MAGNETIC ALLOYS OF IRON, NICKEL, AND COBALT 459 



Other interesting evidence is obtained from the study of the 

 hysteresis loops. For the perminvar alloys the constricted loops are 

 very marked and easily produced. In the permalloys containing 

 between 60 and 80 per cent nickel also there is a tendency to con- 

 striction in the slowly cooled alloy, not so marked but sufficiently^ 

 prominent to lead me to believe that the same general changes occur 

 in both groups of alloys, differing only in the nature of the segregates 

 and the ease with which segregation occurs. Now we know that homo- 

 geneous magnetic materials have a characteristic type of hysteresis 

 loop. For such materials there is no constriction in the middle, but 

 the widest part of the loop is generally at that point. We also can 



r 



r 



Fig. 19 — Hysteresis loops: a, perminvar (45 per cent Ni, 25 per cent Co, 30 per cent 

 Fe); h, bi-metallic rod. Loop traced with cathode ray oscillograph. 



construct hysteresis loops which have constrictions by making up 

 cores of several materials in a parallel or parallel-series arrangement.^ 

 This is illustrated by the hysteresis loop h in Fig. 19. This loop is 

 traced by a cathode ray oscillograph for a bi-metallic rod, 15 in. long 

 consisting of a core of .04 in. diameter unannealed piano wire and a 

 .006 in. wall permalloy tube, heat treated to give high permeability 

 and fitting closely to the wire. Curve a is a loop similarly traced for 

 a perminvar core. Though the magnetic circuit conditions for the 

 two cores are not the same, the marked similarity of the two loops 

 favors the view that the constricted loop of the perminvar results 

 from segregation. 



It is interesting to note in this connection that the examination by 

 X-ray crystal analysis methods of these alloys has not given evidence 



6 E. Gumlich, Arch. f. Elektrotechnik , Vol. 9, p. 153, 1920. 



