﻿110 Dr. W. H. Eccles on the Effect of Electrical 



the outermost layers of the iron wires employed. The cores 

 of the iron wires have therefore not been used. 



Other and great difficulties arise in the matter of pro- 

 ducing oscillations of determinate and invariable character. 

 Maurain, in the greater part of his work on this subject, 

 appears to have used the oscillations that passed through 

 a helix in series with a Leyden jar kept sparking stronglv 

 and continuously by means of an induction-coil. Russell*, 

 in some recent experiments, applied to his iron the oscillations 

 passing through a coil connected directly in series w ith a small 

 induction-coil. This last method appears to the writer to 

 subject the iron to very violent treatment of a nature not 

 easily described accurately ; for how far the mere surgings 

 of secondary current overwhelm in importance the genuine 

 oscillations, it will be difficult to say. Piola f, again, worked 

 with highly damped oscillations, because, he found, such 

 oscillations produce the highest effects. He has, indeed, 

 following Rutherford, used this fact in determining the damp- 

 ing factor of a circuit. But in the present investigation 

 these sources of indeterminateness have, as far as might be, 

 been avoided by using a single train of waves instead of a 

 continued torrent of such trains, and by using oscillations 

 as little damped as possible. 



It would not be to the point to extend this brief survey of 

 previous work till it included that of Walter, of Ascoli, 

 of Arno, and of Foley ; for all these have paid especial 

 attention to oscillations whose magnetic field was in some 

 degree transverse to the main magnetic field. The present 

 paper deals solely with oscillations whose magnetic field is 

 along the direction of the principal field. 



In these experiments fairly soft Swedish charcoal iron, aged, 

 and not freshly annealed, was taken a large number of times 

 round any chosen magnetic cycle (as in ordinary magnetic 

 testing) till what may be called a cyclic state was attained. 

 The field was then given any desired value, and the iron 

 submitted to a single train of oscillations. This was managed 

 by generating the oscillations on a helix surrounding the 

 iron wire. The consequent alteration in pole strength was 

 observed by the deflexion of a magnetometer mirror. These 

 processes were all repeated a number of times for each 

 selected point of the cyclic curve. The figures given in the 

 tables below are thus each the mean of a number of 

 observations. 



In detail, the apparatus adopted consisted of two straight 



* Proc. Roy. Soc. of Edin. Nov. 20, 1905. 

 t Elettricista, iv. p. 145, May 15, 1905. 



