By accurately closing the nose, and bringing the lips tight on the mouth-piece, after a few trials I was able to breathe oxygene or common air in this machine for two minutes or two minutes and half, without any other uneasy feeling than that produced by the inclination of the neck and chest towards the cylinder. The power of uniformly exhausting the lungs and fauces to the same extent, I did not acquire till after many experiments. At last, by preserving exactly the same posture after exhaustion of the lungs before the inspiration of the gas to be experimented upon, and during its compleat expiration, I found that I could always retain nearly the same quantity of gas in the bronchial vessels and fauces; the difference in the volume expired at different times, never amounting to a cubic inch and half. By connecting the conducting pipe of the mercurial airholder, during the respiration of the gas, with a small trough of mercury by means of a curved tube, it became a perfect and excellent breathing machine. For by exerting a certain pressure on the airholding cylinder, it was easy to throw a quantity of gas after every inspiration or expiration, into tubes filled with mercury standing in the trough. In these tubes it could be accurately analised, and thus the changes taking place at different periods of the process ascertained. Whenever I breathed pure nitrous oxide in the mercurial airholder, after a compleat voluntary exhaustion of my lungs, the pleasurable delirium was very rapidly produced, and being obliged to stoop on the cylinder, the determination of blood to my head from the increased arterial action in less than a minute became so great, as often to deprive me of voluntary power over the muscles of the mouth. Hence, I could never rely on the accuracy of any experiment, in which the gas had been respired for more than three quarters of a minute. I was able to respire the gas with great accuracy for more than half a minute; it at first, rather increasing than diminishing the power of volition; but even in this short time, very strong sensations were always produced, with sense of fulness about the head, somewhat alarming; a feeling which hardly ever occurs to me when the gas is breathed in the natural posture. In all the numerous experiments that I made on the respiration of nitrous oxide in this way, a very considerable diminution of gas always took place; and the diminution was generally apparently greater to the eye during the first four or five inspirations.