174 Rev. P. J. Kirkby on Union of Hydrogen and Oxygen 



enough to reduce the errors of observation to within a very 

 few per cent. In this way the quantity of electricity passed 



Fig. 2. 

 A G 



r 



through was known, while the pressures before and after 

 its passage were determined by one of the McLeod gauges. 

 If these are denoted by p , p h the mean pressure p during 

 the experiment is given by 



P=i(Po+Pi), 

 while the fall of pressure Ap is given by 



A P=Po~Pi- 



Sufficient electricity was, as a rule, passed through the gas 

 to make Ap large enough not to involve errors of more than 

 2 or 3 per cent. At pressures lower than those recorded below 

 this was not possible without sacrificing the accurate determi- 

 nation of the quantity in coulombs : for the current ceases at 

 these low pressures to be steady for more than a very short 

 time. 



In some of the earlier experiments a voltmeter of resistance 

 510 ohms, occasionally halved by a shunt of 510 ohms, was 

 used as the ammeter. For the rest a high-resistance voltmeter 

 of 38,000 ohms was employed. The current was, therefore, 

 considerably varied. 



If X is the potential-difference of the parallel plates during 

 the passage of the steady current C, L the voltage of the 

 battery, and R the whole ohmic resistance of the circuit 

 distinct from the gas, then obviously 



X = Z-CR, 



the battery having no appreciable internal resistance. 



The first set of experiments were made with two parallel 

 zinc electrodes AB, CD, the circle of metal surrounded by the 

 guard-ring being 3 cm. in diameter. These are recorded in 

 Tables I., II., III., IV., in which, to recapitulate, 



