decreased for decay of the fission isotope in question in accordance with the time elapsed since mid-June 1958. . 137 181 . Such correction is not necessary in the case of the Cs iw 8 ratio, Support of the essential correctness of this approach to the determination of Hardtack debris may be shown by examination of the activity ratios between the fission products themselves. The Zr? jes?" ratio is a par- ticularly good indicator of the age of nuclear debris in that it varies by only 10 percent between fast and thermal fission of uranium and changes rapidly with time. shown in Figure 6. Values of this ratio for Hardtack and non-Hardtack debris are Additional evidence that the assumption made in determing the Hardtack fraction are reason- able is given by the ar? ?icst?! ratio of the non-Hardtack fraction. The ratio for this portion of the debris follows a line of similar slope from November 1958 into the spring of 1959, indicating essentially a single component during this period, the apparent time of origin of which is October 1958, suggesting that this debris came primarily from the large Soviet series conducted at that time. By the fall of 1959 the non-Hardtack values depart from the initial slope, in a direction which indicates the presence of increasing amounts of older debris. Some caution is in order, however, in interpreting this change of slope of the non-Hardtack component, particularly in light of the appearance of rn'0? in surface air at this time. The fact that the non-Hardtack Zr }cs 13" ratio implies the presence of solely Soviet October debris does not preclude the possibility of as much as 15 percent of older debris being present during the spring of 1959. The magnitude of Soviet debris relative to the total activity decreased very rapidly during the summer of 1959 as will be shown subsequently. Thus the relative importance of older debris, coming down more slowly, becomes greater with the passage of time. Figure 7 shows the breakdown of debris into that coming from Soviet October, high altitude, and residual, or pre-Hardtack, By the spring of 1960 the amount of zr? had fallen to such low levels that accurate measure- ment by these means was no longer feasible. end of 1960 by using the ce! #4 e518" Extension of this method of partition of debris was extended to the ratio as shown in Figure 8, where the components due to Hardtack and high altitude are indicated as well as the residual debris. As in the case of the zr? jes)?" ratio data, there is evidence of pre-Hardtack debris still present in surface air during 1960. There is also evidence in Figure 8 of smali amounts of activity from the French test in February 1960 as indicated by an elevation in the value of the ratio for residual debris during February and March. On the basis of tracer and activity ratio determinations it is now possible to assign the observed Cs activity to various test series as is shown in Figure 9. 137 In the spring of 1959 slightly under 20 percent of the total cst 9? came from Hardtack as indicated by wi8t measurements, the remaining 80 percent coming primarily from the Soviet October series. Perhaps the most striking feature of the data shown in Figure 9 is the rela- tively short interval during which cs 194 from the Soviet October tests was present in surface air. of debris from this source subsequent to the end of 1959 suggests that it has all been deposited. The absence In light of the fact that the decrease of Soviet debris in surface air proceeds at the rate ascribed to tropospheric clearance (approximately 30 days) after the spring maximum, it may be argued that activity from this source seen after May 1959 actually entered the troposphere during the spring break. Radioactivity from the high altitude shot does not even appear at the surface until one year after detonation, rises fairly rapidly and manifests a broad maximum during the spring of 1960. Hardtack debris shows a broad maximum in the spring of 1959 as well as one of lower magnitude in the spring of 1960. Considerable information may be gleaned from consideration of Figure 9 as well as referring to Figure 4. First, the fact is clearly established that the spring maximum is indeed a meteorological phenomenon as 121