The seasonal variation of the Sr®° is remarkedly similar to the seasonal variation of the total ozone in the atmosphere which has been observed by many workers. Based on measurements of the distribution of ozone and water vapour in the atmosphere, Dobson’ has proposed a model for the general circulation which offers a satisfactory explanation of the observed sr*° data. In Dobson’s model, a very cold pool of air forms above the winter pole during the late winter months when the air lies in shadow. The ultimate sinking of this pool, carrying ozone- rich air to lower levels in the stratosphere, is believed to be the cause of the rapid increase in total ozone in early spring in high latitudes. Since the Sr®° concentration in the stratosphere increases rapidly with height! this subsidence would also bring Sr*°-rich air into the lower . stratosphere in early spring, leading to a seasonal variation of the concentration in this region of the atmosphere. The interchange of air between the stratosphere and the troposphere has been discussed by Brewer’ who, in order to explain the form of the water-vapour curve in the stratosphere, has suggested a circulation system in which tropospheric air enters the stratosphereat the equator, travels in the stratosphere to temperate and high latitudes and then sinks again into the troposphere. This circulation provides the means for bringing stratospheric Sr®” down into the troposphere where the concentration would be expected to show the seasonal features discussed above. The form of the global deposition curve (Fig. 5b) supports the view that the stratospheric air enters the troposphere in the middle latitudes, bringing with it Sr®° which is progressively washed out of the troposphere by rain water as it travels north and south from the region of entry. Finally, the subsidence of the belt of cold air in the stratosphere above the winter pole would be expected, from continuity considerations, to initiate a meridional circulation in the stratosphere in which there would be a more or less continuous flow from the summer to the winter hemisphere.® This flow might provide the explanation for the presence and deposition in the southern hemisphere of Sr®° from clouds which were generated in the northern hemisphere. 9 PRESENT AND FUTURE LEVELS OF Sr’ IN U. K. SOIL Mostof the Sr°’ deposited has fallen since May 1954 and, in considering future levels on the ground, computations will be based on experience since that date. In the successive yearly periods starting in May 1954, the sr°° deposition at Milford Haven has been 2.06, 2.24, and 2.55 mc/km? respectively and the cumulative level in May 1957 was 7.5 mc/km?. These annualfig- ures show an increasing trend but since it is unlikely that the controlling meteorological factors remain constant from year to year, a forecast based on the small differences between these figures would be inaccurate and possibly misleading. It is considered preferable to calculate what the future levels of Sr’® might be from some simple assumptions about the frequency of nuclear tests in the future. The formulae used are given in Appendix 3. The simplest assumption to make is that weapons will be tested at such a frequency that the mean ratio of deposition of Sr*? will maintain the value it has had in recent years (2.3 mc/km?/year). On this basis the level of Sr®® on the ground at Milford Haven will reach an equilibrium value of 92 mc/km? in about 100 years time. In this simple model, no assumption has had to be made concerning the size of the stratospheric reservoir of Sr*®. A second simple modelis that in which firing is presumed to cease in the middle of 1957. The size of the stratospheric reservoir is now relevant but this cannot be estimated unless the rate of deposition of stratospheric dust is known. A figure of 12% per year was previously obtained! from an extrapolation of the fission product content of the lower stratosphere but this figure cannot be regarded as more than a rough estimate. The results plotted in Fig. 2, showing substantial deposition from the 1954 tests up to 1956 at least, suggest that the rate of deposition is certainly not very high and it would appear to be reasonable to choose an upper limit of 25% per year. With this assumption, the size of the effective stratospheric reservoir, i.e., the total amountof Sr*° available for deposition at Milford Haven, may be calculated to be about 10 mc/km’ in 1957. If tests now cease, the ground concentration of Sr®° at Milford Haven would be expected to increase and pass through a maximum value of 14 mc/km? in 1964. If a deposition rate of 12% per year is used, the maximum valueis 18 mc/km?, to be reached in 1969. 243