In the quest for improvement of radiation therapy of malignant tumors the utilization of fast electrons has appeared to us particularly attractive (1). The knowledge that the biological effects of x-rays and radium are due to secondary electrons is, of course, not new; electrons with an equivalent of 1.2 MEV have been used for more than fifty years in the so-called beta plaques of radium sources. Free beams of electrons in the multimillion volt range, however, have found only limited use because until very recently the technical facilities for their production were not available. The advantages of utilizing electrons directly for treatment of malignant tumors localized in the depth of the body result from their characteristics. First, their range, and consequently their penetration, are determined by the primary energy. Second, the ionizing property in a free beam of electrons is not concentrated at its source, but is more homogeneously distributed and, therefore, more favorable than in the conventional beams of radiation produced by x-ray apparatus, radium, and cobalt-60 sources. The penetration of electrons was roughly calculated as being about half the distance in centimeters as the creating energy in MEV. For practical purposes it was assumed that the formula (MEV/2) — 1/2 would be correct, indicating that a beam produced with an energy of 10 MEV woul dpenetrate 4.5 cm., while a beam with an energy of 20 MEV would penetrate 9.5 cm., etc. A confirmation of such calculated data could, of course, be expected only after electrons in the multimillion-volt range had been produced. This was first accomplished in 1946, when a free beam of electrons was liberated from a 20-MEV betatron placed at our disposal by Kerst for experimental purposes. Electron beams exceeding energies of 16 MEV were produced by Skaggs (2) and co-workers, who also undertook the collimation of such beams and the determination of isodose curves. Continued research in this field convinced us that electron beams within this energy range were helpful but not sufficient for the clinical use which we visualized, and that electron sources with energies of at least 35 MEV and preferably higher should be available. The 20-MEV betatron used by us had never been intended for application in medical practice and offered no flexibility. Since we were fortunate in having the benefit of Kerst's knowledge, experience, and cooperation, we decided to develop a betatron with a capacity of 40 MEV or higher, specifically for medical purposes. This apparatus was intended to provide energies between 5 and 40 MEV and to have sufficient flexibility to permit convenient exposure of patients; it also should allow for easy determination of energy output and contain simple controls of the mechanical and electrical parts. Blueprints for such an electron betatron were developed in our laboratory by Skaggs and Mueller, with the advice of Kerst and his staff, and were finished in 1951.