PEIRCE. — MAGNETS OF HARDENED CAST IRON. 703 



if the mass be reheated and when it is just below the melting point be 

 suddenly chilled, the whole interior becomes hard. 



It is a good deal easier to harden a lot of straight, round pieces of good 

 gray cast iron, say 20 centimetres long and 1 centimetre in diameter, so that 

 they shall all be nearly alike magnetically, than it is to do the same with an 

 equal immber of pieces of drill rod. Six pieces of Crescent Drill Rod each 

 IG centimetres long and 8 millimetres in diameter, cut from the same 

 specimen, were made glass-hard for me by a skilled worker in steel; 

 these were placed successively in a properly oriented solenoid and ex- 

 posed, first to the action of an alternating current of intensity gradually 

 decreasing from an initially high value to a very low one, then to a steady 

 field of 147 gausses applied first in one direction and afterwards in the 

 other. As a consequence of the preliminary treatment* with alternating 

 currents, the magnitudes of the moments acquired by the pieces under 

 the action of the steady field were quite independent of the direction of 

 the latter. These moments were approximately 2280, 2395, 2495, 2326, 

 2325, and 2360, but when the field was removed the residual moments 

 were 1058, 1074, 1136, 1066, 1050 and 1097 respectively. The same 

 pieces were then placed together in a solenoid made of many turns of 

 large wire and the ends of the whole bundle were connected by a massive 

 yoke; when a current of about 45 amperes was sent through the wire 

 the pieces became charged practically to saturation. When they were 

 removed from the solenoid the average moment of the six was about 1240, 

 the highest 1290, and the lowest 1170. Such uniformity as is indicated 

 by these numbers is, I believe, as great as one can exjiect to get unless 

 one has an elaborate plant ; no such agreement can be hoped for from 

 pieces of different rods of the same brand. Some kinds of special magnet 

 steel give rather better results. 



Although it is obvious that there is no advantage in using cast iron for 

 straight magnets, I have had a number of such magnets made, of each of 

 three shapes, for purposes of comparison with steel bar magnets of 

 the same dimensions. These were all rather short, because we had no 

 means of treating satisfactorily very long, slender pieces, which are apt to 

 warp if not properly supported. It is, of course, impossible to calculate 

 the demagnetizing effects of the free ends of such pieces as I have used, 

 but it has seemed to me legitimate to draw some infereiic -s from the 

 hysteresis curves and from the temperature coefficients of rods of difTereut 

 materials if they are geometrically alike. 



* J. A. Ewing, JIagnetic Induction in Iron and other Metals, 1892, p. 328. 



