SOME ASPECTS OF FALLOUTIN BRAZIL 797 where the rainfall rises to a maximum twice a year, the fallout curve parallels the rainfall curve. In the two stations at Itaici (to the northwest of Sao Paulo) and Sao José dos Campos (to the northeast of Sado Paulo), the curves can be interpreted as showing the first faint indications of a spring maximum. The fallout does not reach a maximum in the summer months as does the rainfall, but in October and November. The spring maximum is more definitely noticed in Sao Leopoldo, close to 30°S, although it is much less pronounced than in the northern hemisphere. A large percentage of the land mass of the southern hemisphere lies between 0 and 35°S, with relatively small land masses between 35 and 70°S. The different black-body characteristics of land masses and oceans introduce a great cooling-factor difference between the northern and southern hemispheres, particularly in the 35 to 70° zone. We should therefore expect a quite different set of effects in the thermo- dynamic systems of the northern and southern atmospheres. The station on Trindade Island was begun in 1962 to study the possible ocean—continent effect. The monthly data for this station, together with that of the coastal station of Rio de Janeiro, the mountain station of Nova Friburgo, and the inner plateau station of Brasilia, are given in Fig. 2. The island and coastal stations show higher levels than the inland stations at higher altitude. The ratio between the two sets of stations, however, is not as great as was predicted by the hypothesis. No definite conclusions can be drawn, moreover, from only four stations. The monthly averages for ground-air contamination, as measured by Staplex filters, for Rio de Janeiro are given in Fig. 3. The most prominent effect is that observed during the seasons in which nuclear tests were conducted, as can be seen in the years of 1956, 1957, 1958, and 1962. What might first appear as winter—early spring maximums for the months of August, September, and October are really seasonal variations due to the testing periods. One might expect that during the nontesting years of 1960 and 1961 a slow increase in ground contamination would be observed as the northern and southern stratospheres cameinto equilibrium. This was not observed in the Staplex-filter data because of low counting rates. The effect was noticed, however, on the NRL filter data taken at the same location. A comparison between the filter and column data for Rio de Janeiro is given in Fig. 4. Although a positive correlation can be noted in these curves, no conclusion can be suggested from the data of a single station. The data that are given here should be combined with the HASL studies from many stations on the subject.

Select target paragraph3