DEFORMATION AND RUPTURE 



311 



by the stresses of a bent beam, compressive below and tensile above. To 

 the tensile stress along the upper part of the wedge I ascribe the vertical 

 fracture which completed the gouge. It is possible that more than one 

 vertical fracture ordinarily occurred, dividing the wedge into several 

 parts. 



The position of the vertical fracture, as thus explained, is conditioned 

 (in part) by the distance to which the conoid crack penetrates the rock. 

 Toward the horns of the crescent, where the conoid crack vanishes, the 

 crack probably penetrates less deeply, and the vertical fracture there 

 traverses a thinner part of the wedge ; hence the curve given by the inter- 

 section of the vertical fracture with the surface is not concentric with the 

 corresponding curve for the conoid fracture, but meets it. Some of these 

 relations are diagrammatically shown in figure 2. 



It is worthy of note that the two fractures, referred respectively to 

 shearing and tensile stresses, differ notably in the textures of their sur- 



A 



F 



ICE 



WMW/A 



c- 



-j— -r 



tf- 



ROCK 



B 



Figuhe 9. — Theoretic Origin of Fractures 

 producing the Crescentic Gouge. 



This ideal section illustrates the theory 

 of origin of fractures producing the cres- 

 centic gouge. AE, original (longitudinal) 

 profile of rock bed ; AFDB, deformed profile 

 of rock bed (with exaggeration of curva- 

 ture) ; DGC, conoid fracture ; FG, vertical 

 fracture. 



Figure 10. — Theoretic Origin of the ou- 

 Hque and vertical Fractures of 

 the Crescentic Gouge. 



This diagram illustrates the alternative 



hypothesis for the origin of the oblique 



and vertical fractures of the crescentic 

 gouge. 



faces. The conoid fracture, as seen in figure 1 of plate 37, gives a rather 

 smooth surface, and that produced by the vertical fracture is compar- 

 atively ragged. The vertical fracture also departs from regularity more 

 widely than the conoid. In the figure just mentioned the outlines of 

 two of the vertical fractures more nearly resemble right angles than 

 circular arcs. 



Since the writing of this paper it has been examined by several friends 

 well qualified to discuss its theoretic part, and as a result of various sug- 

 gestions and criticisms a number of passages were modified. One of the 

 most important suggestions — but one which I was unable to accept — in- 

 cluded the following alternative hypothesis for the origin of the two 

 fractures delimiting the crescentic gouge. I quote: "We will suppose 



