STATUS OF GLOBAL FALLOUT PREDICTIONS 389 altitude sources. Although the more abundant weather data in the lower atmosphere should make the low-altitude trajectories more accurate than high-altitude ones (at least after the fact), local topographic irregularities can more than nullify these benefits, Without verification, predicted and after-the-fact reconstructed trajectories are computedin blissful ignorance. It is this gap for which constant-volume balloons, the “tetrons,” could potentially provide data. Atpresent, it is likely that most meteorologists feel that the path of a cloud of pollutants at the surface can be as reliably tracked with meteorological data as in the upper troposphere, but detailed proof is virtually nonexistent. Paralleling the absence of trajectory verification data at low altitudes is the lack of experimental data on horizontal and vertical mixing after long travel times. In the writer’s opinion, the long-range transport, diffusion, and removal of contaminants in the lower atmosphere deserve prime at- tention, Sources higher in the troposphere also present transport and diffu- sion problems but, surprisingly, more is known concerning long-range atmospheric behavior, and terrain-induced mesoscale systems may play less havoc with the predictions, One should incorporate, as some investigators have already done, the three-dimensional motion via isentropic analysis and take advantage of high-speed computers. These steps may improve ground-level-contamination predictions during nonprecipitating weather. However, for many, if not most, of the mid- or upper-tropospheric radioactive clouds, the pollutant enters man’s environment during precipitation. Although there have been many excellent studies of precipitation scavenging, the fallout forecaster still cannot make a quantitative prediction of deposition in rain; therefore predictions are limited to qualitative statements. The fate of debris added to the lower stratosphere can be pre- dicted, in a broad way, with more confidence than for debris injected elsewhere in the atmosphere because there has been more experience with such sources, The physical models producing the stratospheric transport have been narrowed to two: one is a version of the Brewer— Dobson circulation and the other is the model for down-the-gradient mixing along sloping surfaces. Both appear to yield similar fallout pre- dictions for injections in most parts of the lower stratosphere. This favorable state of affairs mainly stems from the multitude of nuclear clouds injected into the lower stratosphere and followed by sampling programs of the AEC, the Department of Defense, and the Public Health Service. There is still much to be learned from tracers in the lower stratosphere and much that can be gleamed from meteorological reSearch to improve man’s ability to predict stratospheric fallout. However, one has the distinct impression that atmospheric transport and