1882.] 73 [Crosby. 



formed which, considering that all degrees of regularity exist 

 among joints not due to shrinkage, must in many cases be ex- 

 tremely difficult to distinguish from the latter. 



If all shrinkage-cracks in sediments were merely superficial sun- 

 cracks, the problem would be comparatively simple ; but, taking 

 things as they exist, we shall be obliged, if we adopt the classifi- 

 cation proposed by Professor Dana and Mr. Gilbert, to coin a new 

 term, to be used, provisionally at least, in those very numerous 

 cases where the origin of the planes of division cannot be deter- 

 mined. The distinction of ordinary joint-structure from stratifi- 

 cation and cleavage is rarely difficult, but its distinction from 

 shrinkage-cracks is often a very puzzling and difficult problem, 

 especially in coarse-grained and crystalline rocks. 



Although it is true, as Mr. Gilbert states, that the shrinkage- 

 cracks and blocks tend to be circular ; yet it is also true, as an 

 examination of actual examples will show, that there are circles 

 within circles, the centres of contraction being comparatively few 

 and distant at the beginning of the shrinkage; and, when the 

 process is complete, the primary cracks are, in their length and 

 straightness, in striking contrast with those formed later. And 

 it is important to observe that the size of the contracting circles 

 or polygons must be, approximately at least, proportional to the 

 thickness of the deposit, which is equivalent to saying that they 

 must be large in most examples of geological importance. 



It seems better, therefore, to class shrinkage-cracks as joints, 

 and then to distinguish, where we can, between the joints having 

 unlike origins. This plan obviates the necessity, which w^ould 

 otherwise exist, of adopting a new name for what geologists now 

 almost universally call the prismatic jointing of basalt. Besides 

 if fractures due to shrinkage are to be separated from the class of 

 joints, then the concentric fractures parallel with the surface in 

 granitic rocks, and which it is now generally agreed are the result 

 of expansion caused by the sun's heat, should also be classed 

 separately. 



A joint is a plane of division, not often merely a theoretical or 

 potential division, as in crystalline cleavage, but an actual physi- 

 cal break — not a plane where division may occur, but the division 

 itself. In this respect joints agree with shrinkage-cracks, but 

 differ widely from slaty cleavage and the planes of weakness 



