viously described by using the data in Section C.2 and assuming that the total incicent rv diatic was contained within the Same angle at each detector. The integrated response for various a sumed angles was then normalized to a response of unity for a completely surrounded detecto Since the ASEL and sec-GITR response is equivalent to a figure of revolution whose ams is parallel to the ocean surface, only variation in the horizontal angle subtended by the source causes a difference in response. Conversely, since the std-GITR response is analogous to a figure of revolution perpendicular to the ocean surface, only variation in the vertical angle subtended causes a change in response. For greater convenience, the assumed angles have been converted to the distance at which base Surges of various assumed dimensions would sut tend a sumilar angle at the detector. These calculations are presented as alternate scales be neath the scale of assumed angles. The importance of the gamma-intensity time records requires a more detailed description of the limits of accuracy tmposed by the detecting system and readout procedure. The stated limits of accuracy are +30 percent except at peak dose rate where the accuracy becomes inde terminate. Because the std-GITR is a recycling dosimeter, it tends to average out rapid changes in dose rate. This defect is an essential characteristic of the detecting device and cannot be rectified; thus, all peak dose rates reported are probably lower than the actual peal rates by an indeterminate amount. The ionization chambers can respond accurately to dose rates as high as 500,000 r/hr; however, their associated electrometer circuits begin to intro. duce appreciable recycling errors at rates above 87,500 r,hr for the 12-hour GITR. In most instances the GITR record can be used to rates of about 100,000 r/hr with errors ranging he- tween 2 and 7 percent. At higher rates the spacing of individual radiation pulses on the magnetic tape becomes too close for resolution, and the record is said to be saturated. The information on these tapes is in the form of two channels of radiation pulses (Appendi: F) and one channel of timing pulses. Each radiation pulse on the high-range channel represe: a dose increment of 0.243 r; for the low-range channel the value is one-thousandth of the high range increment. The time channel consists of a square-wave pulse created by a mechanical timing motor every 3.75 seconds. Dose rate was obtained from GITR tapes utilizing the GIT in one of two possible ways: (1) the fixed-interval-counting method, and (2) the time-between pulses method. Since the GITR tape transport operates at a nonstandard speed and since the GITOUT was constructed of standard commercial elements, the slowest transport speed for readout is 3,75 in/sec or 15 times the speed at which the 12-hour GITR records. Therefore, when considering GITOUT procedures, a careful distinction must always be made between plz back time and real time. The fixed-interval-counting method was used most frequently. The length of the counting interval is determined by the timing channel, the shortest interval being 3.75 seconds of real time. During each counting interval, all radiation pulses are summed by a digital counter. At the end of each interval, the cumulated total is printed out, and the summation operation i simultaneously switched to a second digital counter so that the tapes can be monitored contim ously. Average dose rate in r/hr over a counting interval of 3.75 seconds is obtained by mul tiplying the sum of the radiation pulses accumulated by 233 for the high-range channel and by one-thousandth of this value for the low-range channel. The GITOUT can reproduce dose rat to an accuracy of +1 radiation pulse per counting interval; thus, the accuracy of the fixed- interval counting method for dose rates represented by less than 10 radiation pulses per inte: val is no better than +10 percent. The time-between-pulse method of readout is highly accurate at any dose rate but has disa vantages in that it is more time consuming than the fixed-interval counting method and frequently requires electronic tape stretching. Dose rate is determined directly by measuring the time required to accumulate the preset dose increments, i.e., 0.243 ror 0.243 mr. Tota running time, however, must now be determined by summing all time intervals instead of Simply multiplying the mmber of intervals by a constant as in the fixed-interval method. Fw thermore a minimum of 200 msec is required to complete any given print-out cycle. As the dose rate increases, the spacing between radiation pulses on the tape decreases. At the min 81