[ aoi ] 



XXIV. On the Origin and Mechanism of production of the 

 Prismatic (or columnar) Structure of Basalt. By Robert 

 Mallet, F.R.S. 



[Concluded from p. 135.] 



Vl/'E have now to consider what will be the form of the trans- 

 verse fractures producing these joints. Referring to 

 fig. 6 (being an axial section of the topmost portion of a prism 

 down to the first joint Im, the axial surface of section in the 

 FiQ. s 



figure being divided by nearly parallel lines, the distances be- 

 tween which represent successive approximately isothermal 

 couches), the prism is at once in progress of cooling and of con- 

 tracting by loss of heat from its top surface $ S, and by that 

 from the sides of the prisms — which are at the lowest temperature 

 at the top (where cooling has been going on longest), and higher 

 at every point below down to the splitting level B B, fig. 4. In 

 a line round the prism, as Im, fig. 6, every particle of the ex- 

 terior couche is subjected to two orthogonal strains, due to con- 

 traction of the material of the prism in its axial and transverse 

 directions respectively ; and the contractile pull in each of these 

 directions is proportional to the length of the contracting column 

 which produces it. Thus at the point o (taken very near to /) 

 the vertical strain op is proportional to Is, and the horizontal or 

 transverse component o q is proportional to Ix, or nearly to the 

 radius of the prism. The resultant, or, of these two compo- 

 nents is the direction of the contracting force at the point 

 o tending to initiate fracture at that point j and if we assume 



