production of the Prismatic Structure of Basalt. 217 



be produced whether the form of the bottom upon which the 

 mass of basalt reposes be convex (as in fig. 22), or concave (as in 

 22 A), or a plane (as in fig. 22 B). Curvature in the prisms 

 must also take place when the top and bottom surfaces are both 

 planes (as in fig. 15) but inclined to each other at an angle large 

 enough to give sufficiently abrupt change in depth to the cooling 

 mass, cooling in all cases being supposed to take place from the 

 top surface or from one of the two surfaces only. The precise 

 curve that will be assumed by each prism must vary in different 

 parts of the mass, and may vary also in different parts of the 

 length of each prism — that is to say, depends upon the greater 

 or less rapidity with which the isotherms vary in divergence 

 from parallelism. The concave side of the curvature of each 

 prism will always present itself towards the deeper part of 

 the mass, in which direction also the isothermal planes con- 

 verge. The cases illustrated by the figs. 22, 22 A, 22 B 

 refer to change of depth or of volume in the mass in one di- 

 rection only, in which case the curvature of each prism will be 

 in one plane only ; but, as is obvious, the depth or volume of the 

 mass may vary in orthogonal directions, and hence the prisms 

 split from such a mass may be curved in more than one plane ; 

 and with certain forms of irregularity of the bottom the curva- 

 ture of some of the prisms may have points of contrary flexure. 

 It is further obvious that curved prisms thus produced by cooling 

 from the top only will be very slightly taper, their smallest dia- 

 meter being at the bottom. As in nature the curvature is 

 usually small in proportion to the length of the prism, so the 

 taper may be so small as to be inappreciable. It is also plain 

 that if from any causes the rate of cooling varies at different 

 points of the cooling surface, corresponding changes may result 

 in the curvature or divergence of the isothermal planes and, by 

 consequence, in the curvatures of the prisms themselves. 



Innumerable variations and combinations of curvature, often 

 so perplexed and confused as to defy analysis, are to be found in 

 many basaltic regions, due generally to successive and superim- 

 posed outflows locally occurring, and to dykes filled with trans- 

 verse prisms cutting through prismatic masses previously deve- 

 loped, and to violent diversities of cooling due to water-channels 

 formed on the surface, or to violent mechanical efforts (such as 

 those of tilting or dislocating of the bottom, or of denudation, 

 &c.) having concurred with, or acted posteriorly to the pro- 

 duction of the prisms themselves ; but into these details the 

 writer cannot here enter. 



No intelligible solution of the mode of production of curved 

 basaltic prisms has, so far as the author is aware, been produced 

 by former authors. De la Beche (' Geological Observer/ 2nd edit. 



