430 Scientific Proceedings, Royal Dublin Society. 



region where the linear variation of magnetization with sin a begins to fail. 

 It must be admitted that this is a stringent test of the formula, though the 

 same symmetry with stronger fields in neighbourhood of 45° was not obtainable, 

 even over a smaller angular range. This is explained, however, by the 

 variation in susceptibility with direction, noticed in the observations on 

 magnetization, becoming more important with stronger fields. 



The value of s = 2 - 5 gives a second value for the demagnetizing factor, 

 since we have NI L = s = 2-5, and from fig, 4, I L = 710, giving iV= 3'5 x 10" 3 , 

 which is in satisfactory agreement with the value 3 3 x 10~ 3 , obtained from the 

 observations on anhysteretic spiral magnetization, thus confirming the validity 

 of this method of correcting for the demagnetizing factor in experiments on 

 spiral magnetization, since the latter value was obtained from observations 

 on longitudinal magnetization alone. 



SUMMAKY. 



Observations have been made on the "Wiedemann effect in a steel tube in 

 which the twist is produced by a combination of circular and longitudinal 

 fields, which gave a spiral field whose axis coincided with the axis of the tube. 

 The intensity of the field was kept constant, while the pitch-angle of the spiral 

 was varied. It is shown that the twist varies as the sine of twice the pitch- 

 angle, thus verifying a formula given by Knott in 1888. 



The longitudinal magnetization of the tube was examined under the same 

 type of spiral field, and was shown to vary approximately as the sine of the 

 pitch-angle. Deviations from this law were shown to be due to the demag- 

 netizing effects of the ends, and in a lesser degree to the fact that the tube 

 is not magnetically isotropic. 



Two methods, which show a satisfactory agreement with each other, are 

 given for obtaining the demagnetizing factor for a steel tube. 



The author has much pleasure in expressing his obligations to Professor 

 Brown, for the use of the solenoid described in this paper. The initial stages 

 of the work owe much to the assistance of Mr. O'Callaghan, A.E.C.Sc.I. ; and 

 Mr. Moore, A.E.C.Sc.I., materially contributed to its progress by a series 

 of tedious magnetic measurements. 



