980 



THE BELL SYSTEM TECHNICAL JOURNAL, OCTOBER 1951 



A£ ^ Es - £o ^ 2.381 - 2.312 

 Kq Eq 2.312 



= 0.03 = 3% 



(11) 



which is much smaller than that given by low frequency measurements. 

 To check our results, and be sure that the crystals were free from imper- 

 fections and strains, two other crystals were prepared and carefully an- 

 nealed at 1100°C. The values found for the changes in elastic constants were 

 considerably less for these crystals. The Q's of the crystal were also higher. 

 Table II shows the measured values and the equivalent AE/E values. The 

 table shows also the measurements for the demagnetized crystal of two 

 Japanese workers^^-" and the equivalent AE/E assuming that the saturated 

 elastic constants are the same as those found for the other three crystals. 

 Since these vary by only ±0.5 per cent among themselves, this appears to 

 be a good approximation. 



Table II 

 Elastic Constants (in 10" dynes/cm^) and A^-Effect in Single Crystals of Nickel 



The lower values of AE/E for the second and third crystals are probably 

 due to larger domain sizes, caused by the longer anneal. 



Damping and A£-Efeect in Polycrystalline Rods 



To test the theory of micro-eddy-current shielding (see Introduction), 

 the velocity and attenuation of elastic vibrations in well-annealed poly- 

 crystalline nickel rods were measured over the frequency range of 5 kilo- 

 cycles to 150 kilocycles. In the method of measurement,^^ shown by Fig. 

 9, two matched piezoelectric crystals of resonance frequency corresponding 

 to integral half wavelengths along the rod, are attached to the ends of the 

 rod. Phase-amplitude balance was obtained by critical adjustment of fre- 

 quency and output of the calibrated attenuator. The corresponding level 

 was then compared with that obtained when the two crystals were cemented 

 directly together. With little error, the velocity of propagation is given by 



V — 



2//o 



n= 1, 2, 3 



(12) 



