HEMATOLOGIC OBSERVATIONS at doses well below the lethal range, and increas- ing dose results in no or minimal further de- crease in count. Lymphocyte depression appears to have no causal relationship with acute radiation deaths. c) Platelet counts have been studied most extensively in dogs (20). As with neutrophiles, the rapidity and magnitude of depression is a function of dose below the lethal range. Maximum depression occurs by the 9th or 10th day with doses in the high lethal range, by the 10th to 15th day at sublethal levels. Recovery begins during the 3d week, but is not complete by the 30th day when most studies have been terminated. Insufficient data are available to indicate the time required for complete recovery. Considerable evidence including studies in the mouse using splenic homogenates, induced bacterial infections and spontaneous infections have indicated that critical neutrophile levels exist, below which survival is correlated with the absolute neutrophile count following wholebody irradiation (15). From data on dogs, it appears that survival is likely unless neutrophile counts remain below 1,200 cells for a period of time. Platelet data on dogs indicate that animals with external purpura have platelet counts of 50,000 or below. Sufficient data on large animals are not as yet available to quantify the extent of maximum depression of either the neutrophile or platelet counts as a function of dose in the sublethal range. The response of the platelet count in the present study.was much less subject ta fluctuationthan were the neutrophile or lymphocyte counts. For the preceding reasons, systematic investigation of the platelet and leukocyte counts in large animals as a function of dose in the sublethal range are indicated. It is not possible to say at present whether severity of exposure, or of radiation damage correlates better with absolute levels of peripheral blood count, or with degree of change from control or pre-exposure levels. Some evidence on this point can be gained by comparing the deyree of depression of the neutrophile counts in Groups II and III, both of , a9 which had essentially the same calculated exposure but for which control hematological values were considerably different (the lymphocyte count is not suitable for comparison since degree of depression was essentially the same in these groups and the higher-dose Rongelap group). At the time of peak depression for each element, the neutrophile counts were essentially identical in terms of absolute counts, but considerably different in terms.of the respective control values. Thus, some evidence is afforded that absolute counts, rather than counts relative to control values, may be the more reliable index of exposure in this dose range. 4.45 Approximation of Minimal Lethal Dose for Man Someindication of severity of exposure can be gleaned from a comparison of minimum individual counts in Japanese groups in whichfatalities occurred. In general, a significant number of deaths was encountered only in individuals whose neutrophile count fell below 1000. In Group I, 42 or approximately 50 percent had neutrophile counts below2000 at some time during the observation period, and 10 percent had counts below 1000. By this criterion, then, the effective dose received by the Rongelap people approached the lethal range. In the dog (Cronkite and Bond, unpublished data), approximately an additional 50 to 100 r are required to lower the neutrophile count by 1000 cells/mm?in the high sublethal dose range. If these data can be applied to man, an additional 50 to LOO r would have placed the dose well in the lethal range. On the other hand, however, it is clear from the present data and from clinical experience with therapeutic radiation that neutrophile counts between L000 and 2000 in human beings are in general well tolerated. Human beings with these levels of neutrophiles show no clinical evidence of illness, are physically active, and generally do not need prophylactic antibiotic therapy. The people of Group [ are estimated to have received 175 ras calculated from dose rate read-