322 Prof. J. J. Thomson on the Discharge 



ference between the dry and damp gas is much greater in 

 the case of nitrogen than in that of hydrogen, the cathode 

 fall with platinum electrodes for damp nitrogen being 260 

 volts, that for the dry 410 volts, while in damp hydrogen the 

 cathode fall was 352, in dry 329. As far as these experi- 

 ments go, the presence of aqueous vapour seems to retard the 

 passage of the electric discharge through hydrogen; they 

 relate, how r ever, to the special case when a steady discharge 

 is passing through hydrogen at a low pressure. The effect, 

 however, is comparatively small, and it seemed worthy of 

 trial whether this comparatively close agreement between 

 the properties of the damp and dry gas was maintained under 

 other circumstances of the discharge and under higher pres- 

 sures. For this reason I have compared the potential-dif- 

 ferences required to produce a spark of given length through 

 damp and dry hydrogen at different pressures. I prepared 

 the hydrogen in the same way as Warburg, using a Kipp's 

 apparatus with zinc and dilute sulphuric acid (the proportions 

 of acid to water being about 1 to 8 by volume). The gas thus 

 obtained w 7 as passed through an alkaline solution of potassium 

 permanganate, which, as well as the acid in the Kipp, had 

 previously been w r ell boiled so as to expel the air. Leduc 

 has shown that this simple process gives very good hydrogen. 

 The hydrogen was allowed to run for some hours through the 

 tubes in which the sparks w<ere to pass so as to expel the air. 

 I was never able, however, to get rid of the last trace of air, 

 as I always found that the gas, if allowed to bubble for some 

 time through an alkaline solution of pyrogallic acid, produced 

 a brown coloration. 



In order to compare the spark-potential through the damp 

 and the dry gas two bulbs communicating with each other 

 were filled with hydrogen, each of these bulbs contained a 

 pair of equal electrodes placed at the same distance apart ; the 

 distance between the electrodes varied in different experi- 

 ments, ranging from about J to -^ of a millimetre. One of 

 these bulbs, which we shall call A, was furnished with a train 

 of two drying-tubes, one containing stick caustic potash, the 

 other phosphorus pentoxide, the hydrogen passed through 

 these tubes on its way to the bulb A. The phosphorus 

 pentoxide had been previously heated to redness in an 

 earthenware crucible, the crust at the top was taken off', and 

 only the residue, which acquired a granular structure as it 

 cooled down, was used; the floor of the bulb A was also 

 covered with this pentoxide. The hydrogen in the other 

 bulb (B) had passed straight into the bulb after bubbling 

 through the permanganate solution, and no drying materials 



