226 On the production of the Prismatic Structure of Basalt. 



and in the same way as described in Fig. 24. 



the preceding pages with respect to /pJJJjl 

 the curved prisms of basalt, and illus- 

 trated by fig. 22 B, which may be 

 viewed as a longitudinal section of 

 one half of the projectile, B B being 

 its axis. But these appear to be mere 

 resemblances and analogies; and if there is any physical connexion 

 between the splitting forces by which prismatic basalt has been 

 produced and those crystallogenic forces first pointed out by the 

 writer in 1855 and here referred to, it will be hereafter to be 

 sought for by the laws ascertained by Mitscherlich with respect 

 to the different dilatation (or contraction) of crystals in different 

 directions with respect to their principal axes, which seem to 

 point to the conclusion that the formation of the crystalline 

 planes of separation in masses of crystallizable substances as they 

 cool may be due to differences of contractility of their constituent 

 crystals in different directions. If so, the production of crystal- 

 line planes of separation parallel to the principal axes in a mass 

 crystallizing from cooling may prove to be as much the result 

 of contraction as the splitting-up of an uncrystallizable mass of 

 basalt into prisms is so. All bodies which contract by heating, 

 by cooling, and by drying, such as sand baked into stone, starch," 

 mud, coke, &c, tend to separate, as has been often remarked, 

 into more or less prismatic masses ; upon the particular condi- 

 tions of which in different cases it is not necessary here to en- 

 large, further than to remark that in the case of mud, which as it 

 occurs in nature has almost invariably been deposited from water 

 in shallow hollows or cavities, the deposit thins out towards its 

 edges. In drying, therefore, by sun and air, the stratum is ob- 

 served to crack first at and near the edges and in directions 

 more or less orthogonal to these. The irregular ribbons thus 

 produced between two contiguous cracks again split transversely 

 by further drying into more or less square pieces — a form which 

 thus results from two successive stages of contraction, in con- 

 tradistinction to the one which produces the hexagonal division 

 of the basalt, as enunciated in the preceding pages. It must be 

 borne in mind that the contractile forces produced by baking, 

 desiccation, &c. must act locally and in succession on different 

 parts of the mass, and can never produce the same regularity of 

 separation as is found in a mass of cooling basalt. 



