702 PROCEEDINGS OF THE AMERICAN ACADEMY. 



of steel foi- permanent magnets of complex forms, without spoiling it, 

 demands a kind of skill which most toolmakers, even in the largest 

 establishments, have not acquired, and it is generally difficult to get 

 satisfactory specimens of any very special shape of curved-steel magnets 

 unless one has access to such facilities as a few of the manufacturers of 

 electrical measuring instruments have provided for themselves. It is true 

 that the hardening of iron castings for magnetic purposes also requires 

 such skill as few persons possess, if the very best results are to be ob- 

 tained, especially when the pieces to be treated weigh more than a pound 

 or two; but a little practice will enable any good workman, who has a 

 gas forge with blast powerful enough to raise the temperature of the iron 

 uniformly nearly to the melting-point, to make good gray-iron castings of 

 moderate size, hard enough for strong magnets, which will leave little to 

 be desired so far as permanence is concerned. It has been my good 

 fortune to have the help of Mr. G. W. Thompson, the mechanician of the 

 Jefferson Laboratory who has had long experience in treating cast iron, 

 and who has made for me, by a process of his own, massive magnets 

 with extremely low temperature coefficients. It is to be noticed that 

 some of the secret methods of hardening cast iron, used by makers of 

 small parts of machinery, do not fit the castings for making good magnets, 

 and that case hardening, which affects the surface only, is useless. The 

 character of the cold bath into which, while it is kept in violent agita- 

 tion, the strongly heated castings to be hardened are plunged, seems to 

 have considerable influence upon the result. 



At the very high temperature, just under the melting-point, to which 

 the cast iron must be raised before it is suddenly cliilled, the metal loses 

 much of its tenacity, and slender pieces must be handled carefully lest 

 they break like chalk. The chilled casting should be hard enough to 

 scratch window glass, if not so readily as hardened tool steel will do 

 it. It is vain to attempt to make any such gray-iron castings as I have 

 used, magnetically hard by chilling them after they have been heated to 

 the comparatively low temperatures that one would use in making steel 

 glass-hard. Everyone who has attempted to harden a thick mass of tool 

 steel uniformly, knows how difficult the task is : it is easy enough to get 

 the outer layers glass-hard, while the interior is much softer ; or, some- 

 times (by overheating the steel), to get the inside hard while the outside 

 is blistered and cracked. If a casting is iieated to a very bright red and 

 then plunged into the bath, the outside may become hard to the file, while 

 the interior, as magnetic tests clearly show, remains soft ; in this case, 

 however, the material will stand a higher temperature without injury, and 



