136 



NATURE 



[July 28, 1923 



provided with modifications for the extraction of the 

 rare gases from the atmosphere. In operating this 

 apparatus the air was compressed to 40 atmospheres, 

 and in a series of tests it was found that about 300 

 kilograms of liquid air could be made per day. With 

 such a supply of liquid air available, ample provision 

 was made, it will be seen, for meeting the needs in 

 regard to liquid air of all departments of the Uni- 

 versity. 



Liquid Hydrogen Apparatus. 



The equipment for liquefying hydrogen included a 

 four-stage belt-driven compressor built by the Burck- 

 hardt Engineering Works of Basle, Switzerland. Its 

 cylinders were water-cooled, had a forced lubrication, 



Hydrogen liqiielier (as installed). 



and were fitted with steel piston rings. The pistons 

 were all in line and constituted one shaft. The gas 

 was cooled after each compression by means of a 

 number of heat exchangers immersed in a tank of 

 running water. The compressor was constructed so 

 as to prevent any loss of gas, and with this end in 

 view, the piston rods were provided with special 

 stuffing boxes in which the packing was sealed with 

 oil contained in specially designed holders. 



The space behind each piston as well as the safety 

 valves was directly connected with a gasometer and 

 through the latter to the intake of the compressor. 

 The compressor had a capacity of 60 cubic metres of 

 free gas per hour and required a motor of 30 kilowatts 

 to operate it when delivering at 200 atmospheres 

 pressure. Twenty litres of water per minute were 

 disposed of by the heat exchangers. 



The hydrogen liquefier is shown schematically in 



NO. 2804, VOL. 112] 



Fig. 1 and as it was installed in the laboratory by 

 Fig. 2, The regenerator coils indicated were similar to 

 those used in the well-known Hampson apparatus for 

 liquefying air. In operating the liquefier, hydrogen 

 specially purified was compressed to 150-200 atmo- 

 spheres and cooled to - 205'^ C. by means of liquid 

 air boiling under reduced pressure. 



The compressed hydrogen passed successively through 

 the coils Lj, Lj', Lj, L3, L4, and L5. The coils L, and 

 Lj' were arranged in parallel and the valve Z served to 

 regulate the proportion of gas that went through 

 each of them. This ensured the proper interchange 

 of heat between the oncoming compressed gas and 

 the outgoing low-pressure vapours. The coils Lj, L,, 

 and Lj were cooled by gaseous hydrogen returning to 

 the gasometer from the expansion nozzle Cj, and the 

 coils Lj' and L3 by the evaporated air drawn off by 

 the vacuum pump. The coil L4 was partly immersed 

 in a bath of liquid air held in the flask Mj. 



The valve A served to admit more liquid air from 

 the reserve supply whenever the indicator Ej of the 

 cork float E showed that it was required. To add to 

 the efficiency of the liquefier, the expansion coil L5 

 was provided with a close-fitting German silver 

 envelope which when properly wrapped with flannel 

 permitted a good junction to be effected between the 

 inner wall of the silvered vacuum flask Mg and the 

 coil. This ensured that the expanded gas passed over 

 the closely wound tubes of the coil and so brought 

 about a good exchange of heat. 



The liquid hydrogen as it formed passed through 

 the opening in the bottom of the flask Mj and was 

 collected in the silvered flask M3. The float indicator 

 D, Dj, Dg, served to show the level of the Hquid in 

 this collecting flask. The weight D was connected 

 with the thin German-silver float Dj by means of a 

 silk thread running over three pulleys D, provided 

 with jewel mountings. The valves B and B^ were 

 used for drawing off the liquid. These were arranged so 

 that they could be pre-cooled by cold gaseous hydrogen 

 as it was returned to the gasometer. The stuffing 

 boxes and screw thread of the valves B, Bj, C and A 

 were so arranged that they were not exposed to 

 cooling and in this way the danger of a freeze-up was 

 eliminated. 



The insulation of the apparatus was specially 

 studied. Vacuum flasks were used where possible, 

 and wherever parts were cooled below the temperature 

 of liquid air they were surrounded by an atmosphere 

 of dry hydrogen or by a partial vacuum in order to 

 avoid unnecessary condensation. All parts were con- 

 structed of German silver where it was an advantage 

 to do so on account of its low thermal conductivity. 

 The entire apparatus was packed in natural wool and 

 enclosed in a thin brass case that was sealed except 

 for the drying tubes H and H^. These tubes served 

 to equalise the internal and external pressures on 

 the case and at the same time prevented water vapour 

 from entering and condensing inside. Fig. i shows 

 plainly the arrangement for supporting the apparatus 

 together with the scheme of the pipe connexions. 

 Mercury traps J and Jj served to protect the apparatus 

 at all times from any sudden but moderate excess of 

 pressure, while the large rubber safety valves G and 

 Gi served to accommodate any sudden but violent 



