243 



In the hypothesis of Mr. Kirwan, the most obvious difficulty 

 seems to be in admitting the prompt subsidence of the finely 

 divided particles of so light a substance as bitumen, through the 

 waters of the ocean. Mr. Kirwan supposes tlie carbonaceous, as 

 well as the bituminous particles, to be separated by the disinte- 

 gration of the antediluvian mountains ; and then to enter into 

 reunion, and suffer a precipitation through the water. Such a 

 combination I have endeavoured to prove is constantly taking 

 place, in every case where the thickening of the substance and the 

 darkening of the colour, of any bituminous substance is going 

 on. Thus, in consequence of the addition of oxygen to the thin 

 and limpid naphtha, a combination of that principle with the 

 carbon ensues ; and the fluid becomes thicker and darker, assum- 

 ing the form of petroleum. By a continuance of this process, 

 these effects are produced in a still higher degree, and the hard, 

 dark bitumen, asphaltum, or the still more highly carbonaceous 

 bitumen jet, is formed. But every one of these substances, we 

 perceive, are of such a degree of specific gravity, when pure, as 

 not to allow us to assume the possibility of their sinking through 

 the moist, pulpy, argillaceous mass, as Mr. Kirwan describes it; 

 neither could it possibly be admitted, even if such a subsidence 

 were granted, that such materials should form, as Mr. Kirwan also 

 conjectures, the lowest stratum. 



Whilst on this subject, I must make one more observation, which 

 appears to me very forcibly to contradict this supposition of the 

 formation of coal by deposition. The specific gravity of the sub- 

 stances forming the roof of coal, constantly exceeds in a very high 

 degree that of the bituminous mass they cover; they containing 

 very frequently so much metallic matter as to render them exceed- 

 ingly ponderous. Now the disintegration, of which Mr. Kirwan 

 speaks, he supposes to have been so complete, as to have converted 



