Experimental Biology and Medicine

Previous studies by Harde, 1 and Harde and Philippe, 2 Harde and Benjamin, 3 Harde and Greenwald, 4 have shown a lowering of vitamin C content of the tissues of laboratory animals in many infections and intoxications. King 5 and his associates noted a similar fact in studies of human autopsies, and later in work on guinea pigs. 6

These observations suggested to us that vitamin C in infectious diseases unrelated to scurvy, such as pneumonia, may be lowered. The examinations of urines for vitamin C from pneumonia and other pathological conditions permitted verification of this hypothesis.

A reducing substance in the urine of normal individuals, probably in large part cevitamic acid, has been studied by Hess and Benjamin 7 and by Birch, Harris and Ray. 8 The latter authors found a relation between vitamin C content of the diet and the urinary excretion of the reducing substance.

Color reactions of the keturonic acids have recently become of importance to biology and medicine. In previous papers we have shown that a series of these acids can be prepared from carbohydrates by oxidation of their aqueous solutions with bromine. 1 , 2 The solutions of the keturonic acids used in the experiments reported here were prepared as follows:

1% solutions of pure carbohydrates were placed in glass-stoppered flasks together with enough bromine to provide a small excess of liquid bromine throughout the experiments. These mixtures were kept in the dark for 42 days at 25 °C. They were then aerated with washed air (and at times with carbon dioxide) to remove the excess bromine and then neutralized with potassium hydroxide to pH 7, using a spot plate. The color tests were applied to aliquots, with the results given in Table I.

In the ferric chloride test of Fenton and Jones 2 we added one drop of 10% ferric chloride solution and 3 drops of 10% potassium hydroxide solution to 5 cc. of neutral sugar solution. The purple color obtained from oxidized glycerol solution indicates that hydroxypyruvic acid is formed 3 and the similar colors from oxidized erythritol, inositol and the pentoses suggest similar products in these cases. These keturonic acids, and also those from glucosamine and ascorbic acid, give green Molisch 4 tests. The Molisch colors from oxidized α- and β-methylhexosides are distinctly different.

Oxidation products from triite and tetrite alcohols give no ketose reactions; those from pentites, the Tashiro-Tietz reaction 4 ; those from hexites and heptites give both the Selivanoff 4 and Tashiro-Tietz reactions. Only keturonic acids from aldoheptoses give positive ketose reactions. Oxidized ketoses, together with their oligo- and polysaccharides, still give positive ketose reactions. The behavior of glucosans and ascorbic acid with these reagents is note-worthy.

A study of the plasma cholesterol fractions in acute infections has been reported. 1 This previous investigation included 12 children ranging in age from 18 months to 13 years. It was observed that the total cholesterol values were much less at the height of the infection than during the period of convalescence and that this reduction was due to a marked fall in the ester cholesterol. The free cholesterol changed very little.

Repeatedly, however, reference has been made to the statement that the response of the blood lipids to various factors might not be the same in the infant as in the older child. Plasma cholesterol is considered to be quite low in many infants. 2 Furthermore, a recent study has shown that infants with eczema tend to have a low serum cholesterol. 3 The influence of acute infection upon the already low values of cholesterol in normal and eczematous babies has not been recorded.

Fourteen infants ranging in age from 3 months to 13 months were chosen for this study. Six of the babies had infections of the upper respiratory tract but no eczema. Three had eczema and repeated acute respiratory infections. The remainder did not have eczema but were ill with pneumonia which represents a rather severe form of infection in the infant. The first blood sample was collected during the height of the disease. The second sample of blood was obtained during the period of recovery after the seventh day of normal temperature. All blood samples were drawn between 12 and 16 hours after a meal. Bloor's methods were followed to determine the total, ester, and free cholesterol values. 4 , 5 , 6

The changes which take place in the cholesterol content of the serum during the acute infections occurring in the infants observed in the preceding paper 1 are accompanied by a fall in the plasma total fatty acids. The serum unsaturated fatty acids have already been found to be replaced in part by more and more saturated acids during acute respiratory infections in older children. 2 In infants the observation has been made that eczema is associated with an abnormally low serum content of unsaturated fatty acids 3 but no study has been recorded of the influence of acute infection on the unsaturation of the serum fatty acids.

This paper presents the observation made in 14 infants with acute illnesses. They are the same babies which were chosen for the cholesterol study. All blood samples obtained during and after the height of each infection were drawn between 12 and 16 hours after a meal. The first blood sample was collected as soon as possible after the infant was admitted to the hospital at the height of the disease and the second blood sample after the seventh day of convalescence. Bloor's methods 4 were followed to determine the blood lipids. The Rosenmund-Kuhnhenn method 5 as modified by Page, Pasternak and Burt 6 was used to determine the iodine absorption of the serum fatty acids.

The results are summarized in Tables I and II.

The fall in plasma total fatty acids which takes place during the height of acute infections in infants is paralleled by a drop in the iodine absorption values. This signifies that in most instances the iodine number of serum fatty acids is definitely lower in the febrile period of the disease than during the period of recovery. Acute infection produces in the infant the same changes previously found to occur in the older child.

There is considerable evidence that experimental diabetes insipidus is neither strictly an endocrine disturbance nor entirely an involvement of the brain but probably the result of both. Rats can be rendered diabetic by removal of the posterior lobe or by a stab wound in the base of the brain, providing the latter injury is such that it completely severs the stalk from its attachment to the brain. 1 , 2

Fisher, Ingram and Ranson, 3 availing themselves of the Horsley-Clark stereotaxic instrument, investigated the effect of restricted hypothalamic lesions in the cat. Their observations were combined with careful histological studies of the resulting degeneration and they were able to demonstrate that a permanent diabetes insipidus ensued only when there was complete interruption of the hypothalamico-hypophyseal tract.

The literature contains numerous reports on the effect of complete hypophysectomy, hypophyseal injuries, and lesions of the tuber cinereum and the probable relation of these to experimental diabetes insipidus, but the writers are not aware of any investigations in which an effort was made to determine the effect of removal of the anterior lobe without concomitant injury to the posterior lobe or stalk.

The literature contains numerous reports on the effect of complete hypophysectomy, hypophyseal injuries, and lesions of the tuber cinereum and the probable relation of these to experimental diabetes insipidus, but the writers are not aware of any investigations in which an effort was made to determine the effect of removal of the anterior lobe without concomitant injury to the posterior lobe or stalk.

The experimental material included 47 female rats, varying in age from 90 to 120 days at the time of operation. The results secured with completely hypophysectomized rats, rats with the posterior lobe alone removed, and the rats with the anterior lobe removed, are summarized in the accompanying chart. It will be noted that following complete hypophysectomy there is an immediate and transitory polydipsia, a condition now known for some time. 4

Methods for quantitation of bile pigments, with the exception of the methods of Hooper and Whipple, 1 Beccari, 2 Kerppola and Leikola, 3 and Peterman and Cooley 3 have been confined to modifications of the van den Bergh method 4 for the determination of bilirubin. That the pigment in the bile of carnivora is present chiefly as bilirubin and that of herbivora as biliverdin 5 is a conception based partly upon physical appearances of the respective bile and partly upon specific tests for these particular constituents. Modifications of the van den Bergh method are satisfactory for clinical identification or quantitation of bilirubin; however, such methods are without value when used for the quantitation of pigment in bile in which the pigment is present in the form of oxidized bilirubin derivatives. Furthermore, Müller and Engel 6 have shown spectrophotometrically that the diazo reaction of the van den Bergh test detects only 30 to 60% of the bilirubin present in a solution, while Beccari 7 has reported that a number of compounds other than bilirubin, and present both in bile and serum, gave color reactions which were indistinguishable from that given by bilirubin.

In attempting to use the method of Hooper and Whipple 1 for the determination of total pigment, it was found that the method was uncertain and has objectionable features which have been pointed out by Peterman and Cooley. 8 These investigators have shown that the blue-green color taken as an end-point in the Hooper-Whipple and their own methods is a mixture of unoxidized bilirubin and bilicyanin, the combination of which constitutes the green pigment of bile heretofore considered as an individual pigment, “biliverdin.”

The production of water intoxication in human beings is difficult. However, instances of such intoxication have been reported,^1-4 and it has been the subject of experimental investigation in lower animals. 5 , 6 , 7 Recent studies 8 , 9 upon the metabolic effects of water have suggested that a definite relationship exists between the initial metabolic rate and the rise in metabolism induced by water. Following hemorrhage, there is a marked diminution in the metabolic response to water and electrolytes. 10 This study has been undertaken to determine whether or not there is an altered susceptibility to water intoxication in animals subjected to experimental traumatic shock.

Healthy dogs weighing between 6-19 kg. were used. Anesthesia was induced by sodium barbital in a dosage of 250 mg. per kilo of body weight. Blood pressure tracings were made in the usual manner before and during the administration of fluid. The metabolic readings were obtained by a Krogh apparatus before, and at 15-minute intervals after, the production of shock. Shock was produced by: 1. massive hemorrhage to the extent of 1/3 of the calculated blood volume. 2. crushing of the muscles of the hind legs or by prolonged handling of the exposed bowel. 3. ligation of the portal vein. 4. injection of histamine dihydrochloride. The fluid utilized was an isotonic (0.9%) sodium chloride solution heated to 38°C., and given by vein at a constant rate of 16 cc. per minute. The standard amount of solution was 3,000 cc. Ten of the animals were used as controls. The urine output was estimated by a catheter placed in the emptied urinary bladder.

Since Elliot's work in 1907, the increased sensitivity to adrenalin in vessels deprived of sympathetic innervation has been well known. Clinical observations on patients following total removal of the thyroid have suggested a close dependency of adrenalin efficiency on the presence of thyroid hormone. The injection of adrenalin in patients deprived of thyroid secretion seemed to produce less vasoconstriction. This diminished action of adrenalin was very striking in limbs rendered especially sensitive to adrenalin by sympathectomy.

To test the accuracy of these observations the following experiments were performed on dogs:

Dogs were trained to stand quietly in a light overhead harness. Skin temperature curves were made at the same time of each fore leg by means of thermocouples attached to the skin. A number of such skin temperature curves under standard conditions revealed no difference between the fore legs. Being thus assured of a proper control, the sympathetic fibers supplying the left leg were destroyed by the removal of the stellate and each adjoining ganglion above and below. After allowing adequate time for the degeneration of the post-ganglionic fibers, determinations by the thermocouples revealed the customary 4° to 5° difference in temperature between the 2 legs. Skin temperature curves were now taken simultaneously on the 2 forelegs following an intravenous injection of a 1-20,000 solution of adrenalin chloride. The injections caused consistently a much greater and more prolonged fall in temperature on the operated side. When a number of similar curves were accumulated under identical standard conditions, the dogs were subjected to total thyroidectomy. After recovery and during the next 8 months temperature changes in both fore legs were again recorded following adrenalin injection. The post-thyroidectomy curves showed a great loss in the sensitivity of the sympathectomized vessels to adrenalin.

In the dogs deprived of thyroid secretion the former prompt pressor response and resulting abrupt and prolonged fall of temperature on the operated side was almost entirely abolished. The curves for the intact limb were but little changed by total thyroidectomy. These experiments served to confirm our clinical impressions of the direct relationship between the thyroid hormone and the effectiveness of circulating adrenalin.

We have previously reported 1 that injections of large amounts of oestrin into mature female rats produced corpora lutea in the ovaries of these animals equal in size to those of pregnant rats. Similar results have been reported by Hohlweg 2 and Selye, Collip and Thomson. 3 Other experiments 3 , 4 , 5 have demonstrated that injection of oestrin simultaneously with A.P.L. (anterior pituitary-like substance) enhances the luteinizing capacity of A.P.L. and its capacity to increase the ovarian weights of immature rats. Hisaw and associates 6 and Lane 7 have demonstrated that injection of oestrin stimulated the production on the luteinizing hormone in the anterior hypophysis. The data presented below support the view that combined injection of oestrin and A.P.L. exerts a much greater effect on the ovaries of immature female rats than do injections of the A.P.L. factor alone.

In the first series of experiments a group of immature female rats received a single injection of 25 units of A.P.L.† They were killed 10 days later after no further treatment. A second group, littermate sisters to those above, received a single injection of A.P.L. and beginning on the fourth day following the administration of A.P.L. daily injections of 200 units of oestrin,† until the tenth day following the injection of A.P.L., at which time the rats were sacrificed. As controls for these experimental groups a group of untreated littermate sisters were used.

In studying the nerve pathways involved in the reflex inhibition of hunger contractions by distention of a Thiry fistula loop in the dog, 1 it was noticed that a previous long-continued distention of the loop brought about a lower threshold for the reflex.

In the experiments here reported, five animals were prepared with Thiry fistulae in the usual manner, the loops being 15-20 cm. long and taken from the upper ileum. After complete recovery the dogs were trained to swallow a stomach tube and to lie quietly on the table without any restraint. Gastric movements were recorded by the customary balloon-tambour method. The intestinal loops were distended by air inflation of a thin-walled balloon which was over-size relative to the lumen of the gut. The balloon was tied over a glass rod so that the expansion would be entirely lateral and the pressure readings would closely approximate the pressure actually applied to the intestinal wall. All records were made after a fasting period of 18-24 hours. In testing for inhibition of gastric motility the intestinal loops were inflated for a short period, usually 10-30 seconds. In sensitizing or reinforcing the reflex, the loops were subjected to a continuous pressure of 75 mm. Hg. for 24 hours, following which tests were again made.

The essential results may be seen in Table I and Fig. 1. In the control experiments pressures in the loops of 75 to 120 mm. Hg. produced inhibition of gastric movement in 25 out of 55 tests. Each dog responded about half the time, the percentage of responses increasing somewhat with the higher pressures. It was noted that hunger contractions of type 2 were most easily affected. Often there was a marked lowering of tonus with the presence of smaller contractions at this lower level.

The effect of liver injury on the amount and distribution of the liver lipids has been the subject of a limited amount of investigation, with a view to gaining a better understanding of the relation between the phospholipids and the other fatty constituents of that organ. Theis¹ found for the normal liver tissue of beef, rabbits and humans that the relation between the phospholipid and neutral fat is quite constant and may be expressed as an equilibrium, 55 to 60% phospholipid, 45 to 40% neutral fat, an equilibrium which is readily displaced as a result of fatty degeneration, insulin administration and disease. The abnormal organs seldom showed either less or more total lipid than the normal tissue but the percentage of phospholipid was always less. Since results obtained in this laboratory indicated that the ratio of phospholipid to neutral fat for normal liver tissue is about 72% to 22%, higher than that reported by Theis, it was felt that this problem deserved further study.

The white rat was used as the experimental animal and the effect of chloroform poisoning, insulin administration and fungus infection (Aspergillus and Sporothrix) on the amount and distribution of the liver lipids was determined. The experiments were carried out on pairs of litter mates of the same sex and approximately the same weight, one serving as the control and the other as the experimental animal. The rats were killed by a blow on the back of the head and the livers removed immediately.

The reactions of the prepubertal rabbit to gonad-stimulating extracts of anterior pituitary and of pregnancy urine have been extensively studied^1-9 but apparently no ovarian responses have been observed in this species at periods prior to the appearance of antrabearing follicles. Since the rat^10-13 and guinea pig 14 , 15 show luteinization of the theca cells in response to injections of similar extracts at an age when the follicle granulosa is still non-reactive, it was of interest to reexamine the immature rabbit with particular reference to possible reactions of an analogous nature.

White rabbits from a single stock, and representing age groups of approximately 15, 30, 45, 60 and 90 days, were used. (Table I.) Each group comprised 2 litters totalling 10 to 14 females. At the beginning of the experiments, one ovary was removed from 5 or 6 of the animals of each group to provide control material. One-half of the animals received an extract of sheep anterior pituitary, and the remaining half an oestrin-free extract of pregnancy urine. One-quarter cc. of the pituitary extract was equivalent to one luteinizing rat unit; ascending doses from 0.5 to 1.5 cc. were injected subcutaneously twice daily for 14 days. Ten γ of the pregnancy urine extract (A.P.L.) represented one luteinizing rat unit; 100 such units were given intravenously once daily for 14 days.

A few animals were killed at intervals, during the injections, but the majority were killed on the 15th day. The ovaries were sectioned serially, except in a number of instances where they were stained for lipoids by frozen section technique. In addition to these, a large number of control ovaries from mature rabbits (normal, treated and hypophysectomized) was studied.

Kugelmann 1 reported 7 cases of angina pectoris who developed an elevation in blood pressure and substernal pain during hypoglycemia induced by the intravenous administration of insulin. The attacks of substernal pain and symptoms brought about by this procedure were similar to those which occurred spontaneously in these patients under ordinary circumstances. The studies of Kugelmann suggested to us using insulin hypoglycemia as a method of determining blood pressure responses. We have studied the blood pressure and blood sugar changes in normal individuals, in patients with essential hypertension, and in cases of Addison's disease following the intravenous administration of insulin. In each case a blood sugar determination (Folin Micro-method using venous blood) was made following a 12-hour fast; 15 units of insulin (Mulford) were then given intravenously, and subsequent blood sugar determinations made every 10 minutes for 90 minutes. To circumvent the effects due to emotional disturbances from repeated vena punctures only one puncture was made and the lumen of the needle kept unobstructed by a flow of 1 cc. normal saline per minute. Blood pressure readings with the patient supine and resting were made every 10 minutes for one-half hour preceding the test and every 2 minutes following the administration of insulin.

In 25 normal individuals following insulin and at a variable time interval, but usually within the first 40 minutes, there was a drop in blood sugar to an average level of 30 mg.% without any appreciable changes in blood pressure. There then occurred a sharp rise in blood pressure averaging 40 mm. Hg., systolic, and 7 mm. Hg., diastolic, over the previous readings. With this event the blood sugar rose from 30 mg.% to 46 mg.% with a gradual return to the fasting level. In the study of 20 cases of essential hypertension, the same phenomena occurred in a more striking fashion.

Proteins treated with nitrous acid are said to be “diazotized” insofar as, like true diazonium compounds, they couple in alkaline reaction with substances containing an aromatic OH or NH₂ group, often forming a colored compound resembling the azo dyes. This reactivity of proteins is puzzling, since there is no known primary aromatic amine in protein which would react with HNO₂ to form a diazonium compound. Flick 1 contended that the reactivity of wool treated with HNO₂ was due to the formation of a nitroso rather than a diazonium compound. More recently, Morel and Sisley, 2 confirming previous work by Landsteiner 3 with salicylic acid, have reported that tyrosine treated with nitrous acid forms a diazonium compound which couples with naphthol to form azo dyes. They therefore ascribed the reactivity of diazotized protein to its tyrosine content. Their contention remains plausible even if tyrosine forms a nitrosophenol instead of a diazonium compound on treatment with HNO₂, as found in the case of phenol by Baeyer and Caro 4 ; for nitroso compounds might also couple with phenols and aromatic amines.

The following data, however, indicate that tyrosine is not the only constituent of protein responsible for its diazotization. Instead, there is reason to believe that tryptophane is also responsible for the fact that protein treated with HNO₂ couples with aromatic amines and hydroxyls.

When tryptophane (or indole) was treated with nitrous acid in the cold, under the same conditions which lead to the “diazotization” of protein, a highly reactive compound was formed which gave a brilliant red color when added to α-naphthol in alkaline reaction (Table I). The HNO₂, probably reacted with the indole NH to form a nitrosamine.

In the course of experimental work upon cardiac hypertrophy it was found necessary to obtain data upon the normal heart weight-body weight ratio in a large number of guinea pigs. Joseph 1 reported ratios of 4.22 gm. per kilo in 14 males and 3.91 gm. per kilo in 33 females. However, these animals had been used for other purposes in the laboratory, so there may be some question as to whether they were actually normal.

The animals reported in this paper were kept upon an adequate diet until a constant weight was reached. They were finally weighed, then killed by a blow upon the head. The thorax was opened and the heart removed from the pericardium. The great vessels were cut flush with the surface of the heart, and all 4 chambers were opened and washed free of blood. The excess moisture was removed by blotting the organ with filter paper, and the heart was then carefully weighed. The ratio was determined by dividing the heart weight, in grams, by the body weight, in kilograms.

The figures obtained are summarized in Table I.

The fact that the standard deviation is the same for the 2 sexes does not confirm Joseph's 1 report of greater variability among males.

We feel that future work may be based upon these normal figures, since they withstood a rigorous statistical analysis.

With the increase in the number of inbred stocks of mice interesting and important data may be obtained in the study of the incidence of certain types of cancer by making reciprocal crosses. Previously all reports on the development of tumors in hybrid generations have been observations secured by crossing strains of animals which differ greatly in the percentage of animals showing growths. In this paper we wish to consider the spontaneous tumor incidence in mice of the first filial generation made by mating individuals of two strains in which mammary gland tumors frequently develop.

Stocks of Mice. (1) The “D” Strain. A sub-strain of the Little dilute brown race, previously described by Murray⁹ was used as one parental race. (2) The “A” Strain. The other stock of animals employed was originally descended from the Bagg albino strain. Data on the tumor incidence has been reported by Strong¹³ and Bittner. 1 A report comparing the breeding behavior and tumor incidence of the D and the A stocks is in press (Bittner and Murray). 2

Females from each strain were used in breeding the first generation hybrids. When the A stock females were mated to D males, the hybrids were termed ADF₁; when the D females were crossed to A males the resulting generation mice were called DAF₁. All the F₁ individuals were employed as breeders in transplantation studies.

The number of animals dying non-cancerous or developing tumors are grouped in monthly age periods in Table I. The proportion of the total cancerous mice recorded in bimonthly periods is represented graphically in Fig. 1.

In order to see what effect methylene blue had upon the form of the hemoglobin in rabbits poisoned with CO, spectrophotometric analyses were made on blood at regular time intervals up to 20 minutes after removal of the animal from the gas chamber. Each rabbit was allowed to remain in an atmosphere of CO plus air, (% composition not determined) until it was unconscious and barely breathing, but not long enough to cause death. CO₂ was absorbed by soda lime. The animal was then taken out, a heart puncture made and 2 drops of blood immediately placed in a small tube filled to the brim with a measured amount of 0.4% NH₄OH. This was then tightly stoppered with paraffined corks excluding air, and shaken to cause complete hemolysis. A 0.03% methylene blue solution (Merck's medicinal) dissolved in 0.9% NaCl was then injected intravenously, 1 cc. per kg., one minute after the animal had been removed from the CO chamber. Heart punctures were made 2, 4, 7, 11, 16 and 21 minutes later and samples collected as described above. Control animals were either given injections of 0.9% NaCl alone or no injections. There was no difference in the observed values between the 2 types of controls. Ten animals in each series were used. The probable error of the mean readings at each time was less than 1%. The disappearance of CO-hemoglobin and reappearance of oxyhemoglobin was then followed by spectrophotometric readings of the blood samples. The method described in previous papers 1 was used, the ratio of the extinction coefficients at 540 and 560 mμ being determined. This indicates the per cent oxyhemoglobin as compared with CO-hemoglobin present. Table I shows the progressive change in the proportion of the total hemoglobin present, the remainder in each case being CO-hemoglobin.

This shows the rapid change from CO-hemoglobin to oxyhemoglobin in the case of the treated animals as compared with the slower change for the controls.

The increase in the crop weight of the common pigeon, while a valuable qualitative test, was found unsatisfactory in the quantitative assay of galactin due to the variability of groups of birds to the same amount of hormone. As the further purification of galactin requires a reliable bioassay, a more accurate method was sought.

Instead of weight, the minimum proliferation of the crop gland was found more satisfactory. This degree of crop growth is characterized by the presence of transverse strands or lobules of cellular development and considerable opaqueness when the crop is extended.

In the assay of the estrogenic hormone, Coward and Burn 1 made use of the variation in the response of rats as indicated by the percentage showing a positive reaction at various levels of administration. The adaptability of this method of bioassay to galactin has been studied. A characteristic sigmoid curve with maximum sensitivity in the region of 50% response was obtained (Fig. 1). An assay method has therefore been devised using minimum proliferation combined with the principle of percentage response.

The details of the assay method follow: Common pigeons weighing from 260 to 340 gm. are used. The fight young birds and the very heavy birds are less uniform in their response. As no sex difference has been noted, both males and females may be used. Experience indicates that about 20 birds should be used. The extract is injected once daily for 4 days just beneath the subcutaneous tissue into the breast muscle.† The crop glands are examined on the 5th day for proliferation.

Harding and coworkers 1 have devised a method of estimating a number of the common sugars when present in a mixture. They utilized Proteus vulgaris, M. krusei, S. marxianus, and M. tropicalis for the differential removal of sugars in a mixture of glucose, fructose, galactose, sucrose, maltose and lactose. This procedure combined with reduction values before and after hydrolysis provided a reasonably satisfactory method of determining the individual sugars in the mixture.

The writers have found it possible to determine glucose, galactose, and lactose in a mixture by a similar, though simplified technique which avoids the use of organisms other than baker's yeast. An improved method for the hydrolysis of lactose was devised which gives complete decomposition as compared with the 72% hydrolysis by the method of Harding and Grant. 2

Galactose or lactose mixed with glucose were satisfactorily recovered after fermentation with washed baker's yeast. Table I shows recoveries of different quantities of glucose, galactose and lactose from solutions of all three. The data are not selected and represent all determinations made. The recoveries of galactose and lactose are not quite as good as those for glucose as might be expected due to the greater number of procedures involved in their estimation. 1. The titration values for the mixture are determined by a Shaffer-Somogyi 3 reagent (79.5 gm. Na₂Co₃, 1 gm. of K1, and 21 gm. of NaHCO₃, per 1) both (A) before and (B) after fermentation using 5 cc. of a 15% suspension of washed baker's yeast 4 (Fleischmann) and 14 cc. of the sugar solution made first just acid to congo with a drop of H₂SO₄.

The method of purifying the gonadotropic substance in pregnant mares'blood by adsorption of the active material to specially prepared aluminum hydroxide followed by elution 1 is a difficult and time-consuming procedure. The method yields preparations directed to the requirements of clinical investigation but for studies on laboratory animals somewhat less purified solutions of the hormone are entirely satisfactory. For this purpose it would be desirable to have a rapid method of eliminating the bulk of the inert material from the serum or plasma of the mare and of concentrating the activity in a smaller volume.

Although there are many chemical and biological differences between the gonadotropic material present in pregnant mares'blood and that of the urine of pregnant women, we have found that the method of Katzman and Doisy 2 for the preparation of the gonadotropic substance of the latter by adsorption of the active material to benzoic acid may be successfully employed to effect a partial purification of the gonadotropic material in pregnant mares'blood. Solutions of the hormone prepared by this method contain much larger amounts of horse serum proteins than solutions prepared by adsorption to aluminum hydroxide.

One liter of citrated plasma from a pregnant mare 3 was diluted with 9 liters of water and acidified with 10% HCl (pH 3.0 to 2.5). When necessary the acidified solution was clarified in the Sharples supercentrifuge. After chilling, the solution was vigorously stirred while 750 ml. of an acetone solution of benzoic acid, saturated at room temperature, was slowly added, following which the suspension was vigorously stirred for half an hour. The suspended benzoic acid was allowed to settle, the supernatant decanted and the precipitate collected on a Büchner funnel. The filtrate and supernatant from the benzoic acid precipitate were combined and the process of adsorption repeated since a single adsorption rarely recovered all of the active material.

It is well established that injection of oestrin in normal female rats induces changes in the anterior hypophysis comparable to those of pregnancy and that such injections in castrated female rats prevent typical castration changes. Daily injection of large amounts of oestrin into normal female rats for 10 days results in a marked but variable weight increase in the pituitaries. The basophiles are markedly degranulated and reduced in relative percentage. The eosinophils also present loss of granules and a variable reduction in relative percentage. The chromophobes are greatly increased in relative percentage. 1 In the studies recorded below we have attempted to ascertain more exactly the morphologic and quantitative effects of large and small daily injection of oestrin for periods of 5 and 10 days on the various cell types of the anterior pituitaries of immature female rats. We wished to determine if the amounts of oestrin necessary to cause changes in one cell type would induce changes in the other types. Female rats, 21 to 23 days old were used. At autopsy, body, ovary and pituitary weights were obtained. Serial sections of all pituitaries were cut. Five representative sections from each were studied; differential cell counts were made on each; the number of mitoses were also counted.

In the first series, 9 rats received daily injection of 10 units of oestrin† for 5 days; a second group of 8 littermates received daily injection of 200 units for the same period. Eleven littermate controls were available. In the rats receiving, 10 units daily the pituitaries were increased in weight to a mean of 3.5 mg.; the mean in the controls was 2.6 mg. The basophiles of the injected rats were moderately degranulated; the result was a decrease in the relative levels of the granular basophiles to a mean of 2.8% while this mean in the controls was 8.3%. A few of the eosinophiles in the injected rats were swollen and showed loss of granules. The relative percentages of eosinophiles were not decreased when compared with the controls. This may possibly be explained by the fact that in the injected rats the number of mitoses found in the eosinophiles was much greater than in the controls; the mean number per section in the injected rats was 14.7 while in the controls it was only 3.2. The mean number of mitoses per section in the chromophobes was also increased; in the injected rats it was 18.4 while in the controls it was 7.7.

Quantitative considerations of the amount of lipoid in a few hundred tubercle bacilli compared to the amount of tubercular or other lipoid required to produce a noticeable foreign body reaction obviously render inadequate the present more or less generally accepted hypothesis purporting to explain the histogenesis of the tubercle. Moreover, such a theory completely fails to explain the development of closely similar, sometimes almost indistinguishable lesions in other infectious granulomata such as syphilis, typhoid, glanders, or brucellosis.

In a previous paper 1 concerned with the histologic reactions in hypersensitive states, the authors pointed out that bacterial allergy—defined as the first stage of the immune response to parenterally injected antigen, characteristic only of active as contrasted with passive immunity—develops much more rapidly than has generally been believed and determines on the part of the host a tissue reaction characterized by a marked infiltration with large mononuclear phagocytes. Attention was called to the fact that bacterial allergy is particularly well marked in that group of diseases in which focal mononuclear reactions characterize the histologic picture and the suggestion was made that the relationship between the two states might better be explained by the assumption that the allergy determined the type of histologic response than by the more conventional theory that the granulomatous response provoked the allergy. The present experiments were designed to test this hypothesis by a study of the time relationships of the development of allergy and the character of the histologic reactions.

Approximately 50 guinea pigs were infected with large doses of tubercle bacilli (5 to 20 mg.), strains of both low and high virulence being used, in order to insure the rapid and intensive development of generalized hypersensitiveness. The sites of primary inoculation were varied, 2 groups of animals being infected intratesticularly, one group intraperitoneally, and one both intratesticularly and subcutaneously.

The use of the older methods of studying the isolated heart led to contradictory conclusions regarding the effect of alcohol on cardiac contraction. Sulzer, 1 using the Starling heart-lung preparation, at present the most sensitive and reliable method for this purpose, found no stimulating effect in any concentration and showed that concentrations commonly reached in the human blood stream in alcoholic intoxication produce dilatation without an increase in the work of the heart. We have confirmed this effect and studied the changes in energy metabolism associated with it.

We have used a modified heart-lung preparation, which has been previously described. 2 In some experiments the diastolic volume was allowed to increase when alcohol was added (Table I), while in others (Table II) the external diastolic volume was maintained constant throughout the experiment by adjustment of the venous return. In all cases the venous pressure had to be lowered in order to keep diastolic volume constant after alcohol. A piston recorder was used for ventricular volume recording and the heart was held in a glass cardiometer, following the technique of Starling and Visscher. 3

After a control period of 15 to 30 minutes, during which the volume output and oxygen usage remained constant, the desired amount of alcohol was injected into the tubing between the stromuhr and the venous reservoir so as to insure thorough mixing before reaching the heart. Within 30 to 40 minutes after the injection the variables recorded had generally reached a new fairly constant level, and then in some experiments alcohol was injected again and the effect of the higher concentration determined. The concentration of alcohol injected was 20% except in Exp. 4/10/33, in which 95% alcohol was added to the blood undiluted for purposes of comparison. The time of injection was generally 2 to 3 minutes.

It has been reported 1 that the intravenous administration of glucose to dogs under barbital anesthesia reduced the duration of the anesthesia by 50%. The original purpose of the investigation reported in this paper was to determine the effect of glucose administration during pentobarbital anesthesia.

Pentobarbital in doses of 40 mg. per kilo was administered to rabbits, intraperitoneally. At the time of maximum depression, as determined by muscular relaxation and diminished respiratory rate, 10 cc. of 50% glucose was slowly injected into the marginal ear vein. The results are shown in Table I.

It will be seen that the injection of glucose to normal-fed animals did not materially shorten the period of depression. However, fasting for 20 hours increased the duration of the period appreciably. This is in agreement with Reidel, 2 who obtained similar results on rabbits during “avertin” anesthesia. He reported that 20 cc. of 20% glucose did not shorten the period of anesthesia, but fasting doubled it.

Although all of our animals received 40 mg. of pentobarbital per kilo, their response varied between no apparent depression and death. Approximately 110 observations were made on the blood sugar level before the injection of the pentobarbital and the same number of determinations were made at the time of maximum reaction to the anesthetic. The blood sugars were determined by the Somogyi 3 micro blood sugar method. The results are divided into 3 groups corresponding to “no depression,” “slight depression” and “marked depression”. The mean blood sugar levels are shown for each group in Tables II and III.

After having administered dahlia inulin (lot 661-non-toxic) intravenously in large doses to dogs, and to man in 42 instances in doses ranging from 30 to 150 gm., without reaction, an unexplained transient reaction consisting of chills, fever, lumbar pain, nausea, vasomotor depression, herpes and anuria was encountered. The same reaction was encountered simultaneously in another laboratory where an independent sample of inulin was being used. The material that produced the reaction in our laboratory (lot 661-toxic) was a new shipment which had been purified separately by the manufacturer from the same batch of crude inulin as had supplied lot “661-non-toxic.” At our request the manufacturer supplied a fresh, highly purified sample (lot 681) prepared from a new batch of dahlia roots, which proved to be, if anything, more toxic than “661-toxic”, 1.0 gm. sufficing to produce either chill, fever, headache, nausea or lumbar pain. This reactive inulin appeared to be relatively innocuous for dogs, rabbits and guinea pigs, even when administered in very large doses. Boiling for 30 minutes in distilled water did not appreciably diminish the toxicity; partial hydrolysis with dilute acetic acid (sufficient to increase the reducing power from 0.9 to 13% by weight) diminished the toxicity somewhat, and complete hydrolysis with N/10 H₂SO₄ decreased the toxicity considerably. However, a fairly typical reaction was obtained after the administration of 20 gm. that had been hydrolyzed in the latter manner. Spectroscopic examination of the ash of both preparations revealed traces of numerous metals, but showed no Pb or Si, and no significant differences between the toxic and nontoxic preparations.

Through the courtesy of Dr. Eaton M. Mackay and of the Bureau of Chemistry and Soils of the U. S. Department of Agriculture, we subsequently obtained a quantity of crude chicory inulin. This material was purified in the Department of Physiology, and with the exception noted below has proved to be innocuous to animals and to man. This material has, at the time of writing, been administered intravenously to man 28 times in doses of 30 to 40 gm. without reaction, and it produced no reaction in doses up to 80 gm.

Two principal conditions favor uterine growth, (1) hormonal, such as occurs in the presence of oestrone 1 and progesterone, 2 and (2) local, such as is observed in the gravid horn of a unilateral pregnancy. The growth of such a cornu is much greater than that which takes place in the non-gravid horn exposed to the identical hormonal environment. 3 The difference has been ascribed to the presence of the fetuses. Local uterine growth is also observed when, during pseudopregnancy, a uterine cornu is distended with rolled rubber dam. 4 Thus the 2 growth factors may be said to be essentially hormonal and physical. The present note is concerned with experiments in which an isolation of these factors is achieved, and uterine growth induced by physical factors alone in the absence of ovarian hormonal influence.

The procedure is as follows: mature rabbits are ovariectomized at the end of a week of pseudopregnancy, at which time the uteri of the various animals are in comparable structural states. After one week of castration paraffin pellets (m.p. 54°) of 2 sizes (1/8“ and 1/4” diameter, respectively, and 1/2-3/4“ in length) are inserted per vaginam into a uterine horn and anchored there by a loose stitch at each end of the pellet. Control sections of uterine tissue are distended, excised and fixed in formol-acetic fixative solution.

In each of 11 experiments it has been found that, in the absence of ovarian hormonal influences, marked uterine enlargement occurs. The muscle cells undergo hypertrophy, connective tissue increases and dilatation of the small vessels takes place. The endometrium, after 2 weeks of chronic distension, is in a state of active proliferation but contains no secreting cells. These changes take place concurrently with some atrophy in non-distended portions of the same uterus.

The only report available on electrocardiographic changes following experimental coronary sinus obturation is that published by Otto. 1 All of the dogs studied, however, had had section of extra-cardiac nerves.

During the course of experimental attempts to increase the blood supply to the heart 2 , 3 we have studied the electrocardiographic findings in 66 dogs before, during, and at various intervals of time up to 4 weeks after partial or complete occlusion of the coronary sinus. This occlusion was produced by ligation or by the injection of escharotics into or around the mouth of the coronary sinus. Twenty-five dogs showed a completely obturated coronary sinus at autopsy, 11 of these having been produced by ligature alone. In 26 dogs the obturation was partial. In 15, the same procedures were employed but no obturation was produced. The electrocardiographic findings in these dogs therefore serve as controls.

Following obturation of the coronary sinus, the entire heart dilated, the veins became engorged and the left ventricle became cyanotic up to and slightly beyond the interventricular grooves. Occasionally, ecchymotic spots appeared over the left ventricle. The right ventricle maintained its normal color except for a strip adjacent to the interventricular grooves.

That the extreme restriction of the inorganic salts of the diet of the albino rat produces striking changes in the composition of the blood has been repeatedly demonstrated in this laboratory. 1 , 2 , 3 A marked increase in the number of erythrocytes occurs, accompanied by a progressive decrease in the concentration of hemoglobin. As yet, no satisfactory explanation for the phenomenon has been obtained. The increase in erythrocytes cannot be due to a diminution in blood volume since the plasma and total blood volume remain within normal limits. 2 Nor is there any evidence of an increased rate of erythrocyte formation, inasmuch as no apparent reticulosis occurs during the period of the rapid increase in the number of erythrocytes. 3 It seemed desirable, therefore, to seek an explanation of the polycythemia in another direction, and the possibility that a decreased rate of destruction of erythrocytes leading to a “passive accumulation” of cells might be involved seemed worthy of investigation. As an index to the rate of erythrocyte destruction, the bilirubin content of the serum is usually employed since, given a normal liver and biliary system, there appears to be a direct relation between the two. 4 , 5 If the increase in the number of erythrocytes in the blood of the “low-salt” rats is a result of a decreased rate of cell destruction there should be some detectable decrease in the concentration of bilirubin in the serum. If, however, cell destruction proceeds at a normal rate, a concentration of serum bilirubin equal to or perhaps slightly greater than normal is to be expected.

The same procedure was followed in this experiment as in a previous one. 3 Male albino rats of the Connecticut Agricultural Experiment Station strain weighing from 40 to 50 gm. at weaning (21 days of age) were placed in individual cages and fed the stock colony ration. 6 Those animals which attained a weight of 120 ± 4 gm. at 35 ± 2 days of age were divided into 2 groups: one, the control group (12 rats), received an “adequate synthetic” ration; the other (10 rats), received a diet extremely poor in inorganic salts. Both diets were fed ad libitum. The composition of these diets and of the vitamin supplements have been described in an earlier publication. 3 Quantitative analyses, recently reported, 7 of the amounts of the chief inorganic constituents present in both the “adequate” and “low-salt” diets, serve to emphasize the extreme degree of the limitation of inorganic elements in the latter ration.

In a previous publication 1 attention was called to the fact that the dog prostrate from adrenal insufficiency, with greatly diminished volume of circulating fluid, low arterial pressure, marked hemoconcentration and exhibiting symptoms of severe dehydration and shock, retains large quantities of fluid apparently immobilized within the tissues. This tissue fluid was shown to be sufficient in quantity, when mobilized and shifted to the blood stream under the influence of cortical hormone, to revive the animal from collapse to the point where all symptoms disappear and activity, vigor and appetite return to normal.

The present communication is concerned with the relation of this internal fluid shift to serum sodium and chloride changes.

The dog exhibiting symptoms of adrenal insufficiency is unable to dilute its blood, i. e., shift the tissue fluids to the blood stream to bolster up a failing circulation. However, within a few hours following hormone administration blood dilution occurs, the hemoconcentration decreases along with a rise in arterial pressure, all symptoms disappear and activity and vigor return. Accompanying these changes is a marked diuresis with outpouring of a large volume of urine.

We assumed on the basis of the experiments of Loeb, et al., 2 and Harrop, et al., 3 that the cortical hormone probably mobilized and redistributed sodium and chloride along with the tissue fluids. Investigation of this point, however, reveals that following hormone administration to dogs in severe insufficiency the shift of fluid from tissues to Mood stream occurs despite no elevation of the serum sodium and chloride levels. On the contrary, these electrolytes may even decrease. The essential data obtained from a representative experiment are given in Table I.

It is only very recently that anterior pituitary tissue has been successfully grafted. Hohlweg and Junkmann, 1 Gardner and Hill, 2 and Haterius, Schweizer, and Charipper 3 have all obtained viable grafts with varying success. May 4 cited a slight increase in body size and a repair of the testes as evidence of function in 2 hypophysectomized rats bearing intraocular pituitary implants obtained from new-born rats.

In the mouse a genetical uniformity of host and donor is necessary for successful intra-testicular grafts of pituitary tissue (Strong 5 ). Will the anterior lobe become functional when implanted in the testis where pressure is slight and vascularization of the graft normal? A large amount of data has been collected which very definitely establishes function of such grafts.

Brief notes on only 2 representative cases are here presented. Intra-testicular grafts become well established and vascularized within 3 weeks. Grafts which have been in the host animal for shorter periods have not been examined. Mouse 103 received an intra-testicular graft of a sister pituitary when 50 days old. The animal was hypophysectomized at 73 days of age and was killed 120 days later. The mouse ejaculated when killed with gas just as many normal male mice do. The seminal vesicles were of normal size and were filled with secretion. The testes were normal in size, and pituitary tissue could not be observed macroscopically in the sella. Serial sections of the sella showed a very tiny cluster of pituitary cells. The fragment, measuring not more than 1.5% of the volume of a normal mouse pituitary, consisted largely of intermediate and posterior lobe tissue. Such a small piece of pars anterior did not maintain the gonads of controls and other experimental animals. Our data show that somewhat more than 5% of the anterior lobe (exclusive of posterior and intermediate lobes) must be left in the sella in order to maintain the gonads. Such animals may be, therefore, considered as physiologically hypophysectomized as sexual activity and adrenal cortex are not maintained normally.

It has been shown^1-4 that, in a significant proportion of cases, rheumatoid arthritis sera agglutinate certain strains of hemolytic streptococci and precipitate group-specific fractions of these organisms. Lancefield's demonstration of distinct serological groups within the hemolytic streptococci 5 , 6 , 7 has made possible a more detailed study of the nature and significance of these immunological reactions. Agglutination and precipitation tests were carried out with rheumatoid arthritis and control sera using representative strains of Groups A to G as agglutinogens and group-specific extracts as precipitinogens. Similar studies were recently carried out by McEwen, Chasis and Alexander. 8

Agglutination. Seventy-six rheumatoid arthritis sera were examined for agglutinins against Group A organisms, 60 against group B, 47 against group C, 67 against group D, 15 against group E, 15 against group F, and 14 against group G by the technic previously described. 3 The results were classified as “positive,” “doubtful” or “negative”. In classifying the reactions both the character of the agglutination and the titer were taken into consideration. Sera agglutinating in a titer of 1:160 or higher were considered “positive” only if they gave a strong reaction in lower dilutions. Sera agglutinating in a titer of 1:160, 1:320 or even higher were considered “doubtful” unless the reactions in the lower dilutions were definitely more marked than in the higher dilutions. Sera were classified as “negative” when agglutination occurred in dilutions of less than 1:40.

The findings of Todd, 1 Coburn and Pauli, 2 and others indicate that the antistreptolysin test furnishes an accurate method for determining the presence of recent infection with hemolytic streptococci of group A. Coburn and Pauli showed that, following hemolytic streptococcal infections, the antistreptolysin titer of the serum rises rapidly, remains elevated for a variable number of months and gradually tends to return to normal values. In nearly all cases in their series there was an appreciable fall in titer within 6 months and within 2 years over half the cases had reached natural levels. Myers and Keefer 3 studied the antistreptolysin titer in 13 cases of active rheumatoid arthritis and found that, although the titers varied considerably, the average was not greater than in normal individuals. They stated that “in most of these patients the disease had existed for many months.” Blair and Hallman 4 determined the antistreptolysin titer in 45 patients with rheumatoid arthritis and found it to be above the normal range (over 100 units per cc.) in one-third of the cases. In their report no reference was made to the duration of the disease at the time of the observations or to the nature of the onset.

We have determined the antistreptolysin titer in 219 cases of rheumatoid arthritis, including 22 cases of rheumatoid spondylitis (Marie-Strümpell). Preliminary observations, indicated that in the great majority of well established cases the titer was within normal limits, irrespective of the degree of clinical activity present. However, when the sera of early cases with acute onset were examined quite different results were obtained. In the present report, therefore, the cases are divided into 2 groups, “early” and “late” cases. For convenience, cases of less than one year's duration are considered as “early” and those of longer duration are considered as “late”. In addition, the nature of onset, whether acute, subacute or insidious, has been recorded in each instance.

The Gregory-Pascoe 1 reaction was found to be specific for the determination of cholic acid. The Reinhold and Wilson 2 modification of this method was combined with the procedure of Harwood 3 in such a fashion as to obviate both the use of the color filter and the necessity of carrying out the reaction in the presence of alcohol. In addition it was found that the large amount of deoxycholic acid present in human bile, although not taking part in the reaction, formed a precipitate which interfered markedly with the colorimetric procedure. This difficulty was overcome by the addition of alcohol after the Gregory-Pascoe reaction had been carried out to completion.

Method. 1-2 cc. of human or canine gall bladder bile, 3-5 cc. of canine fistula bile, 10-30 cc. of human fistula bile or 20-35 cc. of duodenal drainage material are the usual quantities measured out into a 50 cc. centrifuge tube. Three cc. of 2N KOH is then added and thoroughly mixed with the bile. When small quantities of concentrated bile are used, 15-20 cc. of water is added, the presence of alkali preventing the precipitation of protein. Following this, 3 cc. of 40% zinc sulphate is added drop by drop with constant stirring. When the contents are thoroughly mixed, the supernatant fluid is separated by centrifuging and the precipitate washed 3 times with 15 cc. quantities of hot water. The 4 water washings are poured into a 100 cc. volumetric flask. The precipitate is then washed 3 times with 20 cc. portions of 95% alcohol. For the first washing cold alcohol is used, and for the subsequent washings hot alcohol. The alcohol washings are all transferred to a 100 cc. beaker and dried on a water bath. The residue, taken up with 5 cc. 2N K₂CO₃, is combined with the water washings in the 100 cc. flask, which is made up to volume.

By the combination of 3 different methods the analysis of bile for taurocholic acid, glycocholic acid, total conjugated bile acids, cholic acid, deoxycholic acid, total bile acids, and free bile acids has been achieved. It has also been possible to analyze duodenal drainage material for cholic acid, deoxycholic acid and total bile acids.

Methods. 1. The Cuny 1 modification of the Schmidt-Dart 2 method is used to determine the taurine and glycine nitrogen percentage. The first is multiplied by 33.87 and the second by 28.29 to give the bile acids conjugated with taurine and with glycine. The sum of these 2 figures gives the total bile acids that are conjugated. 2. Cholic acid is determined by a method outlined in the previous paper. 3 3. The total bile acids are estimated by an iron precipitation method based partly on principles devised by Harwood. 4 After the bile acids have been extracted and made up to 100 cc. in a volumetric flask and the cholic acid determined, the amount of solution in the volumetric flask, estimated to contain between 40-60 mg. of bile acids, is measured out into a 150 cc. beaker. This calculation is derived from the known amount of cholic acid present in the volumetric flask. The bile acids of dog bile may be considered to be approximately 100% cholic acid, while those of human bile can be assumed to be 50% cholic acid. Thus if the volumetric flask is found to contain 100 mg. cholic acid derived from dog bile, 50 cc. of the solution is measured in the beaker. If the extract is derived from human bile, only 25 cc. is taken.

The superficial and total blood flows of the forearm and hand were studied by the use of 4 different methods. Over 100 experiments were performed on 15 healthy adults, including 2 women. Each experiment was preceded by a rest of 15 to 30 minutes, and equally lengthy periods of control observation. The experiments lasted from 1 to 3 hours.

The plethysmograph indicates that a deep breath causes a fall in forearm volume of from 5 to 10 cc. The shrinkage begins near the end of inspiration, and reaches a maximum in from 15 to 30 seconds. Recovery is slower but complete.

The fall in volume may be analyzed further by the discontinuous blood flow method of Hewlett and Zwaluwenberg. 1 While the hand and part of the forearm are in the plethysmograph, the venous outflow is obstructed suddenly by abruptly raising to 70 or 80 mm. Hg. the pressure within a small cuff placed just centrad to the plethysmograph. The resultant changes in forearm volume indicate the rate of inflow of blood through the still open arteries. Although this method of measuring is discontinuous, it has the advantage of indicating the total blood flow through the forearm.

This method shows that on deep inspiration the blood flow begins to slow in mid-inspiration; reaches its maximum at the end of inspiration, and disappears rapidly. A record obtained in mid-expiration shows the flow already partly recovered. On deep inspiration, the blood flow through the forearm is stopped completely, for from 1 to 15 seconds, depending upon the rate and depth of the inspiration and the subject of the experiment. Experiments performed at every phase of the respiratory cycle indicate that the diminished flow is due to the inspiratory movement, for recovery becomes complete even when the breath is held at inspiration.

In the course of an inquiry into the significance of the proteinuria observed in a case of lymphogranuloma inguinale, the protein content of the serum was determined. An unanticipated hyperproteinemia was found, confirmed by 2 subsequent examinations. Further investigation in 12 cases of lymphogranuloma inguinale revealed high serum protein values in all but 2 cases. The hyperproteinemia was more marked in the advanced stages of the disease (cases 1-7, Table I), particularly in females with rectal stricture, than in those patients presenting suppurating inguinal buboes or proctitis. We had no opportunity to study patients with initial lesions.

There was a relative and absolute increase in serum globulin in 11 cases, shown to be due to an increase in euglobulin in 2 cases, with reversal of the albumin: globulin ratio in 9 instances (Table I).∗ Serum albumin was less than 3.7% in only 2 cases. The fibrinogen content of the plasma and the non-protein nitrogen content of the serum were within normal limits. The serum calcium was 9.1, 10.2, 11.3 and 10.3 mg. % in cases 1, 4, 6, and 11 respectively. The osmotic pressure of the serum in case 1, against normal saline solution buffered at pH 7.4, was within normal limits.

The erythrocytic sedimentation rate 1 , 2 , was increased in all but 2 of the cases examined, but was within normal limits in defibrinated blood (5, 2, 6, 5 mm. in 1 hour in cases 8, 9, 10, and 11 respectively). The Takata-Ara test was found to be positive in the 3 cases examined. The Ray globulin flocculation test, carried out as with kala-azar serum, was negative. The urine contained traces of protein in 5 of the 10 cases examined.

A. Effect of oestrus on the life-span of the adrenalectomized dog not receiving adrenal cortical hormone. Rogoff and Stewart 1 first demonstrated that the bitch, if adrenalectomized when in heat, could survive in normal health and without treatment of any kind, for as long as 65 days. This observation has been repeatedly confirmed in this laboratory by withholding hormone from the animal in heat. 2

During the past 2 years we have studied 5 dogs in heat. The span of normal health varied, but in all cases it was greatly in excess of the longest time interval we have observed to intervene between cortical extract withdrawal and onset of adrenal insufficiency. The dogs were maintained without cortical hormone as follows: One dog 60 days; one 57 days; one 47 days; one 45 days; and one 40 days. At the end of these intervals adrenal insufficiency symptoms were evident and cortical hormone was injected. The arterial pressure, serum sodium, chloride, blood urea nitrogen, hemoglobin and glucose remained within normal limits throughout the period of good health.

Oestrus was experimentally induced in one adrenalectomized bitch by injections of extracts of menopause urine containing folliclestimulating hormone of the anterior pituitary, followed by Follutein (pregnancy urine extract, E. R. Squibb & Sons). The animal came into heat 10 days following the first injection. Cortical hormone was then withheld while the injections of gonadotropic factors were continued. The dog remained free from symptoms and in good health for 25 days.*

At the present time two methods for assaying progestin are in use, one involving the use of mature female rabbits, and the other, immature female rabbits. In connection with their work which showed that progestational proliferation is due to a specific corpus luteum hormone, Corner and Allen 1 developed a method for the standardization of corpus luteum extracts. The test devised consists of injecting the extract over a period of 5 days into a sexually mature doe castrated 18 hours after mating. A rabbit unit (Corner-Allen unit) is then defined as the minimum quantity of progestin which, divided into 5 equal daily doses, produces on the sixth day a state of the uterus equal to that of the eighth day of a normal pregnancy. A uterus in this state shows a +++ or ++++ proliferation.

Allen 2 discovered that, in the immature rabbit weighing from 575 to 1647 gm., progestational proliferation can be induced in the uterus by injections of progestin, provided the uterus is first influenced by a series of oestrin injections. Only a small percentage of the immature animals responded to the progestin treatment when it was not preceded by the oestrin injections. Clauberg 3 , 4 confirmed the findings of Allen 2 on immature rabbits and established the Clauberg unit which is measured in the same way as the Corner-Allen unit except that 600 gm. rabbits are used and are primed with a total of 80 international units of oestrin, administered in 8 daily injections of 10 international units each, preceding the 5 daily injections of progestin.

It was reported 1 , 2 that unilateral adrenalectomy or interference with epinephrine secretion (by adrenal denervation, etc.), ameliorates diabetes in depancreatized dogs. It was stated that in animals already diabetic the adrenal operations diminished the glycosuria; also, that in dogs previously subjected to these operations, pancreatectomy either failed to cause or caused only a mild degree of glycosuria. Insulin requirement (on a constant diet) for controlling the glycosuria was reported to be decidedly less as the result of the adrenal operations which also were said to render animals more sensitive to insulin. In general, these statements agree with conclusions previously reported by others.

We find it necessary to revise the interpretations made from the earlier experiments and published in preliminary reports. More extensive investigation, under better conditions, and quantitative studies of the rate of epinephrine output (by the method employed by Stewart and Rogoff), on an adequate series of control animals, indicate that we were misled in the previous evaluation of our experimental data. Indeed, we are now convinced that satisfactory experiments lead to the conclusion that partial adrenalectomy, or operations for suppressing epinephrine secretion (adrenal denervation, with or without mechanical destruction of medulla), do not prevent or diminish the experimental diabetes produced by total pancreatectomy in dogs.

Although quantitative determinations demonstrated marked reduction or suppression of epinephrine output, in some of the preliminary experiments on depancreatized animals subjected to the adrenal operations, the assumption that this is responsible for reduced insulin requirement proved to be incorrect when a sufficient number of satisfactory control experiments became available. It is not significant that animals surviving only a short time after operation required less insulin, sometimes little or none, since it is known that hyperglycemia and glycosuria may be low or may fail to develop if the post-operative condition of the animal is not good. Of those surviving much longer, the insulin requirement usually was decidedly less after a few weeks than shortly after pancreatectomy.

It has been recently demonstrated that the soluble specific substance of Pneumococcus type I, as originally prepared by Heidelberger, Goebel and Avery, represented a form of the polysaccharide lacking the acetyl group which, when present, conferred upon the substance distinct physico-chemical and immunological properties. 1 , 2 In the case of the type specific polysaccharide from Pneumococcus type III, purified according to the procedure described by the same authors, certain observations such as those of Ward 3 have led us to believe that an analogous alteration in chemical structure may have occurred during the chemical manipulations required for its purification. We have, therefore, attempted to prepare this material by a method which avoids insofar as possible the use of strong acids.

Six- to 8-day cultures of Pneumococcus type III in dextrose phosphate broth were concentrated over a boiling water bath to one-tenth of the original volume. The concentrate was precipitated several times with about 1.2 volumes of alcohol. Proteins were precipitated by careful addition of 1N acetic acid, until maximal precipitation was attained. Neutralization and reprecipitation with the acid was twice repeated.

After the last acid precipitation, the supernatant was made alkaline and precipitated with alcohol. The precipitate was dissolved in H₂O and 2 volumes of saturated ammonium sulfate added, the scanty precipitate discarded and the sulfate removed by dialysis against distilled water. The solution of polysaccharide in the dialyzing sack was precipitated with 1.2 volumes of alcohol. After resolution in H₂O and precipitation with acetone, it was washed with alcohol and ether, and dried in vacuo.

While studying the biological properties of hemolytic streptococci isolated from different sources in this Island, we grouped 46 strains using the precipitin reaction described by Lancefield. 1

All extracts were first tested against group A serum. Those that gave a positive test were not tried against the other sera because cross reactions have never been observed with particular A antiserum. 2 Those that gave a doubtful or a negative result were tested in the same B, C, D, E, F, G and H antisera.

Out of 11 strains from diseased tonsils, 7 belonged to group A, 2 to group C and 2 remained unclassified. All the strains isolated from chronic discharging lesions of the face, from abscesses, from otitis media, from osteomyelitis and from the throat of patients with a diagnosis of scarlet fever were group A. Strains M₆ and M₇, isolated from the throat of cases of agranulocytic angina and rheumatoid arthritis respectively, remained unclassified, but they were not group A.

Of 15 strains from cases of recurrent lymphangitis, 14 were group A and one, group G. Strain D₁, from eczematoid dermatitis, belonged to Group C. Of 2 strains from septicemia, one was group A and the other, G.

Of the 46 strains tested, 37 were group A, 3, group C, 2, group G, and 4 remained unclassified.

It was noted that rats hypophysectomized∗ for 6 to 8 weeks, with metabolic rates 70% of those of normal controls (as determined gravimetrically by the Haldane method), frequently showed high respiratory quotients (R.Q.) after a fasting period of 24 hours. Values as high as 0.78 to 0.81 were sometimes observed, the average of a series of 19 observations being 0.75, while the average of 19 controls was 0.72. If the hypophysectomized rats were fasted longer than 30 hours, the R.Q. usually dropped to 0.73, though in a few instances it remained high for nearly 40 hours. Changes in the acid-base equilibrium were apparently not responsible for the high R.Q., as shown by determinations of the CO₂ content of the blood. Glycogen determinations after a 24-hour fast showed that the hypophysectomized animals contained considerably less muscle and liver glycogen than did normal controls. Hypoglycemia was always present, the blood sugar being around 50 mg. %.

These findings suggested that the regulatory mechanism which leads to the preservation of glycogen during fasting is disturbed after hypophysectomy. The low glycogen reserves and the tendency to hypoglycemia in hypophysectomized animals might be due, among other factors, to the removal of an inhibitory influence on carbohydrate oxidation. In order to test this assumption, a standard amount of glucose (500 mg. per 100 gm. of body weight) was fed to normal and hypophysectomized rats after a fasting period of 24 hours, the aim being to determine (a) how much glucose was oxidized in the first few hours during the period of active absorption and (b) what length of time was required for the complete disposal of the ingested sugar by oxidation.

In former investigations we have shown that the growth-promoting influence of acid extract of anterior pituitary gland of cattle on bone and cartilage of young guinea pigs is exerted without the intervention of the thyroid gland. 1 Correspondingly we wished to determine whether the accelerated healing tendency of fractured bones, caused by the extract, 2 is also independent of the action of the thyroid.

The experiments were carried out in 26 winter and spring guinea pigs, on the average weighing from 120 to 280 gm. Under deep ether anesthesia the tibia and fibula of the right hind leg were broken about the middle of their shafts and bandaged according to the method described previously. Both lobes of the thyroid were completely removed immediately after fracturing. Injury to the parathyroids was carefully avoided. Sixteen of the animals were injected daily with 1 to 11/2 cc. of extract intraperitoneally for periods varying from 6 to 21 days, the first injection being applied immediately after thyroidectomy. Ten guinea pigs were not injected and served as controls. After 6, 10, 14, and 21 days the fractured bones were removed in toto and specimens prepared for microscopical examination as previously described. 3

As to the gross appearance of the callus in the injected and thyroidectomized animals, it was invariably firmer and harder than that of the non-injected thyroidectomized guinea pigs. This fact was noted as early as 6 days after fracturing, but even more distinctly at subsequent periods. In all cases a complete union by a dense, bony, calcified callus had taken place after 21 days, while in the non-injected animals the callus was looser and softer at that period. There was no marked difference in gross appearance between the callus in injected thyroidectomized guinea pigs and in those with intact thyroid glands, which had been injected.

In the course of some experiments with the Brown-Pearce rabbit tumor 1 , 2 , 3 an unusual inhibitory effect was noted following a single injection of 0.3 cc. of a Berkefeld “V” filtrate of fresh tumor given 2 weeks before the tumor inoculation. The experiment had not previously been tried and has not yet been repeated. It would not be reported here except that it is the only case of complete failure of the Brown-Pearce rabbit tumor to grow following intratesticular inoculation in some 7 years of experience involving approximately 100 different transplantations of the tumor, and approximately 1000 rabbits. This experience includes studies on some 20 pure breeds, and on various types of hybrids, particularly with the type of brown-gray hybrids used in the experiment; and the probability of a rabbit being free of tumor at the end of a 2 months'observation period was calculated to be 0.2306. The occurrence of all 6 animals in an experiment being negative by accident is, therefore, very unlikely (P = 0.01—).

Furthermore the 6 animals of this unusual group were identical as to source, age, weight, coat color, vigor, size of testicles, and body build with 36 other rabbits not receiving the filtrate and composing 6 additional groups inoculated on the same day in staggered order 3 in the same testicle with the same dosage of the same fresh tumor. One of the 6 additional groups was a control group and was not otherwise treated. Analyzed per animal inoculated, the incidence of primary tumors in this group was 50% and the tumors averaged 5.8 cc.; the incidence of metastatic tumors was 83.3%; the metastatic tumor averaged 53.0 cc. in volume per animal and was distributed over 6.3 metastatic foci; the mortality was 61.1% over a period of 59 days. The 5 additional groups were treated with various other homologous materials before tumor inoculation and had an incidence of primary tumors of 40% which averaged 3.8 cc.; the incidence of metastatic tumors varied from 80 to 100% per group, and averaged 88.0%; the volume of metastatic tumor per group varied from 40 to 162 cc. per animal and averaged 94 cc.; the number of metastatic foci varied from 10.2 to 24.2 per group and averaged 13.9 foci; the mortality varied from 60 to 100% and averaged 71.4%; the longevity varied from 48 to 61 days and averaged 54.7 days.