282 K K Rail— Effect of Magnetic Force on the 



attributable to the greater ease with which the thicker and 

 narrow piece of metal becomes magnetized. 



It wonld appear from the table as it stands that the thin 

 cross has a maximum P. P. when M is something between 

 1070 and 5100. but the experiments were so rough that their 

 testimony upon this point is very doubtful. 



It might also be inferred from this table that the actual 

 transverse current obtained from the thick cross had a maxi- 

 mum value when M was something less than 8300. This seemed 

 to me not impossible, but the experiments made could not be 

 regarded as proving it. On Jan. 6, 1888, I returned to this 

 matter and tested the thinner of these two crosses with mag- 

 netizing forces about 13000 and 19000 c. g. s. units respec- 

 tively. The transverse currents produced in these two cases 

 were nearly equal, whence I conclude that if there is a maxi- 

 mum value of the transverse current for any particular high 

 value of M, this maximum must be very obscure. It seems far 

 more likely that the tranverse effect continues to increase with 

 M, although very slowly indeed when M is very great. 



Bismuth. — In the summer of 1887, after I had begun pre- 

 parations for a more thorough study of bismuth than I had yet 

 made, I was obliged to leave Cambridge. Messrs. Coggeshall 

 and Stone, of the Harvard class of 1886, and Mr. W. C. Sabine, 

 of the graduate department, took up the work for me. They 

 sawed in two a thick cross of ordinary commercial bismuth 

 about 28 mm long and 3*6 mm wide, and obtained thus two crosses 

 of which one was about 0'86 mm and the other about 3'13 mm in 

 thickness. Their experiments, like those of Ettingshausen and 

 jS"ernst,* indicate that the P. P. of bismuth is independent of 

 the thickness, and show that when the intensity of the mag- 

 netizing field is increased from 4600 to 12000 the K. P. of 

 bismuth falls off very considerably. 



I have more recently made some experiments with the 

 thinner of these two crosses. I find, as did Ettingshausen and 

 is"ernst, that magnetization in one direction may produce in 

 bismuth a greater transverse effect than magnetization in the 

 opposite direction. This dissymmetry appears from my ex- 

 periments to be very small, if it exists, when the magnetizing 

 field is comparatively weak, 4500 or less. The experiments 

 of Dec. 31, 1887, did not show it in a field of 8300, or plainly 

 in a field of 11800, but those experiments were Less carefully 

 made than subsequent ones and are entitled to less credit. 

 According to these later experiments the dissymmetry is well- 

 marked in a field of 8000 or 9000 and is still greater in 

 stronger fields. It was not greatly changed by reversing the 



* 1 have seen only that abstract of their article which appeared in the Beibldtter 

 Xo. 5. 1887. 



