SINCE 1930, we have been interested in the treatment of cases of angina pectoris with roentgen therapy. It was thought that the effect of irradiation, if any, was upon the sympathetic ganglia, and that pain was relieved in a manner analogous to the relief sometimes obtained after alcoholic injection. A similar line of reasoning was used by others in the justification of roentgen therapy in many diverse conditions, including Raynaud's disease, lichen planus, pruritus, and dysmenorrhea. It soon became obvious that it was extremely difficult to evaluate the results. Patients with angina pectoris frequently pursue a very variable course; they are of necessity treated with several therapeutic agents simultaneously. An improved morale and spirit of optimism may in certain cases mislead both patient and physician as to the true progression of the condition. The technic of roentgen therapy entails manipulations which may act powerfully on the psyche. Thus it was found that, while many cases of severe angina showed improvement, and so stated, they all tended to relapse. Other persons with anginoid pains, definite nervous disturbances, but little in the way of organic cardiac findings, might remain permanently improved. It seemed desirable that some objective experimental data be procured, preferably on animals, bearing upon the basic question: Can roentgen irradiation affect the sympathetic ganglia? The following three possible procedures suggested themselves: (1) Irradiation of the ganglia in situ and subsequent microscopic examination. The objection to this was technical. It is uncertain what degree of damage to a ganglion must exist before it is apparent in the fixed stained preparations now available. It is probable that an attack upon the problem in this manner would require the development of an entirely new technic. (2) After ganglionectomy the temperature of the skin which has suffered an interruption of its vasomotor nerve supply will be relatively high, at least for a time. It is customary to compare corresponding parts on opposite sides. This method is satisfactory and simple following operative procedures, which can easily be unilateral. The proximity of the paravertebral ganglia to each other, especially in small animals where they may lie only from 2 to 3 mm. to each side of the midline, practically precludes the use of this method following irradiation. (3) After ganglionectomy the peripheral sympathetic nerve plexuses, lying in the adventitia of blood vessels, degenerate. This degeneration is complete in from 45 to 60 days. It was realized that failure to obtain this degeneration would not necessarily mean that there had been no effect on the sympathetic ganglia, but only that the effect was not comparable to operative ganglionectomy. This was the method chosen.