3 86 



EXPERIMENTAL INVESTIGATIONS. 



to be experimented on, a graduated test-tube filled with the 

 liquid being inverted on the extremity of each. 



From the Z form of the terminals the extremities pro- 

 jected jths of an inch into the test-tubes, so that no gas could 

 possibly escape. At first I tried distilled water, which had 

 been boiled immediately before the experiment to deprive it 

 of air ; I found, however, that from its want of conducting 

 power, only an infinitesimal decomposition was produced ; a 

 few drops of sulphuric acid were added, and immediately a 

 notable evolution of gas took place, a spark being visible at 

 the end of the wire which was worn beneath the surface of 



Fig. i. 



the glass, whether this wire were positive or negative, and 

 none on that which was even with the glass. This spark had 

 a material influence on the relative quantity of gas given off 

 by each terminal, the gas being generally less from the spark- 

 giving terminal than from that which did not yield a spark ; 

 and when both were constructed so as to give sparks, the 

 quantities of gas were irregular, and bore no definite relation 

 to each other. Thus, in one experiment the gas from the 

 negative was to that from the positive as 2-3 : i ; in another 

 experiment the gas from the positive was to that from the 

 negative as 2 '5 : I ; frequently the quantities from each were 

 nearly equal. It is obvious that when the wire was depressed 

 below the surface of the glass a film of gas must intervene 

 between the wire and the liquid, across which film the spark 

 passes. When the wire is even with the glass the gas escapes 

 more readily, and there are fewer or no sparks. 



