Wealden District and the Bas Boulonnais. 39 



We may now pass to the case represented in fig. 13, which differs from fig. 12 



in having for the axis of the oval a curved instead of a straight line. To estimate 

 the effect produced by this change of form on the direction of the longitudinal 

 fissures, conceive the area AB A' B' uplifted and thus placed in that state. of tension 

 which it can just support without dislocation. If we take a point Q in the portion 

 BAB' of the oval, it is not difficult to see that the line of greatest tension through 

 Q must (to a degree of approximation sufficient for our present purpose) bear the 

 same relation to the curved axis of the oval, as the corresponding line through Q' 

 in fig. 12 bears to the rectilinear axis in that case. Consequently the directions of 

 longitudinal fissures through Q and Q' (which must be perpendicular to the lines 

 of greatest tension) must also bear similar relations to the two axes respectively. 

 Consequently the line of fissure through Q, instead of curving upwards as in fig. 12, 

 may curve downwards, as represented in fig. 13, provided the curvature of the axis 

 be sufficiently great. The curvature of the lines of dislocation on the side of the 

 axis opposite to Q will evidently be increased by that of the axis. A fissure near 

 the axis will obviously be approximately parallel to it. 



In the preceding diagrams, I have represented, for the sake of simplicity, only 

 one central and two lateral longitudinal fissures ; but I have shown in the memoir 

 so often referred to, that under very probable conditions, instead of a single con- 

 tinuous fissure, we should have several parallel fissures ; so that if, for instance, 

 we take the simple case of a uniform mass, of which the surface is a parallelogram 

 of indefinite length, elevated by a uniform force, we should have a system of 

 fissures such as represented in the annexed diagram (14). It should be observed, 



-^ — -. — — ■ 14_ 



