Crosby.] 244 [Dec. 16, 



and diverse theories of their origin have been proposed, but geol- 

 ogists have finally settled down to the conclusion that shrinkage 

 is the main cause, and this is the explanation given in nearly all the 

 standard text-books of geology, the first and second classes of 

 joints not being distinguished. Mr. Gilbert, however, showed 

 very conclusively nearly four years ago 1 , that the parallel and in- 

 tersecting joints possess none of the characteristics of cracks 

 known to be due to shrinkage, such as are exhibited in dessicated 

 clays, and that from the nature of the case it is impossible that 

 they should ; and also that all other explanations of this class of 

 joints proposed up to that time were alike untenable. The way 

 thus appeared open for a new explanation of ordinary jointing ; 

 and the principal object of my paper above referred to was to 

 show that the fractures of the earth's crust, generally believed by 

 geologists to result from the vibratory movements known as earth- 

 quakes, must be plane, parallel, intersecting, and normally vertical, 

 possessing, apparently, all the characteristics of parallel joints. 

 So far as I can learn, no serious or unanswerable objections to 

 this earthquake theory of jointing have } 7 et been advanced. 



In the American Journal of Science for January, 1884, Mr. 

 Gilbert accepts the theory in its main outlines, but calls attention 

 to some minor peculiarities of this class of joints which were not 

 satisfactorily accounted for in my paper. Successive earthquake 

 shocks are known to traverse the same district in various, direc- 

 tions, which often make low angles with each other ; but the sys- 

 tems of parallel joints normally intersect at high angles. I 

 explained this disparity as due to the fact that, after the rocks 

 have been broken by one set of joints, the layers or sheets thus 

 formed will break most easily at right angles, so that oblique vi- 

 brations may give rise to rectangular fractures and blocks. Mr. 

 Gilbert shows the insufficiency of this explanation and suggests 

 that after one set of joints has been established by an earthquake, 

 a second series of vibrations crossing the path of the first oblique- 

 ly may have its strains relieved by slipping along existing joints 

 and thus fail to produce a new set ; while vibrations at right an- 

 gles to the original direction would produce strains not at all re- 

 lieved by the first system of joints, and a second system at right 

 angles to the first would be the necessary consequence. 



But Mr. Gilbert also directs attention to the fact that while, 



1 Amer. Journ. Sci., xxiv, 1882, p. 50. 



