vii THE BURNING OF INFLAMMABLE AIR 127 



The water was collected in a smaller bulb, followed by a 

 series of drying tubes, similar to those used to dry the 

 hydrogen. The water could not be weighed in a vacuum, 

 but the observed weight was corrected for the buoyancy of 

 the air which it displaced. In all, more than a kilogram of 

 water was collected. 



In a typical experiment, using phosphoric anhydride as 

 the drying agent, the figures were : 



Exhausted bulb and copper oxide ... ... 673*280 grams. 



,, ,, copper 613-492 ,, 



Oxygen = 59788 



Bulb and drying tubes 93i'487 ,, 



,, ,, ,, with water 998700 ,, 



67-213 

 Air correction ... ... 0*069 



Water = 67-282 



Hydrogen combined with I part of oxygen =0*12533 ,, 

 Ditto, corrected for air in the sulphuric acid = 0*12508 ,, 



From nineteen experiments Dumas concluded that the 

 ratio of hydrogen to oxygen lay between 0*1250 and 0*1256. 

 This gives the ratio of oxygen to hydrogen as 8*00 to 7*96, 

 or taking a mean value, 



oxygen : hydrogen = 7 '98 : 1. 



Scott's determination of the composition of water. 



The figures obtained from the laborious experiments of 

 Dumas remained unchallenged for almost half a century. 

 In 1893, however, Alexander Scott (Phil. Trans., 1893, 184, 

 543-568) obtained an independent value for the composition 

 of water by reverting to the method of finding the composi- 

 tion by weight from the densities and combining volumes of 

 the gases. The densities had been determined with very 

 great care by Lord Rayleigh. Scott measured the com- 

 bining volumes by exploding the moist gases after carefully 



