284 



t K H. HaVr- 



■Effect of Magnetic Force on the 





P., not in absolute 



measure. 



The values of M are 



in abso- 



e measure, as usual 









Date. 



M. 



Tr. 



p. 



Dec. 31,1887, 



5000 



69-5 



139 



u a a 

 tc a a 

 u u u 

 a a a 



8300 

 11900 

 14700 

 19300 



99' =(50 + 49) 

 125' =(61-5 + 63-5) 

 131- =(60 + 71) 

 147- =(59 + 88) 



119 



105 



89 



76 



Jan. 4, 1888, 



7800 



99- =(46 + 53) 



127 



a a a 

 a a tt 



13200 

 17600 



134- =(59 + 75) 

 ]65- =(67 + 98) 



101-5 

 94 



June 27, 1888, 

 " 25, " 

 " 26, " 



2260 

 4600 

 8840 



34-7 = (l5-9 + 18'8) 

 66* =(33 + 33) 

 103-6 = (48-3 + 55-3) 



154 

 144 

 117 



" 25, " 

 " 26, " 



9800 

 11900 



113- =(51 + 62) 

 128- =(58 + 70) 



115 



K)7 



It appears from this table, which is certainly not entitled to 

 full confidence in its details, that the P. P. of bismuth dimin- 

 ishes continually while M increases from 2260, the lowest value 

 used, to 19300, the highest value used, the decrease being 

 about one-half of the whole. 



During the experiments in June a single loop of wire 

 placed upon the bismuth strip was used, as such a loop had 

 been used with iron, to test the magnetic induction through 

 the metal. This quantity remained throughout proportional to 

 the intensity of the magnetizing field, as it was expected to do. 



In a field of about 5000 a test was made in which the direct 

 current was varied from about *12 to about -5 absolute units. 

 The transverse current was found to be nearly, at least, pro- 

 portional to the direct current. 



An unsuccessful endeavor was made to discover a permanent 

 effect upon the equipotential lines in bismuth. The effect, if 

 it existed, must have been very much less than one per cent of 

 the temporary effect. The strength of field used was not 

 recorded. 



I have made no attempt to determine the effect of change of 

 temperature on the R. P. of this metal. 



In considering the fact that the P. P. of bismuth is very 

 large, it should be remembered that the R. P. states the ratio 

 of the transverse difference of potential to the direct current. 

 This ratio is a useful one and is easy to find, but in order to 

 indicate the amount of rotation, or change of direction of the 

 equipotential lines, we need rather the ratio of the transverse 

 difference of potential to the direct rate of fall of potential 

 lengthwise of the strip examined. Bismuth has a very great 

 electrical resistance, and the rate of fall of potential which 



