photography but as the latterwas not an interferedwith aerial. absolute essentM to scientific recording,this risk couldbe taken. Third,rain showers,which couldbe predictedas “few and scattered”,couldbe accepted,while a predictionof “heavyshowers” ~ was less acceptabledue to possible interference with surface photographyessentialto scientificrecording. Furthermore ~ it WaS essentialthat the drone aircraftbe kept away from all showersso that theirfilterswouldnot be wet. The necessityfor accurateweatherforecast ing is obvious. It was noticeable that duringperiodsof “normal”tradewind conditions~ weatherforecastscouldbe made with a high degreeof certaintybut duringperiodsof littleweathermovement,forecastswere very uncertain. This is coveredin more detailh the reportof the Meteorologist. It was desirableto wait for the very latestweatherfiformation in ader to obtainthe most accurateforecast. However,other considerations made mandatorythe establishment of certaindeadlines beyondwhich decisionsto proceedwith or postponethe tests could not be delayed. The prticiplefactorwas the extentto which operations had proceeded. For instance,if the decisionto poetand preparations pcme was reachedpriorto the take-offof drone aircraftfrom Kwajslein, succeedingdelaysof z hours couldbe acceptedfor an indefinite period. However,if the decisionto delaywere reachedaltertlw drone aircrafthad landedat Eniwetok,no more than two such delayscouldbe 25 Section111 —

Select target paragraph3