Experimental Biology and Medicine

The general lines of inquiry to determine the seat of antibody formation have been to seek the tissue where a given antigen is fixed; to determine the locality where the corresponding antibody may first be detected; or to prevent the formation of antibody either by specific injury (benzol, x-ray) or by extirpation of an organ. It has also been claimed that antibodies may be produced in tissue culture. With the evidence obtained by these methods it has been variously concluded that the essential antibody formers are the leucocytes; the leucocyte forming organs, spleen, bone marrow and lymph nodes; the liver; and the capillary endothelium.

We have probably been led astray by the expectation that some particular organ like the spleen, or some strictly localized cell group, rather than a more universal tissue, is responsible. Many of the phenomena of general immunity can be better explained on the assumption that some widely distributed cell-type is the essential antibody producer. The condition of strictly local immunity in various parts of the body, which we believe has now been proved to exist, can be understood only on such an assumption.

The reticulo-endothelial system of cells (Aschoff) fulfills this criterion of wide distribution throughout the body including as it does one of the constituent elements of connective tissue and the capillary and lymph space endothelium, and being related to adult endothelium and the monocytes of the blood. It includes, moreover, the most markedly phagocytic and resistant cells of the body; indeed its differentiation depends on its ability to take up rapidly and to retain particulate matter and colloidal substances. The undoubted importance of spleen, lymph nodes and liver in the disposal of foreign cells and in antibody-formation, that has already been mentioned, would depend on their content of elements that make up the reticulo-endothelial system; and the failure of removal of one of these organs (e. g. spleen) to prevent antibody formation entirely would depend on the vicarious or increased functioning of other parts of the same system.

The behavior of small solid bodies in contact with two immiscible liquids has been much discussed, especially in connection with phagocytosis. Direct observation of interfacial phenomena is possible with bacteria or other particles in films viewed in the darkfield microscope. A small drop of each phase, one of them containing the suspended bacteria, is placed on a clean slide. A cover slip is laid on top. The boundary between the two liquids appears as a brilliant band and the bacteria as shining motes. Bacteria in two phase films of water with a variety of organic liquids have been studied.

Ordinary Gram-positive or Gram-negative bacteria have been trapped in the liquid-liquid interface of all preparations examined. The trapping mechanism is much stronger in preparations with high (e. g., hydrocarbon-water) than with low (e. g., alcohol-water) liquid-liquid interfacial tension. Bacteria in the interface exhibit characteristic gliding movements along the interface, moving in the boundary line from regions of low to regions of higher liquid-liquid interfacial tension. Escape from the interface is possible in a variety of ways, e. g., by sticking to the glass slide or cover slip, by Brownian movement, strong action of the flagella, jolting by other bacterial clumps, or even by centrifugal force while streaming rapidly around a curve in the interface. The efficiency of the interfacial trapping mechanism may be decreased by adding substances, e. g., Na-oleate, lowering the liquid-liquid tension.

Acid-fast bacteria behave entirely differently. They exhibit little or no stab'ility in the interface and pass readily or even spontaneously into the organic phase.

To account for these phenomena it is necessary to consider three interfacial tensions, Tow the liquid-liquid tension, Tso the solid-organic phase tension, and Tsw the solid-water tension. If neither solid-liquid tension is greater than the sum of the other solid-liquid tension plus the liquid-liquid tension, the bacterium will be trapped in the interface. If either solid-liquid tension exceeds the sum of the other two tensions, one liquid will spread on and engulf the bacterium; i. e., if Tsw > Tso + Tow, the organic phase will spread on and engulf the bacterium. This is the condition with acid-fast organisms in a water-oleic acid interface. The tension between the bacterial fat and fatty acid envelope and the oleic acid is low, Tow is not high, and Tsw is greater than their sum. The bacterium passes spontaneously into the organic phase.

Wheat which was grown in the dark (etiolated) was found to have no antirachitic potency. Wheat which was grown in the light and irradiated with mercury vapor lamp conferred protection when fed to rats (10 gm. daily).

The same difference in regard to protective action against rickets was observed in vegetables which were irradiated after they had been plucked. Green lettuce leaves from the market were of no value in preventing rickets, whereas after irradiation when fed in same amounts (10 gm.) they had become antirachitic.

Therefore, in the plucked as well as in the growing green vegetable, irradiation led to the formation of an antirachitic factor.

The fact that inert fluids can be activated and rendered antirachitic by means of irradiation was reported by us some months ago. 1

Steenbock has reported a similar result in relation to the production of a growth-promoting factor. He has also confirmed our observations in regard to rickets, having been able by means of irradiation to render “fats” active in preventing this disorder. 2 In our experiments various inert fluids were irradiated with the mercury vapor lamp for one hour at a distance of one foot, in order to ascertain whether by this means they could be endowed with antirachitic potency. It was found that cotton seed oil and linseed oil could be rendered specifically active by this means. After these oils had been thus irradiated they were able to protect rats from rickets when 0.1 cc. daily was fed in addition to the standard rickets-producing dietary (No. 84). The irradiated oils were able to store this factor for a considerable period. These experiments demonstrate the possibility of producing an antirachitic factor in vitro.

It seems from a review of the literature bearing on the question of the rôle of vitamins in bacterial growth that nobody has investigated the influence of Hydrogen ion concentration of media on the growth promoting effect of these factors. The purpose of the present work was to study this relationship.

As tomatoes have been shown by Thjötta and Avery 1 to contain food accessory substances for bacteria, the growth promoting effect of tomato extract adjusted to pH 7.0 upon the growth of B. Shiga in media of pH range 5.4-9.0 was investigated.

It was found 2 that this extract is not able to promote the growth of B. Shiga in broth at initial pH 5.2-6.2, that it has a moderate growth accelerating effect on this microörganism at initial pH 6.6-7.8 and has the best effect when added to broth having a pH 8.2-8.6. Furthermore it may be pointed out that the best growth of B. Shiga, in the presence of tomato extract, occurs at a pH (8.2-8.6) which is by itself an unfavorable one for the multiplication of this microörganism in plain broth.

Before going any further with this investigation it was decided to determine the factor responsible for the growth promoting effect of tomato extract in the case of B. Shiga.

Thjötta and Avery, 1 who used the tomato extract for initiating the growth of Pfeiffer bacillus in blood free media, demonstrated that the poor nitrogen content of tomatoes cannot alone be responsible for the acceleration of bacterial growth.

It seemed necessary to determine whether the high sugar content might be the responsible factor, since it was found that the extract used in these experiments contained 3.54 per cent of reducing substance expressed as glucose. But a quantitative titration of the growth promoting factor of this extract demonstrated that the latter had a pronounced effect when added in a dilution 1 :l00.

1. Pepsin'inactivates'insulin but does not digest it.

2. Liberation or dissociation of insulin from pepsin takes place, even after prolonged contact, at a properly adjusted pH.

The Ramon test was evolved from the works of Calmette and Massol, 1 who in 1909 applied the flocculation test for titration in vitro of antivenom serum against Cobra venom. Nicolle, Cesari and Debains 2 in 1919 applied the same principle for titration of diphtheria and tetanus toxin-antitoxin by the method of Ascoli. 3 The reaction consisted in the formation of an opalescent ring in contact with a concentrated toxin and gelatinized antitoxin. Georgi 4 added a suspension of cholesterolized heart extract to the mixtures of toxin-antitoxin to obtain flocculation.

Ramon 5 in 1922 found that diphtheria and tetanus toxin and antitoxin alone, when mixed in certain proportions, will bring about flocculation. The mixtures with a deficiency or excess of either toxin or antitoxin will fail to flocculate. The first precipitate to appear in the mixtures Ramon calls the “precipitate indicateur” and this corresponds to a nearly neutral mixture. From the latter the values of toxin or antitoxin are calculated.

Scholtz 6 in 1923 found that it was noit necessary to employ for this test the enormous quantities of toxin used by Ramon (20 cc. for each tube). He used instead 2 cc. of a stable titrated toxin and corresponding diiutions of serum. Glenny and Okell 7 in 1924 also used from 2 to 5 cc. of a stable titrated toxin and minute quantities of undiluted serum. A11 workers agree that the Ramon titrations agree closely with guinea pig tests and are therefore applicable for practical routine estimation of values of toxin and antitoxin.

Povitzky and Banzhaf working on this test since early spring of 1924 arrived at the same conclusions. They found that the most: reliable and constant results are obtained by the use of a potent stable toxin titrated for its L+ and flocculation values and undiluted serum or (if the latter is sufficiently strong) diluted 1 :I, that is, with an equal amount of physiological salt solution. The amount of toxin was usually 2.5 cc. per tube.

We have succeeded in demonstrating in sterile filtrates of broth cultures of four out of twenty-one strains of staphylococcus aureus isolated from various conditions, a powerful poison with a selective action for the skin. The poison is produced by the growth of these strains in any well buffered broth medium containing only a small amount of glucose. The medium from which our most powerful poisons were obtained was prepared by adding an equal volume of M/15 phosphate buffer solution pH 7.4 to ordinary sugar—free meat infusion broth containing 4 per cent Witte peptone, boiling, filtering and autoclaving. After inoculation with a suitable strain of staphylococcus, the poison can be demonstrated in the sterile Berkefeld filtrates of this broth after 24 hours growth, but our most toxic poisons were obtained after four to six days growth.

The toxicity of the filtrates was tested by intracutaneous inoculation in rabbits. One-tenth cc. of filtrates was inoculated intradermally and at the same time a control of 0.1 cc. of uninoculated broth was always injected in the same rabbit.

The reaction produced by the injection of a toxic filtrate becomes evident from two to six hours after injection as a bluishpurple area of 2 to 5 cm. in diameter, depending on the toxicity of the filtrate. The next day, the purple coltor assumes a yellow tinge, and there is usually added a deep red zone of 0.5 to 3 cm. surrounding the yellowish area. The circumsoribed yellowish area of 3 to 5 cm. becomes progressively yellower-as if necrotic. By the fifth day brown patches appear in the yellow area and increase in size until, at about the twentieth day, the whole lesion has become a dark brown scab. Four to eight weeks later the scab falls off leaving an ulcer. Microscopically the lesions show marked infiltration of polymorphonuclear leucocytes with necrosis of epidermis and underlying corium,

The so called germicidal action of milk has given rise to much discussion and controversy as to its extent, significance and even its existence. All workers are in agreement that the germicidal effect, if such exists, is only slight in action and transitory in occurrence after the milk is drawn from the animal. Some doubt its existence at all; some believe it to be effective so far as some organisms are concerned but to have no effect upon the typical milk bacteria; others believe in a bactericidal action, but hold that it may occur in the milks of some cows and not in others, and may or may not appear in the milk of the same cow at different times.

Studies made by measuring the rate of increase in the natural fortuitous flora of fresh milk have, quite obviously, given contradictory results. When fresh milk is inoculated with known pure cultures, the results are still not definite as the germicidal effect is here complicated by the presence of the natural lag period, in the culture used as the inoculum, before rapid growth ensues.

It seemed to us that the problem could be got at more directly, and more definite information obtained, by the application of modern knowledge of the bacterial growth curve. As is well known, the freshly inoculated bacterial culture passes through a latent or “lag” phase before multiplication begins. It is also well known that transplants taken from a culture only a few hours old, while it is in the rapidly growing stage, will show no lag but continue multiplication at the same rate when seeded in suitable media. We therefore inoculated freshly & awn aseptic milk with young cultures in the period of rapid growth. For this purpose Strept. lactis, probably the most perfectly adapted milk organism, was used. Samples of milk were brought to the laboratory immediately after king drawn and placed in a water bath at 37° C. They were then inoculated from rapidly growing cultures three hours old at 37° C.; sterile milk king used as the growth medium from which the inoculum was taken. In these experiments control tests were also run by inoculating at the same time the same culture into sterilized milk (autoclaved) and incubating as in the case of the raw fresh milk.

In 1922 Koppányi and Weiss 1 reported in a preliminary note regeneration of the spinal cord in the fish (Carassius vulgaris, a small specimen 4-5 inches), and in the larvae of the salamander (Salamandra maculosa). Their method was complete section of the vertebra and cord in the high thoracic region. In one experiment three vertebrae were removed with their corresponding spinal cord. Regeneration occurred in both cases. Regeneration was evidenced by functional return in paralyzed regions and morphologically by histological examination. After the section, the animals were quiescent caudad to the section. Three weeks later the salamander larvae showed motility of the posterior portion. The motility slowly recovered until four or five weeks after the section when there was apparently complete functional recovery; their locomotion was coordinate and regular. During this period some of the salamanders metamorphosed. This confirms Loeb's statement that transection of the spinal cord in Axolotl does not prevent metamorphosis. Regeneration was less rapid in the fish, requiring six to eight weeks. There was apparent functional recovery, for swimming was coordinate, but the animals always lay upon the side and seldom swam in the normal position. Histologically, in conjunction with Kolmer, we traced the fibers across the section. These fibers were somewhat fewer in number than in the normal cord. They were mostly gathered in bundles although there was considerable interlacing.

In the salamander with the three vertebrae removed, the vertebrae and cord regenerated completely.

Our first experiments in the present series were done on the adult newts, Diemyctylus viridescens and Diemyctylus torosus. These animals were unsuitable for the experiments because the cord is so far dorsal that the body musculature is unable to hold the cut ends of the spinal cord in opposition when the vertebral column is transected. We used, therefore, the goldfish (Carassius auratus). These were large specimens fro'm 8 to 14 inches in length. A small transverse section was made through the skin and the muscles spread apart and the entire vertebral column cut in two with scissors. There was no bony continuity remaining between the regions above and below the section. A probe was passed fully from side to side in the section. The spinal cord must have been sectioned. The skin was sutured with silk. Wound healing was slow and was incomplete in about six weeks time.

The eggs and sperm of the sea urchin, Arbacia punctulata, were exposed simultaneously to the same x-rayed treatment and subsequently the x-rayed eggs were fertilized by untreated sperm, and untreated eggs were fertilized by x-rayed sperm. Untreated eggs and sperm from the same sea urchins were used in the control cultures. Two sets of experiments have been completed, one during the summer of 1923 and the other during the summer of 1924. In 1924 the conditions of the x-ray treatment were so arranged that the temperature of the water in which the eggs or sperm were kept during treatment did not vary from that of the room in which all the cultures were kept by more than 1° C. The temperature of the room during the different experiments varied from 19° to 22° C. The x-ray treatment was given with a standard Coolidge tube, tungsten target, at 50,000 volts and 3 milliamperes, the distance from the target to the eggs and sperm being 25 cm. In 1923 a portable radiator type tube with tungsten target was used and run at 50,000 volts and 2.5 milliamperes. The distance from the target to the germ cells was 11 cm. and the glass cups were surrounded by a lead box covered on the top where the x-rays entered by a thin sheet of aluminum 3 mills in thickness.

After the treatment of the eggs and sperm, fertilization was brought about in finger bowls and the larvae were reared up to forty-eight hours after fertilization, at which time samples were fixed in 2 per cent formaldehyde. In these samples where development had proceeded at the normal rate in the controls it was found that a considerable number of the pleutei were in a stage where the larger arms were appreciably longer than the rest of the body. This stage was numbered six and the development previous to this was divided into five stages as follows: 1, blastula; 2, gastrula; 3, gastrula which has become triangular and shows the beginnings of the two larger arms; 4, these two arms have developed but are appreciably shorter than the rest of the body; 5, the arms are approximately the same length as the rest of the body. In the case of each sample the stage in development of one hundred larvae was recorded.

The writers have performed a series of experiments upon the hearts of dogs perfused with oxygenated Locke's solution by a modified Langendorff method. Observations were made upon the effect of changes in (1) hydrogen ion concentration and (2) oxygen content of the perfusing fluid upon the transmission of the excitatory process in the auricle. Electrical records were obtained by placing paired non-polarizable electrodes upon the auricle in such a way that the conduction interval was recorded over a single stretch of 12-16 mm. or over consecutive portions of 6-8 mm each. Each pair of electrodes was connected with a galvanometer string.

The effects of changes in pH of the perfusate upon conduction in the auricle were consistent with those reported by Andrus and Carter 1 on the A-V conduction. An increase in the alkalinity of the perfusing fluid, pH 7.8, was attended by a faster rate of transmission both in the naturally beating and rhythmically driven heart. In the latter the same result was obtained whether the stimulating electrodes were placed upon the base of the auricle or upon the appendix. A pH lower than normal, pH 7.0, caused the excitatory process to be propagated more slowly, and again, this result was obtained in the auricles following the sinus rhythm or rhythmic induction shocks introduced through electrodes placed either upon the base or tip of the auricle. In the naturally beating heart an increase in rate followed change to a higher pH and a fall in rate a lower one. Both the development and propagation of the excitatory process were, then, affected by changes in reaction of the fluid bathing the tissue and in the same direction.

In certain experiments the perfusing solutions were saturated with nitrogen in place of oxygen with striking results. If the auricle was perfused with an oxygen-free solution of pH 7.0 there resulted almost immediately a profound reduction in the transmission rate. Soon there appeared a condition in which the transmission of the excitation became progressively slower the farther the wave traveled from its point of origin, so that records taken with three pairs of electrodes showed a longer transmission interval between the second and third than between the first and second. As this condition progressed the muscle under the distal contacts responded to only every second wave of excitation as recorded at the middle electrodes and finally ceased altogether to become excited.

The experimental anatomical studies of Boeke (1913), 1 Boeke and Dusser de Barenne (1919), 2 and Agduhr (1919) 3 indicate clearly that, in addition to somatic afferent and efferent components of the cerebrospinal nerves, nerve-fibers of sympathetic origin also terminate on voluntary muscle-fibers. Not a few physiological studies tend to corroborate this anatomical finding, others yield only negative results. Consequently, the sympathetic innervation of voluntary muscles is not yet universally accepted even as an anatomical fact.

The material on which the present study is based was obtained from three dogs in which the somatic nerve-fibers supplying certain muscles had undergone degeneration following section of the corresponding nerve roots. 4 In two of the dogs the roots of the seventh, eighth and ninth thoracic nerves were exposed by laminectomy and cut distal to the spinal ganglia and proximal to the communicating rami. These animals were killed four weeks later. To avoid confusion which might arise by reason of overlapping of the areas of distribution of the intercostal nerves or the plurisegmental innervation of muscle-fibers, muscle was taken for study only from the eighth intercostal space. Control material was taken from intercostal spaces in which the nerves were left intact. In the third dog the mandibular nerve was cut intracranially. When this dog was killed twenty-three days later, portions of the masseter and pterygoid muscles on the side of the operation and control material from the corresponding muscles on the opposite side were taken for study. In all cases muscle tissue in which the somatic nerve fibers had undergone degeneration and control material was prepared both by the gold chloride and the pyridine-silver method.

In the control material prepared by both methods the terminal branches of the nerves comprising both myelinated and unmyelinated fibers as well as the terminations of fibers of both types on muscle-fibers may be observed. Our observations on the endplates of the ordinary myelinated somatic efferent fibers corroborate those of Boeke (1921) 5 who described these end-plates as hypolemmal in position and resting on nucleated 'sole-plates'composed of granular sarcoplasm which lies superficial to the myofibrillae and in which is imbedded a delicate reticular structure. As the myelinated somatic efferent fibers approach the end-plate some of them are accompanied by a slender unmyelinated fiber which terminates in a small end-net or end-loop within the area occupied by the 'sole-plate'on which the motor end-plate rests. Doubtless, these are the accessory fibers described by Boeke.

In this study we adopted the plan of extirpating the sympathetic trunk in the lumbar region on one side only, using dogs as the experimental animals, and comparing the tonus, the power of contraction and the resistance to fatigue of the muscles of both hind limbs both immediately after the operation (6 dogs) and after ample time (10 to 36 days) had been allowed for the degeneration of the sympathetic nerve-fibers supplying the muscles of the hind limb on the side of the operation (4 dogs). 1 Only the more important results obtained can be reported in this paper.

In the animals which were allowed to live following the operation a degree of hypotonus of the muscles of the hind limb on the side of the operation could be demonstrated by palpation for some time, but gradually subsided. This is in accord with the observations of Negrin, Lopez and von Brücke (1917) 2 and Dusser de Barenne (1917) 3 on cats. We are not prepared to state how long such hypotonus persists. In one animal it was still demonstrable two weeks after operation, in another it could not be demonstrated by palpation ten days after operation. However, when these dogs were placed under surgical anesthesia, even after the hypotonus referred to had apparently subsided, a marked difference in the flaccidity of the muscles of the two hind limbs became apparent. As the animals lay on their backs with the limbs free the hind limb on the side of the operation dropped to a lower position than the other. When equal weights of 100 or 200 grams were attached to both hind feet and suspended over pulleys at the end of the table the limb on the side of the operation was more fully extended than the other. Weights of 600 grams drew both limbs out equally, both being fully extended.

In order to secure graphic records of the contractions of individual muscles the hind limbs were fixed in moderate extension, as the animal lay on its back under surgical anesthesia. The tendons of the gastrocnemius muscles were isolated and cut at their insertions. A strong cord was attached to each isolated tendon and led over a pully at the end of the table to an ergograph or to a muscle lever with a weight attached. The sfimulating current was supplied by two or three dry cells and an inductorium. The stimulus was applied to the exposed sciatic nerves.

It has been known for a considerable period of time that the female in many breeds of birds takes on certain male characters following the removal of the ovary, either by ovariotomy or pathological absorption. In the latter case the reduction of the ovarian mass may be only partial or for a short period of time, followed by recovery and regeneration of the ovary; or the ovary may be permanently reduced and destroyed. In all such cases the female assumes certain male characters, and the extent to which such male characters are developed and their permanency, apparently depends upon the degree of removal or reduction of the ovary.

It is the purpose of this paper to call attention to experiments being carried on in this laboratory which have a very special bearing on some of the work recently reported by Crew 1 in the fowl, and by Riddle 2 in the pigeon. Investigations along this line were suggested by Professor F. R. Lillie, early in the fall of 1922, and have been carried on under his direction. 3

My work thus far has been confined exclusively to the Light Brown Leghorn race. Sexual dimorphism is this breed is very pronounced, and the birds are more active and perhaps somewhat more hardy than are the birds of some of the heavier breeds. Most of the birds were obtained from a reliable breeder immediately after hatching and reared in the laboratory, where operations could be made at various ages and results observed.

The operation in the female is relatively difficult, owing to the proximity of the great blood vessels. In the young female, prior to the age of laying, the ovary is a flat sheet of tissue attached intimately by one surface to the ventral surface of the left iliac and vena cava. It was found that the ovary is most readily and most satisfactorily removed during this stage of its development. If the operation is delayed until the changes preceding the laying season have set in, the ovary has enlarged and has become more extensively vascularized, and its removal is more difficult.

The bacterial allergies, though they are phenomena of relatively specific hypersusceptibility, differ in certain fundamental respects from true protein anaphylaxis. Our own investigations on this subject have dealt chiefly with the tuberculin reaction in regard to which it seems safe to state that:

1. Tuberculin allergy as manifested by the intracutaneous reaction may occur without general anaphylaxis to tuberculoprotein, and vice versa.

2. The active principle of tuberculin which elicits reactions in the sensitive subject is not a protein.

3. Tuberculin sensitiveness seems inseparably associated with the reactions of the animal tissues to the tubercle bacilli or their constituents; in other words, after prolonged experimentation, we arrive again at the recognition that there is no tuberculin sensitiveness without a tubercle. Here again the fact that the tuberculin reaction is merely the classical example of the bacterial allergies in general is apparent in that our own work with other bacteria as well as especially that of Fleischner and Meyer with bacillus abortus has indicated that bacterial allergy is associated with infection rather than with the type of immunization with soluble bacterial substances which leads to antibody formation without necessarily active focal inflammatory reactions.

In the case of the tubercle bacillus Petroff and the writer have shown that the injection of the dead bacilli induces tuberculin allergy quite effeotively, probably because of the insolubility of these organisms, in consequence of which the dead organisms lead to tubercle formation essentially analogous to that induced by the living. This point, apart from its possible practical imterest, is theoretically significant in that it indicates that, in the association of bacterial allergy with infection rather than with irqmunization, it is the tissue reaction which is of importance rather than any differences in the products which pass into the infected animal from dead and living bacteria respectively.

Figures for blood sugar following decerebration have been given by J. Mellanby, 1 but he used an injection of starch for decerebration and the actual area of brain damage remained very uncertain. He found a hyperglycemia lasting as long as six and a half hours and supposed that this high sugar level might be maintained indefinitely. On the other hand, Bazett and Penfield found a disappearance of sugar from the urine in decerebrate cats within two days.

A few observations on blood sugar have been made on cats decerebrated aseptically by a scooping out method under chloroform anesthesia. The blood has been obtained from a cut in the ear and the sugar estimated by Maclean's method. In one such animal a blood sugar of 0.256 per cent was obtained 24 hours after operation, and the urine secreted during the night contained sugar. In another animal the blood sugar 24 hours after operation was found to be 0.19 per cent. In both of the above animals the pituitary was left intact. In an animal decerebrated acutely with the Sherrington guillotine and chloroform anesthesia in which the whole of the hypophysis was removed and also any accessory pituitary tissue in the pharynx a blood sugar of 0.37 per cent was obtained one half hour after decerebration and the level gradually fell until 7 hours after operation it was 0.20 per cent and finally reached 0.165 eleven and a half hours after the operation. At this time the animal began to have difficulty in breathing from tracheal mucus, and his general condition became much worse and with this change the blood sugar rose again reaching a level of 0.38 per cent just before death, 21 hours after operation. The initial fall of blood sugar could not therefore, have been due to a lack of available glycogen. These few figures confirm the data obtained by Mellanby, but indicate that if he had carried on his experiments for a longer time the blood sugar might have fallen, and that the high blood sugar cannot be considered as maintained at a constant high level.

In further experiments animals have been decerebrated aseptically by a scoop method under anesthesia induced by chloretone given in alcohol and water by stomach tube in a dosage of 0.17 gm. per kilo, the resultant anesthesia king loccasionally supplemented by a little chloroform. In these animals the hemorrhage at time of operation has been much less than with chloroform anesthesia. The chloretone has maintained its effects for 24 to 48 hours or even longer, although some rigidity has usually developed within 6 to 8 hours. In some animals the pituitary has been left intact, in others it has been remolved at the time of operation and in sow experiments an attempt has been made to remove pituitary, but at autopsy it has been found to have been only partially removed.

The previous observations, 5 (2528), have demonstrated that the tomato extract is not able to promote the growth of B. Shiga in broth of initial pH 5.2-6.2, that it has a moderate growth accelerating effect on this microorganism at initial pH 6.6-7.8 and that it has the best effect when added to broth having a pH 8.2-8.6. However, upon studying the end reaction of cultures containing tomato extract, it was observed that the final pH of these cultures, while always markedly acid, showed considerable difference, this difference depending upon the initial pH of the medium. Thus, for example, initial pH 8.2 of the tomato extract broth gives a final reading of 5.8, whereas initial pH 7.0 gives a reading of 5.2.

In this connection the work of Cohen and Clark 1 was of considerable interest. These authors found that on the acid borders of optimum zones of pH, slight change in pH produces a marked inhibitory effect upon the growth of bacteria (4.7 for B. Shiga), but the acid border shifts with the nature of the acid and a distinct difference between the effect of HCl and that of acetic acid was noticed. When adjustments were made with acetic acid, the narrow pH zone 5.5-5.7 may be considered critical, while when HCl was used, this point was 4.8-5.0.

In the author's experiments, the question, therefore, arose as to whether the organic acids contained in tombto extract shift the critical point of B. Shiga from 4.7 to some point quickly reached by a tomato extract culture of initial pH 6.6-7.8, but much slower, or perhaps never at all reached by a culture containing tomato extract of initial pH 8.2-8.6.

Investigation of this point demonstrated in fact that the tomato extract broth, from which the growth promoting factol has been remover by process described elsewhere, 2 is able to inhibit completely the growth of B. Shiga at pH 5.4-5.6 whereas plain broth becomes inhibitory only at pH 4.8.

Since the previous observations have shown that bacterial cultures containing tomato extract undergo a rapid change in hydrogen ion concentration leading to a considerable increase in pH, it was decided to determine whether the pH of surrounding fluid would have anything to do with the fate of growth promoting factors in bacterial cultures.

Preliminary experiments had shown that during the first 24 hours of multiplication of B. Shiga in plain broth of various initial pH no growth promoting nor growth inhibiting substances can be detected in the surrounding fluid which could vitiate the interpretation of the following results.

Cultures of B. Shiga containing tomato extract and of initial pH 5.4, 7.0 and 8.6 and control cultures of the same pH in plain broth were centrifuged at various intervals of incubation time for a period of 24 hours, the pH of these samples recorded, adjusted to pH 8.2 and inoculated with B. Shiga. These experiments demonstrated that the growth promoting factors are present in fluids derived from cultures of initial pH 5.8 and 8.2 for the period of 24 hours of observation, whereas they completely disappear from tomato extract broth cultures with initial pH 7.0 as soon as the latter reaches 5.2 (10 to 14 hours). It may be pointed out that the culture of initial pH 8.2 does not reach a pH higher than 5.8 during 24 hours of incubation, while that of initial pH 5.8 naturally reaches pH 5.2 very quickly (2 to 3 hours).

What, then, are the conditions that lead to the disappearance of growth promoting factors from bacterial cultures with initial pH 7.0?

1. The phenomenon of disappearance of growth-promoting factors does not depend on the amount of bacteria in these cultures, since these factors persist in tomato extract cultures of initial pH 8.2 which, as was shown before, 1 contain a more abundant growth than the cultures under discussion.

Bayliss 1 describes the process of autotomy as follows: “If a crab be picked up by one of its ambulatory appendages, it generally, by a powerful contraction, breaks this leg off at a particular place and so obtains freedom. This mechanism was first investigated by Fredericq, and more recently by Raskam (1913). The second segment of the leg in the crab consists of two parts which are distinct members in most crustacea and united by a movable joint. In this animal, however, in place of a joint, there is a double membrane whose two components are not very firmly united. In the middle of the membrane there is an aperture, through which nerve and blood vessels pass. Certain muscles are so arranged that by a powerful contraction, they separate apart the two layers of the membrane. Thus no soft parts are torn, except the nerve and blood vessels; there is practically no bleeding and the peripheral part of the appendage is rapidly regenerated.”

A similar phenomena has long been known in some vertebrates. If a lizard is held by the tail it frees itself with a quick jerk at the expense of the piece of tail which remains in the hand. This phenomenon has also been termed autotomy, but it is different from true autotomy in that the tail itself is passive and is pulled apart, whereas muscles of the crayfish leg accomplish the separation which occurs there. At Dr. Carlson's suggestion, we wish to make manifest this difference by denoting as autotomy that mechanism by which an animal suddenly discards a portion of its body by a mechanism located in the member itself, and by denoting as pseudo autotomy that process by which an animal is enabled to {lose a portion of its body, the part king lost playing a merely passive rule. There is a possibility that such a portion of the body, the lizard tail, for instance, is especially adaptive (as to be unusually friable at a certain segment) for this purpose so as to facilitate the escape of the animal.

In 1922 Dragstedt 1 and his co-workers were successful for the first time in preventing the onset of tetany in thyroparathyroid-ectomized dogs, their method being to add to the diet large amounts of lactose—from 50 to 100 grams—daily.

It has been our plan to repeat Dragstedt's work and do it more exactly and systematically, taking advantage of Cowgill's 2 method of feeding. In doing so we have been able to vary the foodstuff components in any way we like, keeping the total caloric value nearly the same.

After the animals were fed on a certain diet for 6 to 7 days, they were operated upon and the diet was continued. Severe tetany when it occurred usually developed on the second or the third day following the operation. When animals developed tetany, various attempts were made to cure them and special attention was paid to the efficacy of lactose in the treatment. If the animal did not show tetany under the experimental conditions, the diet was changed until symptoms of tetany appeared; this was done in order to exclude any possible action of accessory glands in the experiments.

In the table are given results obtained with various diets. In each case all constituents except casein and carbohydrate were kept constant.

As will be seen in this table, in a diet relatively low in casein, the amount of lactose which may prevent the onset of tetany is comparatively small, and when casein is increased, more lactose is needed.

The dextrin experiments are significant. It is niuch easier to prevent the appearance of tetany or depression in thyropara-thyroidectomized animals than it is to cure it, once it has become established. Although lactose administration at the time severe symptoms of tetany develop appears to give much relief after a few hours, it has never ken so successful as the administration of calcium salts which cure almost all symptoms in a short time. However, by stopping the symptoms of tetany by calcium administration and then changing the diet to a lactose food mixture, we have been able usually, without giving any more calcium, to save the animals from the recurrence of tetany. Furthermore, in mild cases, we were often convinced that lactose administration alone by stomach was effective enough to cure the symptoms.

As a result of experimental work concerning the healing of the meninges following injury, it was found by one of us (S. C. H), 1 that the dura in the presence of an intact lepto meninx heals without adhesion to the underlying membrane. This is brought about, not by the ingrowth of mesothelium from the surrounding normal dura, but by the direct transformation of mesenchymal cells into mesothelium which comes to lie against the uninjured and impervious lepto-meninx.

On the other hand, 2 when the lepto-meninx is injured with the overlying dura remaining undamaged, an analogous healing without adhesions does not take place. The lining mesothelium of the dura disappears and this membrane enters intimately into the reparative process in the subjacent lepto-meninx, the final result being dense adhesions. This seems to indicate some fundamental histogenetic difference between these two structures.

Harrison 3 has shown by experimental methods that the nerve sheath cells are derived from the neural crest. Inasmuch as these cells are analogous to the lepto-meninx in their relationship to the nerve structures which they surround, it seemed possible that this membrane was likewise derived from the neural crest while the dura probably developed from the mesenchyme.

An examination of pig and chick embryos at and previous to the time of the outgrowth of the nerve root fibers from the neural tube, confirmed this belief.

It seemed possible to verify this supposition by experimentation. This work was carried out by the transplantation of sections of the neural tube of the amblystoma with and without neural crest cells. If the lepto-meninx develops from neural crest cells, then transplanting a portion of the neural tube free from such cells should lead to the growth of neural tissues surrounded only by dura, while such transplantation carrying with it nerve crest cells should lead to the formation of both the dura and the lepto-meninx about the neural transplant. Such experiments were carried out in the Spring of 1921, with the following results.

In the starfish Pisaster ochraceus and Patiria miniata the coelomic fluid was found to have a pH about 7.6, compared with the normal for sea water of 8.3. The pyloric caeca, which are naturally immersed in coelomic fluid, were found to have a pH of 6.7. All determinations were made colorimetrically and are given without correction.

Excised caeca, when suspended in a small amount of sea water, gradually altered the pH toward 6.7. Even when the sea water was changed from its natural pH by acidification and subsequent aeration till equilibrium was established, the caeca altered the pH to 6.7 when the water was initially at any pH between 8.6 and 6.3. An excellent criterion of survival of the caeca is presented by the persistence of ciliary activity of the epithelium. These cilia maintain their activity for about 48 hours on excised caeca where the initial pH of the medium is between 8.6 and 6.3. The optimum pH for survival is 6.7, the same as found normal for the caeca. If the pH was initially less than 6.3, the cilia soon ceased and the caecum disintegrated, and any caeca which fell below 6.7 were dying and approached 6.3 as the apparent natural point for disintegration.

Determinations of the rate of digestion of gelatine solutions introduced into excised caeca showed an optimum at about 6.7. The degree of digestion was followed by Van Slyke's method for amino-nitrogen, taking samples from the medium surrounding the caeca after the gelatine had been introduced. In the controls it appeared that scarcely any amino-nitrogen is produced naturally by the caeca in spite of the intense ciliary activity. As CO₂ is produced quite abundantly, oxidation of fat is suggested as the source of energy.

Using the pH change as a proper measure of CO₂, it was apparent that CO₂, production was most vigorous at pH about 6.7. In the more alkaline media CO₂ production is gradually dim inished with further departure from this normal pH. Below pH 6.7 the production appears practically to cease, a condition coincident with rapid death.

Ca, Mg, and CO₂ in sea water are the common constituents which are most variable in quantity. The condition of these substances is sensitive to the slight changes of the ocean, in contrast to the stability of soluble neutral salt forming substances. They are, furthermore, concerned in a direct way with vital reactions. CO₂ is also one of the most important indicators of vital activity and the suitability of the solution for organisms because it is a product of destructive metabolism and a means for the fixation of energy. If the carbonic acid-carbonate system is considered distinctly, it is conspicuous first that it requires time to equilibrate sea water with a gas phase containing CO₂. If acid or alkali is added to sea water, a number of hours aeration will be required before a constant pH is attained with any tension of CO₂.

Bubbling and shaking sea water with pure hydrogen slowly increases the pH to about 9.2. This is about the point found by Atkins and others as the limit for photosynthetic activity of several marine algae. After acidification and bubbling to remove CO₂, an alkali titration curve lags below the curve for a base of the same concentration as the excess base of sea water (0.0025 N). Such a curve shows the buffer effect of non-volatile buffers alone. A direct acid titration shows the combined buffer effect of both volatile and non-volatile weak acids.

Titration curves at various CO₂ tensions differ from those a - pure carbonate solutions in requiring more acid. Taking this difference as a measure of the buffer effect of other weak acids than carbonic, their apparent dissociation constant is calculated as k = n × 10^-8 L. J. Henderson's artificial sea water was 0.0015 M in boric acid, but boric acid is not reported so abundant in sea water. Silica is reported in a concentration sufficiently large to have this effect if it lxhaves like an acid as weak as boric. This point is not revealed with the existing data, hut there is suggested an important topic because of the presence of silicon in many organisms antl geological processes.

Depancreatized clogs form easily detectable quantities of lactic acid during strychnine convulsions. Even though these animals may be unable to oxidize carbohydrate, still the production of lactic acid seems intimately concerned with muscular contractions.

Several investigators have studied the effect of grinding upon raw starch. Schleiden 1 reported that grinding starch with twice its weight of water gave the mass a salve-like consistency. Kraemer 2 found that grinding with sand and Sponsler 3 that grinding dry in a pebble mill rendered part of the starch soluble in cold water. These observations were verified by the writers. Grinding potato starch for 74 hours with flints in a pebble mill and wheat starch for 122 hours injures all the grains, as may be noted with the microscope. About 60 per cent of each of the starches was then soluble in cold water. The solution could be filtered quite clear with paper pulp, was not viscous, and gave a characteristic blue iodine reaction. Apparently the starch grain is protected by a membrane or membrane network impermeable to the soluble substance. When the membrane structure is injured, the contents may be leached out with the greatest ease. Ground starches suspended in water and heated do not gelatinize, and no paste can be obtained. The viscosity of suspensions of ground starch determined between 20° and 95° by means of a Stormer viscosimeter was relatively little more than that of water.

Stiff starch pastes made by boiling 5 per cent of untreated wheat starch in water were ground cold in a pebble mill from 13 to 16 hours and their viscosity tested before and after grinding over a range of temperature from 50° to 90°. After grinding the viscosity was reduced to about one-fourth of its initial value.

Beijerinck 4 ground boiled potato starch with sand, thus rupturing the membranes. He found that this treatment rendered the starch in part soluble, 60 per cent passing into solution. This agrees very well with the observation above reported in regard to the solubility in cold water of ground unheated starch. The 'solubility of raw starch in cold water when ground indicates that the observation of Beijerinck does not depend upon depolymerization by heat and water but that soluble starch as such is present in considerable amount in the natural untreated starch grain.

All parts of the gastro-intestinal tract except the oesophagus are thrown into contraction, during typical anaphylactic shock in dogs. These contractions usually begin after a latent period of about forty-five seconds. They usually reach their maximum by the end of two and a half minutes.

The contractions vary in intensity in different parts of the gastro-intestinal tract. In the stomach, they usually increase the intra-gastric pressure about 10 mm. Hg. In the small intestines, a pressure increase of about 25 mm. Hg. is usually observed. In the colon and rectum, an increase of at least 50 mm. Hg.

Gastro-intestinal contractions similar to these are produced by the intravenous injection of histamine.

These contractions of the gastro-intestinal tract do not take place on travenous injection of specific foreign protein into dehepatized sensitized dogs. In contrast with this finding, dehepatization does not abolish the typical gastro-intestinal contractions to histamine.

The gastro-intestinal contractions in intact anaphylactic dogs are not secondary to the anaphylactically decreased arterial blood pressure. Rapid exsanguination, reproducing the characteristic blood pressure fall of canine anaphylaxis, does not throw the gastro-intestinal tract into contraction, at least during the period of our tests (five minutes). The gastro-intestinal contractions in intact anaphylactic dogs are also not secondary to local passive congestion. Ligation of the portal vein, producing a more extreme passive congestion of the gastro-intestinal tract than that observed during typical anaphylactic shock, does not increase the gastro-intestinal tone during the period of our tests. We, therefore, believe the gastro-intestinal contractions in intact anaphylactic dogs, similar to the urinary bladder contractions previously reported, 1 are caused by chemical products (hepatic anaphylatoxins), explosively formed or liberated by the anaphylactic liver, products having a histamine-like effect on the gastro-intestinal musculature.

The following is our conception of the fundamental physiological mechanism of canine anaphylaxis, and constitutes our present working hypothesis with other animal species.

We assume that the capillary endothelium is the initial point of anaphylactic attack and defense. We further assume that in dogs the different capillary systems have different susceptibilities to the anaphylactic insult, the capillaries of the liver having the greatest susceptibility.

On intravenous injection of the usual doses of specific foreign protein into sensitized dogs, the hepatic endothelium alone is unable to resist the anaphylactic insult. There is a sudden increase in capillary permeability in this organ, with the rapid passage of foreign protein and altered blood plasma into the hepatic tissue spaces.

The hepatic parenchyma, in consequence, is immediately thrown into exaggerated, possibly atypical functional activity. The chemical products from this activity pass rapidly into the circulating blood, on account of the increased hepatic capillary permeability. These products, acting in conjunction with the circulating foreign protein and altered blood plasma, break down the endothelial defenses in the extra-hepatic tissues, with the resulting rapid passage of atypical products intor the extra-hepatic tissue spaces.

This conception, we believe, fully accounts for the clinical picture of canine anaphylaxis. Thus, the suddenly increased local tissue pressure from hepatic edema with the resulting passive narrowing of the hepatic capillaries, coupled with the suddenly increased local blood viscosity from plasma loss, would account for the characteristic hepatic stasis. The suddenly reduced available blood volume from hepatic stasis and resulting gastro-intestinal passive congestion, coupled with the reduced blood volume from hepatic and extra-hepatic plasma losses, would account for the characteristic fall in arterial blood pressure.

Dyspnea from muscular exercise results when the increased respiratory demands can no longer be met without effort. In the present series of experiments the ability to meet these demands was artificially restricted by having a normal individual (A. W. H.) breathe through small apertures. For this purpose two perforated corks were used, the diameters of the bores being 8 mm. and 6 mm. respectively. The first of these caused no effort so long as the subject of the experiments was at rest. It caused moderate respiratory effort when stairs were climbed at the rate of approximately 80 per minute. The second caused slight effort at rest and considerable distress during stair climbing. The latter culminated in discontinuance of the exercise after two or three minutes.

During rest, the respiratory rate was not influenced by the milder obstruction but in some experiments was somewhat slowed by the more marked obstruction. During exercise, the respiratory rate was definitely slowed by each, average figures being: without stenosis 27 respirations per minute, with 8 mm. stenosis 21 respirations per minute, and with 6 mm. stenosis 17 respirations per minute.

During rest, the minute volume of respiration was not altered by the lesser stenosis but in some experiments became somewhat less with the more marked stenosis. During exercise, the minute volume was regularly reduced. After one to three minutes of exercise pulmonary ventilation became approximately constant at the following levels: without stenosis 52 liters per minute, with 8 mm. stenosis 46 liters per minute, and with 6 mm. stenosis 28 liters per minute. With the more marked obstruction, stair climbing had to be discontinued after 2 to 3 minutes on account of increasing distress. With the less marked obstruction this was not the case. Compensation for the restricted ventilation was here accomplished by increasing the percentage of O₂ absorbed from and of C0₂ given off to the respired air.

The data to date on the respiratory changes resulting from insulin injections are conflicting. It is generally agreed that insulin causes a rise in R. Q. when administered to normal subjects but just how this change is brought about is still a matter of discussion. Is the increase in quotient due to increase in CO₂, decrease in O₂, or both?

Normal healthy rabbits were used as the experimental animals in this series of experiments. All were carried out under the same conditions as to care, diet, and time elapsing between food and experiments. Great effort was made to obtain comparable results. The blood sugar curve coincident with the respiratory change was followed. This also served as a check on the potency of the insulin.

From the findings to date the following summary seems justified:

1. Insulin, when administered to normal animals brings about, first a slight decrease, then a decided increase in the R. Q. curve, reaching the peak two hours after injection and returning to the basal level four hours after.

2. A blood sugar drop accompanies the R. Q. change, the lowest blood sugar occurring at the peak of the curve. The return to normal is less rapid than the return to normal of the R. Q. though the curve is well on its return by the end of the four hours.

3. There is, in the second hour, both a decrease in the 0₂ and an increase in the CO₂, both changes tending to increase the R. Q

4. The 0₂ consumption and the heat production would indicate that the total, metabolism is not markedly increased.

In the hope of finding a more favorable subject than the depancreatized animal for study of the influence of insulin administered orally, twelve dogs were used after the establishment of the typical D:N ratio of Lusk. Insulin produced in the laboratory by the amyl alcohol method, 1 administered subcutaneously, caused not only a sharp decline in the excretion of sugar (and nitrogen), but also a rise in respiratory quotient from diabetic level (0.69) to as high as 0.85. Sugar was administered by stomach both in control and in insulin periods. In several instances after a dose of 1 R. U. per two kilos dog the R. Q. rose within the first two hours only to 0.77 or 0.78 and immediately dropped back to diabetic level. The effect on the excretion of nitrogen was sometimes parallel with and in a few instances greater than the effect on the excretion of sugar.

Insulin was given orally in several different combinations: in cod-liver oil emulsion (no effect); in oleic acid suspension (no effect); in blood serum (some effect). The best results (rise in R. Q. to 0.77 or 0.79) were obtained after the use of 1 R. U. per kilo in the form of an enteric coated tablet containing insulin and malic acid. 2 The effect on the excretion of sugar was not so marked even with this tablet as after subcutaneous administration.

There is evidence of retention of sugar as well as of combustion in some of the experiments both after subcutaneous and after oral administration. Whether the effect on combustion is direct or indirect these experiments do not permit us to say.

Numerous experiments on some twenty-five dogs and half a dozen cats have been performed in the attempt to demonstrate clearly the absorption of insulin from the alimentary tract. This is not difficult when the insulin is placed directly in the intestine. A sharp drop in the hyperglycemia of ether anesthesia is easily obtained. But when insulin is given by mouth to animals treated with ether or adrenalin or both the demonstration often fails because when it is adequately protected against the stomach, absorption of the insulin from the intestine is much retarded and controls are difficult to establish. An interval of at least five days is necessary in order to insure a return of the glycogen storage to normal, and after half a dozen treatments with ether for several hours at a time there seems to be permanent impairment of the capacity to store glycogen. Alcohol up to 20 per cent and in amount sufficient to thoroughly intoxicate does not materially improve the absorption.

Some advantage clearly was gained from the administration of insulin both in solution and in enteric coated tablets, combined with sodium oleate.

An ascospore stage in the life history of the organism of blatomycosis has not hitherto been recognized. In fact as great a mycological authority as Vuillemin 1 has included these organisms in the provisional genus Cryptococcus in which asci have never been observed.

These distinctive forms occurred exclusively in the so-called secondary colonies of the cultures, which also contained “dauernzellen” and pigmented oidia. The prolonged viability of the cultures is associated with these secondary colonies, while the primary non-pigmented growth derived from them is seldom viable after a month.

Certain forms pictured in the tissues by Wade 2 are quite comparable with these forms of special resistance in the cultures and form a plausible explanation of the great difficulty of eradication of the disease by therapeutic measures. Inasmuch as oidial stages are common to so many fungi, it may be in order to suggest allocation of the so-called oidiomyces (blastomyces hominis) with the ascomycetes.

Haldane 1 has shown that if large doses of sodium bicarbonate (0.6 gm. per kg. body weight) are fed to normal subjects receiving normal diets the metabolism of glucose is depressed and acetone bodies appear in the urine. An experiment was planned to determine whether an increased excretion of acetone could be induced by moderate doses of the alkali in an arthritic subject who was receiving a diet which furnished sufficient calories for her needs (the patient gained in weight during the period of study) but which contained such small amounts of carbohydrate as just to prevent any significant ketosis. A markedly increased excretion of acetone was induced by 1.3 gms. (20 grains) of sodium bicarbonate fed in three equal doses after each meal. The increased excretion lasted as long as the drug was continued (August 11 to 18). The diet was approximately constant throughout the experiment except on the 17th and 18th of August, when the carbohydrate was somewhat decreased through an error. After this experiment part of the fat in the diet was replaced by a synthetic fat containing fatty acids with an odd number of carbon atoms (intarvin). A reduction in the amount of acetone excreted was found on the 22nd and the 23rd, as would be expected from the work of Kahn. 2 The carbohydrate content of the diet was then slightly decreased, and again intarvin was substituted for part of the fat (August 28) and an even greater decrease in the excretion of acetone than before was found. In this experiment on a subject maintained at the borderline of ketosis therapeutic doses of sodium bicarbonate caused an increased, and the synthetic fat intarvin a decreased excretion of acetone.

It has been shown by Cori, Cori and Goltz 1 and by Frank, Nothman and Wagner 2 from the Minkowski Clinic that by comparing the sugar concentration in the femoral artery and femoral vein of rabbits starved for 24 hours, an increased disappearance of sugar from the blood into the muscles can be demonstrated during insulin action. In view of these results it seemed desirable to investigate whether animals receiving glucose plus insulin would not show a stronger intake of sugar by the muscles than animals which received the same amount of glucose but no insulin. As a standard dose 5 gm. glucose per kgm. was given by stomach tube to rabbits previously starved for 48 hours. The insulin was injected intravenously in a dose sufficiently large to prevent a rise in bloodsugar after the glucose ingestion. Blood samples were taken simultaneously from the femoral artery and femoral vein of unanesthetized animals in half hour intervals, 1 to 6 blood samples before the experiment was started and 4 to 8 blood samples after giving either glucose alone or glucose plus insulin. In order to minimize the individual variability of the test animal several comparative experiments, first with sugar alone and then with sugar plus insulin, were performed on subsequent days on the same rabbit. In another group of experiments this was reversed so that first sugar plus insulin and then sugar alone was given to the same rabbit. The data are summarized in Table I. The figures represent the difference in sugar concentration between femoral artery and femoral vein that were observed at the time intervals noted after giving either glucose alone or glucose plus insulin. The figures of each vertical column represent one individual experiment. An average difference between arterial and venous blood sugar for each observation time was calculated from the horizontal columns. It will be noted that the average differences fo'r the series with glucose plus insulin are for all observation times greater than those of the series with glucose alone. This would mean that more sugar disappears from the blood into the muscles in the former series than in the latter, or that rabbits receiving insulin plus glucose show a larger intake of sugar by the muscles than rabbits receiving glucose alone.

The insulin content of the pancreas, liver, blood, and skeletal muscles of starved and completely phlorhizinized cats and of normal cats, starved for the same length of time has been determined. The Doisy, Somogyi and Shaffer process with only slight modifications was used for the preparation of insulin, and the insulin obtained was tested on mice. From organs other than the pancreas very small amounts of insulin were obtained. The yields have a comparative value only and have been expressed in mouse units, since it did not seem justified to calculate these very small amounts of insulin in the commonly used rabbit units. The amount of insulin injected intraperitoneally which lowers the blood sugar of a mouse (previously starved for 24 hours, weighing 18 to 22 gm. and kept at a room temperature not below 24° C.) in one hour to 0.038 to 0.044 per cent has been considered as a mouse unit. From the following table it may be seen that a lack of insulin production of the pancreas or a lack of insulin fixation in the tissues cannot account for the incapacity of phlorhizinized animals to metabolize sugar.