3i 



begun at one side of the column, and to have advanced across to 

 the opposite side, we must expect to find the resulting fracture 

 quite unsymmetrical, and presenting very different appearances at 

 the places where it entered the previously unbroken stone prism, 

 and where it came to its termination, leaving the column broken 

 behind its advancing front. We find no such appearance ; but, on the 

 contrary, we commonly find a very remarkable appearance of ap- 

 proximate symmetry of character in the cross-joint, with respect to 

 the different sides and angles of the column. Perfect symmetry is, 

 of course, not to be expected, as the columns themselves are often far 

 from being of any regular or symmetrical form; but so far as Professor 

 Thomson's observations of the stones in the Giant's Causeway 

 have extended, he believes no appearance is to be found indicating 

 an advance of the fissure across the column, from one side to the 

 opposite, in any of the joints which exhibit, in other respects, the 

 usual remarkable features. There may, no doubt, be numerous 

 cases of fractures due. to shattering, by causes different from those 

 which have produced the ordinary remarkable joints. Next, any 

 idea that the cracking of the column could have simultaneously 

 begun all round the circumference, and advanced to terminate in 

 the centre, requires little more than to be brought before the 

 mind for consideration to be rejected as untenable. There 

 seems then to remain nothing to suppose but that the ordinary 

 cross-joint fissures came into existence, first in the interior of the 

 column, and then flashed out towards the circumference. 



In order to produce the cross fractures, commencing in the centre, 

 he supposed that a longitudinal tensile stress must have existed in 

 the middle of each column previously to the cracking of the cross- 

 joints. To account for such a tensile stress, he suggested as a 

 probable hypothesis, that after the column was formed, chemical 

 action, caused by infiltration of water, might cause an expansion 

 of the outside of the column, and that the outer part, thus growing 

 longer, would pull the internal part more and more intensely, until 

 at last the internal part would give way, and break into short 

 lengths. The fissures thus formed, it is obvious, must stop short 

 without extending quite to the outside of the column, as the pull 



