308 Scientijic Intel li gene;. [April 



water and Q-b cubic inches of carbonic acid gas. Now the hydro- 

 gen in 1-35 gr. of water is very nearly -qual to seven cubic 

 inches. The carbon in Qx. cubic inches of carbonic acid is equi- 

 valent to 61 cubic inches. Hence it follows that naphtha is 

 composed of 



64- or 13 volumes of carbon 

 7 or 14 volumes of hydrogen 



By substituting atoms for volumes, which may be done in this 

 case without any error, it follows that naphtha is a compound of 



13 atoms carbon = 9-75 



14 atoms hydrogen = 1-75 



11-50 



The specific gravity of the vapour of carbon is 0416, and that 

 of hydrogen gas 0-0694. Therefore, 



Q\ cubic inches of carbon weigh . . 0-82'2 gr. 

 7 cubic inches of hydrogen weigh 148 



0-970 



There is, therefore, in this analysis, a deficiency of three per 

 cent. I am disposed to ascribe this to a small portion of azote, 

 which naphtha seems to contain. But 1 have not been able to 

 satisfy myself experimentally of its existence. My experiments 

 are conducted in copper tubes, subjected to a red heat. This 

 always drives off a quantity of air, varying from 0-5 to 0*7 of a 

 cubic inch, according to the degree of heat to which the tube 

 and the peroxide of copper is subjected, and which it is not pos- 

 sible to raise always to the same degree of intensity. This air 

 always contains 3-i^d per cent, of oxygen, the rest being azote. 

 The reason of this difference between its composition and that of 

 common air is the length of red-hot copper tube through which 

 it is obliged to pass, and which is partially oxidized at the 

 expense of the oxygen of the common air present. Now 0-03 gr. 

 of azotic gas would not amount to xo^^h of a cubic inch, which, 

 being less than the variation in the quantity of air driven off by 

 heat when nothing is heated but the tube tilled with oxide of 

 copper, I have no means of determining whether so small 

 a quantity of azote is disengaged or not. 



I have observed of late, that in order to ensure accuracy in the 

 quantity of water evolved, it is necessary to expose the peroxide 

 of copper to a red heat just before making the experiment ; for 



f>eroxide of copper has the property of imbibing a little water 

 ^ rom the atmosphere, which it gives out again when heated to 

 ' tedness. 



In order to ensure absolute precision in such experiments, it 



would be necessary to have the means of raising the fire every 



me to exactly the same intensity. It is likewise necessary to 



