CONSUMPTION OF OXYGEN. 137 



of Regnault and Reiset, animals were placed in a receiver filled with air, a 

 measured quantity of oxygen was introduced as fast as it was consumed by 

 respiration, and the carbon dioxide was constantly removed and carefully esti- 

 mated. In most of the experiments, the confinement did not appear to inter- 

 fere with the functions of the animal, which ate and drank in the apparatus 

 and was in as good condition at the termination as at the beginning of the 

 observation. This method is much more accurate than that of simply causing 

 an animal to breathe in a confined space, when the consumption of oxygen 

 and accumulation of carbon dioxide and other matters must interfere more or 

 less with the proper performance of the respiratory function. As employed 

 by Regnault and Reiset, it is adapted only to experiments on animals of small 

 size. These give but an approximate idea, however, of the processes as they 

 take place in the human subject. Pettenkofer constructed a chamber large 

 enough to admit a man and allow perfect freedom of motion, eating, sleep- 

 ing etc., into which air could be constantly introduced in definite quantity, 

 and from which the products of respiration were constantly removed and 

 estimated. This method had been adapted to the human subject on a small 

 scale in 1843, by Scharling, but there was no arrangement for estimating the 

 quantity of oxygen consumed. 



Estimates of the absolute quantities of oxygen consumed or of carbon 

 dioxide exhaled, based on analyses of the inspired and expired air, calcula- 

 tions from the average quantity of air changed with each respiratory act, and 

 the average number of respirations per minute, are by no means so reliable as 

 analyses showing the actual changes in the air, like those of Regnault and 

 Reiset, provided the physiological conditions be fulfilled, Where there is so 

 much multiplication and calculation, a very slight inaccuracy in the estimates 

 of the quantities consumed or produced in a single respiration will make a 

 large error in the estimate for a day or even for an hour. Bearing in -mind 

 all these sources of error, from the experiments of Valentin and Brunner, Du- 

 mas, Regnault and Reiset and others, a sufficiently accurate approximate esti- 

 mate of the proportion of oxygen consumed by the human subject may be 

 made. The air, which contains, when inspired, 20'81 parts of oxygen per 

 100, is found on expiration to contain but about 16 parts per 100. In other 

 words, the volume of oxygen absorbed in the lungs is five per cent, or -fo of 

 the volume of air inspired. It is useful to extend this estimate as far as pos- 

 sible to the quantity of oxygen absorbed in a definite time ; for the regulation 

 of the supply of oxygen where many persons are assembled, as in public build- 

 ings, hospitals etc., is a question of great practical importance. Assuming 

 that the average respirations per minute are eighteen, and that with each act, 

 twenty cubic inches (327'7 c. c.) of air are changed, fifteen cubic feet (424-8 

 litres) of oxygen are consumed in the twenty-four hours, which represent 

 three hundred cubic feet (8*5 cubic metres) of pure air. This is the mini- 

 mum quantity of air which is actually used, making no allowance for any in- 

 crease in the activity of the respiratory processes, which is liable to occur 

 from various causes. To meet all the respiratory exigencies of the system, in 

 hospitals, prisons etc., it has been found necessary to allow at least eight 



