IN several publications I have shown that the erythema dose is a very poorly defined term, and its application is rather uncertain. This is the more deplorable because the erythema dose is considered as a unit, as well for the comparison of surface applications as for the determination of internal applications in deep therapy. Five years ago I compared (1) several erythema doses and detected differences of more than 50 per cent; at the same time I determined that a much larger dose of hard rays should be absorbed if the same biological effect is to be produced as with soft rays. Several years later my measurements were confirmed by Martius and Grebe (2), who discovered ratios of 1 to 4 between the lowest and highest so-called erythema doses. In order to overcome these considerable discrepancies, I have tried to improve the two methods which were available for the determination of dosage. The first is known as the indirect method of using the electrical and mechanical factors of the production of X-rays for the dosage calculation. This method had resulted in several dosage formulas. In previous papers (3, 4, 5, 6) I have criticized these formulas and replaced them by a dosage scheme based upon my own observations. This scheme is given by the following formula: “K” is an empirical factor, the significance of which will be taken up later. It was my intention to put this treatment scheme, which I had suggested provisionally, on a broader foundation. For this purpose I have sent a circular to most of the radiologists in this country, who do X-ray treatment work on a large scale, inquiring about their treatment conditions for the production of several skin reactions. Practically all of these circulars (46) were answered, and I am glad to say that their answers were proof of very careful and responsible observations. I gladly give credit and express my thanks for this collaboration to the radiologists of this country. Altogether I received the data of 250 erythema, epilation and similar doses. All these data have been used for the control of my dosage scheme, by calculating the factor “K” from the treatment data. This represents the strength of the applied dose for a given filter in the terms of gap, milliamperage, focus skin distance, and time, all these factors being properly weighed. As far as different filtration is concerned, it represents the time increase necessary to overcome the decrease of intensity due to the filter absorption. The end-results and the statistics are given in Chart I. The first column gives the filter used, in millimeter thickness. The aluminum filters ranged from 1 to 6 mm. ; the copper filters from ¼ to 4∕4 mm. + 1 mm. aluminum. The approximate aluminum equivalent is given too, for the various copper filters.