Experimental Biology and Medicine

An illness on the island of Jamaica known locally as “vomiting sickness” is thought to be, at least in some cases, a form of poisoning caused by eating the fruit of the akee (Blighia sapida). Although it is probably true that the name “Vomiting sickness” has been applied to etiologically unrelated affections, there is strong evidence 1 implicating the akee in a serious and frequently fatal sickness. Deaths after eating akees are seemingly reported sometimes as vomiting sickness, sometimes as akee poisoning. The total deaths registered annually in recent years under these designations have run as high as 350 (1921). The akee in an immature or unripe condition has been reputed to be particularly dangerous.

Considerable unwillingness to accept these conclusions has been manifest, chiefly on the ground that the akee is a common market article in Kingston and elsewhere in Jamaica, and, in its season (November-March), is eaten daily by hundreds, probably by thousands, of persons with perfect impunity. It is also pointed out that with the improvement in water supplies that has taken place in several localities the “vomiting sickness” has greatly declined.

The edible portion of the akee is the arillus, a fleshy envelope attached to the seed. Arilli boiled or fried have been repeatedly eaten by one of us (E.O.J.) as by other Americans and European residents in Jamaica without any sign of discomfort or injury.

Through the kindness of Dr. B. E. Washburn, representative of the International Health Division of the Rockefeller Foundation in Jamaica, we have been able to secure a considerable number of akees in various stages of development. No signs of illness have been observed in experimental animals (cats and monkeys) fed with the arillus alone. The seeds, however, in both the ripe and unripe fruits contain a water-soluble poison which affects cats and monkeys, causing serious vomiting, and occasionally death when a large amount is given. The ovary wall or husk of the fruit also contains a toxic substance that is extracted with boiling water. When the whole fruit is chopped fine and extracted with boiling water, the extracts from both ripe and unripe fruits are poisonous.

The microscopic anatomy of the incisors was studied in 2 groups of rats:

A. 29 rats that were given 1 to 15 injections of parathyroid hormone. Doses varied from 10 to 150 units. Weight: 85-145 gm. B. 7 controls from same colony.

Significant alterations were observed only in Group A: 1. Enamel hypoplasia was found to be formed at the time of the first injection in 5 animals. 2. The alveolar bone showed an abnormal increase in osteoclasts in 14 and a fibrous change of the bone marrow in 3 of the animals that were given 3 or more injections of parathyroid hormone. 3. The principal changes were found in the dentin. Each experimental animal showed a primary hypocalcified stripe in the dentin that was being calcified during the immediate effect of the first injection and a secondary hypercaleified stripe in the dentin that was being calcified subsequently. The extent of the secondary stripe varied with the number and unitage of the doses injected and the duration of the experiment.

Two animals that were given 12 daily injections of 10 units each of parathyroid hormone showed a normal histologic picture.

The absence of reaction in the 3 animals that were given inactivated hormone preparation gives conclusive proof that the effects observed were due to the parathyroid hormone and not to any foreign protein impurities.

It is believed that the experimental pattern in the dentin represents its response to the changes in the calcium and phosphorous metabolism induced by the injections of the parathyroid hormone. The dentin reaction can be explained best by the theory that the parathyroid hormone controls a fraction of the serum calcium. The findings suggest the feasibility of a method of standardization of parathyroid hormone based on the changes shown to occur in the dentin.

Following the introduction by Fogelson 1 of gastric mucin therapy for gastro-duodenal ulcerative disease, a method for the assay of commercial preparations became desirable.

The commercial gastric mucin is prepared by following the customary procedure for commercial pepsin, in which the mucin dough is separated from the pepsin by precipitation at its iso-electric point. It is then purified by repeated washing with 70% alcohol; 2 mixed with required base, and diluted with peptone to decrease the viscosity to a practical state for administration to patients. Although it is the opinion of Hurst 3 that “mucus differs from other proteins in the extreme slowness with which it undergoes peptic digestion…,” nitrogen and reduction values to be presented for commercial mucin suggest the possibility of a partial digestion, when compared with values obtained for gastric mucin by Webster and Komarov. 4 This phase of the subject is being investigated.

Commercial samples were found to vary greatly as to viscosity, acid combining power, sulphur, nitrogen, and mucoitin content. Methods were therefore sought by which mucin could be separated from the commercial diluents and contaminants before assay. For this purpose we have used an additional peptic digestion followed by precipitation with 70% alcohol. By weighing the precipitated material, comparative values for the mucin content of the samples may be obtained. Our data shows this procedure to give a product with characteristic N₂ content and reduction values. Details of the procedure are as follows: Dissolve 1 gm. commercial mucin and 0.1 gm. pepsin (Wilson's 10,000 units/gram) together in 25 ml. of 0.25% (0.068 N) HC1. This strength acid was found to give a pH approximating 2 with most samples. Digest in the incubator at 37°C. over night, then pipette 10 cc. aliquot samples into weighed 50 cc. centrifuge tubes, and add 28 ml. of 95% alcohol. The contents of the tubes are thoroughly mixed and centrifuged at high speed until the precipitate is well packed (5-10 min.).

It was shown 1 that commercial gastric mucin when subjected to peptic digestion and subsequent precipitation with alcohol at 70% concentration, yielded upon hydrolysis characteristic values for total nitrogen and total reducing substances. This suggested the use of a reduction method for the quantitation of mucin. Obviously before such a method can be applied to gastric juice, it is necessary to remove all other reducing substances, or to use a procedure for obtaining gastric juice not subject to their presence. The former procedure would be capable of wider application and work is in progress with this end in view. Our present interest is in the variations in mucin content of gastric secretion, when juice is obtained from the fasting stomach by means of an alcohol test meal under conditions excluding other reducing substances.

In following the curve of mucin secretion, the stomach is emptied after a night's fast, basal secretory rate established, then 250 cc. of 7% ethyl alcohol is given and samples removed through a Rehfuss tube at 15 minute intervals for a period of 2 hours. Determinations of total and free acidity, and pepsin, are made. The mucin secretion is followed by the determination of total reducing substances according to the following method: The gastric sample (5 cc. if available) is diluted with an equal volume of 4 N H₂SO₄. It is then hydrolyzed in a boiling water bath for 2 ½ to 3 hours using an air condenser. After cooling, 2 cc. aliquot samples are pipetted into test tubes, the acidity neutralized to phenolphthalein with 1 N KOH or NaOH, and the reduction estimated by the Somogyi modification of the Shaffer-Hartmann method. 2 The results are expressed in milligrams of reducing substances (calculated as glucose) per cubic centimeter of gastric juice.

To illustrate values found, the accompanying graphs are appended.

We have shown 1 , 2 that certain proteins and peptones are acidic in liquid ammonia and react with sodium with liberation of hydrogen. Proteins and peptones liberate approximately one-half of a molecular weight of hydrogen for each gram atom of nitrogen in the substance when an excess of sodium is used.

Hemoglobin and hematin react profoundly differently toward sodium in liquid ammonia than do the other proteins which we have studied. They liberate hydrogen continuously when excess sodium is present and in quantities disproportionate to the acidic groups present. Their behavior resembles the catalytic action of a rusty iron nail in decomposing the ammonated sodium atom to give sodium ion, amide ion, and hydrogen.

The analytical procedure was essentially the same as that described previously. The substances were dried for 48 hours in an electric oven at 80°C. and then kept in a vacuum desiccator over sulphuric acid.

Hemoglobin is not appreciably soluble in liquid ammonia. It swells, softens, and the particles tend to coalesce. Within a few moments, the supernatant liquid is yellow, but slowly turns to a reddish color after 20 min. After standing for 4 hours, the solution is a light green. The change in color may be due to impurities present. There is a persistent increase in rate of boiling of liquid ammonia following the addition of hemoglobin suggesting that ammonation of the molecules is taking place exothermieally.

When 0.1 gm. of sodium is added to 0.1 gm. of hemoglobin in liquid ammonia, hydrogen is evolved rapidly during the first 15 min. After that the rate was slowed up considerably, but it proceeded until the blue color of the solution was decolorized. If more sodium was then added, the reaction would start again and continue until the solution was decolorized again. This could be repeated several times. The volume of hydrogen evolved was far in excess of the amount liberated by the same weight of silk, zein, egg albumin, casein, and edestin. The rate is characteristic of a substance containing a carboxyl group, as an amino acid. Hemin contains 2 carboxyl groups, and globin probably contains some free carboxyl groups, although the basic groups in it are in excess. It is possible that our method of drying did not remove all of the water. Water, being acidic in liquid ammonia, reacts rapidly with sodium liberating hydrogen. Hofner reports that 100 gm. of hemoglobin contains 9.25 gm. of water of crystallization.

In view of evidence (1) that the severity of pancreatic diabetes is greatly diminished by previous hypophysectomy, (2) that hypophysectomy leads to some atrophy of the adrenals and (3) that the adrenals are in some way connected with carbohydrate metabolism, it seemed necessary to reinvestigate in more detail the effects of adrenalectomy on pancreatic diabetes. The literature will not be reviewed here but it is sufficient to say that the reports have been conflicting and inconclusive. Nine dogs have been used in our studies, 5 of which have had the pancreas removed.

In 3 cases the adrenal on one side was removed first and the pancreas removed 10-18 days later. These animals received no insulin and 2 of them are still alive, now 22 days since pancreatectomy. The third animal died from an infection. These animals showed a mild glycosuria but not as severe as is usually found after extirpation of the pancreas. Within a few days the sugar excretion fell to the level found in dogs following hypophysectomy and pancreatectomy. One animal failed to eat part of the time (probably due to a respiratory infection) and under these circumstances no glycosuria was present. In 2 dogs the pancreas was removed first and the animals put on a constant diet with insulin twice daily. Later, one adrenal was removed and in one case the quantity of insulin necessary to maintain the same degree of slight glycosuria was approximately one-half as great as before the adrenalectomy. In the other animal insulin was stopped after the removal of one adrenal and the glycosuria was only mild. The animals are gradually losing weight and becoming weaker, although they apparently feel well and have good appetites. The fasting (overnight) blood sugar of these animals varied from 82 to 290 mg. per 100 cc.

Epithelium surrounding the base of the permanent cuspid teeth, and pulp tissue in 18 young dogs were transplanted to the fascia of the abdominal wall as an autogenous graft, and after 9 to 31 days there was found newly formed enamel, dentine, cementum and bone.

Method. The epithelium was obtained in the following way: Under ether anesthesia a cuneiform osteotomy of the mandible, including the deciduous cuspid tooth was made in dogs of between 3 and 6 weeks of age and this allowed access to the permanent cuspid tooth, which was removed. A circular incision was made equatorially allowing the dental epithelium and pulp to be removed en masse from the calcified tooth shell. The tooth shell itself was always discarded. The soft tissues were gently wiped with moist gauze, were X-rayed, and a section was removed for histological control. The calcium-free material was then transplanted to a pocket between the lateral abdominal muscles, with aseptic pre cautions, and the incision closed with silk.

It was found easy to separate (a) the sheet of ameloblasts with the subjacent vascular connective tissue parietes, hereafter called the ameloblast layer from (b) the odontoblasts and their underlying pulp cells, in this paper called the odontoblast layer. In some experiments the ameloblast and odontoblast layers were separately transplanted, and in others they were transplanted together.

Results. In all cases calcified tissue was found in the transplant after 9 days. When the ameloblast and odontoblast layers were transplanted together in a nearly normal relationship, enamel and dentine were found between the 2 layers and the ameloblasts were present in a normal cylindrical form, as in the original tissue.

The opinion that the mouse is practically immune or at least highly resistant to tuberculous infections seems to be quite generally expressed. Kettle 1 states that the tubercle bacillus is practically nonpathogenic to the mouse in the sense that it does not cause the necrotic or other lesions seen in susceptible animals and does not necessarily kill the animal. Trommsdorff 2 observed that mice were much more susceptible to bovine strains than to human strains when injected intravenously and suggested the possibility of using mice for type differentiation. His work was confirmed by Peters 3 and Binder. 4 Lange 5 used 3 human strains and 2 bovine strains of virulent tubercle bacilli for intravenous injections and observed no marked differences in the susceptibility of mice to the two types.

This report calls attention to the possibility of using white mice for certain types of work in the field of tuberculosis. The relatively low susceptibility of the mouse enhances its value in experiments where chronicity of the disease is desired. The rarity of spontaneous infection and the consistent results from intravenous administration of suitable doses suggest the possibility of a usefulness in virulence tests and possibly for experiments in immunity.

Recently isolated human and bovine strains and a virulent avian strain∗ were used for intraperitoneal and intravenous injections in full grown white mice. The expectation that mice would be resistant to infection led us to use unnecessarily large doses in the earlier experiments. Two or 3 weeks old cultures were emulsified by shaking with glass beads and the concentrations were made such that the desired dose was contained in 0.5 cc. of saline. In many cases the concentration was reduced by allowing the coarse clumps of bacilli to settle out so that the actual dose given was considerably less than that indicated in the table.

We have investigated the nature of the relationship between calcium and protein in blood serum and other protein-containing fluids, using the frog-heart method, previously described, 1 for direct estimation of calcium ion concentrations, and have accumulated evidence concerning the hypothesis 2 that there is a substance, like citrate, and other than protein, which binds calcium in serum and other fluids of the body. When purified electrolyte-free serum proteins (beef) are dissolved in solutions of approximately the electrolyte concentration of the serum, calcium combines with the protein in a manner analogous to that previously described for citrate. 3 The combination represents an equilibrium between protein, total calcium, and calcium ions, and this equilibrium may as a first approximation be described by the mass law equation

When normal human serum is diluted with a solution containing approximately the electrolyte concentration of the serum, but without calcium, and when calcium is added to the various dilutions, the same relationship is found. From observations on both serum and purified serum proteins at pH 7.35 (approximately) a value of 2.22 has been derived for pK_CaProt when grams of protein are reduced to base combining capacity by use of the factors given by Van Slyke, Hastings, Hiller, and Sendroy. 4 Agreement of the values for pK_CaProt obtained from protein-containing fluids and from purified serum proteins indicates that it is unnecessary to postulate any appreciable amount of other calcium-binding substances in human serum.

Table I includes all observations on undiluted human material in which total calcium, total protein, and calcium ions have been determined upon the same specimens. In all cases the factor used for reducing grams of protein to base combining capacity at pH 7.35 is 0.243, which assumes a normal distribution of albumin and globulin. Further work may indicate the desirability of applying separate factors to these proteins.

Kyes and Strauser 1 studied the effect of heparin on acute anaphylactic shock in the pigeon and concluded that heparin as a reagent inhibiting fibrin formation so markedly eliminates anaphylactic shock reactions as to give support to the view that there is a causal relation between fibrin emboli and the symptoms of shock. There are conflicting reports on the effect of heparin on anaphylactic shock in the guinea pig, Williams and Van de Carr 2 reporting that heparin prevents or reduces the shock symptoms while neither Reed and Lamson 3 nor Hyde 4 found any such protection. So far as we are aware, similar experiments have not been performed on the dog. Dogs were sensitized to horse serum. After varying intervals of time, the dogs were anesthetized with ether, the femoral vein exposed for injections and the carotid artery cannulated for recording blood pressure. Heparin in saline was then injected intravenously and after varying time intervals, the critical dose of serum was injected. The details of the experiments and the results appear in Table I.

In 59 control experiments in which the amounts of the sensitizing and shocking doses and the time intervals are fairly approximate to those here reported, the distribution of the varying grades of severity of shock were as follows: ++++ 28%, +++ 13%, ++ 48%, + 8%, and 0 3%. In the heparin series, the corresponding ratios are ++++ 6%, +++ 18%, ++ 64%, + 6%, and 0 6%. We do not believe that the difference between the 2 series is significant. Certainly there is not the marked modifying influence of heparin reported by Kyes and Strauser in the case of the pigeon or by Van de Carr and Williams in the case of the guinea pig. Consequently we conclude that these experiments lend no support to the theory that the formation of fibrin emboli condition the symptoms of anaphylactic shock in the dog.

Fresh whole adrenal glands (20 gm. portions) were thoroughly crushed in a mortar with sand (3 gm. sand per gm. of gland) and 50 cc. distilled water, until a very finely divided pulpy mass was formed. This pulp was placed in collodion bags and dialyzed repeatedly against 200 cc. portions of distilled water. Dialyzing for 5 minutes results in a bluish dialysate, 5 cc. of which causes a powerful pressor response when injected intravenously in a 10 kg. dog. The dialysates have been removed and fresh portions (200 cc.) of distilled water added daily at varying intervals for 35 days. Intravenous injections in a 10 kg. dog of 1/2 to 5 cc. of the different dialysates causes a rise in blood pressure comparable to the effect obtained by injecting 1/2 to 1 cc. of 1/20,000 P. D. and Co. adrenalin. Chemical analysis, according to the iodine thyosulphate titration, indicates that the total yield per gram of fresh gland over 30 days'dialyzing is equivalent to 10 mg. P. D. and Co. adrenalin. Biological assay as determined by the blood pressor effect on anesthetized dogs, shows even a greater physiological effect than the quantitative chemical analysis indicates.

If autolysis of the crushed adrenal glands in the dialyzing bag be held in abeyance by potassium cyanide, the dialysate contains no pressor principle. On the other hand KCN did not affect the pressor activity of untreated dialysates nor that of P. D. and Co. adrenalin. Boiling for 1 minute does not destroy the pressor effect of the dialysate; neither does digestion with pepsin or trypsin. Passing the dialysate through animal charcoal removes the pressor principle, as does adsorption on aluminum hydroxide.

The pressor substance contained in the dialysate is not precipitated by concentrated NH₄OH, although a heavy flocculent precipitate forms. The mother liquor from this precipitate contains a powerful pressor substance.

Complete pancreatectomy in the dog causes an increase of over 100% in the non-glucose reducing substance of venous blood. (Table.) This substance is determined quantitatively by the iodine titration method of King et al. for blood glutathione as modified by Mancorps and Schmid. 1 At this stage of the research we assume this substance to be glutathione; however, the method employed is not specific but may include such substances as thioneine and cystein. Ascorbic acid also interferes with this titration. The above mentioned interfering substances are probably not present in quantities sufficiently great to cause appreciable error in our interpretations. Further work on this phase is in progress.

Section of the splanchnics in the dog causes a decrease of approximately 50% in the blood glutathione, but no effect on the experimental diabetes following pancreatectomy. Acetone given by mouth in dosages of 6 ce. per kilo of body weight causes this gluta-thione-like substance to increase approximately 200% above normal and does not return to normal for approximately 3 days. Removal of the pancreas on the blood glutathione has been studied by Dott D. Ferrari. 2 Our results are at variance with his findings. Blood glutathione determinations before pancreatectomy in his dogs are not published.

In the course of a study of serum phosphatase activity in the rabbit, dog and rat, it was found that bleeding, diet, distemper, and starvation exert a marked influence upon the valueis obtained. Bodansky and Jaffe 1 have pointed out that a relationship exists between the type of diet, the nutritional state of the animal, and the plasma phosphatase activity. For convenience of presentation, data is given under the 4 respective headings. Serum phosphatase activity was determined according to the method of Bodansky. 2

1. The influence of bleeding upon serum phosphatase activity was determined in adult rabbits. The serum obtained from 25 cc. of blood, drawn from the heart of rabbits, gave average values in a group of 14 animals for calcium of 13.96 mg./100 cc, for inorganic phosphorus 5.04 mg./100 cc, and for phosphatase activity 4.14 mg./100 cc. A second 25 cc sample of blood drawn 8 to 15 days later showed the influence of the previous bleeding upon serum constituents of otherwise normal rabbits maintained on the same diet. The values for serum calcium averaged 14.13 mg./100 cc, for inorganic phosphorus 3.74 mg./100 cc, and for phosphatase activity 2.25 mg./100 cc. These data show that bleeding caused a marked decrease in serum phosphatase.

2. The influence of a high protein or carbohydrate diet on the serum phosphatase activity was studied. The diet of a number of dogs was alternated from bread and meal mash to one of beef hearts. The influence of these 2 diets on the serum phosphatase and inorganic phosphorus is shown in Table I.

The action of methylene blue in restoring the respiration of cyanide inhibited tissues (Gerard 1 ) has led to the clinical and apparently successful use of the dye as an antidote for cyanide poisoning (Brooks 2 ). It has been urged that methylene blue acts in the intact organism not directly by substituting in the cells for the inactivated respiratory enzymes, but indirectly by removing the cyanide and releasing the enzyme. Cyanide would thus be removed by the formation of the stable cyanmethemoglobin, from it and methemoglobin; formed in turn from the blood pigment by methylene blue (Wendel 3 ). An antagonism of methemoglobin and cyanide has been demonstrated on rat tissues (Rosenthal and Voegtlin 4 ).

It seemed of interest, in this connection, to determine the action of methemoglobin and methylene blue, with cyanide, on isolated invertebrate tissues. Ciliated gill tissue of the quahog (Venus mercenaria) was isolated in sea water and its oxygen consumption followed at 22°C. in Warburg manometers. Alkali in the inset, by absorbing CO₂ and HCN, led to a rise in pH, but always less than 0.5.

The normal Qo₂ (cmm. O₂ per hour per gm. fresh weight) averaged 375 (26 experiments) during the first hour in sea water, about 10% less for the second. Cyanide led to typical inhibition, though rather high concentration was needed. As per cent of the normal, average (4 series of experiments) values in cyanide were : M/11,000, 66; M/5500, 58; M/1100, 56; M/550, 55; M/110, 16. Methylene blue in 0.09% (M/350) concentration, increased the respiration in sea water by one-third. (One-tenth this concentration was without effect, even after cyanide.) Respiration inhibited by M/110 cyanide to 16% normal was more than doubled (to 38%) by the dye, though still depressed. This dye concentration did, however, accurately restore to normal the respiration inhibited (to 57%) by M/1100 cyanide (4 experiments). It is to be noted that considerable restoration occurred even when the added dye was of lower molarity than the cyanide, indicating its separate action as a substitute catalyst in the cell.

Evidence has been presented that stimulation of the hypothalamus causes pupillary dilatation, 1 , 2 increase in blood pressure, 3 , 4 cardiac arrhythemia and ventricular extrasystoles, 4 and contraction of the bladder. 5 The suggestion has been made that the parasympathetic and sympathetic divisions of the autonomic system are separately represented in the hypothalamus. 6 , 7 We have shown that the mechanism yielding pupillary dilatation is not sharply localized at one point but is widespread throughout the hypothalamus and that the most marked reactions are obtained from the lateral part in the region occupied by the medial forebrain bundle. 2

Directing our attention to this problem again, we have explored the hypothalamus millimeter by millimeter in 44 cats using a fara-dic current delivered through a bipolar needle electrode accurately placed with the Horsley-Clarke stereotaxic instrument. The technique was the same as that employed in our previous experiments 2 except that changes in respiration, blood pressure, and bladder tonus were recorded on a kymograph. The results of this new series of experiments confirm our previous observation on the pupil and respiration. Pupillary dilatation and increases in rate and depth of respiration were elicited along with a rise in blood pressure and contraction of the bladder as a result of the stimulation. Better reactions were obtained from the lateral part of the hypothalamus containing the medial forebrain bundle than from the medial part forming the wall of the third ventricle.

Stimulation farther rostrally in the region of the anterior commissure and adjacent part of the septum also yielded contraction of the bladder. It would appear that the cell masses concerned with this reaction are located here, that from these a descending pathway runs caudalward and that bladder reactions elicited from the hypothalamus are due to a stimulation of this descending path. No other parasympathetic responses were elicited by us from the hypothalamus. But often a drop in blood pressure, sometimes associated with a decrease in the heart rate, was obtained from the region of the anterior commissure and adjacent portion of the septum. Stimulation of this same region also caused a decrease in the rate and depth of respiration.

Studies of experimental peritonitis have been almost insuperably handicapped by the fact that the injection of bacteria alone has failed to reproduce the disease. This was noted by Wegner. 1 Confirmation by many others has occurred (Pawlowsky, 2 Steinberg, 3 David.⁴) It is a commonplace today that injection of even overwhelming amounts of bacteria fails to produce a peritonitis. Although the animal may be killed by the overwhelming sepsis, the peritoneum does not exhibit any of the usual signs of peritonitis and may even be sterile by the time the animal has succumbed from the absorption of the organism. Rewbridge 5 found that the lethal dose of Welch bacilli injected into the peritoneal cavity was 18 cc. of 24-hour broth culture, but that bacteria in smaller numbers than this had no effect whatsoever.

The standard methods of production of experimental peritonitis are (1) The injection of bacteria plus some foreign body, as agar culture media, feces, etc. (2) Production of internal fistulas such as opening the gastrointestinal tract. (3) Leaving large amounts of necrotic tissue in an infected abdomen such as is produced by the ligation and subsequent gangrene of an appendix. All of these conditions do not as a general principle exactly reproduce the onset of peritonitis in the human clinically.

The experiments tabulated in Table I are important because they demonstrate the fact that it is possible to produce a peritonitis by bacteria alone. Only small amounts of bacteria were used in most cases and the development of a typical peritonitis occurred in about half the cases.

The only difference between these and previous unsuccessful attempts appeared to be that the dogs used were in very poor condition. They were those half-starved animals which arrive in every experimental laboratory and are usually cared for and properly nourished

Collection of bile from experimental animals has always presented almost insuperable difficulties. Simple intubation of the ducts or the gall bladder has resulted in ascending infection, which has seriously interfered with the results obtained, as the bile then originated in livers more or less damaged by hepatitis. In such preparations leakage and consequent loss of part of the bile nearly always occurs if the experiment is long continued. The Rous 1 type of fistula, in which a long tube is coiled within the abdomen thus putting considerable distance between the bile duct and the skin, has been used extensively in this laboratory with indifferent results. For a few weeks it was often successful, but a very high percentage of cases failed in later months which are the crucial point in these studies.

Dragstedt 2 recently reported a new type of fistula which he had used extensively for intestinal studies and for short periods in the gall bladder. It consisted in a gold-plated metal cannula, with a metal disc at the inner end. This was sutured into the intestine. The distal end was then capped with a special small spear point which was pushed through the abdominal wall at a little distance from the laparotomy wound. The intervening space (2-4 inches) was then wrapped in omentum. A screw cap at the distal end enabled one to attach a container.

This principle applied to the gall bladder has functioned with remarkable perfection. The cannula must be considerably longer than that used for intestinal work, as the liver moves on respiration. The total length of our cannulas was 5½ inches. The disc at the inner end was made smaller (¾ inch) and somewhat cup-shaped. After the common duct has been ligated, the cannula is simply implanted into the gallbladder through as small a hole as possible and tied in place. Inversion is not necessary, and not advisable.

The question as to whether the well known anemia accompanying acholic cachexia is a primary factor in the disease or only a manifestation of a late or terminal cachexia in general is an important one. On this feature alone are based many of the current theories as to the origin of cachexia cholipriva. One assumption often made is that prolonged loss of bile will deprive the body of the opportunity of resorption of blood-forming pigments from the gastrointestinal tract. The statement may easily be disproved by a comparative study of the rate of fall of the hemoglobin and the cells in acholic anemia. Such an assumption would surely predicate a more rapid fall of the hemoglobin than in the red count. In previously reported series as well as in our tables it can be seen at once that this is not the case.

The other theory is based on the fact that the anemia is a manifestation of lack of fat soluble vitamin, due to inability to absorb such material in the absence of bile salts from the intestinal tract. The evidence for this is admirably summed up by Takasu. 1

Studies of this character have frequently been complicated by the fact that such a surprisingly high percentage of bile fistula dogs develop ulcers of the stomach or duodenum, which are liable to massive hemorrhage. Estimates of this percentage have been given as high as 90% by some observers, although in our own experience it has been very much less. In the experiments here reported this factor was ruled out by post mortem examinations as well as daily search for blood in the stools. The figures reported are from 4 dogs with biliary fistulas who lived from 4 1/2 to 5 1/2 months.

The average loss in body weight during this period was 33%. The average loss of hemoglobin was 17.4% and of red cells, 17.8%. Four of the dogs died with hemoglobin and red count within the lower limits of normal and the lowest red count was 4,200,000. These findings are certainly not indicative of the fact that anemia is at all a primary factor in the condition. Furthermore, one dog died with a red count higher than before the operation, after having lost 41% of body weight in 5 months, his appearance being in every way typical of acholic cachexia.

Red cell fragility was tested every 4 days in all these animals and in no case was there any deviation from the normal.

From the accompanying chart it is quite clear that the weight loss ran a curve definitely ahead of the anemia, and that while the cachexia and weight loss were profound, the blood findings not only were of a far lesser degree but also did not in the least parallel the progress of the cachexia.

For many years it has been assumed that since the bile was alkaline, loss of bile would consequently produce an acidosis. The importance of this factor in acholic cachexia has been a disputed point.

Okada 1 studied fistula bile from dogs and found a pH averaging slightly over 8.00, Düttmann's 2 investigations showed it a trifle lower. In the prolonged studies of Drury, McMaster and Rous 3 duct bile averaged pH 8.20. Neilson and Meyer 4 found figures in the same range, but called attention to the fact that on standing, or heating, duct bile would increase its alkalinity, the pH rising to 8.40-8.60. This was presumably due to driving off of CO₂. Drury 3 emphasized the point that the bile is a well-buffered solution, and that addition of considerable amounts of acid were necessary in order to change the reaction. In this connection it is well to remember that considerable fixed acid is lost with the bile, as Wangen-steen 5 found that a human with a biliary fistula secreted about 0.4 gm. NaCl daily.

In many previous studies too numerous to mention here, the presence of a mild degree of acidosis has been quite usual, while a few investigators have found it to be of such a high grade as to regard it as a primary factor in the accompanying cachexia. The material here reported consists of CO₂ and blood chloride estimations on 8 animals, part with biliary fistulas and part with cholecystnephros-tomies. From the accompanying table it may be seen that there are no significant changes in the acid-base equilibrium. The average CO₂ before operation was 39.5, exactly that taken at the conclusion of the experiments (39.5). The average preoperative blood chloride was 290 mg. per 100 cc. and at the conclusion of the experiments was 293 mg. per 100 cc.

Since gastrectomized dogs do not develop pernicious anemia, 1 it is pertinent to determine whether substances effective in pernicious anemia are present in the liver and stomach of dogs.

We have prepared dog's “desiccated whole stomach” according to the standard method and have given the preparation orally to patients with pernicious anemia. We have made extracts of the dog's liver and have injected them subcutaneously. We have also fed dog's liver to pernicious anemia patients. These materials were prepared from recently sacrificed dogs.

Dog's desiccated whole stomach. In a previous report, 2 the desiccated whole stomach of dogs was fed to 2 pernicious anemia patients in doses of 50 gm. (dry powder) per day for from 8 to 12 days without observing a response. Later both of these patients responded typically, one to oral equine liver and the other to liver extract subcutaneously. We have extended these experiments to include 5 additional pernicious anemia patients, making a total of 7 patients. A slight increase in both red cell count and reticulocytes occurred in one patient. All patients, except one, when placed on equine liver extract responded typically.

Isaacs and Sturgis 3 obtained remissions from feeding from 30 to 60 gm. of desiccated hog's whole stomach. Our results indicate that if the active principle is present in dog's whole stomach, it is present only in smaller amounts than found in the hog's stomach. Larger doses of dog's desiccated stomach were not used because of the difficulty in getting patients to take larger doses voluntarily over a period of adequate length to be significant.

Dog's liver. Six pernicious anemia patients received extracts of dog's liver prepared according to the standard method. The extract of from 45 to 155 gm. of liver was administered subcutaneously during a 10-day period. Three of the patients, one receiving the extract of 45 gm., one 90 gm., the other 140 gm. of liver, showed some evidence of a response, which, however, is questionable. (Table I.)

Rony, Mortimer and Ivy 1 reported that the thoracic duct lymph of dogs that had fasted 2 to 14 days contains 250 to 1300 mg. of total fatty acids per 100 cc. lymph with an average value for 60 dogs of 630 mg. %. The total cholesterol content of these lymphs varied from 41 to 114 mg. %.

This work was undertaken to obtain further characteristics of the lipids of thoracic duct lymph in fasting. The lipid phosphorus content, the iodine number and the saponification value of the lymph lipids were examined.

The dogs fasted 4 to 7 days; the thoracic duct was cannulated under intravenous pentobarbital anesthesia and from 20 to 50 cc. of lymph were collected. This lymph was analyzed for lipid phosphorus with the Kuttner-Lichtenstein method, 2 for iodine number with the Hanus method, and for saponification value. Table I gives the results thus obtained on 12 dogs.

The values given in Table I for total fatty acids and total cholesterol are quoted from Rony, Mortimer and Ivy's paper. The values for “lecithin” are obtained by multiplying the lipid-P by 25. The saponification value represents the mg. KOH required to neutralize 1 gm. of the solids in the ether extract of lymph. The calculation of the unsaponifiable portion of the lymph fat was made on the basis of the assumption that all lipids present are C18 fatty acids.

It is known that, in fasting, lipids are excreted by the intestinal mucosa into the lumen of the bowel (Bloor and Sperry, etc.). Rony, Mortimer and Ivy found that in fasting the thoracic duct lymph fat consists chiefly of this fat, the major portion of which is reabsorbed by the intestinal mucosa and passed into the lacteals and thoracic duct.

Blood studies in gastrectomized animals have been done in the dog 1 and in the rat. 2 The presence of substances effective in pernicious anemia in hog's liver and whole desiccated stomach made it seem worth while to study the effect of gastrectomy in the hog on hematopoiesis. We desire to report our preliminary observations during the past 7 months.

Five Poland China pigs 12 1/2 weeks of age were selected. On the 24th of June, 1933, four of the pigs were gastrectomized by us in the Physiological Laboratory of the University of Wisconsin Medical School at Madison,∗ one being reserved as control. The distal end of the esophagus was sutured to the side of the first loop of jejunum after closing the proximal end of the duodenum. The operation is more easily performed in the hog than in the dog.

One animal died on the day following operation due to peritonitis consequent to spillage of material from a stomach distended with food, the stomach of this hog not having emptied after 12 hours of starvation. The remaining animals were transferred from Madison to our laboratories on their 48th postoperative day. They have been kept in open air cages since their second postoperative week. A second animal died of stricture of the lower esophagus 71 days after operation. This stricture occurred at the site at which the clamp had been applied. The remaining 2 pigs and the control are living and well on their 210th postoperative day.

The results of gastrectomy and of therapy with ferric ammonium citrate on the microcytic hypochromic anemia, which developed to a point requiring treatment in their fourth postoperative month, are summarized in the accompanying table. The diet has consisted of a mixed meal (18% P; 3.5%fat; 45% C; and 7% fiber) made into a thick mush and brought to the boiling point in a steam cooker.

In an earlier paper, 1 there was described the separation, from meningococcus autolysates, of fractions exhibiting marked type-specificity toward precipitin-containing sera. It is now possible to report the preparation of the Type I specific substance (SSS) in a much higher state of purification.

Twenty liter lots of 3 week cultures of recently isolated strains of Type I meningococcus in hormone broth are acidified with acetic acid to pH 4.0. The SSS has a marked affinity for the acid-precipitable protein, the so-called “P” fraction, and about one-half of the amount present is carried down with the latter under the above conditions. If the broth is diluted 10-fold with tap water before acidification, almost all the SSS is adsorbed, but there is a loss in selectivity. This point is undergoing further investigation.

The “P” fraction is washed with distilled water, suspended in 1 liter of the same solvent and dissolved at pH 7.0 with the aid of NaOH. The bulk of the protein is then precipitated with acetic acid at pH 4.8 and centrifuged off. The precipitate is washed with distilled water and again submitted to the process of solution at pH 7.0 and precipitation at pH 4.8. The combined supernatant fluids, which contain almost all of the SSS, are concentrated in vacuo to the smallest practicable volume and precipitated with 5 volumes of alcohol. After standing overnight, the precipitate is centrifuged off and dried in vacuo with concentrated H₂SO₄. It is then dissolved in 50 to 100 cc. of distilled water with the aid of NaOH, and freed of the remainder of the protein by 0.7 saturation with (NH₄)₂SO₄. The latter is removed by dialysis, 10% of sodium acetate is added and the SSS is precipitated as a yellowish gum by 4 volumes of ethyl alcohol. Further purification is achieved by repeated precipitation by alcohol in the presence of sodium acetate and through precipitation with basic lead acetate or uranyl nitrate. Finally, the SSS is precipitated as a white powder by 10 volumes of absolute alcohol, from aqueous solution containing a small amount of sodium acetate. In favorable cases the yield from 20 liters of broth is 350 mg.

A study was made of the antigenic changes that result when milks are boiled, evaporated, superheated, dried or acidified. This was done to determine why individuals sensitive to raw or pasteurized milk can generally tolerate milks modified by heat.

The milk proteins were purified and their immunochemical reactions studied by the anaphylaxis method in the guinea pig. It was found that the proteins isolated from raw milk are chemically and biologically distinct.

The immunochemistry of evaporated, superheated, freshly boiled, acidified and dried milks and their immunological relation to the purified proteins from raw milk were studied. It was found that there were no antigenic changes as a result of drying, acidification or pasteurization. Milks that were dried, acidified, superheated or evaporated showed no loss in the antigenic properties of the casein fraction. As a sensitizing agent, given by injection, evaporated, freshly boiled and superheated milks showed practically no loss in the antigenic character of the lactalbumin fraction but as a shock agent these milks showed an unmistakable loss in the antigenic properties of the whey proteins.

Raw and the various modifications of milk described above were fed to a large number of guinea pigs. Evaporated, freshly boiled and acidified evaporated milk, when fed, showed a marked reduction in sensitizing ability. The acidified raw, dried and superheated milks did not differ from the raw milk.

A human case markedly sensitive to lactalbumin was studied immunologically. This patient developed anaphylactic shock with recovery when given 0.02 cc. of a 1% solution of pure lactalbumin intracutaneously. The blood serum containing lactalbumin antibodies passively sensitized normal guinea pigs and the skin of normal human recipients. The child was shown to tolerate evaporated milk in large amounts although raw milk, taken by mouth, in very small amounts produced anaphylactic reactions.

Ever since Cooke's polynuclear count became generally used in physiology and clinical medicine, the figures for the normal count in man, as given originally by Cooke and as republished by Cooke and Ponder 1 have been accepted as correct. These figures are as follows:

Cooke and Ponder's normal count is based on the average of the counts for 90 people of both sexes between the ages of 12 and 55, the greatest possible care being taken to exclude foci of infection. That their standard was exceedingly rigorous is indicated by the fact that these 90 “normals” took many years to collect.

But it has been suspected that the criterion of normality used by Cooke and Ponder was too rigorous, and that the counts of “normal” people, in the sense in which the term is generally used by physiologists and clinicians, are more left-handed than the above figures indicate. Kennedy 2 has reinvestigated the matter, using a less rigorous “health standard” more likely to be met by people “normal” in the ordinary sense. His conclusion, however, is that the figures given by Cooke and Ponder are substantially correct, and that an average polynuclear count with a mean of from 2.62 to 2.72 is met in a population of normal persons, whether normality is judged by Cooke and Pender's very rigorous criterion or by Kennedy's looser one.

The technical problem of finding the rate at which pigment leaves the mammalian red cell was proposed to me by Dr. Hugo Fricke, with the idea that if the rate were known it might be possible to calculate whether the pigment leaves the cell through one or two holes in the membrane, or by diffusing across the membrane generally.

The method which I have used consists in taking motion pictures of the haemolysing cells.∗ The best cell to work with is the red cell of man, which is relatively large. A suspension is made by suspending the thrice-washed cells from 4 cc. of blood in 20 cc. of 1% NaCl, for a suspension of this strength, when diluted with equal volumes of a lysin solution, shows from 5 to 15 cells per frame of the film used (35 mm.). Saponin made up in 1.0% NaCl in such a dilution as produces complete lysis in about 3 min. is a convenient lysin. The optical system used should be one of the highest resolution practicable, e. g., an aplanatic condenser of N.A. 1.4 working at 2/3 cone and critically focussed, a 1/10 inch oil-immersion objective of N.A. 1.37, and a 10x eyepiece. If the illuminant is an arc used without color filters, the exposure is about 1/30 second. A film speed of 16 frames per sec. is quite sufficient.

To 0.5 cc. of the suspension is added 0.5 cc. of the dilution of lysin sufficient to give complete lysis in about 3 min. A drop of the mixture is rapidly placed on a slide and covered with a coverglass smeared with vaseline along the edges. The vaseline seal prevents the cells from drifting. The preparation is placed on the stage of the microscope, and the cells focussed through a side telescope which enables them to be seen while the film is in a position ready for exposure.

In the course of experiments designed to test for the presence of an erythrocytogenic hormone in the serum of rabbits subjected to bleedings, 1 a close correspondence between the number of red cells and reticulocytes of several control animals during the recovery period was observed. These control rabbits were bled a certain percentage of their body weight by cardiac puncture, and red cell and reticulocyte counts made daily for approximately 2 weeks following the bleeding. Because considerable discrepancy was found between counts made on successive samples taken from the ear veins, the red cells and reticulocytes were counted in samples drawn from the heart. Care must be taken that the heart punctures are made cleanly and that the needle is held steadily while the blood is being withdrawn. When these precautions are observed, autopsies performed on animals subjected to daily heart punctures for as long as a period of a month showed no damage to the heart or hemorrhage around it. The red cell counts which were required to agree to within 2.5%, were made in duplicate, a total of 800 squares on a Levy-Hausser chamber being counted for each sample of blood. Reticulocytes were counted in wet mounts according to the method described by Ramsey and Warren.²

In Table I are presented the red cell counts (X 10^-3), and reticulocyte counts (%) for 5 animals bled 1/64 of their body weight on the first day. The reticulocyte peak, ranging from 7.2% to 8.1%, is reached in all the animals on the 4th or 5th day of anemia. There is a fairly close similarity in the behavior of the red cell counts. Some of the animals, however, showed a greater post-hemorrhagic drop (i. e., that occurring after the first day of anemia) than others.

The continued administration of the anterior pituitary-like hormone of pregnancy urine (A.P.L.) to virgin rats leads at first to a great increase in the size of the ovaries; but after some weeks the animals appear to become insensitive and in spite of continued treatment the weight of the ovaries is found to decrease to normal or even sub-normal levels. We have previously shown 1 that such “insensitive” rats will still react to gonad-stimulating extracts prepared from the hypophysis, showing renewed increase in the weight of the ovaries with definite signs of luteinization.

Eight female rats, initially 21 days old, received implants of 1-1 ½ rat hypophyses daily for 68 days. At the end of this period the ovaries were found, on laparotomy, to be normal or sub-normal in size. Single ovaries removed from 2 animals weighed 24 and 9 mg. respectively; one animal was sacrificed, and its ovaries were found to weigh 8.5 mg. each. The 7 survivors were then treated with 250 units of A.P.L. daily for 13 days, the implantation of hypophyses being continued. The animals were then killed, and the ovaries were found to be greatly enlarged, each ovary weighing from 119 to 198 mg. We conclude that rats treated with pituitary implants eventually lose their sensitivity to the hypophyseal gonadotropic hormone, but remain sensitive to the anterior pituitary-like hormone of pregnancy urine. The experiment may also be considered to provide further evidence that these 2 gonad-stimulating substances are not identical.

Since the discovery by Hansen of the acid-fast bacillus in the lesion of human leprosy, many investigators who have attempted its cultivation have considered their “culture” the authentic one. Recently Soule and McKinley 1 in their attempts at cultivation of B. leprae prove no exception to this rule. In an earlier publication², they claim complete success with the CO₂, O₂ method of Wherry 3 regardless of the nutritives employed. McKinley and Verder 4 later claim that they have cultivated B. leprae in 5 days in chick embryo tissue suspended in Tyrode's solution under ordinary atmospheric conditions as well as CO₂ and O₂.

At the suggestion of Dr. Duval I have undertaken to repeat the work of McKinley and Verder. This paper reports results obtained upon that part of the work in which the chick embryo tissue is concerned.

The leprous tissue employed was the subcutaneous nodules and material from the socalled leprous abscess of human lepers, kindly supplied us through the courtesy of Dr. Denney of Carville, La. All the material was extremely rich in acid-fast bacilli as was evidenced upon microscopic examination of stained tissue sections and smear preparations. The Hansen microorganism of leprosy appeared chiefly in densely packed masses (globi) and small packets.

The leprous tissue was first cut into small pieces aseptically, then placed in a mortar and emulsified. A part of the emulsion was treated with a 3% NaOH solution for one hour, and immediately afterward was centrifugally washed several times in sterile Ringer's fluid. Leprous material which was not contaminated was used without treating it with NaOH. In addition, the purported culture of B. leprae in minced chick and Tyrode's solution which was sent to Dr. Duval by Dr. McKinley, was used in parallel experiments. Both liquid and solid media (plain neutral agar) containing the emulsified chick tissue were employed in the experiments.

It has been shown by Fevold, et al. 1 , 2 that the gonad stimulating-principles of the anterior lobe of the pituitary gland can be extracted almost quantitatively from powdered glands by means of 50% aqueous pyridine. On account of the difficulty in evaporation of the pyridine extract and the molesting odor, other methods of extracting have been used in the following experiments.

It was found that a solution of 6% normal butyl alcohol, or a 3% solution of amyl alcohol in water would also extract the active principles in a very efficient manner. The extract can very easily be concentrated by vacuum distillation at low temperature (below 40°C) so it offers advantages over aqueous pyridine in this respect.

A comparison of the efficiency of the 2 solvents, 50% aqueous pyridine and 6% aqueous butanol, is given below. The procedure used was that given by Fevold, et al., and was the same in both cases. One hundred grams of the dried acetone extracted pituitary gland from horses were used. The material was extracted each time for 12 hours with 700, 600, 600, 600 cc. of the respective solvent—50% pyridine or 6% butyl alcohol solution.

Each extract was then purified as follows: After evaporation in vacuo at a temperature below 40° to a volume of 500 cc, 4 1/2 volumes of acetone were added. The precipitate thus obtained was emulsified in 1,500 cc. of water and 750 cc. of acetone were added. The insoluble material was re-extracted with 33% acetone for a second time. The combined 33% acetone solution was evaporated in vacuo. The residue was emulsified in 500 cc. water, and 70 cc. of saturated alcoholic benzoic acid were added. The material brought down with the precipitate of benzoic acid was washed with acetone to remove the benzoic acid, then extracted twice with 150 cc. of water; the aqueous extract was mixed with the filtrate from the benzoic acid and precipitated with 4 1/2 volumes of acetone.

Sincer substances of great biological significance are excreted in the urine during pregnancy, it seems possible that other substances of different biological action might be excreted during active labor. Guerin-Valmale, Loriot, and Verdeuil 1 claimed that labor urine produced contractions of the human uterus in vivo. They were able to induce labor in 2 patients 10 days before term by administering labor urine by proctolysis; the patients were delivered about 12 hours after the instillation. Barjaktarovic 2 used the same method in 10 cases during different months of pregnancy and concluded that to induce labor in the second half of pregnancy amounts up to 4,000 gm. of labor urine were necessary. We have, therefore, attempted to extract oxytocic substances from labor urine.

Urine was collected from women during active labor and combined until a total volume of 10 or 12 liters was obtained. (Pituitrin was not given to any case.) Thirteen such collections were made, and each was worked up separately. The samples were acidified with acetic acid and concentrated in vacuo at a temperature of 30-40°C. The method then used to obtain the oxytocic substance was that described by Kamm and his co-workers 3 for isolation of the active principles of the posterior lobe of the pituitary gland.

The final product was a crude material of a gummy character and colored with urinary pigments, 10 liters of labor urine yielding an average of 0.5 gm. This dried material was taken up in distilled water and dialyzed over night in a refrigerator in a parchment bag. The solution from the bag was then made neutral to litmus with sodium carbonate. The solution was tested, using about 0.4 cc. from a total volume of 25 cc. representing the original 0.5 gm., on human uterine muscle obtained from Caesarian sections at term.

In the course of a correlated series of laboratory and clinical investigations, which will later be reported upon in full detail, with carbohydrate obtained from ragweed pollen, certain results were obtained with the Prausnitz Küstner reaction (local passive transfer) which seem worth recording at this time.

In skin testing ragweed sensitive hay-fever patients to this carbohydrate material it was found that a percentage gave positive skin reactions to the carbohydrate about as large as was obtained with the whole pollen or any protein fraction thereof, and also that a percentage gave no reaction on skin tests to the carbohydrate.

Serum for the Prausnitz Küstner reaction was taken (A) from those who showed the most marked reaction (relative to the results obtained with the pollen or other protein fractions thereof), (B) from those who showed no reaction to the carbohydrate although the results with the other ragweed preparations might be equal or thereabouts to those obtained in Group A.

With the sera of Group A it was found that the sensitized areas in the recipient would give a local reaction to the carbohydrate solution whereas with the sera of Group B no reaction was obtained.

The possibility of some peculiarity or exceeding susceptibility of the recipient was ruled out by obtaining the contrast of serum from Group A on one arm of the recipient (which gave a positive Prausnitz Küstner to the carbohydrate) and the serum from Group B (which was negative) on the other arm. This at the same time that the reactions obtained with both sera to whole pollen solution would be equal.

Because of the recent interest aroused by Frei 1 in demonstrating a diagnostic skin reaction to the whole extracts of glands in the virus disease lymphogranulomatosis (4th venereal disease) it was decided to observe the skin reactions on patients with Hodgkin's disease to the extracts of Hodgkin's diseased glands. Although the etiology of this disease remains obscure there are some 2 who believe it may be due to a virus. Skin reactions can be obtained with the extract of skin lesions in small pox and vaccinia, a virus disease.

The injection of filtrates of Hodgkin's glands has been used in treatment of the disease but the skin reactions to these filtrates have not been described adequately. Skin reactions to a preparation from Hodgkin's diseased glands were reported as negative in 3 cases by Fox and Farley. 3

Here we are further recording the negative results obtained in 11 patients with Hodgkin's disease treated by Roentgen ray in whom the diagnosis was assured by biopsy. All but one patient showed a skin reaction of only a few millimeters, similar to that seen in normal persons. In this one the control, saline and 0.3% tricresol measured 0.5 cm. and the unheated extract measured 1.0 by 1.5 cm. at 24 hours, but after 48 hours measured 1-2 mm.

Extracts were made from the lymph nodes of an untreated case of Hodgkin's in the third month of the disease and also from a case of lymphosarcoma. The clinical course of these cases has verified the histological diagnosis. The fresh nodes were extracted with saline, centrifuged and 0.3% tricresol added to the cloudy, supernatant fluid. One-half of this was heated for an hour at 60°C. on 2 successive days. The final extract was not filtered.

Three beliefs exist in regard to the etiology of whooping cough: That the disease is caused by H. pertussis, that the agent is a filtrable virus, or that both play a part. H. pertussis is practically always present in the disease and absent in health. 1 That the organism plays a most important rôle has been shown by occasional successful transmission experiments 2 , 3 , 4 and by reports of effective protection with suitable H. pertussis vaccine. 1 , 4 , 5 The recent description of intranuclear inclusion bodies 6 taken together with certain other considerations has led some^3,7 to believe that a virus plays either a specific etiological rôle or, accompanied by H. pertussis, acts as primary infective agent. Attempted transmission experiments with filtrates have been equivocal or negative. 3 , 4 Because of too recent isolation of the organism or of inoculation with filtrate shortly before 4 the few successful transmission experiments with H. pertussis are open to the criticism that a virus may have been present. It seemed necessary to devise experiments in which H. pertussis or virus effects were clearly separated. In this communication we are reporting successful whooping cought transmission by means of what we believe to be virus-free H. pertussis.

The newer knowledge of the phases of H. pertussis 8 , 9 and their precise differentiation by electrophoresis 10 provided the desired approach. When first isolated H. pertussis differs culturally, morphologically, serologically, and electrophoretically from older laboratory cultures. Leslie and Gardner call the former, Phase I, and the variations rapidly appearing on subculture, Phases II, III, IV. It is possible 9 , 10 that this change represents the well-known virulent (“S”) to avirulent (“R”) transition of other bacteria.

It has been shown by Hering, 1 Heymans, 2 and others that stimulation of the afferent endings in the carotid sinus by an increase in the endosinusal pressure causes a slowing of the heart. Some have held that this is entirely due to a reflex stimulation of vagal fibers while others have presented evidence to show that there is a reflex inhibition of the cardiac accelerators as well. The second explanation involving a reciprocal mechanism has generally appeared to be the more probable. In the present investigation we have studied certain aspects of this reflex by observing the sympathetic impulses to the heart while varying the pressure in the carotid sinus.

The experiments were performed on cats under nembutal anesthesia. The carotid sinus on each side was isolated and perfused by way of the common carotid and external carotid arteries. Inasmuch as the nerve supply remained intact an increase of pressure within the sinus caused an increase in the number of afferent impulses going to the medullary centers. 3 The effect of these impulses on the cardiac accelerator discharge was determined by recording with a vacuum tube amplifier and oscillograph the sympathetic impulses in one of the small nerves running from the stellate ganglion to the heart.

The general nature of the cardiac sympathetic activity has been described 4 as a persistent tonic discharge of impulses grouped into large waves. We now find that as the pressure within the sinus is raised this discharge decreases until at about 125 to 150 mm. Hg. there is a complete inhibition of the sympathetic impulses.

If the pressure within the sinus be maintained at such a level the period of complete inhibition lasts for some seconds, after which there is an escape with a return of the sympathetic discharge. The duration of this period of inhibition is a function of the endosinusal pressure.

Carotene in chloroform solution reacts with formaldehyde in sulphuric acid to form a characteristic deep violet zone between the chloroform and acid layer. The violet zone on shaking disseminates in the acid giving rise to a violet colored lower layer, while the chloroform layer above remains colorless. The addition of 2 to 3 drops of acetic anhydride causes a transitory blue to appear in the acid layer, provided a high concentration of carotene is present in the chloroform solution.

The test, which is considered positive on the appearance of a characteristic deep violet zone, is very sensitive, serving to detect 0.01 mg. of carotene in 2.5 cc. of chloroform. To carry out the test equal volumes of chloroform and formaldehyde-sulphuric acid mixture should be employed. For convenience we have used 2.5 cc. of the chloroform solution with 2.5 cc. of the aldehyde-acid mixture. This mixture is made up in the proportion of 1 volume of 37 to 40% formaldehyde to 50 volumes of concentrated sulphuric acid. Both the formaldehyde and the sulphuric acid should be chemically pure, and the mixture freshly made before using.

Cholesterol, according to Whitby, 1 gives a characteristic reaction with the formaldehyde-sulphuric acid mixture described above. With 1 to 2 mg. of the sterol dissolved in 2 cc. of chloroform he reported a cherry red color in the upper chloroform layer and a brownish red color in the lower sulphuric acid layer, which also showed an intense green fluorescence. When 2 to 3 drops of acetic anhydride were added to the chloroform layer previously transferred to a dry test tube, a blue color formed. With 0.1 mg. of cholesterol in 2 cc. of chloroform he obtained a deep reddish brown color and a very marked fluorescence in the acid layer, but no color in the chloroform layer. With 0.01 mg. of the sterol he could develop only a faint fluorescence in the acid layer and no coloration in the chloroform laver.

Sterols, 1 terpenes, 2 carotene, 3 vitamin A 4 and 5-membered monoheterocyclic ring compounds 5 have the characteristic in common that they react with the Carr-Price reagent, antimony trichloride in chloroform solution, and with the Shear reagent, aniline-hydrochloric acid. 6 Reasoning on the probability of the existence of some common factor in the chemistry of sterols, terpenes, carotene and vitamin A, we determined to find out whether the Liebermann-Bur-chard reagent, used qualitatively in the detection of cholesterol and quantitatively in its estimation, was also given by the other compounds in the group we have mentioned.

We accordingly studied the Liebermann-Burchard test with respect to certain 5-membered monoheterocyclic compounds. Unless otherwise stated, we adopted the procedure of dissolving 1 to 2 drops of a given liquid compound to be tested or 10 to 20 mg. of a solid compound in 2 cc. of chloroform, adding 10 drops of acetic anhydride and 1 to 3 drops of concentrated sulphuric acid. The color reaction undergoes many variations depending upon the concentration in chloroform of the compound tested and the quantity of concentrated sulphuric acid added.

Pyrrol at first gave a yellow turbidity, which gradually turned to very deep orange. On standing a deep blood-orange precipitate separated out, leaving a colorless liquid above. On the addition of sulphuric acid alone slightly heavy droplets separate at the bottom. On standing these droplets solidified and developed an orange tinge, which gradually darkened. Thiophene on the addition of 5 drops to 2 cc. of chloroform gave yellow oily droplets at the bottom of the tube. On standing over night the droplets disappeared and the liquid contents of the tube turned green. The reaction was very slow. The green color, however, once formed became deeper and lasted more than 2 weeks. Black granules finally precipitated out in the colored mixture. With sulphuric acid alone yellow droplets formed at the bottom. These gradually turned deep reddish brown, and at the end of a week the colorless liquid assumed a very slight green tinge.

The advisability of adding iodine to the diet of man and animals has long been questioned. The herd of Brown Swiss dairy cows at the farms of Mr. William Wallace Kincaid,∗ Ellerslie-on-Niagara, New York, had been fed daily large amounts of iodine during a period of 2 years and 11 months. They were, consequently, satisfactory experimental animals for investigating the effects of prolonged increased iodine feeding. The blood iodine was of particular interest. The normal bovine blood iodine is 13, 1 or 14 2 gamma %. In 19 cows of this herd the blood iodine ranged from 35.9 to 78.1 gamma %, averaging 55.

Analyses of the feed given these cows daily, revealed the increased iodine intake (Table I). Since the high milk producing cows were fed more of the “grain mix” they ingested greater amounts of iodine (Table I, a). The natural iodine content of the “grain mix” had been increased by the addition of potassium iodide, fish meal, and kelp. The iodine content of this thoroughly mixed food was 2.006 mg. per 100 gm. The daily iodine intake of the cows varied from 68.6 to 196.1 mg.

Large amounts of iodine, from 51.7 to 163.4 mg. daily, were eliminated, Table II. More was eliminated in the urine than in either the milk or the feces. The high milk producing cows likewise showed the greater elimination of iodine, (Table II, a).

The milk iodine was increased. It ranged from 15.7 to 181.9 gamma per 100 cc, averaging 66.4. The average iodine content of milk from cows on usual feed is 5 gamma per 100 cc. 2 , 3

All physical examinations showed the cows to be in very good general condition. Cows of the herd previously sterile have become pregnant and borne normal calves during this regimen of feeding.

In the present work observations were made of the growth of certain pathogenic fungi on hair. Bonar and Dreyer 1 used bundles of hair inoculated with trichophyton interdigitale to test antiseptic properties of solutions. Tona 2 observed the growth of trichophyton on hair in the test tube. The growth of the various pathogenic fungi (dermatophytes) on a medium of scalp products and hair seems to have received little attention.

The medium used consisted of barber shop hair cleansed with water, placed in petri dishes and test tubes and autoclaved. It was found necessary to keep the medium moist with distilled water by adding an amount sufficient to cover the mat of hair. Growth on this medium was compared with that on conservation Sabouraud's. Sabouraud's conservation medium was chosen as a standard because of its simplicity. Bacto products were used in making the media.

Transplants were made from 30-day-old growths of the organisms on Sabouraud's proof agar incubated at room temperature.

Approximate time of growth of the following pathogenic fungi and the 2 saprophytes, Lichtheimia sp. and Scopulariopsis brevicaulis were noted: Achorion schoenleinii, Acladium castellcmi, Candida Candida, Endodermophyton tropicale, Endomyces capsultus, Endomyces dermatitidus, Epidermophyton inguinale, Glenospora gammeli, Geotrichum baehmann, Indieila americana, Monilia albicans, Microsporon apiospermum, Microsporon audouini, Oöspora humi, Trichophyton crateriforme, Trichophyton granulosus, Trichophyton japonicum, Trichophyton interdigitate, Willia anomala. Observations were limited to 30 days. Nine organisms which grew well on Sabouraud's medium showed no growth on water hair; Achorion schoenleini, Acladium castellani, Candida Candida, Endomyces dermatitidus, Epidermophyton inguinale, Glenospora gammeli, Microsporon audouini, Oöspora humi, and Willia anomala. Since the pathogenicity of the Microsporon audouini manifests itself before, puberty its growth was observed on a medium of child's hair.

The blood of infants shows significant alterations in the first few months of life. Most pronounced, perhaps, is the drop in the number of red blood cells and in hemoglobin. Another hitherto unrecognized feature apparently characteristic of this period is the rapidity with which the blood decolorizes methylene-blue.

The method employed in this investigation was as follows: A stock solution was prepared by dissolving 200 mg. of methylene blue∗ in 40 cc. of 3% sodium citrate∗ in doubly distilled water at pH 7.4 to 8.1. After filtration 0.07 cc. of this solution was dissolved in 50 cc. of sodium citrate of the same strength, giving a final concentration of methylene blue of approximately 1:140,000. This solution was left at least 24 hours in the ice-box before use. For each test 4 samples of freshly drawn blood was added to 0.1 cc. of the methylene blue solution. The quantities of blood with youngest subjects were .02, .03, .04, and .05 cc. respectively, and a range of .03 to .06 cc. in the older ones. The contents were thoroughly admixed and kept at 15°C. for 1 hour. The tubes were then centrifuged for 2 minutes. The extent of reduction was 0 when the dilute blue color was unchanged, to 4 plus when the dye was completely reduced to that of water.

This study included the blood of 238 healthy individuals, 152 under one year of age, the remainder ranged to 40 years. The results are summarized in Table I.

The reduction of methylene blue proceeded most rapidly and completely in the large majority of observations during the first 4 months of life. After this period, the reducing power was markedly diminished and resembled the blood of the older child and adult.

Tests were also carried out periodically with the blood of 3 young rabbits. The results corresponded with the human infant except that the reducing power of the animals disappeared between 2 and 3 months of age.

Arkwright 1 originally described 2 variant forms of members of the colon-typhoid-dysentery group which he designated by the terms S (smooth) and R (rough). Recently evidence has accumulated 2 to show that members of the colon-typhoid-dysentery group, as well as many other bacterial species, possess a third and distinct variant form, the M (mucoid). Organisms in the mucoid phase possess distinct capsules and form large mucoid colonies.

Griffith 3 described 2 variant forms of pneumococcus and, adopting Arkwright's terminology, designated them as S (smooth) and R (rough). A new variant form of pneumococcus was recently described by the author. 4 This new variant form was shown to be strikingly different from the S and R forms described by Griffith.

A comparison of the characteristic features of the 3 chief variant forms of members of the colon-typhoid-dysentery, etc., group on the one hand, and of pneumococcus on the other hand, reveals a striking parallelism. Thus, (1) the M (mucoid) form of the colon-typhoid-dysentery, etc., group corresponds with the S (smooth) form of pneumococcus; (2) the S (smooth) form of the colon-typhoid-dysentery, etc., group corresponds with the R (rough) form of pneumococcus; and (3) the R (rough) form of the colon-typhoid-dysentery group corresponds with the newly described variant form of pneumococcus.

The striking inconsistency in terminology employed to describe corresponding variant forms of different bacterial species is obviously confusing and should be rectified if possible. Before proposing a radical change in the currently accepted terminology of pneumococcal variants a study of bacterial variation in Streptococcus hemolyticus was undertaken in order to determine whether similar variant forms could be demonstrated in this species.

Meyer, Leonard, Hisaw and Martin 1 demonstrated that oestrin diminishes the gonad stimulating capacity of the anterior pituitary. Inasmuch as the injection of P. U. extract into normal females increases their oestrin output, the potency of their pituitaries should thereby be diminished. Leonard 2 injected a large series of adult male and female rats with P. U. extracts† and subsequently tested their pituitaries for gonad stimulating potency by implantation into immature female mice. Pituitaries of the injected animals showed a marked decrease in gonad stimulating potency. The potency of the castrate's pituitary was not diminished by P. U. injections.

The pituitaries of 32 normal and castrate controls and injected rats, from the experiments of Dr. Leonard, were utilized for cytological study. They were fixed and stained by methods previously described. 3 As expected, marked changes were observed in the basophiles. 4 These cells in large part were depleted of their basophilic granules. Their cytoplasm appeared as clumped, yellowish masses frequently with small basophilic granular areas interspersed. The distinguishing Golgi apparatus was hypertrophied. The cytological picture suggests a cell which has been in active secretory phase. These basophiles range in size from normal basophiles down to cells little larger than chromophobes. An even more striking feature, however, is their obvious numerical increase. There is a corresponding decrease in chromophobes. If changes occur in the acidophils, they are of such character that they could be detected only by careful statistical analyses of cell size and numbers. Technical methods which do not demonstrate the Golgi region either in positive 01 negative form would make difficult an accurate separation of the chromophobes from the smaller of the atypical degranulated baso philes. The injections of P. U. extract, therefore, seem to cause both an increase of basophiles and a rapid atypical degranulation of their cytoplasm. The pituitaries of castrates injected with P. U. extract cannot be distinguished histologically from the castrate controls.

I. A micro method for the quantitative estimation of agglutinins.

The agglutination of bacteria may be regarded as a precipitin reaction at the surface of the cells. 1 , 2 A method had been worked out in this laboratory for the quantitative estimation of precipitins, 3 and it has been found possible to adapt the procedure to the micro determination of agglutinins. Thoroughly washed, heavy bacterial suspensions are standardized by measuring their nitrogen content per cc. by the micro-Kjeldahl method. Accurately measured quantities of bacteria and an appropriate antibody solution or serum are then mixed and allowed to stand, with occasional mixing, at 37° for 2 hours and overnight in the ice-box. The remainder of the analysis is carried out as in the precipitin reaction, 3 except that it has been found convenient to make use of the higher speed possible with the angle centrifuge.† This was run in an ice-box compartment. Agglutinin nitrogen is determined by subtracting from the total nitrogen of the washed precipitate the amount of nitrogen contained in the bacteria used. Agglutinin N may also be determined from the difference between a blank without bacteria, and the supernatant from the agglutination. Both methods are illustrated in Table I. Agglutinin — agglutinin N × 6.25, under the assumption that agglutinin is protein. If the relative amounts of bacteria and serum are so chosen that the supernatant contains no more agglutinin the result obtained (divided by the volume) is the agglutinin content of the serum in milligrams per cc. In the table are given typical data for pneumococcus agglutinins. Application of the method to other systems is in progress.

An immediate outcome of the method is given in the following notice. Quantitative evidence bearing on the mechanism of bacterial agglutination has also been obtained, and will be published later.

The experiments of Brooks 1 based upon theory derived from the investigations of Warburg ct al., 2 Meyerhof, 3 Heymans and Heymans, 4 Barron, 5 and others, aroused interest in methylene blue as an antidote for carbon monoxide poisoning. Many case reports (Geiger, 6 Bell, 7 Nass, 8 Christoferson, 9 Brown¹⁰) and some experimental reports (Brooks, 11 Draize, 12 Haggard and Greenberg¹³) have increased the literature upon this subject. Even the contradictory conclusions of Haggard and Greenberg and the concise summary of Henderson 14 have not sufficed to alter the popular conception of methylene blue as an antidote for carbon monoxide poisoning.

In the following experiments, the effect of glutathione upon carbon monoxide poisoning was first investigated because of Brooks'results with a similar oxidation-reduction catalyst, methylene blue. However, the discouragingly negative results soon prompted an investigation of methylene blue.

White rats (100 to 200 gm.) were placed one at a time in a special chamber through which carbon monoxide (produced by the action of hot sulphuric acid upon formic acid, purified by washing with solutions of potassium hydroxide and with water, and made up to 2% by volume in air) was passed at a uniform rate of approximately one-half liter per minute. The animals were removed from the chamber 30 seconds after the disappearance of the righting reflex and immediately injected intraperitoneally with warmed solution. Three solutions were used: (1) Methylene blue: 0.406 gm. of methylene blue U.S.P. Medicinal and 5.24 gm. of dextrose per 100 cc. of solution. (2) Glutathione: 0.60 gm. of glutathione (Eastman Kodak) and 4.00 gm. of dextrose per 100 cc. of solution. (3) Dextrose: 5.42 gm. of dextrose per 100 cc. of solution.

The anterior pituitary substance responsible for thyroid hypertrophy and hyperplasia has been attributed to various anterior lobe hormones or to a separate thyreotropic hormone. 1 By heating a pituitary extract (Phyone), the thyreotropic substance has been separated from the contained gonad and growth stimulating substances.†

To determine the effect of heat on the activity of the growth hormone, young mice, 18 to 23 days of age, were used. 172 animals were divided into 5 groups: Group I injected subcutaneously with phyone; group II injected with phyone heated 60°C. 15 minutes; group III injected with phyone heated 80°C. 15 minutes; group IV injected with phyone heated to boiling 15 minutes. Group V was used as controls. All animals were injected with 0.1 cc. extract daily for 5 days. The average increase in body weight in 6 days was as follows: group I, 56%; group II, 55%; group III, 37%; group IV, 34%; controls, 35%. These data indicate that the growth hormone contained is not thermostable and is inactivated at some temperature between 60° and 80° C.

To determine the effect of heat upon the activity of the gonad stimulating hormone, female rabbits, 11 to 12 weeks of age, were used. These were divided into 3 groups of 7 each: group I, injected with 0.5 cc. of phyone, twice daily for 5 days; group II, with 1 cc. of phyone heated to boiling 15 minutes, twice daily for 5 days; group III, served as controls.

In group I all animals had hypertrophied uteri, and 6 showed from 2 to 4 corpora hemorrhagica in each ovary. No corpora hemorrhagica were present in group II, and in 6 animals the uteri were small and characteristically infantile, while the uterus of one animal, though small, showed a condition of hyperaemia. It is evident from these experiments that the gonad-stimulating hormone is inactivated by subjecting the extract to boiling temperatures for 15 minutes.

Mucin is a natural constituent of the intestinal contents; hence the assumption that it might be used as a culture medium for intestinal protozoa. So far as can be determined mucin has not been employed heretofore for this purpose. Samples of gastric mucin prepared by both Armour and Company and the Wilson Laboratories were used. These are dry powders, readily soluble in aqueous media and solutions may be sterilized in the autoclave.

A medium, composed of liver infusion agar slants 1 covered with a solution of 0.3% mucin in 0.7% aqueous NaCl with a 2 mm. loop of sterile rice starch to each tube, was used to isolate Endamoeba histolytica from one case of acute amoebiasis, from one human carrier and from 3 Macaca irus that were carriers. The amoeba grew in moderate numbers but cysts were not formed. Subcultures were made at 2 or 3 day intervals.

Trichomonas sp. from Pan niger, Cebus fatuellus, Macaca irus and Tapirus bairdii grew in abundance in a medium of 0.3% mucin in 0.7% NaCl solution. Subcultures were made at 3 to 4 day intervals. Trichomonas sp. and Eutrichomastix sp. from various reptiles have grown in abundance in 0.3% mucin in 0.5% aqueous NaCl solution, with subcultures every 7 days. A 2 mm. loop of rice starch to each tube seemed to increase the rate of growth of the flagellates from both mammals and reptiles. This medium, 0.3% mucin in 0.5% aqueous NaCl solution with rice starch, was used also to isolate 4 strains of Endamoeba sp. from Natrix sipedon and N. taxispilota. Growth and cyst formation was abundant. Subcultures were made every 7 days. Balantidium sp. from the intestine of Testuda galapogoensis has been cultured on this medium also, transfers being made every 7 days. The ciliates grew very well but did not encyst.

It is well established that in senile cataract two of the principal biochemical changes are (1) a great increase of calcium and (2) an apparent total loss of glutathione. 1 Evans and Kern 2 demonstrated that in experimentally produced parathyroid cataract (by extirpation of the parathyroid glands), the calcium content of these cataractous lenses increased to approximately 25 times the amount present in normal dog lens.

It is probable that glutathione plays a definite rôle in the internal oxidation-reduction and enzymie reactions of the lens (initially suggested by Goldschmidt 3 ). We were therefore naturally interested to learn the fate of the glutathione of the crystalline lens after removal of the parathyroids. Preliminary study indicated that the normal dog lens contains approximately 0.385% reduced glutathione, and that any figure considerably below this might be considered as indicating a loss of glutathione from the lens. The opportunity to study the fate of glutathione in parathyroid cataract was afforded by the early development of bilateral lenticular opacities in 5 young dogs (2 mos. old) that had been thyroparathyroideetomized in connection with another study. A complete description of these lenticular changes has been made elsewhere. 4

At the time the dogs were sacrificed because of severe distemper, they had been without parathyroids for 57-152 days. The earliest signs of lenticular degeneration were evident at about the 25th day. The dogs were kept on a sour milk-bread-liver diet; tetany was not observed. 5 The dogs were sacrificed by the intravenous injection of small amounts of sodium cyanide. Immediately after death the eyes were excised, the crystalline lenses removed and weighed on a watch glass.

Since the introduction of irradiated ergosterol and irradiated foods for the prevention and treatment of rickets, several methods have been employed in order to impart anti-rachitic properties to milk. Among these may be mentioned: (1) the irradiation of milk in both the powdered and liquid forms; (2) irradiation of the cow; and, (3) feeding to the cow substances containing large amounts of vitamin D, as for example cod liver oil or irradiated yeast. The irradiation of the cow as a means of imparting vitamin D to the milk is obviously most impractical from the commercial point of view, as is also the feeding of large amounts of cod liver oil. The irradiation of the milk itself has the practical disadvantage of being expensive since it involves the installation, maintenance and operation of expensive apparatus in the farm or dairy where it is employed. The direct irradiation of the milk and the indirect method of feeding irradiated yeast to the cow both have the disadvantage that they are not absolutely reliable methods of holding the vitamin D content at fixed concentration, so that repeated testing of the milk at regular intervals becomes necessary in order to be sure that the vitamin D content of the milk has not fallen below the standard.

In view of these difficulties it occurred to us that the method for adding fats to milk, which has been employed successfully by Holt, Tidwell and their associates, in our laboratories, for the past 4 years, might also be more practical for fortifying milk with vitamin D. Our procedure consists simply of homogenizing, in water, a concentrated oily solution of viosterol of known potency, with the aid of an emulsifying agent. We have tried several emulsifying substances, which are quite feasible as such, but in order to produce an emulsion which will be stable for months or years under ordinary environmental conditions, we have found lecithin to meet this requirement. This stable emulsion appears white as milk, is miscible with it, and has no objectionable odor or taste.

Some time ago one of us observed conditioned visceral responses in rabbits with extensive pallial destruction. 1 Acting upon this evidence and with the knowledge that the cortex was considered the usual if not only locus of the conditioning process by Pavlov 2 on the basis of work by Zeliony 3 we attempted to secure a decorticated animal higher in the evolutionary scale, and here describe the behavior of a decorticated dog with intact thalamus. Almost a hundred animals were utilized in this work which extended over a period of nearly 3 years.

A detailed record of the behavior of the animal was kept beginning with the time of operation (done in 2 stages). The day after operation the animal was able to stand rigidly on its feet when put in that position (extensor hypertonia) but slowly sank to the floor under its own weight without moving. Rigidity of the extremities with clasp-knife reflexes (Creed 4 ) was observed until its death. In 3 days, the animal was able to move with difficulty (which is exactly the time Goltz gives), in 5, it would stand for hours without motion but when in motion moved in a straight line. For 8 days no reaction to pistol shots nor a flash lamp beam was discerned, but on the ninth day a pupillary reflex was present and on the tenth the animal reacted to sound (same day Rothmann's dog registered). It grew to use its muzzle sensibility for purposes of directing its locomotion (Rothmann 5 ) keeping its head low to the floor. The animal reacted to olfactory sensations on the ninth day but never ate or drank spontaneously. On the tenth day circus movements appeared and grew stronger with occasional convulsions and somatic weakening until killed on the 19th day. During its life it showed marked resistance to impressed movements, was unable to respond promptly to stimulation nor could it cease its own activity until this wore off gradually. (Efforts were made to observe sympathetic excitation and phenomena which might be considered sham rage.)

Among the various theories concerning the etiology of orthopnea much support has been given to the idea that pulmonary ventilation is facilitated by the upright position. It has been shown that the vital capacity is generally higher, 1 that the total volume of the lung is greater, 2 that the mid capacity is greater, 3 that the accessory muscles of inspiration can be used to better advantage, 4 and that the diaphragm is lower 5 than in the recumbent position.

Aron 6 found in one normal individual that the intrapleural pressure was less negative when he was recumbent. There has, however, been no systematic study of the intrapleural pressure in pathological states leading to orthopnea. For this reason, we have taken intrapleural pressures in both the recumbent and upright positions in 13 individuals, 6 of whom were without orthopnea. Of 7 patients with orthopnea, 3 had cardiac decompensation, 2 were asthmatic, one had extreme pulmonary emphysema with bronchitis and one had marked ascites due to cirrhosis of the liver.

In all cases we found the intrapleural pressure to be less negative in the recumbent than in the upright position. This shift was more marked in the patients with orthopnea than in the control group. Occasionally the intrapleural pressure rose above atmospheric pressure in the recumbent position.

These observations support the mechanical theory of orthopnea. A less negative intrapleural pressure will limit the pulmonary ventilation and, according to the work of Starling, 7 interfere also with pulmonary circulation. By such a mechanism a vicious circle may be established, further congestion leading to greater interference with the ventilation of the lungs. In those cases in which a positive intrapleural pressure occurred during recumbency, direct pressure on the heart might interfere with its action.

From the classical observations of Lewis 1 , 2 in dogs it has usually been inferred that the initial ventricular deflections of extrasystoles arising from the right ventricle are down in lead I and up in lead III, while those arising from the left ventricle are up in lead I and down in lead III. These observations were generally accepted in the interpretation of the human electrocardiogram until it was shown by Barker 3 that on stimulation of the human heart extra-systoles were produced which were opposite in character to those obtained in the dog.

We found in dogs with an experimentally produced chronic pericarditis that the initial deflection of ventricular extrasystoles may be reversed from the usual type described by Lewis. 1 , 2 In these experiments it was noted that the heart did not fall away from the anterior chest as is always the case in the normal animal, an observation which suggested that the change might be due to the fact that the heart was held in its normal anterior position when the chest was opened.

To test this hypothesis the following experiment was performed on 6 normal dogs. An animal was anesthetized and securely fastened on its back to a board. The chest and pericardium were opened in the usual manner. Electrocardiographic tracings were made and at the same time 7 points on the ventricle designated by Barker 3 were stimulated. The board was then turned over so that the animal was suspended with its anterior surface to the floor. This allowed the heart to fall against the anterior chest wall, was always accompanied by slight rotation of the heart to the right, and represented the more nearly normal position of the heart in the dog. Extrasystoles were again obtained by stimulating the identical 7 points. The animal was then placed on its back and the right bundle was cut. Electrocardiograms were taken in this position and also after inversion of the animal.

Because of the relations of the autonomic nervous system to the thyroid gland, we thought it of interest to determine the effects of certain drugs which are known or believed to affect the autonomic nerves in a specific manner.

We tested the effect of these drugs in two ways: (1) We either injected solutions of these drugs subcutaneously or intraperitoneally, or administered them orally, and then examined the thyroid gland microscopically, noting the various characteristic features considered in the previous investigations from this laboratory; in particular making approximate counts of mitoses in various glands. (2) We determined the effect of these drugs on the hypertrophy and hyperplasia which is produced in the thyroid by injections of extracts of anterior pituitary gland.

Male guinea pigs, the initial weights of which ranged between 200 and 300 gm., were used for our experiments.

Experiments with Ergotamine Tartrate (Gynergen). To 8 guinea pigs ergotamine tartrate was given daily for 6 to 9 days in doses varying between 0.5 mg. and 6.0 mg. The drug was administered partly by mouth, and partly by subcutaneous injections. No toxic effects were noted. With the smaller doses there was a gain in weight, but with the larger doses the weight either remained stationary or showed a slight loss. Microscopically the thyroids were those of normal guinea pigs, as far as amount and character of colloid, relative number of mitoses and lack of hypertrophy are concerned.

In 9 guinea pigs, ergotamine and anterior pituitary extract were administered at the same time, 0.1 cc. or 0.3 cc. anterior pituitary extract being injected daily intraperitoneally, for periods of 6 or 9 days, and simultaneously 5.0 mg. ergotamine being given orally and 0.5 mg. injected subcutaneously. Controls were injected with anterior pituitary extract alone.

The iodate titration method of Hess, 1 with slight modifications, has been applied to the determination of reduced glutathione in tungstic acid filtrates of blood in a series of almost 1000 cases. The method is accurate to 2 -mg. per 100 cc, and the only known blood constituents to interfere are thioneine and possibly vitamin C. Thio-neine is partly precipitated by tungstic acid, and could account maximally for only 2 mg. per 100 cc. The effect of vitamin C was not determined but should be insignificant.

Cell volumes by the hematocrit method were determined on most of the cases and reduced glutathione was expressed also on the basis of 100 cc. of cells.

Ninety-three normals gave a mean value of 27.5 mg. per 100 cc. of whole blood and 69.5 mg. per 100 cc. of cells. Variations were from 8 to 46 mg. per 100 cc. of whole blood.

Three hundred forty-two diabetic patients gave a mean value of 25.1 mg. per 100 cc. of whole blood and 66.1 mg. per 100 cc. of cells. The reduced glutathione did not vary consistently with insulin therapy, and did not vary with the level of blood sugar. In a smaller series of cases, reduced glutathione did not correlate with the saccharoids of the blood.

Eighty-eight nephritis cases showed a mean of 21.8 mg. per 100 cc. of whole blood and 68.4 mg. per 100 cc. of cells. Reduced glutathione did not accumulate in the blood when other nitrogen products were retained.

Series of determinations in cases of prostatitis, circulatory diseases, liver damage, fevers, tuberculosis, pneumonia, cancer and other pathological conditions, showed no direct correlation of reduced glutathione with the clinical diagnosis. In general, when the cell volume was low in anemic cases, the whole blood reduced glutathione was reduced below the normal mean, and the figure per 100 ce. of cells was high (80 to 90 mg.). The opposite was true for polycythemic cases.

Henriques and Roland 1 have found that the daily excretion of iron in the urine of both normal and pathological cases is very small, amounting to only 0.08 to 0.32 mg. These values are much lower than those reported by earlier workers. Lintzel, 2 however, has claimed that not more than 0.02 mg. of iron per liter can be found in normal urine. Likewise, other investigators 3 have found the urine of patients suffering from various diseases to be practically iron-free. Because of its bearing on problems relating to the metabolism of iron we have investigated anew the question of the magnitude of the urinary iron excretion.

The urine was obtained in clean bottles from male subjects who were pursuing their usual routine in the laboratory. In brief, the analytical procedure consisted of the following steps: A measured volume, usually 500 cc, of the fresh 24-hour urine specimen was evaporated to dryness in a silica dish, and then incinerated in an electric furnace at about 500° C. for 8 hours. The ash was dissolved in dilute HC1, brought to a volume of 50 cc, and aliquot portions were removed for the determination of iron by the thio-glycolic acid method. 4

The analytical technic was subjected to repeated tests and found to be adequate. All reagents were shown to be free from iron. Small quantities of iron added to urine were quantitatively recovered, proving that iron was not lost in the manipulations. Finally, a solution of pure substances was prepared in imitation of the main constituents of urine and this solution was evaporated, ashed, and analyzed for iron. None could be detected, proving that the iron found in the analyses was not due to contamination.

Rats were made anemic by restriction to a diet of cow's milk. This milk was obtained directly in glass bottles from Guernsey cows and contained an average of 0.14 mg. of copper per liter. One group of 5 anemic rats was sacrificed. The stomachs and intestines were removed, and the carcasses were analyzed for iron and copper. Another group of 5 rats received daily supplements of 0.5 mg. of iron as pure ferric chloride. A third group of 6 rats received daily intraperitoneal injections of 0.5 mg. of iron, and a fourth group received 0.5 mg. of iron and 0.025 mg. of copper as copper sulfate each day per os.

Hemoglobin determinations were made at intervals on blood samples secured from the tails. In 17 days the animals receiving iron by mouth had increased their hemoglobin from an average concentration of 4.15 gm. to 7.23 gm. per 100 cc. of blood. The rats receiving iron intraperitoneally increased their hemoglobin from 4.03 gm. to 12.92 gm. in the same time, and the rats receiving iron and copper orally increased their hemoglobin from 3.90 gm. to 13.40 gm. per 100 cc. of blood. These data confirm conclusions reached in previous work from this laboratory, 1 namely, that supplements of iron bring about some hemoglobin production in anemic rats and that intraperitoneal injections of iron result in hemoglobin production at about the same rate and to the same extent as oral administrations of both iron and copper.

The average values for the analyses of the carcasses of the rats for iron and copper are recorded in Table I. These figures show that the hemoglobin production in the rats receiving injections of iron is associated with an increase in the copper content of the body over that of control rats sacrificed at the beginning of the experiment. It has been calculated that the average retention of copper in this group of rats was 73% of the total intake in the form of the copper of the milk, while in the group receiving iron orally it was 17% and, in the group receiving both iron and copper, only 12% of the total intake during the experimental period.

In a paper with Hess 1 one of us called attention to the occurrence of carotenemia. At the time it was pointed out that this might in certain cases cause confusion with icterus. Later it was observed that the carotenemia might be very intense in diabetic patients on the high vegetable diet of the time.

It is obvious that carotene must exert some influence on the icteric index of the blood, a fact which has been appreciated by most of the workers in this field. Fiessinger, Walker and Thierry 2 from their work felt that the effect was negligible, Boeck and Yater 3 found that carotene accounted for from 10 to 55% of the icteric index, while White 4 concluded that the influence was negligible except in marked cases of xanthemia accompanying diabetes.

The subject has been reinvestigated on 161 patients, 75 of whom were suffering from diabetes mellitus, the carotene being determined in terms of the icteric index (lipochrome index) and in mg. per 100 cc. of serum. In 43 diabetic cases on which icteric indices were determined carotenemia was found to account for an average of 45% of the corresponding icteric indices, the extremes being 19.3 and 70.5%. The average icteric index was 7.1. The average lipochrome index for the 75 diabetic patients was 3.4, corresponding to an average carotene content of 0.25 mg. per 100 cc. of blood.

The average lipochrome index for the 86 non-diabetic patients was 1.7 with an average of 0.11 mg. carotene per 100 cc. of blood. In 10 of these cases with icteric indices below 10, carotenemia was responsible for an average of 20.6% of the corresponding icteric indices, while in 21 cases with icteric indices between 10 and 100 it averaged 8.6% and in 6 cases with icteric indices above 100 the lipochrome indices averaged only 1.2% of the corresponding icteric indices.

From the above findings it would not appear that carotenemia alters the validity of the icteric index except in cases of marked carotenemia such as may be found in some diabetic patients, and then only if the icteric index is quite low.

The observation by Cutting and Tainter 1 of a marked acceleration of the metabolism of animals injected with small quantities of dinitrophenol stimulated a study of the effect of the injection of the dinitrophenol on the oxygen uptake of rat tissues. Rats were injected with the sodium salt of dinitrophenol (40 mg. per kilo of body weight). After a period of from 30 to 40 minutes the animals were killed by a blow on the head and the oxygen uptake of the tissue determined as quickly as possible, employing the technique previously described. 2 The respiratory quotient was determined as described by Richardson. 3

Table I presents the results which were obtained employing rat liver. It is seen that when phosphate Ringer's solution was employed, the oxygen uptake of rat liver slices obtained from the injected animals was essentially the same as that obtained with the control. The oxygen uptake of the liver tissue obtained from injected animals, when suspended in phosphate Ringer's solution containing glucose, was slightly less than that of the control liver. The respiratory quotient of the liver tissue from injected animals was lower than that of the control liver.

Experiments are now in progress in which observations are being made on the oxygen uptake of other rat tissues and also experiments dealing with the effect of the dinitrophenol on the oxygen uptake of rat tissue in vitro.

The present study was prompted by the findings of R. and S. Weissmann-Netter 1 who have made a series of observations on the carbon dioxide and the pH of the blood of patients with migraine. They found that the hydrogen ion concentration and the alkali reserve are normal in the periods of freedom from attacks but that a tendency to alkalosis develops 48 hours prior to an attack. They also reported that hyperventilation may result in an attack of migraine, probably through an excessive loss of carbon dioxide. As a result of the apparent success of controlling the migraine attacks by treatment with ketogenic diets, 2 it seemed desirable to reinvestigate the acid-base balance of the blood of individuals subject to migraine. To date 10 cases of migraine have been followed at intervals, the blood having been taken as early in the attack as possible, at the height of an attack, and immediately after cessation of the attack. It was found that the serum pH, CO₂ content, chloride and total base remained within normal limits, if one excepts 2 cases in which the pH was elevated slightly, due in all probability to the accompanying vomiting. It is probably of some significance that a fairly definite elevation of the blood cholesterol occurred in 50% of the cases.

The time-dilution curve obtained for most haemolytic systems, which is described by the equation

can be analyzed by a method already described (Ponder and Yeager 1 ). The method is somewhat laborious and the analysis is more easily carried out as follows:

One requires a piece of cellophane, about 40 cm. square. The length of the right hand side is divided in such a way that the divisions represent 12.2, 24.4, 36.6, 48.8, 61.0, 73.2, 85.4, and 100% of that side. A series of lines are drawn on the cellophane joining each of these divisions to the left hand corner.

The time-dilution curve is drawn on a large scale in the usual way, and the asymptote (x) is marked with a horizontal line. Place the lower edge of the cellophane along the abscissa of the curve, and move it from right to left until the oblique line corresponding to the 100% division passes through the point at which the horizontal line corresponding to the asymptote cuts the ordinate of the curve. The oblique lines then cut the ordinate of the curve at a series of points. From the curve, read off the time corresponding to each of these points, add up the times, and divide by the time corresponding to the oblique line for 85.4%. Call the result S. The value of n(= 1/p) is found by consulting Table I.

In this way a time-dilution curve can be analyzed in less than a minute, and, if the curve is of the usual form, the value of n is obtained with an error no greater than ±0.05, for values not shown in the table can be found by interpolation. When large numbers of time-dilution curves are obtained in the course of an investigation, a rapid method of analysis is a great convenience, although it is often necessary to analyze the curves more carefully later on.

In the course of their investigations on the nutritive requirements of the chick Elvehjem, Kline, Keenan and Hart 1 found that about half of the vitamin G potency of desiccated yeast was destroyed by heating at 100°C. for 6 days. With the hope of securing evidence of the supposed duality of vitamin G, as postulated by Sure 2 and others, we have studied the effect of heat on both the growth-promoting and dermatitis-preventing activities of dried yeast. After preliminary desiccation the yeast was heated in an electric oven at either 105 °C. or 150°C. for various lengths of time. The resulting products were fed, at a level of 1.5 gm. per mouse per week, to 3-weeks-old mice maintained on an otherwise G-deficient diet. The basal ration was prepared by mixing 200 gm. of vitamin-free casein, 265 gm. of cornstarch, 205 gm. of sucrose, 200 gm. of lard and 40 gm. of McCollum's salt mixture. Each animal received apart from the basal diet 2 drops of a tested cod liver oil and an ample amount of a vitamin B preparation. The latter was made from wheat germ by extraction with 93% alcohol, distillation of the extract in vacuo, and separation of the potent aqueous residue from an inert oil.

The vitamin G activity of the yeast preparations, as measured by the growth response of the mice, is recorded in Fig. 1. No decrease in the vitamin G potency of yeast heated at 105° was observed. Further experiments with mice fed varying amounts of dried yeast, heated at 105° for 2 weeks, would be necessary to reveal whether a slight inactivation occurs. From the limited data available, however, it may be concluded that the destruction of vitamin G under these conditions is certainly not extensive. Moreover, Block and Farquhar, 3 who used the rat as their experimental animal, have recently reported that they observed no decrease in the growth-promoting activity of yeast that had been heated at 100° for 2 or even 4 weeks.

Our data show further that the growth-promoting activity of yeast is slowly lost as a result of prolonged heating at 150°. There is a little vitamin G remaining even after one week of such treatment, but the vitamin is entirely destroyed in 2 weeks. Each of the 4 mice receiving the yeast that had been heated at 150° for 2 weeks developed typical skin lesions, as described by Bing and Mendel, 4 and all were dead in 7 weeks. At this time the mice in the other groups were still alive and their skins were in good condition. It appeared that the mice which had received sufficient vitamin to permit some growth had likewise received enough of the anti-dermatitis factor to meet their requirements, and the experiment was therefore discontinued.

Our knowledge of some of the many aspects of the problem of anemia has been enlarged recently by studies on the milk anemia of the rat. Thus from the work of the Wisconsin investigators the importance of Cu in Hb regeneration is now generally appreciated. Cu therapy, however, does not present any particular difficulty, while Fe deficiencies and Fe therapy still remain important problems in the clinic. For such investigations milk anemia is a convenient experimental approach. Although much work has been done there is still some uncertainty regarding the response of anemic rats to small doses of Fe. 1 , 2 Certain investigators 3 , 4 have reported that glutamic acid supplements Fe in nutritional anemia, but this has not been confirmed by others. 5

The following experiments were carried out with about 300 rats in order to clarify some of the discrepancies in regard to the effect of Fe in anemia and to determine whether glutamic acid could be employed as an effective vehicle for the administration of minerals in secondary anemias.

Experimental. Most of the rats employed in the experiments were bred according to a modification of Elvehjem's technique. 6 The anemia producing diet was powdered whole milk (Klim, containing 1.8-2.0 p.p.m. of Cu and 14-17 p.p.m. of Fe) which was fed in the solid form. Hb estimations (Newcomer) were made weekly and red blood counts at frequent intervals. The inorganic Fe and Cu salts employed for supplements were of the highest purity; the ferrous glutamate was Cu-free and the glutamic acid contained less than 0.2 mg. Cu per 100 gm.

Some of the experiments with FeCl₃, ferrous glutamate and glutamic acid are summarized in Table I. It appears from this table that Hb regeneration is just as effective with ferrous glutamate (Group 1) as with ferric chloride (Group 2) and that addition of glutamic acid produced similar results (compare groups 2 and 3). When glutamic acid (Group 5) was added to supplements of Fe+Cu (Group 4) the hematopoietic response appeared to be slightly greater.

Before the physiology of sexual dimorphism in birds can be completely understood the normal change from the male winter plumage to the female-like summer plumage in the Mallard drake must be explained. This report embodies certain facts regarding feather pigmentation and growth rate in Mallard drakes which must be considered in future studies of the physiological basis for this “eclipse” or “summer plumage” change of the adult male duck.

The males used in these experiments are semi-domesticated Mallards, the majority of which are genetically “dark-phase mallard” (M Li). 1 These drakes are very uniform in breeding, and in adult body weight, between 980 and 1160 gm. The males are left in outside yards until they are to be used in an experiment, when they are brought indoors and kept in artificially heated and lighted pens, the hours of illumination being approximately the same in all trials. At the beginning of the experiment, 50 to 60 feathers are plucked from the middle of the ventral surface on either side of the keel, the plucked area being in the same location on each drake. All trials were conducted between the months of September and the following April. At any time within this period the growing feathers normally show no detectable variation in pigmentation or growth rate.

The depth of color in the growing gray breast feathers may be increased artificially by 2 different treatments, oral administration of desiccated sheep thyroids (Armour & Co. preparation, 0.2% iodine), or subcutaneous injection of an extract of the urine of young men, prepared according to the method of Funk, Harrow and Lewja 2 for male hormone. The thyroid increases the pigmentation and the growth rate of the feathers, and the urine extract augments the pigment increase stimulated by the thyroid. The increase of pigmentation in the partially-grown gray breast feathers is recorded in the tables by direct comparison of the normal and modified feathers of each drake. Each plus symbol indicates a definite intensification of the melanic pigment. Feathers with 5 plus symbols, in addition to being very dark in color, have a definite edging of light brown and in this respect resemble the feathers of the eclipse (summer) plumage. Each drake was killed at the termination of the experiment and, in parts B and C of Table I, a certain number of growing feathers were removed for measurement.

If individual bones differ in their rates of growth and if each of the major long bones is the product of 3 to 8 centers of ossification one may feel surprised at the linear relationships reported by Scammon and Calkins 1 between diverse external dimensions of the human fetus. We need assurance that the simplicity of these gross observations does not mask a greater complexity in the constituent tissues.

In this fetal series we report the lengths of the bone shafts alone, before the appearance of osseous epiphyses, from dried preparations in the Department of Embryology. The observations are well distributed through fetal life, their variability is small, and the error of measurement is less than 2%, but we do not know what change took place in drying.

The rectilinear relationships in Fig. 1 demonstrate our confirmation, by a different technique, of Scammon and Calkins' 1 belief that the relative growth of human fetal long bones is approximately linear. This interpretation applies also to fetal skull growth in the horse 2 but fetal limb growth in the horse (to be published) required a second degree equation. We have also replotted the data of Schmalhousen and Stepanowa 3 concerning bone growth in the chick embryo. Here we find that lengths, as well as weights are interrelated as power functions and thus exhibit growth partition. 4 , 5 The chick humerus length is a function of femur length with exponent 0.87, and approximately the same exponent, 0.88 describes radius or ulna growth, whereas in the hind limb the exponent for tibia or for tarso-metatarsus is 1.05. In view of these facts we cannot eliminate the possibility that our “straight lines” are but segments of power curves with exponents close to unity.