lil a A in a a a il | id Li) Vee , rire ita ; ’ y*t =e. pe egies : — e - ie: =] = le 2 THE HARVEY SOCIETY THE HARVEY LECTURES Delivered under the auspices of THE HARVEY SOCIETY OF NEW YORK Previously Published FIRST SERIES s < SECOND SERIES . THIRD SERIES . FOURTH SERIES. FIFTH SERIES SACRE SERIES 0 ste SEVENTH SERIES . EIGHTH SERIES . NINTH SERIES TENTH SERIES .. ELEVENTH SERIES TWELFTH SERIES . 1905-1906 1906-1907 1907-1908 1908-1909 1909-1910 IQIO-IQII IQII-I912 IQI2-1913 1913-1914 IQ14-1915 IQI5-1916 1910-1917 ‘ = s fox ans ae 4 2 oe “ wv & hen () fae acthy Me dre ‘suvtya 3 ’ hai DIETARY RELATIONSHIPS 163 acter and amount of the inorganic portion of the food mixture. When this is properly adjusted by means of salt additions growth can proceed on certain mixtures. After prolonged suspension on ration 713 the addition of salts led to growth. Equally pro- nounced failures with rations of this type containing beans or peas are abundant in our records. In marked contrast to our failure to secure satisfactory nutri- tion with any of the many combinations of from two to five kinds of seeds, stand the nearly normal growth of rats which are limited from an early age to a monotonous mixture of a seed and a leaf. We have tested most thoroughly combinations of the alfalfa leaf with several seeds, since it can be secured as a meal and as a flour, the former being much employed in animal production. The products which we have employed were a very fine meal consisting principally of the leaves, but containing visible pieces of stem, and also a flour which was practically entirely derived from leaves. Both products came from imma- ture plants dried in the sun. Chart 5 illustrates the fact that rats can take from the age of 35 days a mixture of 60 per cent of any one of at least three seeds, oats, wheat, maize,’ with 40 per cent of alfalfa leaves, and grow to a size closely approximating the normal, and produce young. The same proportions between polished rice and alfalfa,'* peas and alfalfa,17 and cottonseed flour and alfalfa ‘1’ yield less satisfactory results, but with these growth is far superior to what we have ever observed with the seeds alone in any combinations or proportions, when fed to rats whose water supply was dis- tilled, and therefore salt free. On ration 685, oats and alfalfa,17 14 out of 17 young were reared. The one litter of five produced by rat 687 all died soon after birth. Four of the 7 young in the litter from rat 686 were reared. No young have been reared on the polished rice and alfalfa mixture (rat 478). Two factors contribute toward making these combinations of seed and leaf superior to seeds alone. First, the character of the inorganic content of the leaf is much greater and is of an entirely different character from that of the seed. In the following table HARVEY SOCIETY 164 DIN podd haw 4 i oO DIETARY RELATIONSHIPS 165 are tabulated the average analyses of alfalfa hay, cabbage, wheat, oat and corn kernel, potato and of milk. In 100 Parts Dry SuBSTANCE (From Forbes Bul. 207, Ohio Exp. Station) Tot Kale Ma Cal) \Merl Cl ye s PUTA A Sees lec ictiake the 7.380) 1.641 | 0.097 | 2.146 | 0.223 | 0.298 | 0.338 | 0.172 Cabbage’ +. .).3) +i. « 10.850} 2.510 | .805 | 1.663 | .234] .853] .711) .372 WGA ee nS Se dines 2.080] 0.504 | .026| .041] .141| .058|) .408} .009 (0 SND Aa eu oa 1.830) .361| .058] .097| .086} .098| .3885]| .005 OTA eas ofa ty 1.450} .359| .012|) .023] .136|] .013] .289} .005 POCA joie \=)3:2'= = shsces 3.79 | 1.890] .083] .072| .113|] .131] .279| .099 MMe rect nats ste 5.67 | 1.160] .344] .909| .088| .791]} .650} .057 It is fortunate indeed that the inorganic content of the seed and the leaf supplement each other in a manner which meets the demands of mammalian physiology. The second way in which the leaf supplements the grain is in respect to the as yet unidentified dietary factor, fat-soluble A. This substance is present in the alfalfa leaf to the extent of at least five times the amount in the maize kernel, so that any diet containing thirty per cent or more of this leaf is not enhanced by the inclusion of butter fat. Cabbage and clover leaves also are rich in this dietary essential, but our experience with these is less extensive. There is, of course, to be expected a supplementary relationship among the proteins of a mixture of two or more foodstuffs. It could hardly happen that the proteins from several sources would be particularly low in their content of the same amino acids. That the limiting amino acid is not the same in several of the more important seeds is made evident by the curves shown in Chart 6. Rats 652 and 493 received a similar food mixture except that the former had 10 per cent of gelatin supplementing wheat proteins in place of an equivalent amount of dextrin. The gelatin con- taining diet promoted growth at a rate greater than the normal expectation, as compared with a rate about half normal in the gelatin free control.*® The second pair of curves, rats 756 and 785 received rations alike in all respects, except that in the former the protein con- HARVEY SOCIETY 166 sood f $ 9SL Pe) Oo savjdi uy vad Yim So awps DIETARY RELATIONSHIPS 167 tent (19.9 per cent) was derived solely from a mixture of peas and gelatin '? and the latter from a mixture of navy beans and gelatin.’ In neither case did the rations promote appreciable growth, although the entire group receiving beans grew notice- ably better than did those getting peas. Gelatin does not supple- ment efficiently the proteins of either the pea or bean. In the second period in each ease casein was substituted for the gelatin with the result that growth began at once at a fairly good rate. The proteins of the pea or the bean, when taken as the sole source of nitrogen, are of very low biological value.'* An inspection of the curves of rats 647 and 646 shows that gelatin has a very favorable supplementary relation to the oat proteins.1® Gelatin when combined with the proteins of the maize kernel does not enhance the value of the latter.'® In 1914 we repeated the efforts of Slonaker*® to find the effects of a strictly vegetarian diet of wide variety, on the growth of the rat. We offered 19 different foods of vegetable origin, giving the animals five or six at a time and changing their bill of fare at intervals of a few days. The list of foods included grains, leaves of plants, nuts, corn and wheat germ, peas, and flaxseed meal. None ever grew beyond half the adult size.1* We duplicated in all essentials the results which Slonaker had observed. Having gained some insight into the peculiar dietary properties of certain natural foods I predicted two years ago that a mixture of maize 50, alfalfa leaves 30, and heated peas 20 would adequately nourish a young rat during growth. A trial proved that this strictly vegetarian food mixture induced satis- factory nutrition during growth, supported repeated reproduc- tion and rearing of a fairly large percentage of young.’* Chart 7 shows the curves from left to right of mother, daughter and granddaughter. A litter of great grandchildren were success- fully weaned and appeared entirely normal when we discontinued the experiment. I should like to emphasize the importance of a proper adjustment of the well recognized dietary factors by point- ing out that no young would be reared if this diet were changed to inelude less than 20 or more than 45 per cent of alfalfa, the maize replacing or being replaced by alfalfa in the adjustments. HARVEY SOCIETY 168 ova Td e Ud vA pad Se YD o4 a) aL !ySiuo fn Z "hike 6 ok fol yyasg |ayg SH ONG doug 4 |. =, Paar 0:0€ o'0S OW A Dfav- HIP oe See im 169 DIETARY RELATIONSHIPS an jeins Mai vw he \|Pea | ied ale he e Hirth of Ypuirg ; a | &. Oats 60.0 Bean|S 20.0 Dexth B.Fa 170 HARVEY SOCIETY In Chart 8 are shown representative curves showing the failure of the proteins of either peas or navy beans to supplement very efficiently the proteins of the maize kernel ;’ rats 779 and 780. In either of these rations the peas or beans could be re- placed by dextrin and a curve of nearly the same shape would be secured. Both pea and bean proteins improve the proteins of the wheat kernel,’’ for these curves are much better than would be secured if the legume seed were replaced by carbo- hydrate. (Chart l-rat 619; Chart 2-rat 319.) The most re- markable result of studying these combinations is seen in the very different results of supplementing oat proteins with bean pro- teins as contrasted with peas. Pea proteins are wasted when com- bined with those of the oat (while the proteins of the navy bean and oat mutually make good each others deficiencies.17 It follows therefore that the proteins of the pea and the bean are very dissimilar in respect to their yields of certain amino acids and that the same amino acid is not the limiting factor in each. In Chart 9 are presented curves obtained with a series of diets designed to show the relative richness of the fat-soluble A in several natural foodstuffs. The diet, aside from its content of natural food, consisted of such a mixture of purified food mate- rials as would support growth, when both the unidentified dietary factors, the fat soluble A and water-soluble B were added. The role of the natural food in these rations was to supply these two factors.17 These curves make it clear that about 30 per cent of alfalfa leaf is the minimum amount which will supply the fat- soluble A in amount sufficient to induce complete growth and induce the production of a few young. The young were not reared on this ration. In Charts 1, 2 and 3 curves were shown which illustrate that even 70 per cent of corn, wheat or oat kernel do not furnish enough of the dietary A to support com- plete growth and repeated reproduction and well-being over the full span of life. It is of particular interest therefore to compare the effects of flax seed and millet seed as sources of the fat-soluble A. Both of these are richer in this substance than are the cereal grains, and millet seed proves to be unique among the seeds we have 171 DIETARY RELATIONSHIPS ——J 37 Oo x 172 HARVEY SOCIETY examined as a good source of this dietary factor. When 25 per cent of millet seed was combined with purified foodstuffs growth was completed and one rat has produced two litters of young. After eight months on this diet she appears to be in perfect nutrition. The slow but long-continued growth on these rations whose content of both the unidentified dietary factors was solely derived from such small amounts of hemp seed, flax seed and millet seed shows that these stand intermediate between the cereal grains on the one hand and the plant leaf on the other as sources of the fat-soluble A.*7 We have carried out a series of experiments designed to show the minimum amounts of several seeds which are necessary to supply enough of the water-soluble dietary essential for the com- pletion of growth and the production of young. Chart 10 illus- trates the results of this inquiry. The diet in all cases consisted of a purified food mixture, to which was added 5 per cent of butter fat to insure an adequate supply of the fat-soluble A. The sole requirement of the portion of natural food incorporated in the food mixture was to furnish the water-soluble B. While 15 per cent of either maize or oat kernel fails to induce growth at the maximum rate 15 per cent of wheat not only enables the animals to complete their growth, but induces sufficiently good nutrition to lead to repeated reproduction. However, no young can be reared on either of the two lowest wheat rations here described. The factor which is responsible for this failure is the low content of the water-soluble B (Chart 2-rat 223B). Peas and beans likewise are shown to be rich in this dietary essential, but they are both poor in the fat-soluble one.‘‘ The fact that rats 744 and 743 did not respond promptly with renewed growth upon the addition of an alcoholic extract of wheat germ con- taining additional water-soluble unknown is without significance. After a long period of stunting, response to improved diet is not always immediate. Diets deficient in one factor only and for this reason leading to suspension of growth do not all lower the vitality of the animals in the same degree. The extent of the debility determines whether they can respond promptly with growth when the diet is improved. DIETARY RELATIONSHIPS 173 w SGrams. rie] eis n{S-pooy | part see 174 HARVEY SOCIETY I desire next to attempt to offer an interpretation of certain experimental and clinical evidence which may seem to be out of harmony with the simple working hypothesis regarding the essential factors which are concerned in making a diet adequate, and with which I have attempted to explain the cause of the failure of animals to be adequately nourished on a considerable number of diets from various sources. I refer especially to the theory of Funk, which has found wide acceptance, that no less than four syndromes recognized as distinct pathological states, viz., beri-beri, scurvy, pellagra, and rickets, are all referable to inadequacy of the diet, and that each is the result of the lack of an adequate amount of a particular chemical substance in the food. To these Funk gave the general name vitamines. In seek- ing the causes of the inadequacy of certain diets I have dispensed with all but two unidentified substances. There are two lines of evidence which offer strong support of the validity of the assumption that in the case of the factor which I have ealled the water-soluble B we are dealing with a single substance rather than with two or more. The first is based upon the results of feeding such rations as are described in Chart 10 in which all of this factor is supplied by as little as 15 per cent of wheat, or 25 per cent of peas or beans, alfalfa, ete. We have tested a list of foods which ineludes four grains,—wheat, oat, maize and millet, two legume seeds,—peas and beans, three varieties of leaf,—clover, alfalfa and cabbage, and three oil seeds,—hemp, flax and cotton seed, and find that in all these about 15 to 20 per cent is all that is essential with a mixture of purified foodstuffs and butter fat, to promote growth at the normal rate. While this does not establish the conclusion, we may reasonably expect that if the dietary factor B consisted of two or more chemical substances essential for normal metabolism, we should discover in some of these natural foodstuffs a shortage of one as compared with the other, in greater degree than is actually the ease. The second line of evidence which supports the idea that we are here dealing with but a single chemical substance rather than several, is to me still more conyineing. Itis as follows: A mixture DIETARY RELATIONSHIPS 175 of purified food substances becomes capable of inducing vigorous growth when 3 per cent of wheat germ supplies the factor B and ® per cent of butter fat the factor A. Now the wheat embryo can be completely extracted with ether without losing its potency in the least degree. The ether extracted residue, we have ex- tracted eighteen hours in a continuous extractor with hot ben- zene in one instance and with hot acetone in another. The extracts obtained with benzene or acetone will not induce growth when combined with the purified food mixture containing butter fat, even when the extract equivalent to 28 per cent of wheat germ is added. In either case, however, 3 per cent of the ether- benzene or ether-acetone extracted residue is still as effective as ever. Now rats which are at all depleted with respect to the factor B do not recover when only 2 per cent of wheat germ is present in the ration. They do recover from a condition of extreme weakness, and partial paralysis on our ration with the extracted germ to the extent of but 3 per cent. While large doses of the material extracted from wheat germ by hot benzene or acetone contain enough of the factor B to cure a pigeon of polyneuritis, the amount of the curative substance extracted by these solvents is almost negligibly small. The point I would emphasize as of significance in these results is the very small probability that if there were present two or more chemical substances, indispensable for the maintenance of physiological well-being, both or all should be practically com- pletely insoluble in benzene and acetone as well as ether, and that both or all should be readily soluble in water and in aleohol. It would seem a remarkable coincidence if several compounds should show such complete uniformity in their solubilities, stability toward heat and so close a parallel in the proportions in which they occur in the seeds and leaves of widely different orders of plants. When contrasted with the very great variation in the content of the fat soluble essential in different seeds, and the richness of the leaf as compared with the seed in its content of this substance, it becomes the more improbable that in plant tissues generally there should exist a definite proportion between the remaining essential substances, whatever their number. 176 HARVEY SOCIETY I am conscious of the question which you will ask concerning the cause of scurvy and of beri-beri. Are not these both dietary deficiency diseases? During the past two years Mr. Pitz and 1*° have repeated all the more important work reported by others on the production of experimental scurvy in the guinea pig by diet. We have confirmed the observations of Jackson and Moore,” to the effect that with a diet of oats and milk ad libitum, scurvy invariably results. Rolled oats induce the onset of the symptoms sooner than unhulled oats. Only those diets which produce feces of a character very readily eliminated will relieve or prevent the disease. Milk to the extent of 10 per cent of the solids of the diet supplies all the unidentified dietary factors nec- essary for growth in the rat or swine and probably for other mammals as well. We are convinced that the guinea pig suifers from scurvy on a diet of oats and milk because of the consti- pating character of the diet. Oats produce pasty feces, and the guinea pig, being unfortunate in the anatomy of its digestive tract, is quickly debilitated by its inability to empty the large and very delicate cecum. This harmonizes with our observation that orange juice, the panacea for scurvy in the human infant, gives protection to the guinea pig, but is not so efficient as to enable this species to take an oat and milk diet and grow con- tinuously over a long period. We have furthermore been able to effect a complete cure of guinea pigs on a milk and oat diet after they could no longer walk, and showed badly swollen joints and the hemorrhage of the gums, by liberally dosing them with liquid petrolatum. Cavies so relieved have been as active and healthy as if on a diet of green grass, and have resumed growth and have continued to grow steadily, though at a rate slower than the normal rate during three months following the attack of scurvy, while confined to the same milk and oat diet which gave them the disease. The petroleum oil treatment was of course persevered in throughout this period. I question whether anyone would postulate the existence in liquid petrolatum of a ‘‘vita- mine’’ specific as a protection against scurvy. I venture to suggest that the failure of Holst ** and his co- workers to prevent or relieve scurvy in guinea pigs by feeding DIETARY RELATIONSHIPS 177 dried cabbage, finds its explanation in the failure of the latter to take up water in the intestine and again act as a succulent vegetable. Hess ** has recently, on the basis of anatomical lesions and for other reasons, suggested that scurvy and beri-beri show more points of similarity than have been recognized by clinicians and was inclined to consider the two as essentially the same pathologi- cal condition. His efforts toward curing scurvy with wheat germ, and with yeast were unsuccessful, whereas orange juice was potent as a curative agent, and he was forced to conclude that the two conditions were distinct. I will offer the suggestion that the latter view is correct. Scurvy in the guinea pig is the result of the retention of feces. I do not know whether or not the same is true of human scurvy. Neither do I know the cause of the hemorrhage of the joints and gums, whether they are the result of the absorption of a toxic substance of bacterial origin, which injures the blood vessels, or whether they are due to the invasion of the tissues by an organism, through an injured cecal wall. The recent observation of a streptococcus in the congested joints by Jackson and Moore, is suggestive in this connection. I am inclined to attribute the protective power of orange juice as an antiscorbutic to its content of certain salts of citric acid, rather than to the presence of an unidentified organic substance of the class of the so-called vitamines. Its efficiency for the guinea pig appears to be somewhat less than for the human. This may well find its explanation in the much more delicate and inefficient structure of the digestive tract of the cavy as compared with that of the human, so that less efficient protective agents may serve for the latter than for the former. While guinea pigs are protected against scurvy by orange juice we have not seen them grow to any appreciable extent on a diet of oats and milk fortified with orange juice. If the results of experiments of three months duration are to be trusted orange juice does not appear to be any more efficient than is petroleum oil in protecting guinea pigs against scurvy when the animals are kept on a diet of oats and milk. I have dwelt thus long on the subject of scurvy in the hope that the tentative conclusions I have offered may help 12 178 HARVEY SOCIETY in clarifying the interpretation of a most confusing mass of experimental data. That beri-beri is a disease of dietary origin there is no doubt. That it can be relieved in the fowl by the extracts which con- tain the dietary factor B is equally certain. Much less definite statements can be made regarding the other diseases, rickets and pellagra, which Funk has attributed to lack of specific substances in the diet. The evidence as I understand it seems to point strongly toward diet as at least an important contributing factor. Assuming that they as well as beri-beri are of nutritional origin, I venture to state that they are the result of unsatisfactory adjust- ments among the well recognized constituents of the diet rather than to the lack of specific chemical substanees. I have shown you how diets may be designed so as to induce faulty nutrition as the result of the shortage of each of the two unidentified dietary factors A and B; of the inadequacy of the protein con- tent, because of an unsatisfactory yield of amino acids, and as the result of an unsatisfactory composition or inadequate amount of the inorganic content of the food mixture. I may add to these a fifth, viz., malnutrition as the result of the presence of toxic substances in the natural foodstuffs, particularly cotton seed and wheat products, and perhaps also a sixth cause due to mechanical injury to the digestive tract caused by foods pro- ducing feces of unfavorable character (e.g., oats) and by con- stant distention of the alimentary tract as the result of excessive fermentation of foods such as beans and peas which contain a large amount of hemicelluloses. I have attempted to show, so far as experimental data are available, the specifie dietary properties of our natural foods, and how these may be so combined as to make good each other’s de- ficiencies and shortcomings. In presenting this data it has also been emphasized how great is the need of more knowledge in this field. It is indeed remarkable how difficult it is to secure combinations of natural foods which will serve as monotonous diets over a long period and promote in an entirely satisfactory way the well-being of an animal. The consumption of a varied diet makes in some degree for safety, but will by no means insure DIETARY RELATIONSHIPS 179 safety. It remains to be seen whether or not the animals which we have brought to a condition of nutritive disaster by a series of diets, each of which was faulty in respect to but a single dietary factor, as inorganic salts or protein, or the unknown A or B, will reveal to the histologist and pathologist lesions analogous to those seen in the states of human malnutrition presumably of dietary origin. The inadequate diets of man are probably never highly satisfactory with respect to all factors but one. Rather they are in some degree unsatisfactory with respect to two or three dietary factors simultaneously, and the probability that they will be so increases greatly as the food supply tends to be limited to the seeds of plants, and such manufactured products as are derived from the endosperm of seeds. The time will doubtless come within a few years when very specific advice can safely be given as to the best way to plan a variety of human dietaries, but a consciousness of the paucity of our knowledge concerning the peculiar supplementary relationships among our natural food- stuffs, forbids more than a very general statement as to the safest policy at the present time. A diet in which the seeds of plants form the principal part will be made more safe by the judicious use of milk and of the leaves and probably also of the tubers of plants. These have peculiar dietary properties which the best chemical talent of to- day fails to recognize, but which are readily demonstrable by biological tests. REFERENCES *The earlier experiments in this field are summarized by the author in The American Journal of Physiology, vol. 25, p. 120 (1909). 7W. Stepp: Biochemsches Zeitschrift, vol. 22, p. 452 (1909); Zeit. f. Biologie, vol. 57, p. 135. (1912); 59, p. 366 (1912-13). *T. B. Osborne and L. B Mendel: Publications of the Carnegie Institution of Washington, Bul. 156, Pt. 2 (1911). ‘E. B. Hart, E. V. McCollum, H Steenbock and G. C. Humphrey; Research Bul, 17, Wisconsin Experiment Station (1911). °C. Eijkmann: Archiv. pathol. Anat., vol. 148, p. 523 (1897); 149, p. 187; Archiv. f. Hygiene vol. 58, p. 150 (1906). *H. Fraser and A. T. Stanton: The Lancet, Mar. 12 (1910), p. 733; Studies from the Institute for Medical Research, Federated Malay States No. 12; The Etiology of Beri-beri (1911). 180 HARVEY SOCIETY 7C, Funk: The Lancet, November 4 (1911), vol. ii, p. 1266. ®U. Suzuki, T. Shimamura and S. Odake; Biochem. Zeit., vol. 43, p. 89 (1912). ° Hopkins, F. G.: The Journal of Physiology, vol. 44, p. 425 (1912). »E. V. McCollum and M. Davis: The Journal of Biological Chemistry, vol. 15, p. 167 (1913); vol. 23, p. 231 (1915). “4 T. B. Osborne and L. B. Mendel: Ibid, vol. 12, p. 81 (1912). 2. V. McCollum and M. Davis: Ibid, vol. 23, p. 231 (1915). %*E. V. McCollum, N. Simmonds and H, Steenbock: Proc. Amer. Soc. Biol. Chemists, December, 1916; Journ. Biol. Chem. March, 1916. *E. V. McCollum, N. Simmonds and W. Pitz: American Journ. of Physi- ology, vol. 41, pp. 333 and 361 (1916). * EK. V. McCollum, N. Simmonds and W. Pitz: Journal of Biol. Chem., vol. 28, p. 153 (1916). * E. B. Hart and E. V. McCollum: Ibid, vol. 19, p. 373 (1914); McCollum and Davis, ibid, vol. 21, p. 615 (1915); E. B. Hart, W. S. Miller and E. V. McCollum, ibid, vol. 25, p. 239 (1916). KE. V. McCollum, N. Simmonds, and W. Pitz: Unpublished data. *%E. V. McCollum, N. Simmonds, and W. Pitz: Journ. Biol. Chem. vol. 28, p. 483 (1917). * J. R. Slonaker: Leland Stanford Junior Univ. Pub., Univ. Series, 1912. » KE. V. McCollum and W. Pitz: Unpublished data. * Holst and Frohlich: Zeit. Hygiene Infektions Krankheiten, 72, p. 1 (1912). * A. F. Hess: Amer. Journ. Dis. of Children, vol. 8, p. 386 (1914); Journ. Amer. Med. Assn. vol. 65, p. 1003 (1915); Proc. Soe. Exp. Biol. and Med. vol. 13, p. 50 (1915); p. 145 (1916); Amer. Journ. Dis. of Children, vol. 12, p. 152 (1916). 2, Jackson and J. J. Moore: The Journal of Infectious Diseases, vol. 19, p- 478 (1916). THE INFLUENCE OF NON-SPECIFIC SUBSTANCES ON INFECTIONS JAS. W. JOBLING, M.D. From the Departments of Pathology and Experimental Medicine, Vanderbilt Medical School, Nashville, Tenn. HE therapeutic measures used in the treatment of infections may be divided into two general classes: first, those specific substances which are supposed to destroy the infecting organism or neutralize its toxins, and second, those which aim to strengthen the natural processes that normally bring about recovery. Of the specific substances which either destroy the infecting organism or render toxic products innocuous, we have specific anti-sera, and certain chemo-therapeutic substances. Among the most effective anti-sera may be mentioned diphtheria anti- toxin, tetanus antitoxin, anti-meningitis serum, and a serum for the treatment of one form of pneumonia. Many others have been prepared, but there is still some doubt as to their real value as therapeutic agents. Of the chemical substances which have a direct destructive action on the infecting organisms, quinin, salvarsan, and emetin are probably the most important. Bacterial vaccines, autogenous and stock, have been exten- sively used in the treatment of infections with the hope that they might cause an increased production of immune substances. Curative vaccine treatment rests primarily on the assumption that under certain conditions the production of specific antibodies can be increased by the injection of the bacteria causing the infection. Originally it was the belief that vaccines were indicated in those conditions in which the infecting organisms are localized and more or less encapsulated, and thus unable to stimulate the body to produce a sufficient number of immune bodies to bring about recovery. According to these views it will be seen that the use of vaccines as a curative measure was restricted to local- ized infections. 181 182 HARVEY SOCIETY It is a debatable question whether the practical results support the hypothesis on which specific vaccine therapy is based, and it will be difficult to explain by this hypothesis the results now being obtained in the treatment of acute systemic infections. In these conditions the infecting organisms are widely distributed through the body and should furnish the stimulus needed to eall forth an abundant supply of antibodies. When we compare the number of infectious conditions which a physician is called on to treat with the number of specific thera- peutic substances at his command, it will be seen that in the great majority of instances he is compelled to rely on the use of agents which merely serve to strengthen and to sustain his patient. Recovery in these cases depends almost entirely on the gradually increasing concentration or activation of the normal protective agencies of the body, and the final results depend upon the patient’s ability to resist the infection until this stage is reached. The work done in most of the experiments made to find cura- tive agents for the various infections has evidently been based on the assumption that a substance must be found which will act directly on the infecting organism, or that will cause a mobiliza- tion of those specific immune bodies which we are accustomed to look on as the means nature uses to bring about recovery. And yet one finds a considerable accumulation of evidence, both from the laboratory and from the clinic, which definitely indicates that certain non-specific and as yet ill-defined factors have a large share in bringing about recovery from disease. While the emphasis placed on certain phases of immunology and spe- cificity, more especially the fruitful antigen-antibody conception of Ehrlich, has been of inestimable value, I am inclined to believe that it has resulted in the neglect of certain perfectly obvious lines of approach to medical problems. SPECIFIC VACCINES That specific vaccines are effective in the treatment of acute general infections is shown by numerous reports. In 1893, Frankel? treated 57 typhoid fever patients by subeutaneous and NON-SPECIFIC SUBSTANCES 183 intramuscular injections of typhoid bouillon cultures sterilized at 62° C. with excellent results. Petruschky,? in 1902, and Von Peskarola and Quadrone,? in 1908, obtained similar results. Dur- ing the Balkan war of 1913, Petrovitch,* inoculated subcutane- ously 460 typhoid fever patients with typhoid vaccines, with a mortality of 2.9 per cent. Of 220 patients not treated in this manner, 12.8 per cent died. Biedl® states that Wiel treated 14 children suffering from typhoid fever, without a single death, and that Von Variot treated 69 cases by this method with excel- lent results. It must not be forgotten, however, that the mortality among children affected with typhoid fever is usually low. Goer ® treated successfully nine cases of the same disease with a soluble albuminous product obtained from typhoid bacilli, and Krumbhaar and Richardson’ observed favorable results in 77 eases treated subcutaneously with typhoid vaccines. These authors found in the literature records of 1800 cases of typhoid fever in which the patients were treated with subcutaneous injections of typhoid vaccines with favorable results. During the past two or three years the methods of using vaccines in the treatment of acute general infections have under- gone revolutionary changes. Ichikawa,® in particular, was in- strumental in bringing about this change. This author gave intravenous injections of sensitized typhoid bacilli to 87 patients with typhoid fever. Immediate recovery, so far as fever and general toxic symptoms are concerned, followed the intravenous injection of one to two doses of the vaccines. Slight hemorrhages were observed in a few instances. Biedl® treated 21 patients with typhoid fever by the intravenous injection of typhoid vaccines prepared by various methods. Of these 21 cases, 2 patients died as a result of uncontrolable hemorrhages from the nose. Of the 19 remaining, 17 made immediate recoveries so far as temperature and toxic symptoms were concerned, one had a recurrence with an associated broncho-pneumonia and died, and one patient had a slight recurrence which subsided immediately after a second dose. In discussing the explanation of these results he states that they may be due to the action of protein split products, aided by the mobilization of immune bodies, as 184 HARVEY SOCIETY was suggested by Ichikawa. Gay ® treated typhoid fever patients with intravenous injections of sensitized typhoid bacilli and states that the course of the disease was favorably influenced in 66 per cent of the cases. The above references show the results of the treatment of acute general infections by injections of specific vaccines. While the results may not have been entirely uniform in character, sufficient evidence is afforded to show that the older ideas con- cerning the principles on which vaccine therapy were based do not afford a satisfactory explanation. NON-SPECIFIC Matthes,’® early in the development of tuberculin therapy, demonstrated that to all intents and purposes the reaction con- sidered specific for tuberculin could be produced when deutero- albumose was used, and he showed that whatever difference did occur could be explained by the fact that the tuberculin fraction contained certain toxic peptones in addition. Matthes, later, went even further, expressing the idea that fever in general was pro- duced by protein split products, and he suggested the importance of proteolytic ferments in this connection, thus foreshadowing the work of Vaughn in this country and the later German workers in anaphylaxis. Frinkel,’ in 1893, was probably the first to demonstrate the value of typhoid vaccines in the treatment of typhoid fever, and his report was followed almost immediately by that of Rumpf,™ who obtained similar favorable results with a vaccine composed of the bacillus pyocyaneus. The medical profession, however, could see no merit in a non-specific method of treating infections, therefore the entire subject was dropped for a considerable period. In this country, too, the controversy occasioned by the Schafer vaccines is pertinent. Here was a biologic product, which, according to the observations of many competent observers, did at times produce striking results in a variety of infectious diseases; whether or not it was a safe remedial measure, or whether it had defects, is not pertinent in this connection. The very fact that it did certain things at times should have led to a NON-SPECIFIC SUBSTANCES 185 study of why such favorable changes were brought about. The fact that it was palpably non-specific, however, was sufficient to warrant the stamp of disapproval by the medical profession. Within the last three years, however, this view has been gradually changing, and it is significant to note that Wright,'* than whom, of course, no one man has stood out more emphatically for specific therapy, recently made the following statement in this connection: ‘‘All of those who have had much experience with vaccines will have seen cases where therapeutic effects, lying quite outside the range of the particular vaccine employed, and therefore, as we thought, not quite creditable to science, have been obtained with vaccine therapy.”’ Schmidt '* is among those who have observed that following vaccine therapy of any kind the body becomes resistant to a variety of commoner infections. He also called attention to the relatively low ‘‘infection index,’’ as he termed it, which is present in carcinoma and pregnancy, and refers to Rokytansky’s ideas on the antagonism between carcinoma and tuberculosis. Von Wagner " utilized a similar non-specific method when he obtained favorable results in patients with progressive paralysis treated with tuberculin. He had observed, as had others, that intercurrent infections frequently cause a remission of symptoms in this disease which sometimes lasts for a long period. This led him to experiment first with tuberculin alone, and later with a combined tuberculin-antiluetic treatment. He believed that he got better results with the combined treatment than with the antiluetic treatment alone. Pilez 1° observed similar good results with tubereulin, while Donath,!* noting the same influence of intercurrent infections in paretics, suggests the formation of abscesses in these patients by injecting turpentine into the sub- cutaneous tissues. It is not intended here to recommend the use of tuberculin in the treatment of paretics, because it hardly seems warranted, but to indicate the same general phenomenon of a therapeutic effect from a non-specific method. It is in the domain of coagulation disturbances that therapy of this nature has received particular attention in the past. The 186 HARVEY SOCIETY beneficial effects of subeutaneous serum injections in hemophila is well recognized, although here, too, the exact mechanism is unknown. Originally the slight leucocytosis was regarded as the potent factor, but for this there is no proof. Results have been obtained with homologous and heterologous serums, and with whole blood; but decisive results depend considerably on the dosage, and marked fluctuations in the coagulation time fre- quently occur following these injections. That such injections do not alter the disease process through supplying directly some deficiency, but rather through stimulating the elaboration of some substance before lacking, seems to be the conclusion reached by recent workers. This would agree with the stimulating effects observed by Esch, Busse and Weber following the injection of serum and whole blood subcutaneously in tuberculosis and in pernicious anemia. The dermatologists have had similar results within the past three years in a variety of diseases, including some cases of psoriasis which have proved refractory to other methods of treat- ment. Recently Engman and McGarry “ report favorable results in the treatment of lupus erythematosus with intravenous injec- tions of typhoid bacilli. The therapy of typhoid fever has so far been the chief avenue of approach to the problem. A number of vaccines have been elaborated and used subcutaneously during recent years for therapeutic purposes, and the results have been, in general, very encouraging. It was not, however, until such vaccines were used intravenously that the striking pictures of complete abortion of the disease were obtained. Ichikawa,’ who was among the first to use the intravenous method of administering vaccines in typhoid fever, observed that the results in patients suffering from paratyphoid fever who had been treated with typhoid vaccines were similar in every way to those noted in typhoid fever. Kraus‘® obtained similar results in the treatment of typhoid fever with the intravenous injections of vaccines composed of colon bacilli. He also treated eight patients with puerperal sepsis with excellent results. Kraus was so impressed with this form of treatment that he suggests its NON-SPECIFIC SUBSTANCES 187 use in scarlet fever, plague, septicemia, etc. In addition to vac- cines prepared with cholera bacilli, typhoid bacilli and paraty- phoid bacilli, Liidke '® also used deuteroalbumose. He treated 23 typhoid fever patients, and of these, 19 cases were favorably influenced. There was one death in this series. Reibmayr ”’ obtained similar results with colon and cholera vaccines. Hiss and Zinsser *! were among the first to make use of non- specific substances in the treatment of acute infections. They used extracts of rabbits’ leucocytes in the treatment of epidemic cerebrospinal meningitis, in pneumonia, and in staphylococcus infections, and believe that the course of these diseases was favorably influenced. Floyd and Lucas** used the leucocytic extracts in the treatment of 41 cases of pneumonia, and report that the mortality in their series was about half that obtained in the same institution with other forms of treatment. Miller and Lusk,** during the summer of 1916, treated a series of patients with typhoid fever with intravenous injections of typhoid bacilli and with secondary proteoses. In this series, 20 per cent terminated by crisis, and 20 per cent by rapid lysis. They also treated a series of chronic, sub-acute and acute cases of arthritis with these preparations with excellent results, as relief was afforded in the majority of cases. The authors do not state the number of patients treated by each method, but con- clude that both gave the same results. The same authors *4 recently reported a second series consisting of 85 patients with arthritis treated with typhoid vaccines. Forty-five of these cases were acute, four being gonorrheal. Twenty-nine of the acute cases recovered promptly after 1 to 4 injections, 8 showed marked improvement, 6 moderate improvement, and 2 were unimproved. Nine of these patients had recurrences. There were 12 sub- acute cases, 10 of which cleared up completely within 3 to 5 days, and 2 patients were greatly improved. There were two recurrences in this group, but the patients made complete recov- eries on further treatment. Nineteen chronic cases were treated, 10 of which showed definite improvement. Culver *° reports a series of gonorrheal complications, espe- cially arthritis, epididymitis and acute prostatitis, in which the 188 HARVEY SOCIETY patients were treated with a variety of vaccines composed of gonococci, meningococci, colon bacilli, and with secondary pro- teoses. ‘Twenty-eight of the 31 arthritic patients were either completely cured or manifested a decided improvement. The twelve patients with acute epididymitis presented complete free- dom from pain after the first injection. Ziembowski *° has recently reported on the results obtained in the treatment of 100 patients with intramuscular injections of 5 ¢.c. of boiled milk. He states that excellent results were obtained by this means in the treatment of septic war wounds, in erysipelas, in tuberculous bone and joint diseases, and in three cases of actinomycosis. Matthers, in a personal communication, states that he has used various types of vaccines and pure proteins in the treat- ment of typhoid fever, lobar pneumonia, scarlet fever and acute arthritis. He infers from his experiments that the therapeutic results obtained in erysipelas and lobar pneumonia do not justify the use of this plan of treatment. In the other diseases mentioned his results correspond favorably with those reported by other investigators. Manier, Petersen and I have treated 13 eases of arthritis, of which 3 were acute, 3 sub-acute and 7 chronic. Of the acute cases, two were definite cases of acute rheumatic fever. One of these patients, a child of 12 years, with multiple swollen tender joints, fever, etc., cleared up entirely within 24 hours after a single injection of secondary proteoses, while the other, an adult female with multiple joint involvement, temperature, ete., was entirely relieved after eight injections except for some residual pain in one shoulder. The third acute case, with a gonorrheal joint, received four injections with only temporary subjective relief and with but slight objective change in the joint. The three sub-acute cases received an average of five injections, with complete relief of all symptoms and a return of the involved joints to normal in every respect. In the seven eases of chronic arthritis there were varying degrees of disability, from slight stiffness in the milder cases up to complete ankylosis of the majority of the joints of the NON-SPECIFIC SUBSTANCES 189 body in the most severe cases. The injections given the patients varied in number from two to eighteen. The results were, com- plete relief in three, marked improvement in three and no notice- able change in one. Four patients with gonorrheal epididymitis with acute swell- ing and tenderness of the epididymis, urethral discharge, etc., were treated. In every case of this series there was immediate improvement following the first injection, the epididymis becom- ing smaller and distinctly less tender. In each patient there was complete relief, as evidenced by the return of the epididymis to its normal size and consistency, together with the disappearance of the urethral discharge and of pus from the urine. Two patients with erysipelas were treated. One was first seen on the second day of the disease, at which time the condition involved almost the entire face, and, in addition, the mucous membrane of the mouth and of the pharynx. The patient was markedly toxic, with a temperature ranging from 100° to 103° F., and a pulse rate of 140. He was given two doses of a secondary proteose solution on successive days, following which his tem- perature and pulse rate promptly settled to normal, and his local condition, both on the face and in the mouth, cleared up promptly. The second patient was also seen on the second day of the disease, and while he was not so seriously ill, the symptoms did not clear up until after four injections. The amount of secondary proteoses given by us depends upon the apparent toxicity of the patient. We usually begin with 0.25 ¢.c. of a 1 per cent solution and increase the dose according to the reaction obtained. When it is remembered that Liidke,’® Miller and Lusk,?? and Culver?> gave as high as 2 ec. of a 4 per cent solution, it will be seen that the amounts given by us are small. Subsequent experience, however, may demonstrate that the larger doses are advisable despite the more severe re- actions, and particularly in arthritic cases, where the reactions appear to be associated with little or no danger. Our purpose has been to obtain some increase in temperature with a slight chill, and this reaction we can usually obtain at the first injection with 0.25 ¢c.c. of a 1 per cent solution. Injections were given in 190 HARVEY SOCIETY most instances every day. As individuals vary in their reactions, we believe it better to begin with the smaller dose and thus establish the tolerance of each patient before pushing the treat- ment. In no instance were there any alarming symptoms. It is almost impossible to determine beforehand the degree of reaction which will follow these intravenous injections. Some authors believe that it depends on the severity of the disease, while others consider the concentration of immune bodies in the circulating blood more important. From a general survey of the work done, however, it appears that there is some other factor, as yet unknown, which plays an important part in determining the severity of the reaction. Apart from the action of the vaccines, the protein split products, and the effect of both homologous and heterologous serums, observations are recorded by Smithlen,?? Miiller and Weiss,”* and Saxl, Bruck and Kiralihyda ”® on the effect of intra- muscular injections of milk, by Mitlander * on salt solution, and by numerous observers on the effects of dextrose solution, col- loidal metals, distilled water, ete. It seems probable that the reaction and beneficial effects observed from these substances is based on a similar mechanism in all eases. REACTION Immediately following an injection by this method of treat- ment there is usually a reaction which is sometimes severe. As a rule, there is a chill from one-half to one hour following the injection, and this may last 15 to 45 minutes. With the chill there is an increase in temperature of from 1° to 4° F., followed several hours later by a progressive fall. Associated with the drop in temperature there is a general relaxation, profuse per- spiration and a rapid subjective and objective improvement. The pulse may or may not be increased in frequency. The blood pressure in our own eases was not altered. Some authors report that many of their patients had headache, nausea, ete., and again others state that the symptoms following the injections are not sufficient to cause serious discomfort. These differences may depend on the dosage used. NON-SPECIFIC SUBSTANCES 191 Following the injection the leucocytes are decreased in num- ber, at times as low as 2000 per cubic millimeter. ‘This leuco- penia is followed by a gradually developing leucocytosis which usually reaches its maximum in from five to seven hours. The leucocyte count has usually returned to normal within 24 to 30 hours. Immediately following the injection in acute infections, such as typhoid fever, there may be a permanent return to normal temperature—termination by crisis; the temperature and gen- eral conditions may improve more slowly—termination by lysis; or, all the symptoms may return and the disease progress as usual, uninfluenced in any manner, though usually it pursues a milder course. The temperature frequently drops to sub- normal and remains so for several days in those cases which ter- minate by crisis. COMPLICATIONS AND CONTRA-INDICATIONS The only serious results that have followed the use of this form of treatment occurred in typhoid fever patients, but as a large majority of the cases of acute general infections treated were. patients with typhoid fever, it is important to learn more concerning the danger attached to treating this class of patients with non-specific substances. Hemorrhage is the most serious complication reported in typhoid fever. Ichikawa,*® who used sensitized typhoid vaccines in 87 cases, observed hemorrhages in a few patients, one to three days after inoculation, but the frequency with which these were observed was less than that noted among the uninoculated. R. Schmidt*? advises against its use in patients who have already had hemorrhages, or who give histories of having been bleeders. According to this author, bronchial and pulmonary complications also contraindicate its use. Biedl*® treated 84 typhoid fever patients with typhoid vaccines, and two of these died from uncontrollable hemorrhages from the nose. He also believes that previous hemorrhages contraindicate this form of treatment. Typhoid fever appears to be the only disease in which hemor- 192 HARVEY SOCIETY rhages have been observed following this method of treatment. Kraus '* does not mention it as occurring in his cases of general sepsis; it was apparently not observed in any of the paratyphoid cases, and Miller and Lusk ** do not mention its occurrence in their series of more than one hundred arthritic cases. According to R. Schmidt,** protein substances injected in- travenously or subcutaneously tend to decrease the coagulation time, and it is well known that certain of the lower protein cleavage products also inhibit coagulation. On the other hand, the subeutaneous and intravenous injection of homologous and heterologous serums has been a favorite procedure for some time in the treatment of hemophilia. Ichikawa,® Miller and Lusk,”* and others, advise against this method of treatment in patients with organic heart disease, and Ichikawa warns against its use in pregnancy. In the opinion of Lusk and Miller its use is also contraindicated in hypertension. The severity of the reaction is an important factor in deter- mining whether this form of treatment should be used in any particular instance. When we consider that the strength of the vaccines of specific and non-specific nature which have been used in the treatment of typhoid fever varied from 100,000,000 to 4,000,000,000 bacteria to the ecubie centimeter, it will be seen that the danger to the patient is not so great as the severe reactions would indicate. And now, how are we to explain the action of these non- specific substances? Can it be explained according to our present ideas of nature’s method of bringing about recovery from infee- tion? I believe it is doubtful if our present theories on immunity will enable us to explain the action of these non-specific sub- stances. It might be well, however, to take up in greater detail some of the explanations which have been advanced. SELECTIVE STIMULATION It is now the general belief that the hematopoietic organs are the chief source of antibodies, and not the tissue cells in general. As a corollary of this idea concerning the source of NON-SPECIFIC SUBSTANCES 193 antibodies, it would be reasonable to suppose that any disturbance of the hematopoietic system might alter the antibody formation. In view of these facts, it is possible that the various agents may act as stimulants of the hematopoietic tissue, thus suddenly flooding the body with immune substances, and thereby over- coming the infection. According to Wright, vaccine injections were supposed to be followed by a negative phase, at least so far as the opsonic power was concerned. Contrary to this generally accepted view, Bull** has recently shown that this does not hold true following the intravenous injection of a typhoid vaccine in immunized rabbits. Bull noticed that the antibodies were not diminished; on the contrary, they were rapidly increased following the injection. If this is the mechan- ism involved, it is important to bear in mind that the stimulus itself is not a specific factor, but that the hematopoietic system has been attuned to respond to a non-specific stimulus with the production of a specific substance. - Various investigators have stated that the results obtained with these non-specific substances are due to the mobilization of antibodies. Thus Miller and Weiss** thought this was the explanation of the results which they obtained in treating the complications of gonorrhea with gonorrheal vaccines, but sero- logie tests failed to confirm this view. Ichikawa ® attributed his results to the mobilization of antibodies. Kraus** believed at first that the phenomenon was very similar to anaphylactic shock, but concludes that the lack of specificity contradicts this view. For the same reason he believes that the results obtained are not due to a mobilization of immune bodies. Liidke?® found that the agglutination value of the serum of typhoid patients treated with proteoses was not changed. Reibmayr,”° also, found no changes in the agglutinins following the injections of typhoid vaccines. In this respect the observations of Moreschi ** are interesting. Moreschi noted the persistent absence of agglutinins in leukemic patients who suffered from superimposed typhoid and para- typhoid infections. The patient with typhoid fever recovered, while the one with the paratyphoid infection died. Immune 13 194 HARVEY SOCIETY, bodies could not be demonstrated in either case. The recovery of the typhoid patient indicates that agglutinins may not be essential. Experiments which we conducted last spring caused us to believe that the results obtained in the use of non-specific sub- stances in the treatment of infections, were due to the mobiliza- tion of immune bodies. Dunklin, who was working with us, found a marked increase in antibodies following the intravenous injection of proteoses in immunized animals. A similar increase in agglutinins was found in two cases of typhoid fever in which the patients had been treated with the same preparation of pro- teoses. We have not had the opportunity to make further tests in typhoid fever patients, but repeated experiments made during the past few months have failed to show an increase of antibodies in immunized animals following similar intravenous injections. The arthritic cases do not afford opportunities for this character of investigations, therefore we have been limited in our work to animal experiments. The observations of others, however, appear to show that the mobilization of antibodies must play a minor role in recovery from infection following the use of non-specific substances. HYPERPYREXIA It is a common clinical experience that in some diseases, among them sub-acute joint diseases, neuralgia, diabetes, pernicious anemia, certain dermatoses, sarcoma, etc., distinct beneficial results follow at times on some intercurrent febrile condition. May it not be, then, that these non-specific substances influence the course of the disease by producing a high temperature? Rolly and Meltzer,** Liidke,** and other investigators, have reached the conclusion that high temperatures (from 40° to 42°C.) artificially produced, have a favorable influence on an established infection. Heated animals were distinctly more re- sistant to daily injections of small quantities of bacteria, but no difference was noted when single large doses were given. They also found that agglutinins and bacteriolytic substances are produced more abundantly in animals which are kept over- heated. NON-SPECIFIC SUBSTANCES 195 In his discussion of the influence of high temperatures on infection, Liidke *° suggests, first, the possibility of the infecting organism being killed by the heat, and second, that a more rapid and firmer combination of the antigen and immune bodies is caused at high temperatures. Culver ** describes an instance in which a patient suffering from both acute gonorrheal urethritis and malaria, made a com- plete recovery from the urethritis following chills and fever lasting four days. He also states that one rarely sees a gonor- rheal infection coexisting with fever-producing diseases like pneumonia, typhoid fever and malaria. It is well known that gonococei are particularly susceptible to high temperatures, and therefore this factor may be of importance in gonococceal infec- tions, but it is doubtful if similar importance can be ascribed to high temperatures in infections due to such organisms as strep- tocoeei, typhoid bacilli, ete. Inasmuch as a very sharp febrile reaction almost invariably follows the intravenous injection of specific and non-specific sub- stances, the importance of this phase of the subject cannot be overlooked. In one of our cases we observed a reaction tem- perature of 107° F. within thirty minutes after an intravenous proteose injection. This high temperature was unaccompanied by a change in the pulse rate of any moment, or other untoward symptoms. LEUCOCYTOSIS The importance of the leucocytic reaction has been empha- sized by various authors. Gay ** and his associates believe that recovery in typhoid fever following the intravenous injection of a modified typhoid vaccine is due to a specific leucocytosis. More recently, however, McWilliams **? has observed that this hyperleucocytosis is apparently not specific to the degree indi- eated by the work of Gay and Claypole, and that even normal rabbits respond with a marked leucocytosis to the intravenous injection of typhoid vaccine. This coincides with our experience. We must keep in mind, too, the fact that in typhoid fever, par- ticularly, the normal course of recovery is not marked by a leucocytosis. On this ground alone we might be justified in 196 HARVEY SOCIETY seeking the influence of some other factor or factors in the recovery which follows intravenous therapy. On the other hand, Liidke ® states that there was no leucocy- tosis in his series of typhoid fever patients who were treated with albumoses, and Reibmayr”° makes a similar statement for his series of patients treated with typhoid vaccines. Most of those who believe that the leucocytosis is an important factor in recovery think chiefly of phagocytosis; other possibilities, however, must be considered. Hiss and Zinsser * state that immunity is probably in a large degree cellular in character, not only in the sense of phagocytosis, but also in the neutralization of toxins. They conducted a series of experiments with this idea in mind, using the leucocytes of rabbits, and came to the conclusion that ‘‘ leu- cocytic extracts have a distinct modifying and curative action on infections.’’ They believe that the results are due to the neutralization of endotoxins. Opie ** found that leucocytes in- jected into the pleural cavities of dogs in which a tuberculous pleurisy had previously been produced, tended to inhibit the development of the process, as those animals which received the leucocytes lived longer than the controls. Recently, Bail *® has advanced evidence that supports this theory. He found that a strong anti-chlora serum would not neutralize the endotoxin obtained from cholera bacilli, but that the toxin was destroyed if it was first incubated with a fresh emulsion of leucocytes and the antiserum then added. In these experiments the leucocytes were removed by centrifuging before the serum was added. Con- trol experiments demonstrated that the leucocytic emulsion alone did not destroy the toxin. Heating the leucocyte emulsion 30 min- utes at 58° C. destroyed its antitoxic action. Bull and I *° showed that leucoprotease will destroy the toxic extracts of typhoid bacilli and meningococci, and it is not im- probable that a similar explanation will apply to the results obtained by Bail. It will be seen, then, that in considering the part leucocytosis may play in the recovery from disease, we must consider other factors in addition to phagocytosis. Ped Sn <= NON-SPECIFIC SUBSTANCES 197 MOBILIZATION OF FERMENTS Petersen and I have already pointed out that in experimental animals intravenous injection of bacteria, *! kaolin,** protein split products ** and trypsin ** is almost invariably followed by more or less marked mobilization of serum protease and usually of esterase. Similar reactions occur in patients following the intravenous injection of vaccines and proteoses, but not to the same degree nor with the same regularity as in animals, In considering the possible effects of such a mobilization of ferments, both pro- tease and esterase, we must keep in mind the fact that a variety of reactions may occur. The serum protease, as other tryptic ferments, is without effect on bacteria ;*° but if we consider the source of the intoxication which occurs in the diseased organism as primarily due to protein split products derived from the bac- teria, then such a mobilization of protease may be of consider- able importance in the process of detoxication, as the toxic frag- ments are hydrolized to lower and non-toxic forms. Petersen, Eggstein and I** have discussed this possibility in detail in its relation to pneumonia. As a result of this action of the proteolytic ferments the dis- eased organisms would rid itself for the time being of the toxic substances in the circulating blood, although the disease process itself, and the infecting organisms, would possibly continue in existence, causing further injury. This would seem to be the explanation of the clinical picture obtained after intravenous therapy in those instances in which only an incomplete or transi- tory effect results. In the majority of these cases the patient presents a considerable improvement, both subjectively and ob- jectively, on the day or days following the injection, despite the fact that the temperature may recur or even continue uninter- ruptedly. This hypothesis will not, however, explain the con- tinued well being of the patients treated in the early stages of diseases such as typhoid fever. It is difficult to understand this unless it is due to cellular resistance acquired as a result of the injections. The influence of the mobilized ester splitting ferments is as 198 HARVEY SOCIETY yet obscure. In the ultimate analysis, of course, we must turn to ferments of this nature in order to dispose of the invading organism, because it is more than probable that the actual surface of the bacterium consists largely of lipoids, or intimate lipo- protein combinations, for the destruction of which esterase split- ting ferments are probably essential. Citron and Reicher,*’ Peritz,** and others, state that the serum esterase is increased in patients who suffer from infections due to lipoid rich organisms such as tuberele bacilli and lepra bacilli. They believe that a high esterase titer in these diseases indicates increased resistance on the part of the host. ANTIFERMENTS When discussing the increased resistance to infections said to be present in some conditions, among them carcinoma, preg- nancy, and as a result of vaccine therapy, I called attention to the fact that during such states a relatively high antiferment index is a well recognized accompaniment. The question arises whether this is a mere coincidence, or whether there is some casual con- nection between the antiferment increase and the increased re- sistance to infection. Petersen and I *® have previously shown that the antiferment power of the serum depends on the amount of unsaturated lipoids present in a highly dispersed state in the serum. Consequently, any factor which will tend to increase these lipoids, either by increasing the supply or by decreasing the utilization, will increase the antiferment titer, while conversely, any influence decreasing these poids, their dispersion or their unsaturation, will tend to decrease the titer. Wright °° worked with saprophytic and serosaprophytie or- ganisms, and noted that the addition of antiferment to the culture medium completely checked the growth of the serosaprophytie bacteria. He also found that even the saprophytic bacteria grew less luxuriantly when no proteolytic ferment was added. The direct influence of the antiferment cannot be as simple as Wright would assume, as Rettger, Berman and Sturges *! showed that the ordinary pathogenic organisms do not derive their protein NON-SPECIFIC SUBSTANCES 199 requirements from native or even partly hydrolized proteins, but solely from the lowest split products. The antiferment inhibits only the action of tryptic and not the peptolytic and ereptic ferments. Of course, when we are dealing with a definite trypto- genic organism it becomes apparent that an increase in anti- ferment would offer an increased factor of resistance against its growth. The immediate effect of the vaccine and proteose injections is not an increase, but a distinct decrease, in the antiferment titer for a short period of time, followed later by a rise. The exact cause of these changes in the antiferment index remains unde- termined. PHYSICAL CHANGES IN SERUM We have recently found that there are distinct changes in the viscosity of the serum of animals undergoing immunization, and that similar changes occur in anaphylactic shock. This alteration may be of more than theoretical interest in our interpretation of the results obtained by the intravenous injections of non- specific substances. Weil *? and others have recently shown that antigen and antibody may coexist in the blood of the living animal, while Joachimoglu °* has demonstrated that in anaphylactic shock the precipitins immediately disappear. The disappearance of pre- cipitins, which no doubt is due to their combination with the antigen, would probably bring about conditions favorable to protease actvity, as Bulger ** and others have shown that proteo- lysis in serums is active following the formation of precipitates. It is well to bear in mind the possibility that the changes in viscosity following the intravenous injection of non-specific sub- stances may cause a combination of the antigen and antibody which is already present, and thus duplicate the conditions pro- duced when additional antigen is introduced. Similar changes may occur in the serum of patients with acute general infections when they are treated with non-specific substances. Under these conditions it is not difficult to under- stand how areas of lowered antiferment content would be obtained 200 HARVEY SOCIETY in which proteolysis would occur. In such instances we may have a temporary increase in intoxication owing to the hydrolysis of the native bacterial proteins to toxic substances, and then a detoxication due to their further hydrolysis. This may be the explanation of the chill which is followed by a drop in tempera- ture, that occurs after intravenous injections of various sub- stances. These changes in viscosity may also help to explain the fall in antiferment strength which follows the injections. As stated before, however, these serum changes are more or less temporary in character, and will therefore not explain the permanent recovery of patients treated in the early stages of such diseases as typhoid fever. CONCLUSIONS According to our present views the symptoms of an infection are the result of the struggle between the infecting bacteria and their toxins, and the protective agencies of the host. Theo- retically, then, the results observed might be due either to the destruction of the infectious agent with its products, or to the fact that the cells of the host become resistant to the action of these agents, in either case, from our point of view, the disease ceases to exist. Theobald Smith,°* in 1910, said: ‘*The effective- ness of vaccines applied in the course of acute febrile diseases, such as typhoid and pneumonia, must be accounted for by prin- ciples of which experimental medicine has as yet no definite knowledge,’’ and this view apparently holds true at present. That the intravenous injection of non-specific substances exerts a marked influence on those infections in which it has been tried is very evident. It is not believed, of course, that these newer methods will cure all cases of infections. They do, however, open up new possibilities and suggest new methods for attacking in- fections of unknown etiology, as also those caused by organisms for which we have no specific antiserums. That all cases are not benefited does not necessarily reflect on the value of the treat- ment—there are very few theraputic measures which do not have the same objection. NON-SPECIFIC SUBSTANCES 201 BIBLIOGRAPHY 1¥Frinkel, E.: Deutsch. med. Wehnschr., 1893, xix, 985. ?Petruschky, J.: Deutsch. med. Wehnschr., 1902, xxviii, 212. 3 Pescarolo, B. and Quadrone, C.: Zeit. f. Inner. med., 1908, xxix, 989. *Petrovitch, —: Deutsch. med. Wehnschr., 1913, xxxix, 1576. 5 Biedl, A.: Prag. med. Wochen., 1915, xl, 53. ® Goer, F.: Miinch. med. Wehnschr., 1915, xlii, 1312. 7Krumbhaar, E. and Richardson, R.: Amr. Jour. Med. Sc., 1915, exlix, 406. 8Tchikawa, S.: Zeit. f. Immunititsforsch., 1915, xxiii, 32. ® Gay, F. P.: Arch. Int. Med., 1916, xvii, 301. 10 Matthes, M.: Deutsch. Arch. f. klin. Med., 1894-95, liv, 39. 41 Rumpf, T.: Deutsch. med. Wchnschr., 1893, xix, 987. 2 Wright, A.: British Med. Jour., 1915, No. 2832, p. 625, vol. i. 18 Schmidt, R.: Med. Klinik, 1910, vi, 1690. “Von. Wagner, J.: Wien. med. Wochen., 1909, lix, 2124. 1% Pilez, A.: Wien. med. Wochen., 1907, lvii, 1462. % Donath, J.: Wien. med. Wochen., 1916, Ixvii, 1741. Engman, M. and McGarry, R.: Jour. Amer. Med. Assoc., 1916, lxvii, 1741. 1 Kraus, R.: Wien. klin. Wochen., 1915, xxxviii, 29. 1 Liidke, H.: Miinch. med. Wchenschr., 1915, Ixii, 321. * Reibmayr, H.: Miinch. med. Wcehenschr., 1915, lxii, 610. 2 Hiss and Zinsser, H.: Jour. Med. Research, 1908, xix, 321. Floyd, C. and Lucas, W.: Jour. Med. Research, 1909, xxi, 223. * Miller, J. and Lusk, F.: Jour. Am. Med. Assoc., 1916, xvi, 1756. * Lusk, F. and Miller, J.: Jour, Am. Med. Assoc., 1916, Ixvii, 2010. * Culver, H. B.: Chicago Institute of Med., Nov. 7, 1916, unpublished report. » Ziembowski, S. V.: Med. Klinik, 1916, xii, 1174. 7 Smithlen, F.: Wien. klin. Wehnschr., 1916, xxix, 53. * Miiller, R. and Weiss, A.: Wien. klin. Wehnschr., 1916, xxix, 249. * Saxl, Bruck and Kiralihyda: Miinch. med. Wehnschr., 1916, Ixiii, 511. 8° Mitlander: Budapest letter, Jour. Am. Med. Assoc., 1916, Ixvi, 1321. * Schmidt, R.: Med. Klinik, 1916, xii, 171. ™ Bull, C.: Jour. Exp. Med., 1916, xxiii, 419. *5 Moreschi, C.: Zeit. f. Immunitiitsforsch., 1914, xxi, 410. “Fr. Rolly and Meltzer: Deutsch. Arch. f. klin. Med., 1908, xciv, 335. % Liidke, H.: Ergeh. d. Inner. Med. u. Kinderheilk, 1909, iv, 493. * Gay, F., and Claypole, E.: Arch. Int. Med., 1914, xiv, 662. * MeWilliams, H.: Jour. Immunology, 1916, i, 259. * Opie, E.: Jour. Exp. Med., 1908, x, 419. *® Bail, O.: Zeit. f. Immunitiitsforsch., 1916, xxv, 248. * Jobling, J. W., and Bull, C.: Jour. Exp. Med., 1913, xvii, 453. “ Jobling, J. W., and Petersen, W.: Jour. Exp. Med., 1915, xxii, 590. “ Jobling, J. W., and Petersen, W.: Jour. Exp. Med., 1915, xxii, 597. 202 HARVEY SOCIETY “ Jobling, J. W., and Petersen, W.: Jour. Exp. Med., 1915, xxii, 603. “ Jobling, J. W., and Petersen, W.: Jour. Exp. Med., 1915, xxii, 141. * Jobling, J. W., and Petersen, W.: Jour. Exp. Med., 1914, xvi, 321. “ Jobling, J. W., and Petersen, and Eggstein, A.: Jour. Exp. Med., 1915, xxii, 568. 47 Citron, J., and Reicher, K.: Berl. klin. Wehnschr., 1908, xlv, 1398. “ Peritz, G.: Deutsch. med. Wehnschr., 1910, xxxvi, 483. * Jobling, J. W., and Petersen, W.: Jour. Exp. Med, 1914, xix, 459. © Wright, A.: British Med. Jour., 1915, No. 2832, p. 625, vol. i. * Rettgers, L., Berman, N., and Sturges, W.: Jour. Bacteriology, 1916, i, 15. ‘2 Weil, R.: Jour. Immunology, 1916, i, 47. 8 Joachimoglu, G.: Zeit. f. Immunititsforschung, Orig., 1911, viii, 453. “Bulger, H.: Jour. Infect. Diseases, 1916, xix, 882. % Smith, Theobald: Boston Med. and Surg. Jour., 1910, elxiii, 275. METABOLISM OF MOTHER AND OFF- SPRING BEFORE AND AFTER PARTURITION* JOHN R. MURLIN Cornell University Medical College, New York City N presenting to the Harvey Society the results of several | years’ work on various phases of the physiology of reproduc- tion, I wish first of all to make some acknowledgments. As I proceed you will observe that the researches in which I have par- ticipated represent but a small portion of the total body of infor- mation which I shall endeavor to interpret and, in making these acknowledgments, | hope I shall not seem to magnify the impor- tance of my own work. I experience peculiar pleasure in making use of this public opportuntiy to acknowledge my indebtedness to my chief, Professor Graham Lusk, whose interest and encourage- ment have been never-failing. My thanks are also due to Dr. F. G. Benedict for the courtesies of his laboratory in doing what was, I believe, the most important piece of work in which I have participated; to Dr. J. Cliffton Edgar, head of the mater- nity service of the Cornell Division at Bellevue Hospital, and to Dr. L. E. LeFetra, head of the children’s service there, for the facilities of their wards; and finally to Dr. T. M. Carpenter, Dr. H. C. Bailey, and Dr. B. R. Hoobler, who have been asso- ciated with me in authorship. These last named gentlemen have borne their full share of the work and, I trust, have had their share, also, of the pleasures resulting from the pursuit and acquisition of scientific data. Because of the limitations of time imposed upon me, I shall be obliged to limit my remarks to three phases or chapters of the general subject: (1) the nutritive relations of mother and fetus; (2) the substance metabolism of the mother as modified by the presence of the fetus; and (3) the energy metabolism of mother and child both before and after parturition. * Delivered February 24, 1917. 203 204 HARVEY SOCIETY I. NUTRITIVE RELATIONS OF MOTHER AND FETUS Any adequate comprehension of the metabolic relationship between the mammalian mother and her offspring presupposes a broad view of the whole subject of reproduction. As in so many other departments of the physiology of man, interpretation of vital activities is constantly aided by reference to corre- sponding phenomena in lower organisms. I suppose we shall never understand fully what metabolism, nutrition, respiration, reproduction are—what they are in essence—as applied to our own tissues and bodies, until we understand their significance for the lower orders of life. We must have the viewpoint of the general physiologist and ofttimes of the naturalist. Now the viewpoint of the naturalist regarding reproduction, since Weissmann’s great work,’ has been this: The germ cells or reproductive elements are not, strictly speaking, produced by the adult body; the adult body is produced and reproduced by the germ cells as a medium in which the specific stock can be perpetuated. Seen from this angle, for those forms in which the individual counts for little, reproduction becomes the whole end and aim of life. As Hatschek ? frames the idea, ‘‘ Fortpflanzung ist das Endziel der Labenstitigkeit.’’ Continuing the thought, Hatschek says, ‘‘ All the cells of the body stand at the service of the germ cells because in them is perpetuated their own being.’’ The student of reproduction in man only may easily lose sight of these broad, fundamental principles. It was for this reason, I think, that until a few years ago the mammalian ovum was regarded as nothing more than a maternal cell, just as much at the mercy of the maternal circulation as any other cell of the body; and the complicated provisions made after fertilization for insuring its supply of maternal blood were looked upon as beneficent, not to say providential, adaptations for the special eare of the offspring. Partly through the cytological observations of Boveri,? Hicker, Hegner,® and many others, who have demon- strated the independence of the AKeimbahn, or germinal path, from one generation to the next in various lower forms of life (worms, crustacea, insects, ete.), and partly through the embryo- logical studies of Hubrecht,® von Spee,’ Peters,* Hitschmann and METABOLISM 205 Lindenthal,? Bryce and Teacher,’® Herzog, and Johnstone,’? who have directed attention to the details of the process of implantation of the ovum in the wall of the uterus, it has gradually dawned upon us that the ovum never is a true body cell but is, to a large degree, an independent organism, capable even in the face of difficulties of looking out for its own nourish- ment at every step of its development. In fact, the only possible interpretation of events within the ovary previous to the libera- tion of the ovum is that in each follicle a certain cell is selected to grow and thrive at the expense of its fellows because (accord- ing to Miss Lane-Claypon,!* quite fortuitously; according to John Beard,'* by prearrangement) that particular cell possesses the complement of enzymes which enable it to appropriate the materials supplied by its less fortunate neighbors to its own purposes. It is necessary to get this point of view—that events are from the start under control of the new organism rather than the old. Upon fertilization the processes of assimilation in the ovum receive a fresh impetus, the proteolytic and proteo-synthetic changes receive from the sperm cell an activating effect, some- thing like the effect of a kinase upon a proenzyme, and the result is further growth at the expense of whatever materials the ovum comes in contact with. The follicle cells, which cling to the ovum when it is set free, and which, according to one view, prevent the ovum from becoming attached to the wall of the Fallopian tube, are digested away and, in some mammals at least, the tube sup- plies a nutritive secretion which is in all essential respects the analogue of the white of the hen’s egg.142. Arriving at the uterus, the only reason why a fertilized ovum rather than an unfer- tilized one becomes attached to the wall, is that its cells (for there are many by this time) are hungry and they possess the means of satisfying their hunger. From the histological studies made on an age series of hedge- hog embryos by Hubrecht*® and studies of the earliest human embryos by Bryce and Teacher,?° Herzog,’ and recently by Johnstone,!? it is evident that the process of implantation from the standpoint of the embryo is simply a continuation of the 206 HARVEY SOCIETY proteolysis by which it lays claim to its nutriment. The ecto- dermal cells which mediate this function are known collectively after Hubrecht as the trophoblast, or, a more exact term etymo- logically suggested by Minot, the trophoderm. Although an enzyme has not been demonstrated chemically in these earliest stages, it has been found by Grafenburg as early as the second month in the human embryo and there can be no doubt, from the histological appearances of the earliest stages, that a very active one is at work or that it is produced by these trophodermic cells. Whatever they touch, according to von Spee, undergoes solution. Hence, wherever the ovum happens to come in con- tact with the uterine mucosa after the fringe of follicle cells has been digested and absorbed, it there adheres and soon dissolves a depression; the depression becomes a cavity and the cavity extends as the trophodermie cells increase in number. From the standpoint of the maternal organism, placentation, according to the interpretation first given by Sir Wiliam Turner and con- firmed by most recent students*®> of the problem, represents a reaction designed to protect against the invader or, in modern phrase, to restrict the action of its enzymes. The large, special- ized decidual cells are almost certainly active in the chemical defences of the maternal tissue. By the time a circulation is needed to distribute the products of proteolysis to all the embryonic cells, a circulation is neces- sary also to connect the embryo with its advance lines of attack and we have the first steps in the formation of the true placenta. Stages described by Hitschmann and Lindenthal and by John- stone, somewhat older than the Bryce and Teacher ovum, show clearly how the trophoderm of the primary villi become trans- formed into the double layer (syncytium and Langhans layer) of the definitive villi and how the trophoderm is responsible for the erosion and rupture of the maternal veins, thus establishing the intervillous circulation. ‘‘It is only this trophoblast,’’ say Hitsechmann and Lindenthal, ‘‘which is able to open up the ves- sels. The double layered villi no longer have this power; they serve mainly to extend the absorption surfaces.’’ Hitschmann, in a later article, states that by the end of the third month in METABOLISM 207 human development ‘‘the villi have no further power of invasion of the blood vessels.’’ I must pause here to point out that menstruation is caused by an enzyme of very similar nature produced by the ovary just before the ovum is set free and acting upon these same blood vessels (Young ?°). When the placenta is finished we have the following rela- tionships: Maternal blood is separated from fetal blood by the two-layered trophoderm and by certain mesoblastic structures, making an arrangement practically identical with that of the wall of the intestinal villus. The question now arises, has in fact often arisen since the time of Harvey, whether the placenta acts, as Harvey expresses it, ‘‘by a sort of digestion,’’ or in a purely mechanical manner. Time will not permit a full discussion of this question but I will cite some of the latest evidence and the considerations which must be taken into account in making a decision whether the fetus for a time surrenders control of its nutrition to the mother. Truly diffusible substances, like glucose and urea, readily pass the placenta. Cohnstein and Zuntz ‘‘ in 1884 showed that a hyperglycemia produced in the mother was followed by an in- crease in the sugar of the fetal blood. This has been placed on surer ground by recent methods in the work of Morriss of New Haven reported a few weeks ago in this Academy. Dr. Morriss *® finds that the sugar in the fetal blood at the moment of birth is higher when the percentage in the maternal blood has been raised either by prolonged labor or by the use of anesthetics in delivery. Otherwise; 2.e., after easy labor unaccompanied by the use of anesthesia, the percentages on the two sides of the placenta are the same. It is probable, therefore, that glucose passes as readily through the placental barrier as it passes the intestinal wall. (We shall see later that these facts have their significance for the metabolism of the new-born). Studies on the glycogen of the placenta indicate, however, that there is some regulation of the amount of carbohydrate which is permitted to enter the fetal circulation. Both Chipman’? and Lochhead and Cramer ”° have shown that up to the eighteenth day of gestation in the rabbit 208 HARVEY SOCIETY (the entire period being twenty-eight days) glycogen is not found to any extent at all in the embryonic tissues, not even in the liver, but is found in abundance in the maternal side of the pla- centa. Goldmann *' has shown the same to be true of the mouse but Driessen finds it not so strictly true of the human.** Con- trary to the earlier teachings of Cl. Bernard and Pifliiger, it appears that embryonic cells can not store glycogen ** until they have reached a certain age. Glucose, arriving at the maternal placenta faster than it can be utilized by the fetus, is stored on the maternal side as glycogen until the last one-quarter of gesta- tion when the fetal liver begins to assume its carbohydrate func- tion, whereupon glycogen disappears gradually from the maternal placenta. Now it is scarcely open to doubt that the decidual cells perform this glycogenic function in obedience to some influence upon them by the fetus itself, for otherwise glycogen is not found in this situation. Whether fat can be drawn from the mother’s blood by the fetus raises some very interesting questions. Esterases ** have been found in the placenta but no true fat cleavage has ever been proved to take place there. Ahlfeld,*® Thiemich,*® and Oshima,** all have failed to influence experimentally the percentage of fat in the fetal blood by feeding fat to the mother while S. H. and S. P. Gage,*® likewise Mendel and Daniels,** failed to find stained fat in the embryo after feeding pregnant mothers with such materials. Dr. Slemons, in a personal communication, informs me that he has been studying the lipoids in maternal and fetal bloods for a year and has reached the conclusion that neither neutral fat nor cholesterol esters pass the placenta at all but that cholesterol does. Now cholesterol is a colloid in the blood and proof that it traverses the placental barrier is tantamount to proof that a selective activity is going on.*° Bailey and I ** have found that, while the percentage of total fat in the fetal blood is much lower than that of the maternal, there seems nevertheless to be some relation between the two and we are not sure but that there may prove to be some relation again to the severity of labor. This would not be surprising, in view of the fact that muscular work raises the fat in the blood, just as it does the sugar. METABOLISM 209 TABLE 1.—ToTaL Fat IN FETAL AND MATERNAL BLoops at MOMENT OF BrrtH (Murlin and Bailey, Unpublished). Arm Vein Umbilical Vein WO aBO Mle cate’ sian yar nie ate aucie ls 0.60 per cent 0.22 per cent Waele 2h eierayere) siellerelsile oracate aie 0.87 per cent 0.49 per cent MORRO MSPs Loesch ola aha eure lola aie 0.72 per cent 0.48 per cent [OTE Re ee cea oer Reyes 0.44 per cent 0.26 per cent CAS eR Mere lenne tay Meckene eu sans sti 0.24 per cent CEST Sw 8) are eres Cela Cheseae ea pica 0.54 per cent 0.24 per cent CASE Wiel nvenarel nel isle savas pate 0.49 per cent 0.18 per cent MO ABE AS sbtiatereallny's (o's! aisle eter ats 0.40 per cent 0.21 per cent Casey Oem rlortiichisaieepaeiel 0.36 per cent 0.14 per cent There is scarcely any question about the presence of fat in the walls of the chorionic villi. This has been figured as stained by osmic acid and other reagents by many observers, especially Hofbauer *? and Goldman. Hofbauer’s figure shows fat in the deep layers of the syncytium in Langhan’s layerand in the paren- chyma of the villus, and since this is much the same histological picture as one gets in the absorption of fat by intestinal villi, where it is certain that fat is first split into its components and then resynthetized in the epithelial cells, Hofbauer interprets the picture as indicating the same chemical processes for the chorionic villi. The presence of fat in the villi, however, is not proof that it came through from the maternal blood. It may have been formed from carbohydrate or protein just where it is found. The fact that fat is found in abundance in the fetus at birth is amenable to the same interpretation; namely, that the fetal tissues have the capacity to form fat out of carbohydrate or protein. We have as yet, I should say, no positive proof that fat either is or is not taken up by the placenta. We can say with certainty only that it does not pass readily through the villi if it passes at all and it is never present in the same concentration in the fetal blood as in maternal. But this is not inconsistent with the vitalistic theory. How proteins pass the placenta is the most important ques- tion of all. Grafenburg ** was not able to find evidence of proteo- lytic enzymes in the human placenta after the fourth month. Nevertheless, products of proteolysis, especially albumoses, have been found by Mathes,** Basso,*° Hofbauer and others, both in 14 210 HARVEY SOCIETY fresh placente and in minced placenta, after undergoing autoly- sis for a short time, at all stages of gestation. Hofbauer, although a firm believer in the vitalistic theory, was obliged to conclude that digestion stopped at the albumose stage and that synthesis takes place immediately the trophodermie cells are passed. That amino-acids can readily find their way through the placenta has now been definitely proved and Dr. Morse ** in Slemons’s labora- tory has been studying the monamino-acid content of maternal and fetal bloods taken simultaneously at the moment of birth. He finds them nearly identical. Mr. Bock, in Dr. Benedict’s TABLE 2.—MONAMINO-ACID-NITROGEN IN FETAL AND MATERNAL BLOODS AT MoMENT OF BrrtH (Bock) Arm Vein Umbilical Vein (GERO I Soaegacos osc 6.6 mgm. per 100 ce. 12.15 mgm. per 100 ce. CaSey2 erste ie: soiree 8.9 mgm. per 100 ce. 14.92 mgm. per 100 ce. (CHD. cdlccddiccongc 8.93 mgm. per 100 ce. 11.80 mgm. per 100 ce. Casera ier meray revterers 7.44 mgm. per 100 ce. 9.57 mgm. per 100 ce. laboratory at Cornell, at my request has made a number of exam- inations of the two bloods taken in the same manner and has found the percentage in fetal blood very distinctly higher than that of the maternal blood. ‘These data are not yet complete enough for definite conclusions. Should it develop that the amino-acid content of the fetal blood is constantly, even though slightly, higher than the maternal, the conclusion would inevitably be (1) that these bodies are on their way out from fetus to mother or (2) are being produced in the placenta for use in the fetus or (3) again, that a selective activity is at work. Either of the last two interpretations would favor the idea that protein, the indispensable building material, continues to be modified by the placenta. The transport of iron from mother to fetus can not be ac- counted for on mechanistic grounds. The mamalian ovum being practically devoid of yolk, contains no iron for the manufacture of respiratory pigment. There are two possible sources of this: The hemoglobin of the maternal blood, and certain conjugated proteins of the food. Concerning the transformations of the METABOLISM 211 latter as a source of iron for the fetus we know next to nothing. Many observations on the disintegration of red blood corpuscles by the trophodermic cells, however, make it certain that the main reliance of the fetus for this essential element is the mater- nal blood. Hemolysis produced by placental extract, reported by Veit and Scholten ;°7 the eosin reaction of hemoglobin at the free border of the syncytium, seen by Bonnet; and the demon- stration of loose organic compounds of iron in the deeper layers of the villi, by Chipman, Hofbauer ** and Goldmann; may be regarded as pointing the way to a solution of this matter. In those animals which produce a ‘‘uterine milk’’ phagocytosis *° on the part of the trophodermie cells probably accounts for the transfer of iron. One is obliged to conelude from the weight of evidence at present that there is much more to be said for the vitalistic con- ception of the placental function than for the mechanistic. The very existence of the placental barrier, the fact that the two bloods can not intermingle would seem to imply the necessity of a process of naturalization at the border line of all materials which are used in the construction of the fetal tissues. Materials used only as a source of energy need not be so modified. Under the mysterious guidance of the mechanism of heredity the proteins are built up into a new being which reproduces the essential characteristics of the phylum class, order, genus and species to which the germ cells, belong. Heape’s*° famous experiment in which the fertilized ova of one variety of rabbit were transferred to the uterus of a different variety and were born without showing any effect of the foster mother, and experiments by Castle +4 in which the same independence of the germ cells was demonstrated by transplantation of the ovaries of a black guinea-pig into a white female, leave no doubt upon this point— a human child is born human not so much because it is nourished by a human being as because the germ cells from which it came are human. Nobody knows yet how closely related germ cells and fostering mother must be in order that development may proceed. At all events, it is clear that the same necessity for enzymes to harmonize the building materials of the fetus to its own type O12 HARVEY SOCIETY exists at every stage of development before birth, as after. The early enzymes with which the embryo starts out are just as much a part of the mechanism of heredity as are the enzymes of germinating seed. Indeed, one is tempted to assert that the mechanism of heredity is itself mainly an orderly succession of enzymes. Loeb,*? Robertson,*? Loeb and Chamberlain,** Riddle,*® Goldschmidt, *® and others have adduced evidence that the determiners of heredity behave like enzymes and Reichert,*’ from his study of homologous proteins and starches in different species and genera of animals and plants, has formulated a con- ception of the germ-plasm as ‘‘a complex physicochemic system of which an enzyme that starts the serial changes’’ and others that keep them going progressively are integral parts. A very significant observation by Abderhalden,*® made just before the beginning of the war, has a direct bearing upon this question. Attempting to prove the synthetic action of ferments in the con- struction of proteins, Abderhalden made hundreds of different combinations of amino-acids and tissue extracts, but with no marked success until he tried the following: Digesting the several kinds of tissue, kidney, liver, thyroid gland, lung, ete., with pepsin, trypsin and erepsin until the digests were biuret-free, he added to the mixtures of amino-acids thus obtained a macera- tion juice extracted from the same tissues. Under aseptic pre- cautions these mixtures were allowed to stand for five months at room temperature. At the end of this time there was clear evi- dence of synthesis but only in those tubes which contained amino- acid mixture and extract (enzyme) from the same organ. Kidney amino-acids were built up by kidney enzyme but not by thyroid; thyroid amino-acids by thyroid enzyme but not by kidney enzyme, ete. If this observation is confirmed we shall be obliged to infer that the repair in adult life and the development in embry- onic life of each tissue protein is under the control of a specific enzyme, acting upon a specific substrate. It will be in order, then, to attempt to trace the different enzymes back step by step to the germ cells with the hope there to identify them with the chromosomes which are known on morphological grounds to con- tain the determiners of heredity. A substrate common to all tis- METABOLISM 213 sues after the earliest stages could be found only in the blood proteins. Seen with the eyes of the general physiologist, the nutritive relations of mother and fetus, then, find their explana- tion in the specificity of the proteins and the specificity of enzymes which lie at the basis of heredity—the reproduction of kind. Il. SUBSTANCE METABOLISM OF PREGNANCY Stated in terms of the different combinations of protein build- ing stones, or ‘‘stereoisomers,’’ necessary to set up a new human organism, complete in all anatomical details, the requirements for fetal growth are enormous. Can the mother supply all of the building materials from her food, or must she perforce supply some structural elements, chemically speaking, from her own body? In some lower orders of animals the young, before being hatched or setting out upon an independent existence, consume the maternal body, the individual thus being sacrificed for the good of the species. Is gestation in the mammals in this sense to any degree ‘‘a sacrifice of the individual for the good of the species ?’’ Stated in quantitative terms, the substance requirements of the fetus are not large. According to Sommerfield,*® a normal infant at birth weighing 4.340 grams contained just short of 100 grams of nitrogen, not more than would be contained in its mother’s diet for ten days at most. Michel °° has shown also the composition of the fetus at different stages of development for nitrogen, phosphorus, calcium and magnesium. It is noteworthy, according to these results, that up to the end of the seventh lunar month not more than one-fourth of prenatal growth has taken place.* TABLE 3.—COMPOSITION OF THE Fetus (Michel) Age N. 12) Ga, Mg. weeks grams grams grams gram 1 ek aes icy ea 2.941 0.662 0.419 0.021 PAU pe renbarah ener 6.054 1.448 2.214 0.077 ALS altel toreoa ets 11.048 2.444 4.082 0.133 DAO ras cial Suehasiovm sre 16.005 3.527 5.881 0.190 AO vecsicph ae sihe 72.700 18.673 33.260 0.815 *It would have been somewhat more convincing if Michel had used measurements instead of ages. 214 HARVEY SOCIETY Suppose it were possible from the moment of conception to keep a balance sheet of these substances for the mother, setting her intake as food over against the output through her excretory channels. We should then be able to say whether a sufficient quantity of each substance had been retained to cover the require- ments or whether the pregnancy resulted in a net loss. Experiments of this character, with reference to nitrogen, carried out on animals by Reprew,*! Ver Ecke,°* Hagemann,** Jagerroos,°** Bar and Daunay,®* Murlin,®® and Gammeltoft,°’ have developed the following facts: 1. Upon an adequate diet a dog, rabbit or goat may retain more than sufficient nitrogen to counterbalance the loss at par- turition, plus the quota taken up by the uterus and mammary glands. 2. Upon a diet which is only sufficient to maintain nitrogen equilibrium in the non-pregnant condition, due allowance being made for difference in weight, the pregnancy will result in a net loss. The kataboliec effect of the presence of the fetus is greater than the anabolic effect, taking the pregnancy as a whole. 3. While in the latter half of pregnancy there is always a plus balance, in nearly every instance recorded in animals (including the dog, rabbit, rat and goat), whatever the diet, there is in the first half either an actual negative balance or a strong tendency thereto. The latter point may detain us for a moment. Here appar- ently is something quite unusual. One might reasonably expect that the moment conception occurs, retention of materials for the growth of its product would begin and, since the total quan- tity needed for development to the middle of pregnancy could be taken by the mother in a single meal, the retention of this amount spread over so long a time should be an easy matter. Katabolism, however, has the upper hand and Gammeltoft has shown that it is not possible to overcome the tendency by heavier feeding. Indeed, Jagerroos, Bar and I have each noted that the dog at about the third and fourth week of pregnancy, cor- responding to the third and fourth calendar months in human pregnancy, may show lack of appetite and may even vomit. Bar METABOLISM 215 calls especial attention to the correspondence of this period to the period of morning sickness or the so-called physiological vomiting in women. It is true that a period of negative bal- ance has not been seen in the woman. However, it is very significant that in the only two cases in which a nitrogen balance has been kept as early as the third month (one by Landsburg ** and one by Wilson °° the plus balance should be distinctly less in this month than in the second or fourth although there were no gastrointestinal symptoms. The only reason then must have been increased katabolism. What is the explanation of this greater katabolism? It can be found, I believe, only in the nature of the means employed by the fetus for its nutrition at this stage. Recall the activity of the trophoderm up to the end of the third month in the human, and the fact that a proteolytic enzyme is demonstrable up to the end of the fourth month, a time when the placenta is considered to be completed in all essential structures. The time at which the negative balance gives way to a positive balance in the majority of the dogs studied corresponds well with the time at which, according to Bonnet,®° the placenta in this animal is com- pleted (thirtieth day). During the period of negative balance or tendency thereto, when the mother is losing more nitrogen from her body, the trophodermie cells on behalf of the fetus are producing the enzymes which enable the villi to invade the mater- nal tissues and become securely implanted in the decidua. When this invasion process has come to an end, nitrogen retention is easier. The facts clearly suggest that the extraordinary excretion of nitrogenous bodies is an inevitable wastage incident to indis- eriminate action of enzymes and is closely comparable at this stage with the wastage incident to cancer. The proteolytic action of the trophoderm is more or less unrestricted; its enzymes are not yet confined to a definite locality; they may even be dis- tributed throughout the mother’s body. This hypothesis, which was offered seven years ago*® as an explanation of the negative balance in dogs and of the period of physiological vomiting in women, has been reviewed very favor- ably by Gammeltoft, but he criticizes it as laying upon the mother 216 HARVEY SOCIETY the blame for a state of affairs which proves to be toxic for herself. Gammeltoft thus has overlooked the overwhelming evi- dence that the fetus, the neoplasm if one will, is producing the enzymes for its own use regardless of its effect on the mother. If the mother makes a prompt reaction, limiting the invader by means of the placenta, and possibly counteracting the enzymes with an antiferment produced by the specialized decidual cells, the period of heightened katabolism may be of short duration and of no serious consequence. Products of proteolysis may well be the cause of vomiting. Failure of the maternal reaction for defense and continued intoxication with such products are, not improbably, the cause of hyperemesis, or pernicious vomiting. Young’s * explanation of eclampsia as due to infarction of the placenta with consequent autolysis and intoxication of the mother with toxins of protein nature, is readily accommodated to this view and would account for the high undetermined nitrogen of the urine of that disease first noted by Ewing and Wolf and confirmed recently by Lossee and Van Slyke.** The rapid auto- lysis which Young demonstrates in such placente could be ex- plained only by very active enzymes already present. The tox- emias of pregnancy, then, it is suggested, may be explained as perversions of the chemism underlying the nutritive relations of mother and fetus. When the net result of normal pregnancy is counted up for the woman, as has been done from the seventeenth week to the end of gestation for one case by Hoffstrém,** and from the nine- teenth and twenty-fourth weeks, respectively, to the end for two others by Wilson, we find that the total retention of nitrogen exceeds the requirement of the fetus, uterus, placenta, membranes and mammaries by a very handsome amount.* This accords with the experience of a large percentage of mothers who find themselves physically much better off at the end of gestation than at the beginning. Such advantage to the *The human female enjoys a great advantage over the female of many other species in the fact that her offspring rarely weighs over eight per cent of her own body weight, whereas the dog, rabbit and many others deliver as much as twenty to twenty-five per cent of the body weight. INFLUENCE OF PREGNANCY ON THE PROTEIN _ CONDITION OF THE DOG (SOLID LINE) NITROGEN RETAINED FOR GROWTH OF EMBRYOS nee 4 Busy Rea = Doc A Steet 16 Biack = DosB eae oS me | PUPPY 9 2 7 Ee z a 8 = Eas iB ee 24 > = : oe 32 ae N. ee. the Oss : = Sq Ro seieere 40 WEE ee ie VT VT INFLUENCE OF MENSTRUATION ON THE RETENTION OF NITROGEN. THREE EXPERI- BENS ON ONE DOG WITH DIFFERENT DIETS 15 GMS | MS AED , ve 0 Wey moe ven iy wu 3 MENSTRUATION PREGNANCY Fic. 1. The lower curves she yw the course of the total nitrogen retention during the e- menstruation and early pre enan 2V of the dog. Not the muaus nitrogen balance in ube third uid He vurth weeks of pregnancy. U pper curve show Sane t Ne nitre gain or Jo at pace ah ek roe the USS oE -y in two different do Dog a nishe de eaieane a minus bala yme 20 grammes be irre in a ges cate n with fe ae DURE Dog A finishe oa with: about the same. amount ahead in a gestation with yne pup. METABOLISM Q17 Fria. 2 ia DZ eo i ae es Sanaa a w CCP EEE 2 ea me on CC & = POH Fic. 3 Fic. 2.—Chart showing retention of nitrogen by Slate eae case from the seventeenth to the fortieth week of pregnancy. ao N _Fetained ; N_ retained by fetus (based on Michel’s analysis, table 3 N retained by mother’s own body. Fic. 3.—Chart showing retention of phosphorus by Hoffstrém’s case from the seventeenth to the fortieth week of pregnancy. total P retained; P retained by the fetus (based on Michel’s analysis, table 3) ; — — — , P retained by mother’s own body. 218 HARVEY SOCIETY mother is, no doubt, a purposive one, as both Hoffstrém and Wilson suggest, in anticipation of the demands of labor and the lactation period. The pelvic and abdominal muscles appear to claim a large share of the surplus, but the gain is also more or less general. I have calculated the quantity of milk at 1.5 per cent of protein which might be produced from the nitrogen gained during the pregnancy for the three cases mentioned: Grams Retained Equivalent in Milk Hofisthom-si cases maser eo caterer 209.0 87 liters Wilson's case cD Sea cectice setae. 284.5 114 liters Wilsons case nN sce wee ears 210.9 88 liters It is no discredit to the maternal organism to say that this apparent benefit to her body was brought about through the stimulus of the fetus itself. This has been proved for the mam- mary glands by Herrmann ® and others, the hormone coming from the placenta, and it seems reasonable to hope that we shall some day find a stimulus to growth and development for under- developed women by employment of such extracts made from lying-in material. The story for nitrogen retention may be repeated with some variations for other chemical elements, phosphorus, sulphur, eal- cium, magnesium. Hoffstrém’s curves show a substantial gain for each element except calcium. This should be borne in mind in relation to a possible acidosis to be mentioned later. Hoffstr6m’s beautiful work, which required two years for its completion after the materials were collected, together with Wilson’s work, may be truly said to demonstrate that, so far from being a sacrifice of the individual for the good of the species, gestation normally may be looked upon as a means em- ployed by the species for the good of the individual (mother) .* Further evidence of physiological adaptation may be found in a study of the qualitative effects of pregnancy on protein metabolism. Pregnaney is one of the few conditions studied (fasting is another) in which the distribution of nitrogenous and *It would be interesting and important to know whether at the end of lactation any part of this acquisition on the part of the mother has been retained. METABOLISM 219 sulphur compounds in the urine shows any departure from the usual distribution in normal adults on ordinary mixed diets. All of these changes, however, can be explained on physiological grounds. The following changes in the absolute sense have been demonstrated, all referrring to late pregnancy: 1. Urea nitrogen is distinctly lower (Massin,** Whitney and Clapp,®*’ Mathews,®* Edgar,®® Murlin,”® Murlin and Bailey 7"). 2. Ammonia nitrogen is very slightly, if any, higher ease EGS Ese Bee ASRS aS Seas Seca isso Wi SE oa a al ie ZZ fe | Z| oe Se ae REA esas SiESeREG Bes ehOgEs Chae co —e ie [ AT a: eas alee fe a = ca ad Sot a eee een sin 4% {/ae 4a te e 7] Bea 4 ” 45 u% a7 ey ay Jo ea a wa »* Is ea es ie] a a ae Z] ial ial a Bo Wa Sas so See eence Fic. 4.—Chart showing retention of calcium by Hoffstrém’s case from the Reyenteenth to the fortieth week of pregnancy. total calcium retained ; , calcium retained by the tetus (based on Michel's analysis, table 3) ; — — —.,, calcium retained by mother’s own body. (Slemons,’?? Falk and Hessky,** Murlin,’* Murlin and Bailey,” Gammeltoft,°? Hasselbaleh and Gammeltoft,*> Lossee and Van Slyke,®** Wilson °°). 3. (a) Creatinin nitrogen may rise just before parturition (Murlin for dog and woman; not observed three to six weeks antepartum by Van Hoogenhuyse and ten Doeschate ** or Murlin and Bailey); (b) creatinin may fall just before parturition (Gammeltoft for rabbit). 220 HARVEY SOCIETY 4. Creatin appears in urine even on creatin-free diet shortly before parturition (Heynemann,’ Murlin,” Krause, Van Hoogenhuyse and ten Doeschate, Murlin and Bailey, Gammel- toft). 5. Amino-acid nitrogen, as determined by the Van Slyke method, not increased (Lossee and Van Slyke); as determined by Henriques and Sorensen titration method, slightly increased (Falk and Hessky, Murlin and Bailey, Gammeltoft, Wilson). 6. Total puri nitrogen somewhat increased (Murlin and Bailey). 7. Undetermined nitrogen, including polypeptid nitrogen, slightly increased (Ewing and Wolf, Murlin, Murlin and Bailey, Falk and Hessky, Lossee and Van Slyke). 8. Inorganic sulphate sulphur, distinctly lower, and neutral sulphur, only relatively higher (Murlin for dog and woman; Hoffstrém for woman). The explanation of the lower urea nitrogen and inorganic sulphate sulphur is found in the retention of materials which, in the absence of the fetus, would be excreted in these forms. The relatively higher creatinin and unoxidized or neutral sulphur, products considered as strictly endogenous, are explained by the lower percentage of the exogenous urea and inorganic sulphate sulphur. The presence of creatin is probably due either to the slight acidosis to be mentioned presently or to the demands of the fetus late in pregnaney for carbohydrate, or both. The slightly higher amino-acid and polypeptid nitrogen and undeter- mined nitrogen may not be strictly reciprocal with the urea nitrogen, for they are lowered in the absolute sense immediately after parturition. Slight increase of these substances does not signify deficient deamination by the liver, for if the liver were injured, one could scarcely imagine that it would recover imme- diately after parturition. The sudden drop much more cogently suggests a fetal origin for such bodies. In so far as protein materials must be worked over by the placenta to harmonize them to the purposes of the fetus, there must be rejected materials and these added to the general circulation would in part pass to the kidney before being deaminated, whereas such bodies originating in the alimentary tract have first to pass the liver. “INFLUENCE OF PREGNANCY ON THE COMPOSITION OF THE URINE OF THE DOG 96 PERCENT UREA+NH3 NITROGEN 94 TOTALN. 92 em an 90 a BB oe ‘86 INORGANIC SULPHUR 74 PERCENT 72 TOTAL 5. 76 58 66 —_—_—— DoGA = DOTTED LINE Do6B = BROKENLINE Dog C = SOLIDLINE NEUTRAL SULPHUR 29 PERCENT — 27 TOTALS. 25 23 6 Percenr REO=CREATININ N. BLACK=CREATIN N. — 4 OF ee 4 TOTALN. oa WEEN eee le en IW ae ea Vee. WT eX Fie. 5.—Chart showing influence of pregnancy on the percentage of urea and ammonia nitrogen, creatin and creatinin nitrogen. and the inorganie and neutral sulphur fractions in the urine of the dog. Compare Figs. 6 and 7 The changes are due to retention of nitrogen whith -would be excreted as urea in the non-pregnant condition and of sulphur which would be excreted as inorganic or oxidized sulphur. 221 METABOLISM SULPHUR FRACTIONS OF URINE IN PREGNANGY (HOF FSTROM) Pe FoR Sa a WF 8 Oe BER ADNRA Ra RAS (SRE Vs Pa PH SEARS NEARS eee Nee 2G Wreexsitn ott Stqii2emuaiem 2aleceSunaqeuas COMING Y Saaek ae Se Vee pan | Saint Pan We Fin eel PL PECTS Fic. 6.—Chart showing the distribution of the sulphur fractions in the urine of Hoffstrém’s case from the seventeenth to the fortieth week of pregnancy. The neutral sulphur remains at about the same level (in the absolute sense) through: out. The inorganic sulphate sulphur falls, showing that the sulphur which in the non-pregnant condition would be oxi dized is held back by the fetus. 222 HARVEY SOCIETY Fia. 7. Partus SRR ERE NRE week Sr SUINEELE CECE EEE CECE EEC 0: al) i\ fa Oe SI) i a i 2 a a TT 29 ep a0 Fia. 8. Fig. 7.—Chart showing the relation of the hippuric acid nitrogen to the total nitrogen in the urine of a goat during pregnancy (Gammeltoft). The hippuric acid in herbivorous urine corresponds to the urea nitrogen in the urine of other animals. This chart should be compared with Fig. 5, showing the behavior of urea and ammonia nitrogen in the dog. lic. 8.—Chart showing relation of mon-amino-acid nitrogen and of total peptid-bound nitrogen to total nitrogen in the urine for last two months of a normal pregnancy (Gammeltoft). Note the fall in amino bodies immediately after parturition simultaneously with the great rise in total nitrogen. METABOLISM 223 The ammonia nitrogen deserves more than passing mention. Ten years ago this fraction of the nitrogen in the urine suddenly assumed large proportions in obstetrical literature because of its supposed value as an index of hepatic inefficiency. This esti- mate has not been maintained because, on the one hand, known pathological lesions of the liver have not been shown to produce high ammonia and, on the other, because the researches of Underhill and Rand,’® Heynemann, Murlin and Bailey, and Lossee and Van Slyke, have made it very doubtful whether the high ammonia of pernicious vomiting is ever to be ascribed to anything more than starvation and depletion by the fetus; they have shown further that eclampsia may be accompanied by no increase in ammonia and that large errors may easily arise from faulty methods of collection and preservation of the urines. The significance of ammonia in the urine is, therefore, restored to its old status; namely, an index of the depletion of fixed bases. Is there such a depletion in normal pregnancy, either by over production of acids and excretion of bases in combination with them, or by withdrawal of bases to the fetus? The position of Murlin and of Heynemann that there is in normal pregnancy no absolute increase, but only a relative one due to lower urea nitrogen, has been substantiated by the most recent investigations of Hasselbalch and Gammeltoft, at Copenhagen, by Wilson, at Johns Hopkins, and by Lossee and Van Slyke, here in New York. The effect of the nitrogen retention upon the ammonia percentage is well illustrated by the following figures from Landsberg. The diet was similar in all eases. Also the effect TABLE 4.—AMMONIA NITROGEN IN URINE OF PREGNANT AND NON-PREGNANT Women (Landsberg) Total N NHi-N Er eent Average 10 cases pregnant ........ 12.68 0.786 6.2 Average 6 cases non-pregnant ...... 16.03 0.771 4.8 of a severe catharsis is illustrated in one of the cases reported by Murlin and Bailey. 224 HARVEY SOCIETY TABLE 5.—EFFECT OF CATHARSIS (Murlin and Bailey) Total N NHeN Fetcent cy Normal pregnancy, ninth month.... 5.82 0.57 9.8 0.21 After severe catharsis ............. 3.36 0.58 17.3 0.23 In both instances the total ammonia excretion in grams is the same and the higher percentage is due to lower total nitrogen, caused, on the one hand, by retention, and on the other, by a too severe purging which removed the food as well as the waste from the bowel. The absence of high ammonia in the absolute sense does not, however, preclude the presence of a slight acidosis. ‘The body has other resources than the formation of ammonia for the neu- tralization of acid. A slight excess of organic acid in the cir- culation could be compensated by removal of a little more car- bonie acid from the blood, while a slight relative acidosis due to diversion of bases to the fetus would produce the same net effect indirectly by diminishing the carbon dioxid-earrying power of the blood. Asa matter of fact, Leimdorfer, Novak and Porges *° and Hasselbalch and Gammeltoft find a lower alveolar tension of carbon dioxid in pregnant women and state that it appears as early as the first month following conception. Lossee and Van Slyke confirm this for late pregnancy by Van Slyke’s method for the carbon dioxide-combining power. The delicate regulation of the actual hydrogen-ion concentration without drawing upon the ammonia mechanism is beautifully illustrated in Hasselbalch and Gammeltoft’s work. They report simultaneous determinations upon urine, blood and alveolar air, both before and after par- turition, for eleven cases of normal pregnancy. TABLE 6.—ACIDOSIS OF NORMAL PREGNANCY. AVERAGE OF ELEVEN CASES (Hasselbalch and Gammeltoft) Ante Partum Post Partum Urine: TOUAIMNACFOPENY ice accreted cies ences (grams) 9.2 11.4 INS Ro Pee Aes Camere DIS cae Homes (grams) 0.55 0.57 Per cent OL Lota 2%. scteere son oie eis sentore Greer 5.9 4.9 Blood: Paastrs ON eV CNRIGI saree iepeieie eit) a ere ater tele 7.44 7.49 Corrected for Actual Tension .............. 7.44 7.44 Alveolar Air: CO; Pensions (mmsb yy) eed eaiioklaemisiste 31.3 39.5 METABOLISM 225 The ammonia before parturition is not higher than after but there is at any given tension of carbon dioxid a higher concen- tration of hydrogen (lower Py ). Corrected for the observed car- bon dioxid-tension in the blood drawn, the concentration of hy- drogen is exactly equalized. The slight decrease in carbon dioxid- carrying power of the blood (for any given tension of that gas), as compared with the value which is restored after parturition, is exactly compensated by the greater activity of the respiratory mechanism for removal of the carbon dioxid. Hasselbalch, in- deed, has shown that the respiratory regulation of the hydrogen- ion concentration is brought about by an increased sensitivity of the respiratory center to the presence of carbon dioxid. The regulation, therefore, is perfectly automatic. METABOLISM IN THE PUERPERIUM The first few months of pregnancy may well be described as a contest between the new organism and the old. The new, with its intensely active proteolytic enzymes, is concerned with getting a foothold upon a supply of nourishment. When its activities become too aggressive or the maternal reaction is not quite up to normal, the result may be a severe disharmony. Otherwise, the contest results in a compromise, the mother conferring nutriment and protection in exchange for stimulus to her own organs. But there comes, by what precise causes we shall not stop now to inquire, a limit to this tolerance on the part of the mother, and parturition ensues. What readjustments in the metabolism of the mother do we witness after the separation? The necessity for retention of building materials is, for the immediate future, somewhat less even though the child is to be nourished by the mother; for intrauterine life was dependent upon the mainte- nance of many adventitious structures, some of which, being no longer required, are thrown away, and others are restored to pre- gestation size. The first change we might expect, therefore, would be the disappearance of the positive balance for the several elements, which has been deseribed, and its replacement by a negative balance. Only the nitrogen balance has been studied in women during the involution period but it is safe to say that 15 226 HARVEY SOCIETY what holds for this element will in general be true of the others; sulphur, phosphorus, calcium, magnesium. The chief source at least for the extra nitrogen found in the urine is the uterus itself; for Slemons ** has shown that a woman who withstood Caesarean section and prolonged anesthesia ex- creted only 40 grams more nitrogen in the urine during the puerperal period of twenty days than during the post-puer- perium, while a uterus removed from another woman of the same race contained 38 grams. Making some allowance for the effects of anesthesia, some for the difference between the two uteri, and some for lochia, the correspondence is as close as could be ex- pected. The woman whose uterus was removed by the Porro operation should have excreted about 35 grams less than the one whose uterus went through involution following conservative Caesarean section. Less than this difference would have indi- eated the loss of nitrogen from other tissues. As a matter of fact, while on a similar diet and after anesthesia of similar duration, she excreted 61 grams less. The discrepancy, Slemons thinks, was attributed to other extraneous differences between the two cases. The observation proves at least that nitrogen is not lost from other organs. In view of the enormous total retention of nitrogen outside of the genital tract which has been made out by Hoffstrém, Landsberg and Wilson, the occasion for surprise is that all of it apparently is conserved during the puer- perium, doubtless in the interest of lactation. ‘The nitrogen out- put reaches its maximum on the fifth day for the dog and on the sixth or seventh for the woman. Taking the period just before delivery as a standard, the following changes in the distribution of nitrogen in the absolute sense are observed in the period immediately following: The urea nitrogen rises, the creatin rises, the ammonia remains the same, the creatinin falls, and the monamino-acid nitrogen falls. In percentage of the total the urea nitrogen may either rise or fall, depending upon the adequacy of the diet in the first few days after delivery and upon the rate of retention for milk pro- duction, as against the rate of involution a little later. Where sufficient care is exercised in the collection of the urine, it will be METABOLISM 227 found in normal eases that the percentage of ammonia always falls, owing to the rise in the total nitrogen. There is no acidosis characteristic of the puerperium. The absolute decline in the ereatinin excretion, Longridge ** believes, is explained by the reduction in mass of active muscle. The creatin fraction only calls for more extended remarks. Having in mind that creatin had been identified in smooth muscle ** and that in wasting dis- eases, muscular dystrophy especially, creatin always is found in the urine, Shaffer ** and I ®° independently ascribed the crea- tinuria of the puerperal period to the involution of the uterus, and, this view was generally adopted. Mellanby,*® however, sharply challenged this explanation by showing that a woman may excrete even more creatin after the uterus is removed than another who has been through essentially the same operative ordeal without hysterectomy and the facts have since been cor- roborated by Morse.*? The observations of both Mellanby and Morse leave something to be desired in the way of control of the diet. The exrperimentum crucis is, of course, a very exacting one. Both patients, the one after simple Caesarean section, and the other after the Porro operation, must be upon an adequate diet, which, of course, should be creatin-free, from the moment food can be taken. Most operated patients of this character could not take a diet adequate to protect the body against loss of creatin and if they could the further difficulty would at once be presented as to whether the more radical operation would not of itself cause loss of creatin. Both observers have, in fact, proved too much. In order to show that the creatin does not have its origin in the uterus, it would be sufficient only to prove that just as much creatin is excreted by a patient following removal of the organ as by another in whom it is left in place. But in all three cases, one studied by Mellanby and two by Morse, more creatin was found after the radical operation, a fact which points to less complete nutrition or some other vitiating circumstance of the radical operation. Once this is admitted, does not the validity of the whole com- parison fall down? Soon after Mellanby’s paper appeared, Bailey and I at- 228 HARVEY SOCIETY tempted to test the matter on dogs, by removing the uterus several days after parturition, maintaining the animal both before and after operation on the same creatin-free diet. Out of four attempts only one succeeded; 7.¢., in only one case did we succeed in inducing the animal to consume an adequate diet of creatin-free materials throughout. This dog showed the higher excretion of creatin after the operation, but we are almost con- vineed that the result was due to the acidosis incident to anesthesia. Mellanby offers some evidence that the excretion of creatin is in some way associated with the activity of the mammary glands; for example, when the milk is delayed, creatin in the urine does not appear and the curve of ereatin excretion runs parallel with the curve of milk production. Ill. THE ENERGY METABOLISM Growth and maintenance of cells are dependent upon two fundamental properties of protoplasm which ordinarily are re- garded as quite distinct because in the adult they may vary quite independently. In the developing ovum, however, we are be- ginning to see how they may be very closely related. One of these properties, the ability to attract and incorporate into its own structure and thus vitalize germane materials, has been dis- cussed for the fetus and the effects of this intrauterine growth upon the protein metabolism of the mother have now been passed in review. The other fundamental property is the ability to activate oxygen so that without raising the temperature to the kindling point for the oxidizable materials, energy may be set free by oxidation. Growth is even more dependent upon oxidation than is mere maintenance of the body. Warburg *® has made the interesting suggestion that the purpose served by oxidation in cells which do no external work, is to maintain the internal structure of the eell. Certain properties of semi-permeable membranes, such as the electric charge, are preserved, thereby preventing mixing of the constituents by diffusion. Internal structure, Warburg be- lieves, is necessary also to provide surfaces for condensation of the catalysts which are active in the cell processes. Now Meyer- METABOLISM 229 hof *® has shown that the ‘‘caloric quotient’’ of developing eggs is considerably lower than in adult organisms. By ‘‘caloric quotient’’ one means the number of gram calories of heat pro- duced for each milligram of oxygen consumed. In resting adult organisms it varies from 3.2, where protein is the source of energy, to 3.5, where carbohydrate is the source, the value for fat lying between these extremes. In developing sea urchin eggs it is 2.5 to 2.9, thus indicating that oxygen is being used ex- tensively for some other purpose than heat production. Follow- ing Warburg’s suggestion, we may suppose, then, that more oxygen is needed to maintain the structure because more catalysts are at work or are working more actively, or, as Lyon °°? has suggested, the oxygen may be used directly in the synthetic processes of growth. On either supposition we see how the sub- stance metabolism involving transformation chiefly of proteins may be closely related in the embryo to energy metabolism. Warburg ® has recently repeated many of his earlier studies on the respiratory exchange of the sea urchin egg and has con- firmed them in all essential respects. The oxygen absorption of the unfertilized ovum is about five hundred times that of the sperm cell of the same species. But when the two cells unite in the act of fertilization the oxygen absorption goes up to about 3500 times that of the sperm. This increase to sevenfold the metabolism of the unfertilized egg takes place in ten minutes. At the end of six hours it is twelve times and at the end of twenty-four hours, when the gastrula stage is reached, it is twenty-five times the absorption before fertilization. It is safe to say that something like this happens in the mam- malian ovum upon fertilization. But we can not study the metabolism of a mammalian ovum at such an early stage. It is not until well beyond the middle of pregnancy that the respira- tory exchange of the simple fetus is large enough to be measured by existing means. We shall limit our inquiry here to two questions: (1) what kinds of material are oxidized to furnish the energy in the fetus and newborn; and (2) how much energy is thus set free in the pregnant woman and the new-born child in comparison with the adult? 230 HARVEY SOCIETY THE RESPIRATORY QUOTIENT OF DEVELOPMENT Qualitative differences in the energy metabolism of the em- bryo depend upon the kind of material supplied by the mother. The hen supplies only fat and protein in the egg; hence the respiratory quotient during development into the chick can never be higher than 0.78. Chemical studies of eggs before and after incubation by Liebermann,°*? the calorimetric determinations of the heat of combustion before and after incubation by Tangl,°* and the metabolism studies (using both direct and indirect methods) by Bohr and Hasselbalch,** all agree in showing that the material oxidized in the development of the chick is fat. Regarding the source of energy for mammalian development our information is extremely scanty. Cohnstein and Zuntz,' analyzing the blood of the umbilical vein and artery of the sheep embryo, found respiratory quotients of 1.0 and 1.6, re- spectively, in two eases. Bohr,®> measuring the total respiratory exchange of the pregnant guinea-pig before and after clamping an umbilical cord, noted differences which gave a respiratory quotient for the embryo in the neighborhood of unity. These are all the recorded observations on the respiratory exchange of the fetus directly. Such as they are, they indicate plainly that the source of the energy is carbohydrate, the most readily diffusible of all the food-stuffs. Several observers have noted a rising respiratory quotient during pregnancy in both lower animals and the human subject and there is no doubt, from the observations of Carpenter and Murlin ** and Hasselbalch,® that the quotient is higher just before parturition than just after, but it is not certain to what extent the limited diet usually allowed the puerperal mother in the first days after delivery is responsible for the difference. The respiratory quotient of the new-born has been for some twenty-five years a matter of recurrent interest. The earliest observations by Mensi, Scherer and Babak have turned out to be wholly untrustworthy because of imperfect technique. Murlin °° reported in 1908 that the respiratory quotient of the new-born puppy was in the neighborhood of unity. Hasselbalch,*® in 1904, and Weis,’ in 1908, were the first to observe that the quotient METABOLISM 231 for the new-born infant, also, is high, often in the neighborhood of unity, and indicating the combustion of carbohydrate. With- out knowing of these results because they were published in obscure places, Baily and Murlin? obtained quotients of the same character in two infants observed at six hours of age. They confirmed Hasselbalch’s observation, also, that the quotient falls rapidly after the first few hours, and on the second day before food was given they found it slightly below the quotient of pure fat combustion and indicating a certain degree of starvation acidosis. The interpretation placed upon these observations by Hasselbalch and by Bailey and Murlin was essentially the same; namely, that the child is born with a sufficient reserve of carbo- hydrate to supply its energy requirement for a portion of the first day, but that this supply is quickly exhausted and the child should be fed very early. Benedict and Talbot,’ in a long series of determinations on new-born infants, however, have failed to find the quotients uniformly high in the first hours, although they admit that the majority of the cases observed within eight hours of birth gave quotients above 0.80, whereas the majority of those observed after eight hours gave quotients below 0.80. Two rea- sons for the discrepancy found by the different authors may be mentioned. One of these is given by Hasselbaleh; namely, the state of nutrition and the maturity of the child when born. In his series Hasselbaleh was certain that the better the nutritive condition of the infant, the higher was the quotient, and the average respiratory quotient for prematurely born infants was below that of infants born at term. The other reason, I believe, is found in the level of the blood sugar at the time the child is born. When the mother has a severe labor or when an anes- thetic is necessary, the blood sugar of the mother, as well as that of the fetus, according to Morriss,'* rises. In the former circumstance it may rise in the fetal blood to 0.12 per cent. and in the latter to as much as 0.14 per cent. These are distinct degrees of hyperglycemia and might very well sustain the respira- tory quotient at an unusual level for several hours. Hence, we might well expect the quotient of the new-born, following a pro- longed and severe labor, to be high. Indeed, Hasselbalch draws 232 HARVEY SOCIETY especial attention to one of his prematurely born infants delivered by a forceps operation, because it gave a quotient higher than others of like age. When we remember that large, well-de- veloped babies cause more prolonged labor than small or prema- turely born babies, we find another reason for Hasselbalch’s discovery that the former present the higher quotients. QUANTITY OF ENERGY REQUIRED IN DEVELOPMENT With the exception of a few wholly untrustworthy observa- tions of Cohnstein and Zuntz** on the embryo sheep, in which the authors undertook to measure the total oxidation by analysis of blood drawn from the umbilical vein and artery, the only direct studies of the energy metabolism of the mammalian offspring before birth is that of Bohr.®* Bohr operated a pregnant guinea- pig so as to expose the umbilical vessels. With the anesthetized mother immersed in a warm bath of salt solution he then meas- ured the respiratory exchange through the maternal system before and after clamping one of the umbilical cords so as to exclude entirely the one fetus. From the differences obtained he calculated the metabolism of the young near term at 509 ee. of earbon dioxid given off per kilogram an hour as against 462 ce. for the mother, an increase of about ten per cent. Rubner ?°° in 1908 expressed the belief that his law of sur- face area applied to the embryo as well as to the new-born. Assuming the average weight of each individual at birth to be eight per cent of that of the mother, he calculated that the energy metabolism per unit of weight of any new-born mammal would be approximately twice that of the mother. Because the fetus is much less active than the new-born, its metabolism, so Rubner held, should be considerably less than this, which indeed Bohr’s fragmentary results indicate is the case. We shall return to the new-born later. Meantime one gets very little help either from Rubner or Bohr in forecasting what the effect of the fetus would be on the total metabolism of the mother. Granting that its energy requirement is greater than the same weight of maternal tissue, we must remember that a large part of the increased weight at the culmination of preg- ‘TABLE SHOWING THAT’EXTRA HEAT PRODUCTION OF PREGNANCY _ DAY FROM — WEIGHT — Tempmature CALORIES TOTAL ENERGY PARTURITION IN KG. DF CAGE IN Foon Prooucep RETAINED THiRD Berore Wune23) 14.5 = = 28.0°C 907.4 55s 356.| “PARTURITION [JUNE 26) _ Dwe Puppy Born: Weieut, 280 Gm. irine WTS “SEXUAL REST;AFTER LACTATION 13.78 radiate eel Sarason Vay as: 402.1 PARTURITION (JuLy | 5) SECOND PREGNANCY THirD Berore (Dec. 11) 16.86 CLAGE 907.4 763.8 143.6 Parturition (Dec. | 4) Five Puppies Born: WEIGHT, I560 Gy. ‘3591.3-505.3= 46.0 CALORIES FOR | PUPPY WEIGHING 280 GM. 763.8 —505.3=258.5 CALORIES FOR 5 PUPPIES WEIGHING 1560 GM, PB0n sb 46 258.5 Irig. 9.—Extra heat production in a dog just before parturition in two different pregnancies. I’rom the first. one pup was born; from the second, five. METABOLISM 233 nancy takes little or no part in the metabolism; fluids, none at all; membranes and cord, next to none; placenta, uterus and mammaries, probably not more than so much maternal matter. The net effect, therefore, would be a sort of algebraic sum of high, low and medium metabolism added to that of the mother’s. As a matter of fact, the earlier observations on pregnant animals give conflicting results. While Reprew,°! working with rabbits, guinea-pigs and a dog, reported no increase in metabolism per unit of weight, Oddi and Viearelli,!°* working with mice, found a marked increase. Magnus-Levy '° also, in the first observations on the energy metabolism of the pregnant woman ever recorded, noted an increase in oxygen absorption from 2.8 ec. per kilogram a minute in the third month to 3.3 ce. in the ninth, a rise of 17 per cent. My own observations *°® on the dog made on the third day before parturition on a pregnancy from which only one pup was born, show an increase of six per cent per unit of weight over that of complete sexual rest, while on the corresponding day of a later pregnancy in the same dog from which five pups were born the increase was 28 per cent. The extra metabolism was proportional to the weight of the new-born delivered. In the woman pregnant with a single fetus the observations of Zuntz,’°%* Carpenter and Murlin and Hasselbaleh agree in showing an extra metabolism near term of about four per cent over that of the same woman or other women in complete sexual rest. All of these authors surmise that this is scarcely more than may be accounted for by the increased respiratory activity neces- sary to preserve the hydrogen-ion concentration of the mother’s blood. It is evident, then, that the total product of conception added to the mother’s body functions as so much maternal tissue— the higher metabolism of the embryo being just counterbalanced by the inactive and relatively inactive structures. From observations on the dog above referred to, the extra metabolism on the third day previous to parturition was equi- valent to 164 calories in the first pregnaney and 165 ealories in the second per kilogram of pup delivered three days later. Assuming the applicability of Rubner’s law of surfaces to the 234 HARVEY SOCIETY new-born pup, with the same constants as for the adult dog, it was calculated that the theoretical metabolism necessary to main- tain in muscular rest a new-born pup weighing what the fetus actually weighed would be just equal to the extra metabolism of pregnancy. In other words, if the new-born pup were to lie perfectly still and sleep as quietly as the fetus does, the increased metabolism at room temperature over the metabolism at its mother’s body temperature would just compensate the meta- bolism of the placenta, uterine wall, ete., and the total requisition placed upon the mother for maintenance of the new-born by food from her mammary glands would not, for the first days at least, exceed the requisition made upon her body in the last days of pregnancy by way of the placenta. Here appeared a very important principle of adaptation, the requirements of the new- born being just equal to the requirements of the total product of conception, accessory structures included, just before parturi- tion. It was impossible to demonstrate the principle with absolute certainty on the dog because of the mother dog’s anxiety for the offspring in the first days after parturition. It was demon- strated a year later, however, by Carpenter and Murlin ** on three cases of human pregnancy at the nutrition laboratory in Boston. These three cases, two primipare and one multipara, were observed in the bed calorimeter for some three weeks pre- vious to parturition and mother and child together were placed in the calorimeter again as soon as possible thereafter. The mothers were soon trained to lie perfectly still and by keeping the infants awake for several hours just before the calorimeter periods, they were readily induced to sleep throughout or nearly throughout the observational period of two to three hours. In two of the cases the comparison of antepartum and postpartum metabolism of mother and offspring together showed a difference of less than one per cent. The other case showed an increase of seven per cent, partly because the antepartum observations did not occur closer than the thirteenth day before delivery and partly because the child cried on two out of three occasions while in the calorimeter in the later experiments. Ruling out the factor of muscular activity as we are able to do in the two 235 METABOLISM TABLE 7.—ENERGY METABOLISM OF MOTHER AND CuHILp TOGETHER BEFORE ANp AFTER PARTURITION (Carpenter and Murlin) Case. Mean of all days before and after delivery. Case 1— Ist, 4th and 6th before delivery.......-- masa cua tecs 5a 2d, 5th, 12th, 14th and 17th after delivery.......- : Case 2— 13th, 17th, 19th, 20th and 22d before delivery.....- 2d, 5th and 11th after delivery........---+++++++-: Case 3— Ist, 3d, 17th, 21st and 24th before delivery.......-- 4th, 8th and llth after delivery ........-----+++>: Respiratory Exchange. Energy Production. eee - aise Bd S a 8 -AO a & 2 2 o a, : o oO 3 a & E g a E a ane) a Q I BS ce) es | | +) x ©) (eo) a4 3s a 3 36.75 21.3 18.4 .85 60.0 61.3 60.7 36.9 20.2 18.5 .80 61.2 61.2 61.2 36.68 22.3 19.6 .83 63.6 65.9 64.7 368 21.7 204 .78 71:1 67-56 (69.3 36.64 23.9 20.2 .86 72.2 68.7 70.6 37.23 23.1 20.3 .81 70.8 68.6 69.7 Calories per hour. % difference. +.87 per kg. 0.96 Meu 1.11 1.32 1.02 Bt % difference. +15.6 +18.9 + 88 236 HARVEY SOCIETY cases, the curve of total energy metabolism of the mother and ofispring suffered no deflection at parturition. It is a remarkable fact that the increase in oxidation in the child’s body when it passes from the warm environment of its mother’s uterus to the colder environment of the outside world (in bed beside its mother), supplying its oxygen now by its own lungs instead of from the mother’s placenta, should so nearly compensate the oxi- dation in the accessory structures which supported it im utero. Just how much the child’s metabolism is altered by the changed environment and changed circulation we have no certain means of knowing. That it is considerable, that the change represents, indeed, a turning-point, in the quantitative sense as well as in the mode, of nutrition, is evident from what has been said already as well as from what will follow immediately. The demands upon the mother’s digestive system, however, are not greater. She is called upon to supply the same amount of energy in potential form to herself and child immediately after parturition that she did to herself and child immediately before. The rate of oxidation or heat production per unit of weight for the puerperal woman in these three cases was eleven per cent higher than the average for eight non-pregnant women and seven per cent. higher than that of the same subjects just before deliv- ery, a difference which may be ascribed in part to the increased activity of the mammary glands and in part to the stimulating effect of the products of involution. Since these products are protein in nature they would unquestionably stimulate metab- olism in the same way as Lusk °° has shown for the amino-acids. METABOLISM OF THE NEW-BORN The energy production of a grown person in health and while resting in bed may be stated as approximately 1.0 ealory per kilo of body weight per hour (Du Bois). The average for eight normal non-pregnant women between the ages of eighteen and fifty-five years under these conditions was found by Carpenter and Murlin to be 0.99 ealory per kilo an hour and the average for three normal puerperal women was 1.09 calories. The only comparison ever made between the metabolism of METABOLISM 237 the new-born infant and its puerperal mother was reported by Carpenter and Murlin in the work to which reference has been made. The infant’s metabolism was measured by difference be- tween the metabolism of mother and child taken together and that of the mother taken alone. The average age of the infants at the time of the observations on the mother alone was ten days. The metabolism was found to be 2.8 calories per kilo an hour or 2.5 times that of the mother. When we compare the metab- olism per unit of body surface, as caleulated by Meeh’s formula, we find that of the child somewhat less than that of the mother. Nothing could better illustrate the applicability of Rubner’s law of surface to persons of different size and widely different physio- logical conditions than the data from this comparison. The pregnant woman just before delivery, the same woman two weeks after delivery, weighing 9 to 10 kilograms less, and the child weighing one-sixteenth to one-twentieth the weight of the mother —all produce the same amount of heat per unit of surface. TABLE 8.—METABOLISM BEFORE AND AFTER PARTURITION. THe METABOLISM OF THE CHILD WAS DETERMINED BY DIFFERENCE (Carpenter and Murlin) Calories Calories. Weight Calories per Sq. M. per Ig. Case I: in kg. per Hour. (Meeh). per hour Before parturition ........ 63.0 60.7 31.4 0.96 After parturition ......... 51.4 53.9 31.7 1.05 IDMnIG RONG Bese ocala Holle 11.6 6.8 (ONG IU Apa Aa ek ara cy cece eee 2a Wee 30.5 2.70 Case II: Before parturition ........ 58.0 64.7 35.1 1.11 Ate pactunitiony yi). \..). 48.5 59.0 36.2 1.21 MOITETENCOs helete ri eiere ras cis lai 9.5 5.4 GT sire se tvetotekevetetots, shane 3.4 9.8 34.9 2.88 Case III: Before parturition ........ 69.1 70.6 34.0 1.02 After parturition ......... 60.1 60.4 31.9 1.00 ID ater KO ae Som aea ae Or 9.0 10.2 (CHOC) CO GE Ss Brea RC Etch ORG toe 34 9.3 34.8 2.90 Average: Before parturition ........ 63.0 65.1 33.5 1.03 After parturition ......... 53.0 58.1 33.3 1.09 238 HARVEY SOCIETY As sometimes happens in scientific work, the beauty of a comparison of this sort is marred slightly by more accurate data. Since this work was done, nearly eight years ago, observations by Benedict and Talbot and by Bailey and Murlin have shown that the metabolism of the sleeping, new-born infant is nearer two calories than 2.8 per kilogram and hour and 25 calories rather than 30 per square meter and hour, in both respects distinctly lower than the metabolism under similar conditions for the adult. Note that this fulfills exactly the estimate made by Rubner on purely a priori ground. Bailey and Murlin were fortunate enough to have as subjects two infants of widely different body weight born on the same day just three hours apart, so that it was possible to study them successively at exactly the same age. This comparison illustrates the influence of body fat on the heat production. The larger infant has the lower metabolism on the basis of weight, but the two have nearly the same metabolism on the basis of surface. TABLE 9.—ENERGY METABOLISM OF Two NEW-BORN INFANTS (Bailey and Murlin) Weight, kgm Age 1885(0) Cal. per hour Carter Seco ‘ ; hours i fase as and hour (Meeh) IWRACOHO wid Siete ere enone 6 Neilp4 5.649 1.94 23.67 1 A: SS 8 al eet ate 6 0.85 6.724 1.46 20.43 VUES Ott entra a aA tt 31 0.66 6.255 2.22 26.54 1 Dee: BY: A Cea eS 31 0.67 8.704 1.94 26.87 Weal aircraninuetan © 80 0.70 5.972 2.18 25.57 BSA D eerie cch eet ek 80 0.73 7.101 1.66 22.67 Wr aelDiees terres 104 0.70 §.252 1.83 21.85 1538 Ws BO 7 Ce Baresi LAS OS a 104 0.73 7.500 erie 23.47 We Averages vce cle 5.782 2.04 24.43 IBS Averapelijcrs cnn 7.514 1.70 23.36 Benedict and Talbot '°® explained such differences as this on the assumption that fat replaces active tissue. They said, therefore, that the lean infant has a higher metabolism per unit of weight than the fat one because he has relatively more active tissue. It turns out, however, as we were able to show, that fat does not replace active tissue but replaces water. Hence, we are driven back upon the old explanation which Rubner himself gave ; METABOLISM 239 namely, that the lean infant has the higher metabolism because he loses heat faster. He has a larger surface in proportion to weight and, since it is through the surface that heat is lost, it will be in proportion to surface that heat must be produced if the body temperature is to remain constant. A comparison of the energy metabolism of infants through the first year of postnatal life made by Benedict and Talbot and by Murlin and Hoobler 1*° reveals a rapidly progressing increase. Starting at a level below that of the adult, the nursling reaches the adult level at about the second month, and from this time on, while traversing the period of most rapid growth, the period of highest milk consumption, it arrives at the apex of the metab- olism curve, somewhere between one and two years. From this point on to old age (with the exception of a slight mound at the time of puberty) the rate of oxidation in the resting body is steadily receding. This lecture opened with emphasis upon the independence of the embryo. The enzymes which enable it to secure materials for its own nourishment from the mother are really a part of the mechanism of heredity. After producing the ovum the mother has no further influence on the hereditary factors. The enzymes of the embryo, however, can act only on certain proteins—the proteins of its own species. After a period which may well be called parasitism the new and the old organisms become accommodated one to the other and enjoy a period of what Bar has denominated ‘‘ harmonious symbiosis.’’ The harmony applies to both substance and energy metabolism. The maternal metabolism is nicely adapted to the physiological alteration due to pregnancy and the fetus, with all its adnexa, asks for no more in the way of energy than does the same weight of maternal tissue. We conclude now by ealling attention once more to signs of. an independent behavior in the metabolism of the offspring. The low rate immediately after birth is probably due to the fact that the heat regulating mechan- ism is not yet complete; the higher rate beyond the second or third month is doubtless related to the more active growth. 240 HARVEY SOCIETY ee oe BaSon BOER? Ace 8000 DL, BESS eeee eh CORE SCC Bro ORS 4 ORS pale fod tahel patel LYS ei Tes Ty aa: Fic. 10.—Chart from Murlin and Hoobler, showing relation of heat pro- duction to body weight in infants. Methods of von Pettenkofer or Regnault- Reiset. Via. duction to skin surface in infants. Methods of yon Pettenkofer or Regnault- Reiset. 16 METABOLISM 241 tos t besten l ot ht Ped PTR HOP VAP Baa Pt ele AT aT rRweCee a Fra Pee eyes HES BEB aS esa bt i a Fa a Fam a2 es on F amount (Coma eT TT TY TI epee fm fp | fa a ipA tit ttt tt tt AOUBaS Ht) ah a aa es | LEREDEEDe maha eT oF anes ho TTT T Teal ee Ge BREE DeBA DAB ITT) Oe rae Pett yt tet En Reeeee Mea et tT BEE EEE EE CEL He Ge eee eee 11.—Chart from Murlin and Hoobler showing relation of heat pro- 242 HARVEY SOCIETY LITERATURE +Weismann, August: The Germ Plasm, English Trans., London, 1893. 2 Hatschek, B.: Lehrbuch der Zoologie, Jena, 1888, p. 207. ® Boveri, T.: Die Entstehung des Gegensatzes zwischen den Geschlecht- zellen und den somatischen Zellen bei Ascaris megalocephala, Sitz. Ges. f. morph. Physiol., Miinchen, 1892, vol. 8. “Hicker, U.: Die Keimbahn von Cyclops, Arch. f. mikr. Anat., 1897, 45. ®° Hegner, R. W.: The History of the Germ Cells in Insects with Special Reference to the Keimbahn Determinants Jour. Morph. | 1914, 25, 375. ®*Hubrecht, A. A. W.: The Placentation of Erinaceus Europaeus, Quart. Jour. Micr. Sci., 1889, 30; also, Early Ontogenetic Phenomena in Mam- mals, Ibid., N. S. 1908, 53, 1. ty. Spee Graf: Neue Beobachtung tiber sehr friihe Entwickelungstufen des menschlichen Eies, Arch. f. Anat. u. Physiol., Anat. Abth., 1896. * Peters: Ueber die Einbettung des menschlichen EHies, Leipzig and Vienna, 1899. * Hitschmann, F., and Lindenthal, O. Th.: Ueber das Wachstum der Pla- zenta, Zentralbl. f. Gyniik., 1902, 1167. %? Bryce and Teacher: Contributions to the Study of the Early Development and Imbedding of the Human Ovum, Glasgow, 1908. “ Herzog, M.: A Contribution to our Knowledge of the Earliest Known Stages of Placentation and Embryonic Development in Man, Am, Jour. Anat., 1909, 9, 361. ”% Johnstone, R. W.: Contributions to the Study of the Early Human Ovum, Jour. Obst. and Gynec., 1914, 25, 231. 4 Lane-Claypon, J.: On Ovogenesis and the Formation of the Interstitial Cells of the Ovary, Jour. Obst. and Gynec., 1907, 11, 205. %* Beard, J.: The Morphological Continuity of the Germ Cells of Raja batis, Anat. Anz., 1900, 18, 465. “a Emrys-Roberts, E.: A Preliminary Note upon the Question of the Nutri- tion of the Early Embryo with Special Reference to the Guinea-Pig and Man, Proc. Roy. Soc., London, B, 1905, 76, 164. ** Webster: Human Placentation, Chicago, 1901. * Young, James: Reproduction in the Human Female, Edinburgh and London, 1911. 7 Cohnstein and Zuntz: Untersuchungen iiber das Blut, den Kreislauf und die Athmung beim Siiugethier-Ftus, Pfliiger’s Arch., 1884, 34, 173; Weitere Untersuchungen, Ibid., 1888, 42, 342. * Morriss) Wm. H.: The Obstetrical Significance of the Blood Sugar with Special Referenec to the Placental Interchange, Johns Hopkins Hosp. Bull, 1917., 28, 140. METABOLISM 243 *Chipman, Walter: Observations on the Placenta of the Rabbit with Special Reference to the Presence of Glycogen, Fat and Iron, Labora- tory Reports, Roy. Coll. Phys., Edinburgh, 1903, 8, being studies from the Royal Victoria Hospital, Montreal, vol. 1, No. 4. 7° Lochhead and Cramer: The Glycogenie Changes in the Placenta and the Fetus of the Pregnant Rabbit, Proc. Roy. Soc., London, B, 1908, g0, 263. 7 Goldman, E. E.: Die iiussere und innere Sekretion des gesunden und kranken Organismus im Lichte der “ vitalen Firbung.” II. Zur Bio- chemie der Plazenta, Beitr. z. klin. Chir., 1912, 78, 8. Driessen, L, F.: Ueber Glykogen in der Plazenta, Arch. f. Gyniik., 1907, 82, 278. * Mendel and Leavenworth: Chemical Studies on Growth, Am. Jour. Physiol., 1907, 20, 117; also Adamoff, W.: Ein Beitrag zur Physiologie des Glykogens, Zeitschr. f. Biol., 1905, 46, 281. * See Frank, R. T.: An Experimental Study of the Placenta under Physio- logical and Pathological Conditions, Surg., Gynec. and Obst., 1912, 15, 558. * Ahifeld: Zur Frage iiber den Uebergang geformter Elemente von Mutter auf Kind, Zentralbl. f. Gynik., 1877, 1, 265. *Thiemich, M.: Ueber die Herkunft des foetalen Fettes, Centralbl. f. Physiol., 1898, 12, 850. 7 Oshima, T.: Ueber das Vorkommen von ultramikroskopischen Teilchen im fétalen Blute, Centralbl. f. Physiol., 1907, 21, 297. * Gage, S. H., and Gage, S. P.: Coloration of the Milk in Lactating Animals and Staining of the Growing Adipose Tissue in the Suckling Young, Anat. Record, 1909, 3, 203. * Mendel, L. B. and Daniels, Amy L.: Behavior of Fat Soluble Dyes and Stained Fat in the Animal Organism, Jour. Biol. Chem., 1912, 13, 71. 2° Slemons, J. Morris: Cholesterol in the Blood of Mother and Fetus. A Preliminary Note, Am. Jour. Obst., 1917, 75, 569. ** Bailey and Murlin: Unpublished. 37 Hofbauer, J.: Die menschlichen Plazenta als Assimilations-organ, Samml. klin. Vortr., 1907-1909, N. F. Gynik, 165 to 194, 19 to 34. 83 Griifenburg, E.: Beitriige zur Physiologie der Einbettung, Zeitschr. f. Geburtsh. u. Gyniik., 1909, 65, 1. * Mathes, P.: Ueber Autolyse der Plazenta, Zentralbl. f. Gyniik., 1901, 1385. Basso, G. L.: Ueber Autolyse der Plazenta, Arch. f. Gyniik., 1905, 76, 162. 5° Morse, A. H.: The Amino-Acid Nitrogen of the Blood in Cases of Normal and Complicated Pregnancy, and also in the New-Born. J. H. H. Bulletin, June, 1917. * Veit and Scholten: Synzytiolyse u. Hiimolyse, Ztschr. f. Geburtsh. u. Gyniik., 1903, 49, 210. 244 HARVEY SOCIETY * Hofbauer, J.: Grundziige einer Biologie der menschlichen Plazenta, Leip- zig, 1905. *° See Jenkenson: Notes on the Histology and the Physiology of the Pla- centa in Ungulata, Proc. Zoél. Soc., London, 1906, vol. 1, quoted by Marshall, F. H. A.: Physiology of Reproduction, London, 1910, p. 401. “Heape, Walter: Preliminary Note on the Transplantation and Growth of Mammalian Ova within a Uterine Foster-mother, Proc. Roy. Soc., London, 1890, B 48, 457. “Castle, Wm. E.: Heredity in Relation to Evolution and Animal Breeding, New York, 1911, p. 31. “Loeb, J.: Weitere Beobachtungen iiber den Einfluss der Befruchtung, etc., Biochem. Ztschr., 1906, 2, 34. * Robertson, T. B.: On the Normal Rate of Growth of an Individual and Its Biochemical Significance, Arch. f. Entwceklngsmechn. d. Organ., 1908, 25, 581 and 26, 108. “Loeb and Chamberlain: An Attempt at a Physico-chemical Explanation of certain groups of fluctuating Variations, Jour. Exp. Med., 1915, 19, 559. * Riddle, O.: Our Knowledge of Melanin Color Formation and its Bearing on the Mendelian Description of Heredity, Biol. Bull., 1908, 16, 316. “° Goldschmidt, R.: Genetic Factors and Enzyme Reactions, Science, Jan. 21, 1916. *" Reichert, Science, 1916. * Abderhalden, E.: Synthese von Polypeptiden, Peptonen und Proteinen mittels Ferments, Fermentforschung, 1914, 1, p. 47. * Sommerfield: Arch. f. Kinderh., 1900, 30, 253. ° Michel: Sur la composition de ’embryon et du foetus humain aux dif- ferentes epoques de la grossesse., Compt. rend. Soc. Biol., 1899, 422. * Reprew, A. V.: Die Einwirkung der Schwangerschaft auf den Stoffwechsel bei den Thieren, Wratsch, 1888, No. 16; also Innaug. Dissert., St. Petersburg, 1888. Ver Ecke: Lois des Echanges materiels dans leurs Rapports avee les Phases de la Vie sexuelles, Mem. Cour. et autres Mem., Acadm. roy. de Med. de Belgique, Brussels, 1900, 15, 7. “Hagemann, O.: Zur Kenntniss des Kiweissumsatzes) im Thierischen Organismus, Innaug. Dissert., Erlangen, 1901. * Jiigerroos, B. H.: Der Eiweiss-, Phosphor- und Salzumsatz wiihrend der Graviditiit, Arch. f. Gyniik., 1902, 67, 517. Bar and Daunay: Stoffwechsel wiihrend der Schwangerschaft, Jour. de physiol. et de path, gen., 7, 832. % Murlin, J. R.: Metabolism of Development. I. Nitrogen Balance during Pregnancy and Menstruation of the Dog, Am. Jour. Physiol., 1910, fey MEE METABOLISM 245 7 Gammeltoft, S. A.: Untersuchungen iiber den Stickstoffwechsel wiihrend der Graviditit, Skan. Arch., 1913, 28, 325. * Landsberg, E.: Eiweiss und Mineralsto{fwechsel bei der Schwangeren Frau nebst Tierversuchen mit besonderer Beriicksichtigkeit der Funktion endokrinen Driisen, Ztschr. f. Geburtsh, u. Gyniik., 1914-15, 76, 53. ® Wilson, K. M.: Nitrogen Metabolism during Pregnancy, Bull. Johns Hopkins Hosp., 1916, 26, 121. *° Bonnet, Anatomische Hefte, Ier Abth, 1902, 20, 477. * Young, James: The Etiology of Eclampsia and Albuminuria and their Relation to Accidental Hemorrhage, Jour. Obst. and Gynec., 1914, 26, 1. * Ewing and Wolf: The Clinical Significance of the Urinary Nitrogen. I. The Metabolism in the Toxemia of Pregnancy, Am. Jour. Obst., 1907, 55, 289. ® Lossee and Van Slyke: The Toxemias of Pregnancy, Am. Jour. Med. Sc., 1917, 153, 94. * Hoffstrom, K. A.: Stoffwechseluntersuchung wiihrend der Schwanger- schaft, Skan. Arch. f. Physiol., 1910, 23, 326. ® Herrmann, E.: Ueber die wirksame Substanz im Eierstocke und in der Placenta, Monatschr. f. Geburtsh. u. Gyniik., 1915, 41, 1. *Massin, W. N.: Intermediiire Stoffwechselprodukte des Ursache der Eklampsie, Zentralbl. f. Gyniik., 1895, 1105. * Whitney and Clapp: Urine Changes in Pregnancy and Puerperal Eclamp- sia, Am. Gyncology, 1903, 3, 121. *s Mathews, Frank S.: The Urine in Normal Pregnancy, Am. J. Med. Sc., 1906, 131, 1058. ® Edgar, J. Clifton: Clinical Manifestations of the Toxemia of Pregnancy, N. Y. Med. Jour., 1906, 83, 897 and 956. * Murlin, J. R.: Metabolism of Development. III. Qualitative Effects of Pregnancy on the Protein Metabolism of the Dog, Am. Jour. Physiol., 1911, 28, 422. ™ Murlin, J. R. and Bailey, H. C.: Further Observations on the Metabolism of Normal Pregnancy, Arch. Int. Med., 1913, 12, 288. %Slemons, J. Morris: Metabolism during Pregnancy, Labor and the Puer- perium, Johns Hopkins Hosp. Reports, 1904, 12, 111. Falk and Hessky: Ueber Ammoniak, Aminosiiuren und Peptidstickstoff in Harn Gravider, Ztschr. f. klin. Med., 71, 261. ™Murlin, J. R.: Some Observations on the Protein Metabolism of Normal Pregnancy and the Normal Puerperium, Sur., Gynec. and Obst., 1913, Jan., p. 43. * Hasselbalch, K. A., and Gammeltoft, S. A.: Die Neutralitiitsregulation des graviden Organismus, Biochem. Ztschr., 1915, 68, 206. * Van Hoogenhuyse and ten Doeschate: Recherches sur les Echanges organiques ches les Femmes encientes, Ann. de gynec., 1911, 2d ser. 8, pp. 17 and 97. 246 HARVEY SOCIETY ™ Heynemann, Th.: Zur Frage der Leberinsuffizienz und des Kreatinin- stoffwechsels wiihrend der Schwangerschaft und bei den Schwanger- schaftstoxikosen, Ztschr. f. Gebrutsh. u. Gynik., 1912, 71, 110. * Krause, R. A.: On the Urine of Women under Normal Conditions with special reference to the Presence of Creatin. Quart. Journ. Exp. Physiol., 1911, 4, 293. * Underhill and Rand: The Peculiarities of Nitrogenous Metabolism in Pernicious Vomiting of Pregnancy, Arch. Int. Med., 1910, 5, 61. *° Leimdorfer, Novak and Porges: Ueber die Kohlensiiurespannung des Blutes in der Graviditit, Ztschr. f. klin. Med., 1912, 75, 301. *tSlemons, J. Morris: The Involution of the Uterus and its Effect upon the Nitrogen Output of the Urine, J. H. H. Bulletin, 1914, 25, 195. * Longridge, C. Nepean: Excretion of Creatinin in Lying-in Women, with some Remarks on the Involution of the Uterus, Jour. Obst. and Gynec. Brit. Emp., 1908, 13, 420. ® Saiki: Jour. Biol. Chem., 1908, 4, 483; also Buglia and Constantino, Ztschr. f. physiol. Chem., 1912, 81, 122. * Shaffer, P.: The Excretion of Creatinin and Creatin in Health and Dis- ease, Am. Jour. Physiol., 1908, 23, 1. * Murlin, J. R.: Protein Metabolism in Development, Am. Jour. Physiol., 1908, 23, Proceedings xxxi. *® Mellanby, E.: The Metabolism of Lactating Women, Proc. Roy. Soc., London, B, 1913, 86, 88. §™ Morse, A.: Jour. of A. M. A., 1915. * Warburg, O.: Beitriige zur Physiologie der Zelle insbesondere iiber die Oxydationsgeschwindigkeit in Zellen, Ergeb. d. Physiol., 1914, 14, 253. * Meyerhof, O.: Untersuchungen iiber die WiirmetOnung der vitalen Oxyda- tions v orgiinge in Eiern, Biochem. Ztschr., 1911, 35, 246. * Lyon, E. P.: Rhythms of Susceptibility and of Carbon Dioxide Production in Cleavage, Am. Jour. Physiol., 1904, 11, 52. “Warburg, O.: Notizen zur Entwickelungs-physiologie des Seeigeleies, Pfliiger’s Arch., 1915, 160, 324. * Liebermann, L.: Embryochemische Untersuchungen, Pfliiger’s Arch., 1888, 43, 71. * Tangl, F.: Beitriige zur Energetik der Ontogenese, Pfliiger’s Arch., 1902-3, 93, 362. * Bohr, Chr., and Hasselbalch, K.: Ueber die Kohlensiiureproduction des Hiihnerembryos, Skan. Arch. f. Physiol., 1900, 10, pp. 149, 353; also Bohr, C.: Ueber den respiratorischen Stoffwechsel beim Embryo kalt- bliitiger Thiere, Jbid., 1903, 15, 398. * Bohr, C.: Der respiratorische Stoffwechsel des Siiugethierembryo, Skan. Arch, f. Physiol., 1900, 10, 413. * Carpenter, T. M., and Murlin, J. R.:: Energy Metabolism of Mother and Child just before and just after Birth, Arch. Int. Med., 1911, 7, 184. METABOLISM Q47 * Hasselbalch, K. A.: Ein Beitrag zur Respirationsphysiologie der Gravi- ditit, Skan. Arch. f. Physiol., 1912, 27, 1. * Murlin, J. R.: Total or Energy Metabolism in Development, Am. Jour. Physiol., 1908, 23, Proceedings, p. xxxii. ” Hasselbalch, K. A.: Respfrationsférség-paa-nyfodte Bjorn, Bibliotek for Laeger, Copenhagen, 1904, translated and quoted at length in Benedict and Talbot: The Physiology of the Newborn Infant, Carnegie Institution of Washington, No. 233, 1915, p. 15. 100 Weiss, G.: Bull. de Acad. de méd., 1908, 60, 3d ser., 458. 1 Bailey, H. C., and Murlin, J. R.:: The Energy Requirement of the New- born, Am. Jour. Obst., 1915, 71, No. 3. 2 Benedict and Talbot: The Physiology of the Newborn Infant, Carnegie Institution Pub. No. 233, 1915. 13 Rubner, M.: Das Wachstumsproblem und die Lebensdauer des Man- schens, Arch. f. Hyg., 1908, 66, 180. 1 Oddi and Vicarelli, Influence de la Grossesse sur l'ensemble de l’echange respiratoire, Archives italiennes de biologie, 1891, 15, 367; see also Centralbl. f. Physiol. 1891, 5, 602. 15 Magnes-Levy: Stoffwechsel und Nahrungsbedarf in der Schwangerschaft, Ztschr. f. Geburtsh. u. Gyniik., 1904, 52, 116. % Murlin, J. R.: The Metabolism of Development. I. Energy Metabolism in the Pregnant Dog, Am. Jour. Physiol., 1910, 26, 134. 7 Zuntz, L.: Respiratorischer Stoffwechsel und Atmung wiihrend der Graviditat, Arch. f. Gynik. 1910, 90, 452. 1® Tusk, G.: Animal Calorimetry, fifth paper. The Influence of the Inges- tion of Amino Acids upon Metabolism, Journ. Biol. Chem., 1912, 13, 155. 1° Benedict and Talbot: The Gaseous Metabolism of Infants, Carnegie Institution Pub. No. 201, 1914. “9 Murlin, J. R., and Hoobler, B. R.: The Energy Metabolism of Ten Hos- pital Children between the Ages of Two Months and One Year, Am. Jour. Dis. Child., 1915, 9, 81. CARDIAC DYSPNEA* FRANCIS W. PEABODY, M.D. Boston ACH of the milestones reached in the continually advancing progress of clinical medicine corresponds closely to some forward step taken in what have come to be known as the “‘ funda- mental sciences.’’ A new technical method or a new point of view which opens a fresh way of approach in anatomy, physiology, chemistry or biology is quickly seized upon by the physician in the hope that it may prove to be an addition to his armamentarium which will aid him to gain new knowledge of disease—its mechan- ism—its recognition—and its cure. The era inaugurated by Virchow gave to us as accurate a conception of the pathological morphology of the commoner diseases as the methods thus far developed would allow, and the last decade in clinical medicine has belonged essentially to biology and physiology. The study of dead form has largely given way to the study of living processes, —the growth of microorganisms and the abnormalities of function produced in cells and organs under various conditions of disease. The significant role played by physiology is manifest in many fields of clinical medicine, and the application of the methods and instruments of the physiological laboratory to the study of patients in the wards has broadened and in some instances revo- lutionized our conception of human pathology. In seareely any field has this affiliation between physiology and clinical medicine produced more interesting and stimulating results than in the study of the respiration. The application of modern methods permits the accurate determination of oxygen consumption, carbon dioxid production, the respiratory quotient, and heat production. With their aid we are rapidly gaining * From the Medical Service of the Peter Bent’ Brigham Hospital, and the Department of Medicine, Harvard Medical School. Delivered March 17, 1917. 248 CARDIAC DYSPNEA 249 insight into the more fundamental changes of the intermediary metabolism which are met with in disease. The adaptation of recent physiological researches on the chemical control of the respiratory centre has led directly to the use of methods for determining the carbon dioxid content of the alveolar air in the study of disease. From this, and from analogous methods, we have learned much about those pathological conditions in which acidosis is a significant feature. Neither in the study of the gaseous exchange nor of the alveolar air, however, is the interest focused primarily on the respiration itself. Just as the urine is of value in the investigation of nitrogenous metabolism because of the end products which it contains, so the expired air and the alveolar air are chiefly of interest because they serve as indices of the intermediary metabolism. The respiration itself, the vari- ous forms which may occur in disease, the factors which may influence it and limit its efficiency,—these have occupied com- paratively little attention. It is to some of these changes, and more especially to a consideration of the causes of the dyspnea which occurs in association with heart disease that the present paper is directed. Before proceeding to a discussion of those factors which enter into the production of dyspnea, it will be well to state briefly the exact significance of the term itself. As ordinarily used the word is applied loosely to various abnormal types of respiration. Thus not infrequently rapid breathing or tacchypnea is referred to as dyspnea, while even more often the increase of rate and depth of respiration which constitutes hyperpnea is character- ized as dyspnea. Neither condition is, however, necessarily synonymous with dyspnea. Dyspnea, as the derivation of the word indicates, is a difficult or labored breathing, and there is implied in it an element of subjective discomfort. Hyperpnea on the other hand, merely signifies an increase above the normal value for the subject at rest in the volume of air breathed. Such an increase of the pulmonary ventilation, or as it is com- monly called, of the minute-volume of air breathed may be due to a more rapid respiration or to a deepening of the respiration, but usually both factors take part in it. Whether or not in any 250 HARVEY SOCIETY given instance the hyperpnea will amount to a true dyspnea depends on the degree to which the pulmonary ventilation is increased, and on the ability of the subject to raise his minute- volume to that degree easily. As will be seen, anything which prevents a person from increasing his pulmonary ventilation in a normal manner, will be an element in increasing his tendency to dyspnea. It is extremely difficult to analyze with accuracy the funda- mental cause of the subjective sensation which we know as dyspnea. How much is it due to fatigue of the muscles of respi- ration? How much is it due to a functional insufficiency of the respiration resulting in an inadequate oxygen supply to the tis- sues, and an incomplete removal of the waste products of metab- olism? Without doubt both factors are involved and one is con- fronted by a vicious circle, in which waste products accumulated in the cells and blood augment the stimulus to the respiratory centre, and this in turn makes still greater demands on the already tired muscles of respiration. In a general consideration of the respiration it is customary to subdivide the subject into two broad phases,—the external respiration, and the internal respiration. The former depends largely on the lungs, and the essential feature of it is that the pulmonary ventilation shall be such as to supply oxygen to the blood in the amounts required by the metabolism of the body, and to provide for the proper removal of the waste carbon dioxid. The internal respiration in which the circulation plays a promi- nent role, is concerned with the exchange of gases between the blood and the cells of the body. It is clear that if either the external or the internal respiration is inadequate to the task imposed upon it, dyspnea may result. Even when the external respiration produces a blood which is wholly normal as it leaves the lungs, there may be an improper gaseous exchange between the blood and the tissues owing to an imperfect internal respira- tion. Of the internal respiration, which is possibly the more fundamental phase of the respiration, physiologists and chemists know but little, and of its pathology clinicians know, if anything, somewhat less. The methods for studying even so gross a feature CARDIAC DYSPNEA 251 as the rate of the blood flow are still imperfect, and chemical analyses are limited to blood from the peripheral vessels. The whole field of the internal respiration must, therefore, for the: present, be left open, and we shall be restricted to a discussion of the conditions affecting the external pulmonary respiration. One of the chief factors which have aroused interest in the study of the respiration in disease has been the recent advance made in our knowledge concerning the normal control and regu- lation of the respiration. The old discussion among physiologists as to the nature of the stimulus to the respiration was in a large degree settled by the classical paper of Haldane and Priestley * which showed that carbon dioxid is the essential stimulus, and indicated the extreme sensitiveness of the respiratory centre in that a rise of 0.2 per cent of the carbon dioxid content of the alveolar air caused the ventilation to be doubled. Subsequent investigations have tended to broaden this conception and to Winterstein * and Hasselbalch * is due the chief credit of demon- strating that the respiratory centre responds not to carbon dioxid alone but to any increase of the acid radicles in the blood. Since the presence of carbon dioxid and of other acids in the blood depends in general on the chemical processes in the body, it is evident that the basic factor in the regulation of the respira- tion is the metabolism. The respiratory centre controls the move- ments of the lungs and regulates them so that the pulmonary ven- tilation keeps pace with the metabolism. In a normal individual at rest, a minute-volume of approximately 5.0 liters of air suffices to remove the excess of carbon dioxid, and to supply sufficient oxygen for the needs of the body. If, however, the subject walks about the room his metabolism rises, more carbon dioxid is formed, the respiratory centre is more highly stimulated and the pulmon- ary ventilation is increased. The rise in metabolism associated with the walking may require an increase of the minute-volume of air breathed to three or four times its resting value in order that the needs of the tissues for a proper gaseous exchange may 1Haldane and Priestley: Jour. of Physiol., 1905, xxxii, 225. ?Winterstein: Arch. f. d. ges. Physiol., 1911, exxxviii, 167. *Hasselbalech: Biochem, Ztschr., 1913, xlvi, 403. 252 HARVEY SOCIETY be met. Such an increase in minute-volume is easily brought about by increasing the rate and depth of breathing and, indeed, a normal person is hardly conscious of any change in his respira- tion when he is breathing 15.0 liters a minute. With severe exercise the metabolism rises much higher, and in addition to the carbon dioxid formed, Ryffel* has shown that lactie acid may be produced. Here, then, is an additional stimulus to the respira- tory centre. In an attempt to determine how great a pulmonary ventilation normal persons were capable of, a series of observa- tions have been made in association with Mr. F. C. Hall and Miss B. I. Barker. The experiments consisted in having young men,—doctors and medical students,—ride on a stationary bicycle until they were forced to stop on account of shortness of breath. Some of the subjects were athletes in excellent training while others were accustomed to a sedentary life. The subjects breathed through mouth-pieces, and valves were used to separate the in- spired from the expired air. The expired air was passed through a Bohr air meter, and its volume measured for each half minute of the time during which the subject was riding. The rate of the respiration was counted from a continuous pneumographie record. The data obtained over each half minute consisted of the respiratory rate, the total volume of air breathed, and the average volume of each individual respiration. While there was a certain amount of variation among the different individuals in that some tended to greater increase of rate and others to greater increase of the depth of breathing, a number of interest- ing facts were elicited. Over the last minute and one-half of the ride, thus when dyspnea was most marked, and just before having to stop, the minute-volume of air breathed ranged from 47.6 to 80.0 liters. The larger minute-volumes were, of course, in general found in the larger individuals. Comparing these figures with the minute-volume at complete rest, it is found that these normal subjects could increase their pulmonary ventilation on an aver- age of 10.7 times above the resting value. This gives a fairly accurate idea as to the great adaptability of the respiratory ‘ Barcroft: The Respiratory Function of the Blood. Cambridge, 1914, p- 239. CARDIAC DYSPNEA 253 mechanism to any demands that may be put on it. and one has a quantitative value for what we may call the ‘‘pulmonary re- serve.’’ Since the high minute-volume depends on an ability to increase the rate and especially the depth of respiration, it is not surprising to find a close relation between the highest minute- volume and the vital capacity, or the volume of air which can be expired after the greatest possible inspiration. There is also a relation between the volume of the individual respiration and the vital capacity, and it is rather striking that the deepest respirations while riding averaged only 33 per cent of the vital capacity. Curiously enough no definite differences were observed with regard to the respiratory mechanism between the trained and the untrained subjects. A point of considerable interest was the great difficulty experienced in making the subjects highly dyspneic, because they tended to stop riding on account of mus- cular fatigue rather than on account of shortness of breath. This was in part due to the fact that they were using muscles un- accustomed to heavy work, but it showed that in general the respiratory mechanism can normally adapt itself to any grade of metabolism that the body can produce. Normally then, ‘‘the pulmonary reserve’’ is so great, and the minute-volume of air breathed can be so easily raised to many times the volume at rest, that dyspnea is only noticeable under conditions of rather severe exertion. What, however, are the factors which tend to decrease the ‘‘pulmonary reserve’’ or to make a person more readily subject to dyspnea? What must one consider as possible elements in the cause of any pathological dyspnea? Since the ‘‘pulmonary reserve’’ depends on the rela- tion between the minute-volume of air breathed at rest and the highest minute-volume which the subject is capable of breathing, it will be greatest if the minute-volume at rest is low. Thus, first among the factors which may cause an abnormal tendency to dyspnea are those conditions which produce a high minute- volume at rest. Chief among these are an increase of metabolism and the presence of an acidosis. Secondly there are the factors which limit the ability of the subject to meet a demand for a higher pulmonary ventilation. Since an increase in minute- 254 HARVEY SOCIETY volume depends on an increase of rate and depth of respiration it is evident that a high initial respiratory rate and more espe- cially, anything which interferes with deep breathing will tend to reduce the pulmonary reserve. Finally it will be seen that still other conditions probably underlie the type of dyspnea which is associated with periodic breathing. From this point of view, then, we may approach the question of heart disease in an attempt to determine whether or not these possible factors are present, and in how far they may be con- sidered as elements in the production of dyspnea. It is important to appreciate that dyspnea is one of the commonest symptoms met with in patients suffering from cardiac disorders, and that it appears in a considerable variety of clinical conditions. We shall, therefore, expect to find that the causes of dyspnea are not necessarily the same in different cases, and that while the symp- tom has a comparatively simple basis in certain instances, in others it is complex and depends on a number of interacting factors. The dyspnea which is noticed on ascending stairs by a subject with a compensated valvular lesion is quite a different thing from the continuous dyspnea of the same person when in a state of acute decompensation, and this in turn may have different underlying elements from the dyspnea of the patient with cardio-renal disease, or the nocturnal attacks of paroxysmal dyspnea seen in an old man with chronic myocarditis. Let us first consider the question of the metabolism in cardiac disease. The most satisfactory study of the basal metabolism in patients with heart disease was carried out at Bellevue Hospital, New York, in the calorimeter of the Russell Sage Institute of Pathology by DuBois and Meyer in an investigation in which it was my privilege to take part.’ Of fundamental importance was the demonstration by means of the close agreement between the methods of direct and indirect calorimetry, as well as by the finding of respiratory quotients which were within the normal limits, that the intermediary metabolism in heart disease follows a normal course. Sixteen patients were studied. The results showed that in compensated cardiac disease the metabolism is s'Peabody. Meyer and DuBois: Agen. Int. Med., “1916, XVii, 980. CARDIAC DYSPNEA 255 perfectly normal. Of twelve patients, on the other hand, who had some degree of dyspnea at the time they were studied, three showed a normal metabolism, and nine a metabolism that was distinctly above normal. In five of the latter the metabolism was increased from 25 to 50 per cent above the normal. The cause of the rise in metabolism is not evident. These, and other more recent observations from the same source ® indicate that it 1s not a necessary accompaniment of dyspnea, that it cannot be attributed to acidosis, and that it bears no definite relation to the level of the nitrogen in the blood. The subject has been further investigated in the Medical Laboratory of the Peter Bent Brigham Hospital * in association with Dr. J. A. Wentworth and Miss B. I. Barker. The indirect method of calorimetry was used, the apparatus consisting essentially of a large Tissot spiro- meter for the collection of the expired air and the Haldane Port- able Gas Analysis apparatus. By this method data are obtained regarding the minute-volume of air breathed which are lacking in the observations made with the large bed calorimeter. The re- sults of the metabolism determinations in 24 instances agree essentially with those at the Sage Institute. They confirm the fact that in persons with mild grades of heart disease, in whom the lesion is comparatively well compensated, the metabolism is within normal limits, and they demonstrate again that in more severe cases, with or without dyspnea at the time of observation, the metabolism is variable, being frequently normal, but in some instances as much as 40 per cent above normal. In only two cases was the heat production more than 25 per cent above the normal however, and in general, the rise in basal metabolism is neither a constant, nor a particularly significant feature. Of more immediate interest in the study of dyspnea are the observa- tions on the minute-volume of air breathed. These show that while patients with mild cardiac lesions, and only a slight tendency to dyspnea breathe a normal minute-volume of air, usually between 5.0 and 6.0 liters, the more severely affected patients who are either dyspneic while at rest, or who become so on very slight > Aub and DuBois: Arch. Int. Med., 1917, xix, 865. i ‘Peabody, Wentworth and Barker: Arch. Int. Med., 1917, xx, 468. 256 HARVEY SOCIETY exertion, tend to have a considerably higher minute-volume. In this group of subjects the minute-volume at rest ran as high as 11.6 liters while the average in 12 patients was 8.22 liters. There is, moreover, no definite relation between the minute volume and the metabolism, and a high minute-volume may be found in a subject whose basal metabolism is wholly normal. A similar increase in the minute-volume has been reported by Beddard and Pembrey * and by other observers. As to the cause of this increased minute-volume associated with a normal metabolism we have no absolute proof, but there is a very suggestive relationship between the raising of the minute- volume and the decrease of the vital capacity of the lungs. Prac- tically all cardiac patients with a vital capacity of less than 60 per cent of the normal (see below) show a high minute-volume and a similar observation has been made in a case of pleural effusion. The decrease in the vital capacity of the lungs is prob- ably associated with a lessening of the area of the respiratory surface, as for instance, by the production of atelectasis by col- lections of fluid in the pleural cavity. The dead space, consisting of the naso-pharynx, trachea, and bronchi, would not necessarily be affected and the resulting decrease in the respiratory surface, with a relative increase in the dead space would bring about a rise in the actual minute-volume of air breathed in order that the alveolar ventilation, which is after all the essential thing, should remain constant. In patients with severe manifestations of cardiac disease, then, an increase of the minute-volume of air breathed while at rest is very commonly present, whether or not there is any associated rise in the basal metabolism. In such cases the high initial minute-volume will be a factor in the production of dyspnea in that it limits the ‘‘pulmonary reserve.’’ By diminishing the difference between the volume of air breathed at rest, and the maximum volume the subject is capable of breathing, it makes him more readily susceptible to the production of dyspnea. Let us turn to the consideration of a second condition which causes an increase in the pulmonary ventilation, and which may CARDIAC DYSPNEA 257 thus act as a factor in the production of dyspnea in much the same manner as an increased metabolism. ‘his is acidosis. The re- spiratory centre is excessively sensitive to a shift in the reaction of the blood, and any considerabie accumulation of acids in ihe blood stream causes a greater activity on the part of the lungs. indeed the production of hyperpnea is perhaps the most char- acteristic effect of acidosis. In the recent enthusiastic attention which clinicians have accorded to the subject of acidosis, the condition has been held responsible for a great variety of symptoms. It is not to be wondered at, then, that the relation of acidosis to dyspnea is a problem which has given rise to much conjecture and to a con- siderable amount of experimentation. Some observers, notably Lewis and his co-workers * regard acidosis as one of the chief factors in the dyspnea seen in elderly persons with weak hearts and usually with kidney involvement,—essentially the cardio- renal group. It is important therefore to examine in some de- tail into the conditions associated with cardiac disease in which acidosis is present, and to consider in how far it may be regarded as responsible for the production of dyspnea. As regards pure cardiac disease, one may state as the result of many observations on the carbon dioxid content of the blood and alveolar air, that there is no evidence indicating the presence of an acidosis in compensated cases. In patients with pure cardiac disease in a state of acute decompensation the question is less simple to answer. Not infrequently the alveolar air analyses show a low carbon dioxid tension, while the blood analyses show a normal or high tension. With the regaining of compensation and usually with the disappearance of continuous dyspnea, so that the patient is comfortable while at rest, the alveolar carbon dioxid rises quickly and the relation between the blood and the alveolar carbon dioxid becomes normal. How is this to be inter- preted? It is possible that in these acutely sick persons the samples of alveolar air are not reliable, but this explanation is hardly satisfactory, and it is much more likely that the condition *Lewis, Ryffel, Wolf, Cotton and Barcroft: Heart, 1913, v, 45. 17 258 HARVEY SOCIETY is a real one. Peters ’° who has studied the question at the Pres- byterian Hospital and who has found the carbon dioxid content of the blood considerably higher than that of the alveolar air, concludes, and most probably correctly, that there is an inter- ference with the passage of carbon dioxid from the blood into the alveolar air. There is thus an accumulation of carbon dioxid in the blood, and an acidosis in which an excess of carbon dioxid is the essential feature. The possibility of the presence of other abnormal acids due to incomplete oxidation, a condition similar to the acidosis of asphyxia, cannot be definitely excluded, but at any rate, in the production of dyspnea in pure cardiac disease acidosis is a factor which only occurs in the most severely de- compensated cases, and its influence in cases which recover is of short duration. In cases of cardiac disease associated with renal insufficiency, on the other hand, the rdle played by acidosis is much more sig- nificant. Sellards*! and Palmer and Henderson ** showed the frequency with which acidosis occurs in chronic nephritis, and Straub and Schlayer 7° described the low alveolar carbon dioxid tension in uremia. Observations in our own laboratory have confirmed this work and helped to indicate the close relationship between acidosis and renal function.'* In general, cases of chronic nephritis with a normal phthalein output show no signs of acidosis; with the failure to exerete phthalein satisfactorily an acidosis develops which shows itself by an increase in the “‘alkali- tolerance test ;’’ and when the phthalein output has fallen to zero, there is often a degree of acidosis sufficient to cause a fall in the carbon dioxid tension of the alveolar air. The recent work of Marriott and Howland ** shows that the acidosis is due to the inability of the kidney to excrete acid phosphate. A study of numerous eases of renal and cardio-renal disease ” Peters: Am. Jour. Physiol., 1917, xliii, 113. "Sellards: Bull. Johns Hopkins Hosp., 1912, xxiii, 289; ibid. 1914, xxv, 14]. Palmer: Med. Communicat. Mass. Med. Soc., 1913, xxiv, 133. ** Straub and Schlayer: Munchen. med. Wehnschr., 1912, lix, 569. * Peabody: Arch. Int. Med., 1915, xvi, 955. *% Marriott and Howland: Arch. Int. Med., 1916, xviii, 708. CARDIAC DYSPNEA 259 shows that in the advanced stages, before and after the onset of uremia, and even just before death, the alveolar carbon dioxid tension is usually not below 25mm. This is in itself not a sufficient drop to cause a marked hyperpnea. Indeed in diabetes the increase in ventilation due to acidosis is not particularly notice- able until the carbon dioxid tension is approximately 15 mm. Considering, therefore, the comparatively mild grade of acidosis usually met with in chronic nephritis, one must hesitate to at- tribute to it too great a significance in the production of dyspnea. Occasional rare cases of nephritis present the clinical picture of coma and air hunger just before death, and simulate diabetic coma. In these the carbon dioxid tension is about 10 mm., and the air hunger may be relieved by alkali. Thus in a very small group of cases the acidosis may be the direct cause of a hyperpnea which is sufficient to produce dyspnea. If, however, the acidosis which is commonly met with in chronic nephritis is not of itself intense enough to cause dyspnea, — it is by no means true that it is a factor to be ignored. Its sig- nificance may be made clear by some experiments carried on at the Peter Bent Brigham Hospital’® which were devised as a means of studying the production of dyspnea in normal subjects and in persons with cardiac disease. In order to avoid the dangers and difficulties attendant on the production of dyspnea in per- sons with heart disease by exercise, and to allow of the investiga- tion of comparatively sick patients in bed, the dyspnea was pro- duced by a continually increasing percentage of carbon dioxid in the inspired air. The subjects breathed through valves sepa- rating the inspired from the expired air. The expired air passed through a plethysmograph which was calibrated so that its move- ments, recorded on the smoked drum of a kymograph, gave an accurate index of the volume of each respiration as well as of the rate of respiration. The total ventilation for each minute could thus be caleulated. After leaving the nlethysmograph the expired air was rebreathed by the subject. The carbon dioxid tension of the inspired air rose progressively during the experiment and its percentage was determined by the analysis of samples taken at % Peabody: Arch. Int. Med., 1915, xvi, 846. 260 HARVEY SOCIETY frequent intervals. As the result of a series of observations it was found that in normal individuals a given percentage of carbon dioxid produced a fairly constant rise in the pulmonary ventila- tion. Thus when the inspired air contained from 4.2 to 5.4 per cent of carbon dioxid the minute-volume cf air breathed was approximately twice what it was at the beginning of the experi- ment. Exactly the same relationship was observed in most patients with cardiac and renal disease. Their response to carbon dioxid fell into the normal limits. In a number of cases, how- ever, in which the alveolar air showed evidence of an acidosis, abnormal findings were met with. Instead of the pulmonary ven- tilation being doubled by 4.2 to 5.4 per cent carbon dioxid it became doubled when only 2 to 3 per cent of carbon dioxid was breathed. In other words these patients were unusually sensi- tive to the stimulus of carbon dioxid, and it required much less than normal to cause a considerable increase of the pulmonary ventilation. That this effect was actually dependent on the acidosis was demonstrated by performing the experiment again after enough alkali had been given to overcome the acidosis and to bring the carbon dioxid tension of the alveolar air back to its normal value. Under these circumstances the patients re- acted just like normal subjects. The explanation of these results is simple. With the development of the acidosis the so-called ‘‘buffer action’’ of the blood becomes diminished and the addition to it of small amounts of carbon dioxid which under normal circumstances would produce little change in reaction, causes enough shift in reaction to stimulate the respiratory centre. It seems fair to conclude from these experiments that while the degree of acidosis which is commonly met with in patients with cardio-renal disease is not sufficient to cause any decided increase in the pulmonary ventilation, nevertheless it may render the patients unusually susceptible to the production of dyspnea, and it is to be regarded as one factor in causing them to become short of breath on exertion. In severely decompensated eases, even the comparatively slight increase in the pulmonary ventila- tion while at rest may be sufficient to make the difference between comfort and discomfort in breathing, There is then a rational CARDIAC DYSPNEA 261 basis for the administration of alkali to patients with acidosis, and in certain cases definite relief of symptoms may be observed. Having discussed brietly the two chief conditions which cause an increase of the pulmonary ventilation let us now turn to the means by which the body responds to a demand for a higher minute-volume of respired air, and consider in what way these may be affected in heart disease. Such au increase in the minute- volume of air breathed is brought about by an increase of tie rate or of the depth of breathing. We may first give attention to the question of the depth of respiration and observe in how far a limitation in the capacity to breath deeply is to be regarded as a factor in the production of dyspnea in heart disease. In the experiments just described in which the subjects were made dyspneic by rebreathing air con- taining increasing amounts of carbon dioxid, one striking diifer- ence was noted between the normal subjects and the patients who had cardiac disease.'’ While the former did not become extremely dyspneic until they were breathing from 60 to 80 liters of air per minute, the latter were forced to stop when they were breath- ing only 20 to 40 liters per minute. A study of the graphic records of the respiration during the experiments showed that this difference depended on the fact that the patients with cardiac disease were unable to increase the depth of their respiration as well as the normal subjects could. It is obvious that anything which prevents a person from breathing deeply is of profound importance as a factor in the production of dyspnea, for it imme- diately limits the extent to which the minute-volume can be raised, and this prevents him from meeting such increases of metabolism as he normally could. The inability to breathe deeply was found to correspond to a decrease in the vital capacity of the lungs. It has long been known '* that the vital capacity of the lungs is often decreased in heart disease, but no particular attention has been paid to the fact. It seemed, however, that the con- dition merited systematic investigation, and in association with Dr. J. A. Wentworth a careful study of the subject has been * Peabody: Arch. Int. Med., 1917, xx, 433. 4% Arnold: Uber die Athmungsgrosse des Menschen. Heidelberg, 1855. 262 HARVEY SOCIETY made.'® The vital capacity of the lungs is the volume of air that can be expired after the deepest possible inspiration. In our experiments the observations were made by having the subject breathe in and out as deeply as possible through a rubber mouth- piece connected with a calibrated recording spirometer. The movements of the spirometer were recorded on the smoked drum of a kymograph, and the vital capacity was determined by meas- uring the length of the line which corresponded to the greatest expiration and inspiration. In order to decide whether the vital capacity of any given patient was normal or not, it was necessary to have standards for comparison, and since no wholly satisfac- tory data were at hand observations were made on a considerable group of healthy persons. Ninety-six normal men and forty-four normal women were studied. It was found that standards which were sufficiently accurate could be established if the results were classified according to sex and according to height. Various other factors which influence the vital capacity of the lungs could be fairly neglected as they were not particularly significant in the group of cases which we have studied. Thus old age causes a de- crease in the vital capacity, but the majority of our patients were at a time of life when this did not play an important part. Athletic training increases the vital capacity but this rarely affected our results, for in pathological cases it is the decrease that is sig- nificant. When placed in their appropriate groups according to sex and height it was found that 134 of the 140 normal subjects had a vital capacity of 90 per cent or more of the normal figure. Having thus established normal standards of the vital capacity of the lungs for men and women of different heights, it was pos- sible to compare with them the results obtained in patients with heart disease. One hundred and twenty-four cases have been studied and about 224 records have been made. It is convenient to classify these patients according to the vital capacity into four groups each of which presents rather definite clinical character- istics, and it will be seen that there is a very close relationship between the decrease in vital capacity and the tendency to dyspnea. Briefly summarized the results obtained are somewhat as follows: * Peabody and Wentworth: Arch. Int. Med., 1917, xx, 443. CARDIAC DYSPNEA 263 TABLE I. oa eC emee (Ms dacupaane | VO TT APR ot Dae aS 90+ 25 0 0 ; 92 TLE CoG Ae Se 70-90 4] 5 2? 54 Aa gee eed wo NN. 40-70 67 17 39 7 TEATS Sea Le Be under 40 723} 61 100 0 Certain cases were tested several times and, owing to changes in the vital capacity they appear in more than one group. In the “ Mortality ” column they are included only in the lowest group into which they fell. “Symptoms of decompensation” indicates dyspnea while at rest in bed or on very slight exertion. Under “ Working” are included only those actually at work, and able to continue. Many other patients in Group II were able to work, but they are not included as they were still in the hospital. Group I consists of 25 cardiac patients in whom the vital capacity was 90 per cent or more of the normal standard. Thus in these cases the vital capacity does not fall below the limits found in healthy persons. All of them had well compensated hearts, and dyspnea was scarcely a more prominent symptom in their histories than it would be found to be in a similar group of normal individuals. About 90 per cent of them were working, ana the others were limited in their activities by cardiae pain or palpitation rather than by dyspnea. They were thus nearly all in extremely good general condition, and in many the cardiac lesion was merely an incidental finding. Group II consisted of 41 eases whose vital capacity was between 70 and 90 per cent of the normal. These patients differed from those of the group with a higher vital capacity in that practically all gave a definite history of dyspnea on any unusual exertion. The majority, however, were able to work, and the rest, with two possible ex- ceptions, could lead a satisfactory, though somewhat restricted life. Several of them had passed through periods of more or less severe cardiac decompensation, and they are to be regarded as borderline cases whose activities must be somewhat limited, but who, under favorable cireumstances, show little evidence of cardiac insufficiency. Group III consists of 67 patients in whom the vital capacity was between 40 and 70 per cent of the normal. 264 HARVEY SOCIETY These cases are much more severely handicapped than are the members of Group II and practically all suffer from dyspnea on moderate exertion. Those with a vital capacity only slightly above 40 per cent are confined to bed or can do little more than get about the house, while those with a vital capacity approaching the upper limits can walk fairly easily, but they usually avoid the stairs or hills. Only 7 per cent of this group were still at work. Attacks of severe cardiac decompensation occur with con- siderable frequency among those patients, and 17 per cent of the number have died. Group IV consists of 23 cases with a vital capacity of 40 per cent or less. All of them were severely decompensated and the majority were confined to bed. Dyspnea is either constantly present or it is produced by the slightest exertion. The prognosis for patients who fall into this group is bad. A few patients whose vital capacity has fallen as low as this during their first attack of decompensation have sub- sequently recovered so that they could lead a fairly active life, but most of them made comparatively little clinical improvement and 61 per cent have died. These observations demonstrate the important rédle played by decrease of the vital capacity in the production of dyspnea in heart disease. In a surprisingly accurate manner the degree to which the vital capacity is decreased corresponds to the tendency to dyspnea. Patients who have no unusual tendency to become short of breath almost invariably have a normal vital capacity, and those who become dyspneic readily have a vital capacity which is depressed in accordance with the severity of the symptom. But what, it may be asked, is the cause of the decrease of vital capacity in heart disease? The answer to this question is that there are many causes, some of which are obvious and easy to appreciate, while others still remain obscure. Anything which interferes with the free movements of the lungs, or the entrance of air into them, will decrease the vital capacity. Thus pleural effusions, fluid in the peritoneal cavity, emphysema and _ pul- monary oedema may be reckoned among the more gross condi- tions affecting it. These and other similar factors seem to explain CARDIAC DYSPNEA 265 the more severely decompensated cases, but there is a large group of patients with slight symptoms in whom the physical examina- tion gives no clew to the reason for a decreased vital capacity. Further investigation into the cause of the decreased vital ca- pacity in these subjects is clearly indicated, and the work of Siebeck 2° points to a promising line of approach. His compre- hensive study of lung volumes in heart disease suggests that the low vital capacity depends on a change in the elasticity of the lungs which results from an engorgement of the pulmonary vessels due to back pressure from the left side of the heart. If this conception is correct then the vital capacity of the lungs is an index of the state of the pulmonary circulation, and as such is of considerable clinical signifiance. It is probable that in many cases the earliest evidence of cardiac insufficiency occurs in the pulmonary circuit but the usual methods of examination afford no means of detecting it. One clinical fact which is quite in accord with the theory that decrease in vital capacity with its attendant dyspnea is associated with a disturbance of circulation through the lungs is the common observation that dyspnea is an earlier symptom in disease of the mitral valves than it is in disease of the aortic valves. If this relation between the vital capacity and the tendency to dyspnea is generally true when one compares a large series of eases with somewhat arbitrarily chosen normal standards it be- comes more so when one follows the individual patient and watches the changes in the vital capacity which are coincident with changes in the clinical condition. As long as the clinical picture remajns constant the vital capacity is found to be the same, but when cardiac insufficiency becomes more marked, and dyspnea more noticeable, the vital capacity falls. Similarly, an improve- ment in the general condition and a lessening of the dyspnea is associated with a rise in the vital capacity. This parallelism is, indeed, so definite that the determination of the vital capacity seems to assume a practical significance. Dyspnea is, of course, only one symptom of heart disease but it is a very common symp- tom, and it is an important one because the degree of dyspnea ®Siebeck: Deut, Arch. f. klin, Med., 1910, ¢, 204, 266 HARVEY SOCIETY or of the tendency to dyspnea is a valuable index of the state of cardiac efficiency. The clinical records of cardiac patients abound with statements about dyspnea but these are always of limited worth for they are based on either the history as given by the patient or on the gross examination of the physician. Dyspnea is, moreover, such a difficult condition to analyze or to describe, that any objective method which allows one to obtain accurate quantitative information regarding it will serve a useful purpose. Such information the determination of the vital capacity appears to furnish even if only in a somewhat rough way. In many in- stances the observations have proved to give a more reliable con- ception of the clinical condition of the patient than has been obtained from either the history or the physical examination. This of course is true only in cases in which dyspnea is the presenting symptom, and does not hold for the group of patients whose cardiac lesion manifests itself by other symptoms such as pain or palpitation. However the latter includes only a relatively small number of cardiac cases, and in a surprisingly large pro- portion of cases records of the vital capacity give important and helpful data as to the present status and the prognosis. They are often of much greater significance than are records of the pulse rate or blood pressure, and they seem to be a useful, although indirect index of the cardiac reserve. But, as we have already seen, the increase of the minute- volume of air breathed which accompanies a rise in metabolism is brought about not only by a greater depth of respiration but also by a higher rate of respiration. What, then, is the relation of rate of respiration to the problem of dyspnea in heart disease? The facts are simple and well known by all, so that the subject may be briefly dismissed. With the exception of the extremely mild eases of cardiac disease, which are in a good state of com- pensation, most instances have a respiration rate which is some- what above normal, and the more severely affected the case the more rapid the respiration. Now there is, roughly speaking, a maximum rate to which the respiration can rise without losing much of its efficiency. The extraordinarily rapid breathing seen in some hysterical patients, is of course economically wasteful. CARDIAC DYSPNEA 267 The maximum efficient respiratory rate will vary considerably in different individuals and under different circumstances, but it is interesting for purposes of comparison to note that the average highest rate of our normal bicyele riders was 34 per minute. Assuming some such figure as this for the high limit of efficient respiratory rate, it is obvious that the individual with a low initial rate while at rest has a marked advantage. The greater the difference between the rate of respiration at rest and the maximum rate of efficient respiration, the greater is the reserve. With an initial rate of ten the respiration rate can be raised more than three times before the maximum of efficiency is reached, but with an initial rate of seventeen it can only be doubled. Thus the high rate of respiration which is found in severely affected cardiac patients is a significant factor in decreasing their reserve and increasing their tendency to dyspnea. Having considered some of the general conditions which bear on the problem of dyspnea in heart disease we may now turn to a special type of respiration which deserves mention both because it is common in clinical practice and because its mechan- ism involves other considerations than those which have been as yet discussed. This is the periodic type of breathing, which reaches its highest expression in the classical Cheyne-Stokes respi- ration. Careful observation, and more particularly the studying of records made with the pneumograph impress one with the fact that the association of periodic breathing with heart disease is much more frequent than is generally recognized. It appears in cardiorenal cases, in aortic disease, and in advanced myocar- ditis, and it is most often characteristically seen in patients who suffer from attacks of nocturnal dyspnea. A history of the onset of dyspnea in the evening is often given by patients with myo- cardial weakness, and if they are watched it will usually be found that periods of dyspnea alternate with periods of apnea. During the apnea the patient dozes off and goes to sleep. With the beginning of respiration he rouses a little, and at the height of dyspnea he wakes up to find himself intensely uncomfortable and often gasping for breath. His discomfort disappears with the cessation of dyspnea, and during the period of apnea he falls 268 HARVEY SOCIETY asleep again. Such attacks are sometimes referred to as ‘‘ cardiac asthma,’’ but the name is singularly ill-chosen for one of the most characteristic features of the true asthmatic attack is that the breathing is continuously rapid and labored. The volume of air expired has been measured in a few cases of mild periodic dyspnea and the total minute-volume has not been found to be remarkably high. The chief difficulty, and the reason for the discomfort appears to be that the patient is breathing only part of the time. The periods of apnea may last for half a minute, so that the patient is virtually breathing his minute-volume in the remaining thirty seconds. If he were to breathe the same minute volume of air regularly, over the whole minute, much less discomfort would be experienced. The volume of the individtal respirations rises to much above the normal, and since the vital capacity is usually decreased, the deepest respirations may approach the maximum of which the patient is capable. What can one say as to the fundamental cause of this type of dyspnea? The question is unfortunately one which remains in- completely answered, but some facts have been gathered which throw light on it. The suggestion has been made that the attacks are associated with an acidosis. As opposed to this it is difficult to conceive of an acidosis of such sudden onset, and moreover the typical feature of the respiration in acidosis, such as that seen in advanced diabetes, is hyperpnea with deep regular breathing. The clinical picture is quite different from that of periodic breathing. However, to settle the problem more definitely Dr, F. T. H’Doubler has studied the carbon dioxid content of the blood in a number of cases during the attack of dyspnea and either before or after it. Some of the patients who had advanced cardio- renal disease showed a slight decrease in the carbon dioxid tension, but this was rarely below 25 mm. and not sufficient to account for the dyspnea. Moreover there was no significant fall in the carbon dioxid tension during the attack of dyspnea as would be expected if the attack were dependent on a further increase in acidosis. Douglas and Haldane *! consider that the essential cause of Cheyne-Stokes respiration is oxygen lack, and they state that * Douglas and Haldane: Jour. Physiol., 1909, xxxviii, 401. CARDIAC DYSPNEA 269 “‘the periodic breathing is produced by periodic occurrence and disappearance of the (indirect) excitatory effects of want of oxygen’’ which ‘‘may be due to abnormal deficiency in the alveo- lar oxygen pressure’”’ or ‘‘to effects on the circulation of changes in the breathing or to both causes combined.’’ This explanation accounted satisfactorily for the periodic breathing observed by them on the expedition to Pike’s Peak.” The frequency with which the attacks come on at night is a feature of interest. Periodic breathing is a normal phenomenon which occurs in many healthy persons during sleep, and in hibernating animals. Straub ?* has shown that during sleep the alveolar carbon dioxid tension rises, and he attributes this to a decrease in the excitability of the respiratory center. Morphine, which depresses the respiratory centre, often produces periodic breathing. May it not be that the periodic breathing in heart disease is associated with a change in the excitability of the centre? In favor of this suggestion is its nocturnal occurrence, and the fact that in mild cases it often ceases if the patient is roused or in any way excited. To test the question further, some observations have been made with Mr. F. C. Hall on the effect of caffein, a respiratory stimulant, on Cheyne-Stokes respiration. The number of cases as yet examined is comparatively small but in nearly all a definite, though very transient cessation of the periodicity of the breathing, often associated with subjective im- provement, resulted from the administration of considerable doses of caffein. Several other drugs produced no noticeable effect. Morphine, in the few instances studied, caused no change or increased the periodicity, but its administration was thera- peutically beneficial, for it depressed the central nervous system so that the patients did not rouse during the periods of dyspnea. Whether Cheyne-Stokes respiration and periodie dyspnea in heart disease are due to oxygen lack in the sense of Douglas and Haldane or to a depreen of the excitability of the respiratory a hvaolae Haldane, Henderson and Schneider: Trans. Royal Soc. London, 1912, cciii, Series B, 185. 22 Straub: Deut. Arch. f. klin. Med., 1915, exvii, 397. 270 HARVEY SOCIETY centre, or possibly to a combination of the two, is a problem which still awaits solution. Such then, are at least some of the factors which contribute to the cause of that common but singularly complex symptom of heart disease—dyspnea. In its final analysis the problem resolves itself into the question of what we have called the ‘‘ pul- monary reserve.’’ The degree to which any individual mani- fests a tendency to dyspnea depends on the relation between the volume of air which he breathes while at rest and the maxi- mum volume which he is capable of breathing. The ability to meet adequately the needs of an increased metabolism such as occurs with muscular exercise, depends on the ‘‘pulmonary re- serve.’” In normal persons, as has been seen, the ‘‘ pulmonary reserve’’ is great, and healthy young men ean increase their pulmonary ventilation to approximately ten times the volume required by their resting metabolism. But in patients with heart disease the circumstances are much less favorable, and various conditions arise which cut down the ‘‘pulmonary reserve’’ and make them more readily subject to dyspnea. An increase of metabolism, or the development of an acidosis may raise the volume of air breathed while at rest, while an increased respira- tory rate or a decrease of the vital capacity of the lungs will make the maximum ventilation of which they are capable much lower than the normal. A decrease of the ‘‘pulmonary reserve’’ results, and even moderate exertion causes a rise of metabolism and a pulmonary ventilation which produces the subjective sen- sation of dyspnea. The degree to which these different factors are present in any given case is extremely variable. The earliest and most constant feature in the production of dyspnea is ap- parently a fall in the vital capacity and it is often met with quite unaccompanied by any of the other factors which we have con- sidered. In advanced cases of cardiae disease the situation be- comes much more complicated. The vital capacity drops still lower, the rate of respiration rises, the metabolism increases, and an acidosis may appear. Finally the picture is still further con- fused by the onset of periodic respiration, and it becomes, indeed, CARDIAC DYSPNEA 271 quite impossible to determine which element is most responsible for the patient’s unhappy state. Our conception of the etiology of dyspnea in heart disease is still vague and incomplete. Some little insight we have ob- tained, but further knowledge must come from the careful inves- tigation of the individual case, the discovery of other factors in the cause of dyspnea, and the systematic grouping of the separate types of dyspnea. Only by such studies can we hope to reach our ultimate aim—the proper treatment and the relief of dyspnea in heart disease. THE COAGULATION OF BLOOD WILLIAM H. HOWELL Johns Hopkins University, Baltimore, Md. HE clotting of blood takes place, according to the great ma- jority of observers, in two separate stages or phases—First, the formation of thrombin from some antecedent substance exist- ing in the blood. Second, the action of this thrombin on fibrinogen whereby fibrin is formed and deposited as a gel or clot. Two prominent investigators, Wooldridge and Nolf, have dissented from this view, particularly as regards the second stage, but the evidence in its favor is so conclusive that we are justified in accepting it as a basis for a discussion of the details of the process of coagulation. It is convenient to consider the final phase first— The demonstration that fibrin is formed by a reaction between thrombin and fibrinogen we owe to the investigations of Schmidt and of Hammarsten. In his numerous researches upon the co- agulation of blood Schmidt? laid the foundation for all subse- quent work. The summary of these investigations presented in his book contains a wealth of significant observations. Some of these have been amplified and elaborated by later workers, but there are some that as yet have not been followed out by modern methods. They offer leads for promising .investigations. Schmidt’s final theory of the process of clotting need not be de- scribed, since in some respects it has been made untenable by later work. But we owe to him the conclusive demonstration of the existence of thrombin, and of its essential role in the final act of clotting. His work was supplemented by the careful experiments of Hammarsten.* This observer proved that the substance acted upon by thrombin is fibrinogen, a globulin that exists normally in the circulating blood. Making use of the methods discovered by these investigators, or of the improvements upon these methods suggested by later workers, it is a simple matter to prepare separately these two 272 THE COAGULATION OF THE BLOOD — 273 substances in approximately pure condition, free at least from admixture with other so-called fibrin factors. When thus pre- pared the addition of the solution of one to that of the other, in proper proportions, is followed in a few minutes by the gela- tinization or clotting of the mixture. For the convenience of other workers I may describe briefly the methods that I have found most effective in the preparation of thrombin and fibrinogen. Thrombin is prepared from freshly formed fibrin—Blood from slaughter house animals (1 have used always pig’s blood) is defibrinated by whipping with the hand. The strings of fibrin thus obtained are washed in cold water, with constant kneading and pulling until the hemoglobin is removed. The white mass of fibrin is squeezed dry, minced with scissors and then covered with an 8 per cent solution of sodium chloride. The solution is kept in a refrigerator for 48 hours and is then filtered through cheese- cloth. The somewhat viscous filtrate should be rich in thrombin. It can be tested by adding a drop or two to a fresh solution of fibrinogen or to some oxalated plasma. The preparation may be preserved in this crude form or it may be further purified. If preserved in the crude form the excess of sodium chloride should be reduced to about 1 per cent by dialyzing in a collodion tube against seven times its volume of distilled water. The material in the tube is then filtered, distributed in lots of one or two ee. in watch crystals and evaporated to dryness in a current of air from an electric fan—the watch crystals are kept in a desiccator until needed. If a purer preparation of thrombin is desired the first filtrate from the digested fibrin is precipitated by the addition of an equal volume of acetone. The mixture is thrown on a series of small filters, 25 to 50 ce. to each filter, and allowed to filter completely. Each filter is then opened, the precipitate on the paper is spread as thinly as possible with a spatula and the papers, pinned to a board, are dried quickly before an electric fan. The dry filter papers are kept for a day or so in a desiccator and are then extracted with water by placing them in a flat dish and covering with distilled water— They are allowed to extract for an hour, without stirring, the water is then filtered off—the filtrate should be clear. This filtrate may be dis- tributed in small lots in watch crystals, dried quickly and preserved in a desiccator. Some small traces of coagulable protein are still found in this preparation. This impurity may be got rid of by a second precipitation with acetone, the precipitate being treated as above, or by shaking once or twice with chloroform and filtering. Both of these latter procedures are accompanied by a loss of thrombin, but they give a final preparation that is free of all traces of coagulable protein. Fibrinogen.—Fibrinogen solutions do not keep well. It is advisable as a rule not to use them for more than 24 to 48 hours—for this reason 18 274 HARVEY SOCIETY it is better to prepare it in small lots at a time and the blood of the cat furnishes a convenient source. The method of preparation that I have employed is a slight modification of that devised by Hammarsten, the chief difference being the use of the centrifugal in sedimenting the precipi- tate in place of filtration. The animal after fasting for 24 hours is anes- thetized with ether and bled from the carotid through a paratiined cannula into an oxalate solution, consisting of 1 per cent of sodium oxalate made up in 0.9 per cent sodium chloride, boiled and filtered. ‘The oxalate solution is used in the proportion of 1 to 8 of the blood. The tubes in which the blood is caught are inverted to mix the contents and are then centrifu- galized at high speed for 20 minutes—the clear plasma is drawn off—to this plasma one adds enough of a saturated solution of sodium chloride to produce a good precipitate—usually an equal volume or somewhat more of the salt solution. The mixture is placed at once in centrifugal tubes and centrifugalized at high speed for 5 minutes. The supernatant liquid is poured off, the tubes are drained and rinsed once carefully with a little of a half saturated solution of sodium chloride. The sediment is then covered with some of the half-saturated solution for a few minutes. This solution is poured off, the tube is wiped out with filter paper and the sediment is dissolved in a 2 per cent solution of sodium chloride with stirring. This solution is filtered and again precipitated by the addition of an equal volume of a saturated solution of sodium chloride. This pre- cipitate is then centrifugalized, washed, dissolved and filtered as in the case of the first precipitate, except that in the final solution a 1 to 1.5 per cent solution of sodium chloride is used. Two precipitations suffice, if attention is paid to the washing, to obtain a solution of fibrinogen that clots readily with thrombin, but does not clot spontaneously nor after the addition of calcium chloride or calcium chloride and tissue extract (Cephalin). Is Thrombin an Enzyme?—Schmidt believed that thrombin be- longs to the group of enzymes or ferments. In accordance with the prevalent conception of enzyme action this belief implies that thrombin does not take direct part in the formation of fibrin. Acting as a catalytic agent it induces directly, or through an intermediate reaction, the conversion of all or a part of the fibrinogen to fibrin. Schmidt based his belief upon the two reactions generally considered as characteristic of enzymes— namely their thermolability and the fact that they are not de- stroyed in the reaction they cause. In regard to the former it is quite true that thrombin in its normal environment, in blood plasma or serum, is destroyed by heating to 60° C. for even a THE COAGULATION OF THE BLOOD — 275 short time. But experience has shown that this is not the case under all conditions. Rettger* showed that an aqueous solution of thrombin prepared according to Schmidt’s method will stand boiling for several minutes without completely losing its power to act upon fibrinogen—indeed, Schmidt himself called attention to this fact. Subsequent observations reported by myself * indi- cate that this result is owing to the absence of inorganic salts in such solutions or rather to their low concentration. In dilute aqueous solutions of thrombin made by my method and dialyzed thoroughly against distilled water, boiling for five minutes or more while it weakens the action of the thrombin does not destroy it entirely. When to the same solution one adds sodium chloride to a concentration of 0.5 to 1 per cent boiling for a minute destroys the thrombin completely so far as its action on fibrinogen is concerned. While this fact is interesting it is not determinative in regard to the enzyme character of thrombin. It is known that the enzymes vary rather widely in regard to what may be called their lethal temperature, and that this tem- perature may be much influenced by the presence of other sub- stances. Whether or not the absence of inorganic salts influences the thermolability of any of the typical enzymes in a manner similar to that observed with thrombin has not been determined, so far as I know. In regard to the second and more significant property of enzymes, namely, that they act upon their respective substrates after the manner of catalytic agents, the evidence at hand indi- cates that thrombin does not exhibit this characteristic. The reaction between thrombin and fibrinogen appears to be rather of a fixed or quantitative character in the sense that a given amount of thrombin converts only a definite or limited amount of fibrinogen to fibrin. Thus Rettger was able to show that the amount of fibrin obtained from a given solution of fibrinogen increases with the amount of thrombin added. For example: 5 drops of thrombin yielded 0.2046 grms. of fibrin 10 drops of thrombin yielded 0.3573 grms. of fibrin 20 drops of thrombin yielded 0.6089 grms. of fibrin 40 drops of thrombin yielded 1.5872 grms. of fibrin 276 HARVEY SOCIETY In some later experiments I attempted to obtain more definite results by determining the actual weight of thrombin entering into the reaction.’ A solution of purified thrombin was used and the weight of thrombin was estimated as the difference between the weight of the residue after evaporation and the same residue after incineration, on the assumption that all the organic matter present was thrombin. The experiments gave figures of this kind: 0.05 mgm. of thrombin yielded 10.75 mgm. of fibrin 0.16 mgm. of thrombin yielded 34.00 mgm. of fibrin 0.25 mgm. of thrombin yielded 36.80 mgm. of fibrin 0.64 mgm. of thrombin yielded 42.50 mgm. of fibrin Other observers have also noted the fundamental fact that a sub- maximal quantity of thrombin allowed to act upon a solution of fibrinogen converts only a part of the fibrinogen to fibrin no matter how much time is permitted for the reaction to take place. In view of these results it would seem to be necessary to assume either that the thrombin unites with the fibrinogen in some definite way, or else that in the course of the reaction the formation of the fibrin inhibits in some way the further activity of the thrombin. One might suppose for example that the fibrin as it forms adsorbs the thrombin and thus removes it from the possibility of further action. It is not possible at pres- ent, so far as I can see, to come to any satisfactory decision in regard to this point. As stated below there is little or no evidence that the fibrinogen undergoes any profound chemical change in its conversion to fibrin. The change seems to be rather of a physical character in that the fibrinogen is aggregated out of its colloidal solution in somewhat the same way as certain colloids may be aggregated out by the action of inorganic salts. But in this particular case the fibrin-aggregates, if I may use this term, are different in structure and properties from fibrinogen-aggre- gates produced by other reagents. It would seem quite possible or probable that these fibrin-aggregates represent in fact a real physico-chemical combination between fibrinogen and thrombin. Specificity of the Thrombin.—As far as we know the reaction between fibrinogen and thrombin is specific in the sense that thrombin has no coagulating action upon other proteins and on THE COAGULATION OF THE BLOOD 277 the other hand no agent other than thrombin is capable of con- verting a fibrinogen solution to a fibrin gel. In this particular the action of the thrombin is highly specific, as in the case of the enzymes, but on the other hand there is little or no evidence of any specificity of action in relation to the different species of animals. In my own work it has happened that thrombin has always been prepared from pig’s fibrin, but this thrombin acts apparently with equal facility upon the fibrinogen or the blood plasma of other mammals, or indeed of other vertebrates if one may generalize from experiments made with the blood of the bird and the terrapin. Throughout the vertebrates apparently the fibrinogen is sufficiently uniform in properties to give the char- acteristic fibrin gel with pig’s thrombin, and presumably the thrombin from any other animal would exhibit the same uni- versality in its action. Nature and Properties of Thrombin.—The thrombin as pre- pared by Schmidt’s method is very far from being pure. While its aqueous solutions may give no precipitate on boiling, owing to the small concentration in salts, the addition of some neutral salt will cause the formation of a relatively large precipitate on boiling, indicating the presence of considerable amounts of a foreign protein. In the two methods that I have used for the isolation of thrombin ° I have succeeded in obtaining preparations that were free at least from coagulable protein. Such prepara- tions gave no precipitate nor opalescence upon boiling with or without the addition of inorganic salts, but they exhibited posi- tive protein reactions; for example, they gave the biuret, the tryptophan and the Millon’s reactions, and addition of am- monium sulphate to one-half saturation threw down a precipitate which on resolution showed marked thrombic action. When solutions of this purified thrombin were shaken re- peatedly with chloroform they finally ceased to give detectible protein reactions, but at the same time they ceased to show any thrombic action, so that one must conclude either that the throm- bin itself is a protein, soluble in water and not coagulated by boiling, or else that it is so closely associated with a protein of this character that it is not possible to separate them by any 278 HARVEY SOCIETY method yet suggested. My own opinion is that thrombin is a protein or protein derivative. Its tryptophan reaction is espe- cially distinct and this fact would suggest that the indol grouping is characteristic of and perhaps essential to its structure. Rettger called attention to the fact that putrefaction does not readily destroy thrombin, indeed often seems to increase its activity, and this observation suggests the possibility that there may be an essential grouping, containing probably indol, which is respon- sible for the thrombin action, but that this nucleus may be com- bined in the blood in some larger complex, and that consequently different methods of preparation of thrombin may yield products exhibiting somewhat different properties and activity. The older view advocated by Pekelharing* that thrombin is a nucleo-pro- tein is not borne out by my preparations. The purified thrombin gives no reaction for phosphorus. Origin of Thrombin.—Thrombin as such is probably not a normal constituent of the blood or of any of the body tissues. It is possible of course that it may occur in traces in the blood at times, but if so it is promptly removed or combined. In effective concentrations it occurs only under such abnormal con- ditions as lead to the clotting of blood. What does exist in the blood is a mother substance or antecedent material for which the names of prothrombin, thrombogen, proserozym, ete., have been proposed. In discussing the origin of thrombin therefore what we are really concerned with is the origin of the prothrombin. The earlier speculations upon this point were confused by a fail- ure to distinguish between this essential substance and the acces- sory substances contained in the tissues, or between this substance and a second source of thrombin found in the serum after coagu- lation and known as metathrombin. The work of Schmidt re- vealed the difference between thrombin and the zymoplastic material of the tissues, and the work of Morawitz cleared up the confusion in regard to prothrombin and metathrombin. Pro- thrombin itself is pictured as a substance constantly present in normal plasma, which is converted into active thrombin in the changes preceding coagulation. Morawitz has furnished good evidence for the view that one source at least of this prothrombin THE COAGULATION OF THE BLOOD — 279 is the blood platelets. These elements when obtained in quantity by differential centrifugalization yield on solution in water a sub- stance which by itself will not coagulate fibrinogen, but which ean be made to furnish active thrombin when submitted to the action of calcium salts. The results obtained by Morawitz have been corroborated in my laboratory by Bayne-Jones.* There is a bare possibility that in these experiments the prothrombin was simply adsorbed by the platelets, since it has been shown that prothrombin, like thrombin, is readily adsorbed by precipitates such as calcium fluoride or calcium oxalate,’® but the greater probability is that the prothrombin is a constituent of the blood platelets, and that when these fragile elements dissolve in the plasma they yield to it some prothrombin. At the time of the shedding of blood there is a massive destruction of platelets which must add a large increment to the supply of prothrombin carried by the circulating plasma. The more or less complete examinations made by various ob- servers to detect the presence of prothrombin in other tissue ele- ments have given either negative or uncertain results with the exception of an interesting series of observations made in my laboratory by Drs. C. K. and K. R. Drinker.’ In these experi- ments the bone marrow of the tibia of a dog was perfused through its nutrient artery with solutions of sodium chloride 0.9 per cent, or with a Ringer’s solution. It was found that the marrow gives a large yield of prothrombin when perfused with solutions of sodium chloride, and a yield of active thrombin when Ringer’s solution is used, since in this case the calcium of the Ringer’s mixture serves to activate the prothrombin to thrombin. We must believe, therefore, that the marrow constitutes an important source of production of prothrombin, and since there is much histological evidence to show that the marrow tissue gives rise to blood platelets it may be that the prothrombin is supplied to the blood through the latter element. On the other hand, Hurwitz and Drinker’ have shown that in rabbits treated with subcu- taneous injections of benzol until the marrow is rendered aplastic, there is a marked decrease in the content of prothrombin in the blood together with a decrease in the number of platelets. So 280 HARVEY SOCIETY far as our knowledge goes we must consider the bone marrow as the main source of prothrombin, but whether it is given off from this tissue in solution or is passed out in the substance of the platelets to be eventually given to the plasma when these latter dissolve cannot be determined from the evidence at hand. In regard to the chemical nature of prothrombin, and the difference between it and thrombin we practically know nothing at all. I have described a method of separating prothrombin from blood- plasma ** but not in a form sufficiently pure for chemical study. Preparation of Prothrombin.—The method that I have used to separate prothrombin from blood-plasma while it does not give the substance in pure form does yield a preparation that is useful for experimental and demonstrational work. It is as follows: The blood of a cat is oxalated as described above for the preparation of fibrinogen and the clear plasma is obtained by centrifugalizing. The plasma is heated to 54° C. in a water-bath to precipitate the fibrinogen, and filtered. The filtered plasma is precipitated in lots of 5 ec. by the addition of an equal volume of acetone. The precipitate is thrown at once on a filter and the filtration is effected as rapidly as possible by a water pump. The precipitate is washed quickly with ether with the aid of the water pump, and the filter is then removed, the precipitate is spread as thinly as possible with a spatula, dried before an electric fan and kept in a desiccator. When needed one of these papers is cut into small pieces and extracted for 1/, hour to 1 hour with 10 ec. of distilled water to which has been added 4 or 5 drops of a 0.5 per cent solution of sodium bicarbonate. The filtered solution when added alone to fibrinogen does not cause clotting, but if calcium chloride is also added the fibrinogen clots in from 5 to 10 minutes. Two drops of a 0.5 per cent solution of caleium chloride may be added to 5 drops of the solution of prothrombin. The Nature of the Reaction Between Thrombin and Fibrino- gen.—Schmidt believed thrombin to be an enzyme or ferment and interpreted its action upon fibrinogen in the light of what wa; known concerning the action of the more familiar digestive en- zymes. This view was adopted in the beginning by Hammarsten ** after he had shown that fibrin is formed by a reaction between thrombin and fibrinogen. He believed that the fibrinogen under- went hydrolytic cleavage with the formation of fibrin on the one hand, and a soluble protein, fibrin-globulin, on the other. This latter protein was found in the serum. It is a globulin that has THE COAGULATION OF THE BLOOD 281 a temperature of heat coagulation of 65° to 66° C., and like fibrinogen is salted out of its solution by half saturation with sodium chloride. Subsequently Hammarsten abandoned this view because quantitative determinations revealed the fact that vari- able amounts of fibrin, from 61 to 94 per cent, are formed from a given weight of fibrinogen. Schmiedeberg * adopted the cleavage theory and expressed the reaction in the following equation: 2 ( Cyr HyesN 3050s; ) F H,O = Cyos Heo N 30S Ox a Cy Ayre 395037 Fibrinogen Fibrin Fibrin-globulin According to this equation about 48 per cent of fibrin should be obtained from a given weight of fibrinogen, and Heubner in fact stated *® that when fibrinogen solutions are coagulated by heat (58° to 60°C.) the yield in coagulated fibrinogen constitutes from 48.3 to 49.7 per cent of the fibrinogen used. In some later experiments by Huiskamp ™ it is stated that a fibrinogen may be prepared by precipitation with sodium fluoride, which when heated to 55° C. gives the usual heat coagulum, but without any evidence of a second protein corresponding to the fibrin-globulin. He concluded, therefore, that the fibrin-globulin found in serum exists preformed in the plasma and is not a cleavage product of the fibrinogen formed during coagulation. Hammarsten for the reason stated above abandoned the cleavage hypothesis and sug- gested that the conversion of fibrinogen to fibrin consists in a molecular transformation of the former, and that the fibrin- globulin represents a variable fraction of the fibrinogen which has undergone incomplete transformation. Nothing has been added to our knowledge of this subject in recent years, and it is quite evident that the whole matter of the quantitative relations of the fibrinogen and fibrin needs reinvestigation, although it is probable that the matter cannot be studied successfully until a method is devised for obtaining pure fibrinogen. In the method used at present we have no guaranty that the fibrinogen obtained is free from other globulins. If, for example, Huiskamp is cor- rect in the supposition that fibrin-globulin exists in the plasma of the circulating blood then it would probably be precipitated 282 HARVEY SOCIETY along with the fibrinogen by half saturation with sodium chloride. If the fibrinogen were obtained entirely pure it is quite probable that the yield of fibrin instead of representing from 61 to 94 per cent of the fibrinogen might in fact amount to 100 per cent. The ultra-microscopic observations described below suggest this pos- sibility. At present we can only say that the older view that fibrin is formed by hydrolytic cleavage of the fibrinogen can not be accepted without further evidence. The Fibrin-gel.—When fibrin is formed in normal coagulation it is not simply precipitated or aggregated out of a colloidal solution. On the contrary it forms a gel of an interesting and peculiar character. The former idea concerning the structure of this gel was obtained from a microscopic examination of clotted blood. It was supposed that the fibrin is deposited as fibrils or threads which form a network enclosing the plasma and ecor- puscles, and the solidity of the gel was explained on the sup- position that the network forms a honeycomb structure in which the plasma as an internal phase is confined within solid septa of fibrin. In 1885 Schimmelbusch *§ as the result of careful micro- scopic studies stated that the fibrin in clotting is deposited as separate needles of a thin spindle shape and 5 to 20 miera long. This observation was not corroborated, apparently, by later workers and appears to have been forgotten. In 1914 Stiibel *® making use of the dark field illumination described again the formation of fibrin needles or crystals, and in the same year I confirmed his observations in some independent work done with the ultramicroscope.”?° I employed what is known as the slit form of the ultramicroseope. Using solutions of purified thrombin and fibrinogen, or, in place of the latter the clear centrifugalized oxalated plasma, one can obtain easily a beautiful demonstration of the formation of the fibrin needles. The observation cell is first filled with normal saline (NaCl 0.9 per cent) and this is displaced by a suitable mixture of the fibrinogen and thrombin, the concentration of the latter being chosen so as to cause coagu- lation in a conveniently short time of five to ten minutes. When fibrinogen is used the sequence of events is as follows: The solu- tion shows at first a cone of light in which no particulate structure fic. 1.—The crystalline gel of fibrin as seen with the ultramicroscope— some of the fibrin-needles are seen clearly, others are out of focus and blurred. The preparations were made from the oxalated centrifugalized plasma of dog's blood made to clot by addition of a solution of thrombin. THE COAGULATION OF THE BLOOD — 283 is apparent although it may contain scattered coarse particles. After a certain period varying with the concentration of throm- bin used the cone of light becomes more intense, the whole field shimmers with scintillating points and a mass of particles appear that are formed apparently in the previously homogeneous field of the cone of light. These particles assume quickly the shape of short rods that exhibit very active movements which take them into and out of the focal plane, so that continuous observation of a single rod becomes difficult or impossible. The rods lengthen to needles that show less movement as they increase in length and finally the field settles down to a mass of interlaced brilliant needles. When the proportions of thrombin and fibrinogen are not favorable, for example, when the concentration of the thrombin is too low, or when the fibrinogen is in part denatured, the fibrin instead of depositing as firm well shaped needles takes the form of beaded threads or fibers. As regards the size of the crystals it was found in general that the more rapid the coagulation the shorter and finer were the crystals. When oxalated plasma is used in place of the fibrinogen some of the steps described above may be missed. The plasma without the thrombin shows a uni- form cone of light in which are many larger particles some of them obviously fat granules. After the thrombin is added the beginning of the process of clotting may be indicated by an intensification of the cone of light and then without obvious dis- turbance needles appear here and there in the field and rapidly increase in number until they form a stationary intermeshed mass, among which the coarse particles may be seen in Brownian movements. When for any reason the coagulation is very im- perfect, as, for example, may result from too small an amount of thrombin or an excess of antithrombin in the plasma, so that the mixture fails to give perhaps a visible clot in mass, one can often observe the formation of scattered fibrin needles that float sepa- rately in the field but do not cohere to make a meshwork. Under what may be designated usual or normal conditions the needles are from 10 to 30 microns in length. They appear as separate structures in the irregular mesh that they form, but probably they actually ecohere with one another at points of contact. 284. HARVEY SOCIETY We may consider the blood clot not simply as a gel but as a crystalline gel. So far as I am aware no similar gel has been observed for any other of the colloids found in living organisms, although examples of crystalline gels more or less similar to that formed by fibrin have been described, especially the gel of barium malonate discovered by Flade.** It would seem from this description that when fibrin acts upon fibrinogen two separate reactions take place—First, the fibrinogen particles, with or without some intermediate change, are aggregated into definite crystalline forms owing to the influ- ence of some directive or vectorial force. Second, the fibrin- aggregate, that is to say, the fibrin needles set up a gelatinization of the system. Under certain conditions the second reaction, the gelatinization may occur without the crystallization. When sodium carbonate for example is added to the plasma so as to produce an appropriate concentration of hydroxyl ions the sub- sequent addition of thrombin causes the formation of a clot or gel in which no visible structure can be made out with the ultra- microscope. In this ease the fibrin-aggregates are formed without doubt and in consequence of their presence the system gels, but these aggregates are not massed into crystalline structures, nor indeed into visible forms of any kind. Structureless fibrin gels obtained by this method are usually transparent and non-retrae- tile, whereas the normal crystalline gel is opalescent and very retractile. If loosened from the sides of the vessel in which it has formed it shrinks, squeezing out serum, and if broken with a rod each piece rounds off to form a separate structure. The structure- less clot on the contrary is a soft jelly—when cut by a rod the pieces instead of separating flow together on contact. In the formation of the normal clot two main problems are presented for solution, the phenomenon of erystallization and the phenomenon of gelatinization. In regard to the former it seems evident that both the thrombin and the fibrinogen are essential to the erystalline form of aggregation, and this con- sideration would imply that both substances enter into the struc- ture of the needles. Agents other than thrombin that cause a precipitation of the fibrinogen particles give only amorphous THE COAGULATION OF THE BLOOD 285 aggregates. In regard to the cause of the gelatinization I have taken the general view advocated long ago by Nageli, and ex- pressed in more general terms recently by Hober. As applied to the specific case of fibrin this conception may be stated as fol- lows: Fibrin is a hydrophilic colloid. Its particles attract and bind the surrounding water phase in an especial degree com- pared with similar hydrophilic colloids such as fibrinogen. When the fibrin particles or aggregates are greatly dispersed this action results in giving the system a greater degree of viscosity. In a lesser degree of dispersion a soft gel is formed, but when massed into the coarse aggregates of the crystalline needles the surrounding water films are so firmly bound as to form a solid gel. THE FIRST PHASE OF COAGULATION—FORMATION OF THROMBIN The Role of Calcium.—Arthus and Pagés demonstrated the fundamentally important fact that calcium is necessary to the process of the clotting of blood. If calcium is removed, as by oxalating the blood, no clotting occurs. If the calcium is restored by adding a proper excess of calcium chloride clotting occurs promptly. Schmidt was not willing to admit this important fact. He tried to prove that the action of calcium is not specific, but practically all later work has served to demonstrate that he was in error. The role of calcium is specific and can not be taken by any of the other bases normally present in blood. The theories advanced by Arthus and Pagés ?* and by Pekelharing’ to explain the mode of action of calcium need not be discussed, since the careful experiments of Hammarsten * demonstrated that the cal- cium is concerned only in the formation of thrombin. In the second stage of clotting, the reaction between thrombin and fibrinogen, calcium is not necessary, although in certain con- centrations it may influence the reaction especially as regards its velocity. We have little or no knowledge of the way that ecal- cium acts. There is a difference of opinion in the first place as to whether or not calcium ions can cause the activation of prothrombin to thrombin when acting alone. Some workers believe that the calcium is effective in causing this change only when it acts in cooperation with the thromboplastic material of 286 HARVEY SOCIETY the tissues (thrombokinase, cephalin). This question will be dis- cussed below in connection with a description of the action of tissue extracts. It has been noted by many observers that the influence of calcium varies with its concentration. There is a certain optimal concentration at which the formation of thrombin is most favored, and beyond this point a further increase in con- centration tends to inhibit or even prevent coagulation. Sab- batani,”* believed that there is a certain minimal concentration below which the calcium ions are ineffective and also a maximal concentration that suffices to inhibit the activation. The maxi- mal concentration he placed at 18 grms. calcium chloride per liter (molecular concentration 0.162). In this amount coagulation is prevented. Stassano and Daumas”* state that the minimal con- centration lies between 13 and 21 mgms. of calcium chloride per liter. The action of calcium may be prevented by precipitating it out of the blood in insoluble form, as an oxalate, for example, or by forming salts of calcium in the blood which have a small dissociation, so that the concentration in calcium ions is below » the minimal limit referred to above. Sabbatani explains in this way the effect of citrates in preventing coagulation. Making use of fibrinogen solution with only traces of calcium Hammarsten was able to obtain specimens of fibrin in which the calcium con- tent was only 0.007 per cent. If this calcium were contained in the fibrin as a constituent of its molecule it would imply a molecular weight too large (800000) to be accepted as possible. He concluded therefore that the calcium found was present as an impurity, and that it does not form a constituent of the fibrin, nor probably of the thrombin. Presumably from this point of view the calcium ions act as catalytic agents and not by forming a calcium compound with the prothrombin or any portion of its molecule. While this conclusion may be accepted provisionally it may be noted that it is searcely a necessary result of the experiments reported by Hammarsten. If the fibrin needles consist of an orderly aggregation of fibrinogen and throm- bin particles, the latter existing in small proportions compared with the former, it is possible that the calcium might be an essential constituent of the thrombin molecule without making a THE COAGULATION OF THE BLOOD ~— 287 percentage of the fibrin larger than the minimum obtained by Hammarsten. It would seem desirable to make an effort to deter- mine whether an effective thrombin can be obtained entirely free from calcium. The kole of Tissue Extracts—The plasma of the circulating blood contains fibrinogen, prothrombin and calcium ions, but it does not clot, owing to the absence of active thrombin. If it were possible to get this plasma out of the body without coming into contact with the tissues, and without injury or destruction of any of the formed elements of the blood, then, no doubt, the plasma would remain fluid as it exists within the vessels. ‘This experiment can be made in fact when one uses the blood of the lower vertebrates (bird, terrapin). By bleeding through a paraf- fined canula into a cooled paraffined centrifugal tube and cen- trifugalizing at once a plasma may be obtained which clots very slowly or not at all. A similar result might be obtained with the slowly clotting blood of the horse and possibly with other mam- malian blood if sufficient care were taken. The prompt clotting that we observe ordinarily in shed blood is due to something added to the plasma when the blood escapes from its normal environ- ment, and this something is derived either from the breaking down of some of the corpuscular elements of the blood itself or from the outside tissue with which it comes into contact. This fact has been demonstrated in various ways by many observers. In the mammals the platelets of the blood disintegrate very rapidly when the blood is shed and thus furnish one source for the substance that initiates the process of clotting. In the blood of the bird and the reptile there are no corpuscles as fragile as the mammalian blood-platelets, so that in these animals the initia- tion of the clotting depends mainly on material furnished by the outside tissue-elements, the injured tissue, for example, of the wound. Schmidt designated the material furnished by the blood corpuscles or the other tissue elements as zymoplastic substance, on the theory that it acts on the mother substance of thrombin, the prothrombin, and splits off active thrombin by a process of cleavage. Various other names have been given to this same material. Wooldridge *® spoke of it as tissue-fibrinogen, Mora- 288 HARVEY SOCIETY witz** as a thrombokinase, Fuld?* and later Bordet and Delange ** as a cytozym. Nolf included it under a general desig- nation of thromboplastic substances, and frequently it has been referred to simply as tissue-extract. In my earlier papers I suggested the term thromboplastin on the view that the material is a definite chemical substance of unknown composition, but later when its chemical nature was revealed this term was abandoned. Nature of the Zymoplastic Substance.—Information in regard to the chemical nature of this material has accumulated slowly, but its general trend has all been in one direction. Schmidt found that it is soluble in aleohol and ether, withstands boiling and on analysis shows the presence of phosphorus and nitrogen. He concluded, therefore, that lecithin must be present in the tis- sue extracts and must have some relation to their activity. Wooldridge also believed that the activity of his tissue-fibrinogen was referable to the lecithin contained in it, but he made the significant observation that while lecithin prepared from various animal tissues such as the brain, testis and thymus, is quite active, that obtained from the egg yolk is without effect on coagulation. Bordet and Delange state that their cytozym is thermostable and exhibits the solubilities of lecithin, and Hirschfeld and Klinger *° corroborate this statement. Fonio in his advertised description of his patented preparation, so-called Coagulen-Ciba, which is a zymoplastie substance said to be obtained from blood platelets, speaks of the material as a lipoid. Zak*° demonstrated that phosphatid material prepared from the brain (acetone pre- cipitation of a petroleum ether extract of dried brain) has a strong zymoplastie action. It will be seen from these references that there has been a quite general agreement that the zymo- plastic material is a lipoid substance and that it belongs to that group of lipoids known as phosphatids. My own work has led me to a similar conclusion.*t I found that the active material may be extracted from dried tissues with ether, and that when the various constituents present in this extract are examined separately for their zymoplastie action the result shows that the active substance is a phosphatid soluble with difficulty in alcohol THE COAGULATION OF THE BLOOD .- 289 and corresponding in its reactions with the phosphatid designated by Thudichum as cephalin. I was able to show that lecithin pre- pared from various sources has no zymoplastic action, while the related cephalin, from whatever tissue it is obtained, is, when freshly prepared, very active. The cephalin obtained from ether extracts of brain and other tissues withstands boiling without losing its zymoplastic activity. On the other hand it was found that aqueous or saline extracts of fresh organs which exhibit strong zymoplastic action are precipitated even at temperatures of 56° to 60° C. and lose their activity. If, however, the pre- cipitate is collected, dried and extracted with ether a material is obtained which is zymoplastic. My interpretation of these results was that the active substance in tissue extracts is probably cephalin, but in the normal tissues the cephalin is associated with a protein having a very low temperature of heat coagulation, and very readily soluble in water. In this form the active material is thermolabile, since it is precipitated out of solution on heating, but the essential constituent, the cephalin, when separated off, as occurs in ether extracts, is not destroyed even at boiling tem- perature. In this way we may reconcile the opposing state- ments found in the literature of the subject in regard to the ther- mostability of the zymoplastic material. While the evidence col- lected pointed distinctly to cephalin as the essential zymoplastic substance there remained a possibility that the active substance might be an impurity of some kind adherent to the cephalin. At my request McLean undertook a special study of this point.*? Cephalin was prepared with especial care from brain, liver and heart and in as pure a form as our present knowledge of this substance permitted. The material thus prepared exhibited always marked zymoplastic activity, while preparations of related phosphatids, lecithin, cuorin, heparphosphatid and sphingomyelin were inactive. It is difficult to say with certainty what peculiarity in the structure of the cephalin molecule confers upon it this especial property, but the evidence obtained seems to connect this particular activity with the unsaturated fatty acid group. McLean found that a hydrogenated cephalin furnished by Levene and West was inactive. Cephalin when freshly prepared is 19 290 HARVEY SOCIETY usually very zymoplastic in aqueous solutions of 0.1 per cent or more. But the material, even though kept in the dark and in a desiccator loses its activity either slowly (through months) or rapidly (in weeks) according to its mode of preparation. An investigation of this point by McLean, soon to be published, indi- cates that the loss of activity is due to a process of oxidation and that parallel to this disappearance of its zymoplastic action there is a corresponding diminution in its iodine number. The Mode of Action of the Cephalin.—The addition of zymo- plastic material to the blood, from whatever source it is derived, initiates the process of clotting by leading to the development of active thrombin from the inactive prothrombin. How does it effect this result? The nature of the answers that have been given to this question may be illustrated by a brief statement of some of the theories proposed. Schmidt believed that the zymo- plastic material acts on the mother substance of thrombin, the prothrombin, after the manner of an enzyme. The prothrombin undergoes cleavage with the formation of thrombin. It will be remembered that Schmidt did not accept the view that calcium salts are necessary to the formation of thrombin, and it is im- plied in his theory that the zymoplastic material may split off thrombin from prothrombin in the absence of calcium. This, however, is not the case. In a decalcified (oxalated) plasma no thrombin is formed by the addition of calcium free zymoplastic material (Cephalin). This important fact indicates that the zymoplastie material and the calcium cooperate in the production of thrombin from prothrombin, and most of the later views are founded on this assumption. Fuld *° and Morawitz ** were influenced in their theories by the recently discovered action of enterokinase in activating trypsinogen to trypsin. They suggested that tissue extracts con- tain an organic activator or kinase which Morawitz designated as thrombokinase, and that an essential condition for its activity is the presence of calcium ions. In other words, zymoplastic material is a kinase which in co-operation with calcium activates prothrombin to thrombin. Bordet and Delange ** have accepted this view in its essential features, but they have introduced an THE COAGULATION OF THE BLOOD 291 additional factor beside adopting a different nomenclature. They believe that the zymoplastic material, designated by them as cytozym, in the presence of and presumably with the co-activity of calcium reacts with a substance designated as serozym to produce thrombin. Their serozym corresponds with the pro- thrombin (thrombogen) of Morawitz, but they assume further that the serozym exists in the circulating blood in an antecedent stage, a proserozym. At the time of clotting the proserozym is converted to serozym by some unknown reaction, possibly the re- moval of an antagonistic substance. According to their view this latter reaction is the initial step in the process of clotting, but so far as I know, they give no theory or fact in regard to the cause of this change. Morawitz in his numerous publications has expressed no definite opinion in regard to the nature of his thrombokinase. The typical kinase, the enterokinase of the small intestines is supposed generally to be an enzyme, but if we accept the results of the work described above thrombokinase is a definite chemical substance, cephalin, whose action can scarcely be inter- preted as that of anenzyme. Detailed study of its action indicates that a given amount of cephalin oceasions the conversion of a definite amount of prothrombin to thrombin (Gasser). There is no evidence of an enzyme characteristic in this reaction, so that the nomenclature of thrombokinase or cytozym as applied to cephalin would seem to be inappropriate. In common with other workers I am convinced that the conversion of prothrombin to thrombin requires, under the conditions existing in the cireu- lating blood, the combined action of the calcium ions and the eephalin, but I have attempted to explain the action of the latter from a different point of view. Instead of assuming that the cephalin acts upon the prothrombin as a kinase or enzyme or in any other way I have adopted the other possible suggestion, namely, that under normal conditions the prothrombin is pro- tected from the activating influence of the calcium ions by a combination of some kind with an inhibitory agent or antisub- stance, and that the cephalin exerts its accelerating effect upon coagulation by neutralizing the influence of this inhibitory sub- stance thus liberating the prothrombin so that the calcium can 292 HARVEY SOCIETY convert it to thrombin. The basis for this point of view is found in the fact that there exists in the blood a substance which pre- vents or tends to prevent coagulation. This substance has been referred to by many observers, usually under the name of anti- thrombin. I have described a simple method, referred to below, by means of which a relative estimate may be made of the amount of this substance present. Observations by this method have shown that in the slow coagulating bloods, such as those of the lower vertebrates or the blood of a peptonized dog, the amount of antithrombin is increased and that the greater the amount of this antithrombin the more stable or less coagulable is the blood. Tissue extracts or cephalin solutions added to such bloods cause a shortened coagulation time or induce coagulation in those speci- mens in which spontaneous coagulability is no longer exhibited. It can be demonstrated moreover that the addition of cephalin to such plasmas causes a more or less complete disappearance of the detectible antithrombin. This fact may be illustrated by the following experiments made upon the plasma from oxalated blood. The blood of the animal was oxalated as it flowed from the artery. The mixture was centrifugalized and the clear plasma pipetted off. With this plasma two mixtures were made: Mixt- ure A, consisting of plasma one part and water one part—and Mixture B, consisting of plasma one part and an aqueous solu- tion of cephalin one part. The mixtures were allowed to stand for 30 minutes and were then heated to 54° C. to precipitate the fibrinogen. The filtrates from these precipitates were then tested for antithrombin according to the following schema: I. Mixture A, 1 drop + Thrombin 5 drops — Incubation of 15 mins. + Fibrinogen 10 drops = Clot in 5 mins. Mixture A, 1 drop + Thrombin 4 drops — Incubation of 15 mins. + Fibrinogen 10 drops = Clot in 10 mins. Mixture A, 1 drop + Thrombin 3 drops — Incubation of 15 mins. -++ Fibrinogen 10 drops = Clot in 15 mins. Mixture A, 1 drop + Thrombin 2 drops — Incubation of 15 mins. -+ Fibrinogen 10 drops = Membrane-clot in 20 mins. THE COAGULATION OF THE BLOOD — 293 In the similar series made with Mixture B all of the tubes clotted within 3 minutes showing that the antithrombic action of the plasma had been completely removed by the cephalin. A more striking result may be obtained by using pig’s plasma which contains a larger amount of antithrombin. For example— with a freshly oxalated specimen of pig’s blood some clear plasma was obtained and two mixtures A and B were made as above, with water and with an aqueous solution of cephalin. After standing for 30 minutes these mixtures were heated to 54° to remove the fibrinogen and the filtrates were tested for antithrom- bin with the following results: I. Mixture A, 1 drop + Thrombin 5 drops — Incubation of 15 mins. + Fibrinogen 10 drops = Partial clot in 65 mins. Mixture A, 1 drop + Thrombin 4 drops — Incubation of 15 mins. + Fibrinogen 10 drops= No clot in 2 hrs. Mixture A, 1 drop + Thrombin 3 drops — Incubation of 15 mins. + Fibrinogen 10 drops = No clot in 2 hrs. Mixture A, 1 drop + Thrombin 2 drops — Incubation of 15 mins. + Fibrinogen 10 drops = No clot in 2 hrs. II. Mixture B, 1 drop + Thrombin 5 drops — Incubation of 15 mins. + Fibrinogen 10 drops = Clot in 5 to 10 mins. Mixture B, 1 drop + Thrombin 4 drops — Incubation of 15 mins. + Fibrinogen 10 drops = Clot in 5 to 10 mins. Mixture B, 1 drop + Thrombin 3 drops — Incubation of 15 mins. + Fibrinogen 10 drops = Clot in 5 to 10 mins. Mixture B, 1 drop + Thrombin 2 drops — Incubation of 15 mins. + Fibrinogen 10 drops = Clot in 10 to 15 mins, When blood-serum is used instead of blood-plasma a different result is obtained; addition of cephalin solutions causes little or no diminution in the amount of free antithrombin. The differ- ence in the behavior of the antithrombin in plasma and in serum is difficult to understand. It depends apparently upon the fact that in serum fully formed thrombin is present; and, as will be described further on in discussing metathrombin, thrombin and antithrombin tend to combine to form first a loose and finally 294 HARVEY SOCIETY a firm compound, and upon this combination the cephalin has apparently little effect. These considerations led me to believe that in the circulating blood antithrombin protects the prothrombin by some means and that the efficacy of the cephalin resides in its property of neu- tralizing the antithrombin. The evidence that I have been able to present in favor of this view has been indirect only, namely, the existence of antithrombin in the blood and the power of the cephalin to neutralize this antithrombin. Some recent work in my laboratory has brought to light the additional fact that in some of the tissues, liver, heart-muscle, lymph glands, there is a phosphatid material which has a quite specific power of acting upon or combining with prothrombin so as to prevent its activa- tion to thrombin. Whether or not this new substance, which provisionally I have designated as antiprothrombin, will be found as a normal constituent of the blood can not be stated at present. It seems very possible that this may be the ease, and in that event it will be necessary to consider its especial function. It may be that the inhibiting substance which according to my view protects the prothrombin from activation and thus safeguards the fluidity of the blood may be this antiprothrombin rather than the antithrombin. Experiments that are described farther on show that cephalin is capable of liberating prothrombin from the inhibitory influence of the antiprothrombin and therefore has the property of neutralizing the action of both of the inhibitory substances, the antiprothrombin and the antithrombin. Further experiments must be made to determine definitely whether the action of cephalin in clotting is to be explained along these lines or, according to the hypothesis of Morawitz, as the result of some direct action exerted by it upon the prothrombin. Cephalin as a Hemostatic—On any theory of their action tissue-extracts are of the first importance in starting and accel- erating the clotting of blood, and one might suppose therefore that such extracts could be used to control hemorrhages. This thought has occurred no doubt to many workers and some of the attempts to prepare the zymoplastic substance in a form suitable for such uses have been put on record, for example, the prepa- THE COAGULATION OF THE BLOOD — 295 ration of thrombokinase described by Horneffer and Battelli * and the patented preparation of Kocher and Fonio placed on the market for this purpose. With regard to the latter it may be observed that it professes to be a preparation of lipoid material obtained from blood-platelets. Since the same lipoid material exists in much greater abundance in tissues like the brain which are much more accessible it is difficult to understand the reason or value in selecting platelets as a source of supply. According to my views the active constituent of zymoplastic substance is cephalin and since this material may be prepared very readily in active form, the question arises whether this cephalin may be put to some practical use in controlling hemorrhage. It is a question which I am having investigated with care as I believe it promises useful results. Meanwhile I should lke to put on record some incomplete results obtained by myself and others that give some indication of the lines to be followed. Impure but very active preparations of cephalin may be made by the method used in my laboratory as follows: Preparation of Cephalin.—F resh brains are obtained from the slaughter house, cleaned from membranes and blood, ground to a pulp, spread thin upon glass plates and dried in a current of warm air. When thoroughly dry, powder in a mortar and then extract with ether for 3 or 4 hours. Filter off the ether in a closed space and repeat the filtration until the filtrate is clear. Evaporate the filtrate to dryness before an electric fan. Extract the residue thoroughly with acetone (to remove cholesterol and cholesterol esters). Drain off the acetone and extract twice with excess of aleohol (to remove most of the lecithin). Drain off the alcohol, allow to dry and keep in a desiccator in the dark. For use a small portion of the material is stirred in water until a milky solution is obtained—a 0.1 per cent. solution is convenient for use. If necessary it may be sterilized by boiling. Whether or not these solutions can be used advantageously in controlling hemorrhage must be determined by experience. Two general facts, however, must be borne in mind. In surgical operations in which there may be much laceration of tissue there is probably a liberation of much zymoplastic material, and addi- tion of cephalin solutions may therefore be superfluous. Sec- ondly, cephalin in solid form or in solution deteriorates, slowly 296 HARVEY SOCIETY or quickly according to circumstances, and every preparation should therefore be tested before using. One relatively simple method of testing the cephalin which is employed in my labora- tory makes use of its great activating influence on fresh serum. An animal is bled—part of the blood is oxalated and centrifu- galized to obtain oxalated plasma, and a part is clotted and centri- fugalized to obtain serum. In dog’s or cat’s serum the amount of effective thrombin is not large, so that if say 3 drops of the serum are added to 8 drops of the plasma clotting of the latter takes place slowly and often imperfectly, requiring 30 minutes or more. If, however, one adds to the 3 drops of serum 3 or 4 drops of the cephalin solution and then adds the plasma the latter will clot firmly as a rule in one minute or less. Making use of aqueous solutions of cephalin I have been able to control obstinate hemorrhages in hemophilic eases by applying a dressing of gauze soaked in the solution, although prompter and more satisfactory results were obtained by using both throm- bin, in powder, and the cephalin. Cecil has described ** a special method of using cephalin to control hemorrhage after prosta- tectomy with which good results have been obtained, and no doubt under proper conditions this material may be employed to ad- vantage in controlling external hemorrhage. I have been inter- ested also in another possibility, namely, the control of internal hemorrhages in cases with a more or less marked hemophilic tendency. It would seem possible that cephalin introduced into the blood directly or by absorption from the intestines might be used safely to lower its coagulation time. With this idea in mind I have made a large number of experiments on dogs in which cephalin in aqueous solution was injected intravenously, intraperitoneally or subcutaneously or finally was fed by mouth. The results of these experiments have not been published since they showed many irregularities that will require further experi- mentation to explain. Some general results, however, may be referred to briefly. Intravascular injection of cephalin in dogs in the proportion of one decigram per kilogram of animal causes a shortening of the coagulation time of the blood by 14 or % of the normal time, and this effect may endure for at least one ee THE COAGULATION OF THE BLOOD 297 or two hours without any evidence of intravascular clotting and no evidence of a general reaction upon the animal, so far as respiration, heart-rate or blood pressure is concerned. When cephalin is added to blood outside the body it brings about a rapid production of thrombin. It is to be presumed that when it is introduced into the circulating blood there is also a pro- duction of thrombin, but there is no intravascular clotting. The explanation that I would suggest is that this free thrombin is bound by the antithrombin, especially as it has been shown “* that at the body temperature this property of the antithrombin is much augmented. Gasser ** has made it very probable that the antithrombin plays this réle of safeguarding the blood from the action of free thrombin. This method of using cephalin needs more careful study since the reaction differs somewhat in differ- ent animals and the nature of the after-effects have not been determined. Very suggestive results were obtained by giving the cephalin by mouth. Evidence was obtained that in this way enough may be absorbed to have a distinct and prolonged effect in lowering the coagulation time of the blood. As an example, the following experiment may be quoted: A dog weighing 6 + kilograms was placed under morphia. Specimens of blood were taken from the jugular vein for the normal coagulation time. 200 c.c. of water were then given by stomach tube and observations were made on the coagulation time during the subsequent two hours. The stomach was then emptied and 200 e.c. of a 0.5 per cent. aqueous solution of cephalin were introduced, and observations on the coagulation time were made during the next four hours. The following results were obtained. The coagulation times were determined in duplicate. — . Normal specimen Before injection. a = 34 to 35 mins. Coagulation time of 2 cc. He Pea Gave 200 ce. water by stomach tube 2. Specimen 1 hr. ; : a = 36 mins. AGies wate Coagulation time of 2 cc. b= 40 mins. 3. Specimen 1 hr. 40 mins. 5 : £2 a = 34 to 35 mins. After water, Coagulation time of 2 ce. b= 36 to 38 mins, 298 HARVEY SOCIETY Stomach emptied (recovered 85 ec.) and then filled with 200 cc. of the solution of cephalin 4. Specimen 40 mins. 4 : a= 30 mins. After Cephalin POSEY ONS tice Voss mins. 5. Specimen 1 hr. 40 mins. ; : : a —25 mins. After Cephalin Coagulation time of 2 ce. |b=20 mine. 6. Specimen 3 hrs. ; ; i a = 28 mins. After Cephalin Coagulation time of 2 ec. | b= 28 aaa ee 7. Specimen 4 hrs. Conpolutionttimercn ice eat mins. After Cephalin b= 24 mins. On the strength of the results obtained from such experiments I have given cephalin by mouth to hemophilic cases to arrest hemorrhage. The results have seemed to be favorable. For example, one of the boys with congenital hemophilia whose cases I have described previously ** reported with a bleeding tooth. The tooth was badly decayed so that the crown broke off and obstinate bleeding set in. Efforts to control this bleeding by local applications of thrombin and cephalin were not entirely successful owing to the difficulty of keeping the packing in position. He was given 100 ee. of cephalin solution twice daily, before breakfast and at bed time. The bleeding stopped promptly and no further difficulty was experienced. I have had occasion to treat in the same way a number of persons suffering from hemophilie joints. Unfortunately most of these cases were at a distance and observations on the blood were not possible, except in two instances. One of the latter was a boy of 10 years ad- mitted to the Harriet Lane Home with a history of joint troubles and long continued bleeding from slight injuries. Examination of his blood showed a hemophilie condition, although the family history gave no indication of a hereditary factor. The coagula- tion time and the prothrombin time were taken. By prothrombin time is meant the time of clotting of the oxalated plasma when recaleified with an optimum amount of calcium. In normal human beings with the procedure I use it is equal to 10 mins. plus or minus. The boy was given 100 ec. of a solution of eephalin daily for 3 weeks while in the hospital, and subse- quently the same amount daily on alternate weeks over a period of 6 months. The blood examinations were as follows: Oe THE COAGULATION OF THE BLOOD — 299 i ti 2 ec. =75 to 80 mins. Before treatment ........... SAO ah et ¢ Tee Prothrombin time = 55 to 60 mins. ¢ 5 Coagulation time 2 cc. = 80 mins. After 3 weeks’ daily treatment. . Bea yen HE ee After six months’ treatment Coagulation time 2 cc. =50 mins. alternate weeks ............- 4 Prethrodbit time = 28 mins. The boy’s condition had improved and although the examina- tions were fewer than desirable they seemed to show that the cephalin had influenced the condition of the blood in the right direction. A second case was treated in a similar way under the direction of Dr. C. R. Drinker who kindly gave me a complete history together with the results of the blood examinations. The patient, a young man 20 years of age, entered the Peter Bent Brigham Hospital with a long history of swollen and painful joints and frequent severe and dangerous hemorrhages. He was given cephalin solution by mouth daily over a period of 13 weeks. Four examinations of the blood were made. Coagulation time — 47 minutes Prothrombin time = 32 minutes Coagulation time —50 minutes ** ) Prothrombin time = 28 minutes Coagulation time = 35 minutes Prothrombin time = 23 minutes eee time — 30 minutes Before treatment..January 22.... | March 23. After 7 weeks’ treatment ....... After 9 weeks ’treatment ....... Proline eG edinioe oa anaatea Coagulation time —34 minutes After 13 weeks’ treatment ...... Vee time = 19 minutes The patient reported a marked improvement in his condition. These incomplete observations need further confirmation— they are reported simply as an indication that the coagulation time in hemophilic bloods can be moved toward the normal by the ingestion of solutions of cephalin. The treatment is simple and not attended by any untoward symptoms and the results obtained so far are sufficiently encouraging to warrant a care- ful trial of the method under conditions more favorable for observation. The Means for Retarding or Preventing the Coagulation of Blood.—There are many methods known by which the coagulation 300 ' HARVEY SOCIETY of the blood may be retarded or prevented; for example, by cooling, by the addition of neutral salts to a certain concentration, by the precipitation of the calcium of the blood, ete. Under cer- tain pathological conditions such as phosphorus or chloroform poisoning or hepatic cirrhosis, which are associated with liver injury or insufficiency, there may occur a marked diminution in the fibrinogen content of the blood and a consequent retarda- tion and imperfection in the clotting, as has been shown by Whipple and Hurwitz ** and Whipple.*® But among the factors of this kind which occur normally in the body and are subject to possible variations under pathological conditions the most in- teresting at present are the so-called antithrombin and a second substance, lately brought to light, which for want of a better name may be designated as antiprothrombin. I should like to call your especial attention to these two substances. Antithrombin.—Nearly all of the important contributors to the literature of coagulation have referred to the presence in the blood of an inhibiting factor, or have at least considered the possi- bility of the existence of such a factor. Schmidt? in the final summing up of his views on coagulation suggests the idea that there may be a something normally present in the blood that hinders the action of thrombin, a substance therefore which might be designated as antithrombin. In his book he devotes much space to a description of the anticoagulating action of two proteins or conjugated proteins, cytoglobin and preglobulin, which he obtained from many tissues. He believed that these proteins contain an inhibiting group or radicle which possibly may be spht off and be found in the blood, but his further description of the action of these substances indicates that they would fall into the group of the antiprothrombins rather than the antithrombins. Nolf in his theory of coagulation *! recognizes the existence in the blood of an antithrombin or, as he prefers to call it, an antithrombosin, and he attributes to it an impor- tant part in the maintenance of the fluidity of the blood. In Nolf’s theory thrombin does not assume the importance attributed to it in other theories. Like fibrin it is the product of a reaction of thrombogen, thrombozyn and fibrinogen, It con- THE COAGULATION OF THE BLOOD 301 stitutes in fact an intermediate or imperfect form of fibrin. Con- sequently the action of the antithrombin is interpreted not as preventing the action of thrombin on fibrinogen but as retarding or preventing the reaction of the three factors just mentioned. Bordet and Delange ** in speaking of the conversion of the pro- serozym to serozym refer in a guarded way to the possible removal of an antagonistic substance normally present in blood. Morawitz recognizes the existence of a substance or substances in plasma and serum which inhibit coagulation and considers it probable that they play a roéle in the circulating blood— although in his theory of coagulation this réle of an antithrombin is not taken into account. In peptone blood and in avian blood the existence of an antithrombin is accepted by many authors. Schickele *? asserts that in the tissue juice obtained under pressure (press-saft) from a number of tissues, particularly from the uter- ine mucous membrane, there is contained an antithrombin. Ina long series of papers Doyon ** with a number of co-workers has de- scribed several methods by which an antithrombin may be ob- tained from various organs. In their later papers the claim is made ** that this antithrombin is a nucleic acid or nucleic acid complex. They state that nucleic acids of animal or vegetable origin prevent the coagulation of blood in vitro. In looking over the literature bearing upon this point it will be noticed that one difficulty and cause of confusion is that few, if any, authors, except possibly Schmidt, make a distinction between antithrombin proper, that is to say a substance that prevents or retards the action of thrombin on fibrinogen, and inhibiting substances that prevent the formation of thrombin by combining with or neutralizing in some way the prothrombin. The importance of this distinction is shown in the next sec- tion. Throughout my work on this subject I have used the term antithrombin to designate a substance that prevents the action of thrombin on fibrinogen, in the way illustrated so markedly by hirudin which is the typical antithrombin. I have devised a simple method of detecting antithrombin and of expressing with some accuracy the relative amount present. The method depends upon the use of purified thrombin and fibrinogen prepared 302 HARVEY SOCIETY according to the methods described at the beginning of this paper; it is carried out as follows: In a series of four tubes (homeo- pathic vials 75 by 15 mm.) one places respectively 2, 3, 4 and 5 drops of a suitable solution of throbbin—one then adds to each tube 10 drops of a fibrinogen solution and notes the time of coagulation. This series constitutes a control, and in my experi- ments the strength of the thrombin solution was such as to cause coagulation in all the tubes within five minutes. For the liquid supposed to contain antithrombin one makes a similar series of four tubes containing respectively, 2, 3, 4 and 5 drops of the thrombin solution. To each of these tubes one adds a single drop of the liquid under investigation, allows a certain definite period of incubation (I have used generally 15 minutes) and then adds 10 drops of fibrinogen to each tube—the time of coagulation is noted and the delay over the control will be in proportion to the antithrombin present. It is quite important in making this test to give a certain period of time for the combination or action, whatever it may be, of the antithrombin and thrombin. The power of the antithrombin to combine or neutralize thrombin is a function of the time, as is illustrated by the accompanying figure. | When the solution investigated contains small amounts of antithrombin this fact may be missed if the fibrinogen is added at once to a mixture of the thrombin and antithrombin. This factor of time in the reaction between these two substances has been noted by a number of observers who have dealt with the power of serum or plasma to neutralize free thrombin added EO at. * This precaution has been entirely overlooked in some recent work published by Dale and Walpole (The Biochemical Journal, 1916, 10, 331). These observers prepared a solution of prothrombin which they examined for the presence of antithrombin. They found none—whether or not their result is correct I cannot venture to say until I have had an opportunity to repeat their work. But it may be said positively that the method they adopted to detect antithrombin was not adapted to reveal its presence, unless there was a large quantity. On the contrary their method would have destroyed any if it had been present. They neglected in the first place to give a period of incubation before adding the thrombin, and in the second place they heated the solution of prothrombin to 60° C. for twenty minutes, a procedure which would have weakened greatly or destroyed the anti- thrombin. In the third place they did not purify their thrombin, but used an impure preparation containing tissue extract (Kinase). THE COAGULATION OF THE BLOOD — 303 Using this method I have been able to show the presence of an antithrombin in the plasma and the serum of blood of a num- ber of mammals (man, dog, eat, rabbit, pig). The amount of antithrombin present varies somewhat in the animals of any one species, and to a more marked extent in the animals of different species. Asa rule the cat’s blood contains the least antithrombin among the animals examined ; next to the cat comes man, and then Fic. 2.—To illustrate the effect of the period of incubation of a mixture of thrombin and antithrombin upon the neutralizing action of the latter. The figures on the abscissa indicate the number of drops, of a thrombin solution to which was added one drop of the antithrombin solution. The ordinates indicate the time in minutes necessary for coagulation when fibrinogen was added to the several mixtures after periods of incubation of ten, thirty and sixty minutes. The ordinates represent therefore the extent of the antithrombiec action. in series, the dog, rabbit and pig. The usual relation of the antithrombin in the blood of man, dog and rabbit is illustrated by the accompanying figure. Minot and Denny *° have attempted to study the antithrombin quantitatively by my method in various pathological conditions in man. In each determination it was necessary to take the blood 304 HARVEY SOCIETY of a normal man as a standard for comparison. Their results are expressed in terms of what they call the antithrombin factor. This factor was obtained from the tubes in which the clotting occurred within a convenient period of time (8 to 25 minutes). The sum of the added times for the several tubes of the control was divided into the added times for the corresponding tubes of the case examined. The normal individuals showed some varia- Fic. 3.—To illustrate the relative amounts of antithrombin in the blood plasma of man, dog and rabbit. The figures along the abscissa indicate drops of thrombin solution to which were added one drop of the oxalated plasma, after heating to remove fibrinogen. The ordinates represent the time of clotting in minutes after addition of ten drops of fibrinogen solution to each mixture. In each case the period of incubation of the thrombin and anti- thrombin solutions was fifteen minutes. tions, the limits of which were expressed in a similar factor by dividing the added times of the case with the highest antithrombin by that of the lowest antithrombin content and the reverse. Expressed in this way the normals ranged from 0.82 to 1.21. Among the pathological cases studied by them some had a distinct increase in antithrombin, e.g., hemophilia with a factor of 2.3 THE COAGULATION OF THE BLOOD — 305 to 2.4 and acute splenomyelogeneous leucemia 1.9 to 2.4, while others showed a subnormal amount, e.g., thrombosis, 0.55 and jaundice, cirrhosis, 0.55. In the blood of the bird and the reptile (terrapin) the amount of antithrombin is much larger than in the mammal, and this fact may be assumed to furnish the explana- tion of the longer clotting time of these bloods when obtained from the vessels without contact with tissue juice. In the blood of dogs that have been peptonized successfully so as to obtain an incoagulable blocd the amount of antithrombin is increased very greatly. This fact may be illustrated by the figures of one cf my earliest experiments. The specimens of blood were oxalated, eentrifugalized and the clear plasma was heated to 60° C. to remove the fibrinogen. _Antithrombin before peptonization: 1. Thrombin 5 drops + Heated plasma 1 drop — incubation of 10 minutes -+ Fibrinogen 10 drops, clotted in 10 mins. 2. Thrombin 4 drops + Heated plasma 1 drop + incubation of 10 mins. + Fibrinogen 10 drops, clotted in 15 mins. 3. Thrombin 3 drops + Heated plasma 1 drop + incubation of 10 mins. + Fibrinogen 10 drops, clotted in 25 mins. 4. Thrombin 2 drops + Heated plasma 1 drop + incubation of 10 mins. + Fibrinogen 10 drops, clotted in 1 hr. 45 mins. Antithrombin 30 mins. after peptonization: 1. Thrombin 5 drops + Heated plasma 1 drop + incubation of 10 mins. + Fibrinogen 10 drops. No clot in 24 hrs. 2. Thrombin 4 drops + Heated plasma 1 drop + incubation of 10 mins. + Fibrinogen 10 drops. No clot in 24 hrs. With 3 and 2 drops of thrombin the same result was obtained. A reéxamination of the blood of the dog after 24 hours showed that its antithrombin content had returned nearly to the normal level. The Nature of Antithrombin—Its Thermolabilty.—Nothing is known definitely concerning the nature of the antithrombin of blood. The active principle known as hirudin secreted by the salivary gland of the leech is a very strong antithrombin, more powerful apparently than the substance found in blood. Accord- ing to the observations of Franz *® hirudin is a protein belonging to the proteose group. It has a high degree of thermostability. Heating to 80° C. has no effect on its activity ; heating to 100° C. 20 306 HARVEY SOCIETY while it weakens its action does not destroy it completely. Ac- cording to some authors (Dickinson) prolonged boiling does not destroy the activity of leech extracts. I have made a number of efforts to separate antithrombin from the blood, but so far have not been successful. After pre- cipitating blood plasma with acetone, for example, I have not been able to find antithrombin in either the precipitate or the filtrate. Apparently it is destroyed by the acetone. As stated above, Doyon and his co-workers have obtained a substance from the tissues which they designated as antithrombin. His earlier experiments indicated that he was dealing with a stable substance which withstands boiling, precipitation by strong acid, ete. In his later experiments he identifies his antithrombin as a nucleic acid or a compound of nucleic acid, but, as stated in the next section, there is good reason to believe that the material he was studying was not an antithrombin proper. The antithrombin of blood is distinctly thermolabile. I have not made a special study of the degree of its thermolability in different bloods. In this respect the time factor is a matter of great importance. Strong peptone plasmas heated to 80° to 85° C. lose their antithrombin activity completely. According to Gasser *? the antithrombin in mammalian plasma is nearly com- pletely destroyed by heating for five minutes to 65° C. In my own experiments I have found that it is destroyed at 70° C., and that heating to 63°-°64° C. for five minutes in some cases also practically removes this action of plasma. In fact heating to only 60° C. for a minute or less weakens distinctly the anti- thrombie action of plasma. In my earlier experiments all plasmas in testing for antithrombin were heated slowly to 60° C. to remove the fibrinogen and any traces of thrombin that might be present, but in my later experiments I have found it preferable to heat the plasma slowly to the temperature just sufficient to coagulate the fibrinogen. In oxalated plasmas this coagulation occurs at 53° to 54° C. and the plasmas so heated show a greater anti- thrombiec action than if heated to 60° C. These observations all show that the antithrombin of mammalian blood is markedly thermolabile, and in this respect differs from solutions of pure THE COAGULATION OF THE BLOOD — 307 hirudin or leech extracts. It is possible that the thermolability of the mammalian antithrombin may be due to the condition in which it exists in the plasma, and that if isolated it would possess a greater degree of thermostability. But so far as our knowledge goes it is more probable that mammalian antithrombin is not identical in composition with the leech hirudin. Source of the Antithrombin.—A number of authors—Dele- zenne,*’ Nolf,*® Popielski **—have stated that the antithrombin in mammalian blood, particularly the antithrombin formed after the injection of peptone solutions is produced in the liver. This conclusion is corroborated by the experiments of Denny and Minot *° in which the determinations of antithrombin were made by my method. They found that long stasis of blood in the liver, as well as perfusion of the organ, was accompanied by a distinct increase in the antithrombin content, and that similar results were not obtained from other organs. Destruction of liver tissue, as in phosphorus poisoning, occasions on the contrary a marked decrease in antithrombin. Outside the variations noted under pathological conditions, which have been referred to briefiy above, changes in the antithrombin content of the blood may be induced by several experimental methods. Peptonization in the dog, as previously stated, causes a very marked increase in the anti- thrombin.