124 Mr. S. Bidwell. On the Changes 'produced by 



Fig. 3. 



The points corresponding to the maximum elongation in the three 

 curves are marked with the letter E. At first sight thej seem to 

 possess no particular distinguishing characteristic in common. In- 

 deed the only points in the curves which appear to be marked by any 

 special property are those which are called by Chrystal (" Eric. Brit. v ) 

 after Wiedemann, the " turning points." Up to these points the 

 temporary moments increase with the magnetising force, or even more 

 rapidly ; after the points are passed, the rate of increase in the tem- 

 porary moments is less than that of the magnetising force. When the 

 curve does not begin to ascend in a straight line, the turning points 

 are found by drawing tangents from the origin : they are indicated in 

 the curves by the letter T,* but it is not easy to determine their posi- 

 tions with perfect accuracy. 



From a careful examination of these curves it appears probable that 

 a simple relation does exist between the turning points and the points 

 of maximum elongation, the abscissae of the points of maximum elonga- 

 tion being almost exactly equal to twice those of the turning points. 



* According to Eowland ("Phil. Mag.," 1873, vol. ii, p. 155), "the temporary 

 magnetism increases continually with the current." This may be strictly true (up 

 to the turning point) for rods or rings having the diameter of those used by 

 Rowland. Thus the curve for my thickest iron rod ascends in a perfectly straight 

 line ; but a slight convexity towards the axis of x may be suspected in the medium 

 one, and in the thinnest such convexity is quite evident. It is even more marked in 

 the curve for a wire, 0"77 mm. in diameter, given in fig. 5. 



