OBSERVATIONS ON ^ASALT, &C. 175 



thrown down by the failure of the ground on which it ftood, 

 fomelimes by the violence of the waves, and not nnfrequenlly 

 by the working of quarries. In moft infiances, thefe operations 

 have only removed the covering from one fide of the colonnade; 

 and it remains crowned, and generally furrounded, by an 

 immenfe amorphous mafs. Where there are two ranges of 

 column?, with an intervening amorphus ftratum, it is probable 

 lhat the upper is the refult of a fecond inundation of fluid 

 bafalt. It is well known that bafaltic columns are moft (olid 

 at the bottom ; and their convex articulations have been re- 

 peatedly obferved. Since thefe confederations occurred to me, 

 I have had no opportunity of examining, whether the divifions 

 approach nearer to the plane furfaces as they recede from the 

 centre from which the prifms was generated, nor whether 

 below that centre the convex furface of the articulations is 

 inverted ; but I think it by no means improbable, that fubfe- 

 quent obfervations may eftablifh this to be the cafe, and thus 

 confer on this hypothesis nearly all the demonftration of which 

 it is fufceptible. I may however add, that the phenomena Bafaltic veins, 

 of prifmatic divifion in bafaltic veins, perfectly coincide with 

 what might be inferred from the data upon which my reafoning 

 has proceeded. In veins, it is obvious that the refrigerating 

 or abforbing caul'e muft operate with nearly equal force on 

 each fide of the vein ; and it follows, lhat two fets of prifms 

 would be generated, which would be horizontal inftead of 

 perpendicular, and that, unlefs a mafs of amorphous bafalt 

 was interpofed between them, they muft form a divifion in 

 the middle of the vein, as, from the mutual impenetrability 

 of their fibres, they could not incorporate. The coincidence 

 of the exifting phenomena with thefe conclufions, is fuffici- 

 ently remarkable ; for, in numerous obfervations I have made 

 on the bafaltic veins which efFe£i the prifmatic configuration, 

 I found the prifms were always horizontal, and often, that 

 there were two ranges of them. One of their ends applied 

 to the wall of the veins, the other frequently united to an 

 amorphous mafs which feparated them; and, when no fuch 

 intermedium occurred, there was invariably a divifion in the 

 middle of the vein. Not unfreqnently, the veins contain three 

 fets of prifms; a range of final I ones on each fide, and of 

 much larger ones in the middle. In this cafe, the little prifms 



are 



