Although it has been emphasized that the removal rate is 20 per cent per year or more on the average, it should be pointed out even more strongly that we are entirely unsympathetic to the use of a fixed percentage removal independent of the latitude or altitude at which the material is present in the stratosphere. The implication of such a model is that mixing in the stratosphere is very fast and that the hold-up in the stratosphere is due to the difficulty of material in penetrating the tropopause. We are almost positive that there is nothing unique about the tropopause except that it is the bottom of the stratosphere. We are also quite sure that the hold-up in the stratosphere stems from the slow vertical mixing throughout the lower stratosphere in just the same way that smoke emitted from a chimney on quiet nights shows no measureable vertical mixing. We believe that the latitude as well as the altitude of injection will de- termine the amount of Sr*’ removed at various times. 6 FUTURE PREDICTIONS The predictions of future fallout depend on the rate and nature of atomic tests and some of the meteorological features just described. Two types of future testing are usually treated: first, cessation of all tests right now and, second, continuation of tests at the same rate and in the same fashion as the past. The purpose of the discussion below is to compare predictions by two models; certain features will be common to both computations. 1. The observed accumulated deposition, from whatever source, equals 25 mc/sq. mile in the latitude band from about 30° to 50°N. 2. Future predictions will be limited to this same band. 3. There have been four years of testing up to the end of 1957. This is not strictly true, but this will be assumed for purposes of comparing the two models. 4, The rate of injection of Sr® into the stratosphere is 9 Mt/year or 4.5 mc/sq. mile/year averaged over the earth. The two models to account for north temperate latitude fallout are as follows: (1) Stratospheric fallout is uniform over the earth and is being removed at the rate of 10 per cent per year, (2) Stratospheric fallout is being deposited preferentially in the north temperate latitudes; the 30°—50° band is three times the world average, and the stratospheric removal is at the rate of 20 per cent per year. Table 2 shows the maximum deposition in the 30—50°N latitude band and its time of occurrence if tests are stopped at the end of 1957 (already in error since there have been tests sub- sequent to that date). Our estimate is just less than twice the December 1957 fallout value, and the alternate is about the same as the present value. Neither, as we shall shortly see, is very high compared with the fallout amounts if tests continue. Table 3 shows the comparison for a continuation of tests at the hypothetical past rate. Once again the model involving nonuniform stratospheric fallout gives higher predictions, but by only 50 per cent even though the nonuniformity factor for stratospheric fallout is three times greater. At times prior to equilibrium the differences between the two models will be even smaller than 50 per cent — in fact, as of today, the difference is zero since both must ac- count for the observed fallout. At other latitudes, the two models also give different forecasts. In the equatorial region, for example, the nonuniform stratospheric fallout theory gives about 100 per cent morefallout at equilibrium than the uniform theory, but it is, of course, much smaller than the fallout in the 30—50°N band. At about 45°N,the latitude of the heaviest fallout, the nonuniform stratosphere model predicts about 75 per cent more than the uniform model. It should be noted that the percentages reflect differences between two models and not differences from the correct answer. One would then like to convert these fallout forecasts to Sr®’ bone content. However, in doing so, one finds much more uncertainty in the conversion than was found in the estimates of the amountof Sr*® fallout. Predictions of this kind must be accepted with considerable reservation for many reasons besides the fact that 100 years is a pretty long-range forecast. They may be too high because of the exchange of food between different geographical regions, because of unavailability of the Sr® to agriculture due to aging and plowing of the soil as de- 336 Best