E. W. Morley — Volumetric Composition of Water. 283 



its connections, could be computed the reduced volume of the 

 gas in n ; then r was heated, and the valve o was opened. 

 When about a liter of hydrogen had been taken from n, o 

 was shut, and the volume remaining was determined. The 

 heating of r was continued till the gas remaining was reduced 

 to some such volume as ten cubic centimeters, when r was 

 cooled, and the gas was extracted by the Sprengel pump. A 

 suitable excess of oxygen was extracted from its store, the 

 two gases were measured, mixed, exploded, and the residue 

 measured. From this was computed the amount of hydrogen 

 found by analysis, whence was learned by difference the 

 amount of nitrogen which was originally contained in the 

 volume known to have been extracted from n. 



To illustrate by an actual experiment : In n, before extract- 

 ing any hydrogen, the temperature and pressure were 766 '5 

 millimetres, 20*5 degrees ; after, 662*0 millimetres and 20*9 

 degrees. Hence it was computed that 823 cubic centimeters 

 had been admitted to r. When r was cold, the hydrogen 

 remaining was extracted, and transferred to the apparatus 

 shown in tig. 3, and found to be 6*722 cubic centimeters at 

 standard temperature and pressure. Oxygen was added, and 

 the sum found to be 17*101 cubic centimeters. After explo- 

 sion there remained 7*019 cubic centimeters. Hence the 

 hydrogen found in the 6*722 cubic centimeters taken for 

 analysis was 6*721 cubic centimeters. A duplicate analysis 

 agreed well with this; so that this hydrogen was practically 

 free from nitrogen. 



Determination of the volumetric composition of water. — 

 Two jars of hydrogen were extracted from n without heat- 

 ing r, and a jar of oxygen from the store of oxygen. I meas- 

 ured a convenient volume of hydrogen ; for ease of explanation, 

 suppose it was 180 cubic centimeters. About 120 cubic centi- 

 meters were transferred after measurement to a jar in the 

 cistern, and the other 60 to a second jar. Then a volume of 

 oxygen either a little smaller or a little larger was measured ; 

 suppose it was 175 cubic centimenters. After measurement, 

 it was transferred to three jars, 60 cubic centimeters to the jar 

 having 120 of hydrogen, 60 to the other jar of hydrogen, and 55 

 to a small graduated jar. Another volume of hydrogen was 

 next measured, say 179 cubic centimeters. One-third of this 

 was put into the jar into which the smaller quantity of hydro- 

 gen had been put before, and the remaining 120 cubic centi- 

 meters were left in the eudiometer. 



It will be noticed that the three measurements are made at 

 as nearly the same point as is consistent with the fact that 

 there must be a slight excess of either one gas or the other. 

 It would probably have been better to have made the three 



