1899.] on Liquid Hydrogen. 7 



such a jet could be used to cool bodies below the temperature that it 

 is possible to reach by the use of liquid air, but all attempts to collect 

 the liquid hydrogen from the jet in vacuum vessels failed. No other 

 investigator improved on my results,* or has indeed touched the 

 subject during the last three years. The type of apparatus used in 

 these experiments worked well, so it was resolved to construct a much 

 larger liquid-air plant, and to combine with it circuits and arrange- 

 ments for the liquefaction of hydrogen. This apparatus took a year to 

 build, and many months have been occupied in the testing and pre- 

 liminary trials. The many failures and defeats need not be detailed. 

 On May 10, 1898, starting with hydrogen cooled to — 205°, and 

 under a pressure of 180 atmospheres, escaping continuously from the 

 nozzle of a coil of pipe at the rate of about 10 to 15 cubic feet per 

 minute, in a vacuum vessel doubly silvered and of special construc- 

 tion, all surrounded with a space kept below - 200°, liquid hydrogen 

 commenced to drop from this vacuum vessel into another doubly 

 isolated by being surrounded with a third vacuum vessel. In about 

 five minutes, 20 cc. of liquid hydrogen were collected, when the 

 hydrogen jet froze up, from the accumulation of air in the pipes 

 frozen out from the impure hydrogen. The yield of liquid was 

 about 1 per cent, of the gas. The hydrogen in the liquid condition is 

 clear and colourless, showing no absorption spectrum, and the menis- 

 cus is as well defined as in the case of liquid air. The liquid must 

 have a relatively high refractive index and dispersion, and the density 

 appears at first sight to be in excess of the theoretical density, 

 namely 0*18 to 0* 12, which we deduce respectively from the atomic 

 volume of organic compounds, and the limiting density found by 

 Amagat for hydrogen gas under infinite compression. A preliminary 

 attempt, however, to weigh a small glass bulb in the liquid made 

 the density only about ■ 08, or half the theoretical. My old experi- 

 ments on the density of hydrogen in palladium gave a value for the 

 combined element of 0*ti2, Not having arrangements at hand to 

 determine the boiling point other than a thermo-junction which gave 

 entirely fallacious results, experiments were made to prove the ex- 

 cessively low temperature of the boiling fluid. In the first place 

 if a long piece of glass tubing, sealed at one end and open to the air 

 at the other, is cooled by immersing the closed end in the liquid 

 hydrogen, the tube immediately fills where it is cooled with solid air. 

 A small glass tube filled with liquid oxygen when cooled in liquid ny- 

 drogen is transformed into a bluish white solid. This is a proof that 

 the boiling point of hydrogen is much lower than any temperature pre- 

 viously reached by the use of liquid nitrogen evaporating in vacuo, 

 seeing oxygen always remains liquid under such conditions. A first 

 trial of putting liquid hydrogen under exhaustion gave no appearance 

 of transition into the solid state. When the vacuum tube containing 

 liquid hydrogen is immersed in liquid air so that the external wall 



* 'Proceedings of the Chemical Society' (No. 158), 1895. 



