When we began to use rotation therapy at 250 kv. early in 1952, dosage determination methods by Hawley (1), Nielsen (2), and by Kligerman, Rosen and Quimby (3) were available. We adopted the method of the latter group, and it remains our basic method of dose assessment to date. A short time later, Wachsmann (4) suggested the measurement of the dose rate of the transmitted radiation as a measure of the tissue dose at the axis of rotation. We set out to confirm the validity of this system for 250-kv. radiation. The physical arrangement of the rotation chair, x-ray tube, and transit chamber are shown in Figure 1. The x-ray machine is operated at h.v.l. 1.7 mm. Cu, at a target-axis distance of 85 cm. The transit detector consists of a Victoreen Radocon with a 25-c.c. thimble chamber placed in the front surface of a presdwood phantom at 150 cm. from the target. A condenser r meter chamber of equally high sensitivity, such as the Victoreen 2.5-r chamber, would be satisfactory in a similar presdwood phantom, which we considered would minimize differences in the amount of scatter picked up from objects around the transit chamber in different rotation therapy rooms. A 25-r Victoreen chamber was used in a plastic jacket for making the axis measurements in water phantoms of various shapes and sizes. All measurements were expressed as per cent of the air dose rate at the axis or transit positions, as appropriate. Measurements were made simultaneously of axis and transit dose rates in air and at the center of a series of water phantoms, ranging in diameter from 11 to 35 cm., and in a series of elliptical base phantoms of equal base areas but with diameters in ratios of 1:1, 3:2, and 2:1. These studies were repeated for a range of portal areas, for elongated portals, and for non-central tumors, since these gave values which did not fall on the curves for central tumors. Further, the density of the phantom was changed by adding paper and air, bone, etc. Results Figures 2 and 3 show the variation, in per cent of air dose rate with portal area, with thickness of phantom as parameter, at the axis and transit positions respectively. From these curves, a family of curves was drawn (Fig. 4) relating the percentage air-dose rates at the axis and at the transit positions for various areas. The relationship shown was demonstrated to be valid for the elliptical phantoms (i.e., to be independent of body shape), and for phantoms containing water and air or water and bone. Observations on the effect of elongated portals on tumor dose are in agreement with those of Clarkson (6), and this correction should be applied for portals whose sides differ in length by a factor of 2 or more. This situation is not often encountered in our department.