III-25 



Figure III- 9 shows the experimental results for 6 at resonance as de- 

 termined by several authors as well as the corresponding theoretical values ob- 

 tained from Model III. The contributions to 6 due to viscous damping, radiation 

 damping, and heat conduction damping are shown separately. Allowing for differ- 

 ent experimental techniques and limits of error, the results generally confirm 

 the theory. Exner and Hampe seem to have done the best and most extensive ex- 

 perimental work, and their results are probably the most accurate. Unfortunately, 

 this work was done at very high frequencies. Lauer's work and Exner' s work at 

 low and intermediate frequencies, respectively, though done under less carefully 

 controlled conditions, also agree with the theory in these ranges. 



The only serious disagreement with the theory is in the work done by 

 Meyer and Tamm, Carstensen and Foldy, and Fox, Curley and Larsen. In all 

 cases the authors found values for the damping constant which were larger than 

 those predicted by the theory. Fox, Curley and Larsen' s single observation at 

 65, 200 cps is about five times that predicted by theory. Carstensen and Foldy, 

 and Meyer and Tamm report values of 6 for low and intermediate frequencies 

 which are roughly two to three times the theoretical values . 



These differences should not be considered as contradictions of the 

 theory. The higher values were generally obtained indirectly in the course of at- 

 tenuation studies and were needed to account properly for the observed attenuation. 

 The theory is otherwise well supported; therefore, it seems that some significant 

 factors were not considered when working backwards from the attenuation experi- 

 ments . 



In order to see how the variation in 6 ^ compares with that of 

 we have plotted in Figure III- 10 the ratio 



(^■4 



A - 



(?-.-)■/• 



as a function of f for a = 1.0, 0.10, 0.01 and 0.001 cm. One would like to be 

 able to conclude that in the neighborhood of f = f the value of 



(f ■■) 



varies much more rapidly than 6 , i.e., that 6 may be considered relatively 

 constant in the neighborhood of the resonant frequency. Outside this neighborhood, 

 6 is not very important. In later calculations of cross sections, we would then 

 be able to view 6 as being constant over the full range of frequencies. Unfortunately, 

 the horizontal nature of the curves in the neighborhood of the resonant frequency pre- 

 cludes this possibility. 



;artbur Hl.littleJjtf. 



S-7001-0307 



