at the Royal Institution, 1908-1916 751 



together, the fact was demonstrated that the nitrogen boils off from 

 the liquid before the oxygen. The separation of the two gases by 

 fractional distillation has long been carried out on the large scale^ 

 both as a means of producing oxygen and also in preparing nitrogen 

 for use in the production of cyanamide. 



The most striking demonstration given in this lecture was that 

 showing the attraction of liquid oxygen to the poles of a magnet. 

 A cup made of rock-salt, containing some liquid oxygen, was placed 

 between and a little below the poles of a large electro-magnet ; 

 whenever the electric circuit was completed the liquid rose from the 

 cup and connected the two poles. To those of us who first witnessed 

 the demonstration, though in accordance with our theoretical expec- 

 tations, the way in which the liquefied oxygen suddenly leapt at and 

 clung to the poles of the electromagnet was a very wonderful sight 

 — a marvellous and fit culmination of advance in low-temperature 

 technique. 



The lecture on "Liquid Atmospheric Air" (January 20, 1893) 

 is memorable as the occasion on which the Thermos FlasJc was first 

 introduced to public notice and its efficiency demonstrated. The con- 

 struction of the flask was explained and the principles underlying 

 the production of high vacua were fully discussed,- especially the 

 vacua secured with the aid of mercury vapour. The period was also- 

 entered upon during which such remarkable use has been made at 

 the Institution of liquid air, as a cooling agent, in the investigation 

 of physical properties generally at low temperatures. On this occa- 

 sion reference was made to the results obtained, in conjunction with 

 Prof. Fleming, in an investigation of the electric conductivity of 

 metals at very low temperatures ; also to an investigation, carried 

 out in conjunction with Prof. Liveing, of the optical properties of 

 liquefied gases. The direct liquefaction of air was eft'ected before 

 the audience by means of liquid oxygen boiUng under a pressure equal 

 to that of about half an inch of mercury. 



The subject of chemical action at low temperatures was also 

 broached and, by one of those ionrs deforce so characteristic of the 

 lecturer, sulphur was shown to burn below liquid oxygen. Altogether, 

 on account of the variety of topics considered and the striking novelty 

 of the effects demonstrated, the lecture was one of the most remark- 

 able of the unrivalled series of discourses given by Prof. Dewar in 

 illustration of his pioneer low-temperature investigations, for which 

 the Institution is so noted. 



In the lecture delivered a year later (January 19, 1894) on 

 " Scientific Uses of Liquid Air," in which again a very wide field was 

 covered, many illustrations were given of further uses made of liquid 

 air and of Hquid oxygen, of which the latter was often obtailied from 

 the former by allowing the more volatile nitrogen to evaporate away. 



