3. _Marshall Hall's._--This consists in placing the patient in the prone position, with a folded coat under the chest, and rolling the body alternately into the lateral and prone positions. 4. _Howard's._--This consists in emptying the thorax by forcibly compressing the lower part of the chest; on relaxing the pressure the chest again fills with air. Here the patient is placed in the supine position. The objections to the supine position are that the tongue falls back, and not only blocks the entrance of air, but prevents the escape of water, mucus, and froth from the air-passages. 5. _Laborde's Method._--This consists in holding the tongue by means of a handkerchief, and rhythmically drawing it out fully at the rate of fifteen times per minute. This excites the respiratory centre, and this method may be employed along with any of the other methods. XX.--DEATH FROM STARVATION The post-mortem appearances in death from starvation are as follows: There is marked general emaciation; the skin is dry, shrivelled, and covered with a brown, bad-smelling excretion; the muscles soft, atrophied, and free from fat; the liver is small, but the gall-bladder is distended with bile. The heart, lungs, and internal organs are shrivelled and bloodless. The stomach is sometimes quite healthy; in other cases it may be collapsed, empty, and ulcerated. The intestines are also contracted, empty, and translucent. In the absence of any disease productive of extreme emaciation (_e.g._, tuberculosis, stricture of oesophagus, diabetes, Addison's disease), such a state of body will furnish a strong presumption of death by starvation. In the case of children there is not always absolute deprivation of food, but what is supplied is insufficient in quantity or of improper quality. The defence commonly set up is that the child died either of marasmus or of tuberculosis. In cases where it is alleged that a child has been starved and ill-used, one must examine the body for signs of neglect--_e.g._, dirtiness of skin and hair, presence of vermin, bruises or skin eruptions. Compare its weight with a normal child of the same age and sex. If the disproportion be great and signs of neglect present, then the probability is great (provided there be no actual disease present) that the child has been starved. XXI.--DEATH FROM LIGHTNING AND ELECTRICITY The signs of death from lightning vary greatly. In some cases there are no signs; in others the body may be most curiously marked. Wounds of various characters--contused, lacerated, and punctured--may be produced. There may be burns, vesications, and ecchymoses; arborescent markings are not uncommon. The hair may be singed or burnt and the clothing damaged. Rigor mortis is very rapid in its onset and transient. Post mortem there are no characteristic signs, but the blood may be dark in colour and fluid. The presence or absence of a storm may assist the diagnosis.