1885.] strength of mixtures of gases. 327 



charge passed in the tube. This mode of investigation was used 

 in all the experiments, whose results are given below. 



There is another point which may be noticed. Faraday 

 originally observed that after a discharge is passed across a gas it 

 becomes electrically weaker than it was before. It may be easily 

 shewn that this alteration in the electric strength of the gas lasts 

 for some time. For instance I made the distance between the 

 movable electrodes 52 mm., and having filled the tube with 

 nitrogen I found that until it was pumped out to 10 mm. of 

 mercury no discharge appeared in the tube. After having intro- 

 duced a little more nitrogen, I passed the discharges from a 

 Ruhmkorff coil through the tube for 10 minutes ; after waiting 

 for 2 minutes, I found the corresponding pressure of nitrogen 

 12'5 mm., which shews that nitrogen is weakened by the passage 

 of discharges, and does not acquire its previous properties after 

 a rest of 2 minutes. To avoid errors arising from this cause, after 

 every determination some new gas was introduced into the tube 

 (when the gas was very rarefied sufficient to increase its pressure 

 about 30 — 40 mm.). In this manner the gas enclosed in the tube 

 was never sensibly changed by previous discharges. 



Experiments made at different times have not given exactly 

 concordant results in absolute numbers for the same gases, never- 

 theless the order of gases as regards their electric strength was 

 always the same and the ratio of numbers found for several gases 

 was nearly equal. Even for absolute numbers the accordance was 

 much better at low pressures when the discharge had the form of 

 a glow, than at higher ones when the passage of electricity is 

 accompanied by a spark. These differences are doubtless princi- 

 pally caused by the different states of free atmospheric air, by the 

 varying quantity of dust in it, and perhaps by some differences in 

 the working of the Ruhmkorff coil. On the contrary, measure- 

 ments made some hours one after another gave numbers which 

 never disagreed 10°/ even for pressures larger than 100 mm. of 

 mercury. Therefore finally I made experiments with gases, whose 

 electric strength I compared, on the same day, and I shall give 

 below only such measurements or means of such measurements 

 which were executed with this caution. 



The following is an example of measurements executed one 

 after another with oxygen, a mixture containing 61% vol. oxygen 

 and 39°/ vol. nitrogen, air and nitrogen under low pressures. 

 Oxygen which was prepared from chlorate of potash was led 

 through water and a solution of caustic potash. The carbonic acid 

 contained in the air was also removed by means of caustic 

 potash. Finally the gases were dried by passing through two 

 tubes with CaCl 2 and the third with phosphoric anhydride, each 

 of which contained a plug of wool to retain the dust. 



