568 BOND, CRONKITE, SONDHAUS, IMIRIE, ROBERTSON, AND BORG for rabbits or monkeys only (18, 41-44). These data, though meager, are in agreement with what must follow from geometrical considerations (2). No data were available for ring or 42 exposures. If exposures are expressed as midline tissue doses and possible intrinsic energy dependence is neglected, all such complicated laboratory sources and fallout y-radiation must yield essentially identical results. LDDs for man, The consideration of the geometry of exposure and beam energy bear heavily on the LD, value for man. This value has been assumed to be in the range of 400 to 500 r (45); however, the figure is of course an estimate, and the conditions of exposure are not specified. It will be evident from the present paperthat any value assigned is meaningless unless the conditions of exposure and the exact method of expressing dose are specified. There are no adequate sources of human data from which the LD, for man can be satisfactorily estimated. Neither biological nor physical dose data? from the bombings of Tiroshima and Nagasaki are sufficient to allow more than an orderof-magnitude estimate (46-48). Data from human exposures in laboratory .accidents (49-57) are not sufficient to bear heavily on the problem. Total-body exposure of patients in cancer therapy* indicates that very severe hematologic depression results from doses of 150 and 175 r (midline tissue dose, bilateral exposure). This would place the LDse value below 300 r. From the sublethal blood changes of the Marshall Islanders (7), the LDso for man exposed to fallout y-radiation was estimated to be approximately 350 r, midline tissue dose. The possible error in this estimation is verv large. The LD sy values for dogs and swine (midline tissue dose, bilateral exposure) are of the order of 250 1; that for the monkey may he as high as 500 r. It is thus apparent that the LDso value for man cammot be accurately fixed at present. The problem may be further compheated in that the bulk of acute mortality in most laboratory animals occurs within 30 days of exposure, whereas many deaths in man occur between the thirtieth and sixtieth days. It would thus appear that a 60-day mortality value rather than a 30-day value should be used for deseription of the LDs59 in man. SUMMARY AND CONCLUSIONS The influence of the geometry of exposure and of beam energy on the depth-dose pattern obtained in tissue-equivalent material simulating a large animal or man was determined for a variety of exposure techniques used in the laboratory, and for the initial and fallout y-radiations from the atomic bomb. The available LDs, values for large animals obtained under various conditions of exposure have been compared to those predicted from considerations of exposure geometry and beam spectrum. LD, values are given in terms of air dose, as well as tissue dose, since 7 Thefalloff in tissue of a flux of high-energy neutrons is rapid under laboratory conditions (17, 52), and the considerations of geometryoutlined in this paper must be taken into aecount in considering their contribution to the total biological effect from nuclear devices. ‘J.J. Niekson and H. ik. Bane, personal communication.