order to unfold the raw data and to present an energy spectrum for the emitted neutrons, further analysis is necessary. A considerable amount of work has appeared in the literature on the so-called buildup factor. This relates the observed total neutron flux to that expected in good geometry, but does not suffice to calculate time-of-flight corrections. In other words, one must - account for those neutrons of a given energy that are lost from the direct beam and do not reach the detector at the proper time, as well as those inscattered neutrons of higher energy that arrive late due to their longer path and falsely simulate lower energy neutrons. Suppose that the distribution in arrival time of a group of neutrons of energy E, which has been scattered and has finally reached the detector can be calculated (an approximate method for doing this will be described below). The function K (E, ¢') is defined as the number of such neutrons reaching the detector at time ¢’ per emitted neutron of energy E and, hence, falsely simulating neutrons of energy The system efficiency factor, Q(E), which specifies the signal recorded per neutron of energy E, is defined next. Since Q(E) varies slowly with E, the energy degradation that the neutrons undergo during collision with the air can be neglected. This anergy degradation is small for most of the neutrons, since the mean’ numberof scatterings ex- perienced by a neutron ig small. If A(E’) is the attenuation of the direct beam of energy E', due to scattering, absorption, and inverse-square effects, and I(t’ ) is the recorded signal at ¢’, then N(E’) = I(t") im 2/2 £3 Q(B") A(E’) “GE Q(E) K(E, t’) N(E) FE! This is an inhomogeneous integral equation for the desired quantity N(E), the source intensity per unit of energy. The integral represents the total effect of all inacattered neutrons of energy higher than &’, hence capable of being confused with direct neutrons of energy E’. The problem of data reduction from a high-altitude detonation then is reduced to: (1) an evaluation of the kernel K(E, t'), which represents geometrical and physical properties of the atmosphere and its interaction with neutrons of all energies; (2) a calculation of the quantity Q(E); and (3) solution of the foregoing integral equation. Problem 3 seems most susceptible to a stepwise numerical procedure, beginning with the highest energy groups and working downward. Experimental values of I(t’) are inserted, and successively lower values of E’ are reached. Problem 2 is essentially solved by the work on detector, recorder, and electronic calibrations given elsewhere in this report. A simple digital-computer program has been set up to calculate the kernel K(E, ¢’), assuming single scattering of the neutrons and, again, asauming no energy degradation due to scattering in the air. The vaiidity of this assumption and its applicability in the present case might well be questioned, but as a temporary expedient it seemed worthwhile pursuing. Since there is little change of velocity during collision, use can be made of the focal il