PEIRCE. — MAGNETS OF IIARDENKD CAST IRON. 707 



Most of the steel made specially for magnets, which I have been able 

 to get, has come in the form of long bars of rectangular cross-section, 

 about 2 cm. wide and 1 cm. thick ; of such steel 1 have had pieces, 

 18 cm. long, hardened in some of the various ways used by professional 

 magnet-makers, to compare with hardened castings of the same dimen- 

 sions. One of tlie hardened but unmagnetized castings was placed in a 

 solenoid, exposed to the action of a demagnetizing alternating current, 

 and then put rapidly two or three times through a hysteresis cycle, using 

 a maximum field of 145.2 gausses. After this preliminary treatment — 

 which I used in the case of every piece that I examined — the hysteresis dia- 

 gram remained unchanged however many times the iron went through the 

 cycle. I first obtained twenty-two points on half the diagram, laid those 

 down carefully upon a piece of coordinate paper on such a scale that 

 the diagram was 41 cm. long, and found that it was possible to draw a 

 smooth curve which would not lie away from any one of the points by 

 80 much as a quarter of a millimetre and would apparently pass through 

 almost every point ; then I completed the cycle and found that the final 

 reading of the mirror magnetometer did not differ by more than a sixth of 

 one per cent, if by so much, from the initial reading. In the course of its 

 second journey around the cycle, the iron made subsidiary loops on op- 

 posite sides of the origin and then returned again sensibly to its original 

 condition. No diagram obtained from a long , slender wire could iiave 

 been smoother than this one which belonged to this short bar. All the 

 hysteresis curves in this paper are reduced from large drawings ; gen- 

 erally about twenty-four points (though sometimes more) were found for 

 each half diagram, and the curve was drawn through practically all of 

 these. 



Fig. 1 shows an instance where the points were apparently not so well 

 determined as in other cases. This figure represents hysteresis curves of 

 (A) a piece of hardened cast-iron (18 cm. X 2 cm. X 1 cm.), and (B) of two 

 pieces of Seebohm and Dickstahl magnet steel (of the same dimensions), 

 which makes the strongest saturated magnets of any of the special steels 

 which I have used. The maximum field in the case of the A curve and the 

 larger B curve was nearly 166 gausses, and under this field the cast iron 

 and the special steel acquired moments of about 8900 and 15400 units 

 respectively : when the field was removed the residual moments ivere 

 about 3120 and 3550, The piece of steel just mentioned was hardened 

 in a water bath ; the smaller B diagram was obtained at another time 

 with a similar piece of the same steel chilled in one of the baths used 

 by Mr. Thompson. Although the maximum fields were unfortunately 



