the pulse-repetition rate was so great that the recorded marks overlapped and could not be resolved. At that time, Stylus No. 2 would be used, with each mark representing 11 pulses from the detector head. The chart drive that supplied the time base was’ calibrated with a Watchmaster before each event. By means of the Watchmaster, the chart drive could be set to have a maximuin error of 1 minute in 24 hours, or 40.069 percent. This was not the optimum recording system for use with this detector but rather a compromise forced by a lack of funds and time. 2.2.3 Initial Instrument System, Gustave | Detector. For the high-range, fast- resolution detector, the basic circuit of Figure 2.1 was used with a scintillation detector as the sensing element. The latter consisted of an RCA 929 phototube and a National Radiac Scintillon Branch plastic phosphor mounted in an electron-equilibrium thickness of bakelite to provide an air-eyuivalent response (Reference 8). The purpose of the electron-equilirium layer was to present a source of electrons that might be scattered into the crystal to replace those electrons produced by radiation absorbed near the crystal surfaces and lost without being detected. These detectors were constructed to cover three ranges: 10° to 10° r/hr, 107 to 10’ r/hr, and 104 to 108 r/nr. The overall detector response is given approximately by: f=kr Where: (2.3) f= the pulse repetition rate r =the gamima-exposure rate in r/hr k = a parameter chosen to meet specific design objectives The maximum pulse-repetition rate of these instruments was 1,000 pulses/sec, the maximum rate that could Le resolved by the recorder (a Cook Research Laboratory MR-33 eight-channel magnetic-tape recorder). Typical calibrations for these detectors are shown in Figure 2.3. Figure 2.4 shows the energy dependence of the Scintillonphosphor Gustave I detector, relative to Co” gamma radiation at a rate of 100 r/hr. To reduce the errors due to flutter and wow, a 1,900-cycle-time base was recorded on the tape simultaneously with the gamma-exposure-rate datz. An American Time Products transistorized-frequency »tandard with an accuracy of +U.02 percent was used to provide the time base. 2.2.4 Photomultiplier Feedback Circuit, Initial Instrument System. This system was essentially the same as that used during Operation Castle (Reference 2). The detecting element, a Scintillon phosphor 2.75 inches in diameter by 0.5 inch in height mounted in a bakelite block for electron equilibrium, was placed inside a blast-resistant housing at the top of a light pipe. The output of the crystal after passing through the light pipe was detected by an RCA 6199 photomultiplier tube. The photomultiplier tube was used in a 100-percent-feedback circuit which held the photo-multiplier-tube-anode current nearly constant, regardless of the incident light flux, by reducing the dynode voltage (Figure 2.5). The gain of a photomultiplier tube with constant anode current was approximately proportional to the antilog of the dynode voltage. In this manner, a useful dynamic range of about a factor of 10' was realized. 2.2.5 Calibration. Three radiation sources (a 250-kv X-ray generator, a 2.5-Mev Van de Graaff generator, and a 200--curie Co® source) were used in the calibration of 22