150 Dr. F. Auerbach on the Passage of 



nounced in iron the first time it is used galvanically, and 

 mostly vanishes after the tenth to the hundredth closing. 

 Thus the result with an iron wire (7=1670, d=0*20, H = 2D, 

 M=1D) was:— 



Initially ?<* = 7-3501 



After the 10th closing . . 7-3539 



„ 100th „ . . 7-3552 (const.) 



Longitudinally magnetized . 7*3547 

 Magnetized ten times . . 7*3541 

 Demagnetized . . . . 7-3559 



These numbers show the afteraction with transverse mag- 

 netization by the current as well as with longitudinal mag- 

 netization. That the decrease of resistance in consequence of 

 the latter appears so trifling (it became much greater at equal 

 values of H, M, n after the relations of the wire had become 

 stationary), evidently depends on this — that the two after- 

 actions, as indeed is mostly the case with these phenomena, 

 do not go on independently of one another, and therefore, in 

 consequence of the accelerated afteraction of the frequently 

 repeated transverse magnetizing, a part of the decrease of re- 

 sistance is concealed. 



According to the preceding, it appears a probable supposi- 

 tion that the permanent alterations of resistance ivith the mag" 

 netic state, produced by magnetizing, may also be taken as the 

 expression of certain performances of work. In fact, on a cur- 

 rent passing through an iron wire, constant rotations of the 

 molecules will be produced, which at an alteration of the 

 arrangement, such as is produced by magnetizing, cannot 

 remain unchanged. If these speculations are of a very hy- 

 pothetical character, they nevertheless, I think, afford more 

 fixed points than those of Beetz. 



Attempts have been made by several to gain new fixed 

 points by artificial transverse magnetization of the iron passed 

 through by the current. I have not hitherto pursued this 

 idea experimentally, — first, because we have not, as might be 

 inferred from the hypotheses tacitly assumed by the authors 

 in question, simply transverse opposed to longitudinal mag- 

 netization, but much rather here again, according to the ar- 

 rangement, many different kinds of magnetization are conceiv- 

 able (wherefore I have, for the sake of definiteness, designated 

 the one here discussed as circular magnetization); and, next, 

 because these experiments require still greater delicacy of 

 measurement than experiments with longitudinal magnetiza- 

 tion. The theory requires that artificial circular magnetizing 

 (that is, additional to that produced by the principal current), 



