production of the Prismatic Structure of Basalt. - 125 



unshattered lava -streams have cooled pretty symmetrically. 

 Local and occasional admissions of surface-water through some 

 of those cracks is one of the agents in their highly irregular dis- 

 tribution. These conditions may be well seen in some of the 

 great lava-flows of Auvergne, which have been largely and deeply 

 quarried for building-stones. There is therefore no analogy or 

 true relationship between the capricious and accidental forces 

 and conditions which have broken up the mass of lava-flows into 

 blocks of the most diverse forms and dimensions, and those quite 

 symmetric and determinable strains which have produced the 

 splitting of basalt into jointed prisms. Yet instances occur in 

 great lava-masses of the irregularly broken up and the pris- 

 matic structure in different parts of the same mass. The 

 conditions producing this diversity need not be here enlarged 

 upon. 



The primary conditions to the production of straight pris- 

 matic structure in basalt are general homogeneity of the ma- 

 terial, a tolerable regularity in the general form of the mass, and 

 inequality in the rate of its cooling in one or more directions, 

 i. e. from one or more of its bounding faces. The more or less 

 horizontal and tabular masses of basalt as they occur in nature 

 are usually cooled most rapidly from their upper or lower face — 

 or if the basalt be found filling a dyke, then from the two opposite 

 surfaces of the adjacent rock. Let us suppose a perfectly homo- 

 geneous, isotropic, and tabular mass of basalt resting upon a 

 horizontal floor, all its dimensions being great and its length and 

 breadth much greater than its depth. Let us suppose it cooled 

 entirely from the upper surface, from every portion of which there 

 is an equable loss of heat, such as would result from the super- 

 position of a uniform mass of conducting material constantly 

 maintained at a temperature inferior to that of the upper surface 

 of the tabular mass of basalt. We have now to trace the pro- 

 gress of refrigeration and the progress and results of the internal 

 strains due to its contraction by cooling. The mass, assumed as 

 originally in liquid igneous fusion and nearly at the same tem- 

 perature throughout, becomes less fluid, and finally rigid at the 

 surface of cooling. Below this it is in a more or less plastic 

 condition ; and as the wave of heat passes outwards always 

 parallel to the cooling surface, so the thickness of the rigid and 

 of the viscous horizontal couches tend constantly to descend 

 deeper into the mass of liquid basalt beneath. So long as nearly 

 the entire mass remains liquid or viscous, the internal strains 

 produced in it by the contraction due to cooling will be resolved 

 into and met by a vertical descent of the whole mass and by in- 

 ternal movements in the still imperfect liquid. No splitting or 

 Assuring can take place except in such upper laminse of the mass 



