ON THE EFFECTS OF HEAT. 1(J7 



Thenard themselves, that potassium decomposes almost all 

 the metallic oxides with inflammation. Perhaps this in- 

 crease of" specific gravity, which potassium undergoes, may 

 be farther explained as follows. It was remarked by 

 Messrs, Gay-Lnssac and Thenard, in the experiments just 

 alluded to, that potassium decomposes the black oxide of 

 iron without inflammation. Let it be granted, that, during 

 the combination of iron with oxigen, the latter only is con- 

 densed ; and that the metal suffers no condensation* : and it 

 may, I think, be easily shown, that the specific gravity of 

 potash must exceed that of potassium ; supposing the 

 oxigen to be in the same state of condensation, as in the 

 black oxide of iron. The specific gravity of iron is 7*8 ; 

 and that of the black oxide, as nearly as I can determine it, 

 is 4*5.— -Now if the black oxide be composed, as Dr. 

 Thompson has shown, of 785 iron, and 21-5 oxigen, it 

 may be ascertained, by a very easy calculation, that the 

 sp. gr, of the oxigen, as it exists in black oxide, is nearly 

 1*7. — If then we suppose, th?(t the oxigen nnderg9e8 

 no farther condensation, wiien it combines with potassinm, 

 (which it certainly must undergo, or it would not decom- 

 pose the black oxide) still it is evident, that a mere mixture 

 of potassium, of the sp. gr. 0'6, and of oxigen in this 

 condensed state, must be of a greater sp. gr. than the potas- 

 sium alone. I have stated this, merely to show the possi- 

 bility of the case, without referehce to any theory whatever ; 

 but when we add the consideration, that the oxigen is still 

 farther condensed, and that potash is combined with water, 

 every difficulty must be entirely done away. — Permit me to The difference 

 suggest, that, as in alkalis the sp. gr, of the oxigen always oxire?andal- 

 exceeds that of the metal, and in oxides falls short of it ; kali» owing to 

 this may possibly be the cause of their possessing such dif- '^J^^ <^op^^"^a° 

 ferent properties, and, in fact, constitute the difference oxigen, 

 between theni. 



L. O. C. 



* I think no one will refuse to grant this; for it is scarcely con- 

 ceivable, that the sp. gr, of iron is greater in the black oxide, than in 

 the meuUic state. 



V. 



