MOTION OF THE GAS SPHERE S49 



graphs, discussed in Chapter 6, make necessary carefully controlled 

 light conditions which are not readily achieved for small scale explosions 

 and are very difficult to obtain for large charges. As a result, not very 

 many experiments have been done except on a very small scale. Some 

 recent experiments (105), done in extremely clear water off Cuba, should 

 provide useful information about migrations of bubbles from charges up 

 to }/2 pound in weight, but at the time of writing these data had been 

 only partially anah^zed. Photographic records were also obtained of 

 the bubble from 300 pound TNT charges, but uncertainty as to the 

 camera position and orientation make analysis of migrations a very 

 dubious possibility. 



C. Measurements of migration by sound ranging. The difficulties in 

 obtaining photographic records of bubble motion have led to the use of 

 less satisfactory methods based on determining bubble position by sound 

 ranging methods. In such work, arrival times of the secondary bubble 

 pulses emitted at successive bubble minima are recorded; the only in- 

 formation so obtained is the position of the pressure source at these 

 times. The method is limited in accuracy by uncertainties as to gauge 

 positions and suffers from more fundamental errors. IMost sound rang- 

 ing methods depend on measurement of time differences in reception of 

 the secondary pulses at two or more different gauges (see section 10.3). 

 As contrasted with the shock wave, however, bubble pulses ideally have 

 no abrupt changes or discontinuities in outhne and frequently have 

 durations comparable with the time differences in feasible gauge ar- 

 rangements, which inevitably limits the accuracy obtainable. 



Frequently, recorded bubble pulses show small spikes or other minor 

 discontinuities superimposed on the general profile, and these local dis- 

 turbances have commonly been used as reference points. This pro- 

 cedure is not free from objection, however, as gauges differently oriented 

 often show rather different forms of bubble pressure-time curves, even 

 though corresponding reference points are quite obvious. These dif- 

 ferences impl}^ a lack of spherical sj^mmetry in the pressure wave, and 

 make dubious the assumption that the spikes have their effective origin 

 at the bubble center. Measurements on different pairs of spikes in two 

 gauge records confirm the likelihood of such difficulty by sometimes 

 giving quite different indications of the bubble positions. Examination 

 of such data makes reasonable the proposition that these spikes may 

 have their origin in the observed irregularities of the bubble surface 

 when near its minimum size. The computed source of sound may thus 

 differ in position from the actual center b}^ distances of the order of the 

 minimum bubble radius. Both these inherent difficulties and errors in 

 establishing the geometrj^ of the experiment limit the accuracy and 

 reproducibihty of sound ranging data, and these limitations should be 

 kept in mind when considering the results. A further hmitation, com- 



