182 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1936 



a discontinuity as the metal passes over into the supraconducting 

 state. 



I ought here to recall that it was in the cryogenic laboratory of 

 Leiden that supraconductivity was first discovered in 1911 by Kam- 

 erlingh Onnes. He saw that a resistance apparatus of mercury 

 transformed itself practically into a short circuit at a certain tem- 

 perature. He then observed the same phenomenon for other metals, 

 as lead, tin, and thallium, and found that the resistance could be 

 reestablished by placing the supraconducting metal in a sufficient 

 magnetic field. 



Since then many metals and alloys have been examined. At 

 present one studies principally those monocrystals which give the 

 clearest results. It is found that generally the thermal conductiv- 

 ity of a substance reaches a maximum at nearly the same temperature 

 at which supra-electrical conductivity appears. 



Until 1934 no other method of lowering the temperature of helium 

 was known except that of boiling its liquid under reduced pressure. 

 In that way the temperature 0.°7l K. had been reached, and it ap- 

 peared to be exceedingly difficult to go lower. But then following 

 an idea of Debije, De Haas, Wiersma, and Kramers utilized the 

 phenomena of adiabatic demagnetization. This was also done simul- 

 taneously and independently in America by Giauque. The phen- 

 omenon appears as follows: A certain quantity of a salt of iron or of 

 a rare earth metal (paramagnetic) is placed in an enclosure refriger- 

 ated by liquid helium. On applying a magnetic field the salt be- 

 comes magnetic, and thermal energy appears which is dissipated in 

 the liquid helium. If now the salt is thermally insulated from the 

 helium bath, and the magnetic field is turned off, the salt is cooled 

 by becoming demagnetized. For if its temperature were to be pre- 

 served constant it would be necessary to supply the energy which 

 was lost in the form of heat during magnetization. The lowering 

 of temperature which thus occurs is of course a function of the na- 

 ture of the salt and of the magnetic field. It remains therefore to 

 determine the temperature thus attained, for it is impossible to be 

 measured by the pressure of a gas or of the vapor of helium, for these 

 are altogether too small at these excessively low temperatures. 



As a measure of temperatures between 1° and 4° K., the magnetic 

 susceptibility, x, of the salt is employed. In general, x follows the 

 law of Curie, x^'^C for small changes of temperature. Hence one 

 may extrapolate the curve (x, T) up to T=0, and having measured x, 

 determine T thereby. More definitely, if the law of Curie is fol- 

 lowed, and xo2^o be the values of x and T before cutting off the 



magnetic field, and if x, T, are the ultimate values, then T^ = To - • In 



this way with a chrome-potassium alum there has been obtained a 



