748 Professor Sir James Dewar [June 7, 



possess metallic lustre. The density of liquid hydrogen at its boiling 

 point ( - 252° * 5 C. or 20° • 5 Ab.), was found to be about * 07, so that 

 liquid hydrogen was about six times lighter than the lightest liquid 

 hitherto known, namely, marsh gas, whose density is 0*4. It is 

 sixteen times lighter than an equal volume of liquid oxygen, or in 

 other words the two l)odies show a greater difference in density than 

 water and mercury. The boiling points and densities are as follows : — 



In my experiments with helium in 19ol, it was expanded adia- 

 batically, using the Cailletet method, from a pressure of 100 atmo- 

 spheres at the temperature of solid hydrogen down to one atmosphere, 

 without showing any temporary mist during expansion, from which it 

 was inferred that the gas had been cooled to at least 9° and still no 

 liquefaction had occurred. Similar experiments were made by 

 Olszewski in 1005, and he has expressed the view that helium may 

 be practically a permanent gas which we should hardly ever succeed 

 in liquefying. Without offering any opinion other than what I have 

 already expressed as to the probability of helium being liquefiable, let 

 us study the application of the method to helium which was found 

 successful with hydrogen. This method, first used technically by 

 Linde, consisted in the use of a regenerative circuit along with 

 expansion through a fine nozzle (pin-hole). 



A pipe fixed in the King's Well at Bath enabled the escaping 

 gas to be collected and sent to London. This gas consists almost 

 entirely of nitrogen, but contains about one two-thousandth part of 

 its volume of helium — in other words, 2000 cubic feet of the gas 

 would contain 1 cubic foot of helium. The helium was concentrated 

 from the crude gas by partially liquefying out the nitrogen, marsh 

 gas and other impurities until it contained only about ;->0 per cent, of 

 nitrogen along with helium and neon. In this condition it was put 

 through the regenerative circuit under a pressure of about 100 atmo- 

 spheres, yielding a quantity of liquid nitrogen, which was removed. 

 The nitrogen remaining in the helium mixture had now fallen to 

 5 per cent., and on continuing to circulate the regenerator tubes got 

 plugged with solid. To the helium gas, whicli was now too small 

 a quantity to circulate well, wjis added 25 per cent, of hydrogen, 

 and tliis mixture, on circulating, froze also in the regenerator tubes. 

 A further addition of hydrogen, up to 50 per cent., Avas made to the 

 gas, which after passing through the regenerating circuit, gave a 



