Peirce — Permeabilities and Reluctivities for Steel. 277 



reversals was obtained in the usual manner. It is to be noticed 

 that the demagnetizing process does not succeed unless the 

 rod is practically homogeneous throughout its length. 



The ballistic galvanometer employed in this work has been 

 already described* in a previous paper, and it is only necessary 

 to say here that its period was so long that no detectable error 

 was introduced into its indications by the fact that four or 

 five seconds were necessary to make the magnetic changes 

 corresponding to a reversal of current in the exciting coil of 

 the yoke. For the largest currents a battery of about 120 large 

 storage cells was used ; a battery of 30 or 40 cells furnished the 

 weaker currents. 



When it is necessary to create inside an iron rod a fairly 

 uniform magnetic field of intensity much greater than, say 

 500, some kind of yokef is almost indispensable, and many 

 kinds of yoke-permeameters are now used successfully in 

 studying the magnetic properties of short rods at commercial 

 excitations. For very high excitations, at which the air is 

 nearly as permeable as the metal, the leakage becomes very 

 troublesome and depends upon matters which cannot be easily 

 controlled. I hoped, notwithstanding this fact, to be able to 

 calibrate the yoke we had (figure 3) by means of standard 

 pieces and thus make it available for studying short pieces of 

 iron at excitations of about iZ=1000, but Mr. John Coulson 

 and I worked for a long time on this problem without finding 

 any satisfactory solution. We found it possible, however, to 

 determine the length of a half-inch rod of iron which, mounted 

 between the jaws of this particular yoke, would cause the mag- 

 netic field in the air just without the iron near the middle of the 

 specimen to be practically straight and parallel to the rod for 

 a considerable distance. A piece about 15 centimeters long, 

 accurately fitted for about 3*5 centimeters at each end into 

 the taper holes in the jaws and leaving 80 millimeters of the 

 rod free, satisfied the conditions, and this length might be 

 slightly varied, but a much shorter rod violated the conditions 

 in one direction (the determination of fi being too large) and 



* Peirce, Proc. American Academy Arts and Sciences, xliv, 1908. 



f H. E. J. G. DuBois, Phil. Mag. 1890, The Magnetic Circuit in Theory 

 and Practice ; Shuddemagen,- Proceedings American Acad, of Arts and. 

 Sciences, xliii, 1907 ; J. Hopkinson, Philosophical Transactions, clxxvi, 1885; 

 Drysdale, Electrician, xxviii, 1901 ; Thornton, Electrician, xlix, 1902 ; 

 Ewing, Electrician, xxxvii, 1896; Schmidt, Ann. der Physik, liv, 1895 ; 

 Chattock, Electrician, xxxvii, 1896 ; Lamb and Walker, Electrician, xlvii, 

 1901 ; Baily, Electrician, xlviii, 1901 ; Kopsel and Kobinson, Electrische 

 Zeitschrift, xv, 1894; Kath, Electrische Zeitschrift, xix, 1898 ; Kennelly, 

 Electrische Zeitschrift, xiv, 1893; Blondel, Comptes Eendus, cxxvii, 1898 ; 

 Stoletow, Ann. der Physik, cxlvi, 1872; Eowland. Phil. Mag., xlvi, 1878 ; 

 Ewing, Magnetic Induction, 1892 ; Behn-Eschenberg, Electrische Zeit- 

 schrift, xiv, 1893 ; Kapp, Electrical Engineer, xxiii, 1894. 



