2 uecece en nn ween enti, AAO ea REOPENAR A Stara se Rete ‘uc 4 pai amare se he Mi aEala Kista set inletMah te a a : , “~ oe - eee ° 4. 4 Calibration. mt | ke Radium wag used in the calibration of the high level unita, Co ths TH=7-4 scale. The original aster scale calibration on the TiejJ-B and the THe7-A wnite remained unchanged throughout CASTLE. The THe3<¢ was found +%o saturate above 20 ar/he requiring a special calibration curve * Sor correct interpretation of the scale above that value. the latter unit vas used o aly where intensities were expected to be lese than 20 ar/hr, 4.0., Clighta originating at Guaaz and Oahu, + SCOMAMETER:calibration vas generally checked before eacs use by Te.29 at Kwajalein and WILSOW cloud traciring aircraft at Enivetok. . brett o Be - » SCIVTAMETER acintenance was performed at Biivetok, > and Quam. Air conditioned wrking space. wae evailabie- ae ‘| only ab Entwetok. Because of the high resistances employad in ” ¢he cireuit, the excessive hunidity altered their value wrencrer the SCINTANETER case was opened. the § + mei ae % point was accomplished by trial ant error. With t out of ite case, the float point would be set so meter indicated the properradiation intensity; then 28 +h. would be reasserbled with a package of desicant with::. and efter & few hours the error (difference between =etor and true r adiation background) was noted. - The case was: i 3 NNT Unioat LpLnhn wom Bosses: pooch eb rece fore oes ¥: s eane ee vere oa cee mmendvenebnmnnt oe daw cones cag wees . SITY! ee ot - Sg eg = NySictmim ipecginsn: cheTUL arteus t Hevads on different cocasions agree remrkably well with each other but differ from both sete previously mentioned. The averagee of the GENE and Nevada data are plotted on Figure 36. The Peasone for the discrepancies are not clearly understood, any coebination of the possible errors aentioned above may bv res« ponsible, Tt ie felt that the effects of fission product ase é type of bosb, and tastrument energy dependence are factors w hich require further investigation. and the float point adjusted to compensate for ths The procedure was repeated until agreenzint of innt reading with true background was achieved, Whenever the float point was checked, a reading was taken in a Padiation Meld equivaient to greater than half scale deflectior te insure that the remainder of the circuit was operating properiy. wibee tie ALTITUOE Ange CrQUNG :