) aa reflecting eurface in all eespetia- i} dire Duvivg the gatiy irre [ phase, the rate of work by the shetk f¥ornt cn air es compars i to rate or soil or water is in a ratio were than S(h t& 2 an caver oo aftr. It follows that less than 1 ner sent of vhe emerpy wiki be Racer. atomic transmitted to sell or water during tnese ste,es radiatior. in thermal of effect the tests have been comserned with attenuating the peak pressuTes in a blast wave j unis effezt ML be at a minimum on Mike shot because of the giawine angles of inciden-s of thermal radiation; however, toe "Shera. effect’ wih not be completely absent because the firerali attains 6 larpe vert ica height? in a short tim. Other factors lead to difficulty in estimating the effects, The rise of the fireball and consequent afterwind leaa te an att@isation of the blast wave at close distances wiach fe diffie wt tw ostimite: thie effes? is a a maximum because of the Yow Te ym of Burshe heat, considerably higher tamperatures may be achieved in this axplosics: than on an ordinary nuclear explesion; this leads to greater Llossas in ewargy through irreversible heating, te 2 diffesernt "partition of energy", to the possibility of a greater fraction of energy appearing as thermal radiation, and to the poasibiiit; of a amslier comparable blast yield, Again, the expicsion is so large that the atmospnere Gan ne loner be considered as homgenesue; the top «5 the blast wave will be in rarefied atmosphere at a tine when ground pressures are still in the regicn of practical interest. Coneideratle blast experimentation wili be devcted te thia point whier may is ar te @ variation cf 25 per cent in yleld. spain, atmosoteeic inversion may focus energy upward or downward at icone distances; but for an ordinary bomb, this effect is usualiy at pressures reat ©. os. Gn Mike shot, the scaled height eof these inversion jayers are such that sone focussing (or defecussing} of energy may occur at pressures of interest. Finally, or the space scac.e involved here, layers of clouds are close enough te be of some conern beth frem the standpoint of energy reflection as wel’ as from the standocint thermal radiaticn. ef previding a shield fron For the mest part the uncertainties liuted are expected tc be in the order cf 25 to 50 per cent in blast yield, anc small compared te the design umcervainty of S te 5C MT, and not suffielently large to require spetif.e numerical treatment For the mst pari. the data usod in making thease estimates were taken from IEY prebliem M. which wags assumed to be JO KT. isconservative because it implies that blast efficiency of an atomic bomb is 0 65 compar:with TNT This The ectual efficiency my be as Tew as 0 5 for a convarntiosai weapen and vernapa lower for avery large weapor. Moreorer, experien4 on structures are usually based on pressure gauge realings, and these are gensrally lower, perhaps 20 per cent, than tha "ideci" values quoted here.