32 REPORT— 1902. 



use in the circuit, that such machines are not capable of giving. While 

 it is theoretically clear that such machines ought to be capable of main- 

 taining the lowest temperatures, and that with the least expenditure of 

 power, it is a very different matter to overcome the practical difficulties 

 of working such machines under the conditions. Coleman kept a machine 

 delivering air at minus 83 degrees for houi's, but he did not carry 

 his experiments any further. Recently Monsieur Claude, of Paris, 

 has, however, succeeded in working a machine of this type so efficiently 

 that he has managed to produce one litre of liquid air per horse power 

 expended per hour in the running of the engine. This output is twice as 

 good as that given by the Linde machine, and there is no reason to doubt 

 that the yield will be still further improved. It is clear, therefore, that in 

 the immediate future the production of liquid air and hydrogen will be 

 effected most economically by the use of machines producing cold by the 

 expenditure of mechanical work. 



Liquid Hydrogen and Helium. 



To the physicist the copious production of liquid air by the methods 

 described was of peculiar interest and value as affording the means of 

 attacking the far more difficult problem of the liquefaction of hydrogen, 

 and even as encouraging the hope that liquid hydrogen might in time be 

 employed for the liquefaction of yet more volatile elements, apart from the 

 importance which its liquefaction must hold in the process of the steady 

 advance towards the absolute zero. Hydrogen is an element of especial 

 interest, because the study of its properties and chemical relations led 

 great chemists like Faraday, Dumas, Daniell, Graham, and Andrews to 

 entertain the view that if it could ever be brought into the state of liquid 

 or solid it would reveal metallic characters. Looking to the special 

 chemical relations of the combined hydrogen in water, alkaline oxides, 

 acids, and salts, together with the behaviour of these substances on electro- 

 lysis, we are forced to conclude that hydrogen behaves as the analogue of 

 a metal. After the beautiful discovery of Graham that palladium can 

 absorb some hundreds of times its own volume of hydrogen, and still 

 retain its lustre and general metallic character, the impression that 

 hydrogen was probably a member of the metallic group became very 

 general. The only chemist who adopted another view was my distinguished 

 predecessor, Professor Odling. In his ' Manual of Chemistry,' published 

 in 1861, he pointed out that hydrogen has chlorous as well as basic 

 relations, and that they are as decided, important, and frequent as its 

 other relations. From such considerations he arrived at the conclusion 

 that hydrogen is essentially a neutral or intermediate body, and therefore 

 we should not expect to find liquid or solid hydrogen possess the appear- 

 ance of a metal. This extraordinary prevision, so characteristic of 

 •Odling, was proved to be correct some thirty-seven years after it was 

 made. Another curious anticipation was made by Dumas in a letter 

 addressed to Pictet, in which he says that the metal most analogous to 



