ATTAINMENT OF VERY LOW TEMPERATURES. 29 



communicates directly with the pipe C leading from the gasometer to 

 the compressor, and through it the hydrogen which returns from the 

 liquefier through the tube G, figs. 3, 4, 6, L fig. 8 and a fig. 9, can 

 be returned to the circulation system. The hydrogen which escapes 

 from the liquefier when the expansion is first commenced, is allowed 

 to escape through the cock O', which is closed before the cock is 

 opened. The pipe y serves not only to discharge the water delivered 

 through a, but also as a safety valve. 



XII. The Preparation of the Hydrogen. 



In my earlier experiments I prepared the hydrogen by the action of 

 dilute (25 percent) sulphuric acid on commercial zinc, contained in a 

 beer barrel, which had received several coats of enamel paint. Since, 

 however, it appeared that the gas suffered both loss and contamination 

 by diffusion through the walls of the barrel, I replaced it by a lead 

 vessel. The apparatus for generating the hydrogen is shown in fig. 10. 



The vessel A, which was 500 mm. high and 300 mm. in diameter, 

 was made of sheet lead, " burnt " at the junctions in the oxy-hydrogen 

 blow-pipe. At the commencement of each operation it usually con- 

 tained about fifteen kilograms of pure granulated zinc. Dilute sul- 

 phuric acid containing little copper sulphate was run into ^ in a steady 

 stream from the reservoir B, and as the zinc sulphate solution accumu- 

 lated it was run off through the stop-cock E. The hydrogen could 

 either be allowed to escape through the pipe C, which was plunged into 

 a deep vessel filled with water, and acted as a safety valve, or to pass 

 along the pipe D into the purifying towers F, F'. 



The towers F, F', F" were filled with broken pumice, which was 

 kept wet by means of solutions of the reagents contained in the reser- 

 voirs G, G'. The first of these, which supplied the first two towers, 

 contained an acid solution of chromic acid, the second contained silver 

 nitrate solution, which served to remove arseniureted hydrogen from 

 the gas. The excess of the reagents flowed through inverted syphons, 

 which acted as gas traps, at the bottom of the towers, and was collected 

 in the vessels H, H'. 



The gas finally passed through a solution of caustic potash contained 

 in the vessel K and entered the pipe leading to the gasometer through 

 the cock L. 



If the gasometer already contained some hydrogen this gas could be 

 used to wash out the generating apparatus before introducing the fresh 

 quantity of gas. This could easily be done by first lowering the reser- 

 voir /, so as to empty the vessel K, and thus raising the tube C till the 

 gas bubbled through the water in which it was immersed, and 

 escaped. 



