166 Dr. E. H. Hall on the " Rotational Coefficient " 



Laying off the values of 33 on the base-line, and taking the 



values of -= ? plotted on a convenient scale, as ordinates, we 



have curve (1). It will be seen that this curve is nearly straight 

 for a considerable distance, and that if this portion were ex- 

 tended backward it would pass very near the origin. Between 

 the points corresponding to 23 = 5000 and 53 = 8000 the line 

 tends strongly to the right, and thenceforward it continues as 

 if asymptotic to some horizontal line not very far above. 



Base-line of r$\ 

 curve for ^L 



Base-line of ( 2 ) 

 curve for3 x 



Base-line of (1) 



E' 



!b4bh 



curve for 



23 



3000 4000 5000 6000 



V' 



7000 



E / 



9000 10000 J 1000 



The points marking the highest values of -^ do not fall so 



well in line as one might wish ; but by looking at the table it 

 will be seen that there were considerable variations of tempe- 

 rature accompanying these observations ; and to these varia- 

 tions the irregularities can perhaps be in some part attributed. 



W 

 We see now at once from the diagram that ^ is not pro- 

 portional to S3, the magnetic induction in the nickel. Can we 

 find any magnetic quantity to which it is more simply related ? 

 If we turn to the observations of Prof. Rowland on nickel*, 

 we find that they can, as he says, be plotted in several ways. 



Phil. Mag. Aug. 1873 and Nov. 1874. 



