63G Mr. Ellis Williams on the Influence of 



to the pressure, and he proposed to make use of this property 

 in the measurement of hydrostatic pressure. 



The results obtained by Lussana are very different from 

 these. He also found that the resistance was diminished by 

 pressure, but that the diminution was very far from being 

 proportional to the pressure, the rate of diminution with 

 pressure falling off rapidly as the pressure increased. 



The actual values of the decrease are also different from 

 those given by Lissell, and for manganin Lussana found 

 that the resistance was diminished by about 5 . 10 ~ 7 of its 

 value per atmosphere, whereas according to Lissell it is in- 

 creased by 23 . 10 ~ 7 per atmosphere. 



In view of these conflicting results it was suggested to me 

 by Prof. Rontgen that I should attempt to investigate the 

 question, and the experiments described below were under- 

 taken for that purpose. I also took the opportunity of in- 

 vestigating the influence of stress on the resistance of bismuth, 

 and obtained some interesting results which are described 

 below. 



Apparatus and Method of Experimenting . 



The hydrostatic pressure was obtained by means of a 

 Schaffner and Budenberg screw-pump furnished with a 

 Bourdon gauge reading up to 1500 atmospheres, made by 

 the same firm. The gauge was calibrated by comparison 

 with a standard manometer whose behaviour had been care- 

 fully investigated by Dr. Wagner *. 



The pump was connected to a steel cylinder 4 cms. x 80 cms. 

 inside dimensions. Into the open end of the cylinder was 

 fitted a brass piece, which served to carry the terminals of 

 the wire whose resistance was to be measured, and which 

 could be screwed down on a leather collar so as to form a 

 watertight cover to the cylinder. 



In order to diminish as much as possible the rise of tem- 

 perature which takes place on compression, water, which is 

 advantageous both on account of its small coefficient of ex- 

 pansion and its high specific heat, was used in the pump and 

 cylinder. Hence it was necessary that the wire and its con- 

 nexions should be carefully insulated. The apparatus for this 

 purpose is shown in fig. 1. A is the brass piece forming the 

 cover of the pressure cylinder, through which passes an 

 ebonite rod B, serving to insulate the copper wire C, to 

 which one end of the wire under investigation is soldered. 

 A piece of the ebonite rod about 2 cms. long and 2*5 cms. 

 wide projects into the cylinder, and a leather collar between 



* Wagner, Wied. Ann. p. 906 (1904). 



