496 



Mr. S. Bidwell. Changes produced by [May 19. 



that they sufficiently account for the twist which is produced in an 

 iron wire when magnetised circularly and longitudinally at the same 

 time. The resultant lines of magnetisation, as he points out, take a 

 spiral form ; the iron expands in the direction of the lines of mag- 

 netisation, and thus the wire becomes twisted. Professor G. Wiede- 

 mann, however, to whom the discovery of the magnetic twist is due, 

 appears not to be satisfied with this explanation,* believing the effect 

 to be caused by unequal molecular friction. 



The subject of magnetic twists has been very fully and carefully 

 investigated by Professor C. Gr. Knott, and in a paper published last 

 year in the ' Transactions of the Royal Society of Edinburgh ' (vol. 

 36, Part II, p. 485) he indicates many details in which the phenomena 

 of twist closely correspond with those of elongation and retraction. 

 Assuming their essential identity, and noting that " an increased 

 current along the wire affects the points of vanishing twist in a 

 manner opposite to that in which an increased tension affects it,' 7 

 Professor Knott is " inclined to conclude that the pure strain effects 

 of these influences are of an opposite character." Now, since the 

 magnetic elongation of an iron wire is known to be diminished by 

 tension, the remark above quoted amounts to a prediction that in an 

 iron wire carrying a current the magnetic elongation would be 

 increased. " We know nothing so far," Professor Knott observes, 

 "regarding the changes of length when an iron wire carrying a current 

 is subjected to longitudinal magnetising forces," and it was with the 

 object of acquiring some information on this point and testing Pro- 

 cessor Knott's prediction that the experiments described in the present 

 paper were undertaken. The results show that it was amply verified, 

 and thus Maxwell's explanation of the twist receives still further cor- 

 roboration. 



The apparatus used and the methods of observation were the same 

 as those described in my former papers. Each specimen of wire ex- 

 amined was 10 cm. long between the supporting clamps, and the 

 magnetising coil, weighing neaiiy 3 lbs., was as usual supported by 

 the wire itself, an arrangement which, for reasons before given, was 

 essential. The indications of the instrument were read to one ten- 

 millionth part of the length of the wire, and the wire was demag- 

 netised by reversals before each single observation. 



Exp. 1. — The wire first used was of soft commercial annealed iron y 

 0*75 mm. in diameter. The changes of length which it exhibited 

 under the influence of magnetising forces gradually increased from 

 13 to 315 C.G.S. units, are indicated in the second column of Table I, 

 in which the unit is one-millionth of a centimetre or one ten- millionth 

 of the effective length of the wire. The magnetising forces given in 

 the first column are those due to the coil only, no account being taken 

 * ' Phil. Mag.,' July, 1886, p. 50. 



