Appendix 2 THE ORIGIN OF THE Sr*® DEPOSITED IN THE U. K. AND IN OTHER PLACES REMOTE FROM TESTSITES 1. It is comparatively simple to show that the sr” deposition in the U. K. is primarily of stratospheric origin. It is generally agreed that tropospheric dust is deposited in a matter of weeks and a mean atmospheric residence time of 31 days has been obtained by measurement.' An upper limit can therefore be set to the amount of Sr®® of tropospheric origin in a sample by attributing all the observed Sr®to that source, assuming mean age of, say, 35 days, and computing the associated Sr®*, It has been shown by this method that tropospheric Sr*® con- tributes less than 5% of the total observed during the 1955 peak period in Fig. 3. In more detail, if one assumes that sample 17 (Table 2), which gave a particularly high Sr®®/Sr’® ratio, was a mixture of 1954 dust and dust from the U. S. spring tests of 1955 and that the latter was 35 days old, the contribution of the 1955 tests can be shown to be only 8% of the total, and this is certainly an upper limit. Similar reasoning can be applied to the more complex 1956 peak andit can be shown ...2t an upper limit to the contribution of tropospheric dust to the total deposited Sr® is 12%. 2. It is possible to demonstrate the importance of old fission products in the rain water studies by examining the Sr®/Sr®° and Ce!44/sr*® ratios in Table 2. For fresh fission products these ratios would have values of 154 and 34 approximately and these would decay withhalf lives of approximately 55 days and 285 days respectively. The ratios in Table 2 are in general much smaller than their theoretical initial values, indicating the presence of old fission prod- ucts. The highest Sr®*/Sr®° ratio was obtained on sample 38 and the method of para. 1 shows that the short-lived material contributed 45% of the Sr’in this sample; the sample was of low specific activity, however, and the correction serves only to increase the maximum to minimum ratio in Fig. 3. 3. Computations based on the measurementof the gross fission productactivity in rain water also argue in favour of the stratospheric origin of the deposited Sr®*’, The effective age of the fission products in all daily rain water samples has been determined from the slope of the decay curves measured soon after collection. Using the general argument of para. 1, all samples whose effective ages were less than 50 days were selected as of tropospheric origin only and the corresponding amounts of sr®? calculated, using a fission yield of 4%. The result again showsthat tropospheric dust has contributed less than 10% of the total Sr*° deposited. 4. An argument supporting the hypothesis that the Sr™ has its main origin in large-scale nuclear explosions lies in the magnitude of the deposition, which reached a value of 7.5 mc/km? at Milford Haven in April 1957. Up to January 1955, it was possible to distinguish between the fission products from the various series of tests with sufficient accuracy to allow the associated Sr®® to be calculated. By this means it has been shown that the total amount of sr*? deposited at Milford Haven as a result of all weapon tests in the nominal range of sizes, prior to January 1955, was 0.19 mc/km?, Although the exact number of such weapons exploded since then is not known, it is unlikely that they can have contributed more than a few percent of the total Sr? deposited. This argument can be strengthened by considering the total world deposition of Sr’. By extrapolating the curve of Fig. 5a to zero at the poles, and integrating over the surface of the earth, it has been calculated that the total Sr®° deposited in 1956 was 9 x 10° curies. Since the rainfall at each of the measuring stations represented in Fig. 5a is not necessarily representative in amount of the rainfall within its own belt of latitude, a better estimate of the total sr*° deposited per year in recent years can be obtained by combining the specific activity curve of Fig. 5b with the mean annual rainfall in various latitudes’ and integrating over the earth’s surface as before. The value thus obtained for the annual deposition of Sr® is 6.7 x 10° curies which, assuming a 4% fission yield for Sr*’, is the amount associated with the fission products from a 5 MT explosion, or from 2,500 nominal explosions. This number alone suggests that nominal explosions can have contributed a small fraction only of the annual deposition of Sr? in recent years. 246