Dr. C. Fromme on the Magnetism of Steel Bars. 299 



in the k corresponding to an equal magnetizing force X, while 

 II. 5 decidedly shows an increase in h when RM rose from 

 1931000 to 6213000. The reverse is the case with II. 1, 

 where k diminishes. 



Since the different series of observations were not carried 

 out under absolutely similar external circumstances, especially 

 not with equal distance of the helix from the magnetometer, 

 no conclusion concerning the behaviour of k can be drawn 

 from such slight variations, which are moreover not all in the 

 same direction, and we shall be able to regard k as constant 

 for the harder sort II. 



In sort I. it can be proved more certainly that, when RM is 

 quintupled, the h of 1. 1 is considerably augmented, just so that 

 of I. 3 with RM doubled, while for I. 4 and I. 5 nothing cer- 

 tain can be affirmed. 



If we consider the results before us in their relation to Neu- 

 mann and Kirchhoff 's developments, we perceive that in the 

 present case these are no longer valid: to the same argument 

 K, with the different ellipsoids the most different values of h 

 correspond. 



Neumann and Kirchhoff' s developments hold only for steel bars 

 which possess no permanent magnetism and are submitted to but 

 slight magnetizing forces ; they at once become inadmissible as 

 soon as the steel is permanently magnetic. 



This is much more strikingly apparent with the softer than 

 with the harder kinds of steel. 



When, lastly, we examine the series of observations of Table 

 IV. singly, we see the analogy with soft iron carried out also 

 in the behaviour of the magnetization-function : — 



For the forces which are incapable of altering the permanent 

 magnetism of a steel bar, the number k becomes less as the force 

 diminishes ; the magnetization-function of steel exhibits the same 

 behaviour as that of soft iron under the action of the always 

 slight magnetizing forces which have here been employed. ' 



(For intenser forces, as with soft iron, an increase of h 

 would of necessity enter with a diminishing force.) 



In what manner this diminution of k takes place, I have en- 

 deavoured to ascertain by a graphic representation of the de- 

 pendence of k upon its argument K. I found, however, some- 

 what complicated curves with several turning-points, which 

 appeared to become the more regular (viz. straight lines) the 

 more I augmented the permanent magnetism. 



Now, since by this, on account of the diminished constancy 

 of the quantity of the permanent magnetism, the difficulties 

 of observation are increased, and therefore properly a still 

 more irregular course of the curves would have been to be 



