264. Hints for the Formation qf.a Theory of the Earth. 



or full cells that may be obferved in them ; of their different 

 accidents and decompofition. 



41. A. The phenomena they exhibit in the fire, either 

 cxpofed to, or flieltered from the a6lion of the air. But 

 before any arguments are drawn from thefe experiments, it 

 will be neceffary to refolve the queftion, Whether it be true 

 that a ftone may have been melted by fubterranean fire, and 

 yet none of its charafters exhibit thofe indications of fufiou 

 which the fire of our furnaces would have given to the fame 

 ftone. 



42. Their conneclion, if there be any, with lava, fully af- 

 certained as fuch : if it be true, for example, as M. Faujas 

 aflfirms, that currents of lava are feen terminated by columns 

 of bafaltes. 



43. The nature of the bafes on which the bafaltes refts: 

 whether, as M. Werner favs, any is found refting on wacke 

 or hornllone, of an earthy and compact frafture, which it- 

 felf repofes on fand or freeftone. 



44. Whether at other times bafaltes is feen to repofe on 

 beds of coal {houille) which preient no indications of com- 

 buftion. 



45. In a word, to examine whether the foil which bears 

 them, or the fides of the veins in which they are contained, 

 prefent indications of the aftion of fire, or at leaft of having 

 been cxpofed to the conta6l of an incandefcent raafs ; or, on 

 the contrary, whether there appear indications of dcpofits of 

 a fubftance which had been in a ftate of aqueous fluidity. 



46. Whether there are in the bafaltes veftiges of organifed 

 marine bodies or others j and in what ftate thefe veftiges are 

 found. 



47. Wliether there are obferved, as M. Faujas fays, ba^ 

 faltes which feems to have formed a pafiage for itfelf fron> 

 top to bottom through maffes of granite. 



48. In doubtful cafes of this kind it would be neceflarj', 

 were it poffible to be at the expence, to puih a gallery under 

 a rock of bafaltes, to examine \vhether the columns defcend 



below 



