on the Atomic Weights of Bodies. 185 



tempted to determine the specific gravity of oxygen gas, 

 have come to this result. 



The first experimental determination of the specific gra- 

 vity of oxygen gas, which has any pretension to accuracy, 

 is that of Kirwan, related in his "Essay on Phlogiston" p. 25. 

 He procured the oxygen gas from the red oxide of mercury, 

 and dried it by leaving it for 24 hours in contact with con- 

 centrated sulphuric acid. He obtained 1*103 for the specific 

 gravity. 



He informs us that the goodness of this oxygen gas was 

 such, that, when one measure of it was mixed with two 

 measures of common air over water, the gaseous residue 

 unabsorbed amounted to j^ths of a measure. If this trial 

 was made (as it is probable it was) in a glass tube, whose 

 diameter was under an inch, it will indicate that the oxygen, 

 whose specific gravity was taken, contained the tenth part 

 of its volume of azotic gas. Or, admitting that half the 

 residue was nitrous gas, the oxygen would still contain 

 5 per cent, of azotic gas. Now, the specific gravity of such 

 a mixture would be 1-1041, which is almost exactly that 

 obtained by Mr. Kirwan. 



About the year 1806, Biot and Arago determined appa- 

 rently with great care the specific gravity of five different 

 gases, and they state that of oxygen gas to be 1-10359, 

 which coincides with the determination of Kirwan. But we 

 have complete evidence that the determination of these dis- 

 tinguished philosophers cannot be perfectly accurate. For 

 they give the specific gravity of azotic gas 0-96913. Now, 

 if air consists of 21 volumes of oxygen and 79 volumes of 

 azotic gas, according to the common determination, the 

 specific gravity of common air deduced from that of the 

 two gases, as determined by these gentlemen, would be not 

 1 as it Ought to be, but - 9973766 ; shewing clearly, that the 

 specific gravity, either of the oxygen or azotic gas, or both 

 of them, is somewhat below the truth. If air be a com- 

 pound of 20 volumes of oxygen and 80 volumes of azotic 

 gases, (as I have shewn from experiment), the specific 

 gravity of air deduced from that of the oxygen and the azotic 

 gas, as determined by these gentlemen, would be 0-9960220. 

 It is clear, therefore, that the specific gravities, as deter- 

 mined by Biot and Arago, cannot be perfectly exact. 



