Crosby.] 74 [October 4, 



coinciding with and arising from stratification. Joints and 

 shrinkage-cracks agree, too, in their independence of the stratifica- 

 tion, in running in more than one direction in the same mass of 

 rock, and in being usually approximately vertical, at least in sed- 

 imentary rocks. They differ chiefly in this, that shrinkage-cracks 

 are usually short, are not arranged in sets by their parallelism, 

 and rarely cross each other. 



Broadening the definition of joint-structure in the way here 

 proposed, we are obliged, of course, to recognize several distinct 

 classes of joints — distinct both in characteristics and origin. 

 These are, first, the usually short, non-parallel, and non-intersect- 

 ing joints dividing the rocks into polygonal blocks the sides of 

 which are not, as a rule, continuous in the same straight line with 

 the sides of adjacent blocks. The joints and blocks are usually 

 extremely irregular, though the prismatic jointing of basalt is a 

 notable excej:>tion. Geologists are generally agreed that the joints 

 of this class are due to contraction, the contraction arising from 

 the clessication and consolidation of sedimentary rocks, the cool- 

 ing of eruptive rocks, and the crystallization of rocks of either 

 class. Hence these may be properly called contraction or shrink- 

 age joints. Next in order come the much less important fractures, 

 observed chiefly in granitic rocks, which are parallel with the 

 surface of the ground and appear to die out at a depth of one 

 hundred to two hundred feet. These have been explained by Pro- 

 fessor Shaler, probably to the satisfaction of most geologists, as 

 due to the expansion of the surface under the influence of the 

 sun's heat ; and therefore, may be properly known as expansion 

 joints. We see something analogous in the exfoliation of gran- 

 ite blocks when exposed to fire. 



The third and most important class embraces the straight, par- 

 allel and intersecting joints, the joints properly so-called, accord- 

 ing to Mr. Gilbert. Concerning the origin of these, several 

 opinions have been entertained by geologists ; but, as already 

 stated, Mr. Gilbert has demonstrated that the two causes most 

 commonly appealed to — shrinkage and compression — are entirely 

 inadequate. Besides the hypotheses that the joints of this class 

 are, in their origin, akin to shinkage-cracks or to slaty cleavage, 

 Mr. Gilbert cites two others, both of which he likewise properly 



