364 



Prof. J. Dewar. 



and temperature within the limits we are considering is unknown, and 

 no thermometer of this kind can be relied on for giving accurate 

 temperatures up to and below the boiling point of hydrogen. The 

 curves are discussed in the paper, and I am indebted to Mr. J. H. D. 

 Dickson and Mr. J. E. Petavel for help in this part of the work. 



Helium separated from the gas of the King's Well, Bath, and 

 purified by passing through a U-tube immersed in liquid hydrogen, 

 was filled directly into the ordinary form of Cailletet gas receiver used 

 with his apparatus, and subjected to a pressure of 80 atmospheres, 

 while a portion of the narrow part of the glass tube was immersed in 

 liquid hydrogen. On sudden expansion from this pressure to atmo- 

 spheric pressure a mist from the production of some solid body was 

 clearly visible. After several compressions and expansions, the end 

 of the tube contained a small amount of a solid body that passed 

 directly into gas when the liquid hydrogen was removed and the 

 tube kept in the vapour of hydrogen above the liquid. On lowering 

 the temperature of the liquid hydrogen by exhaustion to its melting 

 point, which is about 16° absolute, and repeating the expansions on 

 the gas from which the solid had separated by the previous expansions 

 at the boiling point, or 20° '5, no mist was seen. From this it appears 

 the mist was caused by some other material than helium, in all 

 probability neon, and when the latter is removed no mist is seen, 

 when the gas is expanded from 80 to 100 atmospheres, even although 

 the tube is surrounded with solid hydrogen. From experiments made 

 on hydrogen that had been similarly purified like the helium and used 

 in the same apparatus, it appears a mist can be seen in hydrogen (under 

 the same conditions of expansion as applied to the helium sample of 

 gas) when the initial temperature of the expanding gas was twice the 

 critical temperature, but it was not visible when the initial tempera- 

 ture was about two and a-half times the critical temperature. This 

 experience applied to interpret the helium experiments, would make 

 the critical temperature of the gas under 9° absolute. 



Olszewski in his experiments expanded helium from about seven 

 times the critical temperature under a pressure of 125 atmospheres. 

 If the temperature is calculated from the adiabatic expansion, starting 

 at 21° absolute, an effective expansion of only 20 to 1 would reach 

 6° *3, and 10 to 1 of 8°'3. It is now safe to say, helium has been realJy 

 cooled to 9° or 10° absolute without any appearance of liquefaction. 

 There is one point, however, that must be considered, and that is the 

 small refractivity of helium as compared to hydrogen, which, as Lord. 

 Kayleigh has shown, is not more than one-fourth the latter gas. Now 

 as the liquid refractivities are substantially in the same ratio as the 

 gaseous refractivities in the case of hydrogen and oxygen, and the 

 refractive index of liquid hydrogen is about 1*12, then the value for 

 liquid helium should be about 1*03, both taken at their respective 



