In the course of his researches on the last-named air he showed, that when a succession of electric sparks is passed through this gas a great increase in the volume of the gas occurs. This fact was further examined at a later time by Berthollet, who, by measuring the increase in volume undergone by a measured quantity of ammonia gas, and determining the nature of the gases produced by the passage of the electric sparks, proved that ammonia is a compound of hydrogen and nitrogen, and that three volumes of the former gas combine with one volume of the latter to produce two volumes of ammonia gas.
Priestley's experiments on "inflammable air"--or hydrogen--are important and interesting. The existence of this substance as a definite kind of air had been proved by the accurate researches of Cavendish in 1766. Priestley drew attention to many actions in which this inflammable air is produced, chiefly to those which take place between acids and metals. He showed that inflammable air is not decomposed by electric sparks; but he thought that it was decomposed by long-continued heating in closed tubes made of lead-glass. Priestley regarded inflammable air as an air containing much phlogiston. He found that tubes of lead-glass, filled with this air, were blackened when strongly heated for a long time, and he explained this by saying that the lead in the glass had a great affinity for phlogiston, and drew it out of the inflammable air.
When inflammable air burns in a closed vessel containing common air, the latter after a time loses its property of supporting combustion. Priestley gave what appeared to be a fairly good explanation of this fact, when he said that the inflammable air parted with phlogiston, which, becoming mixed with the ordinary air in the vessel, rendered it unable to support the burning of a candle. He gave a few measurements in support of this explanation; but we now know that the method of analysis which he employed was quite untrustworthy.
Thinking that by measuring the extent to which the _phlogistication_ (we would now say the _deoxidation_) of common air was carried by mixing measured quantities of common and inflammable airs and exploding this mixture, he might be able to determine the amount of phlogiston in a given volume of inflammable air, he mixed the two airs in glass tubes, through the sides of which he had cemented two pieces of wire, sealed the tubes, and exploded the mixture by passing electric sparks from wire to wire. The residual air now contained, according to Priestley, more phlogiston, and therefore relatively less dephlogisticated air than before the explosion. He made various measurements of the quantities of dephlogisticated air in the tubes, but without getting any constant results. He noticed that after the explosions the insides of the tubes were covered with moisture. At a later time he exploded a mixture of dephlogisticated and inflammable airs (oxygen and hydrogen) in a copper globe, and recorded the fact that after the explosion the globe contained a little water. Priestley was here apparently on the eve of a great discovery. "In looking for one thing," says Priestley, "I have generally found another, and sometimes a thing of much more value than that which I was in quest of." Had he performed the experiment of exploding dephlogisticated and inflammable airs with more care, and had he made sure that the airs used were quite dry before the explosion, he would probably have found a thing of indeed much more value than that of which he was in quest; he would probably have discovered the compound nature of water--a discovery which was made by Cavendish three or four years after these experiments described by Priestley.