414 



Professor Sir James Dewar 



[June 5, 



laboratory a degree of cold relative to the average temperature which 

 is exactly comparable to the intensely high temperature of the electric 

 furnace or even that of the sun. 



When liquid air is dropped on to the surface of fluids at the 

 ordinary temperature, the liquid air at once assumes the spheroidal 

 state, and rushes about on the warmer liquid which has a temperature 

 three times higher. Projected by the lantern the liquid air drops 

 look like tadpoles or nuclei with cometary tails formed of the con- 

 densing mist of the vapour of the warmer liquid. This experiment 

 was first shown some two years ago. If, now, liquid 

 hydrogen is poured on to the surface of liquid air, 

 the temperature ratio of the liquids being as great 

 as 1 to 4, we anticipate the liquid hydrogen would 

 equally float about in the spheroidal state. The 

 experiment is quite successful although somewhat 

 less striking than that of liquid air on the surface 

 of carbon tetrachloride, because it all passes in a 

 few seconds owing to the small latent heat of liquid 

 hydrogen, although 10 c.c. of liquid hydrogen is 

 used in each operation. The boiling liquid air 

 surface is clouded over when the liquid hydrogen 

 is poured on to it, owing to the formation of a 

 snow of solid air, and all boiling of the liquid air 

 is arrested. The rapidity of the passage of the 

 phenomena is suggestive to some extent of the 

 difficulties which surround experimental work when 

 attempting to approach the nadir of temperature. 



As an illustration of the importance of ques- 

 tions of radiation in low temperature research, two 

 vacuum-jacketed vessels, alike in all respects except 

 that one is coated inside with silver and the other 

 with a film of highly reflecting lead sulphide 

 (Fig. 1), are connected to the same liquid-air 

 cooled charcoal exhaust, thereby ensuring and 

 maintaining an identical high vacuum in each ; 

 and both afterwards charged with the same volume 

 of liquid air. The air is allowed to distil from 

 each vessel, and the rates of distillation compared. 

 It takes thirty-five seconds for the silvered vessel 

 to boil off a volume of air measured at the ordinary 

 temperature of 50 c.c, while the lead-sulphide 

 coated vessel only takes seven seconds to evaporate a similar amount. 

 Thus the influx of heat from the sulphide of lead coated vessel 

 is just five times greater than that entering by the silver covered 

 surface. A film of nickel deposited from Mond's Xickel carbonyl is 

 nearly as good as silver. Of course the influx of heat is not merely a 

 question of radiation, but also involves gas convection ; which again 



Fig. 1. 



