" ARLLOOS: (contimed) fhe third alternative, which is, I think, obvious, is that one gould pump up a tank with alr and eollect the whole sazple. Whether thie would bring you back your partiale sises in such a& way that you could find out anything sbout then, I don't know. Presumably, be keeping to the sides of the tank after bringing thea dom, CADLEs I'd like te point out that the thermal presipitater is called the thermal diffusion device, and I think the fastest break of sampling that has been achieved to date in something Like 30 co's per minute. So it's terrifically lew; or, in other words, it's enly really effective for vary high concentrations for particles, I don't think it will be usefal for what you have in wind. KELLOOGs Supposing ene imagined a constant~level balloon leveling out at the altitude ef the cloud and running one of these things for several hours. CADLEs Even at several hours at X) co's per sinute you eti1] haven't got very much. Weil, they would be better off using the power requiresents authority te run an slectrostatic presipitater. [ think it would be mch vere economical te handle. Filtration would be out alec, because of power requirements in ordar to get air through any kind ef filter. DOE ARCHIVES