NOV. 19, 1921 bridgman: the discontinuity of resistance 455 



The foregoing illustrations show a few of the equations for specific 

 cases of spurious correlation that can be derived from the general 

 formula. Similar expressions may be obtained from equation (8) 

 for the spurious correlation between any two functions. The value 

 of the spurious correlation involved should always be considered 

 when drawing conclusions from the coefficient of correlation of any 

 two index numbers regardless of their functional form. 



PHYSICS. — The discontinuity of resistance preceding supraconductiv- 

 ity} P. W. Bridgman, Jefferson Physical Laboratory, Harvard 

 University. 



The phenomenon of supraconductivity has to the present time been 

 found in mercury, tin, thallium, and lead.- The resistance of the first 

 three of these metals changes discontinuously by a factor of the order 

 of 10** at a definite critical temperature. This temperature is 4.2° K. 

 for Hg, 3.78° for Sn, and 2.3° for Tl. The probability is that the 

 resistance of lead decreases discontinuously in the same way on 

 entering the supraconducting state, but the temperature has not yet 

 been definitely measured because it is in the range between 4.3° 

 and 20° which cannot as yet be controlled by a suitable cryostat. 

 Onnes estimates that the temperature of discontinuity of lead is about 

 6°K. 



The discontinuous change of resistance has often been supposed 

 to be in some way intimately connected with the supraconducting 

 state. For instance, J. J. Thomson has a theory accounting for the 

 sudden disappearance of resistance.^ Now if such a phenomenon 

 of discontinuity occurred at an ordinary temperature, one would 

 almost certainly look for a polymorphic change as the cause of the 

 discontinuity. The thesis which I wish to support in this paper is 

 that the discontinuous change of resistance on entering the supra- 

 conducting state is also a mark of a polymorphic change, and that 

 discontinuity and supraconductivity are not as intimately* related 

 as sometimes supposed, but are due to quite distinct mechanisms. 



According to this view, the normal condition in an individual 

 crystal grain of any metal at very low temperatures is supraconduc- 

 tivity. Any ordinary metal is an aggregate of crystal grains in ran- 



' Received October 12, 1921. 



' C. A. Crommelin. Phys. Zeitschr. 21: 274, 300, and 331. 1920. 



3 J. J. Thomson. Phil. Mag. 30: 192-202. 1915. 



