artery can be drawn from the pulse or the pressure within it. SUMMARY I. Direct measurement of arterial flow is desirable. At present, electromagnetic flowmeters appear to be the best instruments for this, but are still beset with difficulties.2. Contrary to the usual assumption, pulse, pressure, and temperature are not directly related to arterial flow.3. 'Teflon' prostheses and vein grafts of a narrowing shape were inserted in dogs to by-pass arterial occlusions. Whether the blood entered at the wide or the narrow end did not make any appreciable difference to the resulting flows and pressures.4. The presence of a pulse beyond an arterial obstruction may have an apparently paradoxical relation to the distal flow. It has been confirmed by experiment in dogs that such a pulse may disappear as flow improves and distal pressure rises, and may reappear as flow decreases and distal pressure falls as the result of distal obstruction.