VII.] COMPOSITION OF PURE WATER. 113 



in a dry vessel, its interior becomes bedewed with the con- 

 densed moisture. 



It may seem unsatisfactory to look at a few drops of 

 limpid liquid obtained by the union of oxygen and hydro- 

 gen at the lecture-table, and believe that they are really 

 pure water. Experiments on a large scale have, however, 

 been made, and sufficient water collected to place the 

 matter beyond possibility of doubt. The grandest experiment 

 of this nature was made by three eminent French chemists, 

 Fourcroy, Vauquelin, and Seguin. The experiment com- 

 menced on the 13th of May, 1790, and was completed on 

 the 22nd of the same month. During this time the ap- 

 paratus was constantly watched, the experimenters sleeping 

 alternately, for a few hours, on mattresses in the laboratory. 

 The combustion was maintained almost uninterruptedly 

 for 185 hours; 25,964 cubic inches of hydrogen and 12,571 

 of oxygen were consumed, and the union of these gases 

 produced 7,244 grains of liquid. With this quantity at 

 their disposal, they tested the liquid in every way that 

 could be suggested, and found it to be identical with dis- 

 tilled water. 



There is yet another means of determining the chemical 

 composition of water, which needs to be briefly explained, 

 since it furnishes the most accurate of all methods for 

 determining this important question. Transmit a stream of 

 pure dry hydrogen gas over a weighed quantity of pure oxide 

 of copper (a compound of oxygen and the metal copper) 

 heated to dull redness. Under these circumstances the 

 hydrogen takes oxygen from the oxide, and forms water, 

 which can be collected and weighed, whilst the loss of 

 weight which the oxide suffers must represent the quantity 

 of oxygen which this amount of water contains. It is 

 needless to enter into the details required to secure accuracy 



T 



