great enough to discharge a propagated disturbance, responds with its maximum activity for the conditions.

Whether the actual propagated disturbance is smaller at low temperatures than at high is not known. Boruttau (19) has suggested that this is the case. Assuming this to be true, on cooling, enough afferent fibers carrying weakened impulses would have to converge on a single efferent fiber to produce a discharge through this path. In other words, there would be need for a sort of summation of stimuli through afferent fibers on a single efferent fiber. Very little is known about the passage of impulses from one neurone to another. The phenomenon of "afterdischarge" suggests that some sort of summation may take place. And the greater number of afferent fibers than efferent fibers demands. that more than one sensory fiber converge on a single motor neurone.

Adrian (20) has shown that anaesthetics affect the conduction of a nerve impulse by causing a progressive diminution in the intensity of the impulse until it finally ceases to exist. If cooling a nerve should cause a decrement in the size of the impulse as it passes away from its point of origin, this alone might account for the greater threshold necessary for the reflex response, because of the greater distance to the effector organs. A greater threshold stimulus would be necessary in order to affect a large enough group of fibers in which a certain proportion failed to conduct impulses to the muscle and in which enough fibers conducted the entire distance to produce a minimal response.

The results presented above may be interpreted according to the "all or nothing" principle by assuming that at low temperatures a greater threshold stimulus is necessary in order to include a sufficient number of functioning synapses, a certain number being entirely functionless at these temperatures. Or if it should be that the nerve impulse is much smaller at low temperatures, a greater number of fibers must be stimulated in order to conduct the weakened impulses to a smaller number of motor neurones, allowing summation and reinforcement to occur.

SUMMARY

1. The effects of changes of temperature on reflex arc and nervemuscle preparation in the same animal are compared in spinal frogs. Changes of temperature were obtained, ranging from 4°C. to 30°C., by immersing the animal suspended by its jaws in normal saline or Ringer's solutions of various temperatures. The Lucas fluid electrodes were applied directly to the nerve-trunks.

2. Cooling depresses reflex irritability and nerve-muscle irritability, as indicated by a rise of the threshold strength of stimulus. Warming increases the irritability of both, as shown by a lowering of the threshold strength of stimulus.

3. Changes of temperature affect the reflex are more than the nervemuscle preparation. The average change per degree for the reflex is 1.01 Z units; and for the nerve-muscle, 0.11 Z units. The ratio is about 9 to 1.

4. The greater effect of cooling on the reflex arc is in accord with other differences between conduction in the reflex arc and conduction where no synapse is involved; and the results presented here suggest that the place of incidence of this greater effect is at the synapse.

5. The conclusions reached are reconciled with the "all or nothing" principle for nerve.

I wish to acknowledge my indebtedness to Dr. E. G. Martin, who gave me instruction in the use of his method, and to Dr. Alexander Forbes for his many suggestions.

BIBLIOGRAPHY

(1) KUNDE: Virchow's Arch. f. path. Anat., 1860, xviii, 357.

(2) VAN LEEUWEN AND VAN DER MADE: Arch. f. d. gesammt. Physiol., 1916, clxvi, 37.

(3) KANITZ: Temperatur und Lebensvorgänge, Berlin, 1915.

(4) SHERRINGTON: The Integrative Action of the Nervous System, New York,. 1906.

(5) HEINZMAN: Arch. f. d. gesammt. Physiol., 1872, vi, 222.

(6) GOLTZ: Den Functionen der Nervencentren des Frosches, Berlin, 1869.

(7) FOSTER: Journ. Anat. and Physiol., 1873, viii, 45.

(8) SEDGWICK: Studies from Biol. Lab. of Johns Hopkins Univ., 1883, ii, 385.

(9) GITHENS: Journ. Exper. Med., 1913, xviii, 300.

(10) LUCAS: Journ. Physiol. (Proc. Physiol. Soc.), 1913, xlvi, xxxii.

(11) SEDGWICK: Loc. cit.

(12) PORTER: This Journal, 1917, xliii, 497.

(13) HOWELL: Text-book of physiology, 6th ed., Philadelphia, 1915, 85.

(14) BETHE: Allgemeine Anatomie und Physiologie des Nervensystems, 1903, 79. (15) SHERRINGTON: Loc. cit.

(16) BETHE: Arch. f. Mikro. Anat., 1897, 1, 589.

(17) STEINACH: Arch. f. d. gesammt. Physiol., 1899, lxxviii, 291.

(18) LANGLEY: Journ. Physiol., 1901, xxvii, 224.

(19) BORUTTAU: Arch. f. d. gesammt. Physiol., 1901, lxxxiv, 309.

(20) ADRIAN: Journ. Physiol., 1914, xlvii, 460.

THE EFFECTS OF ADRENIN ON THE URINE FLOW OF ANESTHETIZED AND UNANESTHETIZED DOGS

R. E. LEE GUNNING

From the Laboratory of Physiology of the Northwestern University Medical School and the Hull Physiological Laboratory of the University of Chicago

Received for publication January 14, 1918

The existing mass of data concerning the physiologic and pharmacologic action of adrenin on the various functions of the body is, with but few exceptions, complicated by the concomitant effects of general anesthesia. The data on the effects of the various anesthetics per se are lacking in many details. A possible synergic action of the drugs, which is entirely different from the action of either, may exist. Just what this action may be cannot safely be predicted. That anesthesia does affect the adrenin reaction has recently been shown in this laboratory by Berry (1), in his study of the effects of adrenin on the blood pressure reaction of the unanesthetized dog. The most striking effect observed was that ether anesthesia markedly depressed the blood pressure reaction to adrenin. The question then arises, Is this de.pression exerted on the vasomotor system as a whole or as is entirely possible, are the vasoconstrictor nerves selectively affected?

The extent of the data on the adrenin problem has reached such proportions that it would seem that the time is at hand for more intelligent generalization on the rôle played by the drug in the various normal and pathological activities of the body. Before such generalization can safely be attempted, however, the possible variable of anesthesia must be eliminated. The whole problem of the pharmacology of adrenin in the unanesthetized animal is therefore in need of study. Certain aspects of this problem are now under investigation in our laboratory. The first specific function to be considered is that of the effects of adrenin on the renal secretion.

The generally accepted action of adrenin on the secretion of urine is that of marked inhibition. The recorded literature on the subject, however, does not conclusively support this fact. Bardier and Fraenkel (2), who were the first to study the effects of adrenal extracts on the

urinary secretion, observed a marked reduction in the amount of urine passing from the ureters during an injection. With larger doses there was complete cessation of urine flow. These reactions were then followed by marked polyuria which lasted for some considerable time. Occasionally the injections were followed at once by polyuria. Their experiments were made on anesthetized dogs apparently under the influence of curare. Their extracts were made either from desiccated glands or from fresh glands macerated for twenty-four hours at body temperature. Judging from the effects on the arterial pressure, relatively large doses of the drug were employed. Whether the use of curare or the presence of protein decomposition products in their extracts played any part in their results was not determined. Schlayer (3) in his study of experimental nephritis observed that adrenin, when administered intravenously to animals containing much fluid, acts under certain conditions as a diuretic. In some unpublished studies on the effect of adrenin on the intestinal and renal secretion of the dog during hydraemic plethora, we failed to observe this effect. After repeated injections of varying dosages of adrenin the urinary output was decreased and was often one-fifth to one-tenth of that of the small intestine. Schlayer used large doses of adrenin without blood pressure control and the majority of his animals were renally abnormal. Biberfeld (4) reported that the subcutaneous injection of adrenin in doses of 1.5 to 2.5 mgm. per kilo, produces a marked diuresis in rabbits. Pollak (5) also observed diuresis to occur from adrenin injection. Sollman (6) observed that in the perfused kidney, when adrenin was added to the perfusate to make a dilution of 1-50,000, there was a marked decrease in the rate of urine flow. The concentration of adrenin used by this investigator, however, was such as to more nearly produce a toxicologic rather than the physiologic action of the drug. Meltzer and Auer (7) observed that adrenin restricts elimination and suggested that the substance probably interferes with the eliminating power of the kidney. Again, however, the effects of fairly large doses of the drug were studied whereas there is every probability that under physiologic conditions only minute quantities are ever present in the blood stream. Kleiner and Meltzer (8) later, in an extended comparison of the effects of subcutaneous and intramuscular injection of adrenin in rabbits, showed that the absorption of the drug must be at a very slow rate to produce diuresis. Diuresis occurred only with subcutaneous doses of 0.7 to 1.0 mgm. per kilo. Intramuscular doses of this size or larger doses subcutaneously were found to inhibit the flow of urine. Cow (9)

THE AMERICAN JOURNAL OF PHYSIOLOGY, VOL. 45, NO. 4

established the anatomical relationship of an anastomotic branch of the adrenal vein with the cortex of the kidney. From a series of perfusion studies he concludes that under certain conditions adrenin in appreciable amounts is poured directly into the kidneys from the suprarenal bodies of the intact animal, producing a diminution in the flow of urine. Failing to find reported in the literature the effects of threshold dosages and of slow infusions of adrenin on the urine flow, which probably more nearly approach the normal discharge of the gland, it was necessary to study the effects of these dosages in the anesthetized dog for comparison with their effects in the unanesthetized animal.

METHOD

Medium sized dogs were used as experimental animals. Quiet, good-natured dogs were selected and great care used to avoid exciting them. After some petting, the dog was carefully and quietly laid on the operating board and strapped back down. In the early experiments one-eighth to one-fourth of a grain of morphine was administered but later with careful handling this was found to be unnecessary. One to two cubic centimeters of 2 per cent cocaine solution or an equal amount of 1 per cent quinine-urea bimuriate solution were injected intradermally over the femoral artery and vein just below Poupart's ligament. After waiting a few minutes for the anesthetic to take effect, arterial and venous cannulas were set. Next the skin and subcutaneous tissue over the lower median line of the abdomen were injected with the anesthetic. Again after waiting a few minutes an incision 5 or 6 cm. in length was made into the abdominal cavity, beginning at the upper border of the pubic bone and extending cephalad. The urinary bladder was then aspirated if necessary and drawn through this incision. The ureters were each isolated and cannulated close to the bladder and the viscus was replaced in the abdomen. The abdominal incision was then closed, leaving the cannulas protruding from the wound. All manipulations were made with as little trauma as possible.

The ureteral cannulas were led into a Y tube which extended over the edge of the table and the system was filled with water or salt solution. The blood pressure was recorded from the femoral artery with an undamped mercury manometer. The ureteral outflow was recorded in drops by means of a key and signal marker. Adrenin (Parke, Davis and Company's "Adrenalin") in varying doses was injected into the femoral vein, using a small amount of warmed 0.8 per cent sodium chloride solution to flush it in.