RELATION BETWEEN THE VARIATION OF RESISTANCE IN BISMUTH. 



245 



Plate IX. 



Length, 



Breadth, 



Thickness, 



60-1 

 30-3 

 3-1785 



n = 1-4052 

 Temp. 14° C. 



Field. 



1,340 



3,350 



5,030 



6,700 



8,800 



11,300 



14,750 



17,780 



Trans. Effect. 



- 



0469 



- 



0943 



- 



1202 



- 



1505 



- 



1700 



- 



1879 



- 



1999 





2072 



An 



J An 



•0240 

 •1020 

 •1760 

 •2975 

 •4190 

 •5823 

 ■7735 

 •9870 



•1549 

 •3194 

 •4195 

 •5454 

 •6473 

 •7631 

 ■8795 

 •9985 



Average 



Cj(A?i)2 + c. 2 (An) : i = E. 



C, C 



•30 



+ •097 



•30 



+ •107 



•30 



+ •098 



•30 



+ •101 



•30 



+ •101 



•30 



+ •100 



•30 



+ •098 



•30 



+ •100 



transverse effect and of the variation of resistance must be made ; with these we obtain 

 two points on our line. For any other variation of resistance value we can then find 

 the transverse effect and resolve it into a pure Hall effect and this second effect. 



Evidently the graphic method could also be applied to determine c x and c 2 in the 

 first instance, instead of the method of equations used in this paper. 



The next plates used were two in which the transverse effect attains a maximum 

 numerical value. 



In these two plates also, which were made from the same two supplies of bismuth, 

 we have two effects. 



Finally, two plates were considered in which the second effect is so great as to 

 completely mask the true Hall, so that even with the low fields at disposal the 

 transverse effect changes sign. 



The results of experiments with Plates Ib. and VI. by a different method : — viz., 

 that described in paper, " On Hall Effect and some Eelated Effects in Bismuth," were 

 qualitatively the same. For Ib. the values of c x and c 2 were respectively d =' — "45 

 c 2 = +4*4 and for VI. c x = -'238 c. 2 = +3-3. 



We may sum up the results so far obtained by saying that in the pure bismuth 

 plate the Hall effect alone is present : it is proportional to ( A nf and is negative in sign. 

 In the other plates the transverse effect is composed of two effects, the pure Hall and 

 another, positive in sign and proportional to (Aw)-. In different plates this effect 



