528 - " - 



Braced longitudinal angle iron members B. To the uprights are CoUed nine pieces of i inch thick 

 plywood cut to a calculated curvature and bevelled. Narrow strips of 16 gauge duralumin plating are 

 screwed on to the plywood sections, the strips having been cut and fitted by trial. 



Ideally the reflector should form part of an ellipsoid of revolution, with the lamp and 

 camera at the two foci. This requires that the position of both lamp and camera should always be 

 fixed in relation to the reflector, and that the lamp should operate underwater. owing to the shape 

 of the experimental tank these conditions are not easily satisfied over a range of operating depths. 

 AS a compromise, the present reflector has been rede with sections calculated for a paraboloid of 

 revolution with axis horizontal the lamp being operated just above the water surface. Owing to the 

 finite size of the car headlamp reflector in which the Arditron is housed, this optical system 

 produces a satisfactory bright background for a range rf operating depths down to 5 feet, as may be 

 seen from typical photographs in Figure 8. It will o? seen from these photographs that the illumination 

 tends to be patchy; this is due to the anticlastlc curvature cf the strips cf duralumin which form 

 the reflecting surface. The reflector will withstand repeated explosions of a l z. charge 3 or tt 

 feet away. 



Timing Controls . 



The firing of the two groups of sparks, the opening and closing of the two shutters in the 

 underwater camera, and the firing of the charge are performed in their correct sequence by a rotary switch 

 carrying 6 brass sectors mounted in adjustable rotating ebonite discs. The period of the bubble motion 

 being determined by previous oscillographic records, or by calculation, the sectors of the rotary switch 

 are each preset to operate at the correct time, allowance being made for the known time lags in the 

 spark relay and the shutter mechanisms. 



The rotating spark gap is driven by an induction motor working off the 50-cycle mains. as the 

 motor is practically on "no loan" its speed is always very close to 3,000 r.p.m. (this has been verified 

 by a tackometer). The speed of the drums in the underwater camera has to be adjusted to give proper 

 spacing to the photographs, and for this purpose a remote reading electrical revolution counter has been 

 devized whose operation is shown diagrammat ically in Figures 7a and 7b. 



In Figure 7a, 8 is the spindle whose rotational speed is required. The Brass disc A carries 

 five small equally spaced soft ir.-,n inserts C which pass in turn between the poles cf a permanent magnet 

 D causing a series of voltage pulses t',- be induced in the solenoid E. This pulse consists t'f a positive 

 and an identical negative portion and lasts for about one twentieth of the interval between successive 

 pulses. 



Figure 7b shows the electrical circuit used to indicate the frequency_of arrival cf the voltage 

 pulses produced by the solencid on the camera. Valves vi and V2 form a straight forward "flip fKp" 

 circuit in which there is normally no current flcwing in the ancde circuit of V2. The arrival of a 

 negative pulse cf sufficient magnitude from the camera solenoid causes a square pulse of current t? fl^w 

 in the ancde circuit of V2, the duration and magnitude of this pulse being governed solely by the circuit 

 constants and not by the shape or magnitude of the triggering pulse. At normal drum s^eeds the needle 

 of the milliaimeter in the anode circuit of V2 cannot follow the Individual (ulses it receives But gives 

 a deflection equal to the time average cf the current jassing through it. The deflection is therefore 

 directly proportional to the frequency of the triggering pulses. The circuit shown has a response of 

 1 milliamp per 1,000 r.p.m. cf the camera drum, linear from 500 r.p.m. up to at least 3.500 r.p.m. 



V3 is a thyratron and the associated circuit produces a very short negative pulse at the 

 frequency of the A.C. mains, thus enabling the overall current-frequency sensitivity of the "flip-flop" 

 circuit to be standardized just before use. 



Typical Photographs . 



in Figures 8 and 9 are shown sor? typical photographs obtained from the explosion of a single 

 1 oz. charge of Polar ammon gelignite at a depth of 3 feet. The interval between pictures is 1 millisecond. 

 The group shown in Figure 8 was recorded On one drum of film and was timed to Include the first bubble 

 minimum; the pictures in Figure 9 include the second minimum. The general appearance of the bubble 



