casts were made. A slow ship, therefore, cannot afford to undertake too many stations and must rely heavily on other means for covering large areas of the sea in a reasonable time. A gap in the measurements made by the towed instruments appears on the chart at about midnight of the first day; activity at that time became so concentrated that all of the towed instruments deflected off scale. Fortunately, during this period, the pot instrument continued to indicate gamma intensity; nevertheless, its readings had to be cor- rected to eliminate a continuous drift error. Figure 3.2 gives the behavior of this ionization chamber type of gamma instrument (AN/PDR-T1B) which was supported about 6 feet outboard and about 6 feet over the sea, and was protected from spray by a pot-shaped, steel tank having '4-inch thick walls. This instrument had been sealed inside its protective pot at about 1200 hours, 5 May; unfortu- nately, no provision had been made for the zero knob to be adjusted repeatedly to compensate for drift, and the drift had to be allowed to accumulate for many hours. The actual readings are indicated by circles on Figure 3.2, and a straight line extending back to the time the instrument was last zeroed before its being sealed up is drawn to indicate what is believed to be the drift of the instrument’s zero. The net gamma dose-rate reading of the T1B instrumentinside the pot after being corrected for drift is given by the solid curve below. Beyond the time 1800 on 6 May, the drift unbalance is so large that no confidence at all can be placed in the readings. Measurements summarized by Figure 3.2 serve mainly to interpolate the measurement of surface activity through that period when the most intense peak value of the latter occurred. 3.3 REDUCTION OF READINGS TO ROENTGENS PER HOURIN SITU 3.3.1 Correction for Instrument Contamination. The metal instruments collected measurable amounts of activity on their external surfaces while being towed through contaminated water. This was demonstrated by removing instruments, one at a time, from the sea and cleaning their surfaces with sand paper and with chemicals; the signals almost always dropped after these cleanings, giving evidence that part of the signal came from surface contamination. Figure 3.1 indicates where and when the instruments were cleaned and how muchthe signal decreased consequently. It can be safely assumed that the residual signal, immediately after a thorough cleaning, was due solely to the activity in the sea. However, the law governingthe rate at which an active contamination of this sort accumulates is not at present known, so that the contribution to the signal due to contamination can only be estimated except at the few points where an actual washing was carried out. It is likely that the rate of accumulation is a function of time and is also a function of concentration of active material. It is unlikely that the accumulation process is completely reversible, and it is unlikely that the surface contamination will wash away in clean water at a rate related in any simple way to that at which it has accumulated. No data was recognized as giving any lead to the nature of contamination buildup, so that a simple accumulation proportionality with the time of exposure was assumed. In Figure 3.1, the dashed curves are the results of subtracting from the raw measurements a contamination~produced signal which increased directly with time and which was independentof activity concentration in the sea. Alternative assumptions concerning the rate of surface contamination were later considered and extensive computations made and the results then compared with the simple running correction shown in Figure 3.1. It was found numerical results were not greatly different when the correction was assumed to depend upon concentration also. 3.3.2 Running Plot of Relative GammaIntensity In Situ. Figure 3.3 is a plot of the relative gammaintensity in the surface water derived from the readings of the towed in-~ 27