1882.] On the Movement of Gas in " Vacuum Discharges" 453 



Experiment. 



( h ^ Theory. 



I. II. III. 



Carbon 35-03 .... 34-54 .... 35 09 . . . . 34-82 



Hydrogen 3 41 .... 3 35 .... 3-35 .... 316 



Nitrogen 370 .... „ .... „ .... 370 



Chlorine „ „ 27*4 28-09 



Platinum 26'09 .... 26-8 .... 26'9 .... 26'0 



The Specimen I was the pulverulent yellowish chloroplatinate, 

 simply precipitated and dried in the vacuum. The others had been dis- 

 solved in boiling water, and were thereby slightly altered, as the ex- 

 cess of platinum proves. When quinoline is heated with an equivalent 

 quantity of ethylene chlorhydrine without adding water, a dark purple- 

 reddish mass is obtained. Extracted with ether, dried and treated 

 with absolute alcohol, this mass gives a dark violet solution, which, if 

 ether is poured on it, deposits an almost black mass, which finally 

 crystallises. The crystals, pressed between sheets of paper, are sensibly 

 less coloured than the mother-liquor, which imparts to the paper a 

 dark violet colour. 



I have not yet concluded the analysis of this product, and I propose 

 to continue these researches in various directions. 



IV. " On the Movement of Gas in * Vacuum Discharges.' " By 

 William Spottiswoode, P.R.S., and J. Fletcher Moulton, 

 F.R.S. Received March 25, 1882. 



In the preparation of tubes for our experiments it was often noticed 

 that, after the exhaustion had been carried to a certain degree, the 

 passage of a strong current had the effect of increasing the pressure. 

 This appeared to be due to an expulsion of gas from the terminals 

 themselves by the passage of the discharge. And accordingly the use 

 of such currents from time to time during the process of exhaustion 

 was adopted for making the vacuum more perfect and more permanent 

 than otherwise would have been the case. On the other hand, it was 

 also noticed, that after the tube had been taken off the pump and sealed 

 in the usual way, the passage of a strong current had in some instances 

 the effect of decreasing the pressure. We thus met with two effects, 

 apparently due to the same cause, but diametrically opposite in 

 character. 



The fact of the tube being on the pump or off it did not appear to 

 be at all material to the question, because the first effect could be 

 obtained when the tube was temporarily shut off by a stopcock. Nor 

 indeed did either the first or the second effect depend upon the absolute 



