312 Mr. E. H. Hall on the new Action of 



In order to reduce those results to the form since adopted, it 

 would be necessary to know the thickness of tho gold strip. 

 This thickness might be determined roughly if we knew the 

 specific resistance of the material and the actual resistance of 

 the strip, which is now destroyed. The latter value is known 

 approximately ; and by assuming the specific resistance to have 

 been that of pure gold, we might arrive at a value of the ratio 



M x V 



— p7~ • This value, however, would be very much larger than 



that obtained when thicker strips of metal are used; and facts 

 to be hereafter mentioned make it appear quite probable that 

 the thickness of the strip, as above arrived at, is several times 

 smaller than the true thickness *. 



"Without attempting, therefore, any accurate determination 

 of the constant of this first strip (A), I pass on to 



Gold Leaf, Plate (B). 



This plate also is of very thin metal ; and in general I shall 

 use the term gold leaf when speaking of the metal in this shape, 

 and use the term gold foil to denote the strips of considerable 

 thickness. 



This second plate of gold leaf was not constructed until after 

 several thick plates had been tried and found to give very dif- 

 ferent results from those obtained with the first thin plate in 

 the manner described above. Thinking that some experimental 

 error in the first measurements might account for the discre- 

 pancy, and the first plate being destroyed, I constructed the 

 second one. In making observations with this plate I first used 

 the high-resistance Thomson galvanometer, whereas the low- 

 resistance instrument had been used with the thick plates. 

 Thinking that I might in changing instruments have fallen 

 into some error, I afterwards made another series of observa- 

 tions with the same plate, but using the low-resistance galva- 

 nometer. The results were (the thickness here also being 

 estimated as above described): — 



M x V 

 March 18, with high-resist, galv., w = 622 x 10 10 



jj )) >> v » =oo< „ 



„ 19, „ low-resist. „ „ =681 „ 



Mean „ =647xl0 10 



This result is about four times as large as those found with 



* See also Albert v. Ettingshausen, " Bestimmung der absoluten Ge- 

 schwindigkeit," &c, Sitzunysberichte Akad. JVicn,To\. lxxxi. p. 446 (1880). 

 Pie found the value of the thickness indicated by the weight in similar 

 cases to "be from four to ten times as great as that indicated by the resist- 

 ance. 



