It is a well-known fact that when a beam of radiation strikes matter, some of the rays are scattered in all directions. The amount of scatter depends upon the quality of the radiation, the nature of the material upon which it impinges, the volume of the material, and the area of the beam. The relative amount of scattered radiation is different in different directions, being greatest in the forward and least in the backward direction. The portion scattered backward is of particular interest in radiation therapy, because of its contribution to the dose delivered at the surface of the body. This dose consists, in part, of radiation from the primary beam, and, in part, of that scattered back by the underlying tissues. Since all depth doses are determined in relation to the surface dose, it is necessary to know this accurately. Moreover, it is not possible to determine the erythema dose in terms of physical units (roentgens) until the intensity of a beam as measured in air can be related to that on the surface of the body. Many attempts have been made to measure the back-scatter experimentally, by means of ionization chambers, photographic films, biological media, etc., placed upon the surfaces of water, paraffin, and the human body itself. The results obtained have disagreed widely. Most of the differences may be explained by variations in experimental conditions. This is particularly true in the case of physical measurements, in which instance the nature and material of the ionization chamber are extremely important. When approximately the same experimental conditions are used, there is fair agreement in the results obtained. The method of choice so far has been the use of the small ionization chamber of organic materials, the assumption being that its walls have no effect on the results. Some of these chambers, when calibrated against a standard open air chamber, agree with it over a wide range of quality of radiation. However, the scattered radiation contains components of very long wave length, which may be more or less completely absorbed in the walls of the chamber. At the same time, the secondary radiation from these walls and from the inner electrode may introduce a disturbing factor. The observations here reported were made with a new type of ionization chamber, which is shown diagrammatically in Figure 1. This consists of three parallel plates, about 40 cm. square, of which the two outer ones are made of very fine silk net, impregnated with India ink, while the inner is of fine silk threads, similarly treated, stretched 1 cm. apart. The net, when inked and ready for use, weighs 14 mg. per 100 sq. cm.; this is about one-third the weight of cellophane. The frame for each plate is a square of four strips of very light wood, 1 cm. wide and 2.5 mm. thick, painted with India ink. The two outer ones are maintained 2 cm. apart by plugs of hard rubber at the four corners, and are charged to 300 volts by dry batteries.