Intelligence and Miscellaneous Articles, 



75 



commercial kinds do not contain more than 1 per cent, or carbon, I 

 had prepared by M . Dalifol a series of steels which have been sub- 

 mitted to repeated cementations, then cast, in which the proportion 

 of carbon progressively increases, and the last members of the series 

 will not bear the forge. Chilled at a white heat they are more and 

 more hard, not attacked by acids except aqua regia, less and less 

 conductors of magnetism, and, what is more, less and less magnetic. 

 The last members are scarcely attracted by an electromagnet, scarcely 

 magnetized by a powerful coil either during or after the passage of 

 the current (Comjotes Rendus, vol. lxxvii. p. 91). 



I have explained these facts by assuming that those steels which 

 are bad conductors of magnetism in the direction of their length are 

 the same also in the direction of their depth — that their magnetiza- 

 tion, considerable at the surface, decreases thence to the centre, and 

 that for the most cemented the magnetized layer does not exceed 

 the thickness of a varnish. I have, besides, established that the 

 law of the decrease of magnetism is given by the preceding formula. 

 These ideas having been recently contested on the ground of some 

 experiments neither very precise nor correctly interpreted, I pro- 

 ceed to defend them and give them precision by more thorough in- 

 vestigation. 



I take as an example a bar of steel prepared by M. Dalifol, very 

 homogeneous, well tempered, very hard, and having very little con- 

 ductivity. Its dimensions were — length 280 milhms., breadth 50 

 millims., and thickness 10*6 millims. After magnetizing it to satu- 

 ration, and keeping it at rest for some days to bring it to its defini- 

 tive equilibrium, I dissolved it gradually in cold aqua regia, taking 

 it out from time to time to measure its thickness e and its total 

 magnetism M, continuing the operation without interruption, as 

 rapidly as possible, during one whole day. Here are the results : — 



e. 



M. 



Difference for 



Fraction removed. 



0-1 millim. 



Thickness. 



Magnetism. 



millim. 

 10-6 

 105 

 10-4 

 10-3 

 102 

 10 



9-9 



9-4 



8-9 



86 



8-4 



millim. 

 25 00 

 22-90 

 21-26 

 2005 

 19-00 

 17 25 

 16-62 

 1408 

 12-20 

 11-45 

 1100 



210 

 1-64 

 1-21 

 105 

 0-85 

 0-63 

 0-43 

 0-35 

 0-25 

 020 



0009 

 0018 

 0-028 

 0-037 

 0-056 

 0-066 

 0-113 

 0-160 

 0-188 

 0*207 



0-084 



0-142 



0-198 



0-240 



0-310 



0-335 



0-436 



0-512 



0-550- 



0-564 



The values of M sensibly satisfy the equation M= Ae + Be 3 . The 

 third column contains the losses of magnetism for successive dimi- 

 nutions of thickness equal to 0*1 millim . These losses, at first equal 



