104 BELL SYSTEM TECHNICAL JOURNAL 



It is obvious that only a small part of any sample can be given the 

 most favorable treatment, since the interior portions of the sample 

 cool at rates which are dependent upon the geometrical configuration 

 and thermal properties of the material and are only indirectly under 

 the control of the experimenter. For these reasons each shape and 

 size of sample will have its own best heat-treatment and it is obviously 

 difficult to establish the correct heat-treatment for a small element of 

 volume, characteristic of permalloy as a material. By the use of thin 

 tape, however, we secure fairly uniform treatment of the whole volume 

 so long as the cooling is not too rapid, and fortunately the best cooling 

 rate is not much different from the normal cooling rate of the tape 

 in the open air. It has been found that temperature changes below 

 300°C. have very little effect upon the resultant properties of permalloy, 

 but the rate of cooling from just above the magnetic transformation 

 temperature down to about 300°C. is a controlling factor. By a long 

 series of experiments a heat-treatment has been established which is 

 especially well adapted to the permalloy test rings already described. 

 They are first heated at about 900°C. for an hour and allowed to 

 cool slowly, being protected from oxidation throughout these processes. 

 They are then reheated to 600°C, quickly removed from the furnace 

 and laid upon a copper plate which is at room temperature. 



Not only does each size and shape of sample require its own special 

 heat-treatment, but samples differing only in composition also differ 

 in their most suitable heat-treatments. In our investigation of the 

 nickel-iron series we have not, however, attempted to determine the 

 best heat-treatment for ring samples of each of the many alloys 

 studied. By careful exploration we located the region about 80 

 per cent nickel, 20 per cent iron as the one promising the highest 

 initial permeability and established the best heat-treatment for this 

 composition. Keeping this treatment unchanged we then relocated 

 the best composition, finding it to be at about 78.5 per cent nickel, 

 21.5 per cent iron. There is a maximum temperature in the equilib- 

 rium diagram for this binary at about 70 per cent nickel, 3 and it was 

 natural to suspect that the maximum in initial permeability which 

 we had found at 78.5 nickel might be displaced to 70 nickel by proper 

 treatment. The 70 per cent nickel alloy was accordingly subjected 

 to a great variety of heat-treatments, but no method was found 

 capable of producing in it an initial permeability as high as that 

 readily obtainable in the 78.5 per cent nickel alloy. 



Fig. 1 shows the general way in which initial permeability has been 

 found to vary throughout the nickel-iron series when the heat-treat- 



3 Bureau of Standards Circular No. 58, April 4, 1916. 



