436 THE BELL SYSTEM TECHNICAL JOURNAL, MARCH 1957 



08 



0.6 



cr 



o 

 z 

 < 



0.4 



0.2 



Fig. 4 — The functions Ri and R2 versus the ratio of disc diameter do to cavity ■ 

 diameter D. " 



to be entirely within the circularly polarized region. However, this re- ^ 

 quirement appears to assume that there is some interaction, such as 

 spinwave coupling, between adjacent regions in the disc. It is possible 

 that such interaction exists in single crystals and leads to multiple 

 resonance effects. However in polycrystalline material it is safe to assume 

 that each crystallite reacts independently with the RF magnetic field, 

 enabling us to sum these effects by simple integration. 



This integration has been performed in the derivation of (19) and (20) 

 under the assumption that the disc is thin enough to introduce only a 

 first order perturbation into the cavity field. It will be shown presently 

 that these assumptions are consistent with experimental results. 



The measured effects Acu± and A(l/Q)± depend on the volume of the 

 perturbing body, hence one would expect that these effects are at least 

 an order of magnitude larger for thin discs than for small spheres. This 

 permits very accurate measurement of ferrite parameters in the region 

 below saturation and below and above resonance. In particular, the loss 

 parameters Xm" ± k" of modern low-loss ferrites can be determined ac- 

 curately in these regions. However, in the resonance region, the meas- 

 ured effects become so large that measurement becomes difficult. Hence, 



