Barrett — On the Magnetic Moment of Steel, etc. 471 



position, with the exception that one has 12-3 °/ of manganese, 

 and the other 13*75 of manganese ; yet the magnetic moment of 

 No. 552, with l"4°/ more manganese, is double that of 547. It 

 will further be noticed that a nearly similar composition of the 

 alloy, in the form of ingot and of forged bar, have different mag- 

 netic characters : this, as already remarked, is no doubt due to the 

 mechanical treatment the test bars have received. Looking over 

 Table YI. it will be seen that the specimens 6A and 291 have very 

 low magnetic moments, corresponding to 9*4 and 20 per cent, of 

 manganese respectively. It is possible that we have here a chemi- 

 cal union of the manganese and iron ; in fact these alloys would 

 correspond to a compound having the formulae MnFe 8 and MnFe 4 . 

 Further information on this point can be obtained by a careful 

 determination of the specific heat of the different specimens. Un- 

 fortunately the different treatment the bars received after forging 

 confuses the results. By Mr. Hadfield's permission I cut off from 

 each of the bars similar cylindrical portions. Each of these smaller 

 rods were raised to a white heat, and then dropped in water. The 

 result of this, in manganese steel, is to soften the alloy : hence by 

 diminishing the coercive force to lessen the magnetic retentivity. 

 Each of these portions was re-magnetized in a similar and power- 

 ful field, and the magnetic moment per gramme of each again 

 determined. The moment found in each case was extremely 

 small : nevertheless No. 552 still retained its pre-eminence, having 

 far higher moment than any other specimen, and many of the 

 irregularities in the curve, showing the relationship of the mag- 

 netic moment to the composition, were smoothed away; the magnetic 

 moment rising up to 13'7 per cent, of manganese, and then failing from 

 this point to the 21*7 per cent, specimen. The following experiment 

 will illustrate the magnetic change produced by the different treat- 

 ment of the same specimen : — Two similar portions, each 8 centims 

 long, were cut off the bar No. 552. One portion was heated to 

 whiteness, and then dropped in cold water ; the other was left 

 untouched. Both were then magnetised in the same field. The 

 untouched portion had a moment per gramme of 0*1398 (practi- 

 cally the same as the test bar from which it was cut, as shown in 

 Table VI., viz. 0*1408) ; the other portion, annealed by sudden 

 cooling, had a moment per gramme of 0*0456, or one-third less. 

 The specific gravity of the two portions were not quite the same : 



SCIEN. PROC. li.D.S. VOL. VI., PART VIII. 2 N 



