PEIRCE. — MAGNETIC TESTS UPON IRON AND STEEL RINGS. 87 



Physical Laboratory, Teddington, Middlesex, England, has so success- 

 fully done. If in such a case great accuracy is required, the work has 

 to be carried out with considerable care and some attention has to be 

 paid to the fact that there is a real, if usually small, difference between 

 the value of B corresponding to the mean //, and the mean value 

 of i^. 



I have had occasion of late to determine the permeability of a small 

 ring of extremely pure soft iron, and have found it helpful to compute 

 by the aid of accurate HB diagrams, previously made for two or three 

 different kinds of iron and steel in the form of long rods, what the dis- 

 crepancy {B' — B") would be for these materials at different excita- 

 tions, if they were made into rings of the dimensions of the one I was 

 compelled to use. This paper gives some results which seem instruc- 

 tive, for a very soft kind of Norway wrought iron and for a specimen 

 of Bessemer steel fairly typical of what one meets with in practice. 



The straight rods used were magnetized and demagnetized in a uni- 

 form solenoid about five meters long, consisting of 20904 turns of well- 

 insulated wire wound on a stout, solid-drawn brass tube through which 

 a stream of tap water could be kept running about the rod to prevent 

 any sensible rise of temperature. The axis of the solenoid was hori- 

 zontal and perpendicular to the meridian. The flux of induction in 

 the rods was measured by means of a test coil of fine wire wound on 

 the rod at its centre. This coil was protected by rubber tape and its 

 leads were insulated from the water by rubber tubes of fine bore slipped 

 over them. The ballistic galvanometer employed had a period so long * 

 that no detectable error was introduced into the readings by the fact that 

 a measurable time was needed to make the magnetic changes incident 

 to a reversal of current in the solenoid. The rods were demagnetized by 

 means of a long series of currents in the solenoid, alternating in direc- 

 tion and gradually decreasing in intensity ; and the fact that this pro- 

 cess was successful showed that the rods were practically homogeneous 

 throughout. The rods were so long that the corrections for the ends, 

 as given by du Bois or by Shuddemagen,^ were very small. 



Tables I and II give the results of determinations ® of corresponding 

 values of // and B made by the method of ascending reversals by Mr. 

 John Coulson and myself. A number of diagrams were obtained for 

 each rod to make sure that the rather elaborate apparatus for demag- 

 netizing the specimens was effective and that the metal was practically 



* Peirce, These Proceedings, 44, 283 (1909). 



» C. L. B. Shuddemagen, These Proceedings, 43, 183 (1907). 



* Peirce, American Journal of Science, 27, 273 (1909). 



