fs a | Ve retained and otudied, et lest initially, in the seme crec. This will nescsssit:te a ecabined laboretcry building end animal house at e shere bicod wee installation in the test erea. Such a feceility could be established In sone lerge Quoncet hut or T-~53 Buildings srd would provide working ssece for the entire biclosical sroject deserisad above. “ condition the building for the test It might be necescary to air t umisms. This should invclve a con~ sicsrable amount of pre- and post-test work in the test erea by indivicurlis conceztied with the virloeus experiments. The number of memicls contemcicved er test is around cne thousand and n-t over two thousend. Mice don't tike up much room or require tremendous quantities of food, so the logistic xrcblem se oration for any vresesed tests should begin with cn Gully ace ption of « plan : ef ex.erimernts to be done, with the jmediate inlvioGlor of vor: ti: secure -relsently unknown deta, and to crovide eeerunts fool ler and serviennuzl to vigorevsly execute the acopted plen of sctioa. x duties must be on a ful. time besis avid have no eduiticne? funetions to monitoring. exrform, such as rediolorierl ssfety It is estimuted 1% + the comelete ororrem outlined in this paper could ke au:scuted oy no ucre then tventy blologists, and srobsbiy less, in the tess area,