1895.] Argon, a new Constituent of the Atmosphere. 277 



is thus freed from any possible contamination with oxygen, hydrogen, 

 or hydrocarbons, and nitrogen is gradually absorbed. As the amount 

 of gas in the tubes and reservoir diminishes in volume, it draws sup- 

 plies from the gas-holder, and, finally, the circulating system is full 

 of argon in a pure state. The circulating system of tubes is con- 

 nected with a mercury pump, so that, in changing the magnesium 

 tube, no gas may be lost. Before ceasing to heat the magnesium 

 tube the system is pumped empty, and the collected gas is restored 

 to the gas-holder ; finally, all the argon is transferred from the mer- 

 cury reservoir to the second small gas-holder, which should prefer- 

 ably be filled with water saturated with argon, so as to prevent 

 contamination from oxygen or nitrogen ; or, if preferred, a mercury 

 gas-holder may be employed. The complete removal of nitrogen 

 from argon is very slow towards the end, but circulation for a couple 

 of days usually effects it. 



The principal objection to the oxygen method of isolating argon, 

 as hitherto described, is the extreme slowness of the operation. In 

 extending the scale we had the great advantage of the advice of Mr. 

 Crookes, who not long since called attention to the flame rising from 

 platinum terminals, which convey a high tension alternating electric 

 discharge, and pointed out its dependence upon combustion of the 

 nitrogen and oxygen of the air.* The plant consists of a De Meritens 

 alternator, actuated by a gas engine, and the currents are tranformed to 

 a high potential by means of a Buhmkorff or other suitable induction 

 coil. The highest rate of absorption of the mixed gases yet attained is 

 3 litres per hour, about 3000 times that of Cavendish. It is necessary 

 to keep the apparatus cool, and from this and other causes a good 

 many difficulties have been encountered. 



In one experiment of this kind, the total air led in after seven days' 

 working, amounted to 7925 c.c., and of oxygen (prepared from 

 chlorate of potash), 9137 c.c. On the eighth and ninth days oxygen 

 alone was added, of which about 500 c.c. was consumed, while there 

 remained about 700 c.c. in the flask. Hence the proportion in which 

 the air and oxygen combined was as 79 : 96. The progress of the 

 removal of the nitrogen was examined from time to time with the 

 spectroscope, and became ultimately very slow. At last the yellow 

 line disappeared, the contraction having apparently stopped for two 

 hours. It is worthy of notice that with the removal of the nitrogen, 

 the arc discharge changes greatly in appearance, becoming narrower 

 and blue rather than greenish in colour. 



The final treatment of the residual 700 c.c. of gas was on the model 



of the small scale operations already described. Oxygen or hydrogen 



could be supplied at pleasure from an electrolytic apparatus, but in 



no way could the volume be reduced be]ow 65 c.c. This residue 



* ' Chemical News,' vol. 65, p. 301, 1892. 



