The various hypotheses of the mechanism of coma resulting from cerebral trauma are discussed. Experimental evidence shows that there are two kinds of transient abolition of cerebral function by trauma-acceleration concussion, and compression concussion. The former is a passing paralysis which occurs in all brain-stem mechanisms examined and is brought about at and beyond a threshold value of change in velocity. A lesser degree of change causes momentary depression of function, a greater degree prolongs the phase of paralysis before recovery occurs. Movement of the head is necessary for the usual kind of concussion to occur, as also for contre-coup injury. Compression concussion has much more selective incidence on the respiratory centre, and appears to require an extreme crushing injury or penetration of the skull by a relatively large object. Acceleration concussion is that accompanying closed head injury and is not accompanied by any significant change in C.S.F. pressure. The cerebral blood flow is verv greatly increased in this variety owing to stimulation of the vagoglossopharyngeal nerves at the foramen magnum. With more severe blows subpial or intramedullary lesions may occur in this situation, indicating that distortion at the foramen magnum occurs. No evidence of vascular spasm or paralysis is found. An immediate brief rise of blood-pressure is due to stimulation of the vasomotor centre. Vagal effects may not appear until the traumatic paralysis of all centres begins to pass off. A delayed fall in blood-pressure lasting many minutes may follow severe vagal effects, and appears comparable to acute surgical shock as produced by intense stimulation of any other visceral nerve. Death occurs from failure of blood-pressure, an intensification of shock with shallow respiration and intense constriction of viscera.It is concluded that sudden failure of what has been called the veno-pressor system is important in death shortly following experimental concussion. No macroscopical lesions are found, and histological examination shows no change with ordinary tissue stains. The unconsciousness of coma is believed to be related to direct traumatic paralysis of the cortical neurones and if their sensitivity to physical violence is similar to that of anaesthetics, prolonged impairment of function from this cause alone is possible. Clinical observations were related in support of these conclusions.