42 BRIDGMAN. 



to tension ; the resistance of lithium, calcium, and antimony increases, 

 as is normal, under tension. 



When I started these measurements, only one metal was known 

 which was abnormal with respect to tension; this was nickel. The 

 measurements were made by Tomlinson ^ in 1876, on nickel which 

 presumably had several per cent of impurity. Since it was my good 

 fortune to obtain through the kindness of the Leeds and Northrup Co. 

 some nickel of exceptionally high purity, I repeated the measurements 

 of Tomlinson, and also extended them to find the effect of temperature 

 and cyclic changes of tension. I have verified the sign of the effect 

 found by Tomlinson, although there is not close numerical agreement, 

 as was to be expected. 



Since cobalt is in many respects closely related to nickel, I also 

 determined the effect of tension on the resistance of it. The effects 

 are entirely normal. 



In addition to the five abnormal pure metals mentioned above, the 

 alloys manganin and therlo are also abnormal with respect to their 

 pressure coefficients. I have determined the tension coefficients of 

 these also, and find them to be normal in sign, but to be very small. 



My previous measurements of the pressure coefficient of resistance 

 have suggested certain views as to the nature of the conduction 

 mechanism.^ In the following I shall discuss how far these new facts 

 are in accord with these views. 



This discussion demands a knowledge of Young's modulus. I have 

 determined this for most of these metals. 



Description of Method and Apparatus. 



The apparatus and method were very simple, and for the most part 

 were similar to those already adopted for the measurement of the effect 

 of pressure. 



The resistance was measured by a potentiometer method, the drop 

 of potential between two potential terminals attached to the specimen 

 being balanced against the drop due to the same current flowing 

 through an appropriate combination of known resistances. The 

 details of the apparatus were the same as those previously used in 

 measuring the pressure coefficient, and have been fully described 



2 H. Tomlinson, Trans. Roy. Soc. 174, 1-172, 1883. 



3 P. W. Bridgman, Phys. Rev. 9, 269-289, 1917, and 17, 161-194, 1921. 



