Nonuniform distribution of absorbed dose is frequently encountered in the irradiated mammal; the degree of nonuniform distribution is generally more severe as the size of the animal increases and the energy or penetrating power of the radiation decreases. However, acute mortality under these conditions, e.g., from the hematopoietic syndrome, appears not to be consistently predictable from the dose at any given location or locations within the animal. It is thus reasonable to seek a biological quantity that may be adequate for this purpose. Accordingly, it was postulated that, in animals dying from the bone marrow syndrome, survival is determined by the total number of viable stem cells remaining in the entire body, independent of their distribution. To test this hypothesis, the ${\rm LD}_{50/30}$ value for mice exposed to nonuniform irradiation of varying degrees of severity was obtained, as was that for mice receiving uniform total-body irradiation. The distribution of bone marrow in transverse segments of tissue along the spinal axis was determined, as was the dose to each of the segments. The data were analyzed by multiplying, for each segment, the fraction of stem cells in the fraction of cells surviving, as determined from the dose and a survival curve for stem cells determined separately. The sum of these products yielded the surviving number of stem cells in the total mouse, for both the uniformly and nonuniformly exposed animals. The surviving fraction was found to differ by no more than 20%; this was taken to be reasonable evidence that, based on the number of surviving stem cells, it is possible to predict the mortality rate for both uniform and markedly nonuniform irradiation.

A probabilistic approach has been developed to relate microdosimetry, biological effects, and radiation quality. It is used to derive, and subsequently apply, microdosimetry-based cellular response functions for different biological end points of relevance for radiological protection. The approach makes use of measurable microdosimetry spectra and avoids assumptions concerning the course of mechanisms of radiation action. Instead, it postulates a response function that is, and behaves like, the cumulative probability that a subcellular target structure will respond to a specific target-averaged ionization density. Statistical distributions are applied and their parameters are evaluated to characterize the randomness involved in the localization of sensitive sites and in the reactivity of the whole sensitive structure. The resulting response functions can be used for prediction of the effects of low-level radiation. Such predictions for some selected effects of a stochastic nature (mutagenesis, chromosome abnormalities, etc.) are presented as relative biological effectiveness values based on low doses of radiations with a wide range of linear energy transfer and compared with various quality factor specifications. Cellular response relationships, termed hit-size effectiveness functions, can also be applied directly in radiation protection metrology by incorporating them into the software used to process the readings of microdosimetric spectrometers. The derivation of the functions, rather than their uses in radiation protection, is the principal subject of this report.

The microdosimetric distribution of event sizes, especially for small exposures and high-LET radiation, represents both a fractional involvement of the exposed cell population and variable amounts of energy transferred to the "hit" cells. To determine the fraction of cells that will respond quantally (be transformed) after receiving a hit of a given size, a hit size effectiveness function (HSEF) which appears to have a threshold has been derived from experimental data for pink mutations in Tradescantia. The value of the HSEF at each event size, multiplied by the fractional number of cells hit at that event size, and summed over all event sizes, yields a single value representing the fractional number of quantally responding cells and thus the population impairment for a given exposure. The HSEF can be obtained by unfolding (deconvoluting) several sets of biological and microdosimetric data obtained with radiation of overlapping event size distributions.

Microdosimetric event size spectra (for simulated mean chord length of 0.66 μm) were measured using a cylindrical "wall-less" and a spherical walled proportional counter at primary photon energies of 660 and 1250 keV. These spectra were also measured in a water phantom (30 × 30 × 30 cm) at depths of 2, 5, and 10 cm. Heavily filtered 250-kVp X rays (providing a mean effective energy of 215 keV) and 60-keV <tex-math>$({}^{241}{\rm Am})$</tex-math> photons were also used to measure the event size spectra in air and at depaths of 2, 5, and 10 cm in water for a simulated site diameter of 1 μm using only the walled proportional counter. These measurements were compared with previously measured spectra for 60-, 660-, and 1250-keV primary photon energies. The differences between wall-less and walled measurements were studied. Mean chord length of 0.66 μm was simulated either by Rossi-type tissue-equivalent (TE) or propanebase TE gas. The values of microdosimetric quantities <tex-math>$\overline{Y}_{{\rm F}}$</tex-math> and <tex-math>$\overline{Y}_{{\rm D}}$</tex-math> from these measurements were compared. Our results without a phantom (determined from measurements obtained with a walled counter) showed approximately 10-30% higher values for <tex-math>$\overline{Y}_{{\rm F}}$</tex-math> and <tex-math>$\overline{Y}_{{\rm D}}$</tex-math> than those determined from measurements obtained using the cylindrical wall-less counter. However, ratios of <tex-math>$\overline{Y}_{{\rm F}}(\text{water})/\overline{Y}_{{\rm F}}(\text{air})$</tex-math> and <tex-math>$\overline{Y}_{{\rm D}}(\text{water})/\overline{Y}_{{\rm D}}(\text{air})$</tex-math> were within ±4% for both the wall-less and walled counter measurements.

Visual sensations were experienced when bursts of high-energy pions passed through the dark-adapted right eyes of three human subjects. The threshold for a visual sensation was typically 1-3 μrad at the retina. Data are presented to show that the mechanism is Cerenkov radiation generated within the vitreous humor. Threshold measurements agree with published optical data. A comparison is made between our observations and the light flashes observed in deep space by Apollo astronauts.

Opacification of the optic lens in mice has been evaluated after exposure to the following radiations and doses: 250 kVp x-rays, 4.5-1,000 rads; 14 MeV neutrons, 0.5-60 rads; 1.8 MeV neutrons, 1-36 rads; and 0.43 MeV neutrons, 0.02-104 rads. Highly refined techniques of serial in vivo lens examination and scoring of minute defects (opacities) were employed for this very radiosensitive system. Data analysis was based on the relative accumulation of opacities in irradiated compared to untreated lenses. (Both revealed the same degenerative process.) A nonparametric analysis of the observation discloses that the RBE is inversely proportional to the square root of the neutron dose and that it can reach values in excess of 100. The findings are in accord with the theory of dual radiation action.

Data are presented on negative π mesons from measurements in a low intensity beam of the Cosmotron. Analog data for protons and positive π mesons are discussed for comparison. The energies of the particles were selected to give a range in water of 15 cm. A Čerenkov counter used in the negative beam reduced the background of muons to 15% and electrons to 20% of the registered particles. Pulse-height spectra from scintillation detectors permitted a determination of the depth dose in polyethylene relative to the dose without absorber. The relative heights of the ionization peaks measured with a CsI(Tl) crystal were 3.1, 2.4, and 2.1 for protons, π - , and π + mesons, respectively. Subtraction of the muon-electron beam contamination results in a peak height for π - mesons of 3.0 relative to the plateau of the ionization curve. Plotting the counts registered in the high amplitude channels of the pulse-height analyzer vs absorber depth shows, at greater depth, the contribution by π - capture processes.

BNL Swiss Albino mice were exposed (five in tandem) in a 2.5-cm I.D. Lucite tube to a parallel beam of 2.2-BeV protons. The LD 50 was 1.81 ± $0.03\times 10^{10}\ {\rm p}/{\rm cm}^{2}$ , or 641 rads. The corresponding LD 50 for 250-kVp x-rays was 557 rads, yielding an RBE of 0.87. No difference in time pattern of death was observed between the x-irradiated and proton-irradiated animals. It is concluded that, with the exposure geometry used in these experiments, ionization by primary and high-energy secondary protons was the major dose constituent. A comparison is made with other experiments on the lethal effects of protons in which different geometries were employed. There is evidence that, with exposure in material of larger diameter in which there is a larger contribution to dose from lateral scatter, high-LET components of the beam may play a more dominant role. It was also observed in these experiments that the presence of Pseudomonas aeruginosa may result in a lower LD 50 and "early death," following either x-irradiation or proton radiation. This may have accounted for some of the "early deaths" following proton irradiation reported earlier.