e@ from an aluminum block supported by end pieces which were notched peveral discrete positions to accommodate a standard planchet holder, ‘jhe detectors afforded a fair range of geometry variation, All pieces ‘were made of aluminum and machined to close tolerances so that geometri- eet orientation of the counting samples could be readily reproduced and yeentical operating characteristics could be assured from detector to gatector. The detectors were the continuous gas-flow type, using a gas igonsieting of 90 per cent argon and 10 per cent carbon dioxide at a flow fete of 10 to 15 cu cm per min. Doubly aluminized Mylar (0.95 mg/sq cm) sgge used for the window which was nominally 1 in. in diameter and colli- fpated by 1/16 in. aluminum and 1/16 in. lead. A pre-amplifier of wide ‘@yuemic range and gain of approximately 1000 was built on a chassis aseomdled directly with the detector in such a way that the entire assem‘ply fitted into a commercial lead castle. The detector assembly had ypaout a 200-v plateau with less than 1 per cent slope per 100 v, and operated at about 1900 v. Decade scalers NRDL Model 2 with self-contained register, high woltage, anc automatic clock were used. Time accuracy was about 1/2 sec. A scale factor of 1000 was employed, The dead time of the beta counting system was 5 «sec which gave a coincidence loss of 1 per cent at a counting rate of 100,000 events per minute. 2.303 Counters The gamma detectors consisted of an RCA type 5819 photomultiplier with a light-pipe adapter and a commercially mnounted* cylindrical crystal of sodium iodide, 1-1/2 in. in diameter by 1/2 in. thick. A removable aluminum absorber (1600 mg/sq cm) was used to shield out beta rays. The shelf geometry was the same as that described for the beta counters so that the same planchet holders could be used in both systems, The detec- tor assembly was attached directly to an amplifier chassis, and tue whole assembly was mounted within a commercial lead castle. The amplifier was of a wide dynamic range design with a nominal gain of 1000. The scaler, (Navy Model AN-4DR19) supplied high voltage and power needed for the detector unit. Two types of timers were provided with this scaler; an electromechanical automatic-termineting timer with an accuracy of 1/2 sec, and an electronic timer, The electronic timer* accepted pulses from a crystal oscillator within the scaler on al sec schedule, scaled these down by means of glow-transfer type d.cades with provisions for presetting any scale factor from 1 to 9999 and feeding the carry-over pulse into the gating jack on the scaler which stopped the scaler with an accuracy of about 100 usec. With this timing device short time intervals of counting could be used while maintaining a counting error within allowable counting statistics (i.e., minimum timing error). * ** Supplied by Harshaw Chemical Co., Cambridge, Mass. Model GS-10C, menufactured by Atomic Instrument Co., Cleveland, Ohio. 21