Point-Discharge in Mid'tures of Gases. 17 i^ 



The formation of a positive brash was observed on introduc- 

 ing vapours of CSo, ether, ethyl alcohol, benzole ; acetone, no 

 effect, or only a very slight one, was produced by chloroform 

 and methyl iodide. Taking into account the vapour-pressure 

 of the liquids, alcohol, benzole, and acetone are most effective, 

 CSo and ether must be added in greater quantity. 



The actions of two ditferent admixtures to a third o-as 



o 



are not independent of one another. Thus, if a brush is 

 produced in air by adding a little GO2, this brush will 

 disappear again if so much CI2 is added as would form the 

 brush in absence of CO2, so that the addition of CL has the 

 same effect as adding more CO2. So it is not possible to 

 lower the potential-difference still more by adding a third gas. 



If a small platinum disk as cathode was placed opposite to 

 the point, which was made movable by passino- the glass tube 

 carrvino- it throuo^h a bit of rubber tubinor drawn over the 

 narrow end of the vessel, a well-developed brush could be 

 obtained by just once quickly turning round the stopcock 

 separating the vessel from the Kipp's apparatus in which the 

 CO2 was generated. The electrodes could be separated as 

 much as 20 mm. without the brush ceasing to appear, whilst 

 in pure air the electrodes had to be brought to a distance of 

 2 mm. from one another before a brush-discharge occurred, 

 and then sometimes sparking would set in. Thus this small 

 amount of CO2 gives a brush about ten times as long as that 

 in air. If air is drawn throuo-h the vessel, the lono- brush 

 disappears immediately. When the brush is well-developed, 

 one sees that it starts from a bright point and, widening a 

 little and losing in intensitv, ooes straioht across to the 

 cathode, where there is an increase in luminosity. In the 

 way the brush goes straight to the part o£ the cathode just 

 opposite it differs from the brush in pure air, which can be 

 produced to considerable length by slightly lowering the 

 pressure and introducing a small spark-gap into the circuit. 

 In this case the brush, ^hich is very unsteady, forms a curve 

 having its end on a prominent point on the sharp edge of the 

 cathode. Adding a little COo will immediately make the 

 brush go straight to the middle of the cathode. 



It would be rash at present to venture an explanation of 

 the behaviour of the mixtures. I would only like to point 

 out that the potential-difference of the discharge depending 

 upon the amount of ionization and the mean free path in the 

 gas, an easily ionized gas, which requires a higher potential- 

 difference because of its smaller mean tree path, might be 

 expected to lower the potential-difference in another gas of 

 longer mean free path bv furnishing more ions, if it is only 

 ' N2 



