PHYSICS, CHEMISTRY AND ENGINEERING 153 



rare cases, and have been regarded as exceptional. It has 

 been the privilege of the Engineering Experiment Station of 

 the University of Illinois recently to produce iron and iron 

 alloys with permeabilities all the way from 20000 to 50000 

 and with hysteresis losses of one-half to nearly one-tenth that 

 of the best commercial silicon steel. It is unsafe at the pres- 

 ent time to state exactly what the values are, as the methods 

 of testing that have hitherto been regarded as standard have 

 proved to be inadequate for this high permeability iron. The 

 Burrows compensated double bar and yoke method gives too 

 high a maximum permeability, too high retentivity and co- 

 ercive force, too low hysteresis loss for low densities and too 

 high for high densities. The ring method, while more satisfac- 

 tory in this respect, is open to the objection that the flux dis- 

 tribution is not uniform, and besides is awkward to work 

 with. The subject is at present being investigated both here 

 at the University of Illinois and at the United States Bureau 

 of Standards, and a more perfect method will undoubtedly 

 be developed. 



The method employed for the production of this high per- 

 meability material consists in melting electrolytically refined 

 iron in a vacuum furnace, the absolute pressure being 

 0.5 mm. of mercury. The iron is allowed to cool in the 

 furnace, and when removed has an appearance like that of 

 nickel. The ingots thus produced are forged into rods 

 and machined into proper test pieces. In this state, how- 

 ever, the magnetic properties are very poor, chiefly on ac- 

 count of the molecular strain caused by the mechanical treat- 

 ment, and it is necessary to anneal the rods before the un- 

 usual properties are obtainable. This is done by heating the 

 rods to 900° or 1100°C in vacuo followed by cooling at the 

 rate of 30 °C per hour down to room temperature. 



Thus far we have investigated pure iron, iron-boron alloys, 

 iron-carbon alloys, and iron-silicon alloys, besides the iron- 

 cobalt alloy Fe 2 Co. The iron-aluminum series is being in- 

 vestigated at the present time. It gives the average 

 for a large number of samples of Vacuum Iron, the best 

 of which is below what we can produce with certainty today. 

 The curve just below that for the "Vacuum Iron" proper 

 represents Swedish charcoal iron remelted in vacuo, showing 



