206 Mr. R. Mallet on the Origin and Mechanism of 



ending in a thrust outwards upon the detached fragment, the 

 fracture of which maybe longer or shorter between x and y, de- 

 pendent upon the rigidity of the basalt and its coefficient of con- 

 traction, which no doubt varies a good deal in different examples 

 of that rock. 



From what precedes, it is obvious that in a perfectly homo- 

 geneous and isotropic mass of basalt cooling symmetrically, all 

 the prisms should be divided transversely by joints, each range of 

 which should be at the same level in all the prisms — an arrange- 

 ment which observation shows is very nearly approached in many 

 instances in ponderous masses of basalt ; but a very slight amount 

 of heterogeneity or admixture of adventitious matter, or difference 

 in the rate of cooling, will cause a certain amount of irregularity 

 in the relative levels of the different prisms or length of the joint 

 pieces. It is obvious also that the imbedding of foreign mate- 

 rial, or even the existence of empty cavities, which are often of 

 considerable size in certain basalts, will not only produce irre- 

 gularities in the distances between these joints, but also more 

 or less disturb the regularity of their cup-shape. This last 

 source of disturbance in the cup-forms of the joints has been 

 pointed out by Professor James Thomson in his paper already 

 referred to. On referring to those papers it will be remarked 

 how widely his notions of the production of the cross joints 

 themselves differ from the solution here given, and how entirely 

 inadequate they are to account for the facts. He supposes cross 

 fracture to commence at the axis of the prism and to spread 

 outwards towards its surface. He assumes tension, therefore, 

 greatest about the axis or hottest part of the prism ; and to ac- 

 count for the equal and opposite force, he imagines a couche 

 produced about the exterior of the prism by calling in some 

 wholly imaginary force of expansion produced by some molecular 

 or chemical change in the basalt itself at and near the surface 

 of the prism by which its length there is increased. Even were 

 we to admit for a moment these vague and unsupported hypo- 

 theses, there is nothing indicated to account for the production 

 of the cup-form of the joints, nor for their convex sides being 

 presented in one direction rather than another, nor why they 

 should actually be presented in opposite directions. 



Hitherto we have been considering perfectly homogeneous and 

 isotropic basalt ; and although such basalt is closely approached 

 in nature, observation shows that almost all basalts are more or 

 less heterogeneous as regards the mineral constituents of which 

 they are made up or which are imbedded in them, and to a 

 greater or less extent are not isotropic anterior to cleavage into 

 prisms. The main constituent minerals, namely felspar, py- 

 roxene, oxides of iron, &c, vary much in proportion and in the 



