yield energy in the form of heat render it of peculiar interest in this respect and suggest the value of accurate information of its effects on the economy.

The study of the action of any agent on thermal processes involves the consideration of its effects on the quantity of heat produced, the quantity dissipated, and on the bodily temperature. Investigators have been led to determine heat production by indirect or direct calorimetry. Indirectly, it is estimated through a knowledge of the quantities and chemical composition of the substances entering the body as food, and of the effete principles resulting from their oxidation, and from this data computing the amount of heat evolved; or, by regarding temperature changes in the organism as indicating similar alterations in thermogenesis. Directly, the heat formed is measured by means of an instrument specially adapted for direct determination, and hence known as the calorimeter. Practically, all the work done with alcohol has been accomplished by the indirect method, and is accordingly open to many sources of error and is inconclusive. For while there can be no doubt that CO,, urea, phosphates, and other effete matters arise through destructive tissue metamorphosis, by which heat is formed as the chief result, just what importance any or all of these products possess in indicating the degree, or even the direction, of the activity of the processes concerned in heat production is far from being determined. Desplatz's experiments on rats, guinea-pigs, and small birds, and Senator's on dogs, in which the heat formed was measured by a calorimeter, and the CO, by established methods, would lead us to believe that there exists no direct relation between the quantity of heat formed and of CO, eliminated; while in connection with urea, which was once looked upon as almost an absolute index of tissue waste, recent experiments show quite clearly that the quantity eliminated is indicative of the amount of nitrogenous food and not of retrograde metamorphosis. Finally, that there is no definite relation between bodily temperature and heat production is clearly evident whenever the two are measured simultaneously for consecutive hours. Until, however, the precise values which these effete matters possess in indicating the nature of the changes in thermogenesis are definitely established, their usefulness in suggesting the probable direction of thermal alterations may be accepted with caution and reserve.

The present experimental opinions of the actions of alcohol on the heat functions being founded on the CO,, and urea and other salts eliminated, and upon alterations in bodily temperature, we may consider these for convenience and clearness seriatim.

Since the chemical processes in the body are essentially those of oxidation, and as the heat evolved is almost wholly due directly or indirectly to such changes, the measure of the amount of oxygen consumed is equally, if not, indeed, more important than that of CO, given off. Under normal conditions more oxygen is consumed than represented in the CO, eliminated, the relation between them CO = .906, being the "respiratory quotient." The excess of oxygen goes to form other oxidative products, and as the respiratory quotient is subject under certain conditions to marked changes, it is clear that the quantity of CO, formed may be far from indicating the quantity of oxygen consumed, and, as a consequence, if considered alone, may give rise to obvious sources of error. The older observers did not realize the importance of measuring the oxygen in connection with their CO, estimations, consequently Prout, Vierordt, Böcker, Bernard, Hammond, Smith, Berg, Davis, and others studied only the CO, eliminated, and even these calculations are generally, because of various other defects in the method of experiment, of but little intrinsic value. Of the recent investigations, some have taken account only of the O consumed, others of both the O consumed and CO, eliminated. Thus, Rumpf asserts, without sufficient grounds, however, that in guineapigs in alcohol narcosis the quantity of oxygen consumed is diminished. Henrijean, in experiments on himself, noted an increase. Boeck and Bauer, in a few and unsatisfactory experiments on dogs and cats, found that small doses diminished both the consumption of O and elimination of CO,, while larger doses, not sufficient to cause narcotism, notably increased both quantities. Experiments on rabbits by Wolfers show that when animals are placed in a bath of constant temperature, O and CO, were increased and the respiratory quotient generally lowered. Bodländer, in two dogs, found the consumption of O reduced 11.72 and 19.10 per cent. respectively, and the elimination of CO, reduced 10.78 and 19.16 per cent. In a rabbit the alterations were respectively 3.13 and 7.74 per cent., the respiratory quotient not being especially affected. Zuntz and

2

Berdez, in observations on man, using doses of 20 to 30 c.c., and maintaining complete muscular quiet, noted an increase in both O and CO, of 3.5 per cent. Other experiments made by Geppert led him to assert that alcohol, in moderate quantities, had no decided effect on the amount of oxygen consumed, while the quantity of CO, eliminated was unaffected or slightly diminished. It will be observed that the evidence regarding the consumption of O and elimination of CO, is conflicting. The most carefully performed experiments indicate a diminution of both, but this cannot be considered definitely settled.

The results of observations in connection with the elimination of urea and certain other effete matters found in the urine and fæces are, on the other hand, quite uniform. Thus, Böcker, Fokker, Hammond, Parkes and Wollowicz, Munk, Riess, and Desplatz all affirm that the quantity of urea is diminished, although most of the work is all but valueless, owing to the neglect of the determination of the precise amount of nitrogenous matter entering the body as food, as well as the quantity of water drunk, or because of the periods of experiments being too brief, or for other reasons.

Studies of the action of alcohol on bodily temperature, of both man and animals, under normal and abnormal conditions, are numerous and the results contradictory. The feeling of warmth in the stomach, and afterwards in the skin, after the ingestion of this agent, naturally suggest that it increases temperature, and while we are not without evidence to show that such, under certain circumstances, may be the case, there can be no question, however, that the characteristic action is to produce a distinct diminution. Differences in the results are largely due to variations in the quantity of alcohol taken and upon the attendant circumstances of method or otherwise. But few observers have noted a distinct rise. Lallemand, Perrin, and Duroy state that large doses always cause a primary increase, followed by a diminution. Lichtenfels and Fröhlich, in two experiments on man, one with four ounces of eight per cent. alcohol and the other with two ounces of seventeen per cent.,-found a primary increase in the former of .5° and in the latter of .32°. In the first, the temperature returned to the normal in forty-five minutes, and in the latter in fifty minutes. Riegel states, as a result of a careful series of experiments, that only exceptionally is an ele

vation of temperature observed. Riegel's results I have confirmed by twenty-two experiments on dogs, each covering an average of several hours, with doses varying from 1.25 c.c. to 5 c.c. of alcohol per kilo of body weight. In only five did the tempera. ture at any time go above the normal, then only from .1° to .5° C., commonly but .2° or .3°, and generally immediately after the dose. The action of alcohol on temperature in the opposite direction is far better marked and very constant. Tscheschichin, in experiments on rabbits, always noticed a fall. Lallemand and others, in dogs, found a fall following the primary rise. Bouvier, states that in experiments on different animals, small doses diminish temperature not attended by any increase; larger doses diminish to a greater degree; in fever, a similar diminution occurs. Binz, Daub, Parkes and Wollowicz, Ringer and Rickards, Riegel, and others have all noted a diminution more or less marked or even profound; occasionally, however, as in certain of Parkes and Wollowicz's observations, no appreciable change occurred even after doses of from 10 to 20 ounces of claret in man, or from 1 to 6 ounces of brandy.

The careful studies of Riegel, embracing eighty-six experiments on man, lead to conclusions which doubtless closely represent the peculiar actions of alcohol on animal temperature. In his summary he states, among other conclusions, that-1, alcohol, even in moderate doses, in many cases causes a lowering of the temperature, the amount of diminution averaging, as a rule, a few tenths of a degree C.; 2, only exceptionally is an elevation of temperature detected, and not unfrequently after small doses there is no noticeable change; 3, the frequent repetition of doses of alcohol diminishes the lowering effect; 4, the amount of diminution is directly proportional to the dose; 5, the depression for the most part is of short duration. My own experiments on dogs agree in their results with the conclusions of Riegel. The intensity of the action, however, is far from always being in accord with the dose, a given dose in one case producing a profound effect, and in another little or almost none. Illustrations of this may be seen in records of experiments in this paper.

Briefly, the evidence, then, seems to indicate that the quantity of oxygen consumed and CO, and other effete matters given off are diminished and the bodily temperature lowered. All these changes being in harmo

nious relation strongly suggest that alcohol diminishes the normal activity of tissue changes. This finds support in the experiments of Nencki and Sieber, and Simanowsky and Shoumoff, in their studies on the influence of this drug on the oxidation of certain substances injected into the organism when they found the change diminished.

The fact that CO, is not increased is one of great practical economic importance, for it must be evident that either this gas does not constitute one of the chief products resulting from the decomposition of alcohol, or else general tissue metamorphosis is so diminished that the quantity of CO, eliminated largely represents an alcohol derivative.

That this agent rapidly undergoes chemical change after absorption, that but a small percentage is eliminated as such, and that we do not find either in the system or excreta any immediate oxidation products, is clearly established. What becomes of it and in what form eliminated are unsettled problems. The experiments of Schulinus and Salzynski show that living blood causes about ten per cent. to disappear immediately. The work of Anstie also demonstrates that this substance undergoes rapid alteration. Thus, to

small

dog was given an ounce of brandy daily for eleven successive days. On the last day, after receiving the dose, he was killed, quickly chopped up, and the minced tissues extracted for alcohol. But only one-fourth of the drug represented in the last dose of brandy was recovered, while the total quantity eliminated the day before was only 1.13 grains. Dupré noted that but a small percentage of the alcohol taken was excreted as such, this being supported by the earlier and less valuable investigations of Lallemand, Perrin, and Duroy. Parkes and Wollowicz found that practically none was eliminated after twenty-four hours, and Anstie observed that after the ingestion of three or four ounces of alcohol only one or two grains could be found in the urine. Heubach, in experiments on individuals with fever, to which were given doses of from 18 c.c. to 325 c.c., determined, as a result of twenty-two experiments, an elimination in the urine of 1.12 per cent.

That the skin and lungs are not more important than the kidneys in this elimination is evident from the recent work of Bodländer on himself and a dog. On himself, after taking from 50 c.c. to 100 c.c. of absolute alcohol, diluted with from twenty to fifty per cent. of water, the quantity recovered in the urine varied from .17 to 1.86 per cent., or an

average of 1.18 per cent. The maximum was recovered during the first hour, a small quantity during the second, and the merest traces during the third. In the dog, after doses of from 20 c.c. to 30 c.c. of absolute alcohol, 1.57 per cent. was recovered in the urine. Through the skin he recovered from himself .14 per cent., and through his lungs 1.6 per cent.; through the dog's lungs 1.95 per cent., but none from the fæces and intestines. The total amount recovered from himself was 2.92 per cent., and from the dog 3.52 per cent. Larger quantities have been recovered by other investigators, as, for instance, by Subbotin in experiments on rabbits in which excessive doses were used, the quantity in such observations amounting to about sixteen per cent. It seems very clear from this work that alcohol, in moderate quantity, rapidly undergoes alteration in the system and that but a small percentage is eliminated as such.

While we do not know the precise nature of this change, there is no reason to believe that a more complex molecule is formed; but, on the other hand, the indication is of a breaking up of the molecule into simpler bodies, such as CO, and H,O. As it is obviously impossible to satisfactorily detect in the excretions such small quantities of H,O as would thus be formed, we are dependent alone on estimations of CO,. Here we meet with the formidable indication that the quantity of CO, eliminated after alcohol is actually less than normal. This, however, is explicable in the possibility of the drug acting directly upon the heat processes and lessening normal tissue metamorphosis, thus diminishing the amount of the gas arising in this way, and supplying a portion of the deficiency from its own decomposition. That it does diminish destructive tissue change seems supported by the lessening of the amount of nitrogenous and other effete matters, and also by the diminution of the animal's temperature. But this can only satisfactorily be determined by direct measurement of the quantity of heat produced and dissipated.

Alcohol, as is well known, yields an enormous amount of energy in the form of heat during its complete oxidation, one gramme of absolute alcohol setting free nine kilogramme degrees. Should the alcohol thus be oxidized in the body, the heat evolved from a quantity, such as taken by Bodländer (100 c.c.), would equal about seven hundred and fifteen kilogramme degrees, being equivalent to about one-fourth the total approximate heat production of an average man per

diem. The amount of CO, formed would be equal to about one hundred and fifty-two grammes, or one-sixth the mean quantity eliminated per diem. That this agent actually does yield energy, or else conserve the tissues, or both, is also evident in its power of sustaining the system in disease, in the increase of body weight often observed in drunkards even on restricted diet, and the known ability of the economy to maintain an equilibrium of metabolism on limited food. Should alcohol thus yield energy it is obvious that the total output of heat must be increased or less tissue change occur to compensate for it. Calorimetrical experiments alone can determine this, and the only observations of the kind recorded are those of Bevan Lewis (Journal of Mental Science, 1880-81, p. 20), Wood and Reichert (Journal of Physiology, iii. p. 321), and Desplatz (Maly's Jahresberichte ü. d. Fortschritte der Thierchemie, 1886, s. 365). The experiments of Bevan Lewis were performed on rabbits with a calorimeter such as described by Burdon-Sanderson but somewhat modified. The observations were made during quarter hour intervals and commonly for six consecutive periods. From a number of experiments, only a few of which are given, he reaches the following conclusions: 1. A primary check to heat formation is most marked and protracted when small doses of alcohol are given.

2. A pronounced fall of body temperature is most marked during the first quarter hour and coincident with the primary check to thermogenesis.

5. The increase of heat production is gradually augmented from time to time until the heat climax is reached, a period usually coincident with the registry of the lowest bodily temperature.

6. The heat climax is more protracted or postponed and is greater in degree with larger doses of alcohol.

7. The greatest loss of heat units from temperature (diminution of temperature) occurs during the first interval, subsequent intervals being marked by a still progressive loss, which, however, becomes less towards the period of heat climax, when a restitution to the normal of temperature begins.

8. With small doses of alcohol this restitution of bodily temperature is usually sudden or comparatively rapid in operation; after larger doses the return to normal temperature is spread out over a longer period, being extremely tardy when very large doses are administered.

9. The characteristic action of alcohol is that of greatly increasing the heat production, while a dispersion of fresh formed heat is facilitated by peripheral vaso-motor paresis, and that only in very small doses we get a temporary lowering of the heat functions.

In the experiments of Wood and myself on dogs five observations were made with different doses, and with results in accord generally with Bevan Lewis's, inasmuch as on the whole heat production and dissipation were both increased. Our experiments were conducted differently, the normal production

4. The increased heat production is manifested over a more prolonged period after

larger doses.

of from one to two and a half hours, then the alcohol given, and a second observation made for a similar length of time. These results are here tabulated :

Desplatz made four half-hourly experiments on rats, and in each found a diminution in heat dissipation and of CO, and nitrogen eliminated.

In Lewis's experiments a check to heat formation was detected during the first quarter hour; following this the increase was decided. Thus, after moderate doses, in an animal in which the normal production per kilo was.95 gramme-units, during the first interval after alcohol it was .46 gramme-units, during the second 2.7, the third 1.3, and the fourth 1.5 units, the increase after the first interval being nearly threefold. The mean heat production following alcohol was more than normal,viz., equal to 1.49 units. Our own experiments cover hours; hence we obtain mean figures for the periods during which Lewis made observations for short intervals. Therefore, if a primary check commonly occur, it is lost in the relatively larger heat production following. As a consequence, the mean in our experiments always shows an increase. Lewis always noted a diminution of temperature, dependent, of course, upon the excess of dissipation over production. In our experiments similar results were obtained in two, while in the other three production was in excess, with a consequent rise of bodily temperature. The doses varied within wide limits, but no relation is evident between these differences and the varying results. In Lewis's experiments the temperature curve diminished from the first, showing that heat dissipation was always in excess of heat production. Curiously, the greatest depression of temperature occurred coincidently with the greatest heat production. The results of the experiments of Bevan Lewis, and Wood and myself on heat production are contradicted by those of Desplatz and the evidence indicating a diminution in the amount of tissue metamorphosis.

Whether a drug affects primarily the process of dissipation or that of production must perhaps always remain a matter of speculation. It seems fair to infer that when heat production is changed more than dissipation, in either direction, whether increased or decreased to a greater extent, the former would be the part of the mechanism primarily affected, and that the alteration occurring in the latter would be in the way of compensation to prevent undue increase or diminution of the animal's temperature. With alcohol Lewis always found an excess of dissipation, while in only two of our five experiments was this noted. The indication here would be

that the function of dissipation is the one primarily affected, and that the increase of production was to make up for this excessive loss. This is certainly not without support in the belief that this agent causes peripheral vaso-motor dilatation, and thus facilitates the loss of heat. The increase of thermogenesis, on the other hand, is certainly not in accord with what we would be led to expect from studies made in the elimination of CO2, urea, etc. Apart from this inconsistency, there are ample reasons why, in the premises, both the experiments of Lewis, Wood and myself, and Desplatz are of questionable value. Our own experiments are certainly too few and inconclusive, while those of Lewis lose much of their weight in the fact that the accurate working of his instrument is open to suspicion, for during some intervals absolutely no heat was shown by the calorimeter, although, as is obvious, the animal must have radiated an appreciable quantity. In Desplatz's experiments the heat dissipation only was recorded.

Thus the absence of accurate and satisfactory knowledge of the actions of alcohol on the animal heat functions, of the successive phases of its influence during consecutive hours, together with the value of such data to the therapeutist and dietician, prompted the present inquiry. These experiments were performed by the aid of a new and improved calorimeter (University Medical Magazine, January, 1890), which, by rigid testing, was proven an instrument of precision. Dogs were always used. During the first hour the animals were studied without any drug and the result taken to represent a normal standard, although an uncertain one, since the heat processes are subject to considerable variation from hour to hour. Following this normal hour the alcohol was given and the animal subjected to study for from five to six consecutive hours. Since the normal standard thus obtained is of doubtful value as an accurate mean of heat production, it is sometimes too low or too high. Such errors are eliminated by a frequent repetition of the experi

ments.

Three series of experiments were made. In one commercial alcohol (sp. gr. .835) was given by the stomach, and in a second hypodermically. In the third series absolute alcohol was used to eliminate effects which might possibly be dependent upon, foreign substances, like fusel oil, in the former. The doses employed were always in direct relation to the body weight, being in proportion