Under this head I shall decide upon the method of drawing blood by means of cups and leeches, in the inflammatory state of fever. Where an inflammatory fever arises from local affection, or from contusion in the head or breast, or from a morbid excitement in those, above other parts of the arterial system, they may be useful; but where local affection is a symptom of general and equable fever only, it can seldom be necessary, except where bleeding from the arm has been omitted, or used too sparingly, in the beginning of a fever; by which means such fixed congestion often takes place, as will not yield to general bleeding. XVI. Much has been said likewise about the proper time for bleeding in fevers. It may be used at all times, when indicated by the pulse and other circumstances, in continual fevers; but it should be used chiefly in the paroxysms of such as intermit. I have conceived this practice to be of so much consequence, that, when I expect a return of the fever in the night, I request one of my pupils to sit up with my patients all night, in order to meet the paroxysm, if necessary, with the lancet. But I derive another advantage from fixing a centinel over a patient in a malignant fever. When a paroxysm goes off in the night, it often leaves the system in a state of such extreme debility, as to endanger life. In this case, from five to ten drops of laudanum, exhibited by a person who is a judge of the pulse, obviate this alarming debility, and often induce easy and refreshing sleep. By treating the human body like a corded instrument, in thus occasionally relaxing or bracing the system, according to the excess or deficiency of stimulus, in those hours in which death most frequently occurs, I think I have been the means of saving several valuable lives. XVII. The different positions of the body influence the greater or less degrees of relief which are obtained by blood-letting. Where there is a great disposition to syncope, and where it is attended with alarming and distressing circumstances, blood should be drawn in a recumbent posture, but where there is no apprehension or dread of fainting, it may be taken in a sitting posture. The relief will be more certain if the patient be able to stand while he is bled. A small quantity of blood, drawn in this posture, brings on fainting, and the good effects which are often derived from it. It should therefore be preferred, where patients object to copious or frequent bleedings. The history of the success of this practice in the British army, recently mentioned from Dr. Sydenham, furnishes a strong argument in its favour.