Marine Biological Laboratory Re.e,veH ^^^7 ^5, 1945 Accession No. ^0544 Given By ^^ * Blakiston ' s Son and Go*, Inc. Place Philadelphia, Pn^ PHYSIOLOGY OF BACTERIA PHYSIOLOGY OF BACTERIA BY OTTO RAHN PROFESSOR OF BACTERIOLOGY, CORNELL UNIVERSITY, ITHACA, N. T. With Forty -two Illustrations PHILADELPHIA P. BLAKISTON'S SON & CO., Inc. 1012 WALNUT STREET Copyright, 1932, by P. Blakiston's Son & Co., Inc. PRINTED IN U. S. A. BY THE MAPLE PRESS COMPANY, YORK, PA. THIS BOOK IS DEDICATED TO THE MEMORY OF CHARLES E. MARSHALL of Michigan and of Massachusetts Agricultural College, my teacher and my friend, who re- alized the need for theoretical bacteriology, and found ways and means to further it be- fore its value was generally appreciated PREFACE This book originated from a regular course on the physiology of bacteria given to advanced students at the University of Illinois in the years 1913 and 1914, and again at Cornell University since 1927. It would have been published long ago had not the consequences of the war interfered. This required a complete rewriting of the manuscript; new facts had been discovered; new theories had been substituted for old ones. But the principal conception and the general outline have remained the same. In the seventy years since bacteriology became a science, only a few books on the physiology of bacteria have been published. Duclaux' ''Traite de Micro- biologie" (1900) was followed ten years later by Kruse's ''Mikrobiologie." Then came Euler-Lindner's '^Chemie der Hefe " in 1915. Recently, we have the three volumes of Buchanan and Fulmer's '^Physiology and Biochemis- try of Bacteria," and Marjory Stephenson's '' Bacterial Metabolism," the first one being meant essentially as a reference book, and the second emphasizing only one phase of bacterial physiology, namely, the metabolism, though all other functions are also discussed. It is safe to say that all but the last two books are antiquated. My own intention in writing this book differed some- what from that of the last named authors, and that might give it a right for existence. It is an attempt to co-ordinate the various simplest functions of life, to VIU PEEFACE study each function in itself and in its effect upon the other functions. This study is limited to the necessary functions of life; therefore, motility, phosphorescence, tropisms, and similar conspicuous but unimportant functions have been omitted. By confining the discussion to the indispensable functions of the simplest forms of life, I hope to arouse an inter- est for this kind of work not only among bacteriologists, but among plant and animal physiologists as well. While practically all the discussion in the book refers to bacteria, the principles developed reach beyond the domain of bacteriology, and apply to biology generally. More than that, I believe that some of the principles of biology can be found and studied only with the simplest forms of life, and that general physiology has much to learn from the physiology of bacteria. This book is not intended to be a review of literature on the subject. Such compilations exist already in the books mentioned above. It contains a number of unpublished experiments made with the object of deciding certain doubtful points in physiology. The work is meant to be critical. I have tried to present all theories, and to mention their strong and weak points. To counterbalance any personal view- point, each chapter and subchapter contains a summary, usually separated according to facts and theories. An effort has been made to separate logically, and, if possible, experimentally, growth from fermentation; also to differentiate between the rates of the processes and their end points. Many arguments on growth have arisen because one investigator measures growth by generation-times and another by the maximum number of cells produced. No new definitions are required; no new words have to be coined to distinguish between rates PREFACE IX and endpoints, or to keep fermentation separate from growth. All that is necessary is just a habit of thinking, which is easily acquired. The literature available for the support of many points discussed herein is inadequate. Allowing that a considerable number of publications might have been overlooked by me, the fact still remains that the avail- able material is scanty. And yet, many experiments have been made in the past which could have thrown light on some debated points had they been recorded fully instead of incompletely. Some essential facts, as, the number of acting cells, the length of time of a fer- mentation, or the quantity of inoculum in a growth experiment have not been determined, or at least not stated. Then, too often, data are given only in relative numbers, without the key to absolute measure; while results are given in curves only, from which it is impos- sible to reconstruct the data with sufficient accuracy to use them for any purpose. The making of a book usually involves not only the author, but a number of his colleagues and associates. This has been the case with this book, too. I am obUged to Dr. J. M. Sherman, as head of the department, for providing the opportunity for completing the manu- script. I am obliged to him, as well as to Dr. F. W. Tanner of Illinois University, for reading critically some of the chapters. Several of my former and present associ- ates, especially Mrs. Margaret N. Barnes, Miss A. Jean Ferguson, M. J. Foter, P. Arne Hansen, and J. A. Woerz, have been very helpful in correcting the style and read- ing proof. Further, I have to thank Dr. H. H. Boysen, Dr. D. C. Carpenter, Dr. Hermann Claassen, Mrs. V. A. Coffeen and Dr. G. L. Peltier for permission to use some of their unpublished data, and Drs. W. M. Clark, C. B. X PREFACE Coulter, L. G. Gillespie, E. G. Hastings, H. R. Thornton, and R. W. G. Wyckoff, for permission to use some of their graphs. Finally, I wish to express my special gratitude to the publisher who undertook the printing of this textbook, because he had the confidence that the book will have its place in bacteriological and biological literature. Otto Rahn. Ithaca, New York. TABLE OF CONTENTS Page Preface vii Introduction 1 (a) Significance of the physiology of bacteria 1 (b) Partial functions 4 PART A. ENDOGENOUS CATABOLISM (a) Endogenous catabolism and autolysis 9 (b) Endogenous catabolism of yeast 11 (c) Chemical conception of "wear and tear" 16 (d) Auto-oxidation of the cell 18 (e) Summary 20 PART B. ENERGY SUPPLY OF THE CELL I. Physical Considerations 22 (a) The two sources of energy for life 22 (b) Amounts of energy 25 (c) Methods of measuring energy yields 26 {d) The need for energy 32 (e) Summary 34 II. Chemical Considerations . 35 (a) Enzymes of fermentation 35 (b) The known zymases, or fermenting enzymes 36 (c) Complexity of zymase action 38 (d) Equations of fermentation 40 (e) Successive stages in the fermentation of sugars .... 44 (/) Types of fermentation (sugars, proteins, amino acids) . 49 (g) Interaction between nitrogenous and non-nitrogenous food compounds 62 (h) Mechanism of fermentation, 64 (i) Summary 67 III. The Role op Oxygen in All Fermentations 69 IV. The Oxidation-reduction Potential 82 (a) Electrometric measurements of potentials 82 (b) Reduction potential and acidity 88 xi V(^fff Xii CONTENTS Page (c) Reduction potential indicators 92 (d) Buffering or poising the system 94 (e) Factors involved in reduction 96 (/) Summary 101 V. The Rate op Fermentation 102 (a) Methods of measuring the rate 102 (b) Rate of fermentation of different organisms 104 (c) Decrease of the rate of fermentation 107 (d) Influence of the concentration of microorganisms . . .114 (e) Influence of the concentration of the substrate . . . .114 (/) Influence of temperature 119 (g) Chemical stimulation and inhibition 137 VI. The Endpoint op Fermentation 140 (a) Chemical considerations 140 (6) Constancy of the endpoint 146 (c) Influence of the concentration of cells 146 (d) Influence of temperature 151 VII. Changes in Type op Fermentation 155 (a) Changes due to the chemical composition of the medium 155 (6) Changes due to temperature 156 (c) Influence of concentration of substrate 157 (d) Selection of foods 158 (e) Summary 160 PART C. GROWTH I. General Considerations 162 (a) Measurement of growth 162 (&) Building materials of the cell 163 (c) Sources of carbon 164 (d) Sources of nitrogen 168 (e) Sources of oxygen 172 if) Waste products of construction 173 (g) Balance sheets of feeding experiments 175 II. Relations between Growth and Fermentation .... 180 (a) Efficiency of utilization of energy 180 (6) Efficiency of young and old cells 183 (c) Summary 188 III. The Growth rate 189 (a) Methods of measuring the growth rate 189 (b) The lag period 192 (c) The decreasing growth rate 199 CONTENTS Xlll Page (d) Concentration of food and the growth rate 203 (e) Temperature and the growth rate 213 (J) Chemical stimulation of the growth rate 227 IV. The Endpoint op Growth 230 (a) Causes of the endpoint 231 (6) Endpoint and chemical composition of food 237 (c) Endpoints with an abundance of food 244 (d) Influence of the concentration of food 245 (e) Influence of temperature 251 (f) Influence of chemical stimulants 254 V. Mechanism of Growth 257 (a) Molecular structure of the cell 257 (b) The growth process 264 (c) VariabiUty of the growth rate 266 (d) Summary 269 PART D. MECHANISM OF DEATH I. Definitions of a Dead Cell 270 II. Methods of Measuring Death 272 Endpoint method, plate counts, staining method . . . 272 III. The Logarithmic Order op Death 274 (o) Order of death of higher organisms 274 (6) Order of death of bacteria 278 (c) Mathematical formulation 280 (d) Interpretations of the order of death 282 (e) Physiological youth 288 (J) Decreasing death rates 290 (g) Frequency of the logarithmic order of death 293 (h) Adaptation and selection of resistant strains 298 (i) Summary 301 IV. Death of Dry Bacteria 303 (a) Order of death 303 (6) Oxidation as the cause of death 305 (c) Temperature coefficient of death 308 (d) Summary 312 V. Death by Freezing 313 VI. Death by Heat 317 (a) Methods of measuring 317 (b) Order and rate of death at different temperatures . . .319 XIV CONTENTS Page (c) Cause of death by heat 326 (d) Death by heat in concentrated solutions 330 (e) Thermal deathpoints 332 (/) Summary 334 VII. Death BY Poisons 335 (a) Methods 335 lb) The order of death 339 (c) Chemical considerations 342 (d) Effect of concentration of poison 346 (e) Mechanism of poisoning 355 (J) Temperature coefficient of poisoning 363 (g) Other theories of chemical disinfection 366 (h) Summary 368 VIII. Death by Surface Tension Depression 369 IX. Death BY Light AND Other Rays 371 X. Death through Age 374 XI. Death by Starvation 377 (a) Starvation in water 377 (6) Starvation in salt solutions 380 (c) Starvation from lack of nitrogenous food 383 (d) The fundamental reaction in starvation 384 (e) Summary 385 XII. Death by Suffocation 387 XIII. Summary on Death 392 APPENDIX I. The size of microorganisms 395 II. Multiplication of bacteria 398 III. The fermenting capacity of the cell 404 Author Index 409 Subject Index 429 PHYSIOLOGY OF BACTERIA GENERAL INTRODUCTION (a) THE SIGNIFICANCE OF THE PHYSIOLOGY OF BACTERIA Life processes are largely chemical or physical proc- esses, and it would be wasted time and effort to try to understand the very complicated and interlinked actions and reactions, the total of which we call life, unless the best physical and chemical methods were used for their analysis and interpretation. There may be processes other than physical and chemical involved in life. Most of those changes brought about by life processes which we, from our human viewpoint, consider most important, namely, the prog- ress of civilization, the development of creative thoughts, cannot as yet be explained by chemistry and physics, and perhaps never can be. Whether there is a soul, an agent independent of the body, in man or animals may ever be an open question. So much is certain, however, that it will not cause any actions contrary to the laws of physics or chemistry; else, its existence would have been established long ago. Just as physics is more than mere mathematics, and chemistry cannot be explained altogether by physics, so biology cannot be accounted for exclusively by physical and chemical equations. This book deals only with those aspects of biology that can be treated from the physico-chemical viewpoint. Its intention is to give a conception of the life mechanism of the simplest organisms, and no other group of organ- 2 PHYSIOLOGY OF BACTERIA isms is more fit for this purpose than the lowest fungi which Hve on dissolved food. That some of the human physiologists do not agree entirely with this viewpoint, may be seen from the following statement in Bayliss' ^'Principles of General Physiology" (1924): ''Without denying the great value of the comparative method in ehminating merely incidental phenomena, it must be pointed out that this very simplicity is, in the majority of cases, a disadvantage. The same organ, or even cell, fulfils a variety of purposes which, in the higher organisms, are relegated to distinct groups of cells. Moreover, the size of the organism is of much importance. The science of nutrition would be almost impossible without the larger, warm-blooded animals. The advantage of the increased rate of reactions, owing to the higher temperature, is not to be undervalued. The physiology of unicellular organisms, although of considerable importance in special aspects, is not to be regarded as a ''general physiology." Indeed, if the choice had to be made between the investigation of simple or complex organisms alone, there is no doubt that a much more general and fundamental body of doctrine would be obtained from the latter." Of course, it would be ridiculous to try to belittle the value of the physiology of man and of mammals as a guide to the understanding of the functions of the body of all kinds and types of animals and bacteria, for it is quite evident that the original conception of the nutrition of bacteria was based exclusively upon its analogy with animal nutrition. There are, however, certain facts which could not be observed in the highly specialized physiology of animal tissues, consequently, new concep- tions have been brought into general physiology through the study of bacteria; e.g., the conceptions of anaerobio- sis; of thermophilic organisms; of nitrogen fixation; of fermenting endo-enzymes; the understanding of the organic chemistry of alcoholic fermentation with its intermediate stages; and a new understanding of the laws of death. In fact, the conception of higher plants GENERAL INTRODUCTION 6 and animals as well organized cell federations was possi- ble only after the existence and physiology of unicellular beings had been established. Considering the very little attention paid hitherto to the physiology of bacteria, it seems quite possible that many new contributions to general physiology might originate from the study of the simplest forms of life. Most favorable to such studies is the attitude of Putter in his book on Comparative Physiology (published as early as 1911): "It might perhaps surprise many to find so much space given to the discussion of the biochemistry of molds, yeasts and bacteria; however, with just these organisms, a clear conception of the different metabolic processes can be obtained, and the aim of a comparative physiology seems to me to consist rather more in working out the various principles with the organisms best suited for this purpose than in enumerating a larger number of experiments from all the groups of organisms existing." It is not conceivable that any scientist would con- sider it possible to substitute physiological experiments with the liver or the thyroid gland by experiments with bacteria. Nor is it likely that anybody would believe nutrition experiments with higher animals superfluous and replaceable by experiments with lower fungi. The very simplicity of the metabolism of some microorgan- isms makes them fit for a good many experiments which cannot be carried out with higher plants or animals. That is why the physiology of microorganisms is of con- siderable importance in general physiology. Some of the advantages of using microorganisms in physiological experiments of a more general nature, are: (1) the absolute control of temperature during the entire experiment, and the possibility of working at several different temperatures; (2) the absolute control of food which may be chosen from chemically pure and well 4 PHYSIOLOGY OF BACTERIA defined compounds only; (3) the control of the concentra- tion of food; (4) the complete analysis of all metabolic products; (5) the circumstance that all experiments are averages of millions or billions of individual cells which are all of the same kind and the same age; (6) the very rapid rate of growth; (7) the possibility of separating to a considerable extent the food for energy supply from the food needed for building material. There is one disadvantage in all experiments with lower fungi, i.e. their minute size. (A general concep- tion of the meaning of the size of microorganisms can be obtained from Table 132 of the ' ' Appendix.'') This small size makes the determination of the ''crop" diffi- cult, especially with bacteria, while with molds and yeasts, it is fairly easily accomplished. But even with bacteria, direct and indirect methods have been devised which will be found in the chapter on Growth. Not only from the viewpoint of general physiology is the physiology of microorganisms important; it is also of great value in applied bacteriology. There can be no doubt that future developments in the latter field will depend to a large extent upon a better understanding of the physiology of microorganisms. The canning indus- try, as well as chemical and physical disinfection, need urgently a deeper knowledge of the laws of death of bacteria, while the fermentation industries require more detailed information regarding the different types of fer- mentations, and of the factors influencing them; the bread yeast industry is likewise interested in further advances in the studies of growth and nutrition of microorganisms. (6) PARTIAL FUNCTIONS Primitive Functions. — The only way we can study the mechanism of life is to separate life into a number of GENERAL INTRODUCTION 5 individual processes, and then study each process, or group of processes, by itself (if this be at all possible) and to determine also the relations between various processes. The number of separate processes, distinguishable to the scientist, is practically unlimited, owing to the fact that we can subdivide, again and again, and consider as a separate process each detail of a chemical reaction, as well as the diffusion of each compound from one cell to another. For scientific study, we divide life manifesta- tions into groups which are logically comparable, and which fulfil the same purpose. Such groups of processes, we call functions, or more correctly partial functions, since each of them is only part of the complete life proc- ess. The eating of meat by a cat and the eating of grass by a cow are such separate processes, classified under the same function; while the running of a horse, the flying of a bird, and the swimming of a bacterium give us an idea of the variability of the function which we ordinarily call motility. The distinction of functions is primarily a matter of convenience or viewpoint, and this varies with one's personal attitude, as well as with the progress of knowl- edge. The viewpoint of Pasteur that fermentation and respiration were but two modifications of the same function has been contested most persistently for a long time. Further, it must be stated that we distinguish more functions, ordinarily, in the higher organisms; in other words, what is only one function in the protozoon, may be divided into four or five functions in man. The terms function and partial function relate, therefore, not to well defined groups of processes; and they cannot be well defined, because they are interlinked in such a way that some of the processes are necessary prerequisites for the others, — e.g., respiration is the prerequisite for 6 PHYSIOLOGY OF BACTERIA heat formation in the body. We must distinguish, there- fore, between dependent and independent functions. If the death of a cell by poison is independent of the rate of growth, or of the rate of fermentation, we may call it an independent function. Growth, however, depends upon the amount of available energy. Further progress in knowledge may disclose to us that the assumedly independent functions are, after all, dependent ones. A most natural question arises in this connection: Which functions are found in all living beings, i.e., which functions are essential for life? This question leads us immediately to the simplest forms of life. The most characteristic quality of a living thing which distin- guishes it from all dead matter, is growth. Growth, meaning the construction of complex molecules from less complex molecules, is, however, a dependent func- tion and requires energy, and energy is, with chlorophyl- free organisms, provided by a function of the body which we call respiration, or, in the case of bacteria, fermenta- tion; here we are using the term in its widest meaning. Growth is not identical with multiplication. How- ever, bacteriologists commonly use these terms as synonyms, probably because growth of a culture is identical with multiplication of bacteria. The difference is very great cytologically, but in these biochemical studies, no emphasis shall be placed upon the difference, and the two functions shall be treated as one because we have hardly any means to differentiate them chem- ically (see also Introduction to ^^ Growth'' p. 162). Besides energy formation and growth, all living organ- isms share one more property: they all must die. Star- vation, poison, excessive heat will kill all living cells, and we know that organisms will die of old age in a natural way. Though there are great variations in GENERAL INTRODUCTION 7 resistance and longevity, a living cell that cannot die is unthinkable. The mere absence of growth and energy formation does not necessarily mean death; the same symptoms are characteristic for dormancy. We can imagine con- ditions where we have neither growth nor fermentation and yet no death. This is probably the case with dry bacteria in a vacuum at very low temperatures. The discussion of the causes of death will show that dying is a chemical process. Another function is also found regularly, or at least assumed for all cells, i.e. endogenous catabolism. Catabo- lism is assumed for all cells because all cells die from starvation, death by starvation meaning that a deteriora- tion of some cell compounds occurs which, in normally nourished cells, is counteracted by constructive processes. Endogenous catabolism shall mean here the deteriora- tion of cell compounds, but not of the food taken in. This definition may eventually lead to difficulties, for we may assume that the food combines with some cell compounds before being decomposed. But we shall try to avoid hair-splitting as much as possible. We have, then, endogenous catabolism, death, energy formation and growth (and, eventually, multiplication) as the four functions which are indispensable for life, the primitive functions of hfe, so to speak. This book is limited entirely to the discussion of these four functions. Of these four functions, endogenous catabolism is the least dependent of all. The process of dying may, in some cases, be interlinked with catabolism, as in death by starvation. Growth is the most dependent function, being possible only when energy is furnished, and inseparable from catabolism. It would be impossible to discuss growth without having first studied energy 8 PHYSIOLOGY OF BACTERIA production. For this reason, the discussion begins with the independent, though least-known function of endog- enous cataboUsm. The chapter on growth is preceded by a discussion of the energy Uberation. The causes of death might have been placed before the chapter on growth, but it seemed too odd to discuss death before growth. Hence, the chapter on death has been placed at the end. PART A ENDOGENOUS CATABOLISM, OR NORMAL DETERIORATION OF LIVING MATTER (a) ENDOGENOUS CATABOLISM AND AUTOLYSIS As all living cells will die if kept without food for a varying number of days (except when in a dry state), it must be assumed that deteriorating changes in living cells take place continuously, and can be counteracted only by assimilation of food. These changes will ultimately cause death of the cell if continued for too long a time. Bacteria are no exception to this rule as will be shown in the chapter on Death by Starvation (p. 377). As defined by Folin (1905), this gradual and constant decomposition of living matter may be called endogenous catabolism. All measurements are limited to protein changes, but other compounds may undergo similar changes. Endogenous catabolism has been studied for more than one hundred years with man and animals. It is usually measured by the quantity of nitrogenous matter excreted during starvation or with protein-free diet. The quantity of nitrogen excreted by a starving man seems to vary considerably, but the average is about 6 to 10 gm. of nitrogen per person per day. This means about 80 mg. per kg. of body weight, or, estimating on 10 kg. of protein per average person, it amounts to about 1 gm. of nitrogen for each kg. of dry protein, or to 0.6% of the protein nitrogen. 9 10 PHYSIOLOGY OF BACTERIA Normally chlorophyl plants do not excrete protein cleavage products. There are many observations indi- cating that green plants in the dark not only break down their carbohydrates, but also their proteins. Yet, their synthetic action seems to be so powerful that all the products of nitrogenous catabolism are re-utilized somewhere in the plant, and do not leave it. Differing from this endogenous catabolism of the starving organism is the autolysis of the dead organism. When tissues of a dead animal are held at 37°C. under aseptical conditions, the protein readily breaks down to simpler compounds. This self-digestion or autolysis of tissues is due to proteolytic endoenzymes. The products of autolysis are in a general way compar- able to the products of peptic or tryptic digestion, and are quite different from the products of endogenous catabolism. Another difference between the two processes of protein decomposition lies in the rate of the process. As an example, Jacoby (1900) observed that normal liver of a dog, after twenty-four hours of autolysis, had lost 17.9% of the total nitrogen, or 20% of the protein nitrogen. This is an enormous rate compared with the 0.6% of nitrogen loss by endogenous catabolism by a starving man per day. It is quite necessary to differentiate sharply between these two different processes. This distinction was not made by Rubner (1913), but is drawn very sharply by Euler (Euler-Lindner, 1915): "Yeast contains plenty of proteinases and peptases. In the living cell, the destructive action of these is in adjustable equilibrium with the construction of protein. If, however, synthesis is retarded in comparison with the cleaving action of the digestive enzymes, we observe the phenomenon which is usually called autolysis . . . ENDOGENOUS CATABOLISM 11 Autolytic action occurs, as far as is known, after the 'death' of all cells and tissues; but it is not only in the dead cells that the own protein is attacked . . . . . . Several investigations tell us about the behavior of yeast in relatively large amounts of water, with or without disinfectants. If no nitrogenous food is offered . . . the yeast loses its own protein. This loss may be temporary and need not lead to the death of the cell. It may be supposed that the same reactions take place even under normal conditions, but are covered up partly or entirely by synthetic processes. This process shall be called 'endoproteolysis.' If this endoproteolysis is not compensated by constructive proc- esses, then, finally, those protoplasmic groups which are indispensable for cell life will be attacked. The cell loses the power of reproduction and undergoes complete destruction. This latter process of dissolu- tion shall be called autolysis. This autolysis which starts at the death or afterwards, is essentially an irreversible process initiated by endoproteolysis . ' ' (6) ENDOGENOUS CATABOLISM OF YEAST The amount of nitrogenous living matter changed per day has been found with man to approximate 0.1 gm. of nitrogen per kg. of body weight. If this same ratio would hold with microorganisms, the task of deter- mining catabolism would seem hopeless, for the total solids of the cells in one liter of broth culture of Bad. coll amount to only 220 mg. and their catabolic prod- ucts per day would total only 0.2 mg. of nitrogen per liter. But this ratio does not hold. Rubner (1908) has pointed out that in animals, the relative amount of endogenous catabolism increases with the decrease in size, and Rubner believed the surface to be the important factor determining the rate of endogenous catabolism. But even under these favorable circumstances, the amount of catabolized nitrogen is so small, owing to the minute quantity of living matter in a bacterial culture. 12 PHYSIOLOGY OF BACTERIA that, so far, endogenous catabolism has been studied only in yeasts. As early as 1803, Thenard observed that yeast fer- menting in pure sugar solutions loses protein. Yeast may lose 30% or more of its nitrogen, and finally loses the ability of rapid fermentation (Schiitzenberger, 1876). Pasteur and many others have observed a decrease of nitrogen in yeast, but few of these experiments have been made quantitatively. Though endogenous catabolism of yeast is more rapid than with higher organisms, autolysis is still swifter. Rubner (1904) observed in living yeast kept in a 20% sugar solution, a loss of nitrogen of 58% in six days, while the same yeast suspended in water + toluol, i.e. after being killed, lost 90% of its nitrogen in four days. Effront (quoted from Euler-Lindner) found that 500 gm. yeast in distilled water lost 39 mg. of nitrogen in twenty-four hours while the same yeast in water -|- alcohol + hydrofluoric acid lost 72 mg. After five days, the latter liquid, from the dead yeast, contained ten times as much nitrogen as the liquid from the living yeast. Rubner (1913) made extensive experiments on the nitrogen catabolism of fermenting yeast cells; the main results are given in Table 1. The first two experiments refer to yeast kept in 20% sugar solution at 38°C. for three and six days. Different yeast was used in these two experiments. For the third experiment, the yeast remained in a 10% sugar solution at 28°C. for twenty- four hours, was then centrifuged out, and suspended in new sugar solution; after another twenty-four hours, the sugar solution was renewed in the same way, and this was repeated five times. In connection with the last experiment, it will be of interest to know that the total quantities of alcohol ENDOGENOUS CATABOLISM 13 produced by the same yeast on each of the successive days were 6.70%, 4.16%, 2.30%, 1.19%, 0.74% and Table 1. — Nitrogen Losses by Fermenting Yeast Time Mg. nitro- gen in yeast Relative N quantities Micro- scopic counts (millions) Plate counts (millions) Relative numbers of viable cells Yeast in 20% sucrose solution at 38°C. for three days Start After 3 days.. 62.7 100 72,000 51,850 52.2 83 24,100 40,600 100 78 Yeast in 20% sucrose solution at 38°C for six days Start 430 65 100 15 176,200 173,000 133,000 100 After 6 days 4,975 3.7 Yeast changed every twenty-four hours to a new solution of 10% sucrose, at 28°C. Start After 1 day 2 3 4 5 6 93.5 78.9 66.3 55.9 47.1 39.6 28.2 100 84.4 70.9 59.8 50.3 42.3 30.0 84,600 62,845 66,605 66,950 88,600 105,400 101,600 0.52%. Towards the end, the yeast had almost entirely lost the power to ferment. A similar observation was made by Meyerhof (1916a) in his studies of the Nitrobacter. This organism builds its entire cell from nitrite, minerals, oxygen and CO2. Meyerhof observed that the rate of oxidation slowly decreased if CO2 was removed from the atmosphere. 14 PHYSIOLOGY OF BACTERIA The nitrate bacteria were starving for CO2, and that prevented a normal formation of the oxidizing agent in the cells. Table 2. — Oxygen Consumption by Nitromonas, with and without Carbon Dioxide (mm. 3 oxygen consumed per 2 hour period by 2 c.c. of culture) After 4 hours After 6 hours After 8 hours WithCOz.... Without CO2. 45.5 44.5 49 39 50 mm. 3 O2 31mm.3 02 The chemical nature of the nitrogenous compounds excreted by living yeast is quite unknown while the autolytic products have been studied extensively. It is very probable that most of the excretions of living yeast cells cannot again be assimilated by the same yeast while by far the largest amount of the autolytic products can be used again for cell construction (Lindner, 1905). Probably, all data on the endogenous catabolism of yeast have quite a high degree of error, because they may represent not only the products of living cells, but also of a certain percentage of dead cells. This per- centage is practically unknown in all experiments. Since autolysis of dead cells proceeds much more rapidly than that of living cells, a relatively small number of dead cells will cause a considerable error. The contention of Rubner (1908), that the more rapid rate of endogenous catabolism in small organisms is caused by the relatively larger surface, seems rather improbable since the compounds to be decomposed are already within the cells. Two further explanations seem possible: either the higher organisms re-utilize a large amount of the decomposed protoplasm, or the yeast ENDOGENOUS CATABOLISM 15 protoplasm is more readily decomposed than that of mammals. Both explanations have some facts in their favor. In starving animals, different organs suffer in different degrees; the heart does not lose weight at all (Bayliss p. 274), and seems able to use part of the catabolic products from other tissues. Creatine appears in the urine during complete starvation. It disappears, i.e., it is re-utilized when carbohydrates are given, but not when fats are given. The carbohydrates enable a synthesis which retains the creatine in the body. With yeasts, we need not consider different tissues. All cells are alike physiologically, and what has been excreted as useless waste by one cell cannot be utilized by any other cell of that same culture. Table 3. — ^Loss op Nitrogen by 5 Gm. op Yeast at 28°C. Mg. N in yeast Loss of nitrogen I II III I II III Fresh 24 hours in water 24 hours in sugar solution 108.0 93.9 90.8 109.1 100.0 92.4 113.0 99.8 103.7 14.1 18.8 9.1 16.7 13.2 9.3 Rubner (1913) believed that endogenous catabolism of yeast goes parallel with the rate of fermentation. He gives some data on the loss of nitrogen by yeast in water and sugar solution, and in two of the three experiments, the loss in water is larger. But they were not carried out with pure cultures; there was putrefaction in the culture without sugar, i.e., a synthesis of bacterial pro- toplasm from yeast excretions. This makes the experi- ment quite doubtful. 16 PHYSIOLOGY OF BACTERIA In another place, Rubner mentions that yeast in sugar solution lost 58% of its nitrogen, while the same yeast without sugar lost 89%. (c) CHEMICAL CONCEPTION OF "WEAR AND TEAR" The constant excretion of nitrogenous products by the starving organism has been explained as the result of the continuous '^wear and tear" of the cells. It seems that most physiologists consider this to be a mechanical wearing of the organism, and what originally was to be just a simile is now considered the true explanation. The books of Bayliss and of Cathcart give no other. Yet, this interpretation seems rather unsatisfactory. A mechanical wearing of a resting man or animal during a starvation experiment is difficult to conceive. Many experiments have shown that mechanical labor, unless it is overdone, does not increase, to any marked degree, the wear and tear, as measured by the nitrogenous excretions. This is quite contradictory to any mechani- cal conception of the wear and tear (Bayliss p. 267, Cathcart p. 133). Euler (Euler-Lindner, 1915) offers a biochemical theory already quoted on p. 10. It is essentially the same as that of Rahn (1915) who explains endogenous catabolism as the result of the instability of living protoplasm or parts of it. All cells contain enzymes. All enzymes are unstable and deteriorate with time. While it is generally admitted that enzymes in solution cannot be kept permanently active, it is often stated that they do not undergo deterioration inside of the cell. The only evidence to substantiate this claim is the fairly constant enzyme content of the cell. This proof is not conclusive since we know that all cells can produce the enzymes which are found in them. ENDOGENOUS CATABOLISM 17 Rahn assumed that the enzymes within the cell behave as they do outside. The deficiency in the cells from enzyme deterioration is made up through the synthesis of new enzyme by the cell itself. The cells have the power to produce enzymes not once only, during growth, but as often as is necessary, to keep the supply normal, as has been shown by many experiments with yeast zymase by Buchner, Buchner and Hahn. A certain amount of the cell enzymes thus deteriorates at a constant rate under constant conditions. The deteriorated enzyme, being a waste product, is decom- posed by an autolytic enzyme; perhaps parts of the products are utilized for the synthesis of new enzyme; the rest of the cleavage products leaves the cell. Other unstable cell constituents will have a similar fate, and the total sum of all such deterioration is what we call endogenous catabolism. Meyerhof's experiment with the Nitrobacter (see Table 2) is very interesting in this respect. The oxidizing enzyme seems to be very readily deteriorated, and can be rebuilt only from CO2, not from its own cleavage products. This chemical interpretation of the wear and tear gives a simpler explanation to the endogenous catabolism of the working muscle. The work of the muscle requires additional energy, i.e., a faster combustion of food in the cells. This combustion is brought about by enzymes in the cells, but it is a well known rule of enzyme action that they do not deteriorate faster when acting than when resting. Whether the muscle works or rests, the rate of enzyme deterioration is not influenced. But the rate is greatly influenced by an increase in temperature (see p. 124). As soon as the work of the muscle comes to a point where there is a considerable 18 PHYSIOLOGY OF BACTERIA increase in temperature, we should expect a much more rapid deterioration of the enzyme. This is actually the case. And it has also been found that endogenous catabolism increases in fever, when the body tempera- ture is raised without any increase in mechanical work (Bayliss p. 272). If it can be shown that endogenous catabolism has a very high temperature coefficient, the above theory would be well substantiated. Rubner's few data give little hope that this may be the case. However, the results of a more exhaustive study must be waited for before conclusions can be drawn. A decided reduction of the nitrogen content of starving Asper- gillus mycelium has been observed by Terroine, Wurmser and Montane. In 5 days, it dropped from the normal 6% N of the mycelium solids to about 2.6% in the presence of glucose, and to 2.3-2.8% in its absence. (d) AUTO-OXIDATION OF THE CELL It seems most probable that oxygen will play a con- siderable role in the normal breakdown of living matter. Most cells have the ability to use oxygen for the produc- tion of energy, and it is likely that starving cells will apply oxidation to the less essential cell constituents in order to obtain energy. Oxidation seems a simple way of using waste products of endogenous catabolism to greatest advantage. It will be shown later that the life mechanism is an intricate combination of oxidation and reduction proc- esses. Growth is essentially a reduction process, energy formation is primarily an oxidation process, though the oxygen used for oxidation need not be molecular oxygen. Part of the energy obtained by direct or indirect oxida- ENDOGENOUS CATABOLISM 19 tion is required to furnish the energy necessary for reduction. Table 4. — Influence of Oxygen under Various Pressures at 37°C. UPON Different Bacteria (24 Hours' Exposure) Oxygen pressure in atmospheres 1 IK 2 3 4 10 20 30 40 50 60 70 75 Effect on growth (0 means no growth during twenty-four hours' exposure) B. anthracis Bad. alkaligenes, 3 strains. Vibrio cholerae Ps. pyocyanea Bad. typhosum, 3 strains. . Bad. paratyphosum Proteus vulgare Bad. coli, 4 strains Bad. enteritidis Micr. pyogenes aureus. . . . 0 0 0 0 0 0 0 0 0 0 + 0 0 0 0 + 0 0 0 0 + + 0 0 0 + + 0 0 0 + + 0 0 0 + + + 0 0 + + + 0 0 0 + + + + 0 0 0 Effect on viability (0 means dead after twenty-four hours' exposure) B. anthracis Bad. alkaligenes, 3 strains. Vibrio cholerae Ps. pyocyanea Bact. typhosum, 3 strains. . Bad. paratyphosum Proteus vulgare Bad. coli, 4 strains Bad. enteritidis Micr. pyogenes aureus. . . . + 0 0 0 0 0 0 0 0 0 0 0 + + + 0 0 0 0 0 0 0 0 0 0 + 0 0 0 0 0 0 0 0 0 0 + + + + + + + + + + + + + + + + + + + + + 0 0 0 + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + The cell disposes of molecular oxygen for two reasons: to obtain energy, and to prevent oxidation of vital cell constituents. If the oxygen concentration is increased, the cells may not be able to prevent oxidation of the essential parts, and the cell dies. Table 4 gives some 20 PHYSIOLOGY OF BACTERIA experiments by Berghaus (1907) showing how different organisms can resist different oxygen concentrations to which they had been exposed for twenty-four hours. Many years later, Callow (1924) measured quanti- tatively the amounts of oxygen taken up by washed living bacteria suspended in buffer solution. The amounts varied from 5 to 25 c.c. of oxygen (7-35 mg). per gram of bacteria solids per hour, and remained con- stant for many hours. Even after drying in vacuo and re-suspending, three species took up as much oxygen as before, while Ps. pyocyanea used very much less oxygen. The experiments included Ps. pyocyanea, Ps. fluorescens Bad. prodigiosum, alkaligenes, coli, proteus B. megatherium, suhtilis Micr. pyogenes aureus, Sar. aurantiaca Mycohact. phlei The experiments were also extended to Strept. lactis and Clostr. sporogenes; these took up less than 0.5 c.c. during the first hour, and practically none thereafter. This seems to contradict the fact that these organisms are very sensitive to oxygen. An explanation will be tried on p. 74. (e) SUMMARY OF FACTS Yeasts and bacteria if left in moist condition without food for a number of days will lose nitrogenous com- pounds, which can be found in the liquid surrounding them. The same process takes place if the organisms are given carbohydrates, but no nitrogenous food. The decrease of nitrogenous material in the yeast cell is accompanied by a decrease in its fermenting power. Ultimately, the cells die. Nitrobacter loses very promptly part of its power to oxidize if carbon dioxide is ENDOGENOUS CAT ABOLISH 21 withheld. This is the only carbon compound from which this organism can build its cells. Aerobic and facultative bacteria, suspended in buffer solution after washing, will take up from 7 to 35 mg. of oxygen per gram bacteria solids per hour for two, and even for ten hours at a constant rate. Streptococcus and Clostridium took up less than 0.5 c.c. during the first hour, and practically none later. SUMMARY OF THEORIES The loss of nitrogenous matter by starving cells is supposed to be a normal process of all living cells which is not ordinarily conspicuous because the opposite proc- esses, growth and reconstruction, are more prominent. The nitrogenous products of endogenous catabolism are the cleavage products of some thermolabile com- pounds in the cell, such as enzymes, and living proto- plasm. These compounds deteriorate at a definite rate inside the cell as well as outside. The cell has the power to replace these compounds from new food. In the absence of food, or in the absence of appropriate food, the cell has ultimately not enough of these essential molecules left, and dies. In the case of Nitromonas, the lack of non-nitrogenous food caused the same symptom because with this bacterium, all the carbon in any of the cell constituents must come from carbon dioxide. PART B ENERGY SUPPLY OF THE CELL I. PHYSICAL CONSIDERATIONS (a) THE TWO SOURCES OF ENERGY FOR LIFE In the entire living world, only two sources of energy- are known to be available to living cells, namely light and chemical energy. Light can be utilized only by the chlorophyl plants, the bluegreen algae and a small number of prototrophic bacteria. Chemical energy is utilized by all living organisms. Light is of importance in the growth of certain plants which possess no chlorophyl; but it acts there only as a stimulus. Rhizopus grows better in diffuse light than in dark, but it requires the same amount of food to produce the same amount of mycelium; no energy of light is ''stored" in the cells of the mold. Such stimulation by light causes the daily rings in cultures of bacteria and molds (Hutchin- son, 1907). Heat, electricity, or other forms of energy cannot be utilized by organisms as available energy. A higher temperature may cause a more rapid functioning of the cell, yet it is merely a difference in the rate of the process, and a cell cannot economize on food by utilizing the heat energy of the surrounding medium. A starving cell is not benefited by an increase of temperature. Probably the cells lack the ^transformers" or ^^acti- vators" for all these forms of energy. Electricity, heat, and probably even light are not used as such by the cells. They must be transformed into another kind of energy, and we know only the transformer for light, i.e. the chlorophyl. 22 ENERGY SUPPLY OF THE CELL 23 From this, we must conclude that energy liberated by one cell cannot be utilized by other cells, and that even in tissues, neighboring cells have no means of exchanging any surplus energy directly. It follows further that the chemical energy must be liberated within the cell, for if it were liberated outside, it would have to assume one of the forms mentioned above (heat most probably, or eventually light or electricity) which cannot be utilized by the cells. Whether the mitogenetic rays (Gurwitsch, 1929) act merely as stimulants, or whether their energy can be actually utilized is entirely unknown. This must be kept in mind with bacteria acting upon insoluble food. Only that part of the energy becomes available which is produced in the cell; the energy liberated outside of the cell merely raises the temperature of the medium, without any further benefit to the organism. Table 5. — Energy Liberated by Enzymes from 1 Gm. of Substrate Secreted enzymes Endo-enzymes Pepsin, Trypsin, Rennet 0 calories 4 calories 9 . 3 calories (Rubner) 10 calories 23 calories Lactacidase Alcoholase Urease 82 calories 149 5 calories Lipase (Rubner) 239 calories Invertase Maltase Vinegar-oxidase. . . 2, 530 calories Lactase Since only that part of the food, which is decomposed in the cell is of benefit to the bacterium, it is essential that the food pass through the plasma membrane. The compounds need not necessarily be water-soluble to accomplish this; a lipoid-soluble compound might 24 PHYSIOLOGY OF BACTERIA pass through also. As a rule, we find that the cells change insoluble compounds into a soluble form by secreted enzymes. The energy liberated by secreted enzymes usually is small, that from endo-enzymes is quite large. (Table 5.) The chemical decomposition of compounds within the bacterial cell, for the purpose of liberating energy, shall be taken in this book as a definition of fermentation. It is practically the same definition which was given by Hugo Fischer (1902) : 1. Fermentation is an intracellular process. 2. It yields products essentially different from the materials fermented and which are not mere parts of this material. 3. The end products are useless or harmful to the cell. 4. It produces ''vital" energy (i.e., energy available for cell processes). The definition includes all processes, organic and inorganic, aerobic and anaerobic, with and without gas formation, of proteins, fats, carbohydrates, acids, etc., if only they fit the above definition. That fermentations provide cell energy as respiration does with animals, was recognized by Pasteur (1876), though he did not furnish absolute proof for this statement. He proved that yeast cells grew in the absence of oxygen, and under these conditions caused fermenta- tion, but he did not exclude other possible sources of energy, especially that from the proteins. Since then, several cases have become known where fermentation is the only source of energy, under condi- tions where no other source is imaginable. To this group belongs the fermentation of a mineral-ammonia-sugar solution by yeast, of a mineral-urea solution, with traces of organic acids, by urea bacteria, and of a mineral-ammonia solution by nitrate bacteria. The growth of Oidium lactis and of Mycoderma in a lactic acid- mineral solution -fNHs, and the growth of most prototrophic organisms are examples of simple oxidations furnishing the energy for growth. ENERGY SUPPLY OF THE CELL 25 In discussing the energy supply of cells, it must be considered that most cells can utilize various kinds of food, and may even be able to use the same kind of food in two different ways. It is easily possible to cultivate some typical sugar-fermenters in the absence of sugar. Yeast will grow in sugar-free broth; lactic acid bacteria can grow in peptone solutions; Hydrogenomonas as well as urea bacteria are commonly cultivated on ordinary nutrient gelatin. Most molds seem to be omnivorous, while some of the prototrophic bacteria are limited to only one source of energy. Yeast will ferment sugar to alcohol, but can oxidize it completely if an abundance of oxygen is at hand. Most organisms show certain preferences or, more accurately speaking, certain com- pounds are more readily fermentable by certain organ- isms than others. It is a common experience that protein decomposition (e.g. gelatin liquefaction) is retarded by addition of sugar. Apparently, the energy needs of a cell can be more readily supplied from sugar than from protein (see p. 205). (6) AMOUNTS OF ENERGY The amounts of energy liberated by the various fer- mentations differ widely. As a rule, oxidations yield the greatest quantities, while hydrolyses yield but very little, if any. Only in one case, cells provide for their energy by hydrolysis, namely in the urea fermentation. Hydrolysis is very commonly caused by bacteria, but in practically all cases, it raerely serves as a food- preparing process. The enzymes causing these hydrol- yses are mostly soluble. Table 5 (p. 23) shows the soluble and the endo-enzymes arranged according to heat-units produced from 1 gm. of substance. It is very evident that the enzymes with small heat liberation are soluble 26 PHYSIOLOGY OF BACTERIA while those with a large energy yield are endo-enzymes. Lactase with twenty-three calories is sometimes soluble; sometimes, it is an endo-enzyme. Miquel's claim that urease is a soluble enzyme will be discussed on p. 36. Attention should be called to the fact that the entire amount of liberated energy does not always become available for growth, for thermodynamic reasons. These will be discussed in the chapter on Growth. Ordinarily, in organic fermentations, the difference between avail- able and total liberated energy is small; furthermore, only a small fraction of the available energy is used for growth, even under ideal conditions. (c) METHODS OF MEASURING ENERGY YIELDS We have various means of measuring the amount of energy liberated in a fermentation. It is possible to com- pute it as the difference of the heats of combustion of the fermented material and of the products. This compu- tation is not very accurate since the error in combustion heat determinations becomes greatly increased if we have to consider the difference of two combustion heats of nearly the same magnitude. The inaccuracy is increased by the necessary corrections for the heats of solutions of products and for gases escaping. A good compilation of combustion heats is given by Kharasch (1929). This method may be illustrated by the following example: The combustion heat of glucose is 3,739 calories per gram; if a mold in a very well ventilated culture oxidizes glucose directly to carbon dioxide and water, the total amount of energy Uberated amounts to 677.2 cal. per gram molecule. If oxygen is not available, yeast could change the sugar to alcohol +CO2 according to the equation CeHuOe = 2C2H5OH + 2CO2 There is a considerable amount of potential energy left in the two molecules of alcohol formed, and the yeast can liberate only the ENERGY SUPPLY OF THE CELL 27 difference between the potential energies (or combustion heats) of sugar and alcohol. 1 gm. molecule of glucose = 677.2 cal. 2 gm. molecules of alcohol = 651.2 cal. 26.0 cal. The difference of 26.0 cal. is all the energy liberated in the alcoholic fermentation. It is only one-twenty-sixth of the amount obtained by oxidation. This means that for the production of 3,739 cal., the mold will need only 1 gm. of sugar if plenty of oxygen is available, but that 26 gm. will be needed by yeast if no oxygen is available. This is one of the reasons for the surprisingly great rate of decomposi- tion of food by microorganisms. The equation of the alcoholic fermentation should be written more completely and more correctly: CeHisOe = 2C2H5OH + 2CO2 + 26.0 calories But even that would not be really correct. If sugar is dissolved in water, the temperature decreases, and 2.25 cal. are disappearing for every gram molecule; if alcohol dissolves in water, heat is liberated to the extent of 2.54 cal. per gram molecule. These corrections should also be included in the above equation. Tangl and Rubner tried to get more accurate data by measuring the amounts of energy liberated in the cultures themselves. Tangl (1903) used the combustion heat method, as shown in Table 6. A broth culture was dried on standard cellulose blocks and its combustion heat determined directly. Part of this culture was filtered and treated in the same way. The difference of solids gives the solids of the bacteria grown in this culture; the difference of combustion heats gives the potential energy or combustion heat of these bacteria. The potential energy of the sterile medium is 4.39 cal. for each gram of solids, which is very nearly that of the bacteria themselves. In spite of this similarity in com- 28 PHYSIOLOGY OF BACTERIA m a» r-l Oi ^ 00 * * C3 tS 80 o (M i-H CO 00 t^ t>. >i'o ^ 13 o c> 1 73 oq 1— 1 1— 1 1-H CO CO (N CO CO 00 Tj< CO o * ' 00 1— ( S (N CO CO rH Tjl 05 lO 1— 1 c t^ o • . O OQ 1 (N (N (N d a-s Q $^ CB 0? cT J 1 CO Ci O 00 00 o »o 43 S O o 05 00 UO CO CO O 05 1-1 o 05 Is GO 3 ^§^^ : ; >> 05 CO GO iC • 3 1 tt5 1 r-l lO T-l CO CO (N (N C • .2 O lO lO C lO 03 ^ & 'S '^ 05 |> CO tH (M TtH ■"2 CO (N r-I ^ '"i^ T}H o rH T—l I— ( -s t^ OS CO cc ^-1 m •J t-H 05 t^ (M s oq ^ CO (N O l-H 1 1 CO (U 03 HH a ^ 1^ ca o: 03 : III ■as g 1 ii c. ; 5 -1 a rH ENERGY SUPPLY OF THE CELL 29 bustion heats, all experiments show that the energy actually required was in every experiment at least as large again as the potential energy of the cells produced. Why so much energy is needed, will be explained on p. 32. Rubner in his extensive experiments on the energy requirements of microorganisms used, occasionally, the flil QjtL O.7.* oJjcut. cc£i Fig. 1. 3 t ^^^^^^^UMi Daily temperature increases in cultures of Bad. coli (6% peptone solution), Strept. lactis (milk) and Bad. vulgare (6% meat extract solution), same method employed by Tangl, but in most cases considered the direct method of temperature measure- ment to be more accurate. By keeping large amounts of the culture, usually 250 c.c, in a thermos bottle which was placed in a carefully controlled incubator, the temperature increases could be read accurately by means of a Beckmann thermometer. The specific heat of the medium and the heat conductivity of the system were determined and this permitted the thermometer readings to be computed into calories. Some typical curves are given in Fig. 1, representing the lactic fermentation in milk, the growth of Bact. 30 PHYSIOLOGY OF BACTERIA coll in 6% peptone solution, and the decomposition of meat extract by Bact. vulgar e, taken from Rubner (1906b and c). The results obtained by this method show a fairly good agreement. The average of twelve determinations of the alcoholic fermentation JjO 4 "A /i 10 zoo 300 ^opAoarsX Fig. 2. — Rate of heat formation in 1 liter of soil. at different temperatures, with sucrose or maltose, with large and small sugar concentrations, with large and small amounts of yeasts, gave the heat liberated from 1 gm. of sugar decomposed by yeast to 149.5 cal., the largest deviations being +2.4% and —2.81% (Rubner 1904b). This energy corresponds to 26.9 large calories per gram molecule. Van Suchtelen (1923, 1927, 1929) improved this method by measur- ing the temperatures thermo-electrically, and by reducing heat conduction. His primary object was the determination of the availa- bility of humus and the rate of its deterioration. He used about 1 liter of soil for each experiment, and observed for the first four days a rapid increase in temperature, due to the excess of oxygen intro- duced by the stirring of the soil (see Fig. 2). After twenty days, the temperature had come down to a constant, but was still slightly above the outside temperature, and stayed about 0.3 to 0.6°C. ENERGY SUPPLY OF THE CELL 31 above normal through the entire duration of the experiment which was continued for fifteen more days. This meant about 5 or 6 gm.- calories per hour and per Uter of soil. At this rate, about 4% of the total humus would be decomposed during the first year. A similar method, has been applied by Bayne-Jones and Rhees (1929) to study the energy production by BacL coli during the first hours of growth. The cells were counted every hour, and this, together with the calorimetric measurements, gives the best data available for direct energy measurements. Some of the data are reproduced in Table 7. The energy per cell, either calculated by dividing the total calories produced by the average of the acting cells, or by applying the formula of Buchanan (see Appendix p. 407) shows a sharp maximum during the first hours, and soon drops to a low and fairly constant level similar to that found by van Suchtelen (Fig. 2). Bayne- Jones and Rhees believe that this is characteristic for young cells. It may be, however, that the great amount of heat at the start is due to an oxidation process; after a few hours, the oxygen dissolved Table 7.- -Calories Produced per Cell per Hour Time Calories produced in the medium Cells per flask Calories per cell per hour Total Increase per hour by the formula Hours S AS a, bi and 62 2 AS' 2.303As(log *-;) At(b2 -bi) At{b2 + 61) 0 1 2 3 4 5 6 0 1.485 17.68 50.29 76.65 92.73 108.60 0 1.485 16.20 32.61 26.36 16.08 15.87 360 X 106 420 X 106 2,160 X 106 9,600 X 106 33,000 X 106 96,000 X 106 129,000 X 106 3.81 X 10-9 12.56 X 10-9 5.55 X 10-9 1.22 X 10-9 0.25 X 10-9 0.14 X 10-9 3.72 X 10-9 15.25 X 10-9 6.54 X 10-9 1.36 X 10-9 0.27 X 10-9 0.14 X 10-9 in the medium will have been used up, and no more oxidation takes place except on the very surface of the culture. According to A. Miiller (1912), multiplying cells of Bad. coli require from 0.10 32 PHYSIOLOGY OF BACTERIA to 0.16 • 10~' mg. of oxygen per cell and hour, in a very poor medium (see p. 188). Figuring on 20,000,000 cells per c.c. as the average during the first three hours of the experiments of Table 7, the amount of oxygen consumed by these cells per hour would be 20 X 106 X 0.1 X 10"^ mg. = 2 X 10-^ mg. per c.c, or 2 mg. per liter. Since at 37°, water can dissolve only 6.8 mg. of oxygen, it seems quite probable that in about two to three hours, the oxygen of the medium is all used up, and energy can be produced subsequently only by anaerobic processes which yield much less energy (see p. 27). (d) THE NEED FOR ENERGY It seems justifiable to question why microorganisms need energy for growth. Yeast cells are made from the malt protein and eventually from sugar, and the com- bustion heat of these foods is about the same as that of the new cells. The same is evident in the examples of Tangl's experiments given in Table 6 p. 28. Since the cells are not built directly from the peptone or protein, these materials must be first reduced to much smaller units, amino-acids or, sometimes, even smaller molecules. The architecture of yeast protoplasm is quite different from that of barley protoplasm, and the architecture of protoplasm of B. anthracis is widely different from that of a commercial peptone. The breaking down will produce some energy, but this small amount is not nearly sufficient to weld the small particles into the new structure. Extra energy must be added to bring about new construction. This accounts for some of the energy wasted by bacteria, as well as by all other organisms. However, it does not account for all the waste. gi After growth has ceased, the cells continue for a considerable period to liberate energy. Meyerhof (1914) has shown that life, as such, is not based upon a higher potential energy in the living cells. He killed a thick ENERGY SUPPLY OF THE CELL 33 suspension of living red blood corpuscles instantaneously by adding acrolein, and observed no change of temperature. Nevertheless, life processes depend upon a source of energy. Blood corpuscles do not multiply in vitro, but show distinct respiration. The energy thus produced must be used for some purpose because all vital processes come to a standstill if oxygen, the source of energy, is excluded. Pfeffer (Pflanzenphysiologie II, 885) assumed that the cell is like a heated engine, using up coal all the time in order to be ready to supply energy the moment it is wanted. It seems probable, however, that the cell is more nearly comparable to an electric engine than to a heat-propelled machine. This would make the simile less appropriate. More probable seems the assumption that energy is needed continuously to prevent, or make up for, certain chemical processes which would destroy the living protoplasm. Warburg (1912) believes that this destruction is due largely to diffusion processes while Meyerhof (1914) considers chemical processes more probable. Meyerhof formulates his theory briefly as follows: For life functions, growth and cell activity, a mixture of substances within the cell must be prevented from ever reaching their physical and chemical equilibrium, and this cannot be done by merely slowing up the rate of reaction; it requires cyclic changes to re-establish again and again the continuously destroyed energy potentials. While we know very little about these processes, attention should be called to the reduction-potential of living cells, which is discussed in detail on p. 82. The maintenance of this potential requires a considerable amount of energy. Since potentials of this type in the cell are continuously destroyed, all the energy used for their 34 PHYSIOLOGY OF BACTERIA maintenance will ultimately be transformed into heat, and this becomes evident by a temperature increase of the medium. In most fermentations, our customary methods of cultivation show no increase of temperature because the heat of fermentation is conducted and radiated away almost as fast as it is produced. Even if we use thermos bottles, the increase of temperature rarely exceeds 2°C. (see Figs. 1 and 2). Exceptions are only the oxidative fermentations. In these, the temperature can rise very high if the heat has a chance to accumulate. The heat produced from the oxida- tion of alcohol by vinegar bacteria is sufficient to keep the tempera- ture in the vinegar generators high, therefore care must be taken to control the oxygen supply lest the temperature rise so high as to kill all bacteria. This would correspond, biologically speaking, to death by fever. Such temperature increases due to bacteria seem especially high when we consider that less than 0.1% of the liquid is bacteria bodies. Some bacteria produce light. This is another form in which energy leaves the bacterial cell. The amounts of energy are very small when compared with the amounts that leave the cells in the form of heat. Ultraviolet radiation is claimed by Baron (1928) for several yeasts and bacteria. The energy required by bacteria to move through a liquid has been calculated by von Angerer (1919). He considers his estimates, based on Stokes formula, only as very rough approximations which give merely the order of magnitude. The calories required for one hour's continuous swimming are calculated for one cell of V. cholerae to be 216 X 10-16 cal., for Bact.'typhosum 10 X lO-^^ cal., for Bacillus subtilis 25 X 10-^^ cal. The total energy liberated by one cell in one hour (calculating from lactic acid fermentation) is of the order of 10-1*^ calories. The energy required for motion is of the order of 10"^^ cal., and is negligible in comparison with the total energy liberated. (e) SUMMARY The potential energy (combustion heat) of yeast and bacteria cells is frequently similar to that of the food from which they are made. Nevertheless, a considerable amount of energy is required to produce these cells from food. ENERGY SUPPLY OF THE CELL 35 Fermentation continues, and energy is liberated, after growth has ceased. The surplus energy thus produced appears largely in the form of heat. Motility and production of light do not require measurable amounts of energy. The surplus energy which appears ultimately in the form of heat should not be considered as useless waste. It may have served a purpose in the cell before appearing as heat. Probably, energy potentials of some kind must be maintained in the living cell to prevent detri- mental reactions. The continuous equalization of these potentials would result in the ultimate transformation of the utilized energy into heat. II. CHEMICAL CONSIDERATIONS (a) ENZYMES OF FERMENTATION It has been proved in at least four different cases that fermentation is caused by enzymes. These examples are the urea fermentation, the alcoholic fermentation, the lactic fermentation and the vinegar fermentation. It is assumed by most bacteriologists that all fermentations are due to enzymes, though it seems to be extremely difficult to bring about the experimental proof. The statement that fermentation is caused by enzymes, means primarily only the achievement of a mechanical separation of the cause of fermentation from the living cell. Even under the most favorable circumstances, this separation has been very incomplete. The cell juice pressed out from yeast is quite weak in fermenting power as compared with a corresponding volume of living cells. The significance of the separation first accom- plished by Buchner (1897) lies in the proof that the energy-yielding processes are mere chemical processes. 36 PHYSIOLOGY OF BACTERIA They are brought about by a chemical compound (zymase), which is produced by the cell and which, after once having been produced, acts independently of the cell. The cell may control to a certain extent the amount of enzyme produced, but the action of the enzyme once produced is beyond the control of the cell. The enzyme follows always the laws of enzyme action, even if it should be against the vital interests of the cell. The fermentation of sugar to alcohol and carbon dioxide will continue for some time after the multiplication of yeast cells has ceased; there is already too much alcohol present to allow growth, and very little energy is needed by the resting cells, yet the enzyme is present, and it continues to act until it is checked by the maximum amount of alcohol, or by the disappearance of the sugar. This is against the best interests of the cell (as we understand them), but is in accordance with the laws of enzyme action. Another example is the excessive heat production by vinegar bacteria mentioned before. (6) THE KNOWN ZYMASES OR FERMENTING ENZYMES Urease. — The oldest known zymase is doubtless the urease discovered in 1876 by Musculus who noticed it in the slimy urine of some patients and believed the enzyme to be of human origin. Miquel (1904) sum- marizes his own earlier experiments in the statement that cultures of urea bacteria can be brought to the point where they decompose in one hour 100-120 gm. of urea per liter of culture. Such cultures can be filtered through porcelain filters without much loss of urease if large amounts of culture are used. From this, it would seem that the urease were no endoenzyme. This is very- improbable however because the cells would derive no benefit from urease outside of the cell. It is more probable that a large number of ENERGY SUPPLY OF THE CELL 37 cells are dead in a fairly old culture (at least three days at 30-35°C.) and that the enzyme which is quite stable leaves the dead cells, but not the living ones. The proof that this fermentation can be used as the only source of vital energy by the urea bacteria, has been given by Sohngen (1909). Alcoholase. — Much more interest was displayed when Buchner (1897) finally proved that alcoholic fermentation could be separated from the living yeast cell. The old famous dispute between Liebig and Pasteur had not been forgotten and the discovery of the yeast-zymase, or alcoholase, was a great step forward in biochemistry. Buchner obtained the zymase by tearing the cell walls of the yeast cells through grinding with quartz sand, and by pressing the ground yeast + quartz under high pressures. Thus he obtained the cell sap of the yeast, and this had retained the power of producing alcoholic fermentation. Even after filtration through porcelain, it was still able to bring about the formation of alcohol and carbon dioxide from sugar. Later, another method was used more commonly. The yeast was treated with alcohol and ether, or with acetone; this killed the cells, i.e. it destroyed definitely the power of yeast to grow and multiply, but not the power to ferment. Lactacidase. — After the establishment of the enzy- matic nature of the alcoholic fermentation, it was natural that the same technique was applied to other fermenta- tions as well. Herzog (1903) was the first to grind the cells of Bad. acidi lactici Hueppe, and to obtain a cell sap which would change lactose to lactic acid. This was confirmed later by Buchner and Meisenheimer (1903) who worked with Lact. Delbrilcki. 38 PHYSIOLOGY OF BACTERIA Alcohol-oxydase. — The efforts of Buchner and Meisen- heimer (1903) to obtain from ground vinegar bacteria an enzyme which would oxidize alcohol to acetic acid were without success. Treatment of vinegar bacteria with acetone gave, however, dead cells which would convert alcohol to acetic acid to a small extent. This was confirmed later by Rothenbach and Eberlein (1905). Buchner and Gaunt (1906) gave more detail regarding this enzyme. The enzyme nature of this fermentation seems to be established, but the enzyme itself appears to be even more sensitive than the alcoholase. (c) COMPLEXITY OF ZYMASE ACTION The enzyme content of the yeast cell is not limited to the alcoholase. Almost simultaneously with the alcoholase, the endotryptase was discovered, a pro- teolytic enzyme which can liquefy gelatin. This proves its nature as endo-enzyme since living yeast cells of the Saccharomyces type do not attack gelatin. This enzyme gradually destroys the alcoholase, especially in the absence of sugar. The alcoholase requires special conditions for its action. An essential factor is the presence of phosphate. Ordinarily, the cell juice of the yeast contains enough phosphate to bring about a visible fermentation, but the addition of phosphate often accelerates the reaction greatly and increases the total production of CO2. It is generally assumed that the alcoholase does not act upon the glucose molecule directly, but upon the hexose-phosphate, a phosphoric acid ester of the sugar. Probably, an enzyme of the yeast brings about the esterification. Another essential substance is the so-called co-enzyme of alcoholic fermentation. It was observed that heated ENERGY SUPPLY OF THE CELL 39 yeast juice increases the rate of fermentation of a normal yeast juice. Normal yeast juice can be separated by ultrafiltration, or by dialysis, into two fractions neither of which can cause the fermentation, while, combined, they again have nearly their original fermenting capacity. The nature of this co-enzyme has not been agreed upon. Kluyver and Struyk (1927) believe that different experimenters meant entirely different phenomena when speaking of co-enzyme effects. These authors think it to be a hydrogen acceptor which is necessary to start the first production of acet aldehyde. After this has been done, the fermentation can go on automatically as will be shown on p. 49. Nilsson (1930) believes the co-enzyme enters into various phases of the fermentation process. Thus we have 4 factors connected in the alcoholic fermentation. 1. Alcoholase 2. Phosphate-carbohydrate esters 3. Co-enzyme 4. Endotryptase The latter causes a deterioration of the alcoholase as well as of the co-enzyme. Heat will destroy alcoholase and endo-trj^tase, but not the co-enzyme. To complicate matters still more, the endotryptase is inhibited to some degree by the presence in the yeast juice of another compound : 5. Antiprotease and this compound No. 5 is destroyed by compound No. 6 6. Lipase which also attacks the co-enzyme. The actual amount of alcohol and CO2 formed by the enzjone is very small compared with that of the yeast itself. Harden (1923 p. 30) estimates that the living yeast produces about forty times as much alcohol as the 40 PHYSIOLOGY OF BACTERIA equivalent of cell juice. The acetone yeast is more active, and yields about one-eighth of the amount pro- duced by an equivalent amount of living yeast cells. For this reason, the reality of a fermenting enzyme is, occasionally, questioned again. Rubner (1913) could obtain no appreciable fermentation by treating living yeast in sugar solution with toluol. He believes that only a small portion of the enzyme is separated from the living protoplasm while most of it is interlinked with protoplasm and is inactivated by the death of the cell. Euler and his associates have proved this in a number of pubhcations. The essential thing, from the energy viewpoint, is the endo-enzymatic nature of the fermenting agent; it must work within the cell in order to make the liberated energy available for cell purposes. It is doubtless a complex mechanism which brings about this change of sugar to alcohol and carbon dioxide. Alcoholic fermen- tation occurs in several steps, with intermediate products, some of which have been actually isolated (see p. 49). Whether fermentation is separable from the protoplasm or not, has no great bearing upon the problems treated in this volume. A further discussion of the enzymatic nature of fermentations will be found on p. 64. (d) EQUATIONS OF FERMENTATION If fermentation is nothing but a chemical process activated by the influence of an enzyme, it is necessary that it obey strictly chemical laws, and that we may represent the process accurately by a chemical equa- tion. A considerable amount of effort has been made to prove this for a number of fermentations. Most attention has been given to the alcoholic fermentation; the lactic and propionic fermentations and the acid- ENERGY SUPPLY OF THE CELL 41 gas fermentations of the colon group are also well stud- ied. Data upon other sugar fermentations are rather incomplete. One single equation to cover the entire action of any species of bacterium upon the entire available food is evidently impossible. It is even impossible to cover the decomposition of just the sugar by yeast in one formula because Pasteur found that about 5% of the sugar is used for the construction of new yeast cells, and we could not possibly write this in a chemical equation. It seemed possible for a while to describe, in one equation, that part of sugar dissimilation which is not used for constructive purposes, but even that is inac- curate and incomplete, as a number of old and new experiments have shown. Most organisms have the ability to decompose sugar in more than one way, and this is due to the fact that all sugar fermentations are the result of a number of interlinking reactions. By a change of environment, the one or other of these reac- tions might be prevented, or changed, and the result is a different product, or a different proportion of the prod- ucts obtained. Besides, interreactions between sugar and nitrogenous material are always possible, so that any equation that does not consider the nitrogenous material may not show the real facts (see p. 62). A type of fermentation may be described by one general equation with the reservation that the formula holds true only for the conditions of the one experiment, and that even then, it accounts only for about 90% of the sugar decomposed. If a niore accurate understand- ing of the process is wanted, the different successive stages of the change from sugar to the final product must be worked out, and their interrelations must be studied. isobutyl glycol —0.158* amy 1 alcohol —0.051 n-propyl alcohol —0.002 isobutyl alcohol —0.0015 42 PHYSIOLOGY OF BACTERIA The general formula for the alcoholic fermentation is C6H12O6 = 2C2H5OH + 2CO2 Pasteur hesitated to accept this formula because of other compounds found regularly among the fermen- tation products which were not given by the above equation. The following list gives the approximate amounts of products found in normal industrial alcoholic fermentations : Table 8. — Products of Alcoholic Fermentation from 100 Parts of Sugar Alcohol 48 . 4 parts CO2 46.5 Acetic acid 0.05-0.25 Lactic acid 0 . -0.2 Fusel oil 0.05-0.35% Succinic acid 0.5-0.7 Glycerol 2.5 -3.6 Acetaldehyde 0 -0.08 Furfural traces Yeast cells 1.2-5 Boussignault had the correct conception that in order to establish equations of fermentation processes, it is essential to find out which products are always formed in the same ratio, because only these could derive from the same type of decomposition. It was shown much later by Ehrlich that some of these products, though nitrogen-free, derive from protein material. He found (1905) that leucin was the mother substance of iso-amyl alcohol, and that iso-leucin could be converted into optically active amyl alcohol. This was accomplished by a hydrolysis of the amino acid liberating NH3, and by a decarboxylation of the resulting a-hydroxy acid. * According to Duclaux III, 434 (1900). The amount of by-products is smaller with yeast juice than with living yeast. ENERGY SUPPLY OF THE CELL 43 (CH3)2-CH-CH2-CH(NH2)-C02H + H2O -> Leucin (CH3)2CH-CH2-CH(OH)-C02H i (CH3)2-CH-CH2CH20H + CO2 Iso-amyl alcohol CH3 C2H NcH-CH(NH2)C02H + H2O ^ _/ Iso-leucin ' NcH-CH(0H)-C02H + C2H5/ CH3 \ * >CH-CH20H + CO2 C2H/ Optically active amyl alcohol NH3 Ehrlich proved this theory by demonstrating that an increase of leucin or isoleucin in the medium caused an increase of fusel oil, while asparagin or ammonium carbonate did not cause such increase. In a similar way, Ehrlich (1909) showed that succinic acid is a protein derivative. It is not a reduction product from aspartic acid, however, but derives from glutamic acid by de-amidization and oxidation. CH2— CO2H CH2CO2H I I CH2 CH2 +02-^ I I CH— CO2H + H2O -^ CH— CO2H NH2 OH + NH, Glutamic acid Glutamic acid CH2— CO2H I CH2 CO2H + CO2 Succinic acid 44 PHYSIOLOGY OF BACTERIA Other a-amino acids also are changed to alcohols, e.g. tyrosin OHC6H4-CH2CH(NH2)C02H gives OHC6H4CH2CH2OH, tyrosol, a bitter tasting compound, probably the cause of the bitter taste of certain beers (Ehrlich, 1911). Tryptophane is changed to tryptophol by the same process, and these and similar compounds or their esters may be one of the causes of the differences in the taste of dif- ferent wines and beers. After this evidence that some of the products of yeast fermentation are of protein origin, the products derived from sugar can be traced with more certainty. (e) SirCCESSIVE STAGES IN THE FERMENTATIONS OF SUGARS The first great advance in the study of intermediate products and successive steps in fermentation was Neuberg's work on the fermentation of pyruvic acid by yeast (Neuberg and Hildesheimer, 1911) and his further developments from this starting point. The English biochemists with Harden studied, in great detail, the Colon fermentations as well as those by yeast. At the same time, Warburg concentrated on the chemistry of the respiration process, and Wieland developed his theory of the activated hydrogen as the essential part in all organic oxidations as well as reductions. Oxygen is a '' hydrogen acceptor, '^ but in many cases it can be substituted by other hydrogen acceptors, such as nitrate or methylene blue. Warburg had found that oxygen can oxidize only if ^' activated,'^ and the activator in biological oxidations is iron. A combination of Warburg's and Wieland's theories proved possible; Kluyver and Donker (1926) have applied this combined theory to fermentations and respiration in general, and the discussions of fermentations offered here follows closely their interpretation. As an illustration of Wieland's mode of presentation, the Cannizzaro reaction between methyl glyoxal and ENERGY SUPPLY OF THE CELL 45 acetaldehyde shall be shown here because it is part of the mechanism of alcoholic fermentation. Methyl gly- oxal is written in the hydrated form; the activated H -atoms are shown in heavy type. This compound is the ''hydrogen donator" giving two H-atoms to the acetaldehyde which is the ''hydrogen acceptor.^' CH3-C0-C— OH-H2 + CHs-CHO + 2H = Methyl glyoxal \ Acet aldehyde hydrate ^QH CH3COCOOH + CH3-CH20H Pyruvic Alcohol acid This is an example of the "oxido-reductions," the reduc- tion of one molecule by the oxidation of another, which is the essential part of all fermentations. By this emphasis on the hydrogen atoms as the essential parts of oxidation and reduction processes, the reactions which we commonly call oxidations would become more precisely "dehydrogenations," and the term dehydrogenase in place of oxidase is used by some biochemists. The first step in the fermentation of sugar appears to be the formation of hexose-phosphate, an ester of the sugar with phosphoric acid. This esterification is probably brought about by an enzyme. The glucose-phosphate is then split in the middle of the carbon chain, yielding two molecules of glyceric aldehyde, which, according to Kluyver and Struyk (1926), takes place in two steps: CeHiaOe + HK2PO4 = C6Hn05(K2P04) + H2O CeHuOsCKsPOO = C3H6O3 + C3H502(K2P04) Glyceric aldehyde 46 PHYSIOLOGY OF BACTERIA C3H502(K2P04) + H2O = C3H6O3 + HK2PO4 Glyceric aldehyde The glyceric aldehyde, by activation of a central H-atom which then goes to the end-carbon, is changed to methyl glyoxal. This process is an oxido-reduction, just as the splitting of the sugar molecule into glyceric aldehyde mentioned above must also be considered as a process of this type. CH2OHCHOHCHO -> CH20H-CH0HC— OH -> Glyceric aldehyde Hydrate form \ CH2-CH0H-C + H2O -> H /H ^H CH2-C0H-C -^ CH3-C0H-C -> /OH /H /H CH3-C • C— OH -> CH3-C0-C— OH ^OH "^OH "^OH Methyl glyoxal hydrate Up to this step, to the formation of methyl glyoxal, all glucose fermentations seem to be alike. The next steps decide about the different types, and methyl glyoxal can be decomposed in a number of different ways. But the number of possibilities is limited, and Kluyver and Donker show that a few main types of oxido- reductions of methyl glyoxal and its derivatives can account for most of the well known fermentations. The same reactions are used by different organisms in ENERGY SUPPLY OF THE CELL 47 different combinations, and the number of different reactions is quite small. The following list of basic reactions is given by Kluyver and Donker: BASIC REACTIONS IN SUGAR FERMENTATIONS I. CHANGES OF METHYL GLYOXAL HYDRATE /OH /OH la. CH3-C0-C— OH = CHs'CO-C— 0— H = Methyl glyoxal V \. CH3-CH0H-C00H Lactic acid /OH /OH lb. CHs-CO-C— OH = CH3-C0-C— 0— H = \h \h ch3cho + hcooh Acet aldehyde Formic acid /OH /OH Ic. CHs-CO-C— OH = CH3-C0-C— 0— H = \h \h CHs-CO-COOH + 2H = Acceptor Pyruvic acid = CHg-CHO + CO2 Acet aldehyde II. CONDENSATIONS ^OH H\^ Ila. CH3-C— OH + CCH3 = /^Acet a H O /Acet aldehyde Acet aldehyde /OH Us CH3C— OH + C-CHs = CH3COCHOHCH3 \h 0 ^ Acetyl methyl carbinol 48 PHYSIOLOGY OF BACTERIA lib. CH3C + H— CCHO = Acet aldehyde ch3c + h— c-cho = ch3-choh-ch2cho = \h h/ .OH ch3-choh-ch2-c— oh = \h ^OH CH3-CHOH-CH2-C— 0— H = CH3-CH2-CH2-COOH \ Butyric acid \H lie. CH3-C + H— CCOOH = \0H cetic acid CH3-C + H— CCOOH = CH3-CO-CH2-COOH = \0H H^ CH3-CO-CH2-C = CH3-CO-CH3 + CO2 \ Acetone \0— H III. DEHYDROGENATIONS Ilia. HCOOH = HCOOH = CO2 + 2H ^OH ^OH Illb. CH3CHO = CHs-C— OH = CHg-CO— H = Acet aldehyde Hydrate \ \ \H \H \0H U^ Acetic acid Acetic acid ENERGY SUPPLY OF THE CELL 49 CH3-C00H + 2H = Acceptor Acetic acid ^OH .OH IIIc. CHs-CO-COH = CHsCO-CO— H = \h \h CH3-C0-C00H + 2H = Acceptor (same as reaction Ic.) IV. HYDROGENATIONS IVa. 2H = H2 Hydrogen gas IVb. 2H + O = H2O IVc. CH20H-CH0H-CH0 + 2H = CH20H-CH0H- Glyceric aldehyde Glycerol f^TT OTT IVd. CHg-CHOHCOOH + 2H = CH3-CH2-COOH Lactic acid Propionic acid IVe. CHa-CHO + 2H = CH3CH2OH Acet aldehyde Ethyl alcohol IVf. CHa-COCHOH-CHs + 2H = CHa'CHOH- Acetyl methyl carbinol OTTOTT«r^TT 2, 3, Butylene glycol IVg. CH3CH2-CH2-COOH + 4H = CH3-CH2-CH2- Butyric acid OTT OTT Butyl alcohol IVh. CeHnOe + 2H = CsHuOe Fructose Mannitol IV. etc. (/) TYPES OF FERMENTATION The alcoholic fermentation starts from glucose-phos- phate over glyceric aldehyde to methyl glyoxal as out- lined above. The next step is equation IIIc and IVe 50 PHYSIOLOGY OF BACTERIA IIIc. CHa-CO-C— OH = CHg-CO-COOH + 2H \ Pyruvic acid Methyl glyoxal hydrate IVe. CH3-CH0 + 2H = CH3CH2OH Acet aldehyde Alcohol The combination of these two equations has been given on p. 45 as an example of a typical oxido-reduction. The pyruvic acid which comes from methyl glyoxal, is changed to acet aldehyde according to equation Ic Ic: CH3COCOOH = CH3-CH0 + CO2 The acet aldehyde thus formed can react again as acceptor for the hydrogen from a new methyl glyoxal molecule and the final sum of these reactions is the sum formula of the alcoholic fermentation. Only at the very start, a certain amount of acet aldehyde, or other hydro- gen acceptor, must be present to bring about the first pyruvic acid molecules. There are other possibilities hidden in the yeast mechanism. Small amounts of glycerol are formed, and this is due to the glyceric aldehyde acting as hydrogen acceptor in place of the acet aldehyde : Ic: CHsCO-C— OH - 2H = CH3COCOOH Methyl glyoxal \ Pyruvic acid ^OH IVc: CH2OHCHOHCHO + 2H = Glyceric aldehyde CH20H-CHOHCH20H Glycerol While acet aldehyde seems to be more efficient as hydrogen acceptor, glycerol will be formed in large ENERGY SUPPLY OF THE CELL 51 amounts when the acet aldehyde is ^'trapped/' i.e., combined with NaHSOs to a compound which does not give the above reaction. This process has been used commercially to manufacture glycerol from sugar. The general equation in this case is C6H12O6 = C3H8O3 + CH3-CH0 + CO2. There is still another possibility of reactions hidden in the yeast mechanism: acet aldehyde might act as a hydrogen donator as well as acceptor; the result is the formation of acetic acid: Illb: CH3C— OH - 2H = CH3COOH \ Acetic acid \0H Acet aldehyde hydrate IVe: CHs-CHO + 2H = CH3CH2OH This process takes place in alkaline media. Under this condition, the above combination seems to be favored which is really nothing but a Cannizzaro reaction between two acet aldehyde molecules. This leaves no acet aldehyde to act as acceptor for the hydrogen from methyl glyoxal, and, as before, glyceric aldehyde takes its place forming glycerol. The general equation of the alkaline yeast fermentation is : 2C6H12O6 + H2O = 2C3H8O3 + CH3COOH + C2H5OH + 2CO2 Glycerol Acetic acid Alcohol Neither the alkaline nor the sulfite fermentation of yeast follow the above formulas quantitatively because a certain amount of acet aldehyde will always react with methyl glyoxal before being trapped, and the result is a combination of two types of fermentation. 52 PHYSIOLOGY OF BACTERIA The Lactic Fermentation is essentially type la which is a typical intramolecular oxido-reduction. la: CsHeOs = CsHeOs Methyl glyoxal Lactic hydrate acid This describes the fermentation of about 90% of the sugar by Streptococci, and by some of the Lactobacilli. There are other Lactobacilli, however, which form also alcohol and CO2. This means that these organisms have also the power (i.e. the enzymes) to bring about reactions lb, Ilia, IVc and IVe producing formic acid as intermediate product which is decomposed further, liberating CO2, ethyl alcohol and glycerol. To which type of fermentation the volatile acids which are formed in relatively large amounts by certain groups of streptococci (Ham- mer, 1920, Hucker 1928) should be ascribed, cannot be stated until we know whether this fermentation is accompanied by corresponding amounts of glycerol and ethyl alcohol. The presence of propionic acid with some species suggests also the fermentation type IVd, while others producing formic acid apparently include type lb. The Propionic Acid Fermentation has been summarized by van Niel in a monograph (1928). Several possi- bilities are presented among which no experimental decision has as yet been possible. It may be that lactic is formed as such (la) and fermented further according to equation IVd: CH3-CH0H-C00H + 2H = CHa-CHz'COOH + Lactic acid Propionic acid H2O where the hydrogen derives from a different mode of attack, perhaps like this: CH3-CH0H-C00H + H2O = CH3COOH + HCOOH + 2H ENERGY SUPPLY OF THE CELL 53 the formic acid acting also as hydrogen donator (Ilia) to reduce another lactic acid molecule. It is also possible that lactic acid is not formed as such from glucose in this fermentation, but that glyceric aldehyde or methyl glyoxal undergo an oxido-reduction directly, van Niel has made it quite probable that acet aldehyde is not an intermediary product in this fermentation. The Colon-T3rphoid Group has a much more varied way of attack. Lactic acid is always formed (la) and there is another process of fermentation going on at the same time, independently. This second process resem- bles that of the yeast fermentation to a certain extent, with acet aldehyde as an intermediary product. It starts with reaction lb, followed up by IIIc and IVe lb: CH3-C0C0H = CHs'CHO + HCOOH Methyl glyoxal Acet aldehyde Formic acid IIIc: CH3-CH0 + H2O = CHg-COOH + 2H IVe: CHs-CHO + 2H = CH3-CH20H This accounts for the formation of lactic acid, and without relation to this, for the formation of two mole- cules of formic acid to one each of acetic acid and ethyl alcohol. This would be the type of fermentation by Bad. typhosum. Occasionally glyceric aldehyde acts as hydrogen acceptor and small amounts of glycerol are formed. In the colon-group, reaction Ilia causes the for- mation of H2 and CO2 in equal volumes from formic acid. Succinic acid is also formed in small amounts, and doubtless is derived from sugar (Ayers and Rupp, 1918). Kluyver and Donker believe that glucose which is not in the form of a phosphate ester might be split unevenly into a four carbon and a two carbon chain. 54 PHYSIOLOGY OF BACTERIA The origin of the extra amounts of CO2 in fermentation by Bad. aerogenes (Rogers, Clark and Davis, 1914) is not so easily accounted for. From the larger total volume of gas produced, it does not seem that the increased CO2 : H2 ratio of this group is due to a diversion of hydro- gen; it appears more like an additional source of CO2. An inter-reaction between methyl glyoxal and acet aldehyde might be assumed, as in the case of the alcoholic fermentation, but this would result in a larger amount of alcohol, for which evidence seems to be lacking. Furthermore, this group shows the formation of acetyl- methyl carbinol, probably according to reaction Ila, in the aerogenes group, and of butylene glycol (reaction IVf) in the colon group. Acetone Fermentation by spore-forming rods of the type of B. macerans Schar dinger, and B. aceto-ethylicus Northrop shows also a very varied list of products, adding acetone to all the products observed in the colon-groups. The types of reaction involved are the same as above (la, lb, Illb, IVe, Ilia) and in addition, we have lie. lie: CH3COOH -I- CH3COOH = CH3'CO-CH2- COOH = CHa-CO-CHa + CO2 The subgroup of B. asterosporus and B. polymyxa Beijerinck forms butylene glycol besides alcohol, CO2 and H2. The origin of butylene glycol has already been mentioned in the colon-typhoid group as being due to reactions Ila and IVf while the rest of the products are mainly produced by reactions lb and Ilia. Butyric Acid Fermentation probably starts with reac- tion Ic, methylglyoxal giving acet aldehyde and formic acid. The acet aldehyde goes through the aldol stage and finally into butyric acid (reaction lib) lib. 2CH3CHO = C4H8O2 ENERGY SUPPLY OF THE CELL 55 Part of the acet aldehyde reacts according to Illb and IVa, producing acetic acid and hydrogen, or according to IIIc + IVe, giving ethyl alcohol besides acetic acid. The formic acid from Ic is largely split according to Ilia. With some organisms, this hydrogen is liberated as such; others can activate it to reduce either butjn-ic acid, or an intermediary stage of reaction lib, to butyl alcohol; further acetic acid may be coupled according to reaction lie to form acetone. It seems probable that a high hydrogen ion concentration is necessary to perform these latter changes, and this acidity seems to be pro- hibitive for the species forming butyric acid only. Other Fermentations shall not be discussed here. It is beyond the scope of this book to describe all different types of fermentations. Such descriptions might be found in the book of Fulmer and Werkman (1929) or in vol. Ill of Buchanan and Fulmer's, ^^ Physiology and Biochemistry of Bacteria." This chapter was meant to give a conception of how a fermentation takes place, and what an equation of fermentation does and does not mean. Complete Oxidation of glucose is typical for certain groups of microorganisms, among which the Myco- bacteria are perhaps the most outstanding. No inter- mediary products of any kind could be found by Merrill (1930) in a careful investigation. In the oxidation of glucose by molds, intermediate steps are well-known, though the oxidation will finally become complete. Oxidation follows a hydration of the aldehyde group of the glucose molecule: CeHi^Oe + H2O = CsHuOs-C— OH = CsHnOsCOOH + 2H = Acceptor 56 PHYSIOLOGY OF BACTERIA A dehydrogenation follows, and the acceptor for this hydrogen is oxygen activated by the mold. The remain- ing gluconic acid can be hydrogenated further in various ways, and citric acid as well as oxalic acid are well- known intermediary products. The final result is CO2 and H2O. Fermentations of Proteins. — The protein molecules are so large and so entirely unknown that it is entirely out of the question to present equations for the fermenta- tion of proteins. These large protein molecules can be attacked only by a limited number of bacteria. The common division of bacteria into those that liquify gelatin, and those incapable of doing this, is no reliable indication for their ability to attack protein molecules in general. While the specificity does not seem to be as strict as with carbohydrates, there is a certain amount of specialization the chemical foundation of which is not as yet known. (Rettger, Berman and Sturges, 1916, Berman and Rettger, 1918.) The large protein molecule is broken up by those microorganisms which have the power to do so, into smaller molecules of similar nature which are still too large to be described by a chemical formula. These compounds, called albumoses, hemi-albumoses, pep- tones, peptids and various other names, are usually defined by the precipitants which will make them insol- uble. A number of different compounds are used for this purpose, such as dilute acids, copper salts, lead salts, zinc salts, concentrated ammonium sulfate, tannic acid, phosphotungstic acid, etc. The proportion of the total nitrogen contained in these fractions gives a general conception of the course of protein decomposition, and is of valuable information. ENERGY SUPPLY OF THE CELL 57 e.g., in the study of cheese ripening. In this chapter deaUng with the equations of fermentation, they are of no special interest. The next step in the degradation of proteins is the formation of amino-acids. In a very complete study of the decomposition of casein by B. mesentericus vulgatus, Grimmer and Wiemann (1921) found the following amino-acids: alanin, valin, leucin, tyrosin, aspartic acid, glutamic acid, prolin, arginin, lysin, histidin, trypto- phane, and probably phenylalanin. Of the decomposi- tion products of amino-acids, the following compounds could be identified: putrescin, cadaverin, tryptamin, tjo-osol, ammonia, and probably p. oxybenzoic acid. In the presence of lactose, histamin and 7-amino- but3n:'ic acid were also formed. This is an example of what can be expected in an aerobic decomposition of a protein. Fermentations of Amino Acids. — The further decom- position of these amino acids is something quite definite, chemically speaking, and the equations can be given. A good survey of all products derived from amino-acids is given by Stephenson (1930) in tabulated form. Perhaps the most common type of decomposition of amino acids is the splitting off of NH3. This can be accomplished in four different ways, namely by hydroly- sis, by reduction, by oxidation and by leaving an unsatu- rated double bond (desaturation) . The liberation of ammonia by hydrolysis has already been described in the formation of alcohols from amino acids by yeast, especially with the amyl alcohol (page 43). Leucin was hydrolyzed to a-hydroxy-caproic acid. C4H9-CH(NH2)-C02H + H2O = C4H9CH(OH)C02H + NH3 58 PHYSIOLOGY OF BACTERIA and then the hydroxy acid gave off CO2 and thus became amyl alcohol. C4H9-CH(OH)-C02H = CO2 + C4H9-CH20H In a similar way, yeast can hydrolyze other amino acids (see page 44). The de-aminization by reduction will be mentioned on page 63 where the propionic acid bacteria reduced aspartic acid to succinic acid. This process is very probably not a source of energy. The third possibility, the de-aminization by oxidation, has been used to explain the formation of succinic acid by yeast. It was shown on page 43 that yeast does not produce succinic acid from aspartic acid like the propionic bacteria do, but from glutamic acid. It is, of course, impossible to state whether the removal of NH3 is accomplished by direct oxidation or by hydrolysis, immediately followed by oxidation; the result is the same. This type of oxidation process is also found with fungi. Very probably it will yield energy. The last possibility of deaminization, the splitting off of NH3 leaving an unsaturated carbon chain, seems to have been observed only once, with histidine, though performed by a number of species of the colon-aerogenes group. Raistrick (1917) found the following decom- position of histidine : (C3H3N2)-CH2-CH(NH2)-COOH = Histidine (C3H3N2)-CH:CH-COOH -1- NH3 Urocanic acid Most probably, this process does not yield energy. Perhaps, neurin is formed from lecithin by a similar desaturation. ENEKGY SUPPLY OF THE CELL 59 Another very common type of decomposition of amino acids is decarboxylation, i.e., the splitting off of the CO2 group. This is the way in which amins are formed. E.g., glycine is decomposed into methyl amin by Ps. fluorescens (Emmerling and Reiser, 1902). CH2(NH2)COOH = CH3NH2 + CO2 In this same way, cadaverin may be produced from lysin. CH2(NH2)-CH2-CH2-CH2-CH(NH2)-COOH = Lysin CH2(NH2)-CH2CH2-CH2CH2-NH2 + CO2 Cadaverin Putrescin, or tetra methylene diamine, probably origi- nates from arginine in a similar way. Quite often, decarboxylation and deaminization occur simultaneously. Examples of this have been shown already (p. 43) in the decomposition of leucin, isoleucin, tryptophane and tyrosin by yeast. Aside from these decompositions of amino acids, which can be grouped into distinct classes, many other changes may occur which are typical for one certain amino acid only. This is especially true with oxidative changes. Hydrolysis of amino acids will yield hydroxy acids, and these may be used as substrates for anaerobic fermentations (p. 76). Another group of protein cleavage products are the acid amids. The energy yield of their hydrolysis is probably quite considerable as will be shown later (see p. 77). Anaerobic Putrefaction. — It has been generally assumed that anaerobic putrefaction of proteins as we find it most pronounced in the group of the anaerobic spore formers, follows essentially the same type that has 60 PHYSIOLOGY OF BACTERIA been outlined in the preceding pages. The recent investigations of Sturges and his associates indicate, however, that with some species at least, the type of decomposition is different. Parsons and Sturges (1927) studied the volatile acids produced by Clostr. putrefaciens from pork, and found that acetic, butyric and valeric acids were pro- duced at the molar ratio of 85 : 6 : 9. In gelatin, the relative amount of acetic acid rose even to 91-92.5% of the total acidity. This dominance of acetic acid contrasts Clostr. putrefaciens to most other Clostridia which show a much higher ratio of acids with longer carbon chains. The only exception is Clostr. histolyticum (Sturges, Parsons and Drake, 1929) which produces acetic acid only; non- volatile, ether-soluble acids could not be detected. This holds true for such widely different media as milk, beef heart, brain, liver, gelatin and peptone. The large amount of acetic acid formed by Clostr. histolyticum cannot have been derived from deaminiza- tion of glycine, because even in a gelatin medium (con- taining 25 % of glycine) the acetic acid formed is almost twice the theoretical amount from the glycine, and in a pork medium, it is more than ten times the theoretical amount. Sturges, Parsons and Drake conclude that this organ- ism, and possibly all putrefactive anaerobes, either have some highly specific method of splitting proteins, differing radically from the usually accepted hydrolysis, or that they can split long-chain amino acids into acetic acid and some still undiscovered residues. Urea Fermentation. — The urea fermentation is usually summed up in the formula: CO(NH2)2 + 2H2O = C03(NH4)2 ENERGY SUPPLY OF THE CELL 61 Really, the process must occur in two steps, one being the hydrolysis, and the other being the neutralization of the ammonia and the carbonic acid formed. CO(NH2)2 + 2H2O = CO3H2 + 2NH3 CO3H2 + 2NH3 = (NH4)C03 Sohngen (1909) has proved that urea is entirely sufficient as a source of energy for urea bacteria, and also as a source of nitrogen for cell construction, but not as a source of carbon. While urea is fermented by a large number of species from different genera and families, uric and hippuric acid are fermented mostly by spore formers. The hip- puric acid is hydrolyzed into benzoic acid and glycine : CeHsCO-NH-CHz-COOH = CeHsCOOH + NH2CH2COOH Uric acid is split into its natural components, two molecules of urea and one of a tri-carbon compound which is readily oxidized. Prototrophic Fermentations. — A large number of prototrophic fermentations are known, all of which are oxidations. The largest number of species occur in the sulphur bacteria which oxidize H2S primarily to S, and later to H2SO4. While all members of this very large group behave ahke in their energy acquirement, the bacteria acting upon ammonia have clearly divided their task, as the one group can only oxidize ammonia to nitrite, and the other only nitrite to nitrate; not ammonia to nitrate. The other various oxidations of methane and higher hydrocarbons up to the paraffin, of carbon monoxide, of coal, etc., offer no special interest in their equations. The 62 PHYSIOLOGY OF BACTERIA only interesting species is the Hydrogenomonas, which requires CO2 for the oxidation of hydrogen, and makes formic acid as intermediate product, according to Kaserer (1906) CO2 + 2H2 = HCOH + H2O + 6 cal. HCOH + 02 = H2O + CO2 + 132 cal. (g) INTERACTION BETWEEN NITROGENOUS AND NON-NITROGENOUS FOOD COMPOUNDS We have seen several examples of intermolecular oxido-reductions, i.e. an exchange of hydrogen atoms between molecules of different composition. The Can- nizzaro reaction between methyl glyoxal and acet aldehyde is perhaps the best known case. Since these reactions are essentially caused by the activation of certain hydrogen atoms, it seems probable that many other compounds capable of reduction could act also as hydrogen acceptors, and that the hydrogen acceptors will not be limited to cleavage products of sugar. A very simple and instructive case of this type is reported by van Niel (1928) who studied the action of twelve strains of propionic acid bacteria upon lactic acid and found that the products did not correspond to the equation: 3C3H6O3 = 2C3H6O2 + C2H4O2 + CO2 + H2O Lactic acid Propionic Acetic acid acid The ratio of propionic to acetic acid, instead of being 2:1, was very persistently 1.8:1. Besides, some succinic acid was formed which had to be accounted for. Van Niel could demonstrate that the succinic acid did not come from the lactic acid, but from aspartic acid or some similar amino acid in the rather concentrated ENERGY SUPPLY OF THE CELL 63 nitrogenous material of his medium. The aspartic acid acted as hydrogen acceptor and was reduced to succinic acid and ammonia. COOH-CH2-CH(NH2)-COOH + H2 = COOH-CH2-CH2-COOH + NH3 By adding aspartic acid to the medium, still more hydrogen could be '^deviated" and the ratio of pro- pionic to acetic acid fell to 1.6:1. We might conclude from this that the equation for the fermenta- tion of lactic acid is correct, and that the aspartic acid interfered. Such a statement, however, might be thought to prescribe a definite course to a biological reaction, and to establish standards rather than to give an unbiased interpretation of the facts. The unprejudiced description of the process would give each of the partial reactions (as mentioned on p. 47) and add the above reduction of aspartic acid. This would leave to the bacterium the ''choice" of the hydrogen acceptor, instead of introducing the human element of ''correct equations" and "deviations." This is probably the best known, quantitatively studied example, but many others are known, in which the protein food acts either as hydrogen donator or as hydrogen acceptor. Another very commonly found hydrogen acceptor is the elementary oxygen. In the common types of alco- holic fermentations (beer, wine, distilled alcohol), the influence of oxygen is negligible because the continuously formed CO2 prevents appreciable amounts of oxygen from being taken up by the yeast. In the aerated tanks used for the production XDf bakers yeast, the ratio of alcohol to CO2 is changed considerably (see p. 176). A further good example of oxygen as a hydrogen acceptor is its influence upon the gas ratio of Bad. coli, i.e., the ratio C02:H2. Keyes and Gillespie (1912, 64 PHYSIOLOGY OF BACTERIA 1912a) found that a Bad. coli having the gas ratio 1.03-1.09 in vacuo showed the gas ratio 5.4 in presence of oxygen. Another strain, with the gas ratio 1.77 in vacuo, changed in air from 3.4 to 590. The gas ratio of Clostr. Welchii belonging to the group of anaerobic butyric organisms, was not changed essen- tially by the presence of air (1.42-1.53 in vacuo, 1.44 to 1.85 in air) though some oxygen was adsorbed by the culture. Oxygen, apparently, could not be activated to act as acceptor. Even in vacuo. Bad. coli will change the gas ratio with the medium. With ammonium lactate as the only source of nitrogen, the gas ratio was 1.03, 1.06 and 1.09 in three parallel experiments, while with 1% peptone Witte, it increased to 1.31, and with beef infusion and peptone Witte, it was 1.28. {h) THE MECHANISM OF FERMENTATION The Mechanism of Fermentation has been claimed to be essentially the activation of certain hydrogen atoms in the substrate molecule. This activation is brought about by a catalyst in the cell. The catalyst is sup- posed by some to combine temporarily with hydrogen and then give it up to some hydrogen acceptor. Such catalysts are used commercially; the best known inor- ganic hydrogen catalyst is metallic nickel, which is used in the industrial hydrogenation of fats. Palladium will dehydrogenate ethyl alcohol or hydrated acet aldehyde and transfer the hydrogen either to oxygen or to methylene blue, thereby changing the dehydro- genated product to acetic acid and acting in the same way as do the vinegar bacteria. Possibly, the catalyst or activator only loosens the bond with which the hydrogen atom adheres to the ENERGY SUPPLY OF THE CELL 65 molecule. The catalyst in either case must be an unsaturated compound. If the hydrogen would com- bine with the catalyst permanently, it would lose its power as activator. A third compound steps in to prevent this, namely the hydrogen acceptor which com- bines with the loosened, activated hydrogen atoms thus becoming itself reduced while oxidizing the hydrogen donator. The hydrogen acceptor may have to be activated as well. We are certain of this in the process of respiration where oxygen is the final hydrogen acceptor, and will act only if activated by some complex organic iron compound. The gradual breaking up of the carbohydrate through a number of intermediate steps leads to the question whether for each of these steps, a separate enzyme is necessary. It would seem that what we had called zymase (p. 36) must be a large group of individual enzymes. In fact, several of them have already received names, such as catalase, reductase, alcohol oxidase, lactozymase, lactacidase, carboxylase, carboligase, methylglyoxalase, mutase, dismutase, etc. We have no criterion for the purity of an enzyme and therefore might always assume that a supposedly pure enzjone contains another enzyme if it gives a new reaction under changed conditions. However, the num- ber of different enzymes in one bacterial cell would have to be so large under the circumstances that it seems rather impossible to concentrate so many molecules of the size of enzyme molecules in such a small space. (The dimensions of bacteria will be discussed on p. 397.) Kluyver and Donker (1926) have endeavored to show that Wieland's theory of hydrogen activation, as the underlying cause of all oxido-reduction, brings a unifying principle into biochemistry. It does not seem necessary 66 PHYSIOLOGY OF BACTERIA to assume a different enzyme for each individual step in each fermentation. Kluyver and Donker point out that different intensi- ties of the affinity of the catalyst for hydrogen, i.e. different intensities of the activation of hydrogen atoms might bring about quite different reactions. They further call attention to the great influence of hydrogen ion concentration upon the reduction potential (see p. 88). This potential will very probably be inter- linked somehow with the activation of hydrogen. Another cause for differences of intensity of activation by the same catalyst might be their arrangement within the cell, and the distance between activator (catalyst) and substrate (hydrogen donator). It will be shown on p. 263 that a very definite arrangement of the essential molecules in the cell is necessary. If the distance between catalyst and substrate is larger in one species than in another, the loosening of the bond of a certain hydrogen atom will not be so great, and this may bring about a different type of reaction. Besides, the type of reaction always depends upon the hydrogen acceptors present. If oxygen and an oxygen activator are available, the reaction must be different from that under anaerobic conditions, even with the same hydrogen activators (see the gas ratio of Bact. coli, p. 64). While our knowledge is not sufficient to make accurate statements, the present outlook in biochemistry indicates that the complexity in fermentation processes does not lie so much in the innumerable number of different com- pounds reacting, but more in the adaptability of the reacting compounds to all different conditions. Another step in this direction is the attempt of Utkina-Ljubowzowa and Steppuhn (1929, 1930) to prove that the autolytic enzymes of ENERGY SUPPLY OF THE CELL 67 different tissues and of different organisms are not specific, but that in all organs of all plants, animals and microorganisms, the same three proteolytic mechanisms are found which are characterized by their optimal pH of 3.5, 5.5 and 7.5 respectively. (i) SUMMARY OF FACTS The large majority of bacteria can exist only if they have at their disposal organic or inorganic food which they can decompose. The amounts of energy produced by such decomposition can be measured or calculated. Such decompositions, we call fermentations, using the broadest meaning of this term. In several cases, it has been possible to separate from the living cell the agent that causes fermentation. In other cases, it could be shown that fermentation may continue after the cells have lost the power of reproduc- tion by drying, acetone or alcohol treatment, or ultra- violet irradiation. Ultra-filtration and dialysis have shown that fermen- tation is brought about by the inter-reaction of several agents. In the fermentation by yeast, colon bacteria, butyric acid bacteria, acetone bacteria or others, intermediary products of fermentation have been isolated which demonstrate that the fermentation is not one process, but a succession of several chemical reactions. The type of fermentation may be changed altogether by changing the reaction of the medium, or by adding sulfites, or by changing the oxygen supply, and even by changing the concentration of the substrate. SUMMARY OF THEORIES Bacteria require, for maintenance as well as for growth, a source of energy which, for most species, can 68 PHYSIOLOGY OF BACTERIA come only from the decomposition of organic or inor- ganic compounds. Fermentation (in the broadest meaning of the term) is not absolutely connected with living organisms. The vinegar fermentation can be accomplished by finely divided palladium in the same way as by bacteria. Fermentation may be accomplished by cells after their death, i.e., after loss of reproduction. The fermenting agent could be separated from the cell in some instances. We are forced to the conclusion that fermentation is brought about by enz3rnies. A number of different enzymes are necessary to bring about fermentation. We are at present quite unable to state how many different kinds are needed. Each fer- mentation is not one single process, but a combination of a number of different simple chemical reactions. The fermentation mechanism of different genera and species does not involve absolutely different reactions. Different combinations of the same simple steps in the gradual breakdown of the food molecule may lead to quite different fermentation products. These differ- ences in combination may be caused by differences in pH, by the entering of other compounds into the reaction (such as oxygen), and by differences in the arrangement of the enzyme molecules in the cell. Most of these reactions can be explained best by the assumption that the enzymes of the cell act upon the substrate by activating certain hydrogen atoms. These then combine with a ^'hydrogen acceptor.'^ The accep- tor might be another part of the same molecule, or another organic or inorganic molecule, or oxygen. Oxygen, in order to function as hydrogen acceptor, must be activated by a complex iron compound. ENERGY SUPPLY OF THE CELL 69 III. THE ROLE OF OXYGEN IN ALL FERMENTATIONS Hydrogen and oxygen are chemical contrasts to a larger degree than any other two elements. Oxidation and reduction are diametrically opposed reactions; dehydrogenation means oxidation; reduction means either the addition of hydrogen, or the removal of oxygen. In organic compounds, hydrogen is not as conspicuous as oxygen. There are usually a number of hydrogen atoms, only a few of which can be easily activated. If an oxygen atom is present in an organic compound, it marks reactivity, a step away from the '^parum affinis'^ of the inert hydrocarbons. Thus we have come to consider oxygen as an indicator of biochemical usefulness. The degree of usefulness of a substance for a certain organism is dependent upon the number of oxygen atoms as well as their position in the molecule. Aside from this, the role of oxygen in respiration, the abundance of oxygen in the air, the ubiquity of it make it appear of paramount importance in life processes. Even though the basis of all oxidations and reductions may be the change of hydrogen atoms, we notice such reactions primarily through their counterpart, oxygen. There is no need, as far as the author can see, to eliminate the word oxidation from our biochemical vocabulary. The primary purpose of fermentation is apparently the energy supply of the ce^. It is brought about in practically all cases by oxido-reductions. The one outstanding exception is the urea fermentation which must probably be classified as ^ hydrolysis. Hydrolyses, as a rule, yield so little energy that they cannot be used as source of energy by organisms. (See however p. 76.) In all oxido-reductions, some hydrogen is removed from the substrate, and is placed in the hydrogen 70 PHYSIOLOGY OF BACTERIA acceptor. We should distinguish, for practical pur- poses, three different cases: (1) The hydrogen acceptor is molecular oxygen. Then we have a direct oxidation, as in the vinegar fermentation, or in the action of molds or mycobacteria upon sugar. (2) Hydrogen acceptor and hydrogen donator are the same molecule. This we call intramolecular fermenta- tion. The chemical equation shows this most simply: C6H12O6 = 2C3H6O3 No oxygen and no hydrogen from outside are apparently needed to bring about this fermentation. It consists of nothing but a rearrangement of the H- and 0-atoms (see p. 75). The same is true with all the typical sugar fermentations. As a matter of fact, the case is not quite so simple because we must consider the separate stages of fer- mentation. But the term '' intramolecular fermenta- tion" indicates that one substance alone is sufficient as source of energy. (3) Hydrogen acceptor and donator are different molecules. Two different compounds are required to bring about the change. The acceptor may be oxygen; this proves type (1) to be just a special case of type (3), sufficiently speciaUzed, though, to remain classified by itself. Examples of type (3) have been given in the discussion of interaction between proteins and carbohydrates. Another example which brings the oxygen into the foreground again is the anaerobic fer- mentation at the expense of oxygen from nitrates, sulphates, methylene blue and similar, highly oxidized compounds. Allyn and Baldwin (1930) could make ENERGY SUPPLY OF THE CELL 71 Rhizobium grow as far as 10 mm. below the surface in an agar medium oxidized by addition of 0.015% KMn04 while in the untreated; normal medium, they grew on the surface only. For thermodynamic reasons, processes of type (3) can take place only if the energy liberated by dehydrogena- tion (oxidation) is greater than that required for hydro- genation, or if the total energy gain of the acceptor is smaller than the total energy loss of the donator. This eliminates a priori a number of processes as impossible. Putter, in his book on comparative physiology (1911) has computed the calories produced by 1 gm. of oxygen in oxidizing various organic compounds. Table 9 is an extract from Putter's data. The energy liberated is very uniformly 3.3 calories per 1 gm. of oxygen, regard- less of the type of compound oxidized. Only oxalic acid deviates considerably from this average. With inorganic compounds, the liberated energy varies greatly. Table 9 gives a few examples, computed also on the basis of energy per gram of oxygen. It is very evident which compounds are good hydrogen Table 9. — Calories Liberated by 1 Gm. of Oxygen A. In the oxidation of inorganic compounds Kg. calories per gm. oxygen H2 + O = H2O 4.25 H2O + O = H2O2 -1.44 C + O = CO 1.80 CO + O = CO2 4.24 HoS + 0 = H20+S 3.87 HoS +40 = H2SO4 ^ 3.18 S + H2O + 30 = H2SO4 2.96 NH3 +30 = HNO2 + H2O 1.63 HNO2 + O = HNO5 1.16 H + N +30 = HNO3 1.02 72 PHYSIOLOGY OF BACTERIA B. In the oxidation of organic compounds Kg. calories Methane 3.20 Acetic acid 3 . 272 OxaHc acid 3 . 80 Palmitic acid 3 . 259 Dextrose 3.508 Maltose 3.520 Starch 3.530 Asparagin 3.218 Uric acid 3 . 198 Legumin 3.291 Elastin 3.240 Fat (animal) 3.293 Glycerine 3.5 Alcohol 3.4 acceptors if we consider only the energy viewpoint. The best is doubtless hydrogen peroxide because energy could be gained by reducing it, i.e. by adding hydrogen to it. The other compounds offer a certain resistance to reduction. The energy required to remove 1 gm. of oxygen from these compounds is lowest in the nitrates. Let us assume a bacterium under anaerobic conditions in a medium containing oxalic acid and nitrate. Nitrate could act as hydrogen acceptor, oxalic acid as hydrogen donator. C2O4H2 -^ 2CO2 + 2H + 3.80 cal. per gram oxygen HNO3 +2H -^HN02 +H2O - 1.16 cal. per gram oxygen C2O4H2 + HNO3 -^ 2CO2 + H2O + HNO2 + 2.64 cal. per gram oxygen Such an oxido-reduction is possible because it leaves a positive energy balance. The energy required to fix the hydrogen to the acceptor is less than the energy gained by dehydrogenation. This latter value (energy from dehydrogenation) is quite uniformly 3.3 cal. per gram of oxygen added (or for Ke gm. of hydrogen removed), provided that the oxidation is complete. ENERGY SUPPLY OF THE CELL 73 Table 9 shows that H2O and CO2 and sulfur could not be considered as hydrogen acceptors, and that H2SO4 is a very poor acceptor, while nitrate and nitrite are excellent. Better than all of these (except H2O2) is oxygen gas which requires no energy whatever and, therefore, gives the highest energy yields. The computation of the energy balance shows us which reactions are possible, and which reactions can take place more easily than others. But it is not sufficient that a reaction is thermodynamically possible. It is the type of catalyst which decides whether the possible processes actually happen. Very few organisms can reduce nitrates to nitrogen gas, though this seems the most economical type of reduction from the energy viewpoint. No organisms are known to reduce carbon monoxide, though it would be quite profitable with the proper mechanism. Eventually, the catalysts for such oxido-reductions are already present in the culture medium. Demeter (1930) has shown that sterile broth, asparagin solution, etc. will reduce nitrates to nitrites at 45°C. The importance of chemical structure in addition to energy con- siderations is most striking in the anaerobic species. Though their energy supply depends upon oxido-reductions hke that of all other organisms, they cannot utilize the best of all hydrogen acceptors, i.e. oxygen. This is not merely due to their sensitivity to oxygen because some species which are rather indifferent to it show the same lack of oxygen utihzation. The author cultivated Strept. cremoris at 30°C. in a flask half full of milk and closed by a rubber stopper with a large inverted U-tube whose other -end opened into a glass of water. While with Bad. coli and B. cereus in peptone solution, the water was always drawn into the tube as a result of consumed oxygen, the level of water never changed with milk cultures of Strept. cremoris, not even when a tube of soda-Ume was suspended in the flask to absorb eventual CO 2 formed. Addition of methylene blue made no 74 PHYSIOLOGY OF BACTERIA change, though the streptococcus reduces methylene blue readily. The organism is in need of a hydrogen acceptor, as this reduction proves, but cannot make use of oxygen gas. Methylene blue should have acted as a catalyst, but the small amount of this compound used (0.5 mg. in 100 c.c. of milk) could transfer to the bacteria or their products only 0.028 mg. of oxygen = 0.01 c.c; the process would have to be repeated several hundred times before the oxygen consumption could be detected by this method. This agrees with the findings of Callow (see p. 19) that cells of Clostr. sporogenes and Strept. lactis suspended in buffer solution, without food, take up hardly any oxygen, while aerobic and faculta- tive organisms consumed 5 to 25 c.c. of oxygen per hour. The anaerobic organisms lack the faculty to activate oxygen so that it can serve as hydrogen acceptor. Quite different was the experience of Meyerhof and Finkle (1925) with Lact. Delhrucki. This organism, though practically anaerobic, consumes oxygen and produces an equal volume of CO 2. For each molecule of sugar oxidized, the fermentation of about six molecules of sugar to lactic acid is prevented. This corresponds with experiences on the metaboHsm of the muscle. As in most respiration mechanisms, oxygen consumption was greatly increased by methylene blue, and was retarded by cyanides. With small amounts of cyanide, oxidation could be repressed so that almost equal amounts of lactic acid were formed with air and with- out air. The relative amounts were MiUimols KCN Exp. I Exp. II 0 0.7 0 1.0 Without air With air 112 20 33 32 72 7 30 28 The cyanide had also damaged the fermentation mechanism, but not nearly to the same extent as the respiration mechanism. The data of Table 9 can be used only with complete oxidations. For incomplete oxidations, and especially ENERGY SUPPLY OF THE CELL 75 for intramolecular oxido-reductions of type 2 (p. 70) where acceptor and donator are the same molecule, the above data cannot be applied. In intramolecular fermentations, energy is liberated by shifting the oxygen towards the end-carbon, and by massing the oxygen on one carbon rather than having it distributed evenly to the various carbon atoms. As an example, we can write the glucose molecule and its products thus: H H H H H H HC . C . C . C . C . C glucose 0 0 0 0 0 0 H H H H H O H H H H O C HC . CH HC . CH C alcohol + CO2 O O H H O O H H O H H H H O C . C . CH HC . C . C lactic acid O O H H O O H H H H The shifting of oxygen from the inner to the outer carbon atoms liberates energy. A compound with oxygen on the outer carbons only can not undergo an intramolecular oxido-reduction. All the simple alcohols, all fatty acids (and, of course, all hydro-carbons), as well as the divalent acids like oxalic, malonic and succinic acid can not be fermented except by a different type of hydrogen acceptor, such as oxygen gas or nitrate or perhaps certain protein products. Organic acids may undergo decar- boxylation, but that does not, as a rule, lead to a gain 76 PHYSIOLOGY OF BACTERIA of energy, as may be seen from the following combustion heats per gram molecule : CHa-CHOH-COOH -> CHs'CHsOH + CO2 326 cal.* 326 cal.* Lactic acid Ethyl alcohol COOHCHOHCHOHCOOH -> CH2OHCH2OH + 287 cal.* 287 cal.*2 p^ Tartaric acid Ethylene glycol L/W2 The uniform distribution of oxygen in the carbohydrates makes them especially fit for intramolecular oxido-reduc- tions. The same holds for glycerol. Hydroxyacids (e.g., lactic, tartaric) and amino acids which by hydrolysis may be changed to hydroxy acids, can also be fermented anae- robically according to type 2. The organisms (i.e., the catalysts) which have the faculty to bring about intramolecular changes of this type, can exist anaerobically only if the medium contains compounds of the just-mentioned type. Bad. coli dies without oxygen in a solution containing lactate only, though with access of air, it can utiHze the lactate by using oxygen as a hydrogen acceptor (p. 187). Some peptones contain a compound which permits anaerobic growth of Bad. coli. According to Treece (1928), Bad. coli and aerogenes formed acid and gas in anaerobic solutions of Difco's, and Parke, Davis' peptone, but not with Armour's or Witte's peptone. The products do not seem to be derived from carbohydrate radicals. A peculiar exception to the oxido-reductions of organic and inorganic compounds is the urea fermentation which is a hydrolysis. As a rule, the energy yield of hydrolysis is almost negligible (see p. 25). The cause of this exception of urea fermentation is the circumstance * Combustion heat. ENERGY SUPPLY OF THE CELL 77 that by hydrolysis, an acid and an alkali are formed. We can write the urea fermentation in two steps : CO(NH2)2 + H2O = CO2 + 2NH3 2NH3 + CO2 + H2O = (NH4)2C03 The energy yield of the second process has been deter- mined, and Landolt-Bornstein mentions two values, 10.7 and 15.9 cal. per gram molecule of ammonium car- bonate. The energy yield of the first process, the real hydrolysis, can not been measured separately. Com- bustion heats of urea are given only with nitrogen gas as final product. By combining the values for heats of formation of urea, ammonia and water, the computed energy yield of the hydrolysis proper is negative, —1.5 cal. This value is quite uncertain, representing the difference of various, none too exact, heats of formation; but it is probably near zero. The only energy available to the cells of urea bacteria is that of the second process; namely, the heat of neutralization. This does appear a strange source of energy. The same principle probably applies to the organic acid amids which are contained in the protein hnkages. The heat of neutrahza- tion of acetic acid by ammonia approximates 12 calories per gm. molecule. These values are quite considerable if we realize that the energy yield obtained from the alcoholic fermentation of sugar is only about 26 cal. per gm. molecule. The heat of neutralization also must enter into some of the inorganic fermentations (formation of nitrous acid from ammonia, of sulfuric acid from hydrogen sulfide or sulfur, and in denitrification and sulfate reduction).* This discussion of the role of oxygen for the energy requirements leads us to wonder to what extent our distinction between aerobic, facultative and anaerobic bacteria is justified. Since we judge this property primarily by the growth of bacteria, it really should * Another peculiar type of oxygen-free fermentation is the spUtting of fatty acids into CH4 and CO2 (Thayer, 1931). 78 PHYSIOLOGY OF BACTERIA not be discussed in this part of the book. However, the source of energy is one of the most important factors in growth, and so the two might well be combined and considered here. It might seem that we could define strict aerobes as bacteria which must have oxygen for hydrogen accep- tors. One may object that this definition will fail as soon as new hydrogen acceptors are found which permit growth without oxygen gas. Since vinegar bacteria might obtain their oxygen for fermentation from methy- lene blue, and since oxidation of a medium with per- manganate makes aerobic bacteria grow farther down below the surface (see p. 71), we are warned against making our statements too positive. Another factor enters in, namely, the growth require- ments. Sohngen (1909) has shown that urea fermenta- tion (which needs no hydrogen acceptors) is entirely sufficient for the energy requirements of some urea bacteria, and that very small amounts of organic acids suffice for the carbon of the cell construction; no growth will take place, however, in the absence of oxygen. In the construction of cell compounds from organic acids, dehydrogenation must become necessary for certain parts of the building material, and oxygen is the only acceptor for this hydrogen. Urea cannot possibly act as hydrogen acceptor. When it comes to the differentiation between faculta- tive and strict anaerobes, another factor enters, namely, the toxic effect of oxygen. Really, the only definition that holds with all anaerobes is that oxygen of the concentration found in air is toxic, and prevents growth. Anaerobiosis has no direct connection with the source of energy, and some Beggiatoa species have been claimed to grow only with reduced atmospheric oxygen pressures, ENERGY SUPPLY OF THE CELL 79 while they obtain their energy from the oxidation of hydrogen sulfide. On the other hand, the Streptococci are not able to utilize oxygen (p. 20), but are not especially sensitive to it; we do not ordinarily consider them as anaerobes. Further, the sensitivity of bacteria to oxygen depends upon the number of cells present and upon their physio- logical condition. Large inocula of anaerobes will frequently grow when small inocula of the same culture in the same medium will not. Finally, the kind of medium is of great influence. It has been shown that by oxidizing a medium, or by adding appropriate hydrogen acceptors, aerobic bacteria may develop without oxygen gas. It has also been shown that the addition of pyruvic acid or cysteine to a medium will permit growth of some strictly anaerobic species without special protection from oxygen. Our differentiation between aerobic, anaerobic and facultative organisms is quite useful as a working definition, but it involves a number of different unrelated principles, such as the need of oxygen for energy libera- tion, the possibility of substitutes for this free oxygen by organic or inorganic hydrogen acceptors, the require- ment of oxygen for growth when it is not required for energy, the toxicity of oxygen to the cell, and the over- coming of toxicity by special media. There are no clear-cut groups set off distinctly by their reaction towards oxygen; all are allied in varying degrees, accord- ing to the composition of the culture medium. In discussing oxygen requirements, it should be remembered that the solubility of oxygen is very small, as Table 10 shows. There is only between 0.0007 and 0.0010% of oxygen in solution, and considering oxygen as a food, a medium with such a small amount of food 80 PHYSIOLOGY OF BACTERIA would be considered very poor. It is evident that this supply must soon be exhausted. Table 10. — Solubility of Oxygen in Water Temperature, °C. Oxygen atmosphere, % Air, % 0 0.006941 0.006067 0.005364 0.004799 0.004335 0.003928 0.003585 0.003312 0.003078 0.002857 0.002654 0.000000 0.00145 5 0.00126 10 0.00112 15 0.00100 20.... : 0.00090 25 0.00082 30 0.00075 35 40 0.00069 0.00064 45 0.00060 50 0 . 00055 100 0.00000 A. Miiller (p. 187) has shown that Bad. coli growing in a poor medium requires about 1.6 X 10~^° mg. of oxygen per hour; Pseudo- monas fluorescens needs considerably more oxygen, A full-grown culture of Bad. coli contains about 10^ cells per c.c, or 10^^ cells per 100 c.c, and these cells would require 1.6 X lO-^" X lO^i = 16 mg. of oxygen per 100 c.c. of medium per hour. Since only 1 mg. is available in the medium, the supply will be exhausted in a short time. Diffusion of oxygen is far too slow to provide the bacteria in the lower layers. There can be no doubt that in a testtube culture or flask culture of aerobes, all cells will exist under prac- tically anaerobic conditions except those in the very top surface layer. SUMMARY OF FACTS The processes that yield energy to the cell involve shifting of oxygen, either within the molecule of the substrate, or between two different kinds of molecules. ENERGY SUPPLY OF THE CELL 81 The only exception is the urea fermentation, and, possibly, some similar decompositions of acid amids in protein cleavage. Oxygen exchange between two different types of molecules depends upon the special cell mechanism as well as upon the energy liberated. Oxidation does not always take place when oxygen is available. Some strictly aerobic bacteria can be made to exist anaerobically by supplying them with highly oxidized compounds acceptable to the species. Anaerobic growth of facultative organisms is possible only with special substrates. Many substances not permitting anaerobic growth will allow abundant growth of the same species in air. Some strictly anaerobic bacteria may be made to grow in unprotected testtube cultures if certain chemical substances are added. SUMMARY OF THEORIES Fermentation depends upon activation of certain hydrogen atoms of the substrate as well as upon an appropriate acceptor for the activated hydrogen. The acceptor may be the same substrate, or part of it, or a different substance, or elementary oxygen. An exception to this type of fermentation is the urea fermentation. Energy in this process is not liberated by the hydrolysis proper, but by the heat of neutraliza- tion between ammonia and carbonic acid. Streptococci and Clostridia cannot activate oxygen to function as hydrogen acceptor. This lack has no direct relation to anaerobiosis, for streptococci grow fairly well in air. The distinction between aerobic, facultative and anaerobic species is indefinite. Besides the properties of 82 PHYSIOLOGY OF BACTERIA the species, the composition of the medium decides whether an organism can grow without oxygen, and whether it can tolerate the toxicity of oxygen. IV. THE OXIDATION— REDUCTION POTENTIAL (a) ELECTROMETRIC MEASUREMENTS OF POTENTIALS In 1911, Potter conceived the idea that, since bio- chemical manifestations are frequently accompanied by electrical phenomena, bacterial cultures might also show a change of potential. He tested this assumption by filling a porous cylinder with a liquid medium and placing it into a glass jar filled with the same medium. One of the two portions of medium was inoculated with a microorganism. Platinum electrodes were then placed into both liquids and an electric current of a certain voltage was actually observed, the whole resembling a galvanic cell. The differences in potential were found to be with Volts Bad. coll in tartrate-asparagin solution 0 . 308 Bad. coli in tartrate-starch solution at 30°C 0.349 Bad. coli in tartrate-starch solution at 20°C 0.534 Yeast in glucose or sucrose solution 0 . 300-0 . 400 Potter concluded from this evidence that the disintegra- tion of organic compounds by microorganisms is accom- panied by the liberation of electrical energy. Nine years later, Gillespie (1920) showed that the electric potentials observed by Potter were only a special case of the oxidation-reduction potential. If two plati- num electrodes are placed in the two open ends of a U- tube filled with acid, and one electrode is saturated with hydrogen and the other with oxygen, the difference in potential of the two electrodes would be 1.23 volts. ENERGY SUPPLY OF THE CELL 83 Instead of using hydrogen and oxygen, less extreme reagents might be used, and the potential difference would be smaller. An oxidation-reduction potential is established when an electrode is placed in a liquid which does not react with it chemically. This potential indicates the intensity of reduction. It does not tell anything about the amount of substance that can be oxidized or reduced. This potential is comparable, as a simile, to the hydrogen ion concentration which tells us the intensity of the acid, but not the amount of acid present. ^"'i ^.^ ^X~ r 3 ^ct- • / Z ^ / -rD ^0... a n. r" 60 IZO /so ZkO 300 fnc'ncct^!> Q I I I I I I I I * t I I I I I ' - Fig. 3. — Differences in potential between a mercury electrode in bacteria cultures, and a calomel half cell: A, B, and C: a mixed culture with different amounts of oxygen; D: B. subtilis in dextrose-peptone solution. A single potential cannot be measured; we have to combine two potentials and thus obtain an electric current which gives us the difference between the two potentials. Potter in the above experiments measured the differences between the oxidation-reduction potential of the sterile medium and that of the culture. Gillespie coupled the electrode in the culture with a calomel half cell, and thus measured all his potentials against a stand- ard electrode of definitely known potential. The dif- ferences are recorded in volts.. Figure 3 represents a putrifying solution of peptone in a deep layer, containing a mixed culture of bacteria carefully protected against agitation. Curve A gives the potential difference of an old culture when the access of air was carefully avoided. 84 PHYSIOLOGY OF BACTERIA In B, a little air entered accidentally, causing a low start of the potential ; after about 1 hour, the original potential was again established. Curve C shows the result of vigorous shaking with air; the potential remains low and does not recover. The culture was too old for new growth to occur and those old cells which produced the original potential difference (probably anaerobes) were apparently unable to function. In some other experiments, Gillespie measured also the hydrogen ion potential by bubbling hydrogen through the medium. The following values were obtained: Table 11. — Hydrogen Electrode and Reduction Electrode Potentials of Various Cultures of Bacteria Hydro- gen elec- trode, volts Reduc- tion elec- trode, volts Differ- ence, volts Mixed culture in Peptone Witte, 17 days Mixed culture in Peptone Witte, 20 days B suhtilis in Peptone Witte 0.707 0.715 0.753 0.622 0.594 0.554 0.594 0.700 0.652 0.630 0.616 0.314 0.275 0.580 0.516 0.594 0.370 -0.040 0.077 0.099 0.439 B. suhtilis, in Peptone + sugar Bad coli, in Peptone + sugar 0.347 0.014 Bad coli in Peptone + sugar 0.038 Bad. coli, suspension in 0.9% NaCl. . . . B. mycoides, suspension in broth Soil aerobe, suspension in phosphate sol. 0 0.330 0.692 The reduction potentials of the aerobic spore-formers, B. suhtilis, B. mycoides and one unnamed species, are quite low in comparison to the facultative Bad. coli and the putrifying solution containing anaerobes. The hydrogen electrode potential is the potential we measure in determining pH, except that it is recorded here in volts. We can also define it as the reduction potential of hydrogen gas at 1 atmosphere of pressure ENERGY SUPPLY OF THE CELL 85 at the pH of the culture under test. If the hydrogen pressure is less than 1 atmosphere, the reduction intensity will be lower, and it can be computed how much lower it will be for any given hydrogen pressure. This leads to the conception of Th (see p. 90). Gillespie's data show that in one case, Bad. coli produced the same potential as the hydrogen electrode; in other words, this cul- ture reduced as strongly as hydrogen gas at 1 atm. pressure, and, consequently, could develop hydrogen gas. The differences between hydrogen potential and reduction potential of the other two cultures of Bad. coli correspond to hydrogen pressures of 0.34 and 0.052 atm. respectively. The potentials mentioned above are differences in potential between the culture and an arbitrary standard, in one case the sterile medium, in the other case a calomel half cell. There is no absolute zero point for electric potentials, but in order to compare potentials, a zero point is fixed arbitrarily. The potential of the normal hydrogen electrode is generally accepted as the standard zero point. This may be compared to fixing the zero point of the thermometer scale by the freezing point of water. The difference in potential between the normal hydrogen electrode and the saturated calomel electrode is 0.247 volts, and this amount must be added to measure- ments with the saturated calomel electrode to make them comparable with the standard scale. The EMF on the standard scale is usually expressed as Eh. The Eh of most sterile media in air is about +0.2 to 0.3 volts, and becomes negative by the action of bac- teria. Therefore, the curves' of potential change in bacterial cultures will show a decrease of potential if plotted on the standard or Eh scale. The lower the potential, the lower is the oxidizing intensity, and the stronger is the reducing intensity. If we plot potential 86 PHYSIOLOGY OF BACTERIA differences, however, as in Figs. 3 and 8, the difference becomes larger as the Eh potential drops. This accounts for the two different types of curves representing reduc- tion potentials. Cannan, Cohen and Clark (Clark, Cohen and assoc. 1928, X) obtained the curves shown in Fig. 4 representing the reduction potentials of a culture of Bad. coli and of Bad. vulgare, on the stand- ard or Eh scale. The potential of the colon culture drops considerably +0.15 •^^^^^ ^ r-^^. \ ■^ -^ -0.Z5V- A \ - .0,¥5 3 3> ^_ ===* ^ 0 2 o. V o b o a I.O 1 7. /.¥ 1 t /. 1 a. Fig. 4. -Reduction potential of Bad. coli (A); hydrogen potential of Bad. coli (B), and reduction potential of Bad. vulgare (C). lower than that of the proteus culture. The time is plotted logarith- mically. The third curve, B, is the potential of a hydrogen electrode in the colon culture. After ten hours (log 10 = 1.0), the reduction potential of the colon culture drops sHghtly below that of the hydro- gen electrode. This is in agreement with Gillespie's data on p. 84. Hydrogen gas can be formed only if these two potentials are the same. Several interesting potential-time curves are also given by Thorn- ton and Hastings (1929a and b). Raw milk showed a potential between -1-0.200 and -1-0.300 volts when fresh, and was finally reduced to about —0.200 volts, probably by the action of streptococci (see Fig. 9). The change of potential with time, with the medium, with aeration and with different species has been studied in some detail by Hewitt (1930a, b, c, d). Streptococci and Pneumococci produce peroxide; with the latter, aeration brings the potential above that of the sterile ENEEGY SUPPLY OF THE CELL 87 medium. Neither of these two species maintain the reduction potential very long; it rises slowly in ordinary aerobic cultures. Micr. pyogenes aureus and Corynebacterium diphtheriae have a lower Fig. 5. — Relation between reduction potentials and acidity. potential in aerobic cultures than in anaerobic. Perhaps they need oxygen to produce the energy required for the maintenance of the potential. With glucose, the potential is not as low as without. Serum prevents a low potential; it probably acts as an oxygen carrier. 88 PHYSIOLOGY OF BACTERIA (6) REDUCTION POTENTIAL AND ACIDITY The interpretation of reduction potentials is made quite difficult and complicated by the fact that it varies with the hydrogen ion concentration. This variation is not always regular. It depends upon the state of ionization in the system, and different systems are not affected in the same way by the same change in pH. The system FeCU — FeClg is practically uninfluenced by acidity. Ordinarily, however, the poten- tial drops with increasing pH, i.e., with increasing alkalinity. This is especially true with the hydrogen and the oxygen electrode. The system cysteine-cystine, which will interest us later, has been investigated by Dixon and Quastel (1923), and they found a simple rectilinear relation between potential and pH up to pH 9.5, as may be seen from Fig. 5. The same is true with glutathione though the range investigated by the authors is smaller. The effect of pH upon the potential of the hydrogen electrode is also indicated in this figure. The relation between potential and pH is not always linear, as may be seen from Fig. 6, which represents the system methylene blue — methylene white taken from Clark, Cohen and associates (1928, VIII). A definite potential is required to change the blue dye to the colorless leuco-compound, but this potential varies greatly with the acidity of the solution. At pH 7, a potential equal to our arbitrary zero point is sufficient to turn the color, at pH 10 it requires a potential of — 0.100 volts to do the same, while at pH 4, a potential of +0.150 volts would be sufficient. This figure also shows short stretches of the potentials of the hydrogen electrode and the oxygen electrode and demonstrates the ENERGY SUPPLY OF THE CELL 89 linear relation of these potentials to pH. The hydrogen electrode indicates at what potential hydrogen can be formed. At pH 0, the potential 0 is sufficient; at pH 4.5, which is about the acidity in cultures of Bad. coli, the required potential for hydrogen liberation Fig. 6. — Reduction potential of methylene blue at different acidities. is about —0.4 volts and in alkaline solutions, it becomes so strongly negative that hydrogen formation by alkali-forming bacteria appears very difficult. The reduction potential as ^uch, even if expressed in volts on the Eh scale, tells us very little about the reducing power if we do not also know the pH of the culture. This was the aim of Gillespie in determining the hydrogen electrode potentials (Table 11). The 90 PHYSIOLOGY OF BACTERIA hydrogen ion potential is probably the lowest potential a bacterial culture can reach, and the last column of Gillespie's table indicates how far removed the reduction intensity of the culture is from that of hydrogen. If the potential is measured in this way, it is independent of the pH, or rather, it is already corrected for pH. The same principle was employed by Clark, Cohen and associates (1928, II) to elaborate the reduction potential, rH. This might be best illustrated graphically. In Fig. 5, the change of potential of the hydrogen electrode with pH is given. The hydrogen electrode corresponds to a hydrogen pressure of 1 atmosphere. If the pressure is less, the reducing tendency will be less, the oxidizing intensity will be stronger, and so we obtain, for lower hydrogen pressures, parallel lines above that for the hydrogen electrode. This pressure can be calculated from the formula for the potential of the hydrogen electrode : F = -^] Vp ^^ 1 Vp ' F [H+] 0.434P ^ [H^] where P is the hydrogen pressure, R the gas constant, F the Faraday constant, T the absolute temperature. For a constant temperature , Tirp the expression ^ 404/p is constant, and for 30°C., it becomes 0.06. En = -0.06 log j^ = -0.06(log VP - log [H+]) — log [H+] = pH by definition Eh = -0.06(log VP + pH) For pH = 0, we obtain Eh = -0.06 log VP 0;^ = - log Vp ENERGY SUPPLY OF THE CELL 91 This formula makes it possible to express the reduction potential in terms of hydrogen pressure. As an example, the cysteine potential (Fig. 5) is 0.36 volts above the hydrogen electrode (measured at pH 5). Assuming that the two curves are parallel up to pH 0, we have P = 10~^2 atm. pressure This figure seems quite meaningless. But the reduction potential rH is the negative logarithm of this number, in strict analogy to the definition of pH. Thus, the rH of the cysteine-cystine system is 12 or lup- The general formula for the computation of rH from measurements at any pH is derived from the above & = -.^^ log i^ = -o:^(log VP - log [Hi) RT 0.434F ^^^ [H+] 0.434F 0.434i^ 0.434 Eh ' F (log Vp + ph) RT -f pH = - log Vp for sec, rH = 2(5^ + ph) for pH 5, the cysteine potential is B* = +0.06 volt rH = 2(S-;SI + 0 = - The higher the rH, the lower is P, and the farther is the system removed from the reduction intensity of the hydrogen electrode. The computation of rH is permissible only if the potential — pH curve is parallel to that of the hydrogen 92 PHYSIOLOGY OF BACTERIA electrode. It is nothing but an expression of the distance of the system from the hydrogen electrode. Therefore, no rH can be computed for methylene blue which does not go parallel to the hydrogen electrode, nor for ferrous- ferric systems. Clark, Cohen and associates (1928, supplement) warn against the general use of the rH because it may not always be useful, and may be even wrong. (c) REDUCTION POTENTIAL INDICATORS The electrometric measurement of reduction poten- tials requires a number of carefully adjusted instruments, and there are occasions, such as living protoplasm, where direct measurements could not be made at all. It is a very old experience that bacteria have reducing properties; indigo, litmus and methylene blue are the best-known test substances for microbial reduction. But little attention had been paid to differences in the reduction of these compounds. The conception of reduction intensity as contrasted against reduction capacity was lacking then. Clark, Cohen and various associates studied the reduction potentials of a number of dyes. They measured the potential of the dye in the completely reduced stage, and in the various stages of oxidation. It was found that the oxidation took place within a comparatively small range of potential. Decolorization occurred when about 70-95% of the dye is reduced. With different dyes, the relative position of the potential differed widely. The reduction potentials of a few of the tested indicators are shown in Fig. 7. It is seen that 50% of the methylene blue is reduced at an electrode potential of —0.002 volts. A bacterial culture which produces exactly this potential, will show a blue color ENERGY SUPPLY OF THE CELL 93 with methylene blue because only half of the dye is reduced. But if the potential goes down to —0.020 volts, the culture will be colorless. It would also reduce of dye reduceoL Fig. 7. — Reduction potentials of different indicators at pH 7.4. dichloro-indophenol and naphthol-sulfonate indophenol, but would not change indigo tetra sulfonate nor disulfonate. 94 PHYSIOLOGY OF BACTERIA Thornton and Hastings (1929b) compared methylene blue with janus green B both of which have been recommended for the reductase test in milk, and found that janus green B indicates a potential about 0.1 volts lower than methylene blue. (See points A and B in Fig. 9.) They also found that methylene blue in milk was reduced at a poten- tial higher than that given by Clark, Cohen and associates. The indicators are greatly influenced by the acidity of the medium, as can be seen from Fig. 6 illustrating the influence of pH on methylene blue. (d) BUFFERING OR POISING OF THE POTENTIAL It has been stated before that the reduction potential measures the intensity, but not the quantity of reduction. The term quantity is now usually substituted by the :^ in HE HZ- -tatf- y^ — ^ " -tai, 1 --^^ ifdaus f ,<:^, Z A c/ays 2, ^ elays Fig. 8. -Differences in potential between 4 water-logged soils and a calomel electrode. more significant term ^^ capacity." A bacterial culture will not only show a certain intensity of reduction (which might be measured by reduction indicators) but it has a definite capacity for reduction. We can imagine a medium containing a large amount of a substance which acts as hydrogen acceptor at a Eh of about +0.05 volts or less. The potential of a bacterial culture will remain near this point until all of this compound is reduced, and then will fall to a lower potential. This corresponds ENERGY SUPPLY OF THE CELL 95 exactly to the buffers observed and used extensively in working with hydrogen ion concentrations. In reduc- tion potentials it is spoken of usually as the ''poising" effect. The first example for such buffering or poising is probably that by Gillespie (1920) who measured the reduction potentials of various waterlogged soils. The " 1 — 1 A/ "" -'I . 4/ 7 >/ . , " v -- + 0 / - k^ '>S k 3 V- "-N ^ mf% 1 y ^ ^ -02.- V, / -s I 2 » < ^ iou rs i f Fig. 9. — Reduction potential of raw milk with two different indicators. results of the daily measurements for the calomel electrode as zero are shown graphically in Fig. 8. While the last three soils show a potential greatly differing from that of the calomel electrode, with a strong reducing intensity on the second day, the first, soil remained much less reducing for four days, and on the fifth day was still less reducing than any of the others on the second day. Repeated experiments proved this to be a constant character of this one soil which is evidently well poised. 96 PHYSIOLOGY OF BACTERIA Another example of poising is shown in Fig. 9 which represents potentials in the same milk with two different dyes as measured by Thornton and Hastings (1929b). Methylene blue has only a negligible poising effect as was shown in other experiments. Janus green Bj however, though it stimulated reduction at first, did not let the potential drop as suddenly as is the custom in milk, or in milk with methylene blue ; it came down very slowly, and not to as low an end point as with methylene blue. Cohen (1931) added substances of the type of potas- sium ferricyanide or benzoquinone which, after partial reduction by the bacteria, maintained a readily reversible oxidation-reduction system in the medium. (e) THE FACTORS INVOLVED IN REDUCTION All living protoplasm seems to have the power to effect chemical reductions, and this holds true also with bacterial cells, and consequently with bacterial cultures. With some of them, this power is very pronounced, as with anaerobic bacteria. The power to reduce is identical with the power to oxidize, both being accomplished simultaneously by the transferring of hydrogen from one molecule to another. The reduced substance may eventually be the oxygen gas dissolved in the medium. Aerobic bacteria will soon exhaust the oxygen dissolved in a medium, and the medium will then remain free from oxygen as long as the metabolism of the bacteria remains active, excepting a very thin surface layer into which the oxygen diffuses from the air (see p. 80). Many anaerobic bacteria can accomplish the same if they are present in sufficient quantity at the start to overcome the harmful effect of the oxygen. ENEEGY SUPPLY OF THE CELL 97 Most culture media are reducing solutions, even when sterile. All organic compounds can be oxidized, and it might be expected that they all produce a reduction potential. Only reversibly oxidizable compounds can produce a measurable potential, however. Most organic compounds are quite stable; a sterile solution of sugar or of amino acid remains unchanged. This is only a false equilibrium, however, for Warburg (1921) could demonstrate that amino acids are slowly oxidized in aqueous solution at body temperature if Merck's blood charcoal is added, and Meyerhof and Weber (1923) could oxidize, in the same way, not glucose as such, but the glucose phosphate. At the same time, Hopkins (1921) had found gluta- thione to be the cause of auto-oxidation in cells, and the SH group to be the essential part of the oxidizing property of this molecule. In the presence of glutathione, or cysteine, or thioglycollic acid (all possessing the SH group) organic matter, such as yeast juice, or other cell debris, takes up much more oxygen than corresponds to the amount of sulfhydryl-compound added. In other words, these compounds act as oxygen-catalysts, and overcome the false equilibrium of organic compounds in solution. Culture media are mixtures of many organic compounds, and it has been observed quite early (Th. Smith, 1896) that sterile broth will combine with oxygen, and will reduce methylene blue if protected from air. Coulter (1929) measured the reduction poten- tial of a meat infusion broth (pH 7.6) electrometrically, and found that the potential in air was between -1-0.150 and +0.250 volts. When H2O2 was added, it rose to more than 0.400 volts. When nitrogen gas was bubbled through the broth to remove the oxygen, the potential gradually decreased (see Fig. 10) and in all of his experi- 98 PHYSIOLOGY OF BACTERIA ments with different specimens of broth, it approached a value between —0.050 and —0.060 volts. This is the potential of cysteine which is likely to be present in meat infusion in very small quantities. s (X/X OJOO «s^^ -0,0^ ^\^ 0 .^0,050 SL 20 .-JS.^ 4o SO Aours f20 Fig. 10. — Gradual drop of potential in sterile broth after 5 hours of deaeration with nitrogen gas. Dubos (1929) determined by means of indicators the reduction potential of meat infusion broth made with Fairchild's peptone, and he observed decolorization as far as indigo disulfonate (see Fig. 7) which means a potential lower than —0.125 volts. The quantity of dye that could be reduced was only 0.0017 molar. Thornton and Hastings (1929) found the potential of fresh milk to be between +0.200 and +0.300 volts. The final potential of deaerated milk is sufficiently low to decolorize methylene blue. The addition of cysteine brought the potential of raw milk from 0.2 to 0 volts in fifteen minutes, and to —0.2 volts in three hours. At this time, the check sample without cysteine had just started on the decrease, and had gone from +0.2 to +0.15 volts. The final potential of both samples was —0.2 volts, but this may have been due to bacteria and not to the milk as such. Since each medium has a characteristic reduction potential which establishes itself some time after the removal of dissolved oxygen, and since all aerobic and ENERGY SUPPLY OF THE CELL 99 facultative microorganisms will remove dissolved oxygen from the medium in which they grow, the potential difference between a sterile medium and a culture in the same medium may be due only to the removal of oxygen which brings the natural reducing properties of the medium into play. Thus, Thornton and Hastings (1929) confirm the conclusion of Barthel that the reduc- tion of methylene blue in the so-called reductase test of milk takes place in two stages: (1) the removal of dis- solved oxygen by bacteria, (2) the reduction of the dye by constituents of the milk. Only if the potential of a culture drops below that observed in the oxygen-free medium, are we justified in stating that we are dealing with a reduction potential characteristic for that microorganism. It seems, from all these experiments, that the oxida- tion-reduction potential in bacterial cultures is compara- tively simple. The cells of bacteria, as such, do not establish the potential. If cells are washed once and suspended in salt solution, they are still able to hold a distinct potential, but after repeated washing, it becomes indefinite. The potential must be produced by some metabolic products of the bacteria. It cannot be doubted that a definite potential is established inside of the cells, but this cannot be measured in the cultures. An oxidation-reduction system must contain a reversi- bly oxidizable substance. This substance may act as a catalyst and transfer the oxygen to another substance which is irreversibly oxidized. The reversibly oxidizable compound is represented by compounds like glutathione, cysteine, thioglycollic acid, or perhaps by a dye (all oxidation-reduction indicators being reversible) or even by an oxidizing enzyme. Methylene blue is a very active catalyst of cell respiration. 100 PHYSIOLOGY OF BACTEKIA The kind of the compound determines the potential. The quantity of this compound is one of the main factors in determining the rate of reduction. The reduction capacity of the system is determined by the sum of the reversibly and the irreversibly oxidiza- ble substances. This might be illustrated by an experiment of Dubos (1929) who was the first to emphasize the rate of reduc- tion as an essential factor. He measured the rate of methylene blue reduction by cysteine in vaseline-sealed tubes at 37°C. Table 12. — Time in Minutes Required for the Reduction op Methylene Blue (0.0001%) by Cysteine Concentration of cysteine 0% 0.01% 0.02% 0.03% 0.05% Phosphate bufifer (pH 7.5). No reduction Partial reduction 150 120 60 Phosphate buffer (pH No Partial 40 30 12 7.5) +2% broth. reduction reduction Phosphate buffer (pH No Partial 40 25 12 7.5) +2% yeast ex- reduction reduction tract. The rate of reduction was determined by the quantity of cysteine, but also by the amount of irreversibly oxidizable organic compounds from broth and yeast extract. The meaning of the reduction potential of a bacterial culture is, at the present stage of developments, rather uncertain. If a bac- terium produces a potential lower than the final potential of the oxygen-free medium, it indicates primarily nothing but the presence of a cell product with a lower reduction potential. Next comes the question: If the cell forms a compound which creates a potential of —0.100 volts in the medium, must this cell produce within its protoplasm this same or a still lower potential in order to produce such a compound? Must the reduction potential ENERGY SUPPLY OF THE CELL 101 within the cell of Bad. coli drop to —0.4 volts in order that hydrogen can be formed from sugar ? If not, then the low reduction potential of Bad. coli does not really indicate that it has strongly reducing powers, but it is low because the bacteria form hydrogen, and the reduction electrode is simply changed to a hydrogen electrode by the gas formed. Or the potential —0.06 simply indicates that cysteine has been formed by the bacterium under test. There seems to be some reason for doubt as to what is the cause, and what, the effect. The introduction of oxidation-reduction potentials, though not quite clear as yet in all its aspects, will influence quite considerably our conceptions of the mechanism of life, and of biological oxidations. It will prove useful to explain many reactions hitherto difficult to account for. It will tell us why certain reactions take place in the cells, and our conception of microbial activity will be shifted to a somewhat different basis. (/) SUMMARY The common culture media made from meat extract have a reduction potential of about +0.2 to +0.3 volt taking the normal hydrogen electrode potential as zero. Milk has a similar potential. If these media are deaerated by vacuum or by a neutral gas, the potential drops below zero. They contain reducing compounds which come into play if the dissolved oxygen is removed from the medium. They will then decolorize methylene blue. This reducing property of our common media might be caused by traces of glutathione, cysteine or similar com- pounds in the meat extract or milk. Bacterial cultures will usually bring about a decrease in potential. There is not sufficient material on hand to say whether all bacteria will produce a potential lower than that of the deaerated sterile medium. 102 PHYSIOLOGY OF BACTERIA With some anaerobic and facultative bacteria, the potential is very low. With those organisms producing hydrogen gas, the reduction potential equals or approxi- mates that of the hydrogen electrode at the same pH. The reduction potential is quite meaningless unless we also know the pH of the culture. V. RATE OF FERMENTATION (a) METHODS OF MEASURING THE RATE OF FERMENTATION The very rapid rate of multiplication of bacteria makes it difficult to perform experiments with a constant number of acting cells. The respiration of a leaf or the oxygen consumption of a mouse can be measured over a period of several hours without the organism changing considerably in size or weight during the obser- vation. With bacteria, we may have a doubling of the active mass in less than half an hour. It requires special precautions, therefore, if the rate of fermentation, i.e. the amount of products per cell, or per gram of cells, or per gram of cell solids, is to be determined. The choice of the unit is rather difficult to make. Physiologists working with larger organisms usually take the weight of the organism or tissue as the basis of measurement. Feeding experiments with animals are computed on the basis of kilogram body weight. Since this includes bones and fat, the proportions of which vary greatly, the basis is not really ideal, but it is more accu- rate than the basis of the single organism. With microorganisms, the basis by weight would be possible, but rather difficult to ascertain. It is compara- tively easy with yeast, but much more difiicult with bacteria. The determination of the weight of bacterial cells is possible only if they can be filtered or centrifuged from their medium. It is impossible with cultures in ENERGY SUPPLY OF THE CELL 103 milk or blood or on solid media. In these cases, the single cell is the best unit. Since the determinations will always be averages of many billions of cells, individual differences will be eliminated entirely. The number might be ascertained by plate count, direct count, turbidity measurements or other methods. A simple method for studying the rate of fermentation undisturbed by multipUcation of cells is to use such large numbers of cells that they will not multiply. It is a rule to use more cells than will naturally develop in such medium. This method was suggested by Duclaux (1901, vol. IV p. 328), and applied by Rubner (1904), Slator (1906), and others. Rubner improved upon this method by studying the alcoholic fermentation with large quantities of yeast in pure sugar solution; since there was no nitrogenous food available, the yeast cells could not multiply to any appreciable degree. Slator (1906) found that the presence or absence of nitrogenous food did not influence the rate of fermentation. He measured the decrease of vacuum through CO2 produced by the yeast; this method made it possible to determine the rate of fermentation in a very short time, five to twenty minutes. Slator's technique was later slightly changed by Rahn (1929) who measured pressure increase instead of vacuum decrease. Another method which was used occasionally by Chassevant and Richet (quoted after Duclaux, 1901) was to add slight amounts of a weak disinfectant to the culture, sufficient to suppress growth, but not sufficient to retard fermentation. This method does not seem very advantageous because the disinfectant is likely to affect gradually the fermenting enzyme. An entirely different principle is involved in the computation of the average number of acting cells by 104 PHYSIOLOGY OF BACTERIA considering their mode of multiplication. The formula for the '^fermenting capacity" of the average single cell i.e., the amount of products formed per cell and per hour, is ^ 2.301 ' AiSqog h - log a) ^ At{b - a) where AS is the increase in products during the time interval under consideration, a the number of cells at the beginning, b the number of cells at the end of this interval, and A^ the time of the interval. The origin of this formula is developed in Appendix p. 404. This formula holds true only as long as the bacteria are multiplying rapidly. When the rate of growth decreases, the arithmetic average gives more nearly correct rates than the above formula (see p. 407). (&) THE RATE OF FERMENTATION OF DIFFERENT ORGANISMS The rate of fermentation is not constant during the entire development of a culture. The rate is fastest when the culture is young; with increasing age, the rate gradually decreases until it reaches zero. The causes will be discussed on p. 107. The rates of fermentation mentioned and compared here refer only to young, actively fermenting cells. The rate of fermentation is of great practical impor- tance in the bread yeast industry. Several different methods are used for determining the fermenting capacity for baking purposes. One of the old methods still in use is that of Kusserow where 10 gm. of yeast, 40 gm. of sugar and 400 c.c. of water are mixed, and where the developing CO2 forces water out of a container; this water is measured; its volume corresponds to the CO2 volume produced. The total gas formed in two hours ENERGY SUPPLY OF THE CELL 105 at 30°C. by a good compressed yeast varies from 570- 660 c.c. (Henneberg, 1926, I). This corresponds to 30 c.c. of CO2 per gram of yeast per hour, or about 120 mg. of CO2 (or of alcohol) per gram per hour. Henneberg gives also the following data for other types of yeast: Table 13. -Volume of CO2 Produced by 10 Gm. op Yeast in 400 C.c. OF A 10% Sucrose Solution Beer yeast, bottom yeast Distiller's yeast Wine yeast First half hour 55 70 210 100 400 200 400 260 60 60 160 260 280 310 280 280 63 38 Second half hour Third half hour 153 142 340 330 Fourth haK hour 324 408 Total 1065 630 210 125 780 910 154 180 880 918 Mg. CO2 per gram yeast per hour 174 182 For further data with bread yeast see p. 238, Table 65. Rubner's (1913) experiments give about 0.4 gm. of sugar fermented per gram of yeast per hour; this corre- sponds to 200 mg. of CO2 per gram yeast per hour (see Table 15, p. 108). From Rubner's data that 5 gm. yeast = 38,700,000,- 000 cells, we obtain a CO2 production of 258 X 10~^° mg. or a sugar consumption of 510 X 10~^° mg. of sugar per cell per hour. Another fermentation whose rate has been studied in some detail is the lactic fermentation. Rahn (1911) mentions a number of experiments by Marshall and Farrand (1908) from which he computed the fermenting capacity by the formula mentioned above. He also made experiments of his own. The data given in Rahn's 106 PHYSIOLOGY OF BACTERIA paper are here recalculated for the more correct Buchanan formula. More recent experiments by Baker, Brew and Conn (1919) are based on microscopic counts, and are, therefore, to be multiplied by 1.8 if they were to Table 14. — Fermenting Capacity of Lactic Streptococci (Mg. of acid formed per average single cell per hour) Investigator Number of experiments Average fermenting capacity Limits of fermenting capacity Marshall & Farrand Strain 1 7 10 8 8 14 3 6 12.2 X 10-i«mg. 5.2 9.2 13.6 23.4 15.8 11.4 8.8 9.1-20.7 X 10-10 mg. Strain 2 2.3-10.9 Strain 3 6.2-18.7 Strain 4 5.9-17.5 Rahn Dairy starter 10 . 4-34 . 7 Strain II 9 . 5-23 . 5 Strain IV 7.3-13.8 Baker, Brew, Conn 6.55-11.4 64 12.45 2.3-34.7 be compared with the others calculated from plate counts. This has not been done in Table 14. This Table 14 shows that the average cell (by plate count) of StrepL lactis produces between 10 and 15 X 10"^° mg. of lactic acid per hour. This amount seems ridiculously small, but it is enormous if compared with the volume of the cell. The average cell of Bad. coli has been calculated to weigh about 8 X 10"^*^ mg. (see Appendix). The cells of streptococci are considerably smaller and cannot weigh more than 5 X 10~^° mg. This means then that each cell produces two to three times its own weight of lactic acid per hour, or that each cell ferments two to three times its own weight of lactose every hour. This is a really enormous rate of fermentation, and is possible only by the enormous surface of the cells (50 square feet is the surface of the bacteria in 1 liter of a full grown culture, see Appendix) which allows an almost immediate ENERGY SUPPLY OF THE CELL 107 discharge of fermentation products into the medium (see also Slator and Sand, 1910, on the rate of diffusion into yeast cells). We can further compute the rate of fermentation of Lad. acidophilus from the observation of Rahn (1929b) that 0.045% of lactic acid are formed per hour by 420,- 000,000 cells (microscopic count) of Lact. acidophilus per c.c. This amounts to 45 mg. produced by 42 X 10^ X 102 cells, or almost 10 X lO'^^ mg. per cell per hour. The count probably included many dead cells, which would increase somewhat the fermenting capacity per viable cell. Only one datum on another fermentation is known to the author, namely, the rate of urea fermentation by Micr. ureae. Burchard (1899) used a wrong formula to compute fermenting capacities, but his data show 66 X 10~^^ mg. of urea per cell and hour, four to five times as much as the fermenting capacity of lactic streptococci. SUMMARY The rate of fermentation is known for only a few microorganisms. It is enormously high if compared with larger organisms. High-bred commercial yeasts (beer, wine or bread yeast) ferment about 30-40% of their own weight of sugar per hour. Bacteria being still smaller decompose still more food. Lactob. acidophilus uses at least 50% of its body weight of sugar per hour, the lactic streptococci about two to three times their own weight, and Micr. ureae at least five times its own weight of urea per hour. (c) THE DECREASE OF THE RATE OF FERMENTATION It is a general experience that fermentations come to end, sometimes even before all the fermentable substrate 108 PHYSIOLOGY OF BACTERIA is used up. There are various explanations for this gradual retardation and ultimate cessation of the rate of fermentation; it may be due to lack of substrate, or to toxic action of fermentation products, or to some biological change of the cells. While all these reasons might occasionally fit, the common cause of the decrease in rate ordinarily is either the accumulation of fermenta- tion products, or the decrease of fermentable substrate. Table 15. — Alcoholic Fermentation of Sugar Solutions of Different Concentration by 5 Gm. of Yeast Fermentation constant Sugar Per cent sugar 1 li 2L concentration 20 10 5 2.5 1.25 20 10 2.5 1.25 % % % % % % % % % Hours 2 1.88 1.69 1.83 1.50 0.98 0.0207 0.0187 0.0203 0.0165 4 3.33 3.32 3.29 1.97 1.17 0.0193 0.0192 0.0181 6 4.63 4.64 4.18 2.16 1.24 0.0187 0.0187 8 5.72 5.90 4.46 2.23 1.28 0.0179 0.0185 10 6.68 7.01 4.55 2.31 1.30 0.0172 0.0183 12 7.57 7.73 4.70 2.41 1.33 0.0168 0.0172 14 8.34 8.34 4.80 2.48 1.35 0.0163 0.0163 16 9.05 8.78 4.81 2.54 1.37 0.0160 18 9.63 9.11 4.86 2.60 1.38 0.0154 20 10.27 9.32 4.89 2.63 1.39 0.0152 22 10.88 9.45 4.92 2.64 0.0151 24 11.42 9.52 4.93 0.0144 30 12.90 9.55 0.0119 40 14.98 0.0144 50 16.56 0.0145 60 17.70 0.0147 70 18.34 0.0142 80 18.71 Average 0.0160 0.0181 0.0192 0.0165 90 18.80 ENEEGY SUPPLY OF THE CELL 109 The concentration of the substrate must become very low before it exerts a retarding influence because the rate of fermentation is almost independent of the concentra- tion of the substrate, as will be shown in the next sub- chapter, p. 114. Some very good material on the gradual decrease of the rate of fermentation we owe to Rubner (1913) . Rubner's original data are expressed in calories, as he studied the heat produced by yeast in sugar solutions free from nitrogenous materials. Rahn (1929b) computed from these data the amount of sugar fermented, and Table 15 shows one of these sets, where 5 gm. of yeast fermented 250 c.c. of sugar solution. This table shows the same yeast in sucrose solutions of 20%, 10%, 5%, 2.5% and 1.25%. The retarding effect of alcohol upon the fermentation can be formulated definitely. The simplest form would be the assumption that decrease in the rate is proportional to the amount of alcohol, or that the rate is proportional to the original rate decreased by a certain amount which is proportional to the alcohol already present. This might be written in the following way: Rate at any time = Initial Rate (1 — k X alcohol concentration) Since a fermentation comes to a stop at a fairly definite concentration, which is characteristic for the organism used, the alcohol concentra- tion can be best expressed in terms of this limiting concentration which we will designate as L. Rate at any time = Initial Rate (1 j — ] If the alcohol concentration has reached the limiting concentration, the fraction — j — equals 1, and the rate of fermentation equals Initial Rate X (1 — 1) which is zero; the fermentation comes to a stop. 110 PHYSIOLOGY OF BACTERIA The rate of fermentation can be written in its mathematical form: dx J ^/ alcohoK ax , ,,/. aiconoix which means that the rate with which the sugar is decreased, is proportional to the amount of yeast Y, and is independent of the sugar concentration which, therefore, does not enter the equation at all, and is retarded by alcohol in the manner discussed above. Since one gm. of sugar yields practically one-half gm. of alcohol, we can substitute: alcohol = one-half of sugar, or X alcohol = X because x was the amount of sugar decomposed. Thus we get the equation = ^y(2L-x) Y, the amount of yeast, is constant in Rubner's experiments because more yeast was added than would normally grow in such a solution. k Therefore, the entire expression • F is constant, and may be sub- stituted by another constant, K. The equation -J = K{2L - X) gives upon integration 2L Kt = In 2L -X The accuracy of this equation can be tested by com- puting the X-values for all the above data. The fermentation constants in the right section of Table 15 have been computed with the value L = 10.2 % alcohol. ENERGY SUPPLY OF THE CELL 111 They express the rate of reaction per hour, corrected for the retarding effect of alcohol, and they should be constant during the entire course of the fermentation until the sugar concentration falls below 1.5%. This is not quite the case. The computed data show a slow but distinct decrease. Probably this is not due to the inaccuracy of the data on fermentation, but to a gradual deterioration of the zymase in the nitrogen-free sugar solution. Other experiments by Rubner also show a fair agree- ment with this formulation (see Table 18). Rahn (1929b) verified it in a different way. He measured the rate of fermentation in the presence of increasing amounts of alcohol, using the CO2 pressure as indicator. Table 16 shows the average pressures obtained in three experiments, and, further, the relative rates, based on the unretarded rate as 100. Table 16. — Rates of Alcoholic Fermentation as Inpltjenced by Added Alcohol Experiment II Experiment III Experiment IV Alcohol in solution, mm. mercury Rela- tive Alcohol in solution. mm. mercury Rela- tive Alcohol in solution, mm. mercury Rela- tive per cent per minute rates per cent per minute rates per cent per minutes rates 0 6.58 100.0 0 8.20 100.0 0 6.10 100.0 4.93 3.95 60.0 9.7 2.41 29.4 4.77 4.28 70.4 7.40 2.52 38.3 12.15 1.10 13.4 9.55 1.48 24.3 9.86 1.79 27.2 13.35 1.23 15.0 11.92 0.96 15.7 12.5 0.47 7.1 .... 13.12 14.31 0.87 0.74 14.3 12.1 If the percentage decrease of rate of fermentation is proportional to the concentration of alcohol present, there must be a straight-line relationship between the 112 PHYSIOLOGY OF BACTERIA relative values. Figure 11 shows that this is actually the case until very high values are reached. It seems that from 12% alcohol upwards, there is a very slow yet noticeable rate of fermentation which is almost as high with 12% as with 14.5% alcohol. ■so \ i 4 N. V \ X S \ • ■ZO t 4 ^ i ^ s o Fig. 11. — Rate of alcoholic fermentation in the presence of alcohol. Several series of experiments on the lactic fermentation by Rahn (1911) and by Baker, Brew and Conn (1919) can also be used for a test of the general principle. Here again, the computed values are not absolutely constant, but are sufficiently close to support the principle involved. Rahn (1929b) repeated the experiment shown in Table 16 and Fig. 11 with Lactoh. acidophilus and lactic acid. The results in Table 17 show the rate, i.e., the percentages of acid formed by 420,000,000 cells per c.c. If plotted against the amount of acid added, a straight line is obtained, as in Fig. 11. Even the ENERGY SUPPLY OF THE CELL 113 continued small amount of fermentation at very high concentrations is obvious. Table 17. — Rate of Lactic Fermentation per Hour (By 420,000,000 cells of Lad. acidophilus per cubic centimeter) Initial Rate of fermen- Initial Rate of fermen- acidity, tation in per cent acidity, tation in per cent per cent lactic acid per cent lactic acid 0.16 0.045 0.63 0.0065 0.25 0.0342 0.73 0.0054 0.35 0.0288 0.80 0.0018 0.44 0.0216 0.89 0 0.54 0.0162 0.90 0 SUMMARY OF FACTS The rate of fermentation decreases with the age of the culture. If sufficient substrate is available, the decrease in the rate is caused by the fermentation products. Addition of these to a rapidly fermenting culture will decrease the rate of fermentation. If the added fermentation products are as high as the limiting concentration, there still seems to take place a slow but noticeable fermentation provided that the number of fermenting cells is very large. SUMMARY OF THEORIES The retardation of a fermentation by its own products has been expressed in a simple formula on the assumption that the decrease of the rate of fermentation is propor- tional to the concentration of products divided by the limiting concentration; in other words, the decrease is proportional to the products already formed, expressed in terms of the total possible products. This assumption leads to a formula for a '^ fermentation constant" representing the rate of fermentation corrected 114 PHYSIOLOGY OF BACTERIA for any retardation by its own products. The computa- tion of this constant from available data on the alcoholic and lactic fermentations gave values of sufficient con- stancy to consider them a proof of the general principle involved. (d) THE INFLUENCE OF CONCENTRATION OF MICROORGANISMS It is generally understood that the fermentation produced in any medium is the sum of all fermentations going on in each individual cell. This assumption is the basis of the formula for the fermenting capacity (see p. 404). A number of data are available to show such propor- tionality of cells and products. Slator (1906) found by his method (p. 103) the following relative rates of fermentation : Relative yeast concentrations = 1 : 3 : 5 : 10 : 20 : 25 Relative rates of fermentation = 0.99 : 3.04 : 4.94 : 10 : 19.8 : 24.7 The same can be seen from Rubner's experiments in Table 18 showing the fermentation of nitrogen-free sugar solution by varying amounts of yeast. The fermentation constants, computed with the formula on p. 110, give the rate of fermentation corrected for retardation by alcohol. These rates show good propor- tionality with the quantities of yeast used. (e) INFLUENCE OF THE CONCENTRATION OF SUBSTRATE Considering that fermentation is brought about by enzymes in the cells, we should expect fermentation to ENERGY SUPPLY OF THE CELL 115 Table 18. — Fermentation by Different Quantities op Yeast Percentage of sugar fermented 20 per cent sucrose 8gm. yeast 4gm. yeast 2gm. yeast 1 gm. yeast 10 per cent sucrose 10 gm. yeast 5 gm, yeast 1 gm. yeast After 2 hours. After 4 hours. After 6 hours. After 8 hours. After 10 hours. After 12 hours. After 14 hours. After 16 hours. After 18 hours. After 20 hours. After 22 hours, After 24 hours, After 2 hours After 4 hours After 6 hours After 8 hours After 10 hours After 12 hours After 14 hours After 16 hours After 18 hours After 20 hours After 22 hours After 24 hours Averages . . . 60 96 12 13 6.08 7.00 7.86 8.67 9.47 10.18 10.80 11.45 80 64 30 96 59 4.18 4.75 5.22 5.76 6.20 6.62 7.01 0.25 0.76 1.20 1.61 2.03 2.46 2.84 3.21 3.57 3.92 4 22 4.53 0.03 0.28 0.48 0.74 0.99 19 40 60 80 97 12 2.29 1.61 3.32 4.64 5.92 7.03 7.75 8.35 8.80 9.12 9.32 9.45 9.57 1.09 1.85 2.58 .26 .92 .54 .10 .64 6.12 6.58 7.01 7.42 0.46 0.76 0.98 1.18 1.38 1.56 1.73 1.92 2.09 2.24 2.39 2.54 Fermentation constants 0.0177 0.0170 0.0163 0.0157 0.0154 0.0163 0.0151 0.0150 0. 015110 0.01500 0.0148l0 0.0149 0 008710, 0092 0 0087|0 0085 0 00840 0083 0 0082 0081 0080 0078 0077 0076 0028 0039 0044 0044 0045 0046 0046 0046 0046 0045 0045 0045 0002 0014 0016 0020 0021 0022 0022 0022 0022 0022 0021 0021 0177 0193 0187 0187 0183 0183 0163 0154 0143 0132 0123 0114 0118 0102 0104 0094 0092 0091 0089 0087 0086 0085 0083 0082 .0045 .0041. .0035 .0032 .0030 .0029 .0028 .0027 .0026 .0025 .0025 .0024 0.0157 0.0083 0.0043 0.0021 0.0182 0.0093 0.0031 follow the laws of enzyme action. There might be a possible deviation due to the time it takes the substrate to diffuse into the cell, and the products to diffuse out. 116 PHYSIOLOGY OF BACTERIA Slater and Sand (1910) have shown it to be very unlikely that conditions could be experimentally realized under which diffusion becomes a controlling factor of the rate of fermentation. The influence of the concentration of the substrate upon enzyme activity is rather complex as may be seen from the discussion by Waldschmidt-Leit z ( 1 929) . There is no need of going into the details of this discussion because it concerns largely the rate at very low concen- trations of the substrate, while in experiments with rates of fermentation, we are dealing mostly with fairly high concentrations. Under these conditions, the rate of enzyme action is almost independent of the substrate concentration. The enzyme combines rapidly with the substrate, and this compound decomposes but slowly; therefore, the concentration of the enzyme-substrate compound determines the rate of products formed. As long as there is enough substrate available to combine readily with any enzyme molecule that is set free, the concentration of the enzyme-substrate compound will be constant, and consequently the rate of enzyme action is constant. Nelson and Vosburgh (1917) found the rate of invertase action constant if the sucrose concentration was above 5%. Northrup (1923) found the rate of trypsin action upon casein constant if the casein concentration was above 2%. The rate of alcoholic fermentation has been found to be independent of the sugar concentration by Dumas, as early as 1874, and this result has been verified by a number of investigators (Literature, see Slator, 1906). More accurate measurements were made by Slator (1906) who found the following relative rates of fermentation at 30°C. by the same amount of yeast: ENERGY SUPPLY OF THE CELL 117 Table 19. •Influence of Sugar Concentration upon the Rate of Alcoholic Fermentation Sugar con- centration, per cent Cm. mercury- column in 10 minutes Sugar con- centration, per cent Cm. mercury column in 10 minutes 0.16 2.9 4.05 4.7 5.1 4.8 5.2 4.0 5 5 0.28..... 5.0 5.4 0.52 8.0 5.05 0.66 12.0 5.05 1.00 20.0 4.4 2 00. . . . There is a certain degree of proportionality between the rate of fermentation and the sugar concentration at ai atm. /tressurt o.ai.% ofiS% ajt Q/5%A/aA/0a. Fig. 12. — Influence of the con- centration of oxygen upon the rate of oxidation of Nitrosomonas. Fig. 13. — Influence of the concen- tration of nitrite upon the rate of oxidation of Nitrobacter. when the concentration is low. Above 1.5%, the rate is constant and it decreases very slowly when the con- centration goes above 5%. ^ Very nearly the same result was obtained by Meyerhof (1917) with a culture of Nitrosomonas by varying the oxygen concentration. The rate of oxidation was con- stant above 0.5 atmospheres of air pressure while below 118 PHYSIOLOGY OF BACTERIA 0.3 atm., it seemed nearly proportional to the oxygen concentration. The results are shown graphically in Fig. 12. The influence of the nitrate concentration upon the rate of oxidation by Nitrobacter, measured by the same author, is shown in Fig. 13. The curves are nearly alike, and agree well with those from Slator's experiments. A good verification of the same principle can also be found in Table 15. The sugar concentration was varied while the yeast concentra- tion was constant. The amount of sugar fermented was the same in all concentrations until so much had been fermented that less than 1.5% of sugar remained. This moment is indicated in the table by a cross line. No corresponding data for other fermentations have been found in the literature by the author. The unpublished experiments by Peltier which are mentioned in more detail on p. 207 bring out the same independence as regards the lactic fermentation. Peltier studied the lactic fermentation in milk diluted with varying amounts of water, the highest concentration being twice that of normal milk. From the number of cells and the amounts of acid formed, he com- puted the fermenting capacity per cell. The following data were obtained: Milk concentration 2 1 0.5 2.5% 0.25 1.25% 0.125 0.6% 0.062 Lactose concentration 10% 5% 0.3% „ ^. .^ / 12-24 hrs Fermenting capacity < 24-36 hrs Average 4.7 2.7 3.7 5.4 3.7 4.6 4.2 4.0 5.1 6.5 2.4 4.5 1.9 2.8 2.4 1.6 2.5 2.1 The result is the same as in the alcoholic fermentation: the rate of fermentation is fairly constant above 1 % of lactose, but decreases as the sugar concentration falls below 1%. SUMMARY The rate of fermentation depends upon the concentra- tion of the substrate only if this concentration is low. ENERGY SUPPLY OF THE CELL 119 In solutions containing more than 1 % of sugar, the rate of alcohoHc and of lactic fermentation is practically independent of the sugar concentration. The rate of oxidation of ammonia by Nitrosomonas is also constant if the oxygen pressure is more than one-half of that of the atmosphere, and the rate of nitrite oxidation by Nitro- bacter is constant above 0.1% nitrite. (/) INFLUENCE OF TEMPERATURE ON THE RATE OF FERMENTATION 1. INFLUENCE OF TEMPERATURE UPON CHEMICAL PROCESSES Those chemical processes of which the rate is measura- bly slow, are accelerated by an increase in temperature. The rate of the processes is increased, and the rate of increase is called the temperature coefficient, usually designated by the letter Q. It is obtained by dividing the rates of reaction measured at different temperatures, Ki&t T, andjK:2 at T -{-AT: O - — ' For most chemical processes, at temperatures of biological activity, this coefficient for 10°C. increase, Qio, lies between 2 and 3. van t'Hoff formulated this temperature relation where Ti and T^o are absolute temperatures (degrees Centigrade +273). This shows that the temperature coefficient is an expo- nential function, and this must be taken into consideration when Qio is computed from any temperature intervals other than 10°C. The following reasoning will demonstrate the principle: Assumed that the rate of a process is increased by the coefficient Qi for 1°C. increase. 120 PHYSIOLOGY OF BACTERIA At20°C rate = A^o At 21°C rate = KoQi At 22°C rate = (i^oQi)Qi = KoQi^ At 23°C rate = KoQi' At 30°C rate = KoQi^^ The temperature coefficient for 10°C. increase equals the temperature coefficient for 1°C. increase raised to the 10th power. Qio = Oi 10 For any temperature interval AT, Qio can be found from the formula AT «■« = yji^) 10 The van t'Koff formula shows that the temperature coefficient is not really constant. The expression ^1- To T,To will be different at different temperatures. If Qio = 2.0 at 20 to 30°C., it must change as follows: Qio between 0 and 10°C. = 2.218; A = 6153; /x = 12306 20 and 30°C. = 2.000; A = 6153; m = 12306 40 and 50°C. = 1.838; A = 6153; ju = 12306 90 and 100°C. = 1.575; A = 6153; /x = 12306 In fact, at the absolute zero, at — 273°C., Qio = infinite, and at very high temperatures, Qio approaches 1. The van t'Hoff equation may be written in the form of a differential equation: d\nK fi_ dT 2T2 Upon integration between the limits Ti and T2, the van t'Hoff formula is obtained, with fi = 2A. The general integration yields the equation lnX= -^ + C ENERGY SUPPLY OF THE CELL 121 which means that the rate is a logarithmic function of the reciprocal of temperature, or the logarithm of the rate is a linear function of the reciprocal of temperature. Fig. 14 shows the rate of alcoholic fermentation to be a simple logarithmic function of -^^ according to data by Aberson (1903). 2.*^ } X •^l ^ y^ ^/r li' /f*° IV' UO' I ar 2f z%i' X Fig. 14. — Relation between the rate of alcoholic fermentation and the recipro- cal of the absolute temperature. In biological literature, the statement is frequently- found that biochemical processes of organisms, such as motility, respiration, growth, etc., follow the same temperature laws as chemical processes. This state- ment has its very great limitations as the following reasoning will readily show: Strept. thermophilus will 122 PHYSIOLOGY OF BACTERIA produce measurable amounts of acid at +10°C., but not at 0°C. The temperature coefficient of fermentation at this point equals the rate at 10°C. divided by zero, which is infinite. The same organism will produce acid at 45°C., but no more at 55°C. The temperature coefficient at this point is zero divided by the rate at 45°C. which is zero. Thus, in the very narrow range between 0 and 55°C., the temperature coefficient of fermentation varies from infinite to zero. It is unavoid- able that in this range, there will be a point where it is between 2 and 3 as in normal chemical reactions, but this enormous range is anything but a resemblance to that found in chemical processes. That the temperature coefficient of fermentation really increases with decreasing temperature, is shown by Table 20 and Fig. 15 representing the data obtained by Slator (1906) with different yeasts. The curve for Table 20.- TATION —Temperature Coefficients of Alcoholic Fermen- BY Different Yeasts at Different Temperatures 5°C. temperature coefficients ^^^^7 tilUng y^^^* yeast Wine yeast Average Average Qio 5-10°C 10-15°C 2.65 2.11 1.80 1.57 1.43 1.35 1.20 2.50 1.97 1.98 1.62 1.47 1.36 1.26 2.30 1.85 1.96 1.62 1.41 1.33 1.24 2.50 1.98 1.91 1.60 1.44 1.35 1.23 6.25 3.92 15-20*0 3.64 20-25°C 2.56 25-30°C 2.07 30-35''C 1.81 35-40°C 1.51 Qio shows that extrapolation to lower temperatures must lead to infinity while extrapolation to higher temperatures must lead to zero. A similar situation ENERGY SUPPLY OF THE CELL 123 will be found in the relations between growth and temperature. Fermentation is a chemical process, however, and if it does not follow the temperature laws of chemical processes, the reason for it must be found. It seems % \ s- ^ \ 3 o\ \ 2 \ / ^%^ / 0° 1 r z O'* z 5- 3 0"" 3 5" Fig. 15.- -Variation of the temperature coefficient of alcoholic fermentation with temperature. simplest to assume that the general law really applies here as well, but that above the optimum temperature and near the minimum temperature, a superposition by other processes causes the observed deviations from the rule. 124 PHYSIOLOGY OF BACTERIA 2. INFLUENCE OF TEMPERATURE UPON ENZYME ACTION The cause of the deviation at high temperatures is known. It has been expressed in physico-chemical terms by Tammann (1895) who studied the temperature relations of the action of emulsion upon salicin. He pointed out that increase of temperature not only increases the rate of enzyme action, but also, and to a much greater degree, the enzyme deterioration. The temperature coefficients are 1.26 for enzyme action and 6.36 for enzyme deterioration. The consequence is that at high temperatures, we have a more rapid decomposition of salicin as long as the enzyme is not destroyed. But with increasing tempera- Table 21.— Percentage op Active Enzymes at Different Times AND Temperatures Temperature. . 25''C. 35''C. 45°C. 55°C. 65°C. 75<'C. ^66° 1.84 20,000 1.84 20,000 1.84 20,000 1.84 20,000 1.84 20,000 1.84 B 20,000 t = H hrs. . . . 99.92 99.28 94.74 69.18 10.67 t = 1 min. 66.70 1 hrs. . . . 99.84 98.61 89.76 47.85 1.14 2 min. 44 . 49 2hr8.... 99.68 97.24 80.58 22.90 0.01 5 mm. 13.20 3 hrs. . . . 99.52 95.89 72.33 10.96 0 10 min. 1.74 4 hrs. . . . 99.36 94.56 64.92 5.24 20 min. 0.03 5 hrs. . . . 99.20 93.04 58.25 2.51 10 hrs. . . . 98.42 86.96 33.96 0.0006 20 hrs.... 96.86 75.62 11.54 0 30 hrs.... 95.33 65.75 3.92 40 hrs.... 93.82 57.18 1.33 50 hrs. . . . 92.33 48.59 0.45 tures, this destruction will occur more and more rapidly as shown in Table 21 until at 65°C., 90% of the ENERGY SUPPLY OF THE CELL 125 enzyme is destroyed in half an hour, and at 75°C.; 98% is destroyed in ten minutes. The quicker rate of action, combined with the more rapid rate of destruction, pro- duces the result shown in Table 22 and in Fig. 16 which is a plaster model of the first five hours of Table 22. Table 22. — Amount of Salicin Decomposed by Emulsin at Various Times and Temperatures, in Percentages Amounts calculated Hours 25°C. 35°C. 45°C. 55°C. 65°C. 75°C 1. 2. 3. 4. 5. 10. 20. 30. 40. 50. 6.91 13.41 25 . 02 35.08 43 . 79 51.33 75 . 88 94.08 98.50 99.99 99.99 8.89 16.45 30.01 41.22 50.51 59 34 81 48 95.72 98.81 99 . 75 99.98 00 100.00 100.00 10.46 14.30 6 . 35 21.18 17.92 7.22 33 49 25.31 7.29 44 01 28.62 7.29 53.13 30.15 7.29 58.39 30.86 7.29 75 . 00 31.52 7.29 84.40 31.52 7.29 86.70 31.52 7.29 87.40 31.52 7.29 87.64 31.52 7.29 87.76 31.52 7.29 0 . 0002 0 . 0002 0.0002 0 . 0002 0.0002 0 . 0002 0.0002 0 . 0002 0 . 0002 0.0002 0.0002 0 . 0002 Hours Amounts found bv Tammann 1 13 18 11 3 32 51 16 5 58 61 19 8 65 76 68 73 21 12 23 26 91 1 84 i 28 50 98 .89 30 74 99 92 31 The physico-chemical equations are given by Tammann as follows : 12G PHYiSIOLOGY OF BACTERIA The rate of decomposition of the substrate (sahciii) is proportional to the concentration of unchanged substrate (a — x), and also pro- portional to the concentration of the enzyme, E. k ■ E(a - x) Fig. 16. — Influence of temperature and time upon enzyme action; the percentage of salicin decomposed by emulsin at different temperatures during the first 5 hours. This equation is correct for any constant temperature. If the tem- perature changes, the general law remains unchanged, only k reveals itself as a function of the temperature. This is expressed l)y the van t'Hoff rule koc T-To T-To ENERGY SUPPLY OF THE CELL 127 when ko is the reaction constant at the standard temperature To, recorded as absolute temperature. T is the experimental tempera- ture, ^ is a constant characterizing the process involved. Thus, the equation: — -Tj- = k^e E{a — x) expresses the rate of enzyme action at any temperature. For simpler calculation, we substitute A{T-To) T'To q = e k = koq —^ = koqE{a - x) This would be a simple formula if the amount of enzyme remained constant during the experiment. But this is not the case. Enzymes are unstable compounds, readily decomposed in watery solution, and, therefore, decrease during the experiment. In most cases, their deterioration follows the monomolecular reaction -I = /..(.-,) where b is the original enzyme concentration, y the amount of enzyme decomposed. Here again, K is a function of the tempera- ture, following the same law, but with a different constant. K=Koe ^^° For easier calculation, we introduce again the temperature coefficient. Q = e^ K = K,Q -^ = KoQib - y) 128 PHYSIOLOGY OF BACTERIA Integrated, this equation changes to KcQt = In ^ h - y h b - y h - y = ^KaQt KoQt b b -y = b-e-^"^^ The expression b — y is the amount of enzyme still active after the time t, and is the same as the letter E in the first equation. Table 21 gives the calculated amounts of enzyme which remain active after exposure to different temperatures for different times. The data K^^o = 1.84 and B = 20,000 are taken from the experi- mental data of Tammann. Substituting this value for E in the equation representing the rate of enzyme action, we obtain — -^ = koqE{a — x) = k^qbe- ^°^'^{a — x) Upon integration, the amount of salicin decomposed at any time and temperature is found to be = aLl e This expression has been used by Rahn (1915) for the computation of the data in Table 22 which also gives the experimental data obtained by Tammann. Tammann explains the disagreement between calculated and measured amounts at 65°C. by the protective action of the substrate upon the enzyme. The ^-values of the formula have been calculated from data obtained by heating emulsin in water. In the presence of salicin, the emulsin is more resistant. That is the reason why the calculated amounts are a little too low, even at 45°C. The principle remains unaltered by the protective action of the substrate. Computation of the temperature coefficients reveals Qio = 1.26 for enzyme action, and Q\q = 6.36 for enzyme deterioration. Raising ENERGY SUPPLY OF THE CELL 129 the temperature from 25 to 65°C., i.e. 40°C., would result in a rate of enzyme action 1.26'' = 2.52 times as rapid. The same increase in temperature would cause the rate of enzyme deterioration at 65°C. to be 6.364 = 1^540 times as high as at 25°C. Only 10.67% of the enzyme would be left intact after half an hour's exposure to 65°C. The increased rate of enzyme action is of no avail at very high temperatures because there is no longer any enzyme. At 45 and 55°C., where the enzyme destruction is not quite so rapid, the result of the increased rate of action becomes very noticeable. But as all enzyme is destroyed before all saUcin is decomposed, the reaction never becomes complete. This is accomplished only at the lowest temperatures. As a result of this interlinking of two opposite effects, the optimum temperature of enzyme action is not definite. At one-half hour, the salicin is most rapidly decomposed at 55°C.; after two hours, 45°C. is plainly the optimum; after ten hours, 35°C. shows the largest amount of decomposed salicin; and after forty hours, the lowest temperature appears to be the best. This is also shown in Fig. 16. The heavy ascending line indicates the optimal rate at different temperatures. It is seen to shift towards the lowest temperature with increasing time. This is a very characteristic property of the Tammann principle. The high temperature coefficient of deterioration is the cause of the thermolability of enzymes. For most enzymes, this coefficient is still higher than with emulsin. Arrhenius (1907) gives the following values for Qio: Table 23. — Temperature Coefficients of Enzymes Of enzyme action Of enzyme destruction Rennet. Trypsin , Pepsin . . Qio = 2.61 1.61 2.06 Qio = 65.8 17.8 33.7 130 PHYSIOLOGY OF BACTERIA The decrease of optimal enzyme action with time is very apparent also in the action of pressed yeast juice upon sugar. The following table shows two experiments Table 24. — Influence op Temperature upon Fermentation by Pressed Juice of Yeast (Total CO2 in grams) Time, days S-T-'C. 12-14"C. 22°C. 28-30°C. 35-37°C. 1 0.09 0.18 0.26 0.31 0.49 0.62 0.75 0.91 0.97 0.25 0.49 0.70 0.85 1.11 1.20 1.22 0.56 0.93 1.03 1.06 0.71 0.76 0.80 0.82 0.86 0.88 0.53 2 0 57 3 0 58 4 . . . 0 60 6 0.64 8 0 64 10. . . 14 16 1 0.06 0.13 0.19 0.29 0.39 0.48 0.58 0.21 0.42 0.57 0.65 0.71 0.41 0.59 0.62 0.63 0.63 0.44 0.47 0.47 0.49 0.49 0.34 2 0.35 3 0 36 4 0.37 6 0.39 8 10 by Buchner, Buchner and Hahn (1903) in which the change of the optimum from 30°C. during the first day to 12-14°C. after six days is very distinct. INFLUENCE OF TEMPERATURE UPON FERMENTATION BY LIVING CELLS The result of Table 24 that the optimum temperature of fermentation by pressed juice of yeast approaches 12°C. seems to be contradictory to the old established fact that living yeast ferments best near 30°C. The necessary conclusion must be drawn from this dis- ENERGY SUPPLY OF THE CELL 131 crepancy that the temperature law of enzymes does not apply to living cells. All enzjnues deteriorate, even at low temperatures. Endogenous catabolism in all living cells (p. 16) is almost certainly due in part to deterioration of enzymes in the cell. In normally nourished cells, deterioration probably takes place at the same rate and by the same processes as outside, but new enzyme is being formed continuously to keep a fairly constant enzyme concen- tration in the cell. Deterioration will proceed with greater speed at higher temperatures, and as far as enzymes are concerned, we know that their temperature coefficient of deterioration is enormously high. The production of new enzyme in the cell will also be accelerated by a temperature increase, but its temperature coefficient will be normal, i.e., low. In a fermenting liquid heated slowly, the enzymes deteriorate more rapidly, and are likewise being formed more rapidly. There must always be a temperature at which the originally very slow rate of deterioration overtakes the rate of reconstruction. Heating above this temperature must cause a gradual decrease of the enzyme content of the cell. The higher it rises, the greater will be the difference between the rate of deterioration and the rate of reconstruction, the quicker will be the decrease of enzyme in the cell. If the rate of deterioration is only slightly greater than that of reconstruction, the cell may come to a constant enzyme concentration below normal, but if the difference is considerable, the enzyme will soon be completely deteriorated. Because of the great difference in temperature coefficients, an increase of only a few degrees will change a rapidly fermenting culture into one that has lost the power of fermentation altogether. 132 PHYSIOLOGY OF BACTERIA As long as the enzyme content of the cell remains constant, the increased temperature will cause an increased rate of fermentation corresponding to the normal temperature coefficient of the enzyme action. The highest temperature at which the production of enzyme is still equal to its deterioration, is the optimum for fermentation. Beyond this point, the enzyme con- tent of the cell must gradually decrease and the result is the Tammann principle, i.e. the shifting with time of the optimum from high to lower temperatures. In fermentation by living organisms, the optimum does not shift to the lowest temperature, but to the optimum temperature at which the rate of fermentation remains constant. The Tammann principle at temperatures above the optimum can be seen in Table 25, which presents some unpublished results by the author, and shows the CO2 pressure produced by bakers' yeast in pure sugar solu- tion. The result is not very clear because the rate of fermentation increased during the first hour. But the gradual shifting of the optimum temperature from 44 "^ to 32.5°C. is quite conspicuous. Table 25. — Rates op Fermentation at Super-optimal Temperatures (mm. of mercury pressure produced by yeast in sucrose solution in 5 minutes) Average rates 29°C. 32.5°C. 35.5°C. 40°C. 44°C. From 5-30 min 27.7 32.0 34.0 32.0 31.2 33.0 33.2 39.5 42.7 40.6 42.2 41.6 38.2 40.5 50.0 47.4 48.2 31.0 51.4 57.0 56.7 50.4 50.2 11.4 55.7 From 60-70 min 56 0 From 185-200 min 55.3 From 360-385 min 44.8 From 540-565 min 33 8 After 22 hours 0 ENERGY SUPPLY OF THE CELL 133 It is also worth noting that at 29°C. and 32.5°C., the rate of fermentation remains really constant during 22 hours, and does not decrease as in the case of zymase (p. 13^). This constancy of the optimum temperature after the Tammann principle has been overcome by time, differentiates the fermentation by living cells from that of a mere enzyme action. Below the optimum, the temperature coefficient is normal. From the rates at 29''C. and 32.5°C. in the above table, we obtain Qio = 1.98. Rubner (1913) found lower values, 1.40 and 1.58, between 23°C. and 38°C. because the latter temperature is already too high. Aberson (1903) observed an average coefficient of 2.72 between 12°C. and 30°C. Slator's data ri906) have already been given in Table 20, p. 122. The assumption has been made above that the temp- erature coefficient of fermentation is that of a normal chemical reaction except for a superposition of another process near the maximum and the minimum tempera- tures. This assumption seems justified as far as supra- optimal temperatures are concerned. No good explanation has as yet been given for the existence of a minimum temperature. It is considered an established fact that with most organisms, fermenta- tion ceases at a temperature well above the freezing point. This might be a wrong conclusion, however. If no fermentation products can be found in a medium inoculated in the ordinary way, and held at low tempera- ture, it is proof only that the bacteria did not grow. The number of transferred cells is too small to produce sufficient fermentation products to be detected in the ordinary way. The minimum temperature of fermentation can be established only if very large numbers of cells are trans- 134 PHYSIOLOGY OF BACTERIA ferred to the cold medium. The author has tested the minimum temperature of fermentation of streptococci by transfering the centrifuged bacteria from 250 c.c. of a full-grown lactose broth culture to 150 c.c. of sterile milk, after both milk and bacteria had been cooled to the temperature of the experiment. The acid produced by these bacteria is given in Table 26. Table 26. — Lactic Acid Produced in Milk by Large Numbers of Cells (Strept. cremoris No. 23) Days 0 3 4 7 10 14 21 26 34 At -rc... At +5.5°C. At +10°C. 0 0 0 0.05% 0.27 0.42 0.06% 0.37 0.07% 0.55 0.08% 0.09% 0.54 0.56 0.09% 0.60 0.07% 0.59 0.08% 0.58 {Strept . cremoris No. 18) Days 0 3 6 10 17 22 31 45 59 78 At -rc At +5.5°C 0 0 0.05 0.19 0.07 0.34 0.08 0.44 0.10 0.48 0.10 0.49 0.10 0.43 0.49 0.495 0 51 {Strept. lactis) Days 0 Q 14 23 37 51 65 At -1°C 0 0 0 0.07 0.24 0.54 0.09 0.30 0.55 0.11 0.33 0.54 0.13 0.40 0.55 0.40 moldy At +5°C 0.39 At +10°C All three streptococci produced acid at -|-5.5°C. and even at — 1°C. Strept. lactis and Strept. cremoris 23 multi- plied at 5.5°C. if inoculated into milk in small numbers, at the rate of about one generation in from twenty-four to forty-eight hours; Strept. cremoris 18 did not multiply. ENERGY SUPPLY OF THE CELL 135 At — 1°C., all strains decreased slowly in plate count (see p. 224). From the first period of observation, the temperature coefficients Qio were found: Between — l^C. and Between 5.5°C, and 10°C. Strept. cremoris 23 Strept. cremoris 18 Strept. lactis 2.7 6.1 The experiment was discontinued on account of a break- down of the lowest incubator. The few experiments are not sufficient to draw conclusions regarding the cause of the minimum temperature. They suffice entirely, however, to demonstrate that fermentation will take place at temperatures considerably below the minimum for growth. Whether these results can be generalized so far as to state that fermentation takes place at all temperatures until the freezing of the medium, seems very doubtful. An explanation may perhaps develop from Crozier's (1924) analysis of the temperature relations of biological oxidations. He demonstrated graphically, by plotting rates against the reciprocal of absolute temperature (see p. 121) that at different points, different processes may dominate the rate of life functions. The critical thermal increment fj, = 2A for biological oxidations is frequently 11,500 above 15°C., and 16,100 or 16,700 at the lower temperatures. The reduction of methylene blue by Bad. coli in the presence of succinic acid shows )u = 16,700 (Quastel and Whetham, 1924). For the alcoholic fermentation, the ju-values are quite different. There is apparently no oxidation connected with alcoholic fermenta- tion. Slator's data, according to Crozier's computation, show /x = 12,250 above 21°C., and M = 22,200 below 2rC. which means a higher temperature coefficient at the lower temperature. Why this is the case, and why 21°C. is the critical temperature, is still unknown. 136 PHYSIOLOGY OF BACTERIA No other data on rates of fermentation are known which could be used for a similar analysis. SUMMARY OF FACTS The rate of enzyme action is accelerated by an increase of temperature. At high temperatures, the deterioration of the enzyme becomes also an important factor. The optimum temperature of enzyme action is nothing definite; it shifts gradually, as the experiment is pro- longed, to lower temperatures. This holds true for all isolated enzymes, including yeast juice. For living cells, this holds true only above the optimum temperature. At and below this point, the rate of fermentation is constant. Fermentation will take place at very low temperatures where growth has ceased altogether. Large numbers of cells must be used to prove this. With some organisms, e.g., with thermophilic bacteria, it does not seem probable that fermentation will occur at very low temperatures, though no experiments have been made. SUMMARY OF THEORIES The temperature coefficient of the rate of enzyme action is normal. It is independent of that of enzyme deterioration. The temperature coefficient of enzyme deterioration is much higher than normal. The simultaneous action of these two processes brings about the complex of symptoms briefly described as '^Tammann principle," i.e., high initial rates of enzyme action at very high temperatures rapidly decrease with time and soon come to an end; slower rates at medium tem- peratures, lasting longer, decrease more slowly, but give ENERGY SUPPLY OF THE CELL 137 an incomplete decomposition; quite slow rates at low temperatures decrease but very little, and bring about a practically complete decomposition. In living cells, the deterioration of the enzyme in the cell is counteracted by the cell's faculty to produce new enzyme. The normal cell works with a constant enzyme concentration. In such a cell, an increase of temperature will increase the rate of fermentation, without any later decrease, until a point is reached where the enzyme deteriorates more rapidly than the cell can replace it. Then, the Tammann principle becomes evident. The existence of a true minimum temperature, i.e., a temperature above the freezing point of the medium below which no fermentation can take place, has not as yet been proved. Fermentation will continue at temperatures considerably below the growth minimum. (g) CHEMICAL STIMULATION AND INHIBITION OF FERMENTATION It is quite generally claimed and has been proved in many instances that some poisons in very small quantities act as stimulants to certain life functions. In 1915, Rahn offered an explanation for this by an analogy between the actions of temperature and of chemical agents. In the case of enzyme action, we may assume that the toxic com- pound acts as a catalyst accelerating the action of an enzyme as well as enzyme deterioration. Enzyme action (see p. 126) may be expressed as -w = *^(" - ^) and under the influence of the catalyst whose concentration may be given as c, the acceleration could be represented by c« where n may 138 PHYSIOLOGY OF BACTERIA vary for each enzyme and each chemical compound, but is constant under a given set of conditions. Thus we get -^ = k-E(a-xKl-\-c^) At the same time, the deterioration of the enzyme is increased by the catalyst at the rate of c"». The deterioration is given by the equation -I = ^'(^ - ^) and, in the presence of the poison, or catalyst, -f = k'ib -y){l + c") b — y is the amount of enzyme present at any given time, it is the same as the expression E in the first equation. By substituting the value for E from the last equation in the second, we have a mathe- matical formulation which will essentially work out like the Tammann principle. There appears a false optimum, shifting with time from higher to lower concentrations, and an increased rate of reaction immediately after the poison begins to act, followed by a more rapid destruction of the enzyme, which, ultimately, leads to complete inactivation. No data sufficiently complete to try the mathematical treatment are known, but the stimulation of enzyme action by poisons is known and the shifting of the optimum with time from the higher to the lower concen- trations of the poison has also been established in some cases. A very striking one is the addition of arsenate to zymase (Buchner, Buchner and Hahn). The respiration of mold mycelium seems to be readily stimulated by chemicals. Watterson (1904) observed the C02-production by Aspergillus niger to rise from 439 to 868.5 mg. and from 885.4 to 1,138.2 mg. by addition of 0.004% of zinc sulfate to the medium. Gustafson ENERGY SUPPLY OF THE CELL 139 Table 27. — Influence op Arsenic upon the Fermentation by Zymase (Total CO2 in grams) Experiment No ^ r II III Arsen added 0 2% 0 2% 0 2% After 16 hours 0.50 0.73 1.15 1.73 0.89 0.95 0.99 1.01 0.16 0.24 0.42 0.73 0.35 0.37 0.38 0.41 0.16 0.24 0.42 0.73 0.34 After 24 hours After 40 hours After 64 hours 0.36 0.36 0 38 (1920) found the rate of respiration of Penicillium to be constant between pH 8 and 4. If the acidity was increased beyond pH 4, the rate of respiration increased at once, but dropped soon below normal. As the acidity increases, the time to drop below normal becomes shorter and shorter. Though Gustafson gives only curves and no data, there is an indication of the Tam- mann principle. A. Koch (1912) found decided stimulation of the yeast fermentation by ether. The Tammann principle could not be established with certainty. In some experiments, the flasks with the largest doses of ether gradually gained in fermentation, and finally fermented most rapidly. Koch believed that the yeast became gradually accustomed to the ether, but it may well be that part of the ether had evaporated, and thus the concentration of the stimulant approached the optimum concentration. Stimulation of the rate of alcoholic fermentation by using metal vats instead of wooden ones has been claimed repeatedly. Zikes (1913) found that the CO2 produced by 1,000,000 cells was 0.748 in aluminum vessels against 0.700 in glass vessels. 140 PHYSIOLOGY OF BACTERIA A large number of experiments on stimulation of alcoholic fermentation by poisons has been published by Branham (1929) . These data do not show the Tammann principle. Perhaps the time intervals chosen in these experiments were not appropriate to bring forth this effect. SUMMARY The action of enzymes, zymases and living cells can be stimulated by certain poisons, i.e., by compounds which, in slightly higher concentrations, would be distinctly toxic. In some instances, a shifting of the optimum concen- tration for stimulation with time is quite evident, as in the Tammann principle. It seems possible to explain chemical stimulation on the same basis as temperature effects, assuming the poison to act as a catalyst for enzyme action as well as for enzyme deterioration. VI. THE ENDPOINT OF FERMENTATION (a) CHEMICAL CONSIDERATIONS All fermentation in a given volume of fermentable solution will ultimately come to an end. This endpoint is an important and usually quite definite property of a culture. It is caused by one of two possibilities; either the fermentable substrate is exhausted, or the accumula- tion of products prevents any further enzyme action (or possibly enzyme regeneration). In most fermentations, we can bring about at will the first or second type of endpoint, by varying the concentration of the ferment- able material. It still remains to be shown what happens in the case where all fermentation products are volatile, as in complete oxidation of lactic acid by Mycoderma. ENERGY SUPPLY OF THE CELL 141 In the first case, if fermentation ceases through lack of fermentable material, there is not much to be said concerning the endpoint. It will be in proportion to the amount of fermentable material present, and we can easily compute the maximum amount of fermentation products if we know the initial concentration of the substrate. We know, that from 1 gm. of sugar, we cannot possibly expect more than 0.49 gm. CO2 in an alcoholic fermentation. The few experiments in dilute solutions with which the writer is familiar, indicate that the amount of CO2 or alcohol is considerably less than should be expected, and this is probably due to the utilization of a comparatively large amount of fermentable material for cell construction in a poor medium. This may not be the case, however, if an abundance of non-fermentable construction material is supplied while the fermentable material is scarce. More important and more interesting is the question of the endpoint of fermentation when an excess of fermentable material is present, and the fermentation Table 28. — Influence of the Sugar Concentration upon the Endpoint of Fermentation of Streptococci in Peptone Solution (Numbers mean per cents of lactic acid formed) Organism Glucose concentration 0.5% 1.0% 2.0% 5.0% 10.0% 15.0% Strept. lactis 4 0.41 0.41 0.43 0.34 0.18 0.41 0.38 0.45 0.34 0.20 0.41 0.38 0.43 0.34 0.20 0.38 0.36 0.38 0.34 0.23 0.38 0.34 0.37 0.20 0.21 0.34 Strept. faecium 8 Strept. liquefaciens 1 Strept. mastitidis 3 Strept. cremoris 1 0.27 0.32 0.14 0 18 ceases before all of it is consumed. Since we are dealing with enzyme reactions it seems logical to discuss first 142 PHYSIOLOGY OF BACTERIA the general laws of the endpoint of enzyme action. A brief review will soon convince us, however, that the material on hand is of little help. Even though, on general principles, enzyme reactions are considered incomplete, the endpoint of the action of most enzymes is reached only when the decomposition is practically complete. In the case of maltase and lipase, we have true reversible action. However, in the fermentation processes, there is no reversion known. It seems quite impossible that streptococci produce dextrose from an excess of lactic acid, or that yeast produce sugar from an excess of alcohol and CO2. It is probably safe to assume that the process is irreversible. This assumption that the mass law does not enter into consideration is well substantiated by the fact that a certain strain of Strept. lactis will produce 0.4% of lactic acid in broth and no more, regardless of whether the nutrient medium contains, 2, 5, 10 or 20% of lactose (see Table 28 by Orla- Jensen, 1919). According to the mass law, an increase of substrate should always bring about an increase in the products, because the equilibrium would require the coefficient: '7 -^ — 7-^ or f -T — rjn to be constant. It is not necessary, there- fore, to consider the inhibition of the alcoholic fermenta- tion by alcohol from any different chemical view point than the inhibition by phenol or other disinfectants. The proof that a fermentation is stopped by the accumulation of its products has been given in some instances by the observation that fermentation starts anew if the harmful products are removed. It is generally known that acid forming organisms will pro- duce more acid if the medium is neutralized. ENERGY SUPPLY OF THE CELL 143 wo o o © o T T T T o o o o I— 1 tH 1—1 rH X X X X CO OO CO 00 t> rJ4 M Tt< d d d ^ 2 1 1 1 CO CO O d 6 S I— 1 1 02 bO oooooooo oooooooo XXXXXXXX 1-ioddddoo g ^ a)05iocooot^(Mco T*OSTjHCO CO CO o ooooooooo o a u 1 TtlTjHTtlCOfNCOOOOO'* 1 i000e0000b«0"5 O500COC0COt>Tt<00l>- (£)COCOcOcOtJ1COC0(N 1 ooooooooo :5 £ -^ -J3 COOOCOCOOOOOOOO 05 d o O 00 .-H • d 0 + : : : 1 CO : : : d 00 • • d o lO • CO 00 • 00 0^ to T-H CO • 00 t^ • 1 05 O • ci CO • S5 CO 00 1—1 ! • o «0 154 PHYSIOLOGY OF BACTEKIA cooling to — 1°C. does not seem to injure them, and they produce lactic acid at a slow rate, as would be expected, but still at a rate reasonable for such a low temperature. But the rate of fermentation soon begins to decrease; something seems to interfere. It can- not be the fermentation products because only 0.3% have been formed when the fermentation ceases alto- gether. Nor are other products interfering for fermen- tation begins again at higher temperatures and reaches a normal endpoint. It must be that some change in the cell prevents fermentation. This may be the deterioration of the fermenting enzjnne which cannot be regenerated at temperatures below the minimum of growth. Above this temperature, the enzyme is grad- ually reconstructed, and normal fermentation starts again. Some doubt in this explanation is justified. The endpoints of growing cultures, or of heavily seeded cultures just above the growth- minimum, seem to have a lower endpoint than at higher temperatures while according to the above assumption, it should be at least as high or higher. This is shown by an experiment with Strept. lactis. Four milk cultures were kept at 30°C. until they were just beginning to coagu- late. Then, they were cooled rapidly in water, one to — 1°C., one to 5°C., one to 10°C., while the last one remained at 30°C. They required 5.5 c.c, 5.8 c.c, 5.7 c.c. and 6.1 c.c. of n/10 normal KOH for neutralization, immediately after cooling. Fourteen days later, the neutralization required 6.1 c.c, 6.8 c.c, 8.0 c.c, and 8.8 c.c. respectively. Later titrations showed no increase in acidity. This strain of Strept. lactis grew distinctly at 5°C. in cultures with small seeding. The low minimum of 6.8 c.c n/10 KOH at 5°C. against 8 cc at 10°C. and 8.8 c.c at 30°C. cannot be explained simply by lack of reconstruction power of the bacteria. SUMMARY OF FACTS At temperatures well above the growth-minimum, the final amount of fermentation products decreases with ENERGY SUPPLY OF THE CELL 155 increasing temperature. Above the growth-optimum, the drop of endpoint with rising temperature is very rapid. At temperatures near the minimum, the final concen- tration decreases with decreasing temperature. At temperatures below the growth-minimum, the endpoints are quite low. SUMMARY OF THEORIES The decrease of the endpoint by an increase in tem- perature is well accounted for by the assumption that the products which inhibit further fermentation are more harmful to the cells or to the enzymes in the cells at higher temperatures. This is in accordance with the action of disinfectants. The explanation of the decrease of the endpoint below the minimum as due to deterioration of the enzymes and lack of formation of new enzymes, is not entirely satisfactory, as it leaves some facts unexplained. VII. CHANGES IN TYPE OF FERMENTATION (a) CHANGES DUE TO CHEMICAL COMPOSITION OF MEDIUM Many microorganisms, perhaps all except the strictly prototrophic species, can use more than one kind of chemical compounds for the production of the necessary energy. Quite often, an organism will produce different fermentations in different foods. Yeast growing in a sugar-free medium will form products different from those in a sugar-containing medium. In solutions of dextrose or fructose or maltose or sucrose, however, a beer yeast will form the same endproducts from different 156 PHYSIOLOGY OF BACTERIA substrates. These substrates are similar, but not iden- tical. Other microorganisms produce different fermen- tations in different sugars: mannite bacteria reduce fructose, but not glucose, to mannitol; Leuconostoc produces slime only in sucrose solutions; the Pneumococ- cus group ferments different sugars to different products. On the other hand, the same substrate may be fer- mented by the same organism to different products, owing to outside conditions. Some of these are due to the chemical composition of the medium. The best known example is that of the different fermentations by yeast (p. 50), in a normally acid medium, in an alkaline medium, and in a medium containing Na2S203. The change of the gas ratio of Bad. coli by the presence of oxygen (p. 64) might also be included here. A change of acidity effects some other fermentations as well, as might be seen from the importance of pH control in the acetone fermentation (Arzberger, Peterson and Fred, 1920). (6) CHANGES DUE TO TEMPERATURE It has been shown before (p. 151) that the endpoint of fermentation is greatly influenced by the temperature. Pronounced changes of the type of fermentation by temperature are not known, but the by-products of fermentation are influenced. It is well known that the ripening at high temperatures of the cream for churning gives a sharp, undesirable flavor. Bioletti (1908) stated that the quality of California wines could be greatly improved if the wine manufacturers would employ lower temperatures. Sarkaria and Hammer (1928) observed distinct but not very pronounced changes in the type of metabolism of proteolytic bacteria if they were cultivated at different temperatures. ENERGY SUPPLY OF THE CELL 157 Changes of type might be brought about by the fact that the same organism has different minimum tem- peratures for different substrates. Monilia sitophila grows well with citric acid at higher temperatures, but not at 15°C. or lower, where it still thrives with dextrose (Went, 1901). Similarly, Penicillium, according to Thiele (1896), cannot grow with dextrose if the tempera- ture rises above 32°C., but can still live there with glycerol, while at lower temperatures, sugar is by far the better food. For Aspergillus, the same author found the minimum temperature to be 6°C.-8°C. if grown on dextrose, but it was 10°C.-12°C. if grown on formic acid salts. (c) INFLUENCE OF CONCENTRATION OF SUBSTRATE A rather unexpected experience is the more recent observation that the same substrate at different con- centrations yields different products, or at least different ratios of the products. The first experiment carried out on a large scale is that of Mendel (1911), who studied the products formed by various gas producing bacteria from dextrose, maltose, lactose and sucrose in concen- CO trations from 1 to 20%. He found that the -Tr~ ratio -tl2 was in some cases very distinctly influenced by the concentration. These data, interesting as they are, cannot be taken at full value because the gas ratios CO2 : H2 contradict all other experiences. The hydrogen and carbon dioxide headings of the tables probably became confused in printing. But whatever it be, there is some influence of concentration upon the gas ratio in some instances, and upon the ratio between volatile and non-volatile acids as well, which deserve attention. 158 PHYSIOLOGY OF BACTERIA A much clearer case is that of the citric acid fer- mentation by Aspergillus niger. This Aspergillus forms oxalic acid only when the sugar concentration is low, and citric acid besides oxalic acid when the sugar con- centration is high. Currie (1917) who was the first to make this observation, gave, along with others, the data compiled in the following table. The best medium for a good citric acid formation is one high in sugar and low in nitrogen. Table 37. — Formation of Oxalic and Citric Acid by Aspergillus Niger 1,000 c.c. of medium contain Products after 7-8 days Sucrose, gm. Nitrogen, gm. KH2PO4, gm. MgS04, gm. CO2 Oxalic acid Citric acid Mycelium 50 l.SNaNOs 1.0 0.2 25.08 9.82 2.44 7.44 50 2.0NaNO3 1.0 0.2 27.28 12.09 0.88 10.40 50 S.ONaNOa 1.0 0.5 31.03 12.80 0.00 12.82 50 1.2 NH4NO3 1.0 0.2 25.44 4.33 6.08 13.52 50 2.0 Asparagin 0.2 10.51 3.48 8.23 100 S.ONaNOs 1.0 0.25 39.51 8.26 11.62 15.50 150 2.5NH«N03 1.0 0.25 24.12 0.19 70.00 19.38 In butyric acid fermentation, a gradual change takes place with time. Perdrix found that the formation of acetic acid ceases after five days, while butyric acid continues to be formed in large quantities. (d) SELECTION OF FOODS The selection of foods by microorganisms must be discussed in two entirely separate chapters, one dealing with food for energy purposes, and the other dealing with food as construction material. ENERGY SUPPLY OF THE CELL 159 The study of food selection with microorganisms was started by Pasteur's observation (1860) that Penicillium used d-tartrates in preference to the 1-form. Duclaux (1889) and Pfeffer (1895) studied the choice which Aspergillus and Penicillium made from different sources of energy offered, such as salts of organic acids, glycerol and sugars. They found that, as a rule, both foods were attacked, but usually the one more completely than the other. Acetates were used more readily than either butyrates or tartrates. Dextrose was used in preference to glycerol. Pfeffer found that in a 4.5% peptone solution, glycerol was not touched by Aspergillus. A very highly specialized case of food selection is that of Penicillium palitans which, according to Stokoe (1928) destroys practically all the caprylic acid of coconut oil while comparatively little of the other fatty acids had disappeared. The caprylic acid is oxidized at the /3-carbon as usual, but then, instead of complete breakage of the carbon chain at this point, the keto-acid is decar- boxylized leaving amyl methyl ketone which is the cause of the rancid odor of coconut oil. C5HnCH2-CH2-COOH + 02 = CsHn-CO-CHg-COOH = Caprylic acid Keto acid CsHnCOCHs + CO2 Amyl methyl ketone While molds have been used very extensively for such experiments, we know a good deal less about the pref- erences of bacteria with regard to their sources of energy. The most typical and best known case is the protection of proteins by sugar. There is much less protein decom- position in media containing sugar (Literature and explanations, see Berman and Rettger, 1918b). Most of these facts lack a good explanation. It does not seem that the available energy of a food compound determines the preference. Some compounds are simply 160 PHYSIOLOGY OF BACTERIA ''less digestible," though this term is rather meaningless with bacteria acting upon dissolved food. The so-called selection of foods is probably linked closely with the molecular structure of the cell, and a comprehensive quantitative study of food selection by bacteria might give us a clearer conception of the cell mechanism. (e) SUMMARY OF FACTS The type of fermentation produced by a certain species of bacteria depends above all upon the substrate. Similar substrates, e.g., different sugars, may give the same products with one species, and may give entirely different products with another species. The same substrate may be decomposed by the same organism in different ways, which depend upon the environment. Such differences in type of fermentation may be brought about by changes in acidity. Yeast in alkaline solution forms large amounts of glycerol. Clostr. aceto-ethylicum produces more volatile acids, more alcohol and less acetone if the medium becomes alkaline. Temperature changes cause variations in the by-prod- ucts, but not in the main type of fermentation. Differences in the concentration of the substrate may bring about a different type of fermentation, as in the citric acid fermentation by Aspergillus in sugar con- centrations above 10%. SUMMARY OF THEORIES The basis for many of the possible different types of fermentation has already been given extensively in the chapter on Equations of Fermentation. Why one type changes into another, can not usually be explained. It is understood only in the case where ENERGY SUPPLY OF THE CELL 161 the change is brought about by a change in acidity. But how a change in concentration, or in temperature, can bring about a different type of fermentation, is not easily conceived. There remains always the assumption that inter- mediate products are excreted under special conditions which ordinarily would be decomposed further. Here as well as in the general problems of food selec- tion, we are probably dealing with phenomena relating to the chemical structure of the bacterial cell which is practically unexplored. PART C GROWTH I. GENERAL CONSIDERATIONS (a) MEASUREMENT OF GROWTH Growth, here, shall be understood to mean the total increase of the cell substance of living matter. No distinction will be attempted between the two processes comprised in the increase in size of one cell, and the splitting of this enlarged cell into two smaller cells. It would clear up matters if such distinction could be made, but it cannot, as the factors concerned in the process of cell division have not, as yet, been approached by chemical analysis. All we may do is measure growth by the increase in total weight or volume of living matter, or by counting the increase in the number of cells, with the assumption that the newly grown cells weigh as much in the average as the ones at the start. We know that the young and old cells differ in volume (Henrici, 1928), but we know nothing about their moisture contents. Other indirect methods have been used to measure growth. Rubner (1906a) determined the nitrogen con- tent of the bacteria as a measure of the '^crop,'' and in one experiment, even the sulphur content of the cells was taken as the measure. Turbidity has been used to estimate the number of bacteria (Strauss, 1929), and so has the volume of the cells obtained by centrifugation (Richards, 1928). None of these methods gives an absolutely correct conception of the increase of living matter. Nor can any method be suggested 162 GROWTH 163 which would give a true measure of living matter because we do not know enough about it. Cells may increase in weight by adsorbing more water, or by depositing fat, without increasing the living matter. They may increase in numbers without increase in weight. They may increase in nitrogen content by storing reserve substances of protein nature. In measuring growth, it should be kept in mind that all living cells undergo endogenous catabolism. Therefore, the measured growth is the difference between total growth and endogenous catabolism. The latter, however, is usually so small as to be neglected without appreciable error; but this may not be true at super-optimal temperatures. Growth of cells requires material for cell construction; it also requires a definite amount of energy to weld together the building material which is of a simpler nature than the living matter; finally, it requires a very intricate building plan to have all the construction done in the correct way and at the correct place. It is possible to study the materials needed for cell construction though it is rather difficult in some cases, on account of the extremely small quantities needed; the advance of this line of research depends primarily upon the improvement of the methods of analytical chemistry. It is also possible to study the sources and amounts of energy available to the cells; the entire preceding section on fermentation was devoted to this study. It does not seem possible that we shall ever understand the actual process of growing, the mechanism of the production of new protoplasmic molecules, the principles which keep order in the microcosmos of the cell, arranging all molecules in their proper places and proper sequences after they have been created. (6) BUILDING MATERIALS OF CELLS A study of growth should begin with the materials of cell construction. Unfortunately, the requirements are 164 PHYSIOLOGY OF BACTERIA different for every species of bacterium, and while some general statements may be made concerning the necessity of K, P, Ca, S, also probably of Fe, Mg and Na, and possibly of Mn and eventually even of Zn and Cu, the organic requirements are much more complex. The minerals are usually resorbed as ions, eventually as salts, and it makes little difference in what form they are given as long as they are at all soluble. Carbon and nitrogen on the other hand are capable of forming innumerable different compounds which do not ionize. A good deal of attention has been given to the finding of good, synthetic media, i.e., media of which every component is chemically well defined and known. While good growth has been obtained with some bacteria in certain media containing amino acids and sugars, the media containing peptone, meat extract, or milk are still the only standard ones. It is probable that the reason for this is not the unfitness of the food as such, but the lack of something else. It is difficult to see why bacteria should not grow in synthetic media to the same large numbers as in peptone media because we can be certain of providing the same source of energy. A systematic study considering hydrogen donators and reduction potentials (see p. 82) might lead to better results now than the efforts of 20 years ago. (c) SOURCES OF CARBON The element carbon may be used by one species of bacterium, according to Potter (1908). Carbon mon- oxide is a source of carbon for the Carboxidomonas; it oxidizes CO to CO2, and can grow in the absence of any organic matter. Carbon dioxide can be changed to cell material by all prototrophic bacteria. This group of bacteria, simpler in their metabolism than any other known GROWTH 165 organisms, includes the nitrite and nitrate bacteria, sulphur bacteria, the thiosulphate bacteria and the Hydrogenomonas group (see also p. 168). Generally speaking, these prototrophic bacteria are very similar to algae, the difference resting mainly in the substitution of the chlorophyl mechanism by a chemical method of energy supply. The amount of energy required to make organic matter from inorganic is quite large, much larger than that needed by the common saprophytes. Another compound with but one carbon atom is the formic acid. This is often stated to be a rather unsatis- factory source of carbon for growth, though it can be used by some molds and bacteria. Most experiments on the decomposition of formic acid, (and of many other compounds) have been carried on in a medium containing peptone as a ^^ source of nitrogen"; these results indicate the utilization of the acid for energy, but not for cell construction. The fitness of a com- pound as building material is not dependent upon the amount of energy needed to reduce it to cell con- stituents, but is largely determined by its chemical constitution. There is for example the carbonic acid, with three oxygen atoms coupled to the carbon. This compound can be used by a goodly number of bacteria. Most of these can use neither carbon-monoxide nor methane, though these compounds are much more easily reduced, and methane would even liberate energy if changed to carbohydrate or organic acid. The structure seems to play the essential role. Formaldehyde is a strong disinfectant, much stronger than formic acid. But there is evidence of its being used as a source of carbon. Kaserer (1906) observed that Hydrogenomonas required both hydrogen and carbon 166 PHYSIOLOGY OF BACTERIA dioxide besides oxygen for growth, and that the bacteria first reduced CO2 to HCHO which could be found to exist as such in the culture. Urea does not seem to be suitable as a source of carbon. At least, Sohngen (1909) could not cultivate urea bacteria on urea alone, while very small amounts of organic acids (10-20 mg. of ammonium malate or asparagin in 50 c.c. solution) sufficed to bring about normal development of Bad. erythrogenes and of Uroh. jackschii. Methane can be used only by the Methanomonas (Sohngen, 1906) and it is not certain that the methane as such is the building stone to be hewn. It might well be that these bacteria use CO2 as the starting point for shaping the parts of molecules for growth as do all other prototrophic bacteria. This carbon dioxide might origi- nate from methane oxidized completely by the same cells. The organisms which can use compounds with only one carbon atom must possess the ability to couple carbon atoms. It is interesting to notice that while the coupling of single carbon atoms is possible to all chlorophyl plants, but only to a very limited number of chlorophyl- free organisms, a large number of the latter are able to grow on compounds with two carbons, e.g., alcohol or acetic acid. Many molds can do this; also vinegar bacteria, some yeaSts, Oidium, all Mycodermae and many bacteria can live on acetates. It appears to be much easier to couple two carbons with two carbons than to couple one carbon with another. How this coupling is actually done we do not know with certainty, but it seems possible that we have processes similar to the formation of aldol from aldehyde: CHs-CHO + CH3CHO = CH3CH(OH)CH2CHO GROWTH 167 We must consider in the further discussion that the number of carbon atoms in most of the compounds of the cell is an even one. We find most dominantly six carbons (carbohydrates, leucin, benzene ring) but four carbons are not uncommon in bacterial metabolism (butyric, succinic and hydroxy-succinic acids) and the fats, as is well known, contain only the acids with even numbers of carbon atoms. The most common excep- tions to the dominance of even numbers of carbons in cell compounds are lactic acid and glycerol. The compounds with three carbons are very commonly used, perhaps because they give, when coupled, the carbon skeleton of the sugars. Lactic acid is sufficient for the entire carbon and energy supply of Mycoderma, Oidium and probably of many other molds and bacteria like Bact. coli. The assimilation of higher alcohols and higher organic acids is shown in two tables by Buchanan and Fulmer (II p. 431 and 438) after data from de Jong (1926). Even with sugars, we have not in all cases the proof that they are sufficient as a source of carbon, for in most experiments, an organic source of nitrogen is offered, which may supply carbon as well as nitrogen. In order to be really certain that sugar is sufficient for the carbon supply, we must give the nitrogen either in mineral form, or as an amino-derivative of sugar, or eventually as amino-derivative of the decomposition products of sugar (e.g., amino acetic acid in the case of Bact. coli, or of vinegar bacteria) . That yeasts can grow with ammonia nitrogen from sugar alone has already been shown by Pasteur. Most molds behave in the same way. The difficulty in settling the question of carbon supply really rests with the nitrogen source. If an organism cannot use ammonia, nitrate or urea, but requires an organic source of nitrogen, we can tell but little about the carbon supply. The customary conception of carbon requirements for growth as presented above has become quite questionable 168 PHYSIOLOGY OF BACTERIA by the observation that complete removal of carbon dioxide prevents the growth of all bacteria. This has been asserted by many investigators and denied by many others, but the work of Valley and Rettger (1927) leaves hardly any doubt of the fact. The role which CO2 plays is entirely unexplained. Since the amounts present in air are sufficient for normal growth, it would seem that the carbon dioxide enters in the process of cell construction rather than in energy liberation. With the latter process, larger amounts would probably be required to produce noticeable effects. The author is of the opinion that we stand here at the already opened door to an important biological principle. (d) SOURCES OF NITROGEN Elementary nitrogen is used by the three well-known groups of nitrogen-fixing bacteria, the Rhizohium group, the Azotohacters and certain Clostridium species. Less intensive nitrogen fixation has also been claimed for some other bacteria, and even for yeast (Fulmer and Christensen, 1925). All organisms mentioned can take the nitrogen in other forms. Ammonia and nitrates, also nitrites in low concentra- tions, are used by many molds and many bacteria. Considerable discussion has taken place concerning the preference of ammonium salts and of nitrates, especially for the growth of molds. The results by different investigators are in some cases contradictory. It seems quite probable that we have the superposition of another factor which was not generally recognized at the time when most of these experiments were conducted, i.e., the antagonism of certain salts, the change of hydrogen ion concentration or the oxidation-reduction potential. GROWTH 169 In comparing nitrates with ammonia, it is customary to add to the same mineral-sugar solution in one case a nitrate (KNO3 or NaNOs most commonly) and to the other solution an ammonium salt (NH4CI or (NH4)2S04). If the mineral solution had been a balanced salt solution, the addition of a potassium salt might make it slightly poisonous; if it had been unbalanced, the potassium salt might balance it more perfectly and cause a better growth. The same explanation probably accounts for the fact that some investigators find Ca(N03)2 to produce a much better growth than KNO3 while others find the opposite. It is not very probable that the ion of a soluble ionized salt has different food value because it was originally combined with a different base. Nitrates and especially ammonia are for some bacteria the only source of nitrogen, and organic nitrogen com- pounds cannot be utilized by them. In this group we find the organisms with a very simple, mineral form of fermentation, such as the nitrifying organisms, the sulphur bacteria, and, perhaps, the iron bacteria. They are the same species which can build their cells only from CO2 and not from organic compounds. Nitrate-assimi- lating bacteria are known in great numbers. The denitrifying bacteria, Ps. fluorescens, Ps. pyocyanea, B. suhtilis are able to live on nitrates and some nitrogen- free organic matter. In soil, such species are found abundantly. Some representatives of the colon group can live with nitrate as the only source of nitrogen (Kisch, 1919). Ammonia is also used commonly (e.g., by the colon- paratyphoid group, and by many soil bacteria) and, perhaps, more commonly than nitrates. It is a fairly good source of nitrogen for yeasts which refuse nitrates as a rule (Lindner, 1905). Many of the bacteria which use nitrates, can also use ammonia, though some excep- tions are mentioned. In experiments with ammonium salts, it must be kept in mind that the consumption of 170 PHYSIOLOGY OF BACTERIA ammonia will make the neutral solution acid, and Wehmer found that a Penicillium species was readily killed by prolonged growth on ammonium sulphate merely by accumulation of acid. Accumulation of alkali by nitrate-consuming organisms has never been reported harmful, perhaps on account of the formation of the organic acids. As a curiosity, B. hiltneri which can oxidize HCN to CO2, H2O and N, according to Kaserer (1907), should be mentioned. The organisms which depend upon organic matter for their energy supply, seem to prefer organic nitrogen also. The usefulness of amines for cell construction has not been studied extensively, but we know that methylamine can be used by Aspergillus and Penicillium; that the mono-amine is the best, and trimethylamine the poorest source of nitrogen, and that tetramethyl ammonium hydroxide is not suitable for cell construction, den Dooren de Jong (1927) has given the genus name Protaminohacter to a group of bacteria which were able to assimilate amines. Urea can be used by quite a large number of bacteria as a source of nitrogen. This faculty is not limited to the urea-fermenting organisms. Some yeasts and many molds can also live on urea as the only nitrogen source (Lindner, 1905). Next in simplicity of composition are the amino - derivatives. Alkaloids do not seem to be sufficient for growth, at least not for the common molds, though they are decomposed, perhaps for energy purposes only. Amino-acids such as glycol, leucin, aspartic acid etc. are good nitrogen sources for many organisms but some thrive well only if they have a better source of energy (or of carbon) such as dextrose or mannite. That the GROWTH 171 amino-acids are not always used as such, is quite evident from Ehrlich's observation that the leucin assimilation by yeast is decreased in the presence of ammonium salt. Amino-acid organisms in a strict sense of the word are microorganisms which can utilize nitrogen in no simpler form than that of an amino-acid. This would mean, chemically speaking, that these organisms cannot synthetize the group — NHCHCO — which is so com- mon in the proteins of all living beings. They must obtain this combination ready-made, but can do the rest of the synthesis themselves, eventually only with an especially good source of energy. Most commonly used for experiments of this nature is asparagin, and the good results are perhaps due to the fact that we have here not only the amino-acid group, but, at the same time, the very frequent acid-amid group preformed. If we consider the great variety of nitrogenous cleavage compounds of protoplasm, we realize at once that the synthetizing power of a cell that can grow on one amino- acid only, is very remarkable, since all other amino- acids, carbohydrates, and the cell wall are all formed from the one material. The term '^ peptone bacteria" is sometimes used and is meant to indicate that these organisms can use nothing simpler than peptones. That such organisms exist, is questionable. If we consider that even the mammals can be kept alive, and probably be made to grow with amino-acids only, and that probably all the body cells are fed with amino-acids and not with peptones, it seems quite probable that this would be true also with bacteria. If they refuse to grow on tyrosin, asparagin or leucin solutions, this is no proof that they require peptone. It shows only that the proper amino-acid had not been offered. Certain species may require a mixture of two 172 PHYSIOLOGY OF BACTERIA or more amino-acids, because they cannot form one from the other. In this connection it should be mentioned that Abderhalden and Rona found no influence of the source of nitrogen upon the composition of the protein of Aspergillus niger; whether potassium nitrate, glycocoll or glutaminic acid was given, the protein of the mold gave the same cleavage products. There is no doubt, however, that the properties of yeast are greatly influ- enced by the source of nitrogen (p. 238). It is imaginable, that certain parasitic forms take large molecular complexes out of their host's native protein, and have lost the power of synthesizing these groups. (e) SOURCES OF OXYGEN In the discussion of the needs of oxygen for fermenta- tion, the requirements of oxygen for cell growth have been already mentioned for the sake of a more accurate definition of anaerobic and aerobic bacteria (p. 78). It was pointed out there that an analogy exists between the carbon, nitrogen and oxygen requirements, and that the structure of the molecules is probably quite as essential, or eventually more so than the energy relation. Carbon, nitrogen and oxygen may be compared in regard to their fitness for growth by arranging them according to the energy required to change them into organic material. Carbon Nitrogen Oxygen Completely oxidized CO2 HNO3 O3, H2O2 Partly oxidized CO HNO2 Molecular C N2 O2 Partly oxidized | HCOOH NH2CH2CO2H HCOOH Partly reduced J Completely reduced CH4 NH3 H2O GROWTH 173 The analogy is quite close between carbon and nitrogen. While the completely oxidized and the completely reduced compounds are used by some microorganisms, and the partly oxidized, partly reduced com- pounds by many others, the elementary stage is least suited for assimilation, though less energy would be required to build cells from N than from HNO3. It seems very difficult for cells to break up the affinity between the atoms of the same element. With oxygen, however, the opposite seems to be true; at least this is the case in fermentation, and we may well assume that it holds true for growth. Elementary oxygen is readily used by most living beings. Oxygen fixation is the most common biological reaction. There are exceptions, however, such as the lactic organisms and some strictly anaerobic bacteria which depend exclusively upon oxygen from organic compounds, and cannot use elementary oxygen if it is offered them. We may assume in these cases an analogy to the nitrogen and carbon sources, i.e., the inability to break up the oxygen molecule into oxygen atoms. There is no use in speaking about hydrogen sources, because hydrogen is so common in water as well as in practically all organic compounds that its source could not be traced. Attention should be called only to the hydrogen-fixing bacteria (Kaserer, 1906). (/) WASTE PRODUCTS OF CONSTRUCTION When prototrophic bacteria build their cells from carbon dioxide as the only carbon source and from nitrates as the only nitrogen source, the growth process is entirely a process of reduction and condensation. The only part of the building material that is not used is a good share of the oxygen in these compounds. This oxygen is most probably used for the oxidation of the energy furnishing substrate and cannot be traced. Other organisms build their cells from very complex material, such as amino acids, peptones, etc., and these 174 PHYSIOLOGY OF BACTERIA large molecules are not always put into the cell as such, but are split and reshaped before fitting into the cell architecture. This shaping must cause some waste products. In many cases, these waste products of cell construction may be used as food for energy; but there are some instances known where this is not the case. The only definitely established example is probably that of the fusel oils in alcoholic fermentation which is discussed in detail on p. 43. The yeast needs ammonia for building its own proteins, and takes it from leucin, leaving the rest of the molecule in the form of amyl alcohol. The main reason why more such examples are not known, is the small amount of these products. The total dry weight of the cells of Bad. coli in 1 liter of a full grown broth culture is only 220 mg. (Appendix p. 397). The waste products from the construction of 220 mg. of cells are likely to be less in quantity than the cell weight; this means less than 0.02% of the weight of the medium. It seems probable that the carbon dioxide produced by some streptococci belongs into this category of products. Hucker (1928a) mentions that Strept. lactis produces about 70 mg. CO2 per liter. The dry weight of the cells probably does not exceed 200 mg. The CO2 comes from peptone, and not from sugar; it must originate from the decarboxylation of amino acids. With other species of streptococci which produce more carbon dioxide, there seems to be a distinct protein metabolism established, and more amino-acids are split than are needed for growth. Balls and Brown (1925) and Claassen (1927) during the production of bread yeast observed a small amount of a reducing substance which was not sugar, nor was it GROWTH 175 assimilated by the yeast (see p. 211). This product amounts to about 1% of the fermented sugar, or to about 4% of the yeast solids. The small amount, in comparison to the crop, makes it appear possible that it might be a waste product of growth. (g) BALANCE SHEETS OF FEEDING EXPERIMENTS While an enormous literature on growth of micro- organisms exists, there is very little material available which shows complete balances of all foods used, or of all products, including growth. The most complete data available are those of the bread yeast industry. This is the only large industry which is interested in growth as such, and not in the products of fermentation. Even in this industry, the number of complete experi- ments is not very great. It is regrettable that the carbon dioxide and the quantity of the evaporated alcohol are measured only exceptionally. But enough material is available to get a good conception of the balance. A first attempt at such an account of all foods used and all products formed is the study of yeast metabolism by Brown and Balls (1925) who determined alcohol, CO2, decrease of sugar and amount of yeast growth in a well aerated culture. About 750 c.c. of the mash or wort (consisting of 48 gm. molasses, 1.2 gm. H2(NH4)P04 and 1.6 gm. (NH4)2S04 per liter) were kept at 28°C. and aerated with 5 liters of air per minute, or about seven times the volume of the liquid per minute. The prod- ucts formed were determined after 3, 6, 8, 16.5 and 24 hours. The results of some experiments are given in Table 38. There is a noticeable amount of sugar not used for the formation of either alcohol or CO2, and this amount, in the aerated cultures, is about one-half of the yeast 176 PHYSIOLOGY OF BACTERIA growth. We shall see later, from Claassen's data, that about one-half of the yeast solids is non-protein in nature. An interesting observation is the alcohol: CO2 ratio. This should be according to the simple fermentation formula (p. 42), 1.04. In the experiments without aeration, this ratio is a little higher, Table 38. — Balance Sheet op Yeast Growth (All data are gm. per liter) Experiment number. Aerated II III Not aerated II Duration of experiment. Total carbon dioxide formed . Total alcohol formed Alcohol plus carbon dioxide. . Sugar used Sugar minus alcohol and CO 2. Yeast growth 8 hours 11.39 8.02 19.41 23.34 3.93 5.48 24 hours 14.78 5.17 19.95 23.76 3.81 8.85 28 hours 12.96 4.83 17.79 17.96 0.17 6.73 hours 6.58=* 8.00 14.58 19.01 4.43 2.40 48 hours 7.03* 8.15 15.18 22.20 7.02 (1.21)t * These CO2 values are too low. t This is the value for 6.75 hours; no later value is given. 1.21 and 1.16, owing to an incomplete elimination of CO2 from the wort. In the aerated experiments, it is much smaller, 0.70, 0.35 and 0.37 respectively. The reason for this is either the complete oxidation of part of the sugar, or the utilization of alcohol by yeast, or both. Doubtless, some alcohol is used by the yeast. In Expt. II, all sugar had been used after the eighth hour; but during the next sixteen hours, 5 gm. of alcohol per liter disappear, 1.25 gm. CO2 are produced and the yeast crop increases 2.5 gm. In Expt. Ill, 1.2 gm. alcohol are used after the eighth hour, together with the remaining 0.3 gm. sugar, to produce 2.4 gm. of yeast and 3.0 gm. of CO 2. Special experiments showed that yeast in sugar-free aerated alcoholic solution with ammonium salts gained in weight 25 to 29 %, while the alcohol decreased and CO2 was produced. GROWTH 177 More detailed are the data of an experiment under plant conditions by Claassen (1927) with 77,000 kg. of wort which consisted of molasses, malt sprout extract Table 39. — Balance Sheet of Yeast Growth Solids Of the organic matter Total weight of products kg. Total kg. Inor- ganic kg. Or- ganic kg. Sug- ar kg. Pro- tein kg. Not pro- tein kg. Alco- hol kg. C02 kg. Composition at start Water 70,000 3,000 525 120 350 4,400 2,475 200 61 82 57 336 20 61 6 22 2,139 180 76 35 1,545 44 2 318 60 49 19 276 76 27 14 26 Molasses Malt sprouts Phosphate Seed yeast Wort of seed yeast Total 2,875 445 2,430 1,591 446 393 26 Composition after 11 houre Yeast crop . . 1,875 77,000 823 98 516 1,011 36 393 480 618 22 229 223 251 373 350 98 Fermented wort Carbon-dioxide 823 Alcohol evaporated Total 1,527 429 1,098 22 452 624 448 823 and acid phosphate. To this, after filtration and sterili- zation, the seed yeast was added with some wort. The balance sheet of Table 39 shows the composition of these materials. The air forced through the medium totaled 44,800 cubic meters, or six hundred times the volume of the^medium, = 12,800 kg. oxygen, for the entire time of growth. 178 PHYSIOLOGY OF BACTERIA The lower half of this table shows the composition of the medium and yeast crop, eleven hours later. The balance is not perfect, owing to the enormous bulk of the liquid, and to inaccuracies involved in determining the evaporated alcohol and carbon dioxide. There is an unaccounted loss of 16 kg. of inorganic matter, and of 87 kg. of organic matter, or about 4% of the total organic matter involved. A summary of all analyses made by Claassen in short intervals bears out the claim of Balls and Brown that alcohol can be used for growth by the yeast. After eight hours, the wort contains only 0.054% sugar. During the next three hours considerable growth still takes place, while the alcohol concentration decreases much faster than can be accounted for by evaporation. Table 40. — Utilization op Sugar and Alcohol during Yeast Growth 4 hours, kg. 6 hours, kg. 8 hours, kg. 11 hours, kg. CO2 formed 227 230 185 520 395 189 721 531 304 823 Alcohol formed 423 Sugar used for yeast and by-prod- ucts 329 Total sugar used 642 1,104 1,556 1,575 Yeast and by-products in % of sugar used 28.8% 17.1% 19.5% 20.9% The alcohol :C02 ratio, for this reason, is only 0.52 instead of the theoretical 1.04 for alcohol fermentations. The yeast solids (516 kg. — 82 kg. seed yeast) contain 229 — 49 kg. of protein which may derive from the nitrogenous matter of molasses and malt sprouts, and 251 — 27 = 224 kg. non-nitrogenous GROWTH 179 matter which may derive from the sugar. The subtraction of alcohol and CO 2 from the total sugar leaves 329 kg. to be accounted for by "growth and by-products." Of this amount, 224 kg. are found in the non-nitrogenous part of the yeast. The remaining 105 kg. may have been at least partly used in cell construction, but it cannot be proved. The observation that of the 1,575 kg. sugar used, at least 224 kg. = 14.2 % had been used as construction material, in addition to a simi- lar quantity of nitrogenous matter, is of considerable interest, as not nearly such large amounts have ever been found to be assimilated by bacteria. The efficiency of energy utihzation must be quite high. According to Lindner and linger (1919), yeast will produce fat from alcohol if plenty of oxygen is available. If the increase in non- nitrogenous matter were largely fat, produced from alcohol, it is evident that fairly large quantities of alcohol must be used because, according to the simplest formula 8C2H5OH -f- 0 = C16H34O2 + 7H2O Alcohol Palmitic acid it takes about 3 kg. of alcohol to produce 2 kg. of fatty acid. Both authors give special nitrogen balances for their experiments. That of Claassen is the following: Table 41. — Nitrogen Balance in Yeast Growth Nitrogen in medium : 3,000 kg. molasses with 1.67% N = 50.10 kg. N 525 kg. malt sprouts 1.70% N* = 8.93 kg. N 59.03 kg. N Nitrogen in seed yeast : 360 kg. yeast with 2.18% 7.85 kg. N 4 , 400 kg. wort of seed yeast with 0.0682 % N 3 . 00 kg. N 10.85 kg. N Total N at start .^ 69 . 88 1,875 kg. total yeast crop 1.95% N 36.56 kg. N 77,000 kg. fermented wort 0.045% N 34.65 kg. N Total N at finish 71.21 71 . 21 * Soluble nitrogen only. 180 PHYSIOLOGY OF BACTERIA There is a nitrogen gain of 1.33 kg., doubtless due to the great probable error in analysis. Of the total nitrogen offered, 59.03 kg., the new growth of the yeast contained 36.56 — 7.85 = 28.71 kg. or 48.6%. II. RELATIONS BETWEEN GROWTH AND FERMENTATION (a) EFFICIENCY OF UTILIZATION OF ENERGY Most bacteriologists hold the viewpoint that fer- mentation (as defined by Hugo Fischer, see page 24) is the source of energy which is needed by the cell for repair and for new construction, i.e., growth. The amounts of energy necessary for growth will vary con- siderably with conditions. It might seem that bacteria growing in a peptone solution would need no energy at all, because the combustion heat of the food is as high as that of the bacteria (see Table 42 from Rubner, 1904a) ; and bacteria growing on fat would even have a surplus by just converting every bit of the fat into carbohydrate and protein. Table 42. — Combustion Heats of 1 Gm. of Solids Calories Calories Peptone Gelatine Meat extract Dextrose Sucrose Fat, about Penicillium glaucum . . 5,492 4,992 3,514 3,740 3,950 9,200 4,753 Penicillium glaucum with spores Top yeast Bottom yeast Proteus vulgare Bad. prodigiosum Bad. prodigiosum, old cul- ture 5,359 4,475 4,554 4,643 4,764 4,442 This assumption has two errors. It does not take into consideration that the architecture of the bacterial GROWTH 181 protoplasm is quite different from that of the peptone molecule, and that the peptone molecule has to be torn into very small units before it can be used in the con- struction of protoplasm. The tearing down will yield energy which may not be available, and the building up will require energy. The combustion heat of the proto- plasm tells only how much energy is contained, and not how much energy was necessary to construct it. Growth processes do not work with 100% efficiency as we shall see later. The other error is made by the assumption that the combustion heat (or the balance of combustion heats in anaerobic fermentations), represents available energy. According to thermodynamic laws, the ^'free energy," i.e., the energy that might be used to do work, is not identical with the combustion heat, or differences of these. These values represent the total energy Hberated by the fermentation, but part of this goes towards the increase of entropy, and is lost for any work. This amount varies greatly with the kind of process. Let the combustion of glucose serve as example (Baas-Becking and Parks, 1927) CeHisOe + 6 O2 = 6 CO2 + 6 HoO Solid Gas Gas Liquid Under the above conditions (solid sugar; oxygen and CO 2 at pressures of 1 atm.; water of combustion in liquid form), we have Calories Heat of combustion ^ 685,800 Free energy 674,000 Entropy increase 11,800 The heat going towards the increase of entropy which cannot be used for any work, is 1.7 % of the combustion heat. A mold oxidizing one 182 PHYSIOLOGY OF BACTERIA mol of sugar has not the full benefit of the 685,800 calories, but can use only 98.3% of that amount. In the oxidation of methane to COo and H2O, the loss is 9%. While these losses towards the entropy may eventually be fairly large in inorganic processes, it seems that they are rather small in the organic reactions ordinarily met with in biological processes. Thus, the error in assuming the free energy to equal the differences of combustion heats between substrate and products is not very great, and amounts perhaps to a few per cents of the total energy liberated. For most fermentations, this per- centage is not known. A good survey on the autotrophic bacteria in their energy requirements is given by Baas- Becking and Parks (1927) and a detailed discussion of the theory is offered by Buchanan and Fulmer (1928- 30). Realizing that the energy liberated is not identical with the available energy for construction purposes, but in most cases approximately so, the next question will be that of efficiency: how much of the available energy is actually used for growth? This has been computed for some of the autotrophic bacteria by Baas- Becking and Parks (1927). The percentage of the available energy stored in the new growth is the efficiency. This efficiency seems very low. The following maximum efficiencies were found by Baas-Becking and Parks : Table 43. — Amount of Available Energy Utilized for Growth Per cent Hydrogenomonas 26 Methanomonas (maximum efficiency) 33 . 8 and 25 . 5 Methanomonas (minimum) 0.6 Nitrohacter and Nitrosohacter 6 Thiobacillus denitrificans 9 Sulfur bacteria, probably about 8 GROWTH 183 Linhart (1920) found for Azotobacter an efficiency of only about 1%, but Baas-Becking and Parks believe that 7% is more probable. Similar calculations based on the total energy liberated rather than on the "free energy" have been made by Tangl as early as 1903. One compilation of his data has already been given on p. 28. From his data on B. anthracis, we obtain an energy utiUzation of 49%. This seems very high in comparison with the above data calculated by Baas-Becking and Parks. It must be remembered, though, that the former data referred to synthesis of protoplasm from CO 2, H2O and either NH3 or HNO3 while here, the food has even a higher energy level than the cells. The energy content of 1 gm. of broth solids was 4.4 calories while 1 gm. of bacteria solids represented only 4.23 calories. It is quite evident that even with the same species of bacteria, the energy required for synthesis will vary with the kind of building material used. More data on this subject may be found in Putter's book (1911, p. 121-129). (6) EFFICIENCY OF YOUNG AND OLD CELLS All these experiments, from Tangl to Baas-Becking and Parks, disregard the time factor. It is generally- known that fermentation continues for a considerable time after the multiplication of cells has stopped. This must be of considerable influence upon the efficiency of energy utilization. If one experimenter discontinues his experiments with beer yeast after 2% of alcohol has been formed and multiplication has practically ceased, and the other waits until the fermentation is over, and 6% of alcohol has been formed, the former's efficiency will be three times that of the latter, both using the same organism in the same medium. After growth has ceased, fermentation will be needed for nothing but repair and upkeep; after completion of growth, the efficiency must be practically zero. This would be the one extreme ; we are more interested in the other extreme, i.e., the maximum efficiency. 184 PHYSIOLOGY OF BACTERIA Most probably, this would be found in the very young, actively growing cells. Quite instructive are three series of experiments on the nitrogen fixation by AzotohacteVj made by Koch and Seydel (1912). They determined the nitrogen fixed and the sugar used by Azotohacter at different times, and computed the mg. N2 fixed per gram of dextrose. Nitro- gen is not a direct measure of growth, but is as good as any of the other methods. Table 44. — Nitrogen Fixation and Sugar Consumption by azotobacter Experiment II; 2.383 gm. dextrose in 50 c.c. Days IH 2 3 4 5 7 8 9 11 13 17 32 Mg. N fixed 3.29 3.7 5.3 6.6 7.3 6.7 9.1 8.0 10.0 7.3 8.8 7.6 Mg. dextrose used 0 69 72 162 323 283 706 1,276 1,335 1,287 1,401 1,351 Mg. N per 1 gm. dextrose 00 53.6 75.0 48.6 22.9 24.7 15.0 6.3 7.5 5.6 6.2 5.6 Experiment III; 7.445 gm. dextrose in 50 c.c. Days 6 7 8 9 11 47 Mg. N fixed 1.17 70 17.0 1.80 224 8.7 1.49 191 7.4 1.62 344 5.2 435 4.8 2 47 Mg. dextrose used 494 Mg. N per 1 gm. dextrose 4.3 Experiment IV; 2.316 gm. dextrose in 50 c.c. Days 6 8 9 11 47 Mg. N fixed 1.85 41 30.3 3.90 181 21.1 2.06 747 4.5 2.73 918 3.0 3.56 Mg. dextrose used 1,016 Mg. N fixed per 1 gm. dextrose 3.5 GROWTH 185 Table 44 shows how the growth produced (= mg. nitrogen fixed) by the utilization of 1 gm. of dextrose, decreases rapidly with the age of the culture. The amounts of the first two days are too small to be determined accurately, but the efficiency measured after six days is three to ten times as large as that measured after the twelfth day, in all three experiments. The absolute efficiency can be estimated from the data of Gerlach and Vogel (1902) that 1 mg. of N = 9 mg. of dry cells of Azotobader, and that 1 gm. of dry cells has a combustion heat of 4.6 Cal. (Table 42) . This gives a utihzation of about 50 % for the first four days of Experiment II, and of 30% for Exp. IV. After ten to twelve days, the efficiency is only 3 to 8% (see also Linhart, p. 183). In well aerated cultures of Azotohacter, Hunter (1923) found a constant utilization for the first eight days. The mg. N fixed per gram of glucose consumed were, for four successive two-day periods, 6.9, 9.37, 8.64, and 9.74. The average efficiency of utilization of energy in this culture is much smaller than that of the previous experiment. From Rahn's data on the lactic fermentation (see p. 400), it is possible to compute the amounts of lactose required for one genera- Table 45. -Mg. Sugar Fermented during One Generation op Strept. lactis Hours Culture I, mg. Culture II, mg. 9-12 12.5 X 10-10 12-15 14.6 X 10-10 15-18 9.1 X 10-10 17.6 X 10-10 18-21 14.2 X 10-10 49.8 X 10-10 21-24 12.4 X 10-10 173. X 10-10 24-27 24.9 X 10-10 00 27-30 37.2 X 10-10 00 30-33 00 00 186 PHYSIOLOGY OF BACTERIA tj 0} f-l B P. SVm ^^ 0 5=3 fl o ■* 8. U w o o o ^ ^- t-H O o o got odd d d d d 43 >S >. o -o olved en per ', mg. ■«# o »c i-< »H 00 CO lO CO lO 00 (N «0 l^ lO tH OS .-1 CO CO § O O t^ O CO l> CO o o o o M^s d d 00 00 >0 CO* 1-H d d d d d Q g^ ^ ". s t* «0 00 00 00 CO t^ O CO t^ t' l^ o t- l>. CO CO a a O CO t^ 00 ■* CO CD CO lo ^ (N ec 00 CO o »o IC CO CO 05 1-H t^ ■* Tj< t^ 00 TS CO O "O d CO T(< lO CO l-H Tj< lO 00 lO CO a3 S CO CO CO CO CO CO CO CO CO CO CO 0. O a o 3 o ^ o 1 t^ l^ (N O o • 00 • t^ 1> ^ CO 00 a> lo ic o CO ^ 0) 00 lO Tj( t> iC l-H lO lO 05 t^ CO O O •* CO •w (N CO t- rH lO t- • OS "* 00 c^ C^ CO CO •-• CO CO CO 03 fl S V 1-4 CO • co" • d d lo" ^'' d d d d co" oo" i> a (N CO to lO iO t^ W CO CO CO O £1 a 2 ^^ ^ ;:^ :f^ ^ ;:?; :?^ ^ 3 3 O iC O iC O CO ■* CO 00 o CO 00 CO 00 CO "* LO CO r>. 05 TJ4 lO OS ■* 05 CO O H- -• O CO CO CO CO CO CO CO CO ■* rfl ^ rt CO Tjt -S ^ a a 3 it tH «o N. CO t- CO lo o o 13 C !^ . r-^ o c> o ', 1 d d d d d d d 0) , ^ 2 bC j^ >> >> « J 3 O -n H > C<1 00 lO 00 lO * o 00 t- -** CO "* 00 a 1 ^ 00 "* o iO l^ lO l>'*OCO.-l-»JiOTj(cOO V. ^ CO CO t* t^ CO* d^ d co" d d d oo" lo" r-" d d i CO CO COCOr-tCOCOCOCOfH,-! s U 3 o ,i« o g CO • «o t* C^ 00 lO • ■«J< • IC ?§ § § g S3 s § § § ;^ 0) CO • Ci iO o o 00 • ■* CO • CO rHCO<-iCOOCOCOi-HTj<00 .' T-H co" co" t^" d d d co" t>" co" lo" a rt t» lO CO lO CO CO o a 2 ^ :j^ ;?: ^ :^ :^ :^ :5^ :^ :j^ :5j :?: ^ 2 O CO t^ CO t^ CO tJ< >0 CO 00 o CO lO Oinoiocoi>>oeooi.-i H - o ^ CO CO CO CO CO CO CO eo ^■»*(kO»OOS'*05'*C0»-H 43 rt rH CO CO '^ GROWTH 187 o 1 a* a 1 o bb S (N lO 05 O (N CO Ti< »0 lO (Z> d> d> a> a> II T o 1—1 tH ,-1 t^ 00 Tj< l>. O '-< '-H r-1 rH O o 1 2??§-S5 03 1 Tt< CO CO CO T-H o 1 1 1 ^1 1 1 T o i-H 05 00 «0 UO d d d d d T o 1— H 00 uo »0 tH !> d (N »-H i-H i-H o r 1 ^°^??S22 2 1 CO (M i-H O (N 1 bb O (2 1.87 10-» 0.99 0.80 0.60 O) o at 1 o O CO 00 rH lO (N (N ^ I'' 1 ?3 5;SS 2 1 Cq ,-< rH r-l CO t^ CO O "^ 00 o Ttl 05 05 ^ CO CO CO O O O 1-1 o o o S o o 00 ^ lO 05 00 t^ iO (N (M (N lO »0 lO CO CO (N CO "«i< CO -^ CO CO CO o o o o CO O CO r-< |> TJH CO Tt< <6 (6 c^ S 00 CO CO (N cq (N CO CO »0 (M 1-1 rH 188 PHYSIOLOGY OF BACTERIA tion. Table 45 shows that 9 to 14 X 10 -^^ mg. are sufficient for the doubling of young cells, and that the required amount increases with increasing age of the culture, approaching infinity. Meyerhof (1916a) determined how many mg. of nitrite nitrogen must be oxidized by Nitrobacter in order to increase the organic nitrogen of the culture 1 mg. This increase is a good measure of the growth rate. Young cells accomplished this with 107 mg. of nitrite while older cells required 135 mg. to accomplish the same amount of growth. While with Azotobacter, no accurate data could be obtained for cultures younger than three days, and with Strept. lactis, not younger than nine hours, the very accurate methods of oxygen determination make it possible to estimate the energy utilization right after inocula- tion. A good material for this are the experiments by A. Miiller (1912) with cultures of Ps. fluorescens and Bad. coli in a very dilute medium, containing only 0.004% of asparagin and 0.006% of ammo- nium lactate and minerals. The organisms were grown in closed bottles filled completely to the stopper without leaving an air space. For each test, a new bottle was used. Some of the data on Ps. fluorescens will be given in Table 130, p. 390 while the data on Bad. coli are shown in Table 46. From these data, the oxygen consumption per generation has been computed (Table 47). Though there is a good deal of fluctua- tion in these data, it cannot be said that there is less oxygen needed at the start than a few hours later. This should be the case if the very young cells could use their food more efficiently than the older ones. (c) SUMMARY OF FACTS All microorganisms liberate energy during their life processes. Only a fraction of this energy is accounted for by the energy content of the newly formed cells: the larger part is transformed to heat. Even when cells are formed from building material which has a higher combustion heat than the cells, energy is liberated by fermentation. Part of the total energy liberated by fermentation can not be used for cell construction, for thermodynamic GROWTH 189 reasons. This part is usually quite small with the type of organic changes usually met in fermentations. During the period of rapid growth, more energy is utilized in cell construction than in the later stages of development. SUMMARY OF THEORIES It is generally believed that growth is impossible without a source of energy. The amount of energy required for growth seems quite considerable. In peptone media and in milk, the total energy content of the construction material is similar to that of the cells, i.e., their combustion heats are similar, but a special source of energy seems still necessary. Even when the energy content of the building stones is much higher than that of the cells (methane, amines) growth takes place only when accompanied by oxidation or fermentation. Perhaps, the formation of certain essential cell com- pounds requires a definite potential which can be main- tained only by a special and continuous source of energy. Since fermentation continues after growth has ceased, young cultures must show a better efficiency in the utilization of energy than the older ones. There is no indication that during the lag phase, or at the very earliest period of rapid growth, energy is utilized better than during the immediately following period of rapid growth. in. THE GROWTH RATE (a) METHODS OF MEASURING THE GROWTH RATE The growth rate, or more correctly, the rate of multi- plication, is ordinarily measured by a computation 190 PHYSIOLOGY OF BACTERIA of the generation time (Appendix p. 399). Since the generation time means the time required by bacteria to double in number, it is smallest when multiplication is most rapid. This necessitates thinking in reciprocals. The formula is Hog 2 g - log 6 — log a where a is the number of cells at the start, and 6 the number after the time t. A direct method for measuring the growth rate has been suggested by Slator (1916) who, by the application of calculus (see p. 403), derived the following formula for a growth rate constant : K = 1^ ^ ~ ^^ ^ = log 6 — log g ^ log 6 — log g t tloge 0.434^ Comparing this with the formula for the generation time, we see that the growth rate constant K of Slator is nothing but the reciprocal of the generation time multiplied with a constant factor: K = 1 !2S2 _ 0.694 ^ log e g The disadvantage of the growth rate constant is that it cannot be defined descriptively; it has no easily expressed meaning. It is just a number indicating the relative growth rate. For comparison, the following relation between growth rate constants and generation times might be useful: The generation time of 15 minutes corresponds io K = 2.776 The generation time of 20 minutes corresponds io K = 2.082 The generation time of 30 minutes corresponds io K = \. 388 The generation time of 60 minutes corresponds io K = Q. 694 The generation time of 120 minutes corresponds io K = 0.347 GROWTH 191 If growth is determined by any method other than counting cells per c.c, the growth rate can be computed by the same formula, just as the formula for the generation time would apply. As an example, Slator (1916) measured the growth rate of Lact. Delbrucki of which plate counts would give no accurate measure, by comparing the turbidity of the culture with a standard asbestos suspension, and obtained the following results: Age of culture, in hours b a OAMK 2.53 3.48 5.02 119 800 18,100 0.82 0.83 0.85 The average for 0.434K = 0.84 corresponds to a generation time of 21.5 minutes (see also the computations for yeast growth p. 211). Slator went even further than this and claimed that, since the fermentation is proportional to the number of cells, the growth rate can be computed from the increase of the fermentation products. This will be permissible only if all conditions are entirely under control for we have seen (p. Ill) that the rate of fermentation is easily influenced by the products already formed. This same principle has been used as early as 1877 by Nageli and Schwendener to deter- mine the number of cells. Meyerhof (1917) used it to compare the numbers of active cells in Nitrosomonas cultures. The growth rate of bacteria under optimal conditions is very large. With common saprophytic bacteria in a suitable medium at room temperature, the genera- tion time is about one hour. At high temperatures and in excellent media, especially with a good source of energy, the generation time may drop to fifteen minutes. This means that one cell can multiply to 16 cells in one hour. This rapid rate can be maintained only for a few hours. With yeast, the fastest rate is about 1 hour. However, some microorganisms grow very slowly. Among the slowest is My cob. tuberculosis requiring about 192 PHYSIOLOGY OF BACTERIA six hours for one generation, and slower yet is the organism of Johne's disease. No food has as yet been found to speed up the growth of these organisms to approximately the same rate as saprophytes. In the organism of the host, these bacteria also multiply extremely slowly. The data obtained by the plate count method are averages. The growth rate of individual cells varies greatly even if the cell material, as well as the environ- ment is very homogeneous as will be shown on p. 266. (h) THE LAG PERIOD It has become customary to speak of ^^ growth curves'' of bacteria or yeasts which means curves representing the changes in the numbers of living organisms in a given time. Often, instead of the numbers of the indi- viduals, the logarithms of these numbers are plotted against time. Frequently, these '^growth curves'' include the stage where the individuals die from old age, and sometimes this period is spoken of, and even calculated, as '^negative growth." The conception of death as negative growth cannot be considered a bio- chemical definition; it is used only as a simile, for death is not a reversed growth process. Since death in old cultures will be treated together with other causes of death, this phase will not be considered here. In Fig. 42 of Appendix (p. 404) a growth curve of Ps, fluorescens is given. A set of logarithmic growth curves is shown in Fig. 17 representing the multiplication of a Bacterium A at different temperatures, as measured by Max Miiller (1903). The data for these curves are to be found in Table 48. In the logarithmic growth curve, the generation time is constant as long as the curve is a straight line. This GROWTH 193 period of growth at a constant rate is therefore sometimes given the strange name ^^ period of logarithmic growth"; the bacteria really increase exponentially and not logarithmically. It is a common observation that the rate of multiplica- tion of a culture is not uniform throughout the entire period of development. A decrease in the rate towards Fig. 17. — Logarithmic growth curves of Bacterium A at different temperatures. the end would be expected, but there is also a period of slow development at the start, during the first hours after inoculation. This "Isig period" is visible in the curves of Fig. 17 and is especially conspicuous in the high generation times during the first hours of development in Table 48. The ''lag period" has been the cause of much investiga- tion. As early as 1895, Max MuUer explained it through 194 PHYSIOLOGY OF BACTERIA fl . O TO •J3 f-i 05 g-^- Tti rH CO lO o l> CO (N 05 05 o a T-H (N rH d 05 00 00 00 t> t>- § a r-t rH h o^ .fj o o o o o o o o o o o ti P i> >o CO CO o o o o o o o to o> r^ o o CO o l-H 00 O CO s rH 1-H l> Ci (D o" 1-i rH (N t> CO 00 03 i-H 1-H -(-3 c3 s J. a l> (N tH (N CD T}< 05 OS (X> (© o • • s a 1—1 >— 1 CD CO 1—1 05 00 »o Tt< rH Ti< CO (N CO CO o t>. o o o o o o o T— i fl CO o »o o iO o 1^ o o o o o 8 ^ (N (N CO CO 1— 1 1-H t~>» O o o o T— 1 C^ CQ o o o o 00 o i>« T^ zo o .^ >. •s ^ ■s Tj< (N CO rj< ^ 1-H no 00 PL, r-^ 1 ^. • l> t^ »0 (M O 00 00 t^ lO 00 T}1 lO Til Tjl : ^. '^ ' 00 to lo 00 d lO CO Tti TjJ CO l-H rH Co' d '. ^ Tfi o) oT oj .S • 00 00 i> «D CO CO lo lO lO »o lO lO 00 05 • 1-H T-t o o -^ 1 o o o o o o o o o o o o o o o • O O ^ .^J t>- G5 (N Tt< O O o o o o o o o o o • O o (N i CO rH CO^ ^^ l> 00^ 05^ O o o o o^ o^ o^ o^ ; O^ O r-T (m" Co" t^" lo" Co" CO CO <£ G G c^ lo^ d^ o • o" d^ T-H Til O (N CO o (N O O 1-H lH O ' o o c^ CO CO rH O • ^ 00 a> -M rH CO 00 (N lO 05 • O lO c3 (N »H CO • (N "^ s (N (M • CO 05 •2| ■8 a • (M t-l rH 00 CO l> CO O 00 O CD CD 00 Oi • Oi CO • CO l> o o t^ (N rH rH d d t^ io (^ lii • d to • 05 00 00 CO lO to to lO «o lO "* TjH 00 05 . rH r}< • I-H rH 0 O'^ o o o o o o t^ lO CO 00 Oi CO o o o o o o ill • o o .f^ lO o o o o o • o o CO o o (N lO o^ o o o^ ! *"! ^ T-T (N Co" Co" rH ^" co't^" d" o" 00 05 O r-t (N ^ 00 (N CD 00 (N CD l-s rH rH 1-H (N (N CO CO -* to to GROWTH 195 d . 1 O 05 ■^-^ 05 (N (N (M lO Tt( Oi TtH t^ l> l> l> GO 05 O gs '"' o o « ^^ o o o o o o o o ;3 o o o 8 o o 8 o o o o § cc 00 CO lO o o o o t^ lO iC ^ o o o o 1— 1 O PLH r-( O 00 O O l-l rH CO to ^ §.s 05 lO lO ^ • • • ^ H Jj -^ o o ^ g O Tj^ O CO s ^ (N ^ I>- s co • 1—1 iT . O (M CC > CO O -^ 00 (M S 2 CO t* l> . Oi (M T*H CO Ci e H r- O ro O to CO O l> CO o '^ CO to t^ o to t^ Tf^ (N (N r-^' d (N C^ (N (N (N (N -(J c 8 to' o d" (N TtT CO to CO to C^ CO CO a 1 CO i> 00 05 o i— • Generation j time, hrs. CO Ttl (M r-( to J ' Plate count days ^^ t-h" (n TiT t>r ?:? Tj< O C o o o c CO lO »c IC lO ^ CO CO lo CO rH o o c O 1-H ?C CO o c CO l> 00 l> I> CO CO CO 05 cc c »o ^ i> CO oc oc 1> IC CO CO cr oc CO c 00 o a cr Oi CO »0 (M cr o t^ o c cr o c o 05 »a l> 05 o If: i> o CO (M O lO lO o c C O 05 iC '^ O (N lO lO ?^ »o O C o s c 1-1 lO iC O: l> CO 05 C (T 05 CO O iC OC lO lO c^ »o 05 a a 05 05 »^ i> 00 CO ic t— 1 O: Tti t^ oc oc 0:1 • »J^ O CO c -^ Tt^ oc tH l> oc l> 00 00 »c CO »o i> iC to • IC (N Tj< »C cc CO • ^ iO (N iC oc CO Ttl CO cr X ^ 1 x._ CO GROWTH 219 the growth rates decrease from day to day. With Phomopsis citri, the maximum growth rate, i.e. the optimal temperature is constant at 27.5°C. Above 31°C. the growth rate decreases from day to day. With Diplodia natalensis, the optimum temperature shifts from 30°C. to 25°C. and at 30°C. and above, the growth rate decreases with time. Considering, then, the Tammann principle established, we have a good explanation why the upper range of temperature for growth is limited, why there is a maxi- mum and an optimum temperature which are unknown with common chemical reactions. For the range of normal growth temperatures, the most accurate and extensive data of the older research are those of Ward, Max Miiller, and Barber. All three differ in their technique. Ward measured the increase in the length of the cells of B. ramosus under the micro- scope; Miiller counted his bacteria by the plating method; Barber grew one single cell of Bad. coli in a hanging drop and counted the number of cells under the micro- scope after a given time. Table 56. — Growth op B. ramosus at Different Temperatures Tempera- Generation Growth Tempera- Generation Growth ture, time, rate ture, time, rate degrees minutes constant degrees minutes constant 10-12 300-400 0.10-0.14 26 40 1.04 14 192 0.22 28 33 1.26 15 138 0.30 31 34 1.22 16 108 0.38 34 33 1.26 17 90 0.46 35.5 31 1.34 18 76 0.55 36.8 29 1.44 19 70 0.60 38 30 1.39 21 60 0.69 39 30 1.39 22.6 52 0.80 39.5 35 1.19 24 40 1.05 40 50-00 0.83-0 25 40 1.05 220 PHYSIOLOGY OF BACTERIA ^-^""^ ..1 •^ ^('^''^ ' 5> OJ^ ^^* 0.6 oA V 0.Z ,X fO^ . ZP' . . 30' . . . . ^\>-\ Fig. 22. — Growth rate of B. ramosus at different temperatures. The data of Ward are computed in Table 56, as growth rate constants (see p. 190). These constants are plotted in Fig. 22. It shows the asymptotical approach to zero of the rate of growth at low temperatures, and the abrupt approach to zero at high temperatures. Table 57. — Generation Times op Five Bacteria at Different Temperatures (Shortest times only) o°c. 6°C, 12''C. 25°a 30°C. Bact. A 19^^57' 22n8' 33''38' 15''55' 21''26' 7^4' 6*25' 7n' 5^3' 7^*33' 115'38" 190'30" 145'54" 125' 0" 161'30" 51 '24" 53'9" 53'48" 45'6" 75'36" 46'38" Bact. B 50'56" Bact. C 46'15" Bact. D 44'38" Ps. fluorescens 64'11" Temperature coeflBcient Qio Bact. A 5.4 9.1 1.9 1.2 Bact. B 3.1 2.7 1.1 1.08 Bact. C 13.6 5.8 2.1 1.4 Bact. D Ps. fluorescens 6.8 4.4 2.2 1.01 5.7 5.4 1.8 1.4 GROWTH 221 Q,o \ \ i i\ 3 C 1 9 "^•^u. "^"^•^ 1 N N, 25 30' 35" W M' Fig. 23. — Temperature coefficients of growth of Bacterium coli (upper curve) and of B. ramosus (lower curve). The following table shows the generation times and temperature coefficients of five bacteria at the time of most rapid growth, after Max Muller (1903). The detailed data for Bacterium A will be found in Table 48, p. 194. The temperature coefficient is quite evidently not constant, but decreases with increasing temperature. To emphasize this decrease of the temperature coefficient Table 58.— Temperature Coefficients of Growth B, ramosus Bact. coli 14°-24°C. 4.8 15^-25°C. 4.2 15°-25°C. 3.4 17°-27°C. 2.8 16°-26°C. 2.8 19°-29^C. 2.7 18°-28°C. 2.3 2r-3rc. 2.3 21''-31°C. 1.8 23°-33°C. 2.1 24"'-34°C. 1.2 25°-35°C. 1.9 26°-35.5''C. 1.2 27"'-37°C. 2.0 28°-38X. 0.9 29^-39°C, 1.8 3i°-4rc. 1.4 33°-43°C. 1.3 35^-45°C. 1.04 37°-47°C. 0.42 222 PHYSIOLOGY OF BACTERIA with the increase of temperature, Table 58 represents the temperature coefficients of B. ramosus computed from Table 57, and those for Bad. coli obtained by- Barber (1908). This relation is shown graphically in Fig. 23. It should be mentioned that Barber in his direct microscopic observations with Bad. coli made no statements indicating that he ever found the Tammann principle. He never observed more than ten generations, and the generation times were at 40°C. from 17-25 minutes at 42°C. from 20-23 minutes at 45°C. from 18-30 minutes at 47°C. from 42-68 minutes with no differences whether ten or only two generations were observed. We had started with the assumption that the growth rate is increased by temperature increases in the same way as any chemical process, and that the deviations at the maximum and minimum temperatures (as evi- denced by Figs. 21 and 23) are to be explained by a superposition of other factors. The existence of a true optimum and a less definite maximum temperature for the growth rate could be accounted for very satisfactorily by the Tammann principle. The difficulty lies in explaining the minimum temperature. Growth is the result of many interlinking processes. If one of these many processes is extremely slow, the entire growth is slow. If one of the processes ceases to function, there will be no growth, regardless of the con- dition of all other involved agencies. At the minimum temperature, there is evidently some limiting factor, but we have not sufficient knowledge of the growth mechanism to locate this factor. GROWTH 223 A possible explanation is the assumption that the available energy- is no longer sufficient for synthesis. Each synthesis requires a minimum amount of available energy, not only in quantity but probably also potentially. This is furnished readily in suitable media at normal growth temperatures. Even then, a considerable amount of the available energy is not utilized for synthesis (see pp. 180-188). This apparent waste might be due to radiation or conduction of heat, or to the general effort of keeping up a certain energy potential. In the first case, the absolute loss would be greater at lower temperatures; the relative loss would be very much greater, because the amount of available energy is greatly reduced by low temperatures. If the maintenance of a certain energy poten- tial is necessary to bring about growth, there may be a temperature well above the freezing point where certain organisms lose so much of the small amount of energy furnished them by their fermentation that the potential in the cell cannot be maintained, and the cell ceases to grow. It has been shown on p. 134 that streptococci in milk at — 1°C., gradually lose the power to ferment. Their endo- enzyme deteriorates and cannot be replenished at such low tempera- tures. If the cell ceases to be able to make repairs, it must gradually die from the slow, but certain deterioration of its essential, ther- molabile components. (See also influence of kind of food, p. 157.) The data obtained by the author with streptococci indicate that bacteria die at temperatures too low for. growth. The death rate is very slow, but distinct. Table 59 shows the number of cells in milk cultures of three streptococci. Strept. cremoris No. 18 is the only one which does not grow at H-5°C. The generation times at 5°C. for the first time interval were 40.7 hours for Strept. cremoris 23 and 55.5 hours for Strept. lactis. It takes about 2 days for one cell to double. At 10°C., the generation times are 24.2 hours (Strept. cremoris 23) and 12.5 hours (Strept. lactis). It seems remarkable that a growth process can be so slow and yet not come to an end. At — 1°C., all three cultures showed a decrease in numbers. A gradual decrease of the number of cells of Lactobacillus acid- 224 PHYSIOLOGY OF BACTERIA ophilus in milk at low temperatures has also been recorded by Kopeloff (1926). Graham-Smith (1920) also reports a slow decrease in the number of viable cells of Micr. pyogenes in broth at 8-10°C. Table 59. — Increase and Decrease of Streptococci at Low Temperatures (Colonies per c.c.) Strept. cremoris No. 18 °c. Start 10 days 31 days 45 days - 1 + 5 +10 20,000 20,000 20,000 4,500 40,000 > 1,000,000 1,000 (135,000)? 80 8,900 Strept. cremoris No. 23 °c. Start 7 days 14 days 21 days 34 days 48 days - 1 + 5 5,400 5,400 5,400 2,000 62,000 >1, 000, 000 2,300 1,085,000 75,000,000 700 240 1,500,000 63 5,500,000 + 10 59,000,000 Strept. lactis ""C. Start 9 days 23 days 37 days - 1 + 5 + 10 77,000 77,000 77,000 22,000 630,000 835,000,000 113,600 21,000,000 20,700 6,000,000 Another fact must be mentioned here which is still more difficult to explain biochemically, namely the adaptation of organisms to certain temperatures. This is shown very plainly in Table 60 representing some experiments of Zikes (1919b) with yeast. Of each of the six yeasts tested, one culture was grown at 8°C., and another at 25°C., for several weeks previous to the final test. The temperature of precultivation had a very decided influence upon GROWTH 225 Table 60.— Growth Rate of Yeast at Different Temperatures Temperature, "C. - Generation time of yeast precultivated at No. of yeast 8»C. 25''C. 1 12 6 hrs. 19 min. 17 hrs. 31 min. 15 4 hrs. 45 min. 9 hrs. 58 min. 19 3 hrs. 16 min. 4 hrs. 38 min. 24 3 hrs. 31 min. 3 hrs. 17 min. 30 2 hrs. 19 min. 2 hrs. 16 min. 38 33 hrs. 50 min. 24 hrs. 51 min. 2 9 11 hrs. 47 min. 30 hrs. 24 min. 12 5 hrs. 48 min. 13 hrs. 40 min. 15 4 hrs. 04 min. 7 hrs. 02 min. 20 4 hrs. 13 min. 5 hrs. 02 min. 25 4 hrs. 04 min. 3 hrs. 59 min. 30 2 hrs. 39 min. 2 hrs. 29 min. 39 58 hrs. 10 min. 30 hrs. 25 min. 3 8 5 hrs. 18 min. 14 hrs. 24 min. 15 3 hrs. 29 min. 3 hrs. 52 min. 21 3 hrs. 02 min. 2 hrs. 20 min. 29.5 2 hrs. 02 min. 1 hr. 54 min. 4 9 16 hrs. 30 min. 24 hrs. 41 min. 12 9 hrs. 43 min. 10 hrs. 07 min. 14 5 hrs. 02 min. 6 hrs. 39 min. 20 4 hrs. 43 min. 3 hrs. 55 min. 24 3 hrs. 46 min. 3 hrs. 15 min. 30 2 hrs. 53 min. 2 hrs. 28 min. 40 118 hrs. 03 min. 68 hrs. 07 min. 5 10 5 hrs. 12 min. 7 hrs. 28 min. 15 3 hrs. 40 min. 5 hrs. 12 min. 24 3 hrs. 48 min. 2 hrs. 52 min. 29 2 hrs. 54 min. 2 hrs. 09 min. 6 11 6 hrs. 11 min. 13 hrs. 36 min. 15 4 hrs. 19 min. 4 hrs. 19 min. 20 4 hrs. 19 min. 4 hrs. 00 min. 25 3 hrs. 09 min. 3 hrs. 02 min. 30 2 hrs. 48 min. 2 hrs. 36 min. the rate of growth at different temperatures. Yeast No. 2 shows the greatest adaptation. At 9°C., the culture adapted to low tempera- tures grew almost three times as fast as the other, while at 39°C., it required almost twice as long a time to double. 226 PHYSIOLOGY OF BACTERIA To account for such adaptations biochemically seems almost impossible with our present knowledge of the cell mechanism. SUMMARY OF FACTS Temperature influences the growth rate greatly. At very low temperatures, the time required for a cell to double may be more than forty-eight hours. The same organism could double at the optimum tem- perature in half an hour. There is a definite low temperature at which a culture will cease to grow. Near this point, the time required for a cell to double increases rapidly with a decrease of temperature and finally becomes infinite. At tempera- tures below this minimum, the cells die slowly. From a point about 10°C. above the minimum to the optimum temperature, the rate of growth will be increased from two to three times by a 10°C. rise in temperature. Above the optimum temperature, the growth rate is not constant, but decreases with the time. During a first short interval, it may be higher than the rate at the optimum temperature, but it does not last. A culture grown for a while at low temperatures will grow more rapidly at low temperatures than a culture of the same strain precultivated at high temperatures. Bacteria and yeasts can be adapted within limits to low and high temperatures. SUMMARY OF THEORIES Growth must be considered as the sum of a number of slow chemical reactions, all of which are accelerated by an increase in temperature. Growth must have a fairly normal temperature coefficient. The maximum and minimum temperatures of growth are brought about by the superposition of other processes. GROWTH 227 Above the optimum, we can assume the deterioration of some essential part of the growth mechanism with a high temperature coefficient. Such assumption would lead to the Tammann principle which has actually been observed in a number of cases. Growth rates are con- stant below the optimum temperature, and decreasing with time above the optimum, decreasing the more rapidly the farther the temperature is above the opti- mum. The decrease will ultimately lead to death. The minimum temperature is not as yet explained in a generally accepted way. One possible explanation is the assumption that the slowness of energy liberation does not allow the cells to accumulate enough energy to accomplish some essential synthesis, or repairs. (/) CHEMICAL STIMULATION OF THE GROWTH RATE In the study of chemical stimulation of fermentation, an analogy was drawn between the action of poisons and the action of temperature. With isolated enzymes, an immediate acceleration of enzyme activity was found, followed by a greater decrease due to a more rapid enzyme destruction. In fermentation by living cells which can regenerate the enzyme, a continuous accelera- tion of the fermentation by stimulants was assumed to be possible. In the case of chemical stimulation of growth, the analogy with temperature may well be continued. Let us assume, for the sake of simplicity, that the rate of fermentation remains uninfluenced by the concentration of the stimulant which affects growth. Two processes are stimulated by the poison acting as a catalyst: the cell sjmthesis, and also the deterioration of the growth- producing agent. This agent is deteriorating normally in all cells, though quite slowly, and is constantly being 228 PHYSIOLOGY OF BACTERIA regenerated by the cell. This regeneration has a limited speed. As long as the deterioration of the growth agent does not exceed its regeneration, everything goes well, and a stimulation up to this point will be well tolerated by the cell. Beyond this point, cessation of growth is inevitable if the stimulation continues. This deduction, which is merely based on analogy, leads to the conclusion that the Tammann principle should become evident at higher concentrations of a chemical stimulant. Very little experimental evidence which includes the time factor is available. A good example is an experi- ment of Hline (1909) given in Table 61 which shows, at first glance, a good agreement with the influence of temperature on fermentation and on growth. It is a fine illustration of the existence of the Tammann princi- ple in chemical stimulation. This example may be objected to for the reason that the bacteria are multiplying, and that, in the same culture, the same amount of formaldehyde is acting upon much larger numbers of cells as the time advances. However, the theory of stimulation by poisons assumes that the action is catalytic, and therefore, its action is independent of the concentration of the substrate i.e. the number of cells. This assumption is justified because it will be shown on p. 341 that the death rate depends upon the concentration of the poison, but is independent of the numbers of cells acted upon, within very wide limits. The experiment quoted above is not sufficient, however, to make any further deductions. There is a very long lag period, the control requiring seven hours before the most rapid growth is reached. Besides, the generation time of all cultures but the last shows a distinct increase between two and four hours which suggests some unknown outside retarding influence. Therefore, the generation times mean little. All other material on growth stimulation refers to final crops, and will be discussed later (p. 254). No GEOWTH 229 O) ® 1 S5 o o o o o o o .^ .> O 00 Tt^ '-4-3 '-t-^ T-H rl (M Tl< o3 c3 o CO ^ ^^ d iz;^ o o o o o o o Tf CO c^ o o o c • • ^ CO (N^ O .'' ^" ^ d" d 8 8 a 'a a ^^ '^S^^^P CO CO d r-T ''iT IC I— I 1— I o o o o o o o ^ Co ""^i 00 CO o o o t- CO -," t-h" lo" oo" (n o d >^ ^ T-( i-H CO 00 O 8 si 'a a Q p 00 00 o O 00 CO rfi d ,-i Tt< 1— 1 ^ o o o o o o o -P o e> 00 00 o o o , . , £2 ;3 00 "^ (M <0 O O^ O ■-+3 d fl fl el o O CD <^ TjT (N d~ d" d rH >~l -^ C ■^ o3 lO lO (M l> d si (N oo" oo' s T-l Ttl CJ S C! o o o o o o o — 0) ^ (M CD O O O O O 00 ,-H >-i CO o o^ o . . . fl £3 C fl fl 1 oT TjT lo oi' '^'" <::r o rH IC) ^ 00 ^ ^ a *a "a • fH .1-1 a a >-i . CO CO T^^ o o^ o o 05 CO !>• CO O O O 8 s a a 'a tH CO ^ o o tH O O CO CO CO "^ 0> <30 l> o 1-H IC o TS >-> ,£5 O d T3 a 'd -3 rji rH -^ ^ 1 1 2 t I H C 5" Tt 1 c^ »- "* cs 4 r>i 1 230 ■ PHYSIOLOGY OF BACTERIA further data concerning the stimulation of the rate of growth are known to the author. SUMMARY The rate of growth can be increased by certain concentrations of chemical poisons. If a number of different concentrations are tried, stimulation is greatest at first with the higher concentra- tions, but the optimal stimulating concentration becomes smaller and smaller as the time passes. This behavior suggests that poisons effect organisms similarly as high temperatures, accelerating catalytically the rate of action, as well as that of deterioration of some essential growth factors. IV. THE ENDPOINT OF GROWTH All growth in a given amount of medium will ulti- mately come to an end. The final amount of growth, i.e., the total crop, will depend upon numerous factors. With the same organism in the same medium, it will be proportional to the volume of the medium (unless the volume has had an influence upon the rate of oxygen penetration into the medium). It will be independent of the size of the inoculum unless this has been very near to or above the number of cells that could develop in the medium. (Rahn, 1906; Graham-Smith, 1920; Buchanan and Fulmer, 1928.) The composition of the medium will have the greatest influence. In our standard media, we are likely to get quite uniform crops. It is generally considered a rule that most of the common saprophytic bacteria, the organisms of the colon-group, the streptococci (if sugar is present) and quite a large number of other parasites GROV/TH 231 will grow in nutrient broth to about one billion cells per c.c. This may be a question of size or total surface: Table 132 on p. 397 gives the following final numbers of cells per liter: Yeast: 50,000,000; Bad. coli: 1,000,000,- 000; Bacteriophage: 100,000,000,000. But the weight of these cells ranges in the reverse order. The total crop in 1 liter of medium weighs in moist condition: with aspergillus, 7,000 to 19,000 mg.; with yeast, 6,500 mg.; with Bad. coli, 800 mg.; with bacteriophage, 0.42 mg. The concentration of food will influence the final crop. Addition of chemicals (not food) may influence the crop favorably, by acting as buffer, or providing a more suitable osmotic pressure, or they might decrease the crop if they have a toxic action. (a) THE CAUSES OF THE ENDPOINT OF GROWTH Before the factors of concentration of food, of tempera- ture and of chemical influences can be discussed, it seems essential to try to come to a clearer understanding of the cause of cessation of growth. This question has already been raised regarding the decrease of the growth rate (p. 199). It was not answered in a satisfactory way. Many conflicting experimental data have to be analyzed, and many conflicting theories must be considered. It should be realized from the beginning that different organisms may cease to grow from different reasons, and that the growth of the same organism in different media may stop from different reasons. Theories on the limit- ing factor of Streptococci cannot be disproved by experi- ments with Bad. coli or with yeasts. It must be further kept in mind that the rate of growth is independent of the flnal amount of growth. While the rate was found to be but slightly influenced by the 232 PHYSIOLOGY OF BACTERIA concentration of food, the total crop shows distinct relation to the quantity of food (see next chapter). Probably the most common assumption is that the accumulation of the products of fermentation prevents further growth. This can be most easily proved by removing the fermentation products from a full-grown culture. This has been done by Rahn (1911) by neutralizing the acid in milk cultures of two streptococci. The result was increased growth with Strain II, but no increase with Strain IV. Table 62. — Increase in Numbers of Cells of Acid Forming Streptococci after Neutralization of Culture (Numbers indicate millions per c.c.) Strain II II II II IV IV Just before neutralization 1,790 3,270 1,190 3,290 640 2,235 765 2,010 1,340 1,195 2,065 After 8 hours After 24 hours 1,590 Strain II is not affected by peptone (see p. 210). It is capable of utilizing the protein matter of the milk, and its growth limit is primarily fixed by the hydrogen ions. If these are removed, growth continues until another inhibiting factor intervenes. Strain IV does not find sufficient building material in milk; the final number of cells is increased if peptone is added. Neutralizing a culture which contains no peptone does not further the growth because there is no building material. Lack of food is the limiting factor of Strain IV, fermentation products are the limiting factor of Strain II. Rogers and Whittier (1928) give the following final crops for their strains of Strept. lactis: GROWTH 233 Cells per c.c. Unbuffered broth 60,000,000 Buffered broth 480,000,000 Broth with pH held constant 1 ,000 ,000 ,000 Milk 1,000,000,000 Milk with pH held constant 3,600,000,000 Milk, aerated and pH held constant 5,200,000,000 If the inhibiting product is removed, growth will continue until another product of the cells inhibits growth. In the case of lactic acid bacteria, the growth- inhibiting compound after removal of the hydrogen ions might be thought to be undissociated lactate which was found to be the cause of cessation of fermentation (p. 145). But the experiments of Rogers and Whittier show no influence of lactates upon growth. With yeast in malt wort, alcohol is not the limiting factor of the final amount of yeast, according to A. J. Brown (1905). He found that it required very high amounts of alcohol to decrease the final crop, as may be seen from Table 63. Table 63. — Average Number op Yeast Cells per C.c. in Wort WITH Increasing Amounts of Alcohol Experiment I Experiment II Inoculum . . 15,600,000 Inoculum 6 , 880 , 000 W^ort a.lone 72.800,000 78,400,000 72,800,000 Wort alone 79 , 200 , 000 Wort + 07% alcohol . Wort + 1.4 % alcohol Wort + 2.8 % alcohol Wort + 4.2 % alcohol Wort + 5.6 % alcohol .Wort + 7.0% alcohol Wort + 8.4 % alcohol 68 , 800 , 000 Wort + 1,4 % alcohol 70,400,000 57,200,000 38,400.000 24,800,000 9,200,000 The added amounts of alcohol do not represent the total alcohol concentration because some more is formed 234 PHYSIOLOGY OF BACTERIA by the yeast. The final concentration in normal wort was 3.6% alcohol. Special experiments demonstrated that no other volatile or non-volatile fermentation products were the cause of growth inhibition. Good aeration, however, increased the final crop considerably, and it seems cer- tain that the first limiting factor of yeast growth in wort is lack of oxygen which is nothing else than lack of food. Another example where lack of food is the cause of cessation of growth will be shown on p. 248. It concerns the growth of Micr. pyogenes in a meat infusion medium. Aside from these two factors, accumulation of fermen- tation products, and exhaustion of food supply (includ- ing oxygen as a food), a third possibility has been repeatedly claimed, and has been disclaimed as often by others ; namely, the existence of a more or less specific cell excretion which is thermolabile and of colloidal nature. Eijkman (1905) claimed such a compound for Bad. coll. Full-grown gelatin cultures, after steam heating, allowed normal growth to occur again while the same gelatin, remelted without heating to high temperatures, showed no growth upon re-inoculation. This experiment was modified in various ways, always with the same result. Conradi and Kurpjuweit (1905) enlarged upon this theory. Later authors (e.g., Manteuffel, 1907) disproved this by the claim that the absence of growth was merely due to lack of food. Wherever large numbers of living cells were left, the food was divided between so many cells that none could multiply visibly. If, however, heat destroyed the old cells, the newly inoculated cells had sufficient food for growth (see also Henrici, 1928). Rahn (1906) observed with Ps. fluorescens that new growth developed in old cultures upon re-inoculation after GROWTH 235 steaming or filtering. This process could be repeated several times with the same culture, thus producing Table 64. — Repeated Growth of Ps. fluorescens in Broth Cells per c.c. Generation time, Hours after inoculation Heated to 100°C. Heated to 68°C. minutes 100°C. 68°C. 48 hours 2,200,000,000 2,000,000,000 Heated Second growth 0 2 6 12>i .24>4 100,000 300 , 000 1,620,000 31,000.000 760,000,000 37,000 70,000 650,000 16,000,000 800,000,000 76 99 89 156 135 76 82 128 Third growth r 0 3 6 12 24 He 8,000 44 , 500 100,400 12,000,000 470,000,000 ated 10,800 35,000 98,000 6,000,000 250.000,000 73 153 52 136 106 121 61 134 Fourth growth <, 0 3 6 12 24 48 He 1,000,000 2,570,000 4,070,000 40,000,000 ated 1,000,000 2,200,000 1,940,000 26,000,000 976,000,000 4,200,000,000 2,800,000,000 Fifth growth 0 24 He 28,000 39,000,000 600,000,000 800,000,000 ated 42.000 72,000,000 1,800,000,000 1,700,000,000 [72 I 24 1 48 [72 He 4,800 45^,000,000 400,000,000 500,000,000 ated 12,700 189.000.000 1.100.000.000 900,000,000 182 173 Total number of cells grown in 1 c.c. 7,600,000,000 10,100,000,000 236 PHYSIOLOGY OF BACTERIA again and again a normal crop from the same amount of medium (Table 64) . After sterilization of the culture with ether, no growth occurred; growth in the ether- sterilized culture became possible after steam heating, or after adding pieces of a Chamberland filter to adsorb the inhibiting substance. Avery and Morgan's (1924) inhibiting substance of pneumo- coccus culture finally proved to be hydrogen peroxide, a well- defined cell product. Hajos (1922) obtained an inhibiting substance from cultures of Bad. coli and related bacteria by repeatedly centrif- uging the bacteria out of the culture and re-inoculating the medium until no further growth occurred. The chemical constitution remained unknown so far. Since a few drops of it prevented growth in 10 c.c. of fresh broth, the toxic compound seems to be quite con- centrated. But this substance is heat-resistant and is perhaps a fermentation product. Rogers and Whittier (1928) appear to have established a growth- inhibiting factor in Street, lactis which is a cell product, but not acid. They grew Strept. lactis in double strength broth in a collodion sac suspended in a flask of the same medium, the volume of the outside medium being about ten times that in the sac. After two days of growth, the outside liquid was also inoculated with Strept. lactis, and the bacteria were counted from hour to hour. The cells devel- oped to about 70-71 millions per c.c. while in a normal check, they grew to 300 millions. Repeated addition of concentrated broth brought the number of cells in the double culture up to 98 millions, while the normal culture increased to 550 milHons. Rogers and Whittier beheve this compound to be identical with, or similar to, the agent by which Strept. lactis inhibits Lactobacillus bulgaricus; it is also believed to be the substance controlling the final amount of growth in solid or liquid media which contain but little carbohydrate. SUMMARY There are various reasons why growth of a bacterium will cease in a given volume of culture medium. It has been shown that certain streptococci in milk cease to grow because of too high acidity. It has also been GROWTH 237 demonstrated that other streptococci cease to grow in milk for lack of available nitrogenous food. It has further been found that Ps. fluorescens in broth ceases to grow on account of accumulation of a thermolabile cell secretion, and the same is true for streptococci in broth. (6) ENDPOINT AND CHEMICAL COMPOSITION OF FOOD The final crop or the endpoint of growth will depend largely upon the kinds and amounts of construction material if a reasonably good and continuous source of energy is taken for granted. The type of construction material seems quite essential. Most streptococci grow to much larger numbers in milk if peptone, meat extract, or yeast extract is added (see p. 149). Very extensive quantitative studies on the influence of the kind of building stones upon the crop have been made by the yeast industry. The yeast factory tries to obtain the largest possible amount of cells from a given amount of sugar. Since molasses is the cheapest source of sugar available, the manufacture of yeast from molasses is a common process. The building material consists of different organic or inorganic nitrogenous compounds, including the nitrogenous material in molas- ses. While most of the results of these studies are considered trade secrets, the following two tables show a number of unpublished experiments which I owe to the kindness of Dr. H. Claassen, of the ^'Rheini- scher Aktienverein ftir Zuckerfabrikation/' Dormagen, Germany. These experiments were carried out with 5 to 7 liters of wort under plant conditions, being well aerated from perforated copper tubes at the bottom of the vessel. The normal aeration was 15 to 20 liters of air per minute during the first hour, 50 to 60 liters during the next seven hours, and 15 to 20 Uters during the last hour. The 238 PHYSIOLOGY OF BACTERIA S^ o 1 sid .-100 0000 «0 00 CO ^ OJ ^SSi W ^ .i. 1^ HI 2 ^ o •*00 0000 CO ■'i^ 000 (D lO M l-Ht-HCOM IN — ' OS •-ii t^ CO 00 CD^OO oDiN m MT) o • • • on • ^ o 05000 COTjtrf* Oi-i (N 0 ^iS IC CO rH0O(N (NIN -^.-H t^ (N 00 0 o (N OSOiiO (N ^ O 00 »ooo IN 0Ot>CO .-H 00 (Ncor^rt* (O-^ O C. (N 0 Oir-( l^ -* (NiOOCC (N 4 .-KNIN (Dr-t CO «5 t- (N >O(N00 0 00 O r-* 1000 IN ! CO CO 00 CO (N OOCOIN^ t^Oi ^'S O (N t>. OOr-tOSTt 00 00 aj (NIN 00 00 — n ?o >-^ (N IN ©■<1<0 CO t^ O (N«> W5 00 (N Oit^os 0 IN ^rH^05 (NO ^1 (N (O^ 00 t^ COt^iCtC COOJ C^ rHC^CS, t- •"^ — . . . CO CO t^OO-H !> §0 Q iC (N05 OSOOO Tffosr^ OS IN OOiCO 1000 ^« .c 05CD ^Sb T}< t- (N •ap a 0. ~ ?? to ■^*(N .-H "'R IC (N05 OSiOOiO OCOM «O00 INIC ■^ 00 rHCO^ rHco eo 0 lOiOOO l>iO CO •>* .-H iNooeo 0 (N .—1 CO CO OOOCD ■* ■* o 10 IN 00 05000 CC-ico IC (N O5'«*00(N oot^ IJi o 0 CO t- coiocgc^ lOOO cc Tj* (N ^(N(N I© 00 ooicoo »o M o »n (NOO OiOOiO 05Tt -^ 0 OSrHlNeO 00 lO o 0 eo ^CO-H 0O(N Tj* 0 >OTt<-HCC TtHI© cc ■* rH (NCOCO OS CO 00 * COOOtOb- ■*t^ 003 00 V ed C3 S^ (U CO 0 oooS 0 _, o «o« (Mod OSiCOiO 06 (NOO oi (N .-HOSiN-* COOi -^ o 00 ,-iCO^ 0>(N CO i> ^>»o«o(^ r-4«3 CO IN coco 0 7:2; : £S ■ CO ■ 00 ^ 0 S oT Isl- :2 : 0 11 ^1 •■J . a : a • s at^Sg £,, M ^ : ii -'^'' „, S^'i.5 ■■ ^18888 . 3 F^^-;§ § ^ y 0— (U -*J a a OS « 3 § '^ e « S 2|.2m.s 111 5I g a fi 1 1 11 II 5, « d ^fe 0 PL| BJ9 in I 13 dl DO oj [ paaS J IV 8? onp< "d sai^j jd OJ d qeB8A GROWTH 239 IC03 1-1 CQ o OOi o OO I o Tft-Ht^ IX> (N00O5 0 5DI> I IMC) COCO-* ic i-H t>.(NOO —I (NCCCO 00 «* 00 00 IN coco •so — o oi « 3 o^-^W b3 Or 03 ^ IN (3 0 X a 2 t; >> ^o o 5-* 00 )Z^ .o^ W 5 ti e^f S g c3 2 o " csiS S fi S S ss««s«aa oooooooo sj3;jit2,-i9dpooL3 P^^S ^W 8:jonpoj(j sai'^jadojd ^sbo^ of- : a " M : ;>i <^o 2> 32 o aj s, f « ;^ «: ^ fo §3 « " « " B fl ?2J3j3j3 " a« £ ;'2'P-a a c ha ^ § a ■Si 'g 6g g- 3 240 PHYSIOLOGY OF BACTERIA temperature of fermentation was 24°C. at the start, and gradually increased to about 31°C. The acidity of the solution was kept at pH 4.8 to 5.2, and was allowed to increase towards the end of the fer- mentation to pH 5.5 to 5.8; if necessary, it was regulated by H2SO4 or Na2C03. The wort was allowed to run slowly into the container filling it in five to seven hours, as is the custom in plant procedure. Acid calcium phosphate is added to molasses at the rate of 2.0 gm. P2O5 per 100 gm. sugar; the molasses is diluted, heated to 85°C., acidified with H2SO4, and filtered. Malt sprouts are extracted with hot water. Peanut meal is heated with dilute H2SO4 until about 90 % of the nitrogen has become soluble. In order to offer equal quantities of assimilable nitrogen, the following assumptions, based upon long experience, were made: Molasses: 0.7-0.8% of the molasses weight is assimilable N. Peanut meal extract: 65-70% of the soluble nitrogen is assimilable N. Malt sprout extract : 65-70 % of the soluble nitrogen is assimilable N. Ammonia, asparagin, urea etc: 100% assimilable. When sucrose was offered in place of molasses, K2SO4 and MgS04 were added in appropriate amounts. The washed yeast and the wort were analyzed at the end of the experiment, and the yeast was tested for its fermenting capacity because all these experiments aimed at the production of a good bakers' yeast. The volume of CO2 produced by 10 gm. of yeast was determined as described on p. 104, and the time which 5 gm. of yeast required to raise a standard flour dough to a standard volume was measured. The keeping quality was tested by observing the time required for the yeast to soften in a 35°C. incubator. In all these experiments, the substrate, i.e., the sugar, is limited, and is used up completely during the nine to twelve hours of the experiment, the initial concentration varying between 1 . 5 and 2.5%. The biochemical problem is then to explain why yeast, with the same energy supply, will give different crops with different building materials. The crops vary considerably. The first six experi- ments contain the same amounts of assimilable nitrogen in the wort, 2.4 gm. from molasses and 2.5 gm. from GROWTH 241 other sources. All conditions being exactly alike, the crops obtained per 100 gm. of sugar used were: From molasses + ammonia, gm. From molasses + peanut meal, gm. From molasses + malt sprouts, gm. With slow aeration 98 . 8 With rapid aeration. . . 113 . 2 139.8 216.8 110.4 144.7 With ammonia as the only source of nitrogen, molasses being substituted by sucrose, the crop was only 69.0 gm. (Exp. 7). Asparagin (Exp. 8) and aspartic acid (Exp. 9) gave better crops, 83.5 and 93.9 gm. The total available nitrogen was only 4 gm. against 4.9 gm. in the six previous experiments. Another set of comparable data are experiments 18 to 21, with 3.6 gm. of total assimilable nitrogen and 1.8% sucrose instead of molasses. The crops were: Gm. With ammonium phosphate 75 . 8 With peanut meal 140.2 With malt sprouts 104 . 5 Half ammonium phosphate, half peanut meal 136 . 7 These results probably mean that the protein material of the peanut extract was most similar to that of the yeast, and that with the same amount of energy, the yeast could grow more rapidly, and therefore reach a higher final crop. The surprising fact is that with half the amount of peanut extract, and the rest of nitrogen substituted by ammonia, the yeast reached practically the same weight, 136.7 gm., and had about the same nitrogen content, the same keeping quality and a better fermentative action. This shows that while it is uneco- 242 PHYSIOLOGY OF BACTERIA nomical to feed the yeast on ammonia alone, it is quite advantageous to feed ammonia besides organic nitrog- enous matter. A great drop in the yeast crop, if the last of the organic nitrogen is taken away, is very conspicuous in Table 67. These results are enlarged by a number of other experiments. Numbers 16 and 17 are directly com- parable with the above, having a total available nitrogen of 3.5 gm. Numbers 7, 8, and 9 are comparable with each other, showing the differences in the availability of asparagin, aspartic acid and ammonia. In nitrogen-free sugar solution, (Exps. 10, 11, and 12) the yeast loses nitrogen. Exps. 22 and 23 show that in ''spent" wort, practically all available nitrogen had been assimilated by the first crop, and that after addition of more sugar, only a small amount of the total nitrogen could be used by the second crop. When ammonia is used as the only source of nitrogen (Exp. 18), the crop is small, but the yeast is high in protein content. This is typical for all yeasts grown with ammonia. Table 66. — Nitrogen Content in Yeast Solids Yeast with ammonia Exp. 1 8.53% N Exp. 2 8.39 Exp. 7 9.06 Exp. 16 10.00 Exp. 17 8.58 Exp. 18 8.94 Exp. 21 7.94 Exp. 24 9.11 Exp. 25 8.00 Average 8.73% jast without ammonia Exp. 3: 7.84% N Exp. 4: 6.63 Exp. 5: 8.25 Exp. 6: 7.17 Exp. 8: 8.86 Exp. 9: 6.50 Exp. 13: 7.91 Exp. 14: 7.86 Exp. 15: 7.16 Exp. 19 7.96 Exp. 20 7.23 Average 7.58%, Whether this higher nitrogen content is stored protein, we have no means of telling. It does not mean an increase in zymase, for the GROWTH 243 amounts of CO2 produced average 800 c.c. for yeast grown with ammonia, and 813 c.c. for yeast without ammonia. The average time for softening varies greatly; it is fifty-two hours for yeast grown with ammonia, and ninety-two hours for yeast without ammonia. The two samples with urea are the worst ones to soften. These results are verified by another experiment of Claassen's (1928) where increasing amounts of the assimilable nitrogen of malt sprouts were substituted by ammonia. Table 67. — Yeast Crop and Properties with Increasing Amounts op Ammonia Ammonia in % of Yeast %N Dough fermenta- Keeping quality at assimilable nitrogen crop per liter, gm. in yeast solids tion, minutes 35°C. after 72 hours 0 19.56 6.74 67 solid 10 18.92 6.98 64 solid 20 15.98 7.66 75 soft (in 24 hours) 30 15.88 8.60 76 soft (in 48 hours) 40 14.36 8.13 72 soft (in 24 hours) 50 17.85 8.81 59 soHd 60 17.02 8.86 59 soft (in 24 hours) 70 16.30 9.15 62 soUd 80 16.82 8.95 51 soUd 90 17.40 8.88 51 soUd 100 12.89 9.58 56 solid Poor keeping quality and slow fermentation are characteristic for yeast with a mixed nitrogen diet of from 20 to 60% ammonia. The nitrogen content of the solids increases with the percentage of ammonia in the food, and the total crop decreases. Whether this rapid softening is due to a general weakness of constitution, or to an over-supply of endo-tryptase cannot be ascertained from these data. But there can be no doubt that ammonia, though it gives a good crop with normal fermenting capacity, does give a yeast differing physiologically from that grown on organic compounds only. With yeast, the final crop, as well as its nitrogen content, its fermenting and keeping qualities, are decidedly influenced by the nitrogenous food. This 244 PHYSIOLOGY OF BACTERIA seems to disagree with the finding of Abderhalden and Rona (1905) that the protein of Aspergillus niger was the same regardless of the source of nitrogen. Terroine, Wurmser and Montane (1922), working with the same mold, found little difference in the nitrogen content of the mycelium when the source of nitrogen was ammonium sulfate, peptone, urea, or sodium nitrate; but it dropped decidedly with guanidine. Perhaps it is the proportion of the different kinds of proteins within the cell which brings about the differences in different yeasts. A well-known illustration of the influence of the kind of food upon the final amount of growth is the appearance of pinpoint colonies on agar. Many milk streptococci give such colonies on standard, i.e., sugar-free agar, while the colonies on glucose agar are of normal size. SUMMARY The final number of cells which can develop in a given volume of medium depends to a large extent upon the chemical composition of the food. With different types of material for cell construction, yeast will produce not only different quantities of cells, but also cells of considerable variation of physiological characters. (c) ENDPOINTS WITH AN ABUNDANCE OF FOOD In all these experiments, the source of energy was used completely; the nitrogenous building material was used almost completely (see Exp. 22-23 of Table 65). It is generally taken for granted that a poorer food, even if given in abundance, will produce a smaller crop. The facts seem to bear out the assumption. Really, the situation is not as simple as it seems at first glance. Our easy acceptance of the facts is probably based upon an analogy (perhaps subconsciously) with higher organisms. A poor soil will GROWTH 245 give a small crop, but a poor soil is either a fair one deficient in plant food, or one which contains compounds injurious to growth, such as acid, or alkali. Such poor soil is not comparable with our case, which allows liberal amounts of a poor food. With animals, a large supply of poor food will bring about under- sized growth. But the reason is different. The capacity of the stomach and intestine is limited, and if they are filled to capacity with a food low in nutriment, the cells of the growing animal do not obtain all the nourishment they could utilize. The body does not develop to its maximum size, because the rate of growth is abnormally slow, and sexual maturity develops before the normal size is reached : After this stage, there is hardly any more growth, even with better food. This analogy applies to our problem only if we assume that there is such a thing as an adult stage in bacterial cultures. The compari- son with animals can be carried quite a long way. In this case, we would have to assume that each cell, regardless of its rate of growth, secretes into the medium some substance which, in sufficient concentration, will prevent further growth. In a well nourished culture, this limiting concentration is reached only after perhaps one billion cells per c.c. are formed. Poorly nourished cells secrete this harmful compound at the same rate, and the Hmiting concen- tration is reached long before the more slowly growing bacteria have had the chance to reach the billion mark. This explanation would be satisfactory if the existence of this, peculiar secretion could be proved. This seems possible; a full grown culture fed on ammonia as the only source of nitrogen should then show no further growth if asparagin or peptone were added. No such experiment is known to the author. If, however, there is new growth, then the cause of cessation of growth cannot have been a secretion; it must be in the cell mechan- ism. Then, the cell can use only a Umited amount of poor food, though there is plenty of it available, but will continue to grow with a better food, i.e., if the cell works under higher pressure, or with higher intensity. Only an experiment can decide between these two possibilities. (d) INFLUENCE OF THE CONCENTRATION OF FOOD We must again distinguish between energy-producing food and building material. To the energy-producing 246 PHYSIOLOGY OF BACTERIA foods belong not only carbohydrates and the like, but also oxygen. Attention has already been called to the experiments of A. J. Brown (p. 233) proving the oxygen concentration to be quite essential for a large crop of yeast. Of the large amount of literature on the yields of yeast under varying conditions, very little can be used here, because it is cus- tomary in the yeast industries to record the crop in relation to the seeding, and not in absolute values; it is stated, e.,g. that the yeast multiplied five to eight times. This method of measurement is meaningless except for the direct comparison of two experiments ^ 4% 3u.qt S% SttqcLr- ar Z% Suf€er /U Suyocr ti\ i/rie Fig. 24. — Theoretical growth curve of yeast with abundant building material and increasing amounts of sugar. with the same seeding in the same medium. Even the scientists resort to this method when working on yeast problems (A. J. Brown, Lafar, Henneberg). The theory of the influence of food concentration is fairly simple. Building material that does not contain a source of energy will produce no growth; nor will a source of energy that contains no building material induce growth; both are needed. GROWTH 247 The rate of energy supply is constant as long as there is a fair amount of substrate (e.g., sugar) available. With a low amount of energy food, and sufficient building material, there will be a definite growth. If the energy food is increased, growth will not be more rapid, but will continue longer, because there is an energy supply for a longer time, which will ultimately produce a larger crop. Assuming sufficient building material, further increases of energy food will bring about further increases of the crop as is illustrated schematically in Fig. 24. In actual practice, the accumulating fermentation products or cell secretions will retard the growth rate after a while, and will make the utilization of the food less effective. One of the most commonly quoted experiments is that of Rubner's (1906a) with BacL vulgar e grown in different concentrations of Table 68.— Yeasi Crops Obtained with Different Sugar Concentrations from the Same Building Material Increase for After 7 days After 10 days each additional Glucose Glucose % sugar offered, not used, % % n Nin ^^°P' crop, mg. *^' ^ mg. Crop, mg. Nin crop, mg. Yeast, mg. Nitro- gen, mg. 0 0 40 2.8 1 0 94 8.1 54 5.3 3 0.045 166 12.9 . . 36 2.4 5 0.14 226 17.8 30 2.5 7.5 0.24 280 20.0 . . . 22 0.9 10 0.42 314 21.6 14 0.5 12.5 0.75 352 24.6^ 311 23 0 15 1.2 15 0.98 382 27.5 326 23 8 12 1.1 20 7.48 248 16.7 322 25 2 25 12.90 211 14.8 30 19.60 160 12.2 248 PHYSIOLOGY OF BACTERIA meat extract. Rubner determined the nitrogen assimilated by the bacteria cells as measure of the crop. His quite complicated method of determining the nitrogen of the bacteria gave rather improbable data, and his results are not only contradictory within themselves, but also with the work of others who found the law of diminishing returns established in bacteria. An experiment very similar to the theoretical discus- sion above is that by Stern (1901) who grew yeast in a mineral solution with 0.3% asparagin and increasing amounts of glucose. Table 68 gives the crops from 100 c.c. of the medium. The yeast can grow without sugar. Addition of sugar increases the crop, but the increase per gram of sugar becomes smaller and smaller as the food concentration increases, and above 15%, no further gain is observed. Table 69. — Cells per Standard Loop of Micr. pyogenes in Meat Extract Relative concentration of meat extract 10 25 50 75 100 Start 868 3,566,000 4,273,000 3,392,000 868 6,296,000 9,336,000 9,416,000 5,736,000 868 8,880,000 13,288,000 15,760,000 13,696,000 868 10,664,000 16,960,000 20,300,000 20,864,000 17,936,000 868 22hrs 40hr3 64hr3 88hrs 11,808,000 18,960,000 23,424,000 25,840,000 112 hrs ... 22,816,000 Another frequently quoted experiment is that by Graham-Smith (1920) who grew Micr. pyogenes at 37°C. in meat extract of varying concentration. He used the extract of 400 gm. meat in 1 liter of water, filtered, without any other addition except neutraliza- tion. Compared with standard nutrient broth, this is a rather dilute medium. This meat infusion was diluted GROWTH 249 with varying amounts of water. In Table 69, the original diffusion is given the relative concentration of 100. ^^^^ ^^-..^ A r^ 1 /6 I / y\ X ■lit v\ /^ ^"'^'N ^50 // u "\ >ZS w. .2 H 0 i s 8 nours Fig. 25. — Growth curves of Micrococcus pyogenes in different concentrations of meat extract (the strongest concentration bearing the relative value 100). The outstanding result is that in the more concentrated solutions, growth still continues when it has passed the maximum number (marked by heavy print) in the lower 250 PHYSIOLOGY OF BACTERIA dilutions. This is entirely in agreement with the theory (p. 246) and is graphically shown in Fig. 25. The total crop at the different concentrations shows the following relative values : Relative con- Maximum num- ber of cells per standard loop, millions Relative numbers of cells centration of meat extract Highest concen- tration = 100 Lowest concen- tration = 10 100 75 60 25 10 25.8 20.9 15.8 9.4 4.3 100 81 60 36 17 61 49 37 22 10 There is a general parallelism between food and crop, but it is not a direct proportion between the two. While the food concentration drops from 100 to 10, the crop drops only to 17. Or, while the food concentration is increased from 10 to 100, the crop increases from 10 to only 61. This is the well-known '4aw of diminishing returns." It is quite evident also in the yeast experiment described in Table 68 where the increase in growth became smaller for each additional gram of sugar. There seems to be, for some bacteria at least, a limit of food below which they do not multiply at all. Car- penter and Hucker (1927) give a number of such exam- ples. Peltier's data (p. 207) also show a very decided drop when the milk is diluted farther than 1:16. The problem of growth in very dilute media does not seem to have been investigated to any extensive degree though the bacteriology of surface waters, city water supplies, and even of soil would be greatly benefited by such an investigation. GROWTH 251 SUMMARY With increasing amounts of food, the crop, i.e., the final number of cells in a culture, usually increases. With a liberal supply of building material, increase of energy food will cause no acceleration in growth, but a longer growth period, and, therefore, a larger crop. This is limited by accumulation of fermentation products or of inhibiting cell secretions. With all food components increasing, the crop will still be limited by fermentation products or cell secre- tions. The law of diminishing returns becomes quite evident. Many species require a fairly well defined minimum concentration of food for growth. (e) TEMPERATURE AND ENDPOINT OF GROWTH Our knowledge of the growth mechanism is too limited to permit general predictions regarding the influence of temperature changes upon the final crop. For tem- peratures above the optimum, it is easily seen that the crop must decrease with increasing temperature. This is verified by the experiments of Ward (p. 216). In those cases where growth is limited by accumulation of fermentation products, a prediction seems possible. Since the fermentation products act as poisons, the same concentration will act more strongly at higher tempera- tures, or a smaller amount at high temperature will have the same inhibiting effect as a larger amount at a lower temperature. Consequently, a smaller crop should be expected at higher temperatures. Where lack of food is the limiting factor, our basis for prediction is much less sound. The preceding 252 PHYSIOLOGY OF BACTERIA chapter showed a better utilization of food when the concentration was low. If this should mean that a slower fermentation causes a better utilization, we should expect a larger crop at the lower temperatures. How- ever, there is probably a limit to the advantage of slow fermentation. Where some specific compound (autotoxin or whatever it may be called) is the limiting factor it may be supposed that it will act similarly to the fermentation products. But because it is a specific compound, the amount pro- duced may have a temperature law of its own. Summarizing, the general prediction tends to the assumption that the crop will decrease rapidly as the temperature rises above the optimum, and will increase Table 70. — Influence op Temperatuee upon the Micr. pyogenes in Meat Extract (Cells per standard loop) Crop" of 17°C. 27°C. 37'*C. Start After 1 day. After 2 days After 3 days After 4 days After 5 days After 6 days After 7 days After 8 days After 9 days 632 2,064 73,850 3,120,000 8,128,000 12,400,000 14,512,000 17,072,000 18,272,000 17,592,000 632 5,528,000 10,918,000 12,568,000 14,288,000 15,918,000 15,664,000 13,568,000 632 5,968,000 8,744,000 10,448,000 9,968,000 8,688,000 slowly as the temperature falls below the optimum. There is probably a limit to this increasing crop with decreasing temperature ; we would not expect a maximum crop at the minimum temperature. Very few data are available on the influence of tem- perature upon the maximum number of cells. GROWTH 253 Two good experiments have been given by Graham- Smith (1920). Table 70 shows the slow growth to high numbers at low temperatures, and the rapid growth to to less high numbers near the optimum. In another experiment, the maximum numbers were reached at 33°C. after 4 days 10,610,000 27°C. after 4 days 18,671,000 At 8-10°C., no growth occurred, and a very gradual decrease of the numbers of living cells was observed. Henneberg (1926, p. 179) found that yeast will give larger crops by weight at low temperatures, but that the total amount of yeast protein formed is higher at high temperatures. The following data are given: a. Multiplication at 20°C.: 6.4 fold, at 30°C.: 9 fold Culture Then without aeration Multi- pUcation % pro- tein in yeast Total N in yeast For At m gm. 30° for 8 hrs 12 hrs. 12 hrs. 24 hrs. 24 hrs. 30°C. 10°C. 30°C. 20°C. 7.4 fold 7.1 fold 8. 8 fold 8. 3 fold 57.8 47.7 51.2 50.3 1.01 8 hrs 0.88 22 hrs 22 hrs 1.01 0.98 19° for 24 hrs 24 hrs 24 hrs 24 hrs. 24 hrs. 24 hrs. 19°C. 12°C. 2.5°C. 8.4 fold 8. 4 fold 8. 4 fold 45.8 43.6 42.3 Not quite comparable with these data is the experiment by Stern (1901) who obtained the following final yeast crops in 100 c.c. of a 10% glucose solution with 0.3% asparagin and minerals (Table 71). 254 PHYSIOLOGY OF BACTERIA The yeast cultures were not aerated. At 21°C., the maximal crop is reached, and also the maximal produc- Table 71. — Yeast Crops at Different Temperatures Temperature, Time of growth, Sugar not used, per Crop weight. N in crop, Per cent N days cent mg. crop, mg. in crop 12 13 1.47 252 19.4 7.7 15.5 16 0.36 291 19.8 6.8 18 12 0.47 337 22.1 6.6 21 9 0.92 339 22.4 6.6 25 7 0.44 310 21.2 6.9 30 5 0.56 243 17.3 7.2 37 3 4.01 110 8.5 7.8 tion of yeast protein, though this yeast contains less nitrogen than that grown either at higher or at lower temperatures. The temperature of the maximal crop is a good deal lower than the optimum temperature of growth. SUMMARY The temperature which gives the largest crop is con- siderably lower than that giving the most rapid growth rate, but it is still far above the minimum temperature of growth. An increase above the optimum temperature causes a very rapid decline of the crop. The temperature of the largest crop weight and that of the largest amount of protein in the crop are not exactly the same. (/) ENDPOINT OF GROWTH AND CHEMICAL STIMULANTS It is generally understood that the final growth in any culture will depend not only upon the amounts and GROWTH 255 kinds of food offered and upon the temperature used, but eventually also upon other factors, e.g., upon chemical substances which might increase or decrease the crop, or upon physical factors such as high osmotic pressures in concentrated salt or sugar solutions. The most common factor in chemical respects will be the acidity, or more correctly for most cases, the hydrogen ion concentration. Frequently, acid is a product of fermentation, and these cases fall under the general heading of the influence of fermentation products upon growth. The same holds true with alkali formers. Some organisms, like many yeasts and molds, have the abihty to adjust the medium to their optimum acidity, i.e., they can increase the acidity if sugar is present, or eventually decrease it by complete oxidation of the acid. These are not included in the treatment of this chapter. Growth comes to an earlier end if toxic substances are present in the medium. This is readily understood. It merely means the addition of one more unfavorable factor. But with the same substances which decrease the crop, an increase is often observed if the concentra- tion of the toxic compound is chosen correctly. In other words, the stimulation of the rate of growth by small amounts of poison, as shown p. 227, may result also in a greater crop. Of the older research on chemical stimulation, the papers by Koch (1912) and by Fred (1911) are of special bacteriological interest. The most comprehen- sive modern paper on this topic is that of Hofmann (1922). By dosing the toxic compound exactly right, Hofmann obtained not only a larger number of colonies from the same amount of culture, but the colonies were also larger. Of the six experiments recording the size of the colonies, only one shall be reproduced here (Table 72) . 256 PHYSIOLOGY OF BACTERIA Assuming the colonies to be spherical, the total volume of all on any plate is obtained by multiplying the number of colonies in each group with the corresponding volume, and adding all volumina. The result thus obtained is given in the lowest line of Table 72. The control, without silver nitrate, had produced a total volume of 171 mm. 3 of bacteria. The maximum with AgNOa was 755 mm.^ or 4.4 times as much. Since these last figures represent, at least approximately, the volume of all bacteria produced in the same amount of agar from the same amount of seeding, the only difference being the presence of 0.00075 % AgNOs in the one medium, a great influence of chemical stimulation upon crop production of bacteria cannot be denied. Table 72. — Numbers and Sizes op Colonies of Bad. typhi murium, ON Agar + AgNOa AgNOs added 0 0.0005% 0.00075% 0.001% 0.002% 0.01% Diameter of colonies, in mm. Number of colonies of the various diameters 1.030-1.140 1.140-1.250 1.250-1.370 1.370-1.480 1.480-1.600 1.600-1.710 1.710-1.820 1.820-1.940 1.940-2.050 2.050-2.170 2.170-2.280 12 34 26 26 7 4 4 8 23 24 41 28 14 5 5 12 19 34 61 48 19 13 5 13 13 27 54 16 17 6 2 6 14 16 29 55 16 18 3 7 10 21 23 8 7 7 Total colonies 113 143 211 153 157 83 Total volume of bac- terial colonies per plate, in mm.». . . . 171 218 755 359 352 157 Hofmann mentions a large number of substances which acted as stimulants in similar experiments. Among them GROWTH 257 are HgCl2, CuCh, ZnCl2, AS2O3, H2Cr207, HCHO, CeHsOH, lysol, atropin, saponin, malachite green, methyl orange, methylene blue. With some toxic compounds, however, no stimulation could be noticed in any concentration; among these were FeCls, KMn04, iodine, acetate of lead and of aluminum. Hofmann's data show further there markable fact that the number of colonies from the same amount of bacterial suspension is increased by certain doses of poison. This can mean only that certain cells too weak to reproduce on normal agar (see p. 271) can be stimulated by a small dose of toxic substance sufficiently to produce normal colonies. A general explanation for the stimulating action of chemical poisons has been given on p. 137. The evidence given here is quite sufficient to show that even through a large number of generations, chemical stimulation does not necessarily mean a final detriment to the cell. V. MECHANISM OF GROWTH (a) MOLECULAR STRUCTURE OF THE CELL Growth, chemically speaking, is essentially a process of dehydration and of reduction. Doubtless, breaking- down processes such as hydrolysis and oxidation will also take place, but the outstanding, dominant problem is that of construction either by removal of water, or by reduction, or by other means. While the preceding pages have shown how the growth mechanism reacts upon certain environmental conditions such as concentration of food, temperature, kind of food and stimulants, the mechanism as such has not been discussed. In this chapter, an effort is made to obtain a conception of it. The difficulty of a clear understanding 258 PHYSIOLOGY OF BACTERIA lies primarily in the fact that the problem is not only one of physics and chemistry, but also one of cytology, of systematic arrangement. We cannot understand growth without first under- standing the energy forming processes, for we can hardly imagine growth without a supply of energy. The liberation of energy in bacterial cells is fairly well understood, comparatively speaking. The form of energy made available through cell activity is unknown. It may be heat, or it may be radiant energy; or we may call it ^'chemical energy" which is merely a good word for an unknown form. Number of Enzyme Molecules per Cell. — While the form of energy is unknown, data on its amounts are available. The average cell of Streptococcus lactis, during its period of most rapid growth at 20°C., ferments about 12 X 10~^^ mg. of sugar per hour (p. 106). The weight of one molecule of glucose is g-^^3gm. = 30XlO-2«mg. Therefore the number of molecules of sugar used per hour by a single cell is 12 X 10-10 30 X 10-20 0.4 X lO^o or four billion molecules per hour, or about one million molecules per second. This number is so large that we must consider the fermentation to be for all practical purposes a continuous process. The number of enzyme molecules required to handle one million sugar molecules per second is not known, but it must be fairly large. Most cells can use more than one source of energy, e.g., streptococci and yeasts can grow without sugar. GROWTH 259 and therefore, they must have different sets of enzjones for each source, and probably a fairly large number of molecules of each kind. By applying this same method of calculation to the data on p. 105, it can be shown that a yeast cell will ferment about fifty million glucose molecules per second, and Micrococcus ureae as many as eighteen million urea molecules per second. Number of Growth Catalysts per Cell. — The real growth mechanism of the cells is centered in the chromo- somes. These are the growth catalysts. They are believed to consist of sub-units, called genes, which are beyond microscopic visibility. Each gene is supposed to be the catalyst for one specific reaction or property in the organism, and therefore, each gene must be chemically different from all other genes (unless we are dealing with multiple sets of chromosomes). The size of the genes has not been measured. We know only that it is below the dissolving power of the microscope, i.e., smaller than 100 m/z. The size of the molecules of which the chromosomes are composed is not known either, but of many other proteins, the molecular radius has been determined by the Svedberg method. The author is indebted to Dr. D. C. Carpenter, of the Geneva (N. Y.) Experiment Station, for the compilation of Table 73. The proteins marked with a star are known to have nearly spherical molecules. The diameter of these more common proteins goes as high as 24 m^. The molecules of the chromosomes are much more remarkable than these, because they control the most complicated synthesis of the many different cell substances from simple food. It is probable (though it is not necessary) that molecules which bring about the synthesis of protein are more complex than 260 PHYSIOLOGY OF BACTERIA protein molecules. It seems permissable to assume that the molecules of which the chromosomes are composed, are of the same order of magnitude as the genes, and that a gene consists of only a few very large molecules, perhaps only of one. Table 73. — Size of Various Protein Molecules Calculated from Sedimentation Velocity Measurements Protein Molecular radius in m/i Bence- Jones* Egg Albumin* Insulin* Hemoglobin Serum Albumin Serum Globulin Casein 2X Amandin* Edestin* Excelsin* Legumin* Phycocyan Phycoery thrin * Casein 4X Hemocyanin (Limulus). Hemocyanin (Helix)*.. 3.94 3.94 3.96 3.96 3.94 3.95 4.18 6.96 12.00 * Molecules nearly spherical. The enzymes of dissimilation can be produced by the cell; a cell can control its enzyme content to a certain degree (p. 131). It seems, however, that a gene can be produced only by another gene of the same kind. If a gene is destroyed, the cell cannot usually replace it, and if such a cell can divide at all, the two daughter cells will lack this gene and will be abnormal. More com- monly, the cell will not divide. Chromosomes have never been observed in bacteria. There is some indirect evidence for their existence, however. Lacassagne GROWTH 261 (1928) exposed bacteria to X-rays and observed that a cell could be killed by being hit by one electron, but only if it was hit in a definite part of the cell which he called the sensitive zone. WyckofT (1930) computed the volume of this zone from experiments with X-rays of various wave lengths and found it to be not more than one- hundredth of the cell volume. Nothing can be said about the form or distribution of this sensitive zone in the cell, but the fact of its existence suggests strongly a special segregation of the growth function corresponding to the chromosomes of larger cells. Arrangement of Molecules in the Cell. — A computa- tion of the ^ intensity of energy" or the ^'potential of energy" will permit us a further insight into the rela- tion between the assimilative and the dissimilative mechanism. It has just been mentioned that the average cell of Strept. lactis will ferment about 12 X 10~^^ mg. of glucose per hour. The heat liberated in the lactic fermentation of 1 gm. glucose is 82 calories (p. 23). The sugar fermented by a single cell produces 12 X 10~^° X 82 X 10-3 = 984 X 10-12 calories. Taking the weight of this cell to be 5 X 10"^^ mg., the energy liberated in one hour is sufficient to heat the cell to 200°C. Considering the enormous surface (about 60 square centimeters per mg. of moist cells, p. 397), radiation and conduction must play a great part in relieving the cell of this amount of surplus energy. Possibly, the energy in the cell is not liberated in the form of heat, but in some other, more harmless form. If the entire cell assumed a higher uniform temper- ature, there would be no potential inside the cell. In order to get a conception of the possible energy potentials within the cell, the heat produced per molecule of sugar fermented must be calculated. A single glucose molecule weighing a ^ ^r)23 S^-y ^^ 262 PHYSIOLOGY OF BACTERIA 180 X 82 being fermented to lactic acid, will liberate n ^ -.rs23 calories = 2,460 X 10"^^ cal. If this energy, in form of heat, would distribute itself only upon the two lactic acid molecules formed whose weight would be §~^, = 30 X 10-2' gm., these two molecules would show a rise in temperature of 82°C., (assuming, for simplicity's sake, the specific heat of lactic acid to be 1). But the enzyme which brought about the change of sugar to acid must be directly connected with these molecules, and must receive its share of the energy. If this energy is to be used at all for growth, it must be used at once, before it is radiated or conducted away. A potential of energy is probably required to bring about any synthesis in the cell. Synthesis cannot be accom- plished by a uniform rise of temperature of the cell. To utilize the energy of this potential established at the ''molecule tips" of the enzyme, the synthetizing mecha- nism (such as the chromosomes, and probably other preliminary catalysts as well) must be in close proximity to the source of energy. It does not seem probable that energy could be sent only in one direction, as a beam, to the mechanism of synthesis. Our knowledge of the growth process would be greatly benefitted, if we could get a conception of the number of different chemical processes necessary to change the food to the compounds which make up the cell wall, the protoplasm, and the reserve substances of the cell. Many different steps would be necessary, but with a given building material, it should be possible to estimate the number of steps approximately. It would then be possible to sort out these different steps in groups of identical or homologous reactions, such as the condensation of formaldehyde and of acetaldehyde. It is imaginable that these GROWTH 263 two different, but homologous processes might be accomphshed by the same part of the cell mechanism. Thus we could get a con- ception of the number of different catalysts needed for the growth process. Remembering the enormous surface of bacteria, it seems surprising that intermediate construction products have never been observed with bacteria, yeasts, or molds. The construction of cell matter from building material appears to be a continuous process comparable to the moving belt method of automobile manufacture. The molecules singled out for building stones have no chance to escape or to get side-tracked. Once a certain CO2 molecule has been drawn by a nitrate bacterium into the growth machinery, it cannot leave the cell; it is destined to become a part of cell wall or protoplasm. This is perhaps the most remarkable part of the entire growth process. The absence of partly synthetized compounds in bacterial cultures makes us believe that in one continuous process, the molecules are changed at least to a state where they do not diffuse through the cell wall. This would require a very definite arrangement of the molecules in the cell. There are the molecules of the energy producing zymase, present in quite large numbers. These enzyme molecules are probably not scattered around by chance, but are located at some definite parts of the cell, or at least in definite position relative to the synthetic catalysts. The molecules of these catalysts are not all of the same type; there are different kinds for different synthetic processes. These catalysts must be arranged in some definite order so that the molecule under construction can be ''passed on." Cytologists have found from experiments on inheritance that each gene has a definite location in its chromosome. All of 264 PHYSIOLOGY OF BACTERIA these necessities make the cell appear as a factory well arranged after a carefully thought-out plan. This definite arrangement is exhibited in the cells of higher plants by the symmetry of the chromosomes in cell division. Still, there must be considerable allowance for individuality. There will be different degrees of complexity in the different types of bacteria. A prototrophic bacterium which builds its cells from CO2 and HNOsmust go through more steps than a cell which uses sugar and amino acids. The higher the claims of an organism are in respect to its food, the simpler will be the growth apparatus, the less will be the number of synthetic steps. We can imagine the food requirements to become so highly specialized with certain parasites that the food is almost identical with the protoplasm of the host. The growth mechanism of these organisms would be extremely simple. They would simply take part of the host protoplasm, and incorporate it, after a few simple changes, into their cells where it will be the parasite's protoplasm. Such parasites could not have a cell wall; it would prevent the direct ^^ intake" of such large molecules. It would seem possible that these organisms might consist of just one giant molecule. This might fit the case of filterable viruses and bacteriophages (Rahn, 19316). The knowledge of the filterable forms of well-known bacteria is still too scanty to be included in this discus- sion, but entirely new viewpoints are likely to be introduced in our conception of the cell mechanism by these studies. (6) THE GROWTH PROCESS After having considered the probable arrangement of the most essential molecules in the cell, we return to the GROWTH 265 original problem, the chemistry and dynamics of the growth process. Some information might be obtained from the obser- vation that growth is hardly accelerated by increases of concentration of either substrate or building materials while it is greatly influenced by a change of the substrate, or by a change of temperature. Bad. typhosum in peptone solution had a generation time of about forty minutes, with 0.4% peptone as well as with 1% peptone (p. 204). But with the addition of glucose, the growth rate doubled. Bad. coli in 1% peptone solution showed a generation time of thirty-five minutes and quadrupling this amount gave only 13% increase in the growth rate; the same effect which this additional 3% of peptone produced, could be obtained by as little as 0.1% of sugar (Curran, 1925). Quite different is the effect of temperature upon growth. It influences fermentation, increases the energy yield per unit time. Concentration of food cannot do that. A certain food will give a definite amount of energy. The enzymes in the cell can ferment only a limited number of substrate molecules. For Strept. ladis at 20°C., this number was estimated to be about 1,000,000 molecules per second. Increasing the sugar concentra- tion will not cause more molecules to be decomposed. The enzymes are working with maximum capacity. But an increase in temperature brings about the decom- position of more molecules by the same number of enzymes; the mechanism is accelerated; everything moves more rapidly. A change of substrate changes the rate of energy formation. We have no means of telling exactly what happens when a streptococcus cell is changed from glucose 266 PHYSIOLOGY OF BACTERIA broth to sugar-free broth. The only source of energy then is some peptone or amino acid. The available amount is less per molecule of substrate. How many molecules per second of the new substrate can be utilized by the cell has not been determined as yet. Doubtless, the enzyme furnishing energy from peptone is a different one from that which ferments sugar. These new enzymes must also be arranged in definite position to the synthetic catalysts in order to make them profit from the energy potential. We saw that Bad. typhosum grew twice as fast with sugar + peptone than with peptone alone. It may be that sugar gave a larger amount of energy per molecule. It may be that the energy yield per molecule was not greater, but that more enzyme molecules of this type were present, producing a larger total of energy per unit time. It may be that even the total energy per unit time was not greater, but that the enzyme attacking the sugar is situated in a more favorable position in the cell, closer to the catalysts, while with the more distant protein-attacking enzymes, which produce perhaps more energy, most of the energy is lost by conduction or radiation. By selecting the right kind of organisms and of energy sources, and by applying calorimetry as well as chemical analysis, these questions could be decided to a certain extent. The probability of obtaining a better understanding of the growth processes is greater with bacteria than with any other experimental organism. (c) VARIABILITY OF THE GROWTH RATE Starting from the facts that in most cells, there is only one gene of each kind per cell, and that all genes must have doubled before the cell can divide, the author (1932) developed a theory of the variability of the growth rate. If a large number of uniform cells (i.e., cells with the same kinds and numbers of molecules in the same arrangement) are kept in a uniform, favora- ble environment, there are two possibilities for the order GROWTH 267 of their growth: either they all multiply at the same moment, or their reacting molecules follow definite chemical rules, and the order of multiplication is governed by these laws. The latter case is more probable, and the theory is supported by the facts. It may be assumed that the molecules in a number of uniform cells react as if all cell contents were one con- tinuous fluid. This assumption seems permissible at least for the molecules of the genes because genes do not react with one another. In this case, the doubling of any one certain type of gene molecule must follow the mass law, and as there is only one such molecule for each cell, it can be computed in what order this gene in the different cells will double. It can also be computed in which order any of the other genes, or each molecule of each gene, will double. From this, the probability that all genes in a cell have doubled, can be ascertained. This probability gives us the variation of the rate of cell division. The calculation has been carried out for various numbers of genes (or molecules) in the chromosomes. Table 74. — Frequency of Fission Times of Bacterium Percents aerogenes in Minutes | Over 5 10 15 20 25 30 35 40 45 50 55 60 60 I. 323 observations II. 168 observations III. 242 observations 0.6 1 0.3 2.4 3.4 4.1 17.0 7.8 18.5 48.0 13.030.0 31.023.8 24.0 7.0 20.7 14.9 0.8 13.9 3.6 0.8 6.2 1.2 2.5 1.8 0.6 0.6 2.1 Figure 26 represents one such series. It shows that uniform cells under uniform environment will not all divide at the same rate, but that for chemical reasons, i.e., on account of the mass law, they must display con- 268 PHYSIOLOGY OF BACTERIA siderable variability of the growth rate, some of them dividing four to five time units later than the others. This computation is an over-simplified case, making several assumptions which are not probable. For example, this computation assumes, that all genes multiply at the same rate, and that a cell divides as soon as the last gene has doubled while we know that, with higher organisms at least, some time elapses after chromosome division before cell division occurs. / 2. Z t 5 6 7 8 9 fO ti 12. a f¥ tS /6 Fig. 26. — Percentage of cells dividing in successive time units computed for g = 0.5, and for various numbers of genes, g. The observed facts (Kelly and Rahn, 1932) agree with the theory in a general way. Bad. aerogenes as well as Sacch. ellipsoideus showed a great variability of growth rate. Table 74 shows three series of data with Bad. aerogenes. The frequency curve is skewed to the left as in the theoretical curve of Fig. 26. The results with yeast corresponded to the data computed for a larger number of genes while Bad. aerogenes appears to have a smaller number. A special investigation showed that the offspring of the most rapidly multiplying individuals was not rapid, but average, and the same was true with the progeny of the slowest cells. GROWTH 269 With colonies of Bad. aerogenes, the variability decreased as the size increased. This is due to the fact that fast and slow reproduction is a matter of chance, and not an inherent quality of the individual cell. The more generations we observe, the smaller will be the probability of the average being influenced by chance. With mammals, the gestation period shows a relatively small variation, and no asymmetry. id) SUMMARY Fermentation and respiration is brought about by catalysts (enzymes) of which each cell contains a con- siderable number. They can be regenerated by the cell. Multiplication is brought about by very many different catalysts (genes), each cell containing usually only one of each kind. A gene, once lost, cannot be regenerated by the cell. The molecules which bring about growth, i.e., synthe- sis, must be arranged in the cell in a definite position to each other to prevent the loss of partly finished build- ing material. The enzyme molecules that cause fermentation and liberation of energy must be in a definite position to the growth-producing catalysts to prevent loss of energy by radiation or conduction. By making several simplifying assumptions, it can be shown that uniform cells in uniform environment will not all divide at the same time, but must have a definite frequency curve of the growth rate which is asymmetric and askew. The theory demands that the progeny of rapid and of slow growing individuals should have the same average growth rate, and that cells with only a few genes should vary more than those with many genes. All of these demands are borne out by the experiment. PART D MECHANISM OF DEATH I. DEFINITIONS OF A DEAD CELL Death has been defined in the introductory chapter as a function by itself; it consists not merely in the absence of fermentation and growth, but is a chemical process. We might perhaps define it more accurately by saying that dying is a chemical process which becomes irreversi- ble after a while; from this time onward, the cell is considered dead. The customary conception of dead bacteria is that they can not grow on the agar plate, nor in broth; in other words, that they have permanently lost the power of reproduction. This definition is different from those of other biological sciences; it is distinctly a bacteriologi- cal one. Cells which have lost the power to reproduce may still ferment; their protoplasm may still appear normal under the microscope, and may still be capable of plasmolysis. Such cells would be considered alive by most biologists. It may be well to realize from the very beginning that many definitions of a dead cell exist which might give cause for considerable confusion unless their meaning is kept in mind all the time. The bacteriological definition is not more exact than any of the others, as may be seen from the experiments of Supfie and Dengler (1916) who dried anthrax spores on silk threads, heated them in steam for various lengths of time, and then placed each thread in broth. They took dupUcate samples, and tested one in standard 270 MECHANISM OF DEATH 271 broth, the other in the same broth + 3% glucose + 5% horse serum. The results (Table 75) show that, according to the standard broth test, the spores of strain SM were apparently killed between four and six minutes, and of strain Og between ten and twelve minutes of heating. However, the test in the better medium proved that this conclusion was wrong, as some of the spores revived and ger- minated after twenty-five minutes of heating in the richer medium. The destruction of the spores by heat was the same in both SM and Og cases because the tests were made with duplicate samples. Life in the spores was suspended irreparably in standard broth after four to twelve minutes of heating, while in the better medium, some of these spores recovered, and the fatal stage was not reached until they were heated for about thirty minutes. Table 75. — Viability of Spores of B. anthracis, after Heating in Steam (+ means growth; 0 means no growth; — means not tested) Strain Test medium Heating time in minutes 1 2 3 4 6 8 10 12 15 20 25 30 40 Standard broth. . + + + + + + + 0 + 0 + 0 0 + 0 + 0 + 0 + 0 + 0 SM Same + 3 % glucose + 3 % horse serum . . 0 Standard broth + + — + + + + + + + + 0 + _ 0 + 0 + 0 0 Og Same + 3 % glucose + 3 % horse serum . . - From this experience, it may well be postulated that with a still better test medium, it might be proved that even forty minutes is not sufficient to really kill all spores. Similar experiments have been recorded in Hterature. Death, as defined above, is, therefore, a function of the test medium. It might be recalled here that in the earlier development of agricultural bacteriology, much time was spent on finding the best medium which would give the highest plate count in milk, or in soil. Since the numbers of colonies developing from the same sample on different media are quite different, some cells are counted as "living" on one medium while they are ''dead" on the other. 272 PHYSIOLOGY OF BACTERIA Attention may also be called to the observation of Hofmann (p. 256) that from the same dilution of a pure culture, a larger plate count might be obtained by adding small amounts of poisons like AgNOs or CuCU to the medium. According to our definition of death, the chemical stimulus brought some of the "dead" cells back to life. As long as a cell can recover from injury, the death process is reversible. It seems that the point of irre- versibility is not reached very soon after the process of dying begins. This becomes evident in chemical disinfection. Bacteria treated with a poison may remain incapable of reproduction even when all poison has been washed away from the cells. An antidote that combines with the poison which has already reacted with the protoplasm, may be able to make the cell grow again (p. 358). This discussion leads to the conclusion that our defini- tion of death means the fixation of one point on the continuous line of a gradual chemical process. Dying is a gradual time process, of measurable velocity, and may be reversible, after removal of the cause, during the first stage by the mechanism of the cell itself. Gradually, it becomes irreversible for the cell, but may still be made reversible by some outside influences such as antidotes. Finally it reaches a state where it becomes irreversible under any condition. This may be due to secondary chemical changes which could take place because some part of the cell was not functioning suffi- ciently to ward it off as in normal cells. II. METHODS OF MEASURING DEATH (a) The Endpoint Method. — The oldest method of measuring the power of a disinfectant was to determine the shortest time in which all bacteria added to a certain concentration of the poison would be dead. Naturally, MECHANISM OF DEATH 273 the time differs with the species used. The tuberculosis organism is extremely resistant to alkali; molds can tolerate more formaldehyde than most bacteria; spores are much more resistant than vegetative cells. The experiment is usually carried out by taking samples after certain intervals, transferring them to a good nutrient medium, and waiting for a certain length of time to see whether growth develops. The results obtained are of the nature shown in Table 75. The accuracy of this method depends upon the frequency of the time intervals. In the above example, the interval is relatively large; the span between four and six minutes is 33% of the longest time. The last one of the four experiments of the above table is more accurate. The correct death time hes between twenty-five and thirty minutes. The span within which we know the exact time to be, is only five minutes out of thirty, or 18%. This relative accuracy will prove important where the endpoint method is used for computing death rates. It has been known for many years that more time is needed to kill a larger number of bacteria. This is important for the technique of the endpoint method. In order to obtain comparable data, it is essential that the same number of cells is used for all data to be compared. If that proves to be impossible, the initial number of cells acted upon should be determined. It has become quite cus- tomary to count the initial number of cells. (b) The Plate Count Method.— The plate count for determining the efficiency of a disinfectant has been in use for a very long time. It has the advantage of showing more of the kinetics of the disinfection process than other methods, and requires a smaller number of samples, but more time, and more media; therefore it is used mostly for studies on the laws of death, and less frequently in applied bacteriology. (c) Staining Methods. — Most cells cannot be stained easily as long as they are alive, but take the dye readily 274 PHYSIOLOGY OF BACTERIA when they are dead. The bread yeast industry has made use of this fact for determining the dead cells in a batch of bread yeast, by suspending the cells in a counting chamber filled with a very dilute solution of methylene blue (Lafar V, 173). Fulmer and Buchanan (1923) applied this method to measure quantitatively the number of dead cells in a yeast culture exposed to some disinfectant. In using this method, it should be borne in mind that it means a new definition of death. Considering the death process as a continuous curve plotted against time, the point where the protoplasm begins to take the dye is probably not identical with the point where the faculty of reproduction is lost. Generally speaking, bacteria are too small to lend themselves to the staining technique. This applies also to most other methods for proving the death of cells in botanical or zoological experiments, such as plas- molytic reaction, loss of motility, response to stimuli, etc. The inaccuracy of the various staining methods has been pointed out by Bickert (1930). III. THE LOGARITHMIC ORDER OF DEATH (a) ORDER OF DEATH OF HIGHER ORGANISMS When large organisms, such as green plants, or vertebrates, or insects, are exposed to a detrimental influence, some time will pass before the first organism dies. They will not all die at the same moment; some will die soon, others will be more resistant. The order of death will be represented by a chance curve, due to the variation of resistance among the individuals. This can be shown by the order of death of seeds, or of fruit flies exposed to high temperatures. Table 76 MECHANISM OF DEATH 275 records two such experiments, the one with seeds being performed by Groves (1917), the other, with fruit flies (Drosophila), by Loeb and Northrop (1917). If we plot the curves of the survivors, we obtain the inverted S-shape, the typical curve for chance distribu- tion. If we plot the organisms dying per unit time as Table 76. — Order of Death by Heat Wheat seeds at 87.5°C. Fruit flies, at 39.45°C. Time in minutes Per cent Time in minutes Per cents Per cent Survivors dying per of dying each minute survivors interval 0 98 0 0 100 7 74 3 25 95.5 1.8 8 60 14 30 83.0 12.5 9 38 22 35 65.4 17.6 10 11 27 40 45.4 20.0 11 5 6 45 28.0 17.4 12 2 3 50 17.1 10.9 13 0 2 55 11.8 5.3 60 8.8 3.0 65 2.5 5.3 70 1.1 1.4 75 0.3 0.8 Table 77.-1,000 Mustard Seeds in 0.2% HgCla Time, min. Survivors Seeds killed in 15 minutes 1, a 30 940.0 (30) 0.0018 45 895.8 44.2 0.0016 60 790.6 105.2 0.0023 75 486.6 ^ 304.0 0.0052 90 220.6 266.0 0.0087 105 163.8 56.8 0.0087 120 146.0 17.8 0.0080 135 39.0 107.0 0.0117 276 PHYSIOLOGY OF BACTERIA Fig. 27. — Order of death of Bacterium typhosum exposed to 49", ^5" 35" ffS SS TninuieS fS* Fig. 28. — Order of death of Drosophila exposed to 39.45°. The black line in figures 27 and 28 demonstrates the numbers of survivors; the black blocks represent the organisms dying per unit time (1 and 5 minutes respectively). MECHANISM OF DEATH 277 2. H- 6 10 iS -TrUTijLaeS ^ Fig. 29, — Order of death of the spores of B. anthracis by HgCL. /df Fig. 30.— Order of death of mustard seeds by HgClo. The black line demonstrates the numbers of survivors; the black blocks represent the number of organisms dying per unit time (2 and 15 minutes respectively). 278 PHYSIOLOGY OF BACTERIA indicated by the black blocks (the unit being five minutes with fruitflies), we get a probability curve. (Fig. 28.) Quite similar curves are obtained when death is caused by chemical poisoning, by drying, or other outside influences. As example may serve (see Table 77 and Fig. 30) the death of mustard seeds from treatment with HgCl2 by Hewlett (1909). (6) ORDER OF DEATH OF BACTERIA With bacteria, the order of death is different. If the outside influence (heat, drying, chemical poison, light, etc.) is sufficient to kill the cells at all, the first unit of time will show the largest number of deaths. Madsen and Nyman (1907) and, independently. Chick (1908) were the first to recognize that the order of death of bacteria could be described quite simply by the definition that the rate of death is proportional to the number of living cells. In other words, the per- centage of the survivors dying per unit of time is constant during the experiment, regardless of the number Table 78.- —A Theoretical Case op Disinfection Time Survivors Dying per unit time Total dead 0 1 2 3 4 5 6 1,000,000 100,000 10,000 1,000 100 10 1 0 900,000 = 90% 90,000 = 90% 9,000 = 90% 900 = 90% 90 = 90 % 9 = 90% 0 900,000 990,000 999,000 999,900 999,990 999,999 of cells present. Let us assume a very simple case: we have 1,000,000 bacteria; the death rate is 90%; this leaves, after the first minute, 10% or 100,000 cells alive. MECHANISM OF DEATH 279 During the second minute, again 90% of all living cells die, which leaves 10% or 10,000 cells alive. In this way, the number of cells decreases as is shown in Table 78. If we compare the number of cells dying per unit time, with the examples given for seeds and fruit-flies, a difference in principle will be noticed. That this order is typical for most causes of death of bacteria, will be shown all through this part of the book dealing with death; the evidence regarding the order of death is also summarized in Table 84, p. 295. Only two experiments shall be mentioned here (Table 79), as parallels to the afore-mentioned experiments with multi- Table 79. — Death of Bacteria by Heat and by HgCl^ Bad. typhosiim at 49°C. Spores of B. anthracis in 0.5 per cent HgCU Per cent Spores Time, min. Surviv- ors of total dying per minute 1, a Time, min. Surviv- ors killed in 2 min- utes 1, a 0.28 2,008 0 9,500 1.00 1,198 56.0 0.311 2 4,860 4,640 0.146 2.05 925 13.6 0.190 4 2,964 1,896 0.126 3 755 18.5 0.156 6 1,408 1,556 0.138 4 542 10.6 0.153 10 304 552 0.149 5 488 2.7 0.130 15 2.6 120 0.204 7 289 4.9 0.125 20 1.8 0.3 0.186 10 112.8 2.9 0.129 25 2.0 0 0.147 15 24.2 0.9 0.130 20 3.0 0.1 0.143 cellular organisms, one on the death of Bad. typhosum by heat, by Chick (1910), and another on the death of spores of B. anthracis by mercuric chloride, by Madsen 280 PHYSIOLOGY OF BACTERIA and Nyman (1907). The fundamental difference in the order of death is well shown in the survivor curves of Figs. 27 and 29 as well as in the black blocks represent- ing the number of organisms dying per unit time. (c) MATHEMATICAL FORMULATION OF THE ORDER OF DEATH The regular and simple order of death makes a general mathematical treatment possible. The initial number of cells shall be called a, and the number still living after an exposure for the time t shall be h. Since the rate of death at any time is proportional to the number of survivors, this can be written in the form dt -? = Kdt 0 Integrated, this expression changes to -\nh = Kt + C for ^ = 0, we have h = a, and consequently -In a = C We substitute this value for the integration constant into the preceding equation and obtain — In b = Kt — In a K' t = In a — Inh K - 1 log Q^ - ^Qg ^ ^ ~ t 0.434 This last expression gives the formula for the death rate, K^ for decimal logarithms. If the number of cells dying per unit time is at all times the same percentage of the number of living cells, K must be constant. The MECHANISM OF DEATH 281 computation of i^ in Table 79 shows that it fluctuates around the average value of K. In the experiments with mustard seeds (Table 77), the death rate is com- puted by the same formula; it is not constant, however, but increases continuously. This is sufficient evidence that the death follows a different law. 3,0 % t 4^ 2.0 1.0 K ^0 0 JV ^<. ^< \ \ \ \ \Sh' 4 \ ( 1 \ 10 Fig. 31. 20 30 minutes 60 -Logarithms of survivors of anthrax spores plotted against the time . of exposure to different temperatures. The computation of the death rate will prove very useful when we try to understand the mechanism of the processes bringing about death. It is practically impossible to define K in words that would mean anything more than a description of the mathematical formula. Even the relation between this death rate and P, the percentage of cells surviving per unit time, is not very simple. We find P = 100 ]^Q0.434X If we wish to compute the number of survivors b for any time, we have the equation log 6 = log a - OASAKt 282 PHYSIOLOGY OF BACTERIA log a is constant within the experiment. Therefore, we have log 6 = C - 0.434Z^ which means that the number of survivors is a logarithmic function of the time. For this reason, this regularity of death is often referred to as the ^logarithmic order of death." If we consider log 6 to be a variable, instead h itself, we get the general formula of a straight line y = C - 0Ad4:Kt This means that if we plot the logarithms of survivors against time, all points will be on a straight line. Figure 31 shows the survivors of spores of B. anthracis exposed to three different temperatures plotted in this way, from an experiment by Eijkman (1912-13). The logarithmic survivor curves are straight lines, the different angles indicating different death rates. In recent years, several investigators have been satisfied that the logarithmic order is proved whenever the logarithms of survivors are found to be on a straight line. The most essential point in this kind of graphic proof is the inclu- sion of the initial number of cells in the curve. If this is not included in the graph, a straight line relationship proves nothing. As may be seen from Fig. 32, most of the curves indicating quite different orders of death are so nearly a straight line that they would be considered so from experimental data, if the first horizontal part of the curve were omitted. It is very essential to ascertain whether or not there is an initial period without deaths. The omission of the initial number makes this impossible. Only a constant death rate can then be considered proof of a logarithmic order of death. (d) INTERPRETATIONS OF THE ORDER OF DEATH This order of death is, in its mathematical aspects, a complete parallel to the monomolecular reactions. What the chemist calls ''rate of reaction" or ''reaction veloc- MECHANISM OF DEATH 283 ity/' is mathematically identical with the ''deathrate'' 1 log a - log 6 4. J u A = fTZoI ^s computed above. This similarity has led some bacteriologists to the belief that bacteria are sufficiently small to react like molecules. Others, however, have assumed that the logarithmic order of death is ultimately due to the logarithmic order of some chemical reaction causing the death. Some bacteriologists and biologists have been very much opposed to such mechanistic views and claim that the order of death is due to a graded variation of resistance of the individuals. However, it is not easy to explain the logarithmic order in terms of a graded resistance. Figures 27 and 29 show in their block curves the number of deaths per unit time. The resistance should be graded in a similar way which would require that the most sensitive bacteria are present in the largest numbers. This is quite different from the grada- tion of higher organisms (Figs. 28 and 30). Reichenbach (1911) tried to explain this especially graded resistance by the mode of multiplication. He assumes that from each new generation, a definite percentage ceases to multiply, and goes into a state of dormancy in which it becomes the more resistant the longer it remains in this stage. In this way, the last generation which is the largest, is also the most sensitive. There is no biological reason to assume that part of the cells of an actively growing culture should go into a resting stage. Besides, Kelly and Rahn (1932), in their study of the fission rates of individual cells, could never observe that any cells stopped growing in the early stages of growth after they once had started multiplying. This holds for Bad. aerogenes and B. cereus as well as for yeast. 284 PHYSIOLOGY OF BACTERIA Another attempt at an explanation for this is that made by Henderson Smith (1921). He raises the well-grounded objection that in death by chemical poisoning, differences in resistance might come from differences in the thickness of the cell membrane. Our only measure of resistance is the killing time ; but killing times will correspond to the grades of resistance only if the penetration of the poison is directly proportional to the thickness of the wall. If it would take the poison four times as long to penetrate a membrane twice as thick, the killing times could not be considered as indicators of uniformly graded resistance. However, if retarded penetration were the reason for the logarithmic order of death of bacteria, then they must show an order equal to that of higher organisms when death is caused by agencies which penetrate with immeasurable speed, such as light or ultraviolet rays. This is not the case, however. With these agencies, we find the logarithmic order established just as plainly as in chemical disinfection (p. 371). Quite different again is the interpretation of Rahn (1929c, 30, 31a) who started from the idea that the property of growth is condensed in the chromosomes and genes (p. 259). The genes are very small, beyond the limit of microscopic visibility. Considering the large size of the complex protein molecules of living protoplasm, each gene may contain only one, or a very few, molecules. If only one molecule of one of the genes is inactivated, the cell cannot multiply. Cell division requires the doubling of all chromosomes, and of each gene. In computing the variability of the growth rate (p. 267), Rahn had assumed that the gene-type molecules of a large number of uniform cells under uniform environ- MECHANISM OF DEATH 285 ment react as if all cell contents formed one continuous fluid, as if cell-walls and separating medium did not exist. This seemed justified because the genes do not react with one another, but only with molecules which are plentiful in each cell. Let us assume that a large number of uniform bacteria were heated to 55°C., and at this temperature, the most sensitive gene molecule would be hydrolyzed at K ^^T'^'vX \." \ \ ^\ ^"v ^ ^ ^ \ \ \ \^ ^\. ^\. \ ^ \ \ X ^v ^v ^^ \ \ ^v ^V ^ s. \^ X \ \ 2.S\ \^ \ >v \ \ \ \ \ >w^ ^^ ^V ^k \ \ \^ \ >>>^ ^ S. ^v ^^ \ \ U -2.0 'x ^\ \\\ \ \ t *Q \ \ \. ^^ \^ ^^ \ I 1 N. \ \ v'^xxX \ \ ■1.5^ "V \\\ \ \ \ s X ^V \.\ \ \ 1 *^ >v \. \ \ \ \ 1 V ^v \/V\ \ \ 1 -io.<^ ^^x^ X. n/\\ \ \ 1 v|>^ "*-.^ ^s/v\\ \ 1 ^ ^ ^ ^v \\\ \ 1 ■O.s ""^^.\s^\\l ^ "^o^\l Fig. 32. — Order of death computed for organisms which are killed by the reaction of r = 1, 2, 3, 4, 6, 12, or 100 molecules per organism. The dotted line is the order of death for r = 1 molecule, but with three different grades of resistance. the rate of 90% of all such molecules per minute. At the end of this time, only 10 % of the cells will have this gene intact ; the 90 % will have lost the power to multiply because one of the genes is damaged beyond repair. During the second minute, 90% of the undamaged cells will again suffer the hydrolysis of the special gene, i.e.. 286 PHYSIOLOGY OF BACTERIA 9% of the original number of cells will again become sterile, unable to multiply, and only 1% of the original cells is left unchanged. After the third minute, only 0.1% of all cells will be able to grow, and so on. This order of death is identical with that of a monomolecular reaction (see p. 278). For chemical reasons, then, uniform cells cannot all die at the same time, but must follow the mass law. If the cells are not uniform, the order of death is different, as shall be shown presently. This order of death does not apply to death by physical causes (e.g., freezing). There are some other exceptions, e.g., death by starva- tion. Why the effect of poisons must be considered monomolecular, will be explained on p. 341. From this one extreme, where the inactivation of only one single molecule causes death, to the other extreme of the higher organisms, where many cells must be inactivated before the organism dies, there are many intermediate steps. Rahn (1929c) computed a general formula for the order of death for any number of '^react- ing molecules," i.e., molecules which have to be changed to bring about death. Figure 32 gives the logarithms of survivors plotted against time. It shows a straight line only for one reacting molecule. For all other cases, the line is not straight, but shows at first a period of no deaths, and then turns sharply down in a curve which is not really straight, but might be mistaken as such. Rahn found (1931a) that death of Colpidium by HgCl2 seems to be brought about by the reaction of two molecules per cell, and that avian red blood corpuscles when exposed to ultraviolet light, also dissolve after the inactivation of two special molecules. With yeast, death was brought about sometimes by one, sometimes by two inactivated molecules. Mold spores seem to con- MECHANISM OF DEATH 287 tain quite a large number of reacting molecules. The logarithmic survivor curves for higher organisms are distinctly bulging as may be seen from Fig. 33 repre- senting the death of wheat and mustard seeds. We have seen (p. 280) that for bacteria, the death rate 1 log a — log h K = 0.434 remains constant. This K value must be constant when death is brought about by the reaction of one molecule per cell. If the — ""Tr-^ ^ x% f \\ "\ v»\ \ X \\ vl^ X\ ZO i^O so ti'mt units\ Fig. 33, — Order of death of seeds on standard scale. Wheat seeds, killed by heat, and mustard seeds, killed by bichloride of mercury. number of ''reacting molecules" is larger than one, K increases. When we compare the death rates of higher organisms with those of bacteria, as shown in Table 80, it becomes conspicuous that there is a continuous increase with all examples of the higher animals and plants while with bacteria, the rate is either constant or decreasing. All living beings except unicellular organisms show survivor curves bulging out above the straight Une, indicating more than one reacting molecule per cell (see e.g. Fig. 33). All bacteria (with three exceptions to be discussed later) show either a straight line or a survivor curve sagging below the straight line (see Figs. 31 and 34). 288 PHYSIOLOGY OF BACTERIA This latter type will be explained in the next subchapter as due to a variation in the resistance of the individuals. If bacterial cells contained more than one reacting molecule per cell, their survivor curves should show a bulging out above a straight Hne, and not a straight or even a sagging curve. If they had more than one reacting molecule per cell, their death rates should be continuously increasing, and not remain constant or decrease. Table 80. — Death Rates for Successive Time Units f Computed as fc = -^ log r j Higher organisms Colpidium Bacteria Bad. Fruit Wheat Mustard Anthrax Bad. Bad. flies seeds seeds spores typhosum coli para- typhosum 0.004 0.017 0.0018 0.013 0.146 0.194 0.406 0.748 0.013 0.027 0.0016 0.011 0.126 0.185 0.383 0.355 0.026 0.046 0.0023 0.058 0.138 0.191 0.452 0.222 0.043 0.095 0.0052 0.077 0.149 0.188 0.484 0.166 0.061 0.117 0.0087 0.108 0.204 0.350 0.133 0.077 0.141 0.0087 0.103 0.186 0.366 0.113 0.084 0.0080 0.127 0.147 0.344 0.101 0.088 0.0117 0.145 0.083 0.123 0.070 0.140 0.060 0.167 0.051 (Peters, 1920) (Reichen- (Table 76) (Table 76) (Table 77) (Table 79) (Table 79) (Table 96) bach, 1911) The pronounced difference in the order of death between bacteria and higher organisms indicates that death in the former is brought about by the change of only one molecule in each cell. (e) PHYSIOLOGICAL YOUTH Cells of the same strain show a different degree of resistance at different stages of development. The first quantitative observation in this respect was probably MECHANISM OF DEATH 289 that of Schultz and Ritz (1910). Similar results were obtained independently by Sherman and Albus (1923) who tested Bad. coli in its resistance to heat, to cold, to salt solutions, and to phenol. Some of their results are summarized in the following table. Table 81. — Death of Cells from Young and Old Cultures op Bad. coli Time of exposure Age of culture Plate counts Per- centage of sur- vivors Death rate Exposure to Before After Distilled water 1 hour 11 days 6 days 4 hours 4.5 hours 607 720 530 373 606 710 155 247 100 99 29 66 0.0007 0.006 0.534 0.179 53°C 20 min. 7 days 3.5 hours 1,470,000 27,800 97,000 24.6 6.6 1.1 0.059 0.153 0.5% phenol Smin. 24 hours 4 hours 24 hours 4 hours 4,650,000 1,240,000 6,900,000 2,400,000 42,000 300 10,800 430 0.917 0.024 0.156 0.018 0.713 1.202 0.935 1.249 The differences in the percentage and in the death rate of survivors are so striking and check so well with the preceding experiment that a ''physiological youth" of bacteria, as the latter authors termed it, must be considered established. It seems proved that the physiologically young bacteria are more sensitive to adverse conditions than old ones. This circumstance is probably one of the main reasons for decreasing death rates and sagging survivor curves. 290 PHYSIOLOGY OF BACTERIA (/) DECREASING DEATH RATES Decreasing death rates can be accounted for by varied resistance of the cells. If we have a mixture of sensitive and resistant individuals, the death rate at the beginning of the experiment will be high because all sensitive cells will die rapidly. After most of these are killed, the rate will be largely that of the resistant cells, and this is lower. It is most probable that in a culture which is not especially cared for, there will be organisms of different physiological conditions, and therefore of different resistance to the cause of death. Rahn (1930) computed a theoretical case for bacteria with three different grades of resistance, and the curve of survivors is shown in the dotted line of Fig. 32. This effect of variability of resistance had already been forseen by Harriett Chick (1910) who tried to avoid this decrease of death rate by using very young cultures, three hours old, and by transplanting them repeatedly in order to get rid of old and more resistant cells. The results are seen in Table 82. The degree of constancy or variability of K can be best compared by the variation in the relative death rates based on the average K of each experiment = 100. The lowest line of Table 82 shows that the relative spread of variation in the two experiments with twenty- four hours old cultures is quite large, 137 — 41 = 96% in the one and 140 ~ 56 = 84% in the other. With the three hours old culture, the spread of variation was 108%, i.e., larger than in the twenty- four hour cultures; perhaps there were many old cells left among the new growth. But upon transferring this culture repeatedly every three hours, the uniformity improved, and the spread of variation dropped to 26 and 44%. Reichenbach (1911) tried a similar experiment with the same organism. Bad. paratyphosum, exposed to heat. MECHANISM OF DEATH 291 Table 82. — Death Rates op Bad. paratyphosum in 0.6% Phenol Experiment No Ill V XI XII XIII Age of culture, minutes 24 hours 3 hours 3rd 3 hours generation 4th 3 hours generation 1 0.56 2 0.63 0.46 0.066 0.081 3 0.27 0.61 0.38 0.062 4 0.27 0.42 0.33 0.064 0.086 5 0.26 0.34 0.32 0.053 6 0.22 0.30 0.093 7 0.22 0.25 .... 0.051 8 0.20 0.23 9 0.18 0.22 0.072 10 0.15 0.20 0.058 15 0.12 0.059 0.065 20 0.08 0.054 0.059 Average K 0.197 0.45 0.333 0.058 0.076 Relative Values of K (average K = 100) 1 168 138 113 2 140 107 3 137 135 114 106 4 137 93 99 110 113 5 132 76 96 91 6 112 90 122 7 112 56 87 8 101 69 9 91 66 95 10 76 60 99 15 61 41 101 93 85 20 78 Greatest variation .... 41-137 56-140 60-168 87-113 78-122 He did not transfer repeatedly, variation shall be given here. Only the degrees of 292 PHYSIOLOGY OF BACTERIA Table 83. — Variability op Death Rate op Bact. paratyphosum Age of culture, hours Tempera- ture, ^'C. Average of K Average K = LOO Maximum Minimum Spread 5.5 49 0.179 417 25 392 8 48 0.064 178 41 137 13 18.5 24 50.1 51 50 0.049 0.134 0.046 113 105 (130) 158 (44) 92 88 46 21 (69) 17 (42) 112 28 51 0.177 109 93 13 48 55 49 50 0.066 0.071 113 (122) 109 92 89 21 (30) 20 In this experiment, the older cultures seem more uniform, with less spread than the young cultures; the only exceptional experiment ^>^ * ^^V ^^^'**Vw \ >^s. ^**s^ 1 ^ ^\Nft. ^^^ I * ^**0^V. ^^v \ ^ •ONv ^^^ A ^ ^\\ ^*^ \ •^^. ^^. ^\. \ ^ ^\^*S. ^ ^\jc \ * ^^^N^ ^^^o \. Q^ ^^"""s^^^jP ^CT "^"^^^^ "^^ ^^VV X '5 ^^*>^**-^^ ^ ^^"^^ N. V "^^^^^^---.^ - . ^ ^<^ \ ,-, -»- — ^^^^- ^ ^^55s^ \ v§^ ^^^"^^^^"^^>^^VV 5p tCmc unCts'^^y^^^^ Fig. 34. — Survivors of Bacterium paratyphosum from cultures of different age, drawn on standard scale. is that of a twenty-four hours old culture which was considered abnormal by Reichenbach. The figures in parentheses are unusual MECHANISM OF DEATH 293 values obtained during the first two minutes of the experiment; they might be caused by the change of medium, or by heat shock. The trend of the death rates in all of Reichenbach's experiments is a gradual decrease of K, and the ''spread" of table 83 shows the relative amount of this decrease. If Reichenbach's plate counts are plotted as logarith- mic survivor curves on a standard time scale (Rahn, 1930) the striking curves of Fig. 34 are obtained. The youngest culture contains old and young cells which differ greatly in resistance and the survivor curve sags deeply. With increasing age of the culture, the old cells begin to multiply, and finally, after thirteen hours, all cells are in a state of rapid multiplication and fairly uniform resistance. This is depicted graphically by the gradual straightening of the curves. (g) FREQUENCY OF THE LOGARITHMIC ORDER OF DEATH WITH BACTERIA Some biologists have questioned the logarithmic order of death. Loeb and Northrop (1917) in their study of the death rate of fruit-flies (p. 275 and Fig. 28) make the following statement: '' Miss Chick has stated that bacteria are killed by disinfectants at a rate corresponding to that of a monomolecular chemical reaction, i.e., that in each interval of time the same percentage of individuals alive at this time is killed. She was probably led to such an assump- tion by the fact that the ascending branch of the mortality curve in her experiments was generally very steep. The agencies used by her for killing the bacteria were so powerful that the ascending branch became almost a vertical line, thus escaping attention. Hence she noticed usually only the less steep descending branch which could be interpreted as a monomolecular curve for the reason that her experiments lasted only a short time." This same objection has been raised again and again; it is believed by many biologists that the logarithmic 294 PHYSIOLOGY OF BACTERIA order would not appear at all if the first plate counts had been made sooner. This belief is not based on facts. It has never been proved that a logarithmic order of death would become established e.g. in the experiments p. 275 with seeds, fruit-flies or protozoa, if the first observations would have been omitted. In fact, no logarithmic order would be established if this were done, as any calculation or graphic test will easily show (e.g., see Fig. 33) . The death rate of the higher organisms increases, and the curve continues to bulge, while with bacteria, the death rate is always decreasing if not constant, except in the case of some spore-formers and of staphylococci, which will be discussed on p. 296. It is true, that in most experiments on the order of death, or on the theory of disinfection, almost half of the bacteria or more have died during the first time interval investigated. But this very fact points out that the order of death is essentially different from that observed with higher organisms. Table 84 gives a survey of all experiments known to the author referring to the order of death. They are divided into three groups according as the death rate is dominantly constant or decreasing or increasing. Nearly one-half of all experiments cannot be used to prove or disprove the logarithmic order of death because the most important part of the experiment, the initial number of bacteria, has been omitted. Of the remaining one hundred and fifty-four experiments, only thirty-two have a constant death rate. But we have just seen (p. 290), that a decreasing death rate must be expected to be the most common occurrence because the cultures under test are rarely very uniform in resistance. It should be remembered, too, that the decreasing death MECHANISM OF DEATH 295 Table 84. — Survey of All Experiments on the Order of Death OF Bacteria Author Number of experiments With initial count missing With death rate Decreasing Constant Increasing Death by drying Paul (1909) Paul, Birstein and Reuss (1910a) Madsen and Nyman (1907) . Death by heat Chick (1910) Reichenbach (1911) Eijkman (1912/13). Sattler (1928) Death by chemical poisoning Madsen and Nyman (1907) Chick (1908) Chick (1910) Reichenbach (1911) Paul, Birstein and Reuss (1910b, c) Eijkman (1912/13) Lee and Gilbert (1918) Winslow and Falk (1926)... Myers (1929) 2 17 Death by irradiation Clark and Gage (1903) Lee and Gilbert (1918) Gates (1929) ... 1 1 2 Total in each group 140 83 32 39 296 PHYSIOLOGY OF BACTERIA rate indicates that this order of death is still further removed from that of higher organisms than the constant one. It may well be stated that one hundred and fifteen of the hundred and fifty-four experiments speak in favor of the logarithmic order of death, or at least in favor of an order entirely different from that of higher organisms. There remain thirty-nine experiments with an increas- ing death rate which makes them resemble the order of death of larger organisms. Increasing death rates have been observed occasionally, as exceptions, with several species (e.g. Bad. coli, Bad. paratyphosum) which ordinarily show constant or decreasing rates. But with a few species, an increasing death rate seems to be the rule. The three well established cases are the Staphylo- cocci (Chick, 1910), the large, unnamed spore-former of Table 85. — Disinfection of Micr. pyogenes aureus by 0.6 % Phenol Computed for Standard Scale Chick's Table VI Chick's Table VII Time Survivors Death Time Survivors Death units rate units rate 0 100.0 0 100.0 4.9 92.2 0.035 4.3 84.6 0.073 14.9 80.8 0.031 8.6 81.8 0.043 19.8 73.7 0.033 12.9 72.4 0.047 24.8 64.7 0.052 17.2 63.0 0.050 29.6 42.0 0.063 22.0 52.4 0.056 34.6 31.0 0.073 30.2 26.2 0.083 39.6 18.8 0.091 43.1 10.4 0.098 44.5 15.65 0.089 64.7 2.47 0.107 49.4 12.08 0.092 100.0 0.1 0.115 69.3 3.51 0.121 74.2 1.69 0.118 100.0 0.1 0.149 MECHANISM OF DEATH 297 Reichenbach (1911) and the spore-former of Myers (1929). As an example, two experiments by Chick are given in Table 85. ■•^ — Fig. 35. — Survivors obtained by spreading a culture of Bacterium coli on agar and exposing it there to X-rays. ^The full line shows the survivors when the agar plate is exposed immediately; the dotted line shows the results when the agar plate had been incubated for a short time before exposure. While we have here an approximation to the survivor curves of higher organisms, it is not necessary to assume 298 PHYSIOLOGY OF BACTERIA a larger number of '^reacting molecules" per cell though such cases appear quite possible (see p. 367). Rahn (1930) has pointed out that the staphylococcus gets its name from the tendency to form clusters, and that spore-formers usually grow in threads. Each thread or cluster will contain several cells, yet on the agar plate, it will produce only one colony. The plate count does not give us the number of individuals in these cases, but the number of clusters, and the clusters contain several ^'reacting molecules,'^ namely one for each cell. The fault lies in our inadequate method of counting living cells. A very simple and convincing proof for this assumption is given by Wyckoff and Rivers (1930). They spread a young culture of Bad. coll on an agar surface, exposed it to cathode rays, and con- sidered that the colonies developing on the agar represented the survivors. A microscopic investigation proved that the cells spread on the agar had not been clustered. The survivor curve was nearly a straight line (Fig. 35). A similar plate was held in the incubator for a short time before being exposed. The originally single cells of the bacterium had had a chance to form small groups. The survivor curve of this plate was bulging like those known for larger organisms. (h) ADAPTATION AND SELECTION OF RESISTANT STRAINS According to the strict chemical interpretation of the logarithmic order of death, the last survivor is no more resistant than the first cell that died. Surviving is merely a question of chance. According to the theory of graded resistance, the last survivor is much more resistant than the first, in fact, it is the fittest survivor. Although with higher animals and plants, this method of selecting resistant strains is almost certain to give good results (provided due consideration is given to the laws of inheritance), the experiments with bacteria have not been very successful. MECHANISM OF DEATH 299 This is partly due to inadequate technique. The death rates for two transfers of the same culture on different days vary greatly. Chick (1910) found the death rate (0.434if) of Micr. 'pyogenes aureus with 0.6% phenol at 20°C. to be, on March 13th = 0.32, March 16th = 0.114, March 30th = 0.32 and AprH 6th = 0.136. The same author found that Bad. coli died in water of 489°C. with a death rate of 0.172, and eleven days later at 49.0°C., with a death rate of 0.042. Such variations (amounting in the last case to 400%) are found in the work of other investigators as well (see e.g., Table 82). An explanation shall be tried later (p. 333 and 345). It seems that in recent experiments, it has been possible to reduce this error considerably. Myers (1929) was able to average data from different months without difficulty. But as long as the technique shows deviations within the same strain of the degree just mentioned, there is little hope of proving increased resistance. This may account for some of the failures. It is well known that bacteria can be adapted to poisons, to higher temperatures or other adverse conditions. This is accomplished by subjecting them to gradually increasing intensities of the harmful substances, or conditions. With our problem, however, we are inter- ested only in those cases where the last survivors show greater resistance when compared with their ancestors, or with the average. The experiments of Magoon (1926) which are sometimes cited as such an example, are not at all convincing, Magoon transfered the last survivors of heated spores of B. mycoides to a new medium and determined the resistance of the newly developed spores again. The first culture survived nine minutes at lOO^C; how much longer the spores might have Uved, was not determined. The second culture survived twenty minutes; it was not ascertained how much more heating it might have tolerated. Only with the third culture, it was certain that it could tolerate twenty-eight, but not twenty-nine minutes' heating. There is no evidence whatever that the other two cultures were less qresistant. The fourth culture was tested for thirty-six minutes and all survived; this culture was more resistant 300 PHYSIOLOGY OF BACTERIA than the preceding one. With the fifth culture, Magoon suddenly changes his technique and holds the culture for sixty-four days instead of twenty-nine as in all previous experiments. The result is a greater resistance. If we do not use these incomparable last data, but consider the last experiment finished after twenty-nine days (Magoon's records make this possible), then it becomes evident that the fifth culture is decidedly less resistant than the fourth, and about equal to the third. The experiments do not prove an increase in resistance by selection, but, on the contrary, a fluctuation. The last survivors are not more resistant than the average. Similar are the experiences of Arnold (1929) who sprayed several millions of cells of Bad. coli and Bad. prodigiosum into the nose, made cultures of the few remaining cells and repeated the experiment with these cultures as often as nine times. "There was no evidence of an acquired resistance on the part of those bacteria that had sur- vived for a maximum period of time upon the nasal mucosa." More than twenty years previous to this experiment, Gage and Stoughton (1906) tested Baderium coli. Broth cultures of this organism were exposed for five minutes to temperatures ranging from 45°C. to 100°C. The survivors of each tube were counted by plating. The tube heated to 60° was used as the ''second genera- tion." Subcultures of it were again exposed to different tempera- tures, the survivors were counted, and the culture heated to 60° was used as "third generation." In this way, nine generations were tested, all of which had been heated to 60°C. or 55°C. The results are shown in Table 86. The numbers are the survivors in percents of the inoculum; -j- means growth in the tube, but no colonies on the agar plate; 0 means no growth in the tube. Gage and Stoughton summarize their results as follows: "Experi- ment 195 in which we attempted to produce, by the survival of the fittest, a race of especially resistant organisms, demonstrated a diametrically opposite result." This agrees with the experiences of Gates (1929) whose experi- ments on death by ultraviolet light will be recorded on p. 372. He states that "cocci from colonies of the last surviving organisms have proved to be inherently no more than normally resistant to ultraviolet light." These experiments speak quite decidedly against the assumption that the last survivors are more resistant; MECHANISM OF DEATH 301 Table 86.- —Successive Cultures ob Lasi Survivors OF Bad. coli Percentage op Survivors after 5 Minutes' Heating Generations Heated, "C. First Second Third Fourth Fifth Sixth Seventh Eighth Ninth 45 136.0% 86.0% 100.0% 63.0% 72.0% 63.0% 258.0% 64.0% 243.0 50 70.0 3.6 44.0 46.0 20.0 23.0 4.5 9.7 49.0 65 4.4 0.006 0.050 2.3 0.004 0.020 0.020 0.012 0.200 60 0.001 0.008 0.010 0.001 0.009 0 0.160 0.002 0 65 0.250 0.400 0.001 0.001 + 0 0.020 0 0 70 0.030 0.001 0.002 0 0 0 0.010 0 0.022 75 0.050 0.014 0.003 + 0 0 0 0 0 80 0.015 + 0.003 + 0 0 0 0 0.018 85 0.003 0 + 0 0 0 0 0 0 90 0.003 0 0 0 0 0 0 0 0 95 0 0 0 0 0 0 0 0 0 100 0 0 0 0 0 0 0 0 0 it appears to be much more a question of chance which cells survive. {i) SUMMARY OF FACTS Dying is a gradual chemical change in the cell, and any definition of death is the arbitrary assumption of one point on a time curve. If we define a dead cell as one which has lost the power of reproduction, the death of bacteria through heat, chemicals, light, drying and some other causes proves to be an orderly process. If the logarithms of survivors are plotted against time, bacteria show either a straight line or a curve sagging below a straight line, while larger organisms give a curve which bulges out above the straight line. In only three cases, bacteria are known to show consistently bulging curves. A ^' death rate" can be computed by using the formula for the rate of reaction in monomolecular processes. 302 PHYSIOLOGY OF BACTERIA This rate is increasing with higher organisms, but con- stant or decreasing with bacteria, except in the above- mentioned three cases. Young cells, a few hours after having been transferred into a new medium, are more sensitive to adverse condi- tions than cells of the same culture after growth has ceased. The time required to kill all bacteria in a given volume increases with the number of cells, regardless of the means used to kill them. The last survivors of a disinfection experiment, as a rule, do not give cultures of a higher resistance than the original strain. SUMMARY OF THEORIES The logarithmic order of death cannot be explained by a graded resistance unless several assumptions are made which have no biological justification other than to fit this purpose. If the order of death is computed under the assumption that the destruction of one certain molecule in the cell is sufficient to prevent growth, the resulting order of death corresponds to that observed with bacteria of uniform age. The computation for the assumption that several or many molecules must have reacted before death occurs, gives survivor curves corresponding to those observed with multicellular organisms. The order of death of cells with only one ''reacting molecule" shows a constant death-rate and a straight line in the logarithmic survivor curve. If the cells under test are not all of the same physiological condition, but show differences in resistance, (each grade of resist- ance having a constant death-rate) the logarithmic survivor curve will sag below the straight line and the MECHANISM OF DEATH 303 death rate will decrease. This must happen, e.g., with mixtures of old and young cells. Organisms with more than one such '^ reacting mole- cule" show an increasing death rate, and a survivor curve bulging above the straight line. Bacteria which have a tendency to clump or cluster, and whose plate count does not reveal the actual number of cells, but only the number of clumps or clusters, must give survivor curves corresponding to more than one ^^ reacting molecule" because the only method of counting living bacteria is the plating method, and here, each colony represents a clump of cells whose total number of reacting molecules equals the number of cells in the clump. IV. DEATH OF DRY BACTERIA The following subchapters will treat the various causes of death. That of dry bacteria has been chosen as the first cause to be discussed because conditions here are simpler than with other causes. When bacteria are in a moist condition, endogenous catabolism is decreasing, the cell substance continuously, while in dry cells, lack of moisture prevents this. Thus the uncertain factor of catabolism does not enter into the death of dry bacteria. (a) ORDER OF DEATH OF DRY BACTERIA Dry bacteria will die in the course of time ; some species within a few minutes, most of them within weeks or months, and the spores of some bacilli eventually not for a decade. The viability or longevity of dried bac- teria depends not only upon the species, but also upon the temperature and the stratum on which the cells are dried. 304 PHYSIOLOGY OF BACTERIA While all of this had been known for a long time, no real understanding of the cause of death was obtained until Paul (1909) started his quantitative investigations. Though he used staphylococci which are likely to give abnormal results (p. 296), he obtained a fair constancy of the death-rate in his experiments. Table 87. — Death op Micr. pyogenes aureus at Various Temperatures Time, Number of living cells Death rates for t = 2 days days 16-18''C. 6-8°C. - 190^*0. le-is'^c. 6-8°C. -190°C. 1 90,800 88,800 65,900 2 68,600 77,100 60,900 0.28 0.14 ? 4 48,500 66,400 77,800 0.21 0.10 6 27,200 65,100 98,400 0.24 0.06 8 11,400 37,100 60,700 0.30 0.12 10 5,320 18,700 61,200 0.32 0.17 12 6,680 17,100 69,200 0.24 0.15 14 2,630 10,200 61,500 0.27 0.17 16 900 4,980 73,300 0.31 0.19 18 1,500 3,000 68,100 0.24 0.20 32 300 350 67,900 0.18 0.16 Average. . 0.26 0.15 < 0.001 Paul dried the bacteria on small stones (garnets), of uniform size (details, see Paul, Birstein and Reuss, 1910a), which were kept in test-tubes under varying conditions. From time to time, some of them were removed and shaken with sterile water to separate the bacteria which were then counted by the plating method. The first experiment of Paul, with dry staphylococci at room temperature, at that of the refrigerator, and at the temperature of liquid air, gave the data of Table 87. MECHANISM OF DEATH 305 Another experiment, with another strain, gave similar results. The death rates are sufficiently constant to obtain an average. From the equation on p. 281, the percentage of survivors can be computed: 100 P = 10 0.434/5: = 77.1% at room temperature (16-18°C.) = 86.1% in the refrigerator (6-8°C.) = over 99.9% in liquid air (-190°C). This gives the death rates per unit time (=2 days) as 22.9%, 14.9% and less than 0.1% respectively. (6) OXIDATION AS THE CAUSE OF DEATH The possibilities for the cause of death by desiccation are few. Catabolism is out of the question because it requires moisture. It may be that the colloidal state of the protoplasm is changed irreversibly and irreparably. This does not seem probable because of the very slow death-rate. Such a change might occur during the drying process, and may cause death. After the bacteria are once dry, and still alive, a colloidal change is less probable. Another possibility is oxidation. Paul, Birstein and Reuss (1910a) proved oxidation to be the main cause of death of dry staphylococci. The bacteria were kept in air (20.8% O2), in technical oxygen (96.2% O2), and in mixtures of oxygen and nitrogen. Each of the death rates in the following table is the average of 6-8 indi- vidually determined constants. This table shows that the death-rate increases with the oxygen concentration. The authors found that the rate is proportional to the square root of the oxygen 306 PHYSIOLOGY OF BACTERIA concentration. This might be accounted for by assum- ing either an adsorption of oxygen by the bacterial surface, or by a dissociation of the oxygen molecule into atoms, O2 = 20, as in the slow oxidation of phosphorus. Table 88.— Death of Micr. pyogenes aureus at Different Oxygen Concentrations and Temperatures Experimental series Temperature, °C. Oxygen concentration, % Death-rate per hour Via 18.2 20.8 0.0243 Via 18.2 96.2 0.0530 VIII 37.4 20.8 0.0157 VIII 37.4 96.2 0.0256 VIII 18.0 20.8 0.0017 VIII 18.0 96.2 0.0034 XII 37.4 20.8 0.0264 XII 37.4 54.6 0.0369 XII 37.4 96.2 0.0444 XII 24.9 20.8 0.0107 XII 24.9 63.2 0.0152 XII 24.9 96.2 0.0200 The data of Paul and his associates on the influence of oxygen on dry bacteria have been confirmed and com- pleted by Rogers (1914) who also controlled the moisture content of dried streptococci. It seemed that the less moisture was contained in the cells, the longer they lived. Rogers dried his bacteria by placing the culture in an evacuated desiccator over sulfuric acid at freezing temperature. After being dried, the streptococci were tested for viability in various gases. Only one count was given MECHANISM OF DEATH 307 by the author for each series. The death-rates have been computed in Table 89 for one day as unit time. The observation that more bacteria survived in oxygen than in air, is contradictory to the other data by Rogers as well as to Paul's results. In all other gases, the death-rate was lower than in air, as was to be expected. The vacuum is the best method of preservation. Table 89. — Viability of Dry Lactic Streptococci in Different Gases Oxygen 1 Air Hydrogen Nitrogen CO2 Vacuum At beginning After 132 days 3,270,000 7,750 3,270,000 750 3,270,000 350,000 3,270,000 22,000 3,270,000 276,000 3,270,000 785,000 K = 0.0200 0.0276 0.0074 0.0164 0.0081 0.0047 The difference in the influence of the various ^' inert'' gases is not explained. The average death-rate in all experiments without oxygen is 0.009 as compared with 0.0276 in air. The dry streptococci die in air three times as fast as without oxygen. Paul's computation that the death rate is proportional to the square root of the oxygen concentration does not take into consideration this slow death without oxygen. Possibly, the rate was too slow to influence his calculations. However, the fact remains that even with the removal of oxygen, dry bacteria will still die, though more slowly. The cause of death in the absence of oxygen is unknown. The simpler method of drying bacteria on filter paper used by Brown (1925, 1926) has not been used for quantitative experiments, but could easily be adapted to this purpose. 308 PHYSIOLOGY OF BACTERIA (c) INFLUENCE OF TEMPERATURE Paul's data show that dry bacteria die more rapidly at higher temperatures. At that of liquid air, — 190°C., death was so slow that no distinct decrease could be observed in a month. Since death of dry bacteria has been found to be primarily due to the oxidation of some essential part of the cell, we must assume this oxidation to follow the same temperature laws as other chemical reactions (p. 119). The data of Tables 87 and 88 can be well used for calculating the temperature coefficients of the fatal oxidation. This computation does not require the assumption that the logarithmic order of death is strictly adhered to, nor that the order is due to a chemical process. It is just assumed that the value X is a correct expression of the relative velocity at which the bacteria die. Table 90. — Temperature Coefficients of the Death by Drying OF Micrococci Series Oxygen concentration, per cent Interval of temperature, °C. Coefficient for 10°C. increase VIII* VIII* XII* XII* Table 87 20.8 96.2 20.8 96.2 20.8 37.4-18.0 37.4-18.0 37.4-24.9 37.4-24.9 17 -7 3.16 2.83 2.04 1.90 1.73 Average 2.33 From Table 88. MECHANISM OF DEATH 309 Apparently, the coefficient is not constant, but the data are too few to warrant conclusions. So much seems certain, however, that the temperature coefficient of death through drying is similar to that of common chemical reactions, i.e., it lies between 2 and 3. Rogers (1914) found the temperature coefficient of the death rate of dry streptococci to be 3.6 between 0 and 30°C. It seemed interesting to determine the temperature coefficient at temperatures which exceeded the range of normal growth temperatures. Upon the author's sug- gestion. Dr. H. H. Boysen made three series of experi- ments with dry yeast cells keeping them between 30 and lOO'^C. The distribution of cells in the strata used (infusorial earth or sand) was not sufficiently uniform to use the data directly. They were accurate enough, however, to compute the times required to kill 99% Table 91. — Time Required to Kill 99% of the Cells op Dry Yeast Series Stratum 30°C. ST'C. 42.5«C. 50°C. 60<'C. 80°C. 90°C. 98°C. I II Infusorial earth. Infusorial earth .... 56 d. 21 d. 15 d. lid. 5.8 d. 76 min. 160 min. less than 4 min. 4 min. III Sand 110 d. 75 hrs. 29 min 60 min. (Temperature coefficients) I II Infusorial earth.... Infusorial earth .... • 4.06 3.24 121 no .61p •20/ _^ < 6. r<4.36l \<6.32/ III O TO /5.29\ 14.22J 310 PHYSIOLOGY OF BACTERIA of the initial number of viable cells. From these times, the temperature coefficients of Table 91 could be obtained. The results are fairly consistent with the exception of one too high value in the first series. If the range from 30°C. to 60°C. is taken in toto, the temperature coefficient is 10.2 or 7.96, just which, depends on whether seventy-six or one hundred sixty minutes is used as the correct time. The temperature coefficients for the entire range of the three experiments are 8.34, 5.77 and 3.53 respectively. This difference is very great. There are at least two possibilities that might account for this difference: the stratum on which the yeast was dried; and also, that no effort was made to dry the yeast to the same degree of dryness. The varying moisture content would very likely affect the temperature coefficient greatly, since the coefficient of death by moist heat is very high. Despite several inconsistencies in the results, the data are entirely sufficient to show that the chemical process of dying of dry bacteria is not essentially influenced by the change from '' growing temperatures" to ''killing temperatures." At 30°C. and 37°C., the yeast would have shown good growth in a suitable medium, while at 50°C., in the same medium, the yeast would have died. This great contrast has been entirely eliminated by the absence of water. No inactivation of enzymes or coagulation of proteins by supramaximal temperatures takes place because these processes require moisture. Death is altogether controlled by oxidation, which is a normal chemical process. The dry micrococci of Paul's are reduced from 100 living cells to 1 in thirty-three days at 17°C. (Table 87). Calculating with a constant temperature coefficient of 2.5, the same reduction would require at 67°C. 33 ^ 2.5^ days = 8 hours 7 minutes at 97°C. 33 -^ 2.5^ days = 31 minutes MECHANISM OF DEATH 311 while moist cells would be killed at these temperatures in a few minutes or even seconds. If the coefficient should decrease over this range, the resistance of dry bacteria at high temperatures would be still greater than calculated. This also enables us to compute the time for sterilizing at lower temperatures dry glassware or other dry materials which cannot stand the customary high heat of 160°C. for thirty minutes. With a temperature coefficient of 2.5 (this coefficient has really never been determined for dry spores), it would require at 100°C. about 30 X 2.5^ minutes = five days two hours to obtain a sterilizing efficiency equal to thirty minutes at 160°C. It is generally known that the material on which bacteria are dried has a great influence on their via- bility. Bacteria dried on coverslips die very readily, while those dried in soil remain alive over a long period. The bacteria of legume nodules can be cultivated from herbarium specimens after years, while the attempt to use, commercially, cultures dried on cotton proved to be a failure. A large amount of time has been spent in searching for paints or finishes for walls of hospitals and operating rooms which would kill in the shortest possible time any bacteria smeared on them. But no quantitative work has been done in this respect except just a measurement of the time required to kill all bacteria spread and dried on certain strata. No experi- ments which try to explain the different death-rates on different strata from the viewpoint of death as an oxidation process are known to the author. A very interesting and important question has never been solved or attacked, namely, the cause of death of dry bacteria in the absence of oxygen. It is barely possible that bacteria remain alive for infinite times in the absence of oxygen. Lipman (1931) has succeeded in demonstrating living bacteria in coal. The rate of death becomes smaller when we eliminate the main 312 PHYSIOLOGY OF BACTERIA cause of death. Cells without food live longer when dried; dry cells live longest at the lowest temperatures; dry and cold cells live longest in the absence of oxygen; but they do not live indefinitely, and the cause of death under these circumstances, though unknown, is of general biological interest. The laws of the death of dry bacteria apply also to bacteria in non-aqueous liquids. Bartlett and Kinne (1913) compared the resistance of spores heated in water, glycerol, olive oil, cottonseed oil and paraffine, and found the bacteria in non-aqueous liquids much more resistant. Staphylococci were killed quickly at 100°C. in all media, lower temperatures not being tried. Spores of B. anthracis or B. suhtilis at lOC'C. died in water in a few minutes, but were still alive in glycerol after seventy-five minutes, in oil after fifty minutes. Spores of B. vitalis survived 120°C. (15 pounds pressure) in water ten minutes, in glycerol 90 minutes, and in oil two hours, but not longer. This again indicates that the death process by dry heat is entirely different in its nature from that by moist heat. (d) SUMMARY OF FACTS Dry bacteria die slowly, and there is a fairly definite order of death. The rate of dying varies greatly, even with the same culture dried on different days. Dry bacteria die more rapidly in air than in a vacuum, or in an inert gas; they die more rapidly in an oxygen atmosphere than in air. Oxygen is the most essential cause of death of dry bacteria. Dry bacteria die more rapidly at higher temperatures. Dry bacteria probably die more rapidly when their moisture content is high. MECHANISM OF DEATH 313 Bacteria in oil or other water-free fluids die in the same way as do dry bacteria. The rate of death of dry bacteria is greatly affected by the medium upon which they are dried. SUMMARY OF THEORIES The primary cause of death of dry bacteria is an oxidation process; the death rate is proportional to the square root of the oxygen concentration. The temperature coefficient of this oxidation process is either normal (experiments with micrococci, range 7-37°C. and lactic acid bacteria, range 0-30''C.) or higher, but not above 10 (experiments with yeast, range 30-100°C.). The temperature coefficient probably increases with an increasing moisture content of the test organism. There is no evidence that the cause of death of dry bacteria at temperatures above the maxi- mum for growth is different from that at temperatures of normal life. The cause of death of dry bacteria in the absence of oxygen is still unknown. V. DEATH BY FREEZING Death of bacteria by freezing has been studied to a considerable extent from the viewpoint of typhoid epidemics from river ice, and also on account of food preservation by freezing. The data obtained are not very uniform, sometimes they are contradictory, and make it probable that several factors enter which had not been controlled in the experiments published. If we consider freezing to be a solidification of water, then freezing is identical in its effect with drying, and death by freezing should follow the same general rules and laws as that by drying. However, there 314 PHYSIOLOGY OF BACTERIA is no evidence at all available that death by freezing is due to an oxidation process as is death by drying. All data on death by freezing indicate that it is not brought about by one fundamental reaction. Hilliard, Torossian and Stone (1915) enumerate the following factors which might cause death during the freezing process: inter- ference of low temperature with metabolism; rupture of membrane by pressure of ice formation in the cell; rupture by external pressure or grinding during crystal- lization; pressure by expansion of the freezing medium in closed receptacle; prolonged suspension of metabolism leading to slow death by old age or by starvation. That freezing, i.e., the crystallization of water as such, causes death was shown by Hilliard and Davis (1918) by comparison of the death rate in water which solidified, with that in sugar solutions which did not solidify, at the same temperature. Some of these results are given in Table 92, each datum represent- ing the average of at least four experiments. The Table 92. — Death op Bad. coli by Freezing Temperatures FOR 3 Hours Percentage of cells killed Temperature, °C. In glucose solution, not frozen, % In water, frozen, % -0.5 -1.5 -2 -3 -4 -6 49.6 36.2 40.8 44.2 49.3 49.5 90.3 95.9 93.1 96.9 98.8 99.2 destructive effect of the physical process of freezing was further shown by the very rapid destruction of MECHANISM OF DEATH 315 cells by repeated freezing and thawing as compared with check cultures remaining frozen over longer periods of time (Table 93). It was also found by the same authors that in milk, bacteria are not decreased as much by freezing as in water. Table 93. — Death by Continuous Freezing and by Alternate Freezing and Thawing (Numbers indicate plate counts per c.c.) BacL typhosum Frozen solid Alternate freezing Inoculum 40,896 29,780 1,800 950 2,490 Inoculum . . . 40 896 24 hrs. frozen Frozen 3 times 90 3 days Frozen 5 times 0 4 days Frozen 6 times 0 5 days BacL prodigiosum Frozen solid Alternate freezing Inoculum 339,516 36,410 41,580 14,440 4,850 Inoculum Refrozen once 339 516 24 hrs. frozen 2,570 275 30 hrs. frozen Refrozen 2 times 48 hrs. frozen Refrozen 3 times 15 96 hrs. frozen Refrozen 4 times 0 Recently, extensive studies on the freezing of bacteria and yeasts have been published by Tanner and William- son (1928), not so much with the object of studying the fundamental reactions concerned as to test the differences of resistance of different species. The organisms were frozen in small samples and kept at —13 to — 15°C.; for each test, one of these samples was thawed and a 316 PHYSIOLOGY OF BACTERIA plate count made. Parallel sets in physiological salt solution and in broth were made of all organisms. The survivor curves are quite irregular, indicating the possibility of more than one cause of death. The indication of a period of no death at the start is noticeable in most experiments. This may be due either to cluster- ing of the cells, (see p. 296) or to the slowness of action, for in death by very slowly acting agents, such curves have been observed. The differences in resistance of the different species have been summarized in Table 94 by giving the time required to kill 99.9% of all cells. In salt solution, almost all species can tolerate the freezing better than in broth. Table 94. — Time Required to Kill 99.9% op all Cells by a Temperature op —13 to — 15°C. Time required Species In broth, weeks In salt solution, weeks Bad. coli 8 8-9 Over 80 80 Over 160 5 58 Over 160 9-26 9-10 10-26 46-54 12-16 BacL prodigiosum B. mesentericus 12-16 Over 80 B. suhtilis Over 80 Sacch. cerevisiae Over 160 Sacch. ellipsoideus 18 Sacch. pastorianus Over 160 Sacch. marxianus 14 Pichia membranefaciens 7-9 Torula rosea 10-26 Zygosaccharomyces mongolicus Mycoderma vini . . . ... 46-54 68-160 This very slow rate of death is verified by the obser- vation of Prucha and Brannon (1926) who found that MECHANISM OF DEATH 317 it took one year to kill 99.9% of all typhoid bacteria in ice cream when held at approximately — 19°C. The slowness makes possible the explanation that the fundamental process of death, aside from physical destruction, might be an oxidation process, or, perhaps, a starvation process. No data are available to decide between these two and possibly other explanations. SUMMARY Death by freezing involves a number of different possible causes. While the low temperature, as such, is one of the factors, physical destruction by the forma- tion of ice crystals has been shown to be important, so that repeated thawing and freezing kills more rapidly than long continued cold. The medium in which bac- teria are frozen is also of significance for the rate of death. Death by freezing is slow. The extremes recorded in literature for a reduction of 99.9% are with Bad. coli, eight weeks, B. suhtilis, more than eighty weeks, Sacch. ellipsoideus five weeks, Sacch, cerevisiae, more than three years. VI. DEATH BY HEAT Under ^^heat" in this chapter is understood ^' moist heat," i.e., the application of high temperatures to bacteria in their natural state of living. Death by dry heat has been treated in the chapter on Drying (p. 309). (a) METHODS OF MEASURING While all that has been said about the technique of disinfection experiments in general (p. 270 to 274) applies here, some special features enter into the killing of bacteria by moist heat that deserve special attention. 318 PHYSIOLOGY OF BACTERIA The death rate is influenced remarkably by the medium in which the organisms are heated. In the earlier experiments (Chick, Eijkman) distilled water was used altogether. This introduces considerable uncertainty into the cause of death. Probably, the diffusion of cell constituents into the distilled water is too slow to kill cells in the short time of the experiment. Since the hydrogen ion concentration in distilled water is known to vary greatly, it will not be possible without pH control to obtain data which agree. If these experiments are carried out in a medium suitable for the development of the test organism, there is no chance for variation of acidity, nor for leaching out of essential compounds from the cell. A very remarkable observation was made by Lange (1922) namely, that bacteria killed by heat and added to living bacteria, protect these against the action of heat to a remarkable degree. One of his experiments is Table 95. — Protection against Heat of Living Cells Cells in Broth BY Dead Standard Suspension Water Broth Culture living ( + ) or dead (0) after Living cells Dead cells 5' 10' 20' 30' 45' 60' 120' 180' 240' Bad. coli at 60°C. 0.02 drops 1 c.c 0 . 02 drops 1 c.c. 1 c.c. 5 c.c. 5 c.c. 5 c.c. + + + + + + + + + + + + 0 + + 0 + + 0 0 0 0 0 0 0 0 0 Micr. pyogenes at 56°C. 0 . 02 drops 1 c.c 0 . 02 drops 1 c.c. 1 c.c. 5 c.c. 5 c.c. 5 c.c. + + + + + + + + 0 + + 0 + + 0 + + 0 + 0 0 + 0 0 0 0 MECHANISM OF DEATH 319 reproduced in the preceding table; the addition of 1 c.c. of dead cells doubles the time required to kill Bad. coll, and trebles the time for Micr. pyogenes. It is necessary that the added dead cells come from a young culture. Old cells exert no protective action. (b) THE ORDER AND RATE OF DEATH AT DIFFERENT TEMPERATURES In the killing of bacteria by moist heat, the logarithmic order is generally quite well established. Most of the examples given in the chapter on the order of death refer to death by heat. As typical examples, we may mention (see Table 96) the death of Bad. coli at 48.9°C. and 52.7°C., by Chick (1910) and the death of the spores of B. anthracis at 80°C., 84°C. and 90°C., by Eijkman (1912/13) of which only the survivor curves are given in Fig. 31, p. 281 showing straight lines. Table 96.— Death of Bad. coli at 48.9 and 52.7°C. 48.9°C 52.7°C Time, minutes Number of sur- K Number of sur- K viving cells viving cells 0.25 372 383 1.0 274.3 0.406 208 0.820 2 190 0.383 89.7 0.830 4 68.25 0.452 14.5 0.874 6 23 0.484 1.1 1.017 8 0.55 0.845 10 12.4 0.350 0.066 0.890 15 1.72 0.366 20 0.425 0.344 Mean 0.396 0.878 The computed death-rates are sufficiently constant to be used for the computation of the temperature coef- 320 PHYSIOLOGY OF BACTERIA ficient. We find this, according to the formula from p. 120, for Bad, coli «.. - im" - '- Table 97. — Temperature Coefficients op Death by Heat for Different Species Number Tempera- Author Organism of ex- periments Average Qio ture interval, °C. Gage and van Bad. coli 3 27.9 50-55 Stoughton (1906) 2 26.5 50-60 Chick (1910) Bad. coli 2 12 49-52 Chick (1910) Bad. typhosum 3 110-170 49-54 Chick (1910) Micr. pyogenes 1 29 49-53 Sattler (1928) Micr. sulfureus 4 5.3 63-80 Micr. sulfureus 2 2.3 80-99 Micr. pyogenes 4 2.6 63-80 Micr. pyogenes 4 1.4 80-99 Bad. aerogenes 4 2.5 63-80 Bad. aerogenes 2 1.7 80-99 Ps. fluorescens 4 2.0 63-80 Pink Torula 2 1.2 63-80 (Temperature coeflScient for spores) Ballner (1902) .... B. anthracis 8 10 90-105 Eijkman (1912) . . . B. anthracis 1 40 80-90 Meyer (1906). .. . 22 different bacilli 1 of each 8.3 80-100 Bigelow (1922) .... 15 different thermo- philes (see p. 323, Table 99) 1 of each 8.77 100-130 Williams (1928)... B. suUilis 1 5.8 95-105 In a similar way. Chick (1910) found the temperature coefficients of the death rates for other organisms; so MECHANISM OF DEATH 321 did various authors. The results are compiled in Table 97. We are dealing here with unusually high tem- perature coefficients, corresponding to those for the destruction of enzymes (p. 129). The only exception to these high temperature coefficients are the results obtained by Sattler (1928) on death rates of bacteria in milk. He measured the following temperature coefficients: Table 98. — Temperature Coefficients of Death by Heat in Milk (Temperature interval) 62-80°C. Average 80-99°C. I Average Micr. sulfureus Micr. pyogenes aureus Bad. aerogenes Ps. fluorescens Pink Torula 4.9 2.1 2.1 1.4 1.2 5.1 2.3 2.3 1.7 1.2 5.4 2.6 2.7 2.1 5.6 3.4 3.0 2.4 5.3 2.6 2.5 2.0 1.2 2.1 1.2 1.3 1.71.7 1.4 2.4 1.6 2.3 1.4 1.7 The reason for these low coefficients must be sought in the great range of temperatures. Skrabal (1916) has pointed out that the temperature coefficient is not constant, but must decrease with increasing rates of reaction. The change is slight with normal coefficients, but is great with abnormal coefficients such as those observed with death-rates. Skrabal computed from the results of Arthur Meyer (1906) and of Ballner (1902) the following temperature coefficients : Qio at 0°C. Qio at 100°C. ^^OTQS oi B. suhtilis 19.6 4.9 (A.Meyer) Spores of B. robur 30.8 6.3 (A. Meyer) Spores of B. anthracis 69.7 9.7 (Ballner) A striking experimental example is given by Henderson Smith (1923) for spores of Botrytis cinerea: at 31.0 to 37.0°C., Qio = 690 37.0 to 44.3 132 44.3 to 47.0 92.8 47.0 to 50.3 29.5 322 PHYSIOLOGY OF BACTERIA In Sattler's experiments, the temperature of 80°C. is so high above the maximum temperature that the temperature coefficient of the death rate must have fallen to nearly normal. It is possible to compute temperature coefficients also from endpoint experiments provided that they all have the same initial number of bacteria. Under these circumstances, the initial number, as well as the final number, is the same, the only variable at different temperatures being the time required to reduce the cells to the same number. The formula for the death- rate can be applied to the two temperatures which, for simplicity's sake, shall be assumed to be 10°C. apart, and we obtain where t^ and t'^ are the two disinfection times at the two temperatures. The temperature coefficient is the quo- tient of the two, and since log a and log b are the same in both equations, Qio = Y For temperature differences other than 10°C., the for- mula of p. 120 must be applied. Since the temperature coefficient is an exponential function of the temperature, a straight line should be obtained if the logarithms of death-rates were plotted against temperature. The above formulae show that the death rates are proportional to the times required for killing, and, therefore, the logarithms of these times MECHANISM OF DEATH 323 plotted against temperature should be a straight line in case the coefficient is constant. This is shown to be true with the spores of thermophilic bacteria by Bigelow (1922). In Table 99 are given the data as averaged by Bigelow for the fifteen most typical spore-formers. If we compute the temperature coefficients from these data, it becomes at once evident that they are not very constant, but fluctuate considerably. Table 99. — Time Required to Kill All Spores op Thermophilic Bacteria (Averages of 15 experiments by Bigelow) Heated to, KiUing times in minutes Average Qio Qio calculated from 5°C. intervals 100 105 110 115 120 125 130 135 140 788 -834 383 -405 117 -122 40 - 44 11 - 12 3.9- 4.6 1.7- 2.2 0.7- 0.9 0.6- 0.9 811 394 119.5 42 11.5 4.25 1.95 0.80 0.75 6.8] [9.4 > 10. 4J 1 l^-Q 5.9J 4.24 10.87 8.10 13.34 7.32 (4.75) (5.94) (1.14) ■8.77 The kilhng times beyond 125°C. are so short that the time required for the heat penetration must necessarily mean a considerable part of the total observed. Hence, probably all these data are too high, and the actual time required for kilUng must be much shorter. The average of all temperature coefficients from 100-125°C. is 8.77. These data are plotted logarithmically in Fig. 36. In plotting curves from data by the endpoint method, the definite points are changed to lines, and the correct point lies somewhere on this Une. This adds to the inaccuracy of plotting. Owing to this and to the fact that logarithms make deviations appear much smaller, the curves of Fig. 36 are very straight, showing hardly any fluctuations, while the temperature coefficients computed from the same data fluctuate widely. In the same figure are included some data by Weiss (1921) on the death of spores of Clostridium hotuUnum and also the data of Ballner's 324 PHYSIOLOGY OF BACTERIA (1902) on the heat sterilization of spores of B. anthracis. In one corner of the figure, the theoretical curves for different values of Qio are shown. The angle with the abscissa determines the value of Q. Fig. 36. — Relation between temperature and thermal death times of spores, illustrated by 15 thermophilic organisms, Clostridium botulinum, and B. anthracis. The slope indicates the temperature coefficient, the distance from the base the relative resistance to heat. If this curve were not plotted against logarithms of time, but against time directly, the lines would not run straight, but in parallel logarithmic curves. This type of curve is well known to the dairy bacteriologist from the picture of Fig. 37 by North, which can be found in many textbooks on milk pasteurization. The killing times are exponential functions of the decreasing temperature. Assuming Qio = 11 to hold for all bacteria, the time required to kill a bacterium at 70°C. would be eleven times as long as at 80°C., and if the pasteurizing temperature is lowered another 10°C., the killing time would have to be multiplied again by eleven. MECHANISM OF DEATH 325 The close parallelism of the lines indicates that the bacteria con- cerned as well as enzymes and albumin are all supposed to have practically the same temperature coefficient. This is not correct Fig. 37. — Lines of equal destruction by heat and time, for bacteria and compounds of milk. All temperature coefficients are extremely high (approxi- mately Qio =11) except for Vitamin C which is normal. (see Table 97). Otherwise, this well-known graph represents the behavior of bacteria towards heat quite correctly. Vitamin C has probably a normal temperature coefficient (Rahn, 1925) and its curve crosses all others. Prolonged holding at fairly low temperatures destroys more of this vitamin than short exposing 326 PHYSIOLOGY OF BACTERIA to high temperature. This is quite different from the behavior of bacteria, or enzymes, or coagulating proteins. (c) THE CAUSE OF DEATH BY HEAT In trying to find the actual chemical or physical changes in the cell causing death by heat, it will be necessary to keep in mind the very high temperature coefficient, and also the necessity of moisture. It has been shown that death by dry heat is primarily an oxidation process, with a temperature coefficient not much above normal. This makes it certain that oxidation cannot be the cause of death by moist heat. There are three groups of reactions which show high temperature coefficients: the inactivation of enzymes, the inactivation of toxins, and the coagulation of pro- teins. Let us take the coagulation of hemoglobin which was studied by Martin and Chick (1910) as an example. Table 100 is a summary of their data and shows that the coagulation of hemoglobin by heat is a time process of measurable rate, but with a very high temperature coefficient. The coagulation is a monomolecular process ; this suggests either a hydrolysis, or a splitting off of water. The temperature coefficient is 1.3 for 1°C. increase, or 13.8 for 10°C. increase. The same authors determined the temperature coefficient for the coagula- tion of egg albumin (which is not a simple monomolec- ular process, however) to 1.91 for VC or 646.0 for 10°C. increase. Dried ovalbumin, heated to 120° for five hours, remained perfectly soluble; the presence of water seems to be essential for coagulation, and this again suggests a reaction with water. The monomolecular process of the heat coagulation of albumin has a striking similarity with the death by MECHANISM OF DEATH 327 o OS o ;zi i> n U3 o '^ ^ o o S 02 O w CO W :3 o o ^ « (N o M H C o o ^ <4-l o U < tH lO iO Tfl o (N Ti< o CO »o 1—4 1—1 CO »o CO 1-H ?3 : : : : • o : °° :^ : ; : • o • 1-4 . o . . • . Ti4 . . . o • o rt o 3 « o -a 'o X T-T (M • O) .' i 1 ! ■ o o • (N* 1-1 • '. i6 • • CO . . o . . .1—1 1 • • • 05 • . . lO . • 1-i 5 • ; • CO : : : : ^ ■■■■■. '.s • . • • CO • • CX) 1-1 : ; • ^ ^ • . • 1> lO • • • T-l CO .... CO • • * '■ id • • • • (N ■ ■ ■ -.2 • ■ ■ • CO . . • T^ . • 1-H ... CO . . . . Tt4 . ... CO . . . • . K> . . . . (M . . . . Ui ; ; ; :i - O CO CO t> rt< O CO CO t^ Tti O (N »0 t^ O o CO CO CO i> O (M U3 b- O CO CO CO CO t^ 328 PHYSIOLOGY OF BACTERIA heat of bacteria. The same range of temperatures is involved in both; the temperature coefficients are very high in both cases, well above 10 for a 10°C. increase. The resemblance is also borne out by the circumstance that dry bacteria are much more resistant to heat than moist bacteria. A further similarity is the very great influence of very small amounts of acid upon protein coagulation and upon the death rate of bacteria. The theory of Chick that ''death through heat is due to the action of water upon some one protein which is essential for the life of the bacterium and that the character of this reaction is conditioned by the chemical action of water upon its constituent proteins," seems quite well founded. Temperature coefficients of ordinary chemical reactions are considered fairly constant. High temperature coefficients according to Skrabal (p. 321) should increase with decreasing rate of reaction, i.e., with lower temperatures. This explains why we ordinarily speak of definite coagulation temperatures. A drop of a few degrees makes the reaction almost immeasurably slow. In the discussion of maximal temperatures, (p. 215), it has been shown that growth and fermentation fail at this point because the rate of regeneration is slower than the rate of deterioration. We were dealing there with cell constituents such as enzymes which can be produced by the cell, and which are present in the cell in a fairly large number of molecules. Temperatures just above the maximum may lead to death by endogenous catabolism if prolonged. Death by heat which follows the logarithmic order is probably a different chemical process; it seems to be due to the inactivation of at least one most sensitive gene. It seems reasonable to compare this inactivation with the coagulation of proteins. We know nothing about the borderline between death through the destruction of genes and death through the destruction of enzymes or similar catalysts. The observation that the logarithmic order is not always followed when the rate of death is exceedingly slow, may be due to this change from one cause of death to another. MECHANISM OF DEATH 329 Perhaps, with some organisms, the most sensitive gene is inactivated before all enzyme in the cell is destroyed. The great resistance of spores to heat is accounted for by some authors by the assumption of a very con- centrated cell content. According to Almquist (1898), spores have a specific gravity of 1.35-1.40; this would indicate a very low moisture content. The data had been obtained by centrifugalizing the organisms in salt solutions of known density. In favor of a low moisture content speaks the com- paratively low temperature coefficient of the death-rate of spores, as contrasted with vegetative cells. Data on the death-rates of vegetative cells and spores of the same organism are not known to the author, but the temperature coefficient of spores seems to be fairly uniformly about 10, while that of non-sporulating bacteria is usually higher. It is not necessary, however, to resort to this expla- nation. Thermophilic bacteria are known to grow at temperatures which kill most vegetative forms; they multiply rapidly in milk during pasteurization. Actively growing cells cannot possess concentrated cell contents. Thermophilic bacteria seem to have a protoplasm which is less sensitive to temperatures between 40 and 70°C. The temperature coefficient of the death rate may not differ from that of other bacteria, but the death-rate itself may be much lower, and a higher temperature would then be necessary to bring about inactivation of the growth mechanism. The resistance of spores may be due to this same cause, i.e. to a smaller death-rate. The evidence, so far, seems to be in favor of the assumption of concentrated cell contents. The experiments of Williams (1929) suggest 330 PHYSIOLOGY OF BACTERIA another possibility, namely a lower ash content of the spores which would decrease the rate of coagulation of protein. (d) DEATH BY HEAT IN CONCENTRATED SOLUTIONS That partial dehydration protects the cells against the influence of heat, can be shown in some (unpublished) data which, in 1911, Mrs. Coffeen made upon the author's suggestion. Bacteria were suspended in broth; Table 101.— Retardation op Death in Sugar Solutions (Number of cells per c.c. of broth or broth +50% sucrose) Saccharomyces ellipsoideus Temperature 50°C. 52.5°C. 55^C. Minutes of exposure 0 5 10 20 5 10 20 5 10 20 Broth 741 328 5 402 0 120 0 7 0 367 1 158 0 9 0 275 1 62 0 Broth +50% sugar 1 Streptococcus lactis Temperature 55°C. 60°C. 65°C. Minutes of exposure .... 0 5 10 20 5 10 20 5 10 ■ 20 Broth 198 152 180 140 11 47 0 0 0 39 0 8 0 1 0 0 0 0 0 Broth and sugar 0 Bacterium prodigiosum Temperature 65°C. 67.5°C. 70°C. Minutes of exposure 0 5 10 20 5 10 20 5 10 20 Broth Broth and sugar 29,000 32,000 184 8,000 30 236 4 13 73 1,063 0 25 0 7 135 158 81 8 4 4 MECHANISM OF DEATH 331 in broth + 50 % sucrose ; and in broth saturated with salt, and were exposed to different temperatures for different lengths of time. Tables 101 and 102 give the results. A similar experiment with yeast at different sugar concentrations and at different pH was carried out by Rahn (1928). There seems little doubt that the retarda- tion of death is due to partial dehydration. Yet, the Table 102. — Retardation of Death in Saturated Salt Solutions (Cells per c.c. of broth and salt-saturated broth) Saccharomyces ellipsoideus Temperature 50°C. 52.5°C. 55°C. Minutes of exposure 0 5 10 20 5 10 20 5 10 20 Broth 960 770 61 89 6 43 0 4 3 20 0 6 0 2 6 7 2 3 0 Salt broth 0 Streptococcus lactis Temperature 55"C. 57.5°C. 60°C. Minutes of exposure 0 5 10 20 5 10 20 5 10 20 Broth 44 50 27 17 27 19 4 8 18 27 6 30 1 3 I 0 0 Salt broth 6 1 Bact. prodigiosum Temperature 67.5°C. 70°C. 72.5°C. Minutes of exposure 0 5 10 20 5 10 20 5 10 20 Broth 19,620 19,200 29 0 30 0 12 66 62 199 1 33 1 7 9 133 3 44 1 Salt broth 15 332 PHYSIOLOGY OF BACTERIA cause of death is not that of dry bacteria, it is not oxida- tion. It is some reaction with water, only the rate is decreased because the amount of available water has ^5* SS'' i5° -jiso ^y ' ^^o Fig. 38. — Survivors of young and old cultures of Bacterium coli heated for 5 minutes to different temperatures. been reduced to a point where it can no more be con- sidered as practically unlimited. (e) THERMAL DEATH POINTS The above discussions show very plainly that the old definition of the thermal death point as 'Hhe lowest temperature at which an organism would be killed in ten minutes" is without any real meaning because the killing time increases with the number of cells. MECHANISM OF DEATH 333 The thermal death-point is based on the same wrong conception as the coagulation temperature of the proteins. In reality, we have a reaction of measurable speed, but the very high temperature coefficient obscured this fact for a long time. On the other hand, the high temperature coefficient prevents the probable error in thermal death-points from being very large. If, at a certain temperature, it takes ten minutes to reduce 1,000,000 ceUs to 1 cell, it will take 11.67 minutes to kill 10,000,000 bacteria, and 13.33 minutes to kill 100,000,000, and 15 minutes to kill 1,000,- 000,000 cells. Figuring with a temperature coefficient of 20, it would require an increase in temperature of 0.51°C., 0.96°C., or 1.35°C. respectively if the larger numbers of cells were to be killed in exactly ten minutes. These increases are quite within the limits of error of the ordinary technique of determining thermal death points. The determination of thermal death-points becomes rather uncertain by the observation of Gage and Stough- ton (1906) that in old cultures of Bad. coli, the last survivors are very hard to kill; they behaved very much like spores. Figure 38 gives an illustration of the result. While the great majority of cells (more than 99.99%) is killed at 60°C. in five minutes, a few will remain alive even after five minutes' heating to 85°C. The younger cultures are more uniform in their reaction with heat, and the smoother curve of their survivors indicates a more constant temperature coefficient. This has led to the distinction of a ^'majority thermal death-point.'' Ayers and Johnson (1914) found strepto- cocci with a low '' majority thermal death-point," among which a few cells are able to survive the pasteurization temperature. ^'This ability . . . may be due to cer- tain resistant characteristics peculiar to a few cells, or may be due to some protective influence of the milk." The limitation of the thermal death-point to one experimental time of ten minutes is not very satisfactory for applied bacteriology. For the needs of food bac- 334 PHYSIOLOGY OF BACTERIA teriology, especially for milk pasteurization and for canning of vegetables and meat, this conception has been broadened to the ^'thermal death time," i.e., the time required to kill all cells at a certain temperature, as determined by the endpoint method. The thermal death time is dependent upon all the factors mentioned above. It is not more accurate than the thermal death-point, but it is more useful in applied bacteriology. A detailed discussion has been given by Esty (1928). (/) SUMMARY OF FACTS In the death of bacteria by heat, the logarithmic order is well established. Dead cells give to living cells a certain protection against heat which is too pronounced to be attributed to a decreased rate of heat transmission. At higher temperatures, the bacteria die much more rapidly. There is considerable disagreement in the temperature coefficients of death of vegetative cells. For bacterial spores, the temperature coefficient of death is approximately 10, and is constant over a temperature range of about 30°C. In concentrated solutions, death by heat is retarded. SUMMARY OF THEORIES Death by heat shows such close analogies to the heat coagulation of proteins, in temperature range as well as in the high temperature coefficients, that the general explanation of death by heat being due to coagulation of some parts of the protoplasm seems well founded. The greater resistance of dry cells corresponds well with the absence of coagulation of proteins when heated MECHANISM OF DEATH 335 in dry condition; dry bacteria at high temperatures do not die from coagulation, but from oxidation. The greater resistance to heat of thermophilic bacteria may be due to a protoplasm of a higher average coagula- tion temperature. The rate of coagulation must be considerably lower than with most cell proteins, while the general rule of the high temperature coefficient probably holds true. The greater resistance of spores may be due to the same cause, or it may be due to a high concentration of the spore contents; this would necessitate a low concen- tration of water which seems essential in the coagulation process; it may also be due to a very low ash content. The slower death rates of bacteria in concentrated solutions of sugar, or of salt, coincide well with the assumption of death as a coagulation process, and with the necessity of water for this reaction. There is no real thermal death-point, as there is no real coagulation temperature. The thermal death-point varies with the initial number of cells. But owing to the high temperature coefficients usually found in death by heat, the variation is so slight as to be within the very wide limits of error of this method. A much greater error enters from the medium in which the bacteria are tested. Water has a greatly varying effect, and a nutrient medium would be a much better test medium than water for the thermal death point. Vn. DEATH BY POISONS (a) METfiODS The Medium for Testing. — In the Chapter on Death by Heat, it has been shown that accurate results can be expected only if the bacteria are exposed to the harmful 336 PHYSIOLOGY OF BACTERIA agent in a medium suitable for growth. The same applies, in principle, to death by chemical agents, but it is frequently hard to comply with, since many of the ordinary disinfectants will react chemically with the components of the medium. This reduces the concentration of the acting poison [Kronig and Paul, (1897); Chick and Martin, 1908]. Heavy metal salts will react with proteins and form insoluble precipitates. Formaldehyde also combines chemically with proteins. Acids and alkalis will be decreased in their efficiency through the buffer action of the nutrient medium. While it may be possible to find a good synthetic medium for each kind of disinfectant to be tried, it would be necessary to adapt the media to each bacterium, as well as to each disinfectant. This complicates matters to such an extent that for the sake of uniformity, a suspension of bacteria in water seems preferable for studying the general laws of chemical disinfection. Removal of Poison after Action. — Experiments on the influence of chemical compounds may be carried out for different purposes. The food industry, in testing the efficiency of benzoate of soda, or of vinegar and spices, is not really interested in the killing of bacteria. Their purpose is fulfilled if the agents under test prevent growth and fermentation. In the disinfection of stools and of sputum, and of animal carcasses, actual killing is intended, and is accomplished without removing the disinfectant from the bacteria. For other purposes, the disinfectant is removed after action, e.g. in the disinfec- tion of hands, instruments, tools, furniture and rooms, and fomites generally. There is a great deal of difference in the efficiency of the various methods. The first way, the suppression of growth, shall not be discussed here to any extent, because it is nothing biologically definite, but a mixture MECHANISM OF DEATH 337 of growth and death problems difficult to analyze. There are, in the general technique of disinfection experiments, two distinct methods of procedure which give different results; the one is to remove the bacteria from the poisonous solution, usually together with a small part of the solution, and put them into broth or agar without any further treatment; the other is to remove by chemical or physical means any traces of the poison from the outside and inside of the cells as far as that is possible, i.e., as far as the poison has not formed irreversible compounds with the cell contents. The first method is rather unsatisfactory inasmuch as the amount of poison transferred to the medium for cultivation of the survivors is not always the same. While the amount is not sufficient to prevent growth of healthy cells, it might prevent recovery of injured bacteria. A method of merely physical removal is sometimes used by trans- ferring the cells from the disinfectant into water, centrifugalizing them off and eventually repeating the washing before placing them into the test medium. In this way, all soluble poison is removed from the outside and the inside of the cell, but chemical combinations of the poison with cell contents remain unchanged. If the viability tests on culture media are supposed to correspond with the viability in the animal, the poison must be removed completely by chemical means. The tissue around the wound in which the treated bacteria are placed has a tendency to remove the last traces of poison from within the bacterial cell, and at the same time, it is a very good medium for growth. This is shown very clearly in some experiments of Rodewald (1923). The bacterium of fowl cholera was placed in 0.1% HgCh, removed after certain time intervals, and washed four times in succession with 40 c.c. of water. After seven minutes' treatment with HgCl2, the washed bacteria showed no more growth on sterile culture media, but when injected into mice, killed them; even those Which had been in the bichloride for three hours, still caused the disease. Similar results were obtained with phenol. The action of the living tissues is perhaps partly due to their being an excellent food, but quite likely, their adsorptive properties 338 PHYSIOLOGY OF BACTERIA are also rather essential. Siipfle and Miiller (1920) showed that with HgCh, it is possible to remove the last traces by physical adsorption. Bacteria were washed once or twice after treatment with HgClo, and then were mixed with freshly heated blood charcoal. This charcoal adsorbs the bacteria, and removes from them the last traces of soluble mercuric salts. If broth is added to this mixture of charcoal and organisms directly, bacteria will grow if alive. By this method, it could be shown that Micrococci were still alive after two hours' treatment with 0.1 % HgClo, and the spores of B. anthracis survived treatment with 1% HgCl2 for forty days, with 2% HgCl2 for forty days, with 3% HgCl2 for thirty-eight days, and with 5% HgCl2 for eleven days. These data were confirmed later by Alfred Miiller (1920) who removed the mercury salt from the washed bacteria by ammonium sulfide. Data on the difference of viability of washed and chemically treated bacteria, by Gegenbauer (1921), are given extensively on p. 359: As antidotes in disinfection experiments, the following reagents are commonly used : Against HgCl2 Ammonium sulfide Against Cu salts Ammonium sulfide Against chlorine Na2S03 Against formaldehyde Ammonia Against peroxide Permanganate Against permanganate Sulfides Against phenol, lysol, etc Washing with dilute ammonia (no very efficient antidote is known) In the chapter on Death by Heat, experiments by Lange (1922) were mentioned demonstrating that the addition of large numbers of dead bacteria protected the living bacteria against heat. The same is true with disinfectants; addition of large numbers of heat- killed cells decreases the toxic effect of a disinfecting solution. MECHANISM OF DEATH 339 (6) THE ORDER OF DEATH The logarithmic order of death is the most common form in chemical poisoning of bacteria if the disinfectant is removed at the end of the experiment. The table below shows the action of phenol upon the spores of B. anthracis as determined by Chick (1908). Table 103. — Death of Spores of B. anthracis in 5% Phenol at 20.2°C. Time, hours Survivors E- 1 1 " 0 434 0.5 410 0.049 1.6 351 0.061 2.7 331 0.044 3.92 299 0.041 5.95 241 0.043 25.60 28 0.046 Mean 0 047 A second example shows the very slow death of Bad. typhosum in buffer solutions of different pH, according to Cohen (1922). At pH 5 and 6.4 where the death rate is very low, it decreases while it is quite constant at pH 7.1 and 8.7. This might be expected if the decrease of the death rate is explained by variation in resistance. In very weak disinfectants, the exceedingly slow rate of death gives the most resistant bacteria a better chance to work against the destructive agent. The logarithmic order requires a special justification here because two components are reacting with each other, namely, the bacterial cells (or some essential part of their protoplasm) and the poison. A logarithmic 340 PHYSIOLOGY OF BACTERIA order can be established only if one of the two reacting agents is in such great excess that its concentration Table 104. — Death of Bad. typhosum in m/500 Buffer Solutions OF Different Hydrogen Ion Concentrations pH 3.8 5.0 6.4 Exposure in hours Survivors K Survivors K Survivors K 0 3 6 12 24 48 231,000 200 sterile 1,021 543,000 443,000 365,000 277,000 253,000 160,000 0.0295 0.0288 0.0244 0.0138 0.0111 1,000,000 620,000 568,000 378,000 205,000 0.0346 0.0205 0.0176 0.0143 pH 7.1 8.7 Exposure in hours Survivors K Survivors K 0 3 6 705,000 93,300 0.146 0.099 0.121 0.122 391,000 102,000 30 . 400 0.194 0 1 85 12 24 48 46,000 875 1 2,000 12 sterile 0 0 .191 .188 remains practically constant despite the reaction. Falk and Winslow (1926) come to the conclusion that the bacterial cells are the agent whose concentration remains constant : "If it be assumed that the fundamental reaction of disinfection is unimolecular, we might consider that of the two reacting agents, (1) disinfectant and (2) living cells, one or the other is present in excess and that its concentration is not changing significantly during the MECHANISM OF DEATH 341 course of the reaction. Inasmuch as the constant K obviously changes with a change in the salt concentration, it must be assumed that the reacting substances of the bacterial cell (whether those cells are dead or alive) must always be present in excess if anything approaching the monomolecular curve is to be attained." The author holds the opposite view. The monomolec- ular type of reaction is based on the decreasing numbers of living cells. A decrease of the concentration of poison has never been observed, nor does it enter the mathematical equation at all. The assumption that the reagent whose decrease was observed, i.e., the Hving cells, is, in reahty, constant, and that the other reagent about whose concentration noth- ing is known, is the decreasing compound, seems unjusti- fied. The argument of Falk and Winslow that the rate of reaction changes with the concentration of poison, is no proof for their contention. The change in the concentration of poison molecules can be but slight. To prove this, let us consider the action of 0.01 % of HgCl2 upon a suspension of Bad. coli with 1,000,000 cells per c.c. The amount of HgCl2 per 100 c.c. is 10 mg; the weight of the bacteria (Table 132, p. 397) in 100 c.c. is 100 X 10^ X 8 X IQT^^ = 0.08 mg. Of this weight, 70% is moisture, the dry weight being only 0.024 mg. The ratio of HgCl2 to bacteria bodies is 10:0.024. In order to compare molecules with molecules, these two weights must be divided by their molecular weight. That of HgCU is 271.52; that of the bacteria cannot be stated; the weight given includes cell wall and cell contents; only parts of the latter will react with the bichloride. But assuming the entire weight to consist of protoplasm with a molecular weight of 10,000, we find the ratio of the numbers of acting molecules to be 0.037:0.000,0024, or 15,000 molecules of HgCh for each molecule of bacterial protoplasm. Even if 4 or 10 or 100 of these 15,000 poison molecules should react with one of proto- plasm, their concentration would remain practically unchanged. 342 PHYSIOLOGY OF BACTERIA (c) CHEMICAL CONSIDERATIONS REGARDING THE ACTION OF DISINFECTANTS The study of disinfection on a biochemical basis dates back to the investigation of Kronig and Paul (1897) which might well be called a classic paper on disinfection. This paper shows that the disinfectant power of salts or acids depends upon their ionization. This was proved by exposing spores of B. anthracis to various mercury salts and counting the number of colonies developing after the action of the salts had been interrupted by washing with a 3 % solution of ammonium sulfide. In the following table, the salts are arranged in order of decreasing ionization. The survivors increase with decreasing ionization. Table 105. — Action op Mercury Salts upon Spores of B. anthracis Salt solution, Percentage of salts, % Survivors after molar 20 minutes 85 minutes HeClz Hg 1.68 0.42 0.56 1.57 (16 colonies after 6 minutes) 7 34 00 0 HgCh Ka. . . 0 HgBr2 }^A. 0 Hg(CN)2H6 33 The same importance of ionization was also shown by the decrease of efficiency of these solutions whenever their mercury ion concentration was reduced by addition of other salts with the same anion. In a similar way, the disinfection by salts of silver, gold, copper has been tested, and efforts were made to determine the disinfectant value of the metal salt anions. Further investigations concerned the acids and alkalies. The next table shows that neither molar nor normal MECHANISM OF DEATH 343 solutions of different acids have the same disinfectant value, because their hydrogen ion concentrations differ. But the Ke normal solutions of the stronger acids are completely, and therefore, equally dissociated, their Table 106. — Decrease op Disinfectant Power with Decrease op Ionization (All solutions He molar) Survivors after Salt mixture 8 minutes HgCla 8 HgCla + NaCl 32 HgClo + 2NaCl 124 HgCla + 3NaCl 282 HgCl2 + 4NaCl 382 HgCl2 + 4.6NaCl 410 HgCl2 + 6NaCl 803 HgClz + lONaCl 1,087 Table 107. — Disinfectant Power op Acids Tested on Spores OP B. anthrads Acid Concen- tration Survivors after 20 min. 120 min. 5.42 hrs. 8.25 hrs. Concen- tration Sur- vivors after 2.3 days HCl HCIO4 HNO3 CHCI2CO2H CCI3CO2H HF HBr H2SO4 H2SO4 H3PO4 HCO2H CH3CO2H (C02H)2 normal normal normal normal normal normal normal molar molar normal normal normal 191 385 123 0 243 139 0 0 217 0 0 4 491 424 2,500 97 0 0 1 405 285 1,090 2,280 2,780 34 Ke normal Ke normal 3^6 normal }4 normal He normal 160 283 54 396 382 344 PHYSIOLOGY OF BACTERIA action is fairly alike, and equal that of an n/8 di- chloracetic acid which has the same hydrogen ion concentration. This table shows also that the anions of hydrofluoric acid, nitric acid and trichloracetic acid have a specific toxic action, which becomes less conspicuous upon dilution as compared with the hydrogen ion effect. Further experiments indicate that the alkalies dis- infect in proportion to their hydroxyl ions. A compari- son of the data shows that the hydroxyl ions do not act as strongly as the hydrogen ions; this is especially true for Micrococci. The same authors (Kronig and Paul) studied, further, the effect of oxidants such as chlorine, bromine, iodine, permanganate, chlorate, bichromate, persulfate, and hydrogen peroxide, also of organic disinfectants like phenol and its derivatives. The effect of alcohol upon the disinfectant value has been studied quite extensively. It is not the object of this book to go into these details. Kronig and Paul's paper, now over 30 years old, started a new era in disinfection experiments, and is still a model of carefully planned and executed technique. It was the necessary prerequisite for the discovery of the logarithmic order of death which followed ten years later. After this discovery, various disinfectants were tried on a number of microorganisms in a quantitative way, but the test organisms in all these experiments were limited to a few species, and the kinds of disinfectants used also were few. The final outcome of all these investigations, which were very numerous in the years 1908-1912, has not fulfilled the expectations; they have had very little influence upon practical disinfection. The main reason for this lies in the fact that ordinarily, no result can be duplicated exactly, not even by the same author. While we MECHANISM OF DEATH 345 have learned to eliminate some errors, such as using distilled water of unknown pH, or producing heat shocks and cold shocks, the variation of parallel experiments on different days is still very great. The only certain way to get uniform data is to prepare a large number of bacteria in a uniform way, and preserve them for later experi- mentation. This was done by Paul (1910) in drying a Micrococcus suspension on garnets, and by Myers (1929) in keeping a suspension of bacterial spores frozen. Both these authors obtained comparable data over long periods of time. The old type of disinfection experiments by the endpoint method suffers naturally from the same irregularities, but since the results are not as precise as in plate count experiments, the variation is not so obvious. It is not logical, therefore, to reject data and findings obtained by the newer methods, because their more accurate measure- ments reveals plainly a great variability of the microorganisms while the cruder endpoint method does not show this so distinctly. Very little attention has been paid so far to the specificity of the different species. It is to be expected that among unicellular organisms, differences in resist- ance will be much greater than among multicellular organisms. This is evidenced by bacteria capable of feeding on formaldehyde (Kaserer, 1906), xylol (Soehn- gen, 1906), metabolizing arsenic {Penicillium brevicaule), or living at pH 1 (Waksman's sulfur bacteria, 1922). A first step in this direction is the observation of Cooper and Mason (1927, 28) that the Pseudomonas group (Ps. fluorescenSy Ps. non-liquefaciens and Ps. pyocyanea) is quite sensitive to ^'physico-chemical" reagents causing precipitates of proteins, such as alcohol or phenol, while Bad. coli is less sensitive to these com- pounds, but reacts strongly with ''chemical" disin- fectants such as quinones. The Pseudomonas group which is sensitive to precipi- tants is also much more sensitive to heat; this suggests that the colloidal state of their protoplasm is different from that of the colon group. 346 PHYSIOLOGY OF BACTERIA The analogy between the action of hot water and that of alcohols and phenols is borne out by experiments with thermophilic bacteria. These organisms are somewhat less sensitive to phenol than Bad. coli, but more sus- ceptible to p.-nitrosodimethyl-aniline. The consequence of this specificity is that the phenol coefficient (p. 353) which indicates how many times more toxic a certain disinfectant is than phenol, must be expected to vary considerably with different organisms. This has been observed by several authors . For example, the following phenol coefficients of sodium hypochlorite were obtained with different bacteria by Zoller and Eaton (1923). Mycob. tuberculosis 42 . 8 BacL alcaligenes 100 . 0 B. anthrads 160 Bact. neapolitanum 200 Gonococcus 280 Proteus vulgare 330 {d) EFFECT OF CONCENTRATION OF POISON UPON DEATH RATE The influence of the concentration of a poison upon the death rate has been formulated according to physico- chemical principles first by Watson (1908), after Ikeda (1897) and Chick (1908) had given empirical formulae. Later investigators came independently to the same equations, namely Paul, Birstein and Reuss (1910b) and Phelps (1911). Assuming that n molecules of the disinfectant must react with one cell before the cell loses the power of reproduction, the reaction constant, or death rate, must be proportional to the nth power of the concentration. K = A:c" MECHANISM OF DEATH 347 If the X- values for two different concentrations are known, we have K^ = A;ci^ K, = kCi"" Dividing one equation by the other, we obtain Xt {c.y K, = W and consequently n - ^ - log Zi - log Ci - 10gX2 • log C2 The equation for the logarithmic order of death, which was K't = log5 for any one concentration of a poison, is then changed to i^.^c- = log5 when different concentrations are concerned. Applied to the endpoint method where r^ and consequently log 7 is constant, we must find the relation log ^2 — log ^1 t ' c^ = constant, or tiCi" = ^2^2**, ovn = , — log c This computation of the concentration exponent is made under the assumption that the poison combines chemically with some molecules of the cell to a definite 348 PHYSIOLOGY OF BACTERIA compound, and that this is the cause of death. The next sub-chapter (p. 361) will show that this has not yet been proved. The reaction of one molecule of the cell constituents with n mole- cules of poison does not contradict the analogy with monomolecular reactions. It is assumed that the concentration of the poison mole- cules remains practically constant during the reaction (see p. 341) because they are in great excess over the active molecules of the cell contents. If each cell contained only one '' reacting molecule" as the theory of Rahn (see p. 284) assumes, n must be an integer, providing that the cause of death is due to a compound of this molecule with the poison. If several molecules of the cell must be affected in order to cause death, the value of n need not be an integer (Watson, 1908). A very instructive study of the effect of concentration of different acids was pubhshed by Paul, Birstein and Reuss (1910b). The death rate of Micr. pyogenes Table 108. — Death Rate AND Concentration op Acid Normal- ity H-ions per Hter, mg. Death rate n Acid From normal- ity From H-ions Hydrochloric 1:150 1:100 1:75 1:50 1:25 0.7 1.0 1.3 2.0 4.0 0.035 0.042 0.049 0.062 0.087 0.45 0.54 0.58 0.49 0.45 0.54 0.58 0.49 Acetic 1:31.17 1:17.65 1:2 0.75 1.00 3.00 0.018 0.034 0.234 1.12 0.93 0.89 2.21 1.89 1.75 Butyric 1:26.0 1:14.7 1:66 0.75 1.00 1.50 0.018 0.057 0.294 2.04 2.03 2.04 4.00 4.03 4.04 MECHANISM OF DEATH 349 aureus was measured with different concentrations of hydrochloric, acetic and butyric acid. The data are summarized in Table 108. Fig. 39. — The death rates of Micrococcus pyogenes by increasing concentra- tions of 3 different acids. An important conclusion is evident from these data, namely, that two disinfectants may, at one concentration, have the same disinfecting power while at other con- centrations, they may differ widely, owing to a different concentration exponent n. The extrapolation of the data of Table 108 gives the curves shown in Fig. 39 where the ordinates indicate death rates, i.e., the intensity of the acid as a disinfectant. The concentra- 350 PHYSIOLOGY OF BACTERIA tion is given by weight, and not by hydrogen ions. The death rate by acetic acid, (with n = 1) is propor- tional to concentration, and therefore a straight line. As has been shown before, the death rate in acid solutions depends primarily upon the hydrogen ion concentration. From Table 108, it is evident that in hydrochloric acid, the death rate is pro- portional to the square root of the hydrogen ions {n = 0.5) while with acetic acid, it is proportional to the square (n = 2) and with butyric acid, proportional to the 4th power of the hydrogen ion concentration (n — 4). This is hard to explain, even if we assume a special effect of the kations or the undissociated molecules. Neither these results nor the investigations of Winslow and Lochridge (1906) and of Norton and Hsu (1916) are based on direct pH measurements. Since the development of quicker methods of direct pH measurement, two papers have given extensive data on this point, those of Cohen (1922) and of Myers (1929). Table 104 p. 340 by Cohen covers a wide range of pH . If the death-rate were proportional to the H+ or 0H~ concentration, then the rate divided by this concentra- tion should give a constant. This is not the case; but a constancy over a wider range is obtained if we divide the rate by the square root of H+ or 0H~ concen- tration {n = 3^^). The same is true with Myers' experiments on the death of spores of a bacillus similar to B. cereus, in buffered solutions of highly alkaline washing powders. All of Myers' results are expressed by the time required to kill 99% of the bacteria, since the increasing death rate (see p. 296) made the determination of an average death rate impractical. Excluding the one apparently wrong value of experi- ment No. 10, there is a good consistency of results. Myers found the products of the kilHng-time and 0H~ concentration to be constant for pH 12 to 13. This means that the death-rate is proportional to the 0H~ con- MECHANISM OF DEATH 351 a> 1 '-S 03 O ?* .g j3 --^ ^ C CJ a X o i " u a k.a 0^ '■+3 " a -p o .1^ §i X 8 1 5 ?^ 2 TJ OS a ^ -^ o^ 02 ^ f-l a as o o • - I d 03 S .2 -£3 ft gi^ w a I >> f^ .t^ ^§ « o -(-3 3 o 02 fl o ;a^ o OQ 5 d OS o (M ^ 00 ^g28g^ O i-H (N »C 00 O (N (N O O o o o o o 2^;^?5g OS CO OS OS * CO rj< d (N rH C^ rH i-H OS ">* O (N CO I> 00 1-1 00 (N i-i i-( 1-1 CO rt< O O O 00 "^ O -^ 00 1-1 -* CO »C T}^ CO TjT O O O CO C l> lO lO (N lO »0 lO ic »0 lO o iO »o »0 lO lO lO rH lO o o o o o o o o o o o o o o o o O O 82 03 ^ ^^ w w w w w o o o o o c3 03 ^ ^ ^ ^ ^ ^ ^ 12; 82 2§§§g ^ 03 ce c3 c3 U Ph 03 ea 03 c3 03 03 03 03 tH (N CO "«i< "5 CO t^ 00 OS o I— ( rH (N CO "* lO CO 1—1 1> 00 I— ( tH 352 PHYSIOLOGY OF BACTERIA centration. This proportionality does not hold for the lower pH range. Here, a much better agreement is obtained with the square root of the 0H~ concentration . From pH 9.1 — 12.4, it is satisfactory. This is a wider range of pH than we had before, and even the values for pH 7.5 are not entirely out of range. At pH 13, the agreement does not hold at all. This proportionaHty to the square root of the 0H~ concentration agrees with the just mentioned data of Cohen's. Theoretically, n must increase at very low concentrations. There is a limit for any poison below which it will not harm the cell; it might even stimulate. In other words, K becomes infinitely small when C is still measurable. The value for n is computed from the equation: n = If C2 becomes so small that there is no death, the death rate K2 k\ becomes 0, and 7- = «, and therefore n — w. The change of n K2 from a definite number to infinite is not abrupt but gradual, as will be seen from Fig. 41, p. 358, and from the calculations of n on p. 361. This must be kept in mind when the concentration exponent is determined ; it will be constant only beyond a certain limit. Reichel (see p. 356) called this limit a and did his calculations by assuming K = Ko{c - a)\ A number of concentration exponents have been determined, and they are compiled in Table 110. They are mostly computed from the concentrations by weight though it would be more correct probably, in most cases, to consider ionization. In some instances, both ions will have a deleterious effect, while other com- pounds are not ionized at all. The computation has MECHANISM OF DEATH 353 been limited therefore, to the simplest way of measur- ing concentrations. It seems that different ?2-values can be obtained for the same disinfectant with different bacteria. A comprehensive study of the concentration exponents would seem quite desirable. The table seems to indicate that aside from the value 0.5, the n-values are very nearly integers. Whatever may be the theoretical explanation of these n-values, their existence is beyond doubt. The great differences for various disinfectants and bacteria make comparisons almost impossible. Chick (1908) and Phelps (1911) have both called attention to the fact that the customary method of stating the dis- infectant power by the ^^ phenol coefficient" or ''carbolic acid coefficient" cannot possibly give results which allow a general application. Table 110. — Concentration Exponents for the Potency of Various Disinfectants HgCl2 (anthrax spores) n = 0.25 Alfred Miiller (1920) p. 362 Oxygen (dry bacteria) 0.5 Paul (1910) p. 305 HCl (Micrococci) 0.5 Paul, Birstein and Reuss (1910b) p. 348 H2O2 0.5 Reichel, 1908 HgCU (Micrococci) 0.5 Gegenbauer (1921) p. 360 AgNOs 0.86 Chick (1908) Acetic acid 1.0 Paul, Birstein and Reuss (1910b) p. 348 Formaldehyde 1.0 Gegenbauer (1922) p. 362 HCl (anthrax spores) 1.5 Gegenbauer and Reichel (1913) Butyric acid 2 . 04 Paul, Birstein and Reuss (1910b) p. 348 HgCla (BacL typhosum) 3.8 Chick (1908) HgCl2 (anthrax spores) 4.9 Chick (1908) Phenol (Bad. typhosum) 4.0 Reichel (1909) p. 357 Phenol (Micrococci) 4.0 Reichel (1909) p. 357 Phenol {Bad. paratyphosum) . . . 5.5 Chick (1908) Phenol (anthrax spores) 5 5 Chick (1908) Phenol {Bad. typhosum) 6.0 Lee and Gilbert (1918) 354 PHYSIOLOGY OF BACTERIA The method originally suggested by Rideal and Walker (1903) was to determine the minimum time required to kill a bacterial suspension by a certain concentration of the disinfectant under test, then to determine the phenol concentration required to kill the same organism in the same time. Dividing the one concentration by the other, the ''phenol coefficient" was obtained which stated how many times Table 1 1 1.— Determination OP THE Phenol Coefficient Dilution 2.5 5 7.5 10 12.5 15 minutes Phenol 1:90 + 0 + 0 + 0 0 0 100 0 0 0 Unknown disinfectant . . . 500 + + 0 0 0 0 550 + + + 0 0 0 600 + + + + 0 0 550 : 100 = 5 . 5 is the Phenol coefficient. Table 112. — Phenol Coefficients of HgCh and AgNOa Disinfectant Time of disinfection, min. Concentration per cents Phenol coefficient HgCh 2.5 2.5 10 10 30 30 2.5 2.5 10 10 50 50 0.088 1.20 0.006 1.04 0.0018 0.990 0.0066 1.2 0.0015 1.04 0.0009 0.83 13.6 CoHcOH HgClz 173 CeHfiOH HgCl2 550 CeHsOH AgNOa 176 CeHeOH AgNOa 693 CeHsOH AgNOs CeHsOH 922 MECHANISM OF DEATH 355 more or less efficient the disinfectant under test was than phenol (see Table 111). This comparison was inaccurate. On account of the differences in the concentration exponent n, two disinfectants having equal power at a certain concentration may vary widely if they are both diluted at the same ratio (see Fig. 39). Chick (1908) brought out this point very strongly by showing that the phenol coefficient was a function of the killing time, as is shown in the data of Table 112. As a result of these investigations, all modern methods of com- paring disinfectants including the Rideal-Walker test have standard- ized the time in which the organisms must be killed. This is of limited value only, however, since the phenol coefficient holds true only for the time set in the experiment, e.g. ten minutes with Bad. coll, and gives not even an approximate conception of the concentra- tion required to kill in thirty minutes or in five minutes. The choice of phenol as the standard was especially unlucky since it has the highest concentration coefficient of all disinfectants. If we realize further, that the selection properties of bacteria interfere, and that we obtain different phenol coefficients with differ- ent bacteria (see p. 346), there seems to be little left to rely upon in the Rideal-Walker test. But no better or more practical method has so far been suggested for standardization. About the various kinds of phenol coefficients, their history, their virtues and objections, and the factors influencing their accuracy, see Tanner and Wallace (1929). (e) THE MECHANISM OF POISONING A very thorough attempt to obtain a knowledge of the mechanism of poisonous action was made by Reichel (1909) . He measured the distribution of phenol between oil and water, and the influence of salt upon this partition (Fig. 40), demonstrating that addition of salt increases the phenol concentration in the oil. The same experi- ment was then repeated with coagulated albumin instead of oil, and the same relation was found to take place. The ratio of phenol in the solid albumin phase and in the watery phase was plainly that of a partition by physical solubiUties, and not of adsorption or chemical 356 PHYSIOLOGY OF BACTERIA combination. Bacterial bodies showed this same rela- tion. In all cases, salt changed the ratio, but not the principle of distribution. Uater / i ■30^ .20^ y -to ^ ^^«/ s ,0 IS 10 26-zA^Cf Fig. 40. — Distribution of phenol between oil and salt solutions of different concentration. Then, the disinfecting value of phenol solutions with and without NaCl was tried by the endpoint method. The results show a good agreement with the partition of phenol between water and albumin, especially when the killing time is at least one hour. The relation of the killing times to the phenol con- centration is given by the equation time X {c - aY = K where c is the phenol concentration, a is an initial minimum concentration of phenol showing any effect MECHANISM OF DEATH 357 at all, and n the concentration exponent. The exponent n is found by the equation log t2 — log ti log (C2 — a) — log (ci — a) which corresponds to the formula used for the endpoint method of disinfection (p. 347) except that a was not considered there. Reichel found for Bad. typhosum time . (c - 0.0023)4 = 4.6 X lO'^ and for Micr. pyogenes aureus time . (c - 0.0045)4 = 45.5 X 10-» He came to the conclusion that the bacteria die when a certain phenol concentration is reached inside of the cell. He did not try to really find the fundamen- tal chemical reaction. In 1921, Gegenbauer investigated the fundamental reaction of poisoning by bichloride of mercury. His work started as a duplication of Reichel's experiments, but the results soon proved to be quite different. There was a true partition of the bichloride between coagulated serum albumin and water. HgCU is soluble as such in the solidified albumin. This is not an adsorption process. But besides this, part of the mercury salt reacts with the protein, forming an insoluble mercury- proteinate which cannot be washed out. The amount of Hg bound by the albumin is independent of the concentration; this proves it to be a real chemical compound. Addition of H2S to a piece of albumin which contains dissolved HgCl2 will precipitate HgS at once. But even in albumin which has been washed for eight to ten days after contact with HgCU, and which contains no more 358 PHYSIOLOGY OF BACTERIA dissolved HgCl2, hydrogen sulfide will precipitate HgS from the protein compound, and the protein is liberated again. These experiments were repeated with yeast and gave the same general results. Then, Gegenbauer made disinfection experiments with Micr. pyogenes in different concentrations of HgCl2, <^ / 2 3 tf S 10 iO 30 ¥) 50 75- 100 /toun 2Q0 Fig. 41. — Influence of the concentration of HgCh upon the length of time required to prevent growth of Micrococcus pyogenes when the poison is removed by washing (left curve) and when it is removed by an antidote (right curve). in one series washing the bacteria after disinfection repeatedly with water, in the other treating them with H2S to remove or counteract the poison chemically. The results are summarized in Table 113. There are nine series with chemical treatment, and three series with washing only. Parentheses in the table indicate MECHANISM OF DEATH 359 that the value given is not an average, but is the only datum available. Table 113. — Disinfection op Micr. pyogenes aureus with HgCl2 Times required for complete disinfection Concentration If cells are washed If cells are treated of HgCl2, with water with H2S % Growth, No growth, Growth, No growth, hours hours hours hours 0.01 7 13 208 265 0.025 108 149 0.050 1.3 2.3 72 101 0.100 1.3 2.3 36 59 0.250 (2) (3) 34 56 0.500 1.3 2.3 28 48 1.000 1.0 1.7 16 33 2.000 (1.0) (2.0) 6 11 3.000 2.3 3.7 Figure 41 gives these data in graphic form. The logarithms of time were chosen for plotting to make the relations more outstanding. The two limiting times for each concentration are connected by lines. The correct point must be somewhere between these limits. This graph resembles very much that of Fig. 36, and the influence of Qio is here substituted by that of n. The slope of the curve is an indicator of the magnitude of n. In solutions above 0.05% HgCU, the resistance of the bacteria was not affected by the concentration of the poison, unless the bacteria were treated with an antidote This indicates the formation of a chemical combination of the protoplasm with the mercury salt which prevents 360 PHYSIOLOGY OF BACTERIA reproduction. The independence of the formation of this compound from the concentration of the poison suggests an analogy with the compound observed in serum albumin and yeast. The amount of this com- pound was also independent of the bichloride concentra- tion provided that enough time was allowed for the formation. The minimum time for Micr, pyogenes to lose the power of reproduction under the conditions of this experiment was about one hour. This insoluble compound of some part of the pro- toplasm with mercury did not kill the cell, however. When the compound was broken up chemically by the addition of H2S, the mercury was precipitated as HgS, and the protein was set free again, ready to take up its normal function. The bacteria were not really dead after one or two hours' contact with HgCl2, but reproduced after treatment with H2S, if administered within a certain time. Only when the dormant state was quite prolonged, death would set in, and the cells could no longer recover after chemical treatment. Thus, the action of HgCU is two-fold. First, there is the ''stunning effect"; a dormant stage is produced; the cell ceases to function. It might be compared with a stone in the cogwheels of the cell machinery. The removal of this ''stone" by chemical means proves that the cell is still alive. At the same time, a slower chemical process takes place which kills the cells. This process is independent of the first. The nature of this second process is not clear. It depends upon the concentration of the bichloride. Above 0.1% HgCl2, this relation has been expressed by Gegenbauer in the equation killing time X [HgCU]^-^ = 1.5 MECHANISM OF DEATH 361 The concentration exponent varied considerably, between 0.36 and 1.23 %, and averaged 0.65. Four experiments between 0.025 and 0.1% averaged 0.78, and three experiments between 0.01 and 1% averaged 0.88. The cause of this increase of n at very low concentra- tions of poison has been explained on p. 352. Death is not caused by the formation of the mercury- protein compound. Neither is it caused by the libera- tion of HCI from HgCU. It is not the mere interruption of the cell activity by the ^' stone in the cogwheels," for this would make the killing time independent of the concentration. Any of the above explanations would fit the case of the anthrax spores in Table 114a, but not the Micrococcus. Spores of B. anthracis were tested with HgCU by Gegenbauer with the same technique, and the death time proved to be independent of the concentration, with and without antidote : Table 114a. — Disinfection (End OP Spores of B. Point Method) anthracis by HgCU Concentration of HgCU 0.01% 0.05% 0.1% 0.5% 1.0% 2.0% 3.0% Growth after 110 115 95 100 95 100 95 100 95 100 95 100 95 days 100 days No growth after Mtiller's data (1920) do not agree with those obtained by Gegenbauer. Only one table of Mliller's can be used for this comparison. The critical times for anthrax spores at 37°C., were: Table 114b Concentration of HgCU 0.1% 1.0% 2.0% 3.0% 4.0% 5.0% Growth after 9(15) 12 7 9 7 9 6 7 6 7 5 days 6 days No growth after 362 PHYSIOLOGY OF BACTERIA This table resembles that of Gegenbauer's for the Micrococci. Its concentration exponent is n = 0.25, between 1.0% and 5.0%, or n = 0.23 between 0.1% and 5%. This picture of disinfection is completed by an experiment with cyanide of mercury. The action of this salt upon coagulated albumin and yeast is quite different. In yeast, no definite compound of mercury with protein had been formed in ten days, and with serum albumin, very little was found after fifteen days. The distribution factor for the dissolved salt between protein and Hquid was smaller. The disinfection experiments agree with this behavior. Since no insoluble protein compound is formed promptly, chemical removal of the mercury salt has no effect upon the killing time. Death, except for the very lowest concentrations, is practically independent of the concentration of the poison. The "fundamental reaction" must be of a different nature here than in the case of the bichloride. Gegenbauer (1922) studied further the mechanism of formaldehyde disinfection. He found a true physical partition of formaldehyde between water and oil, as in the cases of phenol and HgCl2 mentioned before; with yeast, however, there was no physical solution at all in the solid phase, but a well-defined chemical compound was formed. The amount of formaldehyde bound by yeast was independent of the concentration of aldehyde. However, for shorter times, less than six hours, the amount of formaldehyde bound by the yeast protoplasm was proportional to the formaldehyde concentration. This is shown in the disinfection experiments. Both with Micro- cocci and with anthrax spores, the rate of disinfection was quite nearly proportional to the formaldehyde concentration, which means n = I. Washing with water or counteraction of the poison with ammonia did not alter the death rate materially. This is to be expected, since there is no dissolved formaldehyde in the protoplasm; however, the chemical compound between formaldehyde and protoplasm is so stable and irreversible that it is not changed back by water or ammonia. In summary, it has been shown in this chapter that some very interesting efforts have been made to obtain a MECHANISM OF DEATH 363 clearer conception of the fundamental reaction between poison and cell, but they did not lead to the end. In ReicheFs experiments, death occurred as soon as a certain concentration of phenol was present in the cells. How- ever, since phenol formed no chemical compound with the cell, we still do not know the real cause of death. In Gegenbauer's research with HgCU, a chemical compound was formed with protoplasm, and this com- pound prevented multiplication; it could be broken up again by H2S, however, and the cell might be still alive, or might be dead, the one or the other depending upon the time of action and the concentration of disinfectant. The real cause of death is still unknown. (/) THE TEMPERATURE COEFFICIENT OF POISONING The influence of temperature upon the rate of chemical disinfection was first studied by Madsen and Nyman (1907) and by Chick (1908). Most of the latter's data were obtained by the endpoint method. Some Table 115. — Temperature Coefficients of Disinfection with Bacterium paratyphosum HgCl2:0.1% HgCl2:0.01% AgNOs: 0.017% AgNO3:0.0017% 25°C. 20°C. 15^C. 10°C. 5°C. 0°C. 3.3 4.0 2.6 5.8 Qio 40°C. 30°C. 20°C. 3.3 3.4 2.8 10°C. 2.7 0°C. 40^C. 3.3 30°C. 2.9 20°C. 2.3 10°C. 40°C. 1 35°C. 30°C. 25°C. 20°C. 15°C. 3.9 2.4 2.6 3.0 364 PHYSIOLOGY OF BACTERIA of the data are shown in Table 115. Usually, the temperature coefficients were but slightly above normal, but occasionally, they went much higher. Chick made the very remarkable observation that the temperature coefficient increased with an increase in the number of cells used for the test. Table 116. — Dependence of the Temperature Coefficient UPON the Concentration of Bacteria 1 % phenol 0.8 % phenol 0.6 % phenol Cells per unit volume Temp. range, °C. Qio Cells per unit volume Temp. range, °C. Qio Cells per unit volume Temp. range, °C. Qio 1,000 750,000 21-11 21-11 3.3 6.0 440.000 76,000,000 187,000 56,000,000 21-11 21-11 21-11 21-11 6.0 >8.2 5.5 12.2 6,600 750,000 110,000 16,000,000 31-21 31-21 31-21 31-21 4.3 10.4 5.5 10.3 Young cultures show a higher temperature coefficient than old ones. This leads to the consequence that at very low temperatures, the death rates of young and old cells must be nearly alike while ordinarily young cells are much more sensitive (see p. 289). These still unexplained complications of the tempera- ture effect upon chemical disinfection should caution us against hasty conclusions. Several very extensive experiments on the influence of temperature have been published by Gegenbauer and Reichel (1913) concerning the action of hydrochloric acid on spores of B. anthracis, including the change of acid concentration, and the addition of NaCl. The temperature coefficient is found to be independent of the tempera- ture, averaging 4.6, but it decreases as the concentration of the acid increases. MECHANISM OF DEATH 365 Paul, Birstein and Reuss (1910c) also give a large number of temperature coefficients for the death of Micrococcus by hydrochloric acid. Recently, this problem has been treated with a different technique by Cooper and Haines (1929), and these authors thought they had proved that three different groups of disinfectants existed, one show- ing no temperature coefficient at all, one with a normal one, and one, very high, ranging from 8 to 20 for an increase of temperature from 20°C. to 37°C. Their technique was practically that of the phenol coefficient, for they measured the concentrations just sufficient to kill the bacteria in thirty minutes. The temperature coefficient was obtained by dividing one concentration by the other. This method of computing the temperature coefficient is not permissable, because it impUes that the disinfectant action is pro- portional to the concentration while we have seen in the preceding chapter that this is by no means always the case. The error of these authors can be shown easily by going back to our standard equation for the death rate, on p. 347 At37°C: KiCiH = log At20°C: K2C2H = log ~ 0 In the endpoint method, log r is constant, t was experimentally, equalized to thirty minutes, therefore KiCx- = K2C2- The relation between the two death rates is given by the formula (p. 127). K\ = KiQn where Qn is the temperature coefficient for the increase in tempera- ture from 20°C. to 37°C. ^'' = ©' 366 PHYSIOLOGY OF BACTERIA C Only a n = 1, we have Q17 = ^r as Cooper and Haines assumed; but (^ 1 for n larger or smaller, this ratio varies. Applying to the experimental data by Cooper and Haines the values w = 4 for phenol and n = 0.5 for H2O2, (Table 110) and assuming arbitrarily n = 0.5 for toluquinone, the following tem- perature coefficients are obtained: for phenol, Qio = 4.2 instead of 1.43 by Cooper and Haines for H.2O2, 1.16 instead of 1.30 for toluquinone, 2.58 instead of 6.6 For the first two compounds, similar values had been claimed which are probably not similar, and the high coefficients for substances like toluquinone might be explained by a small value of n. The claims of Cooper and Haines are improbable, and cannot be con- sidered proved until all w-values have been ascertained. (g) OTHER THEORIES OF CHEMICAL DISINFECTION In 1919, Traube made a very general statement that all toxic action is brought about by surface reactions. He also extended this theory to disinfection: ^'If we disregard the oxidizing disinfectants like ozone, peroxide, and hypochloric acid, it may well be claimed that the action of the other disinfectants is based exclusively on physical processes. Salts of heavy metals, like mercury or copper, cause death by irreversible pre- cipitates, while most of the acids, bases, and organic disinfectants, largely by means of their very strong swelling and solving properties, produce such changes in the bacterial cells that they must perish." The two papers are limited to the discussion of a few series of disinfectants where a proportionality between surface tension depression and disinfecting power exists, but even in these selected examples, exceptions occur. While it cannot be denied that depression of surface tension might be a cause of MECHANISM OF DEATH 367 death (see next chapter), the theory of Traube seems far too generalized to be accepted as such. It is already partly disproved by the researches of Reichel and Gegenbauer (p. 363). The same criticism applies to a similar theory by Davis (1927). A different explanation of toxic action upon cells was given by Rahn (1915) who believed that chemical poisons are catalysts for certain destructive reactions in the cell. He used the well-known parallelism between heat effects and poison effects, (see p. 137) assuming that the poison accelerates the constructive processes as well as the destructive ones, but that the latter are affected more strongly. However, it seems most probable that death i.e. loss of power of reproduction by chemical poisoning is caused by the reaction of at least one gene with the poison. Fermentation and similar life processes might still continue after the cell has lost the power to repro- duce. This is the only cause of death likely to give a logarithmic order of death. Possibly, some chemicals may not affect the chromo- somes, but inactivate enzymes, destroy membranes or produce other damage. In this case, many molecules must have doubtless reacted before the cell is dead, and this must result in a period of no deaths, a bulging sur- vivor curve and an increasing death rate. Gaidukov (1910) defined death as a change from a hydrosol to a hydrogel. He considered the dying of the protoplasm to be a coagulation process. Twenty years later, Bancroft and Richter (1931) came to the same conclusion. They demonstrated the coagulation in the dark field, but failed to prove that the cells showing this symptom were dead. Probably, they had been dead for a considerable time, and coagulation of the 368 PHYSIOLOGY OF BACTERIA protoplasm is perhaps one of the later stages of the death process, and not one of the causes. (h) SUMMARY OF FACTS Bacteria die from chemical poisons in the same general order that is found with death by heat, or by drying. Bacteria, which, by chemical poisoning, have lost the power to multiply, may regain it if an antidote is used. This complicates the definition of death. The same disinfectant does not act equally strong upon all organisms. Certain genera and species are more resistant to a certain poison than others. Well known examples of special resistance are that of tubercle bacteria against alkali, of aciduric bacteria against acid, of many molds against formaldehyde. The toxicity of a disinfecting solution increases with its concentration, but not always in direct proportion. Different disinfectants behave quite differently, and two different disinfecting solutions of equal toxicity may vary widely if they are both diluted with the same amount of water. The disinfecting power of a solution is increased by an increase of temperature. SUMMARY OF THEORIES The fundamental reaction that causes death when a chemical poison acts upon a living cell is still unknown, despite several promising attempts to determine it. Several theories exist; we may assume a chemical com- bination of the poison with essential molecules, or changes of interface tension which bring about a col- loidal state of protoplasm, or a catalysis of some normal catabolic cell process by the poison. MECHANISM OF DEATH 369 With HgCl2, the formation of mercury-proteinate does not seem to be the cause of death, but only of dor- mancy, because the administration of H2S as antidote brings the cells back to life unless the dormancy has been very prolonged. The death rate is sometimes proportional to the concentration, but more commonly, it is proportional to the nth power of the concentration where n is known to vary from 0.5 to 4 or even 6. This means that the relative strength of two different disinfectants cannot be expressed by a simple factor unless the comparison is limited to a definite range of concentration, or, as in the case of the Rideal-Walker method, to only one definite killing time. No better method has as yet been found to compare disinfectants. For low concentrations, the exponent becomes larger. At very low concentrations, the toxic effect disappears entirely, and a stimulation is frequently observed. The temperature coefficient of disinfection shows some queer anomalies. It increases with an increasing concentration of bacteria, and it is higher with young cells than with old ones, but as a rule, is normal, or only slightly above normal. VIII. DEATH BY SURFACE TENSION DEPRESSION The theory that all death by chemical poisoning is caused by surface tension depression has already been discussed on p. 366 and found to be an unsupported generalization. However, in solutions of very low surface tension, death is probably brought about primarily by physical rather than chemical causes. The first experiments of Ayers, Rupp and Johnson (1923) showed no parallelism between surface tension and death rate. These authors, therefore, considered 370 PHYSIOLOGY OF BACTERIA the chemical effect of the surface tension depressant upon the cell to be the real cause of death. However the measurements refer to the tension between medium and air, and this need not be proportional to that between cell and liquid. Pizarro (1927) also believes that chemical reaction is the real cause of death. Some light has been thrown on this question by the experiments of Frobisher (1927) who studied the effi- ciency of disinfectants in the presence of chemically inert surface tension depressants (soaps). He found that 0.1% of sodium oleate had no effect on phenol disinfection, but 0.25% increased the poisonous effect, and 0.5% prevented it. The latter is explained by the law of Gibbs and Thompson that surface tension depress- ants are strongly adsorbed on the surfaces. The layer of adsorbed soap around the cell may become so thick that phenol cannot readily penetrate to the cell. An excess of sodium oleate also decreased the action of hexyl resorcinol. Hansen (1922) found also that saponin or soap had frequently but little or no effect, and some- times even counteracted the efficiency of disinfectants. Leonard and Feirer (1927) experimented with hexyl resorcinol which was the most powerful depressant they could find, reducing the surface tension of water from 73 to 34 dynes, but reducing that of glycerol only from 72.5 to 67. In pure glycerol, this substance did not kill bacteria readily while in water, a 0.1% solution killed bac- teria in fifteen seconds. The authors claim the slower action in glycerol to be a proof that surface tension depression is really the killing agent. With more knowledge concerning the protective action of concentrated solutions against other causes of death (see p. 330), the lower death rate of bacteria in glycerol + hexyl resor- cinol might be accounted for on a different basis. But the pro- portionality between surface tension and disinfecting power of the different substituted resorcinols is a good indication that here we are dealing with a physical rather than a chemical cause of death. MECHANISM OF DEATH 371 For this reason, the order of death by depressed surface tension would be very enlightening, but no data about this are known to the author. IX. DEATH BY LIGHT It is an old experience that bacteria are killed by light and also by rays of shorter wave lengths. The order of death is usually logarithmic. Table 117 gives the data obtained by Lee and Gilbert (1918) on the death of Bad. coli by the light of a 100 Watt lamp. The death rate is a decreasing one and indicates an heterogeneous culture. Header (1926) calls attention to the fact that only certain wave lengths of the sunlight will kill bacteria, and that the action of sunlight is different at different times of the day. Two causes of death are possible: either the rays act upon the cell, or they act upon the medium, producing in it harmful substances, as, e.g., H2O2. In this latter Table 117. — Death of Bad. coli by the Light from a 100 Watt Lamp Time of exposure, Survivors K minutes 0 107 2 92 0.021 3 86 0.031 5 81 0.024 10 74 0.016 15 65 0.014 20 56 0.014 25 49 . 0.014 30 43 0.013 40 33 0.013 50 27 0.012 60 16 0.013 372 PHYSIOLOGY OF BACTERIA case, irradiated water should be toxic. This has been found not to be true (Browning and Russ, 1917, Coblentz and Fulton, 1924). Consequently, the light acts upon the cells directly. The observation of a very low tem- perature coefficient, and the fact that the pH is of practi- cally no influence upon the death rate by ultraviolet light (Bayne-Jones and van der Lingen, 1923, Gates, 1929) point in the same direction. A large number of experiments have been carried out with ultra-violet light. The results by Gates (1929) are among the few derived from monochromatic light. The test organism (usually Micr. pyogenes) was spread on agar surfaces, exposed to the rays, and the survivors counted by the colonies developing. With MicrococcuSj the logarithmic order was not established (see p. 296). Exposure for thirty minutes to wave lengths of 3340 and 3660 A.u. did not decrease the number of living cells noticeably, and even 3130 A.u. seemed^ to have little effect. Wave lengths of 2500 to 3000 A.u. were quite efficient, and even at 2250 A.u., bactericidal action was observed, though the intensity of this wave length was very low with the lamp used by Gates. The temperature coefficient is very nearly 1, indicating that death is primarily due to a photochemical reaction. Another group of rays has been frequently proved to be harmful to bacteria, namely the X-rays. Holweck (1929) and Lacassagne (1929) computed from their experiments with Ps. pyocyanea that this organism must have a sensitive zone of about 0.5/z diameter. If this zone is hit by one ''quantum,'^ the cell dies. If the cell is hit outside of this zone, it does not die. Wyckoff (1930) made computations for Bad. coli from experiments with wave lengths from 0.56 to 3.98 A.u., and came to the conclusion that the measured zone is MECHANISM OF DEATH 373 Table 118. — Death op Bad. coli by X-rays of Various Wave Lengths Intensity of light measured by ion- ization of air 73.1 94.1 173.1 274.5 133.8 e.s.u. Wave length, A.u.. . 0.564 0.710 1.537 2.29 3.98 Line of light used . . . K K K K L Target metal Ag Mo Cu Cr Ag Seconds Percentage of survivors 5 71.4 68.7 62.9 10 86.3 20 67.6 69.3 51.2 39.2 68.9 30 61.2 27.7 29.3 40 44.3 49.1 22.0 18.5 51.8 50 12.8 60 33.4 38.0 14.9 10.0 32.2 80 30.0 90 19.2 26.3 120 12.4 17 3 Seconds Death rates (0.434^) 5 0.0292 10 0.00640 0.0163 0.0201 20 0.00850 0.00796 0.0145 0.0203 0.00809 30 0.00711 0.0186 0.0178 40 0.00884 0.00771 0.0164 0.0183 0.00714 50 0.0178 60 0.00794 0.00700 0.0138 0.0167 0.00820 80 0.00653 90 0.00796 0.00644 120 0.00755 0 . 00635 really the sphere of action of the hitting quantum. This sphere decreases with the wave length and proves the sensitive zone to be quite small. The zone as such, is not more than one-hundredth of the volume of the 374 PHYSIOLOGY OF BACTERIA cell, perhaps much smaller. For a coccus of 1/x diameter, the sensitive zone would correspond to a sphere of 0.215m diameter. The logarithmic order is well established, as may be seen from Table 118. The same is true for cathode rays, as observed by Wyckoff and Rivers (1930) (see Fig. 35, p. 297). If death is caused by the hit of a single quantum, the death rate by a uniform source of X-rays should be the same for all species of bacteria having the same sensitive zone. Wyckoff found it to be identical for the two species he tried, i.e.. Bad. coli and Bad. aertryke. SUMMARY The exposure of bacteria to light will cause them to die if the intensity of the light is strong enough. Ultra- violet light is very efficient in wave lengths between 2500 and 3000 A.u. The effect of light is not brought about by a chemical change of the medium through the light ; the fundamental reaction which causes the death is produced inside of the cell by the light. The temperature coefficient is very low, indicating a physical or photochemical reaction as the cause of death. Death by X-rays and cathode rays is caused by the bombardment of the bacteria with electrons which will kill the cell if they hit its sensitive zone. This sensitive zone has a volume about one-hundredth of the cell. The order of death by rays is logarithmic. X. DEATH THROUGH AGE In the chapter on the endpoint of growth, the cause of death by old age of bacterial cultures has already been MECHANISM OF DEATH 375 discussed, and has been shown to be due to one of three possible causes. Whether the cause of death is starvation, injurious action by fermentation products, or by some specific, thermolabile compounds, the order of death should be logarithmic. As a matter of fact, it is very nearly logarithmic. In Table 130, p. 390, two experiments on the death of Ps. fluorescens by suffocation are given. The two checks, with air, show a normal growth until a maximum Table 119. — Death of Strept. lactis in Milk Culture Age of culture, Plate count Lactic acid Death rate days per c.c. in medium, % 1 1,137,000,000 0.625 2 905,000,000 0.900 0.099 4 572,000,000 0.972 0.099 7 129,000,000 1.026 0.159 10 2,360,000 0.990 0.298 13 3,600,000 0.208 is reached; this is followed by a decrease, which is death by old age. The successive death rates, without omission, after passing the maximum number, are : Exp. 1 0.0051 0.0052 0.0058 0.0045 0.0050 0.0040 0.0036 0.0038 0.0029 0.0026 0.0022 0.0022 Exp. II 0.0115 0.0042 0.0060 0.0054 0.0047 0.0048 0.0043 0.0031 0.0039 0.0034 0 . 0031 0 . 0026 ^ 0 . 0026 0 . 0024 0 . 0024 The order of death allows for no conclusion regarding its cause, but the very poor medium, together with the very low maximum number of only seventy millions per 376 PHYSIOLOGY OF BACTERIA cubic centimeter (compared with about one billion in broth) makes starvation the most probable cause. A case where fermentation products are the most likely- cause of the decrease in living cells has been mentioned by Rahn (1911) for a culture of a lactic streptococcus in milk. The death rates in this experiment (Table 119) increase, but this is probably due to an increase of lactic acid. Even after the maximum num- ber has been passed and the cells already have begun to decrease, the acid still increases, and this increase must necessarily increase the death rate. Another set of data we owe to Chesney (1916) who studied the death of pneumococci in old broth cultures. The death rate in this case is quite constant. Probably, death is brought about by chemical action of the products of metabolism. Buchanan and Fulmer (1928) distinguished between a period of accelerated death rate, and a period of constant death rate. The first-named phase must necessarily exist. Bacteria, after having grown to a maximum number, will not die suddenly at a constant rate. There is a gradual increase of the products beyond the tolerable concentration, and the increase in death rate under these conditions has been shown in Table 119. The period of accelerated death is just an intermediate stage between maintenance and death, and involves no new biologic principle. SUMMARY Death of bacteria in old cultures may be due to one of the following three causes: starvation, fermentation products, or specific thermolabile metabolic products. The existence of the third cause has been questioned for most species, but is not disproved. The order of death gives no clue as to which of the possible causes is acting. The cause of death by old age does not seem to lie in the cell, but in the unfavorable environment. MECHANISM OF DEATH 377 XI. DEATH BY STARVATION (a) STARVATION IN WATER Death of bacteria by starvation has received con- siderable attention on account of its importance in water bacteriology, and in the self -purification of lakes and rivers. Many data have been accumulated for this reason, but the interpretation is very difficult, and in many cases impossible. One difficulty lies in the fact that bacteria, like all other organisms, will be able to live for a certain time on their reserve substances, with- out any food; death by starvation will start only after these stored food substances are used up. This ''period of no deaths" will vary considerably. Another difficulty is founded in the fact that the change from a culture medium to distilled water, or to a balanced salt solution, might injure or kill many cells for reasons other than starvation, e.g., by plasmolysis, leaching out of impor- tant body substances, change of acidity, change of inter- face tension, traces of toxic compounds in water. Theoretically, the order of death from starvation is not logarithmic; there should be a period of no deaths before the cells begin to die. Cohen (1922) calls this the induction period. This initial period of no death or slow death is characterized by an increase of the death rate. A large set of data on the death of Bad. coli and Bad, typhosum is given by Hinds (1916). The death rates decreased quite decidedly. Very probably, this was caused by "cold shock.'' To show this more clearly, the death rates have been computed for each time interval, averaged, and recomputed on the basis that this average is 100. The result is shown in Table 120, which represents 378 PHYSIOLOGY OF BACTERIA the average of 5 to 6 experiments. The bacteria had been grown at 37° for twelve hours before being placed in the water. At 8°C., the death rate was very high during the first six hours, at 20°C., not quite so high, and only for two hours, while there is nothing abnormal at all at 37°C. The numbers in brackets are not averages; each represents but one single datum. The very high death rate for the first few hours, which was observed only at low temperatures, is very probably due to cold-shock. Those cells which survived the cold-shock died more slowly from starvation. Table 120. — Relative Death Rates of Bad. coli and BacL typhosum in Water (Standardized to the average = 100) Death rate during, Bad. coli Bad. typhosum hours 8°C. 20°C. 37°C. 8°C. 20°C. 37°C. 0- 2 2- 6 6- 12 12- 24 24- 36 36- 48 48- 72 72- 96 96-144 211 180 128 72 57 54 45 41 (32) 187 130 115 70 80 90 97 47 (59) 105 70 100 105 (136) (154) (194) 140 117 100 84 85 69 52 154 160 105 92 78 80 73 66 126 84 90 Average 100 100 100 100 100 100 A typical starvation experiment is the following by- Rogers (1918) which shows a rapid decline of bacteria in sewage if dialized in running water, after the first day. Here, we see plainly the short induction period due to reserve food materials in the cells, and the increas- ing death rate. Quite extensive experiments were carried out by Cohen (1922) who studied the death of Bad. typhosum and of MECHANISM OF DEATH 379 Bad. coll in buffered mineral solutions of different pH, as well as in tap water and distilled water. The death rate for Bad. coli is lowest between pH 5 and 6.4, and increases if the pH is increased or decreased beyond these limits (see p. 340). Table 121. — Colon Counts in Dilute Sewage Held in Parchment Bag in Running Water Days Colon cells K After 0 190,000 1 130,000 0.165 2 19,000 0.500 3 9,000 0.441 4 20 0.994 7 30 0.543 0.5 The death rates of Cohen's experiments show very clearly the effect of reserve food material, which causes a period of very slow death, as indicated by an increase of the death rate. At 0°C., this is overshadowed by a cold-shock effect, which even at 10°C. is still noticeable. Table 122. — Temperature Coefficients of Starvation Bad. typhosum Bad . coli °c. K Qio K Qio 0 1.186 1.62 0.0176 2.12 10 1.910 1.63 0.0373 4.36 20 2.928 1.77 0.1654 3.76 30 5.176 .... 0.6214 380 PHYSIOLOGY OF BACTERIA Cohen also computed the temperature coefficients of starvation. These coefficients are too low because at lower temperatures, the cold-shock is superposed, which increased the death rate. It is interesting to note that Hinds as well as Cohen found that Bad. typhosum had a lower temperature coefficient than Bad. coli. (1.71 against 4.30 between 20°C. and 37°C.) (6) STARVATION IN SALT SOLUTIONS While the above data relate to bacteria with practically no food whatever, other data are available where at least the minerals were supplied. Zeug (1920) found that standard physiological salt solu- tion caused some bacteria to die faster than they did with distilled water. He then tried salt solutions which would be less toxic to bacteria, and found that solutions of one salt were quite unsatisfactory, but mixtures of several salts were capable of keeping bacteria alive for a few days without much loss of viability. Most sensitive against 0.85% NaCl were Micrococcus pyogenes, Bad. vulgare, Vibrio Metchnikovi, while Bad. acidi ladici, Bad. septicemiae hemorrhagicae and Ps. pyocyanea decreased but little in three days and Bad. typhi murium even increased. Zeug then set out to find suitable solutions for the three sensitive organisms. The result of this study was that different species required different salt mixtures for long viability. The solution best fitted for one was not satisfactory for the other. This is shown in Table 123, where the optimum mixture for Micr. pyogenes is seen to cause a fairly rapid death of Bad. vulgare, while the solution best suited for Bad. vulgare is not very satisfactory for Micr. pyogenes. Some of the best combinations are shown in Table 124. MECHANISM OF DEATH 381 Zeug did not succeed in finding a suitable mineral solution for Vibrio Metchnikovi, Without calcium lac- Table 123. — Death op Bact. vulgare and Micr. pyogenes in Different Salt Solutions Solution Test organism Survivors after Start 1 day 2 days 3 days 4 days Suited for Micr. pyogenes. Micr. pyogenes. Bact. vulgare . . 235,000 145,000 210,000 100,000 210,000 30,000 200,000 200,000 8,000 0 Suited for Bact. vulgare. Micr. pyogenes. Bact. vulgare . . 145,000 80,000 150,000 76,000 110,000 83,000 18,000 78,000 0 75,000 Table 12 4 For Vibrio Metchnikovi, per cent For Micr. pyogenes For Bact. a h c d e vulgare NaCl 0.5 0.5 0.1 0.5 0.5 0.5 0.1 0.5 0.50.50.5 0.5 0.075 0.6 KCl 0.5 0.1 CaCla 0.6 MgCl2 0.5 0:4 0.1 0.1 0.4 0.1 0.1 0.5 MgS04 Ca-Lactate ... 0 5 0 1 tate, this species died fairly fast. This presence of organic compounds is experimentally objectionable because lactate might be food for the Vibrio; it was suffici- ent to produce multiplication with Bact. vulgare. Zeug emphasizes that there are many different mixtures equally well suited for keeping cells alive. The best mineral medium would be the ideal medium for starvation experiments. Table 123 shows that 382 PHYSIOLOGY OF BACTERIA there is scarcely any death for four days if cells are kept in a well balanced salt solution. This makes it very clear that in most of the experiments mentioned before, death was not really caused by starvation. It should be emphasized here that Zeug always made parallel experiments in plain distilled water and in doubly distilled water; viability was greatly prolonged by the second distillation. Here also should be mentioned the experiments of Wilson (1922) on the viability of Bact. suipestifer in Ringers solution (600 mg. NaCl, 75 mg. KCl, 100 mg. CaCl2, 100 mg. K2CO, 1 liter H2O). There was little Table 125- -Percentage of Survivors of Bacteria Centrifuged in Water and in Broth Washed and B. cereus B. megatherium Bact. ^ ^ ,. J. . Bact. coh prodigiosum Broth Water Broth Water Broth Water Broth Water Before centrifug- ing After centri- f uging One hour later 100 88 43 100 0 0 100 197 300 100 0 0 100 44 37 100 0 0 100 70 70 100 280 131 difference between the death rates in the original solu- tion and in different dilutions, as far as 1 part Ringer solution + 15 parts of water. Winslow and Brooke (1927) verified the results of Zeug's studies. Different bacteria were washed from agar slants into distilled water, centrifuged, resuspended in fresh distilled water, and recentrifuged. MECHANISM OF DEATH 383 Check experiments were made with broth. Table 125 gives the num- bers of survivors, in percentages. It is quite evident that Bad. coli is the only one of the organisms tested which can survive in distilled water for any length of time. It should be borne in mind, therefore, that results on starvation with Bad. coli are exceptional cases, and must not be generalized. These authors then determined the limits of dilution which would keep cells of B. cereus alive for one hour after centrifugation, and found that broth diluted 1:100 protected the bacteria perfectly, broth diluted 1:1,000 only for a short time. Ringer's solution, and sugar solutions of different concentration offered no protection. A peptone concentration of 0.005% or a meat extract concentration of 0.003 % kept the cells alive, but one-tenth of these concentrations caused a rapid reduction of living cells. Death in the last experiments is so rapid that it cannot very well be explained as being caused by starvation. Since Ringer's solution and isotonic sugar solution offered no relief while colloids in very low concentrations sufficed to save the life of the cells, it seems quite possible that some toxic compounds were contained in the distilled water which were absorbed or chemically bound by the colloids. That bacteria will die from starvation within one hour is quite improbable. (c) STARVATION FROM LACK OF NITROGENOUS FOOD Chemical changes taking place in starving cells have already been discussed in Chapter I, on pp. 10 to 18. Those experiments refer mostly to yeast kept in a sucrose solution without nitrogenous compounds. These are the only experiments known to the author as referring to death in the presence of fermentable material, or energy food, but in the absence of nitrogenous material. They correspond to experiments on protein-free diet with animals. Two experiments by Rubner (1913) are sufficiently detailed to give data for the computation of successive death rates. In both experiments, the yeast remained for twenty-four hours in a 10% sugar solution, was then 384 PHYSIOLOGY OF BACTERIA separated from this solution by centrifugation, and was resuspended in a new sugar solution. This was repeated after each twenty-four hour period. The death rates increase in both experiments for the first five days. This is in accordance with our expecta- Table 126. — Yeast in Daily Renewed Sugar Solution Without Nitrogenous Food Plate counts Death rates Rate of nitrogen loss II II II Start After 1 day. After 2 days After 3 days After 4 days After 5 days After 6 days 20,355 20,898 13,728 824 218 8. 38,700 24,800 12,300 2,300 187 80 6.2 0.000 0.086 0.464 0.492 0.679 0.586 0.193 0.249 0.409 0.579 0.537 0.074 0.074 0.075 0.074 0.075 0.087 0.038 0.058 0.062 0.070 0.073 0.074 tion. The rate of nitrogen loss of the yeast cells also increases slightly in one experiment while it is constant in the other. There is no strict parallelism between loss of reproductive power and loss of nitrogen. (d) THE FUNDAMENTAL REACTION IN STARVATION Leaving aside all the possible reactions which might cause death in starvation experiments without being essential to starvation, such as cold-shock, traces of toxic compounds in water, diffusion of electrolytes, etc., two possible causes present themselves : it may be either a hydrolysis, or an oxidation of some essential molecules in the cell. Both of these reactions have been already MECHANISM OF DEATH 385 shown to occur, oxidation in the case of dry cells, hydrolysis (or some similar reaction with water) in death by heat. It seems easy to decide whether oxidation is a cause of death; in this case, starvation should not kill the cells so rapidly in the absence of oxygen. Such experiments (p. 388) have shown Bad. coli to die more slowly in the absence of oxygen, while starving Bad. typhosum died more readily without oxygen. If these experiments can be considered representative, oxidation may be the main death cause of some bacteria, and not of others. This agrees with our experience that some bacteria are killed by oxygen even in the presence of food, namely, the obligate anaerobes (see also p. 203). Oxidation might then be the cause of death of starving cells of Bad. coli. Something is oxidized in the cells which, in normally nourished cells, would be reduced again. Reductions require energy; energy requires food. Table 127. — Rate of Death of Micr. pyogenes at Diffekent Temperatures Death by 49°C. 40°C. 30°C. 20°C. 10°C. 0°C. Oxidation High Low 0.76 0.055 0.247 0.017 0.078 0.0055 0.0247 0.0017 0.0078 0.0005 0.0024 0.00017 Hydrolysis.. . High Low 41.4 7.5 1.60 0.285 0.055 0.0099 0.0019 0.00035 0.000067 0.000012 0.0000023 0.0000005 Typhoid bacteria die faster without oxygen. An hydrolysis is probably the cause, which, in air, can be counteracted by the cells for a while, but not in the absence of air. Possibly, Bact. typhosum has some reserve substance which can be utilized only with oxygen, as e.g., fat. 386 PHYSIOLOGY OF BACTERIA Since the death rate by oxidation is known for dry Micrococci (p. 308) and the death rate by heat has also been determined, (p. 320), the rates from the two causes can be calculated for all tem- peratures, assuming the temperature coefficient for oxidation to be 3.16, and that of hydrolysis to be 28.9. Table 127 gives these. While these calculations have no sound basis at all, and are entirely speculative, they illustrate the possibility that a bacterium in the same medium might die at 20°C. from oxidation (death by oxidation being ten times as rapid as by hydrolysis), while at 40°C., it may die from hydrolysis (death from hydrolysis here being six to twenty times as rapid as that by oxidation). (e) SUMMARY OF FACTS Bacteria and yeasts die if food is lacking. The decrease of living cells is more rapid in water than in balanced salt solutions. In well balanced solutions, which are specific for each species, the bacteria may show hardly any decrease of viable cells for three to four days. If death is accelerated in water, it seems safe to assume that they do not die from starvation, but from other causes. One of these has been found to be cold-shock. The greater viability, if the water used for making salt solutions is distilled twice, suggests the presence of traces of toxic substances in the distilled water. These are counteracted by peptone or meat extract, but not by sugar or balanced salt solutions. Most of the experiments supposed to be starvation experiments really present death by other causes. The true death by starvation must show a survivor curve starting parallel to the abscissa, and must show an increasing death rate. In the absence of oxygen, starving cells of Bad. coli die more slowly, while starving cells of Bad. typhosum die more rapidly. Bad. coli can survive in water very much longer than most other bacteria. MECHANISM OF DEATH 387 SUMMARY OF THEORIES Death by starvation does not show the logarithmic order of death. The survivor curve bulges above the straight line. In balanced salt solutions, the death rate by starva- tion is independent of the salt concentration, within fairly wide limits. The death of yeast cells on a protein-free diet shows the typical bulging starvation curve. The death rate is not proportional to the rate of loss of nitrogen from the cells. The cause of death by starvation may be either an oxidation or a hydrolysis of some cell compound which cannot be counteracted by the cell if food is lacking. With Bad. coli, oxidation seems to be the main cause, while with Bad. typhosum^ oxidation is certainly not the main cause. The temperature coefficient of death by starvation is not known accurately. All experiments have such an evident effect of cold-shock at the lower temperature that the true rate of death by starvation at low tempera- tures is not known. All such temperature coefficients mentioned in literature are very probably too low. XII. DEATH BY SUFFOCATION Suffocation shall be defined here as death by the absence of oxygen. All animals die ultimately in the absence of oxygen, though some lower animals, like the intestinal parasites, can live for a considerable period without it, and thrive better at a low oxygen tension. With bacteria, death from this cause is not so well known. Some experiments by Whipple and Mayer (1906) show that Bad. typhosum in tap water dies more 388 PHYSIOLOGY OF BACTERIA rapidly if air is excluded (see Table 128). Bad. coli was tested by the same authors in nitrogen atmosphere, and displayed no appreciable difference in the death rate. Table 128. — Death op BacL typhosum in Tap Water, with Various Gases Days Air Hydrogen Cells per c.c. Death rate Cells per c.c. Death rate 0 600,000 600,000 2 455,000 0.138 2,400 2.76 4 190,000 0.300 25 2.54 8 120,000 0.200 0 12 67,000 0.185 0 18 25,000 0.176 0 26 9,250 0.161 0 33 2,150 0.171 0 40 132 0.210 0 47 6 0.223 0 54 0 0 0.196 2.65 Air Carbon dioxide Days Cells per c.c. Death rate Cells per c.c. Death rate 0 400,000 400,000 2 265,000 0.206 110,000 1.400/ ^•^*** 4 1,500 8 50,000 0.260 0 This result was verified by Hinds (1916). He studied the effect of nitrogen and hydrogen upon starving bac- teria, and the average death rates found are given in Table 129. The rates are then computed for the death rate in air as standard of 100; we find that BacL typhosum MECHANISM OF DEATH 389 dies more rapidly when the air is replaced by other gases, while Bad. coli dies more slowly. This might be caused by different types of reserve compounds; glycogen could be used as a source of energy in the absence of oxygen, while fat could not. However, it might also be that the cell contents of Bad. coli are more readily oxidized than Bad. tyhposum. The com- plication of suffocation with starvation gives no good data, and more information could probably be obtained with bacteria under normal growing conditions. Table 129. -Death Rates of Starving Cells of Bad. coli and Bad. typhosum in Various Gases Bad. coli Bad. typh osum Air Nitrogen Hydrogen Air Nitrogen Hydrogen Series I Series II Series III ... . 0.336 0.359 0.359 0.285 0.153 0.064 0.238 0.318 0.207 0.100 0.365 0.118 0.817 0.100 0.607 Relative death rates Series I Series II Series III ... . 100 100 100 85 43 18 71 89 88 100 100 117 224 99 166 Average 100 49 83 100 171 133 Whipple & Mayer 100 94 100 1,350 Some excellent experiments on death by suffocation of nourished cells are given by Miiller (1912) who grew Ps. fluorescens and Bad. coli in a medium containing only 40 mg. of asparagin and 60 mg. of ammonium lactate per liter, besides minerals. This medium was 390 PHYSIOLOGY OF BACTERIA Table 130. — Development and Oxygen Consumption op Ps. fluorescens Age of culture hours Plate count per mm. 3 Open flask Closed bottle Dissolved oxygen per liter, mg. Oxygen consumed per hour by 10» cells Age of culture hours Plate count per mm. 2 Open flask Closed bottle 0 3K 8 UK 18 21 23 25 26 27 28 33 37 42 4 4 8.825 5 5 8.825 9 12 8.819 19 23 8.800 53 61 8.695 187 349 8.277 618 621 7.699 1,654 2,200 6.214 3,920 3,596 3.867 4,855 1.759 9,991 0.000 11,436 14,333 45,655 2,654 68,342 396 67,782 260 46>^ 64,700 52 72,823 0.13 mg. 5sy2 69,742 0.30 74K 56,859 0.93 98M 42,574 0.81 1223^ 29,409 0.45 mn 21,429 0.69 1943^ 14,583 0.41 2183^ 15,833 0.50 2i2y2 15,167 (0.25) 266K 11,204 2903^ 14,565 362>^ 11,428 3863^ 12,604 410^ 12,772 80 75 85 27 23 15 68 12 Second experiment 0 2K 7 103^ 13>^ 17 17^ 183^ 20 22 24 27 32 36 41 69 69 8.833 86 104 8.821 109 250 8.660 255 507 8.221 432 1,008 7.315 1,690 3,884 3.355 5,600 1.969 9,847 0.000 9,250 13,167 29,129 8,500 3,452 58,259 1,003 64,701 570 64,420 405 77,306 23 0.059 0.22 0.35 0.42 0.57 0.59 (0.26) 45>^ 51 67M 733^ 973^ 121K U9H 1933^ 217K 2413^ 2653^ 2893^ 3613^ 3853^ 4093^ 78,146 67,221 69,182 53,217 40,613 34,171 19,326 17,926 23,247 13,583 13,892 13,780 11,428 12,268 11,204 38 50 29 3 1.6 0.9 0.1 0.2 SO poor that even Bad, coli could not grow on it in the absence of oxygen. Mliller distributed the inoculated MECHANISM OF DEATH 391 medium over a large number of glass-stoppered bottles which were filled completely so as to leave no air space. The bacteria multiplied readily until the dissolved oxygen in the stoppered bottles was used up. This was deter- mined by chemical analysis. The growth is shown in Table 130 for Ps. fluorescens, and in Table 46, p. 187 for Bad. coli. In both cases, growth continues for about one hour after the oxygen supply is exhausted, and then the bacteria begin to die. The death rates were found to decrease decidedly. Death of Bad. coli by suffocation is very slow as compared with that of Ps. fluorescens, the death-rate being only about one-tenth. The data suggest a logarithmic order of death, with considerable variability of resistance of the individual cells. There is an indica- Table 131. — Successive Death Rates op Suffocating Cells Ps. fluor- escens. I II. 0.146 0.021 0.095 0.050 0.173 0.016 0.145 0.047 0.124 0.160 0.039 0.122 0.135 0.021 0.095 0.095 0.023 0.048 0.131 0.020 0.039 0.100 0.029 0.078 Bad. coli. I II 0.0156 0.0015 0.0050 0.0110 0.0019 0.0121 0.0132 0.0075 0.0032 0.0033 0.0011 0.0018 0.0003 0.0013 0.0005 0.0012 tion of a short initial increase in the death rate; this may be just the period of changing over from no death to the final death rate. The continuation of multiplication for about one hour after the oxygen is completely used up suggests storage of oxygen, either physically or chemically, and deserves attention. Possibly, the absence of oxygen produces a compound in the cell which hinders further cell activity. These experiments offer no opportunity to decide whether this is the case. The fundamental reaction of death by suffocation is still unknown. 392 PHYSIOLOGY OF BACTERIA A possible line of attack of this problem is suggested by the experiments of Meyerhof (1916a) who found that a culture of Nitrobacter after having starved for nitrite for twenty hours at 35°C., will oxidize as rapidly as before starvation began; the same culture, after having been without oxygen for the same time, had lost half of its oxidizing capacity. The cell mechanism had evidently been hurt by the long absence of oxygen, but not by the absence of nitrite. SUMMARY Certain bacteria will die in the absence of food more rapidly when oxygen is removed. This does not hold true for all bacteria. If bacteria are cultivated in a medium which cannot be used by them without presence of oxygen, they will die from suffocation soon after the oxygen supply is exhausted. XIII. GRAND SUMMARY OF THEORIES OF DEATH It can be considered quite well established that dying is a chemical process. On the curve indicating the gradual changes in the cell under adverse conditions, death is not a definite point. Death is a matter of definition. The bacteriologist's definition that death is the loss of reproductive power of the cell is satisfactory for bacteriological purposes; it probably is satisfactory for other purposes as well, as long as the reproductive power of each cell, and not of each organism, is con- sidered. If different definitions of death are chosen, different results might be expected. In most cases, bacteria die in a definite mathematical way approaching the logarithmic order closely enough MECHANISM OF DEATH 393 to allow of the computation of a death rate. This order becomes the more evident the more nearly the bacteria are of uniform age and resistance. The computation of this death rate, as such, does not involve any definite conception concerning the mechanism of the reaction. The utilization of death rates is not different in principle from the utilization of the times required to kill all bac- teria of a given suspension, and makes no more assump- tions. It is more accurate, and more likely to guard the investigator against errors since there are several points to be considered and coordinated, instead only the one datum, i.e., the killing time. If the death rate is not approximately constant, it cannot be used for drawing conclusions. The fundamental reaction, i.e., the reaction in the cell which causes death, is known only in very few instances. The death of dry bacteria is known to be an oxidation process. Death by heat is probably due to the coagulation (or hydrolysis) of a certain compound in the cell. Good progress has been made in the search for the fundamental reactions in chemical poisoning. It would seem possible that most of the causes of death will ultimately be shown to result in the same chemical process, or in the reaction upon the same molecules, but if that should be the case, we are far from having proved it. Of the enormous number of experiments on chemical disinfection, an extremely small fraction only is fit to throw light on the underlying causes of death. On account of the importance of the subject, an enormous number of new experiments is to be expected, and it is hoped that the technique of these future experiments will live up to the standards set by Kronig and Paul more than thirty years ago. 394 PHYSIOLOGY OF BACTERIA The causes of death by old age can be identified, partly at least, with well-known causes. The reactions leading to death by interface tension depression are quite unknown. Death by light and by rays of shorter wave-length has been shown to be due to a photochemical (or physical) process. The reaction itself is unknown. Death by freezing is ordinarily a complication of several factors, which cannot, as a rule, be separated. Experiments on starvation also usually involves several factors, only one of them being starvation proper, while cold-shock and, perhaps, traces of poison, or leaching out of cell constituents, may have a stronger deleterious effect than starvation. Death by suffoca- tion is nothing but a special case of starvation. APPENDIX I. THE SIZE OF MICROORGANISMS One of the common sources of error in studying physiology of bacteria is a wrong conception of the amounts or quantities of living matter or living proto- plasm involved. A feeding experiment with pigs, record- ing accurately the quantities of food used and the amounts of CO2, urea, excreta, etc., given off, but with- out a record of the weight of the pigs before and after the experiment, would appear very unscientific. Yet, that is the customary way of describing fermentation experiments with bacteria. For this reason a computa- tion of the size, weight, surface, and composition of an average representative bacterium cell is given here. Bad. coll has been chosen because it is probably the best known micro-organism, and because it is of average size. The size of Bad. coU varies considerably with age, and also with the kind and amount of food. But we can consider a length of 1.5/i and a diameter of O.Sfj, as representing the average size of cells in a well grown broth culture. For the simpUcity of calculation, we shall suppose the cell to be a cylinder. Its volume is F = 0.4 X 0.4 X T X 1.5m^ = 0.75/i3 The specific gravity is larger than 1, but only shghtly so, for a very high speed of the supercentrifuge is necessary to precipitate the cells. If we assume 1.07 as density of the cell (twice that of skim milk, which contains half as many solids as bacteria), the weight of one cell would be TF = 8.0 X 10-10 mg. 395 396 PHYSIOLOGY OF BACTERIA The surface of the cells is important because all metabolism takes place through this surface only. It is aS = 2 X 0.4 X TT X 1.5 + 2 X 0.4 X 0.4 X TT = 4.77^^ These figures mean little, unless we compare them with things we can really conceive. We cannot imagine the size of ju^ or yu^. One c.c. of a full grown culture of Bad. coli contains about one billion cells. The space occupied by these is 0.75 X 10-9 X 10« mm.3 = 0.75 mm.^ One billion bacteria occupy a space smaller than 1 mm.^, or less than 0.1 % of the volume of the liquid in which they grew. The weight of these is 0.8 mg. or, in other words, 1 mg. of moist bacteria contains about 1,250,000,000 cells. The solids of Bad. coli have been determined to 26.65%, of which 17.2% is protein. The weight of one dry cell would then be 2.2 X 10-1° jjig^ aji(j I jng^ of (jry cells would contain about 4,500,000,000 cells. MacNeal found 5,500,000,000 dry cells of Bad. coli to weigh 1 mg. The bacteria solids of 1 liter of a full grown broth culture amount to 220 mg. Moyer (1929) found for Bad. aerogenes in the average of 8 experiments, 308 mg. of dry bacteria per liter if fed with glycerol, and 562 mg. if fed with glucose. It is easy to see how near this comes to a catalyst, considering the rapid rate of decomposition brought about by only 0.3 gm. of substance per liter. This rate of decomposition is explainable only by the enormous surface. The surface of the bacteria in this one liter of broth culture amounts to 4.77 X lO*^^^ = 4.7m2 = about fifty square feet, which is enormous considering the 800 mgs. of bacteria bodies. The same ratio of surface to weight with man would amount to 0.47 square kilometers or 130 acres. There is one group of organisms still smaller than bacteria, i.e. the bacteriophage. Their size has been estimated to only 20m/x, or 0.02/i. Assuming them to be spheres, their volume would be 0.01 X 0.01 X 0.01 X TT X Vs^JL^ = 4.2 X lO-^M^ = 4.2 X lO-^^ ^m.^. Since one cell of Bad. coli had the volume of 0.75^^ it is 180,000 times as large as a bacteriophage. The surface of a bacteriophage is 4 X tt X 0.01 X 0.01 = 12.5 X 10-V'. One mg. of bacteriophage contains 225,000,000,000,000 cells, and their surface is 0.28 m.^. APPENDIX Table 132. — Dimensions of Microorganisms 397 Sacch. cerevisiae Bad. coli Bacterio- phage Length Diameter Volume of 1 cell Weight of 1 cell Surface of 1 cell Moisture content 1 mg. of dry cells contains . 1 mg. of moist cells con- tains Surface of 1 mg. of moist cells Number of cells per c.c. of full grown culture . . . Weight of cells in 1 one liter of full \ , grown culture j Surface of cells per liter of full grown culture 7.2/i 5.6/i 118ju3 130 • 10-5 mg. 118^2 70% 26 • 10« cells 7 . 7 • 106 cells 30 cm.2 50 . 10^ 6.5 gm. 1.95 gm. 5.9 m.2 (60 square feet) 1.5/* 0.8m 0.75/X3 8 • 10-10 mg. 4.77/^2 73.35% 4.5- 109 1.25- 109 60 cm. 2 109 800 mg. 220 mg. 4.7 m.2 (50 square feet) 20mn 20m/x 4.2- 10- V' 1.25- 10- V 225 -1012 cells 2,800 cm. 2 1011 0 . 42 mg. 0.125 m.2 (1 square foot) Their number in a good medium has been estimated to IQii per c.c. (d'Herelle p. 98). This means 4.2 X IQ-^^ X 10^^ = 4.2 X 10"'' or 0.004 mm. 3 of bacteriophage per c.c. of a full grown culture. Even in a liter of medium, their total moist weight would be only half a milligram, their number totalling about lO^*. The yeasts are much larger than bacteria. An average-sized bread-yeast cell has the dimensions of about 7.2 X 5.6/i, according to Henneberg, the variation with age being very great. The volume of one such cell would be that of an ellipsoid ^i X TT X 3.6 X 2.8 X 2.8 = USjjl' 398 PHYSIOLOGY OF BACTERIA Volume of 50,000,000 cells in 1 c.c. of liquid = 6 mm.^ = 0.6% of volume of medium. All these dimensions are summarized in Table 132 which shows plainly that we cannot generalize from yeasts to bacteria, or from bacteria to bacteriophage. The weights, volumina and surfaces of mold mycelium are again on a different order of magnitude, but the variation is so great and their mode of growth so different that no general statements can be made. II. MULTIPLICATION OF BACTERIA Geometrical Progression: Bacteria divide by fission into two equal cells; the successive generations, 2; 4; 8; 16; etc., can be described by the geometrical progres- sion 21; 22; 2^; 2^ etc. If one cell is placed in a nutrient solution and multiplies, all cells deriving from this one cell will be of the same age, for there is no old cell left when the two young cells have been formed. Each new set of cells is called a new generation. If a cells are transferred to a new medium, the multipli- cation follows the order a;2a;2^a;2^a; . . . 2''a; This rule holds true only for organisms which divide by fission; it is not really accurate with yeasts. There, we have a budding process with the old cell plainly distinguishable from the daughter cells, even if the bud- ding has gone through several generations. Thus, the yeast cells in a fermenting liquid are not all of the same age, and old cells may not bud as freely as the younger ones. Ordinarily, the multiplication will, even in yeasts, follow nearly the progression 2^; but it is never safe to rely upon it. Generation time: The time required by one cell to divide completely in two new cells is called generation APPENDIX 399 time. It may be observed directly (p. 267, Table 74), or may be computed from microscopic or plate counts of the numbers of cells in a liquid at the beginning and the end of an experiment, or by the weight of the cells, e.g. in yeasts. The initial number of cells be a, and the number at the end of the experiment, after the time t, be b. Then, we find b = a2'*, where n is the number of generations in the time t. Since a and b are known, we find n from the equation n = log^- or transformed into decimal logarithms, _ log & - log g The experiment lasted for the time t; in this time, the cells divided n times; therefore, the time for one genera- tion must be t ^ = n Hog 2 log 6 — log o This is the standard formula for the generation time which was given in a somewhat different form as early as 1876 by Pedersen. The generation time is not constant during the growth of a culture; it is, a priori, impossible that it can remain constant, because growth cannot continue in a culture for an indefinite time. The multiplication within a limited space soon comes to point where it is retarded more and more, until it ceases altogether, and the generation time becomes infinite. 400 PHYSIOLOGY OF BACTERIA Table 133. — Development of Two Milk Cultures of Strept. lactis Culture I Age, hours Plate counts per c.c. C.c. n/10 NaOH per 10 c.c. milk % of lactic acid Genera- tion time, minutes Fermenting capacity times 10^0 For- mula Aver- age Cor- rected 0 38 38 1.65 0 79.0 3 160 205 1.65 0 55.0 6 1,720 1,780 1.65 0 52.7 9 21,700 24,800 1.65 0 48.5 12 300,000 302,000 1.65 0 45.1 (103.4) (65.3) 15 4,700,000 4,900,000 1.70 1.70 0.005 18 43,000,000 43,000,000 1.75 1.85 0.011 56.8 88.5 13.3 5.6 9.7 4.9 9.6 21 172,000,000 181,000,000 1.95 1.95 0.027 92.2 9.7 8.1 24 650,000,000 720,000,000 3.20 3.25 0.132 210.3 7.2 7.1 27 1,230,000,000 1,240,000,000 5.35 5.45 0.338 620. 4.7 4.6 30 1,470,000,000 1,560,000,000 7.00 7.15 0.489 00 33 1,280,000,000 7.45 1,500,000,000 7.55 0.527 00 ■ 0.43 0.43 36 1,670,000,000 1,830,000,000 7.50 7.60 0.531 APPENDIX 401 Table 133. — Development of Two Milk Cultures of Strept. lactis. — (Continued) Age, hours Culture II Plate counts per c.c. C.c. n/10 NaOH per 10 c.c. milk % of lactic acid Genera- tion time, minutes Fermenting capacity times 1010 For- mula Aver- age Cor- rected 0 38,000 38,000 1.65 0 3 184,000 216.000 1.65 0 74.8 6 1,600,000 1,820,000 1.65 1.75 0 58.1 9 18,700,000 20,500,000 1.70 1.75 0.007 50.8 (31.9) (21.8) 10.6 12 112,000,000 1.85 70.9 (6.9) (5.5) 114,000,000 1.85 0.018 15 430,000,000 440,000,000 2.60 2.60 0.086 92.2 9.5 8.3 18 1,000,000,000 1,270,000,000 4.55 4.60 0.263 130. 8.1 7.5 21 1,250,000,000 1,490,000,000 6.35 6.60 0.434 663. 4.6 4.5 24 1,490,000,000 1,610,000,000 6.95 7.00 0.479 1010. 1.03 1.03 27 1,300,000,000 1,390,000,000 7.15 7.30 0.502 00 0.53 30 1,170,000,000 1,420,000,000 7.35 7.60 0.524 00 00 0.31 0.55 > 0.31 33 1,380,000,000 1,500,000,000 7.30 7.55 0.520 36 1,5.50,000,000 1,660,000,000 7.55 7.70 0.538 00 ■ 0.21 402 PHYSIOLOGY OF BACTERIA An example of the changes of the generation time is given in Table 133 representing the growth of Streptococcus lactis in sterile milk at a constant temperature of about 20°C. In the first three hour period, seventy-five to eighty minutes are needed for the cells to double, but soon the minimum time of forty-five to fifty minutes is reached, and the cells divide in less than one hour. This period of most rapid growth lasts only six to twelve hours, and then growth slackens, the generation time increases progressively, and soon reaches infinite which means that the culture has ceased growing altogether. Amount of Inoculum : It is not unusual to find state- ments in literature where it is anticipated that by dou- bling the amount of inoculum, the time for completion of the fermentation can be reduced to one-half. In thinking of the geometrical progression of bacterial multiplication, it is evident that this is an error; the time saved by doubling the inoculum is just the time which is required for the inoculum to double, i.e. the time of one generation which is usually not much more then one hour. By using one thousand times the inoculum, the amount of time saved is the time of ten generations, since 2^^ = 1,024. We can test this in Table 133 where the two parallel cultures differ only inasmuch as the first was inoculated with one thousand times as many cells as the other. The average generation time for the first few hours is about sixty minutes. It should therefore require ten times sixty minutes or ten hours for the first culture to come to the point where the second culture started, and from then on, there should be no difference between the two. This is actually the case. After nine hours, culture I has 23,000 cells per c.c. which, after ten hours, would have increased to about 50,000. The time saved in the first experiment by using 1,000 times as many bacteria for seeding the culture was about nine and a half hours. APPENDIX 403 The generation time is the reciprocal of the growth-rate, and is least when the growth rate is greatest. In order to have a direct measure of the growth rate, Slator (1916) suggested the ''growth- constant." He computed this by starting from the assumption that the growth at any time must be proportional to the number of cells present. This leads to the equation db J, 0 Upon integration, this changes to In b = kt + C For ^ = 0, we have kt = 0 and b = a Ina = 0 + C subtracting this equation from the preceding, we ehminate C: \n b — In a = kt I ^n u ^ \ 1 (log ^ — log a) k = -anb-lna) = -^ 0.434 This value of k is called the growth-constant. Its meaning and its relation to the generation time is explained on p. 190. The formula for the generation time is accurate only in the exponential branch of the growth curve, i.e., only as long as the rate of growth does not decrease. If this is the case, the formula is based on the false assumption of an exponential increase. Since we are without accurate knowledge of the cause of the decreas- ing growth rate, we cannot very well put it into a correct mathematical equation. The assumption that growth proceeds in a straight line will be more accurate at this stage than the above formulas, be it the one for generation time, or that of Slator. 404 PHYSIOLOGY OF BACTERIA To show graphically the meaning of this, Fig. 42 has been used which represents the growth of Ps. fluorescens after data obtained by Muller (1903). The dotted lines in this curve illustrate the wrong assumption made in the computation of the generation time at this s ;.^ ^^^5?— 1 /f fa In do V 10 10 / / u 30 ¥0 Houn Fig. 42. — Growth curve of Pseudomonas fluorescens. For the explanation of the dotted and thin lines, see text. stage from the customary formula. It is impossible, however, to compute generation times if we assume an arithmetical increase as shown by the thin straight lines, which represent the arithmetical averages. III. THE FERMENTING CAPACITY OF THE SINGLE CELL The various means of obtaining experimentally fer- mentation without growth have been mentioned on p. 102. It is also possible, however, to eliminate the disturbing factor of the multiplication of cells during APPENDIX 405 fermentation by a mathematical calculation, provided that the increase in cells is known. After an attempt by Burchard (1899) which was based on a wrong mathematical principle, Rahn (1911) devel- oped a formula in the following way: If we assume that the single cell up to the time where it is completely divided produces the amount of products s, then the amount of products in the successive genera- tions will be as 2as 4as Sas . . . a2'^s It is arbitrary to state which shall be the last member of this progression, whether 2"* or 2""^ Rahn decided that 2""^ would be more accurate to assume. The total amount of the products formed, S is the sum of all these amounts. S = as + 2as -\- 4as + 2^as + • • • + 2"-ias Applying the sum formula of geometrical progressions, we get S = as(2" - 1) = s(a2" - a) and since 2"a = b, we get S = s{b - a) or S s = b — a In this formula, s is the amount of products formed by one cell in one generation, i.e., in the time g. Ordi- narily, the fermenting capacity per hour will be more 406 PHYSIOLOGY OF BACTERIA interesting. If we call the amount of products formed per cell per hour, x, we find s X = - 9 Substituting in this equation the formula for the genera- tion time from p. 399, we obtain the final expression for the fermenting capacity of a cell per hour to be ^ " t{h - a) log 2 We had made the arbitrary assumption here that the last member of the progression to form products was a2''~^; perhaps, we might as well have assumed that it was 2"". The more correct way to solve the problem is the application of calculus. This was done by Slator (1913) and independently, also by Buchanan and Fulmer (1918). The rate with which the products are formed is proportional to the number of bacteria. Using the same designations as before, the cells increase to the progression a2'^, and the products increase accord- ingly in the ratio sa2''. The sum of all products formed between the numbers a and a2'" is given by the integral of the rate of increase S = £'sa2''dn between the limits n = 0 and n = n' The solution of the integral gives ^ ^ sa{2'" - 2Q) ^ s(a2^ - a) In 2 In 2 ^ s{h - a) ^ 0.434s(6 - a) In 2 log 2 APPENDIX 407 Computing from this equation the fermenting capacity X per hour by substituting s = - we get ^ ^ X't' 0.434(& - g) log 2 log 2 (log b — log a) or S • (log b - log g) 0.434 . ^ • (6 - g) For the short period A^, with the increase in products AaS, we obtain ^ 2.301 ' AS ' (log g - log b) ^ At'ia -b) This formula has been applied in this book wherever the fermenting capacity per cell has been computed. It holds fairly true as long as the bacteria multiply rapidly. After the curve has passed the point of inflex- ion, however, the products increase no more according to geometrical progression. At this stage, the computa- tion of the fermenting capacity is more accurate if the increase in products is simply divided by the arithmetical average of the number of cells at the beginning and the end of the period. This corresponds to the principle demonstrated in Fig. 42. Table 133 shows some computations by the two meth- ods, and it is evident that the difference is not great. In this table, the initial values obtained for the fermenting capacity are far too high. This is due to the great probable error in the determination of the extremely small amounts of acid formed. For instance, if we assume for culture II that at nine hours, not 0.007 %, but 0.002 % of acid had been formed, the fermenting capacity would have been for the period from six to nine hours 9.1 and for the period 408 PHYSIOLOGY OF BACTERIA from nine to twelve hours 9.4 X 10"^° mg. This difference cor- responds to only 0.05 c.c. of alkah in titration, and is well within the Hmits of error. As soon as the quantities of acid become much larger than the experimental error, the fermenting capacity per cell becomes quite stable. This stabiHzation can also be accomplished by com- putation over larger periods, as is shown in Table 133 in the column marked "corrected." For the first value of the fermenting capacity in these cultures, different zero-points are possible. They gave very similar results. For culture II, the fermenting capacity was for the period 0 to twelve hours: 10.6, for three to twelve hours, 11.25, and for six to twelve hours, 11.3 10-^**. This agreement is a good proof for the accuracy of the formula. AUTHOR INDEX Page Abderhalden, E., and P. Rona, 1905, Die Zusammensetzung des Eiweisses von Aspergillus niger bei verschiedener Stickstoflf- quelle. Z. physiol. Chemie 46, 179 172, 144 Aberson, J, H., 1903, La Fermentation alcoolique. Rec. Trav. Chini. Pays-Bas et Belg. 22, 78 121, 233 Albus, W. R., see Sherman Almquist, E., 1898, Ueber eine Methode, das spezifische Gewicht von Bakterien und andern Korperchen zu bestimmen. Z. f. Hygiene 28, 321 329 Allyn, W. p., and I. L. Baldwin, 1930, The effect of the oxidation- reduction character of the medium on the growth of an aerobic form of bacteria. J. Bact. 20, 417 70 Angerer, K. von, 1919, Ueber die Arbeitsleistung eigenbewegUcher Bakterien. Archiv f. Hygiene 88, 139 34 Arnold, L., 1929, Alteration in the endogenous enteric bacterial flora. Journ. Hygiene 29, 112 300 Arrhenius, S., 1907, Immunochemistry. New York, MacMillan Co. 129 Arzberger, C. F., W. H. Peterson and E. B. Fred, 1920, Certain factors that influence acetone production by B. aceto-ethylicum. J. Biol. Chem. 44, 465 156 Avery, O, T., and H. J. Morgan, 1924, The occurrence of peroxide in cultures of pneumococcus. J. Exper. Med. 39, 275 and 355 236 Ayers, S. H., and W. T. Johnson, 1914, Ability of streptococci to survive pasteurisation. J. Agric. Research 2, 321 333 , and B. J. Davis, 1918, The thermal deathpoint and limiting hydrogen ion concentration of pathogenic streptococci. J. Infect. Diseases 23, 290 146 , and P. Rupp, 1918, Simultaneous acid and alkaline bacterial fermentations from dextrose and, the salts of organic acids respectively. J. Infect. Diseases 23, 188 53 , and W. T. Johnson, 1923, The influence of surface tension depressants on the growth of streptococci. J. Infect. Diseases 33, 202 369 Baas-Becking, L. G. M., and G. S. Parks, 1927, Energy Relations in the metabolism of autotrophic bacteria. Physiol. Review 7, 85 181 409 410 PHYSIOLOGY OF BACTERIA Page Baker, J. C, J. D. Brew and H. J. Conn, 1919, Relation between lactic acid production and bacterial growth in the souring of milk. New York (Geneva) Agr. Exp. Station Tech. Bull 74 106, 112 Ballner, 1902, Experimentelle Studien iiber die Desinfektionskraft gesattigter Wasserdampfe bei verschiedenen Siedetempera- turen. Sitz. Ber. Akad. d. Wissensch. Wien, math.-nat. Klasse 111, 97 320 Balls, A. K., and J. B. Brown, 1925, Studies in yeast metabolism, I and II. J. Biol. Chem. 62, 789 and 823 174, 178, 210 Baldwin, I. L., see Allyn Bancroft, W. D., and G. H. Richter, 1931, The chemistry of disinfection. J. Physic. Chem. 35, 511 367 Barber, M. A., 1908, The rate of multiplication of Bacterium coli at different temperatures. J. Infect. Diseases 5, 379. . . . 196, 219, 222 Baron, M. A., 1928, Bakterien als Quellen mitogenetischer (ultra- violetter) Strahlung. Centr. f. Bakt. II Abt. 73, 373 34 Bartlett, C. J., and F. B. Kinne, 1913, Resistance of micro- organisms in glycerine or oil to the sterilizing action of heat. Science n.s. 38, 372 312 Bayliss, W. M., 1924, Principles of General Physiology. 4th ed. London; Longmans, Green and Co 2, 15, 16, 18 Bayne-Jones, S., and J. S. van derLingen, 1923 a. The bactericidal action of ultra-violet hght. Johns Hopkins Hospital Bull. 34, 11 372 and H. S. Rhees, 1929, Bacterial Calorimetry, II. J. Bact. 17, 123 31 Berghaus, W. H., 1907, Ueber die Wirkung der Kohlensaure, des Sauerstoffs und des Wasserstoffs auf Bakterien bei verschiedenen Druckhohen. Archiv f. Hygiene 62, 172 19 Berman, N., andL. F. Rettger, 1918 a, Further studies on the utiH- zation of protein and non-protein nitrogen. J. Bact. 3, 367 56, 203 , , 1918 b. The influence of carbohydrate on the nitrogen metabolism of bacteria. J. Bact. 3, 389 159 ,see also Rettger Bi kert, F. W., 1930, Zur Dififerentialfarbung toter und lebender Bakterien. Centr. f. Bakt. I Abt. Grig. 117, 548 274 Bigelow, W. D., 1922, The logarithmic nature of thermal deathpoint curves. J. Infect. Diseases 29, 528 320, 323 Bioletti, F. T., 1908, Improved methods of wine making. Cali- fornia Agr. Exp. Station Bull. 197 156 Birstein, see Paul BoYSEN, H. H., 1928, unpublished data 209 Branham, S. E., 1929, The effects of certain chemical compounds upon the course of gas production by baker's yeast. J. Bact. 18, 247 140 AUTHOR INDEX 411 Page Brannon, J. M., see Prucha Brew, J. D., see Baker Brooke, O. R., see Winslow Brown, A. J., 1905, The influences regulating the reproductive functions of Saccharomyces cerevisiae. J. Chem. Soc. London, 87, 1395 209, 233, 246 Brown, J. B., see Balls Brown, J, H., 1026, Vacuum tubes for the storage and shipment of bacteria. Science 64, 429 307 Browning, C. H., and S. Russ, 1917, The germicidal action of ultra-violet radiation, and its correlation with selective adsorp- tion. Proc. Royal Soc. London, Ser. B. 90, 33 372 Buchanan, R. E., and E. I. Fulmer, 1918, Determination of the fermentation capacity of a single bacterial cell. Abstracts of Bact. 2, 11 31,406 and E. I. Fulmer, 1928-30, Physiology and Biochemistry of Bacteria. Baltimore; WiUiams and Wilkins . . 55, 167, 182, 230, 376 -, see also Fulmer Buchner, E., 1897, Alkoholische Garung ohne Hefezellen. Ber. d. Deutsch. Chem. Ges. 30, 117 and 1110 35, 37 , H. Buchner und M. Hahn, 1903, Die Zymasegarung. Miinchen; R. Oldenbourg 17, 130, 138 und R. Gaunt, 1906, Ueber die Essiggarung. Liebigs Annalen 349, 140 38 und J. Meisenheimer, 1903, Enzyme bei Spaltpilzgarungen. Ber. d. Deutsch. Chem. Ges. 36, 634 37 Burchard, a., 1899, Beitrage zur Kenntnis des Ablaufs und der Grosse der durch Micrococcus ureae bewirkten Harnstoffzer- setzung. Archiv f . Hygiene 36, 264 107, 405 Callow, A. B., 1924, The oxygen uptake of bacteria. Biochem. J. 18, 507 19, 74 Cathcart, E. p., 1921, The Physiology of Protein Metabolism. 2nd ed. London; Longmans, Green and Co 16 Carpenter, D. C, and G. J. Hucker, 1927, The relation of hydro- lytic decomposition products of proteins to bacterial growth. J. Infect. Diseases 40, 485 250 , 1931, unpublished data 259 Chesney, a. M., 1916, The latent period in the growth of bacteria. J. Exper. Med. 24, 387 ^ 197, 376 Chick, H., 1908, An investigation of the laws of disinfection. J. Hygiene 8, 92 278, 288, 293, 295, 339, 346, 353, 355, 363 , 1910, The process of disinfection by chemical agencies and hot water. J. Hygiene 10, 237 279, 290, 295, 299, 319 412 PHYSIOLOGY OF BACTERIA Page and C. J. Martin, 1908, The principles involved in the standardization of disinfectants, and the influence of organic matter upon the germicidal value. J. Hygiene 8, 654 336 , , 1910, On the heat coagulation of proteins. J. Physiol. 40, 404 326 Christensen, L. M., see Fulmer Claassen, H., 1927, Stoffwechselvorgange und Stoffbilanz bei der Vergarung der Melasse nach dem Lufthefeverfahren. Z.d. Vereins Deutsche Zuckerindustrie 77, 607 174, 176 , 1928, Versuche iiber den Ersatz der Malzkeime in der Lufthefefabrikation durch Ammoniakverbindungen. Z. angewandte Chemie 41, 1161 243 , 1929, unpubHshed data 237-243 Clark, H. W. and S. de M. Gage, 1903, On the value of tests for bacteria of specific types as an index of pollution. 34th Ann, Report, State Board of Health, Mass. 245 295 Clark, W. M., B. Cohen and associates, 1928, Studies on oxidation- reduction. U. S. Public Health Service, Hygienic Lab. Bull. 151 86, 88, 90, 92, 94 , see also Rogers Coblentz, W. W., and H. R. Fulton, 1924, A radiometric investiga- tion of the germicidal action of ultra-violet radiation. Scien- tific Papers, Bureau of Standards 19, 641 372 Cohen, B., 1922, Disinfection studies. J. Bact. 7, 183 . . 339, 350, 352, 377 , 1931, The bacterial culture as an electric half-cell. J. Bact. 21, 18 96 , see also Clark Coffeen, Mrs. V. A., 1911, unpublished data 330, 331 Conn, H. F., see Baker CoNRADi, H., and O. Kurpjuweit, 1905, Ueber spontane Wachstums- hemmung der Bakterien infolge Selbstvergiftung. Munch, med. Wochenschr. 34, 1761 234 Cooper, E. A., and R. B. Haines, 1928, The influence of tempera- ture on bactericidal action, J, Hygiene 28, 163 365 and J, Mason, 1927, Studies of selective bacterial action. J. Hygiene 26, 118 345 , , 1928, Further investigations on selective bacterial action, J, Path. Bact. 31, 343 345 Copping, A. M., 1929, The effect of ''bios" on the growth and metabolism of certain yeasts. Biochem. J. 23, 1050 203 Crozier, W, J., 1924, On biological oxidations as a function of tem- perature. J. Gen. Physiol. 7, 189 135 Coulter, C. B., 1929, Reduction potentials of sterile culture bouillon. J. Gen. Physiol. 12, 139 97 AUTHOR INDEX 413 Page CuRRAN, H. R., 1925, The growth of bacteria. Thesis, Cornell Univ 206, 265 CuRRiE, J. N., 1917, The citric acid fermentation of Aspergillus niger. J. Biol. Chem. 31, 15 158 VAN Dam, W., 1922, Opstellen over moderne zuivelchemie, 2nd ed. Gravenhagen 144 Davis, B. F., see Avers, and Rogers Davis, M. A., see Hilliard Davis, N., 1927, Interfacial tension and bacterial growth. J. Bact. 13,381 367 Demeter, K. J., 1930, Ein Beitrag zur Kenntnis der fiir die Gras- konservierung wichtigen Mikroorganismen. Centr. f. Bakt. II Abt. 82, 71 73 DEN DoREN DE JoNG, L. E., 1926, Bcijdrage tot de kennis van het mineral process. Dissertation, Delft, quoted after Buchanan and Fulmer, 1928-30, vol. II 167 , 1927, Ueber protaminophage Bakterien. Centr. f. Bakt. II, Abt. 71, 193 170 Dengler, a., see Siipfle Dixon, M., and J. H. Quastel, 1923, A new type of reduction- oxidation system. Part I. Cysteine and glutathione. J. Chem. Soc. 123, 2943 88 Donker, H. F. L., see Kluyver Drake, E. T., see Sturges DuBOS, R., 1929, Observation on the oxidation-reduction properties of sterile bacteriological media. J. Exper. Med. 49, 507. . . 98, 100 DucLAux, E., 1889, Sur la nutrition intracellulaire. Ann. Inst. Pasteur 3, 109 159 , 1898-1901, Traite de Microbiologic: Paris. Masson etCo 42,103 Eaton, S. M., see Zoller Eberlein, L., see Rothenbach Ehrlich, F., 1905, Ueber die Entstehung des Fuselols. Z. Verein Rubenzucker-Industrie, 1905, 539 42, 43 , 1909, Ueber die Entstehung der Bernsteinsaure bei der alkoholischen Garung. Biochem. Z. 18, 391 43 , 1911, Ueber die Vergarung des Ty rosins zu Tyrosol. Ber. d. Deutsch. Chem. Ges. 44, 139 and 45, 883 44 EiJKMAN, C, 1905, Ueber thermolabile Stoffwechselprodukte als die Ursache der natiirUchen Wachstumshemmung der Bakterien. Centr. f. Bakt. I Abt. Orig. 37, 436 and 41, 367 234 , 1912, De reactiesnelheid van microorganismen. Vers. Kon. Akad. Wetensch. Amsterdam, Wis-en Natuurkunde, 21, I, 507 282, 295, 319 414 PHYSIOLOGY OF BACTERIA Page Emmerling, 0., and O. Reiser, 1902, Zur Kenntnis eiweiss- spaltender Bakterien. Ber. d. Deutsch. Chem. Ges. 35, 700.. 59 EsTY, J. R., 1928, Determinations of thermal death time. In Jordan and Falk, The Newer Knowledge of Bakteriology 334 EuLER, H. VON, und P. Lindner, 1915, Chemie der Hefe und der alkoholischen Ganing. Leipzig, Akad. Verlagsgesellschaft . , 10, 12, 16 Falk, I. S., and C-E. A. Winslow, 1926, A contribution to the dynamics of toxicity and the theory of disinfection. J. Bact. 11, 1 295, 340 Farrand, B. S., see Marshall Fawcett, H. S., 1921, The temperature relations of growth in certain parasitic fungi. Univ. of CaUfornia Publ., Agricultural Series 4, 183 214, 217 Feirer, W. a., see Leonard Finkle, p., see Meyerhof Fischer, Hugo, 1902, Ueber Garungen. Centr. f. Bakt. II Abt. 9, 356 24, 180 FoLiN, O., 1905, A theory of protein metabolism. Amer. J. Physiol. 13, 117 9 Fred, E. B., 1911, Ueber die Beschleunigung der Lebenstatigkeit hoherer und niederer Pflanzen durch kleine Giftmengen. Centr. f. Bakt. II Abt. 31, 185 255 , see also Arzberger Frobischer, Martin, 1927, Studies on the relationship between surface tension and the action of disinfectants, with special reference to hexyl resorcinol. J. Bact. 13, 163 370 FuLMER, E. I., and R. E. Buchanan, 1923, Studies on toxicity. J. Gen. Physiol, 6, 77 274 and L. M. Christensen, 1925, The fixation of atmospheric nitrogen by yeast as a function of the hydrogen ion concentra- tion. J. Phys. Chem. 29, 1415 . 168 and C. H. Werkman, 1929, An index to the chemical action of microorganisms on the non-nitrogenous organic compounds. Springfield, C. C. Thomas 55 , see also Buchanan Fulton, H. R., see Coblentz Gage, S., and G. Stoughton, 1906, A study of the laws governing the resistance of B. coli to heat. Technology Quarterly 19, 41 300, 320, 333 , see also H. W. Clark Gaidukov, N., 1910, Dunkelfeldbeleuchtung und Ultramikroskopie in der Biologie und Medizin. Jena; Gustav Fischer 367 Gates, F. L., 1929, A study of the bactericidal action of ultra-violet light. J. Gen. Physiol. 13, 231 295, 300, 372 AUTHOR INDEX 415 Page Gaunt, R., see Buchner Gegenbauer, v., 1921, Studien iiber die Desinfektionswirkung des SubUmats. Archiv f. Hygiene 90, 23 338, S53, 357, 362 , 1922, Studien iiber die Desinfektionswirkung wassriger Formaldehydlosungen. Archiv f . Hygiene 90, 238 353, 362 und H. Reichel, 1913, Die Desinfektion milzbrandiger Haute und Felle in Salzsaure-Kochsalzgemischen. Archiv f. Hygiene 78, 1 353, 364, 367 Gerlach, M., und J. Vogel, 1902, Weitere Versuche mit stickstofif- bindenden Bakterien. Centr. f. Bakt. II Abt. 9, 817 185 Gilbert, C. A., see Lee Gillespie, L. J., 1920, Redueiion potentials of bacterial cultures and of waterlogged soils. Soil Science 9, 199 82, 89, 95 , see also Keyes Grimmer, W,, und B. Wiemann, 1921, Zur Biochemie des B. mesentericus vulgatus. Forschungen a.d. Gebiet der Milch- wirtsch. 1, 2 57 Graham-Smith, G. S., 1920, The behavior of bacteria in fluid cul- tures as indicated by daily estimates of the number of living organisms. J. Hygiene 19, 133 201, 224, 230, 248, 253 Groves, J. F., 1917, Temperature and life duration of seeds. Bot. Gazette 63, 169 275 GuRWiTscH, A., 1929, Ueber den derzeitigen Stand des Problems der mitogenetischen Strahlung. Protoplasma 6, 449 23 GusTAFSON, F. G., 1920, The effect of hydrogen ion concentration on the respiration of Penicillium chrysogenum. J. Gen. Physiol. 2, 617 138 Hahn, M., see Buchner Haines, R. B., see Cooper Hajos, K., 1922, Beitrage zur Frage der wachstumhemmenden Wirkung von Bouillonkulturen. Centr. f. Bakt. I Abt. Orig. 88,583 236 Hammer, B. W., 1920, Volatile acid production of Strept. lacticus and the organisms associated with it in starters. Research Bull. 63, Iowa Agr. Exp. Station 52 , see also Sarkaria Hansen, Th., 1922, Tension superficielle et pouvoir bactericide de divers desinfectants. Compt. Rend. Soc. Biol. 74, 215 370 Harden, A., 1923, Alcoholic Fermentation. 3rd ed. London; Longmans, Green and Co 39, 44 Hastings, E. G., see Thornton Henneberg, W., 1926, Handbuch der Garungsbakteriologie. 2nd ed. Berlin; Paul Parey 105, 196, 246, 253, 397 416 PHYSIOLOGY OF BACTERIA Page Henrici, a. T., 1928, Morphologic variation and the rate of growth of bacteria. Springfield; C. C. Thomas 162, 198, 201, 234 d'Herelle, F., 1926, The Bacteriophage and Its Behavior. Balti- more; Wilhams and Wilkins 397 Herzog, R. O., 1903, Ueber Milchsauregarung. Z.f. physiol. Chem. 37, 381 37 Hewitt, L. F., 1930, Oxidation-reduction potentials of haemolytic streptococci. Biochem. J. 24, 513 and 669 and 676 and 1551 86 Hewlett, R. T., 1909, Disinfection and disinfectants. Lancet, 1909, I, 741 278 Hildesheimer, a., see Neuberg HiLLiARD, C. M., and M. A. Davis, 1918, The germicidal action of freezing temperatures upon bacteria. J. Bact. 3, 423 314 , C. ToROSSiAN and R. P. Stone, 1915, Notes on factors involved in the germicidal effects of freezing and of low tem- peratures. Science 42, 770 314 Hinds, M. E., 1916, Factors which influence longevity of B. coli and B. typhosus in waters. Univ. of Illinois Bull.; Water Survey Series 13, 225 377, 380, 388 HoFMANN, Paul, 1922, Ueber die Giiltigkeit des Arndt-Schulz'schen Grundgesetzes bei der Wirkung von Bakteriengiften. Disser- tation, tierarztl. Fakultat, Miinchen 255, 272 HoLWECK, F., 1929, Production de rayons X monochromatiques de grande longueur d'onde. Action quantitative sur les microbes. Compt. rend. Acad. Paris. 188, 197 372 HoLWERDA, B. J., 1921, Ueber den Einfluss der Milchsaure auf die Milchsauregarung. Biochem. Z. 128, 465 144 Hopkins, F. G., 1921, On the auto-oxidizable constituents of the cell. Biochem. J. 15, 286 97 Hsu, P. H., see Norton HucKER, G. J., 1928 a, Production of carbon dioxide by the strepto- cocci. New York (Geneva) Agr. Exp. Station, Techn. Bull. 142 146, 174 , 1928 b. Certain biochemical reactions produced by strepto- cocci. New York (Geneva) Agr. Exp. Station Techn. Bull. 143 52 -, see also Carpenter HiJNE, 1909, Die begiinstigende Reizwirkung kleinster Mengen von Bakteriengiften auf die Bakterienvermehrung. Centr. f. Bakt. I Abt. Orig. 48, 135 228 Hunter, O. W., 1923, Stimulating the growth of Azotobacter by aeration. J. Agr. Research 23, 665 185, 209 Hutchinson, H. B., 1907, Ueber Form und Bau der Kolonien niederer Pilze. Centr. f. Bakt. II Abt, 17, 602 22 AUTHOR INDEX 417 Page Iked A, K., 1897, Allgemeine Beziehungen zwischen Konzentration und Giftwirkung. Z.f. Hygiene 25, 95 346 Jacoby, M., 1900, Ueber die Beziehungen der Leber-und Blut- veranderungen bei Phosphorvergiftung zur Autolyse. Z. physiol. Chem. 30, 174 10 Jensen, K. A., 1928, Durch direkte mikroskopische Beobachtung ausgefuhrte Untersuchungen iiber das Wachstum des Coli- Bacillus. Centr. f. Bakt. I. Abt. Orig. 107, 1 198 Jensen, Orla, see Orla-Jensen JoFFE, J. S., see Waksman Johnson, W. T., see Ayers Kaserer, H., 1906, Die Oxydation des Wasserstoffs durch Mikroor- ganismen. Centr. f . Bakt. II Abt. 16, 772 62, 165, 173, 345 Kaserer, H., 1907, Ueber einige neue Stickstoffbakterien mit auto- tropher Lebensweise. Z.f. Landw. Versuchswesen Oesterreich 10, 37 170 Kelly, C. D., and O. Rahn, 1932, The growth rate of individual bacterial cells. J. Bact. 23, 147 268, 283 Keyes, F, G., and L. J. Gillespie, 1912, The adsorption of oxygen by growing cultures of B. coli and of Bact. Welchii. J. Biol. Chem. 13, 305 63 Kharasch, M. S., 1929, Heats of combustion of organic compounds. Bureau of Standards Journal of Research 2, 360 26 Kinne, F. B., see Bartlett KiscH, B., 1919, Die Verwertbarkeit verschiedener Verbindungen als Stickstoffnahrung fiir einige pathogene Bakterien. Centr. f. Bakt. I Abt. Orig. 82, 28 169 KopELOFF, N., 1926, Lactobacillus acidophilus. Baltimore, WilHams and Wilkins 224 Koch, Alfred, 1912, Ueber die Wirkung von Aether und Schwe- felkohlenstoff auf hohere und niedere Pflanzen. Centr. f. Bakt. II Abt. 31, 175 139, 255 und S. Seydel, 1912, Versuche iiber den Verlauf der Stick- stofifbindung durch Azotobacter. Centr. f. Bakt. II Abt. 31, 570 184 Kluyver, a. J., und H. J. L. Donker, 1926, Die Einheit in der Biochemie. Chemie der Zelle und Gewebe, 13, 134. .44, 46, 53, 65 and A. P. Struyk, 1926, The functions of phosphates in the dissimilation of hexoses. Kon. Akad. van Wetensch. Amster- dam 29, 322 45 , , 1927, The first phases of the dissimilation of hexoses. Kon. Akad. van Wetensch., Amsterdam, 30, 871 ... . 39 418 PHYSIOLOGY OF BACTERIA Page Kronig, B., und Th. Paul, 1897, Die chemischen Grundlagen der Lehre von der Giftwirkung und Desinfektion. Z.f. Hygiene 25, 1 336, 342, 344, 393 Kruse, Walter, 1910, Allgemeine Mikrobiologie. Leipzig; Vogel. 196, 202 KuRPJUWEiT, 0., see Conradi Lacassagne, a., 1929, Action des rayons X de grande longueur d'onde sur les microbes. Compt. rend. Acad., Paris, 188, 200 260, 372 Lafar, F., 1904-1914, Handbuch der technischon Mykologie. Jena, Gustav Fischer 246, 274 Lange, B,, 1922, Keimmenge und Desinfektionserfolg. Z.f. Hygiene 96, 92 318, 338 Ledingham, J. C. G., and W, J. Penfold, 1914, Mathematical analy- sis of the lag-phase in bacterial growth. J. Hygiene 14, 242 . . 196 Lee, R. E., and C. A. Gilbert, 1918, Application of the mass law to the process of disinfection. J. Physical Chem. 22, 348 295, 353, 371 Leonard, V., and W. A. Feirer, 1927, The bactericidal activity of hexyl resorcinol in glycerine. Bull. Johns Hopkins Hospital, 41,21 370 Lindner, P., 1905, Die Assimilierbarkeit der Selbstverdauungs- produkte der Bierhefe durch verschiedene Heferassen und Pilze Wochenschrift f. Brauerei 22, 528 and 23, 519; see also Chemikerztg 34, 1144 14, 169, 170 , 1919, Zur Verfliichtigung des Biosbegriffs. Z.f. technische Biologic 7, 79 203 , und T. Unger, 1919, Die Fettbildung in Hefen auf festen Nahrboden. Z.f. technische Biologie 7, 68 179 , see also Euler Lingen, F. S. van der, see Bayne Jones LiNHART, G., 1920, The free energy of biological processes. J. Gen. Physiol. 2, 247 183, 185 LiPMAN, C. B., 1931, Living microorganisms in ancient rocks. J. Bact. 22, 183 311 LocHRiDGE, E. E., see Winslow Loeb, J., and J. H. Northrop, 1917, On the influence of food and temperature on the duration of life. J. Biol. Chem. 32, 103 . . 275, 293 Madsen, T., and M. Nyman, 1907, Zur Theorie der Desinfektion. Z.f. Hygiene 57, 388 278, 295, 363 Magoon, C. a., 1926, Studies upon bacterial spores: I. Thermal resistance as affected by age and environment. J. Bact. 11, 253 299 AUTHOR INDEX 419 Page Manteuffel, 1907, Das Problem der Entwicklungshemmung in Bakterienkulturen. Z.f. Hygiene 57, 337 234 Marshall, C. E., 1912, Microbiology. Philadelphia; Blakiston 151, 152 and B. S. Farrand, 1908, Bacterial association in the sour- ing of milk. Michigan Agr. Exp. Station Spec. Bull. 42 105 Martin, C. F., see Chick Mayer, A., see Whipple Meader, F. M., 1926, Sunhght as a disinfectant. J. Bact. 11, 82 . 371 Meisenheimer, J., see Buchner Mendel, J., 1911, Ueber Umsetzung verschiedener Zuckerarten durch Bakterien. Centr. f. Bakt. II Abt. 29, 290 157 Merrill, M. H., 1930, Carbohydrate metaboUsm of organisms of the genus Mycobacterium. J. Bact. 20, 235 55 Meyer, Arthur, 1906, Notiz liber eine die supramaximalen T6- tungszeiten betrefifende Gesetzmassigkeit. Ber. d. Deutsch. Bot. Ges. 24, 340 320 Meyerhof, O., 1914. Zur Energetik der Zellvorgange. Abhandl. der Fries'schen Schule 4, 429 32 , 1916, Untersuchungen iiber den Atmungsvorgang nitri- fizierender Bakterien. I. Die Atmung des Nitratbildners. Pfluger's Archiv 164, 353 13, 17, 188, 392 , 1917, Die Atmung des Nitritbildners und ihre Beeinflussung durch chemische Substanzen. Pfluger's Archiv 166, 240. . 117, 191 und P. Finkle, 1925, Ueber die Beziehungen des Sauerstoffs zur bakteriellen Milchsauregarung. Chemie der Zelle und Gewebe 12, 157 74 und H. Weber, 1923, Beitrage zu den Oxydationsvorgangen am Kohlemodell. Biochem. Z. 135, 558 97 MiQUEL, p., 1904, Die Vergarung des Harnstoffs, der Harnsaure und der Hippursaure. in Lafar, Handbuch der Technischen Myko- logie, Bd. Ill, 82 26, 36 Morgan, H. F., see Avery Moyer, H. v., 1929, A continuous method of culturing bacteria for chemical study. J. Bact. 18, 59 396 MtJLLER, A., 1912, Die Abhangigkeit des Verlaufs der Sauerstoffzeh- rung in nattirhchen Wassern und kiinstlichen Nahrlosungen vom Bakterien wachstum. Arb. a.d. Kais. Gesundheitsamt 38, 294 31, 80, 186, 389 MiJLLER, Alfred, 1920, Die Resistenz der Milzbrandsporen gegen Chlor, Pickelfliissigkeit, Formaldehyd und Sublimat. Archiv f. Hygiene 89, 363 338, 353, 361 , see also Stipfle 420 PHYSIOLOGY OF BACTERIA Page MtJLLER, Max, 1895, Ueber den Einfluss von Fiebertemperaturen auf die Wachstumsgeschwindigkeit und die Virulenz des Typhus-Bacillus. Z.f. Hygiene 20, 245 193, 196 , 1903, Ueber das Wachstum und die Lebenstatigkeit von Bakterien bei niederer Temperatur. Archiv f . Hygiene 47, 127 192, 219, 404 Muller-Thurgau, 1885, Verhandlungen der X. Generalvers. d. Deutschen Weinbauvereins Geisenheim, quoted from Lafar, Handbuch der Technischen Mykologie Bd. V, 435 151 MuscuLus, 1876, Sur le ferment de I'ur^e. Compt. rend. Acad. Paris, 82, 333 36 Myers, R. P., 1929, The germicidal properties of alkaline washing powders. J. Agric. Research 38, 521 295, 297, 299, 345, 350 Naegeli und Schwendener, 1877, Das Microscop, 2nd ed. p. 640; Quoted from Max Muller, 1903 191 Nelson, J. M., and W. C. Vosburgh, 1917, Kinetics of invertase action. J. Amer. Chem. Soc, 39, 790 116 Neuberg, C, und A. Hildesheimer, 1911, Ueber zuckerfreie Hefegarungen. Biochem. Z. 31, 170 44 NiEL, C. B. VAN, 1928, The Propionic Acid Bacteria. Haarlem, J. W. Boissevain 52, 62 NiLSSON, R., 1930, Studien iiber den enzymatischen Kohlehydrat- abbau. Arkiv for Kemi, 10 A No. 7 39 NoRRis, D., see Penfold North, C. E., and al., 1925, Publ. Health Bull. 147, U. S. Publ. Health Service 324 Northrop, J. H., 1923, The kinetics of trypsin digestion; II. J. Gen. Physiol. 6, 423 116 , see also Loeb Norton, J. F., and P. H. Hsu, 1916, Physical chemistry of disinfec- tion. J. Infect. Diseases 18, 180 350 Nyhan, M., see Madsen Orla- Jensen, S., 1919, Lactic Acid Bacteria. Copenhagen; Fred Host and Son 142, 152 Parsons, L. B., and W. S. Sturges, 1927, Quantitative aspects of the metabolism of anaerobes. J. Bact. 14, 181 60 , see also Sturges Parks, G. S., see Baas-Becking Pasteur, L., 1860, Note relative au Penicillium glaucum et k la dissymetrie moleculaire des produits organiques naturels. Compt. rend. Acad. Paris 51, 298 159 , 1876, Etude sur la bi6re. Paris 5, 24 AUTHOR INDEX 421 Page Paul, Th=, 1909, Der chemische Reaktionsverlauf beim Absterben trockener Bakterien bei niederen Temperaturen. Biochem. Z. 18, 1 295, 304 , BiRSTEiN und Reuss, 1910 a, Beitrag zur Kinetik des Abster- bens der Bakterien in Sauerstoff verschiedener Konzentration, und bei verschiedenen Temperaturen. Biochem. Z. 25, 367 295, 304, 345, 353 , , , 1910 b, Beitrage zur Kinetik der Gift- wirkung von gelosten Stoffen: I. Einfluss der Konzentration. Biochem. Z. 29, 202 295, 346, 348, 353 , , , 1910 c, Einfluss der Neutralsalze und der Temperatur auf die Desinfektionsgeschwindigkeit von Sauren. Biochem. Z. 29, 249 295, 365 -, see also Kronig Pedersen, Rasmus, 1876, Beretning der i Carlsberg Laboratoriets fysiologiske Af deling udforte Arbeijder. Meddeleser fra Carls- berg Labor. Copenhagen 1, 55 399 Peltier, G., 1914, unpublished data 118, 207, 250 Penfold, W. J., 1914, On the nature of bacterial lag. J. Hygiene 14,215 201 and D. Norris, 1912, Influence of concentration of food supply on generation time of bacteria. J. Hygiene 12, 527 . . 205 -, see also Ledingham Peters, R. A., 1920, Variations in resistance of protozoon organisms to toxic agents. Jour. Physiology 54, 260 288 Peterson, W. H., see Arzberger Pfeffer, W., 1895, Ueber Elektion organischer Nahrstoffe. Jahr- buch f. wiss. Bot. 28, 206 159 , 1904, Pflanzenphysiologie, 2nd ed., vol. II, 885 Leipzig; Engelmann 33 Phelps, E. B., 1911, The appHcation of certain laws of physical chemistry in the standardisation of disinfectants. J. Infect. Diseases 8, 27 346, 353 PizARRO, O. R., 1927, The relation of surface tension to bacterial development. J. Bact. 13, 387 370 Potter, M. C, 1908, Bacteria as agents in the oxidation of amor- phous carbon. Proc. Royal Soc. London ser. B, 80, 239 164 , 1911, Electrical effects accompanying the decomposition of organic compounds. Proc. Royal Soc. London, ser. B, 84, 260 82 Prucha, M. J., and J. M. Brannon, 1926, ViabiUty of Bact. typho- sum in ice cream. J. Bact. 11, 27 316 PtJTTER, August, 1911, Vergleichende Physiologie. Jena; Gustav Fischer 3, 71, 183 422 PHYSIOLOGY OF BACTERIA Page QuASTEL, J. H. and M. D. Wetham, 1924, The equilibria existing between succinic, fumaric and malic acids in the presence of resting bacteria. Biochem J. 18, 519 135 , see also Dixon Rahn, Otto, 1906, Ueber den Einfluss der Stoffwechselprodukte auf das Wachstum der Bakterien. Centr. f. Bakt. II Abt. 16, 418 201, 230, 234 , 1911, The fermenting capacity of the average single cell of Bacterium ladis acidi. Michigan Agr. College, Tech. Bull. 10; also Centr. f. Bakt. II Abt. 32, 375 105, 112, 144, 149, 210, 232, 376, 405 , 1912, The bacterial activity in soil as a function of grain size and moisture content. Michigan Agr. College Tech. Bull. 16; also Centr. f. Bakt. II Abt. 35, 429 208 , 1915, Der Einfluss der Temperatur und der Gifte auf Enzymwirkung, Garung und Wachstum. Biochem. Z. 72, 351 16, 128, 137, 367 , 1925, Die Bedeutung des Temperaturkoeffizienten fiir die Milchpasteurisierung. Milchw. Forschungen 2, 373 325 , 1928, Incomplete sterilisation of food products due to heavy syrups. Canning Age 1928, 705 331 , 1929 a, The fermentometer. J. Bact. 18, 199 103 , 1929 b, The decreasing rate of fermentation. J. Bact. 18, 207 107, 109, 111 , 1929 c, The size of bacteria as the cause of the logarithmic order of death. J. Gen. Physiol. 13, 179 284, 286 , 1930, The non-logarithmic order of death of some bacteria. J. Gen. Physiol. 13, 395 284, 290, 293, 298 , 1931 a, The order of death of organisms larger than bacteria. J. Gen. Physiol. 14, 315 284, 286 , 1931 b, Betrachtungen fiber die Natur der Bakteriophagen. Centr. f. Bakt. II Abt, 83, 277 264 , 1932, A chemical explanation of the variabiUty of the growth rate. J. Gen. Physiol. 15, 257 266, 284 , see also Kelly Raistrick, H., 1917, The conversion of histidine into urocanic acid by bacteria of the coli-typhosus group. Biochem. J. 11, 71 58 Reichel, H., 1908, Die Trinkwasser-Desinfektion durch Wasser- stoff-Superoxyd. Z. f. Hygiene 61, 49 353 , 1909, Zur Theorie der Desinfektion. Biochem Z. 22, 149 353, 355, 357, 363 , see also Gegenbauer AUTHOR INDEX 423 Page Reichenbach, H., 1911, Die Absterbeordnung der Bakterien und ihre Bedeutung fiir die Theorie und Praxis der Desinfektion. Z. f. Hygiene 69, 171 283, 288, 290, 295, 297 Reiser, O., see Emmerling Rettger, L. F., N. Berman and W. S. Sturges, 1916, The utiliza- tion of proteid and non-proteid nitrogen. J. Bact. 1, 15. . . 56, 203 , see also Valley and Berman Reuss, see Paul Rhees, H. S., see Bayne-Jones Richards, O. W., 1928, The growth of the yeast Saccharomyces cerevisiae. Ann, Bot. 42, 271 162 Richter, G. H., see Bancroft Rideal, S., and J. T. A. Walker, 1903, The standardization of disinfectants. J. Royal Sanit. Inst. 24, 421 354, 369 RiTz, H., see Schultz Rivers, T. M., see Wyckoff RoDEWALD, K., 1923, Ueber die Widerstandsfahigkeit von Gefiiigel- cholera und Streptokokken gegeniiber SubUmat, Carbolsaure und Trypaflavin. Z. f. Hygiene 99, 117 337 Rogers, L. A., 1914, The preparation of dried cultures. J. Infect. Diseases 14, 100 306, 309 Rogers, L. A., 1918, The occurrence of different types of the colon- aerogenes group in water. J. Bact. 3, 313 378 , W. M. Clark and B. J. Davis, 1914, The colon group of bacteria. J. Infect. Diseases 14, 411 54 and E. O. Whittier, 1928, Limiting factors in the lactic fermentation. J. Bact. 16, 211 144, 232, 236 RoNA, see Abderhalden Rothenbach, F., und L. Eberlein, 1905, Zu der Enzymgarung der Essigpilze. Deutsche Essigindustrie 9, 233 38 Rubner, Max, 1904 a, Energieverbrauch im Leben der Mikroor- ganismen. Archiv f. Hygiene 48, 260 12, 103, 180 , 1904 b. Die Umsetzungswarme bei der alkohohschen Garung. Archiv f. Hygiene 49, 354 30 , 1906 a, Die Beziehungen zwischen Bakterienwachstum und Konzentration der Nahrung. Archiv f. Hygiene 57, 161. . 162, 247 , 1906 b, Energieumsatz im Leben einiger Spaltpilze. Archiv f. Hygiene 57, 193 , 29 , 1906 c, Ueber spontane Warmebildung in Kuhmilch, und die Milchsauregarung. Archiv f. Hygiene 57, 244 29 , 1908, Theorie der Ernahrung nach VoUendung des Wach- stums. Archiv f. Hygiene 66, 1 11, 14 424 PHYSIOLOGY OF BACTERIA Page , 1913, Die Ernahrungsphysiologie der Hefezelle bei der alkoholischen Garung. Archiv f. Physiol. Supplement 1913, 1 10, 12, 15, 40, 105, 109, 133, 383 Rupp, P., see Ayers Russ, S., see Browning Sand, H. J. S., see Slator Sarkaria, R. S., and B. W. Hammer, 1928, Influence of temperature on the changes produced in milk by certain bacteria. J. Dairy Science 11, 89 156 Sattler, W., 1928, Untersuchungen iiber die Absterbegeschwindig- keit einiger fiir den Molkereibetrieb wichtiger Bakterien. Milchw. Forschungen 7, 100 295, 320 ScHiERBECK, N. P., 1900, Ucber die Variabilitat der Milchsaure- bakterien in Bezug auf die Garungsfahigkeit. Archiv f. Hygiene 38, 294 151 ScHULTz, J. H., und H. Ritz, 1910, Die Thermoresistenz junger und alter Coh-Bakterien. Centr. f . Bakt. I Abt, Orig. 54, 283 . . 289 ScHtJTZENBERGER, P., 1876, Lcs Fermentations. Paris, Librairie Germer Bailli6re 12 ScHWENDENER, sec NaegcU Seydel, S., see Koch Sherman, J. M., and W. R. Albus, 1923, Physiological youth of bacteria. J. Bact. 8, 127 289 Skrabal, a., 1916, Reaktionsgeschwindigkeit — Temperatur — Stu- dien. Sitz. Ber. Akad. d. Wissensch. Wien, Math, naturw. Klasse II b, 125, 259 321, 328 Slator, A., 1906, The chemical dynamics of alcoholic fermentation by yeast. J. Chem. Soc. 89, 128 103, 114, 116, 122, 133, 135 , 1913, The rate of fermentation by growing yeast cells. Biochem. J. 7, 197 406 , 1916, The rate of growth of bacteria. Trans. Chem. Soc. 109, 2 190, 403 and H. J. S. Sand, 1910, The role of diffusion in fermentation by yeast cells. J. Chem. Soc. 97, 922 107, 116 Smith, J. H., 1921, The killing of Botrytis spores by phenol. Ann. AppUed Biol. 8, 27 284 , 1923, The killing of Botrytis cinerea by heat. Ann. AppUed Biol. 10, 335 321 Smith, Theobald, 1896, Reduktionserscheinungen bei Bakterien und ihre Beziehungen zur Bakterienzelle. Centr. f. Bakt. I Abt. 19, 181 97 Septhenson, Marjory, 1930, Bacterial Metabolism. London; Longmans, Green and Co 57 AUTHOR INDEX 425 Page SoHNGEN, N. L., 1906, Ueber Bakterien, welche Methan als Kohlen- stoffnahrung und Energiequelle gebrauchen. Centr. f. Bakt. II Abt. 15, 513 166, 345 , 1909, Ureumspaltung bei Nichtvorhandensein von Eiweiss. Centr. f . Bakt. II Abt. 23, 91 37, 61, 78, 166 Steppuhn, O., see Utkina-Ljubowzowa Stern, A. L., 1901, The nutrition of yeast. III. Trans. Chem. Soc. 79, 943 248, 253 Stokoe, W. N., 1928, The rancidity of coconut oil produced by mold action. Biochera. J. 22, 80 159 Stone, R. P., see Hilliard Stoughton, G., see Gage Strauss, W., 1929, Objektive Nephelometrie mittels des Mollschen Triibungsmessers, demonstriert am Beispiel der Bakterien- zahlung. Centr. f. Bakt. I Abt. Orig. 115, 228 162 Sturges, W. S., L. B. Parsons and E. T. Drake, 1929, The nature of the volatile acid produced by Clostridium histolyticum from proteins. J. Bact. 18, 157 60 , see also Parsons and Rettger Struyk, a. p., see lOuyver Suchtelen, F. H. van, 1923, Energetik und Mikrobiologie des Bodens. Centr. f. Bakt. II Abt. 58, 413, also 71, 53 and 79, 108 30 StJPFLE, K., und A. Dengler, 1916, Die Bedeutung optimaler Nahrboden zur Nachkultur bei der Prufung von Desinfektions- verfahren. Archiv f. Hygiene 85, 189 270 und Alfred Mijller, 1920, Ueber die Rolle der Adsorption bei der Einwirkung von Sublimat auf Bakterien. Archiv f. Hygiene 89, 351 338 Tammann, G., 1895, Zur Wirkung ungeformter Fermente. Z. physik. Chemie. 18, 426 124, 138, 216, 228 Tangl, F., 1903, Beitrage zur Energetik der Entogenese. Pfliiger's Archiv 98, 475 27, 29, 32, 183 Tanner, F. W., 1925, The "Bios" question. Chemical Reviews 1, 397 202 and G. I. Wallace, 1929, The standardization of disinfect- ants. Centr. f . Bakt. I. Abt. Ref . 114, 161 355 and B. W. Williamson, 1928, The effect of freezing on yeasts. Proc. Soc. Exp. Biol, and Med. 25, 377 315 Terroine, E. F., R. Wurmser and J. MontaniS, 1922, Influence de la constitution des milieux nutritifs sur la composition de 1' Aspergillus niger. Compt. rend. Acad., Paris. 175, 541. . 18, 244 Thayer, L. A., 1931, Bacterial genesis of hydrocarbons from fatty acids. Bull, of the Amer. Assoc, of Petroleum Geologists 15, 441 77 426 PHYSIOLOGY OF BACTERIA Page ThjSnard, L. J., 1803, Memoire sur la fermentation vineuse. Ann. Chim. Phys. 46, 294 12 Thiele, 1896, Die Temperaturgrenzen der Schimmelpilze. Dis- sertation, Leipzig 157 Thornton, H. R., and E. G. Hastings, 1929, Studies on the oxida- tion-reduction in milk. J. Bact. 18, 293 and 319. . 86, 94, 98, 199 ToROSsiAN, C, see Hilliard Traube, J. 1919, Die physikalische Theorie der Arzneimittel- und Giftwirkung. Biochem. Z. 98, 177 and 197 366 Trauner, R., see Rieger Treece, E. L., 1928, Gas production from commercial peptones by B. aerogenes and B. coli. J. Infect. Diseases 42, 495 76 Unger, T., see Lindner Utkina-Ljubowzowa, X., and O. Steppuhn, 1930, Ueber die Identitat der Zellproteasen verschiedensten Ursprungs. Bio- chem. Z. 220, 41 66 Valley, G., and L. F. Rettger, 1927, The influence of carbon dioxide on bacteria. J. Bact. 14, 101 168, 203 VAN and von, see corresponding names. VoGEL, J. see Gerlach VosBURGH, W. C., see Nelson Waksman, S. a., and J. S. Joffe, 1922, Thiohacillus thiooxidans, a new sulfur-oxidizing organism from soil. J. Bact. 7, 239 345 Waldschmidt-Leitz, E., 1929, Enzyme actions and properties. New York, J. Wiley and Sons 116 Walker, J. T. A., see Rideal Wallace, G. J., see Tanner Warburg, Otto, 1912, Untersuchungen iiber die Oxydationsprozesse in Zellen, III. Miinch. med. Wochenschr. 47, 2550 33 , 1921, Physikalische Chemie der Zellatmung. Biochem Z. 119, 134 97 Ward, H. M., 1895, On the biology of Bacillus ramosus (Frankel), a schizomycete of the river Thames. Proc. Royal Soc. London. 58, 265 216, 219, 251 Watson, H. E., 1908, A note on the variation of the rate of disinfec- tion with change of the concentration of the disinfectant. J. Hygiene 8, 536 346, 348 Watterson, Ada, 1904, The effect of chemical irritation on the respiration of fungi. Bull. Torrey Bot. Club 31, 291 138 Weber, H., see Meyerhof Webster, L. T., 1925, Biology of Bact. lepisepticum. J. Exper. Med. 41, 571 202 Wehmer, C., 1913, Selbstvergiftung in PenicilHum-Kulturen. Ber. d. Deutsch. Bot. Ges. 31, 210 170 AUTHOR INDEX 427 Page Weiss, H=, 1921, The heat resistance of spores with especial reference to the spores of B. botulinus. J. Infect. Diseases 28, 70 323 Went, F. A. F. C, 1901, Monilia sitophila, ein technischer Pilz Javas. Centr. f. Bakt. II Abt. 7, 544 157 Werkman, C. H., see Fulmer Wetham, M. D., see Quastel Whipple, G. C, and A. Mayer, 1906, On the relation between oxygen in water and the longevity of the typhoid bacillus. Trans. Amer. Public Health Assoc, 1905 meeting, 31, II. . . 387, 389 Whittier, E. O., see Rogers WiEMANN, B., see Grimmer WiLDiERS, E., 1901, Une nouvelle substance indispensible au developpement de la levure. La Cellule 18, 313 202 Williams, O. B., 1929, The heat resistance of bacterial spores. J. Infect. Diseases 44, 421 320, 329 Williamson, B. W., see Tanner Wilson, G. S., 1922, The proportion of viable bacteria in young cultures with special reference to the technique employed in counting. J. Bact. 7, 405 382 Winslow, C.-E. a., and O. R. Brooke, 1927, The viability of various species of bacteria in aqueous suspensions. J. Bact. 13,235 382 and E. E. Lochridge, 1906, The toxic effect of certain acids upon typhoid and colon bacilli in relation to their degree of dissociation. J. Infect. Diseases 3, 547 350 , see also Falk Wyckoff, R. W. G., 1930, The killing of certain bacteria by X-rays. J. Exper. Med. 52, 435 and 769 261, 372, 374 and T. M. Rivers, 1930, The effect of cathode rays upon certain bacteria. J. Exper. Med. 51, 930 298, 374 Zeug, M., 1920, AequiUbrierte Salzlosungen als indifferente Sus- pensions-flussigkeiten fur Bakterien. Archiv f . Hygiene 89, 175 380 ZiKEs, H., 1913, Ueber den Einfluss von Aluminium auf Hefe und Bier. Centr. f. Bakt. II. Abt. 39, 122 139 ZiKES, H., 1919 a, Ueber den Einfluss der Konzentration der Wiirze auf die Biologic der Hefe. Centr. f. Bakt. II Abt. 49, 174. ... 206 , 1919 b, Ueber den Einfluss der Temperatur auf verschiedene Funktionen der Hefe. Centr. f. Bakt. II Abt. 49, 353 224 ZoLLER, H. F., and S. M. Eaton, 1923, The phenol coefficient and relative disinfecting power of sodium hypochlorite. Abstr. Bact. 7, 6 346 SUBJECT INDEX Acet aldehyde formation by j'-east, 51 Acetone formation, 54 yeast, 37 Acetyl methyl carbinol, 54 Acid accumulation by ammonia consumption, 170 Acid amid fermentation, 59, 77 Acidity adjustment by organisms, 255 Acquired resistance, 299 Adaptation of bacteria to disin- fectants, 299 environment, 298 temperature, 225, 300 Adsorption of poison by char- coal, 338 wound tissue, 337 Aeration of bacterial cultures, 209 Aerobic bacteria, definition, 79 Age and reduction potential, 203 Age, death by, 374 Alcohol oxidase, 23, 38, 65 Alcohol oxidation by palladium, 64 Alcohol utilisation by yeast, 176 Alcoholase, 23, 37 temperature efifect upon, 130 Alcoholic fermentation and sugar concentration, 116 by-products of, 42 mechanism of, 49 retarded by alcohol, 109, 111 temperature coefficient, 1 22, 133 Amino acids, types of decomposi- tion, 57 Amins from amino acids, 59 Ammonia oxidation and oxygen concentration, 117 Amyl alcohol formation, 43, 58 as waste product of con- struction, 147 Anaerobic bacteria, definition, 79 conditions in aerobic cul- tures, 80 Antidotes in disinfection, 338, 358 Anti-oxidants, 79 Anti-protease in yeast, 39 Arrangement of molecules in cell, 261 Aspartic acid, reduced to succinic, 63 Asphyxiation of bacteria, 387 Autolysis of yeast, 12 Autolytic enzymes, uniformity of, 67 Auto-oxidation of cells, 18 Availability of energy, 181 B Bacteria in oil, heat resistance, 312 size of, 395 Bad. coli, types of fermentation, 53 typhosum, types of fermenta- tion, 53 Bacteriophage, size of, 397 growth mechanism, 264 Balance sheets of yeast growth, 175 Balanced salt solutions, 380 Bios, 202 429 430 SUBJECT INDEX Bios problem = oxygen problem, 203 Blood charcoal as oxygen activa- tor, 97 Bread yeast manufacture, 177, 237 {See also Yeast.) Buffering of reduction potential, 94 Building material of cells, 163 and yeast crop, 238 Butyl alcohol fermentation, 55 Butylene glycol, 54 Butyric acid fermentation, 54 C Cadaverin from lysin, 59 Calories, {see Energj^, and Heat). Cannizzaro reaction, 44, 51 Caprylic acid fermentation, 159 Carbolic acid coefficient, {see Phe- nol Coefficient). Carboligase, 65 Carbon dioxide from peptone, 76, 174 required for growth, 168 Carbon, sources of, 164 Carboxylase, 65 Cardinal temperatures, {see Opti- mum, Maximum, Mini- mum Temp.). Catabolism, {see Endogenous CataboHsm). Catalase, 65 Cathode rays, causing death, 374 Cell construction, material for, 163 waste products of, 173 number and fermentation end- point, 146 structure, molecular, 261 volume as measure of growth, 162 of old and young cells, 162 Charcoal as oxygen activator, 97 Chemical disinfection, 335 Chemical stimulation of endpoint of growth, 254 fermentation, 137 growth rate, 227 Chromosomes, 259 in bacteria, 261 Citric acid from sugar, 158 Clostridium, putrefaction by, 60 Coagulation as cause of all death, 367 of protein by heat, 327 Co-enzyme of yeast, 38 Cold, death by, 314 shock, 378 Combustion heat of bacteria, 28, 180 inorganic compounds, 71 media, 28, 180 organic compounds, 72,f 180 Concentrated solutions and death by heat, 330 Concentration exponent of disin- fectants, 347 change of, 352 of cells, and rate of fermen- tation, 114 endpoint of fermentation, 146 substrate and endpoint of fermentation, 141 growth, 245 enzyme action, 116 rate of fermentation, 116 growth, 203 type of fermentation, 157 Creatin as product of starvation, 15 Crops, {see Endpoint of Growth). Cystein as anti-oxidant, 79 cause of oxido-reduction, 97, 98, 100 reduction potential of, 88 SUBJECT INDEX 431 Death as function of the test medium, 271, 337 by age, 374 by cathode rays, 374 by drying, 303 dry heat, 310 freezing, 313 heat in concentrated solu- tions, 330 Ught, 371 moist heat, 317 oxidation, 19, 30, 305 poisons, 335 mechanism of, 355 (See also Disinfection.) starvation, 377, 384 suffocation, 387 X-rays, 372 definitions of, 7, 270, 272 irreversibiUty of, 272 measurement of, 272 order of, 274 rate and concentration of poison, 346 decreasing, 290 definition of, 278, 280 formula of, 280 ionisation, 339 Death rate, increasing, 281, 287, 296, 367 variabihty with same or- ganisms, 299 with drying, 304 freezing, 317 moist heat, 319 poisons, 339 temperature effect, 363 starvation, 377, 384 Decarboxylation, 42, 47, 76 Decreasing death rate, 290 rate of fermentation, 104, 107 growth, 199 Dehydrogenations in fermenta- tion, 48 (See also Oxidations.) Desiccation, death by, 303 Diffusion, influence upon fermen- tation, 116 Dilute solutions, growth in, 250 Diminishing returns in growth, 250 Disinfectants, adsorbed by char- coal, 338 tissues, 337 reacting with medium, 336 Disinfection and ionisation, 342 chemical, 335 concentration of poison and, 346 dormant stage in, 360 measuring of, 272, 335 Disinfection, mechanism of, 355 (See also Death.) temperature coeflBcient of, 363 theories of, 366 Dismutase, 65 Dormancy and death, 7, 360 Dormant stage in disinfection, 360 Drosophila, order of death of, 275, 288 Dry bacteria, death of, 303 heat, death by, 310 steriHsation, 311 Drying, methods for — bacteria, 304, 306, 307, 309 Dyeing of dead cells, 273 E Eh, definition, 85 Endo-enzymes, energy from, 26 Endogenous catabohsm, definition. 7,9 of animal tissues, 15 bacteria, 14, 20 man, 9 molds, 18 yeast, 11 432 SUBJECT INDEX Endoproteolysis, 11 Endotryptase, 38 Endpoint method of disinfection, 272 of fermentation and cell con- centration, 146 fermentation products, 109, 142 food concentration, 141 temperature, 151 cause of, 140 constancy of, 146 of growth, 230 and acidity, 255 aeration, 209, 233, 341 building material, 237, 244 chemical stimulants, 254 composition of medium, 230, 233, 234, 237, 244 food concentration, 245 growth rate, 231 size of seeding, 230 temperature, 251 volume of medium, 230 cause of, 199, 231 constancy of, 231 with abundant food, 244 Energy, amounts from different compounds, 23, 25, 32, 71, 180 available, 181 from enzyme action, 23, 77 oxidations, 71, 180 Uberation per cell, 261 measurement of, 26 need of, 32, 180 sources of, 22 utihsation by young and old cells, 183 for growth, 32, 180 other purposes, 34, 101 Enzyme action and concentration of substrate, 116 Enzyme action and temperature, 124 concentration and equilib- rium, 150 deterioration, 127 in living cells, 16, 131, 215 level in cells, 17, 131 number of molecules per cell, 258 specificity of, 65, 67 stability within the cell, 16, 131 Equations of fermentation, 40 Facultative bacteria, definition, 79 Fat hydrogenation by nickel, 64 Fatty acids split into CH4, 77 Fermentation, alcoholic (see Alco- holic Fermentation), cause of cessation of, 108 constant, 109 decrease of rate of, 108 definition of, 24 endpoint of, 140 intermediate steps in, 44 lactic (see Lactic Fermenta- tion), limiting factors of, 144 mechanisms of, 64 products and fermentation, 109, 142 growth, 199, 232 purpose of, 180 Fermentation, rate of, definition, 102 decrease of, 104, 107 measurement of, 102 and chemical stimulants, 137 concentration of cells, 114 food, 116 fermentation products, 107 growth, 204, 265 temperature, 124 types of, 49, 66 SUBJECT INDEX 433 Fermentation, types of, variation in, 155 Fermenting enzymes, 23, 25 capacity of a cell, 104, 106, 404 Final amount of fermentation products (see Endpoint of Fermentation). Final growth {see Endpoint of Growth). Fittest survivors, survival of, 298 Food concentration and endpoint of growth, 245 lowest, for growth, 250, 383 rate of fermentation, 116 growth, 203 selection, 25, 158 Formic acid as fermentation prod- uct, 62 Freezing, death by, 313 Fruit flies, order of death, 275, 288 Functions, definition of, 5 Fusel oil, origin of, 42, 43 as waste product of growth, 147 composition of, 42 G Gas formation by Bad. aerogenes, 54 coll, 53 streptococci, 76 yeast, 105 from peptone, 76, 174 ratios, 53, 64 Gelatin liquefaction, 56 Generation time, formula, 190, 399 {See also Growth Rate.) Genes, inactivation as cause of death, 284 number of, per cell, 259 size of, 259 Glutathione, as cause of oxido- reduction, 97 reduction potential, 88 Glyceric aldehyde in fermenta- tions, 45 Glycerol formation by yeast, 50 Growth and fermentation, 180, 204, 246, 265 and multiplication, 6, 162 catalysts, 215 number per cell, 259 cessation of {see Endpoint of Growth), constant, 109, 403 Growth curves, 192, 404 mathematical formulation of, 198 endpoint of {see Endpoint of Growth), optimum temperature, 216 maximum temperature, 216 measurenent of, 162 mechanism, 257 complexity of, 264 simpler of parasites, 264 minimum temperature, 157, 223 products, intermediate, 263 temperature coefficient of, 214, 221 Growth-inhibiting cell products, 234 Growth-producing agents, 215, 259 chemical stimulation of, 227 computed from fermentation products, 191 constant, 109, 403 decrease of, 199 inhibition by poison, 227 methods of measuring, 189 rate and building material, 204 different foods, 204, 265 endpoint of growth, 231 energy supply, 205, 211 food concentration, 203 oxygen, 209 temperature, 194, 213 434 SUBJECT INDEX Growth rate of bacteria, 191 B. ramosus, 219 Bacterium ^,194 aerogenes, 268 coli, 206, 222 typhosum, 196, 205 individual cells, 266 Mycohact. tuberculosis, 192 Ps. fiuorescens, 224 Strept. lactis, 207, 209, 210, 220, 400 yeast, 191, 206, 211, 225, 253 variability of, 266 H Heat coagulation of protein, 327 liberation by bacteria, 29, 34, 180 from humus, 30 per cell per hour, 31 of formation (see Combustion Heat). Hemoglobin, coagulation by heat, 327 Hexose phosphate, 38, 45 Hippuric acid fermentation, 61 Histidine, fermentation of, 58 Humus in soil, heat liberation from, 30 Hydrogen acceptors and donators, 44, 62, 68 activation, intensity of, 66 and reduction potential, 66 activators, 64, 68 electrode, 85 fixation, 173 Hydrogenation in fermentation, 49 (See also Reduction.) Hydrolysis, energy from, 25, 77 Inanation, (see Starvation). Indicators for reduction potentials, 92 Interface tension causing death, 366, 369 Intermediate fermentation prod- ucts, 44 Intermediate growth products, 263 Intramolecular fermentations, 70, 75 Invertase, 23 lonisation and disinfection, 342 Lact-acidase, 23, 37, 65 Lactase, 23 Lactic, fermentation and lactose concentration, 118 intermediate products, 52 retarded by lactic acid, 112 Lactozymase, 65 Lag period, 193 and cell concentration, 202 oxygen, 202 resting stage of cells, 201 through poisoning, 196 surface tension depres- sion, 196 Law of diminishing returns, 250 Light causing death, 371 growth stimulation by, 23 Lipase, 23 Lipase in yeast, 39 Logarithmic growth phase, 193 order of death, 274, 282 (*S^ee also Order of Death.) Longevity of dry bacteria, 303 Low temperature, harmful effect upon bacteria, 154, 314, 378 Luminescent bacteria, 34 M Maltase, 23 Maximum temperature of fermen- tation, 131 growth, 157, 216 SUBJECT INDEX 435 Medium, influence on disinfection tests, 256, 271, 337 Methane from fatty acids, 77 Methyl glyoxal in fermentations, 45, 47 Methyl glyoxalase, 65 Methylene blue as activator of respiration, 99 reduction potential of, 88, 93, 95 Milk, reductase test in, 94 reduction potential of, 95 Minimum temperature of fermen- tation, 133 growth, 157, 223 Mitogenetic rays, 23, 34 Molecules, arrangement of — in cell, 261 number of — fermented per cell, 258 Motility, energy required for, 34 Multiplication of bacteria, 162, 189, 398 rate of {see Growth Rate). Mustard seeds, order of death, 275, 287 Mutase, 65 N Neutralisation heat as source of energy, 77 Neutralized cultures of strep- tococci, growth in, 232 fermentation in, 144 Nitrite oxidation and nitrite con- centration, 117 Nitrobacter, cause of suffocation of, 392 energy utilisation by, 188 rate of oxidation, 117 Nitrogen content as measure of growth, 162 fixation, 168, 184 loss of starving cells, 12, 15, 384 Nitrogen, sources of, 168 and yeast properties, 238 Nitrosomonas, rate of respiration, 117 of starving cells, 13, 17 Old age, death by, 374 cause of, 199, 231, 374 Optimum temperature of enzyme action, 129 fermentation, 132 growth, 216 Order of death, definition, 274 by age, 375 chemicals, 339 light, 371 moist heat, 319 starvation, 377 384 X-rays, 372 interpretation of, 282, 293 summary of all experi- ments on, 295 with very slow death rates, 328 growth, 267 Oxidation, increased by methylene blue, 74 of cells, 18, 203, 305, 385 of dry bacteria, 305 old cells, 203 Oxidation retarded by cyanide, 74 Oxidation-reduction potential (see Reduction Potential). Oxido-reductions, 70 definition, 45 Oxygen and growth rate, 209 activation, 44, 68 activators, 65 charcoal as, 97 missing in anaerobes, 64, 74 atoms shifted in molecule, 75 436 SUBJECT INDEX Oxygen and growth rate, 209 consumption by growing bac- teria, 73 of Bad. coll, 186 Ps. fluorescens, 187, 398 fixation, 173 in fermentations, 69 position in molecule, 75 requirements for fermentation, 69, 78 growth, 78, 80 solubility in water, 80 source of, 78, 172 supply and endpoint of growth, 209, 241 toxicity of, 19, 78, 305 uptake of starving bacteria, 20 Protection of living cells by dead, 318, 338 Protein coagulation by heat, 327 content of mold mycelium, 244 yeast, 242, 247, 253 fermentations, 56 Protein molecules, size of, 260 Protein-free diet of animals, 9 yeast, 13, 383 Protein sparing effect of sugar, 159 Proteolysis and autolysis, 10 Prototrophic fermentations, 61 Putrefaction, anaerobic, 59 Pyruvic acid as anti-oxidant, 79 in alcoholic fermentation, 44, 50 R Partial functions, definition, 5 Partition coefficients of disinfect- ants, 356 Pepsin, 23 Peptone, digestibility of, 210 gas from, 76, 174 milk, cause of higher acidity in, 148 Phenol coefficient, limitations, 353 method, 354 variation with species, 346 Physiological salt solution kilUng bacteria, 380 youth of bacteria, 288 Phosphates, importance in fer- mentation, 38 Pin point colonies, 244 Poising of reduction potential, 94 Poisons causing death, 335 stimulation, 137, 227, 254 mode of action of, 355 Potentials in cells, 33, 100, 261 Propionic acid fermentation, 52 Tn, definition, 85, 90 Rancidity of coconut oil, 159 Reductase, 65 Reductase test in milk, 94, 95, 99 Reduction capacity, 94 Reductions, energy balance of, 72 Reduction intensity, 83, 94 potential and acidity, 88 hydrogen activation, 66 pressure, 85 buffering of, 94 capacity of, 100 indicators, 92 loss by old cells, 84, 203 measurements, 82, 92 poising of, 94 of B mycoides, 84 B. subtilis, 84 Bad. coli, 82, 84, 86, 101 Bad. proteus, 86 Corynebaderium, 87 culture media, 85, 97 hydrogen hberation, 85, 89 SUBJECT INDEX 437 Reduction potential of methylene blue, 88, 93. 95 pneumococci, 86 soils, 94 staphylococci, 87 streptococci, 86 washed cells, 99 yeast, 82 within the cell, 92, 99 Regeneration of enzymes in the cell, 16, 131, 215 Rennet, 23 Respiration (see Oxidation or Oxy- gen Consumption). Retardation of fermentation, 107 growth, 199 Salt balance, 164, 380 Salt solution, physiological, 380 Seeds, order of death, 275, 287 Selection of food, 158 of resistant strains, 298 Sensitive zone of bacteria, 261, 372 Shadow forms of Bad. coli, 198 Size of bacteria, 397 of bacteriophage, 397 genes, 259 protein molecules, 260 yeast, 397 Soap, disinfectant value of, 370 Soil, reduction potential of, 94 temperature increase in, 30 Soul and biochemistry, 1 Spores, heat resistance of, 320, 323, 329 Staining of dead cells, 273 Starvation, cause of death by, 384 death by, 377 of bacteria, 14, 20, 377 man, 9 molds, 18 yeast, 11 Stimulation, chemical, of fermen- tation, 137 crop, 254 growth rate, 227 Succinic acid formation by yeast, 43 bacteria, 53, 63 from amino acids, 43, 63 sugar, 53 Suffocation of bacteria, 186, 387 cause of, 392 Sugar retarding protein decomposi- tion, 159 Sulphur content as measure of growth, 162 Surface of microorganisms, 397 tension depression causing death, 366, 369 Survival of the fittest, 298 Survivor curves, 281, 285, 292, 297 Synthesis and energy potential, 262 Synthetic media, 164 Tammann principle in chemical stimulation, 138, 228 enzyme action, 129 fermentation, 132, 138 growth, 216, 228 Temperature, adaptation to, 225 and endpoint of fermentation, 151, 154 growth, 251 enzyme action, 124 growth rate, 194, 213 rate of fermentation, 119, 130 coeflficient, decrease of, 321 definition, 119 of death by chemicals, 363 dry heat, 310 hght, 372 moist heat, 320 starvation, 18, 379 438 SUBJECT INDEX Temperature coefficient of enzyme destruction, 124, 129 fermentation, 122, 130 growth, 214, 221 poisoning, 363 protein coagulation, 326 influence upon yeast com- position, 253 limits, controlled by food, 157 Thermal death point, 332 time, 334 Thermophilic bacteria, death by heat, 329 of spores by heat, 323 Toxicity of chemicals (see Disinfec- tion). Trapping of intermediate products, 51 Trypsin, 23 Tryptophane changed to trypto- phol, 44 Turbidity as measure of growth, 162 Types of fermentation, 66, 155 changed by concentration, 157 kind of medium, 155 temperature, 156 Tyrosine changed to tyrosol, 44 U Ultra-violet Ught, death by, 372 radiated by bacteria. 23, 34 Urea bacteria, fermenting capac- ity, 107 Urea bacteria, oxygen require- ments, 78 fermentation, 61, 77 Urease, 23, 36 Uric acid fermentation, 61 Urocanic acid from histidine, 58 Viabihty of dry bacteria, 303 bacteria in water, 377, 382 Vinegar oxidase, 23, 38 Vitamin C, inactivated by heat, 325 W Waste products of construction, 173 Wear and tear, definition, 16 Weight of microorganisms, 397 Wheat seeds, order of death, 275, 287 X-rays and bacteria, 261 causing death, 372 Yeast composition and tempera- ture, 253 with different foods, 242, 247, 253 Yeast crop (see Endpoint of Growth), dough raising power, 240 fermenting capacity, 105 growth, 175, 211, 238 keeping quahty, 240, 243 manufacture, 177, 237 protein content (see Yeast Composition), quantity and endpoint of fermentation, 147 rate of gas production, 105 size of, 396 softening of, 240, 243 types of fermentation, 49 Yield (see Endpoint of Growth). Youth, physiological, 288 Variability of growth rate, 266 disinfection results, 345 Zymases, 23, 35