154 ON COLUMNAR. EISS1LE, AND SPHEROIDAL STRUCTURE. 



generally form at the points where the spheroid's surface was nearest 

 to the outside of the column, because the shell would be rather 

 weaker there. 



The independence of the structures is to some extent confirmed by 

 the fact that a thin plane lamina of rock sometimes exists between 

 two spheroids in a column. 



All the divisional structures, therefore, which can be observed in 

 igneous rocks are, if the above considerations be admitted, to be re- 

 ferred to one and the same cause, viz. contraction of the mass while 

 cooling. It is not, of course, denied that in some cases effects may 

 have been produced by subsequent chemical or mechanical action (as 

 the true cleavage in certain of the Welsh felstones) ; but it is con- 

 tended that those above described are all the result of one and the 

 same cause. 



Discussion. 



Prof. Ramsay inquired whether Mr. Bonney had observed the cup- 

 and-ball structure in narrow dykes at right angles to the planes of 

 cooling. 



Mr. Koch stated that he had tried many experiments bearing on 

 the subject of prismatic and spheroidal structure. Slag at 1500° C, 

 when suffered to run into water, has a perfectly smooth surface, and 

 exhibits distinct prismatic structure. Fragments of ironstone placed 

 on the sole of a furnace cracked off in shells, leaving spheres ; and 

 quartzite under similar circumstances flaked off in the same way. 



Mr. Judd noticed the interesting bearing of the numerous facts 

 described in this paper on the theories concerning the production of 

 columnar structures propounded by Prof. James Thomson and Mr. 

 Robert Mallet. He, however, deprecated the introduction by the 

 author of the question of the fissile structure of phonolites (which 

 microscopic examination proved to be the result of peculiarities in 

 the arrangement of the crystalline constituents of the rock) as 

 having any relation to the structures resulting from contraction. 



Mr. Rutley stated that one of the felstone dj^kes in the Lake- 

 district showed spurious cleavage in the direction of the dyke. 



Mr. Murphy remarked that as the cooling commenced at the 

 surface, the percolation of water between the first formed columns 

 would aid in the production of the columnar structure lower down. 



The Author, in reply to Mr. Judd, said that he did not regret 

 having introduced the Auvergne phonolites into his paper, for he was 

 convinced that their structure could not be explained by pressure ; 

 he considered that' the intricacy of the arrangement shown by his 

 diagrams negatived the supposition. Separation of the minerals, as 

 described by Mr. Judd, would not necessarily be fatal to his theory ; 

 but in the case of the Roche Sanadoire, he had examined a section 

 to see if it existed, and had not found it. 



