tions may occur at long intervals of time and for short periods; how. ever, such conditions will represent a transition from one of the three categories of weather situation to another and cannot always be counted upon to be of sufficient duration to ensure the success-— ful execution of a complex operational pian for a given detonation, Other meteorological incompatibilities would be the association of strong westerly winds of more than fifty knots above 20,000 feet in the Eniwetok-Bikini belt and clcud ccrditions appropriate for photo“graphy on the ground and in the air. Such conditions have existed during past operations and will occur again but they must be regarded as accidents of nature, accidents of short duration. Their oc- currence and the ability of meteorologists to forecast them cannot be depended upon in planning operations. Finally, it shouid be emphasized that the most that can be done-—-in the present state of tropical meteorclogy--is to forecast broadscale weather situations involving, in the most generai terms, the asscciation of wind systems with average clcud cover and precipitation. It is not possible to say twenty-four hours in advance that an individual cumulus clicud will be located in any particular spot at any given time. The average lifetime of 4 tropical cumlus cloud of any magnitude is only forty-five minutes and its rate of movement depends only partly on the speed and direction of the wind. Its shape and the height to which it will reach; the amount of over- hang; and the rate of dissipation of the tips will all depend on the microstructure of the air--a problem in turbulence theory which is 172