ATTAINMENT OF VERY LOW TEMPERATURES. 27 



(figs. 4, 6, 8) to the main hydrogen supply pipe. The cock r was not 

 opened till the liquefaction of the hydrogen had commenced. 



The tube L was usually surrounded either by a large vacuum 

 vessel or by three concentric glass shades, spaced apart with asbestos 

 mixed with a little pentoxide of phosphorus to absorb moisture from 

 the surrounding air. This arrangement prevented the deposition of 

 moisture on the walls of the tube L and enabled one to see what was 

 taking place inside. 



When the pressure in the apparatus had risen to 150 atmospheres 

 the expansion valve was slowly opened and for a few moments the 

 hydrogen was allowed to escape from the circulation system, either by 

 bubbling through the water in the tank 0, fig. 8, or by opening the 

 cock O' (fig. 9). When sufficient hydrogen had been allowed to 

 escape, and the apparatus was considered to be free from air, the cock 

 O' (fig. 9), was closed, the cock O was opened and the hydrogen was 

 allowed to circulate through the system. 



Shortly after commencing the expansion a small quantity of solid 

 impurity always separated in the bottom of the vacuum-vessel H (figs. 

 4, 6). It is noteworthy that the solid impurity never blocked the coil, 

 though it is probable that the latter was cooled to below the critical 

 point of hydrogen, but seemed to encrust the valve at the point at 

 which the expansion took place. Hannay* pointed out some years ago 

 that gases under high pressure, particularly in the neighborhood of 

 their critical points, were able to hold solids in solution ; considering the 

 complete continuity of the liquid and gaseous states of matter this is 

 hardly remarkable. To this phenomenon, which is allied to that of 

 retrograde condensation in the case of mixed gases, is due the fact that 

 the coils are never blocked by the presence of impurities, which at the 

 temperatures to which the gas is cooled should behave as non-volatile 

 solids, but which are dissolved by the compressed gas in the coil. 



When the liquefier is working well it should require about half an 

 hour from the time of commencing operations to fill a receiving vessel 

 {K, figs. 4, 6) holding about half a liter. When once liquid hydrogen 

 is formed, and the cock r is opened, the liquid falls in large drops from 

 the nozzle of the vessel H. The liquid is usually slightly milky at first, 

 but the solid impurity soon settles to the bottom of the receiving vessel, 

 leaving the liquid clear. 



When the receiving vessel is full the tube L can be lowered, and 

 the receiving vessel drawn out of it by means of a wire attached to it. 

 The receiving vessel is at once plugged with wool and placed inside a 

 second vessel containing liquid air. 



In machines which required solid carbonic acid about four kilos of 

 that substance was required at each experiment, together with about 



* Proceedings of the Royal Society, 1879, Vol. 30, 484. 



