Experimental Biology and Medicine

In most breeds of fowl there is a well marked sexual dimorphism in the color and structure of adult feathers. Experimental or pathological alteration of the endocrine balance has been shown 1 to produce striking reversals and modifications in these characters. Our problem has been to find a new method of attack which would yield further information on (a) the extent to which the output of a feather follicle is controlled by the endocrine relations which obtain at the time the feather is produced, and (b) the extent to which it is determined by characteristics which are inherent in the follicle itself. The procedure has been to transplant pieces of skin from one newly hatched chick to another and, some months later, to compare the feather production of the donor, the host and the skin originally transferred from the former to the latter.

The technique is simple. Young chicks were found to take the anesthetic (ether) very satisfactorily. They recover quickly and do not suffer from infections when moderate care is exercised. The grafts are easily held in place by interrupted silk sutures, which usually fall out spontaneously, but may occasionally remain and serve as good landmarks over a period of several months. The general results obtained thus far may be briefly summarized.

1. The percentage of initial “takes” is high, but many grafts after having become well established are subsequently lost. Such grafts often appear entirely successful at first, and produce normal pin feathers or even fully developed feathers. The cause of these failures has not been studied, but the distribution of cases with reference to sex and breed suggests varying degrees of incompatibility between donor and host. This type of experiment promises to be very favorable for the study of differences in tissue reactions and “individuality differentials.”

The question of a relationship between body weight and blood pressure may be studied from the findings in large numbers of individuals of various sizes essentially normal, particularly as regards the cardiovascular system and nutritional status. Children afford the best material for such a study. In them a large range of body size is available, and arterial disease is so rare that in group studies it can safely be disregarded as a disturbing factor. Satisfactory tables of normal weight or height are at hand for the exclusion of nutritionally unsuitable cases. By utilizing group averages the effects of temporary or accidental variations in pressure, of observational error, and of individual variations in body structure or composition can all be minimized.

The following study is based on observations of about 1,000 approximately normal children, between 4 and 16 years of age, mainly from kindergartens and public schools. By plotting individual blood pressures against body weights∗ it was found that the points of intersection roughly followed a curve of general parabolic form. Dividing the cases into twelve groups by age (without reference to sex) average weights and corresponding systolic and diastolic pressures for each group were determined, with the results shown in Table I. The logarithms of weight when plotted against the logarithms of pressure followed an approximately straight line, from which it was inferred that the relationship between weight and pressure could be expressed by the general formula y=ax^b, or S/D=aW^b. Solving for the six pairs and averaging, the following values for b were found: b_s=0.2055, b_D=0.1900. Assuming a uniform value of 0.2 for b, the average for the constant a was found to be a_s=511.725, or approximately 51.7, a_D=34.730, or approximately 34.7. The formulas derived for systolic pressure and diastolic pressure are therefore S=51.7 W^0.2, D=34.7 W^0.2, and by difference P=17.0 W^0.2, in which S is systolic pressure, D is diastolic pressure, P is pulse pressure, all in millimeters of mercury; and W is body weight in kilograms.

The basis of these experiments is a study of the so-called Koch phenomenon elaborated by Krause 1 on allergy in experimental tuberculosis. The early anatomical response of the tissues of a susceptible animal inoculated for the first time with tubercle bacilli differs from that of a reinfection. In the susceptible guinea-pig a well circumscribed nodular tubercle will develop slowly after a week or more, without any of the signs of inflammation. Only later, after 3 or more weeks, does redness appear around the tubercle, associated with changes that are indicative of a progressive lesion. At this stage the animal has become allergic, i. e., it begins to react in a new way because of the changes set up in the body by the tubercle bacilli. The susceptibility of the animal has been altered. If such an animal now receives an intracutaneous inoculation of tubercle bacilli, the response by the tissue is rapid and local inflammation occurs. Instead of well localized proliferation, diffuse exudation takes place. The inflammation occurs rapidly and subsides as rapidly, and the nodular tubercle now present retrogresses and may disappear completely. All of these changes are contingent, however, upon moderate dosages of tubercle bacilli.

In the first group of experiments, 5 series of normal guinea pigs, 3 animals in each series, were used. Subcutaneous injections of the different materials were given weekly in increasing doses over a period of 10 weeks, the initial dose ranging from 0.1 to 0.2 cc., and the final dose from 2.0 to 3.5 cc. Four weeks after the last injection, the animals were inoculated intracutaneously with 0.4 mg. of a moderately virulent strain of human tubercle bacilli. One group received tuberculin fractions∗; another, tuberculin fractions and toxin filtrates∗ injected on alternate weeks; a third group received toxin filtrates alone; a fourth, the same toxin filtrate heated for one hour at 75° C.; and the fifth untreated group served as a control.

The rôle of the gonadal-stimulating hormone of the anterior pituitary in the development and regulation of the genital system has been previously reported. 1 , 2 That the same stimulation of the ovarian follicle can be effected in ovaries which have been grafted into both normal and castrated males is here reported.

A series of 4 male rats, 96 days of age, may be taken as a typical group. Two of these were castrated and the ovaries of 20-day-old rats were transplanted into the recti abdomines and anchored with silk sutures. Into the other 2 unoperated males, ovaries of 20-day-old rats were transplanted, one ovary into each testis. Eight daily transplants of the fresh pituitary of an adult rat were then given to one of the castrated and one of the normal hosts. The other pair received no transplants. The rats were autopsied on the ninth day of the experiment and the grafts removed and sectioned. The ovarian grafts in the recti muscles of the castrated animals were well vascularized, but considerable degeneration, evidently due to an inadequate circulation, had occurred in the untreated animal. In the graft of the animal which had received the pituitary transplants many primary follicles had disappeared, but a score of large follicles, of mature condition and with apparently normal ova, were found.

In the untreated animal of the second pair of this series, the transplanted ovary had been resorbed. In the treated animal there was a typical bilateral ovariotestis. The vascularization was unusually good, and the follicles, both primary and secondary, were numerous. Many large, apparently normal follicles in the graft were like those in a normal adult ovary, with little atresia of the follicles or ova.

The daily transplantation of the mammalian anterior pituitary into the adult albino mouse or rat results in the development of a greatly increased number of ovarian follicles, 1 and is followed by super-ovulation. The number of ova liberated by one ovary varies from 20 to 48, as determined by a study of serial sections of the uterine tubes. 2

Daily intramuscular transplantations of 2 mouse pituitaries into adult female mice, beginning the first day of the dioestrum, resulted in mating within 12 hours after the second treatment. The animals were killed on the ninth or tenth day of pregnancy. The number of uterine implantations found on the ninth or tenth day was between 19 and 29. Recovery of normal fetuses at this time has not been constant, because of failure of fertilization or implantation in some cases. Another group of similarly treated animals will be permitted to go to term.

Whether or not all glycogens are identical has been investigated repeatedly, 1 but no general agreement has been reached. The report by Samec and Isajevie 2 that dog-liver glycogen contains 0.721% P₂O₅suggests the possibility that the question of the identity of all glycogens may be solved by ascertaining their respective phosphorus contents. The purpose of the present investigation was to compare the phosphorus content of Mytilus glycogen with that of the glycogen examined by Samec and Isajevic.

Samec and Isajevic give no information concerning the method by which their glycogen was prepared. The Mytilus glycogen used in the present investigation was prepared by the method of Pflüger, as modified by Starkenstein and Henze. 3 In addition, it was repeatedly precipitated from slightly acidulated solutions. Difficulties were encountered in filtration so that the material was unavoidably exposed for a long time to strong KOH. The glycogen finally obtained had a P₂O₅ content of 0.1168% as determined by the method of Embden and an ash content of 0.25%. It was white and contained iron, calcium and phosphorus.

Certain investigators have obtained glycogen free from ash. The attempt was, therefore, made to remove the ash of Mytilus glycogen by electrodialysis with ultrafiltration in a Bechold-König apparatus, using a 10% acetic acid-collodion solution for the membranes. Dialysis was continued until the readings of the ammeter connected with the apparatus were constant, which took 4 to 5 hours. The dialyzate contained phosphorus and calcium. Samec and Isajevic also electrodialyzed glycogen and found that it separated into a sol and a gel layer, the material in the sol containing more than five times as great a percentage of P₂O₅ as the material in the gel. Mytilus glycogen did not separate into two layers in this way, perhaps because during dialysis the direction of the current was frequently changed. The solutions obtained could not be precipitated with alcohol until after the addition of a few drops of a 1% ammonium acetate solution.

Van Slyke and Neill 1 have recently introduced an apparatus for the determination of blood gases by manometric methods. This manometric blood-gas apparatus has been used by Van Slyke and Sendroy 2 in determining blood Ca. The Ca is precipitated as the Department of Anatomy, Cornell University Medical College, xddition of acidified KMnO₄. The oxalate is oxidized to CO₂ and H₂O, and the CO₂ so produced determined in the usual way.

A modification of the Clausen method for the determination of small amounts of lactic acid was described recently by Friedemann, Cotonio, and Shaffer. 3 These investigators found that the use of manganous salt increases the rate and yield of aldehyde when lactates are oxidized by acid KMnO₄.

The writers have developed a method for the determination of lactates using the manometric apparatus of Van Slyke and Neill, and in part the modified Clausen reagents of Friedemann, Cotonio, and Shaffer.

In selecting a mixture of MnSO₄ and KMnO₄ that would fully oxidize lactic acid to acetaldehyde and CO₂, an excess of MnSO₄ was found to prevent the oxidation or greatly to retard the process and this was also true for equal amounts of MnSO₄ and KMnO₄. When excess KMnO₄ was present good results were obtainecf over a considerable range of ratios of KMnO₄ to MnSO₄ as shown in Table I. Shaffer gives the oxidation-reduction potential necessary to oxidize lactic acid as +1.36 at 25° C. but carries out oxidation at the boiling point, i. e., at an Eh of approximately +1.74. The Eh of our oxidizing agent was about +1.44 in the concentration of sulphuric acid employed, namely Normal and at 20°.

Procedure: The apparatus was the closed manometer type of Van Slyke gas apparatus. Reagents consisted of 5N H₂SO₄, M/10 KMnO₄, M/10 MnS0₄, Standard Lactic Acid, 5 N NaOH.

It is suggested that inconsistencies in the work of previous investigators on the blood of the normal guinea pig have been due, for the most part, to the use of the usual fixed staining technic. The supravital method of blood study, first used by Pappenheim, 1 and reintroduced by Simpson, 2 overcomes the common difficulties of cell distribution, and identification, as met with in the smear method.

The peripheral blood of the normal guinea pig has been studied by the Neutral Red-Janus Green supravital method. Sixty-four guinea pigs from four different sources, of both sexes, and weighing from 500 to 1000 grams, comprise the series. From 200 to 400 white cells per preparation were counted on each. To compensate for normal fluctuations, counts were made at all hours of the day and night.

The normal guinea pig has been found to have a leucocytic formula peculiarly its own, especially in that an atypical mononuclear is normally present in the circulation. In our counts this will be termed the Kurlow cell, because of its inclusion, the so-called Kurlow body. Considerable evidence has been obtained upholding the belief that the Kurlow inclusion is the result of a symbiotic protozoan.

An average total leucocyte count of 9,600, and a total erythrocyte count of 5,565,000, were established. The differential count showed the following: Neutrophile, 34.9%; Lymphocyte, 49%; Monocyte, 7%; Eosinophile, 3.1%; Basophile, 0.8%; Kurlow cell, 5.0%; Unclassified cells, 0.2%.

It is believed that the above data establishes a fairly accurate means of ascertaining the normalcy of a fully matured guinea pig.

The symptoms following the rapid absorption of tuberculin in the tuberculous guinea pig, in which the subcutaneous inoculation of small amounts of tuberculin invariably prove fatal, are well known.

Fischel 1 has suggested that the severe tuberculin reaction can be mitigated by introducing tuberculin into the skin with a view to retarding absorption and slowly removing the toxic factors elaborated upon the contact of tuberculin with tuberculous tissues. It had occurred to us that if conditions could be set up subcutaneously which would retard absorption of this toxic factor, and at the same time permit the use of larger amounts of tuberculin, some therapeutic results might be developed, without the danger of fatal issue.

To produce these conditions, a local anesthetic, hydrochloride of p-amino-benzoyl-di-methyl-amino-methyl-butanol or by its trade name, the “Tutocain” brand of Butamin, which induces an edema and at the same time renders it painless, was selected. As a rule 4 cc. of a 1% solution was injected subcutaneously. The guinea pigs had been rendered tuberculous by the subcutaneous inoculation of 1 cc. of a suspension of tubercle bacilli (virulent strain No. 427) into the inguinal region 4 weeks previous to the tests. Animals were subjected to an intracutaneous tuberculin test before they were employed in the experiments, only animals showing a strongly positive reaction being used.

One half hour after the application of the drug 0.4 cc. of old tuberculin was injected subcutaneously into the infiltrated area. Previous tests had shown that 0.2 cc. of the same tuberculin was sufficient to cause death of tuberculous guinea pigs within 24 hours.

Guinea pigs, from numbers 1 to 10, were autopsied one week after the test inoculation and the extent of the infection noted. All guinea pigs showed extensive and typical tuberculous lesions in the liver, spleen, lymph nodes, and lungs. The remainder of the guinea pigs were kept under observation for subsequent tests.

Widely conflicting opinion exists with regard to the effect of anesthetics on the resistance of red blood cells to hypotonic saline solutions. Since little of this is based on quantitative data, it was thought worth while to restudy the problem, using the quantitative method introduced by Simmel, 1 in which actual count is made of the number of cells remaining after uniform periods of contact with different grades of hypotonic solutions. The results are conveniently expressed as percentages of the total number of erythrocytes per cubic millimeter, where Hayem's solution has been used as the diluent. The normal variations likely to be encountered by this method have been studied by Leake and Pratt. 2

Dogs were used in all experiments. Blood was drawn by syringe from the femoral artery, through the intact skin. Ether or chloroform was administered by the drop method, and further samples of blood were drawn at various intervals up to one hour after the induction of the surgical stage of anesthesia. These were prepared for determination as soon as drawn.

After about 10 minutes of ether anesthesia, an apparent increased resistance was noted on the part of the erythrocytes to the hemolysing action of the different strengths of hypotonic Simmel's fluid. This may have been due either (a) to destruction of weaker cells, leaving only the more resistant ones in the circulation, or (b) to an influx of cells, from such reservoirs as the spleen, resulting from struggling or partial asphyxiation during induction. Studies are in progress to determine which is the more likely. After 30 minutes of ether anesthesia, the osmotic resistance of the red cells was much below the normal level, indicating injury to the stroma of all cells following prolonged contact with the ether.

Nitrous-oxide or ethylene in combination with oxygen were administered to dogs by the rebreathing technique developed by Dr. Ralph M. Waters, in which the carbon-dioxide produced by the body is removed by soda-lime. Nitrous-oxide with oxygen caused a slight reduction in the osmotic resistance of the red cells to hypotonic Simmel's fluids, which even after an hour's anesthesia, was not nearly as marked as that noted under ether or chloroform. Ethylene with oxygen had about the same effect on the erythrocyte fragility. When ethylene was pushed to asphyxiation after 20 minutes without oxygen, the osmotic resistance of the red cells was reduced to a considerable extent in the lower dilutions of the hypotonic Simmel's fluid (0.4%), but in the higher dilutions (0.7% and 0.6%) it was only slightly below the normal levels.

A mixture of 25% carbon-dioxide with 75% oxygen was administered to dogs without rebreathing. This greatly increased the amplitude of respiration and induced anesthesia with abolition of the corneal reflex in 6 minutes. The pulse was greatly retarded. The mucous membranes remained pink. Clonic convulsions initiated by tremors came on within 10 minutes after induction, and violent tetanic convulsions quickly followed. Upon administering pure oxygen the convulsions promptly stopped, the pulse became very rapid, and the animals recovered. In all cases, except in the weaker dilutions (0.4%), the resistance of the erythrocytes to hypotonic Simmel's solutions was greater than normal.

After the administration of CO₂, much difficulty was experienced in obtaining blood because of the rapid clotting. Hickman noted this action of carbon-dioxide a century ago. Dr. Cline Chipman of Washington, D. C., has observed clinically that a few breaths of carbon-dioxide administered during the anesthetic period will greatly shorten the coagulation time of blood. This has been confirmed by Dr. R. M. Waters of Madison, Wis., and by Dr. M. E. Botsford and Dr. M. J. Price of San Francisco. Such an effect of carbon-dioxide may be of great clinical value in certain cases of capillary oozing at operation. Some phases of the relation of the CO₂ content of blood to its coagulation time have recently been studied by Baumberger. 1

The species of the lower families of termites, so far investigated, have each a characteristic hindgut fauna of flagellate Protozoa, belonging to the orders Hypermstigida and Polymastigida. All individuals of a given termite species over wide geographic ranges present, so far as known, a similar grouping of species. This raises the question whether we are dealing here with specific protozoan faunae, whose species are able to exist only in the one specific termite host, with which they have established obligatory physico-chemical relations, or whether the faunae are distinct because of the isolated habit of life of the termite colony.

A colony of Porotermes froggatti received in normal condition, but which an examination of nearly 100 intestines showed to have lost its protozoan fauna, supposedly due to conditions encountered in transit, was used. Various methods were resorted to in attempts to refaunate the Porotermes individuals with the Protozoa from species belonging to other genera of termites.

All attempts to accomplish this by keeping Porotermes individuals in close association with Termopsis augusticollis, Neotermes malatensis and Reticulitermes hesperus proved failures.

Feeding defaunate Porotermes individuals with the intestinal contents of the 3 species named failed to accomplish transfaunation in most cases. In a few instances Protozoa survived for a few days, polymastigotes in all cases. No signs of increase of numbers or permanent establishment were noted.

All attempts to infect Porotermes indlividuals by applying to the anal aperture a drop of intestinal content of the infective species were failures.

Finally resort was had to injection by pipette through the anal aperture. For these experiments Termopsis was used as the source of infective material. Several cases of persistence for several hours and even a few days were observed, and finally with the last individual what seemed to be successful transfaunation was obtained.

There has been a growing conviction among physiologists and pharmacologists that the reaction of an organ to stimulation will vary, within certain limits, with the state of tonus of the organ at the time of excitation. This conviction applies to drugs involving the autonomic nervous system, but by no means confined to them. A good part of the work has been done on strips of extirpated muscle, or the nerves of the organ have been cut, or numerous drugs influencing the autonomic system have been used in succession, so that, the results must be interpreted in the light of altered physiologic status.

Our work was done on a series of normal dogs trained to lie perfectly quiet under no restrain. A fine balloon at the end of a duodenal tube was passed by mouth (the dogs offer very little resistance after the first 3 or 4 times) and the duodenal tube was connected to a water manometer. The balloon was inflated with 100 cc. of air and the fluctuations of the manometer levels were recorded on a slowly revolving Kymograph. After a suitable control period an injection of 5 m. of 1:1000 Parke Davis Adrenalin was given intracutaneously, in the leg, forming a wheal. (Intracutaneous injection was chosen because a mild relatively prolonged action was desired.)

Controls. The animals' gastric functions were established over a period of one to two weeks under daily observation. They showed: (1) That hunger contractions in a starving animal last from 30 to 50 minutes, and periods of active motility are separated from each other by periods of relative quiescence. (2) These periods of quiescence last from 1 to 1½ hours. In a few instances, however, they were as short as 45 minutes. (3) There was at no time a complete relaxation of the stomach in a starving animal.

An analysis of the contraction curves of the human gall bladder, constructed from cholecystograms taken at short intervals after a meal of egg yolk and cream, indicates that the average gall bladder discharges approximately ¾ of its volume in the first half hour after each meal. The first phase of contraction in which this is accomplished, is usually divided into three parts: the “initial response” (usually of 2 minutes duration), the “2-minute pause”, and the “principal period of discharge”. On the basis of Elman and McMaster's work on the sphincter of the dog, 1 the sudden discharge of the first 2 minutes is interpreted as being due to the opening of the sphincter; the interruption of emptying during the next 2 minutes to the initial increase in resistance offered by the sphincter to the flow of bile (as described by these authors); and the subsequent “principal discharge” to the contractile force of the gall bladder which, after slowly getting under way, is able for a time to overcome the growing resistance at the outlet of the common duct. That even the “initial response” (after ingestion of fatty foods) is partly due to contraction and not wholly to elastic recoil, is indicated by the change in shape which the human gall bladder undergoes during these first 2 minutes—that shape being then assumed which is characteristic of the rest of the first phase of contraction. (See also Ivy's demonstration 2 of the contraction and evacuation of the gall bladder caused by the intravenous injection of highly purified “secretin” preparations.)

Experiments designed to verify this action of the sphincter, show that stimuli other than the ingestion of food may provoke this “initial response”. When a patient, for instance, is suddenly subjected to the odor of fried bacon, there ensues a momentry discharge of bladder bile (1 to 3 cc.)—cornparable in duration to the initial drop in resistance of the sphincter described by Elman and McMaster, for in 2 minutes the gall bladder begins to fill again, in spite of the fact that the patient continues to smell food.

The production of cholecystitis by intravenous injection of a solution of chlorinated soda (Dakin's solution), was first reported by Mann. 1 He found that from 5 to 12 cc. per kg. body-weight, of a solution of chlorinated soda, having an available chlorine content of at least 0.48%, would produce a cholecystitis in dogs in 12 to 24 hours following the injection.

Since Dakin's solution seems to act specifically on the gall-bladder, its injection might prove fatal if the gall-bladder were removed, and, if so, the gall-bladder must act as the specific organ of elimination or detoxification. Such an assumption is justified by the observation of Mann, which we have confirmed, that the solution reaches the gall-bladder via the blood stream and not the bile.

Eleven normal dogs were injected intravenously with 10 cc. per kg. body-weight of a solution of chlorinated soda (Dakin's solution) containing 0.49% available chlorine. One dog died within 24 hours and the autopsy showed a cholecystitis. The other 10 were etherized after from 40 to 48 hours, and necropsies performed immediately. Of these, 3 showed normal gall-bladders, the remaining 7 all had cholecystitis of varying degrees of intensity. Several of these dogs had small hemorrhages beneath the serous coat of the liver adjacent to the gall-bladder; one having a large area of hemorrhage, 5 cm. × 2.5 cm.

Four healthy dogs on which cholecystectomy had been performed 3 weeks previously were then injected intravenously with 10 cc. per kg. body-weight of Dakin's solution containing 0.49% available chlorine. Two hours after the injection all the dogs were noticeably uncomfortable, and 3 of them were vomiting. All of these dogs were alive and apparently in good condition 3 days after the injection.

It was previously reported by one of us 1 that complete picture of uremia in all its clinical, pathological and chemical details could be produced by the injection of hypertonic salt solutions into animals already made acid by the absorption of autogenous edema fluid. In these animals, although the kidneys remained histologically normal, there were evidences of degenerative changes in the liver. It was thought possible that the proteins in the urine in these experiments originated in the liver, and were the exciting cause of the renal changes which afterwards made the kidneys permeable to blood proteins. It is well known that foreign proteins injected into the blood are excreted almost quantitatively and that after a certain period the kidneys are thus rendered permeable to blood proteins as well.

Proteins from these artificial uremias and also those in the urine of dogs with ether nephritis were studied. Rabbits were sensitized to dogs' blood proteins by the subcutaneous injection of dog serum in the following series of doses: 1, 2, 4, 6, 8, and 10 cc. at 3 day intervals. The rabbits were sacrificed on the 21st day and bled. Sera thus obtained will react to dog blood proteins in dilutions of 1/1,000,000.

Urines were collected from our experimental nephritides in small 15 minute samples. Each sample was tested first for traumatic blood, which often spoiled the experiments. Next, all blood-free samples were tested for albumen and on all positive specimens, were tested against the serum. These tests were carried out both on whole urines and on urinary proteins, purified by dialysis and precipitation by ammonium sulphate saturation. From the earlier samples in a large series of experiments over 100 specimens were found in which the urinary proteins were free from serum proteins.

Puppies subjected to common duct ligation and division, develop, in the state of late obstructive jaundice, a marked osteoporotic condition, in which are revealed roentgenographically cortical thinning, relatively wide medullary spaces, and a lack of film contrast generally. A marked lowering of the blood serum calcium accompanies this process. A slight lowering of the blood calcium first manifests itself after about 30 days of obstructive jaundice. Thereafter it is fairly uniformly progressive until it reaches values of between ⅔ to ½ the normal. Growth of skeletal structures, especially of the long bones, and a substantial increase in body weight occurs in the young animal with obstructive jaundice.

The normal blood calciums that obtain for the adult animal throughout the course of obstructive jaundice, where the same serious digestive disturbances arise, would speak against any defect in the capacity of the intestine to absorb calcium. The maintenance of normal blood serum calcium values in the early stages of jaundice in the puppy may be due to mobilization, i. e., bone lysis. The four intrinsic factors in the production of faulty calcification are thought to be: first, small calcium storage; second, the presence of a disturbed metabolic condition unfavorable to osteogenesis; third, an intestinal factor characterized by the absence of some catalytic agent in bile; and lastly, new bone production which exceeds the rapidity with which lime salts are deposited.

The complete report of this paper is in press in the Archives of Internal Medicine.

Harwell, Hoguet and Beekman, 1 in an experimental study of intestinal obstruction, produced obstruction of the duodenum in a series of dogs, and found that “the 3 dogs which lived the longest were all given saline solution subcutaneously.” They injected from 150 to 300 cc. per day. In recent years numerous reports of the beneficial action of saline solution in gastric tetany and intestinal obstruction have appeared. We have undertaken a study of the effect of large quantities of saline solution on the length of life of dogs after the production of an experimental obstruction by the “closed loop” method introduced by Whipple and his associates. 2 We chose this method because it is a very uniform procedure, produces very characteristic results, and has been used widely by experimenters in this field. We have injected large quantities of normal saline solution, 500 cc. for each 7 kilos of body weight, 5 or 6 times a day subcutaneously, to provide fluid and chloride lost by vomiting and to cause a continuous diuresis. By this method we hoped to keep the dogs alive until the “closed loop” became relatively sterile, 3 or some other adaptation might occur.

Under ether anesthesia and using strictly aseptic technique, a “closed loop,” 10 to 12 inches long, of the first part of the jejunum was made beginning 5 or 6 inches below the fixed portion of the duodenum and the continuity of the intestine was reestablished by an end-to-end anastomosis.

The length of life of the untreated dogs confirms the findings of Dragstedt, Moorhead and Burcky, 3 and others, namely that such dogs die in from 24 to 96 hours.

At the present time we have results on six treated dogs, which show that the treatment has proven efficacious. Dog 1 lived 86 hours and possibly should be excluded because of rupture of the loop at the line of closure.

This paper is part of an investigation of the influence of high and low protein diets on basal metabolism, blood chemistry, and the urinary nitrogenous compounds. For a period of 37 days 6 apparently normal women, all engaged in laboratory work, received a high protein diet containing 2 gm. of protein and sufficient carbohydrate and fat to make 40 calories per kilo of body weight. An interval of 3 weeks was then allowed in which the protein intake was lowered until it reached 0.5 gm. per kilo. Although the carbohydrate and fat intake on the low protein diet was adjusted so that the caloric value of the two diets remained the same, all subjects lost weight during the first 10 days. The protein value was then increased to 0.6 gm. per kilo and the weight remained stationary. The low protein diet was continued for 30 days.

All foods were carefully weighed. On 4 days of each week basal metabolism was taken, specimens of blood and urine were collected, and samples of foods were saved and analyzed. Blood was drawn from the cubital vein without turniquet before breakfast, after 30 minutes rest, and was analyzed for sugar, total non-protein nitrogen, urea nitrogen, creatine, creatinine, uric acid, and lactic acid. Total non-protein nitrogen was determined by Koch's method and lactic acid by Clausen's procedure modified by Friedemann. Folin's methods were used for all the other constituents. All determinations were made in duplicate.

No consistent differences were found in blood sugar levels on the two diets. In 3 subjects the blood sugar showed a rise on the low protein diet, but in the other 3 it remained the same on both diets. The average figures for 6 subjects for the 5 weeks on the high protein diet ranged between 86.9 and 80.9 mg. per 100 cc. of blood, while the corresponding values for the low protein diet were 91.6 and 83.5.

For experiments of short duration (6 to 8 hours), total evisceration of a dog may be done successfully by the following simple method, which brings about nearly the same conditions as hepatectomy without the complicated series of operations leading up to the final stage.

Under very light ether anesthesia an incision is made from the xiphoid to a point half way from the umbilicus to the symphysis. The pelvic colon is clamped, ligated and cut, the ligature including enough mesentery to stop the inferior mesentery vessel. The distal end is not inverted. The entire intestinal mass is then delivered and a thin transparent mesentery can be cut, without clamping, to the level of the pancreas, where another set of vessels is met and clamped, the line of cleavage being close to the aorta. The choledochus, hepatic vessels and portal vein are next clamped en masse and the cardiac end of the stomach transfixed and ligated.

The entire mass of viscera can then be rapidly cut loose with one sweep of the knife. A heavy mass ligature is then thrown about the liver pedicle getting all the veins en masse. The three pedicles are ligated and the abdomen closed with a running stitch including all the layers. The procedure, with practice, may be carried out in 8 to 10 minutes, and the animal is left without liver, stomach, intestines, pancreas or spleen. In the few hours of life left, autolytic products of the ligated liver are not absorbed, as its capsule is untouched and the circulation entirely destroyed. The anesthetic is stopped after five minutes of operating, and glucose injections begun at once to counteract the rapidly ensuing hypoglycemia.

During the course of previous experiments, we have reached the conclusion that the autonomic reorientation underlying the production of fever involves splanchnic stimulation (parasympathetic status) with peripheral inhibition (sympathetic status). 1

During the time of chill the temperature of the muscle remains unaltered, or may be diminished at a time when the rectal temperature is rapidly increasing. 2 With this in mind we have carried out a number of observations on dogs which have been eviscerated according to the method described by Andrews. 3

In our previous experiments we have made use of a continuous injection of a dilute suspension of B. coli. When so injected into the anesthetized animal, a short latent period is followed usually by the onset of a severe chill, which may last either continuously or intermittently for several hours.

In eviscerated animals which we have similarly injected (after the period of recovery following the operation) there was no evidence of chilling, nor was there an increase in rectal temperature. Of course these animals cannot be considered normal because they have a low CO₂ combining power and a very low K/Ca ratio.

We believe that the experiments offer additional evidence that the response of the splanchnic organs is of greatest importance in the production of fever. In eviscerated animals the effect of the bacterial injection on the liver is absent, the peripheral orientation (sympathetic) is not established, and heat loss is not retarded.

If the effect of the bacterial injection were directly on the so-called heat regulating center of the brain, the peripheral effect at least should be undisturbed.

The chart which we present is typical for experiments of this type.

In dogs with thoracic incanulation, prepared under local anesthesia, the continuous intravenous injection of dilute B. coli suspension, results in a series of chemical changes in the lymph blood, which can be correlated to the clinical picture.

In the normal animal, a prompt lowering of the K/Ca ratio takes place, followed by a series of fluctuations until a time of maximum injury occurs, when a ratio as low as 0.8 may be reached. The primary drop in the ratio (due to a slight increase in the calcium and a considerable decrease in the potassium) is evidently associated with a primary stimulation of the splanchnic organs. The cellular effort to reestablish the normal equilibrium probably accounts for the fluctuations in the ratio after the primary drop.

During the periods of low K/Ca ratio, an increase in the lymph volume and lymph concentration give evidence of a coincident increase in permeability of capillaries and tissues.

The CO₂ combining power takes a progressive course downward without marked fluctuations of the curve.

The period of maximum injury is usually followed by an antimortem increase in the ratio, due probably to the liberation of potassium from the red blood corpuscles. At this time a most marked gastro-intestinal symptomatology usually occurs (vomiting and diarrhea).

Animals in whom a low K/Ca ratio is present before the injection is started, have an initial response that is paradoxical, i. e., the K/Ca ratio increases primarily. The complete details of the experiments will be published in the Archives of Internal Medicine and the Journal of Infectious Diseases.

For several years we have been studying skin reactions produced by a variety of bacterial filtrates in healthy persons, and in patients with different diseases. The recent typhoid epidemic in Montreal seemed to offer an unusual opportunity for further observations on the result of skin reactions with typhoid filtrates in typhoid fever patients. Through the courtesy of Dr. Ralph Lynch, of the Montreal General Hospital, observations were made on skin reactions of 22 typhoid patients in different stages of the disease. Intracutaneous tests with typhoid filtrates were also observed in other groups; for example, those who had no history of typhoid fever or of antityphoid vaccination, vaccinated persons, patients with other diseases, and 4 typhoid patients in Chicago.

After preliminary tests, the following technic was adopted: The typhoid bacillus employed in the test was the Rawling's strain. The medium used for the bacterial culture was a sugar-free, beef infusion broth 1% peptone, pH 7.4, to which was added 0.1% dextrose; the inoculated broth was incubated for 72 hours; then filtered through a Berkefeld filter and cultured for sterility. One:100 dilution, and in most instances 1:500 dilution in 0.1 cc. amount was injected intradermally in skin of forearm. Uninoculated broth, incubated and filtered in similar manner to the inoculated broth was used for control skin tests. Positive skin reactions appeared within 18 to 24 hours and persisted for at least 48 hours. A strongly positive skin reaction had an elevated, dark red, indurated center at least 2 cm. in diameter, frequently surrounded by a less definite halo of 2 to 3 cm. The severe reactions lasted from 48 hours to a week, then became pigmented, and frequently there was a slight desquamation. In only 2 instances were there systemic reactions.

The numerous reports in the literature showing the marked efficacy of fats, egg yolk and protein in emptying the gall-bladder convinced us that the effect of “secretin” on the evacuation and motor activity of the gall-bladder should be studied, since it is well known that these substances stimulate the pancreas, and that “secretin” (very impure solutions) promotes the formation of bile.

For some time we (Kloster, Ivy and Lueth) have been working on the purification of “secretin”. Starting with the “new secretin” of Weaver, Luckhardt, and Koch, 1 we prepared solutions of such purity that from 1 to 3 mg. of dried material excites pancreatic secretion in 10 to 20 kilo dogs. These purified solutions of “secretin” have been used in this study. They are vaso-dilatin free, and have no objective deliterious effects when injected intravenously in anesthetized or unanesthetized animals.

One series of experiments were performed as follows: A cat was barbitalized, the pancreatic duct was cannulated, and a cannula was placed into the gall-bladder through a small opening in the fundus. On the injection of the purified “secretin” (2 mg.) the intra-gall-bladder pressure rises after a very brief latent period, sometimes before the needle is withdrawn and always before the pancreatic juice and bile begins to flow. Normal saline controls proved negative. The pressure begins to decline in from ½ to 1 hour after the injection.

To avoid the possibility that the increase in gall-bladder pressure was not due to an inflow of bile, we tied the cystic duct or cannulated the bile duct. In the experiments in which the cystic duct was tied, the gall-bladder pressure was increased, but not to the extent of the first series of experiments. This difference, we believe, was due to some interference with the blood supply of the gall-bladder, which is very difficult to avoid in the cat on ligating or clamping the cystic duct.

Some years ago in experiments on pike eggs, I showed that there is an antagonism between the action of Na ions which increased the permeability of the plasma membrane and anesthetics which prevented the increase in permeability. Anesthetics in large doses, however, increased the permeability and killed the eggs. It was not shown that anesthetics ever decreased the normal permeability. At Woods Hole, Mass., I recently performed similar experiments on Fundulus eggs. Isotonic solutions of NaHCO₃ and of the nitrates of Ca and Na were made and mixed so as to reproduce the cations in about the same concentrations as in sea water. In this solution or in distilled water practically no chlorides diffused out of the eggs. A micronephelometer was made by painting black the sides of the plungers of a (B. & L.) microcolorimeter, removing the mirror and illuminating it with parallel, horizontal rays. Ten mg. of KCl in a litre of H₂O was taken as the standard. One cc. of the standard + 1 drop of 10% AgNO³ (acidified with HNO₃) was compared with 1 cc. of the unknown + 1 drop of AgNO₃ sol.∗ 100 eggs, washed until free from chlorides, were placed in 2 cc. of the solution. At the close of the experiment 1 cc. was removed for examination and the chloride content recorded in terms of the standard. Although sea water has a pH of about 8, the exosmosis of Cl was lowest at pH of 7 and less at pH 5 than at pH 8.

From experiments not given in detail it appears that Na⁺ and OH⁻ increase the permeability and Ca⁺⁺ and H⁺ inhibit their action. Alcohol has a similar effect to Ca⁺⁺.

The following determinations of iodine in drinking water of goitrous regions are uniformly low, except in the case of Lasalle, Ill. Lasalle has 2 water supplies. It is stated that there is less goiter among those persons drinking water from these drift wells than the other source of supply, which was not analyzed.

Stimulation by means of condensor shocks and induction shocks was found to be unsatisfactory in the study of the reflex time in non-curarized decapitated or decerebrated tortoises. The fall of a hammer from an electromagnet was used as a mechanical stimulus on the skin of the hind foot. A sharp point attached to the hammer served as a pain stimulus. The skin was covered by tin foil which by contact with a thin piece of copper, closed a circuit through a signal magnet as the hammer struck the tin foil, thus marking the time of stimulation. The leading-off electrodes were non-polariz-able, of the zinc-zinc sulfate-sodium chloride type. An Edelmann string galvanometer was used.

A part of the shell was removed by a rotating machine-driven saw, so that the biceps cruris muscle could be used. The time between the stimulus and the electrical variation in this muscle was considered the reaction time. Nineteen determinations on 5 tortoises showed an average of 38.7 sigmas for reflex time. The reflex time was greater the longer the time after decapitation. In one tortoise the reflex time was 32.5 sigmas on the first day, 38.7 on the second, and 47.5 on the third. In another, it rose from 35.7 sigmas on the first to 49.5 on the second day. It seemed difficult to obtain crossed reflexes, but in a few satisfactory records the time for a crossed reflex was found to be 53 sigmas.

For a study of the velocity of the nerve impulse, the sciatic nerve and gastrocnemius muscle were used. The electrical stimulus was applied to the sciatic nerve through non-shielded electrodes, and the action current of the gastrocnemius was recorded. Single break induced shocks were used, and a signal magnet recorded the time of stimulation.

To extend our knowledge of epidemic diseases, we have undertaken a study of fowl cholera. Its microbic incitant, Pasteurella aviseptica, is similar to B. lepisepticum, the organism responsible for most of the snuffles and pneumonias of rabbits. 1 This latter disease being primarily an infection of the respiratory tract, suggested the possibility that Pasteurella aviseptica infects chickens by way of the air passages, leading to both the general manifestations commonly recognized as fowl cholera and the local involvements of the upper respiratory tract designated as “roup”, “colds”, etc. We have carried out experiments to test this theory, have made serial cultures of a nasal passage of White Leghorn fowls from 3 commercial flocks over a period of 6 months, and have carried out preliminary studies of the bacterial cultures obtained. This report summarizes the results.

A strain of Pasteurella aviseptica, obtained from Dr. Theobald Smith and designated Strain Pa, was administered intranasally to rabbits and mice. 0.01 cc. of an 18 hour rabbit's blood broth culture, dropped through the external nares over the nasal mucous membrane of rabbits, killed invariably within 24 hours. Septicaemia, petechial hemorrhages, and congestion of the lungs were found at autopsy.† Similar procedures usually killed mice; a few survived, however, and carried the virulent organism for more than 6 weeks.

In the natural host, a somewhat different response was anticipated. Groups of chicks, 4 to 5 weeks old, uniform in weight, with no previous exposure to the infection, raised from similar stock and under uniform conditions, were given one drop of culture by way of the nasal cleft. Some died of acute cholera, others survived as carriers; some developed specific blood agglutinins; others seemed entirely refractory. Adult fowls, treated similarly, showed still an additional type of reaction.

Before the mechanism of bone marrow as an organ could be analyzed, certain factors had to be clarified, among which are the structure of its vascular system and the maturation of blood cells, Chart 1 shows maturation as a constantly changing process with more than one variable. There are two critical points, (1) where the specific granules begin, and (2) where the last division of the myelocyte has taken place, giving a reduction to a common size and state of maturity, the leucocyte. The myelocytes are classified according to the number of granules, type A, having the first small clump of granules; type B, showing increasing numbers, and type C, the cell with the maximum number.

Supravital differential counts of blood and bone marrow of 5 normal and 2 abnormal rabbits are given in Tables I-III. Each survey included the epiphysis and shaft of each of the long bones and one rib. Table I shows that the normal rabbit's marrow contains about 70% granulocytes to 25% nucleated red cells. In early embryonic stages, only red cells are formed; during fetal stages, white cells are minimal; in the new-born rabbit the erythroid-mye-loid ratio of the bone marrow is nearly the reverse of that of the adult.

The proportions of the stages of granulocytes are shown in Table II. The first 5 rabbits had normal blood and bone marrow, and in them the major supply of cells was of myelocytes C—(85%) with minimal numbers of types A and B and myeloblasts. The delivery of granulocytes from normal bone marrow is rhythmic, based on rhythmic and orderly division and maturation; the mechanism of delivery depends on a reciprocal relationship between myelocytes C and early leucocytes. The last 2 rabbits had a leucocytosis, pseudoeosinophiles 71% and 68%, with the normal at 44%.

The demonstration by Smith and Hendrick, 1 Goldberger and coworkers, 2 and others 3 in this country that what we have called vitamin B is composed of at least two factors, each with specific functions and properties, necessitates the revision of existing vitamin B tests. In England, Chick and Roscoe 4 have been able to confirm the contentions of Goldberger. They showed that wheat embryo is distinctly richer in antineuritic factor than in antipellagric factor, by utilizing autoclaved yeast as the source of antipellagric and a preparation of Peters to supply the antineuritic. The Peters' fraction was separated by Kinnersley and Peters 5 of Oxford University by controlling the selective adsorption of norite. During the past year Williams and Waterman of our laboratory have also succeeded in separating a yeast fraction which is antineuritic, but is apparently free of the factor (antipellagric) which is present in autoclaved yeast. Their separation was attained by controlling the selective adsorption of fuller's earth. 6

In contrast to the alcoholic extract of corn meal used by Gold-berger, Williams' preparation has supplied a testing fraction entirely devoid of antipellagric factor. To date we have been able to use this preparation in retesting two sources of vitamin B.

Last year we reported that ripe banana tested by the Sherman method must be supplied to rats on B-free, but otherwise adequate diet, in quantities of 8 to 10 gm. daily, to permit a growth gain of 20 gm. in 60 days. Using 5 mg. daily of the Williams' preparation as adequate source of antineuritic, we find that 2 gm. of banana daily supplies enough antipellagric factor to produce 20 gm. gain in 60 days, and that 6 gm. daily is nearly adequate for normal growth. By using autoclaved yeast as antipellagric, we find that at least 6 gm. of banana daily must be added to prevent growth decline.

Overton's lipoid theory 1 which states that only dyes soluble in lipoid can penetrate into living cells, is inadequate in certain respects. 2 This is due in part to the fact that he considered only one partition coefficient, that of the dye between the external solution and the lipoid layer surrounding the living cell. Since it is chiefly cytoplasmic inclusions, granules, fat globules, vacuoles, etc., which are stained rather than the cytoplasm proper, it is necessary to consider several partition coefficients, such as: (1) That of the dye between the non-aqueous layer surrounding the cytoplasm and the external solution. (2) That of the dye between the same nonaqueous layer and the aqueous cytoplasm. (3) That of the dye between the aqueous cytoplasm and the inner non-aqueous inclusions (granules, or fat globules) or the non-aqueous layer surrounding the vacuole. (4) That of the dye between the inner non-aqueous layer of the vacuole and the aqueous vacuolar sap.

Experiments from this point of view have been made on Nitella flexilis and Valonia macrophysa. The results are presented here with Valonia, the cells of which consist of an outer cell wall, an inner protoplasmic layer surrounding a relatively large central vacuole filled with sap, at about pH 5.8 and containing about 0.6 M halides.

We may assume 3 that the dye diffuses through the protoplasm into the vacuole in succession through an outer non-aqueous layer, a middle aqueous layer, an inner non-aqueous layer (surrounding the vacuole), and then into the vacuolar sap.

By applying these ideas experimentally we can predict the behavior of the dyes in penetrating Valonia. The dye in sea water at pH 9.5 or at pH 5.5 is shaken with chloroform so as to represent the two non-aqueous layers. Then the chloroform is removed and shaken with freshly collected Valonia sap.

The histological structure of the glomerulus in the dog, cat, rabbit, and guinea pig, used in various studies of renal function, is, in general, similar. The glomeruli vary in size and number in such animals, and the degree of lobulation of the capillary tufts shows some difference in the various species, yet the finer structure of these bodies remains constant. The epithelium which lines Bowman's capsule consists of a single layer of flattened, non-specialized cells and the reflection of these cells over and between the capillary loops of the glomeruli is made up of an even flatter type of cell.

In 1924 a study was commenced of the histology of the kidney in certain older animals to ascertain if there was any difference in structure in the vascular and tubular units in such ancestral species which might be indicative of a difference in the function of these structures as compared with similar units in more recent species, now used for experimental purposes. The first animal studied was the marsupial, Didelphys virginiana.

Twenty-seven opossum have been used in this study. The animals were placed in large types of rabbit cages and the urine collected. The urine from all of the animals was normal. It was free from albumin, glucose and diacetic acid. The microscopic studies showed an absence of casts. Various types of epithelial cells with occasional red and white blood cells were present. After such initial observations the animals were killed by a stroke behind the head, and the kidneys immediately removed for fixation. The kidneys were sectioned through their long axes, and tissue for fixation removed from each lateral half. The fixatives employed were 95% alcohol, Zenker's fluid, 10% formalin, and a solution of corrosiveacetic. Both paraffin and celloidin sections were made and stained with eosin and haematoxylin and with eosin and methylene blue.

Aside from the classical procedure of Kossel and Kutscher for the fractionation of nitrogenous bases, we possess no methods which permit a systematic biochemical analysis of nitrogenous substances occurring in nature. The Kossel procedure, although extremely useful for the fractionation of stable substances, interferes very often with the activity of relatively unstable substances, such as vitamines, hormones and enzymes. Various aromatic nitro compounds have been extensively used in biochemistry, such as picric, picrolonic and flavianic acids. These reagents often yield characteristic, slightly soluble compounds with relatively pure substances, but often give little or nothing when we deal with complicated mixtures. The idea occurred to me that by using the salts of these acids a series of more specific reagents could be obtained. This expectation was realized. Where the free acid fails to affect a notable separation of active material, this was accomplished by using a salt of a particular acid, or a series of salts, one after the other. The salts used so far were barium, lead, silver and mercury salts, metals which can be removed from the solution without difficulty. In some cases sodium salts were used with success. The method was used either systematically, starting with picric acid and then using the various salts, or one particular salt was used exclusively. It is hoped that by introducing other salts, or salts of other acids, a systematic biochemical analysis of unknown substances can be effected.

Using this new method, vitamine B was separated quantitatively from yeast as well as from rice polishings, but curiously enough, not by the use of one single salt. A combination of Ag- and Hg-picrates was necessary to attain this end, which corroborates my view perhaps as to the complex nature of this substance.

The simple procedure described here has been devised to obviate the necessity of transferring the blood and of using mineral oil, both of which tend to complicate the technique for determining the pH colorimetrically. The only special apparatus∗ required for this consists of a sampling tube identical in shape and size (10 × 1.5 cm.) with the tubes used to contain the standard solutions. It is provided with a double-holed, hollow, glass stopper, which is ground-in and which has sealed to it a small, two-way stop-cock with two holes. (Fig. 1.) The apparatus is made of clear Pyrex or Nonsol glass and connections are made with special rubber tubing, which will not cause excessive film formation on the mercury. Before use all apparatus, including the mercury, is cleaned and tested for neutrality in a manner similar to that outlined by Austin, Cullen, et al. 1

The general method followed is essentially that of Cullen 2 as modified by Hastings and Sendroy. 3 Whole blood is used instead of plasma in accordance with the method of Hawkins. 4 The necessary correction factor for converting the colorimetric readings to the true pH value is found by the method proposed by Austin, Stadie and Robinson. 5

When a determination is to be made, the apparatus is assembled as shown in the accompanying figure. Mercury is run from the bulb B into the sampling tube and connecting tubes until these are completely filled to the three-way inlet stop-cock C, thereby excluding all air. The sampling tube held by a clamp attached to the stop-cock S is then inverted. With the stop-cocks S and C open to make direct communication between T and E, the latter is immersed in a small flask of indicator-saline solution, which has been adjusted previously to a pH of approximately 7.40, and the air in E is displaced with mercury by elevating B.

Changes in the blood calcium and phosphorus in dogs were investigated, both before and after complete thyro-parathyroidectomy. While it was found that tetany occurred at varying levels of calcium concentration, there was always a definite mathematical relation between the concentrations of the two elements when symptoms first appeared. In normal animals the Ca:P ratio ranged between 1.5 and 2.0. After operation there was progressive reduction in this ratio until the first attack of tetany, when it was always approximately 1.0. Occasionally, there was an increase after spontaneous recovery from an attack, but usually there was a progressive decrease with each successive attack until death occurred. Treatment by any of the established methods restored the ratio, but symptoms always disappeared long before there was any extensive readjustment of the ratio. The following table illustrates the relationship:

The statement of Page 1 that during amytal narcosis no perceptible changes in blood sugar content occur was so suggestive for new possibilities in the study of carbohydrate metabolism as to warrant a repetition of the experiments of Page. In our work, the rabbit served as the experimental animal. In a series of 10 experiments 8 gave evidence of a distinct hyperglycemia which usually appears within an hour, or somewhat later, and may last for several hours. The hyperglycemia is induced by doses ranging from 50 to 100 mg. per kilo of amytal whether given by mouth, subcutaneously or intraperitoneally. Since the completion of our work Weiss 2 has shown that amytal produces hyperglycemia also in the cat and dog.

Illustrative examples of the experiments are given in Table I.

A number of investigations∗ have shown that when arsphenamine and neoarsphenamine are intravenously introduced, the arsenicals rapidly disappear from the blood stream, part apparently going to immediate excretion, and part to storage in the tissues. Usually, however, these results have been obtained from 1, or in some instances from 2 or 3 injections of the arsenical. To determine whether the number of injections would play any rôle in the disposition of arsenic we have followed the quantity of arsenic present in the blood after serial administration of neoarsphenamine to dogs, maintained upon a constant fixed diet. The drug was introduced into either the jugular or femoral vein, and blood was drawn from an artery, usually of the leg. Each animal received 0.3 gram neoarsphenamine at intervals of 4 days. The arsenic was determined in 10 cc. samples of blood, by the combined methods of Chittenden and Smith, and Sanger and Black, the total quantity in the body being derived from a calculation of the blood volume from the body weight. The results are detailed in Table I.

From the table may be seen, as others have found, a rapid disappearance of arsenic from the blood within the first one-half hour, after which the decrease is more gradual. It is to be noted that in the case of Dog 1 traces of arsenic can still be found at 24 hours following each of the first 2 injections, but that after the fourth injection there is still a trace present at 48 hours. Following the fifth injection traces can be found at 72 hours and at 65 hours after the sixth injection. With Dog 2 no traces of arsenic can be found after 24 hours after each of the first 4 injections, but after the fifth, sixth and seventh injections traces of arsenic can be shown at 48 hours.

In continuation of studies on cell antigens we examined the production of immune sera for tissues. In the literature there are contradictory reports, but several authors demonstrated marked organ specificity of such sera, e. g., Halpern,¹ Fleisher,² and recently Witebsky and Steinfeld.³ We found that immune sera could be obtained very easily and almost regularly by intravenous administration of small quantities of cell or organ suspensions, i. e., by 3 weekly injections of 10 mg. each of the fresh material suspended in saline. This was done with the hope that in this manner sera of higher specificity could be got than by the injection of larger quantities, as have been used generally. The following materials (ox) were taken as antigens: Tracheal epithelium, thymus cells, kidney and washed sperm cells. The resulting sera differentiated clearly in complement fixation tests between these 4 materials, an example of which is shown in the table.

To 0.5 cc. of progressively doubled dilutions of inactivated serum, starting with 1:10, was added 1 drop of an emulsion of about 2% of the antigen and 0.25 cc. of 1:10 guinea pig serum. After incubation for 1 hour at 37° C, 0.25 cc. of diluted Forssman immune serum containing 2½ hemolytic units and 1 drop of 50% sheep blood were added. The strength of the reactions is indicated as follows: 0 = no hemolysis, tr. = trace, d = distinct, w = weak, str. = strong, v. str. = very strong, ac. = almost complete, c = complete hemolysis.

On testinq, however, with a large number (19) of other tissues group reactions were obtained, more frequrntly with some sera than with others, also in cases where there is neither histological nor embryological relationship, e. g., trachea immune serum on brain.

The micro-technique which we 1 have devised for analyzing small specimens of calcified tissue has been subjected to further test and study. The crushed bones are extracted with ether and alcohol immediately after removal from the body; they are then dried and pulverized. About 10 mg. of bone powder are digested with HCl, trichloracetic acid is added, the solution is made up to 10 cc. and then filtered. Calcium is determined on aliquots of this filtrate. To another aliquot molybdic acid reagent (Briggs-Bell-Doisy) is added, the volume is made up to 10 cc. and then filtered. Inorganic phosphorus is determined on aliquots of this second filtrate. Digestion with HCl of different samples of the same specimen of bone powder gives solutions whose content of calcium and inorganic phosphorus agree as well as duplicate determinations on aliquots of the same solution.

CO₂ is determined on about 20 mg. of bone powder in Van Slyke's manometric gas apparatus.

A specimen of crushed bone was divided into two portions. One portion was extracted with ether and alcohol immediately, the other was allowed to stand unextracted during the summer. After standing for 5 months it was extracted and analyzed. The results for calcium, inorganic phosphorus and CO₂ were the same for both portions. Similar results were obtained with a second specimen of normal bone. The second precipitation with the molybdic acid reagent does not interfere with the accuracy of the method; the values for the ratio residual Ca/P so obtained are normal within the experimental error. When phosphorus values obtained from the first filtrate are employed, low ratios result.

In studying the method 11 specimens of normal calcification and 7 specimens of pathological calcification were analyzed. In all cases the value of the ratio was 1.96 within the experimental error; the theoretical value from Ca₃ (PO₄)₂ is 1.94.

The author previously demonstrated that white rats can be protected against many lethal doses of pneumococci, intraperitoneally injected, if these animals are first fed either the tissues of animals killed by this germ, 1 living pneumococci, 2 or HCL killed germs. 3 It was next decided to determine whether the acid killed germ when dried and mixed with cracker meal and the acid killed germ when still moist and suspended in milk could also be used. The object was to simplify the method of administering the germs as far as possible, consistent with successful immunization. At the same time success or failure accompanying a given modification gradually helps toward an explanation of oral immunization with the pneumococcus.

The organisms were grown, killed and centrifuged as described elsewhere. 3 A thin layer of germs was spread on a watch crystal and warmed on a water bath, so that they were at no time heated above 37° C. Desiccation was complete in 15 minutes. For 20 days each rat was fed the dried germs, 50 cc. of culture per day, mixed with moistened cracker meal. For testing the resistance of control and treated animals, a 24 hour virulent culture was diluted serially and the desired amount injected intraperitoneally in a volume of 0.20 cc. Table I contains the data for this experiment, and gives the rat number, whether control or experimental, its weight, the dose injected and the result. These data show clearly that the rats which were fed the dried pneumococci are considerably more resistant to the virulent germ than the control animals. The experiment was repeated, using larger rats, which responded less regularly than rats of the weights shown in Table I, although to the same degree.

Amoeba were injected with saturated aqueous and alcoholic (95%) solutions of picric acid. When an aqueous solution is injected in amounts equal to half the volume of the amoeba, no toxic effects can be seen. The amoeba is diffusely colored yellow, the intensity of color depending on the amount injected. If the injection be made so as to produce an injury at the point of insertion of the pipette, a local coagulation results. This coagulated portion is extruded by the amoeba, which then resumes normal activity and appearance, except for the yellow coloration.

When a saturated alcoholic solution is injected, there is a tendency for coagulation to occur. Unless an enormous amount is injected, the coagulated portion is localized, and is always extruded. The yellow colored amoeba then recovers completely. Hiller 1 has shown that an injection of small quantities of alcohol in concentrations above 80% will cause a local reversible coagulation. The picric acid dissolved in the alcohol apparently makes this conoagulation irreversible.

These experiments indicate that picric acid solutions which definitely coagulate proteins in the test tube, have no such action on the interior of the living amoeba. However, if there is a local injury, the picric acid will act on the injured cytoplasm, resulting in coagulation.

In contrast to the non-toxicity to the internal healthy cytoplasm is the extreme toxicity of the picric acid when applied to the surface.

By the use of a micro-pipette, solutions of Clark and Lubs' indicator dyes were injected into the subchorionic space, the pericardial cavity, and the yolk of developing Fundulus eggs, 4 to 6 days after fertilization, and the colorimetric pH values of the contents were determined. A few injections were also made into the brain vesicles. Single injections were made in most of the eggs. In several cases an injection was made into the subchorionic space, then the pipette was carried into the pericardial cavity, and another injection made in order to detect simultaneously any possible difference in the pH of different spaces in the same egg. The color tints of the indicators were always determined at the horizon of the egg in order to eliminate interference from the yellowish color of the yolk.

The color of the subchorionic space indicated a pH identical with that of sea-water. The critical indicators used were phenol red and cresol red, but the indicators above and below the ranges of these two were also injected. If the developing eggs were placed in distilled water for 18 hours and then injected, the pH of the subchorionic space was found to be that of the surrounding medium, viz., 5.6±1. Controls in sea-water exhibited a subchorionic pH of sea-water. There was no leakage around the pipette at the point of puncture, because the outward diffusion of the injected dye occurred only through the membrane as a whole, and not around the pipette at the site of the puncture.

In his experiments on the permeability of developing Fundulus eggs to acid, Loeb 1 stressed the importance of placing the eggs in distilled water for 24 hours previous to the experiment in order to rid the membrane of adhering electrolytes.

Ephedrine was successively synthesized by Nagai, 1 Späth and Göhring, 2 and Eberhard. 3 The chemical difference between the ephedrine occurring naturally and that prepared synthetically lies in their optical activity, the natural being laevo-rotatory while the synthetic is optically inactive. From the pharmacological point of view, one would expect synthetic ephedrine to possess an action similar to that of the natural product but to be less potent; for it is known that the racemic form of several well known substances, notably hyoscyamine, hyoscine, homatropine, and epinephrine is weaker than the l-form. 4 After the recent demonstration of the action and clinical uses of the ephedrine isolated from the Chinese medicinal herb Ma Huang, investigators studied the synthetic ephedrine, and showed its comparable effects to those of the natural product. 5 , 6 , 7 , 8 , 9 , 10 , 11

The present work consists of a comparison, both qualitative and quantitative, of the natural ephedrine prepared under the supervision of the author in 1924, and the synthetic ephedrine courteously supplied by E. Merck of Darmstadt, Germany. Its trade name is ephetonine. A preliminary clinical trial is also included. The hydrochloride of the synthetic compound melts at 187° C., is optically inactive, in aqueous solution at 21° C., and gives with cupric hydroxide a purple color extractable by ether, 12 It is apparently stable to light, air and heat.

When applied locally to the conjunctival sac of rabbits and man, synthetic ephedrine produces mydriasis with the preservation of the light reflex. An intravenous injection of the solution in the dosage of 2 mg. per kg. in anesthetized dogs, cats, or rabbits, results in a prolonged rise of arterial blood pressure, lasting for 25 to 30 minutes or longer, much the same as that seen after a similar injection of the natural ephedrine.

In the continuation of the study of ephedrine as a mydriatic, 1 it was found that in Caucasians a 10% solution uniformly produces mydriasis by local application, and that its mydriatic action is further strengthened if it contains, in addition, 0.1% of homatropine hydrobromide, or 1.0% of euphthalmine hydrochloride. These solutions are useful in ophthalmoscopic examination, especially from the patients' point of view, because they cause almost no change in accommodation. There were no untoward symptoms or signs noted. Only in iritis and uveitis do they fail to dilate the pupil.

Ephedrine is, however, much less certain as a mydriatic in Chinese and Negroes. Not infrequently does it fail to produce any appreciable mydriasis. It was also found that pseudo-ephedrine, synthetic ephedrine, cocaine, and euphthalmine, are less powerful in dilating the pupil of the colored races. The results observed in diffuse daylight are summarized in Table I.

Further observations are being made under controlled illumination by accurately measuring the image of the pupil with a filar micrometer attachment in a telescope. These data corroborate the above preliminary measurements and will appear at a later date.

Howard and Lee of Peking 2 reported that ephedrine is more effective as a mydriatic in individuals with light irides than in those with dark. Their results are on the same line with ours.

It has long been known that animals fail to survive extirpation of the liver. The recognized changes in the blood resulting from this procedure show clearly that the liver is essentially concerned with both carbohydrate and protein metabolisms. Does the organ possess a vital function in relation to fat metabolism as well?

For the purpose of solving this and other problems, a successful method of inducing liver insufficiency in rabbits has been devised, whereby slightly more than 90% of the liver is removed without demonstrable interference with the return of blood to the heart. Unless treated, the animals survive but 6 to 18 hours, dying with low blood sugar and convulsions. If given glucose, they live for varying periods up to 5 days, dying eventually in a state resembling that developing in dogs deprived of the entire liver. 1 A study of the respiratory quotient of rabbits deprived of 90% of the liver has brought out the fact that fat metabolism is carried on as readily and as rapidly after the operation as before.

The respiratory quotient of intact rabbits, fasted 3 days, was found to be low, indicating that the body metabolism was largely that of fat. Fasting rabbits were, therefore, employed for the work and, after the preliminary determination of the respiratory quotient, 90% of the liver was ablated. Six to 8 hours thereafter, when the immediate effects of operation and anesthesia had worn off, the respiratory quotient was again determined. No glucose had been given and at this time the blood sugar concentration was found to be 0.60 to 0.70 mg. per cc. In several instances another respiratory quotient was determined on the following day, 24 hours after operation. The animals in which this was done received just enough glucose by mouth to maintain the blood sugar level above the minimum compatible with life.

The method used for the determination of the mucinate content was the nephelometric procedure described by Inouye, 1 as follows: A 10 cc. sample of saliva is made alkaline with 2.5 cc. of 5% NaOH, thoroughly mixed, and promptly filtered through a dry filter. Five cc. of the filtrate, in a collodion bag, is dialyzed against running water until no longer alkaline to phenolphthalein (but still alkaline to litmus), and then against several changes of distilled water. The contents of the bag are then transferred to a 50 cc. volumetric flask and diluted to volume, of which 20 cc. are then treated with 0.02 N HCl from a burette, drop by drop, until an additional drop fails to give more precipitate. This is then compared nephelometrically with a standard solution of Na mucinate solution prepared as follows: A 10 cc. sample of 0.1% Na mucinate solution is dialyzed (if the mucinate is completely soluble in water without the addition of NaOH, the dialysis is omitted), and made up to 50 cc., of which 20 cc. are then titrated to maximum turbidity. If the standard and unknown give widely divergent readings, appropriate dilution of one of them should be made to bring the readings closer together.

The mechanical stimulation of chewing paraffin was used to obtain the “controls”. Twenty minutes after the end of collection of a control sample, the mechanical stimulus was again applied or a chemical stimulus with a few drops of lime juice or crystals of rocksugar on the tongue (in these cases the first expectorations were discarded), or the smoking of a cigarette, and a new sample collected for analysis. Twenty-six samples, from 9 subjects, under the mechanical stimulation, were analyzed.