rebreathed before. The direction of the air current through the apparatus is shown by the position of the arrows in the figure. By changing the position of the aspirating flasks, and turning the stopcocks in the glass tubing inserted through the stoppers closing the upper openings of the aspirators, the current was maintained in the same direction as before, and the entrance of external air was thereby prevented.

The results obtained show that, with the absorption of the CO, as generated, the mouse remained relatively comfortable in the atmosphere present and that no deleterious effects developed from the continued rebreathing of the air confined within the apparatus. The animal seemed to be somewhat oppressed toward the close of each experiment, but revived quickly after removal from the apparatus.

The air contained in the two aspirating flasks was retained each time in the later experiments. Consequently in these experiments the fresh air-supply comprised only that which was enclosed in the Pettenkofer tubes, the rubber connecting tubes, and in the bell-jar containing the animal. several of the later experiments the volume of air within the apparatus was aspirated continuously. through all its parts for some hours before beginning the experiment. In this manner the pure air-supply was reduced to one litre, the amount of air in the bell-jar containing the animal.

V. Experiments to determine the proportions of CO, and of O in the air of a glass vessel in which small animals (mice and birds) had remained until death was produced, and the effects of different temperatures upon the duration of life and on the composition of the residual atmosphere after death in such cases.

The results obtained in these experiments are shown in Table I. At the room temperature death did not take place until the amount of oxygen present was too low to support life. At a higher or lower temperature there was a slightly shorter duration of life, varying with the amount of increase or reduction of the temperature.

TABLE I.

EXPERIMENTS WITH ANIMALS IN CLOSED VESSELS-ATMOSPHERIC AIR.

The effects of temperature upon the duration of life in a confined space (and even in the open air) are better shown in the repetition of Richardson's experiments (8), as presented in Table J. The results obtained in these experiments show that the duration of life is very perceptibly shortened through the influence of a higher as well as of a lower temperature than 18° to 20°C.

An interesting condition noted in autopsies upon a number of the animals that succumbed to the conditions in the "Richardson" experiment was that of the blood in the heart of the animal. In the cases where death supervened in a short time, the heart blood was fluid and seemed to lack the power of coagulation, while in those cases in which death resulted after several hours' confinement in the flask, the cavities of the heart contained firm, dark clots of blood. This condition of the blood was, no doubt, due to the influence of the CO, generated by the animal during the experiment.

CHART I.-SHOWING RELATIVE PROPORTIONS OF CO, AND OF O, AND THE RELATIVE DURATION OF LIFE IN THE EXPERIMENTS IN CLOSED VESSELS.

Chart I. shows the relative duration of life, the relative proportions of CO2 and of O at death of the animal, in the experiments with animals in closed vessels containing atmospheric air.

CHART II.-SHOWING RELATIVE DURATION OF LIFE, PROPORTIONS OF N AND O AT BEGINNING OF EXPERIMENTS, WITH THE TEMPERATURE OF THE ATMOSPHERES IN THE "RICHARDSON

EXPERIMENTS.

Chart II. shows the relative duration of life, the relative proportions of N and of O at the beginning of each of the “Richardson" experiments, also the temperature curve for the entire series.

In order to ascertain whether an atmosphere which had served for respiration, once or oftener, affected an animal differently from an atmosphere made up artificially from pure gases to the same proportions as found in the analysis of the atmospheres in the different experiments reported on, a series of experiments was undertaken to determine the effects of gaseous mixtures made up of varying proportions of CO2, O, and of N. The results obtained in this series of experiments are shown in Table K, giving the capacity of the jar, the weight of the animal, the composition of the atmosphere before and after the experiment, and the duration of life in such an atmosphere. The construction of artificial atmospheres, and the introduction of an animal into such an atmosphere without considerable alteration of the proportions of the different gases, through the accidental introduction of atmospheric air, was not always found an easy matter. The chief difficulty was unfortunately a fundamental one, in that the CO2 was not entirely free from atmospheric air; the oxygen contained more than 10 per cent. of N; while the attempt to obtain pure N from atmospheric air by means usually employed for this purpose-burning out the O with phosphorus-gave variable results with each attempt, the proportion of O remaining after the absorption of the P2O, usually ranged from 2 to 5 per cent. Under these circumstances it will be seen that there was an almost insurmountable difficulty to the construction of an atmosphere having the exact proportions of the different gases predetermined for it, and abundant evidence of this difficulty was obtained from analyses of the mixtures after sufficient time had been allowed, as was supposed, for the diffusion of the gases.

2 5

2

The thorough diffusion of the components of gaseous mixtures appears to be a slow process. Twenty four hours, or longer, was usually allowed for this to take place, yet from the variable lengths of time during which animals of the same size and apparently possessing the same amount of vitality could survive in atmospheres of equal volume made up from the same mixture, and the variable proportions of the different gases found on analysis after death of the animals exposed to these atmospheres, show that perfect diffusion had not always taken place. These discrepancies in the construction of the gaseous mixtures are to be regretted, though they are not great enough to vitiate the value of the experiments taken as a whole. The positive character of the results is too evident to allow these difficulties to have much weight.

There is an uncertain feature in the determinations of the proportions of CO, in the gaseous mixtures, after death of the animal, in those instances where this gas was originally present in high. percentages. On this account it would be well to bear in mind that the third column representing the proportions of the different gases present at death, marked N, represents, in fact, the gases which failed to be absorbed in the gas-burette by the solutions of caustic soda and of pyrogallic acid used to absorb the CO2 and O present. There is no doubt as to the presence of the proportions of CO2, as stated in the different experiments, before placing the animal in the mixture. Whether a large proportion of the CO, was likewise absorbed by the animal, it is impossible to say. There is no probability that such was the case. A part of the loss of CO2 may also be accounted for in the method employed in making the gaseous mixtures. These mixtures were made by displacing water from the jars which were to contain them. The water may have taken up the CO2 more readily than the other gases, especially where this was the first gas introduced into the jar, and may, therefore, have been a slight source of variation in the composition of the mixture; yet, it seems, from analysis made just before placing the animal in the mixture, that the loss in this manner was very small. The desired proportion of CO2 was usually present, even after twenty-four hours had been allowed for diffusion to take place.

2

Chart III. shows the results obtained in these experiments as to the relative duration of life and the relative proportions of CO, and of O at the beginning of the experiment, as well as at death of the animal. In comparing this chart with Chart I., it must be remembered that in this series of experiments the composition of the atmosphere was a different and variable one, while in the series of experiments shown in Chart I., the composition of the atmosphere at the beginning of the experiment was invariably the same-i.e., atmospheric air. This fact, along with the variations in size of jar for different animals, explains the longer or shorter duration of life in this series of experiments as compared with that presented in Chart I. The very important influences of these