These observations tend to support Selye's hypothesis that the effect of the adrenal steroids is to inhibit the capacity of the tissues to respond to a noxious stimulus by inflammatory reaction. In these experiments, the inflammatory response was well under way before the hydrocortisone was introduced into the injured tissues; however, the arthritis caused by the turpentine and the granulomatous process by which the tale was isolated from the joint cavity continued to progress. The gross appearance showed a diminution of the vascular-dilatation element of inflammation in most instances on the treated sides and the microscopic appearance showed as a rule fewer infiltrating phagocytic cells on the treated sides. These findings suggest not only that there is an inhibition to the progress of the inflammatory response, but also that the tissues tend to return to a normal state more rapidly after treatment. In our experiments, the injury was relatively mild and the hydrocortisone seemed to exert a favorable influence on recovery, although the difference between the treated and the untreated sides was not great. If the hydrocortisone had been administered at the time of the injection of the irritant, it is probable that the effect would have been more marked.

There was no tendency for the hydrocortisone to accentuate the damage and to cause necrosis, as was noted by Selye in his experiments with croton oil. It seems probable to us, however, that the thick barrier between the skin and the inside of the granuloma pouch in Selye's experiments, in which the skin did not become necrotic, was a mechanical one and that the steroid really did not actually enhance the necrotizing effect of the croton oil. In more severe injury to tissues, as Selye showed in his experiments, the anti-inflammatory effect of hydrocortisone may result in greater damage to the tissues involved, because the inflammatory reaction may have a protective effect by creating a shield.

In the clinical application of these principles, it seems logical to assume that hydrocortisone would be most useful in those pathological conditions which are inherently localized and self-limited, in which the symptoms are largely due to the inflammatory response per se, and which are either progressing or in which the noxious agent is continuing to act.