THE BRANHAM1 sign is of proved value in the diagnosis of arteriovenous fistulas. It consists of a slowing of the pulse rate on temporary obliteration of the fistula. The response to obliteration is prompt, occurring within seconds, suggesting a reflex arc with the vagus nerve as the efferent limb. It would seem possible, therefore, to eliminate the Branham phenomenon in patients with arteriovenous fistulas by paralysis of the vagus nerve induced by atropine. This has been attempted previously with variable results.2 In the light of our experiences, this variability appears related only to the degree of completeness of the atropine effect attained.
The subjects were soldiers who had sustained wounds which led to the development of arteriovenous fistulas. A solution of atropine sulfate in distilled water was administered intravenously. The doses varied from 1.7 to 1.95 mg. of the drug. In an earlier case, the administration of