52 



W/Q , constant for a given type of explosive, (6) can be made 

 to give results correct to within a fraction of a percent, an 

 accuracy comparable to the degree of reproducibility of 

 experiments. However, the value of W which one must use in 

 (6) to get the best fit to a given range of experimental data 

 will of course differ a little from the correct value of the 

 energy of the motion, because of the sliglit effect of gas 

 pressure on the period. More refined comparisons of theory 

 and experiment than those given by 771111 s have been made 

 subsequently, and are reported elsewhere in this Volume. 



A rough idea of the error involved In neglecting gas 

 pressurR may be conveyed by quoting the results which one obtains 



by substituting into the theoretical calculations of period 



14 

 made by Shiffman and Friedman the empirical value 1.25 



IS 

 obtained by Arons for the adlabatic exponent in the equation 



pV ■= const. and the value obtaine "' by Arons f^r the energy of 



the first pulsation. This gives a period which is lower than 



that given by (6) by less than a percent, if W is interpreted 



as incloding the total energy of the gas relative to infinite 



adlabatic expansion; the ratio of T to a Is greater ohan 



max 



that given by (7) by an amount which varies from about 4^ at 

 sea level to about 9% at 500 feet depth. 



Fig. 2 shows a comparison of an observed radius- time 

 curve with the curve calculated In Appendix 1 by integration 

 of (5) with G = . The observed points were taken by Swing 

 and Crary from a motion picture taken by Sdgerton In 1941 of a 



14 



See the article by Friedman In this Volume. 



15. 



A. B. Arons, J. Acous. Soc. Am. 20, 277 (1948). 



