From recent work it would appear that neither the heart nor the vaso-motor center is the chief factor responsible for the circulatory changes in diphtheria toxin poisoning. The work we have to report is the result of an attempt made to determine what this factor is. We have studied first the reactions of the larger blood vessels in poisoned animals. This was done in the following manner: At the height of the poisoning, at a time when the blood pressure was extremely low, the animal was killed, sections of various arteries removed and placed in cold Ringer-Locke solution, where they were kept until ready for use. For comparison sections of arteries from an unpoisoned animal were removed and preserved. A strip of artery from the poisoned animal and one of similar size from the corresponding artery of the normal animal were then prepared, placed in a vessel containing fresh oxygenated Locke's solution, and attached to an apparatus for recording contraction and relaxation. To the solution was then added either adrenalin I : 1,000,000; barium chlorid I gm. : 255 c.c., or amyl nitrite (.I to .3 c.c. in 250 c.c.) and the contraction or relaxation of the artery strips recorded. Allowing for individual variations, the results we obtained from a considerable number of experiments showed that there was no appreciable difference in the reaction of the arteries of the poisoned and unpoisoned animals. Both series gave the same contraction from adrenalin and barium and corresponding relaxation from the nitrite. We conclude therefore that neither the vaso-motor mechanism nor the muscle tissue in the larger arteries are affected to an appreciable extent by diphtheria toxin. In poisoned animals gastro-intestinal symptoms are prominent. These consist of vomiting and severe diarrhoea.