216 ROBERT B. SOSMAN 



This explanation seems to have been satisfactory to many of 

 the earlier authors of geological treatises, 1 but before many years 

 had passed doubts began to arise as to whether this process could 

 have been an efficient cause of the numerous cases of columnar 

 structure which began to accumulate in geological literature as 

 travel became more extensive and observations multiplied. James 

 Thomson 2 in 1863 urged that contraction of a homogeneous mass 

 was a sufficient cause for all columnar structure, and that the 

 hypothesis of crystallization from centers was unnecessary and 

 improbable. Mallet 3 discussed the contraction hypothesis in detail, 

 showing how it would account, in his opinion, for all of the struc- 

 tures found in columnar rocks. Bonney, 4 Iddings 5 and others have 

 followed the same lines of argument. 



CONTRACTION HYPOTHESIS 



The radial-contraction hypothesis is still the explanation gen- 

 erally accepted by the textbooks, and perhaps applies in the 

 majority of cases of prismatic structure. But a much more com- 

 plete discussion of this hypothesis than has yet been published 

 could be profitably made, for there has been no attempt at any 

 quantitative application of it to actual occurrences. It has served 

 hitherto simply as a qualitative explanation. The relation of the 

 size, shape, curvature, jointing, and other properties of the columns 

 to the original temperature, viscosity, and rate and manner of cool- 

 ing of the rock is capable of more exact definition. 



For instance, the time factor in cooling in its relation to the 

 elastic properties of the rock does not seem to have been considered 



1 More qr less vaguely associated with this definite hypothesis was the idea of a 

 "concretionary force" which is frequently referred to. The idea that columns might 

 be due to the mutual compression of actual spheroids of lava (now understood as 

 "pillow" lava) was also more or less confused with the crystallization hypothesis. 

 Watt's idea of the matter seems to have been perfectly clear, but Mallet, for instance, 

 misunderstands Watt's "mutual compression of spheroids" to mean actual compres- 

 sion (Phil. Mag., L [1875], 221-24); the words "mutual interference of radially 

 growing spheroids" state Watt's meaning more clearly. 



3 Brit. Assoc. Rep., 1863, Abstract, p. 89. 



3 Phil. Mag. (4), L (1875), 122-35, 201-26. 



4 Quar. Jour. Geo!. Soc, XXXII (1876), 140-54. 



5 Amer. Jour. Sci., XXXI (1886), 321-31. 



