EECENT PROGRESS IN PHYSICS. 415 



kept ignited for a longer time, but the gas could not be increased 

 beyond a very limited quantity. 



The experiment just described was repeated and the gas bubble 

 transferred to another tube, the wire was then again ignited in vapor, 

 the bubble formed again removed, until a sufficient quantity of gas 

 was collected for analysis, which required the labor of ten hours. 

 This gas was now detonated in a eudiometer and left a residue of 

 0.35 of its original volume, which consisted of nitrogen. The experi- 

 ment was repeated several times with the same result; sometimes a 

 trace of oxygen was found in the residue. 



Here electrolysis was completely excluded ; the wire was ignited in 

 dry steam. 



When in the apparatus of fig. 55 the 

 sparks of a large hydro-electric machine ^'°" ^''" 



were passed between platinum points 

 through the vapor, a small bubble of de- 

 tonating gas was also formed. 



As in the previous experiments a whole 

 day's work did not increase the bubble, 

 but when it was transferred, another in- 

 stantly formed. The gas similarly col- 

 lected detonated and left a residue of 0.4 

 of its original volume of nitrogen with a trace of oxygen. 



By an estimation, which could of course only be approximate, the 

 detonating gas formed, was found to be about y^Vo" ^^ the volume of 

 the vapor. 



Grove considered this evolution of detonating gas not to be a spe- 

 cific effect of electricity at all, but of heat alone, and indeed, succeeded 

 also in decomposing aqueous vapor merely by heat without electricity. 



Omitting the less successful experiments, we shall at once proceed to 

 those that gave very decisive results. With a constant battery of 30 

 zinc-platinum cells the end of a thick platinum wire was melted into 

 a globule of the size of a pepper corn ; between this and the carbon 

 point of the negative pole the voltaic arc was taken until the gobule 

 was again near its melting point. It was then rapidly plunged into 

 water, freed from air, that was kept boiling by means of a spirit lamp, 

 and into which a tube filled with the same water was inverted. Sepa- 

 rate bubbles of gas rose into the tube. This process was repeated 

 until a sufficient quantity of gas was collected, which, after explosion, 

 once left a residue of 0.4 ; another time only 0.25 of the original 

 volume, consisting, as usual, of nitrogen and traces of oxyen. The 

 galvanic battery here served evidently only to bring the platinum to 

 ignition. When melted and heated by means of the oxy-hydrogeu 

 blow-pipe, it acted exactly in the same manner. In this way more than 

 ^ cubic inch of detonating gas was obtained. 



The heated globule is evidently, when immersed in the water, 

 immediately surrounded by a stratum of vapor, from which then the 

 small quantity of detonating gas is developed. 



