EFFECTS OF EXPOSURE GEOMETRY 569 under manypractical conditions only air dose is available, and since LD5. values have conventionally been expressed as entrance air dose. The following conclusions are drawn: 1. The tissue dose and thus the biological response, for a given exposure as expressed conventionally in terms of entrance air dose, can differ by a factor of greater than 2, depending on the geometry of exposure. It is recommended that tissue dose be used wheneverpossible. 2, With unilateral exposures, the tissue-dose distribution is markedly nonhomogeneous even with the most penetrating laboratory radiations. Laboratory radiation sources have not been used in a way to simulate the measured dose distribution from the initial y-rays of the atomic bomb. 3. Bilateral radiation with 250-kvp X-rays yields essentially uniform tissue- dose distribution, and no appreciable increase in uniformity is obtained with the more expensive higher-energy machines, or with more complicated techniques such as multiport, rotational, crossfire, ring, or 4a exposures. In all such exposures, the depth-dose curve is essentially flat throughout the phantom; however, in all instances the tissue dose is less by an appreciable degree than is the entrance air dose. Use of the midline tissue dose to characterize these exposures is recommended. The pattern of dose distribution from fallout y-radiation can be simulated satisfactorily with bilateral exposure, with the exception of the first few millimeters of tissue. +. There appears to be no satisfactory method of comparing quantitatively unilateral exposure with exposures yielding more uniform dose distribution. Use of gram roentgens, or integral dose, is of little value in this regard. 5. For large animals there appears to be no detectable intrinsic energy dependence of response for X-rays over the range of 250 to 2000 kvp, and for the initial or fallout bomb y-radiations. Reports indicate that undegraded laboratory y-ray sources maybe less effective than lower-energy X-radiations. 6. The application of the considerations set forth in the present paper to the LD, for man is discussed. There do not appear at present to be sufficient data to allow a better estimate of the LDs. for man than the currently accepted +400 to 500 yr (geometry or energy not specified). Present considerations indicate that the LD», expressed as midline tissue dose, may be lower than this for most types of exposure. 7. The relationships of geometry and energy to dose received, discussed in reference to animal mortality, applies also to the practical problem of sterilizing large volumes of food. ACKNOWLEDGMENT The authors are indebted to Mr. James Gilmartin for his aid in carrving out the technical aspects of these studies. REcEIVED: October 8, 1956