Although in clinical neurology peripheral nerve palsies induced by a tourniquet are uncommon, such palsies due to direct pressure of other kinds are frequently met with. Despite this, attempts to define the factors involved in the production of such lesions have been remarkably few. Lewis, Pickering and Rothschild1 studied the paralysis produced by the application of pressure to human limbs both by the sphygmomanometer cuff and by localized pressure on single nerves. They elicited facts of great interest. The paralysis produced by a cuff had centripetal onset and affected touch before pain and pain before motion. The latency of onset with a cuff around the upper part of the arm was almost constant (paralysis at about the twenty-fifth minute) and was the same with pressures of from 150 to 300 mm. The cuff was without effect at pressures below the systolic blood pressure. When paralysis had commenced, the placement