Although 'fresh' cadavers have been used quite frequently to study techniques of artificial respiration and other mechanical problems that are difficult to investigate in the living subject, little information is available on the rate and magnitude of changes in the mechanics of chest wall and lung with death. Carson (1820) and a number of subsequent investigators attempted to measure the elastic recoil of the lungs after death, but conditions of inflation were not controlled sufficiently closely, and conflicting results were obtained. McIlroy (1952) concluded that the distensibility of the lungs varied in a complex, unpredictable manner after death, and the only feature that showed any consistency was the pressure generated in the trachea on opening the thorax.The present report discusses findings in the rabbit and the cat before and up to 18 hr after death. Both open-and closed-chest preparations have been used, and the conditions under which stable pressure-volume relationships may be observed have been defined. Immediate mortal changes are shown to be mainly an exaggeration of mechanical characteristics already present in the living chest. METHODS Material and preparation. The main series of observations was carried out on eleven female rabbits of 1-75-2-25 kg weight; similar experiments were also carried out on an adult cat (female, 2*2 kg) to ensure that the findings were not peculiar to one species. All animals were anaesthetized by intraperitoneal pentobarbitone sodium B.P. (solution 45 mg/ml. in ethanol 10% (w/v)). An initial injection of 3 ml. caused loss of consciousness with minimal disturbance, and a further 1-3 ml. given 20 min later produced deep anaesthesia with rather shallow breathing. A tracheotomy was performed, and the animal was connected to a Palmer pump. Moderate overventilation (stroke 18-25 ml., rate 36/min) then produced apnoea of sufficient duration to permit study of pressure-volume (P-V) relationships. This technique was preferred to the use of curare, which is known to decrease the distensibility of the lungs (Fenn, Otis & Suskind, 1954;Massion, 1957).After 'static' and dynamic curves had been plotted, an overdose of pentobarbitone sodium (10 ml.) was given, and 5 min later the respiratory pump was disconnected. This led to a very quiet death from respiratory failure. Post-mortem gross and microscopic examination showed non.