Point- Discharge in Mixtures of Gases, 177 



small amount of CO2 into the vessel, the star immediately 

 stretched out into a thin, steady brush without ramifications, 

 which reached in a continuous curve to the wall of the sur- 

 rounding cylinder. At the same time the electrometer 

 showed an appreciable fall in the potential-difference. 



The relation of the effect to the amount of CO2 introduced 

 was next investigated. A measured quantity of air was 

 pumped out of the vessel and CO2 let in until the original 

 pressure was re-established. Thus a series of observations 

 reaching from pure air to pure CO2 was obtained, and also 

 one in the reversed order, starting from CO2 and introducing 

 always more and more air. The approximate equality of 

 the potential- differences in mixtures of equal composition 

 obtained in these two ways showed that the gases had really 

 mixed homogeneously ; to effect this one had to wait some 

 time. Removing the CO2 contained in atmospheric air by 

 passing it through caustic potash had no appreciable effect 

 on the discharge, this quantity being too small. But a 

 quantity of CO^ corresponding to something like 1 mm. of 

 pressure produces the long brush and a decided fall in the 

 potential-difference from 4500 to 4200 volt. If more CO2 

 is introduced the brush contracts until, in pure CO2, there is 

 again only a star visible at the point ; the potential-difference 

 rises rapidly until for about 20 mm. of COgit reaches the same 

 value as in pure air ; then it increases, if one allows for the 

 experimental errors, nearly proportional to the amount of CO2 

 present until it reaches the value of 5500 volts for pure CO2. 



Thus it seems that a gas in w^hich the point-discharge 

 requires a higher potential-difference than in air, lowers the 

 potential-difference in air if introduced in small quantities, 

 the electric strength — if this term may be applied to the 

 point- as well as to the spark- discharge — being smaller in the 

 mixture than in either of the constituents. 



Lowering the pressure lessens the effect; the long brush 

 disappears, and the curve giving the connexion between the 

 potential-difference and the quantity of CO2 gets more smooth, 

 until the minimum can no longer be observed. 



Also when the point is negative the mixture of air and 

 CO2 shows a remarkable behaviour. In this case the intro- 

 duction of a small amount of COg causes the potential- 

 difference to rise* from 3700 to 4400 volts. For the interval 



* Warburg (loc. cit.) found a very large increase in tlie potential- 

 difference of the negative point-discbarge in nitrogen by the presence of 

 very small traces of oxygen, but then the potential-difference for pure 

 oxygen is higher than that for pure nitrogen, whilst in the case of CO, it 

 is lower than for air, and yet addition of a little COo increases the 

 potential-difference in air for the negative point-discharge. 



Phil. Mag. S. 6. Vol. 6. No. 31. July 1903. N 



