetric. Moreover, the diagram, by indicating the com- 

 pression of the lowest layer as greater than that of any other layer, assumes that that 

 layer's temporary increment of velocity is greatest ; and this again assumes that the 

 temporary increment of basal friction is not greater than the temporary increment to 

 shearing stress of the ice in planes parallel to the rock surface — a matter as to which 

 I have no information. 



