cally, since counting efficiency is greatly influenced by the soft gamma spectra of the sample. Because the IC data presented in the body of this report has been greatly modified to permit interpretation of the GITR records, all observed tray counts, together with an approximate spectral response of the crystal counter are presented in Sections C.3 and D.2. To comparethe IC collections with the GITR records, both observations had to be brought to a common time, a correction that afforded particular difficulty, since at first inspection each separate collection appeared to exhibit an individual decay. However, after detailed com- parison of the various individual decays (described in Section 3.5.1), it was found that nearly all observed decays for both Wahoo and Umbrella IC collections could be approximated by a family of five curves (Figure 3.36). Lf the apparent dependence upon time of collection is ac- cepted, criteria based on known conditions of sampling can be established for the selection of a specific decay curve for a given IC collection. Thus, these empirically determined decay curves offer a means of correcting the observed tray counts to time of deposition, which is an iimprovement over the application of a single calculated decay curve for all samples. The fact that the observed decay curves when normalized at 22 days again approach each other to within a factor of 3 at 0.2 day, permits the aSsumption that the relative magnitude of the tray counts at 0.2 and 22 days must also be the same within a factor of 3 regardless of the decay curve actually followed by the individual collections. Usually, the tray counts were made at 2 and at 6 days, times when the differences due to fractionation could be as high as a factor of 8.5. Since the decay curves at later times are better known, all IC tray counts were brought to a commontime of 22 days using the following criteria for the selection of decay curves: Curve SI: deposition at 1 minute accompanied by large amounts of water. Curve SIT : deposition at 1 or 2 minutes without large amountsof water. Curve STI Curve SIV : deposition at 3 or 4 minutes. : deposition at 5 minutes or more as long as the GITR record Curve W_ ; ail other deposition. indicates the presence of base surge at the station. The tray counts were then converted to a std-GITR response for a deposit of corresponding magnitude distributed uniformly over the coracle deck, using an empirically determined conversion factor 8 = 0.71 x 107° (r/hr)/epm (Section C.5). This factor was determined using a a® slurry and exactly the same GITR exposure geometry and tray counting equipment. The value of such a conversion factor will, of course, vary as the energy spectrum of the deposited material changes. However, since the base surge samples were known to be enriched in Ba'“°_1a'4° and since these products represent better than 25 percent ofthe totalactivity in normal fission products at 20 days, the application of this conversion factor at 22 days is at least most consistent with the known energy of the material used for its determination. The hypothetical GITR response at 22 days was then converted to the GITR response at 1 minute, using the standard decay curve. The standard decay curve, when normalized with the five empirical decays at 22 days, also passed through the region of closest approach at 0.2 day and in fact agreed within a factor of 2 with an ionization chamber decay curve obtained by Project 2.1 to a time of 6 minutes. This somewhat elaborate technique of correcting the IC collections to a common time was evaluated by examining all trays, which were counted once in the EPG at about 2 days and again at NRDL about 4 to 8 days after zero time. If the decay curve used for each of these individual samples was correct, both counts should yield the same results at any stipulated common time. Therefore, in all cases where trays had been doubly counted, the two counts were converted to the simulated GITR response at 1 minute both by the process just described and by the simple application of the standard decay curve together with the conversion factor 8. In all except two cases, the values determined by means of the standard decay curve alone showed considerably more variation. Specifically, values calculated by means of the empirically determined decay curves showed an average variation of 19 percent of the mean, whereas those determined by means of the standard decay curve showed an average variation of more mA. 109