1908] on the Nadir of Temperature and Allied Problems. 



415 



depends on the perfection of the vacuum ; as well as heat conduction 

 down the necks of the vessels. 



All the oxygen used in London is now obtained from the 

 fractionation of liquid air by the Linde method, and consequently is 

 much purer than the old commercial oxygen, while substantially pure 

 nitrogen can be obtained at the same time if wanted, along with the 

 rare gases. 



The cooling of air by adiabatic expansion is easily shown with an 

 apparatus of the old Cailletet type, a thermojunction in the expanding 

 air is connected to the reflecting galvanometer which shows the altera- 

 tions of temperature on the scale fixed on the wall. The gas is 

 compressed by the pump, cooled to absorb the heat of compression, 

 and then suddenly expanded. Wroblewski in 1888 first saw a mist 

 of hydrogen produced by this instantaneous method of cooling. With 

 helium, I in 1901, and Olszewski, of Cracow^ in 1905, both failed to 

 observe any condensation by adiabatic expansion. The following 

 table summarises the results of these experiments : — 



Adiabatic Expansion of Helium at Temperatuee op 

 Solid Hydrogen. 



Thus Olszewski thought he had reduced his temperature down 

 to 1'7 degree above absolute zero, yet neither he nor I saw any 

 indications of liquefaction. Olszewski, in fact, seemed to regard 

 helium as possibly non-condensable, and such an inference, if true, 

 would be ah important addition to our knowledge. There are, how- 

 ever, several facts that would help to account for the failures to 

 observe any condensation. The refractive index of helium and its 

 latent heat are both very small, the latter probably only one-seventh 

 that of hydrogen, so that it would be very difficult to see and isolate 

 as a mist. When apparatus was arranged for helium circulation, 

 similar to that used in the liquefaction of hydrogen, two difficulties 

 arose, viz. the use of a sufficient amount of pure helium gas, and 

 the supply of enough liquid hydrogen to maintain the cooling until 

 regenerative action on the helium began. For the rapid liquefaction of 

 hydrogen it is necessary to start from a temperature of 65°, and 

 regenerate down to 20\ Assuming the boiling point of helium to 



