20 proceedings: philosophical society 



transformations were demonstrated. These experiments are de- 

 scribed at length in the following: Bureau of Standards Scientific 

 Paper No. 236 (Burgess and Kellberg on Electrical Resistance) and 

 No. 296 (Burgess and Scott on Thermoelectric Power) also in abstract 

 in this Journal (6:650. 1916). 



Discussion. Mr. L. J. Briggs asked what method was used to get 

 the iron pure from the electrode. Mr. Burgess stated that after 

 prying off the iron from the electrode it was melted in a crucible of 

 magnesia and then drawn out after melting in vacuo; the stock of pure 

 iron is kept in a vacuum. Mr. White spoke of the difficulty of ob- 

 taining homogeneous iron and asked whether the iron used was thermo- 

 electrically homogeneous. Mr. Wright referred to recent experiments 

 at the laboratory of the General Electric Company in which the analy- 

 sis of the structure of iron was made by the use of X-rays, and dis- 

 cussed the crystalline structure of silicon-iron and electrolytic iron. 



By invitation Mr. L. H. Adams then gave an illustrated communi- 

 cation on The effect of positive and of negative pressures on the resist- 

 ance of metals. The effect of pressure on the resistance of metals was 

 first noticed by Chwolson. His work and that of Lussana and others 

 were briefly reviewed, and a short account was given of the various 

 attempts that have been made, notably by Koenigsburg and by Griin- 

 eisen, to place the variation of electrical resistance with pressure 

 upon a theoretical basis. Pure hydrostatic pressure apparently al- 

 ways decreases the resistance of pure metals by an amount which 

 varies from about 1 to 30 parts per million per atmosphere. Many 

 alloys, however, exhibit a positive pressure coefficient of resistance. 

 The change in resistance of metals under pressure finds an important 

 practical application in the measurement of very high pressures. 

 Pressure gauges may be constructed of a coil of manganin or of "therlo" 

 wire with an appropriate method for measuring small changes of re- 

 sistance. Such gauges show no hysteresis and there is apparently no 

 upper limit of pressure to which they may be used. Changes in re- 

 sistance are also observed when tensile stresses are applied to metallic 

 wires. Results were shown for the metals copper, platinum, silver, 

 and "therlo." Now, since a tensile stress may be resolved into a 

 negative hydrostatic pressure and two sets of shearing stresses, and 

 since, moreover, preliminary experiments showed that shearing 

 stresses produced no changes in the resistance (at least none greater 

 than 10 -8 per kg./cm. 2 ), it would seem reasonable that there should 

 be a simple relation between the two coefficients of resistance — hy- 

 drostatic and tensile — provided only that each coefficient be properly 

 corrected for the known elastic deformation. Results for four metals, 

 however, failed to confirm this supposition. But it is worthy of men- 

 tion that the corrected tension coefficients are always less and always 

 opposite in sign to the corrected hydrostatic coefficients. 



Discussion. Mr. Hersey asked whether the change of resistance 

 on stretching was permanent. Mr. Sw t ann asked whether any meas- 

 urements had been made in the plane of torsion. Mr. Dickinson 



