i'^\.'\ JOURNAL OF BACTERIOLOGY VOLUME VII BALTIMORE, MD. 1922 UId. CONTENTS No. 1, Januaky, 1922 Certain Genera of the Clostridiaceae. Studies in Pathogenic Anaerobes. V. Hilda Hempl Heller 1 Studies upon Agglutination in the Colon-Typhoid Group of Bacilli. O. Ishii. 39 A Study of Spontaneous Agglutination in the Colon-Typhoid Group of Bacilli. O. Ishii 71 The Sources and Characteristics of the Bacteria in Decomposing Salmon. Albert C. Hunter 85 Viability of the Colon-Typhoid Group in Carbonated Water and Carbonated Beverages. S. A. Koser and W. \V. Skinner Ill A Binocular Microscope Arranged for the Study of Colonies of Bacteria. Guilford B. Reed 121 An Investigation of American Stains. Report of Committee on Bacteri- ological Technic. Prepared by H. J. Conn, Chairman 127 No. 2, March, 1922 Our Society. F. C. Harrison 149 Notes on the Gram Stain with Description of a New Method. Victor Burke. 159 Disinfection Studies. The Effects of Temperature and Hydrogen Ion Con- centration upon the Viability of Bad. coli and Bad. typhosum in Water. Barnett Cohen 183 Microorganisms Concerned in the Oxidation of Sulfur in the Soil. I. Intro- ductory. Selman A. Waksman 231 Microorganisms Concerned in the Oxidation of Sulfur in the Soil. II. Thio- bacillus thiooxidans, a New Sulfur-oxidizing Organism Isolated from the Soil. Selman A. Waksman and J. S. Joffe 239 The Production of Pink Sauerkraut by Yeasts. E. B. Fred and W. H. Peterson 257 Studies on the Biology of Lactic Acid Bacteria: a Summary of Personal Investigations. Costantino Gorini 271 A Method for the Cultivation of Anaerobes. L. D. Bushnell 277 Influence of Vacuum upon Growth of Some Aerobic Spore-Bearing Bacteria. L. D. Bushnell 283 Substitution of Brom-Thymol-Blue for Litmus in Routine Laboratory Work. H. R. Baker 301 The Use of Domestic Methylene Blue in Staining Milk by the Breed Method. W. A. Wall and A. H. Robertson 307 No. 3, May, 1922 Studies on Cultural Requirements of Bacteria. I. J. Howard Mueller 309 Studies on Cultural Requirements of Bacteria. II. J. Howard Mueller 325 iii IV CONTENTS A Method for Counting the Number of Fungi in the Soil. Seknan A. Waksman 339 Studies on Thermophilic Bacteria. I. Aerobic Thermophilic Bacteria from Water. Lethe E. Morrison and Fred W. Tanner 343 An Apparatus for the Rapid Measurement of Surface Tension. Robert G. Green 367 No . 4, JuLT, 1922 Quantitative Determinations of Some of the Biochemical Changes Produced by a Saprophytic Anaerobe. L. D. Bushnell 373 The Proportion of Viable Bacteria in Young Cultures with Especial Refer- ence to the Technique Employed in Counting. G. S. Wilson 405 A Method of Detecting Rennet Production by Bacteria. H. J. Conn 447 No. 5, September, 1922 The Relation of Vitamines to the Growth of a Streptococcus. S. Henry Ayers and Courtland S. Mudge 449 Salt Effects in Bacterial Growth. II. The Growth of Bad. Colt in Relation to H-Ion Concentration. James M. Sherman and George E. Holm. . . . 465 A Note on the Morphology of Bacteria Sjonbiotic in the Tissues of Higher Organisms. Ivan E. Wallin 471 Observations on the Properties of Bacteriolysants (D'Herelle's Phenomenon, Bacteriophage, Bacteriolytic Agent, etc.). Part I. Wilburt C. Davison. 475 Observations on the Nature of Bacteriolysants (D'Herelle's Phenomenon, Bacteriophage, Bacteriolytic Agent, etc.). Part II. Wilburt C. Davison 491 Clostridium Pulrificum (B. Putrificus Bienstock), A Distinct Species. George F. Reddish and Leo F. Rettger 505 Transparent Milk as a Bacteriological Medium. J. Howard Brown and Paul E. Howe 511 The Use of Agar Slants in Detecting Ammonia Production and its Relation to the Reduction of Nitrates. G. J. Hucker and W. A. Wall 515 Methods of Pure Culture Study. Report of Committee on Bacteriological Technic. H. J. Conn, chairman 519 An Investigation of American Gentian Violets. Report of Committee on Bacteriological Technic. Prepared by H. J. Conn, Chairman 529 No. 6, November, 1922 Method for the Isolation of Bacteria in Pure Culture from Single Cells and Procedure for the Direct Tracing of Bacterial Growth on a Solid Medium. J. Prskov 537 Bacterial Autolysis. W. S. Sturges and L. F. Rettger 551 Bacillus Ilemoglobinophilus canis (Friedberger) {Hemophilus cants emend). T. M. Rivers 579 Salt Effects in Bacterial Growth. III. Salt Effects in Relation to the Lag Period and Velocity of Growth. J. M. Sherman, G. E. Holm and W. R. Albus 583 CONTENTS V Further Observations on the "Color Standards" for the Colorunetric Deter- mination of H-Ion Concentration. L. S. Medalia 589 The Cause of Explosion in Chocolate Candies. John Weinzirl 599 Microorganisms Concerned in the Oxidation of Sulfur in the Soil. IV. A Solid Medium for the Isolation and Cultivation of Thiobacillus ihio- oxidans. Selman A. Waksman C05 Microorganisms Concerned in the Oxidation of Sulfur in the Soil. V. Bac- teria Oxidizing Sulfur under Acid and Alkaline Conditions. Selman A. Waksman 609 VOLUME VII NUMBER 1 JOURNAL OF BACTERIOLOGY OFFICIAL ORGAN OF THE SOCIETY OF AMERICAN BACTERIOLOGISTS JANUARY, 1922 EDITOR-IN-CHIEP C.-E. A. WiNSLOW It is characteristic of Science and Progress that they continually open new fields to our visions. — PASTEtm. PUBLISHED BI-MONTHLY WILLIAMS & WILKINS COMPANY BALTIMORE, U.S.A. i-.Dtered Be Becond-clBBBmatter April 17, 1916, at the poetoffice at Baltimore, Maryland, imdertbeact of starch 3f 1870. Acceptencefor mailing at special rate of postage provided for id SMtion 1103. Act of October 3, 1917. Authorized on July 16, 1918. Price, net postpaid Copyright 1922, Williams and WiUdns Comimny $5.00 per volume. United States, Mexico, Cuba $5.25 per volume, Caxiada $5.50 per volume, other countries Made in United States of America The actual direct benefits from Gastron are a known quantity, Unmistakably experienced by the patient, observed by the physician — in disorders of gas- tric function. Inevitably the better digestion promotes better nutrition, strengthens resistance, encour- ages and heartens the patient, thus promotes a condition of body and a state of mind con- ducive to restoration. In these circumstances there is an appeal for a wide application of Gastron; far-reaching indeed be its ultimate good effects. Fairchiid Bros. & Foster NEW YORK Vacuum Devices for Collecting Blood Specimens Supplied with special needles, hand ground and hand finished, which penetrate the skin with a minimum of pain and the vein without tearing. s Keidel Bleeding Tube A glass tube evacuated to an unusual degree, providing a convenient and aseptic method for the collection and transportation of blood samples. Plain, 10 c.c. With potassium oxalate, 20 c.c. Blood Culture Tube An enlarged tube containing either glucose bouillon medium, or ox-bile, glycerin and peptone medium, 50 c.c. Circular on request HYNSON, WESTCOTT & DUNNING Baltimore JOURNAL OF BACTERIOLOGY OFFICIAL ORGAN OF THE SOCIETY OF AMERICAN liACTKRIOLOGISTS DEVOTED TO THE ADVANCEMENT AND DIS- SEMINATION OF KNOWLEDGE IN REGARD TO THE BACTERIA AND OTHER MICRO-ORGANISMS Editorial Hoard Editor in-Chtej C.-E. A. WINSLOW Yale Medical ejchool, New Haven, Conn. A. Parker Kitchens F. C. Harrison, Ex officio Principal, MacDonald College, P. Q., Canada C. C. Bass R. E. Buchanan P. F. Clark F. P Gay F. P. Gorham F. C. Harrison Advisory Editors H. W. Hill E. O. Jordan A. I. Kendall C. B. LiPMAN C. E. Marshall V. A. Moore M. E.Penington F. B. Phelps L. F. Rettoer L. A. Rogers M. J. ROSENAU W. T. Sedgwick F. L. Stevens A. W. Williams H. Zinsser CONTENTS Hilda Hempl Heller. Certain Genera of the Clostridiaceae. Studies in Pathogenic Anaerobes. V 1 O. IsHii. Studies upon Agglutination in the Colon-Typhoid Group of Bacilli 9 O. IsHii. A Study of Spontaneous Agglutination in the Colon-Tjrphoid Group of Bacilli 71 Albert C. Hunter. The Sources and Characteristics of the Bacteria in Decomposing Salmon 85 S. A. KosER AND W. W. Skinner. Viability of the Colon-Typhoid Group in arbo- nated Water and Carbonated Beverages Ill Guilford B. Reed. A Binocular Microscope Arranged for the Study of Colonies of Bacteria 121 An Investigation of American Stains 125 Abstracts of American and foreign bacteriological literature appear in a separate journal, Abstracts of Bacteriology, published monthly by the Williams & Wilkins Company, undei the editorial direction of the Society of American Bacteriologists. Back volumes can be furnished in sets consisting of Volumes I, II, III and IV. Price per set, net, postpaid, $24.00, United States, Mexico, Cuba ; S2,').00, Canada; 120.00, other countries. Subscriptions are in order for Volume V, 1921. Price, per volume, $5.00, United States, Mexico, Cuba; $5.25, Canada; $5.50. other countries. fduTon'ff "o^r^ CATALOGUE is now ready for distribution In this edition there are listed 9,316 apparatus items illustrated by original wood cuts, and 5,549 reagentlitems. A telegraphic code word is given for each size of every apparatus item and for each stock package of every reagent. The prices printed are net, with quantities contained in. and discounts allowed forjoriginal packages. The selection of merchandise listed is based upon thirty years of ex- perience in the supply of apparatus and reagents to chemical, metallur- gical and biological laboratoriesjand has been made with painstaking care. The various items are described and illustrated without exaggera- tion and with as much dicussionfand illustration of underlying principles and details of construction as is necessary for wise selection. The name of the author, with full reference to his publications bearing on the item in question, is given where such data is likely to be of value. We suggest that scientists in requesting copies of tliis catalogue state the capacity in which 'they are connected with organized laboratory work, and mention this journal in writing. In large laboratories and departments where many workers are engaged, we prefer that the request for catalogues come from the professor in charge of the department, laboratory director, chief of bureau, chief chemist, pathologist, bacteriologist or other scientist in charge of laboratory work, with the names of colleagues, asso- ciates or assistants to whom, in his judgment, additional cata- logues should be sent. Requests from firms or institutions should state the name of the scientist or officer for whom the catalogue is desired so that we may check against subsequent requests from individuals. Such cooperation assures adequate distribution and at the same time [reduces unnecessary duplication in the same laboratory. While the sending of this catalogue involves no obligation to purchase of us, the expense of prepara(ion]and publication necessitates a reasonable control of the distribution. Separate pages have been prepared for sending in a loose-leaf binder to those whose interest is temporary or restricted to a given subject or "^ item rather than in the regular use of the complete catalogue. ARTHUR H. THOMAS COMPANY WHOLESALE, RETAIL AND EXPORT MERCHANTS LABORATORY APPARATUS AND REAGENTS WEST WASHINGTON SQUARE PHILADELPHIA. U. S. A. Cable address. "Balance." Philadelphia ERRATUM Journal of Bacteriology, Vol. VI, No. 6, November, 1921. Line 17 from the top of page 560 reads "from pH to 8.0." It should read "from pH 5.0 to 8.0." h::»« 'i't:.'I CERTAIN GENERA OF THE CLOSTRIDIACEAE STUDIES IN PATHOGENIC ANAEROBES. V HILDA HEMPL HELLER From the George William,^ Hooper Foundation for Medical Research, University of California Medical School, San Francisco Received for publication March 1, 1921 In a previous paper a classification was suggested for the group of anaerobic rods which include, roughly, the anaerobic members of the genus Bacillus of former workers. A family was proposed for these organisms, with the name Clostridiaceae ,^ and it was divided into two subfamihes, the Putrificoideae^ or , proteolytic anaerobes, and the Clostridioideae^ or non-proteoly- tic anaerobes. Certain genera for these groups are proposed in the present communication. These in some cases unite various described species, while in other cases the genera themselves correspond to the former idea of species. A key to the genera is also given, which is based on the action of the anaerobes on meat medium and on their general cultural behavior and morphology. Possibly the main lines for tribal organization will to some extent follow this key, but it pretends to be no more than an artificial arrangement. It is in no way complete and cannot be implicitly relied upon for purposes of classification. It is meant more as an index to the forms whose descriptions are sufficiently clear to warrant assigning them de- finite positions. The kej'^ will serve, for a time, as a nest of pigeonholes in which to place new species until material is suffi- cient for a complete reorganization, but workers should not try to place all newly discovered organisms in these genera. The conservative worker, famiUar with aerobic pathogens, who enters the anaerobic field, is all too prone to wish to ' ' identify" ' For definition see Jour. Bact., 6, 536. ' For definition see Jour. Bact., 6, 550. 1 JOURNAL OF BACTERIOLOQT, VOL. VII, NO. 1 Z HILDA HEMPL HELLER his anaerobic strains. A natural feeling exists among sys- tematists, that the description of a new-found tjrpe as a new species is to be avoided unless the describer is very sure that his type differs from all others. There is a fear that a description carefully made will some day be discarded by a future taxono- mist who finds it identical with a former description. This logical attitude has bred and fostered a wholl}'^ unscientific mania for identification of all new strains with organisms al- ready described by others. In the case of parasites or in well- known groups this procedure wiU only occasionally lead the worker astray. In the anaerobic field it will do so frequently. The scientific attitude relative to the taxonomic affinities of anaerobic bacUli is to state the group or generic relationships of the organisms and then to describe them minutely. The routine bacteriologist should be content to assign an organism to its proper genus. Colony formation in deep agar is included in the descriptions of genera, but a restriction of the genera to such t3rpes of colony form as are mentioned would be unwise. The definition of the Gram staining reactions should not be regarded as of great value in these descriptions, as the staining reactions of the organisms and the technique of workers are too variable. Sugar fermenta- tions may be relied upon in the case of actively growing species only, and are valid only for media on which the behavior of a strain is constant. The constancy of the fermentations here quoted has not been verified by myself. Pathogenicity is not to be taken as a criterion for admission to a genus. It is to be feared that bacteriological systematists who are unacquainted with the anaerobic group will object to the creation of so many genera as are here proposed. The Committee on Characterization and Classification of Bacterial Types (1920) propose but 38 genera for the orders Actinomycetales and Eubac- leriales. The influence which the medical history of the science of bacteriology has had on this classification is marked. Two- thirds of the 38 genera recognized contain parasites. If we re- gard the anaerobic bacteria as wild plants growing in soil, as we have every reason for doing because of their many species, we CERTAIN GENERA OF THE CLOSTRIDIACEAE 3 must classifj' them as do the botanists and not as do the bacteriol- ogists. The family of Orchidcae has 334 genera, the Graviineae 298, the Rubiaceae 377, the Leguminoseae 399, and the Compositae 766 (Bentham and Hooker). The following classification imi- tates the taxonomic arrangement of botanical classifications in preference to bacteriological. It must be borne in mind that the date 1920 in bacteriologic systematics corresponds roughly to the date 1760 in botanical systematics. Relative to the evolutionary history of the anaerobes we know nothing, but conjectures are not entirely out of place. Proteins and carbohydrates have existed together and have been destroyed by bacterial action since a remote geologic epoch. We cannot tell whether the organic catalyzers of one type of substance antedated those of the other or whether they are closely related. Botanists are not agreed as to what form of life first tenanted the globe. When looking for ancestral forms it is natural to seek in a group of organisms of great catalytic activity like the Clostridiaceae, for tj^pes which may synthesize simple substances, and if such are discovered it is not out of place to regard them as more closely related to the hypothetical ances'tors of the group than those that do not synthesize inorganic substances. Such reasoning is of course entirely dependent on the homogeneity of the group under discussion. We have no means of proving the homogeneity of the Clostridiaceae. But when classified according to their chemical activities these organisms form a remarkable chain whose links we are all but warranted in regard- ing as the varied end points of an evolutionary process. As our most primitive type we may well choose the nitrogen-fixing anaerobes of the soil. It is to Winogradsky that we turn for the first demonstration of such organisms. He named his anaerobic nitrogen-fixing organism Clostridium Pastorianum. Bredemann declares that nitrogen-fixation is a variable power among all organisms of the amylobacter type, and he includes under the name amylobacter, Clostridium Pastorianum of which bacillus he possesses a strain. These organisms are highly saccharolj^tic, being pectin fermenters, and they do not split gelatin. Farther up the scale we find anaerobic rods which 4 HILDA HEMPL HELLER are actively saccharolytic and do split gelatin, for example the Welch bacillus, vibrion septique, and many other fonns; these organisms do not liquefy coagulated sermn or produce hydrogen sulfide or other protein split products in any considerable quan- tity. Most of them possess no diastatic power toward pento- sans, and many fail even to split pentoses. The group next in the scale are those anaerobes which, though they fail to disinte- grate coagulated serum and muscle particles, are sufficiently proteolytic to free considerable quantities of hydrogen sulfide in media rich in sulfur, such as blood broth. The oedematiens type, and other less well-known organisms may be placed here. Some of these are strongly saccharolytic and others are weakly so. Further advances in attack on the protein molecule are almost invariably accompanied by a decrease in saccharolytic power. B. aerofoetidus, and the two strains which I term Reglillus are apparently slightly proteolytic, and only moderately saccharo- lytic. The more highly proteolytic organisms usually split mono- hexoses, glucose only, or no sugars at all. The most highly adapted catalytic anaerobes are those that produce both acid and aUcaU from their substratum in sufficient quantities to keep their hydrogen-ion end point within their optimum range and which are so highly proteolytic that they disintegrate a great variety of protein molecules and split-products. Such t^-pes are B. histohjticus and B. botulinus. Another type, known as Bifer- mentans, keeps its end point within its optimum range of growth but is not sufficiently proteolytic to continue multiplication for a long period. Such organisms are not highly but widely special- ized, and are adapted to fend for themselves because they are saccharolytic as well as proteolytic. It is to render possible a future scientific and logical classifi- cation of anaerobic organisms that this key and these generic definitions are proposed. They are manifestly incomplete, and it is certain that careful taxonomic investigations will require the emendation of many of the proposed genera. But the present paper aims to blaze a trail where meandering paths have wandered — a trail upon which a highroad may be built in the future. CERTAIN GENERA OF THE CLOSTRIDIACEAE 5 I wish most heartily to thank Dr. Karl F. ^leyer for many helpful suggestions which he has given me during the compila- tion of this classification. Artificial Key to the Geneha of the Clostridioideae A. Do not produce HjS demonstrable by lead-acetate-paper test when grown in blood-broth. B. Do not liquefy gelatin C. Very large rods that form oval spores and store up starch in carbo- hydrate media. Genus 1. Clostridium. Prazmowski em- mcnd. Heller. Type species bulyricum as described by Winogradsky (1895). CC. Rods with spherical spores. D. Do not grow on media containing much protein. Genus L'. Omelianskillus nov. gen. Type species hydrogcnicus as described by Omeliansky (1895 and 1904b). DD. Grow on ordinary anaerobe media. E. Sides of the bacilli parallel, spores strictly terminal. Genus 3. Macinloshillus nov. gen. Type species Ictanomorphus (pseudotetanus bacillus, Mcintosh ('p.32). Bacillus tetano- morphus Committee (p. 41)), as described by the Conmiittee. EE. Spores not always strictly terminal, sides of bacilli may not be parallel. Genus 4. Donglasillus nov. gen. Type species sphenoides (Bacillus sphenoides Douglas Fleming and Colebronk), as described by the Committee (p. 43). CCC. Slender rods with oval endspores, usually Gram-negative. D. Clot milk and attack various sugars, produce much acid. Genus 5. Henrillus nov. gen. Type spe- cies tertius (Bacillus iertius Henry) as described bj' Henry. DD. Do not clot milk, attack few or no sugars, produce little acid. Genus 6. Flemingillus nov. gen. Type species cochlearius (Bacillus cochlearius Douglas Fleming and Colebrook) as de- scribed by the Committee (p. 40). CCCC. Gram-positive rods which are not markedly slender and which produce oval spores. D. Clot milk, saccharolytic. E. Sponilate meagerly, attack a few sugars; occasion- ally moderately pathogenic tissue invaders. b HILDA HEMPL HELLER Genus 7. Vallorillus. nov. gen. Type species fallax (Bacillus fallax Weinberg and S^guin) as described by the Com- mittee (p. 27). EE. Sporulate readily, attack .several sugars; not known to be pathogenic. Genus 8. MuUifermentans nov. gen. Type species lenalhus {Bacillus multifermenlans tenalbus Stoddard) as described by Stod- dard (1919 b). DD. Do not clot milk. Large Gram-positive rods with long elipsoid spores. Genus 9. Hiblerillus nov. gen. Type spe- cies sexlus (bacillus VI of von Hibler) as described by von Hibler (1908). (Re- sum6 by Weinberg and S^guin (p. 202).) BB. Liquefy gelatin. C. Produce stormy fermentation of milk and sporulate on alkaline media only. Genus 10. Welchillus nov. gen. Type species aerogenes (Bacillus aerogenes caps^ilatus Welch and Nuttall), type 1 as defined by Simonds (1915 a and b). CO. Do not produce stormy fermentation of milk. D. Do not sporulate. Genus 11. Stoddardillus nov. gen. Type species egens (Bacillus egens Stoddard) as described by Stoddard (1919 a). DD. Sporulate readily. E. Gram-positive, form woolly colonies in deep agar. Typically highly pathogenic tissue invaders of many species of animals. Genus 12. Rivoltillus nov. gen. Type species vibrion (the vibrion septique of Pasteur) as defined in a future paper. EE. Gram-negative, may contain Gram-positive gran- ules. Form smooth lenticular or modified lenticular colonies in deep agar. Typically pathogenic for cattle, sheep, and guinea- pigs. Genus 13. ArloingiUus nov. gen. Type species Chauvoei (Bacterium Chauvoei Arloing, Cornevm and Thomas) as de- scribed in a future paper. AA. Produce HjS demonstrable by a lead-acetate-paper test when grown in blood broth. B. Produce a large amount of gas from carbohydrates. Heavy Gram- ])08itive rods with little or no tendency to sporulation. CERTAIN GENERA OF THE CLOSTRIDIACEAE 7 Genus 14. Meyerillus nov. gen. Tj^ie species sadowa nov. sp. to be described in a future paper. BB. Produce less gas from carbohydrates. Heavy Gram-positive or Gram-negative rods that form oval spores that are usually subterminal. Genus 15. Novillus nov. gen. Type spe- cies maligni (Bacillus oedematis maligni II. Novy) as described by Novy. Artificial Key to the Genera of the Putrificoideae A. Produce an alkaline reaction in meat medium but do not grossly disintegrate the particles of meat. B. Minute Gram-positive rods which form few spores. Clot milk. Weak tissue invaders, or non-pathogenic. Genus 16. Seguinillvs nov. gen. Type species aerofoetidus (Bacillus aerofoelidus Weinberg and Sdguin) as described by Weinberg and Sdguin (p. 161). BB. Gram-positive rods with oval subterminal or median spores that do not greatly distend the sides of the bacilli. Do not ferment milk. Typically pathogenic tissue-invaders. Genus 17. Reglilhis nov. gen. Type spe- cies progrediens nov. sp. to be described in a future paper. AA. Attack proteins somewhat more energetically than the above and produce a terra cotta coloration of the meat particles and soften and partially disintegrate them. (The terra-cotta color i.s not to be confused with the pink color produced in meat medium by the acid from sugar fermen- tation.) These organisms continue to multiply in meat medium at a moderate rate for months. B. Show no blackening of the meat-particles on prolonged incubation.* Gram-negative or weaklj' Gram-positive rods with oval subterminal or median spores that distend the bacilli. Genus 18. Robertsonillus nov. gen. Type species primus (Bacillus I. Hempl) as described by Hempl. BB. Show after approximately two weeks' incubation a blackening of some of the meat-particles. Gram-negative rods with spherical end-spores. Typically produce a neuro-toxin. Genus 19. N icollaierillus nov. gen. Type species tetani (Bacillus ietani NicoUaier) to be described in a future paper. * "Incubation" refers to anaerobic incubation at 37°C except in the case of deep agar tubes which are incubated in air at 37°C. Meat medium is pH 7.2 to start with. 8 HILDA HEMPL HELLER AAA. Highly proteolytic on meat medium for a short period. Produce a gray coloration or a slight blackening of the medium. Sporulate at an early stage and after three days' incubation vegetate very slowly indeed without so digesting the meat particles that they greatly decrease in size. B. Thick rods with oval or oblong spores which are usually central and do not greatly distend the bacillus. C. Resistant saprophytes. Occasionally invade tissue in company with other organisms or alone in debilitated individuals. Genus 20. Mariellillus nov. gen. Type species bifermentans (Bacillus hifermen- tans sporogenes Tissier and Martelly) as described by Tissier and Martelly. CC. Delicate parasites. Gram-negative tendency. Genus 21. Recordillus nov. gen. Type species fragilis nov. sp. AAAA. Organisms highly proteolytic, producing on three days' incubation in meat medium partial destruction of thrf meat particles which con- tinues on further incubation till the meat particles have greatly diminished in bulk. Coloration of meat usually dark brown or terra- cotta; blackening may or may not take place. B. Slender rods with terminal oval spores. C. Split sugars. Genus 22. Tissierilhis nov. gen. Type species paraputrificus (Bacillus para- puirificus defined by Bienstock 1906) as described by Mcintosh under the name of Bacillus putrificus (p. 39). CC. Do not split sugars. Genus 23. Putrificus nov. gen. Type spe- cies Bienstocki (Bacillus putrificus Bien- stock) as defined by Bienstock (1906)). BB. Heavy rods with subterminal or median oval spores. C. Do not produce balls of amino-acid crystals to a striking degree. D. Not so highly proteolytic as organisms of following groups. Meat particles not much decreased in size, spores often larger, and more nearlj' spherical than in following genera. Isolated colonies in deep agar usually, but not always, large, smooth lenticular or modified lenticular structures that do not become woolly. Typically produce a power- ful neuro-toxin. Genus 24. Ermengemillus, nov. gen. Type species bohdtJius (Bacillus bolulinus van Ermengem) as described by Meyer and co-workers in a future paper. CERTAIN GENERA OF THE CLOSTRIDIACEAE 9 DD. Highly proteolytic, blacken meat readily, produce penctratinK foul odor. Form in deep agar large w( oily or spherical colonies or lenticular colonies that show a tendency to become woolly. Are not pathogenic in pure culture. Genus 25. Metchnikovillus nov. gen. Type species sporogenes (Bacillus sporogcnes Metchnikoff) as described by the Com- mittee as Metchnikofl's race A (p. 36). CC. Produce balls of amino-acid crystals in meat after a com- paratively short period of incubation. Violently proteolytic, may produce an exotoxin, and invade tissue, which is vigorously digested. Genus 26. Weinbergillus nov. gen. Type species hislolyiicus {Bacillus hisiolyticvs Weinberg and Sdguin) as described by Henry, DEFINITIONS OF CERTAIN GENERA OF THE CLOSTRIDIOIDEAE' Genus 1. Clostridium Prazmowski 1880, emmend. Heller. Clostridioideae that do not liquefy gelatin. Most species cause stormy fermentation of milk. Highly saccharolytic, many even splitting pectins. Produce considerable amounts of butyric acid as a split-product of carbohydrate fermentation. May fix nitrogen. Readily derive their nitrogen from inorganic nitrogen salts. Large rods which are frequently polymorphic and form large orgonts and oval spores. Frequently store up starch. Form in deep agar large lenticular or modified lenticu- lar colonies. Common destroyers of plant tissue (not cellulose) in soil. Used in the retting of flax to split pectins. Type species, C. amylobader van Tieghem as described by Win- ogradsky (1896). Probable synonyms: Amylohader Trecul, Clostridium butyricum Prazmowski, Amylohakter Gruber, Gran- ulobakter saccharo-hutyricum Beijerinck, B. saccharohutyricns von Klecki, Oranulohacillus saccharobidyricus mobilis nonliquefaciens Grassberger and Schattenfroh, B. amylobacter von Hibler, B. amy- lobacter Arthur IMeyer and Bredemann. Most of these are today incapable of accurate definition and several of them represent groups and not entities. For this reason the type description chosen is that of Winogradsky instead of that of Prazmowski. ' For definition see Jour Bact., 6, 550. 10 HILDA HEMPL HELLER Trecul (1865) gave the bacteria that contain starch the nanae of Amylobader. He (1867) declared the Amylobader to be heterogenetic, formed of minute particles that organize them- selves into baciUi in decaying plant tissue. Van Tieghem (1877) named Bacillus amylobader bacilli which contained amorphous starch during their growth stage. He believed such organisms to be the agents of cellulose destruction. Prazmowski (1880) de- scribed and figured Clostridiwn butyricum, which though it was in impure culture, was evidently of the type described above. Winogradsky (with Tribes) (1896) first defined a type that can be considered a species ; he assigned no name to the organism. He declares that it does not split cellulose, but pectin. It fer- ments glucose, sucrose, lactose and starch in peptone media. This type of organism has been described by many workers. It has been most thoroughly discussed by Bredemann. Bacil- Itis amylobader A. M. et Bredemann probably includes all the large starch-storing Clostridia described above in the generic definition. In his investigation Bredemann used principally cytological criteria, and essential extensive chemical studies were not made. He regards all differences noted between his strains as fluctuating variations. It would seem illogical from the point of view of the general systematist, to assign merely specific rank to a group of organisms so widespread and of such abundant occurrence as are these soil anaerobes. Bredemann's investigation siinply indicates that he did not find means of distinguishing his strains, or perhaps that by his technique he isolated only a restricted group of the general type. Bredemann declares that the power of fixing nitrogen varies in these organ- isms and cannot be used as a specific character. Bredemann's critique of his cultures is apparently very weak. Thus on page 404 he claims to have changed a Welch bacillus into an amy- lobacter. He quotes seriously the fantastic conceptions of Grassberger and Schattenfroh with regard to the "denaturing" of anaerobic organisms, and his conception of systematic anae- robic work is that of Lehmann and Neumann. He regards as a variant of B. amylobader what appear to be coccus forms contaminating his cultures. CERTAIN GENERA OP THE CLOSTRIDIA CEAE 11 The chemical activities of this type of anaerobe with regard to end-products are described in some detail by Bredemann and by Grassberger and Schattenfroh. The latter find that amy- lase is usually present, sucrase very rarely so. Clostridium Pastorianum is the name given to the anaerobic nitrogen-fixing bacillus discovered by Winogradsky (1896 and 1902). This author distinguishes it from other Clostridia known to him by the fact that the sporangia only partially disappear from about the spores, forming what he terms a "spore capsule," and by the fact that its fermentative ability is less than is that of most soil Clostridia. In peptone media it splits glucose, sucrose, lae\'u- lose, inulin, galactose and dextrin, but not lactose, arabinose, starch, rubber, mannitol, dulcitol, glycerol or calcium lactate. Presumably stonny fermentation of milk does not then take place. Obviously the production of stormy fermentation of milk, depending on the splitting of ore sugar, is not to be regarded as a generic character. Grassberger and Schattenfroh found certain strains of their organism which did not attack milk with energy. Clostridium Pastorianum is sufficiently well dif- ferentiated by Winogradsky from the ordinary amylobacter or hutyricum type to warrant its separation from that type as a separate species. Bredemann regards the "spore capsule" formation described by Winogradsky as a frequent anaerobe character. I have never seen an anaerobe strain which pro- duced the remarkable "spore capsules" figured by Winogradsky. This author isolated C. Pastorianum only a few times out of many samples of earth, and he was familiar with the type usually termed C. hutyricum or B. amylobacter. Gruber distinguished two types of sporulating anaerobic granulose-storing butyric acid bacilU, of which the first is the most like the usual conception of C. hutyricum. Beijerinck differentiated his granulobacilli into an anaerobic and an aerobic form: Grassberger and Schattenfroh were unable to confirm this work. Choukevitch (1911) distinguished three types of amylobacter: one fermented glucose, lactose, starch and hemicellulose; one glucose and lactose only; and a third rarely stored up starch 12 HILDA HEMPL HELLER and had little fermentative power. He (1913) identified as B. amylobader strains which ferment cellulose. Pringsheim described C. americanum, which fixes nitrogen less energetically than C. Pastorianum and ferments the same sugars as that organism, besides mannitol, glycerol and lactose. It will grow in open flasks; Bredemann terms it an anaerobe. Douglas, Fleming and Colebrook have described under the name B. butyricus a medium sized bacillus that should be assigned to this genus or to MuUifermentans. Mcintosh de- scribes what may be the same strain. Prazmowski described under the name Clostridium polymyxa an aerobic organism. The generic name has since been used occasionally for aerobes. It must in future be restricted to the group of anaerobic organisms which most strongly resemble the first type described by Prazmowski, C. butyricum. There apparently remains abundant critical chemical and systematic work to be done in the study of this important genus, and such work should be performed with improved technique and with cultures whose purity will stand criticism. Genus 2. Omelianskillus nov. gen. Clostridioideae that do not grow well in media containing much protein, and may derive all their nitrogen from mineral salts. Split cellulose or hemicellulose. Do not contain starch. Long slender bacilU with spherical spores. Colonies may be produced on potato slants; they are minute, yellowish and transparent. Agents of plant putrefaction, found everywhere. Type species 0. hydrogenicus (the ferment of cellulose which produces hydrogen, of Omeliansky) as described by OmeUansky (1895 and 1904 b). Characters of genus. Probably several species were studied by Omeliansky. In his later work he ad- mits that the cultures studied by him were not pure. Another type, 0. mcthanicus, similar to the above, is said to be the agent of methane formation in cellulose fermentation. This species may be assigned to the same genus. These organisms were isolated by growing them in a medium free of organic nitrogen. There may be other genera of anaerobic CERTAIN GENERA OF THE CLOSTRIDIACEAE 13 cellulose splitters which require some organic nitrogen for their metabolism. Choukevitch (1911) describes under the name B. gazogenes an organism of active growth habit and of very strong fermenta- tive powers, that splits starch and homicellulose. Its morphol- ogy is sunilar to that of Omeliansky's organisms and it may be included in this genus. Choukevitch (1913) considers his type 1 of B. Rodella III as similar to 0. methanicus and type II as similar to 0. hydro- genicus, but his reasons for so doing are not very sound. These two organisms do not split cellulose. Genus 3. Macintoshillus nov. gen. Clostridioideae that do not liquefy gelatin. They produce acid and gas and no putrefaction in meat media. They do not readily attack milk and thej^ ferment few or no sugars. Gram- negative rods with parallel sides and with terminal spherical spores. Colonies in deep agar are small and irregular but not woolly. Frequently found in wounds. Apparently incapable of invading tissue. Type species tetanomorphus (pseudotetanus bacillus, Mcintosh (p. 32), Bacillus tetanomorphus Committee (p. 41)), as described by the Coimiiittee. Glucose and maltose are fermented. Bacillus tetanoides A of Adamson (1919) is to be assigned to this group. Acid and no gas is produced from glucose and maltose by the majority of Adamson's strains, while one strain showed no fermentative abihty. The former type is probably identical with ^Iclntosh's organism. Choukevitch (1913) describes as type 1 of the bacillus known as Rodella III a slender and highly saccharolytic organism with spherical spores, resembling morphologically Omeliansky's cel- lulose fermenters. It does not ferment cellulose nor liquefy gelatin and may temporarily be placed here in spite of the fact that it is said to produce hydrogen sulfide. If the organism does not Uquefy gelatin it is unlikely that it produces any considerable amount of hydrogen sulfide from protein when in pure culture. This organism is a coimnon intestinal sapro- 14 HILDA HEMPL HELLER phyte of cattle and sheep. A heavier bacillus, type II, is equally difficult to place. It also produces hydrogen sulfide and is highly saccharolytic. It may be that another genus should be created for highly saccharolytic spherical end-sporing organisms that grow on ordinary media. Genus 4. Douglasillus nov. gen. Clostridioideae that do not liquefy gelatin. They produce little gas in meat medium. They may clot milk. Gram-nega- tive bacilli which are frequently fusiform and may show peculiar involution forms. Spherical spores are formed in the bacilli; the rods are usually widened by the spores so that their sides are not parallel. Young spores may not be truly spherical. Occasionally found in wounds. Probably incapable of invading tissue. Tj^e species sphenoides {Bacillus s-plienoides Douglas, Fleming, and Colebrook), as described by the Committee (p. 43) : the type which ferments glucose, maltose, galactose, lactose, salicin, mannitol, sucrose, dextrin and starch. The Committee states that the fermentation reactions of this group are variable. Bacillus E of Adamson (1919) is ap- parently closely related to D. sphenoides. It produces peculiar involution forms and though it ferments several sugars (glucose, maltose, lactose and mannitol) it produces very little gas from them, and has not strong fermentative powers. The pointed rod named Coccobacillus proeaciitus by Tissier may be included in this genus. It does not ferment lactose or sucrose but does attack glucose. These organisms are not to be confused with the nonsporula- ting anaerobic fusiform bacilU which show a spotted staining reaction and invade tissue, but grow poorly on artificial media. These latter may be trained to an aerobic habit. They have been placed in the genus Fusiformis in the family Mycobacteri- aceae, order Actinomycetales by the Committee on Characteriza- tion and Classification of Bacterial Types, where they may well be left because of the fact that they are highly adapted parasites. CERTAIN GENERA OF THE CLOSTRIDIACEAE 15 Genus 5. Henrillus nov. gen. Clostridioideae that do not liquefy gelatin. They produce acid and gas in meat medium. They clot milk readily and attack many sugars, producing much acid. Gram-negative or gram- positive slender rods with terminal oval spores. Their colonies in deep agar are large, lenticular and opaque. Very common in soil, often found in wounds, do not invade tissue. Type species H. tcrlius {Bacillus tertius Henry) as described by Henry. Henry believes that the conception B. tertius applies to a group and not to a species. His nine strains which may be taken as a type split the monoses, bioses, mannose, xylose, starch, dextrin, glycogen, salicin, amygdalin and mannitol. This type was described by von Hibler with the number IX. Flemmg (Bac. Y), Rodella (1902) (Bac. Ill), Robertson (1916 a), Mcintosh, Adamson (1919) and the Committee and Weinberg and S^guin also describe it. Choukevitch (1913) identifies spherical sporulating organisms with Rodella III. The original Rodella III did not clot milk and should perhaps be associated with Flemingillus. Genus 6. Flerningillus nov. gen. Clostridioideae that do not hquefy gelatin. They produce little gas or acid in meat medium. They do not grow very abundantly in milk or change it. They do not show any marked tendency to split sugars. Slender Gram-negative rods with oval terminal spores. Colonies in deep agar lenticular, may show an areola of fine radiations. Frequently found in wounds, not pathogenic for guinea-pigs. Type species F. cochlearius (Bacillus cochkarius Douglas, Fleming and Colebrook) as described by the Committee. This type is highly motile and it split none of the carbohydrates that it was grown in. It was described as bacillus III type C by Mcintosh. Bacillus C of Adamson (1919, p. 380) should be referred to this genus. It is slightly motile. Mcintosh's Bacillus III A was also considered by Adamson to belong in such a group as this. It splits glucose and maltose only. We may, until the non- 16 HILDA HEMPL HELLER proteolytic slender end-sporing rods have been carefully investi- gated, include strains that spUt a few easily fermentable sugars and do not produce much acid in genus Flemingillus, and those that split many sugars and tolerate much acid in the genus Henrillus. Choukevitch describes a number of organisms possessing a similar lack of fermentative ability to Flemingillus. Possibly his B. irregularis belongs here; Sireptobacillus anaerobicus-magnus does not resemble this t3TJe morphologically, but does chemically. B. ventriosus of Tissier and B. gracilis-pittidus of Tissier and Martelly resemble this type. Coccobacillus oviformis and B. capillosus of Tissier resemble it in fermentative powers but not in morphology. Genus 7. Vallorillus nov. gen. Clostridioideae that do not Uquefy gelatin. They produce gas and acid in meat medium but no digestion. They clot milk slowly and attack various sugars. Rather slender Gram- positive rods with little or no tendency to form spores. Form lenticular colonies, "coeurs jaunes," in deep agar. May invade tissue, producing oedema and gas. Pathogenicity transitorj'. Type species V. fallax {Bacillus fallax Weinberg and Seguin) as described by the Committee (p. 27), the type which ferments glucose, laevulose and maltose. Henry regards this organism as capable of fermenting manj' sugars and starch. Choukevitch (1911) describes under the name B. bifurcalus gazogenes a large branching organism which has fermentative reactions similar to Vallorillus, but can hardly be included in the genus on account of its unusual morphology. Genus 8. Multifermentans nov. gen. Clostridioideae that do not liquefy gelatin. Produce gas and acid in meat medium. Clot milk readih^, without stormy fermen- tation. Rather small Gram-positive rods with oval central or subterminal spores. Activelj' saccharolytic. Found occasion- ally in wounds, not tissue invaders. CERTAIN GENERA OF THE CLOSTRrOIACEAE 17 Tjrpe species M. tenalbus {Bacillus mttltifermentans-lenalbus Stoddard) as described by Stoddard (1915 b). This organism, of which Dr. Stoddard was so kind as to send me a culture, does not fit into any other of the genera here defined. Stoddard's organism ferments glycerol, maltose, lactose, raffinose, glucose, sucrose, inulin, and salicin, JNIannitol and dulcitol are not fermented. Adamson describes under the name B. bittyricus a "small or medium-sized" bacillus that ferments glucose, lactose, maltose and sucrose but not mannitol and starch. A butyric acid odor is produced. Perhaps it should be assigned to this group. Genus 9. Hiblerillus nov. gen. Clostridioideae that do not liquefy gelatin. They do not clot milk. Large gram-positive rods which form more or less re- luctantly long ellipsoid spores ; they may form orgonts. Colonies in deep agar, small and lenticular or with fine radiations. May be pathogenic for the guinea-pig, producing oedema and gas, or paralysis. Type species H. sextus {Bacillus VI of von Hibler), as described by von Hibler (resiune by Weinberg and S(5guin, p. 202). Von Hibler describes another species which he terms VII. These organisms have much in common and we are probably justified in including them in one genus as Hiblerillus sextus and Hiblerillus septimus. The latter resembles an organism described by Tizzoni and Cattani, according to von Hibler. There are probably many organisms in soil which are pathogenic for rab- bits and guinea-pigs when given certain conditions favorable to invasion, which rarely invade under natural conditions or which on account of shy growth habit are missed when they invade in the company of other organisms. To this genus may be assigned two organisms isolated from the intestine of the horse and described by Choukevitch (1911); they possess similar fermentative ability: the production of acid and no gas in glucose agar. A non-pathogenic one, Strep- tobacillus anaerohicus-rectus may be termed H. rectus; another, H. megalosporus , produced a fibrino-purulent peritonitis in a guinea-pig. 18 HILDA HEMPL HELLER Genus 10. Welchillus nov. gen. Clostridioideae that liquefy gelatin but do not produce hydro- gen sulfid demonstrable by lead-acetate-paper test in blood- broth. They produce much acid and gas on meat medium, but they do not digest it, nor do they digest casein, coagulated serum, or eggvvhite. They produce stormy fermentation of milk and attack many sugars vigorously. Their multiplica- tion is extraordinarily rapid ; they are killed by their own growth products in acid media, in which they fail to sporulate. Deeply Gram-positive non-flageUate rods with square ends. They produce oval subterminal or median spores in alkaUne media, or in media free of fermentable carbohydrate; these spores do not bulge the sides of the baciUi. When growing rapidly the rods are very short, and resemble closely no other type here listed except Stoddardillus. When growing slowly the rods are less abundant and longer, and may be mistaken for MarteU lillus and similar organisms. Typically intestinal saprophytes ; ubiquitous. JNIany strains produce toxin and invade tissue, forming gas, and causing the formation of oedema, and in many cases causing the disintegration of muscle and of connective tissue without the production of a foul odor. This disintegra- tion occurs only in vivo, and is probably due to the enzymes of the host tissue. Welchillus are the most frequent anaerobic invaders noted on the hospital autopsy table: the organisms are present in human intestines. Causative agents of a probably greatly overestunated percentage of gas gangrene cases follow- ing war wounds. Comparatively rare as animal invaders. Type species W. aerogenes (Bacillus aerogenes-capsulatus Welch and NuttaU) type I as defined by Sinionds (type TV of Esty). Ferments, besides other carbohydrates, inuhn and glycerol. Usually pathogenic for guinea-pigs. Synonyms. B. phlegmones-em-physematoseae Fraenkel, B. perfringens Veillon and Zuber, 'Bacille de reumatisme aigue' of Achalme, Butyribacillus immobilis-liquefaciens Grassberger and Schattenfroh. Descriptions and discussions, most of them with large bibliographies, are to be found under the names of the following authors: von Hibler, Simonds, Robertson, Wein- CERTAIN GENERA OF THE CLOSTRIDIACEAE 19 berg and Scl^guin, Henry, jNIcIntosh, Adamson, The Committee, Jablons, Esty. Simonds divided the group into four sub-groups on the basis of the fermentation of inuUn and glycerol. Henry and Esty substantiated this finding. The latter finds his strains divisible into two sub-groups on the basis of the sensitiveness of the spores to heat. The division thus made does not co- incide with those secured by means of sugar fermentation. !Mc- Intosh admits a species which sporulates on ordinary media. The Committee, on which !^fcIntosh later served, do not men- tion such a type. The chemical behavior of Welchillus has most recently been studied by Wolf (1919) and Wolf and Harris (1917, a, b, and c). Agglutinins are extremely difficult to pro- duce with these organisms, and are found to agglutinate only homologous strains. Werner succeeded in agglutinating one non-homologous strain. Robertson (1916 b) failed to im- munize guinea-pigs with bacterial protein. Toxins, according to the work of Bull and Pritchett, are all neutralized by the same antitoxin. Esty immunized guinea-pigs with young whole cultures. Genus 11. Stoddardillus nov. gen. Clostridioideae of energetic growth habit that liquefy gelatin, but do not produce H2S demonstrable by a lead-acetate-paper test in blood broth. Produce abundant gas but little acid in meat medium. Grow very shyly or not at all in milk. Attack a few sugars, but do not produce much acid. Short chunky Gram-positive rods which do not form spores. Colonies in deep agar large, lenticular and opaque. Not easily distinguished from Welchillus. jVIay invade tissue, causing considerable destruc- tion of muscle but no foul odor. One strain, reported from a case of human gas gangrene. Type species S. egens (Bacillus egens Stoddard) as described by Stoddard (1919 a). Splits glucose, laevulose, mannose, mal- tose, dextrin, glycogen, inosite and glycerol. Does not sporulate on six days' incubation in inspissated serum. Such organisms as this are probably not nearly so rare as reports would indicate because they do not sporulate and are 20 HILDA HEMPJU HELLER killed in most anaerobic isolation procedures. Because of their close resemblance to Welchillus and because of the ubiq- uity of organisms of that genus their detection is rendered still more difficult. It may be that later workers will prefer to include this organ- ism in the genus Welchillus. The action on milk is dependent on the fermentation of lactose, and sugar fermentations are not to be regarded as of generic significance. I place it in a genus by itself because spore formation has not been demonstrated for this organism. Gentjs 12. Rivoltillus nov. gen. Clostridtoideae possessing moderately strong saccharoljrtic powers. Liquef}'- gelatin but do not produce US demonstrable by a lead-acetate-paper test in blood broth. Produce in meat medium gas and a pink coloration which does not rapidly fade. Clot milk. Do not liquefy serum or egg or disintegrate meat particles. Gram-positive rods, usually short, with median, sub-terminal, or terminal spores, which usually bulge the sides of the bacillus. Sporangia not often much larger than vegeta- tive rods. In tissue the sporangia may be une\'en in their stain- ing reactions, "granulose" being present; orgonts are long, frequently with parallel sides. Usually form chains on the liver of animals. Colonies in deep agar, though they may start as lenticular structures, consist later of a dense center and a wide loose woolly periphery. They vary in size, etc., according to species. Typically highly pathogenic tissue invaders that produce haemolysis and gas in the animal body. Pathogenic for a wide range of species. Type species R. vibrion (the vibrion septique of Pasteur), as defined in a future paper. Robertson (1920) has divided the group into four sub-groups on the basis of the agglutination reaction. Probably the group of anaerobes whose nature is most fre- quently discussed. The morphological resemblance of individ- uals and of colonies to organisms of the sporogenes type led to the frequent description of mixed cultures of the two tjrpes, CERTAIN GENERA OF THE CLOSTRIDIACEAE 21 usually under the name B. oedematis-maligni. The literature on this group is too extensive to quote. See Ghon and Sachs, Meyer, Weinberg and S6guin, Robertson, the Committee, V\'olf (1918). A review of the annual infections has been made by Heller (1920). Under the name Bacillus tumefaciens (not to be confused with the plant pathogen Bacterium tumefaciens). Wilson describes an organism similar to those of this genus. His description does not convince the reader as to the purity of the culture studied : a mixture of vibrion septique and oedemar tiens-like organisms would behave as did ^^'ilson's bacillus. Genus 13. Arloingillus nov. gen. Clostridioideae that attack sugars with considerable energy but have a somewhat restricted action on proteins. Liquefy gelatin but do not produce HoS demonstrable by a lead-acetate- paper test in blood broth. In meat medium produce gas and a pink coloration that soon fades. Autoagglutinate readily. Clot milk if blood is present. Do not digest serum or egg. Vege- tative forms are small gram-negati^'e rods with even staining; forms about to sporulate are uneven in staining reaction, often far larger than vegetative rods, citron or spindle shaped; or- gonts (see Heller) , show marked tendency to store up granulose. Spores oval, may vary greatly in length. Bacilli do not form long chains on the liver of animals. Colonies in deep agar lenticular, sometimes showing concentric formation, or compound lenticular. Colonies vary considerably according to species. Typically toxic tissue invaders which produce marked haemolysis. Patho- genic for guinea-pigs, cattle and sheep. Type species A. Chauvoei (Bacterium Chauvoei Arloing, Cor- nevin, and Thomas) as described in a future paper. This genus contains several species which will be discussed. These organisms show some similarity to those of the genus Clostridiiun. The B. enteritidis-sporogenes Klein, as described by von Hibler (1908) should probably be included in the genus. The only character which is markedlj'- different from that of the genus is the energetic fermentation of milk shown by von Hibler's bacillus IV. 22 HILDA HEMPL HELLER Gentjs 14. Meyerillus nov. gen. Clostridioideae that produce HaS on blood broth and liquefy gelatin. In meat medium they produce gas but little or no pink coloration; they show no marked proteolytic action. Do not readily attack milk. Large Gram-positive rods with little or no inclination to form spores. Attack a few sugars. Colonies in deep agar large, opaque and lenticular. Typically tissue invaders of marked power, attacking and digesting in vivo the connective tissue more than the muscle. Type species M. sadowa nov. sp. To be described in a future paper. This organism was at first taken for B. Welchii. It does not sporulate. It is one of the four guinea-pig invaders isolated from a case of human gas gangrene. I am inclined to place the sporulating bacillus L of Adamson (1919) in this genus but have not handled that organism. Perhaps it is premature to decide whether sporulation may be used as a generic character. Adamson finds bacillus L very slightly proteolytic on milk and not so on other media. M. sadowa does not grow on milk. Genus 15. Novillus nov. gen. Clostridioideae that in blood broth produce H2S demonstrable by lead-acetate-paper test; they liquefy gelatin. Produce gas and on continued incubation produce a pink coloration in meat, but this color rapidly fades. Autoagglutinate with extreme readiness. Slowly attack milk and a few carbohydrates, but do not form much acid. Heavy rods of apparently shy growth habit on most media; form a few oval spores that may or may not bulge the sides of the baciUi. Colonies usually large and opaque, may form slender projections or even long fine woolly filaments. Frequently yellowish. Typically toxic tissue invaders which frequently cause the formation of a thick gelatinous oedema that does not lose its gelatinous consistency on section. ]May also produce gas and a black haemorrhagic condition of the muscle if rapid invasion by large numbers of bacilli takes place. One strain causes considerable tissue destruction. Pathogenic for guinea-pigs, mice, man, hogs, horses, cattle. Type species A'', maligni {Bacillus oedematis-maligni II, Novy) as described by Novy. CERTAIN GENERA OF THE CLOSTRIDIACEAE 23 The lesions typically produced by organisms of this group were first described by Koch. Gohn and ^Mucha give an elaborate description of an organism which belongs to this genus. Von Hibler studied four strains of "Novy's bacillus." Descriptions of organisms of this type have been given by Kerry, by Rivas and by Diedrichs. Weinberg and Scguin described B. oed-erna- iie?is which is different from A\ maligm but is to be included in this genus. B. bellonensis Saqu^p^e also belongs in this genus, as may also the organism called by Adamson bacillus S (1919, p. 373). I have found a tj^pe that was rapidly fatal to horses, and another different somewhat from all the rest in a septic wound, and hope to define these species accurately in another paper. Wolf (1920) has studied the biochemistry of an organism of this group. DEFINITIONS OF CERTAIN GENERA OF THE PUTRIFICOIDEAE* Genus 16. Seguinillus nov. gen. Putrificoideae that do not exert a marked action on protein. Soften and may blacken meat. jVIilk clotted, may later be digested. Attack a few sugars. Very reluctant to form spores. Minute Gram-positive or Gram-negative rods, uniform in thick- ness but not in length. Spores oval, sub-terminal. Deep colon- ies lenticular or modified lenticular. Occasional tissue invaders which produce a proteolytic gangrene or phlegmon. Type species S. aerofoetidus {Bacillus aerofoetidus Weinberg and Seguin) as described by Weinberg and Scguin (p. 161). The organism forms oval subterminal spores. These may well be called "borderline" organisms. Henry places them in his " saccharolytic group" on account of their sugar-splitting proclivities. A strain given by me by Weinberg does not grossly attack meat particles. Genus 17. Reglillus nov. gen. Putrificoideae that show to the eye little proteolytic action on meat. They may or may not blacken it slightly, but the * For definition see Jour. Bact., 6, 550. 24 HILDA HEMPL HELLER meat particles do not dimmish in size. Sugars are attacked. Milk is digested. Gram-positive rods which readily form oval subterminal or median spores. Colonies minute, opaque, com- plex, yellow. Typically toxic tissue invaders which produce in the guinea-pig a clear white oedema that does not rapidly lose its ©edematous nature on section. Type species R. progrediens nov. sp. To be described in a future paper. Two species, one found in a case of human gas gangrene (see Barney and Heller), the other in a sheep. Genus 18. Rohertsonillus nov. gen. Putrificoideae that produce on meat medium a little gas and a terra cotta coloration, multiplying slowly for a long period. A black pigment is absent. Sugars not fermented. Weakly Gram-positive or Gram-negative rods with oval spores, usually subterminal, that somewhat distend the bacillus. Two spores often occur in one rod. In old cultures enormous snakey Gram- negative rods are produced. One species, found twice in wounds. Non-pathogenic for guinea-pigs in pure culture. Type species R. primus (Bacillus I, Hempl) as described by Hempl. Descriptions of bacilU similar to this have not been noted. Proteolytic organisms producing a terra cotta colora- tion in meat medium are not uncommon, however. Genus 19. Nicolaierillus nov. gen. Putrificoideae that in meat medium produce gas and various color changes: yellowish, pink, grey or mauve, depending on the medium; the particles of meat are gradually suffused with a black pigment, and bleach at the top. The meat is softened but the particles do not greatly dmiinish in size. Do not attack sugars. Gram-negative (weak methyl violet) rods that form terminal spherical spores. Colonies in deep agar diverse. Com- mon in soil, found in horse feces, maj'^ multiply in wounds, but do not normall}'^ invade tissue. Produce a characteristic neuro-toxin. Type species N. tetani (Bacillus tetani Nicolaier), to be de- scribed in a future paper. Tulloch flOl? and 1919) has divided the group into four types on the basis of the agglutination reaction. Adamson (1920) CERTAIN GENERA OF THE CLOSTRIDIACEAE 25 describes the cultural behavior of five strains of B. lelani. Two papers on the behavior of my cultures arc forthcoming. This is not the genus Plectridium of Fischer. Plectridiuin included some butj^ic acid bacteria, the tetanus bacillus, a putrefactive organism, Plectridium putrificum, and other genera not yet known; and it embraces, according to my scheme of classification, elements altogether incoherent. Were the de- scriptions fuller or were I better acquainted with the proteolytic group, two more genera of proteolytic spherical end-sporers might be suggested: Those that blacken meat readily, e.g., B. cadaveris-sporogenes, Klein; and those which do not blacken it at all, e.g., B. tctanoides B of Adamson. Genus 20. Markllillus nov. gen. Hardy Putrijicoideae that in meat medium multiply rapidly at an early stage of incubation, producing a greyish coloration and later a blackish deposit on the meat particles, and after three days' incubation cease to multiply actively. Sporulate early in the de\'elopment of a culture, later cease to do so but vegetate very slowly. Produce very little gas in meat medium. Digest milk. Attack a few sugars. Heavy deeply gram-positive rods, may vary greatly in size. Spores usually cocoon-shaped, usually median or sub-terminal, do not greatly bulge the sides of the bacillus. Colonies in deep agar lenticular, irregular, or stellate. Common putrefactive organisms that readily invade tissue in company with other organisms, producing a greenish proteolytic gangrene. I have found them in the heart's blood and organs of a woman dying of pernicious vomiting and uraemic poisoning and apparently the only invader in a mouse dying from an otherwise unknown cause. The Committee find B. bifermentans in acute cases of gas gangrene. Weinberg and Seguin state that such an organism invades guineor-pigs in company with B. perfringens. It is possible that Martellillus bacilli pro- duce metabolites poisonous to themselves and to animals and that cause early sporulation. They frequently cease active multiplication when the culture has a reaction near pH 7.0. Type species M. bifermentans {Bacillus bifermentans-sporo- geiies Tissier and Martelly) as defined by Tissier and Martelly. 26 HILDA HEMPL HELLER Hempl has described a similar organism not identical with that of Tissier and Martelly (Organism II, which may be renamed Martellillus proteolyticus) . Bacillus II of Chouke\'itch probably belongs in this genus, as does von Hibler's bacillus XV. B. sporogenes B, as described by Choukevitch also belongs here. A rather shyly growing species sent me by Major Nichols which was named B. hellonensis Saqu^pee, was apparently pure, non- pathogenic, and referable to this genus. It was not the organism described by Saquepee as B. hellonensis, nor did it resemble other strains sent me by Saquep6e. Adamson describes two species of this type. One under the name of B. hifermentans- sporogenes, which is non-motile and does not spUt sugars, the other under the name of "Central spore bacillus" which he identifies as Mcintosh's type XII, which is motile and splits glucose and maltose. The type species is, however, supposed to be saccharolytic. Mcintosh's type XII (parasporogenes) does not behave in my hands as does this latter type of Adamson's and I should not include them in the same genus. INIcIntosh's type XIII, a pathogenic proteolytic organism, should probabty be placed here. Some of the organisms here listed are prob- ably identical, but one is not warranted in so considering them without a direct comparison of strains. Study of many strains would probably necessitate the division of this genus into two, on the basis of carbohydrate fermentation. This is one of the important groups that most need syste- matic investigation. Genus 21. Recordillus nov. gen. Puirijicoideae that, though they sporulate, are exceedingly delicate and soon die in meat medium and other media. Their growth in meat medium resembles that of the organisms of genus Martellillus; they do not produce much gas, they color the meat particles a greyish color and form a blackish pigment. Gram-negative rods with central or sub-terminal cocoon-shaped spores. Parasitic forms which infect cattle in California and Nevada. I venture to place such organisms in a separate genus on account of their parasitic habit and on account of their CERTAIN GENERA OF THE CLOSTRIDIACEAE 27 delicacy. It maj' be that they are descended from organisms of the genus Martellillus. Type species R. fragilis nov. sp. Characters of genus. I regret that the strain kindly sent me by Dr. Records died soon after its arrival. As it had sporulated in meat medium this was surprising, but was consistent with Dr. Records' de- scription of the behavior of the organism. The strain was isolated from a liver infarct in a cow. Genus 22. Tissierillus nov. gen. Putrificoideae that attack sugars and clot milk. Slender Gram-negative or Gram-positive rods with oval terminal spores. Colonies in deep agar have radiate periphery and opaque cen- ter. Frequently intestinal saprophytes. Type species T. paraputrificus (Bacillus paraputrificns de- fined by Bienstock) as described by Mcintosh under the name Bacillus putrificus. Ferments glucose, maltose, lactose, su- crose and starch. References to organisms of this type are made by Passini, Moro, Metchuikoff and Kligler. Very likely some descriptions of this type were based on the behaidor of mixed cultures. Genus 23. Putrificus nov. gen. Putrificoideae that do not attack sugars. Slender Gram-nega- tive or Gram-positive rods with oval terminal spores. Colonies in deep agar have radiate periphery and opaque center. Putre- factive organisms found in soil and wounds. Type species P. Bienstocki (Bacillus putrificus Bienstock) as defined by Bienstock (1906). References: Bienstock (1884, 1901, 1906), Klein, Rodella (1905). The latter differentiated three types of what he termed B. putrificus. Tizzoni, Catani and Baquis described two organisms of this general type. B. poslumus of Wiircker is of this group. His B. putrificus is of sporogenes affinities. I possess a strain (lOR) isolated by Miss Robertson or myself from a wound, which is mildly proteolytic, does not split sugars, is Gram-negative, and on serum produces oval terminal spores. Its deep colonies are transparent, len- 28 HILDA HEMPL HELLER ticular, with a protruding fluff. Dr. Meyer tells me that he occasionally encounters such organisms, very slow in initial activity in pure culture, but more active in mixed culture, mildly proteolytic in their action on meat. The colony is not unlike that of B. botulinus. The Committee find that many so-called "Putrificus" strains consist of a non-proteolytic oval end-sporer contaminated with a sporogenes-like organism. In the recent German literature "Putrificus" refers usually to the sporogenes tj-pe. Metchni- koff's strains termed Putrificus do not resemble closely those described by Bienstock. Genus 24. Ermengemillus nov. gen. Putrificoideae that produce a yellowish coloration in meat medimn, and later blacken and digest it. They are more highly proteolytic than the organisms of the foregoing genera. They ferment various sugars. Gram-positive rods which form sub-terminal oval spores. Their colonies in deep agar are, when discrete, lenticular or kidney shaped, and may show tufted smooth or woolly polar projections and infrequently fine loose woolly radiations. Fairly common in soil, grow readily in vegetable material and on meat. Form a characteristic and powerful neuro-toxin. Type species E. botulinus {Bacillus botulinus ^•an Ermengem) as described by K. F. Meyer and co-workers in a future paper. Leuchs discovered that there were two types of toxin produced by different European strains. Burke found two types of toxin produced by different American strains. Probably a nxmiber of species may be distinguished by careful methods. Mcintosh and the Committee find that B. botulinus shows slight proteolytic power, and meat is stated to be a poor medium for the growth of the organism. INIy experience has been very different from theirs. Indubitably pure cultures of Ermenge- millus grow readily on meat medium (pH 7.2) and may show obvious putrefaction in a few daj-s when incubated anaerobically at 37°. But the meat particles are not rapidly diminished in size by the proteolytic action. The observers who did not note the CERTAIN GENERA OF THE CLOSTRIDIACEAE 29 proteolytic action of these organisms on meat must have used an acid meat medium, or a substratum poor in peptones and pep- tids, or must have incubated their cultures at room temperature. Von Hibler found that the organism blackened brain medium. Genus 25. Metchnikovillus nov. gen. Highly proteolytic PiUrificoideae that readilj^ blacken meat. They do not produce in it abundant amino-acid crystals, but digest meat, serum, egg and casein rapidly, forming more alkali or less fatty acid than do the organisms of the succeeding group. They split few sugars. Gram-positive or weakly Gram-positive rods, vegetative forms uniform and considerably smaller than sporangia. Sporulate readily in ordinary media, forming oval spores which are usually sub-terminal, though in some strains median spores predominate. ^Multiplication is exceedingly active, forty-eight hour colonies in deep agar are large and woolly. Fre- quently a few colonies are larger than the others but they do not give rise in a following generation to a preponderating number of large colonies. Ubiquitous. Common intestinal organisms, abundant in soil; very common in infected wounds. Not capable of invading in pure culture in moderate doses, may invade in company with other organisms or alone when given in large doses. Type species M. sporogenes {Bacillus sporogenes type A of Metchnikoff) as described by the Committee (p. 36). Klein described as B. enteritidis-sporogenes a mixed culture which contained a non-proteolytic organism, a pure culture of which was described by von Hibler as B. enteritidis-sporogenes Klein (von Hibler IV) . This tissue invading pathogen, thought by some to have been B. Welchii,\^ most nearly referable to the genus Arlo- ingillus. The strain was derived from patients with enteritis and it was apparently contaminated with a proteolytic organism of the genus Metchnikovillus. Metchnikoff described two types, A and B, of intestinal anaerobes which he thought were similar to the organism of Klein, and which he termed B. sporogenes. His descriptions permit of no identifications. Choukevitch, working in ^letchnikoff's laboratorj^ described B. sporogenes A and B more carefully. Type B should be referred to genus 30 HILDA HEMPL HELLER Martellillus. Weinberg made the identification of type A more exact, and the Committee discuss it at some length. They allow two species, B. sporogenes type A of Metchnikoff, and B. para- sporogenes Mcintosh, which are different in colony formation and serologically. Donaldson describes the "Reading" bacillus, to be assigned to this genus. Superficial acquaintance with many strains of proteolytic anaerobes leads me to suggest that Metchnikovillus may be defined as a genus of many species; Dr. K. F. Meyer is also of this opinion. Henry (p. 361) beUeves it likely that the conception "sporogenes" refers to a group of organisms. Barger and Dale, Weinberg and Seguin, M. E. Bullock and the Committee discuss the poisonous growth product of this type of organism: it is apparently a lower spUt-product than are true toxins. Some of these authors find that the culture filtrate of B. sporogenes increases the toxicity and invasive power of B. Welchii. Donaldson and Joyce placed the Reading bacillus in wounds to digest necrosed muscle and report no accidents due to its presence. Wolf (1919), Wolf and Telfer, and Wolf and Harris (1917, a and b) describe the chemical activities of these and related organisms. Because of their universal occurrence and active growth habits, organisms of this type frequently contaminate anaerobic cultures. Their chemical activities are thus described in con- junction with the pathogenic properties of the cultures in which they are active. The descriptions which tally more or less accurately with the sporogenes type are legion. Thus B. oede- matis maligni Koch of von Hibler was probably a mixture of an organism of this genus with a chain-forming vibrion sep- tique; while bacillus XI of von Hibler (1908) was probably such a mixture with a similar pathogen which formed chains somewhat reluctantly (PI. II, fig. 3). I am today able to discover only the sporogenes type in von Hibler's strain of bacillus XI. Whether bacillus XIX of Mcintosh should be included in the genus Metchnikovillus is difficult to state. It is apparently an active tissue invader and forms smooth lenticular colonies. CERTAIN GENERA OF THE CLOSTRIDIACEAE 31 When surface methods of isolation are followed this type of organism is very difficult to remove from cultures of other anae- robes. The Committee emphasize this point. I have had no more contaminations by sporogenes than bj^ aerobes and by various other anaerobes. The organism is common in nature and gains entrance to pure cultures occasionally. Genus 26. Wcinbergillus nov. gen. Highly proteolytic Putrificoideae that in meat medium cause the formation of balls of amino-acid crystals. They digest the meat particles till their bulk is greatly reduced and form large amounts of ammonia, amino-nitrogen and fatty acids. They digest the casein in milk rapidly. Attack few or no sugars. Produce little or no gas in agar media. Gram-positive or Gram- ,negative rods with sub-terminal oval spores. Colonies in deep agar small, delicate woolly structures. May invade living tissue in company with other organisms, or at times alone, pro- ducing a rapid and complete digestion of muscular and con- nective tissue structures. Type species TT^. histolyticus (Bacillus hisiolyticiis Weinberg and S^guin) as described by Henry (p. 370) . Weinberg and S^guin and the Committee allow considerable variation in characters for the strains termed by them Bacillus histolyticus. It is possible that B. sporogenes-parvus of Choukevitch may belong in this genus. Wolf and Harris (1918) have made a chemical study of what is probably the type strain — they re- ceived it from Henry who received it from Weinberg. REFERENCES AcRALME, P. 1891 Examen bact^riologique d'un cas de rhumatisme artieulaire aigu mort de rhumatisme c(Sr6bral. Compt. rend. Soc. Biol., 43, 651. AcBALME, P. 1902 Recherches sur quelques bacilles ana^robies et leur diffe- rentiation. Ann. de I'Inst Past., 16, 633. Adamson, R. S. 1919 On the cultural characters of certain anaerobic bacteria isolated from war wounds. Jour. Path, and Bacteriol., 1918-19, 22, 345. Adamson, R. S. 1920 On the cultivation and isolation of Bacillus tetani. Ibid., 23, 241. 32 HILDA HEMPL HELLER Abloing, Cornevin, and Thomas. 1887 Le charbon symptomatique du boeuf. Paris. Babnhy, E., and Heller, H. H. The investigation of a case of gas gangrene due to a mixed infection. Studies on pathogenic anaerobes, HI, see Heller. Bbntham, G. and Hooker, J. D. 1883 Genera Plantarum. London. BiKNSTOCK, B. 1884 TJeber die Bakterien der Faces. Zeitschr. f. klin. Med., 8,1. BiENSTOCK, B. 1899, 1901 Untersuchungen iiber die Aetiologie der Eiweiss- faulnis. Arch. f. Hyg., 36, 335; 39, 390. BiENSTOCK, B. 1906 Bacillus puirificus. Ann. de I'lnst. Past., 20, 407. Beijekinck. Quoted from Grassberger and Schattenfroh. BoTKiN, S. 1892 tJber einen Bacillus butyricus. Zeitschr. f. Hyg., 11, 421; Centralbl. f. Bakt., 11, 628. Bbkdemann, G. 1909 Bacillus amylobacter A. M. et Bredemann in morpho- logischer, physiologischer und systematischcr Beziehung. Centralbl. f. Bakteriol. II. Abt., 23, 385. Bull, C. G., and Pritchett, I. 1917 Identity of the toxins of different strains of Bacillus Welchii. Jour. Exp. Med., 26, 867. BtTRKE, G. S. 1919 Notes on Bacillus botulinus. Jour. Baeteriol., 4, 555. Choukbvitch, J. 1911 Etude de la fiore bacterienne du gros intestln du cheval. Ann. de I'Inst. Past., 25, 247. Chodkbvitch, J. 1913 Recherches sur la Acre microbienne du gros intestin des bovid^s et des moutons. Ann. de I'Inst. Past., 27, 246, 307. "The Committee," without other designation, refers throughout this paper to The Committee upon Anaerobic Bacteria and Infections, which served as a special investigation committee for the Medical Research Com- mittee, which sec. Donaldson, R. 1919 Character and properties of the "Reading" bacillus. Jour. Path, and Baeteriol., 22, 129. Donaldson, R., and Joyce, J. L. 1917 A method of wound treatment by the introduction of living cultures of a spore-bearing anaerobe. Lancet, 2, 445. Douglas, Fleming and Colebrook. Studies in wound infections. Medical Research Council. Special Report Series No. 57. London, 1920. VAN Ermengem 1897 Ueber einen neuen anaeroben Bacillus und seine Bezie- hungen zum Botulismus. Zeitschr. f. Hyg. u. Infektionskh., 26, 1. Ernst, P. 1893 Ueber einen gasbildenden Anaeroben im meneschlichen Korper und seine Beziehungen zur 'Schaumleber.' Arch. f. path. Anat. etc., 133, 308. ESTT, J. R. 1920 The biolog>- of Clostridium Welchii. Jour. Baeteriol., 5, 375. Fischer, A. 1903 Vorlesungen iiber Bakterien. Jena. Fleming. 1915 a Some notes on the bacteriology of gas gangrene. Lancet, 2, 376. Fleming 1915 b On the bacteriology of septic wounds. Ibid. 2, 638. Frabnkel, Eug. 1893 Uber die Aetiologie der Gasphlegmonen (Phlegmons emphysematosa). Med. Klin., 13, 13. CERTAIN GENERA OF THE CLOSTRIDIACEAE 33 Ghon, a., axd Sachs, M. 1903 Bcitriigc zur Kentnis dcr anaeroben Bakterien des Menschen. Zur Aetiologie des Gasbrandes. Centralbl. f. Bak- teriol., Abt. I, A., 34, 288; 609. 35, 665. 36, 1 ; 178. Ghon, a., and Mucha, V. 1905 Beitriige zur Kentnis der anaeroben Bakterien des Menschen. Zur Aetiologie der Peritonitis. Centralbl. f. Bak- teriol.,39, 497, 641. Grassberobr and Schattenfror 1902 1'ber Ruttersiiueregahrung. Zur Morphologie des beweglichcn Buttersiiurebacillus. Arch. f. Hyg., 42, 219. Grtiber, M. 1887 Eine Methode der Kultur anaerobischen Bakterien nebst Bemerkungen zur Morphologie der Buttersauregiihrung. Centralbl. f. Bakteriol., 1, 367. Gruber, M. 1905 Beitrag zur Identifizierung und Beschreibung von Clostri- dium polymyxa Prazmowski. Centralbl. f. Bakteriol. Abt. II, 14, 353. Harris, J. E. J. 1920 Contribution to the biochemistry of anaerobes. VIII. The biochemical comparison of microorganisms by quantitative methods. Jour. Path, and Bacterid., 23, 30. Heller, H. H. 1920 The etiology of acute gangrenous infections of animals. A discussion of blackleg, braxy, malignant oedema, and whale septi- caemia. Studies on pathogenic anaerobes I. Jour. Infect. Dis., 27, 385. Heller, H. H. 1921 a Principles concerning the isolation of anaerobes. Studies on pathogenic anaerobes II. Jour. Bacteriol., 6, 445. Studies on pathogenic anaerobes III, see Barney and Heller. Heller, H. H. 1921 b Suggestions concerning a rational basis for the classi- fication of the anaerobic bacteria. Studies on pathogenic anaerobes IV. Jour. Bacteriol., 6, 521. Hempl (Heller) H. 1919 Some proteolytic anaerobes isolated from septic wounds. Jour, of Hyg., 17, 13. Henrt, H. 1917 An investigation of the cultural reactions of certain anaerobes found in wounds. Jour. Path, and Bacteriol., 21, 344. VON HiBLEB, E. 1908 Untersuchungen uber pathogenen Anaeroben. Jena. von Hibler, E. 1909 Zur Kentnis der anaeroben Spaltpilze und deren Differ- entialdiaguose nebst einem Bestimmungsschliissel in 2 Tabellen. Innsbruck, Jena. Ber. d. Naturwissenschaftlichmedizinische Vereines in Innsbruck. 1908-09, 32. Jablons, B. 1920 Studies on pathogenic anaerobes. I. Biology of Bacillus Welchii. Jour. Lab. and Clin. Med., 374. Kerry 1894 Ueber einen neuen pathologischen anaeroben Bazillus. Oesterr. Zeitschr. f. wiss. Veteriniirkunde. 5. Quoted from Centralbl. f. Bak- teriol. Abt. I. 1890, 16, 372; and Fliigge, Die Mikroorganismen, 1896, Ed. 3, 2, 242. Klein, E. 1895 Ueber einen pathogenen anaeroben Darmbacillus. Bacillus enteritidis sporogenes. Centralbl. f. Bakteriol. Abt. I., 18, 737. Klein, E. 1897 Report on the morphology and biology of Bacillus enteritidis sporogenes. Centralbl. f. Bakteriol. Abt. I., 22, 113; 577. 23, 542. JOUBNAL OF BACTERIOLOGY, VOL. VII, NO. 1 34 HILDA HEMPL HELLER Kligler, I. J. 1915 A biochemical study and differentiation of oral bacteria with special reference to dental caries fll). Jour. Allied Dental Societies 10, 282. Koch, R. 1881 Zur Aetiologie des Milzbrandes. Mitt. a. d. kaiserl. Gesund- heitsamte., 1, 49. Lettchs, J. 1912 Beitriige zur Kentniss der Toxin und Antitoxin des Bacillus botulinus. Zeitschr. f. Hyg., 65, 55. McIntosh, J. 1917 The classification and study of the anaerobic bacteria of war wounds. Med. Res. Comm. Special Report Series No. 12, London. Medical Rese.\rch Committee. 1919 Reports of the Committee upon an- aerobic Bacteria and Infections. Report on the anaerobic infections of wounds and the bacteriological and serological problems arising therefrom. Special Report Series 39. London. (Excellent bibli- ography.) Metchnikoff, E. 1908 Etudes sur la fiore intestinale. Ann. de I'lnst. Past. 22, 929. Meyer, K. F. 1915 The etiology of "symptomatic anthrax" in swine, "specific gas phlegmon of hogs." Jour. Infect. Dis., 17, 458. Moro, E. 1905 a Morphologische und biologische Untersuchungen liber die Darmbakterien des Sauglings. I. Die Bakterienflora des normalen Frauenmilchstuhles. Jahrb. f. Kinderhlk., 61, 687. Moro, E. 1905 b Morphologische und Biologische Untersuchungen iiber die Darmbakterien des Sauglings. II. Die Veiteilung und die Schicksale der normalen Bakterien im Siiuglingsdarm. Ibid., 61, 870. NovT, F. G. 1894 Ein neuer Bacillus des malignen Oedems. Zeitschr. f. Hyg. u. Infektionskrankh., 17, 209. Omeliansky, W. L. 1895 Sur la fermentation de la cellulose. Compt. rend. Soc. Biol, 121, 653. Omeliansky, W. L. 1904 a Ueber die histologischen und chemischen Verander- ungen in den Flachsstengln unter dem Einfluss der Bakterien der Pektin- und Cellulosegarung. Centralbl. f. Bakteriol., 11, 561. Omeliansky, W. L. 1904 b Ueber die Trennung der Wasserstoff und Methan- garung der Cellulose. Ibid., 369, 703. Omeliansky, W. L. 1906 Ueber die Ausscheidung des Methans in der Xatur bei biologischen Processen. Ibid., 15, 673. Passini 1905 Ueber Giftstotfe in den Kulturen des Gasphlegmonebazillus. Wien. klin. Wchnschr., 18, 921. Pasteur, L. 1861 Animalcules infusoires vivant sans oxygene libra et deter- minant des fermentations. Compt. rend. Acad. Sciences, 52, 344. Prazmowski, a. 1880 Untersuchungen iiber die Entwicklungsgeschichte und Fermentwirkung einiger Bakterien-Arten. Leipz. Pbingsheim, H. 1906 Ueber ein Stickstoff assimilieren des Clostridium. Cen- tralbl. f. Bakteriol. , 14, 795. RivAS, D. 1902 Beitrag zur Anaerobenziichtung. Centralbl. f. Bakteriol., 32, 403. Robertson, Muriel. 1916 a Notes upon certain anaerobes isolated from wounds. Jour. Path, and Bacteriol., 20, 327. CERTAIN GENERA OF THE CLOSTRIDIACEAE 35 Robertson, Mubiel. 191C b Notes on the vaccination of guinea-pigs with B. pcrfringens. Lancet, ii, 516. lloBERTSON, Muriel 1918 Notes on vihrion septique. Brit. Med. Jour., 1, 5S3. Robertson, Muriel I9lX) Serological groupings of vibrion septique and their relation to the production of toxin. Jour. Path, and Bacteriol., 23, 53. RoDELL.^, A. 1902 tJber anaerobe Bakterien ini normalcn Siiuglingsstuhle. Zcit.schr. f. Hyg. u. Infcktionskh., 39, 201. RoDELLA. A. 1905 Sur la differentiation du 'Bacillus putrificus' (Bienstock) ct des bacilles ana6robies tryptobutyriques (Achalme). Ann. de I'Inst. Past., 19,804. RoDELLA, A. 1906 l^ber die Bedeutung der streng ana6roben Faulnisbacillen fur die Kiisereifung. Centralbl. f. Baktoriol. II. Abt., 16, 52. SiHONDS, J. p. 1915 a Studies in Bacillus Welchii with special reference to classification and to its relation to diarrhoea. Monog. Rockefeller Inst, for Med. Res., No. 5, SiMONDS, J. p. 1915 b Classification of the Bacillus Welchii group of bacteria. Jour. Infect. Dis., 16, 31. Stoddard, J. L. 1918 Points in technic of separating anaerobes. J. Am. Med. Ass., 70, 900. Stoddard, J. L. 1919 a Bacillus egens. A new pathogenic anaerobe. Jour. Exper. Med., 29, 187. Stoddard, J. L. 1919 b B. multifermentans tenalbus. Lancet, 1, 12. VAN TiEGHEM, Ph. 1877 Sur le Bacillus Amylobacter et son r(Me dans la putrd- faction des tissus v6g6taux. Bull, de la Soc. Bot. de France., 24, 128. VA.v TiEGnEM, Ph. 1879 Physiologic vdgdtale. Sur la fermentation de la cellulose. Compt. rend., 88, 205. TissiER, H. 1908 Recherches sur la flore intestinale des enfants &.g6s d 'un an 11 cinq ans. Ann. de I'Inst. Past., 22, 189. TissiER AND Martelly 1902 Recherches sur la putrefaction de la viande de boucherie. Ann. de. I'Inst. Past., 16, 865. TizzoNi, G., Cattani, J., and Baquis, E. 1890 Bakteriologische Untersuchun- gen (iber den Tetanus. Ziegler's Beitrage z. path. Anat. u. allg. Path., 7, 568. Tbecul, M. A. 1865 Production de plantules amylifdres dans les cellules v6g6tale6 pendent la putrefaction. Chlorophylle cristalizde. Compt. rend. Acad. Sciences, 61, 156; 436. Tr^cul, M. A. 1867 Sur le Bacillus amylobacter. Compt. rend. Acad. Sci- ences, 65, 513. TuLLOCH, W. J. 1917 On the bacteriology of wound infections in cases of tetanus and the identification of Bacillus tctani by serological re- actions. Jour. Roy. Army Med. Corps, 29, 631. TuLLocH, W. J. 1919 The isolation and serological differentiation of B. tetani. Proc. Roy. Soc, 90, 145. Weinberg, M., a.nd S^guin, P. 1918 La Gangrene gazeuse. Paris. Werner, G. 1905 Die Agglutination bei Gasphlegmonebazillen. Archiv. f. Hyg. u. Infektionskh., 53, 128. 36 HILDA HEMPL HELLER Wilson, W. J. 1919 Gas gangrene due to infection with a new pathogenic anaerobe — Bacillus tumefaciens. Lancet, i, 657. WiNOGRADSKY, S. 1894 Sup I'assimilation de I'azote gazeux de I'atmosph^re par les microbes. Compt. Rend. Acad. Sciences, 1893, 66, 1385; 1894, 68, 3.53. Centralbl. f. Bakteriol., 1894, 16, 129. WiNOGRADSKY, S. 1896 Sur le rouissage du lin et son agent microbien. Compt. Rend. Acad. Sciences, 1895, 71, 742. Centralbl. f. Bakteriol. II. Abt., 1896, 2, 273. WiNOGRADSKY, S. 1902 Clostridium Pastorianum, Seine Morphologie und seine Eigenschaften als Buttersaureferraent. Centralbl. f. Bakteriol., 9, 43; 107. WiNSLOw, C.-E. A., AND oTHEBS 1920 Final Report of the Committee of the Society of American Bacteriologists on Characterization and Classi- fication of Bacterial Types. "The families and genera of the bacteria." Jour, of Bacteriol., 6, 191. Preliminary report. Ibid., 1917, 2, 505. WoLT, C. G. L. 1918 Contributions to the biochemistry of pathogenic an- aerobes. V. The biochemistry of vibrion septique. Jour. Path, and Bacteriol., 22, 115. Wolf, C. G. L. 1919 Contributions to the biochemistry of anaerobes. VI. The proteolytic action of Bacillus sporogenes (Metchnikoff) and Bacillus welchii. Ibid., 22, 270. Wolf, C. G. L. 1920 Contributions to the biochemistry of anaerobes. IX. Biochemistry of Bacillus oedematiens. Ibid., 23, 254. Wolf and Harris 1917 a Contributions to the biochemistry of pathogenic anaerobes. I. The biochemistry of Bacillus welchii and Bacillus sporogenes (Metchnikoff). Ibid., 21, 386. Wolf and Harris 1917 b Contributions to the biochemistry of pathogenic anaerobes. III. The effect of acids on the growth of Bacillus welchii (Bacillus perfringens) and Bacillus sporogenes (Metchnikoff). Bio- chem. Jour., 11, 213. Wolf and Harris 1917 c The conditions of growth of Bacillus welchii in the presence of oxj'gen. Lancet, II, 787. WoLP and Harris 1918 Contributions to the biochemistry of pathogenic anaerobes. IV. The biochemistry of Bacillus histolyticus. Jour. Path, and Bacteriol., 22, 1. Wolf and Telfer 1917 Contributions to the biochemistry of pathogenic anaerobes. II. The acid production of Bacillus welchii (B. per- fringens) and Bacillus sporogenes (Metchnikoff). Biochem. Jour., 11, 197. WiJRCKER, K. tlber Anaerobiose, zwei Fiiulniserreger und Bacillus botulinus. Jahresber. der phys. med. Soz. in Erlangen. 1909, 41, 209. INDEX TO NAMES OF BACTERIA Actinomycetales, 2; 14 Amylohactcr, 9 ff. Arloingillus , G; 21 ; 29 Chauroci, 0; 21 Bacille de reumatisme aigue, 18 Bacilbis, 1 aerofoclidus, 4 ; 7 ; 23 aerogenes capsulatus, 6; 18 ■ — amylobacter, 9 ff. bellonensis, 23 ; 20 bifermentans sjHyrogines, 4; 8; 25 bifurcatus gazogenes, 16 — — — botuliniis, 4; 8; 28 • — ■ butyricus iClostridiitm),5; 12; 17 of Adamson, 17 ■ — cadaveris sporogenes, 25 capillosus, 16 cochlearius, 5; 15 Bacillus C of Adamson, 15 Bacilhis egens, 6; 19 enterilidis sporogencs, 21 ; 29 Bacillus E of Adamson, 14 Bacillus f alia X, 0; 16 — — ■ — gazogenes, 13 ff. gracilis putidus, 16 histolyticus, 4; 9; 31 hydrogenicus, 12 irregularis, 16 Bacillus L of Adamson, 22 Bacillus megalosporus, 17 ynelhanicus, 12 mullifcrmentans tenalbus, 6; 16 oedcmalicns, 4; 23 oedemalis maligni, 21 ; 30 — — //, 7; 23 parapuirificus, 8; 27 parasporogenes, 26; 30 perfringens, 18; 25 phlegmones emphysemaioseae, 18 Bacillus posltimus, 27 puirificus, 8; 27; 28 Bacillus Rodella III, 13; 15 Bacillus saccharobutyricus, 9 Bacillus S of Adamson, 23 Bacillus sphenoides, 5; 14 sporogenes, 9; 29 ff. B, 26; 29 ff. parvus, 31 tertius, 5; 15 telani, 7; 24; 25 letanoides, A 13 B, 25 telanomorphus, 5; 13 • lumefaciens, 21 ventriosus, 16 Welchii, 4; 22; 30 Bacillus Y of Fleming, 15 II of Choukevitch, 26 I of Hempl, 7; 24 — II of Hempl, 26 — Ill A of Mcintosh, 15 Ill C of Mcintosh, 15 XII of Mcintosh, 26 — XIII of Mcintosh, 26 XIX of Mcintosh, 30 Ill of Rodella, 13; 15 IV of von Hibler, 29 — VI of von Hibler. 6; 17 ■ ■ VII of von Hibler, 17 IX of von Hibler, 15 XI of von Hibler, 30 Bacterium Chauvoei, 6; 21 — lumefaciens, 21 Bifermentans, 4 Butyribacillus immobilis liquefadens, 18 Central spore bacillus, 26 Closlridiaccae, 1; 3 Clostridioideae, 1 ; 9 ff . Clostridium, 5; 9ff. 21 — • americanum, 12 amylobacter, 9 ff. bulyricum, 5; 9 ff. 37 38 HILDA HEMPL HELLER Clostridium, Paslorianum, 3; 11 • — — — polymyxa, 12 Coccobacillus oviformis, 16 • proeacutus, 14 Douglasillus, 5; 14 proeacutus, 14 ■ ■ — sphenoides, 5; 14 Ermengemillus, 8; 28 boiulinus, 8; 28 Eubacleriales, 2 Flemingillus, 5; 15 cochlearius, 5; 15 Fusiformis, 14 Granulobacillus saccharobutyriais mo- bilis nonliquefaciens , 9 Gramdobactcr saccharobutyricum, 9 Henrillus, 5; 15; 16 tertius, 5; 15 Hiblerillus, 6; 17 megalosponts, 17 rectus, 17 Septimus, 17 seiiMs, 6; 17 Macintoshilliis 5; 13 — — ■ — • Ictanomorphus, 5; 13 Marlellillus, 8; 18; 25; 26; 30 bifermenians, 8; 25 proleolylicus, 26 Metchnikovillus, 9; 29 ff. — sporogenes, 9 ; 29 Meyerillus, 7; 22 sadowa, 7; 22 Mullifermentans, 6; 12; 16 lenalbus, 6 ; 17 Mycobacteriaceae, 14 Nicolaierillus, 7; 24 telani, 7; 24 Novillus, 7; 22 Novillus, bellonensis, 23 maligni, 7; 22 oedematiens, 23 Omelianskillus, 5; 12 gazogenes, 13 hydrogenicus, 5; 12 methanicus, 12 Plectriditim, 25 pulrificum, 25 pseudotetanus bacillus, 5; 13 Pulrificoideae, 1; 23 ff. Putrificus, 27; 28 Bienstocki, 8; 27 Reading bacillus, 30 Recordillus, 8; 26 fragilis, 8; 27 Reglillus, 4; 7; 23 progrediens, 7 ; 24 Rivoltillus, 6 ; 20 vibrion, 6; 20 Robertsonillus, 7; 24 primus, 7; 24 Seguinillus, 7; 23 aerofoetidus , 7; 23 Sloddardillus, 6; 18; 19 egens, 6; 19 Slreptobacillus anaerobicus magnus, 16 rectus, 17 Tissierillus, 8; 27 paraputrificus, 8; 27 Vallorilhis, 6; 16 fallax, 6; 16 vibrion septique, 4; 6; 20 Weinbergilhis, 9; 31 histolylicus, 9; 31 Welch bacillus, see II'eZcAiWus Welchillus, 6; 18; 19; 20; 22 aerogcncs, 6; IS STUDIES UPON AGGLUTINATION IN THE COLON- TYPHOID GROUP OF BACILLI O. ISHII From the Deparlmertl of Preventive Medicine and Hygiene, Harvard Medical School, Boston, Massachusetts . Received for publication April 19, 1921 THE PREVENTIVE ACTION OF FORMALIN IN THE AGGLUTINATION TEST Malvoz in 1897 claimed that formalin produced chemical agglutination of Bod. typhosum, but not of Bad. pnratyphosum and Bad. coli, while, on the other hand Beco, Kemy, Widal and Nobecourt were unable to confirm Malvoz's observations. By adding amounts of formalin from 0.05 to .5 per cent to broth or agar cultures, I have obtained no e\adences of chemical agglutination with Bad. paratyphosum A and B, Bad. dysen- teriae and Bad. coli; with concentrations of 0.0.5 to 0.3 per cent formalin spontaneous agglutination in these cultures was almost or entirely prevented while Bad. typhosum and Bad. paratyphosiim A showed spontaneous agglutination without for- malin but none with it, when the above dilutions were used. WTien concentrations of 1 to 5 per cent were used slight spon- taneous agglutination sometimes occurred. Spontaneous agglu- tination was more or less inhibited in all dilutions with Bad. paratyphosum B. With Bad. coli there was absolutely no change in spontaneous agglutination when formalin was added. From the above facts it is evident that a concentration of 0.05 to 0.2 per cent formalin gives the most satisfactory results in the prevention of spontaneous agglutination, correct read- ing being possible at the end of two hours. However, it is a better rule to let the tubes stand over night. 39 40 O. ISHII Formalin prevents spontaneous agglutination, but does not interfere with agglutination by specific immune sera; it does, in fact, increase the agglutinating reaction, when used with specific serum. It has been customary to add formalin as an antiseptic, or to prohibit the growth at certain times of the cultures of the colon- typhoid group, which were to be used in conducting the' agglu- tination test. This method was adapted by Loele, who used 2 per cent formalin, while Porges, Coles, Bass and Watkins, TABLE 1 Standardization of formalin concentration in plain broth culture media to prevent spontaneous agglutination. {Cultures twenty-four hours at 37°C.) Organism < BACr. TTPHOaOM BACT. PAB.iTYPHOSUM A BACT. PARATYPHOSUM B BACT. COLI Number of strain | No. No. 2 No. 3 No. 4 No. 1 No. 2 No. 4 No. 6 No. 1 No. 2 No. 3 No. 5 No. 1 No. 3 No. 5 4 No. 8 Spontanous agglutination 1 on cultures J 4 3 3 3 2 2 3 2 3 3 3 3 3 3 4 Dilutions of formalin : 0.05 per cent 0.1 per cent 0.3 per cent 1.0 per cent 5.0 per cent 0 0 0 0 1 4 0 0 0 0 0 2 0 0 1 1 1 3 0 0 0 0 3 0 0 0 1 2 2 0 0 0 0 1 o 0 0 1 1 1 2 0 0 0 0 1 2 1 1 1 2 2 3 I 2 1 1 1 1 2 2 2 2 2 2 2 3 3 3 3 3 3 3 4 4 4 4 4 4 3 3 3 3 3 3 3 3 3 3 3 Control no formalin. . . 3 tj: 5, complete; 4, almost complete; 3, weak; 2, very weak; 1, trace; 0, negative agglutination. Buxton and Vaugham, Roily, and Asser used 1 per cent, Widal 0.67 per cent, Garrow, Chick 0.1 per cent; Neisser, Proscher, Lion, Martineck, Ehrsam, Flatau and Wilke, and Selter also used formalin and showed, that it had a tendency to slow down or dinunish the reaction. Dreyer has made an especially careful study of the use of fonnalin. He found that 0.1 per cent formalinized broth cul- tures agglutinate more actively than those without formalin, and that 0.5 to 1 per cent formalin cultures agglutmate less actively than those without formalin. STUDIES UPON AGGLUTINATION 41 In my experiments (table 2) using Bad. lyphosum and Bad. paratyphosKtn A and specific immune senmi, agglutination was strongest with tests containing 0.05 to 0.1 per cent formalin; 0.3 per cent showed sUghtly less agglutination and 1 per cent TABLE 2 Standardization of formalin concentration in plain broth culture media to promote specific serum agglutination. (Cultures of twenty-four hours at SyC.) Oroanibm BACT. TTPHOSCM BACT. PAn.VTTPn08DU A BACT. PABATTPBi }8rM R Number of strains' No. 1 No. 2 No. 3 No. 1,2 No. 29 To- tal 5 No. 2 No. 3 No. 6 No. 1,2 No. 4 To- tal 5 No. 1 No. 3 No. 5 No. 1,2 No. 5, 2 To- tal 6 Spontftneous ag-1 glutination on ) cultures j 0 0 0 4 3 0 0 0 2 2 0 0 0 3 3 Dilutions of formal in and specific serum: 0.05 percent. 4 5 5 5 5 24 5 5 5 5 5 25 5 5 5 5 4 24 0.1 percent. . 4 5 5 5 5 24 5 5 5 5 5 25 5 5 5 5 4 24 0.3 percent. . 4 4 5 4 5 22 5 4 4 5 5 23 5 4 4 5 4 22 1.0 percent. . 3 4 4 4 4 19 4 4 3 4 4 19 4 4 4 4 4 20 5.0 per cent.. 2 3 3 4 3 15 3 3 3 3 4 16 3 3 3 4 4 17 Control, 0 3 4 4 4 4 19 4 4 4 4 4 20 5 4 4 4 4 21 Control with- out serum: 0.05 percent. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 3 4 0.1 percent. . 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 3 4 0.3 percent. . 0 0 0 1 0 1 0 0 0 1 0 1 0 0 0 1 3 4 1.0 percent. . 0 0 0 1 0 1 0 0 0 1 1 2 0 0 0 1 3 4 5.0 percent. . 0 0 0 1 1 2 0 0 0 2 1 3 0 0 0 2 3 5 Control, 0 0 0 0 4 3 7 0 0 0 2 1 3 0 0 0 3 3 6 Keyt:5, complete; 4, almost complete; 3, weak; 2, very weak; 1, trace ;0, nega- tive agglutination. produced about the same results as those obtained without forma- lin. With Bad. paratyphosum B, agglutination by specific serum in the presence of 0.05 to 0.3 per cent formalin was about the same as, or slightly stronger than, without formalin; with 42 o. iSHii 1 to 5 per cent formalin the agglutination reaction was less than without formalin. However, it has been difficult to standard- ize this test with formalin, because the reaction depends on the strain of microorganism employed. Many baciUi showed agglu- tination in all dilutions used and only a few showed different degrees of agglutination; I found that 0.05 to 0.2 per cent formalin- ized serum gave the best results for most of the strains employed. TECHNIQUE Since in plam broth or agar cultures spontaneous agglutina- tion interferes with specific agglutination, in such cultures accurate tests are impossible. When spontaneous and non-spon- taneous agglutination types are mixed together, I have observed two types of colonies, (by streak on agar plate) occurring regu- larly throughout the cultures. One type of colony has a smooth surface and regular outline and shows non-spontaneous agglu- tination, while the other type of colony has a rough surface and irregular outline and is spontaneously agglutinating. Either type of organism would, however, produce both types of colonies after growing for some time. Some strains changed in a few daj^s, while others remained tioie to type, producing characteristic colonies even after continual transplanting for over a year. Cultures used in the experimental work were obtained before each set of experiments by plating and fishing a characteristic colony into experimental media. Only neutral broth cultures gave uniform results. Therefore, 18 to 24 hour cultures in this medium at 37°C., could be used as an indicator of spontaneous and non-spontaneous agglutina- tions. The method of my investigations has been to mix 0.3 of each suspension of bacterial cultures in each medium, (such as broth, pepton water and so forth) with 0.3 cc. of the seiinn dilution, with formalin or without, or with other experimental fluids in small test tubes, of the usual Wassermann type. The tubes, were left at room temperature (excepting in the experunents on temperature), and the preparations were examined, both STXJDIES UPON AGGLUTINATION 43 macroscopically (bj'' means of a hand lens) and microscopically. In the latter case, as many as 15 or more mixtures were exannned at the same time, by transferring a small loopful of each to a square on a large thin glass slide, which had been marked off in squares by means of a grease pencil. USE OF FORMALINIZED SPECIFIC SBEUM IN AGGLUTINATION TEST VnTH VARIOUS BACILLI (TABLE 3) I have observed that specific serum to which has been added 0.2 per cent formalin (final 0.1 per cent) in 0.85 per cent salt solution shows stronger agglutination than occurs without forma- lin. The formalinized senam mixed with an ecjual amount of broth culture after three hours at room temperature, shows strong agglutination with Bad. typhosum and Bad. paratyphosum A, and the same is true, if tubes are allowed to stand over night. Bad. paratyphosum B and the Bad. dysenteriae group show a weak reaction, when examined within three hours; but a stronger reaction if kept over night. Bad. coli shows weak formalin readings after three hours; but if kept over night, the result, with or without formalin, is the same. For the graphical agglutination test, we must use non-spon- taneous agglutination bacilli. However, by the addition of formalin, spontaneous agglutination is prevented while at the same time a specific agglutination regularly occurs. As showni in table 3 of the strains of spontaneous agglutination bacilli, Bad. typhosum and Bad. paratyphosum A yielded specially good results; Bad. paratyphosum B showed less influence, while Bad. coli was not influenced at all. When the tests were conducted without formalin, spontaneous and specific serum agglutina- tions usually present the same appearance microscopically and macroscopically, although certain strains of spontaneous agglu- tination bacilli sometimes yield weaker macroscopic reactions. It is evident that spontaneously agglutinating bacilli should not be used for graphical agglutination tests, and that formalin may be used to prevent the complex reaction. O o ►J n <: E- s s B. S 8s;h| anong I tniuss agioddg c ° S -M o M»^°x| -> I 6 ON « f ON I cj I Z -OX r 'ON I uituas ogpedg smox « '£ 'SON I C'S-sonJ^, e ig)ox I 8 "ON I Z 'ON I I 'ON I uinjas ogioadg 38 •w o o 1-H (M -^ Tj< 1-H I-H lO »C 1-H ^^ 28 o e i«»ox I Z 'e 'SON I « I ^ ro r-l rt ?; '2 'BON I 'c I £ mox C ON I o I J -ON I 'ON mniaB ogiaadg Z moj. I C'Z-eoN „ e'l'soN <^ £ injox £-°N| J -ON I 'ON mtuas ogpaclg o8 *J o o8 -^ o o o o 3 Z "52 ■« o c • r 1, c a> en O M 44 ^ 2 ■• E — £ ii 2 ° "s -5 fc< CO o STUDIES UPON AGGLUTINATION 46 ACTION OF FORMALIN ON PSEUDO AGGLUTINATION IN THE CROSS AGGLUTINATION REACTION WITH DIFFERENT IMMUNE SERA (TABLE 4) In conducting cross agglutination tests with various immune sera pseudo agglutination usually did not take place with forma- lin. However, certain strams of baciUi showed marked agglu- tination, as well as specific agglutination, without formalin. Bad. typhosum. Non-spontaneous agglutinating bacilli tested with the formalin showed very marked prevention of pseudo agglutination in these immune sera: Bad. paratyphosum A, and B, in horse serum, and even better in Bad. dysenteriae Shiga serum, but showed strong spontaneous agglutination without it. The serum of polyvalent dysentery (1:2000), with Bad. ty- phosum had almost the same agglutinatmg reaction as the Bad. dysenteriae group, both with or without formalin. Hence, we may use the serum for identification of Bad. typhosum as well as for the Bad. dysenteriae group. Bact. paralyphosxim A. Formalinized Bad. typhosum serum ( 1:200), prevented pseudo agglutination entirely, with non- spontaneous agglutinating strains, while showing strong pseudo agglutination without formalin. In the Bad. paratyphosum B serum (1:50), with non-spon- taneous agglutinating bacilli, both with or without formalia, agglutination does not appear. This serum showed no tendency to cause pseudo agglutination for Bad. paratyphosum. A. Polyvalent dysentery serum, has agglutinated Bad. paraty- phosum A, in almost the same degree with and without formalin; this serum seems to have a tendency to agglutinate this bacillus, but in this experiment the sei-um was highly concentrated 1:100. This reaction does not compare with the agglutinating reaction of Bad. typhosum or the Bact. dysenteriae group, when a dilution of 1 : 2000 or more is used. Bact. dysenteriae Shiga serum (1:50), showed negative agglu- tination, both with and without formalin; this serum has no tendency to agglutinate with Bact. paratyphosum A. Normal horse serum (1:50), shows very weak agglutination when using formalin, and stronger without it, but this serum shows a tendency to agglutinate only Avith Bact. paratyphosum A. o s ©J 8 s . s 9 ["lOX O o M t^ o o CM o -^ s s o ° 1 Suoj^s o o CO •-< o o lO lO 'S^ CO LO l-O o o ssiH o o (M <-> o o IC »o o c lo in o o Jta.ixaiji o o (N .-1 o o •O LO (M .-1 lO lO o o X ';^;q> o o CO rt o o »-0 "O i-O lO o o I K^mS o o CO CO o o lO IC i-O lO o o E I'JJox o o o o o o ^H I— 1 CO to lO >o o c £ON o o o o o o lO o CO CO >0 U3 o o ("•ox o o o o o o >-o IC O o o "5 fOSl O iO o o ^ I" o o N CO o o '» SON O Tt< o o CO TJH o o 1-i I— I o o ■< I ON o -* o o CO Tj< o o CM CO o o g EWOX O rH o »o S 2 O Tt< (N O o o £ ON o ■* O C^l ■O lO O .-H O CM o o S-ON O CO O CO IC lO o o O CO o o u ^ I -ON O f o o >0 lO O CO M ^ o o *o CO o a a c n c B c o c o o o o o -*J -^ -^ a 3 C D C 3 O 3 C 3 C 3 a 3 00 1" O rt "3 g S "3 lg lg £ :S, O c3 lg £ ii o ^ 11 o a fe cc |i< W fr( M fcl CB f^ CO fe CB fe CQ o u o o o • o o w o i* -H o E 1— 1 in e. S 2 CD S ^ ^ ^ g rt >0 ^ _ 3 „ 00 c 6 □ m •a s 3 m o j2 < c3 -a o GO o C. 5 >-. ci 03 • c GO t. ■*J 03 g 00 s t4 a 3 c3 rS C3 o >. o O g; pa m ca CL, Q ^ < h oc ffl M a: :^ a " a § o § o s o s = W3 <5 1 5 T H « a: H k o 0 « ;4 CO 1— < )-H ; £ 1 o 1 46 STUDIES UPON AGGLUTINATION 47 Bad. paratyphosum B. Formalinized Bnct. lyphosum senim (1:200), practically did not show the pseudo agglutination; most cases were negative (with the exception of no. 1 strain, which had strong agglutination with this serum, and weak agglu- tination with homologous serum, — a peculiar strain but never- theless typical Bad. paralypJiosuDi B), while showing strong agglutination without formalin. Bad. paratyphosum A senim (1:50); pseudo agglutination is shown to be negative with formalin, but strong agglutination occurs without formalin; hence, this serum does not have a tendency to pseudo agglutination with Bad. paratyphosum B, if used with formalin. Polyvalent dysentery serum (1: 100), or normal horse serum (1 :50), shows negative or weak pseudo agglutination with forma- lin, and strong agglutination without it. Bad. dysenteriae Shiga senmi; many strains are negative or show a weak reaction with formalin, but show more pseudo agglutination without formalin. This, however, is a strongly concentrated sei-um (1 : 50), if it is diluted 1 : 100, it does not agglu- tinate with Bad. paratyphosum B. Bad. coli. As a loile Bad typhosum serum (from immune rabbit or sheep), with or without formalin, shows no agglutination in high concentrations, as 1:200. Bad. paratyphosum A and B serum (from immune rabbit), generally showed no agglutination with Bad. coli. Polyvalent dysentery serum showed strong agglutination with Bad. coli either with or without formalin, almost equal to that of the Bad. dysenteriae group, as observed by Kligler. This serum we could use for a diagnostic test for Bad. coli. Bact dysenteriae Shiga serum (1:50), showed slight or negative agglutination with formalin, but more without formalin; it is however not a marked reaction. Bad. dysenteriae group. With Bad. typhosum serum (1:200), and Bad.paratyphosumB serum (1:50), with or without formalin, pseudo agglutination was generally negative, but with Bad. paratyphosum A serum 1:50 (obtained from rabbit), there was slight agglutination. 48 o. iSHii Bad. dysenteriae Shiga serum (1:50), shows sUght agglutina- tion with the Flexner and Strong strains; but the reaction is stronger with formalin than without; with the Hiss strain, there is no agglutination either with or without formalin. This serum does not have a tendency to group agglutination as do sera of other Bad. dysenteriae groups, and horse serum serves for agglu- tination as well as specific serum. HORSE SERUM AGGLUTINATION FOR BACT. DYSENTERIAE GROUP AND BACT. COLI. (tABLE 5) Lentz, Park and Williams have found that Bad. dysenteriae is agglutinated by normal horse serum, and Gasiakoski, Park and Williams found that Bact. coli is agglutinated by normal horse serum. In my experiments nearly every strain of Bad. dysenteriae and Bad. coli was readily agglutinated by normal horse serum. All samples of horse serum employed, as sho-wm in the table, produced agglutination of both groups of bacilli ; only one strain of Bact. dysenteriae Shiga no. T obtained from Dr. Shiga, showed negative or weakly positive agglutination. I have also found that the sera of horses immunized with different bacteria as for instance Corynebad. diphtheriae, the meningococcus, Bact. paratyphosum and the pneimiococcus, have the same agglutinating power as normal horse serum. COMPARATIVE AGGLUTINATION TESTS WITH 0.7 PER CENT ACID, NEUTRAL, AND 0.3 PER CENT ALKALINE BROTH CULTURE MEDIA (TABLE 6) Bad. typhosum. Freshly prepared acid broth was used; a heavy growth and actively motile organisms with a tendency to spontaneous agglutination were obtained within 20 hours; with formalin, spontaneous agglutination was prevented almost en- tirely, while without formalin considerable spontaneous aggluti- nation occurred. These cultures were more easily agglutinated with specific serum than cultures grown in neutral or alkaline broth. According to Dreyer, Eisenberg and Volk, Joos, and Weiss acid increases the agglutinating power. STUDIES UPON AGGLUTINATION 49 TABLE 8 Agglutination of Bad. dysenteriae and Bad. coli, with normal and immunized horse serum with different organisms DACT. ^^ CI d Z 1 05 1 .1 d s d B X a 2 1 s 2; .a g Z e 3 2 CD 1 S 3 z 1 Z No. 2' 1:200 5 1 5 5 4 5 5 5 5 5 3 1:1000 5 0 2 2 0 1 3 3 5 2 1 Normal horse No. 3 1:200 5 3 3 3 1 4 4 5 1 3 3 1:1000 5 1 1 1 0 1 1 2 0 0 0 No. 6' 1:200 5 2 5 4 2 5 5 5 5 5 4 Anti-meningococ- 1:1000 5 0 1 0 0 1 5 4 3 2 1 Normal and Immunized cus No.?/ 1:200 1:1000 5 5 0 0 5 4 5 2 5 5 5 2 5 3 5 1 5 3 5 2 4 1 sera of Horse Anti-diphtheria No. sj 1:200 1:1000 5 6 0 0 5 2 5 1 4 1 5 5 5 5 5 2 5 3 2 0 4 1 No. 9 1:200 5 0 5 5 5 5 5 4 5 3 4 . 1:1000 5 0 3 1 2 3 3 2 2 1 1 Anti-pneumococ- f 1:200 5 0 5 5 5 5 5 4 5 4 4 cus \ 1:1000 5 0 3 2 1 4 4 1 3 1 1 Anti-B. para- J 1:200 5 4 5 4 5 5 5 5 5 4 3 typhosum A \ 1:1000 5 1 5 1 4 3 5 5 5 1 1 Control 0 0 0 0 0 0 0 0 0 0 0 I I Key: 5, complete; 4, almost complete; 3, weak; 2, very weak; 1, trace; 0, negative agglutination. In neutral broth the motility, spontaneous agglutination and also specific serum agglutination, were much weaker than in acidified broth. Formalin, prevented spontaneous agglutina- tion entirely. In alkaline broth most strains of Bad. typhosum showed a much weaker growth, motility and spontaneous agglutination. Eh £ 0- o CO c 00 CO IN i-l 1. •^ (M --1 (N ^ IN —1 o 0^ 8 0- W + V O M -H o Cfl + V O .-1 o o M + 05 ^ C<1 r-t e T o + C<3 ■* c< ? ^ ^ + o •* cq o + O ^ IN o d + o in CO o 2 z IS ? ^ s ^ iO + + o in CO o + O iO CO c + o m CO o - -e 1 "a § £ 0- + V O M O o + V V O Tj< r- o + c ^ i:\ o S i- T i + O Tl< r-< ° 1 + V O lO IN c 1 + O Tt< cq o t-; 10 + + t-H U3 CO o + IN to CO c + •-I m o o o Q o o o o o o o „- o o o o " lO o o o tN ■- r-l >o —c in T-H rH •—1 f-H t-( •— I ^ ^ ^- ". C3 0 a c g o o •irJ a -5 ■*^ s S R Cj c3 cS ■i: •*^ i ■s. "3) S g i i ^ 1 s -1 1 a 0 to c 1 o en m 3 •« O O 0 9 =s 2 1 o o OO 00 3 .^ o o a c o M X « 2 r > S o "c 1 <^ s ■u § 2 c Si a ^ o ^ c a s c O- "^ 3 h y. -5 a 1- c_ 2 ^ a C 2 ^ a: ■ r; c 1 50 O Oi r/l r 5 CO -f CO CO 1— 1 e^ -H 1 a 1 O N -H c > S CO in t~- CO e^ 'H (N -H 1 u 1 n to 3 o f n c > T)< lO t^ CO IM .-1 IM -< o a o c > - IN U5 •* ^H e^ f i 1 O ■* rt 3 c - (N lO ■* f t-H tt © lO cq c > - 1 M >0 ■* 1-i o lo CO c ) ^ ■n ■* (N e^ i 1 O U5 M c > .- lO ■* (N EI i ,= O U5 CO c c cs "5 -^ N O M O c c O •* IM O i » ■V i o CO .-1 c 6 o in CO o ' ^ c O •* N c c o lo CO o O U5 P) c ) c o io CO o {■ es S o in c^ c > d c o in CO o ^ 2 o m !N c c O lO CO o O "J" CO c c c in CO o ^ « •^ ^ 1 O lO Tjt c ^ c c in CO o O lO rtl c c a in CO o o o^ Q o o o ira lO -^ m rt *-^ ^H i-H I— t c a c c •*: (S C! .£ a •*: ^ _3 3 "3 ^ § OJ OS o3 5 2 a a B S a 00 = •« :: c o n c o e a a m c o o s: fl , ; « c kl > c o ■c -4J O ■4^ 3 c a o •^ =! - Q o z c c 03 ^ C O 1 a i£ u > c S o in •^ 3 ^ 'So 51 52 o. isHii The specific serum agglutination in this medium was also weak. Therefore the use of alkaline media for Bad. typhosuvi is not favorable for agglutination tests. According to Park and Wil- liams, and Tarchitte, the agglutinating power of the organism is lost in an alkaUne medium. Bad. paratijphosum A. These cultures were grown in acid, neutral and in alkaline broth, the reactions of the media being the same as with Bad. typhoswn. Bad. paratyphosum B. In general the differences in growth and motility in acid, neutral and in alkaline broth were far less marked than in the case of Bad. typhosum and Bad. paratyphosum A. Both spontaneous and specific agglutination were strongest in the acid medium, but the differences were less than with Bad. typhosu7n or Bad. paratyphosum A. However, there is some variation depending upon each stram. Bad. dysenteriae and Bad. coli group. On an average the speci- fic serum agglutination test was sUghtly stronger in acid broth; otherwise the results were almost the same for acid, neutral, and alkaline broth cultures. In the preceding experiments the nutrient broth was prepared and titrated with phenolphthalem, just before using, as I have reported. The mediimi becomes more acid on standing, due to the absoiption of carbon dioxide from the atmosphere. COMPARATIVE AGGLUTINATION TEST W^TH AG.AR CULTURE EMUL- SIFIED IN SALT SOLUTION OR BROTH MEDIUM Bass and Watkins, Buxton and Vaugham, KoUe, Jordan, Park and WilUams, Ritchie, Weil, Wretoski and other workers have used salt solution, for emulsifying the agar culture for the Widal test. Block, Grumbaum, Durham, and Ker used broth medium with the agar cultures; Hiss and Zinsser used agar cultures emul- sified in salt solution or broth medium. In my experiments, I have not found any particular difference between these fluids. But agglutination is sUghtly stronger with broth emulsions of Bad. typhosum, Bad. paratyphosum A, Bad. dysenteriae and Bad. coli; with Bad. paratyphosum B the reactions are about equal. STXJDIES UPON AGGLUTINATION 53 There is slightly less tendencj' to spontaneous agglutination in a broth medium than in salt solution; therefore, I have con- cluded that salt is better for preparing suspensions with agar cultures for Widal test. EFFECT OF SODIUM CHLORID ON AGGLUTINATION TESTS (TABLE 7) jNIalvoz used agar cultures emulsified with distilled water to avoid chemical changes resulting from sodium chlorid. Weil, working with agar cultures, obtained the same agglutinating reaction, when using cither distilled water, or 0.85 per cent sodium chlorid. Asakawa, Bordet, Chick, Joos, Jordan, and Forges claim that sodium chlorid is necessary for the agglutination test. Dreyer, Krumbaar and Smith used tap water in diluting the serum for broth cultures and the results were better than with sodium chlorid. Chick claims 0.42 per cent sodium chlorid gave good results. In my observation 24 hour agar cultures of Bad. iyphosum, Bad. paratyphosum A and B, Bad. dyscnteriae and Bad. coli emulsified with distilled water were agglutinated by specific serum in dilutions of 1:40 or 1:50, as well as when emulsified with 0.85 per cent salt solution; in this case a certain amount of salt contained m the seiiim aided agglutmation. In weak dilutions of 1 : 80 or more, there was no agglutination in distilled water (with one exception), but marked agglutination in salt solution. Evidently at that dilution, the salt content of the serum was not suflficient. Only one strain of Bad. dysenieriae Flexner no. 2, showed constant agglutination in distilled water with a specific serum dilution of 1 : 80. Spontaneous agglutination appeared with spontaneous agglu- tinating bacilli in almost every instance, when using either dis- tilled water or salt solution. Quantity of sodium dilorid. I have tried various dilutions of salt, varying from 0.05 to 5 per cent, with agar cultures, but no apparent differences were observed in their effect on the agglu- tination reaction. It is my conclusion, that only a trace of sodium chlorid is necessary for the Widal test. Since strong solutions are of no advantage, I believe ordinary physiological 0.85 per cent salt is « CO s hi g r a esiti O 1 ic o o m 1 o o 1 < a w g s a 3noJJS O lO o o m o o 0 Janxo[j o mm m m o o ■ 1 I Jauxdij o 1 m lo o m o o X «3!qs o m m o m o o 3 C 'S -SOX CO mm CO m CO CO u n C 'I SON (N 1 m m CO m ^H »-t 9 ON o m m o m o o 5- -ON o 1 m m o m o o n e > -BOK n mm rH m 1-i t-t 5 '5 -SON c^ 1 m m i-H m r-l t-H Ci n Z-oti c > m m o m O O I ON c ) m m o m o o < < Z 't -SON c^ m m (M m I-H 1-i 1 Z 'Z 'SON C-- m m m m o m o o 13 iON c^ m m' ■1 m T-4 I-H S 00 H-fe 0 "I 'SON CS m m rt m 1—1 I-H '^ ^3hoid group various kinds of culture media have been used. The cultures were grown at 37°C. for 24 hours in neutral broth, one per cent glucose broth, one per cent pepton water and on agar, the agar cultures being emulsified with salt solution. Bad. tijphosum. Grown in plain broth, glucose broth, pep- ton water and on agar, it showed the strongest agglutination with specific serum in the first named medium, becoming weaker in each of the others in the order given ; the reaction was very weak with agar cultures. Dreyer claims that glucose broth cultures have a tendency to spontaneous agglutination, but my results hi using glucose broth were quite irregular; spontaneous agglu- tination occurred m some cases, while in others there was no spontaneous agglutination, but there was generally negative or weak spontaneous agglutination with both spontaneous and non-spontaneous agglutinating types. .\s compared with cultures in plain broth those grown in glucose broth showed very irregular and weak motility. Forma- hnized serum showed stronger agglutination than plain serum; especiallj^ marked was the difference with glucose broth. With plain broth or pepton water or agar cultures, the difference was a great deal less marked. Bad. paratyphosum A. Cultures grown in glucose broth showed weak motility, many strains agglutinating spontaneously, even though non-spontaneous agglutinatmg bacilli were used. Form- alin does not prevent this agglutination. In specific agglutination tests glucose broth cultures showed considerable reaction, even more than with pepton water or agar cultures. However, these results are due to the spontaneous agglutination in the glucose broth. It is evident, that this JU3V tn o 00 -- o 2 o o o o JS'jtiM a 01 (N ^ •* o IM to ■* f— 1 -dad^uooiad i s o o q-joiq asoo ^ rt o o o "S oo>* IC -n^S ■^nao jad j rH '~' a e O O ^ o o o to o UVDV + O M rt o + O "S< o o e S l.s aaivAi + rt -«< M o ■fl- + O '^t^ ^ o xoxdad j-NHD aad i 6 2 d S5. g H,ioaa + O Tl< (M o + O lO ■-< o asoaaTio .LNaoaadx V S ^ Hxoaa mvid + O lO M o + O U3 (N o u S avov + O (N O o + V O IN O o „ '-H aaiVAi + O CO o o + O Tfl ^ o oO NOidad iNao sad x 6 - 6 Hxoaa >=; + O -*< (N o 2 + CO lO CO CO asooQio XNao aaa i o --^J s Hxoaa Nivid + O lO CO o + O IC -H o to *» avov + O (M O o + o c-1 o o o S ^ X aaxvAi + rt Tf -H o + o ^ o o ^U KOidad iNao aad i N V Hxoaa z + ^ ■* CQ o •z + ^ ■* 1-1 o -C S; asoomo XNaoaad j V V 1^ Hioaa Kivid + O lO (N o + V o ■* ^ o s s avov + o C^ o o + V O (M O o 8 1 :SO aaiVM + O CC •-( o + C^ -^ C) y-t NOidad inao aad i .-' .-. o 2: Hj.oaa + O •!l< !N o z + IN m CO IN aeoomo ixao aad i V V HXOHB NlVld + O L-J CO o + O lO IN o e -i o o *^ "^ ■^ (^ "TJ e ^ o c o o c >o o c o o o o p^ .^ 13 (N >-i c -H to o o ca .-( T- I— 1 1-H o C3 -4J C3 ^ s g a 3 3 a 2 to > > 1 ^ c. 1 g >" a - 2: oil a "2 - c 1- ^ a D c s o o C m s M m c PQ s CB « u 56 Cl — lO o H 0» >o o 1 r^ 00 1 o 1 « n »-H 1 -1< lO CS N CO «3 0> ^ Nl o 00 o o i ■g i-H s I-H I-H > t^ >o N 00 00 00 o> 1 C3 1 « 1 O lO lit 1 O 1 •V f-H i-t »-H r 1 O r)< 00 o O 00 CO ^1 o o o o l-H f-H 1 (M rt ^ > / / O CO N o O CO w c 3 o ■* .- ■< o + O M -H o c s t-H lO C^ c > O ■* (N O 1 6 a + •^ CO C<1 ^^ iM lO IN c- , ^ O '^ I- < o V + <=> n ci o c s O lO O U3 O ) o + rt CO w o fH M N c 3 O ^ '■ < o V + »-l lO CO l-H 1 -H ■* -^ (V > (■ O -* c^ 1 o 6 Z V M 1 C + V -H •* CO (N Z c N U3 CO c > £ O CO c ) o + o ■* c^ o c O lO M c ) o o c^ 1 o + O CO (N o -i V O C<1 (N c ) O U3 Ti < o 6 + ■I CO IM o - 4 V r-H Tl< tN c ' 1 O >0 r» o + (N ■* CO CO 2: 4 ca lo CO c^ 1 O lO CC o V V K + O CO (N o 4 V o lo cs c O lO ^ o ' 1 1 3 3 3 c O O o « ss n o o fl ,— . o 3 o o o O o o O ^ iT. c M >0 c t^ CO 3 (N - o 'H j-t o •-H I-H O I-H 1— a ^ ^ C3 ^-_,^-^^ si '^— N^-— a a a 3 E 3 3 3 Nl s~. bo bo Ul HI bO bO ta C3 =3 -1 P3 03 3 a -S cc _>;> n o o! O 3 O *^ o CJ c Z) 2 o, > a a; a ;3 > 0 d -^ ? cu 1 c > a ci (U ^ a :» 3 Ul c: t: *^ -*^ fl >- OT ?S •*^ 3 o ei c o 0 o n, — 03 o o -M o o c. o 1 s 1 !» ffl 1 U 1 n 1 s 02 K o « M CU o + a V o -«-> > a « 3 60 Ml + 01 ■■-^ (a bO 57 58 o. iSHii medium is not suitable for the agglutination test for Bad. paratyphosum A. Occasionally spontaneous agglutination in pepton water cultures is difficult to prevent, while in plain broth it is relatively easy to prevent; in pepton water there is but a slight tendency to spontaneous agglutination even with non-spontaneous agglu- tinating bacilli. Specific agglutinations with pepton water cultures were slightly weaker than with plain or glucose broth cultures. With formalin the reaction was slightly stronger than without. With agar cultures the agglutination reaction was generally slow and weak and appeared somewhat incomplete, dependmg however upon the strain and age of the culture. With or without formalin almost the same results are obtained with non-spon- taneous agglutinating bacilli. Bact. paratyphosum B. Glucose broth cultures of Bad. para- typhosum B usually show weak motihty and strong spontaneous agglutination, which is not prevented when treated with formalin ; even non-spontaneous agglutinating bacilli give the same result. This medium does not appear to be satisfactoiy for the Widal test with Bad. paratyphosum B. Cultures in plain broth and pepton water showed more vigor- ous motility and ahnost the same agglutinating reaction with specific serum. But in pepton water with non-spontaneous agglutinating baciUi, there was a slight tendency to spontaneous agglutination even when the cultures were mixed with salt solu- tion or with formalin. It appears that plain broth cultures are more reliable than are those in pepton water. The agar cultures showed weak specific agglutination m com- parison with the other three kinds of media, but did not have the tendency to spontaneous agglutination. Bact. call. Glucose broth cultures of most strains of Bad. coli grown for eighteen to twenty-four hours at 37°C. showed con- siderable spontaneous agglutination with both spontaneous and non-spontaneous agglutmating bacilli; hence this medium can- not be used for the specific agglutination test. STUDIES UPON AGGLUTINATION 59 Certain strains in pepton water cultures showed a slight ten- dency to spontaneous agglutination, while in broth they showed no spontaneous agglutination. The specific agglutinations were weaker in pepton water than broth medium; therefore I consider the latter most suitable for the Widal test with Bad. colt. Agar cultures show very weak agglutination, but there is no tendency to spontaneous agglutination with Bad. coli. Bad. (hjsenteriae group. This group showed strong specific agglutination in plain broth cultures, it was slightly weaker in pepton water and agar and especially weak in glucose broth; but glucose broth did not show spontaneous agglutination. TABLE 9 Relation of age to specific agglutinating power with non-spontaneous agglutination hacilli, grown on broth culture medium Oroaxism BACT. BACT. PARA A BACT DTSENTERIAE ~ f- O o Number of strain i 1 55 CO 3 1 6 1 o ea .SS CO 0) a M r 1 IS .a a 55 a 1 day 5 5 5 15 5 5 5 15 5 5 5 5 20 2 Ages of cultures and 2 days 5 4 5 14 5 4 4 13 5 5 5 3 18 3 specific serum 3 days 4 3 4 11 4 3 3 10 4 4 4 2 14 4 5 days 3 2 3 8 3 2 2 7 3 3 2 1 9 5 Key: 5, complete; 4, almost complete; 3, weak; 2, very weak; 1, trace; 0, nega- tive agglutination. relation of agglutination power to age of cultures (table 9) Dreyer claims that cultures more than one day old had a weak agglutinating power with specific serum. My experiments for a period of five days, with Bad. tijphosum, Bad. paratyphosum A and the Bad. dysenteriae group, indicated that cultures on neutral broth medium one day old were agglutinated best, the reactions become weaker each day thereafter. However, I found that one strain of Bad. dysenteriae Shiga no. T showed just the opposite result, the agglutinating power becoming stronger with older cultures over a period of five days. 60 O. ISHII spontaneous agglutination according to age of cultures (table 10) WTien spontaneous agglutinating bacilli were grown in neu- tral broth or pepton water, spontaneous agglutination was more vigorous when cultures were but one day old, diminishing steadily afterwards and almost disappearing on the fifth day. I concluded therefore that both spontaneous and specific agglu- tinating power diminsh with the age of most strains of the colon- typhoid group. TABLE 10 Spontaneous agglulinaiion according to age oj bacilli growing in plain broth at 37° C. BACT. Number of strain d 3 2 1 0 K 6 Z 3 2 1 0 00 6 2; 3 2 2 1 1 3 2 1 0 i 2 2 1 0 CD 1 3 2 1 0 i 3 2 1 0 C4 i 3 3 2 1 6 Z 9 Ages of cultures < 1 day 2 days 3 days 5 days 3 2 1 0 26 19 11 2 Key: 3, strong; 2, medium; 1, weak spontaneous agglutination. COMPARATIVE RESULTS OF MICROSCOPIC AND MACROSCOPIC AGGLU- TINATION TESTS Chick, Dreyer, Garrow, Jordan, Panton, Walker, Wilson, and several other investigators prefer the macroscopic method; Delepine, Ritchie, and several other workers have used the microscopic method. In using non-spontaneous agglutinating cultures, strong ag- glutination could be seen equally well with either method, and weak agglutination better by the microscopic method. For instance, in some cases which appeared negative macroscopi- cally, small clumps could be seen microscopically when examined at the end of three hours. These small clumps are of great value in certain cases of very weak reactions. This is especially noticeable with the Bad. dysenteriae group, but less so for STUDIES UPON AGGLUTINATION 61 BacL typhosum, Bad. paratyphosum A and B, and the Bad. coll group. In the case of Bad. typhosum and Bad. paratyphosum A and B, spontaneous apghitinalion generally appears weak macroscop- ically and strong microscopically; but with Bad. coli it appears to be almost the same with either method. It is specially to be noted that when using spontaneous agglutinating bacilli for the Widal test, the macroscopic method was found more reliable. For the fundamental or graphic agglutination test, I used non- spontaneous agglutinating bacilli. ^Tien agglutination was TABLE n Agglutination tests at different lemperatvres ORO ANISH BACT. BACT. BACT. BACT. COLI BACT. Number of strain o 5 5 4 3 6 Z 5 5 4 3 6 y, 5 5 4 2 15 i 5 i 5 5 4 4 6 2 5 5 4 3 5 o i- 15 15 12 11 d Z 5 5 4 3 6 !S 5 5 4 3 « 6 y. 5 5 4 3 ec *« I 15 15 12 9 6 5 5 4 4 6 5 5 4 3 6 2; 5 5 4 3 1 E- 15 15 12 10 to 5 5 4 3 s 5 5 4 3 s 5 5 4 3 era 3 With each Room tempera- ture 15 specific serum 37°C. water bath 45°C. water bath 55°C. water bath 15 5 12[ 4 8| 4 15 12 9 Controls of all 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Key: 5, complete ; 4, almost complete ; 3, weak ; 2, very weak ; 1, trace ; 0, nega- tive agglutination. weak or the case questionable, I resorted to the microscopic method for final determination. AGGLUTINATION AT DIFFERENT TEMPERATURES (TABLE 11) Delepine, Jordan, Konrich, Lion, Meyer and Kilgore, Widal, and many others conducted agglutination tests at 37°C. ; Dreyer and Blake, Hetsch, Kolle, Kutscher, Forges, and Weil advise the use of 50°C. to 55°C. ; Joos 35°C. to 40°C. ; BerUner and Cohn, and Durham recommend room temperature. Recently many workers have read the test, after keeping the tubes in the water bath two to three hours at from oO°C. to 55°C. then leaving them at room temperature or in the cold room over night. 62 o. rsHii In my observations on the effect of temperature of the agglu- tination with Bad. typhosum, Bad. paraty-phosum A and B, Bad. coli, and the dysentery group, the tubes kept in a water bath for two hours at 45°C., showed a slightly weaker reaction than those kept at room temperature or 37°C., while at 55°C., the reaction was much weaker than at 45°C. The results, after leaving the tubes for 2 to 3 hours, or even for twenty-four hours at room temperature, were just the same as they were when the tubes were placed in the water bath at 37°C. for two to three hours and then left standing at room temperature over night. TABLE 12 Agglulinalion tests conducted with bacilli after heating at 55°C.for two hours Organism BACT. BACT. BACT. BACT COT.I BACT. CO ec en CO eo Number of strain d d d ^ d d 6 ■3 d d CO d s 6 n 6 a a 1 S 5 /5 5 5 Hi* 5 5 H Z Z 5 5 15 5 5 5 15 5 5 K 5 H With specificf Unheated 15 5 15 5 15 serum \ 55 °C. 2 hours 3 3 3 9 3 2 2 7, 3 4 3 10 3 4 3 10 2 3 2 7 Control of bo th 0 0 0 0 0 0 0 n n 0 0 0 0 0 0 0 0 0 0 0 Kejj: 5, complete; 4, almost complete; 3, weak; 2, very weak; 0, negative agglutination. In the next experiments (table 12), bacilli were heated at 55°C. for two hours before conducting the agglutination test; it became evident that these bacilU had less agglutinating power than the unheated bacilli. On the other hand, the serum which had been heated at 55°C. for two hours gave just the same re- sults as serum which had not been heated. When temperatures from 45°C. to 55°C. were used, the clumps were smaller than when the test was conducted at room tempera- ture or 37°C. This was true when either the microscopic or macroscopic methods were employed. The latter temperature is therefore to be preferred, ^ith the colon-typhoid group with or without formalin; the end results, in relation to the temperature, are just the same. STUDIES UPON AGGLUTINATION 63 THE TIME FACTOR IN THE AGGLUTINATION TEST Afrglutination tests after various periods of time were carried out with the colon-tyiihoid group, in relation to certain titres of the serum. Bad. typhosinn showed the reaction quicker than the other organisms, but the aggkitination reaction of most of the strains was completed only after standing over night, instead of after three hours. However, some strains showed almost complete agglutination in three hours; others only a trace in this time. Of course with strong seioim, agglutmation took place soon after mixing with the bacterial suspension. With Bad. paratyphosum B, the results were of the same order, but less rapid than with Bad. typhosrim. Bad. coli comes after Bad. parnlyphosum B, with regard to the speed of the reaction. Bad. paratyphosum A, and Bad. dysenteriae show relatively veiy slow agglutination; in most instances, the reaction was definitely obtained only after standmg over night. However, the speed of the reaction varies with the strains of bacilli ; some strains were always quickly agglutinated and others more slowly. THE USE OF SEVERAL STRAINS FOR THE WIDAL TEST (TABLE 13) In the Widal test most bacteriologists use only one strain of bacillus with the serum of the patient. This method does not give reliable results. Ordinarily we use polyvalent immune serum for the identification of a bacillus; for the same reason we should use several strams in testing the serum of a patient. A serum agglutination reaction depends upon the organism infecting the patient. If the serum agglutinates the stock cul- ture, it indicates a close relationship between this organism and the organism which infected the patient; a negative test indi- cates distant or no relationship. In a case of typhoid fever (P. B. G. Hospital) the serum showed positive agglutination for three strains (nos. 1, 4, 33), but negative for three others (nos. 3, 7, 29), and doubtful for one more (no. 2j. 64 o. rsHii In performing this test, therefore, I have found it convenient to grow each strain (6 or more) of this organism in broth medium at 37°C., for eighteen to twenty hours. Formaliu 0.1 per cent is added to each of the cultures, which are then all mixed to- gether, and are ready for immediate use; or they may be stored in a cold room until needed, as m Dreyer's method. This prep- aration gives reliable results in diagnosis with an unknown serum. TABLE 13 The serum of typhoid fever, Widal test with different strains BACT. TTPH08DM 7 STRAINS No. 1 No. 4 No. 33 5 4 3 1 0 No. 2 No. 3 No. 7 0 0 0 0 0 No. 29 Serum dilutions of typhoid fever 1:25 1:50 1:100 1:150 1:200 5 5 4 3 1 + 5 4 3 o 0 3 2 1 0 0 1 0 0 0 0 0 3 1 0 0 0 Results + + ? 0 0 Key: 5, complete; 4, almost complete; 3, weak; 2, very weak; 1, trace; 0, negative agglutination; +, positive reaction. SUMMARY 1. Formalin 0.05 to 0.2 per cent prevents spontaneous agglu- tination almost entirely in some instances and entirely in others with Bad. iyphosum and Bad. paratyphosum A and in a lesser degree with Bad. paratyphosum B; with Bad. coli formaUn has no effect in this regard. 2. One-tenth per cent formalin prevents spontaneous agglu- tination and increases specific agglutination, ha\'ing a strong effect on Bad. typhosum and Bad. paratyphosum A, but less on Bad. paratyphosum B and Bad. dysentenae and none on Bad. coli. 3. Formalin prevents pseudo agglutination to a great extent in cross agglutination test with different sera. 4. Cultures in acid medium in which there is heavy growth and active motiHty, yield stronger agglutination than cultures STUDIES UPON AGGLUTINATIO>f 65 in neutral broth with Bad. iyphosiim and Bad. paratyphosum A; this effect is less noticeable in cultures of Bad. paratyphosum B, Bad. dysentcriae and Bad. coli, but the agglutinating power of these organisms is very weak in alkaline media. Alkaline media are therefore not suitable for the agglutination test with cultures of these bacteria. 5. Agar cultures emulsified with 0.85 per cent salt solution or neutral broth are almost equal in susceptibility to agglutination, but certain strains have a tendencj^ to spontaneous agglutina- tion in broth medium; therefore salt solution is better. 6. As to the effect of sodium chlorid on agglutination reactions I could find no difference in the degree of agglutination reaction when using strong or weak solutions of sodium chlorid; only traces are necessary for agglutination. 7. Glucose broth cultures of Bad. paratyphosum A and B and Bad. coli showed a tendency to spontaneous agglutination; Bad. typhosum showed weaker specific agglutination in glucose broth, but stronger than in pepton water, only occasionally show- ing spontaneous agglutination. The dysentery group yielded weaker agglutination with glucose broth cultures than with plain broth, pepton water and agar cultures. Pepton water cultures in general show a slight tendency to spontaneous agglutination, when non-spontaneous agglutinating bacilli are growTi therein. This spontaneous agglutination is difficult to prevent even when formalin is used and the specific agglutination reactions of Bad. typhosum, Bad. paratyphosum A and B and Bad. coli are weaker than in plain broth or glucose broth. On the other hand the dj^sentery group j-ielded stronger specific agglutination reactions in this medium than in glucose broth. Agar cultures showed much weaker agglutination than broth and pepton water cultures. In general, plain broth cultures gave the best results. 8. Susceptibility of most cultures to agglutination decreases with age ; eighteen to twenty-four hours cultures were found best suited for agglutination tests. J017BNA1. or BACTERIOLOOT, VOL. TU, NO. 66 o. isHn 9. The microscopic is more reliable than the macroscopic method for weak or graphical agglutination work ; with strongly agglutinating sera showing the same reactions in both, the macro- scopic method is preferred, because it is easier and quicker than the microscopic method. Therefore the macroscopic method is to be regarded as most satisfactory for usual work and particularly when many reactions are to be performed. 10. Weak agglutination results if bacilli are heated at a tem- perature of 45 to 55°C. for two to three hours; this temperature does not affect the agglutinating activity of serum. Room temperature and 37°C., both gave better results than higher temperatures for the incubation of tests. 11. Two to three hours are not enough for complete agglu- tination; the tests should be set aside and read the next day. 12. Normal horse serum showed strong agglutination reaction with Bad. dysenteriae, Bad. coli and other bacilli of the tj^ihoid- colon group. Polyvalent Bad. dysenteriae serum had strong agglutinating power for Bad. typhosum and Bad. coli as well as for the dysentery group. 13. In serum diagnosis with the Widal test, many different strains should be used since from one strain alone, rehable re- sults cannot be obtained. SUPPLEMENTARY NOTES a. Eighteen to twenty-four hours cultures of the majority of strains of Bad. coli grown at 37°C., show a tendency to spon- taneous agglutination in most culture media; cultures growTi at room temperature show less and are therefore preferable. b. Old stock cultures transplanted to a fresh culture medium at first show weak, or no, susceptibility to agglutination; how- ever, after a few sub-cultures (once or twice a day, if actively growing bacilli), they regain this property. The same holds good for all cultures of new bacilli, and it is necessarj^ therefore, for accurate agglutination work to test the agglutiuabihty of the bacterial culture with a control serum of known agglutinat- ing powers. STUDIES UPON AGGLUTINATION 67 c. Most cultures grown at 37°C. for eighteen to twenty-four hours show a slight pellicle on the surface; to avoid mistaking this for spontaneous agglutination the cultures should be shaken and left at room temperature for a while, after the pellicle falls to the bottom of the test tube. d. Heavy growths yield better results in the Widal test when diluted 2 or 4 times with salt solution. e. Acid broth (0.2 per cent) cultures should be used for routine agglutination tests inasmuch as acidity is useful and greatly helps in weak agglutination. /. It is well known that commercial formaUn contains formic acid. Fearing that the formic acid would interfere with the Widal reaction, I have tried certain strains of Bad. typhosum and Bad. paratyphosian group with various dilutions of formic acid (0.1 to 10 per cent) and found that with these dilutions weak chemical agglutinations are produced, but the specific agglutination is diminished. With a weak solution of formalin (0.1 to 1 per cent) no chemical agglutination is produced and spontaneous agglutination is prevented, as above stated. Evi- dently the quantity of formic acid in these concentrations of formalin is so small that it does not interfere with the reaction. Commercial formalin gave satisfactory results throughout this work. g. The addition of phenol (0.1 per cent) or corrosive subli- mate (0.01 per cent) did not change normal agglutination power. Tricresol (0.1 per cent) slightly weakened agglutination in the colon-typhoid group. All substances produced weak or nega- tive agglutination reactions when used in more concentrated solutions than those above mentioned. This investigation was made possible by a grant from the National Canners xVssociation and has been carried out under direction of Prof. JM. J. Rosenau, of the Department of Preven- tive Medicine and Hygiene, Harvard Medical School, Boston, Mass. 68 o. isHii REFERENCES Aaser, p. 1905 Ueber die Makroskopischeagglutinationsprobe bei Typhoid- fieber. Berl. Klin. Wochnschr., 42, 256-260. AsAKAWA, N. 1903 Ueber das Wesen der agglutination und eine neue Methode die agglutination schnell zu beobachten (Gefriemethode). Ztscbr. F. Hyg. u. Infections Krankn., Leipz, 45, 93-96. Bass, C. C, and Watkins, J. A. 1910 A quick macroscopic typhoid agglutina- tion test. Arch, in Med., 11, 717. Beco, L. 1899 Note sur la Valeur de I'Agglutination par le Serum Antity- phique Experimental Conne Moyeh de Diagnostic entre Lebacille d'Eberth et les races coliformes. Centralbl. f. Bact. u. Parasite. E. A., 26, 136. Behlineb, A., and Cohn, M. 1900 Klinische Beitrage zur Diagnose die Abdom- onal Typhus. Mun. Med. Woch., Sept. 11. Block, E.B. 1S99 Technique in serum. 1899 British Med. Jour., 11. BoRDET, J. 1899 Mecanisme De L'agglutination. Ann. Past., 13, 225. Buxton, B. H., and Vaughan, V. C, Jr. 1904 On agglutination Jour. Med. Res., 7, 115. Chick, H. 1916 The preparation and use of certain agglutinating sera. Lan- cet, London, April 22, p. 857. Coles, A. C. 1916 An easy and rapid method of doing Widal's reaction for typhoid. British Med. Jour., 684. Delepine, S. 1896 On sero diagnosis of typhoid fever. Lancet, London, December, p. 1586. The technique of serum diagnosis with special reference to typhoid fever. British Med. Jour., 1, 967. Dheyer, G. 1909 Widal's reaction with sterilized cultures. Jour. Path, and Bact., 13, 331. Dbeyer, G., and Blake, A. J. 1906 On the agglutination of bacteria. Jour, of Path, and Bact., 11, 1. Durham, H. E. 1896 On a special action of the serum of highly immunized animals. Jour, of Path, and Bact., 4, 13. Ehrsam 1904 Ueber das Fickersche Typhusdiagnostikum. Munch. Med. Wochenschr., 662. EiSENBEBG, P., und Volk, R. 1902 Untersuchungen uber die agglutination. Zeitachr. F. Hyg. inf. Krankn., 40, 155. Flatau, G., and Wilke, A. Ueber Fickers Tsiihusdiagnostikum. Much. Med. Woch., 52, 110. Garrow, R. p. Clinical agglutinometer, 1917, Lancet, London, 17, 262. Gasiakowski. Hand B. Von Kolle Wassermann. 2 Aufi., 2, 543. Grumbaum, a. S. 1898 The agglutination action of human serum for the diag- nosis of enteric fever. Lancet, London, 1898, 2, 806 and 1747. Hetsch. 1904 Ueber die Deflerenzierung der Wichtigsten, Infektionsserreger gegenuber ihnen Wahestehenden Bacterien. Klin. Jahrb., Jena, 12,73. Hiss and Zinsser. 1917 Text-book of bacteriology. IsHii, O. 1920 The fate of bacteria of the colon-typhoid group on carbohydrate media. Jour, of Bact., 5, 437. Joos, A. 1902 Ueber den Mechanismus der agglutination. Zeitschr. f. Hyg. inf. Krankn., 40, 203. STUDIES UPON AGGLUTINATION 69 Jordan, E. O. 1916 Text-book of general bacteriology. Ker, C. B. 1897 The serum diagnosis of typhoid fever. British Med. Jour. 2,1723. Kligler, I. J. 1918 Note on cross agglutination of B. coli communis and B. dysentery Shiga. Jour, of Bact., 3, 441. KoLLE, W., CND Turner, G. 1897 Ueber den Fortgange der Rinderpestfore- chungen in Kochs Versuchsstation in Kimberley. Deutsche Med. Wochenschr. Kl. Jahrb., 11, 793. KoNRiCH. 1908 Ueber den Einflss Von Warme und Zeit auf den ablauf der agglutination. Centralbl. f. Bact. und Parasite. E., 48, 92. Krumbaar and Smith 1918 Repeated agglutination test by the Dreycr method in the diagnosis of enteric fever in inoculated person. Jour. Inf. Dis., 23,126. KuTSCHER, K. H. 1900 Eine Fleischvergeftungscpideniic in Berlin infolge Infektion mit dem Bacterium Paratyphic B. Zcitschr. f. Hyg. inf. Krankn., 55, 331. Lbntz. 1909 Hand b. Kolle Wasserinann 1. Erg. Lentz and Park 1914 Path. Micro. Organism, Park and Williams. Lion, A. 1904 Die Methoden zur Ausfuhrung der Grumber-Widalschen Reac- tion. Munch. Med. Wochenschr., 908. LoELE, W. 1906 Die agglutination in den handen des praktischenarztes. Deutsche Med. Woch., Leipzig, 140. Malvoz, E. 1897 Recherches sur I'agglutination des bacillus Typhosis par des substances clinique. Ann. Past., 11, 582. Martineck 1905 Ein fiir die Praxis geeignetes Besteck zur Austellung der Gruber-Widal Reaction mit dem Kickerrschen Typhusdiagnostikum. Munch. Med. Woch., 701. Meyer, K. F., and Kilgore, E. S. 1917 The agglutinins and complement fixing antibodies in the serum of person vaccinated against typhoid fever. Arch, in Med., 293. Neisser 1904 Neue diagnostische und therapeutische Methoden beim Typhus. Berl. Klin. Woch., 41,821. Panton, P. N. 1916 The technique of the agglutination test. Lancet, London, 750. Park W. H., and Williams, A. 1914 Path, microorganism. 1914. Park. W. H. The differentiation of typhoid and coli bacilli, 1897, British Med. Jour., 2, 1778. Porges, O. 1905 Ueber die agglutinabilitat der kapselbakterien. Wien.Klin. Wochenschrift, 18, 691. Pboscher, E. 1902 Zur austelling der Widalschen Reaction. Centralbl. f. Bact. und Parasite, e, 31, 400. Reiit, L. 1900 Contribution a I'Etude de la Fievre Typhois et de son bacille. Ann. Past., 14, 555. Ritchie, T. R. 1916 On the agglutination reaction of the bacilli of typhoid dysentery group with normal sera. Lancet, London, 1257. RoLLY 1904 Zur Diagnose des Typhus abdominalis. Munch. Med. Woch., 1041. 70 o. isHii ScHBLLEB, R. 1904-05. Experimen telle Beitrage Zur Theorie der agglutination, normal agglutinine. Centralbl. Bact. und Parasite. 36, 427 and 694. 38, 100. Sblteb 1905 Zur Typhusdiagnose mettelst des Typhusdiagnostikums Von Ficker. Miinch. Med. Woschenschr., 108-110. Tabchitte, C. 1898 Contribute alio studio della sierodiagnosi nel'infezione tifoide. Gazzette degli Ospedali, Milano, 2nd Sem., 19, 1411-1414. Walker, E. W. A. 1916 A note on Widal's reaction with standardized agglu- tinable cultures. Lancet, London, 1, 17. Weil, E. 1904 Ueber den einfluss der temperatur auf die spezifische und nicht spezifische Agglutination. Centralbl. f. Bact. und Parasite. Orig- inale, 1st Abt., 36, 677. Ueber den Mechanismus der Bacterienagglutination durch gelatine. Ibid., 37, 426. Weiss, H. 1917 Cultural and antigenic differences in strain of Bacillus typho- sus and studies in the paratyphoid group. Jour. Med. Res., 31, 135. WiDAL 1896 On the sero-diagnosis of typhoid fever. Lancet, London, 2, 1371. WiDAL 1896 Sero-diagnosticdelaFievreTyphoide, Jour.de Med. etChirurgie, Paris, 67, 533. WiDAL AND ^foBECotJRT 1897 Scrum reaction in parapcolibacillary infection. Comptes Rendus Soc. de Biologic, Paris, 842. Wilson, G. S. 1917 The effect of pyrexia on inoculation agglutination. Lan- cet, London, 263. Wbetowski, T. 1908 Zwei neue Agglutinationmethoden. Centralbl. f. Bact. und Parasite. 1 Abt. 47, 513. A STUDY OF SPONTANEOUS AGGLUTINATION IN THE COLON-TYPHOID GROUP OF BACILLI 0. ISHII From the Department of Preventive Medicine and Hygiene, Harvard Medical School, Boston Received for publication April 19, 1921 It is well recognized that spontaneous agglutination may occur in broth cultures of Bad. typhosum and Bad. paratyphosum A and B. Because of this phenomenon, the results of agglu- tination tests with specific sera are not reliable under such con- ditions, since it is practically impossible to distingush between spontaneous and specific agglutination. Block in 1897 found that cultures which were transplanted too frequently, may agglutinate spontaneously. He further observed that when cultures were grown in alkaline broth the same phenomenon occurred. Delepine and Fison (1897) also noted spontaneous agglutination in cultures of Bad. typhosum. Kruse, Rittershaus, Kemp and Metz described spontaneous agglutination of typhoid and dysentery bacilli in plain broth cultures and in very concentrated peptone broth cultures. Nicolle (1898) reported that changes in the bacterial protein occurred in old cultures, and that the organisms became more sensitive and readily agglutinated spontaneously in such cul- tures. Smith and Reagh (1903) reported that the Bad. ider- oides developed granular colonies on gelatin cultures; i.e. were agglutinated spontaneously. Steinhardt (1904) found that in agar cultures, showing spontaneous agglutination, the colonies were irregular in shape and less translucent than the colonies of the type which did not agglutinate spontaneously. Teague and McWilliams (1917) reported that organisms isolated from the blood of a rabbit which had been injected with Bad. typhosum, 71 72 o. isHii produced different kinds of colonies when grown on agar plates, some being opaque and others transparent; in addition there were also differences in their contours, some having irregular outlines while others were round with smooth edges. Further- more some colonies were small and others large. APPEARANCE OF SPONTANEOUSLY AGGLUTINATED COLONIES ON AGAR PLATES Plating many stock cultures upon two per cent agar plates we have found two distinct types of colonies. The first of these presents a smooth surface and a regular outline, and is not spontaneously agglutinated. The other type shows a much heavier growth, has a rough svirface, is irregulariy round, with a serrate border and is much more transparent, both at its cen- ter and border, than the first type of colony. That most of the colonies of this second type undergo spontaneous agglutination, can be noted by means of the low power microscope or hand lens. In some instances, upon sub-culturing, certain strains of those colonies which previously had been transparent became opaque and vice versa; or an organism originally producing small colonies produced large ones. These points of differentiation are not so important as those relating to the outUne of the colony and smoothness or roughness of the surface. The available space for growth on the plate as well as the rate of growth, appear to influence these characteristics. SPONTANEOUS AGGLUTINATION IN BROTH CULTURES (TABLE 1) Steinhardt, Teague and IMcWilliams found in broth cultures of Bad. typhosum, which underwent spontaneous agglutmation, flocculi or pellicle formation as contrasted with a uniform, cloudy growth of other organisms. Smith and Reagh showed that a broth culture of Bad. ideroides became clear after spontaneous agglutination had occurred. We have observed spontaneous agglutination of Bad. typho- sum, Bad. paraiyposum A and B, Bad. enteritidis, Bad. dysenteriae and Bad. coli in pepton water, glycerol broth and glucose A STUDY OF SPONTANEOUS AGGLUTINATION 73 broth. Tn many cases the broth was clear macroscopically, but the clumps were visible when a hanging drop was examined. I\Iore or less of the precipitate collected in the bottom of the tubes. In some instances flocculi and a pellicle were visible on the surface of the culture, and there was always a precipitation in the bottom, either spontaneous or following agitation of the tube. In other cases, small clumps could be seen with the naked eye scattered throughout the culture. With the tjTihoid and paratj'phoid bacilli there was a uniform cloudy growth and apparently an absence of spontaneous agglutination when TABLE 1 Varialionn of growth in broth medium. Cultures spontaneously agglutinated show clear supernatant fluid, precipitation, pellicle, flocculi or small clumps, others appear uniformly cloudy, to the naked eye like cultures not spontaneously agglutinated ORGANISMS GROWN IN BROTH PRECIPITATION, PELLICLE AND FLOOCOLI CNirORMLT CLOUDY TOTAL NUMBER OP STRAINS Bact. typhosum 4 7 15 5 12 9 1 0 16 Bact, paratyphosum A 16 Bact. paratyphosum B 16 Bact. coli 5 A tabulation of our experiments shows the following results: Bact. pa-^a- typhosum B showed precipitation with formation of pellicles and small clumps in clear fluid in all cases with the exception of one strain. Bact. coli, 5 strains exhibited this same phenomenon in every case. Bact. paratyphosum A showed 7 positive and 9 negative, Bact. typhosum 4 positive and 12 negative tests. examined with the naked eye, but with a hand lens, flocculi were clearly visible in the supernatant fluid. We have, therefore, been led to conclude that in broth cul- tures when flocculi are visible, macroscopic observation may be relied upon; but in cloudy cultures, microscopic examination should always be employed. In twenty-four hour cultures at 37°C., of microorganisms which do not undergo spontaneous agglutination, the growth in pep- ton water, glycerol or glucose broth, whether the reaction of the media be acid, neutral or alkaline, is generally uniformly cloudy. In some cases, however, there is a peUicle and SdccuU are seen near the surface of such cultures. 74 O. ISHII , Microscopic examination of the twenty-four-hour broth cul- tures by the hanging drop method showed differences in morphol- ogy and motility between those organisms which agglutinated spontaneously and those which did not. In most cases the bacilli which undergo spontaneous agglutination are longer, sometimes having the appearance of being fungiform and show- ing relatively greater motility. The bacilli which do not agglu- tinate spontaneously are shorter and are less motile. COLONY CHANGES IN ARTinCIAL CULTURE MEDIA (TABLE 2) LoflHer (190G) found four types of colonies of Bad. coli: (1) transparent, (2) flat, (3) thick, and (4) opaque. According to his evidence, these types are constant and do not change. Baerth- lein (1911) reported a cholera vibrio colony that did not change after passage through animals, but in a stock culture twenty- two days afterwards, in several subcultures, the yellow type turned light and twenty-eight days later the light colony changed to a yellow one. Steuihardt found spontaneous agglutination in Bad. typhosum after the twentieth passage through bacteri- cidal serum cultures. Teague and Mc Williams state that they found some changeable bacilli, but they do not draw any definite conclusions from their finding. According to our observation colonies of many strains of the colon-typhoid group may lose the property of spontaneous agglutination, while others that do not show this property at first may show it after some time. From a stock culture we obtained colonies which were, and others which were not, agglutinated spontaneously. These were spread on plates of two per cent agar (using very high dilution of the bacteria in broth or salt solution so that each single colony should be separated from every other colony) and incubated for 24 hours at 37°C. The colonies thus grown were inoculated into broth tubes and placed in the incubator. Sub- cultures were made every two days by inoculating one loopful of the broth culture into a fresh tube of broth. The bacilli were examined after each sub-culture by streaking a loopful of the freshly inoculated broth on agar plates. The following results on agar i)lates and in cultures were obtained: A STUDY OF SPONTANEOUS AGGLUTINATION 75 TABLE 2 Broth cultures of the isolated colony were kept at syC, and transplanted every second day, then examined for any change of colonics by streaking on agar plates OROANIBM NDHBEB or STRAINS TYPK OFCOI.O.NY NIIMflRR OP DATS. WHEN CHAN(JK FOUND. l(KP»ATKn ONB TO TOREK TIMES No. • { Non-spont. aggl. Spont. aggl. 24 10 35 13 Bact. lyphosum No. = { Non-spont. aggl. Spont. aggl. 6 IG 6 14 13 5 No. 29 1 Non-spont. aggl. Spont. aggl. 32 50 12 48 20 47 No. 1 I Non-spont. aggl. Spont. aggl. 4 4 4 4 3 4 Bact . paratyphosum A No. 2 1 Non-spont. aggl. Spont. aggl. 3 -30 3 3 No. 3 { Non-spont. aggl. Spont. aggl. 7 12 6 18 8 15 No. 4 < Non-spont. aggl. Spont. aggl. 5 13 5 25 4 Bact.paratyphosum B No. = { Non-spont. aggl. Spont. aggl. 5 -60 10 No. « { Non-spont. aggl. Spont. aggl. 5 24 8 25 12 No. 1 Non-spont. aggl. Spont. aggl. 2 2 3 3 3 2 Bact. coli No. 2 1 Non-spont. aggl. Spont. aggl. 7 4 10 No. = { Non-spont. aggl. Spont. aggl. 10 30 13 28 Bact. enteritidis < No. ^ { Non-spont. aggl. Spont. aggl. 24 16 76 O. ISHII TABLE 2— Continued ORGANIBM NUMBER OF STRAINS TYPE OF COLONY NUMBER OF DAYS, WHEN CHANGE FOUND. REPEATED ONE TO THREE TIMES Bact. enteritidis No. 3 1 No. 7 1 Non-spont. aggl. Spont. aggl. Non-spont. aggl. Spont. aggl. 7 5 5 5 Bacl. dysenteriae Shiga No. 1 1 Flexner No. 1 < Non-spont. aggl. Spont. aggl. Non-spont. aggl. Spont. aggl. 22 4 8 17 5 Key, — No change. (a) Different organisms showed a differentiation into two types of colony at different periods of growth, Bact. coli showing a change in from 2 to 30 days. (6) Bact. paratyphosum A showed the change within 3 to 14 days (some strains however, remaining constant even after 30 days). (c) Bact. paratyphosum B showed a transition period of from 4 to 30 days, while some strains did not show the spontaneously agglutinated type of colony even after 60 days. (d) Bact. dysenteriae group showed changes in from 4 to 22 days (some strains bemg negative after 30 days). (e) Bact. enteritidis changed in from 5 to 24 days. (0 Bact. typhosum showed the longest period, the time vary- ing from 6 to 50 days in some strains, whUe others did not show changes even after 3 months or more. From colonies of known type (as determined above) isolated on agar plates, the different members of the colon-tj^phoid group were inoculated into broth media and allowed to grow at room temperature without sub-culture, plates being made every five or ten days to determine the type of colony. In 33 strains of Bact. typhosum of both types, changes were observed to take place in 30 strains in fifteen to ninety days, while three strains, A STUDY OF SPONTANEOUS AGGLUTINATION 77 originalh' of a sjiontancou-sly agglutinating variety, showed no change after one hundred days. Eighteen strains of Bad. para- typhosum A showed changes in from ten to eighty-five days, two strains of an agglutinating type, showed no change until after one hundred days. Nineteen strains of Bad. paratyphosum B. showed changes in from fifteen to one hundred days, while two strains of spontaneously agglutinating type did not change even in one hundred days. All strains of Bad. coli showed a change in from ten to thirty days. Bad. enteritidis was examined after GO days growth, and all 8 strains showed a change. With the Bad. dysenteriae group two strains showed changes in thirty- five to forty days, while two strains showed no change after sixty days. The period at which the change in colony occurs depends upon the strains; some strains always change within a short time and others only after many days. ISOLATION OF TYPES OF COLONY FROM STOCK CULTURES (TABLE 3) In our first work with stock cultures three types were found. These were, pure colonies not showing spontaneous agglutina- tion, pure colonies showing spontaneous agglutination and mixed colonies showing the presence of both types. In a study of the colon-typhoid group of bacilli from agar stab cultures which had ordinarily been sub-cultured at inter- vals of two or three months, there were found many cultures producing colonies of only one kind, namely those which agglu- tinated spontaneously; other cultures yielded colonies of bacilli all of which failed to agglutinate spontaneously. In the present work extending over a period of a year, it was found that each pure type after frequent transplantation tended to change into both types; of 29 strains of Bad. typhosum show- ing pure colonies of the spontaneous agglutinating type only 3 strains remained unchanged. Of 4 originally pure strains of Bad. coli used only 2 adhered to the original type. Bad. para- typhosum A (19 strains) Bad. enteritidis (9 strains) and Bad. dysenteriae (3 strains) showed both types of colonies. Bad. paratyphosum B (21 strains) showed both types of colonies, but 78 o. isHn 2 strains (nos. 19 and 21), after frequent sub-culture in broth, pepton water and agar throughout a period of one year, gave 1 type of spontaneously agghitinating bacilli during the entire period. When, however, 1 per cent glucose broth was used for daily transplants, culture no. 19 showed 2 types of bacilU after seven days and culture no. 21 after eighteen days. A new strain of Bad. typhosum was isolated from the blood of a patient at the Peter Bent Brigham Hospital in the spring of TABLE 3 Isolation of colonies from stock stab cultures of agar medium First isolation' Isolation of subcultures after several months OaOANISM Bad. typhosum... Bad. paraty- phosum A Bad. paraly- phosum B Bad. coll Bad. enteritidis . Bad. dysenteriae Bad. typhosum.. Bad. paraty- phosiim A Bad. paraty- phosum B Bad. coli Bad. enteritidis Bad. dysenteriae NtJMBEB OF STRAINS SPONTA- NEOUS AOGLnTI- NATION NON- SPONTA- NEOUS AGGLUTI- NATION MIXED, SHOWING BOTH TYPES 33 4 15 13 19 5 7 7 23 4 3 16 8 1 3 4 9 1 2 6 3 — 1 2 33 3 - 29 19 - - 19 25 23 8 — 2 6 9 — — 9 3 — — 3 DOUBTFDT. AGGLUTI- NATION 1918. The cultures on agar plates gave pure smooth colonies which when transplanted to broth showed non-spontaneously agglutinated bacilli. A single colony was fished from an agar plate and transplanted successively for a period of twelve days in broth. At the end of the twelfth day it showed .the presence of bacilli of the spontaneously and non-spontaneously aggluti- nating types. Most of the colonies which developed after planting cultures freshly obtained from animals showed non-spontaneous-agglu- A STUDY OF SPONTANEOUS AGGLUTINATION 79 tination. When these were sub-cultured at later periods, the colonies showed both agglutinating and non-agglutinating tjqies. In all instances the changes in the strain was from a pure to two types of colonies. The results of this work indicate that every member of the colon-typhoid group is changeable and may develop two types of colonies, dependent upon duration of cultivation or other circumstances of growth, as well as upon the type of artificial culture medium employed. PHENOMENA OF SPONTANEOUS AGGLUTINATION IN VARIOUS FLUID CULTURE MEDIA (TABLE 4) It has been observed, that spontaneous agglutination may occur in all ordinary fluid or solid culture media. Spontaneous agglutination occurs in a greater degree in glycerol broth, pep- ton water and all acid media, than it does in neutral or alkaline broth and it is more marked in concentrated than in dilute, Uquid pepton medium. Cultures in alkaline broth usually remain clear and transparent, the presence of the alkali greatly retarding or completely inhibiting spontaneous agglutination. When glucose broth was used as a culture medium, spontan- eous agglutination of Bad. typhosum was prevented in many instances, but glucose broth cultures of Bad. paratyphosum A and B showed spontaneous agglutination in 24 hour cultures, and this has been even more marked with similar cultures of Bad. coli. Growth in plain broth, either weakly acid or neutral in reac- tion, depends upon the type of colony. In some instances there is an excessive growth as well as a spontaneous agglutination. When such a culture was shaken or mixed with broth or salt solution, it disintegrated being thus differentiated from true spontaneous agglutination, in which disintegration occurs only when special methods are employed. Plain broth or pepton water cultures which became spontan- eously agglutinated were at first clear and after twenty-four hours a sediment was forming. If a cloudy growth is encountered in sub-culture, we are sure to find both spontaneous and non- spontaneous agglutination tyi)es of organisms present. This i i. n WO^I 00 oo 00 CO T-i + Z •* -ON to 1-1 T-4 r-4 + 6 '^ ON IM CO (N CO 5 ■£ ON '^ cq IN c^ 5 'J OM Ci (N (N CO Z l ox 1 "6 s a a c c 0 "b 1 1 0 c < o 0. Ol c £ "o S 9 2 C 1- _c '5 5 ■»- a <= c c > a > t t- > a 1. c } c c ) C , 1 J 5. 1- a C > a i c ) C ) 13 O o. + tiC bO -is &0 a o 5? 80 A STUDY OF SPONTANEOUS AGGLUTINATION 81 can be determined by plating the culture and observing the two types of colonies. If flocculi, pellicles or small clumps are pres- ent in a twenty-four hour sub-culture of a non-spontaneously- agglutinating strain, we can assume that it has changed its characteristics and expect to find both types of growth present. VAHIATION IN GROWING POWER OF THE SPONTANEOUS AND NON- SPONTANEOUS AGGLUTINATION COLONIES FROM ONE STRAIN Dreyer, using stock cultures for the agglutination test, passed a strain through several sub-cultures on broth medium and then treated a vigorously growing sub-culture with formalin. He reports that agglutination reactions with such cultures are more powerful. Block (1897) claims that if too frequent transplantation of the culture is made there will occur in time spontaneous agglu- tination. This is possible only if an easily changeable strain containing spontaneous and non-spontaneous agglutination types is used. After isolating the non-agglutinating type of colony two broth cultures were made from a single colony, one of which was daily transplanted into fresh broth. After several days in the agar plate from the sub-cultures many colonies of sponta- neously agglutinating types were present, while in the first broth culture which was not transplanted, there were only a few col- onies of the spontaneously agglutinating types. Our deduction from these experiments is, that the type of organism showing spontaneous agglutination grows more vigorously than the non- spontaneously-agglutinating type taken from the strain in these experiments. On the other hand, if the growth of the non-spon- taneously-agglutinating strain is stimulated, one will find that the spontaneously agglutinating type is weakened greatly and, in some cases, entirely disappears. Experiments were made on growing the two types of culture together in broth, as follows: Broth cultures were prepared con- taining one loopful of a spontaneous and one, of a non-spon- taneous agglutination type of organism. The results as showTi on agar plates made at the end of each twenty-four hours varied widely. In some the growth of non-spontaneously agglutinating 82 o. isHii bacilli daily exceeded that of spontaneously agglutinating bacilli; others showed a more vigorous growth of the latter type, and a few remained constant, showing parallel growth of the two types. Other cultures which were not transplanted, showed a very slow change, while the daily transplanted culture showed in comparison very rapid changes in the proportion of the colony types. This varied directly with the original strains of the bacilli. SPECIFIC AGGLUTINATING POWER BETWEEN TWO TYPES OP COLONIES OF ONE STRAIN In experiments on specific agglutination with different types of colonies, no difference was found in the power of agglutination with two types of colonies of one strain. Spontaneous agglutination differs chemically from specific serum agglutination, in that spontaneous agglutination can be prevented with most strains of Bad. typhosum and Bad. para- typhosum A and to a lesser extent with Bad. paratyphosum B, by adding 0.05 to 1 per cent formalin. Formalin in 0.05 to 0.2 per cent, does not interfere with specific serum agglutination, but slightly increases it if plain salt solution (0.85 per cent) has been used. Many spontaneously agglutinating strains of Bad. typhosum when grown in glucose broth fail to show spontaneous agglutina- tion. In others spontaneous agglutination is not inhibited by this means; neither does growth in glucose broth destroy the power of specific agglutination. NOTES ON AGGLUTINATION TEST In the agglutination test as applied to members of the colon- typhoid group it is necessary to differentiate between organisms giving spontaneous agglutination and organisms which do not agglutinate spontaneously. Organisms of the first group ap- pear to give specific serum agglutination, but in the control bacillary suspension, when mixed with broth, pepton water, A STUDY OF SPONTANEOUS AGGLUTINATION 83 salt solution or tap water (the diluents usually employed), they also give spontaneous agglutination. Even old stock cultures which fail to give spontaneous agglu- tination at first, will do so, if the culture is inoculated into fresh media or repeatedly sub-cultured on agar, pepton water or broth. Spontaneous agglutination also appears in the suspen- sion of a twenty-four-hour agar culture in water, salt solution, broth, pepton water or diluted senim. If we examine an easily changeable stock culture, i. e.,a culture containing both types of colonies, and streak such a culture on the agar plate, we will find the two types of colonies described above. If upon plating we can isolate colonies of the non-spontaneously agglutinating type, only occasional sub-cultures from these colonies have to be made for the agglutination test, as such col- onies invariably fail to agglutinate spontaneously, and further tests are unnecessary. Agar cultures seem to show greater stability than fluid media cultures; some strains, however, always show one type of col- ony. For example certain strains of Bad. iyphosum which have been handled for a year and a half showed constantly on both solid and fluid media non-spontaneous agglutination. Such a strain is valuable as a stock culture for the Widal test. In many of these cultures, even after being kept for several months in an ice chest, there was only one type of colony present cor- responding with the original form, although some of the cultures were dead. These cultures were made on 0.5 per cent agar medium and the tubes were sealed with paraffin. This investigation was made possible by a grant from the National Canners Association, and has been carried on under direction of Prof. M. J. Rosenau of the Department of Preven- tive Medicine and Hygiene, Harvard Medical School, Boston, IMass. 84 o. isHii REFERENCES Baebthlein 1911 Ueber mutationsartige Wachstumserecheinungen bei Chol- erastammen. Berl. Klin. Woch, 48, 373. Block 1897 Technique in diagnosis. British Med. Jour., 11, 1777. Delepine, a. S. 1897 The technique of serum diagnosis with special reference to typhoid fever. British Med. Jour., 1, 967. Dreter, G. 1909 Widal's reaction with sterilized cultures. Jour. Path, and Bact., 13,331. FisoN, E. T. 1897 Widal's serum diagnosis of typhoid fever. British Med. Jour., 11,266. Kruse, Rittershatjs, Kemp and Metz 1907 On Dysenteriae und pseudo dysenteriae. Zeischr. F. Hyg., 57, 417. LoEFFLER 1906 Plate culture of B. coli on malachitegreen agar. Centralblatt fur Bact. u. Parasit, 38, 101. NicoLLE, M. C. 1898 Recherches sur la substance aggluttne, Ann. Past, 12, and Compt. rend, de Soc. Biol. Smith, T., and Reach, A. L. 1903 Agglutination affinities of related bacteria parasitic in different hosts. Jour. Med. Res., 6, 277. Steinhabdt, E. 1904 Variations in virulence in organisms acted upon by serum and the occurrence of spontaneous agglutination. Jour. Med. Res., 8,409. Teagub and McWilli.\m8 1917 Spontaneous agglutination in typhoid and paratyphoid cultures and its bearing upon absorption of agglutinins. Jour, of Immunology, June, 167. THE SOURCES AND CHARACTERISTICS OF THE BAC- TERIA IN DECOMPOSING SALMON ALBERTO. HUNTER From the Microbiological Laboratory of the Bureau of Chemistry, U . S. Department of Agriculture^ Received for publication May 9, 1921 In the course of an extensive investigation of the bacteriology of decomposing salmon a large number of cultures from various sources have been collected for study. The descriptions and, to some extent, the identification of the bacteria obtained from decomposing salmon caught on the spawning migration have been given in a previous report on this subject (Hunter, 1920). In that report the statement was made that the bacteria isolated from decomposing salmon were found to be those which are de- scribed in the literature as water, sewage and soil organisms. It was also stated that there was apparently no contamination of the fish with spore-forming organisms in the cannery. In order to determine definitely the relation between the flora of sea- water and that of decomposing salmon, experiments on the decomposition of "feedy" salmon, which have been described in a previous report (Hunter, 1921) were conducted at Astoria, Oregon. From the plates and the mixed cultures obtained from the decomposing sahnon 197 cultures were selected for study. Samples of sea-water were collected from various locations near the mouth of the Columbia River. From these water samples, plated on glucose agar, 14 cultures were selected for study. Although this number of cultures appears rather small, ' Published by permission of the Secretary of Agriculture. The writer wishes to express his appreciation of the advice and helpful sug- gestions which have been given by Dr. Charles Thoni under whose supervision this work has been done. During the work in the field, the writer was ably assisted by Mr. B. A. Linden of the Microbiological Laboratory. 85 86 ALBERT C. HUNTER it may be taken as representative of the flora of the sea-water in that locahty. Particular care was taken to avoid the selection of any more duplicates than was necessary and, since the plates from the water samples presented the same types of colonies repeatedly, the 14 cultures collected seemed fairly representative. In order to extend this investigation of the bacterial flora of the salmon industry, a field laboratory was later established at Juneau, Alaska. Samples of sea-water were collected from various locations in southeastern Alaska. A large number of mixed cultures from different parts of the sahnon canneries in this region were also collected to determine whether or not the bacterial flora of the cannery is identical with that of the sea- water and that of the decomposing salmon. From the water samples collected in Alaska 1 1 cultures were selected for further study. Here again the different types of colonies on the plates from the water samples were comparatively few and the 11 cultures obtained seemed representative of the bacterial flora of the sea-water in that region. From the mixed cultures ob- tained from the canneries 94 pure cultures were isolated and pre- served for further study. The organisms from sea-water and decomposing salmon and from the Alaskan canneries have been studied as separate groups and according to their morphology and their cultural reactions duplicates have been checked within each group. This has reduced the original number of 316 cultures to 85. The final designation of each of the 85 cultures with the number of original cultures included under this designation are given in table 1. Each of the 85 cultures has been studied regardless of its action on salmon but particular attention has been given to the character which each organism may or may not possess of pro- ducing foul odors or indol in a specially prepared fish medium. - 2 This medium was prepared in the following manner: To 1000 grams of finely chopped saltwater trout, or weakfish, from which the skin and bones had been removed, was added 1000 cc. of distilled water and 15 grams of pepton. The infusion was made by heating in the Arnold sterilizer or on a water bath at 95° to 100°C. for one hour with occasional stirrings. The juice was strained through cheese cloth with a meat press, filtered through cotton and the reaction adjusted to neutral. The infusion was then heated in the BACTERIA IN DECOMPOSING SALMON 87 TABLE 1 Summary of the number of cultures studied COLTOBB NUMBER OF OniOINAL CULTUnKH IN- CLUDED UNDKK THIS DESIGNATION CULTUHE NDMBER or OHIOINAL CULTDRES IN- CLrOED UNDER Tins DEHIUNATION CULTURE NUMDEB OP ORIGINAL CULTURES IN- CLUDED UNDER THIS DEBIONATION w. 2 360 11 443 2 w,. 1 366 1 444 3 w. 4 374 12 451 1 w, 3 380 12 452 2 w,. 1 390 7 452a 1 W,b 1 395 1 453a 1 w. 1 397 1 454 3 399 1 456 9 w,. 1 400 1 458 2 w,. 2 401 1 459a 1 Wrt 1 403 1 460 7 w,. 1 406e 2 461 1 W,b 1 410 3 462 2 w„ 2 416 4 463 3 417 1 463a 1 w„. 1 419 1 464 3 w„ 3 420o 1 466 2 420E 1 467 1 H. 1 421 1 470a 1 I 1 424 3 471a 1 301a 3 425 4 303 4 427 2 473 2 314a 3 429 1 480 1 316b 1 429a 1 503 6 323 5 430 8 510 7 325 52 431a 1 513a 1 343 3 432 1 516 30 351a 1 436a 2 518 17 354 1 439 4 523 5 440 5 525 5 Each organism was grown in this medium for one week at 30°C. and the presence or absence of abnormal odors noted. The morphology and cultural reactions of 79 of the 85 cultures are Arnold for thirty minutes at 100°C. and filtered, using folded filter papers. About 10 CO. was placed in each tube with about 1.5 grams of raw fish. For anaerobic cultures the surface was covered with a layer of liquid petrolatum. The medium was sterilized in the autoclave for fifteen minutes at 15 pounds. 88 ALBERT C. HUNTER given in tables 2, 3, 4, 5, 6, 7, 8 and 9. For convenience in pres- entation, the cultures from sea- water, salmon and the Alaskan canneries are given separately. The cultures from sea-water are given in tables 2 and 3 according to their ability to liquefy gelatin and regardless of pigment production. The cultures from decomposing salmon are described in tables 4, 5 and 6. Table 4 contains the gelatin-liquefying bacteria exclusive of pigment producers, table 5 the gelatin non-liquefying bacteria exclusive of pigment producers and table G the pigment pro- ducing bacteria. The cultures from the Alaskan canneries are described in tables 7, 8 and 9, and are divided according to gelatin Uquefaction and pigment production in the same manner as those presented in tables 4, 5 and 6. In recording the odor produced by the individual bacteria in fish broth three terms are used. The term "normal" signifies that no abnormal odor was produced and that the organism alone has no physically discernible effect on the fish. The term "off" is used to describe any abnormal odor which is not putrid or exactly foul but indicates that decomposition is taking place. The use of the term "foul" is obvious. "WTiile some of the organisms alone had no apparent effect on the fish, it was possible that these organisms, when grown in fish broth in mixed culture with other organisms might aid in the decomposi- tion. With this in mind several combinations of bacteria were inoculated into fish broth in flasks and incubated for one week at 30°C. All the cultures from sea-water were inoculated into one flask and in one week produced a distmctly putrid odor. Four different combinations of bacteria, the sources of which were decomposing salmon, produced an extremely putrid, foul odor in the fish broth. A rmxed culture of all the organisms isolated from the canneries produced a foul, but not a distinctly putrid, odor in this medium. The odor produced by the organ- isms growing in combination was in each case much worse than when grown individually. Six cultures which have been carried along as individual cul- tures are not included in tables 2 to 9. These are 351a, a pink yeast; 451, 459„, 470. and 480, white yeasts; and 366, a culture BACTERIA IN DECOMPOSING SALMON 89 of Actinomyces. Culture 351a was isolated from the flesh of the back of a decomposing salmon and 366 was obtained from the intestines of a salmon. Cultures 451, 459„ and 480 were isolated from canneries in Ketchikan, ^Vlaska, and 470„ was obtained from the cannery at Yes Bay, Alaska. Of the 79 cultures de- scribed in tables 2 to 9, 72 are rod-shaped organisms without spores, 3 are spore-forming rods and 4 are streptococci. In studying these bacteria but little attempt has been made to identify as specific organisms any except the lactose fermenting organisms and the fluorescent bacteria. The attempt has been rather to determine what relation, if any, exists between the sea-water flora and the flora from decomposing salmon and from the canneries. The extensive work done by several groups of investigators, notably Winslow, Kligler and Rothberg (1919) and Levine (1918), make it possible to identify members of the colon-aerogenes group and, as far as possible, the lactose fer- menting bacteria in this collection have been identified. The descriptions of fluorescent bacteria given by Edson and Carpenter (1912) and by Tanner (1918) also make it possible to identify bacteria of this group. The inadequate descriptions of non- fermenting, asporogenous, gelatin liquefj'ing and non-liquefying bacteria given in the literature, if they are given at all, make it very difficult to identify organisms of this kind. Examination of table 2 shows that of the 9 gelatin liquefying cultures from sea-water 1 is Bad. cloacae and 2 are Ps. fluorescens. Ws differs from W2 only in its faUure to reduce nitrates. Two cultures in this table are unpigmented forms fermenting none of the sugars; one of the cultures, W13, produces spores. Two cultures are pigmented organisms which show no fermentative reactions. One culture, W^, produces spores and ferments glucose and sucrose with the production of acid but no gas and one culture, Wtc, ferments glucose, lactose and sucrose with the production of acid. In table 3 one culture, Wu, was identified as Bact. aerogenes. One culture produces an acid fermentation in glucose; two cul- tures are unpigmented bacteria which do not ferment and two cultures produce yellow pigment and show no fermentative reactions. 90 ALBERT C. HUNTER n a i a 1 s IB - o oj S) £ -S ^ 5 &; o E n fa "3 - a o f^ !z; o o 1 < a 1 j3nrai90J£-Ba3oA o o o o o paj iXq}3M o o o o o 950J3nS o o o o o SSOJOBI o o o o o asoonijf) o o o o ♦ paonpaj a^cj^i^ + + + + + lopai _^ O O O _J. H S i 3 ■a "O -a "o ~ '^ 2 6"" o s M § P-5 J s 1 ^ ^ ^ ^ ^ I a D p d ce M CD ^ ^ ^ ^ ^" BACTERIA IN DECOMPOSING SALMON 91 J3 J3 la O o o .13 a c « a: en 5 OS GO s ■O O a; ^ I O M O i-i J3 o -, o - — C f^ c3 i ■- O M •* o f^ ■^ °! cc n "3 « 5 « o o — ^ n3 ^^ **-« cS g3 (-. -^ «3 .a 05 J3 a o o e o + o o o o o <; <: O < © o .2 M m V en .2 w 'C ■3 ^ ™ 'w n S 2 0 -tj 0 »• ^ b« i 3 m ^ _^ ^ CO _ 3 CO Medi rod Smal "3 S 1^ J3 ^< c: p CO oa C« -i ^ a - S 1 on a >; B 5 S m OJ 1 1 QJ 00 fl 1 "■S 0^ 0 ol 0 * H ^ H^ 3 (S D r from CO, Wa er fr inoo b *«-i 5 00 0 b T3 < 4) c ■IsJ-^l ka atea-fr Harbo ka, an CJ ^ > K-^ »-- "^ »-* cQ 00 o «* 3 * •3 « 03 .X ^ .s ■So BACTERIA IN DECOMPOSING SALMON 93 Of the 10 cultures described in table 4, 3 were identified as Bad. cloacae although in the case of 323 and 380 the methyl-red and the Voges-Proskauer tests are not typical. Repeated plating and testing of these two cultures did not alter the results of these tests and 323 and 380 have necessarily been recorded as atj'pical strains. Culture 31GB cannot be confirmed as BacL cloacae since it does not ferment sucrose. In all other reactions it is similar to Bad. cloacae. One culture, 397, in table 4 is a spore-forming organism which does not ferment any of the sugars; one culture, 395, ferments glucose with the production of acid but no gas; one culture, 354, produces acid in glucose and sucrose and three cultures, 399, 400 and 513^, produce acid in glucose, lactose and sucrose. Culture 513^ is a streptococcus. In table 5 one culture, 525, was identified as Bad. communior, one, H2, as Bad. aerogenes and two, 343 and 523, as Bad. coli, although 343 is atjpical in that it gives a positive result in both the methyl-red and the Voges-Proskauer reactions. Culture 503 produces acid and gas in glucose and gives a positive methyl- red test. Three cultures are streptococci, two of which, 516 and 518, produce acid in glucose, lactose and sucrose and one of which, 510, produces acid only in glucose and lactose. One culture, 406, produces an acid fermentation in glucose. The two remaining cultures in this table, 314a and 410, show no fermentative reactions. Among the pigment producing bacteria described in table 6 are two strains of Ps. fluorescens, 325 and 374. Culture 374 differs from 325 only in its failure to produce indol. Four of the cultures in this table show no fermentative reactions. Culture 401 produces acid in glucose and culture 403 has an acid fermen- tation in glucose, lactose and sucrose. Of the 11 cultures from canneries described in table 7, two, 417 and 419, were identified as Bad. cloacae, although 419 must be recorded as atypical inasmuch as it gives negative results with both the methyl-red and the Voges Proskauer reactions. Two cultures, 471a and 473, produce acid in glucose, lactose and sucrose and one culture, 429, produces an acid fermentation in glucose only. The remaining six cultures in this table show no fermentative reactions. 94 ALBEKT C. HUNTER p S. s S ►J s 9 e :§. cs >1- sc ic it= ic o « — o ia OOO o!z:> (£ O NIVXS WVHO T3 £r ^ '^ 0) 3J a) 5 O a. ^ -3 '^ laD ^S o 03 c. '^ o ©a; U &< Q S C « OJ g J 3 .2 • .H _ C13 CI O o — < ^ o < C' N 3 N a a a c o o o o o + , m m 05 .i- OT V en -T3 O -O T3 T! O S? id £ e p c 2 a> o CD — Si 3 -^ r ti- cs r T3 n c; 03 W 30: -.00 "- 13 -e "o ^ £ o g Ol a> if^ 0 CO CO CO CO 3 -^ ^ ° OJ « § m ra O CO to o ^H i-< ■^ 10 CO CO (N CO o 00 CO o CO d m b( at 0 U) CS T3 g 3 0 ^ -a •a 0 d m lU ■«A C3 0) 0 0 ■0 "O a 0 3 '•* 'r* S» s o o u o u w s » » 1 S oq ft? 0} , _ OS OS si a ■— « __^ a a t^ 3 a o h la h o o 12: [^ o [3^ !? ^ o e u « < Si u S Jans5ie o) oj bC U) U) ^3 a n c 'n CJ C3 « o ^ ^^ ^ — o o o _ _ + + n en -a -c o o o J= ^ .2 o o o o S o o _ „ g J= J3 i ^ CE t» CC tC W C E :2 T3 S ■a o K 03 (-1 C a 5 o CO CQ m .^ ^ „ a» -^ Q> -o J= ■a Cl ^ (1 - c CJ c 03 OQ 1 i fi 3 O O c & o .3 § fl o & 00 ts ■S o « "' a. o >-, a C "a - ■^ i 01 a r^ X: -« r O ^"^■' ,;- J= == (U '^ 03 "^ S Ph 'CQ Ai W I ■a c 3 n oi OJ .S o -a ^ .s ■So 96 ALBERT C. HtTNTER One culture, 424, in table 8 was identified as Bad. aerogenes; one culture, 464, produces acid in glucose and sucrose and two cultures, 454 and 463, produce acid in glucose only. The other eight cultures described in this table do not ferment any of the sugars. Among the twelve cultures in table 9 are two, 431a and 467, which produce acid in glucose, lactose and sucrose; one, 461, which produces acid in glucose and lactose and one, 443, which produces acid in glucose only. The remaining 8 cultures show no fermentative reactions. (■■■. Throughout the investigation no obligate anaerobes were isolated. Anaerobic cultures were made from the material investigated and several cultures isolated which were at first regarded as anaerobic bacteria but subsequent work showed them to be facultative. This was especially true of the strepto- cocci isolated from salmon. These streptococci were originally isolated from anaerobic cultures and they appear to grow best under anaerobic conditions. They do grow fairly well on agar slants, however, and have been carried along in this work as aerobic organisms. As a result of this study of the morphology and the cultural reactions of these 79 organisms it has been possible in several instances to identify as the same organism cultures from the various sources and in this way to estabUsh a partial correlation between the bacterial floras of sea-water, of decomposing salmon and of the Alaskan canneries. This has been done where the morphology and cultural reactions are such that there can be no doubt that the cultures are identical even though it were impossible to identify them as to group or species. This correla- tion is shown in table 10. Cultures Ws and 325 are both the same strain of Ps.fluorescens. Reference to table 1 will show the relative abundance of this organism among the original 316 cultures. This organism appears to be widely distributed in the sea-water since it was obtained from samples collected at Ilwaco, Wash., four miles off the Oregon Coast and Lynn Canal, Alaska (see map, fig. 1). It was also obtained 52 tunes from the various parts of BACTERIA IN DECOMPOSING SALMON 97 n u 1 i, 1 « 1 o 1.2 4 « s a «o u « « a. a. r 1 . < , C3 CU cS H S a ^^ o o g S ta ifl 3 ta ;5^ !? Ec. O o 1^ O <0 z o e < 1 e 0 J9nw:^eojj-«38o^V o o o o o o O o o P9J [A mow o o o o o o o eeojong o o o o o o o < OBOJDBT o o o o o o o ■< 9B03n I Q o o o o < < < < paanpdj s^vj^i^^ + + + + + + + + lopui o o o + + o o + .J T3 -2 •a ■g •a O 03 o 0 fl a c D •X — o O o o ;5 •*d *3 -f^ H V o o a a €• o. S Q^ M CL. fu (£ (S Q aaijaabii Nixv^ao o o o + + + + o NivM wvao ° + + o o o o o & t QQ C ^1 a o S a yello lesh color ° s ° s 3 a o>^ ^ f^ lx< [^ o >H T3 -a CO ■n 00 '1 a^ 00 CO 0. o o o O 2 2 1 2 o in ■— ] "-■ ;— ] ;— ■ 13 S TS S .2 « —1 o c3 C3 03 oi 0) -2 -a .2 rt c c 03 W g a " m 1" a 0) i-g :^'t-9|^" 6 O o C3 O -T^ C. >-, « 2 (U 3 - UC -S « ^^ a a) — o ew woun ack , wound "3 03 B. Stomach ack, b pew wo stomach testines, mouth, 1 kidney ack, p wound, s ach, i tines, h liver, h ills a> > Ph P3 OS PQ m m O ^-1 d M O id ■* ^_^ m O to (N h- O CO CO CO n ■* ■^ JODRNAL OF BACTERIOLOGY, VOL. VII, NO. 1 98 ALBERT C. HUNTER i o u n Q e 3 o o «-*» f^ s u u s n ■fci --o n o u e e cq 05 ^ 5 K. I ^ ra c; — c E _ E ^_^ C 2 c — — . n O o o 3 O O !c 3 sa 3 3 3 O O s CD ^ fe,^ fa ^O ^ O ^ fa fa o I < Jsn^jjeoJd-saaoA o o o O o o o o o + ° 'S. p9j l^qiajv o o o o o o o o o o o ■5> J asoiDng O O 1= o o o o m "m CQ CD E £ E &£ -*^ O s S E la* n •2 -§ = t -§;! to c; , -a -a ~ -i-T CO 3 CD 3 03 -3 ■:= 13 •- 13 e o ■« o c « o CO O oi ^ o t: o ^3 o ^ «• «=5 '^ ^ E 03 !» '- E (m (U L. V »N .& -i -i > C3 CO 00 ^■s S^ t: ~ 5 c;^.-^ =3 t. M pa a a 3 *o .t: -M o o ^ -^ -ft cj u 00 00 O ^^ X u • § f- 0) £ a ~ " ^^ _ '___ _^ _^ o o + + + + ° + 00 2 (opai o o o o ° + + + ° + 0 0 1 i t3 T3 -c -a tj ■a bd u o o o 0 OJ 0 0 S3 * •A eg ^N .2 eao N N be M < S i "V, "C 'n □ 'C ■;: c 'ut H a _o _o c a _o _o « -i; P JZ J "o 7=1 T3 O "o "o "o 0 u 0 75. 3 H I) (U a o ■3 0 0 NIVXS revHo o o o o + 0 000 0 0 0 *^ a a; o (U y ^ c (<• t« t' c _3 m ^^ — DQ _ 3 ^ = =:-- -3 ■£ -C 0 3 » •W -3 T3 "3 o "rt 'O S "a ^ 0 c3 .-a g M CO OD S| 2 CO 0 "a c •1 S « s t. c o t. Is S 0 S 3 1 H^ ^ o •- c! CJ -tJ O -^ 0 — 1i -4^ c c3 ^ CQ ^ o 3 =5" •^^ CJ -C ^ ^ v: « K cs i2 ii ace 5j B « r^ r^ r5 J2 .C -C 0 0 u ■^ -^ -^ 0 ^ ■a^ 0 c; i.s HI *j --^ ^ 0 c 0 ^ w w 0 '^ j3 n 0 W ■ H « a A ii 2 CD 05 O n C<1 CO 00 'if CO ■* •* CO « r: -f t^ »o LO 10 0 CO §5i -* -r ^ rj. •f -r ■^ -^ -^ ^ S rn Ul « 0 bC c x; <»:> e J3 T) T3 u es si no CO 0 -2 S3 33 0 •r! •a 13 r! 0 < < 100 ALBEET C. HUNTER Da _ _ ^^ ^ ,^ , 1 , , g C3 c3 a 03 C3 « 03 cj e H _ a ^^ a a a ^_ a a s t- :; t- 3 C i~ 3 u, ^ k< EQ o o o to o a 0 o o o o o 00 E ^f^^ O (^ ^'^ 12; ;S Iz; 1? 2; janBii o o o o o o o o o o o o § 1 a •i H n 8 -sojj raaoA paj iXq,3H o o o o o o o o o o + o seoiong o o o o o o o o o o < ■< aeo^osq o o o o o o o o o < < < asoonio o o o o o o o o < < < < CC 1^ n paonp -aa a}i!Ji!N ■ + + ■+ + + + + o + + + + u [opni = + + o o + ° o o o o + s •a "S •« "o 1 o T3 -a 0) -a T3 CD ,2 ■s 00 S .2 g.2 ^ a bO s 0} s D o o o o3 o o o3 *5 O o Q 1 ~ o — •*^ o ^ o 3 -a 0^3 0 X C " "o "o o -tJ "o &£ 'o CO O Q, W tc o o a o o3 o a 3 m c •" ^ o (O (U O Q) o o o ^ Q Ph Q M ^ QQ ^ (S Q O Q > -anc = + + + o + '= o + o + o "~. KlVJfi nVHO o o o + + o o o o o + o W h 1 'l t-t ', '^ "§ o "oj « o o O) H 2 "o >. >> "o t^ >> a s CJ C3 OJ o & 2 & J2 O o o § ^1 ■■-J 2 5 »3 ^ 2 iJ S a o a o to ;^ a o 3 ^ a o S :5 OJ — ' o — ' ^ OJ o "^ t-> o 0) — 0) O fe O hJ ij fa >H ij o > h:i >H *^ 0) -C X j3 -S (h a to -^ -»^ |H Q c3 o o « ^ s j= « ;^ '" -S f-H O 0) d of C i^ 0 CD -4^ ■3 ° )— t i = i o ^ ^ « -s ^ ^ ■S " s a 1 o >> a) "a "3 .5 0, O >g .2 & S ° 5 a i5 o o ■^ s •« o o -t^ d S O ^ OQ 03 -a o 1-5 i-i t^ J= J3 QJ ■^ .ti ja o « W Eh W w W O O *— < <5 CO o w s I a u 91 <9 O i-H IQ i 1- D D g^JS! ^ 5§§ 5S 5 to CO s 03 O o 'S a BACTERIA IN DECOMPOSING SALMON 101 the salmon as described in table 10. Cultures Ws and 374 are also identical strains of Ps. fluorescens isolated from sea-water in widely separated localities and from the various parts of decom- posing salmon. As shown in table 1 this organism was obtained from water 3 times and from salmon 12 times. Ps. fluorescens was not found among the 94 cultures from Alaskan canneries. Cultures Wsd and 301 „ are yellow bacteria identical in their morphology and cultural reactions. This organism was obtained once from sea-water collected at Ilwaco, Wash., and three times from salmon. Cultures Wea, 303 and 456 are also yellow pig- ment producing bacteria with identical moiphology and cultural reactions isolated once from sea-water, four times from salmon and nine times from Alaskan canneries. References to table 10 and to the map (fig. 1) will show that this organism was found in widely separated areas throughout Southeastern Alaska. A third yellow organism, represented in the tables as 403 and 467, was obtained once from salmon and once from a cannery at IVIetlakatla. Cultures We, 390 and 417 were identified as a strain of Bad. cloacae isolated once from water collected at Chinook, Wash., seven times from salmon and once from a cannery at Haines, Alaska (see map, fig. 1). Since 316b, 323, 380 and 419 each represent an atypical strain of Bad. cloacae they have not been considered as identical with We, 390 and 417. It is possible, however, that the correlation might be extended to include the sources from which the atypical strains of this organism were isolated. Cultures Wtu, 406 and 454 are identical and this organism was found twice in water from Alaska, twice from salmon and three times from canneries in Ketchikan, Alaska. It will be noted that Haines and Tee Harbor, Alaska, where this organism was found in the water are a great distance from Ketchikan (fig. 1) where it was found in the canneries. The fact that this organism was also isolated from salmon in Oregon indicates that it is widely distributed. Cultures Wrb and 430 are bacteria having thes ame moiphology and cultural reaction, isolated once from the water at Haines, 102 ALBERT C. HUNTER I n CO to • U to tt ►- H o 2 ^ ^ 03 |5? 00 ■— ' >• 3 B ea K — 03 S5 W Oc3 < 3- a" g o a 03 bM s < c ^ - Pi o IS < 03 s IS O CO 0) '3 1 N" t s o s 5 3 c3 > o n :-§^ O ■4^ '■5 CO C CO 3 ~ 3 .a g ?M tT c3 o S o & £ ^ « 2 O a 3 o _>. S *^ 'S _^ & is _>, 'oj 3 (U C "- D. 'Jj 'a3 « s >> OJ ^ Q -^ — c "S "S XI a ft S) J2 OS 'S C3 a 5* g n pq cq n « « s ?j m e « o (^ =3 O O 1^1 ll m f- c3 (rf o o ^ § J3 00 03 c3 00 oi •si ^.^ ^ ^__ gwc O o o o o o =3 a !? °S ») sc C3 o! _g .c & hJ & g o _g _& J2 Ic 03 a> PJ -* i=l H KH )-H u O WEh B c lO « T)< So 1— ' 5^" «. o t~ 5 CO CD " ^u ^S; ^!§ ^S t- o — ^ « ■«< ^^^ BACTERIA IN DECOMPOSING SALMON 103 Ol 2 ^ S '^ o « -t^ O <5 V a o a ►-1 o .a 1 w M ID a" S3 "m 33 oJ OQ ra Sec s g '3 o w >H a w E M ^" •3 " =3 ^ W >< w ^ o o s o o I n a o a 3 O > 3 " -> o- M O c w s i s g o c a S ■— C3 '3 '3 (D ::; ' J5 >. W W r-" ^ £ 3 W kS 0] -H CO t^ So o 05 CO « « ^ <« CO 2 t- 400-47 401-44 403-46 is CO ^ &wg ^ CO P- CO 1. Haines 6. Funter Bay 11. Ketchik.in 2. Excursion Inlet 7. Hawk Inlet 12. Chomley 3. Tee Harbor 8. Hoonah 13. Yes Bay 4. Juneau 9. Chatham 14. Metlakatla 5. Douglas 10. Sitka 104 15. Lynn Canal BACTERIA IN DECOMPOSING SALMON 105 Alaska, and eight times from canneries in Alaska extending from Excursion Inlet on the north to Metlakatla on the south. This culture was not found in decomposing salmon. As seen in the tables, cultures Wtc, 399 and 473 are identical in morphology and cultural reactions. This organism was isolated once from sea-water collected at Haines, Alaska, once from salmon and twice from canneries at Yes Bay and Ketchikan, Alaska. Cultures Wsa, 360 and 421 were identified as a flesh-colored organism isolated once from sea-water collected at Ilwaco, Wash., eleven times from salmon and once from a cannery at Tee Harbor, Alaska. The culture of Bad. aeroqenes isolated twice from sea-water from Alaska, once from sahnon and three tinaes from canneries at Haines and Tee Harbor, Alaska, is designated in the tables under ^^'l2, H; and 424. Cultures Wi2a, 314„ and 458 are identical. This organism was isolated once from sea-water from Funter Bay, Alaska, three times from sahnon and twice from canneries at Ketchikan and Hawk Inlet. The isolation of this culture from material collected from such widely separated areas indicates that it is generally distributed throughout the whole region. Only two spore-forming bacteria were found in the course of the work and one of these, represented in the tables as Wu and 397 was isolated three times from sea-water collected at Ilwaco, Wash., Haines, Alaska, and Lynn Canal, Alaska. It was also isolated once from decomposing salmon. Although this organism appears to be widely distributed in the sea-water of that region it was not common in the decomposing salmon and was not found at all in the various parts of the canneries. When grown in pure culture this spore-former seems to have no effect on fish. Cultures I and 420e are orange-colored organisms identical in morphology and cultural reactions isolated once from salmon and once from a cannery at Haines. It was not found in the water samples examined. .\n orange-colored organism, repre- sented by 401 and 443, was isolated once from salmon and twice from canneries at Sitka and Chatham, Alaska. 106 ALBERT C. HUNTER Cultures 400 and 471a are identical and were isolated once from salmon and once from a cannery at Yes Bay, Alaska. In considering the sources of these various cultures, it may be borne in mind that all the isolations from salmon were made in Oregon and the fact that some cultures isolated there are identical with those obtained in Alaska indicates that such organisms have a wide distribution. The frequency with which some bacteria are found in the cul- tures from decomposing salmon, when considered in connection with their decomposing action on fish, indicates that these organ- isms play an important part in the decomposition of the salmon. As explained in a previous report (Hunter, 1921) the salmon cultures studied in this investigation were obtained on successive days from salmon which were held under known conditions. The predominance of Ps. fluorescens throughout the viscera and the muscular tissue of the decomposed salmon leads to the con- clusion that this organism is an important factor in the decompo- sition of the salmon. Two other organisms appear to play an important part in the decomposition of the salmon, namely, Bact. cloacae and the flesh-colored organism designated as No. 360. These organisms appear repeatedly on the plates and cultures made from decomposing salmon and when grown in fish broth in pure culture they produce very foul odors. As stated before, the other bacteria present may be regarded as accessory in the decomposition of the salmon but from this mvestigation it seems evident that Ps. fltwrescens, Bact. cloacae and the flesh- colored organism (360) are of greater importance than any of the others. This mvestigation has not showTi that there are any different bacteria introduced by the use of the pew, or single-tine fork, than are to be found in the sea-water or m the salmon as it comes from the sea-water. In obtaining cultures from canneries particular attention was paid to the exact location within the cannery from which the material was collected. This was done to determine whether there was any chance of contamination of the canned fish with spore-formers before cooking and also to determine whether there was a bacterial flora peculiar to the cannery or whether BACTERIA IN DECOMPOSING SALMON 107 the bacteria found within the cannery were those which came from the salmon and the sea-water. The investigation has shown that many of the organisms collected are generally dis- tributed throughout the canneiy and arc not restricted to any particular part. In the case of such cultures as 430, 450 and 460, which were found eight, nine and seven times respectively, the organisms were isolated all the way along the canning line from the butchering table to the retorts. Since the location within the cannery, from which the culture was obtained, has no particular significance in a report of this kind, no note of it has been made in the tables. It is thought sufficient to give simply the geograi)hical location from wliich the culture was obtained. Of the 79 cultures reported here only 39 or about 49 per cent were found to have a common source such as sea-water and salmon, sea-water and the cannery or salmon and the cannery. If a larger number of cultures had been collected from a large number of sources throughout the salmon camiing region, it is verj' probable that this percentage of correlation would be in- creased. Of the 15 cultures from sea-water there are only three (Wi, Wia, and Woi,) which were not also found in salmon, in the cannery or in both salmon and the cannery. Fifteen cultures from salmon are identical with cultures from other sources and 14 cultures were isolated from salmon and not isolated from sea-water or from the canneries. Of the 35 cultures re- ported from canneries only 12 were obtamed from other sources. In considering the comparatively small number of cultures collected (316) over such a large area it is not surprising that not more than 49 per cent of them were isolated from the three sources. There are only four cultures in the collection from sea-water (Wi, Wu, Wvb and Wgb) which were not also found in decomposing salmon and it seems probable that if the number of cultures from salmon were larger it would include these four organisms. On the other hand it is also probable that, if it had been possible to collect a larger number of water samples and, hence, a larger number of cultures from water, very many of the 14 cultures from salmon, which it was impossible to correlate with any other source, would have been included in the sea-water flora. This is particularly true of such organisms as Bad. coli 108 ALBEKT C. HUNTER and Bad. communior. It is apparent that, just as suggested previously, the bacteria causing decomposition in sabnon are those forms the natural habitat of which is the sea-water from which the salmon are taken. The correlation between the flora of the Alaskan canneries and the flora of the sea-water and the salmon is not as clear. Only about 34 per cent of the cultures from the canneries can be traced to another source and this leaves the source of 66 per cent of these cultures vmexplained. There is the same probabiUty existing here, however, that, if the number of sea-water cultures could have been increased the number of cannery organisms correlating with them might also have been increased. From this investiga- tion, the outstanding fact about the bacterial flora of the Alaskan canneries is that it consists mainly of asporogenous, non-ferment- ing bacteria which appear to have very little effect on the decom- position of the salmon. This confirms the statement made in a previous report (Hunter, 1920) that the organisms concerned in the decomposition of salmon are those forms which are brought with the salmon from the sea-water and that the decomposition is not due to bacteria which contaminate the sahnon within the cannery. SUMMARY In studymg the distribution of the bacteria concerned in the decomposition of salmon, 316 cultures were collected from sea- water, from decomposing salmon and from salmon canneries throughout southeastern Alaska. By checking the duplicates this number was reduced to 85 cultures, one of which was an Actinomyces, one a pink yeast and four white yeasts. Of the remaining 79 cultures 72 were rod-shaped organisms without spores, 3 were spore-forming rods and 4 were streptococci. The morphology and cultural reactions of these 79 cultures are given. While no attempt has been made to specifically identify many of the cultures, 6 have been identified as Bad. cloacae, 3 as Bad. aerogenes, 2 as Bad. coli, one as Bad. commxtnior and 4 as Ps. fluorescens. The majority of the bacteria collected apparently belong to a large group of non-fermenting soil and water bac- teria. These bacteria are similar to those mentioned by Jordan BACTERIA IN DECOMPOSING SALMON 109 (1903) in his report on the kinds of bacteria isolated from river water and are inchided in his Croups VIII to XIII inclusive. Conn (1917) stated that slowly liquefying or non-liquefying, non-spore-forming short rods such as these make up from 40 to 75 per cent of the organisms developing on aerobic plates inoculated with soil. The results of this mvestigation indicate that Ps. fluorescens, Bact. cloacae and an unidentified flesh-colored organism play an important part in the decomposition of the salmon. In determining the correlation between the bacteria from the three sources, it has been found that 80 per cent of the bacteria collected from sea-water are also found in decomposing salmon, in the canneries, or in both. Approximately 52 per cent of the salmon cultures were found elsewhere and about 34 per cent of the cultures from the Alaskan canneries were obtained from other sources. The results of this investigation confirm the statement pre- viously made (Hunter, 1920) that the bacteria concerned in the decomposition of salmon are those forms the natural habitat of which is the sea-water from which the salmon are taken and that the decomposition of salmon is not due to bacteria which contaminate the salmon within the cannery. REFERENCES Conn, H. Joel 1917 Soil flora studies. Part I. Jour. Bact., 2, 35. EdsoNjH.A., andCarpenter, C. W. 1912 Thegreenfluorescentbacteriaoccur- ring in maple sap. Vermont Agr. Exp. Sta. Bull. 167,521. Hunter, Albert C. 1920 Bacterial groups in decomposing salmon. Jour. Bact., 6, 543. Hunter, Albert C. 1921 The decomposition of "feedy" salmon. Jour. Agr. Res. In press. Jordan, E. O. 1903 The kinds of bacteria found in river water. Jour. Hyg., 3,1. Levine, M. 1918 A statistical classification of the colon-cloacae group. Jour. Bact., 3, 253. Tanner, F. W. 1918 A study of green fluorescent bacteria from water. Jour. Bact., 3, 63. VViNSLOw, C.-E. a., Kligler, I. J., AND RoTHBERG, W. 1919 Studies on the classification of the colon-typhoid group of bacteria with special reference to their fermentative reactions. Jour. Bact., 4, 429. VIABILITY OF THE COLON-TYPHOID GROUP IN CARBONATED WATER AND CARBONATED BEVERAGES S. A. KOSER AND W. W. SKINNER From the Bureau of Chemistry, United States Department of Agriculture, Washington, D. C. Received for publication May 19, 1921 The destructive effect of carbon dioxide on various micro- organisms and the value of carbonation for the preservation of foods and beverages have claimed the interest of a number of workers since the first days of bacteriology. As early as 1885, Leone reported the examination of several commercial mineral waters which were under a slight pressure of CO2. The number of microorganisms found to be present was always low. He also observed that after passmg CO2 gas through a drinking water the total count rapidly diminished. Somewhat later than this a number of investigations were made of the destructive effect of CO2 under relatively high pressures. Schaffer and Freudenreich (1891-1892), after study- ing the effect of pressures of 40 to 50 atmospheres of CO2 com- bined with an increase of temperature, conclude that CO2 has onh^ a feeble bactericidal action. Sabraz^s and Bazin (1893) found that cultures of Bad. typhosum, Bad. coli, Staphylococ- cus aureus, and the anthrax bacillus were able to develop after exposure to 60 to 70 atmospheres of CO2 for several hours. These results are contradicted by D'Arsonval and Charrin (1893) who report that CO2 under a pressure of 50 atmospheres sterilized cultures of Ps. pyocyanea in from six to twenty-four hours. Recently, Larson, Hartzell, and Diehl (1918), in a study of the effect of pressures upon bacteria, found that CO2 under a pres- sure of 50 atmospheres would destroy Bad. typhosum, Bad. coli, Mycobad. tuberculosis, Ps. pyocyanea, staphylococci, strep- 111 112 S. A. KOSER AND W. W. SKINNER tococci, and pneumococci, in a period of time ranging from one and one-half to two and one-half hours. Yeast cells were unaf- fected after an exposure of forty-eight hours. Since these pressures are many times greater than those to which the ordinary carbonated beverages are subjected, there is the possibility that certain organisms may retain their ^dtahty for a longer period. Several reports of the examination of car- bonated beverages purchased in the open market have shown that occasionally there are encountered considerable numbers of microorganisms, including those indicative of pollution. Allen, LaBach, Pinnell, and Brown (1915) report a sanitary survey of the "soft drink" industry of Kentucky. Although carbonation was found to cause a distinct reduction in the numbers present, occasional high counts and the presence of Bad. coli were reported. Stokes (1920) recently examined a great variety of "soft drinks" and noted the frequent presence of Bad. coli in 10 cc. and 1 cc. quantities, with an occasional occurrence in 0.1 cc. The plate counts exhibited great variation, and while the majority of samples yielded counts of less than 100 per cubic centimeter, a few showed surprisingly high numbers. Gershenfeld (1920) reports similar results. Young and Sherwood (1911) have reported an experiment in which they determined the viability of Bad. typhosum Bad. coli, and Enjthroh. ■prodigiosus in car- bonated water to which lemon syrup had been added. Although the typhoid bacillus showed a considerable reduction m numbers after four hours exposure, a few viable cells were found after ten days. Bad. coli and Erythrob. prodigiosus were found to be somewhat more resistant than Bad. typhosum. In the present investigation chief emphasis has been placed upon the colon-typhoid group for the purpose of determining the length of time one may expect the various members of this group to withstand the environment of the different types of commercial carbonated beverages. The following organisms have been employed: Bad. coli (fecal origin). Bad. paratyphosum B, and Bad. typhosimi. Also, as a matter of interest, two common spore forms were included ; B. mesentericus, and a putrefactive anaerobe of the Clostridium sporogenes group. VIABILITY OF THE COLON-n'PHOrD GROUP 113 The beverages were prepared and carbonated in the 7-ounce bottles commonly used in the industry. 8ince they were pre- pared as nearly as possible under commercial conditions and no effort was made to sterilize the various ingredients, control examinations of the product were made previous to experimental inoculation to determine the absence of the particular type of organism used in the investigation. In no instance was any difficulty of this kind encountered. Throughout the work commercial CO2 was used for carbonation. As a test for any impurities in the carbon dioxide which might affect the death- rate of the organisms, tap water was carbonated as usual, then heated in the Arnold sterilizer for a short period to expel the CO2 and finally the death-rate of Bad. coli in this water was comjiared to that in parallel samples of the original tap water. No discrepancies other than those which might be attributed to experimental variation were observed. Small amounts of a suspension of the various test organisms in sterile tap water were used for inoculation. This was accom- plished in one of the two following ways. The first method consisted of adding equal amounts of bacterial svispension to each bottle just before carbonation. In the second method the samples were prepared, bottled, carbonated, and capped as usual. They were then stored at 1°C. for several days until used, when the bottles were re-opened and inoculated. If opened while still cool, there was little loss of CO2 gas. The first method was used for most of the experiments with Bad. coli. The sec- ond method was necessary when working with Bad. typhosum and Bad. paratyphosum B since by the first method there is more or less spattering of the material during the process of carbonation. luMnediatelj^ after inoculation and at definite intervals there- after plate counts were made. To prevent the considerable loss of CO2 upon repeated opening of the same bottles, especially those held at room temperature, a number of bottles were inocu- lated with equal amounts of bacterial suspension and, at each time interval, different sets of two were opened and samples withdrawn for plating. When the numbers had become so 114 S. A. KOSER AND W. W. SKINNER reduced as to give negative results upon plating 1 cc. quantities, larger amounts, 5 cc. and 10 cc, were introduced into broth to determine, insofar as possible, the final disappearance of the organisms in question. This was done by streaking Endo plates from the broth cultures, fishing any typical colonies, and finally applying the usual methods used for the identification of the various members of this group of organisms. EXPERIMENTAL Since temperature may be expected to exert a marked influ- ence upon the death-rate, experimental samples were held at two different temperatures, namely, in cold storage at 1°C. and at room temperature, 19° to 23°C. Tables 1 and 2 present data showing the viability of Bad. coli in carbonated water at several different pressures and also, for purposes of comparison, in plain tap water. It is evident that carbonation causes a speedy de- struction of the colon bacillus and that this effect is dependent upon the temperature at which the samples are held, being much more pronounced at room temperature than at 1°C. Further- more, the different degrees of pressure of CO2 mentioned in these tables apparently exerted little or no influence upon via- bility, for the organisms were killed as speedily in water saturated with CO2 (at both 20°C. and 1°C.), but under no excess pressure, as they were in the carbonated samples under pressures of 28 and 41 pounds per square inch. In fact, where the pressure was released the plate counts frequently were less than those of the samples held under pressure (table 2), a phenomenon which was regularly observed upon several repetitions of the experiment. To gain an idea of the hydrogen-ion concentration of carbonated water the indicators brom-phenol-blue and methyl-red were added to different bottles which were then filled with carbonated water at these several pressures. In this way the value was roughly determined as pH 4.0-4.4. Release of the pressure, as indicated in table 2, was followed by very little, if any, im- mediate change in the hydrogen-ion concentration when meas- VIABILITY OF THE COLON-TYPHOID GROUP 115 ured in this way. A\'hen, however, such samples are held for a period of one or two weeks at 19° to 20°C. there is a gradual escajje of CO2 gas as evidenced by a tlecreaso in the hydrogen- ion concentration. It is believed that under the conditions of our experiments the acidity of the dissociated carbonic acid is TABLE 1 Shounng the comparative viabilily of Dad. coli i?t carhoiMlcd lap ivotcr iindvr pressure and in plain (non-carbonated) lap water CARBONATED TAP WATER TIME- INTERVAI, PRESSnnE 41 POUNDS PER HQUARE INCH (2.78 ATMOSPOERES) AT 18°C. CONTROLS, PLAIN TAP WATER At once 181,000* 156, 000 2ai, 000 190, 000 4 hours 79, 000 80, 000 24 hours 950 6,600 203, 000 194,000 4 days 1 cc. plate 26 34,200 18,000 Held at room tem- negative perature (20-21 "C.) 7 days 10 cc. Of 5 cc. + 1 cc. 0 10 cc. + 5 cc. 0 14,800 10,000 14 days 10 cc. 0 10 cc. 0 2,200 1,000 5 cc. 0 At once 163, 000 181,000 200, 000 260,000 24 hours 51,000 4 days 25, 500 23,000 166, 000 180, 000 Held at 1°C. 7 days 14 days 2,700 1 cc. plate negative 5,400 30 190,000 100, 000 Lost Lost 26 days 5 cc. + 1 cc. 0 1 cc. + 0.1 cc. + 17,000 Lost * Figures represent numbers of Baet. coli per centimeter. t 0 indicates the absence, and + the presence, of Bact. coli as determined by transferring the specified amount of water (10, 5 or 1 cc.) to broth. This was done when the numbers had become so reduced as to give negative results upon plating 1 cc. quantities. the main factor responsible for the death of the bacteria. Other factors, such as differences in osmotic pressure, may also play a part. Several of the simpler carbonated beverages were used in the next experiments. It should be realized that certain acids — 116 S. A. KOSER AND W. W. SKINNER usually citric, tartaric, phosphoric, or lactic — are added to some types of beverages and that these acids may affect the longe\'ity of the organisms in question. A comparison of the viabiUty of Bad. coli in a non-acid and in an acid-containing beverage is shown in table 3. It will be noted that the hydro- gen-ion concentration of the latter is considerably greater than that of the former. The colon bacillus is killed much more speed- ily in the acid-containing beverage, the effect being especially marked at the higher temperature. Apparently the rapid de- TABLE 2 Viability of Bad. coli in carbonated tap water under pressure and with pressure released Held at room temperature (22-24°C.) Held at 1°C. TIME INTERVAL At once 24 hours 3 days 7 days 18 days At once 24 hours 3 days 7 days 18 days pbessube 28 pounds peb square inch (1.9 atmospheres) at 24°C. 440, 000 1,500 13 10 cc. + 5 cc. 0 10 cc. 0 500,000 34, 000 3,600 100 1 cc. + 360, 000 1,360 18 10 cc. 0 10 cc. 0 390, 000 14, 700 900 88 1 cc. + EXCESS CO2 ALLOWED TO ESC.U'E AT THE RESPECTIVE TE.MPERATURE8 BEFORE INOCULATION 440, 000 870 10 cc. + 6 cc. 0 10 cc. 0 300,000 8,200 83 1 cc. + 1 cc. + 500,000 130 36 10 cc. 5 cc. 10 cc. 5 cc. 540,000 10,300 71 1 cc. + 1 cc. + struction of Bad. coli in the acid-containing lemon soda is due mainly to the dissociation of the citric acid present, for in addi- tional experiments in which this acid was omitted the death rate was found to be comparable to that of the non-acid vanilla soda. Also in this case the hydrogen-ion concentration had decreased from pH 3.0 to pH 4.0-4.4 by the omission of the citric acid. Experiments with other acid-containing beverages have dem- onstrated that as the amount of acid is increased above that indicated in table 3, the more readily is the colon bacillus killed. VIABILITJ' OK THE COLON-TYPHOID GROUP 117 In one instance in which 0.156 per cent lactic acid (5 grains per 7-ounce bottle) was employed the numbers of Bad. coli dropped from several hundred thousand to several hundred per cc. in TABLE 3 Showing the viabilily of Bact. coli in a non-acid and in an acid type of carbonated beverage TIUE INTERVAL VANIIXA eODA (NON-ACID), PRESSnRE 24 POC7ND8 (1.6 ATM08P0EnE8) AT 21*C., pH 4.0-4.4 LEMON SODA (ACID TYPE), PBEBSDBE 19 P0DND8 (1.3 ATMOSPHERES) AT 23.5°C., pH 3.0 At once 132,000 136, 000 245, 000 160, 000 24 hrs. 20, 900 32, 700 900 200 XT 1 J i 3 days 1 cc. nega- tive 1 cc. nega- tive Held at room temperature (21-24°C.) 4 days 7 days 140 1 cc. posi- tive 20 1 cc. posi- tive 10 cc. nega- tive 10 cc. nega- tive 14 days 10 cc. nega- tive 10 cc. nega- tive t At once 140,000 127, 000 225,000 215,000 24 hrs. 68,000 46,000 115,000 144,000 3 days 87, 000 25,000 4 days 17,200 13, 500 Held at 1°C. 7 days 14 days 11,100 390 11,200 600 6,800 600 4,900 1 mo. 1 cc. posi- 1 cc. posi- 10 cc. nega- 10 cc. nega- tive tive tive tive 2 mos. 10 cc. posi- tive; 1 cc. positive 10 cc. posi- tive;! cc. negative Composition of the above beverages: Vanilla soda: 10 mgm. c.p. vanillin (0.0048 per cent) and 10 grams sucrose (4.8 per cent) per 7-ounce bottle (207 cc). Lemon soda: 0.5 cc. commercial lemon flavor, 3 grains citric acid (0.094 per cent), and 20 grams sucrose (9.6 per cent) per 7-ounce bottle. Carbonated water added to make the finished beverage. 4 hours at 20°C. When plain non-carbonated tap water was substituted for the carbonated water in the acid beverages, the death-rate of Bact. coli remamed practically the same. That is, in these instances the added acids are the chief causative agents 118 S. A. KOSER AND W. W. SKINNER in the destruction of the colon bacillus, irrespective of the effect of carbon dioxide.^ Since it was found that Bad. coli is able to withstand carbona- tion for an appreciable period, the next step was to investigate the viabiUty under similar conditions of several of the patho- genic members of the colon-tj^ihoid group. In these experi- ments Bad. paratyphosum B and Bad. iyphosum were used. It was at once apparent that both of these types are consider- ably less resistant to the destructive effect of CO2 than is the colon bacillus. Table 4 presents results which are character- istic of a number of similar experiments. One point of particu- lar interest is the persistence, at 1°C., of the last few surviving organisms. These were too few in number to be estimated by plating and their presence could be detected only by the culti- vation of 10 cc. amounts in glucose broth. Additional experi- ments with acid-containing beverages have shown that in these the tjTJhoid bacillus is killed almost instantly. Thus Ln a car- bonated lemon soda containing 0.156 per cent lactic acid, Bad. iyphosum decreased in numbers from an initial inoculum of 27,400 per cubic centimeter to 10 per cubic centimeter within one hour and after two hours its presence could not be detected. It should be emphasized that throughout all of the foregoing experiments the water used for carbonation and for preparation of the various beverages was an ordinary city supply of low mineral content. Under certain conditions, as for example in carbonated water of high mineral content, it is possible that non-spore-forming organisms may remain alive for longer periods than those herein reported. This possible influence of certam inorganic salts upon the viability of microorganisms in a carbonated enviromnent has not been studied in the present investigation. ' The usual methods of bacteriological analysis could not be applied when larger quantities of the highly acid beverages were to be examined. It was found that sufficient amounts of acid were carried over to the culture medium to cause a distinct increase in the H-ion concentration, sufficient, indeed, to effect a retardation or even complete inhibition of growth. By the use of larger quantities of broth in flasks, instead of the usual amounts ordinarilj' contained in test tubes, this difficulty was largely overcome. VIABILITY OF THE COLON-TYPHOID GROUP 119 e B a B o o 88 S q; q; 0) > > > C3 t~ a, > 2 <= O CO CO CO o o I §1 > '■5 U - — 0) a ^ a > . " c- " il O _> cS bC B " O § 2 b to t» 3 ;-i >^ O cJ OS j= -a T3 cc -^ t^ g g o to 1 »o '3 O O bC o CO crj a a a ca « V oj . oj . a> ■ > 3 l^ O > o > ^■^ P w o - O ^^ • > 4 a 3 c» u > '^ > y -^ S'5 ^ 3 1 l-H O^ -i *• ^ =5 O O o * »H rH t-i f-H C3 g ■ 1 S d 05 a e to" i CO o a ° C a> a (V O O 05 V3 o =1 0 2 o 5 S K U s 1-H T-i I— < o" o o o "S. ti 6 a; a, a, *"■ C3 Q) fi 0) g .; .S t; -^ 0 t~ l-H ti « o s -^ S ■*; K r~t tJ «o O O < 1-H s. 1 00 Q 1 1 1 to bfi'n O • ■g o ^" g.£; "'^ 3 ?5 ^ rH o-^ o *^ o *" o « »-l 1-1 •— 1 ;^ u CE] b: •h> 3 O a. 5» m s;;' s CO o i-H CO t O ^' 3 c^ ^ ■-< ~ -^ o o o *■ o " e oq S »-* »-( I-H »-H o, O <0 g" "o> >. £ g J2 ;5 C^ B 73 to ^ .S c? 2 S 2 u t5 S » I— 1 o « o « o =* O »H t-t l-H _ o CO CO 50 CO (O on CO CO 5 o 3 3 3 >^ >. >* >> >> b fi o o O ff c! c3 d 03 i" <© t^ o ■ 120 S. A. KOSER AND 'W. W. SKINNER The resistance of spores to the conditions described in this paper is of some interest when compared to that of Bad. coli and Bad. typhosum. The spores of both B. mesentericus and Clost. sporogenes were found to be quite resistant, for after one month in carbonated water no reduction in numbers could be detected. In one experiment the spores of B. mesentericus survived in a citric acid beverage (pH 3.0) for one month with Uttle, if any, diminution in numbers. It must be stated emphatically that the results obtained in this investigation do not warrant the conclusion that water of a low sanitary quality can be used.by the industry in the prepara- tion of carbonated beverages, or that carbonation can be relied upon to destroy evidence of pollution. In many instances, particularly during the summer months, beverages are consumed within a few hours after their preparation and it is obvious that under these conditions pathogenic organisms, if originally present in the water, may survive carbonation and reach the consumer. SUMMARY Under the conditions of these experiments carbonation exerts a distinctly harmful effect upon the members of the colon-ty- phoid group and their period of viability in carbonated water is much shorter than that in plain tap water. The destructive effect of the CO2 is especially marked at room temperature, 19° to 23°C., and less so at 1°C. In a "non-acid" beverage, the organisms may persist for a slightly longer period than m carbonated water. In beverages containing 0.094 per cent or greater amounts of citric or lactic acids, the death-rate is very rapid and is apparently due to the effect of these acids, irrespective of the CO2. Bad. typhosum and Bad. paratyphosum B are more readily destroyed by CO2 than is Bad. coli. The spore forms of a common aerobe, B. mesentericus, and of a common anaerobe, Clost. sporogenes, were found to be quite resistant to carbonation, surviving one month at room tempera- ture with no apparent diminution in numbers. VIABILITY OF THE COLON-TYPHOID GROUP 121 REFERENCES Allen, P. M., LaBacu, J. O., Pinnell, VV. R., and P.uown, L. A. 1915 Non- alcoholic carbonated beverages, sanitary condition and composition. Kentucky Agr. Exp. Station, Bulletin 192. D'Arsonval, A., AND CuAnni.v, A. 1893 Pression et microbes. Compt. rend. Soc. Biol., 532. Gershenfeld, L. 1920 Bacteria in (so-called) soft drinks. Am. Food Jour., 15, 16-17. Larson, W. P., Hartzell, T. B., and Dikiil, H. S. 1918 The effect of high pressures on bacteria. Jour. Inf. Dis., 22, 271-279. Leone, C. 1885 Sui microorganismi della acque potabili, loro vita nolle acque carboniche. Gazzetta chimica italiana. Vol. 15. Translated by Von Sehlen, Archiv. f. Hygiene, 1886, 4, 168-176. SABRAzfis, J., AND Bazin, E. 1893 L'acide carboniqueA haute pression, peut-il 6tre considerde comme un antiseptique puissant? Compt. rend. Soc. Biol., p. 909. ScHAFFER AND Freudenreich 1891-92 Annalcs de Micrographie, 4, 105-119. Stokes, W. R. 1920 Bacteriological examination of soft drinks. Amer. Jour. of Public Health, 10, 308-311. Young, C. C, and Sherwood, N. P. 1911 The effect of the environment of carbonated beverages on bacteria. Jour. Ind. and Eng. Chem., 3, 495. A BINOCULAR jMICROSCOPE ARRANGED FOR THE STUDY OF COLONIES OF BACTERIA GUILFORD B. REED Queen's University, Kingston, Ontario Received for publication July 24, 1921 The binocular microscope arranged as described below has been used in this laboratory for several years with so much satisfaction and profit in the isolation of bacteria that a descrip- tion of such a simple piece of apparatus may not be out of place. Every one who has examined colonies of bacteria or other or- ganisms on culture media by the use of direct illumination from below has experienced more or less difficulty. The medium is frequently too opaque to admit sufficient light, especially in the case of media contaming blood; some very small colonies have the same refractive index, or so nearly the same refractive index, as their medium that observation may be very difficult even on transparent media. These difliculties ai-e obviated and ob- servation greatly facilitated by viewing the colonies in light reflected from the surface of the culture as is usually done in using a hand lens for this purpose. To make such observa- tions with higher magnification and greater ease w^e arranged a binocular microscope with magnifications of ten and twentj^ diameters in the following manner. The tube and focusing apparatus was removed from the stage and base and attached to an independent support so that its optical axis was at an angle of 45° with the stage, though so arranged that the angle, might be altered as desired (A, fig. 1). A small arc with a condensing lens was supported so as to project a beam of light at an angle of 45° with the stage and at right angles to the opti- cal axis of the microscope. As this gave rather more Ught than was necessary a 75-watt "daylight" lamp was supported about 123 JODBKAl, OF BACrHBIOLOCI, VOL. Til, NO. I 124 GUILFORD B. REED 15 cm. above the stage and surrounded, except for a slit 2 cm. wide, with an opaque reflecting screen in such a position that a beam of light was projected to the stage at right angles to the optical axis of the microscope (B, fig. 1). A solid black stage provided support for Petri dish cultures or a groved block placed on it provided for the support of tube cultures so as to bring the surface of the media parallel with the stage. Fig. 1. A Diagram op a Binocular Microscope Arranged for the Sttjdt or Colonies on the Surface of Media A, Binocular microscope tube and focusing apparatus supported on a base independent of the stage and at an angle of 45° with the stage; B, "daylight" 75 watt light surrounded by a screen except for a slit; C, solid black stage and base; D, glass screen to protect open plate cultures; E, Petri dish culture. The focal distance of the binocular is sufficient to permit of the examination of tube cultures or Petri dish cultures with the cover left on. For the fishing of colonies from cultures which were to be further incubated a plain glass was supported 15 mm. above the stage by strips along three sides leaving the right BINOCULAR MICROSCOPE FOR STUDY OF BACTERIA 125 hand side open. An uncovered plate culture might then be slipped under this glass screen with very Uttle danger of dust or breath contamination and at the same time permit the fishing of colonies, under microscopic observation, with a platinum wire. The most satisfactory platinum wires for this pui-pose were drawn out in the fiame to a fine point which was then turned at a rif!;ht angle about 1 mm. from the tip. The chief value of the apparatus in this laboratory has been in the isolation of organisms producing very small colonies from material containing large numbers of other species, i.e., the isolation of H. infliienzce from sputums. It is a very easy matter to fish the smallest H. influenza: or similar colonies of other species, or to remove half of such a colony and smear for micro- scopic observation and later to use the remaining half colony for the inoculation of subcultures. AN INVESTIGATION OF AMERICAN STAINS REPORT OF COMMITTEE ON BACTERIOLOGICAL TECHNIC Prepared by H. J. CONN, Chairman^ New York Agricultural Experiment Station, Geneva, New York Received for publication October 16, 1921 Early in 1920 the Committee on Bacteriological Technic was asked to look up the matter of biological stains at present available in America and to see what could be done toward standard- izing them. At present it is difUcult, if not impossible, to obtain the (Jriibler stains, and the American products are known to be variable. The impression has even been common that American dyes are generally imsatisfactoiy for staining. Upon looking into the matter, the committee has found that many American stains are as good or even better than the old Griibler stains, except for certain special uses which their pro- ducers did not have in mind when preparing their products. Certain American producers of biological stains are trying very hard to put on the market an entirely satisfactory line of goods. The difficulty comes from the fact that the field is a small one, and so much competition has arisen that no one can make a satisfactory pi'ofit. The danger is that soon the American producers will all be driven out of the busmess and importation of stains will again be necessary. We do not desire to be depend- ent upon foreign production in this line, because of the great importance of stains in public health work and the possibility of another national emergency when importation will be im- possible. To relieve the situation, therefore, some one manu- facturer must be given enough support to make the business profitable for him. This means standardizing on one line of stains, either all produced by the same house, or if produced by different concerns, each particular stain coming from one manu- facturer only. ' For the numerous collaborators in this work, see list at end of the report. 127 128 H. J. CONN In order to survey the field it seemed necessary to test products from as many sources as possible. As the whole field is a large one, the attention of the committee at first was given to but three stains or groups of stains; fuchsin, methylen blue, and gentian violet. About thirty members of the Society, — to whom much credit is due,^ — volimteered to assist. Their names are given at the end of this paper; and without their hearty co- operation, the work would have been impossible. Our study is not complete yet, but a fairly satisfactory survey of the field has been made, so far as concerns fuchsin and methylen blue. The gentian violet situation is more complicated and further in- vestigation is necessary. Upon looking into the commercial situation, it proved necessary to obtain samples of the dyes from two different classes of dealers : the basic dye manufacturers and the so-called "manufacturers and standardizers " of biological stains. The former group are the manufacturers of general textile dyes, and embrace such concerns as the National Anilin Company, the duPont Com- pany, the Calco Chemical Company, and Dicks, David and Company. The latter group generally buy their dyes from the former, make certain tests of them and put them on the market as biological stains, with or -ftdthout modification; their chief function is standardization rather than manufacture. Repre- sentatives of the latter group are the Providence Chemical Laboratories, the H. S. Laboratories, the Coleman and Bell Company (formerly the National Stain and Reagent Company), and the Heyl Chemical Company. Samples from these various concerns were obtained through a jobber who had consented to cooperate in the work. This was done so that no concern would know the reason for which the samples were bought. To make unbiased opinions still more certain, the samples were dis- tributed among the various investigators bj'^ number only, without reference to the names of the dealers. Throughout this work, the conunittee has been in corre- spondence with all of the second group of dealers mentioned above, and with the Calco Chemical Companj' as well. There has seemed no purpose in getting in touch with the other basic INVESTIGATION OF AMERICAN STAINS 129 manufacturers, as the production of biological stains is a very small part of their business and the results of this work mean Uttle to them. In general, hearty cooperation has been secured. All but two of the dealers have told us to feel perfectly free to publish the results. As these two are merely distributors of scientific supplies, and as it is diflicult to tell the original source of the samples obtained from them or to leam whether their present suj^ply is still the same, their names are suppressed here, using instead the designations A and B. P"'or similar reasons the name of a third house of like character is suppressed, although in this case the committee was not asked to do so. The list of dealers (other than these three) whose products have been examined is as follows. In this list the addresses are given for those firms who deal specially in biological stains. Calco Chemical Company, 136 Liberty Street, New York City Coleman and Bell Company, Norwood, Ohio Dicks, David and Company E. I. duPont de Nemours Company Geigj' Chemical Company Goldin Biological Laboratories, Providence, R. I. Harmer Laboratories Company, Lansdowne, Pa. Heyl Laboratories, 437 Baretto Street, New York City Holland .\nilin Company H. S. Laboratories, 6005 Girard Avenue, Philadelphia H. Kohnstam and Company Mallinckrodt Chemical Company Merck and Company National Anilin Company Newport Chemical Works Providence Chemical Laboratories, 51 Empire Street, Providence, R. I. Williamsburg Chemical Company Very few tests were made with the products of the Heyl Laboratories. This was merely because the Heyl products are sold under certain restrictions that make it difficult for jobbers to buy them. We could have obtained all the Heyl stains we JOUBNAL or BACrSRIOLOOT, VOL. VII. NO. 1 130 H. J. CONN wanted direct from the company; but for the sake of fairness we did not want to buy this one brand of stains direct, while picking up the others on the open market. The jobber through whom we dealt had such delays in getting the Heyl samples that they were not ready at the time of our main tests. It might further be remarked that Mr. Heyl himelf says that his crude dyes are generally obtained from abroad, as he does not believe the American products satisfactory. One of the troublesome factors in the stain situation is due to confusion in nomenclature. It may happen that the same name is given to more than one product or that the same product is called by several names. Accordingly it seems worth while to discuss these three stains from the standpoint of their chemical composition and to list their synonyms. In the list of synonyms in each case, preferred names are given in bold-faced type. FUCHSIN (basic) .S'j/nojiJ/rns Rosanllln Diamond fuchsin Magenta Rubin Anilin red The formula of fuchsin, or rosanilin, is (NH2.C6H4)2:C:C6H4" NH. Being a basic dye, this is ordinarily combined with an acid, and the dye with which we are familiar is the chloride (NHs -06114)2 :C:C6H4:NH:HC1. Fuchsin proves the easiest to obtain in a satisfactoiy state of purity of any of the dyes we have so far investigated. There are three chief uses of fuchsin in bacteriological work: general bacterial staining; staining for acid-fast quaUties; and use in the endo medium for colon-typhoid differentiation. A satisfactory stain should give good results for all three purposes. In the present investigation, the acid-fast test has been most frequently used, but enough tests have been made for the other purposes to indicate the general utility of the different samples. Only two investigators have used the stains in the endo medium, I>rVESTIGATION OF AMERICAN STAINS 131 both of them using different technics; although one of these two (Castleman) made his tests on two different occasions, with a different technic each time. Results were reported in two different ways. Some grouped the samples in three or four categories which it has proved possible to denote by four terms: excellent (E), good (G), fair (F), and unsatisfactory (U). Others listed the samples in the order of their excellence, in which case instead of placing them in these four categories, they have been numbered 1, 2, 3, etc., the low numbers indicating the best samples. These two sets of symbols are used in Table 1, where the results of the different investigators with the different samples are listed. It will be seen that there is some variation in the findings of the different investigators, although not as much as in the case of the other two dyes, reported below. When fuchsin was used as a stain it proved difficult to pick out any one sample or any two or three samples that were superior to the others. To make the comparison more definite, each of the four classes was given a numerical symbol, these numbers averaged, and the average converted back into the class symbol again, using the grades: E, E— , G+, G, G — , etc. By glancing at this average grade, it will be seen that the sample from the Providence Chemical Company ranks G + , none of the others ranking better than G. All but three of the others rank G. The samples are rearranged in table 2 in the order of excellence, according to their average grade, this same table listing the number of times each sample was found excellent and the number of times found unsatisfactory. Only three samples were re- ported unsatisfactory by any one, namely those from dealers A, B, and C, each being so reported by two different investi- gators. These three were the only samples to have an average grade lower than G, and yet two of them were reported as good or excellent by some investigator. There seems no reason to feel that any of the samples, with the exception of these three, could not be successfully substituted for the Griibler product. The superiority of the Providence sample is so slight as to be of little significance. JOURNAl. OF BACTEHIOLOOT, VOL. VII, NO. 1 132 H. J. CONN s "a ta o §gg 226! gi| M g] r* 5 > s uinipani opng '?B9^ puoo3e' 'uBoiansB'^ Ot)(=^OOfeOOt= o 5 u I I I I I + I I I o a S ■< a "I i* < fe > O 5 •< < + I I UIB^S \Vian9Q '8a93J0TJ CJt>& O niB^s |qai2 'sja^aj^ OlsOOOHO OWtJ ulB-jsiqaiz '^jaajng OOOOOOCOOO^ o uiB'ts iqaiz 'iiiH CO-H c^>t>- 10 ^-^C! 00 uiB?s iqajz 'P^^oa ooooooooooo o nin^s-ja^nnoj tUBJQ 'UBOiaHSBQ OS^COC^C/J^^lCt^OlM to mv%e iqaig aBinaj^sBQ IMOt^ONCOlO^-HCO— iO> 1 : S : ■ 0 ■ : 0 : ^ • "O ■ e 03 : 13 : '5 ■a a 03 : '> : . 03 . ^ Ih b3 fc« 0 y m OQ : t. » n 0 ^ 2 i3 ^I- t. ^ >, 0 "> a 0 D.-r' S ■'-> U) U QO c] "•S 6-X3 & g^SM-^-S-ga c3(D0303Q)i:oO .QJQJ.iiD- 0 Q K ^ Ih; 04 00 W Q PQ INVESTIGATION OF AMERICAN STAINS 133 When used in the endo medium, the samples were reported quite differently. The result plainly depends on the technic, some methods calling for too concentrated a solution of the dye to be decolorized by the amoimt of svdfite ordinarily used. The following formulae were used : Meyers: 0.5 cc. of 10 per cent alcoholic fiiclisiii added to 10 oc. of 2.5 per cent sodium sulfite solution. Tlli^^ quantity added to 100 cc. of the agar. (Stand- ard method of .Vmerican Public Health .Vssnciation.) TABLE 2 Summorii of reports onfuchsin; samples arranged in order of excellence AS 9TA1N 8AUFLB Num- ber of tests Num- ber of times excel- lent Num- ber of times unsatis- factory 0 0 0 0 0 0 0 0 0 2 3 3 Aver- age grade Conclusions IN THE ENDO MEDIUM Providence [Calco 10 10 11 9 10 8 11 9 9 11 9 8 3 1 2 0 1 2 2 •) 2 2 1 0 G-f F+ F- Very good Good Good Good Good Goo O 00 o M g tS 13 . C.2 ^ 's: b-i > se- es O C3 O C) o c o , SuQQSi^'^a&;oasaK C3 CQ « C. S a CO < si c 0 a •S « -a :s OS i 5 OC ^, INVESTIGATION OF AMERICAN STAINS 137 &d:3So bt ^ 1 o o w I.I 1 + + 1 1 1 («. (I. o o o o o fe w w « O fe H ;3t!fc,owowot300 oooo w O OO o w w O H o O H fLt b, fz, H + K C=< IDH O t) o OH W &00 O oo o O O w H S H H t>oo K := s o c=,o ^3 fo fe O 0^3 ^ t fc. s a ;= "3 n 4i» T3 £? -s c^ S 3 J3 kl d a p: o « rmer llink rck. jvide lema vl a ^ - 1 _» _o .g "3 "3 =: cj ~ " - o C a s <• p. OK c K C C is 1 138 H. J. CONN TABLE 4 Summary of reports on methylen blue; samples arranged in order of excellence AS STAIN SAMPLE Num- ber of tests Num- ber of times excel- lent Num- ber of times unsatis- factory Aver- age grade Conclusions IN THE LEVnJB MEDIUM For bacilli: Coleman and Bell Heyl 9 2 7 9 7 6 7 6 8 6 7 8 7 7 6 7 2 9 10 9 8 7 8 9 5 5 2 1 1 1 0 0 0 1 0 0 0 5 5 3 4 0 o 2 4 0 0 1 0 0 1 0 1 2 2 2 3 2 4 2 3 6 0 0 0 0 0 0 0 0 0 o 5 5 3 E- G- F+ F-l- F+ F F F F F- U E- E- E- G+ G+ G G G- F F- F- F- Very good Very good Fair to good Fair Fair Fair Fair to poor Fair to poor Fair to poor Fair to poor Unsatisfactory Unsatisfactory Very good Very good Very good Very good Very good Very good Good Good Fair to good Fair to poor Unsatisfactory Unsatisfactory Unsatisfactory Excellent National Anilin. . Calco Harmer . . . Fair Good Unsatisfactory Unsatisfactory Unsatisfactory Good Excellent Good H. S. Laborato- Holland Anilin. . . Kohnstam Providence Dealer A . . * Dicks, David . . Medicinal: Unsatisfactory Good Dealer C Good Williamsburg .... Coleman and Bell Heyl Providence Goldin Excellent Good Not tested Unsatisfactory Fair H. S. Laborato- ries Fair Merck Unsatisfactory Excellent Fair Unsatisfactory Good .♦Dealer A * Harmer * Mallinkrodt.... * Dealer B * These samples dissolve milk smears when used by the Breed method. The results are listed in table 3, in which the same set of sym- bols is used as in table 1. A summary of the results is given in table 4. The most apparent fact is that the best results in staining have been obtained with the zinc-free dye. This is presumably because of the insolubility of the zinc salt in alcohol, INVESTIGATION OF AMERICAN STAINS 139 as Macy and Robertson, who used aqueous solutions, obtained essentially as good results with the zinc salt. The striking exceptions to the inferiority of "mcthylen blue for bacilli" are the samples from Coleman and Bell, and from the Heyl Chemical Company. After completing the test, it was learned that these two samples, although not labelled medicinal, and not claiming to meet the U. S. P. requirements, are fairly free from zinc, having been especially purified and corrected for bacterial staining. Of the samples of medicinal methylen blue, six stand at the top, as a stain, with the honors quite evenly divided between them: namely, Calco, dealer C, WiUiamsburg, Coleman and Bell, Heyl, and Providence. Almost as good are the Goldin and H. S. samples. Of these eight companies, we find that the Calco Chemical Companj', the Coleman and Bell Company, the Heyl Laboratories, the Providence Chemical Laboratories, and the H. S. Laboratories are making special efforts to put out a pure product well adapted to both staining and therapeutic uses. It is plain that they are succeeding in this, so far as staining properties are concerned, and provided the quaUty of their products is maintained any of them may be used as a stain without question. One fault found with certain samples of methylen blue is that they have a tendency to dissolve milk smears off the slide when used for the Breed method. The chemical explanation of this phenomenon has not been obtained. The samples giving this trouble were the zinc salt sample from Dicks, David and Com- pany, and the medicinal samples from Harmer Laboratories, MalUnkrodt Chemical Company, and from dealers A and B. It is interesting to notice that these same samples have been graded lowest according to their staining properties. They are plainly not to be recommended. There is little correlation between the results of the staining tests and those obtained by Levine with the eosine-methylen- blue medium. The samples from dealer A, for instance, which stood low as regards staining properties were among the four 140 H. J. CONN best samples in this medium. The medicinal sample from the Providence Laboratories, on the other hand, proved good for staining but unsatisfactory in this medium. The five samples grading good or very good as stains and also proving good or very good as an indicator were the two Coleman and Bell samples and the medicinal samples from the Calco Chemical Company, AMlliamsburg Chemical Company, and from dealer C. Those samples proving unsatisfactory were condemned because of toxicity. Strange to say, as many medicinal samples as zinc salt samples proved toxic, although for medicinal purposes the zinc salt alone is condemned on account of toxicity. In fact, for this medium, the zinc salt seems to prove almost if not quite as good as the free chloride. There can be no question that, so far as our data go, one of the five samples just mentioned as satisfactory for both purposes should be recommended. The Heyl samples seem to rank with these as to staining properties, but for the reasons mentioned in the introduction, could not be included in the early tests, so we do not know how satisfactory they would prove in the Levine medium. We have not learned the original source of the sample from dealer C. The Williamsburg sample was labelled medicinal, but proves not to be entirely zinc-free. As a result, the composition of this product may well have been changed by the present time to meet U. S. P. requirements, and one could not count on getting another sample like the one tested. The two Coleman and Bell samples and the Calco medicinal sample all give very good results, and are produced by companies we beUeve to be reliable and to supply nothing but distmctlj' American products. Both companies we understand to manufacture their owti methylen blue from American-made cnides or inter- mediates. Any of these three samples can be used with entire confidence of obtaining good results. Of these three, the Cole- man and Bell "methylen blue for bacilli" has two pomts especially in its favor: it is not quite so expensive as the U. S. P. preparations, and it is the best of these three samples in the Levine medium. INVESTIGATION OF AMERICAN STAINS 141 GENTIAN VIOLET The term gentian violet is at present very indefinitely used. Grubler apparently originated it, applying it to a certain mixture of dyes all closely related chemically and having similar staining properties. These dyes belong to the pararosanilin series. The pararosanilin base is considered to have the formula: /CCH4NH2 In this formula it will be noticed that there are six hydrogen atoms attached to the three amino-nitrogen atoms. These six hydrogen atoms may each be replaced by a methyl group (CH3), and ethyl group (dRi) or even by a benzyl group (CeHs). The comi^ounds of most importance in gentian violet are those with four, five and six methyl groups respectively, known as: tetramethyl-pararosanilin pentamethyl-pararosanilin hexamethyl-pararosanilin In these three compounds, the greater number of methyl groups present the deeper is the shade of violet. Thus the tetra- methyl compound is a reddish violet while the hexamethyl compound is a bluish violet. The introduction of a benzyl group tends to darken the shade still further. The tetra- and pentamethyl compounds are seldom prepared in a pure state, but the hexamethyl compound, mider the name of crystal violet, is readily obtainable in fairlj' pure form. Criibler's gentian violet, besides these compounds, is supposed to have contained other rosanilins, but these three methylated pararosaniUns were the most important constituents. American gentian violets are sometimes pure crystal violets, sometimes some other dj^e of this series, and sometimes various mixtures of the above mentioned dyes. The term gentian violet is not used by the dealers in textile dyes nor by their dye chemists. This stain is ordered only by the biologist, and hence it is furnished only by biological supply houses and some dye manufacturers that specially cater to the biologist. 142 H. J. CONN These stains may be grouped in the following three categories : Methyl violet Methyl violet B Methyl violet 2B Methyl violet BS Methyl violet BN Methyl violet BO Methyl violet BBN Dahlia B Under these designations are sold certain mixtures of the tetra- methyl and pentamethyl compounds, in varying proportions, sometimes also containing the hexamethyl compound as well. The letters following the name are used in the trade to indicate the shade, methyl violet B and methyl violet 2B, for example, each being progressively deeper in shade than plain methyl violet. The trade designations do not refer to chemical compo- sition except in so far as the deeper shade indicates a more highly methylated compound. ir Benzyl violet Methyl violet 5B Methyl violet 6B Methyl violet 7B These dyes are benzylated compounds, generally mixed with some of the purely methyl compounds above mentioned. The number of B's in the trade designation indicates the depth of violet produced in dying. Ill Crystal violet This dye is the pure hexamethyl compound. It is the most readily obtained in pure state of the whole group, and there is less confusion in regard to its composition than in regard to any of the others. INVESTIGATION OF AMERICAN STAINS 143 For the present series of tests, samples of the following were collected: Methj'l violet, methyl violet B, methyl violet 2B, methyl violet BS, methyl violet OB, crystal violet, and gentian violet. There were over thirty samples in the entire lot, too many, it was felt, to send to any one investigator for the first test. Accordhigly a more or less indiscriminate selection was made, so that each investigator had about twelve samples to test, generally one or two of each type of dye. They were asked to use the samples for the Gram stain, each man to employ his own technic, but to give the exact procedure used. As it turned out, scarcely any two men used the same technic, so the results varied greatly. The plan is to run another test with a standard technic. At present so few men have reported on any one sample of dye and such various technics were used that the results are not regarded as having more than a general signifi- cance. It would not be fair to judge the individual samples on the basis of this preliminary work ; so in this report the names of the manufacturers and distributors are not given. Each sample is denoted by number alone, each number indicating a certain business house. Results are given in table 5. It was not expected when the work was started that the com- pounds of lower methylation would prove satisfactory for the Gram stain. It was somewhat of a surprise, therefore, to find that out of 84 tests of individual samples of methyl violet, methyl violet B and methyl violet 2B, there were 28 reports of good results, and one report of excellent results. Nevertheless, out of these same 84 tests, there were 36 reports of unsatisfactory findings, such a high percentage in comparison with the residts obtained with the higher methylated compounds that no further work is to be done with these dyes. The dyes called aniUn violet, methyl violet BS, methyl violet 6B, and crystal violet, give in general as good results as "gentian violet" and sometimes better. There is some variation, however, in the products of different manufacturers, but on the basis of these few tests, none can be condemned. According to present indications, crystal violet can be substituted very successfully for gentian violet in the Gram stain. Some manufacturers at 144 H. J. CONN " fS. as aavBo aovasAY 1 o + 1 + 1 fil, ft fx* fit lit tH lit 1 + 1 III fa fa O fa (ii (« fa NVKJJOH O CO CO O fa fa t) ca b: o o poqiaui eui3^?v H o o noi-jnjoe onoqJ^O a o o aoi?nios3uiiJiiR o t3 D SNISHVH IvaKNVH SO OH aonnios otioqjEO o &. fc. t> !i< &< O O fa & OfaOOO t3D aoi^njos 3ui]ji)g H (i< fe t3 t> t5 fc. t) fa t) O fa W fa P ts & ouaaf^aeia Id t= P p t) ;= U) avhhuh o o fa o o g ta aoi)n]oe oiioqJSQ t3 o ^ ;=> p :d ao^nios SaiiJlJS ^ fc< fa o fa ;= O aoijnjoe otjoqiBQ t) o t> o noi^nios 8a!iJi?S fa f=( 13 t3 llVAOiS O O t3 t-^ tJ OS NOSIHHVH fa t3 O fa & 'o ■J31B0OH t30 oo e.saHajovjaKVJv CO r^ c^ WD CO t^ CO OS C• 00 t>» CO ■ ■♦» 5.2 INVESTIGATION OF AMERICAN STAINS 145 1 o 1 1 1 1 tK C5 O O II 1 + + OOOOfcOOf»t3 oot>t>oocw E2 t)H a OOfaWOf=3fafaO t)fat3t3Ht3t>H fa wo W W fa t> W W fa faHfaofafawa ^ - t3 JD O &0 H O fa ooowo^wo o o O a o fa &, ;= Wfa fa fa fa b< c t3fa p :=) H w O t3 OO P Ofa & t3 s 0& fat) t> fa& t5 fa fc. o OO o fa o » Ci) 1^ H to W t30 o H D o o faH fa h- Ci CO 00 C) — M 'N — CI CI CI -!• t^ f~ O •■•; t^ 1-" C^ C^l Ol c^ c)i^«o05cci^ooo> "-i CI CJ M O o 146 H. J. CONN present make this substitution for us, sending crystal violet when gentian violet is ordered. The chief cause of variation in results seems to be due to the differences in technic. This shows plainly the need of standard- izing the Gram stain, a task proposed to this committee some time ago. It is hoped as the result of the present work to accom- plish this. Apparently the Stirling technic does not always work with these American products, as they are generally more con- centrated than Griibler's gentian violet. Griibler's stain is well knoT\Ti to have contained a large amoimt of dextrin. The investigators who have used one of the methods calling for a carbolic solution seem to have obtained the best results; although one man who has compared a carbolic solution with the anilin- sulphate solution of Atkins, reports even better results with the latter. In conclusion it can be said that several of the American gentian and cr^-stal violets are as good, and some of them better than the Griibler product, when the Gram stain is used as a criterion. It is still too soon to endorse the product of any one manufacturer or distributor, but good results can apparently be coimted on with any of the following : Methyl violet 6B Coleman and Bell, Heyl, or H. S. Crystal violet Coleman and Bell, Harmer, or Providence Gentian violet Coleman and Bell It is suggested that anyone wishing to order before definite recommendations are made by the committee specify one of these seven products. GENERAL RECOMMENDATIONS There is at present so much confusion in this field that the biologist generally has httle idea what actual dye he is buying under any given name. One practice that has contributed to this is that distributors buy the stains in bulk and then repack in small containers bearing their o^v^l label without mention of the name of the manufacturer or standardizer. Sometimes the distributor renames the stain on his own responsibility. There INVESTIGATION OF AMERICAN STAINS 147 is no objection to the distributor's name occurring on the label, provided the stain is designated by the original name, followed by the name of the manufacturer or standardizer from whom it was obtained. Our chief recommendation at present is there- fore that everyone buying stains from dealers in general labora- tory supplies insist that the source of the stain as well as the distributor's name be printed on the container. It is our hope that the distributors may cooperate with us in doing this. When it comes to selecting between the different brands of stains on the market, it must be recognized that the number of dyes so far tested are so few that no general recommendations can be made as yet to apply to all bacteriological dyes. The present work, however, indicates that the products of Coleman and Bell, the Calco Chemical Company, the Providence Chemical Laboratories, and the H. S. Laboratories, all rank well. The managers of the companies are willing to cooperate, and they are all putting on the market stains that there is every reason to believe are of strictly American make. In particular, we would call attention to the extremely good showing of the Coleman and Bell (formerly National Stain and Reagent Company) products. This company, moreover, has been good enough to furnish the committee with fairly definite specifications for those dyes so far tested, an important point because one of the chief matters to be kept in view is the permanence of the supply of stains. There is every reason to expect that the Coleman and Bell Company, given adequate support by the users of stains and government protection against foreign products, wiU be able to continue in the business; but if it should not, much of the information obtained by this company would be at the disposal of our committee to use in duplicating their products. On accoimt of their willingness to cooperate and their eagerness to be of service to bacteriologists, all of the Coleman and Bell products — even those not yet tested — deserve a thorough trial. 148 H. J. CONN LIST OF COLLABORATORS M. F. Boyd, University of Texas, Galveston, Texas. C. T. Burnett, 608 Majestic Building, Denver, Colo. P. Castleman, Health Department, Boston, Mass. S. H. Craig, H. K. Mulford Company, Glenolden, Pa. A. A. Eisenberg and G. M. Hamel, St. Vincent Hospital, Cleveland, Ohio. M. S. Fleisher, St. Louis University School of Medicine, St. Louis, Mo. F. P. Gorham, Brown University, Providence, R. 1. R. B. H. Gradwohl, Gradwohl Laboratories, 7 W. Madison Street, Chicago, 111. Edith Hannum, H. K. Mulford Company, Glenolden, Pa. F. C. Harrison and E. Hood, Macdonald College, Quebec, Canada. M. J. Harkins, H. K. Mulford Company, Glenolden, Pa. D. J. Healy, Agricultiiral Experiment Station, Lexington, Ky. P. G. Heineman and C. R. Hixon, U. S. Standard Products Company, 111 W. Monroe Street, Chicago, 111. Grace A. Hill, Washington State College, Pullman, Wash. F. W. Hochtel, University of Maryland Medical School, Baltimore, Md. G. J. Hucker, Agricultural Experiment Station, Geneva, N. Y. C. A. Hunter, State College, Pa. F. M. Huntoon, H. K. Mulford Company, Glenolden, Pa. G. F. Leonard, E. R. Squibb and Sons, New Brunswick, N. J. M. Levine and L. H. James, Iowa State College, Ames, la. C. B. Lipman, University of California, Berkelej', Calif. H. Macy, University of Minnesota, St. Paul, Minn. G. McConnell, City Hospital, Cleveland, O. P. Masucci, H. K. Mulford Company, Glenolden, Pa. J. T. Meyers, University of Nebraska, Omaha, Neb. C. Murray, Iowa State College, Ames, la. E. M. Pickens, University of Maryland, College Park, Md. Neva Ritter, Consumers' League, Kansas City, Kans. A. H. Robertson, Agricultural Experiment Station, Geneva, N. Y. C. Roos, H. K. Mulford Company, Glenolden, Pa. W. D. Stovall, State Laboratory of Hygiene, Madison, Wis. E. F. Voigt, Board of Health, Fort Smith, Ark. E. M. Wade, Board of Health, Minneapolis, Minn. JOURNAL OF BACTERIOLOGY Contents NOVEMBER, 1921— VOL. VI— No. 6 George E. Holm and James M. Seierman. Salt Effects in Bacterial Growth. I. Preliminary Paper. Hilda Hempl Hbllrr. SiipKCstions ConccrninR .1 Rational Basis for the Classi- fication of the Anaerobic Becteris. Studies in Pathogenic Anaerobes. IV. I. Prelimin.ary Paper. J. Howard Brown. Hydrogen Ions, Titr.ition and the Buffer Index of Bacterio- logical .Media. J. E. Rcan and G. A. Palmbr. On Decreasing the Exposure Necessary for the Gelatin Determination. n.\ni->i.r> Macv. Chart of the Families .and Genera of the Bacteria. SEPTEMBER, 1921— VOL. VI— No. 5 G. P. Plaisance and B. W. Hammer. The Mannitol-Producing Organisms in Sil.ige. Hilda Hempl Heller. Principles Concerning the Isolation of Anaerobes. Studies in Pathogenic Anaerobes. II. John F. Xoutox and Mary V. Sawyer. Indol Production by Bacteria. Anou.sTo BoxAzzi. On Nitrification. IV. The Carbon and Nitrogen Relations of the Nitrite Ferment. Th. Thjotta AND Odd F.\l8en Sundt. Toxins of Bact. Dysenterias, Group III. JULY, 1921— VOL. VI— No. 4 Laura Florence. Spiral Bodies in Bacterial Cultures. James M. Sherman. The Cause of Eyes and Characteristic Flavor in Emmental or Swiss Cheese. G. J. HncKER. A New Modification and Application of the Gram Stain. Louis J. Gillespie. Color Standards for the Colorimetric Measurement of H-Ion Concentration. Harriet Leslie Wilcox. The Effect of Pepton upon the Production of Tetanus Toxin. K. G. Dernby and J. Blanc. On the Growth and the Proteolytic Enzymes of Certain .\naerobes. MAY, 1921— VOL. VI— No. 3 S. Orla-Jensen. The Main Lines of the Natural Bacterial System. Kan-Icuiro Morishima. Variations in Typhoid Bacilli. Frederick A. Wolf and I. V. Siiunk. Solid Culture Media with a Wide Range of Hydrogen or Hydroxyl Ion Concentration. AuousTO Bonazzi. Studies on Azotobacter Chroococum Beji. ORDER BLANK Williams & Wilkins Company Publishers of Scientific Journals and Books Baltimore, Maryland, U. S. A. Kindly enter my subscription for the Journal of Bacterioloot Volume VII, 1922, for which I enclose $5.00 (Canada, S5.25; foreign, S5.S0), or, send me a state- ment and I will remit. .Signed Address RARE SUGARS and OTHER FINE CHEMICALS "DIFCO" STANDARDIZED Being the first manufacturing firm in the United States to make and extensively advertise rare sugars, we have steadily added to our line, and our list now comprises the following: RARE SUGARS Arabbose Levulose (Fructose) Rafiinose Dextrose (Glucose) Maltose Rhamnose Galactose Mannose Saccharose (Sucrose) Invert Sugar Melezitose Trehalose Lactose Melibiose Xylose OTHER CARBOHYDRATES Dextrin Mannite Inulin Salicin LABORATORY REAGENTS Invertase for Analytical Use Acid Potassium Phthalate Blood Serum, Desiccated Beef Extract Fibrin Gelatin Decolorizing Carbon Tyrosine » DEHYDRATED CULTURE MEDIA Bacto-Peptone Proteose Peptone Carried in slock by the principal dealers in Scientific Supplies Specify "DIFCO" WE INVITE COMPARISON Digestive Ferments Company Detroit, Michigan, U. S. A. VOLUME VII NUMBER 2 JOURNAL OF BACTERIOLOGY OFFICIAL ORGAN OF THE SOCIETY OF AMERICAN BACTERIOLOGISTS MARCH, 1922 EDITOn-lN-CHIEF C.-E. A. WiNSLOW II IS chaTMlerislic of Science and Progress Ihal they corttintially open new fields to our visions. — PASTEim. PUBLISHED BI-MONTHLY WILLIAMS & WILKINS COMPANY BALTIMORE. U. 8. A. i.nt«ed a..t«>rd-cl«f»m»lfpr April 17. 191«. «t thepoetoffice »l B«ltimor«.M»r3fl«ndjOnd«rthe»ct ol * Act of October 3, 1917. Authorued on July 18. 1918. Copyright 1922. WilliamB aad WiUdns Company f $5.00 per volume, I^nitrd States, Mexico. Cuba *• "^fi \ $5.25 per volume, Canada postpaia j 55^fin per volumo. othrr countries Made in United States of America The actual direct benefits from Gastron are a known quantity, Unmistakably experienced by the patient, observed by the physician— in disorders of gas- tric function. Inevitably the better digestion promotes better nutrition, strengthens resistance, encour- ages and heartens the patient, thus promotes a condition of body and a state of mind con- ducive to restoration. In these circumstances there is an appeal for a wide application of Gastron; far-reaching indeed be its ultimate good effects. Fairchild Bros. & Foster NEW YORK JOURNAL OF BACTERIOLOGY OFFICIAL ORGAN OF THE SOCIETY OF AMERICAN BACTERIOLOGISTS DEVOTED TO THE AOVANCEMENT AND DIS- SEMINATION OF KNOWLEDGE IN REGARD TO THE BACTERIA AND OTHER MICRO-ORGANISMS Editorial Board Editor-in-ChxeJ C.-E. A. WINSLOW Yale Medical School. New Haven, Conn. A. Parker Hitchbns F. C. PIarrison, Ex officio Principal. Mac Donald College, P. Q., Canada Advisory Editors C. C. Bass H. W. Hill V. A. Moore M. J. Rosenau R. E. B0CHANAN E. O. Jordan M. E.Penington W. T. Sedgwick P. F. Clark A. I. Kendall F. B. Phelps F. L. Stevens F. P. Gay C. B. Lipman L. F. Rettger A. W. Williams F. P. GoRHAM C. E. Marshall L. A. Rogers H. Zinsser F. C. Harrison CONTENTS F. C. Harrison. Our Society 149 Victor Burke, \otes on the Oram SUxin with Description of a New Method 159 B.1RNETT Cohen. Disinfection Studies. The Effects of Temperature and Hydrogen Ion Concentration upon the Viability of Bact. coli and Bact. typhosum in Water. 183 Selman a. Waksm.^n. Microorganisms Concerned in the O.xidation of Sulfur in the Soil. I. Introductory 231 Selman a. Waksman .ixd ,I. S. Joffe. Microorganisms Concerned in the O.xidation of Sulfur in the Soil. II. Thiobacilhis Thiooxidans, a New Sulfur-o.xidizing Or- ganism Isolated from the Soil 239 E. B. Fred and W. H. Peterson. The Production of Pink Sauerkraut by Yeasts. . . . 257 CosTA.vTiNO GoRiNi. Studies on the Biology of Lactic Acid Bacteria: a Summary of Personal Investigations 27^1 L. D. Bushnell. A Method for the Cultivation of .Vnaorobes 277 L. D. Bushnell. Influence of Vacuum upon Growth of Some Aerobic Spore-Bearing Bacteria 283 H. R. Baker. Substitution of Brom-Thymol-Bluc for Litmus in Routine Laboratory Work 301 W. A. Wall and .A. H. Robertson. The Use of Domestic Methylene Blue in Staining Milk by the Breed Method , 307 Abstracts of American and foreign bacteriological literature appear in a separate journal, Abstracts of Bacteriology, published monthly by the Williams & Wilkins Company, undet the editorial direction of the Society of American Bacteriologists. Back volumes can be furnished in sets consisting of Volumes I, II, III and IV. Price per set, net, postpaid, $24.00, United States, Mexico, Cuba;S2o.OO, Canada; 820.00, other countries. Subscriptions are in order for Volume V, 1921. Price, per volume, $5.00, United States, Mexico, Cuba; $5.25, Canada; $5.50. other countries. NOW READY FOR DISTRIBUTION A NEW 828-PAGE EDITIOX OF OUR CATALOGUE E.\ TITLED LABORATORY APPARATUS AND REAGENTS SELECTED FOR L.VEOR.\TORIES OF CHEMISTRY, METALLURGY AND BIOLOGY DJ THEIR APPUCATION TO EDUCATION, THE INDUSTRIES, MEDICINE .\XDTHE PUBLIC SERVICES We suggest that scientists in requesting copies of this catalogue state the capacity in which they are connected with organized laboratory v.ork, and mention this Journal in writing. In large laboratories and departments where many workers are engaged, we prefer that the request for catalogues come from the professor in charge of the department, labora- tory director, chief of bureau, chief chemist, pathologist, bacteriologist, or other scientist in charge of laboratory work, with the names of colleagues, associates or assistants to whom, in his judgment, additional catalogues should be sent. Requests from firms or institutions should state the name of the scientist or ofRcer for whom the catalogue is desired so that we may check against subsequent requests f ro:n indi\ddual3. Such co-operation assures adequate distribution and at the same time reduces unnecessary duplication in the same laboratory. While the sending of this catalogue involves no obligation to purchase of tts, the expense of preparation and publication necessitates a reasonable control of the distribution. Separate pages have been prepared for sending in a loose-leaf binder to those whose interest is temporary or restricted to a given subject or item rather than in the regular use of the complete catalogue. ARTHUR H. THOMAS COMPANY WEST WASHINGTON SQUARE PHILADELPHIA U. S. A. CABLE ADDRESS, "BALANCE." PfflLADELPHIA tlUI OUR SOCIETY' F. C. HARRISON Principal, Macdonald College, P. Q., Canada INTRODUCTION The constitution of our Society states that the object of the Society shall be the promotion of the science of bacteriology, the bringing together of American bacteriologists, the demonstra- tion and discussion of bacteriological methods, and the considera- tion of subjects of common interest: Thus, in accord with other scientific societies, its ultimate object is to make life a better thing than it is, to help in the service of man, and to attempt in some small measure to attain what Aristotle in Book III of The Republic has expressed that: Our youth will dwell in a land of health, amid fair sights and sounds and receive the good in everything; and beaut}% the effluence of fair works, shall flow into the eye and ear like a health giving breeze from a pure region and insensibly draw the soul from earliest years into likeness and sympathy with the beauty of reason. To accomplish these benefits, our work lies ready to our hands, but our strength may be the greater and our faith the firmer if we spare a moment from present toils to look back upon the achievements of the past, to gather strength and encouragement before confronting the future. The Society was founded in 1900, under the presidency of the late Prof, W. T. Sedgwick, a great teacher, an inspiring investi- gator and a kindlj' spirit that radiated good will and courteous consideration to others. In the two decades just past the Society has increased in stature and in wisdom and in favor with fellow workers. ' Address of the President at the Twenty-third Annual Meeting of the Society of American Bacteriologists. 149 JOCRX.VL or BVCTERIOLOar, VOL. VII. NO. 2 150 F. C. HAKRISON The membership has grown from fifty in 1900 to over one thousand in 1921. The Society has tasted the joy that springs from labor, and perhaps its greatest achievement has been the estabUshment of the Journal and Abstracts which are to our workers ports and happy havens indeed. Further achieve- ments of happy memory and daily use to the teacher and in- vestigator are the Society's card, the standard methods of many routine proceedings; some order out of the chaos of bacteriologi- cal nomenclature; and the commencement of work on standardiz- ing materials and methods. The recital of the past is an augury for the future. Can we, at this time, suggest a policy for our Society? Are we able to plan a cooperation of efforts which will interest all members, for each to give something of himself for the good of the Society, and for the benefit of human life and effort? "The keen spirit seizes the prompt occasion — makes the thought start into instant action, and at once plans and performs, resolves and executes." May I, therefore, take this opportunity of placing before j^ou a few thoughts regarding the future. I. STEADY CAMPAIGN FOR MEMBERSHIP, AND THE ESTABLISHMENT WHERE POSSIBLE OF LOCAL BRANCHES Thanks to the activity of Dr. Ayers and his committee, there has been this year a large increase in membership. But we must have more, if the plans outlined for a larger journal are to be carried out; a large membership is fundamental and when ob- tained many other things will follow in consequence. We have started a number of local branches. No scheme offers better prospects of success for keeping the interest in the Society between annual meetings. Every endeavour should be made to increase these local organizations for, besides interest, they afford opportunity for social meetings, arouse a feeling of professional solidarity and permit of more frequent contributions to our science. Further, by enlarging the field from which members are drawn, the danger of narrow specialization is avoided, and members of the local branches will be given a broader outlook. OUR SOCIETY 151 II. IMPROVEMENT OF THE JOURNAL AND ABSTRACTS When so much has been done in establishing these two im- portant journals and bringing them to their present state of ex- cellence, any drastic criticism wouki be a task of supererogation. I mention this matter, however, because our Secretary, in his circular letter dated October 25, 1921, stated that the Editor must have more space in order that papers offered for publica- tion may appear more promptly. In order to make the Journal OF Bacteriology a monthly periodical, there must be an increase of membership to 1500, or else other means must be adopted, and several alternate suggestions are set forth. All of us would like to see the size of the Journal increased, and regular publication guaranteed; these benefits can best be obtained only by an increase of membership. If our publica- tions maintain a high standard of excellence, their worth will ensure more subscriptions from foreign countries. Each member should be jiersonally interested in the success of these journals, and should see that they attain a wider field of usefulness. Judicious and frequent suggestions to libraries that do not take them, bringing them to the notice of medical men and others interested in cognate subjects would help to adver- tise and undoubtedly secure subscribers. III. PROVISION for critiques AND RESUMES Abstracts at present is filled with references and short re- sumes of work done by bacteriologists all over the world. I suggest that it would be of considerable interest to arrange for critiques or comprehensive r(5sum6s on many subjects, to be written by students of particular groups. P'or example: a comprehensive summary of the literature on the decomposition of cellulose; botulism; the carrier of infection, etc. Such reviews would be of great value to students, save much time in hunting up references, and bring the subject up to date. If written in a judicial spirit, the writer would be able to size up the situation, and give an appreciation of the subject as a whole. If possible, such contributors should be paid. 152 r. C. HABKISON IV. CAED INDEX COMPILED FROM ABSTRACTS The Library of Congress prepares and issues a card index of books and periodicals. The United States OflBce of Experiment Stations issues a card index of all its publications and that of the experiment stations. Would it be possible, and would it fill a need, if the editors of Abstracts prepared a card index of the papers abstracted? The basis of subscription would have to be the individual card. Laboratories might arrange to subscribe according to the titles and sub-titles of the table of contents of Abstracts. General bacteriology should interest all. Agricultural colleges would naturally desire the card index for dairy, soU, and plant bacteriology. Public health laboratories should be interested in water, sewage and food bacteriology' and health board labora- torjr methods and so on. Prompt service as regular as the issue of Abstracts would be possible. How many would be wUling to subscribe for such a service? V. COOPERATION BETWEEN OUR SOCIETl" AND THE AMERICAN PUBLIC HEALTH ASSOCIATION In some regards our Society and the laboratory section of the American Public Health Association cover similar subjects. Both organizations have attempted to standardise technique and methods, both have published so-called standard methods. ^Members of our Society have been prominent in the American Public Health Association and vice versa. I suggest that some form of cooperation be instituted which would prevent any useless duplication of work. The two societies should work together, for in unity there is strength, and projects to be initi- ated by each might well be considered jointty, not necessarily by the whole society in convention, but at any rate by responsi- ble committees of each. otm SOCIETY 153 VI. A COMPREHENSIVE STUDY OF METHODS AND MATERIALS Valuable work has been done by committees of the Society who have given their labor and time ungrudgingly in order to work out methods or examine materials. I mention this here because I believe that a more comprehensive survey of methods and material would be of great value. In biological problems we cannot hope for the minute exactness of the chemist, but the chemist has his standard and authorised methods, which he dare not depart from. When a method is being improved or a new one instituted a committee and official referee is appointed, and a thorough test, often extending over years, is given. It seems that such methods would be helpful in bacteriological technique and help to ward aside many criticisms that are at present levelled at us. The Society might add to the committees already at work, and assign more problems to be worked out for the common good. VII. BITREAU OF EMPLOYMENT The Chemists Club of New York city has managed an em- ployment agency since 1913 with marked success. I have fre- quently availed myself of its services with satisfaction to all concerned, and I suggest that our society might organize and run a similar bureau. If this employment bureau were incorporated as a "membership corporation" no person could benefit by the profits but any such accruing would go to the Society. Such an organization, if carefully organized and conservatively run, should be of great service to those in administrative positions, or heads of departments desiring to obtain assistance, not to mention those aspirants to bacteriological fame who desire to place a foot on the first rung of the ladder of employment opportunity. Certain large employers of trained bacteriologists, such as the health departments of large cities, could fiJe their specific needs with the bureau, and these organizations could be kept informed of any men coming on the list who were specially fitted for their particular line of work. 154 F. C. HABRISON VIII. A BACTERIOLOGICAL MUSEUM The American Museum of Natural History in New York has maintained a bacteriological museum under the skilled super- vision of Prof. C.-E. A. Winslow. This has been a valuable asset to members of the Society, and should have support. Some of the older members may remember the Museum which Krai maintained in Prag, and the fine series of museum speci- mens and photographs he used to prepare. Krai's collection was very helpful to teachers and investigators, and has now been moved to Vienna. The Society should have a museum, which should serve as a repository for all type species of organism descriljed in our literature; further, I suggest that part of the work of such a museum would be to keep up the pathogenicity of organisms of economic importance, or those used for teaching. For several 3''ears I have sought plant pathogens of knowii Wrulence, for class work, but have received organisms devoid of pathogenicity, to the disappointment of the classes, and the possible loss of faith in one's veracity. If the American Natural History Museum will undertake this work, nothing more need be done by the Society, save to let aU members know what services the museum can render. If, how- ever, some measure of support is necessary, I trust that the Society will investigate in what way the interests of its members can best be served, and make proper business arrangements with the Museum. I suggest also another function for a museum, that of a IX. BACTERIOLOGICAL CLEARING HOUSE You are all aware of the fimctions of a clearing house in a large city, an organization the members of which meet daily and pass through the various cheques and transactions of many banks. Something similar would be valuable to the bacteriologist. I presume that many of you have had similar experiences to my own. Working on a particular problem, or doing some routine work, such as water or milk analysis, you find occasionally an OUR SOCIETY 155 organism thut attracts your interest by some peculiarity or abnormality. You isolate it, and put it by, with the intention that at some convenient season you will investigate it further, but alas, procrastination is the thief of time; the convenient season does not come, and the organism probal)ly dies of exhaus- tion, starved by inattention to its material needs. Now, if we had a clearing house the course would be different. You would say, candidly, I have not the time to work out this organism; it is interesting on account of its morphology or some peculiarity of its culture; I shall send it to the clearing house with a note as to whore it was found, and its pecuharities. It is, therefore, duly dispatched, and on arrival at the clearing house, those in charge will read the letter and note the pecuharities, and will say, send this to John Doe, he is interested in this line and is working on this particular group, or investigating this irregularity, etc., or failing a student of this group, the organism might be investigated by those in charge of the clearing house. This is a function that a well equipped and adeciuately staffed museum might undertake — possibly the American Museum of Natural History might institute a department charged with such work. If this were possible it would be a great asset to our Society. Failing such an organization, it might be possible to arrange for some distribution through the Journal. X. A NEW DETERMINATIVE BACTERIOLOGY Those of us who teach and those who are engaged in general or systematic studies, know the difficulties experienced in determin- ing species. Frantic search of Chester, Migula, Matzuschita, of mono- graphs here, of periodicals there, often fail to give any information or assistance on the points we desire, and then we realise the inadequate nature of our descriptions and our classifications, for they are numerous. The Society, through the interest of some of its members, has taken cognizance of the chaos in classification and has adopted certain revisions which are helpful. 166 F. C. HARRISON The Society also has a chart that has undergone a number of revisions, and which in its present form represents what is con- sidered necessary for a proper description of an organism. These two contributions are excellent, but not enough. We want full descriptions as per Society chart of all knouTi organisms, and we want them properly named and classified according to our latest classification. Here then is a splendid task for the Society, a task seemingly of great magnitude, but with proper organiza- tion and cooperation we should be able to surmount all obstacles, and produce a new determinative bacteriology' approved by the Society, and all interested in bacteriology. Such a pubUcation from its intrinsic value would find a place in every laboratory and succeeding generations of students and workers would rise up and caU us blessed. We have a membership of a thousand. If each one would pledge himself to give a full description of an organism assigned to him, what a magnificent start it would be. I trust the Society will accept this suggestion and formulate the necessary conmiittees. XI. TEACHING PROPAGANDA FOR BACTERIOLOGY A subject, the study of which may not only serve on accoimt of its educational value by enlarging our knowledge of nature and training the powers of observation and judgment, but also because of its sheer practical utility as the servant of medicine, pathology, sanitation, industry, agriculture, and household Ufe; should have a well defined place in all our colleges and univer- sities, and possibly in secondary schools. Yet we find that it is not mentioned in the curriculum of eight of our agricultural colleges. In many medical colleges it takes a subordinate place, and in many institutions having comprehensive courses in botany and zoology no course is given in bacteriologJ^ Surely this is not right. Several past presidents of our Society have spoken about the value of the subject from the educational and practi- cal aspects, notably the late Prof. W. T. Sedgwick and Pro- fessors Bergey, Marshall and Jordan, and there are excellent references in our literature as to the place the subject ought to OUR SOCIETY 157 hold in our halls of research and learning. Therefore, all members of this Society should carry on incessantly and enthusiastically an active propaganda for bacteriology, for more bacteriology', and for still more bacteriology until the subject becomes more wide- spread in its benediction and embraces all classes and all institu- tions of secondary and higher learning. A subject, which has achieved so much for the rehef of suffering and the prevention of disease and whose effective progress and definite mission can be measured almost daily merits the early attention of the pupil and student. Let us remember the words of the Greek philosopher, Prodicus, "That which benefits human life is God." CONCLUSION In conclusion, let me state that I have made these suggestions in the hope that they will be freely discussed by our members. We have a live Society, we are all interested in a subject that has advanced in the last forty years by leaps and bounds, there is much to investigate and nothing can provide an investigator with quite so pleasant a decoy as the persuasion that his chosen theme is in the nature of virgin soil. Here is a true adventure of the spirit, for he is adding a few sovereign grains of gold to the riches of his science for the enrichment of present life. If little labour little are our gaines; Man's fortunes are according to his paines. — Herrick. NOTES ON THE GRAM STAIN WITH DESCRIPTION OF A NEW METHOD^ VICTOR BURKE From the Department of Pathology, Stanford University, San Francisco Received for publication June 2, 1921 An examination of the literature concerning the Gram method of stainhig bacteria and of the methods of making the stain in various laboratories reveals the fact that a successful stain, i.e., one giving sharp differentiation between Gram positive and Gram negative organisms, can be obtained by several methods. A comparison of these procedures indicates that in a successful method each step is properly coordinated with and influenced by the other steps in the process. One step can be modified if other steps are also properly modified. This fact has been recognized and made use of in the various processes employed to improve or shorten the process of making the Gram stain. Of all the various methods advanced none has successfully eliminated the personal factor. A satisfactory stain, especially of pus smears, depends to some extent upon the skill of the operator. Any one procedure has to be more or less modified depending upon the type of smear to be stained. For this reason the untrained assistant can not always be depended upon to produce satisfactory results by following any method so far advanced. The skill of the operator, aside from his knowledge of the organ- isms to be examined, depends, upon the proper control of decolor- ization and the ability to detect any defect in the prunary stain used. ' Aided by funds from the United States Interdepartmental Social Hygiene Board for Research in the Prevention and Cure of Venereal Diseases. 159 160 VICTOR BURKE The control of the decolorizer requires experience. Some of the methods used advocate exposure of the smear to the decolor- izer for a definite period of time, other methods recommend exposure to the decolorizer as long as any of the stain comes out. The skilled technician modifies both of these methods as the occasion arises. The time of exposure to the decolorizer must vary with the type of smears to be examined and depends upon the excellence of the violet stain and decolorizer used and upon the nature of the treatment accorded the smear between the removal of the iodine solution and the appUcation of the decolor- izer. It has not heretofore been recognized that the extent of the blotting and accompanying drying of the smear after the iodine solution greatly influences the decolorization of the Gram negative organisms. The process of blotting after the iodine constitutes one of the two critical steps in the Gram stain and will be discussed in detail later. Aside from the blotting after the iodine and the control of the decolorizer, the balance of the process can be carried on by an untrained assistant. No one method so far advanced has been shown to be dis- tinctly superior to all the other methods. The Committee on the Descriptive Chart of the Society of Bacteriologists has not yet selected any one method as the standard method to be used in the study and description of pure cultures. This committee has presented three methods for the consideration and criticism of the members of the Society. (Conn, H. J., 1919; Atkins, K. N., 1920.) While not recommending any one complete method it does recommend that 100 per cent alcohol be used as the decolorizer, that the films be made with distilled water and that the gentian violet, iodine solution and alcohol be drained or blotted off but not washed off of the fihn. The different methods so far advanced vary in the making of the solutions, the strength of the solutions and the time periods of exposure. The most important modifications of the method as presented by Gram have resulted from efforts to improve the primary stain by increasing the penetration and intensity of the stain and the permanency of the solution. In most cases in- creasing the intensity of the stain by the use of mordants such NEW METHOD FOR GRAM STAIN 161 as anilin oil and phenol results in a loss in the stability of the solution. The stability of the solution has been found to vary with the method of mixing the ingredients together. The original anilin gentian violet as described by Gram may deterio- rate in a few days. Sterling's modified solution lasts several months. Other modifications are said to keep indefinitely. Perhaps the most promising of these is the one recently described by Atkins (1920) in which anilin sulphate replaces the anilin water in the primary stain and NaOH is added to the iodine solution. One of the objections to the use of the Gram stain has been the unsatisfactory or unstable nature of this primary solution. Until the discovery of Jensen (1912) that an aqueous solution of methyl violet gives very satisfactory results it was assumed that a mordant such as anilin oil was a necessary factor in making a good Gram stain. Aqueous solutions of methyl violet are coming into general use in Europe. That this has not occurred in America is due to the difficulty of obtaining methyl violet 6b, as recommended by Jensen and to the fact that the American dyes have given unsatisfactory results. With the production of better domestic dye stuffs or improved methods of using the present products we believe that the favored primary stain will eventually be an aqueous solution which will be stable over a long period of time. This stability, combined with ease of prep- aration, are factors in its favor. The present paper describes the results of experiments to compare aqueous solutions of various domestic dye stuflfs and their value as substitutes for the anilin gentian violet solution in the Gram staining method.^ The dyes used gave variable, and in most cases such unsatisfactory results that the experiments were extended to cover an analysis of the different factors deter- mining the Gram reaction in the hope of modifying the staining method in such a way that satisfactory results could be obtained with more of the domestic dye products. The experiments mcluded the determination of the effect of heat and of acid and alkali added to the primary stain on the slide, the effect of ' This work was begun by Mrs. Pearl M. Smith and continued by the writer. 162 VICTOR BURKE washing between the different solutions, the effect of heat on the iodine solution, a comparison of different decolorizers and the effect of water on decolorization. Certain of the experi- ments have a bearing on the nature of the Gram reaction and the characteristics of the Gram precipitate. Organisms used. Unless otherwise stated the smears used in these experiments were made from pure cultures of Staphylococcus aureus; Bacterium typhosum; Neisseria catarrhalis and Neisseria gonorrheae; grown for approximately twenty-four hours on peptic digest agar slants to which 33 per cent hydrocele fluid had been added or Loeffler's blood seium tubes. Films. The films were made in tap water or physiological salt solution. A few preliminary experiments convinced us that no difference resulted from the making of the films in dis- tilled water, tap water or salt solution. It was assumed that if a difference did result this would be a factor that would have to be taken into consideration in staining pus and body fluids. Mounting the films in an acid or an alkali does influence the results as will be described later. DYES USED IN PRIMARY STAINING SOLUTION One per cent aqueous solutions of six different samples of methyl violet, three of gentian violet and one of crj^stal violet were used and compared. The dye was added to the distilled water, shaken thoroughly, allowed to stand several hours and filtered as used. Such solutions remain stable for a considerable time. Some of our solutions were kept two months and showed no deterioration, in fact there seemed to be some improvement with age. Results obtained. According to the method of .Jensen, satis- factory results can be obtained with aqueous solutions of methyl violet 6b, only when using a strong iodine solution* and absolute alcohol as the decolorizer. With our dyes we compared the results obtained with the following decolorizers, 95 per cent alcohol, 100 per cent alcohol, 100 per cent acetone and acetone • One gram of iodine, 2 grams of potassium iodide, 100 cc. of distilled water. NEW METHOD FOR GRAM STAIN 1G3 and ether (equal parts), otherwiHe following the method of Jensen except that safranin was used as a counter stain. Briefl}^ Jensen's method is as follows: Air drj% fix with mild heat, cool before flootling with 0.5 i)er cent methyl violet solution for fifteen to thirty seconds; rinse off methyl violet with iodine solution, flood with fresh iodine solution for thirty to sixty sec- onds, drain off iodine solution and wash with absolute alcohol until stain ceases to come out of film; counter stain for fifteen to thirty seconds with neutral red made up as follows: 1 gram neutral red, 2 cc. glacial acetic acid, 1000 cc. distilled water. With 95 ])er cent alcohol as the decolorizer only one of our dyes gave satisfactory results. Using absolute alcohol four of the dyes, three of the methyl violets and the crystal violet, gave good differentiation. Better results were obtained with acetone or acetone and ether than with absolute alcohol. These results show that all American methyl violet dyes do not give satisfactory results when used according to the method of Jensen. Of the dyes used one of the methyl violet dyes gave distinctly superior results and another distinctly inferior results. The crystal violet dye gave better results than some of the methyl violet and any of the gentian violet dyes used. The dyes were found to vary in the amount of precipitate formed upon the addition of the iodine solution and the rapidity with which this precipitate went into solution in the decolorizer. The dye giving the poorest results produced a heavy precipitate which went into solution slowly. This required longer exposure to the decolorizer which partially accounts for the poor results. One of the essentials of a good dye should be that the precipitate formed with the iodine solution go into solution in the decolorizer very rapidly. An attempt to modify the method of Jensen so that satisfac- tory differentiation can be determined with a larger percentage of American dye products was successful and is given in detail at the end of the article.* We will discuss here only certain •• The methyl violet dye giving the best results, i.e., resisting decolorization the longest, was a sample furnisherl by the Will Corporation of Rochester, New York, and submitted to us as their methyl violet No. 3. With the method of 164 VICTOR BURKE parts of this technique which are deemed worthy of special attention. EFFECT OF SODIUM BICARBONATE WHEN ADDED TO THE PRIMARY STAIN The addition of a few drops of a strong solution of sodium bicarbonate to the dye on the sUde improves the intensity of the stain in the Gram positive organisms. A few drops of 10 per cent lactic acid produces the opposite result. No attempt was made to determine to what these changes are due. We are possibly dealing with changes in osmotic pressure and sub- sequent concentration of the dye in the cell or with a change in the size of the molecule of the iodine-dye precipitate, or simply with a heavier precipitate. The sodium bicarbonate tends to pre- cipitate the dye but the acid does not. The sodium bicarbonate and lactic acid do not produce a permanent change in the cell as is readily shown by exposure of the film to either one of the solutions and then changing the reaction by the addition of the other solution. A film so treated will stain as though the first solution had not been used (Burke, 1921). The effect of the sodium bicarbonate is shown in staining films made from old cultures of Gram positive organisms. Such films stamed by the ordinary methods show many organisms which are Gram negative and some which are Gram amphophile or Gram positive. Similar films stained with sodium bicarbonate added to the violet dye will show a larger percentage of Gram positive organisms. Apparently some of the Gram amphophile organisms have absorbed or retained a larger amount of the dye and appear Gram positive. Sodium bicarbonate does not tend to make a naturally Gram negative organism Gram positive. These facts suggest the possibility, not heretofore recognized, that the loss of the Gram positiveness of organisms in old cultures is not entirely due, as formerly assumed, to autolysis and alt era- staining described in this article satisfactory results were obtained with a crystal violet and methyl violet furnished by the National Stain and Reagent Company of Norwood, Ohio; three methyl violet and a gentian violet furnished by the Will Corporation of Rochester, New York; and a methyl violet and gentian violet furnished by the Harmer Laboratories of Philadelphia. NEW METHOD FOR GRAM STAIN 165 tion of the cell wall but may be due in part to the presence of acid. The application of these facts in the study of slow growing organisms is obvious. Its value in the examination of pus and body fluids is also evident and has been discussed in a separate paper (Burke, 1921). The sodium bicarbonate solution may be omitted from the Gram stain if the best dyes are used. We have found that 3 to 8 drops of a 5 per cent solution of sodium bicarbonate solution is usually sufficient to insure good results. If too much sodium bicarbonate solution is added a heavy precipitate forms and there is an almost complete separation of the dye from the water. This should be avoided. A film forms where a drop of strong sodium bicarbonate hits the dye. This disappears as the dye and sodium bicarbonate are thoroughly mixed. Sodium bicarbonate should not be added to the stock solution of the stain as there results a rapid breaking down of the solution. Whether the sodium bicarbonate is just as effective when added to the iodine solution was not determined. EFFECT OF WATER ON DECOLORIZATION The presence of water on the slide to which the decolorizer is added has a marked effect on the rate and extent of decoloriza- tion. The rate and degree of decolorization in the presence of water depends to some extent, according to the physical concep- tion of the reaction, upon the action of the water on the cell wall of the Gram positive organisms and upon the fact that the precip- itate formed by the dye and the iodine goes into solution in the decolorizer more rapidly if kept moist than when allowed to dry. It follows that in making a stain it is advisable to remove as much water as possible from the slide without allowing the dye precipitate to become dry before adding the decolorizer. Also that after decolorization of the Gram negative organisms is complete water or an aqueous counter stain should not be added to the sUde until the decolorizer has evaporated or been removed. The addition of water to the decolorizer decreases its power to take up the dye precipitate into solution and increases its rate of decolorization of the Gram positive organisms. This 166 VICTOR BURKE is shown by the fact that eighty per cent alcohol decolorizes the Gram negative organisms more slowly and the Gram positive organisms more rapidly than 100 per cent alcohol. If the alcohol is diluted sufficiently both types of organisms decolorize at the same rate. With the addition of more than 50 per cent water the Gram negative organisms do not decolorize over night. The explanation of these facts upon a physical basis is that (1) the precipitate is more soluble in alcohol or acetone than in water and (2) the water alters the cell wall of the Gram positive organism or reduces the size of the molecules of the dye precipi- tate so that the dye is more easily washed out by the decolor- izer. With the addition of suflicient water to the decolorizer the Gram positive organisms or the molecules of the precipitate are so altered that the dye comes out of the Gram positive organisms as readily as out of the Gram negative organisms. Conversely the elimination of water from the decolorizer and from the cell results in the Gram positive organisms retaining the dye more tenaciously than the Gram negative organisms. Unfortunately prolonged exposure to a water free decolorizer will remove the dye from the Gram positive organisms and thorough drying of the precipitate delays its solution in the decolorizer. It follows that if we dry the film too thoroughly before adding the decolorizer the rate at which the dye precip- itate goes into solution ma}'' be decreased to such an extent that before the dye can be washed from the slide and the Gram nega- tive organisms it ma}^, but does not always, begin to come out of the Gram positive organisms. It is not all washed out of Gram positive organisms by acetone in 12 hours. By taking advantage of these facts it is possible by careful blotting of the film before adding the decolorizer to increase the Gram positiveness of the Gram positive organisms without affectmg the rate at which the dye precipitate goes into solution in the decolorizer and the decolorization of the Gram negative organisms. Successful staining with some dyes depends upon the skill with which the water is removed from the film. This constitutes perhaps the most critical step in the process of staining by Gram's method. Success with the poorer .\merican dyes depends upon one's ability to properly gauge the effect of the blotting upon the NEW METHOD FOR GRAM STAIN 167 decolorization of the Gram positive organism and the skilful control of the decolorizer. With the best dj'es it is not necessary to pay special attention to this process although the excess water should always be removed by blotting. There are then, in so far as water is concerned, three facts which should be understood, controlled and utilized in the making of a Gram stain: 1, Water added to the decolorizer increases its power of decolorizing the Gram positive organisms; 2, Water added to the decolorizer slows down the rate at which the dye- iodine precipitate is taken into solution; 3, A dye-iodine pre- cipitate goes into solution before drying much more rapidly than after drying. Since a small amount of water must be left in the film and the addition of water to the decolorizer affects the results it is in- advisable to use the decolorizer more than once as is sometimes done when decolorization is brought about by placing slides in Coi)lin jars. Also the decolorizer takes up but a small amount of the precipitate and quickly becomes saturated and then ceases to decolorize the Gram negative organism. EFFECT OF THE IODINE SOLUTION The addition of the iodine solution causes a heavy precipita- tion of the dye. This precipitate is insoluble in water, but readily soluble in alcohol or acetone and has no staining affinity for cells. It is washed out of the Ciram negative organisms more rapidly than out of the Gram positive organisms. It is washed out of the Gram negative organisms more rapidly and from the Gram positive organisms less rapidly than the unpre- cipitated dye. According to the physical conception of the Gram reaction, the molecules of the Gram precipitate are of such a size that m solution in alcohol or acetone and in the absence of water they do not as readily pass throiigh the limiting membranes of Gram positive as of Gram negative organisms. Neide (quoted by Benians, 1912) states that the potency of the Gram reaction depends largely on the strength of the iodine solution and on the length of the period of application. By using a solution containing double the amount of iodine present 168 VICTOR BXJRKE in "Lugol's" solution or increasing the period of exposure to "Lugol's" solution he was able to make some Gram negative organisms retain the violet dye. He also claimed that heating the iodine solution on the slide tends to make the Gram negative organisms retain the stain . Benians studied the effect of steam- ing the iodine solution for five minutes and found that after this treatment of the Bacterium coli organisms either intact or crushed resisted decolorization with 100 per cent alcohol for five minutes. He assumes that the heat causes a chemical change in the dye- iodine precipitate in the presence of the bacterial cell substance and that the precipitate ceases to be soluble in alcohol. If increasing the exposure to, or increasing the strength of, or steaming, the iodine solution causes the Gram negative organ- isms to retain the dye then these are factors to be controlled in making a Gram stain. However, in our o^vti experiments we were unable to see that increasing the period of application to two hours or doubling the strength of the iodine solution (2^- 100) had any effect on the decolorization of Gram negative organisms. Likewise the steaming of the iodine solution on the slide for five minutes did not effect the decolorization of typhoid organisms. Films so treated decolorized as rapidly as when exposed to the iodine solution for one minute. If, however, the film became dry during the process the Gram negative organisms retained the dye much longer. In our experiments acetone was used as the decolorizer. The former workers used absolute alcohol which may account for the differences in results. Our experiments convince us that if acetone or acetone and ether is used as the decolorizer and the film is not allowed to dry steaming or prolonged exposure to the iodine solution will not materially affect the decolorization of the Gram negative organism. VALUE OF DIFFERENT DECOLORIZERS Alcohol 95 per cent Commercial 95 per cent alcohol is used in many laboratories as the decolorizing solution. It gives satisfactory results with some of the better dyes but it can not be used with the poorer dyes. As we have already shown the NEW METHOD FOR GRAM STAIN 169 addition of water to the decolorizer slows down the rate of the decolorization of the Gram negative and increases the rate of decolorization of the Gram positive organisms. Seventy-five per cent. alcohol decolorizes the Gram positive almost as rapidly as the Gram negative organisms. Fifty per cent alcohol decolor- izes both tyjies of organism at about the same rate. Alcohol absolute. Absolute alcohol gives better results than 95 per cent alcohol as there is a greater margin of time between the decolorization of the Gram negative and Gram positive organisms. The high cost and difficulty of obtaining absolute alcohol miUtate against its use. Alcohol and acetone. The addition of acetone to the alcohol increases the rate of decolorization of the Gram negative organ- isms and slows down the rate of decolorization of the Gram positive organisms. Therefore a decolorizing solution of alcohol and acetone gives better results than absolute alcohol. Acetone. Acetone decolorizes the debris on the sUde and the Gram negative organisms from 5 to 10 times as fast as absolute alcohol and the Gram positive organisms much more slowly. Acetone does not completely decolorize the Gram positive organ- isms if they have been stained with a good dye, in 15 hours. With one of the dyes used by us absolute alcohol decolorized Staphylococcus aureus as much in fifteen minutes as acetone did over night. The addition of water to the acetone has the same effect on the decolorization of organisms as the addition of water to alcohol. Eighty per cent acetone gave as good results as 95 per cent alcohol. Acetone and ether.^ With 100 per cent acetone the decolori- zation of Gram negative organisms is almost instantaneous. If desired this rapid decolorization can be slowed down by the addition of ether to the acetone. One part of ether to 1-3 parts of acetone serves as a very good decolorizer. As ether costs about the same as acetone there is little or no economic advan- ' It is a common laboratory procedure to use a mixture of alcohol and acetone as the decolorizing solution. Lyon (1920) recommends the use of acetone alone as the decolorizer. We have found that a mi.\ture of acetone and ether is just as satisfactory as acetone. 170 VICTOR BURKE tage in adding ether to the acetone. Short exposure of the Gram positive organisms to the ether does not affect their Gram positiveness. ADVANTAGES OF ACETONE AS A DECOLORIZER 1. Reduces time required for the decolorizing process. 2. Gives a greater time period between decolorization of Gram negative and Gram positive organisms. 3. Is cheaper and more readily obtained than absolute alcohol. 4. Gives better results with the poorer dyes. 5. Makes a cleaner slide as the dye is more thoroughly ex- tracted from the debris and clusters of bacteria. 6. Gives better results as it decoloiizes the Gram positive organisms more slowly than alcohol. 7. Makes possible the use of a stronger dye or a mordant such as phenol because any heavy precipitate on the slide is quickly washed away. Alcohol has no advantage over acetone as a decolorizer in the Gram process of staining. When these facts become known we believe acetone will come into more general use as a decolor- izer. As has been emphasized by Lyon (1920) acetone can take the place of alcohol in the preparation of pathological sections and in the cleaning and drying of pipettes and other glassware. Counter stain The choice of a counter stain and the strength of the solution to be used should be determined by a number of factors. With the better violet dyes the choice of a counter stain is relatively unimportant but with the poorer dyes the choice and control of the counter stain largely determines the excellence of the result. Some of the counter stains in general use have a greater tendency than others to mask the violet dye. One has the choice of either, first, shortening the decolorization with the possibility of leaving a trace of the violet dye in the Gram negative organisms and depending upon a strong counter stam or long exposure to cover over any violet dye remaining in the Gram negative organisms or, second, continuing the decolorization until the Gram negative NEW METHOD FOR GRAM STAIN 171 organisms are thoroughly decolorized, with the possibility of reducing the brilliancy of the dye in the Gram positive organisms and relying upon a weak counter stain not to mask the dye in the (jram jiositive organisms. The degree of contrast is the same in both cases but I prefer the former method with a strong counter stain like Safranm O which brings out the Gram negative organisms very distinctly and stains them a color which contrasts more shai-ply with the blackish purple of the Gram positive organ- isms than that of some other counter stains. I find it easier to make a decision and study the morphology if both tjTies of organisms are heavily stained with distinct colors than if both are more weakly stained with distinct colors or if one tjrpe is intensly and the other very faintly stained. XYLOL Clearing the stained film in xylol or turpentine improves the definition, thereby making it easier to separate the Gram positive from the Gram negative organisms and to study the morphology. It is of particular value in the examination of pus and mixed cul- tures. By clearing the organism in this manner instead of depending upon the immersion oil one can more readily determine whether the density of color in some of the questionable organ- isms is due to a mixture of the violet dye with the counter stain, or to a masking of the violet dye by the counter stain, or to an excessive concentration of the counter stain. GRAM REACTION The phenomenon of the Gram reaction has been explained upon both a chemical and a physical basis. The chemical explanation of the reaction is that the dye, iodine and protein of the cell of the Gram positive organisms form a comparatively insoluble compound. The physical explanation is based upon the assumption, supported by some convincing experiments, that the phenomenon depends upon the nature of the cell mem- brane and the size of the molecules in the dye iodine precipitate, the molecules in solution in the decolorizer being unable to 172 VICTOR BURKE pass readily through the cell membrane of the Gram positive organisms owing to the size of the pores. It does not come within the scope of this paper to give a critical analysis of the facts bearing on the above theories. For further discussion the reader is referred to a recent paper by Benians (1920) in which the evidence in support of a physical explanation of the Gram reaction is clearly presented. We wish to describe here the results of attempts to apply to practical staining methods certain of the conceptions presented in Dr. Benians' article. The results obtained have some bearing on the questions involved and may stimulate others to further research along similar lines. Benians divides bacteria into three groups as regards the Gram reaction: (1) Gram positive organisms into which the dye penetrates and from which the dye-iodine precipitate is not readily washed out by the decolorizer; (2) Gram negative organisms hke the gonococcus into which the dye penetrates but from which the dye iodine precipitate is rapidly washed out by the decolorizer; (3) Gram negative organism of the cohform type into which the dye probably does not penetrate and under certain conditions is not even absorbed into the surface of the cell and which are therefore readily decolorized. According to the physical conception of the Gram reaction the cell membrane of the Gram positive organism does not allow the passage of the compound dye-iodine molecule when in solution in the decolorizer. If the dye-iodine precipitate in solution in the decolorizer can not pass out of the cell due to the character of the cell wall then we are justified in assuming that under similar conditions the dye-iodine precipitate can not pass into the Gram positive cells. We can also reasonably assume that since the dye-iodine precipitate readily passes outward through the wall of Gram negative organisms like the gonococcus it will just as readily pass inward through the cell wall in so far as physical conditions operate. It follows then that if we expose films of staphylococcus and gonococcus to an alcoholic solution of the dye-iodine precipitate the dye-iodine compound should penetrate the gonococci but not the staphylococci. Also since 75 per cent NEW METHOD FOR GRAM STAIN 173 alcohol decolorizes the Gram positive organisms very rapidly a solution of the Gram precipitate in 75 per cent alcohol should stain the Gram positive organisms almost as rapidly as the gonococcus like organisms. If this dye-iodine compound could be forced to remain in the gonococci we could stain the films with a weak counter stain which would stain the staphylococci and we woiild have the phenomenon of a reversed Gram stain. The practical application of such a staining method in the examina- tion of pus for gonococci is obvious. Unfortunately the addition of iodine to the dye saturates its affinities so that an alcohol or acetone solution of the dye-iodine precipitate does not stain cells. By simply exposing films of organisms to the solution we can not determine whether the molecules of the precipitate have penetrated into the gonococci and not into the staphylococci. The staining power of the solu- tion can be materially increased by the addition of alkaU but we do not know what affect this has on the size of the molecules. If the molecules are reduced in size they should enter the Gram positive as well as the Gram negative organisms. If the size of the molecular groups is not altered by the addition of the alkali then the gonococci should be penetrated and stained and the staphylococci not penetrated and not stained or only the surface stained. Exposure to a counter stain or weak decolorizer and counter stain should stain the staphylococci and not the gonococci. ATTEMPT TO REVERSE THE GRAM PHENOMENON Experiment 1. Films of pure cultures of Staphylococcus aureus, Neisseria catarrhalis and Bacterium typhosum were made on a slide in physiological salt solution, air dried and fixed by heat. A sufficient amount of iodine solution was added to a quantity of an aqueous solution of methyl violet to cause a maximum precipitation, the precipitate washed to remove excess iodine and dried. The films were flooded with a saturated alcoholic solution of this precipitate. A few drops of a strong mixture of sodium bicarbonate, sodium phosphate and sodium h3'droxide were added to the dye on the slide. After an exposure 174 VICTOR BURKE of 10 minutes or more the slide was dipped in water to remove free dye and lightly counter stained with aqueous Safranin O. Upon examination the staphylococci were Gram negative and the other two types of organisms weakly Gram positive. This experiment suggests that the dj^e-iodine precipitate in solution in alcohol with the addition of alkali more readily penetrates the Gram negative than the Gram positive organisms used. The difference noted apparently is not due to a more rapid decolorization of the staphylococci by the water. If the films are examined after washing and before exposure to the Safranin solution the three types of organisms appear to be equally well stained. Since the counter stain more quickly stains the staphylococci than the other two organisms we assume that the staphylococci are less heavnly stained by the dj'^e-iodine precipitate or are only surface stained rather than that the Safranin has a greater affinity for the staphylococci. Staining by this method, as controlled at the present time is entirelj'' inadequate from a practical point of view for distin- guishing staphylococci from gonococci in mixed infections. The results obtained are not uniform and the degree of differentiation is not sufficient. EXPERIMENTS TO DETERMINE WHETHER THE PRIMARY STAIN PENETRATES THE CELL WALL OF TYPHOID LIKE ORGANISMS Benians' conception of two types of Gram negative organisms as described above is based upon two experiments as follows: 1. If gonococci and Bacterium coli, unfixed by heat, are shaken up in weak solution of methyl violet (1 to 40,000) and then centrifuged, the gonococci are throwni douai well colored and the dye is cleared out of the solution while the Bacteriuvi coli are thrown down uncolored leaving the whole of the dye in solution. If the Bacterium coli organisms were boiled or killed at 65° for thirty minutes they absorbed the dye and were thrown down well colored. The Bacterium coli organisms exposed to 60°C. for thirty minutes did not absorb the dye. If the unfixed bacilli were suspended in strong solutions they l^ecame deeply colored. 2. When films of Bacterium coli organisms and Bacterium coli NEW METHOD FOR GRAM STAIN 175 organisms ground in a mortar were placed on the same slide, fixed by heat in the usual way, treated for two minutes with 0.5 aqueous solution of methyl violet and then deeolorized with 95 per cent alcohol it was found that the dye was held much more strongly by the ground up debris of the organism than by the intact. According to Benians "This seems to provide almost certain evidence that the dye had never really permeated the intact bacilli, to get into their substance, as it had been able to get into the substance of the bacilli when broken up. The dyes were therefore only absorbed to the exterior of these intact organisms." We suggest here that the difference in rate of decolorization between the intact organisms and the amorphous material may have been due to a greater saturation and more rapid drj'ing of the amorphous material rather than to an entire lack of penetra- tion of the dye into the interior of the intact cells. It seemed important to us, from a purely practical considera- tion of the subject, to determine whether all the members of the large coli-typhoid-dysentery group of organisms resisted penetration of the dye as had been demonstrated for Bacterium coli and to such an extent as to permit of differential staining between these organisms and the gonococcus like organisms. Our experiments were based on three assumptions: (1) That if the cell wall of these organisms resisted penetration by the dye it should resist decolorization of the cell if the dye could be gotten into the cell without altering the cell wall and the decolor- izer did not affect the cell wall; (2) That if the organisms were stained with methyl violet, then stained with a dye such as Safranin which slowly covered over the methyl violet and then cleared in xylol the cells should not show a violet center and a Safranin periphery unless the violet had penetrated to the interior of the cells; (3) That if the organisms were stained with methyl violet, partially decolorized with acetone and counter stained with Safranin the center of the cells should not appear violet and the margin of the cell Safranin unless the violet had pene- trated to the center of the cells. 176 VICTOR BURKE Experiment 1. Films of Bacterium typhosum, Staphylococcus aureris and Neisseria catarrhalis were air dried and stained over night in a 0.5 per cent aqueous solution of methyl violet, then exposed to the decolorizing action of water, weak alcohol, 100 per cent alcohol, acetone, and chloroform. In all cases the typhoid organisms decolorized as rapidlj^ or more rapidly than the other organisms. Experiment 2. The experiment was repeated with carbol- fuchsin as the staining solution; similar results were obtained. Increasing the period of exposure over the usual two minutes period, increased the time required to bring about decolorization. Apparently the increased exposure brought about a greater concentration or greater penetration of the dye in the cell. There is of course the possibility of a chemical change resulting from the long exposure. Experiment 3. The above experiments were repeated but with the appUcation of an iodine solution in the usual manner. The typhoid and catarrhalis organisms decolorized at about the same rate and much more rapidly than the staphylococci when exposed to acetone. From these experiments we conclude that after the dye pene- trates typhoid organisms the cell wall offers no greater resistance to the removal of the dye than the cell wall of Neisseria catarrh- alis and Staphylococcus. We have not demonstrated that the dye penetrated the cells but since a similar exposure will result in the staining of acid fast organisms we assume that the dye did penetrate. The following experiments favor this assumption. Experiment 4- Films of Bacterium typhosum, Neisseria catarrhalis and Staphylococcus aureus were air dried, stained for two to three minutes in a 0.5 per cent aqueous solution of methyl violet, washed in water to remove excess stain, stained lightly with 2 per cent aqueous Safranin O, washed and cleared in xylol for ten minutes. Upon examination the staphylococci almost uniformly showed a safranin colored border with a dark center. The safranin had penetrated part way into the cells and masked the violet dye. The Neisseria catarrhalis organism resembled the staphylococci with the exception that the Safranin had not NEW METHOD FOR GRAM STAIN 177 penetrated so deeply. The typhoid organism showed less pene- tration of the Safranin than the staphylococci, in this respect resembling Neisseria catarrhalis, but an occasional organism showed a very definite violet center with a Safranin margin. The Safranin masked the violet dye in the staphylococci more rapidly than in the other two organisms. Experiment 5. The experiment was repeated with films fixed by heat in the usual manner. The results were similar to that obtained with the unfixed films. Experiment 6. The experiment was repeated with unfixed films with the exception that the smears were partially decolor- ized by a brief exposure to acetone and ether (equal parts) before being stained with the Safranin. Upon examination many of the typhoid organisms showed a very definite purple center surrounded bj' a Safranin margin. Some of the typhoid organisms showed irregularly spaced blackish granules in the center somewhat resembling a string of cocci and surrounded by a Safranin colored margin. Other of the typhoid organisms appeared uniformly Safranin colored. As far as the eye could determine the violet dye had penetrated to the center of many if not all of the typhoid organisms. We conclude from these experiments that in ordinary Gram staining the violet dye penetrates through the cell wall of Bacter- ium tijphosum. Our experiments failed to demonstrate that' the cell membrane of typhoid like organisms oilers much if any greater resistance to the dyes than the cell wall of Neisseria catarrhalis. The use of weaker dyes might bring out difference* not e\adent when stronger dyes are used. The solution of methyl violet used by us was of the same strength as used by Benians in one of his experiments with Bacterium coli in which he apparently demonstrated that the dye did not penetrate to the center of the cell.' • With the Gram stain used by us the typlioid organisms decolorized slightly more rapidly than the A^eisKeria catarrhalis organisms but as the latter were distinctly larger the difference in rate of decolorization may have been due to the difference in size. 178 VICTOR BtTRKE These experiments were discontinued because they did not give promise of demonstrating a method by which we could stain the colon-typhoid group of organisms differently from the gonococcus like organisms. A MODIFIED GRAM STAIN The following method of making a Gram stain with aqueous solutions of dyes and without the addition of mordants to the stock solution of the primary stain gives better results than any other method known to us. With this method satisfactory differentiation between Gram positive and Gram negative organ- isms can be obtained with many of the domestic gentian violet, methyl violet and crystal violet dyes. With some of the better dye products certain of the steps in the process can be omitted and others modified but with the poorer dyes strict attention to all of the details given is advisable. 1. Air dry thinly spread film and fix with least amount of heat necessary to kill the organisms and fix them to the shde (A) . 2. Flood smear with a 1 per cent aqueous solution of the dye to be used. Mix with the dye on the slide 3 to 8 drops of a 5 per cent solution of sodium bicarbonate, allow to stand two to 3 minutes (B). 3. Flush off the excess stain with the iodine solution ' and cover with fresh iodine solution and let stand one minute or longer (C). 4. Wash in water as long as described and blot off all free water until surface of film is practically free of water, but do not allow the fihn to become dry (D). 5. Decolorize with acetone or acetone and ether (1 part ether to 1 to 3 parts acetone) until decolorizer flows from slide prac- tically uncolored. This usually requires less than ten seconds (E). 6. Blot dry. The slide quickly dries without blotting (F). 7. Counter stain for five to ten seconds or longer if desired with a 2 per cent aqueous solution of Safranin O (G). 8. Wash off excess stain bv short exposure to water, blot and dry (H). 'One gram iodine, 2 gnims potassium iodide, 100 cc. distilled water. NEW METHOD FOR ORAM STAIN 179 Inunerse in xylol or turpentine for several minutes or until clear. Examine. If the first attempt at staining a smear does not give satisfac- tory results it is advisable to wash off the oil with xylol, wash off the xylol with acetone and restain. It has been our experience that restaining smears gives better results than the original attempt. A. The film can be made in either distilled or tap water or physiological salt solution. B. Some workers recommend cooling the slide before flooding with the dye. With some dyes steaming seems to improve the result or shorten the required period of exposure. Passing the slide through the flame until steaming Iiegins and allowing to stand the two minutes is sufficient. Steaming does not cause the Gram negative organism to resist decolorization. Anilin- gentian-violet can be used in place of the aqueous solution if desired. .Vllowing the stain to dry around the edge makes a dirty slide but does not affect the Gram reaction. The strength of the solution of the dye and the period of exposure can vary somewhat without affecting the result. A saturated instead of a 5 per cent solution of sodium bicarbonate may be used. C. The excess stain can be blotted off or washed off by a brief exposure to water. The exposure to the water should be as brief as possible as water tends to reduce the amount of dye in the cells. Washing with water has the advantage of giving cleaner slides and effecting a saving of the iodine solution and the decolorizer. D. The iodine solution can be blotted from the slide but this has the disadvantage of leaving a small amount of iodine on the slide and with the volatilization following the addition of acetone there is some irritation of the exposed mucous mem- brane of the worker. If for any reason one can not complete the staining process after reaching this stage it is advisable to place the slide in water until the staining can be completed. E. The decolorizer should be placed upon the slide, allowed to stand for a few seconds and drained off. Then fresh decolorizer allowed to flow over the surface of the slide until it drops off 180 VICTOR BURKE clear. Proper control of this process will reduce the amount of decolorizer used to a minimum. Placing the decolorizer in a Coplin jar and dipping the slide up and down in it will not give satisfactory results. F. Drying after the decolorizer is essential as the aqueous counter stain mixing with the decolorizer has an effect on the Gram positive organisms. The smear can now be exainined for Gram positive organisms. The oil can then be washed off with xj'lene, the slide dried and the counter stain applied. G. Various counter stains can be used but I prefer Safranin O, or neutral red as recommended by Jensen. With dyes giving a poor Gram reaction it is necessary to reduce the counter stain- ing to a minimum. H. Washing should be sufficient to remove the dye from the surface of the organisms, flushing for a few seconds will suffice. CONCLUSIONS 1. Steaming the iodine solution on the slide or increasing the period of exposure does not cause the Gram negative organism to resist decolorization. 2. Drying the film after exposure to the iodine solution greatly delays decolorization of the Gram negative organisms. 3. There are two critical steps in the Gram staining method; (a) the removal of the water after the iodine solution and (b) the decolorization. The control of these steps determines to a large extent the amount of differentiation between the Gram positive and Gram negative organisms. 4. The addition of water to the decolorizer, either on the slide or in the bottle, retards the decolorization of the Gram negative organisms and increases the rate of decolorization of the Gram positive organisms. 5. Acetone is superior to absolute alcohol as a decolorizer. It decolorizes the Gram negative organisms and debris on the slide much more rapidly and the Gram positi\"e organisms more slowly than absolute alcohol. 6. Mounting the fihns in distilled water, tap water or physio- logical salt solution does not affect the staining reaction. Mount- NEW METHOD FOR GRAM STAIN 181 ing the films in sodium bicarbonate or lactic acid greatly affects the result. 7. The addition of sodium bicarbonate results in a greater con- centration of the methyl violet dye being present in the Gram positive organisms after decolorization and lactic acid brings about the opposite result. The failure of Gram positive organ- isms in old cultures and in smears from the genital-urinary tract to retain the violet dye may be due in part to the presence of certain acids. These facts suggest the possibility of enhancing the value of gentian violet in selective media and improving dye therapy by the addition of an alkali. 8. All of the colon-typhoid group of organisms do not differ from the gonococcus-catarrhalis group of organisms in their resist- ance to the penetration of the methyl violet dye used in the Gram stain. Results of an attempt to reverse the Gram reaction are described. There is described a modified Gram stain which, without the use of mordants or alcohol, gives very good differentiation when using many of the American dye products. RECOMMENDATIONS 1. That the Comniittee on Pure Culture Study of the Society of Bacteriologists mstead of attempting to select any one Gram method to be used as the standard method in the study and description of pure cultures designate the control organisms to be used in checking up the stain. Any Gram method •jvhich would give sharp differentiation between the two control organ- isms could be considered a satisfactory Gram stain to be used in the study of pure cultures. The selection of the control organ- isms will require considerable study as the organisms selected should have the same size and morphology when grown under identical conditions and should require a first class stain to bring out sharp differentiation. The custom of using staphylococcus and typhoid as control organisms is not satisfactory. The control organisms selected could be sent out to the different lOnSNAI. or BACnSBlOLOOT, VOL. Til, HO. 2 182 VICTOR BURKE laboratories. We believe this method will bring about more uniform results than the selection of any standard Gram method. 2. That the dye manufacturers attempt to improve the dyes used in the Gram stain by the addition of alkali which will not affect the stability of aqueous solutions of the dye. REFERENCES Atkins, K. N. 1920 Jour. Bact., 6, 321. Benians 1912 Jour. Path, and Bact., 17, 201. Benians 920 Jour. Path, and Bact., 23, 401. Burke, 1921 Jour. A. M. A. 77, 1020. Conn, H. J., and others 1919 Jour. Bact., 4, 112. Jensen 1912 Berl. klin. Wochenschr., 49, 1663. Lton 1920 Jour. A. M. A., 75, 1017. DISINFECTION STUDIES' THE EFFECTS OF TEMPERATURE AND HYDROGEN ION CON- CENTRATION UPON THE VIABILITY OF BACT. CO LI AND BACT. TYPHOSUM IN WATER BARNETT COHEN Division of Chemistry, Hygienic Laboratory, United States Public Health Service, Washington, D. C. Received for publication May 21, 1921 Exposure to extremes of temperature and of hydrogen ion concentration produces an accelerated death rate of bacteria. These are typical conditions of ordinary disinfecting procedures. Temperatures and hydrogen ion concentrations in the zones between those which are favorable and those which are distinctly lethal may be expected to produce an adaptation in the organisms or a comparatively slow death rate, knowledge of which may reveal the conduct of bacteria under moderately unfavorable conditions in nature, and add to our understanding of certain phases of disinfection. The object of this study was to investigate the effects of varia- tions in moderate temperature and moderate hydrogen ion con- centrations upon the death rate in water and dilute buffer solu- tions. The experimental method imposed sub-lethal conditions upon the bacteria in contradistinction to the accelerated death induced by ordinary disinfecting procedures. It has been found in unbuffered media like distilled water or tap water that the death rates vary coincidently with apparently unimportant shifts in hydrogen ion concentration. When the pH factor is controlled by means of M/500 buffers, the death rates are stabilized so that comparisons become possible. The ' Published by permission of the Surgeon General. Thesis submitted in partial fulfilment of the requirements for the degree of Doctor of Philosophy at Yale University. 183 184 BAENETT COHEN pH zone of tolerance, or minimum death rate, of Bad. typhosum lies between 5.0 and 6.4; that for Bact. coli is wider, and centered near absolute neutrality. The effect of sub-lethal factors is to make a period of induction apparent before death proceeds at a logarithmic rate. The phenomena of growth, maintenance and death of bacteria are fundamentally important to bacteriology, theoretical and applied; and there are conceivably numerous factors that may influence them. Omitting from consideration in this discussion the obviously large part played by the food supply, we recognize certain predominating physical and physico-chemical factors, chief among which are temperature and the concentration of hydrogen ions. In bacterial death, and in disinfection, which may be regarded as a special phase of bacterial death, their role may indeed become paramount. Temperature plays its important part in the Ufe and death processes of bacteria by controlling the active agencies involved. Referring to the death processes Clark and Lubs (1917) have said: ". . . . in cellular destruction temperature is to be considered as an accelerating condition .... among the active agents concerned the concentration of hydrogen ions may be of great significance." This opinion is amply supported by experimental evidence reaching back to the days of Pasteur. As an aid to the study of certain phases of biological processes, the effect of temperature has been widely used and has yielded facts of importance. It has been found empirically that tem- perature augments physical phenomena in arithmetical progres- sion and chemical phenomena in geometric ratio. As a rule the temperature coefficient of velocity of most chemical reactions for a rise of ten degrees is 2 or more ; and for physical phenomena it is nearer 1. Snyder (1908, 1911) cites numerous examples of these two types of temperature coefficient and shows that in most physio- logical processes the temperature coefficient of chemical reaction velocity appUes. Loeb (1908) ingeniously accounts for the much richer animal and plant life of Arctic waters by reasoning that since a temperature decrease of ten degrees increases via- DISINFECTION STUDIES 185 bility 1000 times, and of twenty degrees, one million times, while the rate of development is reduced one-third to one-ninth, it therefore follows that at 0°C. many more successive generations may exist simultaneously than at 10° or 20°C. Madsen and Streng (1910) found that the conservation of agglutinins was affected in a like manner; and Kanitz (1915) in his monograph on the subject collects many of the data for convenient comparison. The role of temperature in the growth of bacteria has been recognized from the very beginning. Its effects are found alike in those life phases in a bacterial culture characterized by Buchanan (1918) as the logarithmic growth phase, the maximum stationary phase, and the logarithmic death phase. The present experiments extend the application to the period of accelerated death. Ward (1895) by painstaking bacterioscopic studies worked out a curve showing the relation between the rate of growth of B. ramosus and the temperature of its environment. Barber (1908), doing the same thing with Bad. coli, found growth at 30°C. increased two to three times over that at 20°C., and Lane-Claypon (1909) confirmed this observation. Slator (1919) demonstrated the same relationship to hold for the growth and maintenance of yeasts. Houston (1914) showed that as the temperature was decreased the viability of Bad. coli and Bad. typhosum in water was in- creased. The crude experiments of Konradi (1904) showed a similar relationship for staphylococci and other organisms; and Li\'ingston (1921) finds this true for hemolytic streptococci. The magnitude of the temperature effect observed by these authors may only be inferred, however, for their data are mainly qualita- tive in nature. Paul, Birstein and Reuss (1910) give quan- titative data upon the viability of staphylococci and find a tem- perature coefficient of 2 to 3 for a 10° rise. In the process of disinfection we have numerous examples and a mass of exact data that show the general application of this rule. In this con- nection we need only mention the classic researches of Paul and Kriinig (1896), Madsen and Nyman (1907), Paul (1909) and of Chick (1908, 1910, 1912). In the heat steriUzation of bac- 186 BARNETT COHEN terial spores, Bigelow and Esty (1920) find that the time of steri- lization is increased ten times for a 10° reduction in sterilizing temperature. The Committee on Standard Methods of Examin- ing Disinfectants of the American Public Health Association in its last report (1918), has made reconm:iendations for the inclu- sion of the temperature coefficient as one of the three necessary items in the characterization of disinfectants. As has been stated above, the relation between temperature and the speed of chemical reactions is as yet upon an empirical basis, van't Hoff (1896) attempted a derivation from ther- modynamic considerations but obtained no definite solution. Arrhenius (1889) assumed that reacting substances occur in two tautomeric forms, "active" and "passive," and that a certain quantity of heat is involved in activation. He suggests a rela- tion of the form: Vi = vo-e for the velocities of reaction, 2^0 and Vi, at absolute temperatures To and Ti, the quantity q being the constant of activation. Numerous empirical formulae of the same general form have been suggested from time to time; and that of Snyder (1911) is one that he applied to physiological processes. Tolman (1921) dis- cusses the views that have been held and points to the lack of a real fundamental explanation of the temperature effect in monomolecular reactions. Dushman (1921) and Lewis and McKeown (1921) offer theories of chemical reactivity which appear to be based upon fundamental considerations and include the effect of temperature. There is an ever-increasing literature upon the effects of hy- drogen ion concentration' in many biological processes, as a glance at the references quoted by Clark (1920) wiU show. In the field of bacteriology its importance as a controlhng factor is definitely established. This control is exercised in many subtle and unexpected ways — upon the activity of specific ' In this work, the Sorensen symbol pH is used synonymously for the hydrogen ion concentration. DISINFECTION STUDIES 187 enzymes, upon toxin production, upon the dissociation of essen- tial foodstuffs and upon the state of aggregation of cellular pro- toplasm. Our present meager knowledge permits us usually to see only the end result in growth, metabolism or death. Cohen and Clark (1919) found that the acid limit for the growth of Lactobacillus bulgaricus is not identical with that for fermenta- tion, thus suggesting a possible distinction between bacterial growth and metaboUsm. In the experiments to be reported in this work there is a hint that we might extend this distinction to include bacterial death. These factors, temperature and hydrogen ion concentration, have been utilized in the present investigation to study some of the characteristics of that phase of disinfection produced by mild factors, which for want of a better term we have called sub-lethal. Beginning with the first systematic studies by Koch (1881), who was later followed by Paul and Kronig (1896), Kronig and Paul (1897), ^ladsen and Nyman (1907), and culminating in the achievements of Chick (1908, 1910, 1912), there has been evolved a theory of disinfection based upon well known principles of physical chemistry. This may be summarized by the dictum of Phelps (1911): "The rate of dying, whether under the influ- ence of heat, cold or chemical poison, is unfailingly found to follow the logarithmic curve of the velocity law, if the tempera- ture be constant." Lee and Gilbert (1918) come to a like con- clusion after a critical investigation. On the other hand there are some like Reichel (1909), Loeb and Northrop (1917), Brooks (1918), Peters (1920) and Smith (1921) who urge that the logarithmic process is only an apparent one, and that careful study of the intimate nature of the disin- fection curve will show it to be dependent upon the individual resistance of the organisms. Brooks supports the contention of Loeb and Northrop that bacteria are distributed according to resistance upon a probability curve and that bacteria of low resistance die off first. Since the logarithmic law takes cogni- zance of the number of organisms without regard to the distribu- tion of resistance among them, the operation of this law among bacteria would, so they claim, be an unnatural process. 188 BAENETT COHEN To throw some light upon this controversy and at the same time aid in an understanding of the mechanics of the disinfection process, it seemed that a study based upon the following con- siderations might prove useful. ]\Iost pre\'ious studies of disin- fection were based upon results of the appUcation of an intense disinfecting agent, like heat, heavy metals, acids, alkalis, etc. As a result, the response of the organisms was prompt and the period of their adjustment to the new conditions was so small as to be overlooked. Only in exceptional cases, as the one cited by Chick, where old cultures of Staphylococcus aureus were used, was the early period long enough to excite attention; but in that case the cultures used could not be considered as fairly homogeneous. If the conditions causing death of the organism could be toned down in intensity to a degree that might be termed sub-lethal, then there should be an opportunity to observe their adjustment from their manner of response. The reduction of lethal intensity may be satisfactorily accom- plished by maintenance of the organisms at sufficientlj' low tem- peratures on the one hand and by the control of the hydrogen ion concentration of their environment on the other. EXPERIMENTAL The series of expermients reported and discussed below was planned to give information on the following aspects of the behavior of bacteria as represented by the colon-typhoid organisms. 1. The response of bacteria to the sub-lethal factors (in con- tradistinction to disinfection as ordinarily understood) of starva- tion and moderate intensities of hj^drogen ion concentration. 2. The effect of moderate temperatures upon the rate of this response. 3. The analysis of the behavior in the Ught of the physico- chemical concept of the disinfection process. A statistical method of approaching a solution of these ques- tions appears at present to be the only satisfactory one; and briefly stated, the method of study pursued was to expose Bad. coli or Bad. typhosum to a given solution and follow periodically DISINFECTION STUDIES 189 the numbers of survivors capable of forming colonies on nutrient agar. Materials and apparatus Close attention was paid to obtaining unifomiity of conditions throughout the work. Constant temperatures (10°, 20°, and 30°C.) were secured m an electrically heated and controlled, well-stirred air thermostat, with fluctuations probably not greater than 0.5°C. In the experiments conducted at 0°C. the bottles were maintained at the temperature of melting ice. The plating 7natcrials were of the usual kind, consisting of graduated pipettes, petri dishes, dilution bottles and tubes, and nutrient agar. The glassware was cleaned in the usual manner and sterilized by dry heat at 160°C. for five hours or longer. The dilution bottles were of about 250 cc. capacity, and were filled from an automatic burette with 100 cc. of distilled water. The dilution tubes were filled with 10 cc. of water. Sterilization in the auto- clave brought the contents down on the average to 99 cc. and 9 cc. respectively. The procedure in making distilled water the suspension-fluid or the diluting fluid is at variance with the customary practice of using so-called physiological salt solution or mixtures of this solution with nutrient broth. This variation was made advisedly because preliminary tests had shown quite conclusively that for our purposes distilled water was a satisfactory neutral fluid' (cf. Zeug, 1920). The nutrient agar for plating was made by a uniform method at various times from a single lot of Difco proteose peptone and Liebig's beef extract. It contained per litre : grams Peptone 10 Liebig's beef extract 3 Shred agar 20 The reaction of the medium was adjusted colorimetrically to pH 7.0. Occasional tests of the reaction of the nutrient agar at the time it was actually used for plating showed the hydrogen ion concentration to be uniformly between pH 6.7 and 6.9. 190 BAENETT COHEN The distilled water of the laboratory, from a gas-heated Stokes automatic still, was used in all cases except where otherwise mentioned. In experiments requiring water of exceptional purity the latter was prepared by double distillation out of acid and alkali in Pyrex glass, with precautions for the exclusion of car- bon dioxide. The tap water used was Potomac River water which had pre- sumably been treated with alum and filtered through sand. Ficker (1898) has shown that an "oligodynamic" property is readily acquired by water which is allowed to remain in contact for any long time with the metallic fixtures in ordinary plumbing. To prevent this occurrence in our water, the latter was allowed to run freely for several hours before being used for experiment. During filtration, a by-pass provided a constant flow of fresh tap water under a head of one meter. The buffer solutions were prepared with the precautions de- scribed by Clark (1920)* The only divergence made here was to dilute these buffers to a concentration of M/500. The pH values of these dilutions were then determined colorimetrically. Tests were made to determine the minimal amount of buffer necessary to maintain constant hydrogen ion concentrations under the conditions of these experiments; and it was found that the M/500 concentration of buffer answered this purpose most satisfactorily. Tlie bottles in which the bacterial viability was studied were of approximately 1000 cc. capacity, of ordinary glass and with ground stoppers. Before use, they were thoroughly cleansed with fresh chromic acid cleaning mixture, rinsed, steamed in an Arnold sterilizer for several hours, and well rinsed with distilled water. Except for the experiments with double-distilled water, in which cases the containers received an internal coating of puri- fied paraffin, no attempt was made to prevent possible solution of the glass in the contained water. This procedure was followed because in preliminary experiments a comparison of the \'iabiUty curves from water in Jena glass containers and in the above softer glass bottles showed no appreciable differences. A\Tien DISINFECTION STUDIES 191 precautions were taken during manipulation to avoid unneces- sary and excessive heating, the amount of glass constituents dis- solved apparently played no significant part in the experiments as conducted. Paraflin melting at 55°C. was boiled with several changes of distilled water, stirred frequently and finally skimmed off into clean sterile bottles. The latter were then placed in a hot air sterilizer for several hours and later allowed to cool while a uniform coat was deposited internally. The organisms selected for experimentation were two members of the colon-typhoid group, Bad. coli and Bad. iyphosum. These were chosen both because they are well suited for this type of study; and because they have been extensively studied along related lines so that results obtained here may be readily com- pared and may add to the continuity of our knowledge. Both cultures had been grown for a long time upon artificial media but this does not constitute a defect in the present ex- perimental plan. The undoubted acquisition of higher resist- ance to external influences altered somewhat the degree of the mortality rate by accentuating and magnifying the retardation during the early phases of the process, precisely the condition desired. Bact. coli. This organism was isolated from a polluted stream in 191G. Its cultural and morphological characters are typical of the Bad. coli-communis (sucrose-negative) type. Winslow and Talk (1918, 1920) have utiUzed this strain in their studies on salt antagonism. Bad. typhosum. This was a culture of the Rawlings strain (no. (308) obtained from the American Museum of Natural His- tory, New York. Stock cultures were preserved on nutrient agar at 10°C. At least five daily transplants on agar at 37° were made before a culture was used for experiment. The procedure in making the tests. The requisite number of bottles were filled with buffer solution or water, allowing an air space of about 100 cc. and the whole sterilized at one time in the autoclave at 15 pounds pressure for ten minutes. Exceptions 192 BAKNETT COHEN to this procedure were made in the case of the double-distilled water and the filtered tap water experiments. The double- distUled water was freshly distilled directly into sterile paraffined bottles, and the filtered tap water was filtered through a Berke- feld (N) porcelain candle into sterile bottles. The bottles were then placed in the constant temperature box at 0°, 10°, 20° and 30°C., respectively. DupUcate bottles were maintained at each temperature, and kept there at least 12 hours to assume their required temperatures before the beginning of the experiment. Cultures of the organisms to be studied were grown on agar slants at 37° for sixteen to eighteen hours. A platinum loopful of the growth was carefully removed and shaken into 9 cc. of sterile distilled water. Precaution was taken to remove only the bacterial growth and to avoid taking up any of the medium. A homogeneous suspension of the organisms was then obtained by thorough shaking. The tube was allowed to stand for several minutes to permit sedimentation of particles and the supernatant fluid was used to inoculate the bottles. The bottles, innnediately after being inoculated, were thor- oughly shaken to distribute the organisms evenly throughout the volume, and samjiles were taken to determine the bacterial content by remo\Tng 1 cc. of the suspension and plating it in suitable decimal dilutions. Duplicate plates were made of each dilution and, usually, there were 3 or 4 dilutions and sometimes more, of a single sample. After a suitable period of incubation, usually 48 hours at 37°C., the bacterial colonies on the plates were counted. For the organisms maintamed at 0°C. it was necessary to incubate for a longer period to permit growth of colonies of adequate size. The plates were counted and the results of dupUcate plates averaged and noted as the bacterial count for the partic- ular moment that the sample was taken. Bottles were removed from their respective temperature surroundings only for minimal intervals, not more than a minute or two, usually. Otherwise, they were maintained, protected from light, at their respective temperatures. Successive sam- ples to determine the bacterial content were taken at regular DISINFECTION STUDIES 193 intervals, and at each such time the bottles were thoroughly- shaken to secure an even distribution of the bacteria. Samples were also taken regularly for the colorimetric deter- mination of pH and for the analysis of the absorbed gases. The determination of absorbed gases in the water was carried out as follows: 5 cc. of the sample were transferred quickly to a Van Slyke (1917) gasometric CO2 apparatus. The contained gases were extracted, CO2 determined by means of absorption with KOH, the oxygen determined by means of absorption with alkaline pyrogallol and the volume of the residue noted. Pre- liminary analyses of air and of air-saturated water showed the method to be satisfactory for this purpose. By these procedures, a study w-as made of the viabiUty of Bact. tijphosum and Bact. coli at 0°, 10°, 20° and 30°C. in double- distilled water, in tap water and in dilute buffer solutions. The experimental results In early experiments, the mortality of the organisms was often variable, and could be attributed possibly to the presence of soluble constituents from the glass containers. It was first suspected that some salt effect perhaps might be responsible. The results of the first four experiments show clearly the marked effect upon the mortality which was coincident with apparently small variations in hydrogen ion concentration. Behavior in unbuffered solutions. The result of Experiments 1 to 4 may be most conveniently considered under one head. In these experiments Bact. typhosum and Bact. coli were each exposed to distilled water and to tap water at 0°, 10°, 20° and 30°C. The course of events is graphically presented in the accompanying charts (figs. 1 to 4). All the figures are plotted with time as abscissae and the logarithms of survivors as ordinates. The resulting curves show the actual rate of decline in numbers. That is, the slope of a curve (the graphic equiva- lent of the velocity coefficient, k) at any point is indicative of the speed of disinfection and is strictly comparable with the slope of any other curve on the same chart. The experimental points are connected by straight hnes without effort to draw "smoothed" curves through greatly Fig. 1. Experiment 1. The Death Rate op Bact. typhosum in Double- Distilled Water at 0°, 10° 20° 30°C. Logarithms of numbers of survivors are plotted as ordinates against time intervals as abscissae. Duplicates at any temperature are marked (I) and (II). Observed pH values of the water are noted along the curves. o Days Fio. 2. Experiment 2. The Death Rate of Bact. coli in Double-Distilled Water at 0°, 10°, 20°, 30°C. Duplicates at any temperature arc marked: (I) and (II). Observed pH values of the water are noted along the curves. 194 6 i' « ( Sl S \ ^ ^ ^^ 23 > t \ k] ^"**«s^ R "^^^^ 97 * ,1 1 i S \, ^'^"^ ■^ Uf « icr ; ^ > K "^2/ \ 9.3 / «> ^ ^^^ ,(J) 30* ^3 ODzo- N^^ 0 -iJdtrti b/m stf (11) icr Fio. 3. Experiment 3. The Death Rate of Bact. ttphos0m in Autoclaved Tap Water at 0°, 10°, 20°, 30°C. Duplicates at any temperature are marked: (I) and (II). Observed pH values of the water are noted along the curves. o Coyi 6 Fia. 4. Experiment 4. The Death Rate of Bact. coli in Berkefeld-Filtered Tap Water at 0°, 10°, 20°, 30°C. Duplicates at any temperature are marked: (I) and (II). Observed pH values of the water are noted along the curves. 195 196 BARNETT COHEN deviating points. It is true that the error in the experimental method employed is not inconsiderable and that a smooth curve might ade- quately represent the course of events occurring but there is also a danger of the eye being misled by an apparent regularity where none may exist. The well-known higher resistance of Bad. coli is here again exemplified. As will be seen further on, this resistance is of utility in helping to throw light upon bacterial behavior during the early stages of disinfection. Examination of figures 1 to 4 shows certain striking facts that are common alike to both organisms in unbuffered surroundings. The rate of decline in numbers does not always run parallel in duplicate bottles held at the same temperature. This perplex- ing result is consistently found in these four experiments and inspection at once shows that the divergence is the result of a chance distribution, some curves being generally parallel and others far apart. Of course, this divergence if present increases with the time elapsed, and depends upon the organism; for Bad. ty-phosum being a matter of days, and for Bad. coli, one of weeks. These results were further substantiated six months later, when Bad. typhosum was studied in autoclaved tap water. This experiment, no. 12, was carried out in triplicate, and every precaution was taken to treat the samples alike. Yet in spite of these efforts, the mortality in each of the triplicate bottles followed a different course. (See table in Appendix.) What could account for the divergence of dupUcate bottles? They were handled alike as nearly as possible, and the bottles themselves, so far as we knew, were all of the same origin. The dissolved gases could not be responsible, for, as will be seen later, equilibrium with the air was attained within six hours from the beginning of the experiment; and furthermore, in experiment 4 filtered tap water, which presumably retained all of its dissolved gases, yielded the same result. A possible clue is furnished by the pH values of the water determined at successive intervals. These values are noted along the curves on the charts, and it becomes evident that du- DISINFECTION STUDIES 197 plicate bottles giving parallel curves show but insignificant devia- tions in pH from each other. On the other hand, a wide diver- gence of the curves is coincident with appreciable fluctuations in pH. These findings apply alike to double-distilled water and to Washington tap water. In the case of the distilled water, which contained a minimum of dissolved salts, the possible contami- nants are the atmospheric gases of the laboratory which include, among others, ammonia and sulfur dioxide. The tap water contained in addition to the gases, about 163 parts per million of dissolved mineral matter. Both waters are poorly buffered, with their pH equiUbria situated on the steep portion of a curve, so that a trace of strong base or acid will cause a relatively large change in the hydrogen ion concentration. This circumstance makes it difficult to estimate the pH accurately, for the hydro- gen electrode cannot be used, and the addition of free indicators, which are themselves strong acids, may x-itiate the results. We have, however, made a colorhnetric estimation of the pH by utihzing the sodium salts of the Clark and Lubs indicators and comparing the results with those obtained with the free acids. These pH deviations fall within the physiological range of our organisms, as will be sho\vn later, and it could scarcely be urged, therefore, that these insignificant shifts were the cause of the large variations in mortality. It may however be that the slight pH changes observed were indicative of large and more profound local fluctuations in the unbuffered waters surrounding the bacteria ; but it is worthless to pursue these thoughts further for they are purely speculative, and we are ignorant of all the forces operating in the present case. Yet the obvious inference occurs, that the control of the pH with buffers might stabiUze a possible factor in the problem and lead to more consistent results. Later experiments are concerned with this aspect of the problem. Our observations suggest that unbuffered solutions may not be depended upon to retain a constant and uniform hydrogen ion concentration for long, so that we must conclude that results of like experiments in the past may have been vitiated by this factor. Falk (1920) mentions the importance of pH in the study of salt effects upon the viability of bacteria. 198 BAENETT COHEN It is evident from mere inspection of figures 1 to 4 that effects of temperature, whatever they may have been, were completely screened. The air-saturation of the suspension fluids Several considerations made it seem desirable to learn to what extent our water samples were saturated with air gases. The method of analysis used makes no claim to absolute accuracy, but gives a good idea of the actual condition present. Table 1 shows that distilled water, seven days old, was fully aerated at each temperature. To determine how soon saturation would TABLE 1 Dissolved oxygen and carbon dioxide in distilled water held at different temperatures for seven days VOLUMES PER CENT TBUPEBATURB Found Calculated o. COj 0. CO. 0° 0.98 0.08 0 986 0.072 10- 0.80 0.06 0.796 0.047 20° 0.68 0.03 0.649 0.036 30" 0.53 0.526 occur under the conditions of our experiments, the following experiment was carried out. Bottles containing 900 cc. distilled water and an air space of 100 cc. were autoclaved, allowed to cool while closely stoppered and then held at 0°, 10°, 20° and 30°C. respectively for twelve hours. The bottles were then thoroughly shaken, and samples removed for gas analysis as indicated in table 2. The results show that within four to six hours the waters were practically saturated with air. (COj determinations are not given because the minute quantities present were within the limits of error of the method employed.) These observations confirm occasional tests which showed that our bacteria were exposed to an air-saturated en\'ironment. As a consequence we may assume that the gas tensions were practically invariant at each temperature throughout these DISINFECTION STUDIES 199 experiments. That is, the decline in bacterial population at each temperature occurred under practically constant concen- tration of atmospheric gases. It is interesting to note, however, that at 0°, 10°, 20° and 30°C. the concentrations of oxygen oc- curred in the ratio of 980:790:049:526. This observation takes on a certain theoretical significance when con.sidered in the light of the findings of Paul, Birstein and Reuss (1910) that dried staphylococci die off at a rate approximately proportional to the square root of the oxygen concentration. The experi- TABLE 2 The reabsorption of atmospheric oxygen by auloclaved distilled water* TOLnUES PER CENT OF OXTOEN WATER HELD AT Hours after twelve-hour incubation Calculated for air-saturated 0 2 4 6 water 0" 0 54 0.88 0 97 0.97 0 986 10" 0.41 0.70 0.79 0.81 0 796 20" 0.25 0.58 0.64 0.66 0.649 30' 0.11 0.47 0.52 0.52 0.526 * The water was autoclaved for ten minutes at 15 pounds, allowed to cool to room temperature and was held in an air thermostat for twelve hours. The water was well shaken before each successive sample was removed for analysis. ments of WTiipple and Mayer (1900) in this connection are interesting. They found that exclusion of oxygen from water containing colon and typhoid bacteria resulted in a rapid de- crease in the bacterial population. The contradiction between these observers may perhaps be more apparent than real. Jacobs (1920) has sho^vTl for tadpoles that CO2 may produce a lethal effect which is accounted for by a specific penetrability. Koser and Skinner (1921) studying the viability of the colon- typhoid organisms found that carbonation of water causes a rapid diminution in numbers. Our experiments throw no Ught upon this phase of the problem. Carbon dioxide was present in our solutions in presumably constant, small amounts; and its effect should be present to the same degree in all experiments. 200 BAENETT COHEN The influence of soluble glass constituents Chemists from the time of Lavoisier have had to consider the question of the solubihty of the glass vessels in their work. That a marked influence upon biological experiments may be exerted by soluble constituents from the container has been re- cognized now and then, but this aspect is yet too often neglected. Some biological investigators hke Beneke (1895) and Ficker (1898) have observed marked responses referable to glass con- stituents dissolved in the distilled water they used. MoUsch (1895) tried to eliminate this source of difficulty by coating his vessels with paraffin. Houston (1914) attempted a comparison of containers of different materials upon the viability of bacteria and obtained some striking qualitative differences. Bigelow and Esty (1920), Esty and Cathcart (1921) and Fabian and Stull (1921) observe differences in heat sterilization referable in part to the solubility of the glass containers, and suggest that the result may be due to pH effects. Under moderate conditions of temperature, the quantity of material dissolved by water from ordinary laboratory glassware in a short time is comparatively small;' and the most obxious results of such solution are the effect upon the pH of the water and the dissociation of other cations and anions. The results of the experiments reported here show that changes in pH of distilled water may often be quite considerable; and tap water, whether sterilized or filtered, has showTi itself not very different in behavior from distilled water in this respect. There are occasional references in the Uterature to a toxic action exerted by distilled water upon microbes. Laird (1919) found distilled water toxic to staphylococci, which fact he con- siders, as does Burgess (1920), to be due to plasmolysis and sug- gests "equilibrated" salt solutions as indifferent suspension media for bacteria. No observations of a similar nature seem to have been reported for bacteria of the colon-typhoid group. Our own experiments demonstrate the prolonged viabihty of these • It should be noted that our glassware had been in use for some time and had become "aged" to some extent. DISINFECTION STUDIES 201 organisms in distilled water and tap water, and we must there- fore conclude that in this bacterial species plasmolysis is not a very prominent phenomenon. Winslow and Falk (1918) have demonstrated very interesting salt effects and antagonisms upon Bad. coli dying in water; and Hohn and Sherman (1921) note similar effects upon growing bacteria. The suggestion of Zeug should be recalled in this connection. It is conceivable that such effects occurred in our experiments but were overshadowed by other more prominent factors. In the hght of our experience with the activity of unbuffered suspension media upon bacterial death we must appreciate at once the great experimental difficulties involved in the study of pure salt effects, and the need for caution in their interpretation. The behavior of Bad. typhosum at different hydrogen ion concentrations The study of the mortality of bacteria at constant hydrogen ion concentration was made in M/500 solutions of Clark and Lubs' phthalate and phosphate buffers. It was considered desirable to keep the concentration of salt down to a minimum, and this dilution of the buffers was found to be the lowest that would maintain the pH unchanged under our conditions. Experiment 5 {Bad. typhosum at pH 3.8, 5,0, 5.4, 6.4, 7.1, 7.6, 8.7 and 9.5) was performed to determine if constancy of pH would condition a uniformity in the results from duplicate bottles; and to learn the effect of pH at constant temperature (20°C.) upon the viabiUty. Figure 5 shows that the effect of maintain- ing a constant pH throughout the experiment was to produce a high uniformity in the results from duplicate bottles.* As a consequence, duplicate curves, while determined in all cases, are omitted from the rest of the charts for the sake of clarity in presentation. Reference to figure 5 shows that decline occurs at all hydrogen ion concentrations, and is least at pH 5.0 and 5.4. At pH 3.8, ' These are not selected specimens, but include all the tests made. 202 BAKNETT COHEN corresponding roiighly to the acidity of N/10,000 HCl, the de- cline is most acute, and by contrast, at pH 5.0 (equivalent to about N/100,000 HCl), the rate of decline is the smallest. On Fig. 5. Experiment 5. The Death Rate of Bact. ttphosum in M/500 Buffers AT 20°C. The curves actually start from nearly the same origin, but have been spread apart to avoid confusion. Note the close parallelism between duplicates at almost all pH values. the alkaline side, the acceleration in death rate is more gradual. Northrop (1920) studying the acid stability of pepsin found pH 5.0 to favor the greatest stability. He found for this enzyme, however, that increase of acidity did not destroy it so markedly as increase of alkaUnity. DISINFECTION STUDIES 203 Our curves being derived under the same conditions are com- parable and their relations may be designated numerically by their slopes, provided the curves are fairly straight lines, as in the present case. This is another way of stating that we are comparing rates of decHne, or reaction velocities if we prefer to think in terms of the monomolecular reaction. The velocity constants thus derived are given in table 3. Thb mode of expressing differences in disinfection has many advantages and has been recommended by Phelps (1911) as one of the logical procedures in the evaluation of disinfectants. Buchanan (1918), Bruett (1919) and Salter (1921) have utilized it in studies of heat sterilization and thermal death points. TABLE 3 Average velocity constants for the death of Bact. typhosum at different pll values at 20°C. pHa.s pHS.O pH5.4 pH6.4 pH7.1 pH7.6 pH8.7 pH9.5 k 1.055 95.5 0.0134 1.2 0.0110 1.0 0.0138 1.5 0.(M37 4.8 0.1100 10.0 0.2134 22.4 0.2855 •Relative Jt 31.4 * Relative k represents the magnitudes of the various constants compared with that at pH 5.4 taken as unity. The velocity constants we have thus derived show clearly that there is a zone between pH 5.0 and 6.4 in which Bact. ty- phosum declines in numbers very slowly. A small change in pH toward the more acid side of the zone produces an almost hundred-fold increase in the rapidity of death; while on the alkaline side a similar change in pH increases the death rate only four or five times. In brief, a small increase in the concen- tration of hydrogen ions at the acid end of the zone of tolerance produces a profound effect, whereas relatively large increases of hydroxyl ions produce a much smaller effect. This coincides with the findings of Cohen and Clark (1919) who called atten- tion to a similar effect upon the growth of members of the colon group of bacteria. This observation is of some importance theoretically, and may help to throw light upon the mechanism involved in the death of 204 BAKNETT COHEN bacteria. For instance, the optimum for growth of Bad. ty- phosum lies between the pH limits 6.2 and 7.2, though it will tolerate greater extremes. Yet here we encounter the significant fact that its optimum for maintenance under conditions of starva- tion Ues between pH 5.0 and 6.4. It would, however, be unwise at present to dwell upon the possibilities in this direction. Of interest in the same connection is the work by Shohl and Janney (1917) who established the pH limits for the growth of colon and tj^phoid bacteria in urine. Marsh (1918) found that Bad. typhosmn is sensitive to the degree of acidity occurring in sour milk; and Beckwith (1920) reports that this organism is capable of surviving in rabbit bile in vivo even when the hy- drogen ion concentration is depressed to pH 9.4. It may be asked how far we are justified in concluding that the control of the pH with buffers conditions uniformity in results between duplicates. Striking graphic proof is given in Experiment 5 in whifch only one (that at pH 7.1) out of 8 tests failed in this respect. In experiments 6 to 13, the data for which are tabulated at the end of this paper, we find the large majority' of duplicates to run closely parallel. That there are a few ex- ceptions is evident but even in these, the average divergence is very much less than in duplicates from de\iating unbuffered solutions. A further test of this fact was made six months later when Bad. iyphosum was studied in triplicate samples buffered around pH 6.4. In this experiment, no. 13, the technique was especially careful, and the individual bacterial counts are the most accurate we have made, the probable error being below 5 per cent. If the resulting data be plotted against time, it will be found that each of the tripUcate samples yields an almost identical curve. We have reason to believe that non-uniformity of results under these conditions is due mainly to laxity in technique. ' Of 33 such tests, about 4 may be considered to have shown any considerable divergence between duplicates. DISINFECTION STUDIES 205 The behavior of Bad. typhosum and Boot, coli at pH 3.5 at different temperatures Since it became possible to eliminate variability in the results between duplicate bottles by means of control of pH, we could proceed to the study of the effect of temperature upon the death of these organisms. Fig. 6. Experiment 6. The Death Rate of Bact. ttphostjm in M/500 Phthalate Buffer, pH 3.5 at 0°, 10°, 20° 30°C. Closely parallel duplicate curves are omitted Experiments 6 and 7 were conducted at a comparatively high acidity (pH 3.5) in order to obtain a comparison of the relative rates of dying of these two organisms at 0°, 10°, 20° and 30°C. Such a comparison is not feasible at less intense acidities because of deviations of the disinfection curve from the simple logarithmic form, due to the predominance of certain factors referable to the specific peculiarities of each organism. This will be con- sidered later. 206 BAKNETT COHEN Figures 6 and 7 present the curves of the death rates observed. These experiments are strictly comparable for they were done 0 Hours 7 73 IS 2o~ IS 30 Fia. 7. Experiment 7. The Death Rate of Bact. coli in M/500 Phthalatb Buffer, pH 3.5, at 0°, 10°, 20°, 30°C. Closely parallel duplicate curves are omitted at the same time and with identical materials, and the marked difference in resistance of the two organisms is at once apparent. In the case of Bact. typJiosuvi (fig. G) the first portion of each DISINFECTION STUDIES 207 mortality curve is very nearly a straight line, but at the lower end there is a tendency to flatten out somewhat. In the case of Bad. coli (fig. 7) there is a suggestion of this tendency at the beginning of the curve. If we take the average velocity coeffi- cients for comparison, the results may be arranged as in table 4. We note in the case of Bad. typhosum that the increase in the death rate is fairly regular for 10° intervals, being 1.02, 1.53 and 1.77 from 0° to 30°C., while for coli the corresponding in- creases are 2.12, 4.36 and 3.76. With Bad. coli the eflfect of temperature is to increase the velocity for the 10° intervals from 0° to 30°C. in the ratio of 1 : 2 : 9 : 35 which corresponds closely to the exponential series: 2° : 2' : 2' : 2^ Such a relation is not evident in the behavior of Bad. typhosum. TABLE 4 Average velocity coefficients for the death of Bad. typhosum and Bact. coli at pH S.6 at different temperatures TBMPERATDRE BACT. TTPHOSDH BACr. COLI *t *t 0.. *0 Qu *0 0° 1.186 1.62 0.0176 2.12 67 10° 1.919 1.53 0.0373 4.36 51 20° 2.928 1.77 0.1654 3.76 18 30° 5.176 0.6214 8 Qio is the temperature coefficient for the indicated 10° interval. If the ratio of velocity coefficients be considered a measure of the relative resistance of these organisms, we observe that at 0°C. Bact. coli is 67 times as resistant as the typhoid bacillus. As the temperature is increased by ten-degree intervals the rela- tive resistance of Bad coli decreases so that at 30°C. it is only eight times that of Bact. typhosum. 208 BARNETT COHEN The behavior of Bact. coli at different hydrogen ion concentrations and temperatures An inspection of the curves for the death of Bact. coli at pH 3.5 (fig. 7) shows that at 30°C. the curve is ahnost a straight fine with a sUght suggestion of flattening out at the beginning. At 20°C. this flattening out is more pronounced, and at 10°C. and 0°C., still more so. These de\dations cannot be fuUy accounted for by the probable experimental error, though the process of Fio. 8. Experiment 8. The Death Rate of Bact. coli in M/500 Phosphate Buffer of pH 6.1 at 0°, 10°, 20°, 30°C. Parallel duplicate curves are omitted drawing 'smoothed' curves through the points may be made to wipe out this deviation. We have found, in experiments not detailed here, that at higher acidities these de\'iations from the straight-line logarithmic decUne are absent so that the resulting curves resemble those found for Bact. typhosum in experiment 6. If we now refer to figures 8, 9 and 10, it will be observed that these deviations are enhanced, especially for the lower tempera- tures. Such deviations simply mean that as time progresses DISINFECTION STUDIES 209 0 Ooys 6 Fig. '9. Experimext 9. The Death Rate", of Bact. cou iNfM/SOO Phosphate Buffer op pH 7.1 at 0°, 10», 20°, 30°C. Parallel duplicate curves are omitted oTS^ 6 Fio. 10. Experiment 10. The Death Rate op Bact . coli in M/500 Phosphate Buffer of pH 8.0 at 0°, 10°, 20°, 30°C. The pH remained constant until the fortieth day, but was found to be 7.7 on the sixtieth. 210 BAKNETT COHEN the speed with which the bacteria perish increases to a maximum when the curve becomes a straight Une. This phenomenon among bacteria finds its counterpart in certain chemical reac- tions which possess an induction period prior to reaction at the maximum rate. If the time necessary to reach maximum veloc- ity in death rate is short then the calculation of an average velocity coefficient involves an insignificant error. This rep- resents the mathematical equivalent of drawing a straight line as close as possible to the experimental points. If, however, this induction period is prolonged, then such a procedure may involve serious difficulties. TABLE 5 Average velocity coefficients of mortality of Bad. coli at different pH values and temperatures TBMPEBA- PH 3.5 pH6,l pH7.1 pHS.O h 0.0 k 0,. k 0.0 k 0.0 0° 10° 20° SO- 0.0176 0.0373 0.1654 0.6214 2.12 4.34 3.76 0.000206 0.000315 0.000353 0.001611 1.53 1.12 4.56 0.000107 0.000174 0.000520 0.001170 1.62 2.99 2.29 0.000260 0.000314 0.000666 0.001996 1.22 2.12 3.00 An exact numerical measurement of the extent of this devia- tion is unfortunately not possible because of the limitations of the present experimental method. It may, however, be inferred that it is an inverse function of the temperature. E\'idently at 30°C. it lasted about twenty days in experiments 8 and 9; and at 10° and 20°C. the end was not reached in sixty days. As stated above, a comparison of the average velocity coefficients is not justified as a logical procedure under the circumstances; but with this reservation we may consult table 5 to get an idea of the magnitudes involved. This table indicates that the pH zone favoring the lowest destruction rate of Bad. coli is around absolute neutrality. It appears that this zone covers a greater range of hydrogen ion concentration than does that of Bad. typhoswn. DISINFECTION STUDIES 211 We find in these experiments striking examples of that pre- liminary period in the disinfection process which we have referred to as the period of induction. The bacteria do not begin to die off at the maximum rate, but the mortahty increases to the maximum gradually, depending upon the pH and the tem- perature. The lower the temperature and the less extreme the acidity, the greater is the duration of this period. We have succeeded in this case, apparently, in magnifying that early phase in disinfection which ordinarily is so small as to escape observa- tion. Its significance will be discussed presently. DISCUSSION The probable error involved in the method of experimentation leaves much to be desired. Under the most nearly ideal condi- tions some of the results involve a probable error of only 2 to 3 per cent of the mean values indicated. In others, the probable error amounts to 10 per cent of the mean or more. When the bacterial numbers per unit volume are low, then the probable error increases greatly. These considerations led to the choice of an initial number of one million per cubic centimeter as involv- ing the least manipulative error. It might perhaps be well to express here a word of caution in regard to a too ready use of the calculated velocity coefficient, k, in the comparison of disinfections under varying conditions. This mode of expressing differences in disinfection has many advantages, provided its use implies no unwarranted assumptions. The general curve of disinfection, plotted logarithmically, is a straight line in the main, but at each end there are characteristic deflections. If these deflections are of small extent then an average velocity constant calculated for the whole of the disin- fection period will involve no serious error. If, however, there is a lengthened period of induction before disinfection proceeds at the maximum rate, then obviously, such a basis for the calcu- lation of the velocity constant would involve a large error. The same objection would hold for the condition in which the disin- fection curve flattens out toward the end of the process. The latter consideration prompted Phelps (1911) to recommend the 212 BARNETT COHEN determination of the velocity constant up to the point where the numbers of organisms have been reduced by 50 per cent. To avoid error from the former consideration as well, k should be determined for the middle part of the disinfection curve, say, between the points for 25 and 75 per cent reduction. The temperature coefficient (Qio) in bacterial viability We have already mentioned the broad empirical generaliza- tion that the temperature coefficient is about 1 for the accelera- tion of physical processes by a rise of 10° in temperature, and 2 or more for chemical processes. The rule is not a hard and fast one, and its basis is obscure. It is therefore important to re- cognize that the coefficient is at best a suggestive observation until we possess a better knowledge of the mechanism of the temperature effect. Obviously, in strictly controlled experiments where only one reaction is allowed to take place, a temperature coefficient so derived is of some value. On the other hand, in biological ex- periments, the observed effect may be the resultant of an unknown number of independent as well as interdependent reactions, which may be both physical and chemical in nature. A tempera- ture coefficient derived under such circumstances must be of doubtful significance. Changes of temperature may affect consecutive reactions in totally different ways. Mellor (1909) and Osterhout (1917) discuss this aspect of the problem and cite illuminating examples. With these considerations in mind we may note that the tem- perature coefficients found in this investigation are in the main those to be ascribed to chemical reactions. This would indicate that somewhere in the series of consecutive processes ending in death, the slowest or limiting reaction velocity was that ascrib- able to the ordinary chemical reaction. Of more theoretical interest is the finding in experiment 7 that the death rates for Bad. coli increased for 10° intervals from 0° to 30°C. in an ex- ponential ratio of the form: 2° : 2' : 2' : 2^ a regularity that seems more than accidental. Its interpretation is difficult for the present but we have here a suggestion for a possible method DISINFECTION STUDIES 213 of getting a closer view, perhaps, of one of the mechanisms in- volved in bacterial death. In this connection might be mentioned the contribution of Watson (1908) who showed that variation in the rate of disinfection due to change in concentration of the disinfectant could be expressed as an exponential function of the concentration. Another apparent effect of increasing temperature is to de- crease the period of induction j)rior to the logarithmic rate of death. The evidence is incomplete, but it would seem that the duration of induction varies in inverse proportion to some exponent of the temperature. The effect of pH upon bacterial viability We have seen that an uncontrolled hydrogen ion concentra- tion may affect the death rate variably in unbuffered surround- ings, so much so as to obscure the effects of wide ranges of tem- perature. When the pH is controlled by means of dilute buffer solutions the death rates become stabilized. Under the latter conditions there becomes evident a zone of pH in which the death rates of the bacteria are at a minimum. For Bad. ty- phosum this zone Ues between pH 5.0 and pH 6.4 and for Bad. coli it is wider, with the optimum near absolute neutraUty (pH 7.0). These zones may be regarded as optima for tolerance xinder moderate lethal conditions and when compared with the optima for growth show interesting relations, especially in the case of Bact. typhosuvi. The pH zone for optimum growth of this organism Ues between 6.2 and 7.2, while here we have found that the zone for greatest tolerance lies between pH 5.0 and 6.4. Cohen and Clark (1919) showed for the colon-dysentery group that a slight increase in acidity beyond the optimum limit for growth caused a very large effect by preventing growth. The same phenomenon has been found in the present study of response to lethal conditions, where a slight increase in acidity caused a prompt change from maxunum tolerance to high mortality. In practical disinfection the intensity of attack, presumably is so great that such variations in tolerance as we have observed would hardly affect the bacterial death rates. Yet the possibility JOURNAL OW BACTERIOLOOT, VOL. VII, NO. 2 214 BAENETT COHEN remains that if a disinfectant be applied in a medium having a pH favorable to high bacterial tolerance, the result might be appreciably affected. It appears that this tolerance constitutes one element in the general condition termed "resistance." The foregoing observations furnish corroboration of the present weU- estabUshed principle of determining disinfectant values under accurate control of the hydrogen ion concentration (Wright, 1917). The laws governing the disinfecting process We may now turn to a consideration of the process of disin- fection and the bearing of our experiments upon the laws govern- ing it. To treat the subject fairly and avoid the possibihty of misunderstanding, we shaU begin from certain fundamental concepts in physical chemistry. One of them is the assiunption of the vaUdity of the mass law. This states that at any moment, the velocity of a chemical reaction is dependent upon the rela- tive masses of the reacting bodies as well as their nature. If only one substance is undergoing change in a reaction then according to the mass law, the velocity of such change "noil de- pend upon the nature of the substance and its amount at any given moment, temperature and other conditions remaining constant. This statement of the law of monomolecular reac- tions is expressed concisely by the relation V=C-k (1) in which V represents the velocity of reaction, C, the amount of substance and k, a characteristic factor depending upon the nature of the substance undergoing change. dx The velocity may be expressed by — -, in which x represents at the amount of substance changed in the time t; and if the oripnal amount of the substance be designated by a, then a-x will repre- sent the amount remaining after the time t. We may now ^\Tite the above equation in the form DISINFECTION STUDIES 215 't = ^(«-^> (2) This expression on integration becomes fc = flog-^ (3) t a — X which is the familiar equation of the velocity of a monomolecu- lar reaction and often spoken of as the logarithmic law. It is perhaps well to emphasize that this formula represents the course of events taking place and makes no pretence of indi- cating the mechanism involved. It indicates the effect of active mass. Although the value of k will be greatly modified by change in the nature and intensities of the forces concerned, the essential form of the velocity curve will in no wise be altered since it is fixed by itis inherent relation to the numbers of reacting molecules and by this factor only. The factor k may include, by its very definition, many influences. Therefore, to establish experimentally the apphcability of the monomolecular law in any given case, all the influences must be kept constant while the effect of mass is being observed. For the present, our knowl- edge of the mechanism of a monomolecular reaction is nil. We may only guess that it involves complex electronic relations (cf. Tohnan (1921) and Dushman (1921)). Note that k in the final expression happens to be designated as the velocity con- stant since time is inversely related to it. If log (a —x) be plotted against time in the above equation, the resulting graph will be a straight line. The above theoretical considerations have therefore led us to the deduction of a logarithmic equation that should hold good under the ideal conditions imposed. When we turn to actual experimental observations on monomolecular chemical reactions it is found that this logarithmic law holds good, according as we are able to maintain the ideal conditions. There may occur deflections from the true logarithmic rate at the beginning and the end of the reaction so that instead of a straight line plot of the results, we get a somewhat s-shaped curve. If conditions 216 BAENETT COHEN are properly selected, the deflections may be reduced considerably. As a matter of experimental fact, they have never yet been com- pletely eliminated (except possibly for radioactive substances). These considerations are not held to affect the validity of the monomolecular law. They lead us rather to investigate the disturbing influences which, when discovered and eliminated, furnish further support to the law. We now possess a fundamental concept of the origin of the monomolecular law as related to chemical reactions and may turn to the phenomena observed in disinfection. If the disin- fection process be followed by noting the numbers of surviving bacteria at successive intervals we find that the rate at which disinfection proceeds is iii general proportional to the number present. This proportionality is not absolutely true but gener- ally so, and the graph representing the course of the process is more or less a straight Une with deflections at both ends. Under appropriate conditions these deflections may be eUminated so that the curve becomes very nearly a straight Une. The mortal- ity process as observed by Chick and others, when strong disin- fecting agents are used, has been shown conclusively to follow the logarithmic rate, or a sUght modification of it. It is needless to stress the self-evident analogy between the course followed by a monomolecular chemical reaction and the course of the mortality process of bacteria. They represent actual observations. In this connection must be mentioned the statistical deduction of Yule (1910) based upon the theory of probability, that in a population exposed to a single lethal influence, the rate of death will follow a logarithmic course, and when there are a number of sub-lethal causes, the death rate wiU be a modification of the logarithmic one. The above facts have been fully explained by Chick and it would seem superfluous to repeat them were it not for the criti- cisms of Chick's conclusions by Loeb and Northrop (1917), Brooks (1918) and Smith (1921). These authors urge that individual variations among bacteria may be distributed on a frequency curve such that the mortality rate will proceed appar- ently logarithmically. DISINFECTION STUDIES 217 Now it would of course be extravagant to say, since the course of disinfection is found to follow the monomolecular law, that therefore a single molecular species is concerned. We can only say that whatever the mechanisms are and whatever the number of consecutive reactions may be, they leave dominant the effect of bacterial concentration. We may say in other words that we are fortunate in having to deal with a phenomenon in which the effect of concentration can be found, and that we can formu- late the course of the disinfection process in terms of an equation expressing the relation of concentration to the course of the process. We are, aware, however, of some of the factors that affect the course of the reaction. These in the case of chemical reactions, are enumerated and discussed by Mellor (1909). One of them, the effect of successive intermediate reactions, has been apphed by Osterhout (1917) in an interesting manner to biologi- cal phenomena, and by Winslow and Falk (1920) to controvert the notion that disinfection is due to a distribution of variable resistances. We have already indicated that even in the case of a simple chemical reaction known to involve only one molecular species, the mechanisms of disintegration are unknown, or at least the subject of dispute. The monomolecular law in such cases shows only the course of the reaction as it is related to the concentra- tion of reacting bodies. The analogy in the two cases should be plain. In neither case does the monomolecular law tell us anything about the mechan- isms concerned. In both cases the monomolecular law fonnu- lates the relation between concentration and the course of the process. If then we say that the course of disinfection is determined by the distribution of varying resistances we add nothing to the formulation of the experimental facts. We could just as well say that the course of a monomolecular chemical reaction is determined by the frequency distribution of resistance among the individual molecules. In the criticism by Loeb and Northrop, by Brooks and by Smith use is made of the fact that at the beginning and end of the dis- 218 BAENETT COHEN infection curve there are deviations from the monomolecular rate. This is emphasized as evidence that the monomolecular rate at the middle portion of the curve is more apparent than real. However, with cultures having presumably the same distribution of resistances we have found that the shape of the beginning and end of the curve may be modified at will by chang- ing the exterior conditions. Consequently, what force do these objections to Chick's theory possess? The chance distribution of variable resistances assumed for bacteria is paralleled by the chance distribution of variable energy quotas in the molecules of a substance. Furthermore, the logarithmic rate observed, whether in chemical phenomena or in disinfection, is actually a statistical resultant of like signifi- cance in both.* From such considerations we must conclude that the criticisms of the monomolecular theory of the disinfec- tion process are of no fundamental force for they reduce to a matter of definitions only. It appears that Chick's contribution has a larger significance than merely the application to a special phase of bacterial exist- ence. Yule's statistical deduction, the experimental findings of Loeb and Northrop regarding the viability of Drosophila, those of Brooks regarding hemolysis of red blood cells and a number of others of like import, all indicate the operation of a general principle. CONCLUSIONS 1. The mortality at constant temperature of bacteria in un- buffered media like distilled or tap water is variable and coinci- dent with apparently insignificant pH variations. Controlling the pH by means of M/500 buffer solutions stabifizes this variability. 2. Subjecting organisms of the colon-typhoid group to mild lethal conditions under moderate temperatures and hydrogen ion concentrations tends to magnify the induction period prior • It must be remembered that the logarithmic curve merely integrates the results of all factors. It is a statistical summation and gives no information regarding the forces at play. DISINFECTION STUDIES 219 to mortality at the maxiimun or logarithmic rate. This provides an opportunity for studying the early response of the organism to the disinfection process. 3. The period of induction is decreased by higher acidity and by higher temperature. It appears to have^a duration inversely proportional to some exponent of the temperature. It is analo- gous to the induction period occurring in chemical reactions. 4. At constant pH, the relative resistance of Bad. coli to Bad. typhosum decreases with rise in temperature from 0° : 10° : 20° : 30° m the ratio of 67 : 51 : 18 : 8. 5. At 20°C. Bad. typhosum possesses the greatest tolerance within a narrow zone of hydrogen ion concentration delimited by pH 5.0 and 6.4. A slight increase in acidity beyond the zone results in conditions of maximum mortahty. J'or Bad. coli the zone is wider and centered about absolute neutrality. Cohen and Clark (1919) found that the pH optima for growth and fermentation of bacteria may be different. It is now shown that the optimum for tolerance may also be distinct. 6. The mortality of bacteria whether by strong disinfectants or by milder agents follows the laws of logarithmic decline. It is shown that the course of the disinfection process can be expressed by mathematical relations comparable to those used in dealing with monomolecular chemical reactions. I desire to express my cordial appreciation for important criti- cism and suggestions to Drs. C.-E. A. Winslow, E. Elvove, S. C. Brooks and especially to Dr. W. Mansfield Clark. REFERENCES American Public Health Association, Report of Committee on Standard Methods of Examining Disinfectants. 1918. Am. Jour. Pub. Health, 8, 506. ARRHENrus, S. 1889 The reaction velocity of the inversion of sucrose by acids. Zsch. f. physik. Chem., 4, 226. Arrhenitjs, S. 1915 Quantitative laws in biological chemistry. London. Barber, M. A. 1908 The rate of multiplication of B. coli at different tempera- tures. Jour. Inf. Dis., 6, 379. Beckwith, T. D. 1920 The viability of B. typhosus in alkaline bile in vivo. Proc. Soc. Exp. Biol. Med., 18, 36. Beneke, W. 1895 DiezurEmiihrungder SchimmelpilzenothwendigenMetalJe. Jahrb. f . wissensch. Bot.inik, 28, 487. 220 BARNETT COHEN BiGELow, W. D., and Estt, J. R. 1920 The thermal death point in relation to time of typical thermophilic organisms. Jour. Inf. Dis., 27, 602. Brook-s, S. C. 1918 A theory of the mechanism of disinfection, hemolysis, and similar processes. Jour. Gen. Physiol., 1, 61. Bhdett, E. M. 1919 Utility of blanching in food canning. Effect of cold shock upon bacterial death rates. J. Ind. Eng. Chem., 11, 37. Buchanan, R. E. 1918 'Life phases in a bacterial culture. Jour. Inf. Dis., 23, 109. Buchanan, R.E., Thompson, C.E., Orr, P. F., and Brtjett, E.M. 1918 Notes on conditions which influence thermal death points. Abs. Bact., 2, 5. Burgess, K. E. 1920 The toxicity towards Staphylococcus of dilute phenol solutions containing sodium benzoate. Jour. Phys. Chem., 24, 738. BusHNELL, L. D. 1918 The influence of cold shock in the sterilization of canned goods. Jour. Ind. Eng. Chem., 10, 432. Chick, H. 1908 An investigation of the laws of disinfection. Jour. Hyg., 8,92. Chick, H. 1910 The process of disinfection by chemical agencies and hot water. Jour. Hyg., 10, 237. Chick, H. 1912 The factors conditioning the velocity of disinfection. Orig. Communications, Sth Internat. Cong. Appl. Chem., 26, 167. Chick, H. 1912 Thebactericidalpropertiesof blood serum. Jour. Hyg., 12, 414 Clark, W.M. 1920 The determination of hydrogen ions. Baltimore. Clark, W. M., and Lubs, H. A. 1916 Hydrogen electrode potentials of phtha- late, phosphate and borate buffer mixtures. J. Biol. Chem., 26, 479. Cohen, B., and Clark, W. M. 1919 The growth of certain bacteria in media of different hydrogen ion concentrations. Jour. Bact., 4, 409. Dushman, S. 1921 A theory of chemical reactivity. Calculation of rates of reactions and equilibrium constants. J. Am. Chem. Soc, 43, 397. EsTY, J. R., and Cathcart, P. H. 1921 The change in pH of various mediums during heating in soft and Pyrex glass. Abs. Bact., 6, 6. Fabian, F. W., and Stull, R. C. 1921 A study of the hydrogen-ion concentra- tion of different kinds of glassware when sterilized with buffered and non-buffered solutions. Abs. Bact., 6,210. Falk, I. S. 1920 Studies on salt action. III. The effect of hydrogen-ion con- centration upon salt action. Proc. Soc. Exp. Biol. Med., 17, 210. Fickek, M. 1898 Ueber Lebensdauer und Absterben von pathogenen Keimen. Zsch.f. Hyg., 29, 1. Holm, G. E. and Sherman, J. M. 1921 Salt effects in bacterial growth. Abs. Bact., 6,6. Houston, A. C. 1908 The vitality of B. tj^phi in water. 1st Research Report, Metr. Water Board, London. Houston, A. C. 1909 The vitality of B. coli in water. 3d Research Report, Metr. Water Board, London. Houston, A. C. 1911 The vitality of B. typhi in water. 7th Research Report, Metr. Water Board, London. Houston, A. C. 1914 The effect of temperature on the vitality of B. coli and B. typhi in water. 10th Research Report, Metr. Water Board, London. Jacobs, M. H. 1920 To what extent are the physiological effects of carbon dioxide due to hydrogen ions? Am. Jour. Physiol., 61, 321. DISINFECTION STUDIES 221 Kanitz, A. 1915 Die Biochemie in Einzeldarstellungen. Heft I. Temperatur in LebensvorganRp. Berlin. Koch, R. 1881 Uoher Desinfektion. Mitt. a. cl. kais. Gcsundh., 1, 1. KonrXdi, D. 1904 Ueber die Lebensdauer pathogener Baktericn im Wasser. Centralbl. Bakt., I. Abt., Orig., 36,203. KosER, S. A., AND Skin-ner, \V. W. 1921 Viability of the colon-typhoid group in carbonated water and carbonated beverages. Abs. Bact., 6, 12. KrOnig, B., a\d Paul, T. 1897 Die chemischen Grundlagen der Lehre von der Giftwirkung und De.sinfektion. Zsch. f. Hj-g., 25, 1. Laird, J. S. 1920 The chemical potential of phenol in solutions containing salts; and the toxicity of these towards anthrax and staphylococci. Jour. Phys. Chcm., 24, 664. Lane-Claypon, J. 1909 The multiplication of bacteria and the influence of temperature and some other conditions thereon. J. Hyg., 9, 239. Lee, R. E., and Gilbert, C. A. 1918 On the application of the mass law to the process of disinfection. Jour. Phys. Chem., 22,348. Lewis, W. C. M., and McKeown, A. 1921 The radiation theory of thermal reactions. Jour. Am. Chem. Soc, 43, 1288. LrviNQSTON, G. S. 1921 The vitality and viability of hemolytic streptococci in water. Am. Jour. Hyg., 1, 239. LoEB, J. 1906 The dynamics of living matter. New York. Loeb, J. 1908 Ueber den Temperaturkoeffizienten fiir die Lebensdauer kalt- bliitiger Thiere und iiber die Ursache des naturlichen Todes. Pfliigers Arch., 124,411. Loeb, J., and Northrop, J. H. 1917 The influence of food and temperature upon the duration of life. Jour. Biol. Chem., 32, 103. Marsh, P. 1918 The survival of typhoid bacilli in sour milk. Am. Jour. Public Health, 8, 590. Madsen, T., and Kyman, M. 1907 Zur Theorie der Desinfektion. Zsch. f. Hyg., 67, 388. Madsen, T., and Streng, O. 1910 Einfluss der Temperatur auf den Zerfall der Antikorper. Zsch. f. physik. Chem., 70, 263. Mellor, J. W. 1909 Chemical statics and dynamics. London. Northrop, J. H. 1920 The influence of hydrogen ion concentration on the inactivation of pepsin solutions. Jour. Gen. Physiol., 2, 465. Osterhout, W. J. V. 1917 Some aspects of the temperature coeflficients of life processes. Jour. Biol. Chem., 32, 23. Paul, T. 1909 Der chemische Reaktionsverlauf beim Absterben trockener Bakterien bei niederer Temperaturen. Biochem. Zsch., 18, 1. Paul, T., Birsteln, G., and Reuss, A. 1910 Beitrag zur Kinetik des Abster- bens der Bakterien in Sauerstoff verschiedener Konzentration und bei verschiedenen Temperaturen. Biochem. Zsch., 26, 367. Paul, T., .^nd Kronio, B. 1896 Ueber das Verhalten der Bakterien zu chemis- chen Reagenzien. Zsch. physik. Chem., 21, 414. Paul, T., and Prall, F. 1907 Die Wertbestimmung von Desinfektionsmitteln mit Staphylokokken, die bei der Temperatur der fliissigen Luft aufbe- wahrt wurden. Arb. a. d. k. Gesundh., 26, 424. Peters, R. A. 1920 Variations in the resistance of protozoan organisms to toxic agents. Jour. Physiol., 54, 260. 222 BARNETT COHEN Phelps, E. B. 1911 The application of certain laws of physical chemistry in the standardization of disinfectants. Jour. Inf. Dis., 8, 27. Rbichel, H. 1909 Zur Theorie der Desinfektion. Biochem. Zsch., 22, 152. Salter, R. O. 1921 A comparative study of hemolytic streptococci from milk and from human lesions. Am. Jour. Hyg., 1, 154. Shohl, a. T. and Janney, J. H. 1917 "Ihe growth of Bacillus coll in urine at varying hydrogen ion concentrations. Jour. Urol., 1, 211. Slator, a. 1919 Yeast growth and the alcoholic fermentation of living yeasts. Jour. Soc. Chem. Ind., 38, 391. Smith, J. Henderson'. 1921 The killing of Botrytis spores by phenol. Ann. Appl. Biol., 8,27. Snyder, C. D. 1908 A comparative study of the temperature coefficients of the velocities of various physiological actions. Am. Jour. Physiol., 22,309. Snyder, C. D. 1911 On the meaning of the variation in the magnitude of the temperature coefficients of physiological processes. Am. Jour. Phys- iol., 28, 167. Snyder, C. D. 1911 An interpolation formula used in calculating temperature coefficients for velocity of vital activities. Science, N. S., 34, 414. ToiiMAN, R. C. 1921 Note on the theory of monomolecular reactions. Jour. Am. Chem. Soc, 43,269. Van Slyke, D. D. 1917 A method for the determination of carbon dioxide and carbonates in solution. Jour. Biol. Chem., 30, 347. van'tHopf, J. H. 1896 Studies in chemical dynamics. Translated by Thomas Ewan. Easton, Pa. Ward, A. M. 1895 On the biology of Bacillus ramosus (Fraenkel), a schizomy- cete on the River Thames. Proc. Roy. Soc, 58, 265. Watson, H. E. 1908 A note on the variation of the rate of disinfection with change in concentration of the disinfectant. Jour. Hyg., 8, 536. Whipple, G. C, and Mayeh, A. 1908 On the relation between oxygen in water and the longevity of the typhoid bacillus. J. Inf. Dis., Supplement no. 2, 75. WiNSLOW, C.-E. A., and Falk, I. S. 1918 Studies on salt action, I. Effect of Ca and Na salts upon the viabilit}- of the colon bacillus in water. Proc Soc. Exp. Biol. Med., 15, 67. WiNSLow, C.-E. A., and Falk, I. S. 1920 A contribution to the mechanism of disinfection. Abs. Bact., 4, 2. Wright, J. H. 1917 The importance of uniform culture media in the bacterio- logical examination of disinfectants. J. Bact., 2, 315. Ycle, G. U. 1910 On the distribution of deaths with age when the causes of death act cumulatively and similar frequency distributions. Jour. RoyalStat. Soc, 73,26. Zbug, M . 1920 Aequilibrierte Salzlosungen als indifferente Suspensionsfliissig- keiten fur Bakterien. Arch. f. Hyg., 89, 175. APPENDIX In the following pages will be found, in tabular form, the detailed experi- mental results. Most of them have been presented in graphic form in the charts. The values tabulated are the Briggsian logarithms of the numbers of survivors per cubic centimeter of water or buffer solution. Observations of the pH of the suspending fluid are noted wherever made. Experiment I. Morlalily of Bact. typhosum in double-distilled water in paraffm- lined bottles at different temperatures O'C. lO'C. 20°C. 30'C. I I II I II I 11 hoUTt 0 6 0133 6.2122 6.1399 6.0294 6 0509 6.1335 6 0069 pH 6.4 6 3 6.2 6.4 6 2 6 3 6 3 6 6.0038 6.1584 5.9085 6.0128 5.9754 6.0645 6.0645 pH 6.3 6.1 6.2 6.1 6.1 6.3 6.3 12 6.0112 .6.1959 5.9004 6.0453 5 9814 5 9978 6.0086 24 6.0008 6.1038 5.8000 5.7882 6.0128 5.9685 72 5 9981 5.8209 5.5366 6.0064 5.7059 5.1847 5.6609 120 5.9780 5.5933 5.2301 5.7251 5.2072 3.0492 5,0934 pH 0.5 6.5 6.4 0 6 6.5 6.4 7.0 240 5.9664 4.8382 3.2014 4.8919 2.7070 0.7782 3.6201 pH 0 6 6.7 6.5 6.7 6.0 6.9 7.4 336 5.8782 3.4900 1.3802 3.3902 1.2553 Sterile 1.7634 6.7 6.9 6,5+ 6.8 6.6 7.0 7.5 Experiment 2. Mortality of Bad. coli in double-distilled water contained in paraffin- lined bottles at different temperatures o°c. 10°C. 20''C. 30°C. TIME 1 II I II I II I II hours 0 6.1430 6.4116 6.3351 5.7340 5.9619 5.8325 6.1103 5,9474 pH 6.1 6.5 6.4 6.3 6.3 6.1 6.3 6.2 24 6.0000 6.3617 6.1761 5.4J25 5.7782 5.6096 5.8439 5.7076 48 5.9074 6.3234 6.1038 5.3617 5.7938 5.5366 5.8506 5.7235 72 5.8470 6.3962 6.0679 5.3522 5.7160 5.4871 5.7059 5.6794 120 5.8075 6.3O10 6.1504 5.4472 5.6314 5.4848 5.4298 5.6551 240 5.5575 6.1895 5.9143 5.2227 5.3560 5.0170 5.1818 5.5877 pH 5.9 6.1 360 5.1106 6.1096 480 4.9410 6.0852 5.3181 4.8109 5.2330 4.8882 4.8055 4.6911 pH 5.8 6.3 6.4 6.2 6.4 6.3 6.4 6.3 600 4.3766 6.0311 pH 6.0 6.5 720 5.3075 4.3424 5.1367 4.7177 4.4082 4.10(M pH 6.8 6.3 6.3 6.5 6.5 6.6 6.5 1080 4.9455 3.9768 4.7435 4.5988 3.0899 3.1399 pH 6.3 6.3 6.4 6.3 6.1 6.2 1464 4.7931 3.8261 4.0607 4.4843 2.3909 2.724a pH 6.5 6.5 6.7 6.7 6.6 6.5 1680 4.7435 3.5786 3.6263 4.3345 2.0000 2.5611 pH 6.3 6.5 6.7 6.6 6.8 6.7 223 224 BAKNETT COHEN Experiment S. Mortality of Bact. typhosum. in autoclaved tap water at different temperatures o°c. 10°C. 20°C. 30°C. TIME I I II I II I II hours 0 5.9468 6.1875 6.2279 6.1492 5.8779 6.2648 6.2330 pH 9.4 9.3 9.4 9.4 9.3 6 6.1399 6.1553 6.1206 5.8603 6.1732 6.2279 pH 9.1 9.2 9.2 9.2 9.3 9.3 12 5.7374 6.1303 6.1523 6.0494 5.8470 5.8149 5.8938 24 5.8000 6.0170 6.0294 5.9009 5.6821 4.7803 5.4871 72 5.4813 5.2504 4.6444 5.2175 4.9836 2.6551 2.6875 120 5.4009 5.1818 4.1523 4.8414 4.4728 2.1271 1.9542 pH 8.8 8.8 9.1 8.9 9.1 8 9 9.1 240 5.3372 4.3784 2.0607 3.9671 2.9128 1.2014 0.0000 pH 8.7 8.7 9.3 8.9 9.1 8.9 9.1 336 5.3195 4.0569 0 9031 3.6385 1.9031 pH 8.7 8.8 9.3 8.9 9.2 DISINFECTION STUDIES 225 Experiment 4. The morlalily of Bad. coli at different temperatures in Berkefeld- fdtered tap water O'C. 10°C 20°C. 30'C. TIUE I II I II I 11 I II hoUTt 0 6.2156 6 4267 6.0682 6.1345 6.0187 6.0278 5.8319 6.0682 pH 7.5 7.5 7.8 7.8 7.6 7.0 7.5 7.8 24 6.1697 6.4123 6.0800 6.0738 5 9557 5.9863 5.7513 6.0203 48 0.1967 6.3399 5.9782 6.0286 5.9494 5.9474 5.7917 6.0294 72 6.1242 6 3238 6.0064 6 0838 5.9727 5.9978 5.8082 5.9863 120 6.0856 6.3713 6. 0382 6.0569 5.9600 5.8901 5.7745 5.9948 240 6 1196 6 3577 5.8028 5.7185 5.9868 pH 7 3 7.1 360 6.0294 5.5611 5.9360 5.8537 5.4014 5.8692 pH 7.4 7.4 7.4 7.5 7.5 480 6.0766 6.4752 5.3096 5.7612 5.7664 4.7723 5.7543 pH 7.0 6.6 7.5 7.6 7.7 7.8 600 Contami- nated 5.1732 5.5752 5.6335 4.0086 5.5922 pH 7.5 7.6 7.6 7.7 7.6 696 5.9445 pH 6.7 6.6 744 5.0531 2.7528 5.5159 pH 7.6 7.3 7.6 840 5 9777 pH 7.0 960 4.7033 4.6609 5.0569 0.0000 5.0864 pH 7.3 7.4 7.4 6.5 7.4 1200 4.5490 4.0294 4.1399 0.0000 4.9106 pH 7.5 7.6 7.5 7.0 7.5 1440 4.5092 pH 7.7 7.7 Note: The tap water was not subjected to solving glass constituents from containers. high temperature to avoid dia- Experiment B. The effect of hydrogen ion concentration upon the mortality of Bact. typhosum for forty-eight hours in M/BOO buffer solutions at 20°C. pH3.8 pHS.O pH6.4 pH6.4 TIME I II I II I II I II hours 0 5.3636 5.5682 5.7348 5.7825 5.7889 5.9036 5.8579 6.0000 3 2.3010 2.3010 5.6464 5.6212 5.7024 5.7686 6 Sterile Sterile 5.5623 5.5922 5.6875 5.6776 5.7118 5.7924 12 5.4425 5.4409 5.7059 5.7076 5.6637 5.7543 24 5.4031 5.4166 5.5038 5.7267 5.4548 5.5775 48 5.2041 5.0294 5.2672 5 3711 5.2253 5.3118 VELOCITY COEFTICIENT. h 1.0209 1.0891 0.0110 0.0157 0.0109 0.0111 0.0132 ! 0.0143 pH7 1 pH7.6 pH 8.7 pH9.5 TIME I II I II I II I II htmrs 0 5.8195 5.6730 5.8482 5.8414 5.5922 5.7093 5.5752 5.5775 3 5.0088 5.1818 5.1139 4.8082 6 5.8899 5. 6180 4.9699 4.6284 4.4829 4.2148 4.0128 3.8096 12 5.8555 5.5478 4.6628 4.5441 3.3010 2.4771 2.0000 2.3010 24 5.5809 4.3997 2.9420 2.6675 1.0792 0.0000 0.0000 0.3010 48 4.6435 2.6532 0.0000 1.1139 Sterile Sterile Sterile Sterile VEL OCITY COEFFIC lENT, k 0.0244 0.0630 0.1215 0.0984 0.1887 0.2380 0.2979 0 2730 Experiment 6. Mortality of Bact. typhosum in M/SOO buffer of pH S.5 at different temperatures o°c. 10°C. 20''C. 30»C. TIME I II I " I II I 11 minutes 0 5.7760 5.8651 6.1004 6.0170 6.2856 5.9294 6.2041 5.9926 15 5.9106 5.4314 2.3617 2.0414 30 4.4533 4.1335 4.2695 4 6911 2.2625 1.6628 45 3.9854 3.6794 1.2788 1.0000 60 3.5065 3.8882 2.8865 3.3902 3.2455 3.1139 1 . 1461 0.6990 75 3.0000 90 2.5911 2.9420 120 1.8441 1.9294 1.9868 2.3032 180 1.2788 1.0414 0.3010 0.3010 240 1.1761 0.9777 300 Sterile 0.0000 VELOCITY COEFFICIENT, k 1.1500 1.2219 1 .9931 1.9053 3.0401 2.8155 5.0580 5.2936 Note: The buffer consisted of 50 cc. M /5 potassium acid phthalate plus 14.7 cc- M/5 HCI in 5000 cc. distilled water. Tliis had a pH of 3.2 which shifted to 3.5 during sterilization. 226 Experiment 7. Mortality of Bact. coli in M 1600 buffer of pH S.S at different temperatures o*c. lO'C. 20''C. 30'C. TIUB I II I n I II I II houn 0 6.1847 6.2648 6.2304 6 2455 6.0344 6.2455 6.2227 6.1703 1 5.9309 6.0253 2 6 0212 6.0607 5.0128 5.6415 3 4 9699 5.2718 4 6.1553 6.2227 6.1703 6.0192 5.6675 5.9063 4.363G 4.7627 5 3.4456 3.9590 6 2.7186 2.9289 8 6.0828 6.2227 6.0899 5.9590 5 2175 5.5092 0.6021 12 6.0334 6.1599 5.9C99 5.7007 4.0864 4.2253 15 6.0212 6.1761 5 9117 5 3139 24 5.7846 5.9538 5.5092 4.9186 32 5.5478 5.7796 5.2856 4.8055 VELOCITY CO EFFICIENT, k 0.0199 0 0152 0.0295 0.O450 0.1023 0.1684 0.7026 0.5402 Note: The buffer solution was identical with that used in experiment 6. Experiment 8. The mortality of Bact. coli at different temperatures in M/SOO buffer at pH 6.1 o°c. 10°C. 20°C. 30°C. TIME I II I II I II I II Aour» 0 6.2511 6.4578 6.0864 6.5821 6.4133 6.4249 6.1775 6.1790 24 6.1134 6.4305 50 6.1271 6.4301 74 6.0899 6,4085 120 6.0846 6.4698 5.9180 6.5617 6.3858 6.3786 6.0676 6.1265 240 6.0933 6.4288 5.9325 6.5023 6.4171 6 3655 6.0660 6.0702 360 5.9357 6.2703 5.8848 6.4183 6.3416 6.2878 6.022 6.0212 480 5,9722 6.3019 5.7582 6.4232 6.3232 6.2844 5.9058 5.7042 600 5.8082 6.4198 6.2122 6.2148 5 5809 5.2625 720 5.8839 6.3219 5.7627 6.3670 6.2227 6.2068 5.3444 4.9047 840 5.9672 6 3959 5.6839 6.3579 6.0029 6.0128 5 2480 4.7679 %0 5.7143 6.3619 6.1810 6.1422 4.9934 4.5587 1128 5.7152 6.4188 6.0745 6.0993 4.7185 4.3324 1230 5.6821 6.3275 6.0374 4.4440 4.1614 1440 5.9652 1560 5.5416 6.1461 5.8998 5.8357 4 1072 3.8831 VELOC TT COEFKIC lE.VT, k. 10' 0.3454 0.2800 0.3292 0.3775 1.326 1.895 Note: The buffer consisted of 50 cc. M /5 KHiPO, plus 4.75 cc. M /5 NaOH in 5000 cc. distilled water. The original pH of 5.9 shifted to 6.1 during sterilization. 227 Experiment 9. The mortality of Bad. coli at different temperatures in M/500 buffer at pH 7.1 o°c. lO-C. 20°C. 30'C. TIME I II I II I II I II hours 0 6.3701 6.4752 6.3811 6.4514 6.5316 6.2405 6.3113 6.5051 25 6.3304 6.4440 6.4984 6.2896 6.2516 6.4786 48 6.3560 6.3574 6 4422 6.2350 6 2417 6.4735 74 6.3369 6.4734 6.2863 6.4019 6.4393 6.2054 6.2169 6.4824 120 6.2941 6.4914 6.3218 6.4084 6.4764 6.2222 6.2300 6.4433 240 6.3707 6.4436 6.2941 6.3516 6.4777 6.1945 6.1897 384 6.3287 6.4541 6 3240 6.3104 6.3647 6.0682 6.0708 6.2601 480 6.3072 6.4254 6.2397 6 3118 6.3879 6.0792 5.8248 5.9643 600 6.3131 6.4079 6.3(H7 6.3843 6.2471 5.9294 5.4843 5.6712 840 6.2922 6.3728 6 1553 6 2989 6 2034 5.9015 5.0864 5.2363 984 6.1239 6.3560 5.8451 4.9395 5.2455 1200 6.1453 6.0998 5.5366 4.7723 5.0374 1440 6.1106 6.2227 5.9041 5.3874 4.5855 4.8785 VELOCITY COEFFICIENT, li. 10' 0.0928 0.1220 0.188 0.159 0.436 0.593 1.20 1.13 Note: The buffer consisted of 50 ec. M /5 KHjPO, plus 29.63 cc. M /5 NaOH in 5000 cc. distilled water. The original pH of 7.0 shifted to 7.1 during sterilization. Experiment 10. The mortality of Bad. coli at different temperatures in M/SOO buffer at pH 8.0 o°c. IO°C. 20°C. 30°C. TIME I II I II I II I II hours 0 6.1818 6.4526 6.3892 6.3485 6.5498 6.4439 6.4417 6.4882 24 6.1120 6.3141 6.2856 6.3464 6.3575 6.2923 6.2487 6.2777 48 6.1139 6.3395 6.2853 6 2889 6.3187 6.2350 6.1281 6.1351 75 6.0503 6.2413 6.2826 6.2084 6.2011 6.1383 5.9759 6.0374 120 6.0128 6.2227 6.2608 6 2209 6.1351 5.9309 6.0000 240 6.0212 6.2034 6.0958 6.1106 6.0212 5.9624 5.7010 6.0086 360 5.9542 6.2470 6.1367 6.1523 6 0425 5.7825 5.5250 5.8351 480 6.0029 6.2427 6.1470 6.1162 6 0107 5.7657 5.2504 5.6464 600 5.9881 6.0294 5.9661 5.6911 5.1875 5.4518 720 6.0238 6.2365 5.9445 6.0199 5.8865 5.5786 4.8825 5.3243 840 5.8651 5.9633 5.8451 5.5119 4.6503 5.2601 960 5.9206 6.0098 5.9143 5.6981 4.5453 5.2577 1200 5.9243 6 0029 5.8082 5.5933 4.0229 5.1430 1440 5.8938 5 9912 5.7745 5.3032 3 4787 3.7067 VE LOCITYCOEI FICIENT, k. 10' 0.2194 0.3141 0.3441 0.2841 0.5385 0.7925 2.058 1.933 Note: The buffer consisted of 50 cc. of M /5 ICHoPO, plus 46.80 cc. M /5 NaOH in 5000 cc. distilled water. After sterilization, the pH remained at 8.0. Toward the end of the experiment, the pH gradually shifted to 7.7, apparently because of the low concentration of the buffer and the nearness to the limit of its zone of effectiveness. 228 Experiment It. Mortality of Bad. coli in M 1600 buffers at iO'C. pH8.4 pHS.O pH9.5 TIME I II I II I II *our» 0 6.5391 6.2584 6 3874 6.4564 6.1818 6 2577 24 6.5159 ■ 6.2430 6 3973 6.3079 6.1703 6.1206 48 6.4683 6.2414 6.3784 6.4502 6.0719 6.0294 72 6.6021 6.2492 6.4031 6.0374 6.0607 120 6.4969 6.1614 6.4216 6.4654 6.0607 5.9638 264 6.4393 6.2072 6.4089 6.4314 5.9890 6.0086 480 6.5809 6.2227 6 5172 5.9890 6.0098 624 6 5539 6.2227 6.5441 6.4624 5.9912 5.7973 840 6 4955 6.2201 6.4133 6 3711 5.8494 5.8573 pH shifted to 7.9 Experiment IS. The mortality of Bact. lyphosum in autoclaved tap water at SO°C. TIME I II III 0 6.5873 6.5628 6.7284 pH 8.9 8.7 8.4 6 6.5463 6.5721 6.6887 pH 8.9 8.7 8.4 24 6.4742 6.5575 6.6939 pH 9.0 8.8 8.8 48 5.1931 6.5968 6.6532 pH 8.9 8.8 8.7 72 3.4914 5.2742 6.6135 pH 8.9 8.7 8.7 96 2.6096 4.9074 6.5064 pH 8.9 8.7 8.6 120 2.4065 4.6375 6.4389 pH 8.9 8.7 8.5 144 2.2201 4.7771 6.2695 pH 8.9 8.7 8.6 167 1.9243 4.2380 6.1804 pH 8.9 8.5 8.5 216 1.1139 3.6021 5.6821 pH 8.5 8.3 8.2 264 0.6990 5.7275 5.5159 pH 8.2 7.7 7.6 312 Sterile 5.6096 5.2684 pH 8.1 7.9 7.9 384 5.4654 4.9299 8.3 8.1 8.1 229 230 BARNETT COHEN Experiment 13. Morlality of Boot, typhosum in M/BOO buffer of pH 6.S-6.B at SO'C. TIME PHOSPHATE BUTTEE PHTHALATE BUTFEB I II III I II III ' hour a 0 6.6133 6.5376 6.5417 6.6073 6.5234 6.4469 6 6.6314 6.5703 6.5931 6.6703 6.5289 6.4510 24 6.6583 6.5090 6.5916 6.6086 6.5489 6.4713 48 6.5991 6.5714 6.5895 6.6074 6.4821 6.4004 72 6.5705 6.5907 6.5844 6.6425 6.4938 6.4327 96 6.6087 C.5609 6.5585 6.5821 6.5314 6.4284 120 6.5627 6.5038 6.4813 6.4846 6.4705 6.4480 144 6.5714 6.5502 6.5370 6.5555 6.4548 6.3720 167 6.5991 6.5658 6.5658 6.4705 6.4702 6.3988 216 6.5363 6.5502 6.3375 6.4975 6.3756 6.3290 264 6.5647 6.5563 6.5922 6.5350 6.4487 6.4294 312 6.6206 6.5723 6.6157 6.5488 6.4932 6.3639 384 6.5899 0.5447 6.5465 6.5798 6.5452 6.4396 552 6,2876 6.4976 6.6204 6.4343 6.3897 6.2819 763 5.5872 5.9064 5.9694 5.9890 6.0174 5.6945 960 5.0515 5 3345 5.7179 5.7669 5.8873 5.2978 Note: The Clark and Lubs phosphate and phthalate buffers of pH 6.0 were diluted to M/500 and autoclaved, after which their pH remained between 6.3 and 6.5. This experiment was performed six months after Nos. 1-11 and shows that any specific effect of the buffer as a salt is subordinate to the pH effect, at least under these conditions. MICROORGANISMS CONCERNED IN THE OXIDATION OF SULFUR IN THE SOIL I. INTRODUCTORY! SELMAN A. WAKSMAN Received for publication June 16, 1921 Hydrogen sulfide and other sulfides in solution are slowly- oxidized to sulfur, under natural conditions. The sulfur formed, as well as elementary sulfur, particularly when present in a fine state of subdivision and in the presence of certain catalytic agents, undergoes further oxidation, the resulting product being sulfuric acid. In the presence of specific bacteria, this phenom- enon is much more rapid. These bacteria were first studied extensivel}^ by Winogradsky and were designated by him as sulfur bacteria; they possess a strong oxidizing power, in con- tradistinction to the reducing bacteria, which form hydrogen- sulfide from sulfur and its compounds. The sulfur bacteria, originally studied by Winogradsky, contain sulfur granules within their cells, as a result of the oxidation of the hydrogen sulfide. Later investigators, in their studies of sulfur oxidation, also included, under the term sulfur bacteria, organisms which are able to oxidize hydrogen sulfide, sulfur, and thiosulfate, but which do not, however, store any sulfur within their cells. The so-called sulfur bacteria are not related to one another morphologically and belong to widely difTerent genera. By the use of physiological and morphological differences, they can be di\'ided into five groups: (1) colorless, thread forming bacteria, accumulating sulfur within their cells. (2) Colorless, non- thread-fonning bacteria, accumulating sulfur within their cells. (3) Purple bacteria, oxidizing sulfur and accumulating it within 'Paper No. 32 of the Technical Series, New Jersey Agricultural Experiment Station, Department of Soil Chemistry and Bacteriology. 231 232 SELMAN A. WAKSMAN their cells. (4) Bacteria oxidizing sulfur and sulfur compounds, but accumulating sulfur outside their cells. (5) Bacteria oxidiz- ing elementary sulfur and not accumulating any sulfur within or without their cells. The first four groups are mentioned in the latest texts and re\dews on sulfur bacteria, while the fifth group has been suggested on the basis of the results presented below. It is of interest to note that in most of the work on sulfur oxidation by bacteria, with the exception of that of Jacobsen (1912),^ the starting point was not suKur itself, but hydrogen sulfide, sulfides, or thiosulfates. The activity of the organisms was judged not by the oxidation of suKur, as measured by the production of sulfuric acid and sulfates and subsequent change in reaction, but either by the disappearance of the sulfide in the medium or by the appearance or disappearance of the suKur granules within or without the microbial cell. Of importance is also the fact that all the earher work and most of the later work has been done with organisms present in canal water, mud water, curative muds, and sea water and very Uttle attention was paid to the microorganisms concerned in the oxidation of sulfur in the soil. This is the reason for using hydrogen sulfide and sulfides as a source of sulfur, since the latter substance, and not elemen- tary suKur, is present or is produced abundantly under those conditions. A detailed review of the oxidation of sulfur and sulfur com- pounds by microorganisms is given in the papers of Omeliansky (1904), Diiggeh (1919) and in the book by Kruse (1910). A brief historical re\'iew is presented here for a better understanding of the work that is to follow. The first group of suKur oxidizing bacteria (colorless, thread forming) consists of three genera: Beggiatoa, motile, forming no sheaths; Thiothrix, fastened, forming no sheaths, and Thio- ploca, threadforming bacteria, surrounded with a jelly-like sheath. The Beggiatoas were the first organisms to attract attention as having to do with the oxidation of sulfur or its derivatives. Cramer (1870) pointed out that the granules found 'Bibliography is found at the end of the second article, in this series. OXIDATION OP SULFUR IN THE SOIL 233 within the cells of Beggiatoa consisted of sulfur. Cohn (1875) then proposed the theory that the Beggiatoa and the purple bacteria produce hydrogen sulfide by the reduction of sulfates. But it was Winogradsky (1883, 1887, 1888) who demonstrated that the hydrogen sulfide produced by other bacteria is oxidized by the Beggiotoa to sulfur and sulfuric acid. 2HS +02 = 2HoO + Sj + 122 Cal. Sj + 3O2 + 2H,0 = 2H0SO4 + 282 Cal. This oxidation is so important for the very existence of these organisms that, when the hydrogen sulfide is taken out of the medium, they oxidize the sulfur present witliin their cells and, when this is used up, they die out. The presence of traces of organic substances and nitrates in the water is sufficient for the development of these organisms, as long as there is enough hydro- gen sulfide, while the presence of sugars, peptone and like nutrients will stimidate the growth of other microbes but will injure these sulfur bacteria. According to Winogradsky (1888), the sulfuric acid formed is neutraUzed by the calcium carbonate present in the water. H2SO4 + CaCOa = CaS04 + CO2 + H.O H2SO4 + CaH, (003)2 = CaS04 + 2COo + 2H2O The reaction of the water cultures of the acid producing bac- teria was not found to become acid. Beggiatoa has been obtained in pure culture by Keil (1912). _ No definite physiolog- ical studies were made by Winogradsky. In general, the results of this investigator are summarized under the following four headings: (1) The sulfur bacteria oxidize hydrogen sulfide and accumulate sulfur in the form of small spheres, consisting of soft amorphous sulfur which never crystalUzes in the hving cells. (2) Thej^ oxidize the sulfur to sulfuric acid, which is at once neutra- Uzed, by the carbonates present, into sulfates. (3) Without sulfur, the organisms soon die off. (4) They can live and mul- tiply in liquids containing only traces of organic substances. This last point was refuted by Keil (1912), who demonstrated that the organisms are autotrophic and do not need organic 234 SELMAN A. WAKSMAN substances for their growth. Keil claims to have isolated pure cultures of Beggiatoa and Thiothrix, and found these organisms to be able to live in media free from any traces of organic matter, although the presence of small quantities of organic substances is not detrimental to these organisms. Ammonium salts are used as sources of nitrogen and only carbonic acid as a source of carbon. Carbon dioxide pressure may vary within the limits of 0.5 and 350 mm. (25 mm. is the optimum) ; oxygen may vary within 10 to 20 mm. and H2S within 0.6 to 1.7 mm. The presence of carbonates is important for the neutralization of the acids. The organism seems to assimilate carbon at the rate of 1 gram per 8 to 10 grams of sulfur oxidized (see also Hinze and INIolisch). The Thioplaca has been studied in detail by Wisloch and Kolkwitz. The second group of the sulfur oxidizing bacteria consists of colorless organisms forming no threads and containing sulfur within their cells. The following forms belong to this group: Monas (Hinze (1913)), Thiophijsa (Hinze (1903)), Thiovulum (Hinze (1913)), Spirillinn (Molisch (1912)), Thiospirillum (Ome- lianski (1905)), Bacteman bodsta (Molisch (1912)), Bacillus thiogenes (Molisch (1912)), and Achromatium (Nadson (1903), and Griffith (1913)). Some idea of this group of organisms is obtained from a ref- erence to the work of a few investigators. Jegunow (1896) studied the oxidation of the hydrogen sulfide formed in the mud of the Liman in Odessa and in the Black Sea. He described two sulfur bacteria; Thiobacterium a, a motile, colorless, slightly curved organism, 4.5 to 9^ long and 1.4 to 2.3^ wide, containing a finely granulated plasma and large sulfur granules. Thio- bacterium /3, motile, colorless, curved, 2.5 to 5 by 0.6 to 0.8//, and containing a row of shining sulfur granules. Monas MiilleH was described in detail by Hinze (1913), who has shown that this organism belongs morphologically to the flagellates and physio- logically to the sulfur bacteria. jNIost of these organisms were isolated from the water. Gicklehorn (1920) recently described several new sulfur bac- teria, of which tv;o are classified with this group, namely : Spiril- OXIDATION OF SULFUR IN THE SOIL 235 lum agilissimum isolated from river mud in Gratz which measures about 6 to 10 by 1.8 to 2/n, of a rapid motiUty,and filled with black sulfur granules; Chromalium cuculliferum which is round to slightly elhptical, 6 by 4^, of a slow motiUtj^, with black, shining, sulfur drops alwaj'S found in one pole, with one flagellum on the granule-free pole. This form was found in a rotting mass of algae in the garden basin at Gratz. Gicklehorn also described three more forms: Microspira vacillans, Pseudomonas bipundata and Pseudomonas hyalina, observed in the slime of the large basin in the botanical garden at Gratz, which he classified with the colorless sulfur bacteria. It must be noted here that none of the forms studied by Gicklehorn were cultivated in pure culture, which is true of most of those studied by the other investigators. The third group of the sulfur oxidizing organisms is found among the purple bacteria. These are distinguished from the sulfur bacteria described above by the production of a red, red \'iolet or red brown pigment which is unevenly distributed throughout the cell. In addition to the red pigment (bacterio- purpurin), there is also present in all these bacteria a green pig- ment (bacterio-chlorin). These bacteria are found abundantly in sulfur springs and in mud waters. Not all the purple bacteria are able to utilize hydrogen sulfide. Mohsch (1907) succeeded in cultivating some of them in pure culture, but not the sulfur forms. The role of sulfur in the metabolism of purple bacteria is still an open question, since, according to Nadson (1903) and Molisch (1907), the hydrogen sulfide is not required for nutrition and sulfur is not accumulated. These results are in direct opposition to the earlier ideas of Winogradsky (1888) and others. A detailed study of the purple bacteria is found in the work of MoUsch (1907). The fourth group of sulfur bacteria includes colorless organisms that do not accumulate sulfur within their cells. These were first demonstrated by Nathanson (1903) in sea water, and were found to be able, by means of oxidation of hydrogen sulfide or sodium thiosulfate, to reduce carbonic acid and construct from it organic substances. By using a medium consisting of 3 per cent NaCl, 0.25 per cent MgCl2, 0.1 per cent KNO3, 0.05 per cent 236 SELMAN A. WAKSMAN Na2HP04, 0.2 to 1 per cent Na2S203 and some JVIgCOj, he ob- tained a good growth of these bacteria and, on adding agar, he has been able to isolate them in pure culture. In the absence of the carbonate, but in the presence of CO2 containing air, the growth was much slower; in the absence of both carbonate and CO2, no growth took place, even in the presence of various organic substances. The medium did not become acid even in the ab- sence of carbonate, indicating that the oxidation of the thiosul- fate did not take place according to the formula : Na2S203 + H2O + 40 = NajSOi + H2SO4. but according to the following reaction: " 3Na2S203 + 50 = 2Na2S04 + Na2S406. While no sulfur accumulates within the cell, there is an abundant production of free sulfur outside of the cell, not in direct contact with the colony, but at some distance from it, suggesting an extra- cellular oxidation. Beijerinck (1904) confirmed the results of Nathanson (1903) by the use of a medium consisting of 100 parts of water, 0.5 NasSjOs. 5H2O,0.1NaHCO3, O.O2K2HPO4, 0.01 NH4CI and 0.01 part MgCU. The medium was not sterihzed, was inoculated with canal water and incubated at 28 to 30°. In 2 to 3 days, the surface of the medium became covered with free sulfur, filled with bacteria. On making a transfer into a fresh flask with medium, a suKur layer was obtained in 24 hours. According to Beijerinck (1904), the reaction takes place as follows: Na2S203 + O = Na2S04 + S. This reaction is exothermic and functions as a source of energy, which is used for the reduction of NaHCOa and for the building of the bacterial body. CaS and HoS can replace the thiosulfate. H2S+O = H2O +S. Nan S4 Os + Na2 CO3 + 0 = 2 Na2S04 + CO2 + S2 The ammonium salt can be replaced by nitrates. No other of the organic substances tested could replace the carbonic acid as OXIDATION OF SULFUR IN THE SOIL 237 a source of carbon. The organism, Thiobacillus thioparus, is a short rod, 3 by 0.5^, not forming any spores, very motile and very sensitive, so that on plates the organisms die off in a week. By using a medium consisting of canal water 100, powdered sulfur 20, KNOs - 0.5, Na^COa - 0.02, CaCOj - 2.0, K2HPO4 — 0.02 parts, in a closed flask, incubated at 30°C., Beijerinck obtained an oxidation of sulfur accompanied by a reduction of the nitrate to atmospheric nitrogen. 6KNO3 + 5S + 2 CaCO, = 2K0SO4 + 2 CaSOi + 2 COo + SNs + 695.5 Cal. The sulfur is oxidized to sulfuric acid which acts upon the CaCOj giving CaSO^ and CO2. By using a medium consisting of tap water - 100, Na,So03. 5 H2O - 0.5, K.HPO4 - 0.01, NaHCOa — 0.02, agar 2.0 parts, Beijerinck isolated, in pure cul- ture, the organism, Thiobacillus denitrificans, which is a very mo- tile, short rod, hardly distinguishable microscopically from Thio- bacillus thioparus. Thiohacillus denitrificans was further studied by Lieske (1912) and Gehring (1915) and was found to occur in various soils. The organisms on the plate, lose their ability to grow rapidly, long before they are dead. Beijerinck's work was continued further by Jacobsen (1912, 1914), who found a crude culture of Thiobacillus thioparus to be able to oxidize 58.8 mgm. of sulfur to sulfuric acid, in five weeks, out of a total of 648 mgm. added to the medium. Pure cultures oxidized, in eight weeks, 165 mgm. out of 648 mgm. of sulfur added. Gicklehorn (1920) has studied two organisms belonging to the fourth group of sulfur-oxidizing bacteria, found in garden soil, which are able to oxidize K2S with the liberation of free sulfur. The organisms are 1 to 2 by 0.3 to 0.5^ and 2 to 4 by 0.5 to l^i in size. However, he did not isolate his organisms in pure culture and did not record any quantitative physiological data. Finally, we have a fifth group of sulfur bacteria, which are studied in detail in the next paper. Two preliminary reports on this organism by Waksman and Joffe (1921a, 1921b) and a detailed study of the methods used in its isolation, by Lipman, Waksman and Joffe (1921), were pubUshed elsewhere. 238 SELMAN A. WAKSMAN To summarize: (1) The microorganisms concerned in the sulfur cycle are separated into reducing bacteria and oxidizing bacteria, the latter being the true sulfur bacteria. (2) The true sulfur bacteria are divided into five groups: the first three groups of sulfur bacteria are found in sulfur springs, canal and mud waters, curative muds, river water and sea water; they oxi- dize hydrogen sulfide and sulfides, but not elementary sulfur, and accumulate sulfur within their cells; the fourth group of bacteria, consisting of small rod shaped organisms, is found in sea water, canal water and soil; these bacteria are able to oxi- dize hydrogen sulfide and other sulfides, thiosulfates and ele- mentary sulfur, forming a heavy pelhcle on the surface of the me- dium and allowing an accumulation of sulfur outside of their cells; the fifth group of sulfur bacteria occurs in soils to which ele- mentary sulfur has been added, particularly in soil-sulfur-com- posts, oxidizing primarily elementary sulfur, thiosulfates to a small extent, but not hydrogen sulfide or sulfides; these bacteria grow uniformly throughout the medium, not forming any peUicle, do not liberate any sulfur and allow a very intensive production of sulfuric acid, and the necessary carbon is derived entirely from the carbon dioxide of the atmosphere; the fifth group is morphologically related to group four, but includes organisms very small in size and the strongest sulfur oxidizing and acid pro- ducing bacteria known. IMICROORGANISMS CONCERNED IN THE OXIDATION OF SULFUR IN THE SOIL II. THIOBACILLUS TIIIOOXIDANS, A NEW SULFUR-OXIDIZING ORGANISM ISOLATED FROM THE SOIL' SELMAN A. WAKSMAN and J. S. JOFFE Received for publication June 16, 1921 By composting sulfur, rock phosphate and soil it was found (McLean, 1918) that sulfur is rapidly oxidized to sulfuric acid; the acid acts upon the tricalcium phosphate, converting it into di- and mono-calcium salts. In the absence of a neutralizing agent or, after this agent has all been used up, the sulfuric acid formed, in the presence of an excess of sulfur, accumulates in the medium. On inoculating such composts into proper culture media, we finally succeeded in isolating a small bacterium which is active in the oxidation of the sulfur. A detailed study of the composting of sulfur, of the transformation of the tri-calcium phosphate and of the methods used in the isolation of the organism are found elsewhere (Lipraan, Waksman and Joffe, 1921); only a brief review of the process of isolation is presented here. Method of isolation. The following media were originally used for the isolation of the organism: Medium 1 : (NH,), S0« 2.00 gram KjHPO, 1.00 gram MgSO« 0.50 gram KCl 0.50 gram FeSO, 0.01 gram Sulfur 10.00 grams Ca, (P04)j 10.00 grams Distilled water 1000.00 cc. Medium S: Same as no. 1, but with 0.1 per cent glucose. Medium 3: Same as no. 1, but in place of 10 grams only 2.5 grams Caj(PO«)i ^ per liter. 'Paper No. 33 of the Technical series, New Jersey Agricultural Experiment Stations, Department of Soil Chemistry and Bacteriology. 239 240 SELMAN A. WAKSMAN AND J. S. JOFFE The media were distributed in 100 cc. portions into 250 cc. Erlenmeyer flasks and sterilized in flowing steam, for 30 minutes, on three consecutive days. The flasks were then inoculated with various dilutions of the composts. ]\ledium 2 was found to allow a growth of both a sulfur oxidizing bacterium and one or more species of fungi. By omitting the glucose from the medium, the fungi were practically ehminated. It was found later that, by cutting down the tri-calcium phosphate in the medium to 0.25 per cent, a more rapid develop- ment of the organism took place, thus giving medium 3, which is a modification of 1 and 2. Well advanced composts were used for inoculation. The material was diluted 10, 1000, 100,000 and 10,000,000 times with sterile water, then 1 cc. of each dilution was added to 100 cc. of the sterile medium and the flasks incubated, at 25°, for seven to fourteen days. The flasks became turbid on the fourth or fifth day, the amount of turbidity depending upon the dilutions used, the higher dilutions developing slower than the lower ones. A pellicle or fungus mycelium was formed only in the flasks containing glucose. By transferring the cultures into fresh flasks, the same phenomenon was observed A\'ith a uniform turbidity in four to five days. By examining the culture under the micro- scope, it was found to contain a very minute non-motile bacterium present in abundance and accompanied by a few larger cylindri- cal cells which were found to be spores of a fungus occurring abundantly in the compost. The impure culture of the or- ganisms was found to possess strong sulfur-oxidizing properties, about 200 to 300 mgm. of the sulfur being oxidized, in each flask, in fourteen days. In the presence of tri-calcium phosphate more of the sulfur is oxidized, since the acid formed is used up in converting the insoluble phosphate into soluble calciimi-acid- phosphate and calcium sulfate. A further accumulation of the sulfuric acid resulted also in the formation of phosphoric acid and calcium sulfate. The medium had originally a reaction equivalent to pH 5.6 to 6.2. Following the oxidation of the sulfur, the reaction became gradually acid and, at a pH of OXIDATION OF STJLFUK IN THE SOIL 241 2.6-2.8, the reaction remained stationary till all the tri-calcium phosphate had been transformed into mono-calcium salt, after which the reaction became more acid, as shown in table 1 and figure 1. All attempts to grow the sulfur-oxidizing organisms on solid media failed, neither agar nor silica-jelly media allowing any growth to take place. TABLE 1 Course of reaction and accumulation of water soluble phosphates AOB OrCCXTCRE pH PER CENT OP IN80LDBLE PBOS- PHATEa MADE WATER 80LDBLE* dans At start 5.4 1 5.4 2 5.3 0.9 4 4.6 5.5 6 3.5 8 2.6 33.7 10 2.7 27.5 12 2.6 81.7 15 2.4 93.9 19 2.3 86.3 23 2.3 85.9 30 1.8 • 38 1.8 86.1 68 1.7 120 0.8 85.9 •Medium contains originally 1 per cent insoluble phosphate. A pure culture was obtained by continued transfer in fresh flasks with high dilutions, so as to eliminate any contaminating organism, the medium being made acid at the start (pH 2.0- 3.0), by the use of phosphoric acid and mono-potassium phos- phate. The culture was finally obtained in a pure state. Its purity was demonstrated by repeated microscopic examinations, by the uniform growth in the liquid media and by the fact that no organism developed, when the culture was inoculated upon common bacteriological media. 242 SELMAN A. WAKSMAN AND J. S. JOFFE On repeated transfer, the culture was found to deteriorate since it took a longer period of time to develop. It was found, necessary, in order to obtain a good growth, to use a sterile pipette instead of a loop, one or two drops being sufficient to inoculate 100 cc. By buffering the medium with suitable sub- stances, such as phosphates, the organism would develop much more rapidly, particularly at the more acid reactions. The 1^ 0, P^ curve per cent soluble P curve *^ ptr/od of mctft^at/oPi /n ^ay. I Fio.l. CourseofReaction AND Accumulation OF Water Soluble Phosphates IN A Pure Culture of Thiobacilliis thiooxidians n. sp. organism was found to be morphologically similar to the two ThiobaciUi described by Beijerinck, and it is, therefore, classified in that genus, under the name of Thiohacilliis thiooxidans n. sp. Morphology. Vegetative cells, on the synthetic media used, are short rods, with rounded ends, usually occurring singl}', to some extent in pairs and rarely in triplets. The majority are less than 1 micron long and about 0.5 micron in diameter. Spore formation, absent. The majority of the cells are non-motile, although a few motile cells can also be found in young (seven OXIDATION OF SULFUR IN THE SOIL 243 days old) cultures. The organism stains well with gentian- violet and methylene blue. It is Gram-positive. CULTURAL CHARACTERISTICS No agar or other solid medium has been found as yet, upon which the organism would grow. It grows in licjuid media with a strong uniform clouding, without any surface growth or sedi- ment formation. It does not grow on the common organic media, although the presence of glucose or peptone in the medium is not injurious. Inorganic media containing sulfur as a source of energy are suited for its growth. In the presence of tri- calcium phosphate, the growth of the organism is accompanied by characteristic reactions: the sulfur forming originally a layer on the surface of the medium usually drops to the bottom, the sulfuric acid formed from the oxidation of the sulfur dissolves the tri-calcium phosphate giving soluble phosphate and CaSOr 2H2O, the calcium sulfate crj^stallizes out in the form of radiating mono- clinic crystals hanging down from the sulfur particles that are floating on the surface of the medium or protruding upward from the bottom. The reaction of the medium becomes acid as indicated by the change in the hydrogen-ion concentration. At a pH of about 2.8, the reaction becomes stationary till all the calcium-phosphate has been dissolved. In the presence of an excess of this neutralizing agent, or in the presence of rapidly dissolving alkaline carbonates, the culture is injuriously affected. Anything that will tend to change the medium to an alkaline or even a less acid reaction (except, of course, the action of the buffers), such as shaking the culture, in the presence of even smaller amounts of tri-calcium phosphate, will also tend to affect the uniform growth of the organism injuriously. The culture can be kept alive for numerous consecutive genera- tions on the liquid media and when not injured by an excess of alkali or acid, may be as active as a recently isolated culture. The index No. of ThiobaciUus thiooxidans is, according to the new Descriptive Chart of the Society of American Bacteriolo- gists, 5332-5230-2222. 244 SELMAN A. WAKSMAN AND J. S. JOFFE PHYSIOLOGY Source of carbon. The organism derives all its carbon need from the CO2 of the atmosphere. TNTien carbon was introduced into the culture in the form of carbonates and bicarbonates, the presence of the former prevented growth due to the fact that they kept the medium alkaline, thus preventing a normal development of the organism, while the latter, if present only in small amounts, allowed a good growth to take place. But since the growth was not any better, and to some extent even worse than in the bi- carbonate-free flasks, its use is superfluous. At this point, we get a clear differentiation in the metabolism of two important autotrophic organisms, the nitrifying and the sulfur-oxidizing bacteria. While the former thrives best at an alkaline reaction, the latter grow best at an acid reaction. Sodium bicarbonate is considered to be indispensable for the nitrifying bacteria; this was thought to be due to the utilization of the bicarbonate as a source of carbon, but, as recently pointed out by Meyerhof (1916), the bicarbonate merely serves the purpose of a buffer in the medium, to keep the reaction alkaline (optimum pH 8.3- 9..3). In the case of the sulfur oxidizing bacterium, which has its optimum at a distinctly acid reaction (pH 3.0-4.0), the bi- carbonate is not necessary since its buffering properties will tend to make the medium less acid and thus have an injurious effect, while as a source of carbon, the CO2 from the atmosphere seems to be sufficient. Source of energy. Sulfur is the all important source of energy for this organism. The organism is strictly autotrophic and, although glucose did not exert any injurious action, and perhaps its action was even to some extent beneficial, the amount of sulfur oxidized and acid produced were about the same in glucose and in glucose-free cultures. In addition to sulfur, thiosulfate is also utilized, but to a much smaller extent: while, with elementary sulfur, gi-owth appears in four to five days, under favorable conditions, as demon- strated by the turbidity and change in pH value, with thiosulfate, growth appears only in ten to twelve days and is much slower. OXIDATION OF SULFUR IN THE SOIL 245 Hydrogen sulfide and sulfides are not utilized at all, which sharply differentiates our organism from those of Nathanson (1903), Beijerinck (1904), and Jacobsen (1914), as will be pointed out later. Mineral requirements. Mere traces of K, Mg, Ca, Fe, in ad- dition to phosphates, are sufficient for the growth of the organism. As a matter of fact, good growth and good sulfur oxidation were obtained by omitting, in various batches of media, each of the first four minerals, but, of course, no precaution was taken to eliminate any traces present in the distilled water or any sub- stances that might have been dissolved out by the action of the sulfuric acid on the glass of the flask. Source of nitrogen. Due to the very small amount of growth made by the organism, the amount of nitrogen required is very small: without introducing any nitrogen source into the medium, some growth is obtained, the nitrogen being derived either from the contamination of the other salts, the distilled water, or traces of ammonia in the atmosphere. The best sources of nitrogen are ammonium salts of inorganic acids (particularly sulfate), followed by the ammonium salts of organic acids, after which come the nitrates, asparagin and amino acids. Nitrites, in concentrations used (2 grams per liter) are toxic. Good growth is obtained with pepton, but the amount of sulfur oxidized is less than with the other sources of nitrogen. Relation to oxygen. The organism is strictly aerobic, in view of the fact that it derives the oxygen necessary for the oxida- tion of sulfur to sulfuric acid from the atmosphere. Influence of organic substances. As pointed out above, glucose does not act injuriously, neither do other organic substances, like pepton. Substances like glycerol, alcohol, mannitol and glucose seem to have a slight favorable effect in the presence of a good nitrogen source. All these substances either act like stimulants or else take part in the structural requirements of the organism. Influent of stimulants. In addition to the pure organic substances, above mentioned, which may stimulate to some extent the growth of the organism, other substances may exert JOUBNAL or BACTBBIOLOOT, VOL. Til, NO. 2 246 SELMAN A. WAKSMAN AND J. S. JOFFE the same action. A detailed study of the influence of stimulants on the oxidation of sulfur by a pure culture of Thiobacilllus thiooxidans is presented in table 2. The medium was buffered with phosphoric acid and mono- potassium-phosphate to a pH of about 3.0 It was distributed TABLE 2 Influence of stimulants on the oxidation of sulfur TOTAL Ca. (POOj 0,£5 PER CENT STIMULANT, 0.1 PERCENT SULFATES IN 100 CC. OP MEDIUM MILLIGRAMS pH TITH.ITION* mgm. + Control 241.0 3.0 1.0 -1- 1476.0 1.4 2.8 — 912.5 1.6 2.5 CaS04 (0.25 per cent) 1041.0 1.4 2.7 + Glucose 991.0 1.5 2.7 — Glucose 1012,5 1.3 2.6 + Mannitol 1008.0 1.6 2.5 — Mannitol 859.0 1.5 2.5 + Glycerol 858.2 1.6 2.5 — Glycerol 948.5 1.4 2.6 + Alcohol 905.0 1.6 2.5 — Alcohol 994.0 1.4 2.8 + Soil 1019.5 1.6 2.7 Soil 1226.8 1.4 2.6 + Al2(S04)a 1394.0 1.4 2.9 — Alj(S04), 989.0 1.4 2.6 + Thallixun nitrate 1133.1 1.5 2.7 _ Thallium nitrate 883.5 1.5 2.5 + MnS04 1013.4 1.6 2.5 - MnS04 933.8 1.4 2.5 'Titration = cubic centimeter of j^ NaOH necessary to neutralize 1 cc. of culture, with phenolphthalein as an indicator. in 100 CC. portions in 250 cc. Erlenmeyer flasks containing 1 gram portions of powdered sulfur and the proper amounts of Cas (P04)2, where present. The flasks were plugged with cotton and sterihzed in flowing steam, for thirty minutes, on three consecutive days. The organic substances were sterilized separ- ately, then added to the sterile medium. The flasks were all inoculated with one drop of the same pure culture and incubated OXIDATION OF SULFUR IN THE SOIL 247 for twelve days. At the end of that period, the pH was deter- mined by the colorimetric method, titration was obtained from the amount, in cubic centimeters, of j^ NaOH necessary to neutralize 1 cc. of the filtered culture using phenolphthalein as an indicator. The total sulfates were obtained by adding the amounts of soluble and insoluble sulfates : the latter were obtained by digest- ing the filtered residue in acidulated water and determining the sulfates in an aliquot portion. In the presence of calcium phosphate, the largest amount of sulfur oxidized by a pure culture of the organism was obtained in medium 1, to which no stimulating agent has been added. In the absence of the tri-calcium phosphate, the amount of sulfur oxidized was appreciably less, oxidation in this case being stimu- lated by various substances. The most beneficial influence was exerted by the addition of a small amount of soil: this may be due to the introduction, with the soil, of a small amount of the lacking calcium salt or of some vitamine-like substance. The favorable action of the organic substances, aluminum and manganese sulfates, may be of a stimulating nature; however, this beneficial action is only very small and lies within the range of natural variability of the organism. Influence of temperature. The optimum temperature for the activitiesof Thiobacillus thiooxidansn.sp. lies at about 28° to 30°C. Growth and sulfur oxidation are much slower at lower tempera- tures (18°) and at 37°C. Temperatures of 55°-60°C. are sufficient to kill the organism. THE NATURE OF ACID FORMED AND THE INFLUENCE OF REACTION UPON THE GROW^TH OF THIOBACILLUS THIOOXIDANS N. SP. To get an insight into the true nature of the acid formed, particularly in the presence of tri-calcium phosphate, a series of tubes containing 2 cc. portions of the culture were arranged; measured quantities of j NaOH were added to these, then the volume of the liquid was brought, in all tubes, to 3 cc. by the addition of distilled water. The hydrogen ion concentra- tion of the tubes was then determined, by the colorimetric method. The results are tabulated in table 3 and graphically presented in figure 2. 248 SELMAN A. WAKSMAN AND J. S. JOFFE TABLE 3 Tilration and hydrogen^on concentration of a H day old culture of Thiobacillus thiooxidans n. sp. pH VALUES ^NaOH No calcium phosphate in the original culture 0 . 25 per cent of tri-calcium phosphate originally present in the culture cc. 0 1.5 1.5 0.02 1.6 1.5 0.04 1.7 1.6 0.06 1.7 1.7 0.08 1.7 1.7 0.10 1.8 1.7 0.12 1.8 1.8 0.14 1.9 1.8 0.16 2.0 1.9 0.18 2.0 1.95 0.20 2.2 2.0 0.22 2.3 2.2 0.24 2.4 2.2 0.26 2.5 2.3 0.28 2.6 2.4 0.30 2.8 3.0 0.32 4.4 3.6 0.34 6.4 4.4 0.36 6.4 5:6 0.38 6.4 6.2 0.40 6.4 6.6 0.42 6.4 6.6 0.44 6.6 6.6 0.46 6.6 6.6 0.48 7.4 6.6 0.50 7.5 6.6 0.62 9.0 7.2 0.54 9.4 7.2 0.66 8.8 0.58 9.6 OXIDATION OF SULFUR I\ TIIK SOIL 249 It will 1)0 observed, by {ilanciiit? at tiie curves in figure 2, that the hydrogeii-ioii eoncentnition slowly decreases, as manifested by a slow increase in the ]Ai values, with the addition of alkali, till the pH reaches 2.S, then there is a sudden drop in the curve, to i)H ().4, wlien the curve af;ain l)ecomes slanting, followed by a second droj). This o'ves the buffer effect of the cultures; the buffer action is more pronounced in the i)resence of tri-calciuin ])hospha1e, which increases the i)h()sj)hate content of the inediuin. 3' i- 6- 7- g '/f /Vrcenf- of f-r/ei/Zruj/v /lias/'yia^f II or/n/na/ cu/f^or^ Amount of a/ua/i ir!~c ^ Fi(i. 2. TiTKATiox CuRVKS OF THE CULTURE OP Thibacillus Ihiooxiilaiis N. SP. The sulfur is oxidized into sulfuric acid; this acid acts upon the tri-calciuni iiliosjihate transforming it first into the di-calcium salt, tlien the mono-calcium salt, and finally into phosphoric acid, while the calcium is precipitated as calcium sulfate; further oxidation of sidfur results in the production of free sulfuric acid. As to the influence of initial reaction upon growth, we find that a reaction having a hydrogen-ion concentration equivalent to a pH of 2.U-2.S is the most favorable for the growth of the 250 SELMAN A. WAKSMAN .\ND J. S. JOFFE organism. Reactions more acid than 2.0 easily become injurious, although the organisms still continue to live at even as low a reaction as a hydrogen-ion concentration of pH = 0.6, while the medium titrates 0.8 normal acid (with phenolphthalein as indicator, using i^ NaOH: the culture being gro\\Ti on medium 1). Reactions ranging in pH from 4.0 to 6.0 are less favorable. Growth is slower to start, but once the reaction, through a slow oxidation of the sulfur, has reached a pH of about 3.0, the growth becomes more rapid. Reactions equivalent to pH 6,0 and above are unfavorable for growth. When a culture, at a pH 0.8 to 1.6 (these being the limits tested), is filtered free from any unoxidized sulfur, then stoppered and allowed to stand, the liquid is found to clear up, after a period of time, and the bacteria are agglutinated with the formation of flaky masses at the bottom of the containers. The rapidity of agglutination depends on the reaction of the culture, the more acid cultures agglutinating more rapidly than the less acid ones: at a pH = 0.8, agglutination took place in four to five days, while at pH = 1.5, it took more than two weeks for this jahenomenon to appear. It is interesting to note that this phenomenon was never observed in the unfiltered culture, i.e., in the presence of unoxidized sulfur, even if the cultures were kept at ]iH0.6 to 0.8 for a long time. Neutralizing agents. The acid formed rapidly changes the hydrogen-ion concentration of the medium and growth almost ceases. To obviate a rapid change in reaction by the acid pro- duced from the oxidation of the sulfur, neutralizing agents are to be used. These should be of suth a nature as not to make the medium alkaline or tend to change the reaction rapidly: this eliminates, therefore, the use of carbonates and soluble oxides, like CaO. The best substances are buffers, like phosphates, but to keep the reaction above a very high acidity by means of soluble phosphates, high concentrations have to be used, which will exert an unfavorable physical effect upon the organism. Solid salts, insoluble in water which, on dissolving by the action of the acid, will give a soluble substance and an insoluble residue, are best for this purpose. CaCOa and INIgCOs can be used, but OXIDATION OV SULl rii IX THIO SOIL 251 these go rapidly into solution by the action of (iio acid, thus tencUns to change the reaction towards alkaUno. Cas (P04)2 offers the best material for the jiurjwse, because, on dissolving, it gives an acid salt and an insoluble residue (CaSO*2H20). Mechanism of sulfur oxidation. The sulfur is oxidized, accord- ing to the following reaction: So + 30, + 2H2O = 2H..SO4 + 282 Cal. (1) In the presence of tri-calcium ])hosphate: Cas (PO^), + 2H,S04 = Ca (H,P04)2 + 2 CaS04 (2) Ca (HoPOO^ + H,S04 = 2 H3PO., + CaS04. (3) The energy liberated in the oxidation of sulfur is used bj^ the organism for its activities. The acid formed interacts with the neutralizing agents of the medium, giving first mono-cal- cium iihosi)hate, at a pH of about 2.8-3.0, then phosphoric acid. So that, at a condition of equilibrium, we have a mix- tuie of phosphoric and sulfuric acids, and the calcium salts of these acids, the condition of eciuililtriuni depending on the stage of oxidation. Taxonomic considerations. The first paper of this series con- tains a study of the five groups of sulfur bacteria. The organism descri])ed in this pa]ier, ThiobaciUus thiooxidnns n. sp., is placed in a fifth group, which includes members morphologically related to the members of the fourth group, but which are distinctly different physiologically. Group four includes colorless sulfur bacteria which do not accunuilate sulfur within their cells, but which produce an abimdance of sulfur ( from US and thiosulfates) outside of their cells. This groiqi of bacteria is the one closely related to the organism studied in this jniper and will, therefore, be discussed in greater detail. Ciroup four is represented by two bacteria, ThiobaciUus thioparus (Nathanson) Beijerinck and ThiobaciUus denilrificans Beijerinck. Group five, which is so far represented only by ThiobaciUus Ihiuoxidans n. sp., will include colorless sulfur-oxidizing bacteria which do not accumulate sulfur either Fig. 3. Thiobacillus Ihidoxidans, n.sp. (X 1500). Cdlturk Grown i\ Ixougamc Medium Stained with aqueous-alcoholic solution of gentian violet OXIDATION OF SULFUR IN THE SOIL 253 within or without their cells, which are very small in size (a micron or less in length and 0.5 micron in diameter) and which oxidize sulfur rapidly to sulfuric acid with a very acid reaction. ThiobaciUus thioparus was demonstrated by Nathanson in sea water and by Beijerinck in canal water. It was isolated on a medium contahiing sodium thiosulfate as a source of sulfur, in addition to minerals and ammonium chloride (0.01 per cent) and sodium bicarbonate as a source of carbon. In two to three days, the surface of the medium became covered with free sulfur filled with bacteria. This organism is 3 by 0.5^, not forming any spores, is very motile and very sensitive, dying out on the plate in a week. The thiosulfates can be replaced by CaS, H2S and elementary sulfur. Jacobsen dissolved the sulfur in sodium sulfide and precipi- tated with dilute hydrochloric acid, washed and dried, then added to a medium containing 100 parts of water, 0.05 K2HPO4, O.O5NH4CI, 0.02 MgClo, 2 of CaCOs or MgCOs and a trace of FeCls (3 parts of NaCl were used in the case of the organism iso- lated from sea water). The cultures were incubated at 30°C. The organism was found to form a film on the surface of the culture, with sulfur granules surrounding the cells; at the end, instead of sulfur, only a slimy bacterial mass was found to remain. Traces of hydrogen sulfide were always found. Pure cultures of the organism were obtained on agar plates, using 0.5 per cent of sodium thiosulfate and some CaCOs. The carbon dioxide is obtained from carbonates, no growth being obtained and no sulfuric acid produced in the absence of carbonates. The or- ganism is autotrophic since it does not require any organic matter for its development ; it is sometimes motile and sometimes non- motile. ThiobaciUus denitrificans was isolated by Beijerinck by adding to 100 parts of canal water, 10 parts of powdered sulfur, 0.05 KNO3, 0.02Na2CO3, 2CaC03, O.O2K2HPO4 and 0.01 part of MgCl^, and incubating the medium at 30°C. The sulfur was oxidized and growth was accompanied by a reduction of the nitrate to atmospheric nitrogen. The organism was isolated on agar plates and was found to be a motile, short rod, hardly distinguishable 254 SELMAN A. WAKSMAK AND J. S. JOFFE morphologically from the Thiobacillus thioparus. Both organisms use carbonates and bicarbonates as sources of carbon and rapidly lose, on the plate, their ability to grow. The following table gives the salient features of organisms belonging to groups 4 and 5. Autotrophy. Thiobacillus thiooxidans belongs to the autotro- phic bacteria which derive their energy from inorganic substances, TABLE 4 Salient features of sulfur oxidizing bacteria, not accumulating sulfur tinthin their cells Energy Size Accumulation of sulfur outside the cell Pellicle formation Carbon sources Aerobism Growth on agar media. Motility Acid accumulation th. thioparus (nathanson) beijbbince HjS, thiosul- fate, sulfur 3 by 0.5 u + + + + Carbonates, bicarbonates Aerobic + + Active TH. DENITRrFICANS BEIJERINCK H2S, thiosul- fate, sulfur 3 by 0.5 M + + + + Carbonates, bicarbonates Anaerobic + + ? TH. THIOOXIDANS N. 8P. Sulfur, thio- sulfate 1 by 0.5 /I None None COj from at- mosphere Aerobic ± Very strong, pH goes down to 0.6 and its carbon from the CO2 of the atmosphere. This bacterium, which can derive its carbon from the CO2 of the atmosphere, its energy from inorganic sulfur, its nitrogen from ammonium sulfate and other inorganic salts and whose mineral need is very small, was probably among the very first to start life on our planet. The sulfuric acid formed interacted with the in- soluble silicates, phosphates, carbonates, etc., thus helping to break down the original rock and allowing the hfe of other or- ganisms to follow. This organism or, perhaps group of organisms, together with the nitrifying bacteria may thus have formed the initial step in the organic world, manufacturing organic materials for other forms of Ufe to follow. OXIDATION OF SULFUR IN THE SOIL 255 SUMMARY 1. Thiobadllus thiooxidans n. sp. was isolated from composts of soil, sulfur and rock phosphate, by the use of inorganic media. 2. It oxidizes elementary sulfur to sulfuric acid, derives the necessary carbon from the CO2 of the atmosphere and its nitro- gen need from inorganic nitrogen salts. 3. It is responsible for the oxidation of sulfur in the soil and when soil is composted with sulfur or with sulfur and rock phos- phate. 4. The sulfuric acid produced from the oxidation of sulfur by ThiobacUlus thiooxidans n. sp. transforms tri-calcium ])hosphate into soluble phosphates and finally into phosphoric acid. 5. ThiobacUlus thiooxidans n. sp. produces more acid, from oxidation of sulfur, and continues to live in a more acid medium, than any other living organism yet reported, the hydrogen-ion concentration of the medium increasing to a pH 0.6 and less. REFERENCES Beijerinck, M. W. 1904 Dber die Bakterien, welche sich im Dunkeln mit Kohlensfiure als Kohlenstoffquelle erniihren konnen. Centralbl. f. Bakt., 2 Abt., 11,593-599. CoHN, F. 1875 Untersuchungen iiber Bakterien, II, Beitr. Biol. Pflanz., 1, H. 3, 141. CoRSiNi, A. 1895 Uber die sogenannten "Schwefelkomchen" die man bei der Familie der Beggiatoaceae antrifft. Centralbl. f . Bakt., 2 Abt, 14, 272- 289. Crauer, In Muller, C. 1870 Chemischphysikalische Beschreibung der Ther- men von Boden in der Schweiz. DtJGGELi, M. 1917 Die Schwefelbakterien und ihre Tatigkeit in der Natur. Natur wiss. Wochschr., 16, No. 24, 321-328. DtJOGELi, M. 1919 Die Schwefelbakterien. Neujarsblatt d. naturf. Gesell. Zurich. No. 121, 43p. Gehrino, a. 1915 Beitriige zur Kenntnis der Physiologic und Verbreitung denitrifizierender Thiosulfat-Bakterien. Centralbl. f. Bakt., 2 Abt., 42,402^38. GiCKLEHORN, J. 1920 t)ber neue farblose Schwefelbakterien. Centralbl. f. Bakt., 2 Abt., 50,415^27. HiNZE, G. 1903 Thiophysa volutans, ein neues Schwefelbakterium. Ber. deut. bot. Gesell, 21, 309-316. HiNZE, G. 1913 Beitrage zur Kenntniss derfarblosen Schwefelbakterien. Ber. deut. bot. Gesell., 31, 18&-202. 256 SELMAN A. WAKSMAN AND J. S. JOFFE Hetippe, F. 1905 tlber Assimilation der Kohlensaure durch chlorophyllfreie Organismen. Arch. Anat. u. Physiol., Phj'siol. Abstr., 33. Jacobsen, H. C. 1912 Die Oxydation von elementarem Schwefel durch Bak- terien. Folia microb., 1, 487-196. Jacobsen, H.C. 1914 Die Oxydation von Schwefelwasserstoff durch Bakterien. Folia microb., 3, 155-162. Jegunow, M. 1896 Bakterien-Gesellschaften. Centralbl. f. Bakt., 2 Abt., 2, 11-21, 441-449, 487-482, 739-752. Jegunow, M. 1898 Flatten der roten und der Schwefelbakterien. Centralbl. f . Bakt., 2 Abt., 4, 257. Kbtl, F. 1912 Beitriige zur Physiologie der farblosen Schwefelbakterien. Beitr. Biol. Pflanz., II, 335-302. KoLKWiTZ, R. 1912 tlhei die Schwefelbakterien Thioploca ingrica Wislouch. Ber. deut. bot. Gesell., 30, 662-666. Krttse, W. 1910 Allgemeine MLkrobiologie. LiESKE, R. 1912 Untersuchungen liber die Physiologie denitrifizierender Schwefelbakterien. Ber. deut. bot. Gesell., 30, 12-22. LiPMAN, J. G., Waksman, S. a., and Joffe, J. S. 1921 The oxidation of sulfur by soil microorganisms. Soil Sci., 12, 475-189. LocKETT, M. T. 1914 Oxidation of thiosulfate by certain bacteria in pure culture. Proc. Roy. Soc, 87, 441. McLean, H. C. 1918 The oxidation of sulfur by microorganisms in its relation to the availability of phosphates. Soil Sci., 5, 251-290. Meyebhof, O. 1916 Untersuchungen iiber den Atmungsvorgang nitrifizieren- der Bakterien. Arch. Ges. Physiol., 164, 353-127 ; 165, 229-284. MoLiscH, H. 1907 Die Purpurbakterien nach neuen Untersuchungen. Jena. MoLiscH, H. 1912 Neue farblose Schwefelbakterien. Centralbl. f. Bakt., 2 Abt., 33, 55-62. Nadson, G. a. 1903 On the sulfur microorganisms in the Gulf of Hapsala. Bull. jard. Bot. St. Petersburg., 13, 102-112. Nathanson, a. 1903 Uber eine neue Gruppe von farblosen Schwefelbakterien und ihren Stoffwechsel. Mitt. zool. Station, Neapel. 15, 655. Omelianski, W. 1913 Der Kreislauf des Schwefels. Lafar's Handb. d. techn. Mykol., 3, 214-244. Omelianski, W. 1905 Uber eine neue Art farbloser Thiospirillen. Centralbl. f . Bakt., 2 .\bt., 14, 769-772. Sawjalow, W. 1913 Uber die Schwefelwasserstoffgiirung im Schwarzen Heil- schlamm. Centralbl. f. Bakt., 2 Abt., 39, 440-tl7. Waksman, S. A., and Joffe, J. S. 1921 The oxidation of sulfur by micro- organisms. Proc. Soc. Exp. Biol. Med., 18, 1-3. Waksman, S. A., and Joffe, J. S. 1921 .\cid production by a new sulfur oxi- dizing bacterium. Science, N. S., 63, 216. West, G. S., and Griffith, B. M. 1913 The lime-sulfur bacteria of the genus Hillomia. Ann. Bot., 27, 83-91. WiNooRADSKT, S. 1887 Uber Schwefelbakterien. Bot. Ztg., 45, No. 31-37' WiNooRADSKY, S. 1888 Beitriige zur Morphologie und Physiologie der Bak- terien. I. Schwefelbakterien. Leipzig. 1-120. Wislouch, S. M. 1912 Thioploca ingrica nov. spez. Ber. deut. bot. Gesell., 30, 470-473. THE PRODUCTION OF PINK SAUERKRAUT BY YEASTS' E. B. FRED AND W. H. PETERSON From the Departments of Agrictdtural Bacteriology and Agricultural Chemistry, University of Wiscortsin, Madison Received for publication June 25, 1921 Sauerkraut, or sour cabbage as it is sometimes called, is ob- tained by the acid fermentation of cabbage. The process of fermentation and manufacture is simple and the resulting pro- duct is greatly relished by many people. Some idea of the im- portance of this method of preserving cabbage may be seen from a glance at the sauerkraut industry in Wisconsin. In this state alone more than 36,000,000 pounds of sauerkraut are manu- factured annually, in addition to that prepared in small quanti- ties in innumerable households. Normal sauerkraut has a distinctly acid reaction and a faint pleasant aroma. The shredded cabbage after it has turned into kraut loses some of its toughness, but should still retain a com- paratively firm texture ; the white color tends to lose its opaque- ness and the cabbage becomes sUghtly translucent. Kraut with a strong odor and soft texture is of poor quaUty. The preservation of cabbage in the form of sauerkraut is generally a result of natural fermentation. Clean white cabbage is cut into shreds, salt is added, and the entire mass packed into a vat and heavy weights are placed on top. In a few hours fermen- tation begins and the sugars of the cabbage are rapidly converted into lactic acid, acetic acid, alcohol, and small amounts of other products. .Although many kinds of microorganisms may be found in the juice of the kraut, the lactic acid bacteria are the most important. • Published with the permission of the Director of the Wisconsin Agricultural Experiment Station, 257 JODBNAI, OF BACTEBIOLOOT, VOL. VIl, NO. 2 258 E. B. FRED AND W. H. PETERSON In recent years, some of the canning companies have found it difficult always to secure a sauerkraut of good flavor, texture, and color. One of the difficulties encoimtered, has been the occurrence of a sauerkraut with a well defined pink or salmon- pink color. Although not unfit for food, this pink colored sauer- kraut is an undesirable product as it must be sold at a price lower than that obtained for white kraut. From a review of the literature, it seems probable that micro- organisms are involved in the formation of the pigment. In 1904 Butjagin, and a year later, Wehmer called attention to the occurrence of pink producing organisms in sauerkraut. Henne- berg in 1916 reported that the addition of 1.2 per cent of lactic acid to cabbage resulted in the production of a reddish colored kraut. The presence of large numbers of yeasts in sauerkraut, and in a few cases of pink yeasts, has been reported by various investigators. The distribution of these pink yeasts in nature and the factors that influence pigment formation have been the source of much study. Grosbusch (1915) isolated from apples a colorless torula which under certain conditions formed a deep red pigment. Some of the conditions which he found favorable for pigment production, were low sugar content, certain kinds of sugars, and a slightly acid reaction. Beijerinck (1919) decribed a yeast producing a colorless substance which became a deep red pigment in the presence of oxygen and iron salts. A complete discussion of the literature of the pink yeasts vail be found in the papers of Will (1907; 1912), Pringsheim and Bilewsky (1911). In view of the occurrence in practise of the colored sauerkraut it becomes a matter of some importance to discover the cause, and if possible, the remedy for this undesirable t}T)e of fermentation. EXPERIMENTAL Large samples of normal sauerkraut and pink sauerkraut were secured from one of the canning companies and analyzed. This kraut was six weeks old and judged proper for canning. The comparative chemical and bacteriological analyses are given below: PRODUCTION OF PINK SAUERKRAUT BY YEASTS 259 Analysis of sauerkraut Chemical: 1 . Per cent of water in kraut 2. Total titratrable acid in 100 cc. juice. 3. Volatile acid as acetic in 100 cc. juice 4. Non-volatile acid as lactic in 100 cc. juice 5. Alcohol as ethyl Bacteriological: Number of microorganisms in 1 cc. juice.. 90.6 140.0 cc. O.lN 0.247 gram 1.026 grams 0.727 gram 3,600,000 88.0 147.5 cc. 0.1>f 0.255 gram 1.426 grams 0.978 gram 91,000,000 Chemical analj^sis failed to show any striking difference in composition between the normal and the pink sauerkraut. Later analyses of other samples indicate that the figures given above are fairly representative of the two kinds of sauerkraut. The results of the bacteriological analyses are of more significance and show that the colored kraut is much richer in microorganisms. By means of direct microscopic mounts from the normal kraut it was found that there was a preponderance of rod-formed bacteria while the pink kraut contained yeast cells almost exclusively. The enormous number of yeast cells in the juice of pink kraut suggested that these organisms might be the cause of the pink pigment. Prompted by the fact that yeasts are usually present in high numbers in pink sauerkraut a great num- ber of dilution plates were poured. Almost without exception the colonies on these plates consisted of yeasts^ but rarely was any pigment developed. Many samples of pink sauerkraut were plated and from well isolated colonies transfers were made to glucose yeast-water agar slants. In general these yeasts from pink kraut showed little if any color on agar slants. A few transfers gave a pale pink color. From a large number of cultures three pink colored colonies were selected for further study, numbered 24-1, 85-1, and 95-6. These strains show a difference in color; 24.1 and ' The term yeasts in this paper is used to designate the true saccharomycetes and also those which do not form ascospores, the lorulae. It is probable from the results of previous workers that these pink yeasts are properly termed torulac. 260 E. B. FRED AND W. H. PETERSON 95.6 are a salmon shade of pink and 85.1 is a magenta shade. In form these organisms vary from round to oval, but elongated cells were the most abundant. Factors that inflitence color Oxygen. The influence of oxygen on the production of pig- ment by these yeasts was determined by growing the cultures in the air, and in a desiccator where only a limited supply of air was available. In the presence of air the growth and pigment production was good while in its absence a fair growth and no color was obtained. When these colorless tubes were exposed to the atmosphere a sahnon-pink color developed rapidly usually within one to two days. The influence of oxygen was also noted when kraut was exposed to the air, e.g., pale pink kraut turned a deeper color after a few minutes exposure. Reaction. The reaction of the medium for cultivating these organisms was varied between the pH values of 6.5 and 7.2. Pigment was formed in all cases and approximately to the same extent. At a reaction of 5.5 the growth was not so rapid, but the color was somewhat more brilUant. Temperature. The influence of this agent on color formation by these yeasts was studied at 18°, 22°, 28°, and 37°C. Only a very scanty growth was noted at 37°C, a profuse growth at 22° and 28°C, and a fair growth at 18°C. The deepest pigment was found in the tube cultures at 22°C. and next in intensity of color, at 28°C. Although the higher temperature of 28°C. re- sulted in a profuse growth the pink color was not nearly so noticeable as at lower temperatures. Apparently about 22°C. or 71°F. gave the most marked color. Iron and manganese salts. The influence of iron citrate, iron ammonium citrate, iron lactate, iron sulphates, and man- ganese sulphate, on pigment formation was studied. A decided difference in behavior of the various strains of yeast towards the iron and manganese compounds was noted. Strain 85-1 re- sponded far more to the iron salts than either strain 24-1, or 95-6. Of the various compounds, iron lactate proved the best stimulant forpigment production. The other compounds, e.g., PBODUCTION OF PINK SAUERKRAUT BY YEASTS 261 iron citrate and iron ammonium citrate also favored pigment production. The manganese salts apparently promoted growth but did not bring about an increase in the pigment formation. Sugars. In order to overcome as much as possible the break- ing down of the sugars from the high heat, concentrated solutions of sugars in water were sterilized and added to the culture me- dium when cool. The yeast medium contained the following constituents : Ammonium sulphate 3.0 grams Asparagin 1.5 grams Dibasic potassium phosphate 2.0 grams Calcium chloride 0.25 gram Magnesium sulphate 0.25 gram Sugar 20.0 grams Agar 15 0 grams Water 1000.0 cc. Threa sugars were studied; xylose, glucose and maltose in 2 per cent solutions. Each strain of yeast was grown in triplicate on agar slants, containing each sugar, and these cultures were incubated at 20° to 22°C. At regular intervals, usually of 1 week each, these cultures were examined for rate of growth and pigment production. All three strains showed by far the most rapid growth in the glucose medium, with maltose next in order, and xylose last. The decided difference between growth and pigment formation is brought out in a striking manner from the results of this test. Without exception the easily fermentable sugar, glucose, gave a profuse growth and only a trace of pigment. Somewhat similar results were obtained with maltose although a pale pink was noted. In the presence of xylose these yeasts grew slowly but produced a decided pink color. From these results it is clear that the sugar, xylose, which is fermented only with difficulty is the best one of the three for pigment production. After thirty days the glucose cultures entirely lost their color while the other cultures became a deeper pink. Our results agree with those of Grosbusch, who foimd that the best color was obtained in the case of the non-fermentable sugars, arabinose and raflRinose. 262 E. B. FRED AND W. H. PETERSON P5 "e < X I ■< o £ O 1 B a o O 3 o a o >- o O £ B5 0 o H 3 ^ ^ .9 .9 a a -*:> -*J la a c3 C5 J<1 J c3 0 0 OJ 3 3 o 3 3 3 005 000 d Ah S b t4 (« Ph di Ph .^ c a ^ J:( Jd to M M a c g •a .9 .9 CL, Ph PL, Ch Ph PL, 0 0 m m m m m 3 3 3 3 3 00 — 000 Ph FL, fc. Ph PL, Ph _«J 3 & 3 a Q. cS ^ ^ (U ^ ^ =5 .S.S-3 .9 .9 -C fin FL, Ph p^ Ph pa >>!>> <^ ^ -♦J 03 C3 C3 000 _bC M CO M m m CO ;i>>>>> a 3 fl V* U t4 C3 c3 C3 1-H u 0 0 C3 ,,1, mxnm fo P^ fe i 00 Cl) T ■^ 3 U ^ b k< 000 000 « 000 ^ ■^■^ 000 q ^:z;z Ph PL, Ph 0 0 0 >>>>>. cJ 03 t3 3 3 3 n:? -a -a 0 0 ^H *tH M t» 02 <5 <5 «5 IS 0 a 3 0 IN ■( .9 "3 -a ■3.9.9 Ph P-, O, Ph Ph PL, >.>>>> C 3 3 Lri (m Im « 0 w 02 GO t» t^ tx, Ezi k, U. tH Li t.< Im 000 000 000 000 0 « t> 000 000 000 zz^ ^ z :? , >> >> >^ >.>!>> 3 d e 3 3 3 t 03 ei o3 d c3 o3 0 0 « V u u CO 02 M W 03 02 3 3 u ^ ^ -M □ 3 .9 a a a ^ a 0 J.! ^ 0 .9 -3 -a .9.9-3 Pi PL, PIh ft, Ph PL, >.>>>. 3 3 3 C3 C3 C3 Lh h U 02 02 M [i, fe fe 8 0 ^ 0 • I PRODUCTION OF PINK SAUERKRAUT BY YEASTS 263 Sodium chloride. The influence of salt on pigment production naturally suggested itself since it has been reported by many experienced makers of sauerkraut, that high concentrations of salt caused the reddening of the kraut. Sodium chloride in concentration of 2 and 4 per cent was added to triplicate tubes of the glucose, maltose and xylose agar media. These tubes were inoculated with each yeast. 2.5 per cent salt is the approximate amount used in the manufacture of commercial kraut. In table 1 are given a summary of the results of these tests. The cultures plus salt developed much more slowly at first than those without salt, but after two weeks this retardation was not so noticeable. In the glucose series, culture 24-1, no salt, the pink color began to fade to a pale pink at the end of two weeks, while in the presence of salt a brilliant pink pigment persisted. . These cultures were kept for 6 weeks but without any loss of color, save in the no-salt group. Instead of a loss, the older cultures of the salt group showed a greater amount of pigment and a deeper color. Somewhat similar results were secured with the other strains of yeasts although the intensifying effect of salt was not so noticeable. It seems safe to conclude that sodium chloride even in large amounts exerts a favorable effect on pigment production. This favorable influence in not notice- able until the cultures are several weeks old. PRODUCTION OF PINK SAUERKRAUT BY INOCULATION WITH YEASTS In large glass percolators of 2 Uter capacity 1000 grams of cut cabbage were packed. The outside leaves and core of the raw cabbage were removed, it was cut on a small shredder and salt was added. Part of the cabbage was inoculated with cul- tures of the yeast isolated from pink kraut. The entire mass was packed into the percolator and weighted do^Ti with a bottle of sand or mercury, that weighed 1 kilo and which fitted closely in the top of the percolator. The lower end of the percolator was fitted with glass wool, and below this was inserted a one hole rubber stopper fitted with a glass tube. Through this glass tube, which was sealed at one end with a rubber tube and screw clamp. 264 E. B. FRED AND W. H. PETERSON samples of the fermenting liquid were removed from time to time and the total titratable acidity measured. For comparison, samples were also taken from the top of the percolators and titrated. No very decided difference was noted in titration figures from the different parts of the same container. Since it is highly important to note the color changes during fermentation, percolators were found especially suitable for this study. The plan of this experiment follows: 1. 2 per cent salt no inoculation, Control 2. 2 per cent salt no inoculation, Control 3. 2 per cent salt inoculated with yeast 24-1 4. 4 per cent salt inoculated with yeast 24-1 5. 2 per cent salt inoculated with yeast S5-1 After four days the cabbage in percolators, numbers 3, 4, and 5 showed numerous pink colored spots and two days later the the entire mass became pink throughout. The pink pigment of numbers 3 and 4 especially increased with age, until at the end of two weeks the plant tissue was a deep red or a purplish red color. This increase in color as the kraut aged, was not noted in the cabbage inoculated with 85-1. When four weeks old the kraut was removed and examined. Numbers 1 and 2 were sour, rather tough and judged of fair quahty. Direct microscopic mounts showed a preponderance of bacteria. Numbers 3 and 4 were bitter with an unpleasant flavor. Microscopic mounts showed an almost equal number of yeasts and bacteria. Total acid determinations, made at two-day intervals failed to bring out any very striking differences; in general the containers with yeasts showed less acid at the end of the experiment than the untreated controls. The total titratable acidity at the time of opening is given below: PRODUCTION OF PINK SAUERKRAUT BY YEASTS 265 O.JNaeU in too ee. «f juic* ee. 1. Control 218.0 2. Control 162.0 3. Inocvilated with 24-1 73.0 4. Inoculated with 24-1 144.0 5. Inoculated with85-l 179.0 EFFECT OF SODIUM CHLORIDE AND OF TEMPERATURE ON THE PRODUCTION OF PINK SAUERKRAUT Attention has been called to the statement of sauerkraut manufacturers that salt favors the pink color. liaboratory tests with pure cultures of pink yeasts also indicate the favorable TABLE 3 The influence of sodium chloride and iempcralure on the production o/ pink sauerkraut AQE OF SODirM CHLO- RIDE TEMPEBATDBE IN CENTIOH.VDB 28° 21° 16° dnys per cent 7 2.5 A trace of pink No pink color No pink color 7 3.5 A trace of pink No pink color No pink color 7 4.5 Pink at bottom No pink color No pink color 7 5.5 Brilliant pink throughout No pink color No pink color 10 2.5 Spots of pale pink No pink color No pink color 10 3 5 More pink than above No pink color No pink color 10 4.5 Pink at bottom No pink color No pink color 10 5.5 Brilliant pink throughout No pink color No pink color 20 2.5 Pale pink No pink color No pink color 20 3.5 Pink Pale pink No pink color 20 4.5 Brilliant pink throughout Pink spots No pink color 20 5.5 Brilliant pink throughout Pink throughout Pink spots i nfluence of this chemical on color production. To see if the raw- cabbage without inoculation, but high in salt, would undergo a fermentation which produced a pink color, 12 large percolators of cabbage were treated as outlined in the table below and incu- bated at three temperatures, 16°, 21°, and 28°C. The procedure 266 E. B. FRED AND W. H. PETERSON was the same as in the previous experiment. Table 2 gives the general plan and results of this test. The data recorded in the table show: first, the close relation between amount of sodium chloride and formation of pink sauerkraut; second, the decided influence of temperature on color. These two points are so clearly brought out in the table that a detailed explanation is not needed. The findings are in accord with the observations of kraut manufacturers ; namely, that high salt concentration may cause pink sauerkraut. TABLE 3 Effect of sodium chloride on the ■production of acids by lactic acid bacteria; calculated for 100 cc. of culture NUUBER MEDIUM TIME APTEB SODIUM CHLORIDE O.lN ACID days grama CC. 1 Cabbage juice 3 None 70.4 2 Cabbage juice 3 1 55.6 3 Cabbage juice 3 2 49.8 4 Cabbage juice 3 3 32.2 5 Cabbage juice 3 4 1.8 6 Cabbage juice 3 5 1.2 7 Cabbage juice 3 6 0.0 An explanation for this condition may be found from a study of pure cultures of yeasts and bacteria common to sauerkraut. Yeasts can live in the presence of high concentrations of salt while the bacteria commonly associated with kraut are not so resistant. Support for this statement is given in the figures of table 3. Here the influence of varying amounts of sodium chloride on a pure culture of lactic acid bacteria was studied. This lactic acid organism was isolated from normal sauerkraut. About 4 per cent of salt is the critical concentration for this organism. Orla-Jensen (1919) has carried out an extensive study of the influence of sodium chloride on acid production by various types of lactic acid bacteria. He found that the conmion forms of lactic acid bacteria, the Bacterium ladis-acidi, Lactobacillus bidgaricus, and other tjiies are slightly retarded by 2.5 per cent of salt. On the other hand, certain of the rod forms of lactic PRODUCTION OF PINK SAUERKRAUT BY YEASTS 267 acid bacteria isolated from plant tissue are not retarded in the least by this concentration of salt. In the presence of larger amounts of salt (5.5 per cent), all of the rod forms of lactic bacteria are injured. A concentration of not more than 2 to 4 per cent of salt would not seriously retard the growth of the yeast and would favor its production of pigment. EFFECT OF HIGH ACID FORMING BACTERIA ON THE PRODUCTION OF PINK SAUERKRAUT The use of bacteria which would ferment the sugars of the cabbage juice rapidly naturally suggests itself as a means of combating the growth of pink yeasts. If great numbers of acid producing bacteria are seeded in the raw cabbage then it is pos- sible that they will so dominate the fermentation that the wild yeasts will not be able to exert any well defined effect on the kraut. The fallacy of this assumption, however, is clearly seen from the results of table 4. As in the preceding tests large per- colators with 1000 grams each of raw cabbage were used. The data include the results of four separate tests carried out at different times. In every case before the end of the fermentation period the cabbage to which acid was added or which was inocu- lated with high acid-producing bacteria turned pink. Appar- entlj^ the high acid production which followed inoculation favored the gro^\•th of pink yeasts. Direct microscopic mounts made from this pink sauerkraut furnished support for this statement "VNTien opened, the inoculated kraut showed great numbers of yeasts. Some idea of the rate of acid formation may be gained from the figures in this table. In the early stages of fermentation, without exception, the kraut seeded with aciduric bacteria showed a much greater quantity of acid than the control kraut. This difference is especially noticeable in the titration figures of the first four or five days. After eight days, conditions changed and before the end of the fermentation the control krauts were far more acid than the inoculated krauts. Culture 124-1 which is a high acid former from pentoses and other sugars is especially favorable for the development of the pink yeasts. On the other hand the non-pentose fermenter Bad. lactis-acidi which 268 E. B. FRED AND W. H. PETERSON produces acid more slowly than Culture 124-1, does not favor the growth of the pink yeasts. Perhaps the beneficial effect of inoculation on pigment production is due to the rapid destruc- tion of the easily fermentable sugars and to the change in reac- tion. This combination of conditions no doubt reduces greatly the complex of the bacterial life and thus favors the development of the wild yeasts. TABLE 4 The relation of acid forming bacteria and pink yeast to the production of pink sauerkraut TREATMENT 0. In ACID IN 100 OC. OS- JOICE NUMBER Age in days 2 cc. 28 83 23 92 111 114 25 95 84 110 23 110 39 31 4 cc. 79 136 99 120 150 150 79 131 135 121 8 cc. 195 156 165 122 168 121 145 140 125 43 130 115 96 20 cc. 192 155 218 169 152 161 197 153 151 192 114 173 118 142 Quality Appearance 1 Fair Poor Raw Fair Raw Raw Good Poor Poor Poor Fair Raw Poor Good White 2 3 4 Lactob. bulgaricus Control Culture 52-7 Pink Trace of pink Pale pink Brilliant pink Pink 5 Culture 124-1 . . 6 Culture 124-1 7 Control White 8 Culture 124-1 Pink 9 Culture 124-1 Pink 10 11 Control + acids* Culture 85-1 Deep pink White 12 13 14 Culture 85-1 + acids*.. Culture 85-1 and 124-1. Culture 24-1 and Bad. lactis acidi Deep pink Pink White * Acetic and lactic acids equivalent to 109 cc. O.In per 100 cc. of juice. The stimulating effect of acids on pigment production is further indicated by the results obtained in containers Nos. 10 and 12 where acetic and lactic acids were added to unin- oculated cabbage and to cabbage inoculated with yeasts. The quantity of acid added was approximately that found in a moder- ately ripened kraut. Pigment was observed after seven days which is about the earliest appearance of color in any of the experiments. PRODUCTION OF PINK SAUERKRAUT BY YEASTS 269 SUMMARY The pink or red color of sauerkraut is due to the growth of certain yeasts or torulae. Although these organisms are com- monly found in great numbers In kraut, they frequently fail to show any pigment. The production of pigment is not a fixed characteristic but depends upon many factors. The fol- lowing conditions influence the formation of color: kind of sugar, amount of sugar, amount of sodium chloride, reaction, temper- ature, and oxygen supply. The chief factors in the production of pink sauerkraut are high temperature, high salt concentration and high acid content. When cabbage is allowed to ferment at a temperature of 20°C. or above a pink pigment is frequently noted. Almost without exception, cabbage fermented in the presence of large amounts of sodium chloride, 3 per cent or more, showed a decided pink or red color. High acid produc- ing bacteria as well as the direct addition of acids to cabbage favor the development of pink kraut. REFERENCES Aderhold, R. 1906 Sauerkraut. Lafar's Handbuch d. Tech. Mykologie, 2, 31^-322. Beuerixck, M. W. 1919 Chromogenic yeasts. Chem. Absts., 13, 1082. BuTJAGiN, B. 1904 Vorlaufige Mitteilung iiber Sauerkrautgariing. Centralbl. f. Bakteriol., II, 11, 540. Grosbusch, T. 1915 t)ber eine farblose stark roten Farbstoff erzeugende Torula. Centralbl. f. Bakteriol., II, 42, 625-638. HEifXEBERO, W. 1916 Das Sauerkraut (Sauerkohl) Die deutsche Essigindus- trie, 20, No. 28, 194. Obla-Jensen, S. 1919 The Lactic Acid Bacteria. M^moires de 1' Academie Royale des Sciences et des Lettres de Danemark, Copenhague. Sec- tion des Sciences, 8me serie, t. V, no. 2, pp. 81-196, with 51 plates. Pringsheim, E., AXD BiLEwsKT, H. 1911 Uber Rosahefe. Beitrage z. Biologie der Pflanzen, 10, 118-132. Wehmer, C. 1905 Untersuchungen iiber Sauerkrautgarung. Centralbl. f. Bakteriol, II, 14, 682-713; 781-800. Will, H. 1907 Rosahefen. Lafar's Handbuch d. Tech. Mykologie, 4, 280-302. Will, H. 1912 Beitrage zur Kenntnis rotgefarbter niederer Pilze. Centralbl. f. Bakteriol., II, 36, 81-118. STUDIES ON THE BIOLOGY OF LACTIC ACID BACTERIA: A SUMMARY OF PERSONAL INVESTIGATIONS' COSTANTINO GORINI Bacteriological Laboratory of the Royal Superior School of Agriculture, Milan, Italy Received for publication July 30, 1921 In the following pages are summarized the most important results of investigations which I have pursued during the recent war period upon the biological characteristics of the lactic acid bacteria. ACIDO-PROTEOLYTIC PROPERTIES One of the most important characteristics of many lactic acid bacteria is that of possessing acido-proteolytic properties. After the original discovery of this property I have often asserted that this activity must take place in natural milk of acid reaction, but that it is not observed in milk which has received the addition of chalk or other substances, since these additions not only alter the composition of the milk but also so alter its adaptability for the growth of the organisms concerned that the natural functions of the latter are no more exhibited and thus the results obtained in the study are no longer apphcable directly to normal conditions such as occur in the cheese industry. The detection of this acido-proteolytic property is very simple and indeed requires no chemical manipulations. Just as the Uquefaction of gelatin indicates proteolytic properties, the casein cleaving properties of these organisms are indicated by the 'The investigations covered by this summary have been published in the transactions of the Reale Accademia dei Lincei of Rome and the transactions of the Reale Istituto Lombardo di Scienze e Lettere of Milan during the years 1914-1920. 271 272 COSTANTINO GORINI liquefaction of the casein coagulum. I have described, as organ- isms possessing acido-rennin properties, cocci from the udder of the cow, and bacilli from cheese and fermented milk beverages such as Yoghurt and Gioddu. The inception of casein break- down is dependent upon various circumstances; chief among these, and of paramount importance, is temperature. In 1897 1 demonstrated that while high temperatures are appro- priate to the process of lactose fermentation, lower temperatures are more appropriate to the cleavage of casein, and in several additional pubUcations I was able to verify my findings. Finally in 1915 I presented analytical data bearing upon the behavior of lactococci and lactobacilli at temperatures ranging between 25° and 35°C. and 15° to 20°C. Many lactic acid forming bacteria which failed to show lacto-proteolytic properties when cultivated at the higher temperatures were found to possess these properties when cultivated at the lower temperatures, i.e., such tempera- tures as prevail during the process of cheese ripening. Another condition of great importance for the furthering of the lacto-proteolytic activities is the composition of the medium. Already in 1902 I had advanced evidence as to the influence of the nature of the protein upon the activities of the organisms, indicating that while among the udder cocci some dissolved casein as well as gelatin, others only dissolved casein while again others were found that acted only on gelatin. Another factor of great importance was found to he in the quality of the milk itself, which presents great fluctuations the causes of which are to be found in the changes it undergoes pre- vious to reaching the laboratory (besides the influence exerted by the race of the animals furnishing the milk, their physiological state, their feeding regime, etc.) or in the laboratory itself. In addition, in 1915, 1 was able to show the noxious influence exerted by the peptonized constituents of the milk derived from the casein cleavage. These constituents are often to be found in miUc, especially in market milk, and are due to the great develop- ment of organisms before steriUzation. Another factor of great importance, and indeed of an essential character in the behavior of milk, is to be found in the method of BIOLOGY OF LACTIC ACID BACTERIA 273 sterilization previous to use. Milk sterilized in the autoclave acquires a brownish tinge, indicative of a change of state, which makes it unfit for the demonstration of the proteolytic activities of the lactic acid bacteria. For such a study milk should be utilized that has been sterilized at a moderate heat so that its white color has been preserved. By using this white-sterile milk I was in a position to detect casein cleaving properties in some of the lactic acid bacteria which in brown-sterile milk failed to digest this compound. Some of the so-called propionic acid bacteria behaved similarly. Undoubtedly, in this factor of appropriate sterilization is to be found the reason why many authors claim for the lactic acid bacteria negative or only neghgible casein-acido-dissolving prop- erties, and therefore contrary to my opinion have ascribed to them no role in the ripening of cheese. BACTERIAL FLORA OF THE UDDER The importance which the acid forming organisms seem to bear, both from the standpoint of hygiene and in the dairy, gave me occasion for a deep study. It is to be emphasized neverthe- less that investigations seem to point to the fact that from the hygienic and sanitary standpoint it is not only the cleanUness of the stable and of the cow that are to be considered, but also other internal and external factors. A fact of great importance in this connection is, that apparently sound cows may harbor in their udder for longer or shorter periods of time bacteria (cocci as well as bacilli) which sometimes prove beneficial and other times noxious both from the hygienic as well as from the dairying stand- point. Therefore a selection of milch cows on the basis of this udder flora has been proved useful by me and to this I have given new support by carrying out fermentation tests as a means of judging of the quaUty of the milk, especially of such milk as is used for direct consumption by children and invalids. HEAT RESISTANCE It is generally assumed that an exposure of one quarter of an hour at 60° to 80°C. is sufficient to destroy the non-spore-forming JOURNAL OF BACTERIOLOGY, VOL. VII, NO. 2 274 COSTANTINO GORim lactic bacteria. On the contrary I have discovered that the lethal temperature may reach 100°C. and sometimes still higher values. My experiments have shown that this happens through the formation of a sheath of coagulated casein in immediate contact with the acid-rennin forming organisms. These findings allowed me to explain the apparent failures in milk steriUzation, and they furnish the basis for a betterment of the processes in use for industrial milk sterilization. VISCOSITY IN YOUNG CULTURES Many lactic acid forming organisms are capable of inducing ropiness in milk only during the early phases of incubation. The property of making milk ropy, although recognized by various investigators for the lactic acid bacteria, has not generally been considered as a constant and essential character of this group of organisms. In 1912 I first described a lactic acid bacterium wliich constantly induced viscosity in milk, but only on the inception of acidification and before the inception of coagulation. This organism at first induces ropiness, but the increase in acidity of the medium and the increase in the firmness of the coagulum soon obliterate this characteristic result of the growth. When cultivated at a not too high temperature and on white- sterile milk several other lactic organisms proved themselves, in my hands, to possess this propert}^ The reason why many investigators disagree as to the constancy of this rope-producing property in many of the lactic acid bacteria is to be searched for in its transitory manifestation. SPORE-PRODUCING LACTIC ACID BACTERIA In 1904 I studied a spore-forming bacillus from cheese and described it under the name of Bacillus addificans-presamigenes- casei, on account of its acid-rennet and acido-proteolytic charac- ters and found it related to the subtilis or thyrotrix groups. In 1906 I found in silage another similar form and described and pictured it. Additional investigations have authorized me in the assertion that this type is rather widely disseminated in milk and dairy products and is therefore worthy of attentive study. BIOLOGY OF LACTIC ACID BACTE:RIA 275 APPLICATION OF LACTIC ACID BACTERIA IN THE CHEESE INDUSTRY AND IN THE PREPARATION OF ACID SILAGE In 1006 I first drew attention to the advantage to be derived from the utilization of the lactic acid bacteria in cheese manu- facture, and in 1907 I recommended their utilization in the prep- aration of silage. Although from the beginning of my investi- gations I was fortunate in the selection of the appropriate types, so that their practical application now dates back fifteen years, my study was continued upon the various types which from time to time have attracted my attention. The result of these in- vestigations is that several species, or varieties, of lactic acid bacteria can be utilized for the purpose, but that nevertheless it is not a matter of indifference, for the preservation of the charac- ter of determined cheese quality, which type has been used. Two advantages are to be derived from the utilization of the lactic acid bacteria in cheese manufacture, i.e., (1) the elimina- tion of the noxious, putrefying and gas forming organisms and (2) the furthering of the appropriate ripening and an acceleration thereof. The first of the above aims is reached by the utilization of organisms that produce a high degree of acidity, whereas it is necessary for the attainment of the second aim that the organisms utilized possess the faculty of producing the appropriate products in the process of proteolysis. I will repeat here that the applica- tion of inoculation according to my method joins with the hy- gienic production and treatment of the milk in yielding a product practically free from undesirable microorganisms. Similar considerations may be made with reference to the preparation of silage. Silage that has undergone the lactic acid fermentation is by far the best, both from the standpoint of cattle feeding as also from the standpoint of subsequent cheese manufacture. Every other silage is undesirable because, in spite of its satisfactory appearance and apparently low germ content, it is a carrier of butyric acid organisms. The production of a lactic silage is best accomplished by follow- ing the accompanying general rules: (1) The use of siloes with imperv'ious foundations; (2) a semi dried condition of the fodder; 276 COSTANTINO GORINI (3) air exclusion from the mass obtained by the \ise of deep layers of silage and of strong packing whereby the temperature of the silage is kept between 35°C. and 40°C.; and (4) inoculation with lactic acid bacteria to best insure the desired results, espe- cially when fodders are used which are not fit for a spontaneous lactic fermentation. In conclusion it pleases me to be able to emphasize that the majority of my results have been verified in recent years by various authors, i.e., Barthel, Boekhout, and De Vries, Burri, Esten, Evans, Hardings, Harrison, Hoffman, Lohnis, Orla Jensen, and others. A mi:thod for the cultivation of an^verobes» L. n. HUSIINELL Kanxns Agrividlural Experiment Slalion Reccivcil fur pulilioiition September 11, 1921 In tlio course of a studj^ of the bacteria causing spoilage in canned foods it became necessary to place large numbers of cultures under anaerobic conditions, and to devise anaerobic methods which would care for them with a minimum of ex- pense of time and material. A review of the literature showed that the most commonly recommended methods, which would be suitable for these pur- poses, were absorjitioii of oxygen liy alkaline pyrogallate and the rei)laccment of air by an inert gas, usually hydrogen, or a com- bination of the two. The chief objections to these methods are that they are difficult to apply, requiring complicated apparatus, are expensive and time consuming. We have encountered still another difficulty in that even the most carefully made containers are likely to leak air. In case of a leak, the pyrogallate solution is soon exhausted, or the hydrogen passes out very rapidly and is re- placed by air. ^^'e have found that even the all-glass Novy jar would not maintain anaerobic conditions for twenty-four hours in many cases. These jars are also of rather limited capacity and very expensive. The number of methods suggested would lead one to doubt the complete suitableness of any one for all occasions, and experience with their use merely establishes the correctness of this impression. Several years ago Dr. Wolbach suggested the method of Sellards which he had used successfullj' in his work on anaerobes. ' Contribution Number 30 from the I3;ictoriological Laboratories of the Kansas ARncultural Experiment Station. 277 278 L. D. BUSHNELL This method makes use of metallic phosphorus for the removal of oxygen. Most authors mention this method of obtaining anaerobic conditions in their review of the literature, but so far as we have been able to find, none of them except Wolbach have advocated it. Tn the introduction to his article describing this method, Sellards (1904) makes the statement that up to that time no one had recommended the use of phosphorus for anaero- bic work, although phosphorus is one of the most powerful ab- sorbing agents of oxygen. Among the recommendations for its use, this author mentions that phosphorus is very convenient, that it requires no previous preparation, that it keeps well, and that its absorbing efficiency is very easily tested. He mentions two possible diffi.culties; first, that the oxides of phosphorus formed during the absorption of oxygen might change the reaction of the medium employed; second, that the vapors of the elementary phosphorus might injure the nutrient media. He mentions not having had any difficulty from these sources. The phosphorus method has been used in this laboratory for the past three years and we wish to recommend it for the cul- tivation of anaerobic bacteria. We have been able to culture all obhgate anaerobes, which we have tried upon the surface of plates and slant agar. Obligate anaerobes also grow well in liquid media under anaerobic conditions produced by phosphorus. We have had no diffi.culty concerning the injur\' to the media. There is a slight rise in temperature within the jar as the phos- phorus burns; this, however, exerts no detrimental effect upon the cultures. One great advantage of this method over others is that the power of phosphorus to absorb oxygen is so great that a rather small amount will absorb the oxygen within the jar, and enough will remain to absorb whatever may leak in during the incubation period. Sellards recommended the presence of an excess of alkali to absorb the phosphorus acid anhydrides. We have found that water serves the same purpose almost equally well. There is some theoretical objection to burning the phos- phorus in the jar, in that elementary phosphorus is volatilized. We have had no difficulty from this source as far as the growth of bacteria is concerned, but some of the material is deposited CULTIVATION OF ANAEROBES 279 upon the surface of the plates and upon the cotton plugs, thus niakinp; the cull me dishes unpleasant to handle 'I'his has been overcome to some extent by coveriiip; (ho jjlatcs and cotton plugs with i)aper. Another source of trouble was encountered in labeling the tubes with a wax pencil. The phosphorus seemed to soften the wax of the ordinary red wax pencil, causing the labels to become blurred. Later a blue wax ]iencil was found which was not influenccHl in this way. Wo have also resortcil to the use of l)ai)('r labels held in jjlacr liy ruiihcr bands, since the excessive moisture in the jar will soon cause gummed labels to fall off. Some ililliculty was experienced in obtaining an anaerobic jar large enough to hold all (he plates or cultures which we wished to run at one (ime, ami at the same (ime, one which could l)e easily liandlcd and remain air tight under a reduced pressure of three to four inches of mercury foi' several tlays. Quart "Lightning Seal" fruit jars have been used for test tube work to some extent. In (his case, however, the sides of the jar and tubes must be protected from the burning phosphorus. This may be accomi)lished bj' strips of thin asbestos boartl. As a large anaerobic jar, we have used an aluminum pressure food cooker (figs. 1 and 2), manufacturetl by the Pressure Cooker Comi)any, of Denver, Colorado. The pressure gauge on the top is removed and this opening plugged. Just before using, the edge of the cover is heavily smeared with a rubber cement, similar to that used on stopcocks. This jar will retain a vacuum of 4 to 5 inches of mercury for as long as two weeks in the incuba- tor at 37°C. (we have not tested it for longer periods.) Water is placed in the bottom of the jar, cultures are placed on the wire rack (fig. 2), an evaporating dish containing the phosphorus is placed in the jar and covered with a bit of wire gauze having a small opening in the center. This is to prevent the burning phosphorus from siiattering. ^^'hen all is ready, one of the stopcocks in the cover is opened and the cover is held so that it may be replaced quickly A hot needle is passed through the small hole in the wire gauze, and the phosphorus is ignited. The cover is quickly replaced and screwed down firmly. The pet cock is shut off after a few seconds, and a careful watch main- 280 L. D. BXJSHNELL tainecl for leaks. As the phosphorus burns there is considerable positive pressure due to the increase in temperature The pet cock should always be left open until this is nearly neutralized, otherwise leaks may develop. If there are leaks around (he hd, they can be detected by the puffing out of phosphorus fumes. The burning usually ceases in a very few seconds, due to the exhaustion of oxygen; and the jar is filled with fumes. In glass jars these are seen to subside in about fifteen minutes if there is plenty of fresh water in the bottom. Anaerobic conditions are Fig. 1. Anaerobic Jar with Lid in Place Fig. 2. Anaerobic Jar with Lid Remo\'ed thus obtained about the cultures almost instantly. The rate at which oxygen dissolved in the medium, will diffuse out will depend upon the medium. Litmus milk and broth containing methylene blue are decolorized within a few hours. Deep agar tubes containing methylene blue are not decolorized to the bot- tom for about three days. Apparently oxygen diffuses out about as slowly as it diffuses into the media, since oxygen will diffuse into agar at the rate of about 1 cm. per dienl. This is indicated by the fact that most vigorously growing anaerobes will grow to within one or 1.5 cm. of the surface of a deep agar tube. The jar which we have described is of light weight, quickly sealed and very convenient. It may also be used with hj-drogon or illuminating gas. If alkaline pyrogallate is used it must be CULTIVATTON OF ANAEROBES 281 placed in a separate dish and not upon the bottom of the jar since these chemicals attack the metal. The capacity of the size which was used is about twenty-five plates or eighty test tubes of 15 mm. diameter. Everything considered, this is the most convenient and effective method of culturing anaerobes which we have yet encountered. One precaution which must be taken on all occasions, is not to allow bits of phosphorus to remain exposed to the air. These may not burn for several hours if moist, but as soon as they become dry they will bum and may set fire to the laboratory. The phosphorus in stock should be placed in firmly corked glass bottles and kept in the ice box. In this laboratory we have had some difficulty because the tem- perature became so high during the summer that the sticks melted together into a mass in the bottom of the container, thus making it difficult to remove. This trouble may be overcome by keeping the material in the ice box. Sellards recommends drawing out the phosphorus into small sticks. This is not at all necessary, and involves considerable trouble. REFERENCE Sellards, A. W. 1904 Some researches on anaerobic cultiirea with phospho- rus. Cent. f. Bakt. Abt. I. 37, 632-637. INFLUENCE OF VACUUM UPON GROWTH OF SOME AEROBIC SPORE-BEARING BACTERIA' L. D. BUSHNELL Kansas Agricultural Experiment Station Received for publication September 11, 1921 In the course of a study of the bacteria present in canned foods, we found that certain aerobic spore-bearing bacteria were pres- ent in jars which showed no evidence of spoilage, if they were properly sealed. These results were reported from this laboratory in 1918 (Bushnell 1918), but no attempt was made to determine the types present. Vaillard, in 1900 and 1902 examined bacteriologically, many cans of meat, and found living organisms in seventy or eighty percent of them. He believed that the bacteria survived in a dormant state in the cans from five to seven years. Among the spore-bearing types he isolated B. subtilis and B. mesentericus, (three varieties, vulgatus, niber, and fuscus). Deichsetter, in 1901, reported on the examination of preserved food provided for the Bavarian Army during a period of five years and failed to find microorganisms in canned foods, save in cans in which the food was sent in under suspicion. He con- sidered that Vaillard's findings were probably due to faulty technic. Pfuhl, in 1904, examined canned meats from five firms and found bacteria in 29 out of 106 cans. He considered that the findings of both Vaillard and Deichsetter were correct and that the difference in results was due to a difference in the care with which the foodvS were steriUzed. Very little work had been reported upon this point until 1919 when Weinzirl pubUshed the results of his findings on commercial * Gontribution Number 37 from the Bacteriological Laboratory of the Kansa.s Agricultural Experiment Station. 283 284 L. D. BUSHNELL canned foods. He states that in commercial canned foods gi%ang no evidence of spoilage, microorganisms were found in 179 out of a total of 782 cans, or in 23 per cent of the cans. The spore- bearers were practically the only organisms present, due to their superior resistance to the sterihzing process. Viable spores were found in 19.2 per cent of the non-leaking cans. Of the types of bacteria isolated, B. mesentericus predominated, with B. subtilis next. This author concludes that the living spores in commercial canned foods are unable to grow, due to the absence of oxygen, and that the vacuum is essential to the preservation of canned foods under the present method of processing. Chejmey, in 1919, reported B. mesentericus in apparently perfect cans, which were given a standard processing. Hunter and Thom, in 1919, made an examination of 530 cans of canned salmon and found 237 unsterilized. 234 of these cans contained the same organism of the B. mesentericus group, either in pure culture or in connection with other species. Only 13 showed active spoilage. From the above it is evident that the aerobic spore-bearing types predominating in unspoiled canned foods belong to the B. mesentericus and B. subtilis groups of bacteria. We may consider three reasons why these organisms predomi- nate in foods under such conditions : 1. Spores of certain types predominate on the product as it goes into the container. 2. Spores of certain types are more resistant to heat than spores of other types. 3. The spores of certain types are not all destroyed during the processing period and those remaining are able to grow under conditions as they exist in the container. From the results obtained in this laboratory, it is evident that B. mesentericus predominates, with B. subtilis second in number among the aerobic spore-bearing types. We have no idea of the number of each type upon the raw product as it went into the jars, so that it is not possible to consider this point except in so far as we may apply the work of Bruett (1919) upon the death of bacteria. She concluded that the death rate followed INFLUENCE OF VACUUM UPON GROWTH OF BACTERIA 285 the laws of monomolecular reactions. According to this law, if all spores were of equal resistance to heat, those present in the largest number would be the last to disappear. However, spores of different species of bacteria are not of equal resistance to heat, and while those of a particular species may follow this law, it cannot be appUed in a comparison of the thermal death rate of different species. We have found that the several cultures of B. mesentericus with which we have worked, are less resistant than the strains of B. subtilis. Regardless of this decreased resistance B. mesentericus predominated among the organisms isolated. Our cultures had grown for some time upon culture media and the resistance may have changed by this treatment. Weinzirl considers that B. mesentericus predominates in canned foods because of its superior heat-resisting quaUties. Lawrence and Ford (1916) state that the spores of B. sub- tilis survive steaming one and one-fourth hours in the Arnold sterilizer and autoclaving up to and including 19 pounds pressure but are usually destroyed at 20 pounds. The B. mesentericus spores survived one hour in the Arnold sterilizer and autoclaving at 19 pounds pressure, being killed by 20 pounds pressure. These statements would indicate that their cultures of B. subtilis were somewhat more resistant than those of B. mesentericus. From some previous work upon these two types we had con- sidered that B. mesentericus could grow in the absence of oxygen more readily than B. subtilis. Cheyney in 1919, also calls at- tention to this fact in his recent article on the bacteriology of canned foods. We have occasionall}' isolated B. mesentericus from the deeper layers of agar in our search for anaerobes, but we have never isolated B. subtilis under such conditions, although it has been found on the surface several times. It must be admitted that we did not make quantitative determinations of the types present in each jar. The predominance merely means that in the routine isolation of colonies more of the B. mesentericus type were isolated, although we are convinced that this organism did predominate in the jars. 286 L. D. BUSHNELL In an attempt to determine why B. mesentericus predominated, we undertook some experiments, using these two types. The types used were isolated from jars of asparagus and were rapid spore formers, although they had been grown in the laboratory for more than a year. The thermal death point of the spores of the B. subtilis culture used in these experiments was from ninety to one hundred and twenty minutes in steam at 98°C.; for B. mesentericus eighty to one hundred minutes. The time at which all are killed depends to some extent upon the numbers present. For the experimental work, the organisms were grown upon plain extract agar from four to six days. The growth was scraped from the medium and suspended in a small amount of sterile normal saline. This was shaken in a heavy walled bottle with glass beads and filtered through sterile cotton to remove clumps. The suspension was heated at 80°C. for twenty minutes to kill the vegetative cells. Experiment 1 . In this experiment we wished to determine the influence of different amounts of air upon the growth of the organisms. The number of spores indicated in the table were added to tubes of extract broth, 0.50 per cent N/1 acid to phenol- phthalein pH 5.9. In this case the column of air above the medium was about 5 cc. The tubes were exhausted to various points and sealed. The results are shown in table 1. From this table we may conclude that there was sUght growth of both organisms during fifty-one days incubation at room tem- perature in the less exhausted tubes. Experiment II was conducted in order to determine the in- fluence of varying amounts of salt and air upon the growth of B. subtilis and B. mesentericus. In this case, known amounts of air were left above the Uquid. An attempt was made to remove all possible traces of air from the medium. To do this, the tubes were partially filled with a known amount of broth to which varying amounts of salt had been added. The tubes were then drawn out to a slender neck, as close to the liquid as possible, and heated in a seamer for fifteen minutes. The tubes were next cooled in cold water, and INFLUENCE OF VACUTJM UPON GROWTH OF BACTERIA 287 1 oc. of a suspension of spores added. The spores were sus- pended in salt solutions to correspond to that in the tube, so that the salt concentration was not changed. All the tubes were filled to a mark on the constriction, with broth containing cor- responding amounts of salt. A certain per cent of this was then removed and the tubes exhausted and sealed at the mark. The tubes were then incubated at 37°C. for twenty-seven days and plates made. TABLE 1 B . 81IBT1LI8 B. ME8ENTERICD8 DATS INCUBATION Num- ber Tubes exhausted to following mm. Hg. on manometer Num- ber Tubes exhausted to follow Hg. on manomet^ ingmm. r aSded 175 350 525 685 added 175 350 525 685 5 3110 2740 2220 3000 2280 1300 1830 1730 1370 1440 5 31 25 18 43 58 13 32 31 36 30 12 3110 1170 7000 1900 1000 1300 2100 2000 1090 640 12 31 69 60 16 71 13 31 18 19 36 19 3110 2300 3500 2500 8000 1300 3200 1830 1020 120 19 31 50 41 50 30 13 60 38 13 20 41 3110 3200 4000 5000 6000 1300 4000 1990 1260 260 41 31 49 29 11 12 13 90 72 27 8 51 3110 9000 7000 3200 5000 1300 4000 2460 1120 370 51 31 90 120 12 11 13 150 71 18 11 Since each tube had a slightly different amount of broth added, it was necessary to calculate how many organisms were present in each centimeter of liquid in the beginning. These numbers are in one column and the number at the end of the incubation period in a parallel column. The volume of air varied from 0.3 cc, in the tubes containing 1 per cent air to 1.5 cc. in the tubes containing 25 per cent air, each tube being of somewhat different volume from the others. The results are shown in tables 2 and 3. These tables show the same as table 1, that there is some growth in these tubes. There is an interesting point relative to the action of increased amounts of salt. In every case, when the average is taken for all tubes in the same concentration of salt, there is an increase over that of a lower concentration. This point is somewhat more in evidence in connection with 288 L. D. BUSHNELL TABLE 2 Influence of varying amounts of salt and air upon the growth of B. subtilis in broth AIR 1 PEE CENT NaCl 2 PER CENT NaCl 4 PER CENT NaCl EXHAUSTED TO MM. Hg. Bacteria added per cubic centi- meter Bacteria determined per cubic centimeter Bacteria added per cubic centi- meter Bacteria determined per cubic centimeter Bacteria added per cubic centi- meter Bacteria determined X>er cubic centimeter 0 50 100 per cent 5 10 25 5 10 25 5 10 25 167,000 146,000 152,000 146,000 152,000 141,000 149,000 162, 000 162, 000 100,000 150, 000 117,000 210,000 160,000 180,000 111,000 220,000 220,000 149,000 157,000 151,000 151,000 157,000 149,000 151,000 143,000 146,000 190,000 200,000 240,000 102,000 110,000 110,000 160,000 240, 000 260,000 175,000 165,000 151,000 157,000 165,000 162,000 162,000 143,000 149, 000 320,000 250,000 210,000 320,000 240,000 270,000 180,000 270,000 Average 153,000 163,000 150,000 179,000 158,000 258,000 175 • 350 • 525 685 ■ 1 5 10 25 1 5 10 25 1 5 10 25 1 5 10 25 113,000 135,000 135,000 113,000 113, 000 123, 000 132, 000 123,000 113,000 113,000 118,000 123,000 123,000 141,000 135, 000 129,000 76,000 112,000 102,000 89,000 78,000 98,000 55,000 37,000 84, 000 126, 000 76,000 97,000 48,000 145,000 98,000 119,000 129,000 129,000 124, 000 104,000 114, 000 114,000 129, 000 114,000 124,000 119,000 119,000 132, 000 28,000 27,000 158,000 200,000 200,000 348,000 152,000 256, 000 240, 000 220,000 26,000 44,000 100,000 160,000 104,000 124,000 117,000 113,000 124,000 117,000 118,000 107,000 108,000 113,000 104,000 118,000 119,000 113,000 113,000 114,000 164,000 384,000 101,000 424,000 448,000 342,000 476,000 688,000 444,000 436,000 428,000 416,000 516,000 458,000 454,000 472,000 Average 123,000 90, 100 118,000 153, 000 114,000 416,000 INFLUENCE OF VACUUM UPON GROWTH OF BACTERIA 289 TABLE 3 Influence of varying amounts of salt and air upon growth of B. mesentericus in broth AIB 1 PEBCEKT NaCl 2 PER CENT NaCl 4 PERCENT NaCI EXHArffTED TO Mil. Hg. Bacteria added per cubic centi- meter Bacteria determined per cubic centimeter Bacteria added per cubic centi- meter Bacteria det«rmined per cubic centimeter Bacteria added per cubic centi- meter Bacteria determined per cubic centimeter 0 per cent 5 10 25 340,000 307,000 345,000 320,000 220,000 450,000 328,000 314,000 328, 000 280,000 310,000 260,000 323,000 323,000 340, 000 720,000 750,000 860,000 50 5 10 25 328,000 340,000 323,000 340,000 260,000 290,000 328,000 328, 000 345,000 450,000 440,000 480,000 341,000 341,000 328,000 940,000 870,000 Broken 100 5 10 25 323,000 328,000 340,000 Broken 210,000 190,000 314,000 292,000 345,000 440,000 290,000 300,000 328,000 314,000 353,000 560,000 630,000 810,000 Average 331,000 285,000 324,000 367,700 332,300 767, 500 175 1 5 10 25 15,300 14,900 17,400 29,000 24,000 22,000 15,300 15,500 17,700 14,300 35,000 36,000 36,000 60,000 17,400 15,300 15,300 14,500 23,000 25,000 45,000 24,000 350 1 5 10 25 15,200 16,000 17,700 18,100 18,000 11,000 32,000 31,000 15,900 18,500 15,700 26,000 15,000 34,000 17,700 17,400 15,900 17,700 22,200 24,000 48,000 29,000 525 1 5 10 25 16,600 17,600 18,600 16,000 26,000 26,000 18,000 27,000 15,300 14,900 16,000 16,200 27,000 32,000 27,000 22,000 17,400 15,700 16,600 24,000 34,000 29,000 686 1 5 10 25 17,700 16,000 16,100 15,000 13,000 12,000 17,400 16,000 18,600 15,700 29,000 27,000 30,000 20,000 16,000 15,700 17,400 17,400 25,000 30,000 24,000 21,000 Average 16,600 21,500 16,900 30,400 16,700 31,300 290 L. D. BUSHNELL B. mesentericus than with B. subtilis. The very small amount of air remaining in these tubes apparently has no influence upon the amount of growth. There were relatively no more organisms present in tubes merely sealed and with 25 per cent volume of air, that there were in those tubes which were exhausted and with but 1 per cent of the volume of air. We believe that the organ- isms in the tube exhausted to 685 mm. were under as completely anaerobic conditions as it is possible to obtain. According to Bitting and Bitting (1916) an ordinary tin can shows a vacuum of about four inches of mercury when exhausted at a temperature of 130°F. and tested at 85°F, We may conclude from the results obtained, that the degree of vacuum plays no part in the destruction of the spores of these two organisms. "V\1ien one per cent of salt is present, there may be a sUght decrease in the twenty-seven days of incubation. This is somewhat more marked in higher vacuum than in the lower, but the differences are not marked. In the presence of larger amounts of salt, there appears to be an actual increase in the number of viable bacteria. This is apparently not due to accidental conditions, since we have made numerous parallel determinations and find that the averages of those determined at the end of the incubation period are two or three times as high as the number added. Why there should be a decrease in the presence of 1 per cent, and an increase in the presence of 4 per cent salt we are unable to say. Of course there is considerable variation in the determinations, but the averages indicate a real increase. Experiment III. This experiment was set up parallel with that of experiment II, except that varying amounts of acid were added to the broth. The organisms were treated in the same way as those in the last experiment, except that they were suspended in acid broth after heating to kill the vegetative forms. The tubes were incubated at 37°C. for twenty-three days for B. subtilis and twenty-two days for B. mesentericus. The results are shown in tables 4 and 5. Apparently B. subtilis spores are more sensitive to acid than those of B. mesentericus. The degrees of vacuum had no influence INFLUENCE OF VACUUM UPON GROWTH OF BACTERIA 291 TABLE 4 Influence of varying amounts of acetic acid and air upon the growth of B. suhtilis in broth EXHAC8TBD TO MM. Hg. 50 100 per cent 5 10 25 5 10 25 5 10 25 Average . 175 350 525 685 5 10 25 5 10 25 5 10 25 5 10 25 PER CENTS OP N/1 ACID AKD pH 0.5 per cent pH 6.6 Added 715,000 783,000 716,000 813,000 872, 000 783,000 858, 000 813,000 783,000 802,000 Deter- mined 429,000 450,000 457,000 392,000 465,000 472,000 408,000 430,000 472,000 800,000 S5S, 000 737,000 907,000 37,000 813,000 872, 000 858, 000 813, 000 443.000 409,000 383,000 409,000 400,000 374,000 451, 000 440,000 383,000 410,000 463,000 400,000 391,000 250, 000 250,000 303,000 212,000 270,000 303,000 266,400 259, 400 273,600 276,000 258,000 Average , 409,400 per cent pH 5.00 Added 173,000 19,000 265,000 223,000 232,000 175,000 209,000 250, 000 273,000 836,000 224,000 410,000 151,000 383,000 409,000 429,000 440,000 366,000 429,000 409,000 440,000 391,000 216,000463,000 262, 300 Deter- mined 890, 000 872,000 715,000 800,000 828,000 761,000 858,000 847,000 828,000 214,000 270,000 190,000 250, 000 230,000 137,000 128, 000 178,000 210,000 230,000 220, 000 225,000 2 per cent pH 4.30 Added I'^"".- mined 241,000 205,000 256,000 234,000 247,000 198,000 236, 000 234,000 822,000 410,000 409,000 451,000 409,000 440,000 409,000 409,000 463,000 366,000 410,000 488,000 429,000 418,000 206,900424,400 800,000 890,000 828,000 &58, 000 858,000 800,000 813,000 761,000 800,000 231,000 157,000 171,000 192,000 148, 800 141,600 162, 400 146, 400 165, 600 176, 800 162, 400 158, 400 4 per cent pH 4.00 Added Srd 823,000 391, 000 391,000 400,000 409,000 409,000 400,000 451,000 391,000 409,000 400,000 429,000 400,000 162, 100 407, 700 184,000 170,000 145,000 176,000 166,000 186,000 52,000 100,000 130,000 146,000 70,000 170,000 140,000 103,000 210,000 100,000 180,000 110,000 140,000 190,000 131,000 140,000 upon the decrease in numbers, but the higher amounts of acid were more active than the smaller amounts. As with B. suhtilis in the presence of salt, there is a marked decrease even in the tubes with smaller amounts, but unUke the higher amounts of 292 L. D. BUSHNELL TABLE 5 Influence of varying amounts of acid and air upon the growth of B. mesentericus in broth AIR PER CENTS OF N/1 ACID AND pH EXHAXTSTED MM. Hg. 0.6 per cent pH 5.60 1 per cent pH 5.00 2 per cent pH 4.30 4 per cent pH 4.00 Added Deter- mined Added Deter- mined. Added Deter- mined Added Deter- mined 0 50 ■ 100 per cent 5 10 25 5 10 25 5 10 25 116,000 104,000 102,000 108,000 111,000 104,000 111,000 101,000 109,000 90,000 80,000 89,000 54,000 56,000 75,000 45,000 52,000 48,000 114,000 114,000 102,000 108,000 96,000 101,000 114,000 109,000 104,000 120,000 150,000 140,000 140,000 68,000 160,000 110,000 110,000 200,000 114,000 102,000 118,000 102,000 118,000 108,000 91,000 109,000 102,000 240,000 220,000 280, 000 240,000 120,000 190,000 109,000 102,000 101,000 106,000 91,000 111,000 96,000 102,000 108,000 120,000 130,000 230,000 140,000 150,000 150,000 113,000 113,000 140,000 Averag a 107,000 66,000 107,000 199,000 107,000 215,000 103,000 151,000 175 350 525 685 5 10 25 5 10 25 5 10 25 5 10 25 40,000 41,000 47,600 42,900 44,000 40,000 40,000 41,000 39,000 44,000 48,800 45,000 19, 200 16,000 19,200 14,800 15,200 12,000 14,400 16,500 16,000 16,000 18,800 12,800 42,900 33,200 36,000 36,600 45,100 41,000 40,900 44,000 39,100 41,000 40,000 40,000 16,800 22,800 23,200 25,200 26,800 19,200 14,000 16,800 14,000 18,400 21,600 13,600 47,600 45,100 39,000 42,900 40,900 46,300 36,100 48,800 47,600 45,100 45,100 45,100 18,000 20,000 36,000 49,000 35,000 60,000 28,000 30,000 27,000 17,000 34,000 38,300 37,400 42,900 46,300 46,300 45,100 41,000 38,300 37,400 39,100 46,300 44,000 49,600 33,600 47,200 36,800 48,000 65,600 43,300 59,200 40,000 48,000 43,000 51,200 Average 42,800 15,980 40,300 19,360 44,100 31,700 41,800 47,140 salt there is a marked decrease in the higher concentrations of acid. Even sealing seems to cause a marked decrease in the number of viable spores of both types. Perhaps if the incubation period had been lengthened there would have been a still greater INFLUENCE OF VACUUM UPON GROWTH OF BACTERIA 293 decrease in case of B. subtilis. Parallel tubes which were not scaled showed heavy growth of these organisms in all concentra- tions of salt, but onlj' a trace of growth in the higher amounts of acid. In the sealed tubes there was a very faint visible trace of surface growth in the tubes containing larger amounts of air. This was easily broken up and did not re-form on standing. The plates made from these tubes checked much better than would generally be expected, when it is considered that organisms of this type produce such heavy surface growths in open tubes. However, these organisms do not form such adherent growths in sealed tubes, and by vigorous shaking a fairly uniform suspension may be obtained. In the case of B. mesentericus there was much less marked action of acid. Either there is not so much decrease in the higher amounts of acid or there is a sUght increase of this organism after the initial decrease. At the end of the incubation period there were about the same numbers as at the beginning of the experiment. In Experiment /F an attempt was made to determine the in- fluence of the amount of air and salt upon the thermal death point of B. subtilis and B. mesentericus. The results are shown in tables 6 and 7. The tubes and spore suspensions were prepared as above. It is evident from these tables that B. mesentericus spores are somewhat more casly killed by heat than those of B. subtilis. We have found this to be true in numerous other tests upon the thermal death point of these organisms. The amount of salt or the amount of air has practically no influence upon the thermal death point, particularly in the case of B. subtilis, the larger amounts of salt seeming, however, to protect the organisms to some extent. Experiment V shows the influence of acetic acid and varying amounts of air upon the thermal death point of B. subtilis and B. mesentericus. In this experiment the spores were prepared as above described. The Uquid in which the spores were suspended during the heating was extract broth plus 0.5 percent; 1 per cent; 2 per cent; 4 per 294 L. D. BXJSHNELL cent of N/1 glacial acetic acid, giving the pH as shown in the tables. Two centimeters of this acid broth were added to the tubes which were then heated in the steamer for fifteen minutes and cooled as rapidly as possible to remove air from the liquid. To TABLE 6 Influence of varying amounts of salt and air upon thermal death point of B. suhtilis spores. Original numbers of spores added 32,000,000 per cubic centimeter TIME OF HEATINa AT 98°C. MM. Hg. 15 minutes 30 minutes 60 minutes 120 minutes ■per cent Open 460,000 140,000 6,100 15 175 110,400 40,000 7,600 31 1 \ 350 520,000 170,000 6,500 49 525 640,000 200,000 8,300 15 685 400,000 121,000 9,300 32 Average 426,000 134,000 7,560 28 Open 670,000 180,000 8,700 38 175 500,000 250,000 9,800 21 2 350 400,000 120,000 6,400 42 525 580,000 210,000 14,600 71 , 685 700,000 160,000 10,400 24 Average 570,000 184,000 9,980 39 Open 860,000 480,000 18,000 36 175 840,000 200,000 14,500 56 4 1 350 840,000 190,000 33,600 66 525 473,000 200,000 26,000 82 685 860,000 44,000 27,200 72 Average 774,600 302,500 62 General ave raee 590,200 206,800 14,910 44 these tubes was added 1 cc. of a heavy suspension of spores sus- pended in acid broth, similar to that in the tubes. The tubes were then filled to the mark with similar broth and the volume noted. Amounts equal to 5 per cent, 10 per cent, 25 per cent of the total volume were removed from the tubes. They were INFLUENCE OF VACUUM UPON GROWTH OF BACTERIA 295 then cxliausted to the points desired and sealed at the mark. The tubes were all placed in a steam sterilizer and heated for one hour. Tables 8 and 9 show the results obtained. Since each tube was of slightly different volume the liquid remaining after removing the above amounts would contain slightly different TABLE 7 Influence of vari/ing amounts of sail and air upon Ihe Ihermal death point of B. mesenlericus spores. Original number of spores added 3,780,000 TIME or HEATINQ AT 9S°C. 8ALT MM. Hg. 15 minutes 30 minutea 60 minutes 120 minutes per cent Open 8,000 700 17 2 175 2,400 200 8 0 1 1 350 1,000 130 34 0 525 3,600 500 17 0 685 1,600 200 8 0 Average 3,320 346 17 1 Open 2,300 1,100 98 0 175 1,100 800 30 0 2 360 1,400 400 27 0 525 1,000 520 39 0 685 1,000 700 74 0 Average 1,360 704 53 0 Open 1,700 390 69 0 175 1,000 640 56 0 4 ] 350 2,300 700 88 3 525 1,400 320 20 0 685 1,500 200 70 0 Average 1,580 450 61 1 General ave rage 2,087 500 44 1 numbers per cc. The reduction in numbers in this experiment is so striking that the numbers are not included in the tables in each case. The average number for each cc. of the liquid re- maining in the tubes was about 37,400,000 per centimeter for B. subtilis and 1,760,000 for B. mesentericus. 296 L. D. BITSHNELL The above tables show a very marked influence of acid upon the thermal death point of both organisms. The fact that canned fruits and vegetables containing, or treated with acid, kept so much better than fruits and vegetables containing no acid, was TABLE 8 Influence of varying amounts of acetic acid and air upon the thermal death point of of B. subtilis spores MM. Hg. AIR PEB CENTS OF N/I ACETIC ACID 0.5 per cent pH5.5 1 per cent pHS.lO 2 per cent pH4.5 4 per cent pH 4.10 175 per cent 5 10 25 590 490 700 9 0 0 3 0 2 0 0 0 Average 593 3 1 0 350 5 10 25 482 142 336 0 0 0 0 2 6 0 0 0 Average 320 0 2 0 525 5 10 25 43 113 46 0 0 0 0 0 0 0 0 0 Average 67 0 0 0 685 5 10 25 44 600 702 0 3 1 2 4 0 1 2 0 Average 448 1 2 1 Open 372 0 3 1 General avera( ?e 360 1 2 1 formerly thought to be due to the fact that the acid inhibited growth. From the results which we have obtained, we are inclined to believe that the keeping is due, not so much to the influence of the acid in inhibiting growth as to the fact that most or all of the organisms present are killed by the heating process. INFLUENCE OF VACUUM UPON GROWTH OF BACTERIA 297 The few remaining are probably unable to grow to any extent in the highly acid medium. Here, also, the amount of air present in the containers has no influence whatever upon the thermal death point of the bacteria present. TABLE 9 Influence of varying amounts of acetic acid and air upon the thermal death point of B. mesentericus spores AIR PEB CENTS OF N/1 ACETIC ACID MM. llg. 0.6 per cent pHS.O 1 per cent pHS.lO 2 per cent pH 5.60 4 per cent pH 4.10 175 per cent 5 10 25 4 10 8 1 1 0 0 0 0 0 0 0 Average 7 1 0 0 350 5 10 25 11 2 16 0 1 1 0 0 0 0 0 0 9 1 0 0 525 5 10 25 13 10 10 2 1 3 0 0 0 0 0 0 11 2 0 0 685 5 10 25 2 5 16 2 0 2 0 0 0 0 0 0 7 1 0 . 0 Open 12 2 0 0 General avera ee 9 1 0 0 Experiment VI shows the influence upon the thermal death point of B. subiilis and B. mesentericus of several of the more common organic acids found in fruits. The spores were prepared as described in experiment V and placed in the acid solutions 298 L. D. BTJSHNELL after heating at 80°C. for twenty minutes to kill the vegetative forms. The results are shown in table 10. TABLE 10 Influence of organic acids upon the thermal death point of B. subtilis and B. mesentericus spores TIME OF HEATING FEB CENT OF N/1 ACID ADDED TO BROTH 0.5 per cent 1 per cent 2 per cent 4 per cent B. subtilis added 7, 305, 000 per cubic centimeter minutes Lactic 15 30 60 74,400 3,400 80 17,300 2,000 10 3,200 50 0 273 0 0 B. mesentericus added 2,500,000 per cubic centimeter f 15 18,000 100 90 10 Lactic < 30 8,200 0 0 0 60 13 0 0 0 B. subtilis added 17,500,000 per cubic centimeter Tartaric 15 30 60 520, 000 73,000 600 410, 000 54,000 576 52,000 3 2 2,000 3 0 B. mesentericus added 2,145,000 per cubic centimeter [• 15 14,000 1,630 4 10 Tartaric I 30 1,200 220 0 0 . 60 39 3 0 0 B. subtilis added 27,000,000 per cubic centimeter ' 15 2, 000, 000 110,000 200 500 Citric ] 30 240,000 600 200 ■ 60 1,200 4 4 0 B. mesentericus added 17,000,000 per cub c centimeter 15 24,000 60 50 12 Citric 30 2,150 48 0 0 60 0 0 0 0 The results show that there is very Uttle difference in the action of these organic acids. The acetic is perhaps a little more effec- INFLUENCE OF VACUUM UPON GROWTH OF BACTERIA 299 tive, but not enough to be of practical importance. This table also shows the influence of the amount of acid, and also the fact that B. mesentericus is somewhat more easily destroyed by heat than B. subtilis. SUMMARY From the results obtained, we are inclined to believe that B. mesentericus predominates in canned foods because it is capable of growing to some extent in absence of air, rather than because its spores are more heat resistant than some other types of aerobic bacteria. The amount of vacuum under which spores of these organisms are placed during the heating does not influence the thermal death point. The small amount of acid present had but slight retarding influence upon the growth of these organisms in air, but did have a marked influence upon the thermal death point. It may be that the beneficial influence of acid upon the keeping of canned foods is due more to the lowering of the thermal death point than to the inhibition of growth of the organisms. The amount of air remaining above the Uquid has little in- fluence upon the growth of these bacteria, since sealing the tubes prevents all but minimum growth. This inhibiting influence is more marked in case of B. subtilis than in case of B. mesentericus. REFERENCES Bitting, A. W., axd Bitting, K. G. 1916 Canning and how to use canned foods. Bull. 14, National Canners Assoc, Washington, D. C, 191 pp. Bruett, E. M. 1919 Utility of blanching in food canning. Effect of cold shock upon the bacterial death rates. Journ. of Ind. & Eng. Chem. 11, 37-39. Bdshnell, L. D. 1918 The influence of cold shock in the sterilization of canned foods. Jour, of Ind. & Eng. Chem. 6, 431-436. Cheynet, E. W. 1919 A study of the microorganisms found in merchantable canned foods. Jour, of Med. Res., 40, 177-197. Deichsetter, J. 1901 Ueber den Keimgehalt den Fleischkonserven. Zeitschrift fvir Untersuch, d. Nahr und Genussmit., 4, 1115. Hunter, A. C, and Thom, C. H. 1919 An aerobic spore-forming bacillus in canned salmon. Jour. Ind. & Eng. Chem., 11, 655-657. Lawrence, J. S., and Ford, W. W. 1916 Studies on aerobic spore-bearing non-pathogenic bacteria. Jour. Bact. 21, 273-320. 300 L. D. BUSHNELL PruHL, E. 1904 Beitrage zur Bakteriologischen Untersuchungen den Fleisch- konserven, Zeitschrift fiir Hygiene. 48, 121-134. Vaillabd, L. 1900 Les Conserves, Rept. to 10th Congress of Hygiene and Dem., Paris. Vaillabd, L. 1902 Les conserves de la viande: Les accidents qu'elles provo- quent leurs causes; les moyens de les prevenir. Rev. d'Hygiene, 24, T. 17-35 and 109-121. Wenzirl, J. 1919 The bacteriology of canned foods. Jour, of Med. Res. 39, (New Series 34) 349-413. SUBSTITUTION OF BROM-THYMOL-BLUE FOR LITMUS IN ROUTINE LABORATORY WORK H. R. BAlsER Contribution No. 44 from the Department of Bacteriology, Kansas State Agricultural College Received for publication December 10, 1921 One of the common methods for the qualitative determination of acid or alkali production by bacteria is to inoculate nutrient extract broth containing various carbohydrates with Utmus as an indicator. Litmus possesses the disadvantage of being reduced by many organisms to a colorless compound, thus rendering it useless as an indicator. Clark and Lubs (1917) mention that the sulphon- phthalin indicators are much more resistant to bacterial action than indicators like methyl red or litmus. They suggest, be- cause of certain preliminary tests, that indicators like brom- thymol-blue and brom-cresol-purple might be used to advantage in replacing other indicators which are now used in making indi- cator media. With this suggestion in mind, an experiment was undertaken to find a method of preparing media to determine qualitatively acid or alkaU production by bacteria, which would be easy to prepare, and more sensitive than litmus; and one in which the reaction could be quickly determined at any time during the . incubation period. In this experiment, sugar free broth was used, which was pre- pared as follows : One pound of ground lean meat was digested for two hours with 1 liter of distilled water. After cooking, the broth was filtered through absorbent cotton into a flask and steriMzed in the auto- clave at 18 pounds pressure for twenty minutes. When cold the broth was inoculated with a culture of Bad. saccharolyte (Rivas) 301 302 H. R. BAKER and incubated at 37°C. for forty-eight hours to render the medium sugar-free. Then the medium was sterilized in the Arnold for twenty minutes; 10 grams of peptone and 5 grams of sodium chloride were added; the reaction was adjusted to pH 7.0 with brom-thymol-blue; the medium was again steamed for twenty minutes, the reaction readjusted, and the medium filtered. To determine the amount of brom-thymol-blue which would inhibit acid production by microorganisms, fifty cubic centi- meters of sugar free broth were placed in each of fifteen flasks. TABLE 1 The influence of brom-lhymol-hlue upon acid production by Bad. coli-communis. Using 1 per cent glucose broth, initial pH 7.0, increasing amounts of a 0.2 per cent alcoholic solution of brom-thymol-blue inoculated u^lh 0.1 cc. of an eighteen hour broth culture DILUTIONS OF BROM-THTMOL-BLUE LENGTH OF INC0BATION O O ^, nof- f $5.00 per volume, Tnited States, Mexico, Cuba ""'f' °fi $5.25 per volume. Canada posi:paia [55,50 per volume, other countries Made in United States of America PEPTON Perfectly serviceable for the formulas and in all the technic of the bacteriological and antitoxin laboratory. It is employed in the usual proportions and for whatever purposes Pepton of this most desirable quality is required. Manufactured by Fairchild Bros. & Foster New York RABINOSE •01 FMiSTItHL ^ (ql NEMICAl^ U.S.A. Chemicals for Bacteriology THOSE acquainted with the high standard of PFANSTIEHL SUGARS and AMINO ACIDS will be interested to learn of the availability of the following reagents in the Standard of Pfanstiehl: Balsam, Caaada Cfor moaating slides'^. Benzidiue, C. P. (for blood test). Casein, C. P. (for testing pan- creatin). I Cedarwood Oil (for immersion of objectives). Cbolesterin, C. P. (for diagnosis of s^hilis). Creatinine C. P. (for blood tests, etc.). p - Dimethylaminobenzaldehyde, C. P. (for bacteriological tests). Gelatin (in sheet form for culture media). Litmus, highest purity. Phloroglucin, C. P. (for micros- copy). Potassium Sulphate, C. P. (for Kjeldahl determinations). Rosolic Acid, C. P. (for gastric and water analysis). Sodium Nitrate, C. P. (for Amino N determinations. Uric Acid, C. P. (for standard). Also Standard Solutions and Reagents for Blood Chemistry according to Folin and Myers. CARRIED BY THE LEADING DISTRIBUTORS Write for Htt of 600 reagenta availabh in the Standard of P/anatiehl SPECIAL CHEMICALS COMPANY OEVOTCO TO THt CHCMICAL INOETPCNOCNCC OF AHCRICA Highland Park, III. JOURNAL OF BACTERIOLOGY OFFICIAL ORGAN OF THE SOCIETY OF AMERICAN BACTKIIIOLOGISTS DEVOTED TO THE ADVANCEMENT AND DIS- SEMINATION OF KNOWLEDGE IN REGARD TO THE BACTERIA AND OTHER MICRO-ORGANISMS Eililortal lionrd EditOT^n-ChieJ C.-E. A. WINSLOW Yale Medical School. New lUven, Conn. A. I'arker Hitchbns Lore A. Rooer^, Ex ofBcio C. C. Hass R. E. Bdchanan P. F. Clark F. P. Gay Adinsory Editors F P. GORHAM F. C. Harrison E. O. .Jordan C. R. LiPMAN C. E. Marshall V. A. Moore L. F. Rettoer L. A. Rogers M. J. ROSBNA A. W. Williams II. Zinsser CONTENTS J. Howard Mueller. Studies on Cultural Requiroments of Bacteria. I 309 J. Howard Mueller. Studies on Cultural Roquirumonts of Bacteria. II 325 Selmax a. W.^ksmax. A Method for Counting the Number of Fungi in the Soil 339 Lethe E. Morrison and Fred W. Tanner. Studies on Thermophilic Bacteria. I. .\erobic Thermophilic Bacteria from Water 34o Robert G. Greex. An Apparatus for the Rapid Measurement of Surface Tension .. .. 367 Abstracts of American and foreign bacteriological literature appear in a separate journal, Abstracts of Bacteriology, published monthly by the Williams & Wilkins Company, undei the editorial direction of the Society of American Bacteriologists. Back volumes can be furnished in sets consisting of Volumes I, II, III and IV. Price per set, net, postpaid, $24.00, United States, Mexico, Cuba ; $2.5.00, Canada; $2(5.00, other countries. Subscriptions are in order for Volume V, 1921. Price, per volume, $5.00, United States, Mexico, Cuba; $5.25, Canada; $5.50. other countries. JOURNAL OF BACTERIOLOGY AN OUTFIT FOR THE PRECISE ELECTROMETRIC DETERMINATION OF H-ION CONCENTRATION IN SOLUTIONS THE TEST m OUR STOCK FOR IMMEDIATE SHIPMENT OF SERVICE 4882 Type K Potentiometer Outfit, 'with'diagrammatic illustration of'connections. This equipment is suited for practically every hydrogen-ion research problem in bacteri- ology, physiology, and bio-chemistry, and has been arranged as a result of the experience of many competent investigators in this field. It is parlicularly convenient as regards setting up and operation since no adjustments requiring high manipulative skill are necessary. The Clark hydrogen electrode is perhaps the most widely used for biological solutions. With the Enclosed Lamp and Scale Galvanometer as above illustrated, the type K Potentiometer Outfit measures the diiference in potential between a calomel electrode and hydrogen elec- trode to from .02 to .04% (± .2 millivolts), which in ordinary solutions is the "equivalent of 0.0001 of the pH unit. The outfit consists of the following equipment. 4790.'«Type K Potentiometer, L. & N 275.00 4796.IFTSnclosed Lamp and Scale Galva- Jnometer, L. & N 55.00 4797. Lamp Resistance for Galvanometer, for no volts 5.00 4878. Combined Support and Shaker, IL. & N. for calomel electrode, con- E"necting vessel and hydrogen electrode, with motor 75.00 4820. 2092. 4854. 48M. 4866. 9374. Weston Standard Cell, Model 4 25.00 Storage Battery 6.00 Clark Hydrogen Electrode Vessels with two platinum electrodes for same (2) ea. 9.00 Calomel Electrode Vessel 4.00 Clark Connecting Vessel 7.80 Switch, single pole double throw 80 4882 Complete Outfit, as above listed, with Shaker with motor for 110 Word volts, 60 cycle a.c 471.60 Ezimq 4883. ditto, but with [Shaker with motor for 110 volt d.c 471.60 Ezjef Prices subject to change without notice. ARTHUR H. THOMAS COMPANY WHOLESALE, RETAIL AND EXPORT MERCHANTS LABORATORY APPARATUS AND REAGENTS WEST WASHINGTON SQUARE PHILADELPHIA, U.S. A. Cable Address, "BALANCE," Philadelphia irY STUDIES ON CULTURAL REQUIREMENTS OF BACTERIA. I J. HOWARD MUELLER Fro?n the Department of Bacteriology, College of Physicians and Surgeons, Columbia University, New York Received for publication November 1, 1921 INTRODUCTION With the exception of a few such media as Uschinsky's and Frankel's, which are of interest only from the fact that they demonstrate the abihty of certain species of bacteria to syn- thesize protein and other complex physiological substances from simple salts, all the pathogenic bacteria are cultivated for prac- tical purposes on empirical mixtures containing infusions of meat, the digestion products of protein and so forth. Most pathogens either fail to grow, or produce very scant growth on simple synthetic media. Many fail to multiply even on meat extract-peptone media, but flourish after the addition of serum, blood, and similar materials. It occasionally happens that a lot of stock media, prepared with all due regard to known com- ponents and hydrogen ion concentration will fail to grow such organisms as pneumococci, although other lots prepared in the same way have been successful. Numerous attempts have been made to amplify the use of sjTithetic media by the addition of known compounds as a source of nitrogen or of some particular chemical grouping which has been suspected of playing a part, but as far as progress in the direction of routine cultivation of bacteria on media of known composi- tion is concerned, the results have been uniformly disappointing. On the other hand, there seems to have been very little effort made to attack the problem from the angle of an analysis of the basic factors supplied by the physiological mixtures which are 309 JOCBNAI. or BACTBRIOLOOT, VOL. VU, KO. 3 310 J. HOWARD MUELLER known to induce growth. One is struck by the lack of reference to the identity of the components of meat extract or meat infusion which seem to be so universally favorable. Purine bases are mentioned, but apparently on insufficient experimental grounds. In fact the only piece of work which the writer has found in which an effort was made to determine this point was reported by Armand-DeUlle (1913), who claims to have substituted arginine for meat infusion in media used for the cultivation of the tubercle bacillus. Other work (Long, 1919) indicates that this organism is able to grow on a great variety of substances, and one is not justified in appljdng the conclusions of Armand-DeUlle to other types of bacteria without further experimental evidence. The possibiUty of obtaining some light on the requirements of patho- genic bacteria from this point of \'iew is, however, fairly promising, and it is the purpose of the present paper to outUne the method of approach and to cite a number of preliminary experiments which serve to indicate the hues along which work may be done. The results already obtained with two of the factors which it has been possible to single out and follow up has been made the subject of two preliminary reports (Mueller, 1920), and will shortly appear in greater detail. A word regarding the purpose of such an investigation and the results which can be expected from it may not be out of place. The present empirical methods of media preparation are gen- erally admitted to be uncertain from the standpoint of result, and wasteful in materials used. It may be justly questioned, however, whether a more thorough understanding of require- ments will lead to a radical change in our methods of prepara- tion of media. For example, even if certain amino acids and organic bases should prove to be the only essential factors be- side salts and perhaps carbohydrates, the difficulty and expense of obtaining them in pure form might well prevent their ex- tensive practical use, and it is quite possible that peptone and meat will continue to be their most available source. It should, however, be possible to understand the reasons for the uncer- tainty of results and occasional failures now existing, and per- haps to guard effectively against them, even though with ex- perienced workers, they occur rather rarely. CULTURAL REQUIREMENTS OF BACTERIA 311 Perhaps the most important results to which success in such a piece of work might lead, are the applications of the findings to problems of more general biological importance, particularly to those of animal metaboHsm. For, whatever may prove to be the nature of these substances which cause growth of bac- teria, they are largely or entirely components of animal tissue, and it is probable that they are either needed also bj' the animal bodj^ and supplied by plant or other sources, or else are synthe- sized by the animal itself to fill some metabolic requirement. When it is possible to catalogue the substances required by pathogenic bacteria for growth, it will probably be found that most of them are either required by, or important in, animal metaboUsm, and while many of them will surely be compounds at present familiar to the physiological chemist, it is equally prob- able that some will be new, or at least of hitherto unrecognized importance. This point is sufficiently clear in the light of many recent publications in connection with the relation of vitamines to the growth of bacteria and of yeast. Probably of no less importance will be the results from the standpoint of the classification of bacteria. Doryland (1920) has discussed this question at some length, and it is quite possi- ble that unexpected relationships or dissimilarities in bacterial species may develop on the basis of food requirements. That related species among pathogenic bacteria do have similar needs cannot be questioned. The colon-typhoid group grow easily on simple meat extract broth. The streptococci and pneumo- cocci require the presence of an infusion of meat, the meningo- cocci and gonococci usually grow poorly without the additional presence of "hormones," blood serum, etc., while the influenza bacillus needs a substance associated with hemoglobin. It is quite possible that the more fastidious types of bacteria may require some of the factors necessary for the more easily growing forms, plus one or more additional substances. For example, if growth of the typhoid bacillus depends on the presence of three compounds in meat extract broth, A, B, and C, then the pneumococcus will perhaps fail to grow unless A, B, and C are sup- plied together with D and E, and so on to such organisms as the 312 J. HOWARD MUELLER gonococcus and others. Such a conception will at least serve as a working hypothesis upon which to begin investigation. The choice of organisms with which to work must be governed by several considerations. In the first place, since several fac- tors are probably involved in the growth of all parasitic bacteria, it would be well to select a species having a somewhat limited number of requirements. In this way one might hope to single out one or two factors at a time for identification and, using these as a basis, proceed to the study of others. On the other hand, bacteria with too simple requirements should be avoided, because of the probability that such organisms possess the power of using a variety of difi'erent materials ha\-ing no immediate relationship to the components of meat extract and peptone. The colon bacillus, for example, can grow on Uschinsky's or other simple media, about as well as on ordinary extract broth, and for that reason would not be suitable for the purpose at hand. A further point to be considered is that since the work will probably extend over a long period of time, the type of or- ganism selected must be such that its cultural requirements wiU not change materially during preservation on culture media. For these reasons, the pneumococcus-streptococcus group was selected as being probably the most satisfactory. These organ- isms grow well on meat infusion broth, but poorly or not at all on extract broth. While blood or serum improves the growth, neither is essential. In addition, these organisms can be pre- served almost indefinitely and kept at what should be quite uniform food requirements. The method by which this is at- tempted will be described fully below. The original plan of the work was to start with meat infu- sion-peptone broth, to eliminate such factors in its composition as could be managed experimentally, and to substitute known compounds, such as amino-acids, purine bases, etc., or failing in this, to determine, if possible, the chemical nature of the material removed from the media. As will appear, the work has been little more than begun, since the difficulties in substi- tuting known compounds for the meat infusion have proved unexpectedly great. CULTURAL REQUIREMENTS OF BACTERL^ 313 METHODS Preparation of cultures Three types of pneumococci, I, II and IIA, together with a strain of Streptococcus hemohjlicus, were used. The pneumo- coccus strains were each passed through two mice, and the streptococcus through one mouse and stock cultures were made from the heart blood of the second mouse into small tubes con- taining about 1 cc. of sterile human blood. The latter was ob- tained in the usual way from the median basilic vein and trans- ferred from a svTinge to small sterile tubes, each tube containing two or three glass beads. The tubes were than shaken until defibrination was complete, and incubated twenty-four hours to insure steriUty before being inoculated from the heart blood of the mice. After inoculation the tubes were incubated eight to ten hours until smears showed that the organisms were multi- plying, and they were then stored in the ice box. For trans- plants, ordinary meat infusion peptone broth containing 0.1 per cent glucose and brought to pH 7.4 to 7.8, were used. A small loopfuU of the blood culture was transferred to the broth, and after incubating eighteen to twenty-four hours, the culture was used, in the case of the pneumococci for the inoculation of a second similar tube of meat infusion broth. These two cul- tures are called respectively the A and B cultures. Experi- mental media were inoculated from the B cultures in the case of pneumococci, and from the A tubes with streptococci. In this way it is believed that the food requirements of the bac- teria remain reasonably constant. The stock culture remains alive for long periods of time, up to six months or more. After being opened repeatedly, old tubes gradually dry up, and such cultures have been transferred to fresh blood tubes, using the same technic, and always passing the strain through one mouse to guard against a gradual change in cultural requirements in spite of the blood media. It is not unlikely that significant quantities of blood may be carried over into the test media with only a single passage intervening, but it certainly is not enough to produce growth on unsuitable media, although it cannot be overlooked in estimating the value of results obtained. 314 J. HOWARD MUELLER Determination of liyniling hydrogen-ion concentration for strains employed It was found, by inoculating a series of plain broth media with pH ranging from 6.8 to 8.6, that the pneumococci found conditions suitable for growth between pH 7.4 and 8.0, while the hemolytic streptococcus grew well throughout the same range, and even as far down as pH 6.8. For the adjustment of experimental media, therefore, a reaction of pH 7.4 to 7.8 has been used for all the strains of test organisms. In cases where several experimental lots of media are prepared at one time, the adjustment of the reaction is facihtated by the direct addition of phenol red to the entire lot. It is a simple matter to reach the particular pH desired bj^ using a solution of phenol red of 0.02 per cent concentration, in the proportion of 1 ce. to 25 cc. of media. Normal or semi-normal NaOH (moderatelj^ free from carbonate) is then added drop bj' drop until the color is distinctly red, but not purple. One can quicklj- learn to recognize the correct color without the need of standard color tubes. In the case of unusually dark colored media the reaction can be ad- justed approximately by this means and finished by the usual method of dilution and addition of more indicator followed by comparison with standard tubes. Experiment showed that several times this amount of phenol red could be added to media without any influence on the growth of the organisms in question. Limiting range of osmotic pressure Although moderate variations in osmotic pressure are prob- ably of httle importance in culture media, an experiment was carried out in broth made up with varjing concentrations of NaCl. The test organisms grew equally weU in salt-free broth and through several intermediate concentrations up to an equiva- lent of 2 per cent NaCl, so that with ordinary care the question of osmotic pressure apparently need not enter into consideration. Presence of glucose In preliminary work with the pneumococci it was observed that occasionally a lot of meat infusion was met with which CULTURAL REQUIREMENTS OF BACTERIA 315 failed to give growth, although prepared with all due attention to hj'drogen ion concentration, etc. The addition of as little as 0.025 per cent glucose to such media at once improved them to such a degree that marked growth occurred, and with twice this rjuantity, growth was approximately as heavy as in the average broth prepared without glucose. While the point has not been verified experimentally, it is not improbable that most meat contains small quantities of glucose or other carbohydrate, and that occasionally this may be much diminished or absent. For routine purposes, therefore, and in all experimental media, 0.1 per cent glucose is added. This quantity is insufficient to produce acid in amounts great enough to kill the cultures in twenty-four hours, or to interfere w^ith agglutination with specific sera. Inorganic constituents Little is known definitely of the salt requirements of bacteria. There is some e\'idence in the literature to indicate that they are moderately elastic, and in any case such minute traces as may be required, must occur in the peptone and meat infusion used as routine media. Where these or similar substances are omitted, small quantities of inorganic salts may well be required. The nature of these can only be detennined when purified organic compounds can replace the meat infusion and peptone. To guard as far as possible against failure of expeiimental media to support growth through lack of inorganic material, the same salt mixture which is used in Uschinsky's and other similar media has been employed instead of simple NaCl, in the preparation of all experimental media in which the ordinary meat infusion is not used. To simplify the preparation of media, the salts are aU dissolved in twice the concentration required, together with glucose, and phenol red is added to this solution. ^ledia are prepared by adding an equal volume of meat infusion or other solution to this preparation, thereby reducing the concentrations of the constituents to that desired. The composition of the solution is shown below, and it wall be referred to subsequently as "glucose-salt solution." 316 J. HOWARD MUELLER per cent NaCl 1.0 MgSO, 0 .04 CaClj 0 .02 KKiVOi 0 .2 Glucose 0.2 Phenol red 80 cc. of 0.02 per cent solution per liter If all the components except the potassium phosphate are dis- solved, and the solution diluted almost to the final volumebefore this substance is added, calcium phosphate is not precipitated, although a good deal is probably lost in the precipitate which usually forms on adjusting the reaction and boiUng. Sterilization of media Since it has been shown that ten minutes autocla\-ing at 10 pounds pressure is even less destructive to sugars then the Arnold temperature for three quarters of an hour, this method has been used in most of the work. Practically no change in pH occurs, and contaminations are exceedingly rare. Recording of results T\Tien growth has continued for twenty-four hours, the degree of turbidity is recorded by comparing the tube ■with a set of seven tubes containing suspended BaS04 (Koser and Rettger, '19) ranging from the faintest trace of a cloud to a suspension as hea\'y as the best growth obtained with these organisms on good media. Negative growth is shown by " O," and anj' degree of growth by a ntimber corresponding to the series numberof the BaSOi standards. This affords a means of correlating experiments done at different times, although of course it is not quantitative. It usuallj- happens, however, with the pneumococcus cultures, that after optimum growth has been reached, autolysis follows very quickly. Within a few hours, a tube which has shown growth equal to standard tube nos. 4 and 5 will clear up and become simplj' opalescent. With Tj-pe I this happens quite regularly in less than forty-eight hours on good media, since maxunum growth is reached by this strain in fifteen to eighteen hours. Tj-pes II and IIA grow more slowly, but are often autolyzed in forty-eight CULTURAL REQUIREMENTS OF BACTERIA 317 hours. In recording growth, such tubes are marked "A." This autolysis is of course not a criterion of good growth, for in experi- mental media, scanty growth is also followed by clearing up. The streptococcus appears not to show this phenomenon. Preliminary experiments on infusion broth The growth of the test organisms on peptone-free infusion and on peptone water alone was first investigated. I Meat infusion 25 cc. Glucose-salt solution 25 cc. Peptone' 0 .5 gm. J . „ fMeat infusion 25 cc. \Glucose-salt solution 25 cc. I Peptone 0 .5 gm. Water 25 cc. Glucose-salt solution 25 cc. LOT NUMBER pH TYPE I TTPE II TTPE IIA 8TBEPT0CO0CI 1 7.8 A A 4 5 2 7.6 A 5 2 6 3 7.8 4 0 0 0 It appears from this experiment that peptone-free infusion is practically as satisfactory a medium for the growth of the pneu- mococci and streptococci as the usual meat infusion broth con- taining 1 per cent of peptone. On the other hand, with the exception of the strain of Type I pneumococcus, peptone water alone will not support growth. As a working hypothesis, it has been assumed that these organisms have a nitrogen require- ment supphed by peptones or amino acids, and also a need for certain accessory substances supphed by meat infusion. There- fore, on the basis of such a supposition, meat infusion seems to contain not only the accessory substances, but also peptones, amino acids or other sources of available nitrogen. The problem • Throughout the work, "Difco" Peptone has been used for the sake of uni- formity. 318 J. HOWARD MUELLER is complicated by tliis fact, since it is necessary to separate by some means the accessory substances from the nitrogen supply before either one can be studied separatelj'. The need for an accessory factor Before describing the various methods by which a separation of the accessory factors of meat infusion was attempted, an experiment will be described wliich strengthens the probability that such substances, other than protein degradation products, are necessary. It is possible that only amino acids or peptones might be required for growth, but that in the preparation of com- mercial peptone some essential amino acid, as, for example, tryptophane, is wholly, or for the most part, altered or destroyed. It is quite conceivable that unstable groupings other than trj^)- tophane may be present in the original protein molecule, which may withstand moderate heating in neutral solution and thus be present in meat infusion among the anuno acids or peptones in a soluble form, and yet be almost or quite lacking in commer- cial peptone. If such were the case, a whole protein hydrolyzed by trjT)sin or erepsin, together with salts and glucose, would probably serve as a complete culture medium. A specimen of commercial" casern after several day's digestion with trypsin was, as a matter of fact, found to be quite satisfactory without the presence of infusion, for the pneumococcus and streptococcus. However, when the casein was purified by three precipitations from Na2C03 solution by acetic acid, washed with alcohol and ether and then digested as before, growth was negative. The following protocol shows the results of such an experiment, in which casein was prepared directlj* from milk. The "crude" casein is the first precipitate obtained by acetic acid, the "pure" casein has been three times reprecipitated and finally washed in alcohol and ether. The two preparations of casein were dissolved in 0.5 per cent Na2C03 and digested with a small quantity of trypsin (Fair- child) at 37° under toluol for two weeks. At the end of this time, the two solutions were boiled and filtered. Twelve cubic centi- CULTURAL REQUIREMENTS OF BACTERIA 319 meters of each filtrate represented about 0.5 gm. casein. The following media were prepared: I Digest of impure casein 12 cc. Water 13 cc. Glucose-salt solution 25 cc. I Digest of pure casein '. . . 12 cc. Water 13 cc. Glucose-salt solution 25 cc. LOT Pll I II IIA STREPTOCOCCI 24 hours 48 hours 24 hours 48 hours 24 hours 48 hours 24 hours 48 hours 1 2 8.0 SO 6 0 0 0 0 0 6 1 6 0 6 0 5 0 6 0 Control experiments showed that the digest of pure casein was not inhibitory, for the addition of meat infusion produced heavy growth. The above experiment was confirmed several times, using difTerent preparations of casein, always with the same result. The conclusion seems warranted that while crude casein contains some accessory substance, this is not a part of the protein mole- cule (and hence probabh^ not an amino acid or polypeptide grouping), and can be easily removed by purification of the protein by standard means. Possibility of another source of growth accessory substance free from protein nitrogen By the use of the term "accessory substance" in connection with these studies, it is desired to avoid, as far as possible, the conception of vitamines. There is now abundant evidence in the literature that vitamines, particularly the water-soluble vitamine, may be as essential for certain microorganisms as for animals, but there is Uttle probabiUty that the accessory factor or factors of meat infusion is in any way connected with the water soluble vitamine. In the first place, muscle tissue is believed to be low in \dtamine. In the second place, experi- 320 J. HOWARD MUELLER ments with the pneumococcus and streptococcus have shown that "protein-free milk," which does contain the water soluble vita- mine, is almost without activity when substituted for meat in- fusion in media. To economize space, protocols of these ex- periments are omitted. It is not impossible that physiological extracts other than meat infusion might supply the accessory factors in a form more free from other nitrogenous compounds such as amino-acids, than the latter, and in a few experiments it has been possible to show that some other preparations, notably one from blood, and one from spinach leaves, gave growth when mixed with peptone and only scant growth in its absence. The blood was diluted with water, acidified, boiled and filtered. The spinach leaves were dried, ground fine and extracted with water. In the case of the latter, initial extraction of the dried powder with ether did not remove the accessory substances, and boiUng with repeated changes of alcohol for several hours extracted only a small part. It would perhaps be possible to develop a technic along either line for the preparation of a solution of the accessory factors sufficiently free from protein nitrogen to investigate the nature of the requirements of the test organisms by the addition of pure amino acids, but is has seemed more satisfactory first to exhaust as far as possible the more obvious methods for a separation of the meat infusion. Attempts to separate the growth accessory factors from the amino acids of meat infusion 1. Repeated extraction of meat. While carrjdng out some ex- periments along another fine, it was observed that the test or- ganisms grew as well on a trypsin digest of the insoluble meat residue remaining after the preparation of meat infusion which had first been thoroughly boiled out in three changes of water, without the addition of any meat infusion, as upon the usual peptone broth. This suggested the possibiUty that the growth accessory material might be extracted from the coagulated protein with some difficulty, and might be partially separated in this way from amino acids, etc. Chopped beef was, therefore, CULTURAL REQCriREMENTS OF BACTERIA 321 soaked in cold water in the proportion of 1 pound of meat to 500 cc. of water, and heated to 55° for a few moments and strained. The residue was boiled for five minutes with a second 500 cc. of water, and strained, and the extraction repeated a third time. The first extract was boiled to remove coagulable material, and all three extracts were filtered. Media were prepared in the following way, and after inoculation, gave the results indicated: OROWTH 36 aotJBs EXTRACT 1 EXTRACT n EXTRACT III PEPTONE H.O OLtJCOSE SALT I II IIA Strep- tococci CC. CC. CC. fframs cc. CC. 1 25 1.0 25 5 7 7 6 2 25 25 1 1 2 5 3 25 1.0 25 3 3 4 4 4 25 25 0 1 1 3 5 25 1.0 25 5 3 5 6 6 25 25 0 0 0 0 7 1.0 25 25 3 0 1 0 Other experiments of the same kind have given hke results, but there is enough irregularity so that the method is not ideal. It does, however, bring further evidence as to the presence of two classes of compounds in the meat infusion. Further work with the method, substituting known amino acids instead of peptone, may lead to more positive results. 2. Chemical fractionation of meat infusion. From the fact that commercial meat extract cannot replace meat infusion for the growth of the pneumococcus and streptococcus, one may suppose that heating, oxidation or simply long preservation may destroy the active material. If it can be shown experimentally that these or other comparatively simple chemical processes are destructive, plans for chemical separation can then be made in such a way as to avoid or minimize this difficulty. Accord- ingly, a few simple tests were made on the influence of such ob\dous procedures as suggested themselves. It was found that drying on the water bath, long boiling (two to three hours) in neutral or moderately acid solution (2.5 per cent HCl) and 322 J. HOWARD MUELLER exposure to mild oxidizing or reducirg agents in no way dimin- ished the power of infusion or destroyed its activity. It would seem that if hot alkali and strong acids are avoided, it should be possible to preclude loss of activity while carrjdng out simple precipitations on meat infusion. This has been found to be the case, and several different re- agents have been employed, without, however, any marked success in the separation of the constituents of the infusion. Precipitation with alcohol, up to 85 per cent concentration, carried out bj^ preliminary evaporation of the infusion to a small, measured volume, followed by addition of the required amount of 95 per cent alcohol, has in several experiments yielded a pre- cipitate which contains the growth accessory substances, since the test organisms grew on media prepared from the precipitate plus peptone, but not in the absence of peptone. The separa- tion is somewhat more complete if the first precipitate is dis- solved in a little water and reprecipitated. However, in either case, a part both of the growth accessory factors and of the amino acids remains in the alcohol filtrate, and it has not so far been possible to separate them quantitatively in this way. Lead hydroxide, lead acetate, mercuric chloride and silver nitrate and baryta have all been tried, none of them with results satisfactory enough to follow up extensively. In general, the filtrates from these reagents will permit growth to some extent without the addition of peptone, and better growth in its pres- ence, which is taken to indicate that the growth accessory frac- tion is not precipitated readily by these metals, while the amino acid fraction, perhaps in the form of peptone, is partially throwm down. By precipitation with tannic acid, followed by removal of the latter with Ba (OH)- + Pb (0H)o in the usual way, very similar results are obtained. While none of the methods of chemical separation have given successful results, they have at least shown that it is possible to submit meat infusion to such processes without loss of acti\aty. It has, indeed, been observed that where a single preparation was put through several consecutive precipitations, the activity was gradually diminished or lost, but it is at least as possible that CULTURAL REQUIREMENTS OF BACTERL\ 323 this may have been due to the loss of some essential factor in a form which could not be recovered, as for example a tannin precipitate or an extremely insoluble silver combination, as to actual chemical decomposition of the substance. It is, in fact, quite probable that there are many essential factors present, and that progress can be made only as methods are developed which will enable one to single these out for identification. Such a method seems to have been found in the treatment of meat infu- sion with charcoal. By this means, one or more factors are re- moved from the infusion which may be again suppUed by the addition of a small amount of peptone, or of a sulphuric acid hydrolysate of casein. In following up this lead, a considerable amount of work has been done, and a number of interesting observations made which have alreadj^ been briefly reported (Mueller, 1920) and which will be dealt with in the next paper of this series. SUM*L4.RY The purpose of this paper has been merely to outUne the plan of work and describe the method followed in our studies upon the problem of the nutritional requirements of certain bacteria. The procedures as outlined in the section on methods will be used in such work unless modifications prove desirable. Suffi- cient e\'idence has been obtained from the experiments here reported to warrant belief that two classes of organic compounds (in addition to carbohydrates), are required for the growth of pneumococci and streptococci, the first suppUed bj' protein degradation products, the second by extractives of meat. Both occur together in ordinary meat infusion, but they may be separ- ated more or less completely in several ways. The necessity for a non-protein substance is shown most clearly by the failure of a trypsin digest of purified casein to support growth, while that of impure casein is satisfactory. The possibility of a sep- aration of the two classes of compounds, as they occur together in meat is evident from experiments of several tj'pes, particu- larly by repeated extraction of meat, alcoholic precipitation of 324 J. HOWAED MUELLER meat infusion, and charcoal decolorization of heart infusion. There is every reason to beUeve that there may be several indi- \adual factors faUing into each of these two groups. REFERENCES Armand-Delille, Mater, Schabffer, and Terroine 1913 Jour, de physiol. et pathol. gen., 16, 797. DoRTLAND 1916 Jour. Bact., 1, 146. Koser and Rettger 1919 Jour. Infec. Dis., 24, 301. Long 1919 Amer. Rev. Tuberculosis, 3, S6. Mueller 1920 Proc. See. Exper. Biol, and Med., 18, 14 and 225. STUDIES ON CULTURAL REQUIREMENTS OF BACTERIA. II J. HOWARD MUELLER From the Department of Bacteriology, College of Physicians and Surgeons, Columbia Vniversily, New York Received for publication November 1, 1921 In the introductory paper (Mueller, 1922) of this series of studies, it was intimated that by treating an infusion of beef heart with charcoal, two factors necessary in the growth of hemolytic streptococci were removed, and that these factors could be again supplied by the addition to the charcoal treated infusion of commercial peptone or of a sulphuric acid hydrolysate of casein. The main facts learned up to this time about these two growth determining substances have been already briefly reported (Mueller, 1920), and will be here presented in detail. Work upon these substances has not yet been completed in the sense of chemical isolation and identification, and it is hoped that after further investigation which is now under way, it will be possible to give a more definite report of their nature. TECHNIC The general technic of preparation of media, adjustment of reaction, inoculation and recording of results, together with the method of carrying the test strain of streptococcus in culture, etc., has been fully described in the introductory paper (Mueller, 1922), and need not be repeated here. Removal of certain growth determining factors from beef heart infusion by charcoal, and reactivation of the charcoal treated infu- sion by means of peptone Three hundred cubic centimeters of a beef heart infusion, prepared by mixing chopped heart muscle and water in the pro- 325 326 J. HOWAED MUELLER portion of 1 pound meat to 500 cc. tap water, heating slowly to boiling, straining and filtering, was boiled for twenty-five min- utes with 10 per cent "Norit," a commercial grade of wood char- coal used in sugar refining. The mixture was filtered through paper, and the colorless filtrate used in the preparation of the following media: ■J . fDecolorized infusion 25 cc. \Glucose-SaIt solution 25 cc. Lot 2 Same plus peptone ("Difco") 0 .5 gm. . „ fOriginal heart infusion 25 cc. \Glucose-salt solution 25 cc. Lot 4 Same plus peptone 0 .5 gm. fWater 25 cc. Lot 5 < Glucose-salt solution 25 cc. [Peptone 0.5 gm. Each lot was brought to pH of 7.4 to 7.8, filtered if necessary, tubed and sterihzed at ten pounds steam pressure for ten minutes. LOT NUMBEE GROWTH OF STEEPTOCOCCra IN TWENTT-FOOB HOURS 1 2 3 4 5 0 61 6 7 0 It is evident from this experiment, that beef heart infusion, prepared as described above, constitutes a perfectly satisfactory medium for the strain of streptococcus used, without the addi- tion of peptone. Peptone water alone, moreover, even when suppUed with glucose and an inorganic salt mixture will not support growth of the organism. When the infusion has been treated with Norit and filtered, it is no longer suitable for growth, but may be reactivated by the simple addition of peptone. ' Numbers refer to the standard turbidity scale composed of varying suspen- sions of BaSO<. CULTURAL REQUIREMENTS OF BACTERIA 327 Many confirmatory experiments of the same nature have shown that within rather wide limits, the amount of charcoal used and the duration of boihng are without appreciable influence on the result, and as a standard method of producing the decol- orized infusion, 2 per cent of charcoal and fifteen minutes boiling have been adopted. The most obvious interpretation of the experiment is that the charcoal treatment removes certain substances, perhaps amino acids or polypeptides which occur also in commercial peptone, from the infusion. If polypeptides are concerned, they would probably be hydrolyzed by boihng with strong acid, and the resulting hydrolysate would not have the property of reactivat- ing the decolorized infusion, while if one or more amino acids are responsible for the phenomenon, they would perhaps withstand acid hydrolysis. Reactivation of decolorized infusion loith a sulphuric acid hydrolysate of casein A quantity of commercial casein was hydrolyzed by boiUng for eighteen hours with a mixture of six times its weight of water and three times its weight of concentrated H2SO4. The result- ing solution was freed from H2SO4 by Ba (OH) 2, the precipitate washed with water, and the filtrate and washings combined and concentrated. A quantity of the resulting hydrolysate equiva- lent to 0.5 gram of the original casein was used in the preparation of media as follows: J . , fDecolorized infusion 25 cc. \Glucose-salt solution 25 cc. Lot 2 Same plus casein hydrolysate 0.5 gm. [Water 25 cc. Lot 3 < Glucose-salt solution 25 cc. [Casein hydrolysate 0.5 gm. LOT .NUMBER TWENTY-FOUR HOURS' GROWTH 1 2 3 0 7 0 328 J. HOWARD MUELLER It is apparent that the reactivating material will withstand fairly thorough acid hydrolysis, and is, therefore, probably, but not necessarily, not of a polypeptid nature, since it is known that protein hydrolyzed in this way is not completely reduced to the amino acid stage. The further possibihty exists that the reactivating material is not connected with the protein molecule at all, but is present as an impurity. If casein is reprecipitated several times from so- dium carbonate solution, by means of acetic acid and the resulting product washed thoroughly by alcohol and ether, the "pure" casein so obtained will yield a hydrolysate which is just as active as that from crude commercial casein. It cannot, perhaps, be justly concluded from this experiment that the activating material is in fact a part of the protein molecule, for, as Funk (1920) has suggested, it may equally well be explained on the basis of a quantitative adsorption of the material from solution by the casein during precipitation. Since the material is known to be adsorbed quantitatively by charcoal, it is not impossible that it may also be taken up by other finely di\'ided precipitates, al- though it is not adsorbed by such precipitates as BaS04, metallic sulphides, etc., as will appear later. For the present, we must await further evidence to show whether the substance is of protein or non-protein origin. Since many of the experiments reported in the first paper of this series had shown that meat infusion was a rather difficult material to work with chemically, and since there are a number of fairly standardized methods for the partial separation of the amino acids in protein hydrolysates, the simpler experimental course seemed, at the time these preliminary observations were made, to attempt the separation and identification of the activat- ing material from such protein hydrolysates. The possibihty of re-extracting the material from the charcoal was considered, but since about 20 per cent of the total soUds of the heart in- fusion are removed by charcoal under these conditions, it seemed probable that even though a method for re-extracting from the charcoal could be de\ised, the extract might contain a somewhat complex mixture of compounds difficult to characterize and work CULTTJRAL REQUIREMENTS OF BACTERIA 329 with. It was, therefore, assumed as a working hypothesis, that one or more amino acids were involved in the reactivation and further efforts have been directed along the line of the isolation of such compounds. Preliminary observations on distribution of activating material If only a single substance were concerned in the reactivation, and it were one of the known amino acids, it should be possible to gain a clue as to its nature by using the hydrolysates of sev- eral different types of protein, and checking up a possible failure to reactivate in certain cases against a common deficiency, a method which has been widely used in work on animal metab- oUsm. A number of proteins were, therefore, submitted to sulphuric acid hydrolysis, and tested with decolorized infusion. It appeared that the hydrolysates of casein, meat protein, edes- tin, egg white, and to a lesser extent, egg yolk and gelatine, were able to reactivate, while the material from wool, silk, and wheat gluten were inactive. No common deficiency was apparent, and the results in some cases were not always clear cut. Since it will shortly be shown that two substances are probably involved in the reactivation, it is quite possible that certain proteins may lack one and not the other, and it wiU be necessary to run through such a series of proteins again, testing for each substance indi- vidually, when the properties of the two have been more carefully investigated. Separation of an active fraction from casein hydrolysates with mercuric sulphate After trying a number of methods for the separation of an active fraction from hydrolyzed casein, with Uttle success, it was finally found that a solution of mercuric sulphate in 5 per cent sulphuric acid would serve to throw down a precipitate contain- ing most, if not all, of the activating material. This separation was first carried out upon a fraction containing the mono-amino acids of casein prepared by the butyl alcohol extraction method of Dakin (1918) which had been shown in preUminary experiments 330 J. HOWARD MUELLER to contain the greater part of the active material. It was found that the filtrate from the HgS04 precipitate was no longer active, after precipitating the Hg with H2S, and remo\'ing the H2SO4 with Ba (OH) 2, while the mercury precipitate, after freeing from Hg and H2SO4 in the same way, was quite active. Following this observation, preparations of the amino acids known to be precipitated by this reagent, namely, tryptophane, tyrosine, cystine and liistidine, were obtained and their ability to reactivate the decolorized infusion either singly or in combina- tion with each other was tested and found negative. It was then found that mercuric sulphate would precipitate the active material directly from the casein hydrolysate, without resorting to the preliminary separation of the latter by means of butyl alcohol, using Dakin's method. T^Ioreover, the sulphuric acid used in hydrolj^sis did not have to be removed with baryta, but could be neutralized by sodium hydroxide and the precipi- tation carried out in the resulting strong solution of sodium sulphate equally as well as in a solution free from salts. It is rather difficult to determine the optimal conditions of precipita- tion, but a considerable excess of HgS04, and not too high a concentration of H2SO4 in the mixture seem to give the best results. As a standard procedure a weight of HgS04 equal to that of protein taken, make up in 5 per cent H2SO4, and added to a hydrolysate which contains from 5 to 10 per cent amino acids and is nearly neutral in reaction, has been used. Pre- cipitation is complete in about twenty-four hours. The filtrate, after freeing from Hg and H2SO4, may still show a slight ability to reactivate the decolorized infusion, but the precipitate is always strongly active; whether one of the two active materials to be described is removed more completely than the other by this method has not been definitely determined, but it is quite possible. Attempt to purify the active fraction by fractional precipitation with mercuric sulphate In the preparation of tryptophane by the method of Hopkins and Cole, advantage is taken of the fact that cystine is precipi- CULTURAL REQUIREMENTS OF BACTERIA 331 tatod more easily by that reagent, than tryptophane, in purify- ing the latter. It was, therefore, a logical procedure to attempt a separation of the active fraction of the mercuric sulphate precipi- tate by fractional precipitation with the same reagent. The results of a number of preliminary experiments in this direction gave evidence that the first crude mercuric sulphate precipitate contained two substances, both of which, together, were required for reactivation. One of these was easily reprecipitated by the addition of mercuric sulphate solution to the solution obtained from the first crude precipitate. The second was not completely reprecipitated even by the use of a considerable excess of the precipitant. The following experiment will illustrate these facts. It will be observed that in place of an acid hydrolysate of casein, a commercial enzjine digest of milk proteins to the amino acid stage, called "aminoids" has been used. This was done for two reasons. In the first place, the process is less troublesome than the long continued acid hydrolysis, and in the second place the preparations obtained are somewhat more active than those from an acid hj'drolysate. The latter, after a certain amount of chemical manipulation will often give rather weak growth in test media, corresponding to only a "2" or a "3" on the BaS04 turbidity scale. It is reaUzed that such a procedure is not be- yond criticism, and that it perhaps strengthens the possibiUty that one is not dealing with amino acids, but with adsorbed non- protein material. However, it is equally possible that these physiologically active substances are partially destroyed or altered by long treatment with acid. At any rate, the main facts of each point established with aminoids have been checked with an acid hydrolysate of casein, and like results obtained. Ten grams aminoids were dissolved in 100 cc. water and pre- cipitated with 100 cc. of a 10 per cent solution of HgS04 in 5 per cent HjSOj. After standing over night the bulky precipitate was filtered on a Buchner funnel, and washed with water. It was then suspended in water, made sUghtly alkaUne with Ba (0H)2 and decomposed with US, allowing the reaction to take place for some time, with occasional warming. The precipi- 332 J. HOWAED MUELLER tated HgS was filtered off, and the hea\'ily pigmented filtrate freed from Ba with H2SO4, and diluted to 100 cc. This crude preparation was tested for activity as follows: (Decolorized infusion 25 cc. Glucose-salt solution 25 cc. HgSO« ppt. fraction 2.5 cc. ( = 0 .25 gm. casein) Lot 2 Same plus HgSO« ppt. fraction.. .0.25 cc. ( = 0.025 gm. casein) J , o /Decolorized infusion 25 cc. [Glucose-salt solution 25 cc. LOT NUMBER TWENTT-ronR hours' growth 1 2 3 4 3 0 To the remainder of the HgS04 precipitate fraction was added 10 cc. of a 10 per cent solution of HgS04, and after standing over night, it was filtered. Both the precipitate and the filtrate were freed from Hg and from H2SO4 as usual, and brought to a volume of 100 cc. [Decolorized infusion 25 cc. _ . , I Glucose-salt solution 25 cc. Lot 1 < , plus [HgSO^filtrate fraction 1 .0 cc. Lot 2 Same plus filtrate fraction 0 .5 cc. Lot 3 Same plus precipitate fraction 1 .0 cc. Lot 4 Same plus precipitate fraction 0 .5 cc. Lot 5 Same plus each fraction 1 .0 cc. Lot 6 Same plus each fraction 0.5 cc. J . „ fDecolorized infusion 25 cc. \Glucose-salt solution 25 cc. CTJLTURAL REQUIREMEXTS OF BACTERIA 333 LOT NCMDER TWCNTT-rOCB BOOBS' OBOWTH 1 1 2 0 3 0 4 0 5 5 6 5 7 0 This particular experiment has been selected from a number of a similar nature since it shows a fairly clean separation into two fractions. Not infrequently one fraction or the other will by itself reactivate the infusion to some extent and produce slight growth, but almost invariably both together are better. Two points are illustrated in the above experiment. In the first place, an actual separation has taken place. It is not a distribution of the activating material through both fractions with a corresponding dilution, which might reduce the concen- tration below that required. In the second place, it is essential to have some idea about the quantity of material which is being added, best in terms of its equivalent of the original casein. Ob\'iously, if, as is really the case, a quantity of casein hydroly- sate, or of the first HgS04 precipitate, equivalent to 0.25 gm. casein will reactivate 50 cc. of a mixture of decolorized infusion and glucose salt solution, one must not test a later preparation or fraction with a concentrated solution, using perhaps an equiva- lent of 5 grams of casein or more. If such were done, and the preparation found active, one might still have lost 95 per cent or more of the active material and not recognized it, for bej^ond a certain optimum, growth is not increased by multipljing the quantity of reactivating material considerably. It has not been found possible so far to define the conditions of this precipitation so exactly as always to obtain a complete separation, and while it was useful in leading to the discovery that there were two substances involved, it has now been given up in place of a more certain method. 334 J. HOWARD MUELLER Separation into two fractions by means of silver sulphate and baryta Because of the presence of considerable histidine in the crude HgS04 precipitate, as shown by color reactions, a purification of the fraction by means of Ag2S04 and baryta suggested itself. As will be seen, this resulted in a simple method for the separa- tion of the activating material into the two fractions. The method is not attended by the uncertainty of the HgS04 frac- tional precipitation, and leads quite easily to a quantitative separation. The Ag2S04 precipitate, which contains the factor which will hereafter be referred to for the sake of bre\'ity as "X," corresponds to the second HgS04 precipitate, while the filtrate from the Ag2S04 fraction contains the "Y" as does also the filtrate from the second HgS04 precipitation. Aminoids are precipitated as described above with HgS04, and after freeing the precipitate from Hg and from H2S, a hot saturated solution of Ag2S04 (or AgNOa if the nitrate radicle will not interfere with further work) is added until an excess is present as shown by testing a drop with a drop of Ba(0H)2. Cold saturated Ba(0H)2 solution is then added until precipita- tion is complete and the precipitate filtered or centrifuged off. The filtrate is freed from Ag with H2S and Ba with H2SO4, and is concentrated to an equivalent of 10 per cent original aminoids over a low flame, and is found to contain the Y fraction, entirely free from X. The Ag2S04 precipitate contains the X fraction together with traces of Y, and it must be reprecipitated in the same way to free it entirely from Y. The Ag and Ba are re- moved as usual, the H2S boiled out, the solution cooled and more Ag2S04 and Ba(0H)2 added as in the first precipitation. This precipitate, when freed from Ag and Ba is found to contain the active X, quite free from Y. The test for active X or Y is of course made in the usual way, using decolorized infusion and glucose salt, together with a preparation known to contain Y or X as the case may be. Further purification of X fraction The Agi,S04 precipitate is always pigmented, and contains considerable histidine. It is needless to say that histidine, as CULTURAL REQUIREMENTS OF BACTERIA 335 well as tryptophane, tyrosine and cystine have been tested out individually and collectively against known X and Y prepara- tions and found to have no influence in producing growth under these conditions. The constancy of pigment in the X fraction, together with the fact of the disappearance of pigment from the heart infusion on boiling with charcoal, may be suggestive. However, the results obtained by the precipitation of this frac- tion with phosphotungstic acid, indicate that the pigment is not concerned in the action of the X fraction. By the addition of phosphotungstic acid to this fraction, in the presence of 5 per cent H«S04, it was possible in several experiments to obtain a filtrate quite free from pigment, which when freed from phosphotung- stic acid with Ba(OH)? and the excess of Ba removed, contained an active X factor. Such a solution, evaporated on a watch glass j-ielded a semi-crystalline residue. The phosphotungstic precipitate also contains a small amount of the X, but in the single experiment in which it was attempted to learn quantita- tively how it was distributed by using diminishing quantities against a constant amount of Y, the filtrate seemed to have about 75 per cent of all the X in the fraction. Unfortunately, these observations were made with a single solution of phosphotungstic acid, and all subsequent preparations of the reagent have de- stroyed the activity of the X fraction entirely. Up to the pres- ent, therefore, all that can be said of the X fraction is that it is apparently not in any way connected wdth the pigment nor with the histidine which it contains. It is hoped that further work will throw more Ught on this factor. Further ■purification of Y fraction The Ag2S04 filtrate, or Y fraction has proved to be somewhat simpler to work with than the X fraction. When evaporated, after freeing quantitatively from Ag and Ba, it is semi-crystalline. It contains a varjdng quantity of tjTosine and perhaps some trjTitophane. By precipitation with a small quantity' of HgS04 and allowing the material to stand over night, any tryptophane is thrown out, together with part of the tyrosine. The result- ing filtrate, after remo\'ing Hg and H2SO4, contains active Y. 336 J. HOWARD MUELLER It may be evaporated nearly to dryness, and after standing over night on ice to allow tyrosine to separate, may be filtered, the tyrosine washed out thoroughly with cold water, and the filtrate and washings, which still contain the active Y, evaporated fur- ther, with small additions of alcohol, to crystallization. The material which separates in the first crystallization is made up of microscopic round plates or spheres with no definite crystal form. After one or two recrystallizations these are seen to be made up of needles, and finally they crystalUze out as shining, colorless, microscopic leaflets, often with obtuse angles and sev- eral sided, for the most part in rosettes; but when single crystals can be made out, they are hexagonal. These crystals are appar- ently a new amino acid, containing sulphur, and a detailed account of their composition and properties will appear shortly. Unfortunately, it cannot at this time be stated definitely that they constitute the Y substance. The earlier crops of indefinitely crystalhne material are highly active, as are also the mother liquors. A quantity of soUd weighing as Uttle as 0.00001 gram has reactivated 25 cc. of a mixture of decolorized infusion and glucose salt solution in the presence of an active X preparation. However, on further recrystallization the activity is apparently lost, but the activity of the mother liquor also slowly disappears, and it is, therefore, not yet clear whether the sulphur containing crystals are some tautomeric form or oxidation product of the Y, or whether they are in no way related. In the meantime, a study of the properties of the sulphur compound will perhaps lead to methods by which it will be possible to answer this question. DISCUSSION OF RESULTS Since pathogenic bacteria such as the streptococci are biologi- cally adopted to growth on animal tissues, it is more than prob- able that the chemical substances required by them for growth are constituents of the animal body and hence probably of im- portance in animal life. It is evident that studies on the cultural requirements of such bacteria may, therefore, lead to results equally as important to the student of animal nutrition as to CULTUltAL REQUIREMENTS OF BACTERIA 337 bacteriologists. In the case of the two substances which have been described in the present paper, one may anticipate that they may develop significance for animal metaboUsni as further information on their properties is gained. Whether they are related to the vitamines, and constitute, as Funk (1920) has sug- gested a Vitamine D, connected with deficiencies in certain proteins, is a question which can so far not receive an answer. The writer has preferred to avoid the conception of "vitamine" as far as possible in the experimental approach to the problem. In any case, the method as it has developed, offers a simple biological test for the presence of these compounds, which has the very great advantage that it may be quickly carried out. All the test solutions are easily prepared, and the growth test itself requires only twenty-four or at most forty-eight hours for com- pletion. With this advantage, it should prove possible to iso- late and identify these compounds unless their properties are such that decomposition or molecular rearrangement follows their purification. Work is being carried on with both fractions, and it is hoped to have more data available for report in the near future. CONCLUSIONS 1. Peptone-free beef heart infusion plus glucose and inorganic salts constitutes a satisfactory medium for the hemolytic strep- tococcus. 2. Short boiling of heart infusion with 2 per cent wood char- coal ("Norit") removes some component of the meat infusion and renders it no longer suitable for the streptococcus. 3. Such an inactive infusion may be reactivated by the addi- tion of small quantities of peptone or acid hydrolysate of cer- tain proteins, such as casein and edestin. 4. Acid hydrolysates of such proteins, as wool, silk and wheat gluten are not suitable for reactivation. 5. The activating material may be precipitated from hydroly- sates of casein by means of HgS04. 6. It may be separated into two fractions, active only when mixed together, by means of fractional precipitation of the first 338 J. HOWAKD MUELLER HgS04 precipitate by HgS04, or precipitation bj^ Ag2S04 and baryta. 7. So far as has been learned, knowTi amino acids will function in place of neither of these fractions. 8. The silver sulphate precipitate or X fraction does not de- pend for its activity on the pigment. It escapes precipitation by phosphotungstic acid under certain conditions, but is readily destroyed by this reagent. 9. The silver sulphate filtrate, or Y fraction contains a con- siderable quantity of a new sulphur containing amino acid, the relation of which to the active Y has not yet been demonstrated. REFERENCES Dakin 1918 Biochem. Jour., 12, 290. Funk 1920 Proc. N. Y. Pathol. Soc, 20, 119. Mueller 1920 Proc. Soc. Exper. Biol, and Med., 18, 14 and 225. Mueller 1922 Jour. Bact., 7, 309. A METHOD FOR COUNTING THE NUMBER OF FUNGI IN THE SOIL' SELMAN A. WAKSMAN Received for publication November 2, 1921 The numbers of fungi in the soil are usually determined by the plate method used for the determination of the number of bacteria. In view of the fact that the dilution used for the determination of bacteria is necessarily high, due to the large numbers of bacteria in the soil, the fungi are so diluted that very few appear on the plate: of a dozen plates prepared from the same soil, using the same dilution, three or four may be free from fungi entirely, three or four may have only one or two fungus colonies, while three or four may have several colonies, particu- larly in the case of humus-rich and acid soils. It has been pointed out by the author (1922 a) that the probable error in- volved in the determination of the numbers of fungi by this method is so great, as to make the results absolutely worthless. To reduce the variabiUty of the numbers of fungi on the plate and thus obtain a low probable error, low dilutions have to be used, so as to have 30 to 100 fungus colonies developing on the plate; this would necessitate a dilution of only 500 to 2000 for an ordinary fertile soil. But, if the common plate used for the determination of bacterial numbers is employed, so many bac- teria will develop on the plate, as to prevent the development of most of the fungi. To obviate this difficulty, use was made of the fact that fungi can grow readily at a much higher acidity than the bacteria and actinomycetes. The author and others have long made use of the fact, that, when a culture of a fungus is wanted free from bacteria raisin 'Technical Paper No. (61) of the New Jersey Agricultural Experiment Station, Department of Soil Chemistry and Bacteriology. 339 340 SELMAN A. WAKSMAN agar, which is acid in reaction maj^ be used, or a drop of lactic acid added to each tube of the common media. A medium has therefore been devised, havdng a reaction acid enough to prevent the development of the actinomycetes and the great majority of bacteria. At first, raisin agar was used, but, in view of that fact that this medium is not definite in composi- tion and its reaction depends on the acid content of the raisins, the following synthetic medium has been developed. Glucose 10 grams Peptone 5 grams KHjPOj 1 grams MgSC-THjO 0 .5 grams Distilled water 1000 cc. Dissolve by boiUng, add enough ^ acid (H2SO4 or H3PO4) to bring the reaction to a pH = 3.6 to 3.8. This will require from 12 to 15 cc. of N acid per liter of medium. Add 25 gm. of agar, dissolve by boiling, filter, tube and steriUze as usual. The final reaction should be pH = 4.0. The soil is now diluted, in the regular way, to only 5V to -^h^ of the highest dilution used for the determination of bacteria and plates are prepared in the regular way. The plates are incubated for seventy-two hours at 25°C. To obtain an accurate count and a low probable error, 10 plates should be prepared. The colonies may be counted after 48 hours, then after 72 hours, due to the fact that in some soils, rich in mucorales, the spreading forms will tend to overgrow the plate in 72 hours. The following table gives a comparison of the numbers of fungi obtained from the same soil by the regular bacterial plate and the special method hereby suggested. A cultivated field soil was used for this purpose. Instead of an impossible figure of 460,000 fungi per gram of soil, only 29,400 have actually been found by the modified method. The variability of the common method is so great as to make it valueless. Further details on the appUcation of the new pro- cedure will be pubUshed elsewhere (1922 b). This method can also be appUed to the determination of the number of fungi (molds) in various food preparations. COUNTING THE NUMBER OF FUNGI IN SOIL 341 TABLE I .SOIL DILCTIO.N 200,000 1000 Egg- ng: ilbumen r (used for Specii il medium the determina- tion of bac- ter ia) 2 34 1 40 5 40 0 24 7 2 38 22 1 24 3 22 0 20 2 30 Mean 2.3 ± 0.47 2.21 db 0.33 96.1 ±14.3 % 20.4% 29.4 7.9'i 27.1 5.8' ±1.70 a ' ± 1 .19 C.V ±4.0% Em % Number of fungi per gram of moist soil 460, 000 ±94, 000 29,400 ±1,700 REFERENCES Waksmak, S. a. 1922 a Microbiological analysis of soil as an index of soil fertility. I. Mathematical interpretation of the results of a quan- titative bacteriological analysis of the soil. Soil Sci. (To be pub- lished.) Waksman, S. a. 1922b The growth of fungi in the soil. Soil Sci. (To be published.) JOUBNAL OF BACTEBIOLOOT, VOL. VIl. NO. 3 STUDIES ON ther:mophilic bacteria . I. AEROBIC THERMOPHILIC BACTERIA FROM WATER' LETHE E. MORRISON and FRED W. TANNER Department of Bacteriology, Universitij of Illinois, Urbana Received for publication November 26, 1921 I. INTRODUCTION The discovery of microorganisms that are able to Uve at rela- tively high temperatures (60^C. and above) has forced us to change our ideas on the resistance of protoplasm to heat and to admit that life is possible above the generally fixed limit of 42° to 45°C. The term thermophilic was probably first used by Miquel (1879) to describe those organisms that grow at temperatures so high as to be fatal to most microorganisms. This conception seems to have been lost sight of by many more recent workers. In order to have a better understanding of what the term thermophilic means, a number of definitions of the term as found in different texts on bacteriology are included. In his physiological classification of bacteria, Giltner designated as thermophilic those that have a minimum temperature of 45°C., optimum, 55°C. and maximum, 70°C. IMuir and Ritchie define thermophilic bacteria as organisms that grow best at a tempera- ture of from 60° to 70°C. Hiss and Zinsser say that thermophiUc bacteria are high temperature bacteria obtained from hot springs and from the upper layers of the soil. Rahn in Marshall's Microbiology describes thermophiUc bacteria as extraordinary organisms having their maximum between 70° and 80°C., a temperature which coagulates albumin ; corresponding to the high ' Abstracted from a thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Bacteriology. Copies of the original thesis are on file in the Library and Bacteriology Seminar of the University of Illinois. 343 344 LETHE E. MORRISON AND FRED W. TANNER maximum the thermophiles have a verj' high optimum, and the minimum hes with most species above 30°C. According to Hew- lett there is a group of so-called thermophilic bacteria which thrive best at a temperature of 60° to 70°C. Those bacteria whose optimum temperature is above 40°C. and which are spoken of as the "thermophil" bacteria, is the definition given for them by Morrey. Buchanan does not mention thermophiles in his book but speaks of the organisms which produce large quantities of heat as thermogenic bacteria. In his book Chester places thermophilic bacteria in a class that does not grow at room tem- peratures or below 22° to 25°C. II. RELATION OF HIGH TEMPERATURES TO LIFE All h\ang things have their minimum, maximum and optimum temperatures for growth and other functions. The range of temperatures at which they exist may depend among other fac- tors, on the species and on the ancestral history of the indi\-idual. Many investigators have experimented on the growth of organisms at high temperatures with varied and interesting results. The first data concerning organisms that live at high tempera- tures were published by Sonnerat (1774). He reported on fish that hved in water at a temperature of 69°C. Schwabe (1837) reported the growth of algae in a hot spring at Karlsbad at 70°C, Flourens in 1846 mentioned algae which flourished in a hot spring at a temperature of 98°C. Brewer (1866) found some "Nostoc- formen" in a hot geyser at 83° C. Ehrenberg reported the exist- ence of red and green algae from the Island of Ischia which grew at 63° to 65°C. III. HISTORICAL The Uterature on thermophilic microorganisms has already become quite voluminous and in order to save space in this publi- cation, we have summarized in table 1, those papers which a careful search of the Mterature has revealed. STUDIES ON THERMOPHILIC BACTERIA. I 345 IV. EXPERIMENTAL Sources of ciiUures. The present investigation has been hmited to a study of some of the characteristies of 52 cultures of aerobic thermophilic bacteria from water; a more detailed study of forms from other sources is now being made. The cultures of ther- mophilic bacteria used in this investigation were isolated from samples of water furnished by the Illinois State Water Survey. Permission to use the samples was obtained from Professor E. Bartow, while director of the Water Survey. These samples came from different places in the state of IlUnois and had been collected and shipped to the Water Survey according to directions furnished by them. ^lany types of waters from different sources including deep wells, shallow wells, drilled wells, dug wells, springs, raw and treated municipal supplies and springs were used. In this manner it was possible to carry on a more representative study than if the samples had been taken from a restricted area. Out of 224 samples of water, 60 were found to contain thermophilic bacteria according to the method adopted for their isolation. This is to be regarded as a minimum for it is believed that thermophiles are quite abundant in nature and many samples wliich were negar tive when 1-cc. portions were examined would probably have been positive had a larger amount been used. Method of isolation. Agar plates were poured in the usual way using 1-cc. and 0.1-cc. portions of the samples; the plates were incubated at 55°C. for twenty-four hours. Most of the ther- mophiles grow very rapidly at 55°C. and a longer incubation period was unnecessary. Any colonies that had developed in that length of time were transferred to agar slants; later it was found that it was easier to keep the cultures in broth since agar dried so quickly at 55°C. Control plates and agar plates that had been exposed to the air of the laboratory were incubated under the same conditions, but in no case was there any growth of thermophiles shown, either from the agar itself or from the air of the laboratory. 346 LETHE E. MORRISON AND FRED W. TANNER M ~ — S >- C ■2 -^ c o +J *^ ■t 9i S fe » a i I oj ■tJ c; a £ 0.^ ■s " s £ ° S 2 " S o S ■o 2 t' a o -g 5 6 |||h ill ° £1 i.| I "s S g - c ^ J= •r (u c ♦J K-f ".£ 1 "^ rs — : ■-? ^ i 5 a ■ii -w -.J d s c ; 5 :h i . 1 "J * -§ lis- S3 §^ 1. fe • a J 'g J 5 .3 ^1 2 "i ITS - ■£ tr. " 5 g •g " 5 1 I •= S ^l I 5J 3 S o -2 s 0; o g 1 G .5 f^ H O O >, - i; C ^ O g 4> « ^ -S £ 3 a ^- «= i .2 S • o fe c £ g, ■tJ »S 3 03 u O c u a a a -c -.£ £ ; ■£ S -g ° 5 £ s- £ "E C O _ c' .' £ C-1 ITT up ■S5 a *o S •- a g ^3 S 3 — s STUDIES ON THERMOPHILIC BACTERIA. I 347 Si: § 8.1 ! |*£ ill •g * • III ■|l! g J .c S.s"S < •€ E ^8 %. ■5*2 5 E S S^ * r II sl s s a .5 i I o ■« 1 — >>- S S w «■ o w JE e J< 0 3 I S ^ a sill illl S g 1^1 S £ c >. - a o c, -^ B B o zs 2 o = -3 2 i 3 — a § s E a u S p p op p c^ w o «■) CJ w p o ? I si s S a S ? I I . T ^ -is C □ ^- . OD » 3 - fe *j s -^ = :2 S •c .a» 1 ■fl 1 h-J c; s ? a s i? S i< 05 05 'a 8 S P -a a 348 LETHE E. MORRISON AND FRED W. S S£ O J. H ►J m 55 :; 5 S.S =2 -e = .= .2 -5 .£ "^^ — s S .i ^■ S; ^ oj cj t£ ^ 2 < TANNER £ ■c 3 g^ .H ?i a ^ fl "2 £ ll » 1 •3 g > i| CS 5 " II .2 = ^ & C 5 ^ CD b a * li? II c: — J! 3 J5| "o "B. 2 § ^ E fe£ Bf o en "^ li c u .5 C.S = c- i e .2 5 m5 -0 .O £ -H -S cj 03 c; J = ■3 STUDIES ON THERMOPHILIC BACTERIA, 349 s » 4! J II a 1 2 ji '2 a : €=. 2 eg a 5 s i a o ou E S ^5 P oo gi ^ S9 ^ 2 S "S fc. c3 C ^ 3 a ■ 8 o H O " -2 ■-3 »: " o a C3 .3:5 5 •! t fc 3^. 3 I 3 5 ■s s ^ 1 .§■3 = S = 5 5 ^ Si ££ P 2; ^ S3 S o C I '5 C S G a 3 o o e o o *- •5 8 fe b £ b « I I b S b 3 3 £ « § •S ■2 s 350 LETHE E. MORRISON AND FRED W. TANNER lis •5 -c e £ a) 3 ._• a S -2 £ * il CJ ■- T! X Y i ^ c c a S a =5 ^ o c 9 5 •?S M 08 ^ o 2 c 13 . ^ -w S Z S § ■ .£-5 g o £ ^ f. ^ . ^ 5_ !« 33 '^ 3:2 S o i > n: -a .^ ' 2 s ■^ x> 2q ^ £ 3 -o .S ja 3 s o d < C/J O "3 a a. < B ■a P5 § OP o a o "■Sag ■^ ^ Si 7? *^ — -^ o .£ o > .2 ^ i s -^ = a s H E 3 a-2 V 1^ S 2 >> ° i « a .2 o S O O - I 0} t£ S P m O 'S a a c c o £ ^ 5 ri ■^ 111 = I f 1 i g § -5 4 -g '^ .(-■ "o "g S til 2P l"l gl ■^1 S'g g g as a « 5 ^ -s a H 2 S c a; a i = I? a a c ^ 1^ '• 5 -S a" 2 q; P £ ^ ^ .5 I -S •^ a a i • -a S -^ jt ■ t = i § 'si gt • I •= ° a I < a J2 .5 E .''' E «' E T •9 I Is - i p E ►J is I I I J § S J tS fS *3 ^ -«i -c C ». R. o o o g £ £ III *■?■ S aja o a> 4} SCO E- O O a I O 3 I e STUDIES ON THERMOPHILIC BACTERIA. I 353 •c a Qo 3 n s a s 3 3 3 s s a a a Q. poo O i B 3 ^ 00 6 I 354 LETHE E. MORRISON AND FRED W. TANNER Methods of study. Inoculations into the different media used in this work were made either from twenty-four-hour agar slant cul- tures or twenty-four-hour broth cultures. Since this work was begun before the adoption of the new chart, the Descriptive Chart of the Society of American Bacteriologists indorsed in 1914 was used in the study of these thermophiles. The group number for each culture was determined under as uniform conditions as pos- sible. In the work that is now under progress on thermophiUc bacteria from soil, canned foods, and other sources the Descrip- tive Chart indorsed by the Society of American Bacteriologists at the meeting of December 30, 1920, will be used. The index number, it is beUeved, will give a better description of the or- ganisms since it seems to embody characteristics which are more important. Media and technic. With one or two exceptions the media and technic used in this study were those recommended by the Committee on the Descriptive Chart of the Society of American Bacteriologists in their report on INIethods of Pure Culture Study (1920); the cultures were, however, all grown at 55°C. Other exceptions will be mentioned later. Morphologij. All the cultures studied were motile rods and usually grew in chains containing from two to many individuals. Sometimes chains of four or five rods showed an active snake- like movement. The rods were both long and short; some had rounded ends. Carbol fuchsin and Gram stains were used to stain the smears; all were Gram positive except nos. 10, 20, 40. Without exception, the cultures studied formed spores. Some of the spores were central and some polar; some were oval and some round; in a few cases the diameter of the spore seemed to be larger than that of the rod and produced a sort of Clostridium form. Nutrient broth. Witte's peptone was used in the nutrient broth employed in this study, since it seemed to possess certain advan- tages over other peptones. Good growth was secured with all strains in nutrient broth at 55°C. Most of the cultures produced turbidity and sediment in the broth; the surface growth in many of the cultures was membranous or showed a heavy pelUcle. STUDIES ON THERMOPHILIC BACTERIA. I 355 Indol. Tests for indol were made on nutrient broth cultures that had been incubated for four days at 55°C. Both Ehrhch's test and the Nitroso-indol test were used. It was found that EhrUch's test was much more satisfactorj^ when the tubes were heated sU!ihtl}^ All of the cultures formed indol from Witte's peptone in varjdng amounts. Htjdrogen sulfide formation. To determine hydrogen sulfide formation, nutrient broth (made of Witte's peptone) over which a strip of lead acetate paper was suspended by means of the cotton plug, was used. The cultures were incubated for four days at 55°C. The blackening of the paper indicated hydrogen sulfide formation. All the cultures studied formed H2S. Streak cul- tures on "Bacto Lead Acetate Agar" plates also showed that all the cultures formed H2S. This latter medium seemed to be well adapted to the determination of tliis characteristic. Potato slants. The growth of these thermophiles on potato slants was abundant in most cases even after twenty-four hours at 55°C. The type of growth varied from a filiform to spreading growth. The potato was turned gray, brown, or reddish brown. The cultures could not be kept longer than from two to three days since they dried out so quickly at 55°C., but the growth at this temperature was quite rapid on this and other media. Liquefaction of gelatin. The "provisional method" was used to determine this characteristic. The cultures were first accus- tomed to the gelatin medium by preUminary cultivation for eight- een to twenty-four hours in a 1 per cent solution of gelatin at 55°C.; then the surface of gelatin in test tubes was inoculated and the tubes incubated for thirty days at 20°C. All the cultures except nos. 6, 15, 20, 32, 51 had partially or completely hquefied the gelatin at the end of thirty days. Gelatin cultures prepared in the same way and incubated at 55°C. for four days were all hquefied with flocculent growth throughout the gelatin and would not harden when placed in the refrigerator. The fact that all the cultures studied hquefied gelatin at 55°C. and some of them at 20°C. can probably be explained by the fact that 20°C. was below the minimum temperature for growth for those cultures which did not hquefy gelatin at that temperature. 356 LETHE E. MORRISON AND FRED W. TANNER Litmus milk. In this medium azolitmin was used as the indica- tor. The litmus milk cultures were incubated at 55°C. for four to seven days, only, because they evaporated so quickly at this temperature. Peptonization occurred with at least 75 per cent of the cultures; and in each of these cases the medium was alka- line. All of the cultures curdled the milk and in those cases where peptonization did not occur the medium was acid. More work is being done on the growth of these organisms in milk and par- ticularly on the use of brom-cresol purple as an indicator. Fermentation of sugars and glycerol. No gas was formed by any of the cultures. None of the cultures produced acid in lactose; the cultures varied slightly in their formation of acid in glucose, sucrose and glycerol broth. Brom thjTnol blue was used to test for acidity since that indicator was used to adjust the reaction of all the broths when made. The amount of acid formed in the different broths by these cultures may be of signifi- cance and should be determined. Oxygen relation. This characteristic was determined by noting the presence or absence of growth in the open and closed arm, respectively, of fermentation tubes containing glucose broth. All the cultures used in this study were found to be strict aerobes. Reduction of nitrates. To determine this characteristic, both nitrate broth and nitrate agar slant cultures incubated at 55''C. for four days were used. SulphaniUc acid and alphanaphthyl- amine were used to test for nitrites. All of the cultures reduced nitrates. Diastolic action on starch. Two per cent agar containing 0.2 per cent of soluble starch was used for this determination because this stiffer agar seemed to stand the incubation at 55°C. better. Dot inoculations were made in the center of the petri dishes con- taining the hardened starch agar; these were incubated at 55°C. for forty-eight hours since longer incubation dried the agar and made it crack. All the cultures grew well on this media and all produced diastatic action, some feeble and some strong. Temperature relations. Some of the cultures were grown on agar slants at different temperatures and it was found that 50° to 55°C. was the optimum temperature for growth. Since it is STUDIES ON THERMOPHILIC BACTERIA. I 357 believed that the temperature relations of the thermophilic group of microorganisms is a subject worthy of intensive study, a sepa- rate investigation on this subject was initiated which is now near- ing completion in this laboratory. V. DISCUSSION A comparative study of 52 strains of thermophilic bacteria from water indicates a group in which the characteristics are not widel}^ divergent. All of the strains were spore formers and all Uquefied gelatin. When separated into groups according to the "group number" they fell into nine groups. JNIost of these groups were defined by differences in the terminal reaction in glycerol and carbohydrate media. If these determinations are neglected all of the strains would have fallen into one group. A survey of the literature on thermophilic bacteria indicates that many of the strains there described have been superficially studied and that new strains have been named without sufficient data. Consequently many of the names which are used for thermophilic bacteria are being appUed to the same organism. Without exception the 52 cultures which were used in this study formed spores and in this characteristic seem to agree with most of thermophilic bacteria which have been described in the literature. This then seems to be the most common characteris- tic of members of this group. It has also been the basis for in- cluding among the thermophilic bacteria, bacteria which do not belong there. Spore formation when taken into consideration along with the pecufiar reaction to temperature makes the ther- mophilic bacteria a difficult group for canners of foods, for in- stance, to cope with. The abiUty to form spores allows the thermophilic bacteria to survive the process and perhaps to de- velop when the cans are stacked in the warehouse. The recent publications of Weinzirl, Cheney, Bigelow and Esty, and others have indicated the significance of these bacteria. They are also related to certain phases of the dairy industry. Fliigge, Leich- mann, Russell and Hastings have found that they are able to survive pasteurization temperatures. 358 LETHE E. MOERISON AND FRED W. TAXNER The forms isolated from water were Gram positive and in this characteristic, also, agreed with the forms described in the litera- ture. Their destructive nature is indicated by their action on the proteins in milk and on gelatin. Indol and hydrogen sulphide were produced in most media. Rabinowitsch (1895) reported the proteolytic abihties of the thermophiles to be their most characteristic function. Although none of the cultures studied fermented any of the sugars used, many of them did produce some acid in glucose or sucrose broth; this together with the fact that all of them showed diastatic action on starch would seem to indicate that these ther- mophiles decompose the more complex carbohydrate molecules more readily than they do the simple sugars. That many of the thermophiles decompose cellulose quite readily is seen in a review of the literature on those types that function in spontaneous heating during the fermentation of malt, tobacco, cotton, hay, and manure, the fermentation of silage and the decomposition of cellulose. The fact that some of these cultures came from water from quite deep wells and others from surface waters demonstrates that even in water the thermophiles can exist at widely varjdng ranges of temperature. They are probably widely distributed in waters and had larger portions of the samples of water been used for plating, no doubt, many more cultures would have been isolated from the 224 samples of water examined. The widely differing sources of the thermophiles described in the hterature also indi- cate that they are widely distributed in nature and further investi- gations of their temperature relations will explain this distribu- tion. Rabinowitsch (1895) and some others have claimed that this ability of thermophiles to grow at a high temperature was a property of adaptation to en\ironment. Blau and Bruini both claimed that many non-thermophiles had many of the same characteristics as thermophiles. Other investigators have pub- hshed data to support this claim. Bredfeld (1878) gradually developed the resistance of spores of Bacillus subtilis until it took three hours at 100°C. to kill them or five minutes at 110°C. Koch (1876) observed the germination of spores of Bacillus STUDIES ON THERMOPHILIC BACTERIA. I 359 anthracis and Bacilhis siibiilis that had been subjected to 123°C. in dry air. According to Arloing, Cornevin, and Thomas (1882) the spores of Bacillus anthracis-symptomatici would not resist boiHng for more than 2 minutes; but if previously dried, boiling for two hours was necessary to destroy them. This seems to agree with the theory of Davenport and Castle (1895) that by the loss of water, which is a necessary consequence of increased chemical activity resulting from warmth, organisms are able to increase their resistance to high temperatures. If we accept this view thermophiles are explained on the basis of adap- tation to environment. Tsikhnsky (1903) also believed that thermophiles were merely variations of common non-thermophilic microorganisms that had adapted themselves slowly to high temperature. He thought that the length of time necessary for these organisms to adapt themselves to high temperature detennined whether they were facultative or strict thermophiles. Schillinger proposed the term thermotolerant to be apphed to these organisms. Miehe tried to explain the liigh optimum tem- perature of thermophiles by assuming that they might have been brought over from the tropics and have adapted themselves to lower temperatures. By comparing them with other bacteria, Miehe came to the conclusion that all bacteria could be grouped on the basis of their minimum temperatures. He divided the thermophiles into two groups: (1) orthothermophiles with a maxi- mum temperature of 60°-70°C.; (2) thermotolerants with a maximum temperature of 50''-55°C. but which also grow well at ordinary temperatures. Many investigators, among them Rabinowitsch, Schiitze, and others, found a certain parallelism between temperature relations and the relation of thermophiles to oxygen. In most cases those organisms which had high optimum temperatures were strict aerobes. Bergey divided thermophihc bacteria into two groups: (1) true thermophiles, those that grow at temperatures above the maximum temperature for the great majority of bacteria, es- pecially the pathogenic forms; (2) facultative thermophihc 360 LETHE E. MORRISON AND FRED W. TANNER bacteria, those that develop at room temperature, about 20°C., and have their optimum temperature at about 50°C., and their maximum temperature at about 60°C. It would seem to be indispensable to fix clearlj' the limits within which the term thermophilic bacteria should apply. Some divi- sion evidently must be made in this group of organisms that grow at such widely differing ranges of temperature. The division made by Bergey into true thermophiles and facultative ther- mophiles seems to be the most tenable up to the present time. Further work on temperature relations of these thermophiles is being carried out in this laboratorj^ and perhaps when the data from this investigation are available a better differentiation will be possible. Of the 52 waters from which thermophiUc bacteria were isolated 44 (almost 85 per cent) were condemned for the presence of B. coll of fecal origin. Tliis fact suggests a possible sanitarj^ signifi- cance of thermophiHc bacteria in water analysis. The data are insufficient to draw any definite conclusions on this subject but it is a subject worthy of investigation. A similar suggestion was made by Brazzola (1906) when he stated that he thought the thermophiles were of very great importance in the study of the potability of water. VI. CONCLUSIONS 1. The aerobic thermopliilic bacteria studied in this investiga- tion seemed to make up a closely related group when the salient characters only are considered. 2. All strains form spores and are strongly proteolytic which, in connection with their temperature relations, makes them of importance in food preservation. 3. Thermophilic bacteria are widely distributed in nature (soil, water, etc.) and thus may cause serious losses in those industries where high temperatures are used for controlling bacterial development. 4. The abiUty of thermophilic bacteria to grow at high tem- peratures may be due to a particular property of the protoplasm (water content?). STUDIES ON THERMOPHILIC BACTERIA. I 361 5. Further investigations on thoir temporaturo relations may aid in a better understanding of the thermopliilic bacteria and in their separation into more sharply defined groups. This work is in progress in these laboratories. REFERENCES Ambroz, a. 1913 Denitrobactcrium thcrmophilum spec. nova, ein beitrag zur biologie der thermophilen baktericn. Cent. Bact. Abt. II, 37, 3. Ambroz, A. 1910 Uber das phanomen der therobiose bei den mikroorganismen. Cent. Bact. Abt. I, Ref., 48, 2.57 and 2S9. Anitschkow, N. N. 1906 Zur frage uber die rolle der thermophiles bakterien in darmkanal des menschcn. Cent. Bact. Abt. I, Orig., 41, 320 and 426. Arloing, Cornevin, et Thomas 1882 Compt. rend Acad. d. sc, 94, 189. Original not seen. Reviewed by Mace, E. 1912 Traite Pratique de Bacteriologie, v. I. Bardou 1906 fitude biochimique de quelques bacteriacees-thennoplules et de leur r6Ie dans la desintegration des matieres organiques des eaux d'cgovt. These de pharm.acie de I'University de Lille, 1906. Original not seen. Reviewed by Ambroz 1910. Barlow, B. 1912 A spoilage of canned corn due to a thermophilic bacterium. Thesis for Degree of Master of Science, University of Illinois. Benignetti, D. 1905 Di un germc termofilo isolato dai fanghi d'Acqui Riv. d'Igicne a Sanita pubbl. 1905 Original not seen. Reviewed in Cent. Bact. Abt. II, 14, 420. Beroey, D. 1919 Thermophilic bacteria. Jour. Bact., 4, 301. BlGELOW, W. D. AND EsTY, J. R. 1920 The thermal death point in relation to time of typical thermophilic organisms. Jour. Inf. Dis., 27, 602. Blau, O. 1906 tjber temporaturraaxiam der sporenkeimung. Cent. Bact. Abt. II, 15, 97. Brazzola, F. 1906 Significata die batteri termofili, di quella putrefazione e del gruppo coli, nell'esame batteriologico delle acque. Cent. Bact. Abt. II, 16, 582. Bredfeld 1878 Untersuchungen iiber die spaltpilze. Bacillus subtilis. Origi- nal not seen. Reviewed by Mace, E. 1912 Traite Pratique de Bacteriologie, v. I. Brewer, W. H. 1866 The presence of living species in hot and saline waters in California. Amer. Jour. Sc. and Arts, 41, 2nd. ser., p. 391. Bhuini 1905 tlber die thermophile mikroben flora dis menschlichen darm- kanals. Cent. Bact. Abt. I., Orig., 38, 298 BuRRiLL, T. J. 1889 The biology of silage. Agric. Exp. Sta. Univ. Illinois Bull. 7, 1889 p. 177-194. Original not seen. Reviewed in Exp. Sta. Red., 1, p. 200. Cambier, R. 1896 Resistances des germes bacteriens a la chaleur seche. Ann. de Micr., 8, 49. Catterina, C. 190i Contribution to the study of the thermophilic bacteria. Cent. Bact. Abt. II, 12, 353. 362 LETHE E. MORRISON AND FRED W. TANNER Certes, a., et Garrigou 18S6 De la presence constante de Microorganismes dans les eaux de Luchon, recueillies au griffon a la temperature de 64°, et de leur action sur la production de la baregine. Compt. rend. Acad. d. Sc, 103, 703. Cheney, E. W. 1919 Study of micro-organisms found in merchantable canned foods. Jour. Med. Res., 40, 177. CoHN, F. 1876 Untersuchungen uber bacterien. Beit. Biol. Pflan., 2, 271. CoHN, F. 1890 Ueber die warmeerzeugung durch schimmelpiize und bakterien, Vortrag gehalten etc. zu Brieg a. 15/6. 1890 Original not seen. Reviewed in Lafar, F. 1904-^ Handbuch der technischen mj'kol- ogie, V. III. GoHN, F. 1893 Ueber thermogene bakterien. Ber. Deutsch, Bot. Gessel, II, 66. Original not seen. Reviewed by M.\ce,E. 1912 Traite Pratique de Bacteriologie, v. I. Conn, H. J. et al. 1920 Report of the committee on the descriptive chart for 1919. Jour. Bact. 5, 127. Davenport, C. B., and Ca-stle, W. E. 1895 Acclimatization of organisms to high temperatures. Arch. f. Ent. Org., 2, 227. Davis, B.M. 1897 Vegetation of the hot springs of Yellowstone. Sc, 6, 145. DE Krutff, E. 1910 Les Bacteries thermophiles sons les tropiques Cent. Bact. Abt. II, 26, 65. DoNK, P. J. 1920 A highly resistant thermophilic organism. Jour. Bact., 6, 373. DupONT, C. 1902 Sur les fermentations aerobies du fumier de fenne. Ann. Agron., 28, 289. Original not seen. Reviewed in Lafar, F. 1901-6 Handbuch der technischen mykologie, v. III. Ehrenberg 1858 Monatsber, K. P. Acad. Wiss. Berling, 1858, p.448. Original not seen. Reviewed by Ambroz 1910. Falcioni, D. 1907 I germi termofili nelle acque del Bullicame. Arch. d. Pharm. Sperim, 1907, No. 1, Original not seen. Reviewed in Cent. Bact. Abt. II, 20; 164. Flotjrens 1846 Compt. rend. acad. d. sc, 22, 934. Original not seen. Reviewed by Tsilkinskt, Mlle. P. Ann. de I'lnst. Pasteur, 13, 788. FLtJQGE 1894 Die aufgaben und lesitungen der milchsterilisierung gegentiber den darmkrankheiten der sauglinge. Ztschr. f. Hyg., 17, 272. Origi- nal not seen. Reviewed by .A.mbroz 1910. Georgevitch, p. 1910 Bacillus thermophilus Vranzensis. Arch. Hyg., 72, 201. Georgevitch, P. 1910 Bacillus thermophilus Jivoini nov. spec, und Bacillus thermphilus Losanitchi nov. spec. Cent. Bact. Abt. II, 27, 150. Gilbert, Dr. 1904 l^eber Actinomyces thermophilus und andere actino- myceten. Ztschr. f. Hyg., 47, 383. Globig, Dr. 1888 Ueber Bacterien-wachstum bei 50°-70°C. Ztsch. f. Hyg., 3,294. Gorini 1895 Studi critico-sperimentali suall sterilizzazione del latte. Giornale della reale Scoieta d'Igiene 1895, No. 1. Original not seen. Reviewed by Ambroz 1910. Grieg-S.mith, R. 1921 The high temperature organism of fermenting tan-bark, Part I. Proc Linnean Soc New South Wales, 46, Part I, 76. STUDIES ON THERMOPHILIC BACTERIA. I 363 Hardin'o, H. a. 1896 Thesis for Degree of Bachelor of Arts, University of \Visconsin. Original not seen. Personal interview with author. Har.shbarger, J. W. 1897 The vegetation of Yellowstone hot springs. Amer. Jour. Pharm., 69, 625. Jaqer 1909 Die haktcriologie des taglichen lebens, Hamburg und Leipzig, 1909. Original not seen. Reviewed by .\mdkoz 1910. Karlinski, J. 1895 Zur kenntiss der bacterien der thermalquellen. Hyg. Rund., 6, 685. Kedzoir, D. 1896 Uebcr eine thermophile cladrothrix. Arch. Hyg., 27, 328. Kehlgr, W. 1904 Uber methoden zur sterilissation von erdboden und pflan- zensainen und uber zwci ncue thermoresistente bucterien. Disserta- tion Konigsberg, 1904. Original not seen. Reviewed by Ambroz 1910. Koch 1876 Untersuchungcn liber bacterien. Beit. Biol. Pflan., v. I. Original not seen. Reviewed by Mace, E. 1912 Traite Pratique de Bac- teriologie, v. I. Koch, A., .\n'd Hoff.maxx, C. 1911 Uber die Verschicdenheit der Tempera- turanspruchc thermophiler Bakterien in Boden und in kunstlichen Nahrsubstraten. Cent. Bact. .\bt. II, 31, 4.3.3. KoNiNG, C. J. 1897 De gistings onzer inlandsche tabak. Tijdschr voor toege- paste scheikuande en hygiene Deel. I, 1897. Original not seen. Reviewed by Ambroz 1910. KoNiXG, C. J. 1897-98 Hollandsche tabak I and II. De Natuur, 1897-98. Original not seen. Reviewed by .\mbroz 1910. Kossowicz, A. 1912 Mycological notes with reference to provisions. S. Landw. Versuchsw., 15, 737. Reviewed in Chem. Abstr., 7, 2441. Kossowicz, \. 1913 Occurrence of thermophilic bacteria. Deut. Zuckerind, 37, 1019. Reviewed in Chem. Abstr. 7, 2699. Kroulik, .\. 1912 Uber thermophile zellulosevergarer. Cent. Bact. Abt. II, 36, 339. Laxa, 0. 1898 Ueber einen thermophilen Bacillus aus zuckerfabriksprodukton. Cent. Bact. Abt. II, 4,362. Leichman 1894 Landw. Versuchsstationen. Bd. 43, 1894, S. 375. Original not seen. Reviewed by .A.mbroz 1910. McFadten, a., axd Blaxall, F. R. 1894 Thermophilic bacteria. Jour. Path, and Bact,. 3, 87. McFadyen, a., and Blaxall, F. R. 1896 Thermophilic bacteria. British Med. Jour., 2,644. MiCHAELis, G. 1899 Beitrage zur kenntnis der thermophilen bakterien. Arch. Hyg., 31,3. MiEHE, H. 1905 Anhang zu Heft III der .A.rb. der Deutsch. Landwirt. Ges., Berlin, 1905, S. 76. Original not seen. Reviewed in Lafar, F. 1905- 1908 Handbuch der Technischen Mykologie, v. II. MiEHE, H. 1907 a Die selbsterhitzung des Heues. Cent. Bact. Abt. II, 20, 295. MiEHE, H. 1907b Thermoidium suphureum n. g. n. sp. ein neuer Warmepilz. Ber. Deutsch, Bot. Gesel., 24, 510. Original not seen. Reviewed by Ambroz 1910. 364 LETHE E. MORRISON AND FRED TV. TANNER MiQUEL, P. 1879 Bull, de la Statisquemunicipalede Paris, Dec. 1879. Original not seen. Reviewed by Tsiklixsky, Mlle. P. 1899 Ann. Past. Inst., 13,788. MlQTJEL, P. 1881 Annuaires de I'observ. de Montsuris, 1881, 464. Original not seen. Reviewed by Mace, E. 1912 Traite Pratique de Bacter- iologie, V. I. MiQUEL, P. 1882 Les organismes vivants de ratmosphere. These de Paris. Original not seen. Reviewed by Mace, E. 1912 Traite Pratique de Bacteriologie, v. I. MiQUEL, P. 1888 Monographie d'un Bacille vivant au-dela de 70°C. Ann. de Micr., 1888, an. 1, 4. MiTOSHi 1897a Uber das maseenhafte Vorkoramen von Eisenbakterien in den Thermen von Ikao. Jour. Coll. Sc. Imp. Uni. Tokyo, 10, 139. Origi- nal not seen. Reviewed in Cent. Bact. Abt. II, 3, 526. MiTOSHi 1897 b Studien liber Schwefelrasenbildung und die Schwefelbakterien der Thermen von Yumoto bei Nikko. Jour. Coll. Sc. Imp. Uni. Tokyo, 10, 143. Original not seen. Reviewed by Sames, T. 1900 Ztschr. f. Hyg., 33, 313. Negre, L. 1913 a Thermophilic bacteria in the sands of the Sahara. Compt. Ren. Soc. Biol., 74, 814. Original not seen. Reviewed inChem. Abstr., 7, 2410. Negre, L. 1913 b Thermophilic bacteria of the waters of Figuig. Compt. rend. Soc. Biol., 74, 867. Original not seen. Reviewed in Chem. Abstr., 7, 2410. Noack, K. 1912 Beitrage zur biologie der thermophilen organismen Disserta- tion Leipzig, 1912. Opeescu, V. 1898 Studien iiber thermophile bacterien. Arch. Hyg., 33, 164. Patxschke, W. 1919 The resistance of bacteria to high temperatures and the use of the Lobeck Biorizator. Ztschr. f. Hyg., 81, 226. PoTjPE, Fr. 1898 Ztschr. f. Zucher. in Bohem, 22, 341. Original not seen. Reviewed by Lafar, F. 1904-1907 Handbuch der Technischen Mykologie, v. I. Pbingsheim, H. 1911 Uber die .\ssimilation des Luftstickstoffs durch thermo- phile Bakterien. Cent. Bact. Abt. II, 31, 23. Pringsheim, H. 1913 The fermentation of cellulose by the action of thermo- philic bacteria. Cent. Bact. Abt. II, 38, 513. Rabinowitsch, L. 1S95 Ueber die thermophilen bakterien. Ztschr. f, Hyg., 20, 154. Russell, H. L., and Hastings, E. G. 1902 A Micrococcus, the thermal death limit of which is 76°C. Cent. Bact. Abt. II, 8, 339. Sames, T. 1900 Zur kenntniss der bei hoherer temperaturen wachsenden bakterien and Streptothrixarten. Ztschr. f. Hyg., 33, 313. ScHARDiNGER 1903 Ueber thermophile bakterien aus verschiedenen speisen und milch. Ztschr. f . Unters d. Nahr. u. Genus., 6, 865. ScHiLLiNQER, A. 1898 Ueber thermophilen bakterien. Hyg. Rund., 8, 568. ScHLosiNO, T. 1889 Compt. rend. Acad. d. sc, 109, 835. Original not seen. Reviewed by Lafar, F. 1904-C Handbuch der technischen mykol- ogie, V. III. STUDIES ON THERMOPHILIC BACTERIA. I 365 PchlOsino, T. 1892 Contrilnition a I'etiKle dcs fermentations du femier Ann. Agron., 18, 5. Original not seen. Reviewed by Lafar, F. 1904-6 Hnndbuch der technischen mykologie, v, III. ScHt)TZE, II. 1908 Beitrage zur kenntnis der thermophilen Actionomyceten u. ihrer sporenbildung. Arch. Hyg., 67, 35. ScHWABE 1837 Uber die Algen der Karlsbader warmen Quellen. Linnaea 1837. Original not seen. Reviewed by Ambroz 1910. Setchell, W. A. 1903 The upper temperature limits of life. Sc, 17,934. SoNNERAT 1774 Observation d'un phenomenc singulicr sur des poissons qui vivent dans une eau qui a 09° chalcur. Jour, de Physique, 3, 256. Original not seen. Reviewed by Amdroz 1910. Teich, M. 1896 Beitrag zur kenntniss thermophiler bakterien. Hyg. Rund., 6, 1094. Tirelle, E. I termofili delle acque potabili. Riforma Medica, 10, 265. Origi- nal not seen. Reviewed in Cent. Bact. Abt. II, 19, 328. TsiKLlNSKT, Mlle. P. 1S9S O mikrobach jiurischich provisokich tempera tursch. Russ. Arch. F. Path., 5, 189S. Original not seen. Reviewed by Ambroz 1910. TsiKLlxsKT, Mlle. P. 1899 a Sur les mucedinees thermophiles. Ann. de rinst. Pasteur, 13, 500. TsiKLiNSKY, Mlle. P. 1899 b Sur les microbes thermophiles des sources ther- males. Ann. de I'lnst. Pasteur, 13, 788. TsiKLlNSKY, Mlle. P. 1903 Sur la Flore microbienne themiophile du canal intestinal de I'homme. Ann. de I'lnst. Pasteur, 17, 217. Van Tieghem, M. P. ISSl Bacteria living at high temperatures. Bull, de la Soc. Bot. de France, 28, 35. Original not seen. Reviewed in Jour. Roy. Micro. Soc, 1881, 778. Velich, a. 1914 t)ber thermophile mikrorganismen. Caspois ceskych. leka« rur, 53, 1026. Original not seen. Reviewed in Cent. Bact. Abt. II, 44, 174. Vernhout 1899 Onderzock van bacterien bij de fermentatie der tabak, Mede- deelingen uit s'Lands. Plantentium 1899, 34. Original not seen. Reviewed by Ambroz 1910. Weber, 1895 Die bakterien der sogenannten sterilen milch des handels use. Arb. a. d. Kaiserl, Gesund, 17, 108. Original not seen. Reviewed by Ambroz 1910. Weinzirl, J. 1919 The bacteriology of canned foods. Jour. Med. Res., 39, 349. WiNSLOW, C.-E. A., ETAL. 1920 The families and genera of the bacteria. Final report of the Committee of the Society of American Bacteriologists on Characterization and Classification of bacterial types. Jour. Bact., 6,191. WiTTLiN 1897 Bakteriologische Untersuchungen der Mineral-und Thermal- quellen der Schweiz; Ragaz-Pfafcrs. Cent. Bact. Abt. II, 3, 400. Addendum — text books Buchanan, F. D., and Buchanan, R. E. 1915 Household Bacteriology. The MacMillan Company, New York. 366 LETHE E. MORRISON AND FRED W. TANNER Chester, F. D. 1901 A Manual of Determinative Bacteriologj-. The MacMil- lan Company, New York. GiLTNER, W. 1916 Laboratory Manual in General Microbiology. John Wiley and Sons, New York. Hewlett, A. T. 1902 A Manual of Bacteriology. J. and A. Churchill, London. .Hiss, P. H., and Zinsser, H. 1920 A text-book of Bacteriology, 4th ed. D. Appleton and Company, New York and London. Marshall, C. E. 1917 Microbiology, 2nd ed. P. Blakiston's Son and Com- pany, Philadelphia. Morret, C. B. 1921 The Fundamentals of Bacteriology. Lea and Febiger, Philadelphia and New York. MuiR, R., AND Ritchie, J. 1913 Manual of Bacteriology. MacMillan Com- pany, New York. AN APPARATUS FOR THE RAPID MEASUREMENT OF SURFACE TENSION' ROBERT G. GREEN Department of Bactenology and Immunology, University of Minnesota Investigations carried on in these laboratories concerning the role of surface tension in certain bacteriological phenomena (Larson, 1921), have led to the development of an apparatus for the rapid determination of surface tension by the drop-weight method. After using the more common methods of surface tension measurement the belief is expressed that the most rehable and constant results for both pure and biological hquids can be obtained from their drop weights. The apparatus here described repre- sents a means of determining the surface tension of a liquid by the drop-weight method without any mathematical calculation or the determination of the drop weights and is termed a surface tension balance. The apparatus is designed primarily for rapiditj-- of measurement and an accuracy is obtained which is consistent with ordinary experimental conditions. In general the apparatus consists of three mechanical parts, a dropping pipette, a balance beam mounted on a torsion wire and an adjustable scale. The dropping pipette (A) differs in no way from those already in use and all the inaccuracies and corrections, which have been recorded for this method, occur but such errors may be mini- mized and reduced to an order of magnitude which is of no concern to the biologist, by using a dropping surface of the proper dimensions (Morgan, 1911). The rate at which the drops fall may be easily controlled by using a fine capillary tube for an air inlet protected by larger tubing (/). * Aided by a grant from the National Dental Research Association. Presented at Twenty-third Meeting of the Society of American Bacteriologists, December 26, 1921. 367 368 ROBERT G. GREEN The steel torsion wire (D) is tightly stretched between two binding posts and carries the balance beam, one end of which (5) swings over the face of the adjustable scale (F), the opposite end carrying a silver cup (C) swoing on watch jewels. The balance beam can be swung into the horizontal or zero position by means of the bar (H). The adjustable scale consists of a series of arcs of the same radius but the segments are of different lengths and each seg- ment is divided into the same number of units by cross lines. The scale face (F) is movable horizontally on the scale carrier {E) in the visible slot and is fixed in any position by means of a thumb screw behind. This mobiUty allows any one of the various sized arcs to be moved directly under the end of the pointer (B). The scale is constructed in relation to the size of the torsion wire so that the units of the scale represent sur- face tension in dynes per centimeter. The weight of a drop of hquid falling from the pipette (A) into the cup (C) will be proportional to the surface tension of the liquid according to the formula. where y = surface tension w = weight of drop K = constant of dropping pipette. The force producing torsion in the wire will be the drop weight times the acceleration of gravity and the amount of torsion pro- duced will be proportional to this force. The degree of torsion indicated by the pointer is therefore directly proportional to the surface tension of the liquid dropping from the pipette. The mobility of the scale makes it possible to di , ide any arc indicated by the pointer into a desired number of units by moving under the end of the pointer the proper sized scale arc and in this way an arc of torsion that is proportional to a surface tension of sixty dynes may be divided into sixty units and then each unit will represent a djme of force. The surface tension balance is caUbrated by means of a stand- ard liquid of high surface tension as follows : A drop of the liquid of known surface tension is dropped from the pipette into the cup. RAPID MEASUREMENT OF SURFACE TENSION 369 The pointer then swings into some position over the face of the scale. The scale face is then moved horizontally until the reading off the end of the pointer is the same as the surface tension of the liquid used in calibrating. The balance is then ready for use and if a liquid is now used which has one half the surface tension, the weight of the drop will be one-half, the torsion produced in the wire will be one-half and the reading on the scale will be one-half the previous reading. The sensitiveness of the apparatus varies with the mechanical construction depending upon the ratio of the mass of the balance beam to the torsion constant of the torsion wire. The number of drops of hquid used in determinations can vary from a single drop to many if a torsion wire of the proper size be chosen. Five drops have been found very satisfactory in our work. Evaporation is the greatest source of error with this apparatus. It cannot be used for very volatile Uquids. However, if cali- brated with water and used for watery solutions, the error from evaporation is rendered neglible in that under like experimental conditions, the same evaporation takes place in calibration that occurs in actual measurements. This has proven out in that repeated runs of watery solutions give identical results to the limit of readability of the scale. The temperature variation may be rendered negligible by controlling temperature or proper cor- rections may be made. The accuracy of the surface tension balance now in use is plus or minus one-tenth dyne. The pipette is the only part of the apparatus to be cleaned as the cup need only be dried and the pointer reset to zero. The pipette may be cleaned very rapidly if suction and compressed air are available. Since the time consuming weighings are eliminated the measurements may be taken with great rapidity. With this apparatus the surface tension of thirty solutions were measured in one hour. SUMMARY An apparatus termed a surface tension balance is described for the rapid measurement of surface tension by the drop-weight 370 ROBERT G. GREEN method. The apparatus consists of a dropping pipette, a deH- cate torsion balance and an adjustable scale upon which the surface tension is read directly in dynes per centimeter. REFERENCES Larson, W. P: 1921 Proc. Exp. Biol, and Med., 19, 62-63. Morgan-, J. Livixgstox: 1911 Jour. Amer. Chem. Soc, 33, 349-362. JOURNAL OF BACTERIOLOGY. VOL VII PLATE 1 m ^^ -\ 1 1 .^ \ A 1 ^ k F Y c H 1 (Green: Rapid Meoiiurcment of Surface Tension) JOURNAL OF BACTERIOLOGY Contents MARCH, 1922— Vol. VII— No. 2 F. C. Harrisok. 0>ir Society. Victor Buhkk. Notes on the Gram Stain witli Description of a Now Method. Barnett Cohe.v. Disinfection Studies. Tho KiTects of Temperature and Hydrogen Ion Concentration upon the V'iability of Bnct. coli and Bact. typhosum in Water. SEUiiAN' .\. Waksman. Microorganisms Concerned in the Oxidation of Sulfur in the Soil. I. Introductory. Selmax a. Waksman .a.nd J. S. Joffe. Microorganisms Concerned in the Oxida- tion of Sulfur in the Soil. II. Thiobaciilus Thiooxidans, a New Sulfur-oxidiz- inK Orpanism Isolated from the Soil. E. B. Khed and W. H. Peterson. The Production of Pink Sauerkraut by Yeasts. Constantino Gorini. Studies on the Biology of Lactic Acid Bacteria: A Summary of Personal Investigations. L. D. BusHNELi,. .\ Method for the Cultivation of Anaerobes. L. D. Bu.siiNEi.L. Influence of Vacuum upon Growth of Some Aerobic Spore-Bear- ing Bacteria. H. R. Baker. Substitution of Brom-Thymol-Blue for Litmus in Routine Laboratory Work . W. A. Wall and A. H. Robertson. The Use of Domestic Methylene Blue in Staining Milk by the Breed Method. JANUARY, 1922— VOL. VII— No. 1 Hilda Hempl Heller. Certain Genera of the Clostridiaceae. Studies in Patho- genic Anaerobes. V. O. IsHii. Studies upon Agglutination in the Colon-Typhoid Group of Bacilli. O. IsHii. A Study of Spontaneous Agglutination in tne Colon-Typhoid Group of Bacilli. Albert C. Hcnter. The Sources and Characteristics of the Bacteria in Decom- posing Salmon. S. A. KosER AND W. W. Skinner. Viability of the Colon-Typhoid Group in Carbo- nated Waier and Carbonated Beverages. Guilford B. Reed. A Binocular Microscope Arranged for the Study of Colonies of Bacteria. An Investigation of American Stains. NOVEMBER, 1921— VOL. VI— No. 6 George E. Holm and James M. Sherman. Salt Effects in Bacterial Growth. I. Preliminary Paper. Hilda Hempl Heller. Suggestions Concerning a Rational Basis for the Classi- fication of the Anaerobic Becteris. Studies in Pathogenic Anaerobes. IV. I. Preliminary Paper. J. Howard Brown. Hydrogen Ions, Titration and the Buffer Index of Bacterio- logical Media. J. E. Rush and G. A. Palmer. On Decreasing the Exposure Necessary for the Gelatin Determination. Harold Mact. Chart of the Families and Genera of the Bacteria. SEPTEMBER, 1921— VOL. VI— No. 5 G. P. Plaisancb and B. W. Hammer. The Mannitol-Producing Organisms in Silage. Hilda Hempl Heller. Principles Concerning the Isolation of Anaerobes. Studies in Pathogenic Anaerobes. II. John F. Norton and Mary V. Sawter. Indol Production by Bacteria. AuGCSTO IJoNAZzi. On Nitrification. IV. The Carbon and Nitrogen Relations of the Nitrite Ferment. Th. Thjotta and Odd Falsen Sundt. Toxins of Bact. Dysenterias, Group III. ORDER BLANK Williams & Wilkins Company Publishers of Scientific Journals and Books Baltimore, Maryland, U. S. A. Kindly enter my subscription for the Joornal of Bacteriology Volume VII, 1922, for which I enclose $5.00 (Canada, $5.25; foreign, $5.50), or, send me a state- ment and I will remit. Signed... Address. BACTO-PEPTONE FORMULA NOT ALTERED SINCE 1914 MEETS EVERY DEMAND FOR A STANDARD PEPTONE BACTO-PEPTONE, the original peptone designed to meet particularly the varied nutritional requirements of bacteria in culture, is the first peptone on record having a definite requirement as to the H-ion concentration in solution. BACTO-PEPTONE, correct in its chemical and biolog- ical properties, meets the varied requirements of a pep- tone for routine bacteriological^work, the world over. BACTO-PEPTONE is standardized so as to support vigorous INDOL production within 24 hours after inoculation. CARRI ED I N stock by the PRI NCI PAL DEALERS In Seiantino Supplies Specify "DIFCO" The TRADE NAME of the PIONEERS In the RESEARCH and PRODUCTION of DEHYDRATED CULTURE MEDIA and RARE SUGARS DIGESTIVE FERMENTS COMPANY DETROIT, MICHIGAN, U. S. A. VOLUME VII NUMBER 4 JOURNAL OK BACTERIOLOGY OFFICIAL ORGAN OF THE SOCIETY OF AMERICAN BACTERIOLOGISTS JULY, 1922 EDITOR-IN-CBIBP C.-E. A. W1N8LOW It is eharacUTislic oj Science and Progress thai they continxially open new fields to our visions. — PASTBtm. PUBLISHED BI-MONTHLY WILLIAMS & WILKINS COMPANY BALTIMORE, D.S.A. Enteiwd »• Meond-clut mMtcr April 17, 1910, at the^>ostoflice at Baltimore, Marylasd, under the aet of March 8, 1870. Aoecptance for mailinK at epecial rat« of poatan proTided for In Seotion 1103, Aet of October 3, 1817. Authoriaed on July 16, tail. Copyright 1S22, WiOiama and WiOdna Company v-rif not f 55.00 per volume, I'nited States, Mexico, Cub» rrice. nei , »;, ^^ ^„, volume, Canada ' volume, other countriee „„'• °f} < $b'.Zb per volume, Canada postpaid 1 55 fio ^gj. , Made in United States of America PEPTON Perfectly serviceable for the formulas and in all the technie of the bacteriological and antitoxin laboratory. It is employed in the usual proportions and for whatever purposes Pepton of this most desirable quality is required. Manufactured by Fairchild Bros. & Foster New York BIOLOGICAL STAINS A Complete List for Bacteriology, Pathology, Zoology, Botany, Etc. CHEMICAL INDICATORS Indicators for Every Purpose, including the Hydrogen-Ion Indicators (Clark and Lub's List, Sorensen's List, Etc) FINE ORGANIC CHEMICALS Especially Prepared for Laboratory Use. \\'ritc for oar NEW CATALOGUE giving the entire list, together with Methods for Staining, Solubilities, Indicator Chart, Etc. Coleman & Bell Products may be secured from Laboratory Supply Houses through- out the United Slates and in Foreign Countries, or may be ordered direct. The Coleman & Bell Company, Norwood, Ohio, U. S. A. Formerly National Stain & Reagent Co. JOURNAL OF BACTERIOLOGY OFPICIAL ORGAN OF THE SOCIETY OF AMERICAN BACTERIOLOGISTS DEVOTED TO THE ADVANCEMENT AND DIS- SEMINATION OF KNOWLEDGE IN nEGAKD TO THE BACTERIA AND OTHER MICRO-ORGANISMS Editorial Board EditOT-in-Chie) C.-E. A. WINSLOW Yale Medical School, New Haven, Conn. A. Pabker IIitchens Lore A. Rogers, IOx officio Advisory Editors C. C. Bass F. P. Gorham C. E. Marshall M. J. Rosenatj R. E. Bdchanan F (":. llAiiRisoN V. A. Moorb A. W. Williams P. F. Clark E. O. Jordan L. F. Rettger IT. Zinsser F. P. Gay C. B. Lipman L. A. Rooers CONTENTS L. D. BuSHNELL. Quantitative Determinations of Some of the Biochemical Changes Produceci by a Saprophytic Anaerobe 373 G. S. Wilson. The Proportion of \iablc Bacteria in Young Cultures with Especial Reference to the Technique Employed in Counting 40.5 H. J. Co.VN. A Method of Detecting Rennet Production by Bacteria 447 Abstracts of American and foreign bacteriological literature appear in a separate journal, Abstracts of Bacteriology, published monthly by the Williams & Wilkins Company, undei the editorial direction of the Society of American Bacteriologists. Back volumes can be furnished in sets consisting of Volumes I, II, III and IV. Price per set, net, postpaid, $24.00, United States, Mexico, Cuba ; S2.5.00, Canada; $2').00, other countries. Subscriptions are in order for Volume V, 1921. Price, per volume, $5.00, United States, Mexico, Cuba; $5.25, Canada; $5.50. other countries. JOURNAL OF BACTERIOLOGY A SIMPLE ELECTROMETRIC OUTFIT FOR THE MEASUREMENT OF H-ION CONCENTRATION IN SOLUTIONS Operating on the compensation principle but without a potentiometer No. 4886. Hydrogen-ion Outfit, with diagrammatic illustration of connections. A millivoltmeter system — operating on the compensation principle but mthout a potentiometer — for the measurement of H-ion concentrations in solutions and, with the substitution of a plain plati- num or other unattackable electrode in place of the hydrogen electrode, for the measurement of o.xidation and reduction reactions. For the location of endpoints or variations in H-ion concen- tration, the salt bridge shown in the illustration can be omitted. The arrangement is that of Hildebrand (see Journal of the American Chemical Society, 35: S47 , 1913) with a L. & N. d'Arsonval Galvanometer of the pointer type substituted for the capillary electrometer. A stirring device of some convenient form (not shown in illustration) should be used with the outfit. T he components of the outfit, which are sold separately under their respective catalogue numbers are as follows: — 4852. Hildebrand Hydrogen Electrode $5.75 4834. Calomel Electrode Vessel 4.00 4044. Single Contact Key 3.00 4800. Portable d'Arsonval Galvanometer, I.. & N 18.00 8714. Rheostat, 920 ohms 8.00 9810. Weston D. C. Millivoltmeter, 0-1200 millivolts 18.75 8741. Coil, 10,000 ohms, mounted 10.00 9373. Single Pole Single Throw Switch 65 2410. Paired Burettes, ,^0 ml in 1/lOths, for titrating , $9.00 9336. Support, large _ , 1.00 3222, Clamps, (3) ea. .40 2090. Burgess Super-sijc Dry Cells (2) ea. 1.00 9357. Support 1.50 3218. Clamp 70 2128. Beakers, Pyrex Glass, 300 ml (2)....ea. .26 4876. Bent Glass Tube is 4886. Complete Outfit, as above listed $84.25 Code Word Ezkci Prices subject to change without notice ARTHUR H. THOMAS COMPANY WHOLESALE, RETAIL AND EXPORT MERCHANTS LABORATORY APPARATUS AND REAGENTS WEST WASHINGTON SQUARE PHILADELPHIA, U. S. A. CABLE ADDRESS. "BALANCE." PHILADELPHIA LIBRARY NEW YOKK UdTAMCAL fiAKUisti QUANTITATR^ DETERMINATIONS OF SOME OF THE BIOCHEMICAL CHANGES PRODUCED BY A S.\PROPHYTIC ANAEROBE' L. D. BUSHNELL Kansas Agricultural Experiment Station Received for publication December 3, 1921 INTRODUCTION The differentiation and classification of microbial species is certainly the most difficult of all problems in bacteriologj\ "^Tien one finds what he considers a new species he is at once aware of the lack of means to characterize it adequately so that it may be recognized bj^ others. At the same time the similarities that it presents to other types, which are certainly distinct, may lead to confusion, when an attempt is made to place it properly among species already described. This confusion is more marked in the case of the anaerobic than of the aerobic types. This is prob- ably due to the fact that anaerobic types are supposed to be very unstable in their characteristics and less care has been used in describing them; and also to the fact that many mixed cul- tures of anaerobes have been described as pure cultures. Probablj' some of this confusion is due to the fact that too much emphasis has been placed on one factor, such as the form of the colony, or the location and size of the spore. It is a well known fact that colony formation is quite variable, in many cases being completely altered by the consistency of the medium. Again the location of the spore varies in many of the individuals of the same culture. Bacteriological experience has taught us not to rely too much upon one factor, but rather to classify microorganisms accordmg ' Contribution no. 41 from the Bacteriological Laboratories of the Kansas Agricultural Experiment Station. 373 lOCrBNAL OP BACTEBIOLOGT, VOL. Til, NO I CSJ CD 374 L. D. BUSHNELL to a group of factors. The fermentation of both carbohydrates and proteins should be considered, and a certain amount of cultural and morphological variation should be allowed for each group. DESCRIPTION OF THE ORGANISM USED IN THE PRESENT STUDY In the study of anaerobic bacteria isolated from spoiled canned asparagus we have found one type which predominated and which could be disting-uished by colony formation in the deep agar column. This organism was a Gram positive, spore-bearing, anaerobic rod; motile, by means of 12 to 15 peritrichic flagella. The size of the majority upon plain agar was 0.80 to 0.90 by 2.50 to 3.50 microns. They were somewhat longer and more slender upon glucose media. The surface colonies are very thin and transparent at first, but gradually assume a more opaque granular center from which the surface growth extends over the medium for several milli- meters. The edge is quite regular in outline, sUghtly raised from the medium. The colonies later become more irregular with root- like projections extending from the edge of the colony while similar projections extend down into the medium below the colony. These colonies usually show a central nucleus which at first is more opaque than the remainder of the colony. The colonies are transparent but in most cases there is a whitish appearance in older cultures. Under low magnification the surface colonies appear to be finely granular but remain nearly colorless. The deep colonies are spherical at first and more opaque than those on the surface. They later develop root-Uke outgrowths or may become quite woolly in some instances, especially if the medium is soft. Older colonies tend to become somewhat less opaque as they increase in size. Some of the colonies are some- what lenticular in outline and have woolly or root-Uke outgrowths from one side only. This appearance is described as "Eii gre- nade" by Weinberg and Seguin (1918). Some of the cultures of this group appear ragged in outline, not having distinct projec- tions but with a very irregular surface. CHANGES PRODUCED BY A SAPROPHYTIC ANAEROBE 375 The organism would be considered to be proteolytic in its action upon protein-containing media. Gelatin is liquefied rapid- ly and completely. Inspissated serum is rapidly cleared, fol- lowed by nearly complete digestion. In alkaline meat media there is a rapid growth with the development of much ga.s and a distinct!}^ unpleasant but not putrefactive odor. There is slight digestion of the meat at the surface. This, however, was not marked in any of the cultures. The surface layers of the meat are darkened, while the lower layers retain their original color. There is a heavy, whitish growth of bacteria on the surface of the meat in very old cultures. About 1 or 2 mm. below the sur- face is a blackened zone. Traces of hydrogen sulphide are found in meat cultures. The liquid remains neutral or slightly acid and the particles of fat on the surface appear to be partially dis- integrated. Growth in Von. Hibler's brain media is very similar. Growth in milk is rapid. There is very little or no gas formed. The casein is thrown down in a soft clot which is usually com- pletely digested within four days. With some of the cultures, however, digestion is never complete. The whey separates and in a short time becomes of a clear light amber color, or yellowish and turbid, according to the culture. The reaction is neutral or sUghtly acid on standing, and the fat layer becomes disintegrated, probably due to saponification. The cultures have a strong cheesy to butyric acid odor. Some of the cultures are somewhat more putrid and upon long standing, especially if enclosed in a jar, there is a sUght odor of ethyl-but\Tic or of valerianic acid. Cultures seem to differ considerably in this respect with age. In the egg medium, made according to the formula given by Robertson (1916), the organism grows slowly. The first indica- tion of growth is the appearance of blackening on the bottom of the tube. This is followed in some cases by the formation of a soft, yellowish clot which becomes fissured and may settle out, leaving a clear, yellowish, transparent fluid. Some of the cultures do not cause coagulation even on long standing, and those cul- tures which form a clot, fail to produce further change, except in a few cases, in which there is a marked shrinking of the clot. If the medium was made sUghtly acid nearly all the cultures caused 376 L. D. BTJSHNELL the formation of the clot, with darkening at the bottom of the tube; but aside from the shrinking no other change occurred. Indol, phenol, skatol or alcohol were not found. There was a slight reduction of nitrates to ammonia. The pH as determined colorimetrically in Clark and Lubs media ranged from 5.8 to 6.2. Cultures were all negative for the methyl-red and the Voges- Proskauer reactions. Considerable difficulty was encountered in determining which of the carbohydrates were actually fennented. The work was first done by the use of Durham's fermentation tubes, using meat infusion broth made sugar free by fermentation with B. saccharo- lyte. This medium was placed in tubes and autoclaved. To it was added 1 per cent of the carbohydrate to be examined and the tubes were incubated for the detection of contamination. Just before inoculation they were heated and cooled rapidlj^ in running water. They were then inoculated with a large loopful of a four day potato-peptone culture of the organism. The cul- tures were placed over phosphorous in an anaerobic jar and incubated for three days at 37°C. Difficulty was sometimes encountered, however, in determin- ing the fermentation of such carbohydrates as lactose, inulin, etc., due to the fact that the closed arm of the tube would contain a bubble of gas from one to three millimeters in diameter. The reaction of the medium in these cases was but little changed, being sometimes made slightly acid. In other cultures, however, distinct amounts of acid were formed. Hiss's serum water plus these carbohydrates gave similar results. Table 1 shows the essential features of these tests. The " -|-" indicates a distinct amount of acid or gas. If a bubble of gas is present it is marked with a "b." From our preliminary tests it was decided that we were dealing with organisms belonging to the B. sporogenes group (JVIetchnikoff , 1908). There were differences between the various cultures, but these were not marked. Some cultures alwaj's produced a cloud- ing of the medium, others grew mostly on the bottom of the tube. Certain cultures produced more blackening of the egg medium than others. Generally speaking all the cultures exhibited the same characteristics. n Q .2 > 3 O t-H "o "o o o "o o b£ o Irregular not woolly o •no| 1 1 1 1 1 1 •H 1 + + + + + '00 + + + + + noHooiv 1 1 1 1 1 ai3v oiusnvA ^* ^ 1 aiov 3iuxj.aa ^ + + + aiov Diuav ^ + + + VINOKHV ^ + + + + + ioxva« 1 1 1 1 1 1 TONSHJ 1 1 1 1 1 1 S'H ' 1 Ji 1 1 1 aeoiAX t. 1 ^ •^ 1 1 1 1 1 fl 1 1 1 1 1 1 1 1 1 afiONiavuv « 1 -o 1 1 1 1 -O 1 a i 1 1 1 1 1 1 1 1 BOUVIS »| + + + + + Xl « 1 + + + + + 1 NionvB u 1 + + + + + Xi .| + + + + ■4^ + loxioaaa in s> J3 jQ 1 6? 1 a 1 1 1 1 1 1 1 1 TOXINNVK » + + + + + XI 1 = i 1 1 1 noaaono " + + + + + + 1 « + + + O- *3 + 1 3eox^vre « + + + + + + + « + + + + + + + aeoJJVT » -^ ^ ^ 1 XI XI 1 => 1 1 1 -4^ 1 1 1 asoaaas 60 ^ XI -D ^ J2 1 o 1 1 1 1 1 1 1 I asoiovnvo " + + + + + + 1 a + + + + + + 1 aeoaiiAaT « + + + + + + + = + + + + c^. CW. + aeoomo «| + + + + + + + « 1 + + + + + + + NiTaa JO DNinasovaa ^ ^ i i3 a + xvare ao ONiNaaovia ^ + + xiaavo dO Koixsaoia _}_ + + + + + + KQaae so NoixvoMaaOn _j- + + + + + Nixviao ao KoixvDiaaobn -|- + + + + + + eauods ao moixvoot CO .i-3 ■*J iiniioK + + + + + + + Kivxs iMvao + + + + + + + Kaoj T3 o •a -a o o T3 O o 8RSINT Dao Da o lO CO IN D. M t-H o c to HH . »— t C5 o D,i-i to HI . (— I ca B. spor. Henry (4) 377 378 L. D. BUSHNELL All the cultures were completely non-toxic when grown on brain and liver broth and on brain — saline for different lengths of time and injected subcutaneoush^ into guinea-pigs in 1 cc. amounts. Table 2 shows the results of one series. TABLE 2 Residts of animal inoculation cntTnBE AGE OFcni/rnBE DATS TO DEATH OF ANIMAL days 2B 6 Found dead on 30th day (colon bacilli from heart blood) 5 B 16 Alive after 6 weeks 28 B 23 Alive after 6 weeks 37 B 34 Alive after 5 weeks TABLE 3 Agglutination reaction RABBIT IMMU- NIZED CULTURE DILUTIOX OF IMMUNE SERUM KOBMAI. SEBUM CCLTUBE 1::5 1:30 1:100 1:200 1:400 1:800 1:1600 1:3200 1:50 1:200 2B 2B *10 10 10 10 10 10 5 0 2 0 5B 0 B 8 10 10 10 10 10 8 2 0 0 28 B 28 B 10 10 10 10 10 10 8 0 0 0 2 B 5 B 10 10 10 10 10 10 0 0 •1 0 2B 28 B 10 10 10 10 10 5 0 0 2 0 5B 2 B 10 10 10 10 10 10 10 3 3 0 5 B 28 B 10 10 10 10 10 8 0 0 2 0 28 B 2 B 10 10 10 10 10 10 3 0 0 0 28 B 5B 10 10 10 10 10 10 2 0 1 0 5B B. spor. I 10- 10 10 10 10 9 3 0 3 0 5B B. spor. II 8 10 10 10 10 8 0 0 2 0 5 B B. bot. II 10 10 10 5 0 0 0 0 2 0 * Figures show amount of agglutination. 10 = complete. Several of the cultures of this organism were also serologically similar to known cultures of B. sporogenes, as showTi by table 3. Their non-toxic nature and the appearance of colonies in deep agar would differentiate them from B. hotulinus, to which species they are otherwise very smiilar. CHANGES PRODUCED BY A SAPROPHYTIC ANAEROBE 379 QUANTITATIVE ANALYSIS OF BIOCHEMICAL CHANGES Since the qualitative results left us in doubt concerning the fermentatiun ability of these organisms, we attempted to deter- mine these points by quantitative tests. Wolf and Harris (1917a; 1917b; 1918) and Wolf and Telfer (1917) and Wolf (1918; 1919) have conducted exhaustive experi- ments upon the biochemistry of pathogenic anaerobes from war wounds, using quantitative methods. These authors did not attempt to determine the ability of these organisms to ferment many of the carbohydrates ordinarily used for the separation of microorganisms into groups, and studied only those found in wound infection. They did, however, collect some very valu- able information concerning some of the activities of a large number of the more common pathogenic anaerobes. We hoped, by tests of tliis sort, to determine some more accu- rate and rapid method for the classification of anaerobes in general, and those in particular with which we were working. Since the dangers of botulism from the eating of canned foods is more prev-alent than formerh^ it is quite essential that bacteriol- ogists should obtain all the information possible concerning the groups of spore-bearing anaerobic bacteria which may be found in food. As has been mentioned above, the classification of anaerobes is very much confused at the present time; and since their fermentative ability, as determined by the ordinary quali- tative methods is difficult to determine accurately, the only ap- parent method of obtaining real differences is by quantitative methods. We have, therefore, attempted to utilize the following quanti- tative indications of difference: (1) Amount and kind of gas produced from various carbohydrates; (2) amount and kind of acid produced from various carbohj^drates ; (3) amount of pro- teolytic action as determined by formation of ammonia and amino-acids. The first condition to be fulfilled is to obtain a medium which will be as constant as possible in composition, and one in which organisms of this type are able to grow. The second condition 380 L. D. BUSHNELL is to obtain an apparatus by means of which the various changes may be noted and recorded over a series of days, since the fermen- tations change with the age of the culture to a marked extent. We hoped to obtain more accurate information by this method although the amount of work necessary is multiphed in proportion to the number of determinations made. Of course, it is impossible under all conditions to make the medium absolutely the same. However, we have attempted to do this as far as possible, by obtaining enough material in one lot for an entire series of tests. The same asparagus was used throughout; and for the carbohydrate fermentation tests, we used a 2 per cent peptone-water to which, after sterilization, was added 1 per cent of the carbohydrate in question, except in case of glucose of which 0.5 per cent was used. We used peptone- water rather than broth because this could be made more uni- formly, and did not require an adjustment of the reaction. Un- fortunately these organisms would not grow in any of the mineral solutions usually recommended for the culture of nitrogen fixing anaerobes, so their action in a medium of known composition could not be determined. In their experimental work Wolf and Harris used a large 2-liter bottle for culturing the anaerobes. Tliis bottle was fitted with a two-holed rubber stopper through which passed two glass tubes, one of the tubes extending into the space above the liquid and one into the liquid itself and nearly to the bottom of the bottle. The container was exhausted and the culture allowed to develop in vacuum until the space above the liquid was filled with gas. Samples were removed at intervals by means of the tubes. The presence of the gases was utiUzed to force out enough hquid for an analysis. The apparatus described by these investigators could not be used by us, since our cultures did not grow very rapidly, and we invariably found that a leak had occurred and air had entered. Also, we experienced difficulty with contamination, in sampUng according to their method. Aside from the difficulties mentioned, the apparatus recommended by Wolf and his associates would probably give more uniform results for a single series of examina- CHANGES PRODUCED BY A SAPROPHYTIC ANAEROBE 381 tions than the one used in this work, since but one culture con- tainer wa-s used and all the samples taken from it. Results could not be exactly duplicated, however, by the use of this apparatus. All our cultures were placed in a vacuum as complete as we could obtain. Great difficulty was experienced in avoiding; leaks, and in obtaining a culture vessel which would withstand steriliza- tion and still be strong enough not to be crushed under the high external pressure. At the time this work was done, special glass- ware could not be purchased on the market and "pure" gum rubber tubing that would hold a vacuum was not obtainable. Numerous samples of so-called pure gum rubber, were secured, but none of them would hold a high vacuum for twenty-four hours, although they were coated with celloidin and various cements recommended for this purpose. We desired enough cul- ture material for several tests and this required a large culture vessel. DESCRIPTION OF THE APPARATUS The following apparatus (fig. 1) was finally devised and met these difficulties veiy well. A quart milk bottle (N) was used as a culture vessel. To this was added from 300 to 500 cc. of the culture medium and the bottle was then autoclaved at 20 pounds pressure for thirty minutes. The asparagus was merely suspended in tap water, 100 grams in 300 cc. of water. Fresh whole mixed milk was used. The potato medium was made by adding 100 grams of potato to 300 cc. of water. The potato was carefully selected, washed, peeled and passed through a meat grinder; and the potato pulp was washed in running water for several hours. For the carbo- hydrate fermentation tests a 2 per cent peptone solution was used. Three hundred cubic centimeters of this solution were placed in the bottles and sterilized. The carbohydrates were sterilized in 10 per cent solution and added at the time of inoculation. As an inoculum, we used 1 cc. of a four day old potato-pep- tone-water culture of the organism grown at 37°C. The four day culture seemed to give the best results. At first smaller amounts 382 L. D. BUSHNELL were used but in many cases we failed to obtain growth. Care was always used not to add bits of potato. Just before inoculation, the bottles of media were heated in the steamer and then cooled as rapidly as possible to drive out the air. Care was always exercised not to shake up the medium during the inoculation and sealing process. After inoculation, the cotton stoppers were removed and the bottles were plugged with a No. 8, two-holed rubber stopper, fitted with two glass tubes. One of these tubes was constricted near the upper end and served as an attachment for the vacuum pump, and for sealing after the container had been exhausted. The second tube was bent to serve as a manometer tube (P) as indicated in figure 1. The rubber topper was fitted with the tubes, wrapped in gauze and then covered with a thick layer of cotton. The lower end of the manometer tube and the upper end of the straight tube were plugged with cotton and all autoclaved at twenty pounds pressure for an hour. It is necessary to wrap the cotton stopper in the mouth of the bottle, and the rubber stopper with gauze to prevent bits of cotton adhering to them, since this would prevent a perfect seal. After inoculation, the cotton stopper was discarded, and the rubber stopper carrying the tubes, was unwrapped and well flamed with a Bunsen flame. The rubber stopper was then pushed firmly into the neck of the bottle and covered with a thick layer of rubber cement. Thin blocks of wood were placed on the top of the stopper, and all wired in place with No. 16 u'on wire. After the bottle had been fitted, it was placed in a special incu- bator at 37°C. and immersed in a copper tank (AI) containing heavy mineral oil. The lower end of the manometer tube was placed in a dish containing mercurj', after removal of the cotton plug. The bottle was then exhausted as much as possible with a Geryk pump. When the mercury column ceased to rise in the manometer tube the tube attached to the pump was sealed at the constriction with a blast lamp. Bj^ this method we could obtain practically a complete vacuum as determined by the temperature and barometric pressure at the time. Figure 1 shows the details of this apparatus. CHANGES PRODUCED BY A SAPROPHYTIC ANAEROBE 383 A series of cultures was prepared in exactly the same way and an analysis made at intervals as indicated in table 4. METHODS OF ANALYSIS Gas analysis. {Total volume.) The amount of gas could be calculated from the manometer readings and reduced to normal temperature and pressure by using the following formula : 273 F(P-r) ~ 760 (273 + t) in which V = volume, P = barometric pressure in mm. Hg., T = vapor tension of watcr,= 760 = normal pressure, 273 = abso- lute temperature. Determinations of the total volume of gas were recorded upon the bottle which remained in the incubator for the longest time. Manometer readings were taken on all the cultures at each interval but not all of these are included in the tables, since they would give no additional information as to the course of the fermentation. Method for collection and analysis of gas. To obtain samples of gas, uncontaminated with air, the mercury pump was attached to the slender sealed tip (D) shown in figure 1. A tight fitting rubber stopper was first placed over the sealed tip and the rubber connection forced carefully over the tip until the metal tube came well down over the tip. This was wired firmly in place. Next the large rubber tube, which served as a receptacle for mercury, was drawn down and fitted over the stopper and wired in place. This large tube was filled with mercury to cover the connections completely. The mercury pump was filled with mercury as completely as possible and the tube leading from the mercury pump to the con- nection with the culture container was exhausted with the Geryk pump. When the entire apparatus was exhausted, the connec- ' The vapor tension of water was used in all cases. This probably varied to some extent as the fermentation proceeded and new volatile products were formed. From the fact, however, that these were of unknown value, it was pos- sible to use only the vapor tension of water in our calculations. 384 L. D. BUSHNELL tion at D was given a slight bend to break the sealed tip. (If the tip has not been sealed too close to the lower part of the con- striction it is very easily broken.) When the tip is broken there is an immediate fall of the mercury in the pump and a correspond- ing rise of the mercury column in the manometer tube. The gas is pumped out of the culture bottle and collected in a container over the mercury pump. The pumping is continued until the mercury column in the manometer tube becomes stationary. -'=,»•"-/ • ^/"V Fig. 1. Showing Arrangement of Cultures in the Incubator Fig. 2. Showing Details op the Connection Between the Cul- titre Bottle and the Mercury Pump This is usually very close to the theoretical vacuum, and required about fifteen or twenty minutes. By this means it was possible to obtain the dissolved gases as well as those which had collected above the liquid. The liquid boiled vigorously at 37°C. and the water vapor probably aided in washing the gases out of the bottle. After the gas had all been forced into the collecting bottle the connection between this bottle and the gas analysis apparatus CHANGES PRODUCED BY A SAPROPHYTIC ANAEROBE 385 was opened and the mercury in the gas burette lowered. By this means a sample of gas was drawn in, measured and analysed. Figure 2 shows the details of the connection made at (D) figure 1 . The efficiency of our apparatus and pump was judged by the amount of oxygen in the gas obtained for analysis. In a few analyses we obtained a slight reduction in volume after absorbing the gas with alkaline pyrogallate solution. In such a case we calculated the air in the gas by multiplying the amount of de- crease by five and making corrections in the calculation. These amounts were, however, usually nearly within the error in reading our burette. The apparatus used for the analysis of the gas is described by Burrell and Piebert (1913) and is a slight modifica- tion of that described by Haldane. Calculations. The following example will show the calcula- tions necessary in this connection. These were used throughout and for that reason the figures for each analysis are not included. 21.65 cc. of gas taken 5.63 cc. after CO2 absorptioa 16.02 cc. of CO, 5.63 cc. after Oj absorption 0.00 cc. of O2 18.45 cc. of CO2 free air added 24.08 cc. total volume 16.85 cc. after combustion 7.23 cc. decrease due to combustion 7.19 cc. after CO2 absorption 0.04 cc. of CO2 due to combustion 4.82 cc. of hydrogen (| of decrease) 73.99 per cent CO2 22.26 per cent Hj 3.71 per cent residual gas (by measurement) The 0.04 cc. of CO2 due to combustion was disregarded, since this amount is insignificant. After the completion of the gas analysis, air was allowed to enter the apparatus through a cotton plug. The bottle was then removed from the incubator and the examination completed. The bottle was opened as carefully as possible and a sample taken for cultural purposes. From this sample stains, shakes and plates were made. The cultures were incubated several days at 386 L. D. BUSHNELL 37°C. From the shakes we could detect the presence of anae- robes, and plates showed the presence of aerobes. After we be- gan the use of the present apparatus no contamination has been noted and we were always able to culture the anaerobe if growth had occurred. Samples were also taken for the determination of ammonia, amino-acids and volatile fatty acids. Ammonia determinations. The ammonia determinations were made by distilUng in the presence of magnesium oxide. At first we tried the aeration method but even after eight hours we were able to obtain tests for ammonia with Nessler's reagent. After the gas analysis was made there was insufficient time for the aeration . It has been found in this laboratory, in working with fermenting mixtures of soil, etc., that there is a close correlation between duplicate samples run by the aeration and distillation methods. With small amounts of ammonia, the distillation methods, usually gave somewhat higher results than the aeration method; while with large amounts of ammonia present the results were reversed. When small amounts of ammonia (10 to 20 mgm. per 100 grams soil) were present the distillation method gave 3.G per cent more ammonia than the aeration method. "N^Tien large amounts of ammonia (60 to 75 mgm. per 100 grams soil) were present the aeration method gave 4.0 per cent more ammonia then the distil- lation method. Amino-acid determination. Amino-acids were determined by the Van Slyke method. These determinations were complicated by the presence of very large amounts of ammonia. Van Slyke (1911) recommends the following method for the removal of ammonia: The ammonia can be removed bj' distillation with Ca(0H)2 under diminished pressure, using a 10 per cent suspen- sion until a slight excess is present, as shown by turbidity and alkaline reaction of the solution. The apparatus is evacuated to 30 mm. or less of mercury, and distillation continued for one- half hour at 45 to 50°C. Alcohol is added to prevent foaming. We were unable to use this method because of lack of time and the difficulty of removing all the ammonia from the medium, As large amounts of ammonia were present and the temperature CHANGES PRODUCED BY A SAPROPHYTIC ANAEROBE 387 of the laboratory was high (usually 25 to 35°C.), a correction for ammonia was necessary. The following method was used in making these corrections. A 2 per cent peptone solution containing known amounts of standard ammonium hydroxide was placed in the Van Slyke apparatus and the amount of nitrogen determined in the usual manner. The difference between the peptone solution alone and a peptone solution plus ammonia was considered to be due entirely to ammonia. Van Slyke states that about 15 per cent of the ammonia nitro- gen present will be determined as amino-acid nitrogen at 20'C. in fifteen minutes. We obtained the following percentages by averaging the results obtained from duplicate determinations made on different dates. Per cent of ammonia nitrogen that would be recorded as amino- acid nitrogen at different temperatures for the time used (five minutes). At21°C. = 34.2 per cent At23°C. = 39.6 per cent At 25°C. = 44.9 per cent At 27°C. = 50.6 per cent At 29°C. = 55.7 per cent At31°C. = 61.1 per cent At 33°C. = 66.3 per cent At35°C. = 71.6 per cent C. = 77.0 per cent At 37 This means that the above per cents of the total ammonia nitrogen present would be recorded as amino-acid nitrogen at the various temperatures recorded. It will be noted that these figures are considerably higher than those recorded by Van Slyke. The results obtained give a fairly uniform curve and were obtained under our experimental condi- tions. Of course, they were obtained by the use of a compara- tively simple mixture, while in actual tests we were deaUng with an extremely complex mixture. However, all the fermentation mixtures were different and it would have been impossible to simulate them closely. TABLE 4 Showing results of chcTnical examination ACTION ON ASPAHAGC8 (100 GRAMS IN 300 OC. WATER) ACTION ON AiKALINE EGO MEDICU PRODUCTS FORMED Days incubated Days incubated 2 4 6 8 0 5 s 12 Total cubic centi- meters of gas. . . 148.0 251.0 257.0 280.0 66.0 77.0 102.0 112.0 Daily cubic centi- meters of gas . . . 74.0 51.0 3.0 11.0 33.0 4.0 8.0 2.5 Per cent of CO2.. . 98.6 97.2 97.1 95.3 64.9 71.3 70.1 73.7 Per cent of H2. . . . 0 0 0.5 3.3 26.4 17.8 23.9 16.8 Per cent residual gas 1.6 1.9 1.8 1.4 8.7 10.8 6.0 9.5 Cubic centimeters of COj in gas . . . 146.0 244.0 250.0 267.0 43.0 55.0 72.0 83.0 Daily CO2 in gas . . 73.0 49.0 3.0 9.0 22.0 4.0 6.0 3.0 Cubic centimeters of H2 in gas 0 0 1.0 9.0 17.0 14.0 24.0 19.0 Daily H2 in gas. . . 0 0 0.5 4.0 9 -1.0 3.0 -1.0 Ratio H2/CO2 — — 1/250 1/29.0 1/2.4 1/4.0 1/3.0 1/4.4 Ammonia-N milli- grams per 100 cc. 19.0 28.0 33.6 34.4 17.3 54.6 65.4 74.4 Daily ammonia-N milligrams per 100 cc 7.0 4.5 2.8 0.4 5.5 12.4 3.6 2.2 Amino-acid-N mil- ligrams per 100 CO 27.83 35.14 61.86 69.20 67.56 103 .35 93.08 87.75 Daily amino-acid- N milligrams per 100 cc 3.04 3.65 13.36 3.67 26.85 11.93 -3.42 -1.33 Cubic centimeters N/20 NaOH to neutralize vol- atile acids per 100 cc 9.10 11.70 11.00 8.20 — 6.80 9.60 15.60 Cubic centimeters of N/20 NaOH to neutralize volatile acids per 100 cc. daily 3.05 1.30 -0.35 -1.40 — 2.26 0.93 1.50 10 cc. 6.5 5.1 5.4 7.3 — 6.30 8.8 11.5 20 cc. 13.1 10.2 10.8 13.4 — 10.5 14.7 20.5 Per cent of 30 cc. 19.7 15.4 15.3 19.5 — 14.6 19.6 26.9 volatile 40 cc. 26.1 22.5 21.6 27.4 — 17.7 24.6 33.3 acids in 50 cc. 32.5 27.9 27.0 32.3 — 21.7 28.1 38.4 fractions 100 cc. 48.3 44.4 45.1 51.2 — 39 6 48.5 52.5 of distil- 200 cc. 68.1 66.6 65.1 73.1 — 60.4 62.0 65.3 late 300 cc. 79.0 78.6 78.2 85.2 — 72.5 76.7 71.8 400 cc. 85.6 85.4 79.6 90.1 — 80.8 85.6 76.9 500 cc. 89.9 90.5 81.4 92.5 - 87.6 88.5 80.7 Before inoculation: Befor 5 inoculation: AniE lonia-N 5.0 mg m. per Ami ■nonia-N 6.2 m ;m. per 10 0 cc. IC 10 cc. Ami □ o-acid-N 20.7c mgm. Am no-acid-N 13.& 5 mgm. pe r 100 cc. P< ;r 100 cc. Vola tile acids 3.0 c :. N /20 Vol! itile acids 0.0 c c.N/20 pe r 100 cc. P« T 100 cc. 388 TABLE i— Continual PRODUCTS rORUED Total cubic centi' meters of gas. . Daily cubic centi- meters of gas. . Per cent of COa.. Per cent of Ifj. .. . Per cent of resid- ual gas Cubic centimeters of COi in gas. Daily CO2 in gas Cubic centimeters of Ha in gas Daily Ha in gas. . Ratio Ha /COa. . . . Ammonia-N milli- grams per lOOcc Daily ammonia-N milligrams per 100 cc Amino-acid-N mil ligrams per 100 cc , Daily amino-aeid- N milligrams per 100 cc Cubic centimeters of N/20 NaOH to neutralize volatile acids per 100 cc Cubic centimeters of N/20 NaOH to neutralize volatile acids per 100 cc. daily. . 10 cc. 20 cc, Per cent of 30 cc. volatile 40 cc. acids in 50 cc. fractions 100 cc. of distil- 200 cc. late 300 cc. 400 cc. 500 cc. ACTION ON WHOLE UILK HEDIUU Days incubated fi 8 132.0 194.0 26.0 21.0 76.8 78.2 15.0 17.5 8.2 4.3 101.0 152.0 20.0 17.0 20.0 34.0 4.0 5.0 1/5.1 1/4.5 72.6 77.4 12,7 1.6 83.14 92.45 11.80 3.11 39.60 40.90 6.91 0.43 8.5 8.9 15.6 16.2 21.7 22.6 27.3 27.4 32.8 31.4 49.5 48.0 66.1 65.6 74.2 74.9 79.3 80.2 82.6 84.6 12 241.0 12.0 83.6 1.5 124.47 7.99 44.60 0.92 8.9 15.2 21.9 27.9 32.8 47.1 61.0 67.8 73.6 77.6 Before inoculation: Ammonia-N 9.0 mgm. per 100 cc. Amino-acid-N 24.14 mgm. per 100 cc. Volatile acids 5 .05 cc.N/20perl00cc. ACTION ON 2 PER CENT PEPTONE WATER Days incublit«d 43.0 14.0 72.7 17.3 10.0 31.0 10.0 7.0 2.0 1/4.4 98.0 32.0 132 .98 19.59 10.95 3 7 12 19 24 29 42.3 60.4 72.2 81.4 84.6 s 11 61.0 77.0 102.0 9.0 5.0 8.0 76.5 81.7 86.2 15 .2 12.6 5.2 8.3 5.7 8.6 47.0 63.0 88.0 8.0 5.0 8.0 9.0 10.0 5.0 1.0 0.3 -1.6 1/5.3 1/6.3 1/17.6 101.0 10-1.0 108.6 1.3 1.0 1.3 112.04 103 .74 33.48 -10.47 -2.77 -23.42 12.60 15.80 15.35 0.55 1.06 -0.15 6.6 5.8 6.3 11.9 11.2 11.8 16.7 16.3 16.5 21.5 20.9 21.3 26.3 25.6 25.6 41.4 40.3 40.6 58.2 58.1 59.5 69.7 68.2 69.4 77.7 75.9 76.5 83.4 81.8 81.6 110.0 3.0 86.1 8.9 5.0 95.0 2.0 10.0 1.6 1/5.9 111.5 0.9 32.56 -0.64 13.57 -0.59 7.2 13.6 19.6 23.8 27.9 42.8 61.1 71.2 79.2 84.8 Before inoculation: Ammonia-N 2.0 mgm. per 100 cc. Amino-acid-N 74.20 mgm. per 100 cc. Volatile acids 1 .35 cc. N /20 per 100 cc. 389 TA BLE l— Continued ACTION ON 2 PER CENT ACTION ON 2 PER CENT PEPTONE SOLUTION, PLUS PEPTONE BOLCTION, PLU8 r»t»r^ TNTT/'vna c/\t>mi?t^ 0.5 PER CENT GLUCOSE 1 PER CENT LACTOSE PROD UC i o rOnMt.u Days incubated Days incubated 3 6 10 2 6 9 Total cubic centimeters of gas.... 208.0 312.0 333.0 82.0 128.0 133.0 Daily cubic centimeters of gas. . . . 69.0 34.0 5.0 41.0 12.0 2.0 Per cent of CO: 62.2 33.7 68.4 27.3 90.0 75.9 16.9 84.5 8. 8 88.2 Per cent of H2 8.6 Per cent residual gas 4.1 4.3 7.2 6.7 3.2 Cubic centimeters of CO2 in gas . . 129.0 213.0 300.0 62.0 108.0 117.0 Daily CO2 in gas 43.0 28.0 22.0 31.0 15.0 3.0 Cubic centimeters of Hj in gas. . . . 70.0 85.0 14.0 U.O 11.0 Daily !!•> in eas 23.0 5.0 _ 7.0 — 1.0 0 Ratio H2/CO2 1/1.8 1/2.5 1/4.5 1/9.8 1/10.6 Ammonia-N milligrams per 100 cc 22.6 55.0 78.5 76.0 97.0 100.0 Daily ammonia-N milligrams per 100 cc 6.8 10.8 5.9 37.1 5.2 1.0 Amino-acid-N milligrams per 100 cc 79.81 97.95 113.12 109.76 116.14 65.40 Daily amino-acid-N milligrams per 100 cc 7.57 6.04 3.79 21.68 1.59 -16.91 Cubic centimeters of N /20 N aOH to neutralize volatile acids per 100 cc 24.2 32.5 59.8 12.75 11.85 10.30 Cubic centimeters of N/20 NaOH to neutralize volatile acids per 100 cc. daily 7.6 2.8 6.8 6.10 -0.25 -0.52 10 cc. 2.5 4.9 6.0 5.3 6.8 5.9 20 cc. 5.0 9.0 11.4 10.1 12.6 11.2 30 cc. 6.7 14.8 16.4 14.9 17.9 16.2 40 cc. 8.4 19.7 20.4 19.8 22.9 20.8 Per cent of volatile acids 50 cc. 10.3 23.8 24.1 23.7 27.6 24.7 in fractions of distillate 100 cc. 22.6 27.9 35.4 39.5 41.4 39.6 200 cc, 42.7 41.9 54.5 58.4 62.5 61.1 300 cc, 56.1 60 0 66.8 69.4 74.0 62.9 400 cc. 62.8 72.4 75.6 76.0 81.6 69.9 500 cc. 67.7 80.6 81.2 80.8 86.9 75.0 Before inocu lation: Before inoculation: Anmionia-> 2.0 Ammonia-N 1.8 mgm. per 100 cc. mgm. per 100 cc. Amino-acid -N 57.1 Amino-acid-X mgm. per 100 cc. 66.39 mgm. per 100 cc. Volatile aci ds 1.4 Volatile acids 0.55 cc. N/2 ) per per 100 cc. 100 cc. 390 TABLE t—Continutd PR0DCCT8 rORUED ACTION ON 2 PKnCKNT PEPTOSK BOLCTION. PLLB 1 PER CENT SUCROSE Total cubic centimeters of gas. . . . Daily oibic centimeters of gas. . . Per cent of COj Per cent of H; Per cent of residual gas Cubic centimeters of COj in gas.. Daily CO2 in gas Cubic centimeters of H2 in gas. . . Daily Kj in gas Ratio Hj/COj Ammonia-N milligrams per 100 cc Daily ammonia-N milligrams per 106 cc Amino-acid-N milligrams per 100 cc Daily amino-acid-N milligrams per 100 cc Cubic centimeters of N/20 NaOH to neutralize volatile acids per 100 cc Cubic centimeters of N/20 NaOII to neutralize volatile acids per 100 cc. daily ■ 10 cc. 20 CO 30 cc 40 cc 50 cc 100 cc 200 cc 300 cc. 400 cc. 500 cc. Per cent of volatile acids in fractions of distillate Days incubated 5.0 1.0 4.0 0.5 71.81 -0.30 3.00 0.61 5.0 9.8 13.3 15.0 16.7 29.9 46.6 56.6 64.9 71.6 28.0 8.0 18.3 70.9 10.8 5.0 1.6 19.9 6.6 4.0/1 51.5 15.8 77.46 1.8S 17.30 4.77 6.9 12.1 16.8 20,8 24.2 39.8 57.7 68.1 75.0 80.0 ACTION ON 2 PER CENT PEPTONE SOLUTION, PLCS 1 PER CENT SALICIN 42.0 7.0 52.4 35.4 12.2 22.0 8.5 14.9 -2.5 1/1.5 99.5 24.0 83.79 3.16 13.40 -1.95 6.8 12.8 18.4 23.1 27.3 44.9 63.7 77.3 87.1 94.8 Before inoculation: Ammonia-N 1.8 mgm. per 100 cc. Amino-acid-N 73.0 mgm. per 100 cc. Volatile acids 0.55 cc. N /20 per 100 cc. Days incubated 43.0 11.0 21.0 4,8 61.81 1.18 5.90 1.34 1.7 3.4 5.1 6.8 7.6 12.7 19.5 25.5 28.9 32.3 208.0 55.0 61.9 34.1 4.0 129.0 4.3 71.0 24.0 1/1.8 45.0 8.0 77.46 5.21 20.20 4.77 5.5 9.5 13.5 16.9 19.4 33.3 56.6 62.5 69,9 75,3 228,0 10,0 65.8 30.7 3.5 150.0 16,0 70.0 -0,3 1/2.1 81.5 18.3 83.79 3.16 42.60 11.20 5.8 11.5 16.4 20.8 24.8 40.2 59.3 70.8 77.4 82.5 Before inoculation: Ammonia-N 1.8 mgm. per 100 cc. Amino-acid-N 57.08 mgm. per 100 cc. Volatile acids 0..35 cc. N/20 per 100 cc. 391 TABLE i—Conlinued ACTION ON 2 PER CENT ACTION ON 2 PER CENT PEPTONE SOLUTION. PLUS PEPTONE SOLUTION, PLUS nnnnrTTA vnTtxtvn 1 PER CENT M.VNNITOL 1 PER CENT GLYCEROL tr ^i\JU \-/\^ XO rvtk3fl£jU Days incubated Days incubated 2 6 9 4 7 9 Total cubic centimeters of gas... . 51.0 99.0 137.0 110.0 227.0 255.0 Daily cubic centimeters of gas.... 25.0 12.0 13.0 28.0 39.0 14.0 Per cent of CO2 65.3 27.2 7.5 - 85.0 10.8 4.2 - 46.5 44.5 9.0 45.4 Per cent of H2 4 .8 Per cent of residual gas 11.8 Cubic centimeters of CO2 in gas.. 33.0 — 116.0 — 106.0 116.0 Daily CO2 in gas 16.0 27.0 35.0 5.0 Cubic centimeters of H2 in gas. . . 14.0 — 15.0 — 101.0 109.0 Daily H? in cas 7.0 _ 0.3 _ 34.0 4.5 Ratio Hj /COa 1/2.4 76.0 1/7.8 1/1.0 1/1.0 29.5 Ammonia-N milligrams per 100 cc. 99.0 109.0 4.0 35.0 Daily ammonia-N milligrams per 100 cc 37.1 7.7 3.3 0.55 10.4 -2.8 Amino-acid-N milligrams per 100 cc 113 .79 116.36 66.80 73.74 92.81 86.51 Daily amino-acid-N milligrams per 100 cc 23.70 0.64 -16.51 0.18 6.35 -3.15 Cubic centimeters of N /20 NaOH to neutralize volatile acids per 100 cc 22.45 27.50 32.90 4.00 9.40 10.50 Cubic centimeters of N/20 NaOH to neutralize volatile acids per 100 cc. daily 10.90 -1.26 -1.80 0.86 1.80 0.55 10 cc. 6.7 5.2 4.6 2.5 5.3 6.9 20 cc. 12.2 10.3 8.8 5.0 9.6 12.6 30 cc. 16.9 15.1 12.8 6.3 12.8 17.4 40 cc. 21.4 19.6 16.9 7.5 15.9 21.2 Per cent of volatile acids 50 cc. 25.4 23.6 19.1 17.5 19.3 25.0 in fractions of distillate 100 cc. 40.5 38.4 33.4 32.5 32.9 39.3 200 cc. 58.8 57.3 52.9 45.4 58.0 55.5 300 cc. 68.1 67.9 65.3 55.0 69.1 67.8 400 cc. 74.8 75.4 73.5 62.5 76.6 74.8 500 cc. 80.1 81.3 79.6 67.5 81.9 79.5 Before 3 inoculation : Before inoculation: Ami nonia-N 1.8 Ammonia-N 1.8 m gm. per 100 cc. mgni. per 100 cc. Am no-acid-N Amino-acid-N 66 .39 mgm. per 73.02 mgm. per 10 0 cc. 100 cc. Vols itile acids 0.65 Volatile acids 0.55 ■ cc . N /20 per cc. N/20 per 10 0 cc. 100 cc. 392 TABLE i— Concluded VRODCCTS rORUCO Total cubic centimeters of gas Daily cubic centimeters of gas Per cent of COs Per cent of Hj Per cent of residual gas. . Cubic centimeters of COj in gas Daily CO3 in gas Cubic centimeters of H; in gas Daily Ha in gas Ratio H./CO. Ammonia-X milligrams per 100 cc Daily ammonia-N mill grams per 100 cc Amino-acids-N milligrams per 100 cc Daily amino-acid-N milli- grams per 100 cc Cubic centimeters of N/20 to neutralize volatile acids per 100 cc Cubic centimeters of N/20 to neutralize volatile acids per 100 cc. daily. . ■ 10 cc 20 cc. 30 cc. 40 cc. 50 cc. 100 cc. 200 cc. 300 cc. 400 cc. 500 cc. Per cent of volatile acids in frac- tions of distil- late ACTION ON 2 PER CKNT PEPTONC BOLOTJON. PLUS 1 PKRCENT SOLUBLE ftTARCU DATS INCUBATED 18.0 4.0 44.1 53.6 2.3 8.0 2.0 10.0 2.5 1.2/1 40.0 9.5 07.26 0.68 3.50 0.70 17.1 31.2 42.8 54.3 62.8 88.5 97.1 99.9 113.0 47.0 73.6 23 .5 2.9 S3.0 38.0 27.0 8.0 1/3.1 50.0 5.0 82.50 7.62 16.15 6 33 11.1 20.8 28.9 35.4 41.9 63.0 90.1 97.1 213.0 33.0 71.2 24 .7 4.1 151.0 23.0 53.0 5.0 1/2.9 52.5 0.8 89.90 2.43 23.95 2.60 20.8 39.9 50.1 58.2 64.2 90.2 98.1 316.0 34.0 79.3 18.0 2.7 251 .0 33.0 57.0 1.0 1/4.4 81.0 9.5 108.51 6.20 31.65 2 .58 23.6 41.8 53.9 60.0 66.1 96.4 99.4 ACTION ON 2 PER CENT PEPTONE BOLUTION, PLUS 1 PERCENT INULIN DAT8 INCUBATED Before inoculation: Ammonia-N 2.0 mgm. per 100 cc. Amino-acid-N 64.54 mgm. per 100 cc. Volatile acids 0.70 cc. N/20 per 100 cc. 4 8 28.0 30.0 7.0 1.0 20.6 29.3 — 70.0 - 0.7 6.0 9.0 1.5 1.5 21.0 — 10.0 — 2.4/1 47.5 60.0 11.4 6.3 61.52 64.20 -0.75 -1.34 1.10 - 0.10 _ 4.5 — 9.1 — 13.6 — 22.7 — 50.0 — 68.2 — 77.2 — 81.8 — - - 33.0 1.0 42.4 50.6 7.0 14.0 1.5 17.0 -1.0 1.2/1 69.0 2.2 82.54 4.58 11.9 2.7 6.7 13.4 16.3 24.4 29.4 51.2 73.1 88.2 94.1 97.4 Before inoculation: Ammonia-N 2.0 mgm. per 100 cc. Amino-acid-N 6-1.54 mgm. per 100 cc. Volatile acids 0.70 cc. N/20 per 100 cc. 393 394 L. D. BUSHNELL Since there were such large amounts of ammonia present, and because it is so difficult to remove it from the sample, corrections by the use of the factors have probably given results as nearly correct as could have been obtained by any other method. Volatile fatty-acid determinations. For the determination of volatile fattj^-acids we have followed the Duclaux method as modified and described by Dyer (1916). This author suggested that the method might be employed in the studj' of acid produc- tion by bacteria, and Wolf and his associates used it for the study of the acid production by certain pathogenic anaerobes. While the results which we have obtained show that there is a difference in the fermentative activities of this organism upon the different carbohydrates the differences are not great. The distillation of volatile fatty acids is not an easy task and while its use may be of value in detecting in a general way what prod- ucts are formed during a given fermentation, their value as a diagnostic biochemical test for bacterial differentiation is limited. Our attitude toward this test cannot be better expressed than by quoting from the summary of Wolf and Telfer. An experimental critique of the Dyer method of estimating volatile fatty acids has been made. This method, while perfectly satisfactory in the form stated by its author in dealing with a mixtm'e of two vola- tile acids, the nature of which is Icnown, fails when a mixtm-e of unknown acids is to be analyzed. The color tests as proposed by him are satis- factory when dealing with pure acids, but not as positive as could be desired for the identification of an acid in the mixture. The separation of the acids is necessary before any reliance can be placed on these color tests. From a study of table 4, it may be seen that in most cases the fractionation curves lie very near to those of propionic and buty- ric acids. By the tests devised by Dyer we could obtain no evi- dence of propionic acid, though various fractions were refraction- ated as indicated by Wolf and Telfer. We could obtain some fairly good tests for acetic and butyric acids after repeated rc- fractionations; in the case of milk, the odor would indicate vale- rianic acid though no quaUtative test could be obtained for this or for caproic acid. CHANGES PRODUCED BY A SAPROPHYTIC ANAEROBE 395 The results indioatod that the curve was (hie not to propionic acid but rather to a mixture of acetic, buytric and perhaps to valerianic acid. We felt, however, that work of this sort led to rather uncertain results, since we were able to simulate almost any acid curve desired by mixtures of other acids in the proper proportions. The refractionation method proposed by Wolf and Telfer was not exhaustive enough to obtain various acids in a form suf- ficiently pure for accurate tests. The fractional distillation of the cultures was completed in an attempt to find differences wide enough to be of diagnostic value, even though the acids present were not completely identified. DISCUSSION OF RESULTS An examination of table 4 will give an idea of the action of these orgainisms upon media rich, both in protein and carbohydrates. The cultures were very closely related in all respects in their bio- chemical reactions. In most respects the quantitative results compare favorably with the qualitative results, but unfortunately do not give much additional information. Those showing large amounts of gas and acid in fermentation tubes, and Hiss serum water media con- taining the various carbohydrates and grown over phosphorus in an anaerobic jar, also show quantitatively larger amounts of gas and acid when grown in vacuum. It was impossible to obtain duplicates which checked exactly. Variation is inherent in all bacteriological work and the results obtained in this work check about as closely as those in many other bacteriological deter- minations. Probable error. Two series of cultures were set up to determine the probable error. One series of five bottles containing two percent peptone-water after four days contained the following amounts of gas 92. 86, 61, 73, and 87 cc, respectively. A series of six bottles containing whole milk after two days contained the following amounts of gas 237, 243, 232, 203, 196, 186 cc, respec- tively. By using the formula: E = 0.6745Y A^ (iV - 1) 396 L. D. BUSHNELL in which E = probable error, S = sum of residuals, N = number of parallel determinations. The probable error as determined for the first series is ±4.78, and for the second series is ±3.07. The probable error was not determined for ammonia, amino- acid, or volatile-acid production, but dupUcate determinations usually varied about as much as in the gas determinations. Nature of fermentation. It will be noted that this organism produced considerable amounts of gas from peptone solution. This probably accounted for the presence of bubbles of gas in fermentation tubes containing non-fermentable carbohydrates. The higher negative tension and the larger amounts of media in the bottles tended to magnify the gas production from this medium in bottles, as compared to that in tubes in the anaerobic jar. (The latter usually showed a tension of about 350 mm. of mercury.) When we mention the fact that an organism ferments a certain carbohydrate we usually refer to the fact that the change results in the production of gas or acid. These are products easilj' determined and are evidently due to deep-seated changes in the product. It is quite probable, however, that there may be or- ganisms which bring about but shght changes in a carbohydrate, not extensive enough to lead to the formation of acid or even gas. It is also possible that these products may be produced and again utilized in the synthesis of cell protoplasm, or changed to other products of fermentation . It is a well known fact that the nitrif j'- ing bacteria utilize CO2 as a source of carbon in the oxidation of ammonia and nitrites to nitrates. Nikitinsky (1907) showed that there were certain anaerobes which were able to utilize hydrogen. For this purpose he used "konzentrierten kanalisationsflussig- keit" and "Schlamm aus einem Absitzbecken" placed in an atmos- phere of hydrogen. In one case he found that 500 grams of "Schlamm" was able to combine with an average of 30 cc. or a maximum of 70 cc. of hydrogen per day. It is a well known fact that many bacteria can utilize acids and alcohols as a source of energy. It may also be possible that bacteria utilize only a minimum amount of the carbohj'drate, enough to establish their initial development, but do not produce CHANGES PRODUCED BY A SAPROPHYTIC ANAEROBE 397 changes, deop-scated enough to be detected l\v the ordinary chemical methods. After growth has been estabUslied they may then utilize some other clement in the medium as their chief source of energy. It may bo noted that the anaerobes studied by us grow verj' well in a solution of peptone and water. These organisms could not utilize simple nitrogenous foods to any extent; nitrates, ammonia salts, urea and asparagin in a sim- ple solution or in the presence of glucose give practically no de- velopment. They were able to grow well upon simple peptone solutions and attack the albumins of egg-white, blood serum, milk and meat, and also liquefy gelatin, either in the presence or absence of glucose. The changes in all cases are deep-seated since gas, volatile fatty acids, amino-acids and ammonia are formed from all of these media in fairly large amounts. Traces of hydrogen sulphide are also found in some of the meat cultures. The proteolytic action is tryptic in nature, since it does not take place in a medium acid to litmus and is more active in a medium neutral to phenolphthaline than in one neutral to litmus. Gas production. In cases in which large vohuucs of gas were present we considered that active fermentation had occurred, and this was correlated with the results of the fermentation tube tests. From these results, we are safe in considering that an anaerobe has attacked a carbohydrate vigorously only when it produces more than 1 per cent of gas in the closed arm of the fermentation tube, and a reaction distinctly acid to brom-thymol-blue. A bubble of gas might be taken to indicate slight fermentation of the carbohydrate, if it were present in the tube containing car- bohydrate medium but not in a tube containing the media minus the carbohydrate. We have made as many as twenty parallel cul- tures of these anaerobes and treated all exactly alike; in some tubes there may be as much as one per cent of gas, in others a small bubble, and in others none at all, and all containing about an equal amount of growth. Since irregularities are unavoidable with these organisms, some arbitrary hmit should be set. In this work we recorded a positive fermentation when more than one per cent of gas was formed and a questionable result when less than that amount was present if there were no marked change in the reaction. 398 L. D. BUSHNELL In the case of lactose, inulin, arabinose and xylose there appears to be no doubt that these carbohydrates were not attacked to any extent. With mannitol there was an increase of gas and but httle increase in acid above that in peptone solution alone. We have concluded that there was a slight fermentation of mannitol. In most experiments the predominating gas was CO2. Hy- drogen was obtained in all cases except \\dth the cultures upon asparagus at the beginning of the fermentation. One point of interest in this connection is the ratio of Ho to CO;. In all our tests the amount of CO2 increased with the age of the fermenta- tion, while the Ho remained practically stationary after the first few days. This caused a change in the ratio of H2 to CO2 from day to day. Most authors mention a ratio between these gases as determined at a certain time. By determining it at intervals the ratio might be somewhat different as shown in this case. Apparently the ability of the organism to form hydrogen was lim- ited to the beginning of fermentation; or the substance from which this gas was formed was used up in the earUer stages; or the hydrogen was oxidized as. it was formed, as has been suggested by Nikitinsky, (1907) and Kaserer (1906). According to Perdiix (1891) B. amylomyze produced a gas of variable composition with the different carbohydrates at different ages of the culture. He gives the following table to show this (for glucose). \ R D Hydrogen Ac. carb. Hydrogen Ac. carb. days 3 175 85 2.0 175 85 4 275 145 1.9 100 60 5 350 220 1.6 75 75 11 670 450 1.5 120 130 V = volume, R = ratio, D = difference. Grimbert (1893) found that his B. orthobuUjlicus growing in glucose solution without chalk produced the following results: CHANGES PRODUCED BY A SAPROPHYTIC ANAEROBE 399 nroRooEN AC. CABB. R Up to and including 4th day 11.06 11.24 1.90 10.00 32.76 C.90 1.16 4th to 13tli dav 0.34 13th to 22nd dav 0 28 24.80 49.00 0.50 In our work the irregularities in the total amounts of the various gases were due to the fact that the amounts of gas were deter- mined from the bottles which remained in the incubator for the longest time, and only part of these were analyzed at each period. An examination of the gas volumes as mentioned on page 398 will explain this ^'ariation. In all cases there was considerable residual gas which could not be accounted for by the amount of air remaining in the bot- tles, since these were exhausted to nearly the theoretical vacuum. No doubt this was nitrogen gas which had been hberated from the medium. In no case was there more than a trace of CO2 formed by combustion with oxygen. For tliis reason we may say that no methane or ethane were formed by these organisms. It will be noticed from table 4, that gas is produced slowly in all cases except in presence of glucose and glj^cerol. These sub- stances are easily fermented and all cultures produce gas in large amounts. In the presence of lactose, there is about the same action as upon peptone solution except that slightlj^ more hydro- gen is produced in some cases, and the growth is slightly more vigorous. A sUght action upon the lactose may have occurred in the latter case. Ammonia. The ammonia production is not always in inverse ratio to the production of gas and acid by these organisms. The cultures showing larger amounts of gas show smaller amounts of ammonia in some cases but not in all. Glucose reduced the ammonia production for this organism. In the presence of lac- tose, mannitol and sucrose there was practically no reduction in the ammonia formation. In the presence of saUcin there was a decrease, and the organism is to be considered a salicin fermenting type. Peptone solution plus glycerol shows much less ammonia 400 L. D. BUSHNELL produced and there was clear e\adenee that the organisms at- tacked the glycerol. There seemed to be vigorous growth in inulin-peptone solution but the action upon the medium was not so vigorous. The proteolytic action of these organisms is also shown by the production of ammonia from asparagus, alkaline egg medium and whole milk. Amino-acids. In the Uberation of amino-acids from the pep- tone solution plus carbohydrates there is a correlation similar to that seen in the production of ammonia from similar solutions. From peptone alone there is a marked decrease in the amino-acid content after the first three days. The freeing of amino-acids does not correspond to the ease with which a carbohydrate is fermented. There is an increase in the amino-acid content for a certain period in all cultures. In the presence of glucose, salicin, sucrose, starch and inulin this increase continues to the close of the experiment. With lactose and gh'c- erol there is a slight decrease as the culture ages. In case of mannitol there is a decrease from the beginning. There is a marked production of amino acids from asparagus, alkaline egg medium, and whole milk. In a purely nitrogenous media there is a marked increase in the amino-acid nitrogen, followed by a de- crease as seen in alkaUne egg and peptone solutions. TMs point is not so marked in media containing fermentable compounds such as asparagus, milk, glucose, salicine, glycerol and starch solutions. Apparently the organisms may not attack the amino- acids as readily in the presence of fermentable carbohydrates as they do in their absence, or they may be produced more rapidly than they are used and thus accumulate in the medium. It is impossible to elucidate this point at present since total nitrogen determinations were not made on each sample at various periods. Volatile fatty acids. Fairly large amounts of volatile fatty acids were produced in most cases. The figures in table 4 rep- resent cubic centimeters of N/20 alkali used to neutralize the acid obtained by complete distillation, as above described. The per cent of volatile acid in each fraction was obtained by dividing the amount by the total acid. Usually there was a change in the CHANGES PRODUCED BY A SAPROPHYTIC ANAEROBE 401 total amount of volatile acid, and in most instances some change in the fractionation constant from day to day. With peptone alone, there was little change, but with glucose, salicin, glycerol, and soluble starch, there was an increase, while mannitol caused a decrease. These differences were constant and are probably due to a sUght difference in the fermentation reaction which occurred in these cultures at different ages. The increase in the amount of acid obtained in the first fractions would indicate an increase in the higher acids; a decrease in the amount of acid obtained in the first fractions would indicate an increase in the lower acids over those originally present. Less acid was produced in the presence of glycerol and soluble starch than was produced from peptone alone. Apparently the gas and acid production are not always related in the fermenta- tion of the various carbohydrates. This is shown to be especially true of glycerol and soluble starch in which there are very large amounts of gas and only small amounts of acid formed. The distillation curves for glucose, glycerol and inulin indicate the presence of an acid of low molecular weight, probablj^ acetic acid; while curves obtained from starch indicate large amounts of acids of high molecular weight, probably butyric. Usually the same tjT^e of curve was obtained throughout the entire incubation period, indicating that the difference in the gas ratio is not closely correlated to the production of different kinds of acid. Ap- parently the CO2 in the later stages of fermentation comes from the amino-acids present and not from the carbohydrates. CONCLUSIONS 1. We have described a method adapted for the quantitative stud J' of the decomposition products formed by anaerobic bacteria. 2. The method described has been found very satisfactory for the cultivation of anaerobic bacteria. It has also proven very satisfactory as a means of measuring the amount and kind of gas produced. 3. The objection to the use of such an apparatus is that it does not give the anaerobes an oxygen tension which might be more favorable to their development. Even the "obligate" anaerobes 402 L. D. BUSHNELL probably utilize a certain amount of free oxygen when it is available. 4. The quantitative results obtained in this particular case do not happen to yield results of significance in the subdivision of the group of organisms studied. They do, however, give us a clear picture of the biochemical behavior of these types which will make possible an accurate comparison of their physiology with that of other forms which maj^ later be investigated. 5. It may be hoped that chemists will interest them.selves in the improvement and simplification of chemical methods neces- sary to determine quantitatively and biochemically changes due to the growth of microorganisms in various culture media. As such methods are devised the bacteriologist will obtain an in- creasingly sound and logical basis for the differentiation of micro- bial species, and will obtain much valuable information concern- ing the nature of the life processes of microorganisms. The author is indebted to Professor P. L. Gainey and Professor 0. W. Hunter of this laboratory for assistance in this work. REFERENCES BuBRELL, G. A., AND SiEBERT, F. M. 1913 The sampling and examination of mine gases and natural gas. U. S. Dept. of Int., Bu. of Mines, Bui. No. 42. Dyer, D. C. 1916 A method of steam distillation for the determination of volatile fatty acids, including a series of colorimetric qualitative reactions for their identification. Jour. Biol. Chem., 28, 445-473. Grimbert, M. L. 1893 Fermentation anaerobic producte par le Bacillus orthobulylims, ses variations sous certaines influences biologiques. Ann. de I'lnst. Pasteur, 7, 353^03. Hexkt, H. 1917 An investigation of the cultural reaction of certain anaerobes found in wounds. Jour. Path, and Bact., 21, 344-38.5. Kaserer, H. 1906 Die Oxydation des Wasserstoffes durch Mikroorganismen. Cent. f. Bakt., Abt. II, 16, 681-696. Metchnikoff, E. 1908 Etude sur la flore intestinale. Ann. de I'lnst. Pasteur, 22, 929-95.5. NiKiTiNSKT, J. 1907 Die anaerobe Bindung des Wasserstoffs durch mikro- organismen. Cent. f. Bakt., Abt. II, 19, 495-499. Perbrix, L. 1891 Sur les fermentations produites par un microbe ana^robie de I'eau. Ann. de I'lnst. Pasteur, T. 5., 287. Robertson, M. 1916 Notes upon certain anaerobes isolated from wounds. Jour. Path, and Bact., 20, 327-349. CHANGES PRODUCED BY A SAPROPHYTIC ANAEROBE 403 Van Slyke, D. D. 191 1 Tlie analysis of protein by determination of the chemi- ciil groups characteristic of the different amino-acids. Jour, liiol. Chcm., 10, 15-55. Weinberg, M., and Segui.x, P. 191S La Gangrene Gazeus, pp. 444, Masson et Cie, Paris. Wolf, C. G. L. 1918 Contributions to the biochemistry of pathogenic an- aerobes. V. The biochemistry of Vibrion scjHiquc. Jour. Path, and Bact., 22, 115-128. Wolf, C.G. I*. 1919 Contribution to the biochemistry of pathogenic anaerobes. VI. The proteolytic action of li. sporogcties (Metcliinkoff) and li. xrelchii, ibid., 22, 270-307. Wolf, C. G. L., AND Haruis, J. E. G. 1917a Contribution to t!ic biochemistry of pathogenic anaerobes. III. The effect of acids on the growth of Bacillus wclchii (B. pcrjringens) and Bacillus sporogenes (.MetchnikofI). Biochem. Jour., 11, 213-245. Wolf, O. G. L., AND Harris, J. E. G. 1917b Contribution to the biochemistry of pathogenic anaerobes. I. The biochemistry of B. wclchii and B. sporogenes. Jour, of Path, and Bact., 21, 386^.52. Wolf, C. G. L., and Telfer, S. V. 1917 Contribution to the biochemistry of pathogenic anaerobes. II. Acid production by Bacillus u'clchii (B. perfringens) and Bacillus sporngcnes (Metchnikoff). Biochem. Jour., 11, 197-212. THE PROPORTION OF VIABLE BACTERIA IN YOUNG CULTURES WITH ESPECIAL REFERENCE TO THE TECHNIQUE EMPLOYED IN COUNTING G. S. WILSON From the Institute of Pathology, Charing Cross Hospital, London Received for publication December 16, 1921 PART I During the course of some work on the morphological changes occurring during the life history of certain bacteria, it became desirable to ascertain the true relationship existing between the number of living organisms and the total number of cells in a young broth culture. A few preliminary experiments were suf- ficient to show that a certain discrepancy existed between these numbers in spite of the fact that it has been frequently stated, and in many cases taken for granted, that inayoung broth culture — up to twenty-four hours — all the bacilli are viable. On look- ing up the literature it was found that of the many observers who had made a comparison of the two counts — Hehewerth (1901), Zelikow (1906), Eijkman (1904), Winterberg (1898), Hein (1902), Wright (1902), Anderson, Fred and Peterson (1920) and Glynn, Powell, Rees and Cox (1913) — not one had succeeded in proving the two to be identical. In fact, in the majority of cases, a de- finite numerical inferiority was found in the case of the viable count. In nearly every case this discrepancy was passed over with but Uttle comment, the results being explained by errors in the technical procedure adopted. The argument appears to have been that because in the early stage of a bacterial culture all the bacilli are living, then any failure in the agreement of the total and viable counts must be due to technical error. To determine 405 JOUBNAL OP BACTEHIOLOOY, VOL. VII, NO. 1 406 G. S. WILSON this point exactly it was clear that a technique would have to be evolved which would permit of as small an error as possible. A careful study of the Uterature seemed to show that few of the methods which have been employed hitherto could claim to conform to such a degree of accuracy. Generally speaking, the methods which have been employed may be classified into (1) the direct and (2) the indirect. In the former the organisms are counted directly under the microscope, in the latter the number of bacteria present is calculated from an enumeration of the colonies which develop when an aUquot part of the emulsion in question is mixed with a nutrient medium in a Petri dish, and incubated for a variable period of time. The former is designed to record the total number of organisms present, the latter only the mmiber which happens to be \'iable at the moment of sampUng. With regard to the direct or total count, Klein (1900) appears to have been one of the first to realize the value of such a method of estimation. His technique consisted in staining a moist prep- aration of organisms with aniline gentian violet, spreading a loop- ful of known capacity on a coverslip, drying, and clearing in xylol balsam. The total number of fields on the coverslip was deter- mined for a definite combination of lenses, fifty fields were counted and the number of stainable organisms per cubic centimeter in the original culture was calculated. This technique was followed by Hehewerth and Eijkman, and in a somewhat modified form by Anderson, Fred and Peterson. Zelikow introduced a method whereby the number of bacteria was determined by estimating the amount of dye they were able to adsorb from a solution of fuchsin, the strength of the latter before and after adsorption being determined by means of a Duboscq colorimeter. Winter- berg conducted his count in a Thoma-Zeiss chamber. Wright's method of counting against red blood cells is too well known to need description. A modification of it was described by Harrison (1905) who observed the mixture of bacteria and red cells in a moist film instead of in the dried condition. Finally, a new form of counting chamber — the Helbe — similar to the Thoma-Zeiss, but measuring only 0.02 mm., in depth, and fitted with an opti- VIABLE BACTERIA IN YOUNG CULTURES 407 cally piano cemented coverslip for use with a ,!, inch objective, was employed by Glynn, Powell, Rces and Cox, the organisms, being examined in the stained condition under open illumination. The indirect or viable count has, as a rule, been performed by a modification of Koch's original plating method. The only exceptions which need be noted are those of Naegeli and Schwen- dener (1877) who took the amount of fermentation as their basis for the computation of the number of growing organisms, and Winterberg who accepted motihty as his criterion of liability. The modifications of Koch's method have been concerned with the medium used, the question of prehminary dilution, the methods of dilution and the exact technique of counting the plates. The majority of observers appear to have used agar, but Buchner, Longard and Riedlin (1887), Hehewerth, Eijkman, Zelikow, and Chick (1912), seem to have preferred gelatin, though in some cases both media were employed. With regard to pre- liminary dilution, the earUer workers generally preferred to plate out the original emulsion, while of late the tendency has been in the opposite direction, as in the case of Winterberg, Zelikow, Mullcr (1895), Rahn (1906), Madsen and Nyman (1907), Lane Claypon (1909), Penfold (1914), Coplans (1909), Chesney (1916), and Noyes and Voigt (1916). The method of dilution has been subject to considerable variation; on the whole volumetric pipettes have been most popular, but Lane Claj^Don, Ficker (1898), Chick, and Penfold used dropping pipettes, w^hile Hehe- werth, and Graham Smith (1920) resorted to the use of a standard platinum loop. Gotschlich and Weigang (1895) elaborated a special technique in which dilution was performed by a combina- tion of gravimetric and volumetric methods. The important question of the counting of the plates has naturally depended largely on whether or no a preliminary dilution of the emulsion was made. Where the number of colonies was very great, micro- scopic counting was adopted, usually with the aid of a Wolf- hiigel's plate, as in the case of Buchner, Longard and Riedlin, Where on the contrary, dilution was employed, the use of the microscope was no longer necessary, and counting was perforaied with the naked eye or with a magnif jdng glass — Kronig and Paul, 408 G. S. WILSON Zelikow, Neisser, Lane Claypon, Penfold, and ]\Ioore (1915). Graham Smith used a dissecting microscope, while Frost (1921) has evolved a microplate method for purposes of rapid enumera- tion of organisms in milk. The foreging summarj^ of the litera- ture is only intended to indicate the main outlines which have been followed, a complete survey lying beyond the scope of this paper. In perusing the results of previous workers, it was striking to observe the peculiar lack of attention which was paid to the estimation of the experimental error involved in the methods employed. Probably this is to be attributed to the fact that in many cases in which the enumeration of bacteria was undertaken, a relative, rather than an absolute accuracy was essential. It was felt that the successful accomphshment of this object could only be attained by working out a technique in which the errors inherent in every step should be known with certainty. The ful- filhnent of this requirement has been kept constantly in mind throughout the course of the present work. Technique of the total count After careful consideration and personal experience of several methods — especially those of Wright and of Brown and Kirwan (1915) — it was decided to adopt the method which has been in use in this laboratory during the past two years. Practically, this resolves itself into the adaptation of the Helbe counting chamber to dark-ground illumination. Instead of examining the organisms in a stained condition with open illumination they are observed in their natural state against a dark background. For use with a parabolloid condenser, the only alteration of the cham- ber which is required, is a selection of a shde of such a thickness that the distance between the lower surface of the chamber and the upper surface of the condenser shall lie between 0.9 and 1.1 mm. With a slide of other dimensions than these the correct focussing of the rays of light becomes impracticable. The depth of the chamber is 0.02 mm., while the surface is ruled into small squares.^ The best combination of lenses has been found to be '■ A slide of the dimensions quoted was made by Messrs. Hawksley & Son. VIABLE BACTERIA I.\ YOUNG CULTURES 409 a two-thirds objective and an eighteen compensating ocular. Preliminary dilution of the emulsion is made — where necessary — by means of droppinp; pipettes — to be afterwards described — the diluent used being a 1 per cent solution of phenol in 0.9 per cent saline. If the preparation is examined immediately after ad- justment of the covcrslip, the bacilli stand out as light, refractile rods having a yellowish color, and showing a certain amount of Bro\\-nian movement; after a lapse of ten or fifteen minutes, the bacilli begin to settle on the floor of the chamber and of course lose their molecular motion; though in this condition they are quite easy to recognize, it is better to examine the preparation directly after it is put up so as to have the advantage of the Brownian movement, the degree of which was soon recognized to be characteristic of bacilli, as opposed to that possessed by minute particles of albumins and other material which is of a far more active nature. The only serious drawback that has been encountered in the use of this method is that it is unsuitable for dealing with broth emulsions containing fewer than fifty million organisms per cubic centimeter. This objection is generally of little importance, for the majority of counts one desires to make are concerned with much thicker emulsions, but the force of the disadvantage be- came clear when we wished to perform total counts on the early stages of a young broth culture before any visible turbidity had appeared. For this purpose we at first resorted to the use of a Thoma-Zeiss slide of 0.01 mm. depth, the bacilli being suspended in a weak solution of meth3dene blue, and examined by open illumination. Experience showed, however, that such a method was unreliable, errors of 50 per cent in comparative counts being met with; it was therefore abandoned altogether. So far, no satisfactory solution of this difficulty has been found. Estimate of the error of the total count Throughout most of the work to be described in this paper, a particular strain of Bad. snipestifer has been used, but for some of the earlier counts the stock strain of Bad. typhosum was selected. 410 G. S. WILSON In order to ascertain the error involved in the actual process of counting, three experiments were performed on three separate occasions. On the first occasion an emulsion of Bad. suipestifer in 1 per cent phenoUsed saline was counted eight times in succes- sion, and the results compared. They are shown in table 1, the counts being in terms of the original emulsion. On the second occasion, using an emulsion of Bad. typhosum, five counts were made, with a resulting percentage mean error of 4.04 per cent. On the third occasion ten counts were per- formed with a resulting percentage mean error of 4.95 per cent. TABLE 1 NUMBER OF COUNT COtINT PEB CDBIC CENTIMETER 1 13,830,000,000 2 14,400,000,000 3 15, 130, 000, 000 4 14, 730, 000, 000 5 15,830,000,000 6 14,100,000,000 7 15, 000, 000, 000 8 14,830,000,000 14, 730, 000, 000 464, 000, 000 Percentage mean error 3.15 From these three experiments it is seen that the mean error of performing the total count under the conditions described may be taken to be not greater than 5 per cent. It is interesting to compare these results with those obtained by Glynn, Powell, Rees and Cox. Working with staphylococci, streptococci, and Bad. coli they found that in the case of an 0.02 mm., chamber the average percentage deviation from the arithmetic mean was 3.1 per cent; with an 0.1 mm., chamber it was 5.4 per cent, and with Wright's blood film method it was 34.1 per cent. With regard to Klein's method, Gotschlich (1912) quotes the mean error of the total count as 19 per cent. VIABLE BACTERIA IN YOUNG CULTURES 411 The method of dilution Before proceeding further it will be necessary to give details of the methods of dilution which have been employed in both the total and the viable counts. For this purpose dropping pipettes have been preferred to volumetric pipettes, chiefly because in following out the growth of a bacterial culture such a large number of volumetric pipettes are required that the price of the latter becomes prohibitive : moreover dropping pipettes are more convenient to handle and more accurate in their delivery. For full details with regard to the preparation of these pipettes and for an account of the precautions to be observed in handUng them, the original article by Donald (1915) should be consulted. Suffice it to say that his technique has been used throughout with the exception of one or two modifications. Thus, for the purpose of cahbration, an Imperial Standard wire gauge has been employed and, in order to insure accuracy, the pipettes have been subsequently tested by means of an ordinary screw microm- eter measuring to 0.01 mm. Further, the drop values have been estimated by gravimetric, instead of by volumetric methods. The actual size selected was no. 22 on the wire gauge. After working out the drop volumes of various fluids at the temperature at which they were to be used, one factor remained to be deter- mined, namely the effect of variations in density of a bacterial emulsion on the size of the drop deUvered. On investigating this point, it was found, contrarj^ to expectation, that there was no appreciable difference in the drop values of emulsions ranging in thickness from 500,000,000 to 10,000,000,000 organisms per cubic centimeter. The explanation of tliis would appear to be that as the emulsion increases in density, the weight of the organisms in the drop just suffices to counterbalance the increasing viscosity of the Hquid, so that the actual value of the drop remains ap- proximately constant. Errors of the dropping pipettes In order to determine the errors of the dropping pipette a con- siderable number of experiments were made, the details of which 412 G. S. WILSON it is not proposed to enter into here as the space required would be greater than would be warranted by their relative importance. The results arrived at may be set out in the following order: (1) In making any dilution not less than four drops should be used, as with a smaller number sufficient accuracy is difficult to obtain. With four drops, however, or more, the error of dehvery does not exceed 1.2 per cent. (2) The bacterial content of drops delivered in succession appears to be uniform; no difference between indi- vidual drops could be substantiated. (3) So long as the interior of the pipettes is clean- -and all pipettes should be washed through with alcohol and ether — only a comparatively small number of bacteria remains adherent to the walls; this error of chnging does not appear to be greater than 1 per cent. (4) In making serial dilutions a separate pipette should be used for each emulsion. Under these circumstances the deviation of the actual from the calculated dilution was found to be not greater than 3 per cent. The employment of roll tubes for the estimation of the liable count Before entering on a complete description of the method em- plo3^ed in the estimation of the viable count, it is thought ad\-is- able to give a brief resume of some of the technical details which had to be worked out in order to insure the greatest degree of accuracy. In the first place Petri dishes have been discarded in favor of roll tubes. These are prepared in the following manner: Test tubes measuring 6 inches by f inch are selected; after sterilization in the hot oven, about 2 cc, of nutrient agar is placed in each. They are then autoclaved for twenty minutes at 120°C. When required for use the agar is melted and allowed to cool to 45°C., in a water bath; the inoculum is dehvered directly into the tube, the contents of which are niked bj- gentle shaking. The tube is then rolled between the fingers — as in the case of the Esmarch roll tube — the agar being allowed to pass about halfway up the tube. As a rule sohdification is complete in half to one minute, and the tubes are then incubated in an inverted position for three days at 37°C. At the end of that time, when they are removed, it will be found that the agar is studded with colonies varjdng from round to lenticular in shape, evenlj^ distributed VIABLE BACTERIA IN YOUNG CULTURES 413 throughout the tube. In order to count the number which has developed, a series of circles, 1 cm. apart, is liphtlj^ drawn around the tube with an oil-marking pencil, and a longitudinal line drawn from top to bottom intersecting each circle at right angles. The actual counting is performed by means of a small magnifj'- ing glass, the tube being examined against a dark back-ground illuminated by a partially concealed electric bulb. The contents of each circle are observed in turn from above downwards, till the total number of colonies in the tube has been enumerated. This procedure is greatly facilitated by the registration of each colony on a small counting machine — a process which has the double advantage of not only saving one the trouble of memorizing the figures, but also of eUminating the personal factor in counting, for one has no idea of the actual number till the tube has been completely examined. In using roll tubes for this purpose one or two small points should be noticed. The temperature of the agar at the time of inoculation may be allowed to vary between 41° and 50°C.; no deleterious effect of a temperature even as high as 55°C. could be noticed in experiments made to determine this point. Further, it is well to allow the agar tubes after melting to remain in a water bath at 45°C. for five to ten minutes, so as to allow of the con- densation of some of the excess moisture suspended in the air of the tube, otherwise it may be deposited on the surface of the agar after rolling, and lead to an undesirable amount of spreading. It is to allow any such moisture to run down on to the cotton plug and so be absorbed that the tubes are incubated in an inverted position. Finally to prevent excessive condensation, it is ad- visable to place the tubes in the incubator as soon after rolling as possible. Comparison of tubes and plates The question will naturally arise as to the reason for the abandonment of plates in favor of tubes. In reply the following factors may be adduced: (1) The prohibitive cost of plates when large numbers are required for use. In many experiments 40 and 50 tubes have been used at a time, and as these experiments 414 G. S. WILSON have been repeated from day to day, the acquirement of a stock of 200 or 300 plates would have been necessitated, had these been employed. (2) There is considerably less risk of contamination in the case of tubes. Absolute sterility is obtained by autoclav- ing the agar. Contamination, in fact, does not occur. In deal- ing with several thousand tubes during the course of this work, only one single case of contamination has been encountered. With plates, on the contrary, the risk of contamination is by no means negligible, especially if incubation has to be continued for three days. (3) Less media is required in the case of tubes than with plates; for the former 2 cc. are sufficient, for the latter — if a 5-inch plate be employed — 16 cc. or so must be prepared. (4) Tubes are easier to count than plates. They possess no corners; one is dealing with a relatively flat surface throughout, whereas with plates there is frequently an undesirable amount of spread- ing of the colonies at the junction of the bottom with the sides which renders counting a difficult and uncertain procedure. (5) Tubes can be incubated immediately after rolling, instead of having to be kept immobile for twenty minutes till the agar is set, as in the case of plates. Where time is a consideration the practical value of this point will be appreciated. After finding that tubes were perfectly well suited for the pur- pose of performing the viable count, it was necessary to compare the results obtained with those obtained by means of plates. For this purpose two distinct sets of experiments were undertaken, differing in the way in which the plates were poured. In the first series viable counts were made on various broth cultures employing both tubes and plates. The tubes were put up and rolled in the way already described. In preparing the plates the emulsion to be counted was delivered into a test tube containing 15 cc. of melted agar which was then poured into a plate and allowed to soUdify. Each was incubated for three days before counting. In all cases the conditions were strictly comparable. During the course of several exijeriments of this nature 69 tubes and 69 plates were examined. The total number of colonies in each set were added together : the results were as follows : VIABLE BACTERIA IN YOUNG CULTURES 415 MUHBER or COLONIBB IN TDBEa NOUBER or COLONIES IN PLATES PEKCENTAOE Or COLONIES IN PLATES TO THOSE IN TCDES 8741 7527 86.1 In reviewing these figures one was naturally struck by the distinct numerical inferiority of the plate colonies. It was sur- mised that part, at least, of this discrepancy might be attribu- table to the method of pouring, since in this procedure a certain unknown quantity of agar is bound to be left behind in the tube. To test this a second series of experiments was carried out, similar to the first, except that the drops of emulsion were de- livered directly into the plate, and the agar poured on top of them, mixing being performed as well as possible by gentle tipping to and fro of the plate. In the course of three experiments a comparison of 14 tubes and 14 plates was made, with the follow- ing results: NUMBER OF COLONIES IN TCBES NUMBER OF COLONIES IN PLATES PERCENTAGE OF COLONIES IN PLATES TO THOSE IN TrBES 8041 7636 per cent 94.93 Comparing these two sets of experiments it would appear that approximately 9 per cent of the contents of the tubes must be left behind when these are used for pouring agar into the plates. This is somewhat higher than the figure quoted by Winterberg, who likewise estimated this error, though in a different way: he found it to be about 5 per cent. At any rate it may be concluded that this method of inoculating plates does introduce a consider- able error, the efTect of which will be to give a uniformly lower count than should actually be the case. On the other hand it would be thought that when the emulsion is delivered directly into the plate, and the agar poured over it, the count should be the same as that yielded by the roll tube method; it is seen, how- ever, from the second protocol that the plate count is still 5 per cent below the roll tube count. The explanation of this may lie in the possibility that too small a number of experiments was 416 G. s. "mLSON made, but, while granting that this may be true, it would appear more probable that the difference is due chiefly to a lack of uni- form mixing of the emulsion with the agar in the plate prepara- tions. An examination of these showed that the colonies tended to be aggregated in groups, while in the roll tubes they were dis- tributed equally throughout the agar. The effect of this crowd- ing is, as is afterguards demonstrated, to lead to the failure of certain bacilU to develop into single colonies; hence a smaller number of colonies will result than would otherwise be the case when the bacilli are separated by a sufficient space from each other. If, in the plate preparations, uniform mixing could be insured, this difference would probably disappear; incidentally, however, the difficulty of insuring such a mixture is by no means imaginary, and whatever method be employed for gaining this purpose an unnecessarily large amount of time must be spent over each plate. To sum up briefly, then, it is claimed that roll tubes possess certain advantages over plates, the chief of which is undoubtedly that the count tends to be higher — probably at least 5 per cent — and therefore presumably more accurate. The only objection to their use is that they take a slightly longer time to put up ; this is more than counterbalanced, however, by the greater rapidity and ease with which they can be counted. The nature of the diluent For counting emulsions dilution has been preferred to direct sowing. For this purpose a suitable number of drops is delivered into a flask containing the diluent, and if necessary a further dilution is made by the use of a second flask. In every case a fresh pipette is used for each dilution. In the early stages of this work attention was directed to the nature of the diluent used in preparing the emulsions for counting. A review of the literature shows that the most popular fluids employed for this purpose have been saline, distilled water, and tap water. A preliminary experiment undertaken to ascertain whether any material difference could be discerned between these VIABLE BACTERIA IN YOUNG CULTURES 417 threo diluents showed that the organisms remained alive longest in tap water, while in distilled water they rapidly died out. With regard to saline, Flexner (1907) found that a pure solution of sodium chloride was distinctly inimical to the life of the Meningococcus. More recently Shearer (1919), while confirming this result, pointed out that the deleterious action of this salt could to a certain extent be neutralized by the addition of a small amount of calcium. To investigate this effect more closclj', a saline emulsion of a five and one-half hours' broth culture of Bact. suipestifer was prepared, and roll tubes were put up from it at various intervals, 3 to 5 tubes being employed for each count. The emulsion was put up at 3 p.m., and the counts were made each hour till 6 p.m., and again the following morning at 9.30 a.m. The results were as follows : TABLE 2 TIME COUNT PER CUBIC CENTIMETEB 3 p.m. 125, .500, 000 4 p.m. 120,400,000 5 p.m. 101,100,000 6 p.m. 45,380,000 9.30 a.m. Sterile "Sterile" = no organisms developing in a tube seeded with 0.1 cc. of the emulsion. The effect, then, of eighteen and one-half hours' contact of the bacilli with saUne, was to sterilize the emulsion. Considering that this effect might be due to absence of certain salts in the diluent which were necessary for the continued viability of the organisms, comparisons were instituted between tap water, saline, and Ringer's solution. Three sets of two flasks were taken, and equal quantities of these three fluids placed in them, respectively, in this order. The primary and secondary dilutions were made as nearly simultaneously as possible. After performing a viable count on each of these secondary dilutions, the flasks were lightly plugged with cotton wool to protect them from contaminating organisms. Viable counts were made at intervals up to twenty- 418 G. S. "mLSON one hours on each of them. The results have been plotted in the form of a graph to render them more striking (fig. 1). The figures are given in terms of the original culture. From this it will be seen that while all three fluids have a deleterious effect on the organisms, that of Ringer's solution is the least marked, and saline the most pronounced, that of tap water occupying an intermediate position. A fresh experiment was performed with the idea of discovering what proportion of Ringer's solution was necessary to prevent COUNT IN riiLkioNS 200 . \ \ ^^**'*-.»„,^^ > - >. ** •*• ioo 400 too aoo ioiT.CC. LMiMMimiai ^^^^B Set of Clark and Lubs Indicators, standardized dry powders *CLARK AND LUBS INDICATORS, Standardized Dry Powders .\ set of indicators covering the important zones of pH, carefully selected for their brilliancy, and which are relatively unaffected by conditions other than Hydrogen-ion concentration, such as sjilt and protein errors. The Thymol Blue of this series is a three-color indicator, recommended for differential titrations; see Clark and Lubs, Journal of the American Chemical Society, Vol. 50 [1918). The Brom Cresol Purple covers practically the same range as litmus, and is more brilliant and reliable. Common Name Thymol Blue — acid range Brom Phenol Blue Methyl Red Brom Cresol Purple Brom Thymol Blue Phenol Red pH range 1.2—2.8 2.8-^.6 4.4—6.0 5.2—6.8 6.0—7.6 6.8—8.4 7.2—8.8 8.0—9.6 8.4—9.2 Chemical Name Thymolsulphonephthalein Tetrabromphenolsulphonephthalein Orthocarboxybenzeneazodimethylaniline . . Per 0.1 gram .25 .25 .25 .25 .25 .25 .25 .25 .25 .25 Per 1 gram 2.00 2.00 .50 2.00 Dibromthymolsulphonephthalein Phenolsulphonephthalein 2.00 2.00 o-Cresolsulphonephthalein 2.00 Thymol Blue — alkaline range Phenol Phthalein Thymolsulphonephthalein Phenol Phthalein 2.00 .50 Cresol Phthalein Cresol Phthalein LOO • Prepared for us in the laboratories of Hynson, Westcott & Dunning. THE TEST IN OUR STOCK FOR IMMEDIATE SHIPMENT OF SERVICE ARTHUR H. THOMAS COMPANY WHOLESALE, RETAIL AND EXPORT MERCHANTS LABORATORY APPARATUS AND REAGENTS WEST WASHINGTON SQUARE PHILADELPHIA, U.S.A. CABLE ADDRESS, "BALANCE," PHILADELPHIA UAROIN THE RELATION OF VITAIMINES TO THE GROWTH OF A STREPTOCOCCUS^ S. HENRY AYERS and COURTLAND S. MUDGE From the Research Laboratories of the Dairy Division, United States Department of Agriculture The importance of vitamines in animal nutrition has led to a little experimental work with bacteria. Results of various in- vestigators have caused the belief that certain materials such as animal and plant tissues contain growth accessor}' substances which stimulate the growth of microorganisms and which with some bacteria are essential for growth. Among the studies along this line may be mentioned the work of Cole and Lloyd (1917), Paccinni and Russell (1918), Hall (1918), Agulhon and Legroux (1918), Kligler (1919), Bachman (1919), ^Yilliams (1919), Willaman (1920), Davis (1921), Rivers and Poole (1921), and MacLeod and Wyon (1921). In some papers the growth-promoting substances are spoken of as growth-accessories substances, and in others as vitamines. The work in general clearly indicates that for some microorganisms there are present in certain materials growth-promoting sub- stances. As long as this term is used, one is not committed to the assumption that these substances may be vitamines. Glu- cose in small amounts is a growth-promoting substance for many bacteria, in fact any easily available source of carbon can be considered a growth-promoting substance. If, however, the substances which have been found to be growth promoters are classed as vitamines, then there is a great possibility of investi- gators being misled as to the real connection between the recog- nized vitamines and the growth of microorganisms. In this paper we shall present the results of a few experiments • Presented at twenty-third annual meeting, Society of American Bacteriolo- gists, Philadelphia, December 27, 1921. 449 JOUnXAL OP BACTEBIOLOQT, VOL. Til, NO. 5 450 S. HENRY AYERS AND COURTLAND S. IIUDGE with the water-soluble vitamine B and the fat-soluble vitamine A in their relation to the growth of a streptococcus. EXPERIMENTS WITH WATER-SOLUBLE B Throughout our experiments we have used a culture of a patho- genic streptococcus which grew slowly in a peptone yeast medium. This organism would not grow in a peptone medium without broth, and the fact that it would grow with the addition of yeast extract brought up the possibility that the water-soluble B might be the reason for the growth. Several experiments were conducted in order to throw some Ught on this point. In the first experiment 10 grams of autohzed yeast were extracted in a Soxhlet apparatus for eight hours with 95 per cent alcohol. Fresh alcohol was then added and the ex- traction continued for another eight hours. This process was repeated until the extraction had been continued for forty hours. During the extraction the yeast was removed several times from the thimble and the clumps were thoroughly disintegrated. As the alcohol was removed during the extraction it was evaporated at a temperature of G0° to 80°C. The combined residues from the successive extractions were added to 500 cc. of distilled water and heated for a half hour. To this was added an equal amount of 2 per cent Difco peptone and the reaction adjusted to pH 7.2. This medium was then steamed for fifteen minutes, filtered, placed in flasks, and sterilized. The residue from the original 10 grams of yeast which had undergone extraction by the 95 per cent alcohol was added to 500 cc. of distilled water and another medium prepared with Difco peptone as described. A third medium was prepared by using 10 grams of the regular autolized yeast which had not been extracted. This was added to 500 cc. of distilled water and a medium prepared with Difco peptone, the same as with the other media. In all the experiments the media were distributed in 50 cc. amounts in 100-cc. Erlenmeyer flasks. These flasks were approximately of the same shape so that the depth of the medium and the surface area exposed was about the same in each flask. VIT^UIINES AND GROWTH OF STREPTOCOCCUS 451 The flasks containing each of (he three media described were inoculated with a twenty-four hour culture of the streptococcus and incubated at 37°C. Examinations for growth were made after seventeen, twenty-four and forty-eight hours. Assuming that the water-soluble B can be extracted by hot 95 per cent alcohol, the medium made up with the alcohol extract should TABLE 1 Growth of a streptococcus in various media 95 per cent alcoholic yeast ex- tract -H peptone Extract from yeast residue -1- peptone Extract of regular yeast -f peptone Fraction I + peptone Fraction II + peptone Filtrate A -f- peptone Yeast extract treated with Lloyd's Reagent + peptone Regular j'east extract + pep- tone IIOCU9 OF INCUOATION 18 24 48 + * -I-* -1- + + + * + + + - + + + + + -1- - + + -1- REMARKS Should contain water-soluble B Should not contain water-solu- ble B Contains water-soluble B Should contain some water-solu- ble B Should contain most water-solu- ble B Should contain little if any water- soluble B Should not contain water-soluble B from yeast Should contain water-soluble B from yeast * Slight. contain this vitamine. The medium prepared from the alcohol- extracted yeast should not contain the water-soluble B, or at least the amount should be greatly reduced. Further, the me- dium made from the regular yeast should of course contain the vitamine. The results shown in table 1 were of considerable interest to us. The growth was recorded by -f signs which were increased 452 S. HENRY AYERS AND COTJRTLAND S. MXTDGE in number as the growth seemed heavier. It will be observed that the growth was equally good in the medium prepared from the regular yeast and from the yeast extracted with the alcohol, whereas only a slight growth was manifest after forty-eight hours in the medium containing the alcohoUc extract of yeast. The results would seem to indicate that water-soluble B is not the growth-promoting substance of yeast at least for the streptococcus used in this work. There is, of course, a possibihty that the water-soluble B was not soluble in the 95 per cent alcohol and was not contained in the extract, and for this reason another experiment was conducted in which a method similar to that of Osborn and Wakeman (1919) was used for the preparation of fractions containing dif- ferent amounts of water-soluble B. Some of the same autolized yeast was used, 10 grams being added to 500 cc. of boiUng water containing 0.01 per cent acetic acid. This was stirred for some time and filtered, the filtrate being concentrated to 300 cc. at a temperature of about 80°C. To this 300 cc. of filtrate sufficient 95 per cent alcohol was added to get a concentration of about 52 per cent. The solution was allowed to stand over night at about 5°C. and the precipitate which had appeared the next day was filtered off. This precipitate was known as fraction I. The filtrate was then concentrated to 200 cc. and sufficient 95 per cent alcohol added to give a concentration of 79 per cent. This solution also was allowed to stand over night at a temperature of 5°C., and the precipitate filtered off. This was known as fraction II. The filtrate was again concentrated, this time to 100 cc, and sufficient 95 per cent alcohol added to give a concen- tration of about 90 per cent. After standing over night at 5°C. the precipitate that formed was filtered off and kept as fraction. III. The filtrate from fraction III, which we shall term filtrate A, was evaporated to dryness, and dissolved in 1000 cc. of distilled water, to which 1 per cent of Difco peptone was added, the reac- tion adjusted to pH 7.2 and the medium placed in flasks and sterilized in an autoclave. Fractions I and II were each dissolved in 1000 cc. of distilled water and 1 per cent Difco peptone was added to each lot. These media were adjusted, placed in flasks VITAMINES AND GROWTH OF STREPTOCOCCUS 453 and sterilized as described above. Osborn and Wakcman found that fraction II contained most of the water-soluble B, and that fractions I and III contained some. According to this we should expect that the peptone medium containing fraction II should contain most of the water-soluble B, that the medium containing fraction I should contain less, and that filtrate A should contain the least or none at all. On this assumption, that water-soluble B is the growth-promoting substance of yeast, it would be expected that media containing fraction II would give the best growth with the streptococcus. From the second section of table I it will be seen that such was not the case. There was no growth in peptone media containing fraction II or fraction I, but there was growth in the peptone medium containing filtrate A. These results seem to confirm those obtained in the first experiment and again indicate that water-soluble B is not the growth-promoting substance of yeast. A third experiment was conducted in which the water-soluble B was considerably reduced and possibly entirelj'' removed from the j'^east extract. Seidell (1916) has shown that, by shaking autolized yeast extract with Lloyd's Reagent (Fuller's Earth) using 50 grams per liter of extract, only an inconsiderable amount of the vitamine appeared to remain in the filtrate. Ten grams autolized yeast were added to 200 cc. of distilled water. This was allowed to stand for about one hour with frequent shaking before filtering. To this filtrate 50 grams of Lloyd's Reagent were added together with HCl to make the solution 0.01 normal. The yeast extract was then shaken every half hour for four hours and the Lloyd's Reagent removed by filtration. The Lloyd's Reagent was supplied to us through the kindness of Dr. Seidell and was known to be an active grade of Fuller's Earth. The filtrate was made up to 1000 cc. to which 1 per cent of Difco peptone was added and the reaction adjusted to pH 7.2. After filtration this medium was put in flasks and sterilized in the usual way in the autoclave. Since it has been found that the Lloyd's Reagent removes a large part, if not all, of the water-soluble B from yeast extract it would be expected that a peptone medium containing a yeast extract treated with this substance would not 454 S. HENRY AYERS AND COURTLAND S. MXJDGE support growth of the streptococcus. The results shown in the third section of table I indicate that the growth in such a medium was just as good as the growth in our regular peptone medium containing the extract of autolized yeast. We have only shown in our tables the result of growth in a single flask, but these results have been duplicated many times so that it is felt that the results presented express the general average. The results of MacLeod and Wyon (1921) are interesting in connection with our experiments. They conducted some experi- ments on the growth of a streptococcus with extracts from various substances. The used extracts from hver, kidnej^, egg yolk, yeast, muscle, bran, and milk, and found that with the exception of milk and bran the results seemed to favor the vitamine hy- pothesis so far as small amounts were active. They found how- ever, that yeast extract had but httle more growth promoting power than muscle extract. In view of the high water-soluble B content of yeast, and of its small amount or even absence in muscle, these results appear to us again to indicate that the water-soluble B is not the growth-promoting substance of yeast. MacLeod and Wyon also worked with the pneumococcus and meningococcus and they point out that the growth-promoting property of certain extracts did not bear a direct relation to the known vitamine content, and that yeast had little or no effect in promoting the growth of these organisms. A survey of the second and third experiments shows that an attempt was made in two distinct ways to obtain data on the relation of water-soluble B to the growth-promoting substance or substances of yeast. In the second experiment accepted methods were employed for removing a fraction from yeast extract containing a large amount of water-soluble B. This fraction with peptone did not support growth of the streptococcus and this fact is very significant. In the third experiment ac- cepted means were used for removing or at least greatly reducing the water-soluble B content of the yeast extract. This extract which should have been free or very nearh^ free from water- soluble B, with peptone, supported growth of the streptococcus in VITAMINES AND GROWTH OF STREPTOCOCCUS 455 a normal manner. The results of these experiments when in- terpreted in hght of our present knowledge of water-soluble B, and apjilied to the streptococcus studied, permit only the con- clusion that this vitaminc is not the srowlh-promotinp; substance of yeast extract. It is interesting to note that Funk and Dubin (1921) working on the vitamine reciuircments of yeasts and bacteria, have isolated a substance which they believe to be definite and specific for the stimulation of the growth of yeast. This substance, which was separated from vitamine B, they call vitamine D, EXPERIMENTS WITH CABBAGE EXTRACTS It has been found that extracts from plant tissues apparently contain growth promoting substances for microorganisms, and cabbage seems to have given particularly good results. We tried one experiment in order to determine whether or not cabbage ex- tract was valuable for the growth promotion of the streptococcus. One hundred cubic centimeters of finely minced cabbage were added to 300 cc. of distilled water and allowed to stand for forty- eight hours in the ice box. The suspension was then steamed for thirty minutes and filtered. Fifty cubic centimeters of 2 per cent Difco peptone solution were distributed into a series of flasks. To these flasks were added increasing amounts of the cabbage extract, these amounts being 1 cc, 5 cc, and 50 cc. Whenever necessary the medium was made up to 100 cc. and dis- pensed into tubes. Thus we had a series of media, one contain- ing 1 per cent peptone alone and the other three having 1 per cent, 5 per cent, and 50 per cent of cabbage extract. Series of each of these media were inoculated with severail different cul- tures (one of them being the culture used in the other experi- ments), were incubated for twenty-four hours, and then examined for growth and reaction. It was found that in the plain peptone without cabbage extract there was a fair growth of some of the streptococci, but with others the growth was questionable. In all of the tubes containing cabbage extract, growth was observed and the increase was more or less proportional to the increase in percentage of cabbage extract. The acidity also increased in a 456 S. HENRY AYERS AND COURTLAND S. MUDGE similar manner. These results showed quite plainly that the cabbage extract contained some growth promoting substance, but the increase in acidity with the increase in percentage of cabbage extract also indicated that there was sugar present. This fact led us to consider the effect of sugar as a growth promot- ing substance. An analysis of the cabbage extract for which we are indebted to Doctor Rupp of these laboratories, showed that it contained 1.4 per cent of reducing sugar. Calculations showed that when the cabbage extract was incorporated with the peptone solutions the resulting media contained 0.014, 0.07 and 0.7 per cent of sugar for the 1, 5, and 50 per cent cabbage media, respec- tively. In order to show the effect of these percentages of sugar on the growth of the streptococcus a series of media was prepared as previously described except that a 1.4 per cent glucose solution was incorporated in the media instead of the cabbage extract. The growth in these last media was compared with that in the cabbage extract media mentioned. From the results of this comparison, shown in table 2, it will be seen that growth in peptone media with cabbage extract was practically identical with that in the peptone media ■nith the glucose solution. The growth increased, generally speaking, with the increase in the percentage of cabbage extract, and, in a similar manner, with the increase in the percentage of glucose solution. It shovdd be remembered that the glucose solution contained 1.4 per cent reducing sugar so as to correspond ■with the cabbage extract. It is particularly interesting to note the increase in growth due to the incorporation of only 1 per cent of the glucose solution which gave a sugar content in the medium of only 0.014 per cent. From these results it seems evident that a very small amount of sugar acts as a growth-promoting substance for some of the streptococci. We wish to emphasize that Avhen extracts of plant tissues are used as a source of growth-promoting substances, the possible effect of sugar and other reducing substances present must be given considerable thought before the growth promotion can be attributed to vitamines. This is not only true when plant tissues are used but holds equally well for any extracts which may contain sugars. VITAMINES AND GROWTH OF STREPTOCOCCUS 457 u w S a .2 1 3 5 a Tt* -(J* lO W3 U5 lO lO ■* ■* ■* 1 4. + + + + 1 + + + + ! + + + + + + + + s H 6 0. o •>»< lo lo in to U5 ■* Tli •^ .3 c It 2 O + + + + + + + + + + + + + + + + + + + + g u o ca u § s 5 a to lO O lO U3 o § 1 S + + + + + "3> -a c 1 1 o. (N <0 00 lO ■* to uj •* ■* lO 1 .£3 ic 2 O + + + + + + + + + + + + + + + + + + + + u o s: u P4 i g 3 O 0. O (N (N O ■* b^ f- N^ t>^ t^ e o ■r* 2 o + + + + + + + + + + •a 6 00 O 0> GO OS CO t^ «> O O "S. .a 2 o + + + + + + + + + + J5 H O 2 C .2 1 _3 5 a to o to ■* 00 1-^ b^ t-^ t^ t-^ s 2 o + + + + + 1 ^ 1 M 1 a a to to 00 ■«< 00 t~ t^ b- t^ t^ J3 2 o + + + + + rt C) CO ■>»< lO 458 S. HENRY AYERS AND COXJRTLAND S. MUDGE EXPERIMENTS WITH FATS AND OILS We became interested in the effect of fats and oils as sources of growth promoting substances from the results obtained by Dr. Sherman of these laboratories, \Yho found that fat-soluble A was apparently necessary for the growth of some of the organ- isms of the high acid group. In all of our experiments the following basic medium was used: Peptone (Dif co) 10 grams Autolized yeast 10 grams Distilled water 1,000 cc. pH 7.4 Fifty cubic centimeters of this medium were placed in 100 cc. Erlenmeyer flasks and 0.5 cc. of the fat or oil to be studied was added before sterilization at 15 pounds for thirty minutes. This gave a concentration of 1 per cent of the oil in the medium. In the first experiments butter-fat and cod-liver oil were used. The butterfat was prepared from sweet cream churned into butter and washed with warm water several times. As a final process the fat was filtered through paper into a clean, dry beaker in which it was stored at about 5°C. during the course of these experiments. Using the same culture of streptococcus employed in the experiments with water-soluble B it was found that both butterfat and cod-liver oil stimulated the growth considerably during the first twenty-four hours. These experiments were repeated several times and the same results were obtained. WhUe these results indicated that fat-soluble A might be respon- sible for the stimulation it was decided to try ohve oil which is beUeved not to contain this vitamine. The same stimulation occurred however as with butterfat and cod-Uver oil. This seemed to present evidence that fat-soluble A was not responsible for the stimulation. While it is generally considered that oUve oil does not contain fat-soluble A, the work of Drummond and Coward (1920) indicates that it may be present in small amounts. For this reason it was decided to use oil which could hardly be suspected of containing fat-soluble A and white mineral oil was therefore VITAMINES AND GROWTH OF STREPTOCOCCUS 459 selected. ]\Iuch to our suri)ii.sc the same stimulation of growth of the streptococcus occurred as with butterfat and cod-liver oil. In these prelininary experiments there was no accurate measure of the degree of stimulation of growth, for it had been judged only b\' the appearance of the turbidity of the cultures after shak- ing. A more accurate measure was desired and the following method was adopted. Flasks containing 50 cc. of the yeast peptone medium with 1 per cent of a number of different oils- were inoculated with a loopful of a water suspension of a twenty- four-hour growth of the streptococcus on agar. After twenty- four hours' incubation at 37°C. the amount of growth was deter- mined bj' plating on infusion agar. The results of these experi- ments are shown in table 3. It will be seen that the inoculated control peptone yeast medium without oil contained 170,000 bacteria per cubic centimeter after twentj^-four hours of incuba- tion. With sesame oil there was apparently little if any stimula- tion, while chaubnoogra oil seemed to be toxic. The rest of the oils and fats increased the count to a considerable degree. Some of the vegetable oils were considerably active, as for example okra-seed oil. This result is based on only one experi- ment, and in a repetition of the experiment a count was obtained with okra-seed oil similar to those observed with the other oils. The most interesting feature of the results was the marked stimulation of growth by mineral oil, vaseline, and even solid paraffin. Microscopic examinations were made in some cases to prove that the plate count represented a difference in the actual number of cells in the cultures. Since we were working with a chain-forming streptococcus, the influence of the oil might have been to decrease chain formation, and therefore apparently increase the count. Microscopic tests showed, however, that the plate count represented the real difference in growth. The results so far were obtained in media containing 1 per cent of the various oils but it was thought that this percentage could be lowered. To this end oils were selected from the three ' We are indebted to Doctor Jameison of the Bureau of Chemistry for many of the oils used in our experiments. 460 S. HENRY AYERS AND COURTLAND S. MUDGE major gi-oups (animal, vegetable, and mineral) and the amount of fat diminished to 0.2 per cent, 0.1 per cent, 0.02 per cent and 0.002 per cent. In table 4 the results are given. It will be seen that the counts tended to decrease with the lessened amount TABLE 3 Effect of various fats and oils on the growth of a streptococcus PAT OR OIL (I PER CENT USED) Control without oil Sesame Chaulmoogra Castor Com Rape seed Lumbang Coconut Peanut Peanut Soy bean Olive Chia seed Mustard seed Okra seed Linseed Lard Lard Lard , Butterfat Cod liver Mineral Mineral Vaseline Vaseline Paraffine, solid Paraffine, solid BEUABK3 BACTERIA PER CUBIC CENTIUETEB TWENTY-FOUR HOURS 170,000 400,000 less than 100, 000 1, 100, 000 Crude 2, 600, 000 Crude 3,500,000 Candle nuts 4, 100, 000 Refined 6, 000, 000 Crude 8,800,000 Refined 11, 000, 000 10, 400, 000 9, 700, 000 10, 400, 000 14,000,000 33,000,000 Raw 5, 600, 000 Prime steamed 3, 900, 000 Open kettle ren- dered, direct heat 1,300,000 Prime steamed, re- fined 5, 100, 000 From sweet cream 14, 100, 000 8,100,000 1 per cent added 5, 700, 000 2 per cent added 5, 100, 000 1 per cent added 9, 100, 000 2 per cent added 9,700,000 1 per cent added 5, 900, 000 2 per cent added 11, 100, 000 of fat or oil but even small quantities caused a marked stimula- tion. In plating these oil media it was difficult to keep the 1 cc. sample drawn from the flask free from drops of oil and per- haps this is the explanation of some of the inconsistencies seen VTTAMINES AND GROWTH OP STREPTOCOCCUS 461 in the tables. In makinp; the dihitions of 0.02 per cent and 0.002 per cent of the fats and oils the following methods were used: To dilution blanks containing 100 cc. of 1 per cent peptone, 1 cc.of the oil was added and an emulsion obtained by thorough shaking. By adding 1 cc. and 0.1 cc. to our media we obtained the percentages of fat mentioned above. The main point to be observed however is that even down to 0.002 per cent of oil in the medium there was a stimulation of growth. This small TABLE 4 Effect of small amounts of butterfat, rape and mineral oils on the growth of a streptococcus CONCENTRATION BtriTEnKAT RAPE OIL UINERAl. Oil. Control without fat or oil Bacteria per cc. 680, 000 10, 000, 000 2, 140, 000 1,260,000 6, 900, 000 5,000,000 10,300,000 Bacteria per cc, 680, 000 13, 600, 000 12,000,000 17,900,000 13, 900, 000 5, 600, 000 5,600,000 Bacteria per cc. 080, 000 1.0 32, 000, 000 0,2 25, 000, 000 0.1 17, 900, 000 0.02 6, 900, 000 0.002 3, 100, 000 Extracts (7) from fats and oils 3, 600, 000 Butter 1 per cent. Butterfat 1 per cent. Bacteria per cc. f first experiment 430, 000 ■\second experiment 0 in 1 : 10, 000 dilution ffirst experiment 11, 000, 000 '\second experiment 14,000,000 amount of oil could hardly be distinguished on the surface of the media even on careful examination. At the same time other media were made in which the oils butterfat, rape and mineral or extracts from them, were added, in the following way: 100 cc. of water was shaken up with 5 cc. of the fat and oils for a period of three hours. The bottles were than set aside in the laboratory for several days after which time was a layer of oil at the surface of the water but the water itself was slightly cloudy. Two cubic centimeters of this emulsion, if it was an emulsion, since we dealt with pure oil and distilled water, were added to 462 S. HENRY AYERS AXD COURTLANd S. MUDGE the basic media and inoculated with the streptococcus. In table 4 experiment 3 we see a most astonishing stimulation of growth. Another interesting phenomenon was observed which might well be mentioned at this point. In 1918, Mr. Johnson of these laboratories prepared in the field, some pasteurized-cream butter without salt and also separated some of the butterfat free from casein. These were sealed in sanitary cans and sent to Washing- ton. Since then they have been in our incubators at tempera- tures averaging 30°C. These cans were opened recently and the fat found to be in a melted condition. Samples were taken of the supernatant fat from a butter can and also some of the oil from the can of butterfat free from casein. Media were made with these two fats using 1 per cent. The results which are also shown in table 4 are peculiar to say the least. The fat from the butter itself seems to be toxic but the stored fat free from casein stimulated bacterial growth the same as our fresh butter- fat. One other point of interest is that the stimulating effects of fats and oUs was not manifest to any degree in a plain peptone medium but was evident in the yeast peptone medium. It was found by the addition of methylene blue that an anero- bic condition did not exist in the media with fats and oils. It cannot be assumed therefore that an anaerobic condition was responsible for the stimulation of the growth of the strepto- coccus. The experiments with fats and oils show one thing definitely which is that very small amounts of these materials stimulate in a most remarkable manner the growth of a streptococcus growing in a yeast peptone medium. If our studies had been limited to butterfat and cod-Uver oil strong evidence could have been presented as to the stimulation by fat-soluble A. But the stimulating effect of mineral oils and even soUd paraffin change the aspect of the situation. Our results as we see them permit of one of three possibilities, first, that the growth-promoting property of fats and oils is not due to the fat-soluble A, or second, that if fat-soluble A is re- sponsible then it must be contained in mineral oil. The second VITAMINES AND GROWTH OF STKEPTOCOCCUS 463 possibility suggests that it may he dosinible to study further the vitamine content of mineral oil. Third, that the stimulation with fats and oils containing vitamine-A and that with mineral oils is not due to the same thing. Facts in our possession at present do not suggest which possibility is the correct one. SUMM^UIY AND CONCLUSIONS 1. The results presented in this paper apply only to the growth of a culture of a pathogenic streptococcus and do not necessarily apply to bacteria in general. 2. Autolized yeast extract contained a growth-promoting sub- stance or substances for the streptococcus studied. Water- soluble B did not, however, appear to be the significant substance. 3. Cabbage extract was found to promote growth, but a glucose solution containing the same amount of sugar as the cabbage extract showed a similar growth-promoting effect. It is evident that when extracts of plant or anim.al tissues are used the sugar content must be given consideration in connec- tion with their growth-promoting properties. 4. Fats and oils, vegetable, animal and mineral even in very small amounts were found to stimulate the growth of the strep- tococcus. Either the growth-promoting property of fats and oils is not due to fat-soluble A, or this vitamine is present in mineral oils, or the stimulation is due to different causes in the case of the vitamine-containing fats and oils and the mineral oils. REFERENCES Agulhon, H., et Leguoux. R., 1918 Contribution a I'etude des vitamines utilisables i la culture des microorganismes. Application au bacille de I'influenza. (B. de Pfeiffer). Compt. Rend. Acad. Sci. [Paris], 167, 597-600. Bachmann, F. M. 1919 Vitamine requirements of certain j'easts. Jour. Biol. Chem.,39,no. 2,23.5-258. Cole, S. W., and Lloyd, D. J. 1917 The preparation of solid and liquid media for the cultivation of the gonococcus. Jour. Path, and Bact., 21, no. 2, 267-280. Davis, D. J. 1921 Food accessory factors in bacterial growth. III. Further observations on the growth of Pfeiffer's bacillus (B. influenzae). Jour. Infect. Diseases, 29, no. 2, 171-177. 464 S. HENRY A'i'EES AND COHRTLAND S. MTJDGE Drummond, J. C, AND Coward, K. H. 1920 Researches on the fat-soluble accessorj- substance. V : The nutritive value of animal and vegetable oils and fats considered in relation to their colour. Biochem. Jour., 14 , no. 5, 668-677. Funk, C. and Dubin, H. E. 1921 Vitamine requirements of certain yeasts and bacteria. Jour. Biol. Chem., 48, no. 2, 437-443. Hall, I. W. 1918 On the amino-acid content of nutrient media. Brit. Med. Jour., no. 3015, 398. Kligler, I. J. 1919 Growth accessory substances for pathogenic bacteria in animal tissues. Jour. Exp. Med., 30, no. 1, 31-44. M'Leod, J. W., AND Wton, G. a. 1921 The supposed importance of vitamins in promoting bacterial growth. Jour. Path, and Bact., 24, no. 2, 205-210. Osborne, T. B., and Wakeman, A. J. 1919 Extraction and concentration of the water-soluble vitamine from brewers' yeast. Jour. Biol. Chem., 40, no. 2, 383-394. Pacini, A. J. P., and Russell, D. W. 1918 The presence of a growth-producing substance in cultures of typhoid bacilli. Jour. Biol. Chem., 34, no. 1, 43-49. Rivers, T. M., and Poole, A. K. 1921 Growth requirements of influenza bacilli. Johns Hopkins Hosp. , Bui. , 32 , no. 364, 202-204. Seidell, A. 1916 Vitamines and nutritional diseases. Reprint no. 325 from U. S. Pub. Health Rpts. , 31 , no. 7, 364-370. Willaman, J. J. 1920 The function of vitamines in the metabolism of sclero- tinia cinerea. Jour. Amer. Chem. Soc, 42 , no. 3, 549-585. Williams, R. J. 1919 The vitamine requirement of yeast. A simple biological testfor vitamine. Jour. Biol. Chem., 38, no. 3, 465-486. SALT EFFECTS IN BACTERIAL GROWTH' II. THE GROWTH OF BACT. COLI IX RELATION TO H-ION CONCENTRATION JAMES M. SHERMAN and GEORGE E. HOLM Prom the Research Laboratories of the Dairy Division, United States Department of Agriculture, Washington, D. C. Received for publication, January 7, 1922 It has been shown in a previous publication (Holm and Sher- man, 1921) that sodium chloride and various other neutral salts in 0.20m concentration affected the rate of growth of Bact. coll. Using neutral salts with a common cation (sodium) but with various anions a marked difference was observed between the action of the various salts. The effect of the chlorides of sodium, potassium and ammonium seemed to be approximately the same while the calcium and iron salts tested retarded greatly or inhibited growth. These experiments were carried out at a pH of approximately 7.0 and with a salt concentration of 0.20m in 1 per cent'pepton. In as much as we know that there are limiting pH values for bacterial growth, varying with different organisms, it would be of interest and value to know just to what extent this neutral salt action is affected by various H-ion concentrations In the following experiments, as in our former communication, the rate of growth was determined by the time that expired between inoculation and the first sign of turbidity. The medium used was 1 per cent pepton to which had been added various amounts of salts, and the H-ion concentration adjusted by the use of concentrated HCl and NaOH solutions. The effects of various concentrations of NaCl at various H-ion concentrations were first tried. The H-ion and salt concentra- ' Published with the permission of the Secretary of Agriculture. 465 JOURNAL OF DACTFRIOLOOT, VOL. VII, NO. 5 466 JAMES M. SHERMAN AND GEORGE E. HOLM tioiis, and the time for growth in each series, are shown in' table 1 . This table shows a decided accelerating effect upon growth with added NaCl in low concentrations. Although there was little difference between the effects of 0.10, 0.20 and 0.30m NaCl media upon growth, there seems to be an optimum effect at about 0.20m. The optimum H-ion concentration for growth either in controls or in pepton containing NaCl at various con- centrations seems to be about the same, approximately pH 7.8. At optimum salt concentrations there is very little difference in the rate of growth over a wide range of H-ion concentration, TABLE 1 The rale of growth of Bad. colt in various concenlrations of NaCl in 1 per cent pep- tone and at various H-ion concentrations NaCL SERIES I SERIES II BERIE8 III SEBIES IV 8EBIE3 T pH Hours pH Houra pH Hours pH Hours pH Hours Control 5.3 36 6.3 lOJ 7.0 7 7.7 6 8.3 7 O.OOM 5.3 61 0 3 Si 7.1 5! 7.7 3i 8.3 4 0.10m 5.3 4 6.4 4 7.1 4 7.8 Si 8.3 3J 0.20m 5.3 4 6.5 3^ 7.2 3i 7.8 3i 8.3 3i 0.30m 5.3 4i 6 5 3f 7.3 3f 7.9 3J 8.3 3^ 0.40m 5.3 5i 6.5 4i 7.3 4 7.9 4 8.3 4 varying from 5.3 to 8.3 on the pH scale, while in the pepton solution alone the range is somewhat narrower. Beyond the range for optimum growth there seems to be a decided retarda- tion for each small change of H-ion concentration. These results are brought out more clearly in figure 1 which shows a pronounced widening of the limiting pH values for growth with added NaCl, especially in optimum salt concentrations, and a retardation of growth for each small change in pH near the luniting values. In order to ascertain if there was actually a shifting of the limits of growth, or merely a widening of the optimum range for growth, the effect of NaCl was tried at pH values representing the approximate limit of growth in the acid region. At a pH of 4.8 it was found that only rarely would Bact. coU grow in 1 per cent pepton at 37°C., but that it did grow quite readily in the SALT EFFECTS IN BACTERIAL GROWTH 467 4 1, i- r ■ .- i; ■•: : :■..:. '\.-' ' v,5 '— ■:•■■•'; ■■ ^9 1» // Jr^ - ^- /2 r- 13 -- 14 15 h : . -r ,rixl;0 r sted sted sted •a (0 r - " - '■ ., [^ '^ >. f- ^ O tc ^ 1^ »J X. ^ o is >• fc « h. O r; CT H B! ■ , •■ l-l C w -: *; = -J B ; « K 0. H < p * B w H K S 2 " j; ^- O H 5 E h « M o r - (1> V u w V •J o o coOwJJ^iJtwwO O m caw, 1.0_'J-J-:a g s g S :: g ;> f= s o OCOOrtTlrrcOiO o l-H UJ 1 u «2 2 a o > « S CO eo CI c-1 M o ^ o o ■>!• CO «. u iJ^gsei? s gfeS*50Bfc;Suo t^OJ-H— 'Cn — OC^I ^H >o ? 1-H .-* CO ^ -^ ^H ^ I- C-1 ^ u K°H Ofet-jemOp "o 1 a Ci^O^CO^iOC: ^ O o n ;& •«* w *o >> >. ;». >. >. X >> >> >> e u U ^ t^ t^ 1^ t- b> L. 0) O CJ O O O QJ « o •S >.>>->>>j3> > > C^' O S9 oooooo-»jo o o Si ocoooocJo o o 3 a>(DCjoa>cj^o (U (U o rtP^C^C^p^p^Qp:? ^v^ rt 1 1 1 3 03 h 'zr'tT'Zr'tr'fcr'^ tr'^ ■3 K a>a>aj(i;GJC c bi bO ■< .2 acccccc— a so X X X >^ X K X — T3 s "S (uccoojo^ci ri C -3 O a ^ feBSSfefe-'-S _c c3 (P a; tt *o ^o3 ^fc^lr^^&b^^l rfi C3 t^ CI <^ o — a .2 ■- .2 e c :° b. « flflcaccflo "t^ aj "t^ CO o o o Q a^>>%>>>^>^>^>>t> Ml 5 bC-^ — >. OCCPQQO ■3 "a g 2 0 a 0 CO "o g " 3 .. o o ■*^ s '3 a o -*j 0 *j - IE go m ;,5 ■a T3 .2 -E 3 " C -a '3 (3 0 0 0) m Z^ (-1 3 03 0 -3 S £ J ii o ^ 'S 0 *^ « 0 '3 0 ■" "^ -a 0 0) S '^ S ° !3 0 3 -2 =3 ^ CO ^ 0 S 0 0 2 5& 3 r& 3 '0 0 t: 1. 4; ^ 3 O "3 3 0 3 "5 Ei rs 3 s ii c3 cj to CC M M CO 02 H -'»= = g s s^ 0 < " « ? of^ ».^ I. J 01 ^ h fc'^ K S W P. f^ J H ca c3 03 cS M i~t ^ (-• COLONI SUBCUL PLATE A TWENTY HOURS A o o »-H o f-H a» Qi V > > > > OJ a> 0 ID CD DQ m M n 1 , 1 , 1 , U 1 o 0 C £3 3 3 3 3 3 3 ^ o o o o 0 0 0 t, t4 b 0 t4 l-H jj tD « Q p td 3 D a; U3 2 3 PH o £ a o c o o c 0 0 3 ^ 2 '5 ?1 ^ 03 o D, a, c 0 5 " < a, OJ 3 o a> (D 3 O 0) 0. 3 0 OJ 0. c 0 0 .Sec 0 2 0 03 C. 0 3 t. >) 1-. J= t!:: ja u> JS '-' >-. t-i C c3 t- a c3 O 3 o3 =3 1 3 3 1 3 c! 3 cu ^ 3 0 T3 Ed ja ■« ^ -t^ «o <•-> lO -4-* M -^ T3 U 1 .*J 0 1 O CO 'J' lO 1 CO S5 -«J CQ •^ »-t 0 ■* 'm 1 3-S 1 I "3 1 x> o :3 ^ 3 I 0 33 1 X! 3 *-> ^^ ^ CO ,-. 00 ^^ a)" ,-v "3 -i-i C- 3 '© w 1— ( 1 1.1 en S3 *"* ^■| CC CO 3 m m -u « CO 3 S H :3 u 3 Pi h 3 i-i 3 CO 3 a :3 t^ < Eg CO fl^ « « -4-3 rt^S )— t P^ -2 «*: ■^^ a a 3 13 ^3 -a c 0 CO ^^ T3 S 13 2 £^ « o cl © 03 03 c! a> a .2 03 3 '"^ g =';j Ses CO m m OT 0 CO 'S CO 0 cc (fi CO ■n 0 CO ^ CO ** w ^ We o s 1 £ s o 3 a o £5 ^ a w a Sa 0 Sa f-i t^ u bl 0 (m 0 o >H - «!- "" 3^ t*H -^^ «*H 1^^ >- u 3>^"- a^'-^ SI J s -a t3 1- TS -§ s -^ IH "0 0 I. -0 -3 S ^^ 0 ^^ -o c ■c 0 2 "=> •< 0) (U O 0 1 n '3 a "5 a ■3 S c 2 i-i _« d ci X 0 '3 H c .5 9 >! C3 C c =5 c q X -3 0 E ■*" S s PiH E ^ E fe NATURE OF BACTEniOLYSANTS 497 I j: - j= : J3 ; J3 ; j3 : X : ^ ; t4 CJ M U fct o "fc.1 o Vi o ^ CJ u u *i-i eS "5 rt c! « 03 Cj 03 a C3 d 03 03 03 _d "a U4 it CJ — 03 3 CO (» £ 3 bo £ OJ — £ t OJ 3 to £ 0) =3 .^ 3 •" 3 •4^ 3 T" '-^ 3 '~ CJ •S 3 o a ■0 S 00 O 0] -T3 3 CO 2 u a 3 o 3 2 CJ 3 " CJ J3 — 3 rt CO ^ CJ 1 "a u 3 CO 3 u CJ u •a 't-t '3 c o -^ '^ >■ ci > a OJ i; c3 ® > d 1 " 'a5 o > CJ CQ o 3 CJ m £. OJ .£ tn 00 1" to ■n " S S CJ B a 2'^ .2 E en OJ CJ o T3 £ 3 £ 3 CO 03 — - d to a- T3 a en 0) T) G o -0 a en '3 o to o5 g g m '5 to o 3 to B '3 3 CO CJ CJ 3 w _a "3 B Ol CJ CJ "30 3 "" '3 O CJ -g o £ o CO a; 1 c 3 CO CD OJ .a 3 a to" _aj 3 3 CJ 3 CO XI 3 ■q 3 o O *© o o 'o ■•^ o O o a o O S CJ 'o o 3 •^ o t) t- o o j; o o O (5 CJ CJ S CJ o ^ CJ CJ o M 5S H H :?; H H H H y. "3 "5 "3 "5 "3 3 "S 3 ^ b Ih u> Im (H (h (H ri o > 02 f-t > eg 0} OJ > o CJ CO OJ > OJ 02 3 3 1- IS. 1 3 ~ o o CJ 5 o " u bl t- •"* Ui .— 1 £ fc- ^ 1-i *— * £ a! "5 "S " £ 3 OJ d OJ 11 OJ 3 fs 5= i.s !* .a 3 ? ^ t- 3 CJ 01 a o o 11 1^ «.a c -^ CD 3 CJ B O a 5 a a 3 CJ d be -♦J Q. a O _2 o u OS bC a a a O b£ d d >. « >^ 0) w ® .4J >. » <»:> >. 1 >, OJ >^ QJ .- CJ -1-3 >^ * d h c5 t- >^ t- c ■^ t: fl C5 ^— ca M 03 t- >> l- a d C~' ■? 3 ■a 3 cS 3 o ■r 3 1 -«-* 03 •a ■? 3 ■72 « 3 -T3 -^ 1 »o 00 «> o > >l >* >; >1 >> X ■~^ >^ >^ !^ ^^ >^ cj a C3 c3 C5 ■s ee 03 .-^ c3 d ^ d b. L- '•^ »-. '— ^ L- tH Ui t- ^^ ^ i;:^ u u — - t- •a 3 E 3 II S 3 E cS 3 3 a c 3 «i5 II 3 a 3 3 5 a^ be •3 bC B t£ bU bt O o • d O - Q Q l« •5 O Q c5 d o 03 . d P d« ►S d ^ d ^d Jd OJ d CJ d d -go "id d d "Sd "— ' — ' ■*■ — ■ M M ■ k; m w . c4 p4 w ""h w a ^ — H •^ w ^w « -w >< Ih* K-l (-•* >" ^ > -• >- u >i t^' >H t-<* >-l t," Lh C >H u ^P' Q Q Q 0 o 0 Q p p p Ih U M tH Lri u U t. Pi (-) X E o a y. E o is 22 c E o o a •y. s o CJ B y. E o ii e £ OJ a 1 « »« a> 0) .„ ,£ *^ o CJ u-l OJ 0) C2 OJ c: CJ OJ C3 ^ fc< U^ ►^ fe P^ P^ fa S fa fa JOURNAL OF BACTERIOLOQT. VOL. Til, I70. 5 498 WILBURT C. DAVISON « < n "^ J2 3 .3 w "3 ^ 3 *" o < - gS oj- : o D J s J I •*^ <^ "3 3 3 03 0! 3 hn t4 § a ci c3 ca ^ 4J Si >-, >-> >■ o. a a >> •a T3 C3 -a s. Q. p. T3 c^ c— c- >> oJ 2 3 §> £ 3 o3 s 3 e^ -*-> -*-) -n -^ o O o 1 O CM CM c-i *-i CO =3 & e o CO P c3 O. c3 P. -2 3 CO o 3 " .9 •o „-£ S o .2 i- c3 a CO WW ^ , SK HI W o O a) *A 0 . E •a S: i W -' ^ ■; E c3 O « CJ — - E s3 t^ e3 . b< cj .9 2 5 5 <= NATURE OF BACTERIOLYSANTS 499 STUDY OF THE COLONIES OF NORMAL STRAINS OF BACT. DYSEN- TERIAE (fLEXNER) In sixteen out of twenty agar subcultures of four old laboratory strains of normal Flexner bacilli as well as in three out of eighteen agar subcultures of twelve freshly isolated strains of normal Flexner bacilli, I have noted one or more colonies with irregular TABLE 13 liaclerioli/tic aclivily of fdlrates of peptone water, cultures of normal dysentery bacilli (Flexner) STRAIN or rLEXNSB DT8ENTERT BACILLUS TTPE OP COLONT NUM- BER OF CUI^ TUBES FIL- TERED LENGTH OF INCUDATION AT Sy'C. DEFORE FIL- TRATION NUM- BER OF ACTIVE FIL- TRATES PER- CENT- AOB OF AC- TIVE FIL- TRATES per cent Flexner Y (Hiss and Russell) Regular 8 5 hours to 4 50 (obtained from the Lister 152 days Institute, London) Flexner Y (Ledingham) (ob- [ Regular 5 1 to 91 days o 40 tained from Dr. E. G. D. ] Irregular* 2 1 day 2 100 Murray, London) [ Flexner Y (Sada Gopall) f Regular .5 1 to 91 days 4 80 (obtained from Dr. E. G. I Irregular 3 1 to 15 days 3 100 D. Murray, London) [ Flexner 106 (isolated from Regular 1 5 days 1 100 patient in II.L.H.) Flexner 120 (isolated fromf Regular 1 1 day 0 0 patient in H. L. K.) \ Irregular 1 1 day 1 100 f Regular 20 5 hours to 11 55 Total ] 152 days Irregular C 1 to 15 days 6 100 * Colonies with irregular edges that are sometimes present in subcultures of normal strains. edges among the mass of regular, round, typical, dysentery colonies (table 12). These colonies were not as "irregular" and "moth eaten" as those which were noted in subcultures of "sensitive" strains but nevertheless resembled them. Successive 500 WILBXJRT C. DAVISON subcultures of some of these spontaneously "irregular" colonies contained "irregular" colonies for as long as ten generations. Subcultures of others contained nothing but regular colonies. The "irregular" colonies did not appear to be as viable as normal colonies and subcultures occasionally failed to grow. Inasmuch as these "irregular" colonies occurred more frequently in sub- cultiu-es of old laboratory strains than of freshly isolated strains, it is possible that aging may influence their development. Growth in media containing lactose also appeared to favor their development. Filtrates of cultures of these "irregular" colonies were somewhat more lytic than filtrates of normal cultures (table 13). It is possible that these "irregular" colonies of normal strains and the "moth eaten" colonies of "sensitive" strains are related or even perhaps identical. The different degrees of lysis produced in different cultures by the same filtrate (part I) apparently did not depend altogether on the presence or absence of "irregular" colonies in these cul- tures as only in six strains were "irregular" colonies noted. FILTRATES OF PEPTONE "WATER CULTURES OF NORMAL STRAINS OF BACT. DYSENTERIAE (fLEXNER) Inasmuch as the bacteriolytic activity of stool filtrates, which had become contaminated with stool or air organisms and had then been refiltered, was sometimes increased, it seemed possible that filtrates of normal cultures theinselves might be bacteriolytic. I therefore inoculated regular and "irregular" colonies of several strains of normal Flexner bacilli into peptone water and filtered these cultures after various periods of incubation at 37°C. (table 13). Eleven of the twenty filtrates (55 per cent) of pep- tone water cultures of regular colonies of five normal strains of Flexner bacilU and all of the six filtrates (100 per cent) of "irreg- ular" colonies were slightly bacteriolytic. The degree of bacter- iolysis was much less than that of stool filtrates and of filtrates of cultures of "sensitive" strains. In none of the instances in which subcultures were made of dysentery bacilU which had been incubated with these filtrates of normal cultures were "sensitive" colonies found. It is therefore probable that although filtrates NATURE OF BACTERIOLVSANTS 501 of normal strains of dysentery bacilli were bacteriolytic their activity was extremely weak. BACTERIOLYTIC ACTIVITY OF FILTRATES OF A PEPTONE WATER CULTURE OF BACT. SUBTILIS NicoUe (Davison, 1922) reported that cultures of B. subtilis and their filtrates, were bacteriolytic for several types of organ- isms. As he had not tested the bacteriolytic activity of B. subtilis against Flexner bacilli, I did so. As shown in table 14, the filtrate of a culture of B. subtilis (obtained from a contam- inated plate) was not bacteriolytic for Flexner bacilli but the TABIj; 14 Dacleriolytic activily offiUrates of peptone water cultures of B. subtilis FILTRATE OB SUflPEKSION t'SED Filtrate of four-day peptone water culture of B. subtilis Filtrate of peptone water culture of normal dysentery bacilli which had been incubated forty-eight hours with the filtrate of B. subtilis culture (above) Filtrate of eighteen-day peptone water culture or normal dys- entery bacilli which had been intentionally contaminated with B. subtilis BACTERIOLTTIC ACTIVITY Oy FILTRATE OB SUS- PENSION AUAIN&r NORMAL FLEXNER BACILLI 0 ± filtrate of a normal dysentery culture to which the filtrate of a B. subtilis culture had been added, or which had been inten- tionally contaminated with B. subtilis, was ' sUghtly bacterio- lytic. However this bacteriolytic activity was no more marked than that of filtrates of normal dysentery bacilli to which nothing had been added (table 13) so it is improbable that B. subtiUs or its filtrates can lyse Flexner bacilU. ABSENCE OF BACTERIOLYTIC ACTIVITY IN TRYPSIN As the optimum reaction for bacteriolysis apparently coin- cides with that of the enzyme, trypsin, I tested the bacteriolytic activity of solutions of trypsin and also compared the action on gelatin of trypsin, bacteriolysants and "sensitive" and normal strauis of Flexner bacilli. 1 per cent and 5 per cent solutions of 502 WILBTJRT C. DAVISON trypsin (Fairchild and Foster) were made in normal saline and 20 per cent alcohol. These were proven sterile by culture. As TABLE 16 Absence of bacteriolytic activity in trypsin SOLUTION OR FILTBATK USED BACTERIOLTTIC ACTIVITT OF SOLUTION OR FIL- TRATE AGAINST CULTURES OF NOR- MAL DYSENTERY BACILLI PRESENCE OP "sensitive" COLO- NIES IN Aoxa StJBCt-LTl'BES OF DT8ENTERT BACILLI AFTER THET HAD BEEN INCUBATED WITH SOLUTION OB FILTRATE FOR TWENTT-POUR HOCKS AT 37°C. 1 per cent alcoholic solution of trypsin 1 per cent saline solution of trypsin 5 per cent alcoholic solution of trypsin 5 per cent saline solution of trypsin F 104 = 5 cc. 5 per cent alcoholic solution of Trypsin added to 100 cc. of an eighteen-hour peptone water culture of Flexner Y bacilli; incubated eighteen hours and then filtered F 103 = 2 cc. 5 per cent saline solution of Trypsin, added to 100 cc. of an eighteen- hour peptone water culture of Flexner Y bacilli; incubated eighteen hours and then filtered 0 0 0 0 0 0 0 0 0 0 0 0 TABLE 16 Absence of gelatin liquefying enzyme in bacteriolysants and cultures of "sensitive" and "normal" strains of Bad. dysenteriae {Flexner Y) {Hiss and Rxissell) BACTERIOLYSANT, CULTURE OR SOLUTION USED F 100 (actively bacteriolytic filtrate) "Moth eaten" colonies of "sensitive" strain of Flexner Y bacilli Regular colonies of normal strain of Flexner Y bacilli 5 per cent alcoholic solution of trypsin 5 per cent saline solution of trj'psin LIQUIFICATION OF GELATIN AT ROOM TEMPERA- TURE 0 (34 days) 0 (48 days) 0 (48 days) + (24 hours) + (24 hours) shown in table 15, neither these solutions, nor filtrates of dys- entery cultures which had been incubated with these solutions were bacteriolytic. That the trypsin used was active and could liquefy gelatin is shown in table 16. NATURE OF BACTERIOLYSANTS 503 ABSENCE OF GELATIN LIQUEFYING ENZYME IN BACTERIOLYSANTS AND CULTURES OF "SENSITIVE" AND " NORMAL" STRAINS OF BACT. DYSENTERIAE (fLEXNER Y, HISS AND RUSSELL) As shown in table 16, gelatin was not liquefied by a bacterioly- sant nor by cultures or "moth eaten" colonies of a "sensitive" strain of Flcxner bacilU nor by regular colonies of a normal strain. SUMMARY "Moth eaten" or "sensitive" colonies were obtained in sub- cultures of dysentery bacilli which had been attacked by a bac- teriolysant. They were subcultured for forty-two successive generations. The media in which "sensitive" strains of Flexner bacilli were grown, the solutions in which they were washed, and the suspensions of disintegrated or ground up "sensitive" baciUi, were all strongly bacteriolytic. In agar subcultures of old laboratory and freshly isolated normal strains of Flexner bacilh "irregular" colonies were occasionally found, which resembled the "moth eaten" colonies of "sensitive" strains. Filtrates of peptone water cultures of normal strains and ex- pecially of their "irregular" colonies, were shghtly bacteriolytic. B, subtiUs was not bacteriolytic for Flexner bacilli. Trypsin was not bacteriolytic and bacteriolysants and "sensitive" and "normal" strains did not liquefy gelatin. CONCLUSION It would seem possible that the lytic principle in d'Herelle's phenomenon is an enzyme. This enzyme is not trypsin. A small amount of the lytic principle is contained in, or produced by, normal strains of Flexner bacilli which have been out of contact with the human body for many years. It is possible that the amount of the bacteriolytic enzyme produced by a culture can be increased by aging, by growth in special media or by contact with external influences such as intestinal secretions, tissue extracts, leucocytes, etc. The action of these external influences is probably to favor the develoment of lysogenic organisms at the 504 WILBURT C. DAVISON expense of the non-lysogenic. This enzyme not only dissolves organisms but also favors the multiplication of bacteria which produce this enzyme. In this way the bacteriolytic principle is carried from generation to generation. It is highly improbable that this phenomenon represents a defense mechanism on the part of an animal against bacterial invasion. REFERENCE Davison, W. C. 1922 Filterable "substance" antagonistic to dysentery and other organisms (d'Herelle's phenomenon, bacteriophage, bacterio- lytic agent, etc.). A bibliographic review. Abstracts of Bact. (in press). CLOSTIUIMT-M IHTHIFICrM (B. PUTRIFICUS BIEN- STOCKj, A DISTINCT SPECIES GEOHGE V. REDDISH and LEO V. KETTGEK From Ihe Sheffield Laboralonj of Bncleriology, Yale University Received <«v puMicMlioii Fcl>ru;iry 3, 1922 There lias been luiuli coiilusiini re^anliiif;; the identity of the organism w hicli w as discovered by liienstock (1884) and which he named Ji!> < KonavH •< euaTiMrf < 600H < < soojjina lOVH I I o o I + ^ w w i^ ++++ + I 4-1 + 11+4- J- I 4-1 III 4-4-4-1 I + fe I o HXXNaH -< ^ M E^ W fc. o » o fe fe fc. P O t> t3 &< tJ t) fc< fc. fc. En fa ow oooowc o w C t- &,o w o o H » W fa w O W fa w w o H o w o w a H O t) O H a a K a o o ^ a !-j a a fa a t> a u> o O ^ fa a ^3 a a a a a fa a a a a ^ a a i:ia a a a o a a o t) a oa oa ofaO o aa o o o fa o a o fa a a a a a a o a a fa fa fa fa o o a fa fa o o fa fa fa a o fa oo fa 0^3 OP a fa t3 o o t3 o taC o fa p fa a a a a P a a a fa a oo o o faO a a a fa fa fa c a a o u S a 3 M ■^ -9 1-3 iJ c c c _; c ;:: V aw a N J2 — .li :3 Si ■^ a S o ^ fa 1-! o ~ :^ cu w ■a c o a .^^ _« 2 !» _^ "o o • '3 _ O O ^J-( I-! O c 5 fl3 0^ fa X 5 c '"' § is CO o o ft >, "E a 3 d m N ^ ja S -S fa hJ ^j- C> 534 H. J. CONN are averaged, there is very little choice between crystal violet, methyl violet 6B, and gentian violet. A few more F grades appear with crystal violet than with either of the others but not enough to be of any significance. For our practical purposes these results indicate that the Gram stain can be made with a dye bearing any one of these three names, provided it comes from a satisfactory source. Quite in contrast to the uniformity of these average results is the variation in the reports from the individual investigators. In spite of the standardized technic used in the work there is not a single sample in regard to which entirely consistent results have been made and practically every sample that has been tested more than four or five times has been reported good or excellent by some investigator and unsatisfactory by some other. This suggests that either the personal equation is one that cannot be eliminated entirely or else that some point in the technic requires further standardization. The chief points which have been mentioned so far as susceptible to further standardization are: temperature at which the cultures are incubated, strength of alcohol used for decolorization, length of time of decoloriza- tion, and nature of the counter-stain used. It is very doubtful whether variation could be avoided by controlling all these factors but the committee hopes to plan a further investigation along this Une simply to test out some of these points in regard to the Gram technic. In a recent paper Burke (1922) has sug- gested the importance of some of these factors, but it seems as though a cooperative experiment might give more light on the subject than any study made by a single observer. In looking up the results of the individual manufacturers of these stains, it is to be noticed first that nearly every sample which has ranked as high as G -|- or E — is one that has been tested so few times in this work that the results are not especially significant. For this reason it is regarded as impossible to de- cide which is the best of the samples tested. It is perhaps of some significance, however, to notice those few samples that have ranked below G — in the general average. These samples are crystal violet of Harmer, H. S. Laboratories, Leitz and INVESTIGATION OF AMERICAN GENTIAN VIOLETS 535 Griibler, the methyl violet GB of Providence, and the gentian violet of Harmer. With the exception of these few samples there seems no question but that the rest are entirely satisfac- tory for the Gram stain, and the American samples have ranked practically as well as the Griibler samples. In fact, the Griibler crystal violet ranks the lowest of any of the samples tested. The chief conclusion from this work seems to be that samples of these three dyes are now produced in this country which are perfectly satisfactory for the Gram stain, apparently just as good as the Griibler samples, and that either crystal violet or methyl violet 6B can be substituted for gentian violet in the gram stain. The committee especially recommends the use of crystal violet for this purpose as it is the one dye of this series for which a definite chemical formula can be indicated, although commercial samples of it are not necessarily pure and may vary in their composition. This variation is probably less than in the case of methyl violet 6B, where the name denotes merely a certain shade of violet, and in the case of gentian violet, where the name is not used in the dye industry and is not mentioned in Schultz's index of dyes. For this reason the committee recommends that those ordering stains to be used in the Gram technic should specify crystal violet instead of gentian violet. Turning to the results with the different samples of crystal violet, it will be seen that only one of those investigated was from a basic manufacturer, namely the Du Pont sample. This sample and the Goldin sample have given the highest average grade of any of the crystal violets tested. All of the other crystal violets seem to fall slightly below the Du Pont sample which is rather surprising because there are indica- tions to suggest that nearly all of these dealers in biological stains merely rebottle the Du Pont crystal violet and sell it under their own name. If this is the case, the sUghtly higher rank of the Du Pont sample may be considered to be pure accident. The only other actual manufacturer of crystal violet in this country of which we have knowledge is the National Anilin and Chemical Company. Their product was not included in 536 H. J. CONN this work but there is reason to beUeve that it is carefully pre- pared and cursory examination of slides stained with it by the gram technic indicates that it is as satisfactory as the Du Pont product. In this connection it should be remarked that the Du Pont Company does not sell crystal violet except in bulk, whereas the National AniUn Company is now making a speciaUty of biological stains and sells them in small containers. If the Du Pont product is desired, it should be ordered through some supply house that can be trusted to supply the actual products specified. REFERENCES BtjRKE, V. 1922 Notes of the Gram stain with description of a new method. Jour. Bact. 7, 159-182. Conn, H. J. 1922 An investigation of American stains. Report of committee on bacteriological technic. Jour. Bact. 7, 127-148. INFORMATION FOR CONTRIBUTORS Although there are numerous journals in the United States that deal with various ■peoial phases of bactcrioloRy (as appliot to Medicine, Sanitary Scicncf;, Agriculture ami the like), The Journal of Bactebiology is the unly journal in the ICnglisli language to represent the science as a whole. The Society of .Vmerican Bnctcriologistfi has established the Journal op Bacterioloqt •s its official organ and an a medium for tho discussion of the more general problems of the science — Ihc structure and physiology of the microbes, the inter-relationships of mi- crobio types, the effects of physical and chemical agents upon microbic life, the mutual interactions of microbes growing together in various media, the nutritional needs and products of metabolic activity of various microbes, and new methods of laboratory tech- nique— and similar advances in knowledge which are so fundamental as to be of vital interest to workers in all ports of this groat field. The Jodunal of Bacteuiolooy includes in its scope not only the bacteria but other related micro-organisms, yeasts, molds, pro- tozoa, etc. While it is planned to make the Journal in particular an organ for the more fundamental and general aspects of bacteriology, it will necessarily include many papers whose interest is mainly technical, parlic\ilarly in those fields of bacteriology which have now no satisfactory organ of publication at their disposal. Manuscripts should be sent to Prof. C.-E. A. Winslow, Yale Medical School, New Haven, Conn. AH other communications pertaining to editorial work should be addressed to Major A. Parker Ilitchens, M. C, Army Medical School, Washington, D. C. Twenty-five reprints, without covers, of articles will be furnished gratis to contributors when ordered in advance. A table showing cost, with an order slip, is sent with proof. INFORMATION FOR SUBSCRIBERS The Journal of Bacterioloot is issued bi-monthly, in January, March, May, July, September, and November. A volume consists of approximately 600 pages. Subscriptions to the Journal are taken for the volume only. Under the present plans, one]volume is issued a year. Subscription Price: 1922 volume. Vol. VII, $5.00, United States, Mexico, Cuba; $5.25, Canada; $5.50, elsewhere, net postpaid. Foreign subscriptions to the current volume accepted at par of exchange. Back Volumes are supplied on orders for Vols. I to VI, inclusive. Price, $36.00, United States, Mexico, Cuba; $37.50, Canada; $39.00, elsewhere. Extra single volumes will be supplied on request when in stock at $5.00, United States, Mexico, Cuba; $5.25, Canada; $5.50, elsewhere. Prices are net postpaid. Single numbers can be furnished, in many in- stances, at SI .00 a copy. Foreign ■'!ubscriptions lo the back volumes accepted at the current rale of exchange. Correspondence concerning business matters should be addressed to the Williams & Wilkins Company, Publishers of Scientific Journals and Books, Mount Royal and Guilford Avenues, Baltimore, U. S. A. Subscriptions are received: For Argentina and Uruguay : Beutelspacher & Cia., Sarmiento 815, Buenos Aires, Argentina. For Australia: Stirling & Co., 317 Collins Street, Melbourne. For Belgium: Henri Lamertin, 58 Rue Coudenbcrg, Bruxelles. For the British Empire, except Australia and Canada: Address any British bookseller or send your orders direct to the publishers. For Canada: Wm. Dawson & Sons, Ltd., 87 Queen Street East, Toronto, Canada. For Denmark: H. Hagerup's Boghandel, Gothersgade 30, Kobenhavn. For France: Emile Bougault, 48 Rue des Ecoles, Paris. For Germany: R. Friedlander & Sohn, Buchhandlung, Carlstrasse 11, Berlin NW. 6. For Holland: Scheltcma & Ilolkema, Boekhandel, Rokin 74-76, Amsterdam For Japan and Korea: Maruzen Company, Ltd. (Maruzen-Kabushiki-Kaisha), U to 13 Nihonbashi Tori-Sanchome, Tokyo; Fukuoka, Osaka, Kyoto, and Sendai, Japan. For Spain: Ruiz Herraanos, I'laza de Santa Ana, 13, Madrid. For the United S'ates and other countries except as above: Williams & Wilkins Company, Mottnt Rotal and Guilford Avenues, Baltimore, U. S. A. Claims for copies lost in the mails will not be allowed unless such claims are received within 30 days (90 days, foreign) of the date of issue. Claimants must state that the Publication was not delivered at their recorded address. The publishers will not be responsible for loss due to change of address unless notification is received at least two weeks in advance of issue. ■i CO STANDARDIZED DEHYDRATED MEDIA Prepared According to the Formulae of the A. P. H. A. "Standard Methods" 1920 Bacto-Nutrient Broth Bacto-Nutrient Gelatin Bacto-Nutrient Agar 1% Bacto-Dextrose Broth Bacto-Lactose Broth Bacto-Litmus Lactose Agar OTHER MEDIA IN GENERAL USE Bacto-Nutrient Agar 1.5% (Blood Culture) Bacto-Purple Lactose Agar Bacto-Endo's-Agar Bacto-Eosine Methylene Blue Agar Bacto-MacConkey's Agar Russell Double Sugar Agar Bacto-Lead Acetate Agar Bacto- Starch Agar Bacto-Lactosc Peptone Bile Bacto-Lltmus Millc Bacto-Purple MiUi Bacto-Loeffler's Blood£Serum Krumwledc Triple Sugar Agar MEDIA FOR FERMENTATION STUDIES Andrade Dextrose Broth Andrade Lactose Broth Andrade Saccharose Broth Andrade Maltose Broth Andrade Mannite Broth Andrade Dextrose Agar Andrade Lactose Agar Andrade Saccharose Agar Andrade Maltose Agar Andrade Mannite Agar Carried in slock by the principal dealers in scientific supplies Specify "DIFCO" The trade name of the Pioneers In the research and development of Bacto Peptone and Dehydrated Culture Media. DIGESTIVE FERMENTS COMPANY DETROIT. MICHIGAN, U.S.A. THIS NUMBER COMPLETES VOLUME VII VOLUME VII NUMBER G JOURNAL OK BACTERIOLOGY OFFICIAL ORGAN OF THE SOCIETY OF AMERICAN BACTERIOLOGISTS NOVEMBER, 1922 KDITOB-IN-CHIEF C.-i;. A. WiNSLOW It is characteristic oj Science ajtd Frogrtss that they continually open new fields to onr visions. — Pasteur, PUBLISHED BI-MONTHLY WILLIAMS & WILKINS COMPANY BALTIMORE, U. S. A. Entered as sccond-claas matter April 17, 1016, at the poatoffice at Baltimore, Maryland, under the act ol March 3, 1879. Acceptance lor mailing at special rate of postage provided for in gection 1103, Act of October 3, 1917. Authorized on July 16, 1918 Copyright, 1922, Williams & Willsina Company T>rinfl not C $5.00 pei volumc, United StatGS, Mexlco, Cubd nostnaid i S5.25per voliime, Canada posTi)aia I j5 QQ pgj. volume, other countries Made in United Slnif-^ r,< America PEPTON Perfectly serviceable for the formulas and in all the technic of the bacteriological and antitoxin laboratory. It is employed in the usual proportions and for whatever purposes Pepton of this moat desirable quaUty is required. Manufactured by Fairchild Bros. & Foster New York '■*-«&«.■ '-ARAB1N0SE ' ' , . frANSTIEHL MADE IN U.S.A. AMINO ACIDS Pfanstiehl The Standard for Research THOSE who know from experience the high standard of Pfanstiehl Sugars, Laboratory Reagents and Blood Test Solutions, will be interested in the availa- bility of the following PFANSTIEHL AMINO ACIDS: a-Alanine Glycine (GlycocoU) b-Alaniae Histamine Asparagin Histidine (dichlorid*) 1-Aspartic Acid leucine Cysteine Hydrochloride Lysine Picrate Cystine Phenylalanine Glutamic Acid Tryptophan Glutamic Acid Hydrochloride Tyrosine CARRIED BY THE LEADING DISTRIBUTORS Write for Hat o' Rare Sugars, Laboratory Reagents and Blood Teat Solationt aoailable in Standard of Pfanalimht SPECIAL CHEMICALS COMPANY DEVOTED TO TMt GMCMICAL INDEPENDENCE Or AMERICA Highland Park. III. JOURNAL OF BACTERIOLOGY OFFICIAI. ORGAN OK THli .SOCIKTY OF AMERICAN UACTF.RIOI.OGISIS DEVOTED TO THE ADVANCEMENT AND DIS- SEMINATION OF KNOWLEDGE IN REGARD TO THE BACTERIA AND OTHER MICRO-ORGANISMS Editorial Board Edilor-in-ChieJ C.-E. A. WINSLOW Yale Medical School, New Haveo, Conn. A. Parker Kitchens Lore A. Rogers, Ex officio Advisory Editors C C lUss F. P. GoRHAM C. E. Marshall M. J. Rosenad r'. e'. Buchanan F. C. Harrison V. A. Moore A. \V. Williams P F. Clark E. O. Jordan L. F. Rettger H. Zinsser F. P. Gat C. B. Lipman L. A. Rogers CONTENTS J. 0RSKOV. Method for the Isolation of Bacteria in Pure Culture from Single Cells and Procedure for the Direct Tracing of Bacterial Growth on a Solid Medium 537 William S. Sturges and Leo F. Rettger. Bacterial Autolysis 551 T. M. Rivers. Bacillus hemoglobinophilus canis (Friedberger) . (Hemophilus canis Evxend) 579 J. M. Sherman, G. E. Holm and W. R. Albus. Salt Effects in Bacterial Growth. III. Salt Effects in Relation to the Lag Period and Velocity of Growth 583 Leon S. Medalia. Further Observations on "Color Standards" for the Colorimetric Determination of H-Ion Concentration 589 John Weixzirl. The Cause of Explosion in Chocolate Candies 599 Selman A. Waksman. Microorganisms Concerned in the Oxidation of Sulfur in the Soil. IV. A Solid Medium for the Isolation and Cultivation of Thiobacillus thio- oxidans ^^ Selman A. Waksman. Microorganisms Concerned in the Ox,idation of Sulfur in the Soil. V. Bacteria Oxidizing Sulfur Under Acid and Alkaline Conditions 609 Abstracts of .\merican and foreign bacteriological literature appear in a separate journal, Ahslracls of Bacteriology, published monthly by the Williams & Wilkins Company, under the editorial direction of the Society of .Vmerican Bacteriologists. Back volumes can be furnished in sets consisting of Volumes I, II, III and IV. Price per set. net, postpaid, S24.00, United States, Mexico, Cuba; S25.00, Canada; $26.00, other coimtries. Subscriptions are in order for Volume V, 1921. Price, per volume, $5.00, United States, Mexico, Cuba; $5.25, Canada; $5.50, other countries. JOURNAL OF BACTERIOLOGY STOKES AUTOMATIC WATER STILLS FOR GAS, GASOLINE, KEROSENE, ACETYLENE, STEAM OR ELECTRIC HEATING THE TEST IN OUR STOCK FOR IMMEDIATE SHIPMENT OF SERVICE 4670, 6 burners 4674 4670. Distilling Apparatus, Stokes, for gas heating, with a(ljustal)le burner permitting the use of any kind of gas. Capacity per hour, gallons i | 2J Number of burners i 2 6 Each 4671. 4672. 4674. 25.00 27.00 60.00 Distilling Apparatus, Stokes, for electric heating, identical in appearance with No. 4670 but without gas burner. The heating element is enclosed in a copper tube in the bottom of the distilling chamber. Capacity per hour, gallons 5 § 2\ Each, for 110 volts 45.00 47.00 90.00 Distilling Apparatus, Stokes, for steam heating, small model, similar to No. 4670, but for heating by means of ste.im coil inside of boiling chamber. For operation on 25 lbs. steam pressure. Capacity per hour, gallons 1 3 Each 30.00 75.00 Distilling Apparatus, Stokes Combination, U. S. Army Medical Department model, consisting of Still with a capacity of 1 gallon per hour, with copper steam coil inside of boiling chamber for heating by connection with high-pressure steam line. With clamp on condenser column for supporting an acetylene or kerosene burner for healing where steam connection is not available. With gas burner as supplied with No. 4670 31.00 Prices on small model for gasoline heating and on large models for steam heating quoted on request Prices subject to change without notice ARTHUR H. THOMAS COMPANY WHOLESALE, RETAIL AND EXPORT MERCHANTS LABORATORY APPARATUS AND REAGENTS WEST WASHINGTON SQUARE PHILADELPHIA, U. S. A. CABLE .ADDRESS. "li.VLANCE," PHILADELPHIA METHOD FOR THE ISOL.\TION OF BACTERIA IN PURE CULTURE FROM SINGLE CELLS AND PROCEDURE FOR THE DIRECT TRACING OF BACTERIAL GROWTH ON A SOLID MEDIUM J. 0RSKOV Assistant Bacteriologist of the Slate Scrum Inslilutc, Denmark (Dr. Th. Madsen) Received for publication January 17, 1922 By a pure culture we understand, as is well-known, a culture consisting of individuals of which we know with certainty that all are descended from one single cell, and from one only. As all bacteriologic work, of whatever kind it may be, depends on our working with such rehable pure cultures, many efforts have, of course, been made in the course of time in order to devise reliable methods of isolating a single bacterium. The first investigator who solved the problem in a satisfac- tory way, — although not in regard to the bacteria proper — was Emil Chr. Hansen. The principle of his method was, briefly stated, to observe directly under the microscope the growth of the individual yeast-ceU until it has formed a small colony in a gelatin droplet on the lower surface of a coverglass in a moist chamber. Yeast-cells are however far bigger than most bac- teria, and there is no possibility of tracing with any certainty the growth of a bacterium, as for instance a colon bacillus, in a similar way in gelatin. Of methods that have been proposed and employed for single cell cultivation of bacteria, the best known are those of Schouten, Barber and Malone, none of which have however attained any extensive application, no doubt partly owing to the intricate apparatus they require, and partly to the difficulty involved in picking up such minute objects as bacteria with such relatively coarse implements as pipettes and loops; and when Barber states that he is able to pick up successively each single one of four 637 THE JOURNAL OF BACTERIOLOOT, VOL. VII, NO. 6 538 J. 0RSKOV bacteria, which he views in a small hanging-drop of broth, and to inoculate four broth test tubes with each of them separately, one feels predisposed to doubt the possibility of ever acquiring such practice. The method most generally used at present is, no doubt, Burri's India ink method, which is the starting point for the pro- cedure described in the followdng communication. As is well known, the principles of the India ink method are, briefly stated, these: the bacteria are emulsified in diluted India ink, of which emulsion minute droplets are deposited on a gelatin plate in a Petri dish by means of a mapping-pen. Those droplets which, by microscopical investigation with a high power dry lens, prove to contain only one single cell, are noted and allowed to stand until a small colony has developed, from which subcultures are prepared ; or, the India ink droplet is removed, together with the bacterium, by means of a sterUe coversUp that is superimposed on the black spot of the gelatin plate, removed again together with the India ink and the bacterium, and dropped into an appropriate fluid nutrient medium. This is, as has been said, an excellent method, by means of which, with some practice and patience, rehable single-cell cultiires of most species of bac- teria can fairly easily be produced. (It is, however, not all bacterial species that will stand the India ink.) A drawback in Burri's method is the necessity of having the unhandy Petri dish standing on the microscope stage during the examination. Therefore, I devised a modification: by means of sterile Pasteur pipettes I poured liquid gelatin upon previously sterilized slides. On the gelatin surface three rows of India ink droplets were deposited, which could now be much more easily and rapidly examined by shifting the mechanical stage, the se- lected India ink droplets being subsequently removed as usual by means of sterile coverslips. If it is desired to trace the development of the new formed elements on the gelatin, the slides are placed in a sterile Petri dish with a piece of moist filter paper at the bottom. In this way the India ink spot can be examined at intervals, and the development can be observed. The image however, will rapidly ISOLATION OF BACTERIA IN PURE CULTURE 539 become blurred under such conditions. The India ink will be broken up, the new-formed elements pushing in beneath it; even though the India ink be highly diluted, which facilitates observation, the image will rapidly lose its sharpness. Our objective being, in my case, partly to obtain single-cell cultures from some atypical bacterial elements, and partly to trace their development, it seemed natural, to attempt to do completely without the India ink, the multiplication of the bacteria being much easier to follow, the more dilute the India ink. When this was done I admit being surprised at seeing how well the organisms showed up as sharply defined, highly refractive ele- ments, readily distinguishable from other chance particles or impurities on the surface. Gelatin is however not a particularly suitable medium for most bacteria, and growth at 37°C. covild not be observed in this manner, so I tried whether the bacteria were as clearly visible on an agar surface. In order to study this point, agar ^common filtered broth agar) was melted, and by means of a coarse Pas- teur pipette poured over a sterile slide. The bacteria were even more readily discernible on this surface than on the gelatin. It is difficult however to procure a perfectly level agar surface in this maimer. This difficulty was overcome by abandoning the pouring of the agar on the slide and, instead, excising the medium out of an agar plate in a Petri dish by means of a knife, pre\'iously sterilized in a flame and cooled, lifting the excised cube of agar on the blade of the knife and depositing it on the steri- lized slide to which it will immediately adhere. The essential conditions were now provided for tracing bac- terial growth on a solid medium, especially if certain difficulties could be overcome in regard to ensuring reliable and readily obtained single-cell cultures. At this point in my investigation a paper appeared in The Journal of Hygiene by Hort, in which he describes a method, the underlying principles of which are the same, namely, the observation of unstained bacteria without a contrast on an agar surface, partly by means of oil-immersion lenses, partly by a system of high-power dry lenses. Having given a review of the usually employed methods of isolation, 540 J. 0RSKOV all of which, including Burri's method, he rejects, he suggests a new method of isolation ha\ing two modifications, one in which examination is undertaken with an oil-immersion lens, and one in which a dry lens is employed. In bacteriologic Uterature, Hort is thus the first to point out that it is possible to see bacteria distinctly on the surface of an agar plate and to watch their growth by means of a high-power system of dry lenses without any staining or contrast. It is true that HiU, in his work on the morphology of the diphtheria bacil- lus, mentions that diphtheria baciUi can be seen distinctly on agar by means of a high-power dry lens, but he does not seem to reaUze the possibUities involved in tliis fact. It is no doubt possible, and I think, probable, that others too have been aware of this fact; Hort is however, as stated, the first who has defined it and understood how to make use of it. The medium which Hort employs in his examination, he pre- pares in a manner similar to the one originally appUed by me. He pours the hot agar over sterile slides in as level a layer as possible, taking care to keep the agar well within the edges of the shde. The mode of procedure now varies as to whether he employs the oil-immersion lens, or the chy lens system for further examination. In the first case a series of sterile covershps have been previously prepared, each with a small circle etched on its surface by means of a diamond. In the center of this circle, a minute droplet of broth is now deposited, taken from a broth culture containing the bacteria under investigation in a suitable emulsion. The inoculated coverslip is now placed face down- wards on the agar surface; the area within the small circle is thoroughly examined with an oil-immersion lens, and, in case only one single organism is found witliin the circle, the slide is placed in a Petri dish which is then placed in the incubator. The circle is examined at short intervals, until a small colony has formed from which subcultures can be prepared. Hort himself remarks, in regard to tliis procedure, that great care must be taken to ensure that the droplet does not run out- side the etched circle when the coversUp is appUed to the agar, adding, however, that with some care this is easily avoided. It ISOLATION OF BACTERIA tN PURE CULTURE 541 is evident that we must feel perfectly sure on this point, consid- ering that the area within the circle is the only one examined. However, it seems difficult to understand how one can be sure that the small droplet keeps within the etched circle, as a small liquid layer will form between the coverslip and the agar, the moment these two are directly applied to each other, which will immediately make the droplet invisible. The possibility will like- wise always exist, of various currents arising between coverslip and agar, both when the slip is placed and removed; and, even though the growth of a colony from one single cell has been observed ^vdthin the circle, this may become contaminated from a small colony immecUately outside the circle, the moment the cover- slip is removed. As a rule, there is little chance of this happen- ing, but it does compromise the reliability of the method. Moreover, there is the question of the power of certain motile bacteria to move on the uncovered agar plate. The possibility of such active motilit}' on the part of the organisms is, of course, increased by the placing of the coverslip on the agar, by which a small liquid-filled space is formed. For his second method Hort emploj^s a medium prepared in a similar way. A highly chluted emulsion of bacteria is spread over the agar plate by means of a glass rod, the inoculated plate being now covered ^^ith a thin sterile strip of celluloid, which has been previously perforated. Small sterile covershps are placed over the holes so as to form minute moist chambers. These are now searched with a high-power dry lens and the cham- bers containing one single cell only, are marked. The examina- tion is now continued as described above. (It is not evident from Hort's treatise, whether the coverslips are removed during the examination under the microscope; if not, one would think that the dew on the coverslip would be obstructive.) The col- onies ha\ing reached an appropriate size, subcultures are pre- pared by means of a special apparatus consisting of a tube with a needle adjusted in a special way and screwed on to the nose-piece like an objective. This method is reliable, but as Hort himself says in his final remarks: "In conclusion it is necessary to point out that cultivation of bacteria from single cells is, even when 542 J. 0RSKOV emplojdng a good method, a most tedious procedure, invoh-ing several hours' close work for each organism isolated, if the re- sults are to be relied on." I beheve the difficulties to have become considerably reduced in the procedure described in the following pages, and, the simpler a method is, the more reUable will it generally be. The prin- ciple upon which the method of isolation described below has been based is, as mentioned, the fact that a bacterium — possibly the minutest ones excepted — can readily be distinguished on the surface of a clear transparent medium, such as for instance agar, gelatin, or ascitic agar. The mode of procedure is, briefly stated, as follows: A young bacterial culture, such for instance as a twelve-hour old broth culture of colon bacilli, is inoculated on the agar plate in a Petri dish. The agar had better not be more than a few millimeters thick (bacteria can also be dis- tinguished on very tliick agar, but less sharply). The upper and lower surfaces of the agar should be parallel so as to ensure that the excised bits shall be of equal thickness everywhere, partly in order to obtain plane images, partly to avoid the risk of running down with the objective into the agar, which is in its immediate vicinity during examination. It is important, in inoculating the culture, to be fairly clear at the outset as to the density of bacteria desired on the plate. In the diagram, figure 1, some dotted Unes show how I am ac- customed to proceed. A big droplet of the broth culture is deposited in the center of the circle, and, by means of a glass rod, bent at a right angle, the drop is pressed down between the parallel dotted Unes. Now the glass rod is moved from side to side across the first inoculated area, and, finally, the remain- ing half of the dish is inoculated and it is placed in the incubator for about one hour at 37°C. (Inoculation can of course also be performed in a streak, which some will perhaps find more to the purpose, and in this way an appropriate difference in the density of the bacteria can likewise be obtained.) Tliis measure is taken because the development in the case of the colon bacillus begins just after the expiration of one hour, and because bac- teria are more readily discernible when in development, o^\'ing ISOLATION OF BACTERIA IN PURE CULTURE 543 to their increased refractive power. As previously described, a suitable scjuare of the agar is now excised and placed upon the previously sterilized microscope sUde, which is most con- \iently sterilized by flaming (cf. fig. 2). The microscope slide plus agar, which, for convenience sake, I shall term "slide" in the following discussion, is placed on the stage of the microscope, and an area is chosen where the organ- isms are placed at a convenient mutual distance, commencing Fig. 1 the examination where thej'' are Ijing close, and thence, by means of the mechanical stage, proceeding to where they are lying more scattered.' Having now come upon an area where there is one organism, only, within the field of vision, and this single bacterium having 1 The objective exployed by me was a Zeiss Apochromat; any suflSciently powerful objective will however do. The magnification, at which I usually worked, was 750 diameters. Illumination isa very important factor. The source of light must be uniform. I used a powerful metal filament lamp with frosted bulb, the light of which was considerably reduced by means of the diaphragm of the illuminating apparatus of the microscope. 544 J. 0RSKOV been centered, the area is noted by means of the mutual rela- tion of the regular scale and the vernier attached to the mechani- cal stage, and one should now be able to focus exactly in the same place again. At this point there is however a great diffi- culty to be overcome, as the least inaccuracy in the re-adjust- ment may have the effect of causing the selected organism to vanish from the field of vision, others being substituted, and, Fig. 2 even though the most painful care be taken, the same fatal accident may happen o\\dng to quite negUgible displacements of the scales, which it is often quite impossible to control. In order to be able to find a particular bacterium again, I have proceeded in the following manner: Prior to placing the agar plate on the slide, a complex of fine lines are by means of a diamond scratched criss-cross, preferably on the lower surface of the shde, over the area to be covered by the agar (cf. fig. 2, above). The lines will become less fraj^ed in the glass if the ISOLATION OF BACTERIA IN PURE CULTURE 545 scratcliing be performed in a drop of immersion-oil. In an objective (preferably a different one from that with which the growth is observed, as the micrometer lines will disturb observa- tion) is placed a squared eyepiece micrometer, which is cemented on to prevent displacement. If we now focus sharply on the scratches on the slide with the low power of the microscope, the lines in the eyepiece micrometer will be intersected by these scratches in a quite specific manner (cf. fig. 3), and thus we obtain two distinguishing marks instead of one. The course of procedure will now be as follows: the agar is placed on the scratched area of the slide, and we search for a place where there is only one organism within the field of vision. The spot is marked by means of the scales of the me- chanical stage, the objective with the attached micrometer is placed in the tube, the microscope is adjusted to low magnifica- tion and focussed sharply on the scratches. Careful drawings are made on squared paper of the position of these scratches in relation to the eyepiece micrometer, the slide being now placed in a sterile Petri dish with a piece of moist filter paper at the bottom. (The filter paper must not be too wet as this may cause the development of so much aqueous vapour during incuba- tion that the glasses become wet enough for the agar squares to sUde, when the whole experiment is ruined.) By means of thus marking the position of the organism we have always succeeded in hitting upon exactly the same spot for repeated examination. The growth is now watched at proper intervals, the adjust- ment being performed so that, firstly, the scales of the mechani- cal stage are placed in the proper mutual relation, which we have noted do^NTi, secondly, with the low power of the microscope we make the scratches on the slide correspond to the proper points in the eyepiece micrometer, and, finally, eyepiece and objective are changed, and we can now easilj^ observe the alterations in the small colony in development. (The slides must of course previously be cooled to the same temperature as that of the objective, in case examination takes place at a lower tempera- ture than that of incubation, as, othel•^^'ise, condensed moisture 546 J. 0RSKOV will gather on the front lens and obstruct \'ision altogether.) In this way, the agar will preserve its shape for at least twenty- four hours, even though subjected to several exanainations, provided these be not of too long duration. Now, do not these repeated examinations involve a great danger of contamination from the air? A risk there is, of course, but it is apparently, insignificant. In the first place it can easily be ascertained that only very few "alien" colonies will be seen Fig. 3 Fig. 4 to develop in the Petri dish by contamination from the air, in spite of repeated examinations within forty-eight hours, and it would be a stroke of very bad fortune if such a germ from the air should settle just within the field of vision to be examined. If this should occur, it would soon be discovered; it could only escape detection in case a germ from the air dropped upon the selected colony immediatelj^ before inoculation was undertaken from it, and then it would most probably be disclosed in further investigations of the bacterial species in question, as it would ISOLATION OF BACTERIA IN" PURE CULTURE 547 be almost inconceivable ill-luck if the "alien" microorganism should be one that was closely related to the isolated one. Now, the new-formed colony having reached a convenient size, sub-cultures should be prepared from it. At this point the colonies have grown so large as to be distinctly visible with the low power of the microscope, the time and mode of re-inocu- lation depending on the relative situation of the colonies. In case the colonj^ the shape of which we recognize with the low power of the microscope, is placed in a sufficiently isolated position, we may defer inoculation until it has grown big enough to allow of our conveniently inoculating from it by means of a fine inoculation needle under the microscope at low magnifica- tion. If there is any danger of neighbouring colonies impinging upon it, we must undertake inoculation while it is yet smaU. As mentioned, Hort used a special apparatus for this purpose. A small harpoon, which one may prepare oneself, will how- ever do. On the front lens of an objective is placed a small lump of modelling wax to which is attached a fine thin platinum wire not thicker than 0.15 mm. with a blunt end (cf. fig. 4). Previous to "harpooning" the colony some preparatory practice is nec- essary. A small agar square is excised and fitted in the usual way and placed on a slide. A droplet of India ink or some other staining fluid is deposited on the agar with a mapping- pen. This spot is now centered in the field of vision at the low power of the microscope, and the objective ^ith the attached platinum needle, which is screwed on to the nose-piece, is directed across the spot, the needle being adjusted by a pressure from the fingers so as to be mounted exactly above the spot, and gently depressed so as to touch the agar. The point of contact is readily discerned with the low power of the microscope and marked by means of the eyepiece micrometer. We know now exactly where the needle will hit, being able then to inoculate from the colony by adjusting it to the exact point in relation to the eyepiece micrometer at which the needle hit last. The needle is carefully lowered into the colony until it touches the agar plate. The point of contact is most readily noticed by 548 J. 0ESKOV following the reflection of the needle in the agar; when the needle and its image meet, contact with the colony has been estabhshed. The point of the needle is now touched with broth in a small loop which is raised so as to encompass it several times, agar is inocu- lated from the loop, and finally the needle is washed in a tube of broth. The "harpoon" is sterilized by flaming, and now it only re- mains to examine the area where the colony was previously situated. The bacteria from the colony will be seen to have been scattered somewhat, and we note whether the adjacent colonies are totally intact, both with the low and with the high power of the microscope. In case growth results from the inocu- lation we know that we have obtained a reliable pure culture. Compared with Hort's dry-lens method, the procedure described presents several advantages. Firstly, it is difficult to pour agar over the shdes so as to obtain an even layer, and it takes a long time. Petri dishes with agar are always at hand in any bacteriologic lab- oratory; these should however be freshly poured to avoid the risk of obtaining pure cultures from chance microorganisms from the air which, being overlooked at inoculation, may have formed small colonies. Secondly, we avoid the perforated celluloid plate which is a hindrance to free operation and means a con- siderable hmitation in the possibility of finding conveniently placed microorganisms. While Hort spends several hours on the pure cultivation of a bacterium by his method, I beheve that the total work in my isolation method \\-ill, in most cases, only amount to a fraction of an hour for each single bacterium. As has been shown, we are able to trace the growth from the single cell until a small colony has developed. Details can of course only be observed as long as the colonies are small and single-laj'ered ; so soon as the colonies are crowded in several layers, exact examination is of course out of the question. If, for instance, we desire to ascertain whether a morphologically at3T)ical element is viable, and to follow its development, we isolate it as described above and observe its growth at proper intervals. If it is desired to get a su^^'ey of the way in which colony formation proceeds we need only, at proper intervals, ISOLATION OF BACTERIA IN PURE CULTURE 549 to excise small bits of the agar plate inoculated with a bacterial culture, and we obtain in this way a far better picture of the actual morphology of the bacterium than by producing prepara- tions according to the usual methods, whether it be the milder procedure of emulsifying the bacteria in a drop of a staining fluid, or one of the various staining methods with previous fixation. If, owing to the minuteness or too crowded placing of the microorganisms, we should fail in distinguishing what we de- sired to see, such as for instance the cell division lines, by means of the dry lens system, we need only place a coverglass on the surface of the agar. Tliis will immediately adhere to the agar, and, by means of the oil-immersion lens we shall be able to de- tect the bacteria quite distinctly and also to trace, the growth of a single element. Any one can readily be persuaded as to the faciUty with which bacteria are distinguished on an agar surface without staining or contrast, by examining an inoculated agar plate after a few hours' incubation. REFERENCES Barber, M. A. 1908 The rate of multiplication of Bacillus coli at different temperatures. Jour, of Inf. Dis., 6,379. BuRRi 1909 Das Tuschverfahren. Jena. Hansen, Emil Chr. 1883-1888 Methode til Fremstilling af Renkulturer af Sacchaocjmaeter og lign. Mikrober. Medd. fra Carlsberg Lab. 2, p. 152. Hill, H. W. 1902 "Hanging Block" preparations for the microscopic obser- vation of developing bacteria. Jour. Med. Research, 8, 202. HoRT, E. 1920 The cultivation of aerobic bacteria from single cells. Jour. Hyg., 18, 361. KtJSTER, ScHOUTEX 1921 Anlcitung zuT Kultur derMikroorganlsmen. Leipzig, p. 59. Maloni:, L. H. 1919 A simple apparatus for isolating single organisms. Jour. Path, and Bact., 22, 222. BACTERIAL AUTOLYSIS WILLIAM S. STURGES and LEO F. RETTGER From the Sheffield Laboratory of Bacteriology, Yale University Received for publication January 28, 1922 The term "autolj'sis" has quite generally been used to desig- nate a process of cell or tissue deterioration in which the more complex nitrogenous substances, particularly proteins, are re- duced to simpler form through the agency of enzjTiies which have been elaborated by the cells or tissues themselves. The self-digestion of yeast, and of Uver and other animal or- gans was recognized by Salkowski (1891) to be due to enzymatic action. An increase in the soluble nitrogenous constituents and the presence of leucine and tyrosine were observed. Jacoby (1900) showed that there is a consistent increase in ammonia during the process and advanced the hypothesis that this con- stituted the mechanism for the elimination of soluble kataboUc products of the normal living cell. Fermi (1890, 1891 and 1894) was among the first to demonstrate the presence of enzymes in bacteria. Extensive comparative studies with pepsin and trypsin showed that the proteolytic principles of bacteria resembled the latter. Our knowledge of the proteolytic enzjones of bacteria was further advanced by Wood (1890), MacFadyen (1892), Vignal (189G), Eijkmann (1901) and Cacace (1901). By the employment of Buchner's " Hef epressaf t" method Hahn (1900) obtained an extract from Bad. typhosum and Mycobact. tuberculosis which he claimed possessed autolytic properties. Brieger and Mayer (1903 and 1904) reported that they had removed agglutinogens and other specific bodies from Bad. typhosum by autolysis. It appears quite probable, however, that the processes which they employed were such 5S1 552 WILLIAM S. STURGES AKD LEO F. RETTGER that osmosis cannot be excluded as an important factor. Con- radi (1903) described an "autolysis" of Bact. typhosum and Bad. dysenteriae whereby endotoxins were liberated. He be- lieved that all bacteria possess autolytic enzjones. Neisser and Shiga (1903) reported similar observations. They found, however, that the sterile filtrates of twenty-four hour cultures which had been heated at 60°C. for one hour and kept at 37° for two days showed the presence of free "receptors" which had the power of absorbing agglutinins and depressing agglutination. The filtrate from the dysentery bacillus culture was as to:dc as that obtained by Conradi. Hence, in view of this heating, one is not justified in saying that the toxin is liberated by autolysis. The autolysis of B. anthracis was described by Levy' and Phersdorff (1902). Microscopic changes were noted in alkaline suspensions to which toluol had been added. Gelatin-liquefy- ing and other enzymes were observed, and the authors claimed that the autolysate was toxic to white mice. However, as much as one thirteenth of the body weight had to be injected in order to obtain these toxicity results. Rettger (1904) appUed chemical criteria to the study of the autolysis of microorganisms and pointed out qualitative changes of considerable significance. He noted the liberation of coagu- lable protein and subsequently the formation of leucine and tyrosine in water suspensions of Erythrobacillus prodigiosus. The biuret test was found to be very valuable in following the course of autolysis. Microscopic changes in stained mounts were followed also. During the incubation of water suspensions of Bact. coli to which toluol had been added there was apparently some liberation of coagulable proteins from the cells. It is impossible to state whether this was the result of real autolysis or of osmotic changes. In Flexner's study of the meningococcus (1907) the interest- ing observation was made that this organism resists disintegra- tion longer when kept at 37°C. than in the refrigerator. Con- centrated suspensions in physiological sahne solution were found by microscopic examination to undergo rapid autolysis. Woll- stein (1907) showed that the meningococcus and gonococcus BACTEUIAL AUTOLYSIS 553 were very much alike in tlieir behavior under the various ex- perimental conditions. Burgers, Schermann and Schreiber (1911) claim to have observed autolj'sis by Bad. coli, Bad. typhosuvi and the pneu- mococcus after kilhng with chloroform. According to them, however, B. megatherium, Staphylococcus and Streptococcus do not autolyze. jNIcClintock and Clark (1909) satisfied them- selves that the rapid lysis which suspensions of the gonococcus undergo is due to enzymatic activities. Heating at 70°C. for one hour prevented the cellular changes from taking place. Rosenow (1912) maintained that substances capable of causing anaphylaxis are Uberated by the pneumococcus, streptococcus, gonococcus, meningococcus, Bad. coli, Bad. typhosum, and to some extent by the staphylococcus. Warden (1913, 1915, and 1917) made a rather extensive study of the autolysis of the gonococcus. In his earlier work he looked in vain for any external factors which might be responsible for the disintegration of the cells, but did observe an "enzyme," trjTDtic in nature, which was always present in autolyzing sus- pensions. In a later paper (1915), however, he concludes that "lysis of gonococci. ... is probably due, not to activities of enzymes, but to other causes, among which water permeability and solution of fatty substances play an important part." AUlau-e (1913), NicoUe (1913) and SaUmbeni (1913) studied the process of so-called "autolysis" by determining the "soluble" and "insoluble nitrogen." They claim to have observed in- creases in "soluble nitrogen" during the autolysis of Bad. coli, Bad. typhosum, Proteus vulgaris, Pseudomonas pyocyanea, the gonococcus, meningococcus and pneumococcus. Corper (1916) and Corper and Sweeny (1918) demonstrated that suspensions of the tubercle bacillus in physiological saline solution, with or without the addition of antiseptics, undergo autolysis at 37°C., as evidenced by marked increases in non-coagulable and amino nitrogen. Dernby (1917 and 1918) working with yeast, and later with animal tissues, was able to show the presence of various proteo- lytic enzymes having specific hj-drogen ion concentration re- THE lOVBHU. or BACTEBIOLOOT, VOL. Til, NO. 6 554 WILLIAM S. STURGES AND LEO F. EETTGER quirements. A pepsin-like enzyme was observed whose activity depended upon a hydrogen ion concentration greater than pH 5.5. In ever}^ instance there were also present enzjTnes of a tryptic nature, most active in a fairly alkaline medium. The acid end of the range of the latter, however, overlapped the alkaline end of the range of the former ferments and a digestion (or autolysis) carried on in the pH zone common to both tended to be much more complete than one carried out in the alkaline range so often chosen for such experiments. Our knowledge of bacterial autolysis is as yet far from com- plete, and the Uttle information which has been acquired is con- fined, with very few exceptions, to the changes which take place in the protein of the bacterial cell. In some instances the in- vestigations have not gone beyond a morphological study, and in very few cases were the changes determined by modern chemical methods. There is considerable difference of opinion in regard to the definition of the term "autolysis." The word has been variously used to express changes in gross and microscopic appearance, loss of vitality, changes in solubiUty and actual changes in chemical composition. From a careful consideration of the phenomena which Salkowski originally observed and for which Jacobi first introduced this term, and in view of its extensive usage in the biochemical field, the authors would define "autoly- sis" as "the breaking down and solution of some of the essential chemical constituents of the cell by agencies (enzymes) originat- ing within the cell." As bacteria are composed mainly of proto- plasm it may be assumed that in ordinary autolysis the hydrol- ysis is chiefly of a proteolytic nature. It is for this reason that the present investigation has been concerned wholly with the changes which the proteins and related complex nitrogenous substances undergo. The logical basis for any serious consideration of the phenom- enon of autolysis must be a study of the chemical changes involved. The hydrolysis of protein or protein-hke substances consists in the decomposition of the more complex molecules into simpler components. This constitutes a change in the BACTERIAL AUTOLYSIS 555 absolute and the relative amounts of the various forms of nitro- gen. The present investigation is essentially an attempt to follow these changes by simple and practical chemical and physico-chemical methods. METHODS AND EXPERIMENTAL PROCEDURE The quantitative biuret test as employed by Vernon (1903) was found to be a valuable method of determining the relative amounts of protein and other closely related nitrogenous sub- stances. Increases in amino and ammonia nitrogen were followed by the Sorensen titrations (1908). Amino nitrogen was determined also by the Van Slyke method (1911 and 1912). Observations were also made of the changes in conductivity of the autolyzing suspensions of bacteria, a procedure used by Sjoquist (1895) and later by Bayliss (1904) in their studies of digestion. The organisms which were selected for the investigation may be placed in three distinct classes. The first is made up of those which are powerfully proteolytic, B. subtilis, Erythrohacillus prodigiosus, and related species. These are all gelatin-Uquefiers. The second class is non-proteolytic and includes the colon- typhoid group. The third includes certain of the pathogenic cocci which possess little or no proteolytic power, namely Stct^ phylococcus aureus, Streptococcus pyogenes, the pneumococcus, gonococcus and meningococcus. The non-pathogenic organisms were grown on plain two per cent agar. Some of the pathogenic cocci required special media. A luxuriant growth of the pneumococcus, gonococcus and men- ingococcus was obtained by the addition of one part of ascitic fluid to two parts of three per cent agar. An even more satis- factory medium was the testicular extract agar which Hall (1916) recommends for the cultivation of the gonococcus. This medium has the advantage of not being affected by ordinary sterilization methods. The meningococcus was grown on liver extract agar in the experiments which involve amino nitrogen determinations. In one instance the pneumococcus was grown in fresh extract broth and the organisms removed by centrifuging. 556 WILLIAM S. STURGES AND LEO F. RETTGER Kolle flasks were empioj'ed in order to supply a maximum surface for growth. The condensation fluid was removed with a pipette and the surface allowed some time to dry, before inocu- lation. The inoculum was taken from j^oung test tube cultures of the same media, and spread over the surface with a bent glass rod. The flasks were incubated for varying lengths of time, usually until the maximum growth was obtained. Bad. coli and the different cocci were grown at 37°C. The other cultures were incubated at 24 to 30°. The harvesting was accomplished by scraping the growth from the surface with a stiff steel Tvire and transferring to sterile glass-stoppered bottles. Suspensions were then made by adding sterile O.S per cent sodium chloride solution, or in some instances distilled water, and shaking vigorously. After filtration through sterile cotton two to five per cent of toluol was added and the bottle well shaken. After the withdrawal of a small portion of the suspension for preliminary examination, the bottles were placed in the incubator. They were shaken frequently and samples withdrawn from time to time for analysis. The different analyses were made as follows: The biuret test Twenty cubic centimeters of 4 per cent NaOH were measured into a Nessler tube and 2 cc. of f^ normal CUSO4 added. A measured amount (usually 0.1 to 0.5 cc.) of the supension was introduced with a sterile pipette. The amount was gauged by the intensity of color that developed. The most accurate readings are made if the test is so regulated as to give a com- paratively faint color. Color comparisons were made with standard tubes to wliich definite but varying amounts of Witte's peptone had been added. The color strength was expressed in the terms of the percentage of "Witte's peptone to which the unknown was equal in biuret-giving properties. It was found advantageous to use 0.1 per cent "Witte peptone solution which (within the range of 0.1 to 2 cc.) in the standard tubes gave readable differences in 0.1 cc. gradations. These represent gradations of 0.1 mgm. of Witte's peptone. One cubic centimeter is a fair amount of the unknown to use for a test. One tenth BACTERIAL AUTOLYSIS 557 millipram of peptone in 1 oc. is a 0.01 per cent solution; 0.2 mgm. is a 0.02 per cent solution. It is possible, therefore, to read differences of 0.01 per cent. Slight cloudiness occasionally obscured the color to some extent. It could often be removed by centrifuging or allowing the turbidity to settle out. The Sorensen test Five cubic centimeters of the bacterial suspension were added to 30 cc. of distilled water. After neutraUzation, 5 cc. of neutral formaldehyde were added, and the resultant acidity titrated with N/20 sodium hydroxide, phenolphthalein being used as an indicator. The titrations were made immediately after the formaldehyde was thoroughly mixed with the test sample. As the results were to be used for comparative purposes only, the titrations were expressed in the number of cubic centimeters required to neutraUze 100 cc. of the undiluted suspension. The amount of nitrogen in milligrams may be determined by multi- plying this figure by 0.7. Conductivity determinations These were made with 10 cc. portions of the autolyzing ma- terial and in a conductivity cell having bright electrodes. Ap- paratus and methods conform with those in use in ordinary class room work. The electrodes were not platinized, owdng to the possibility of a disturbing catalytic action being exerted by the finely divided platinum black. The zero pomt on the Wheatstone Bridge as determined by the minunum buzz in the telephone receiver was consequently not as distinct as when electrodes are platinized, but it was found that with a little practice duplicate readings could be made which checked within 2 mm. Even this slight variation was reduced by taking at least six readings within the middle third of the Wheatstone Bridge scale, emplojdng different resistances. The determina- tions were made at 25°C. Tlie Van Slyke method for the determination of amino nitrogen This method was found to yield valuable results when cer- tain necessary precautions were observed. The possibilities 558 WILLIAM S. STURGES AND LEO F. RETTGER of variation could never be sufficiently reduced, however, to place much reUance on a single determination. At least two determinations were always made, and when these gave results differing from each other bj^ more than 10 per cent the process was repeated until satisfactory checks were obtained. The difficulties encountered were in all probability due to the presence of bacterial cells and other suspended matter in the test material. It soon became apparent that the number of oscillations had to be regulated (240 per minute) and that more than five minutes was required to obtain a complete reaction. Ten minutes was finally adopted as the routine time. This was followed by an additional shaking by hand for one or two minutes while the motor was being used to shake the Hempel pipette. The addi- tional gas Uberated at this time was also analyzed and its volume of nitrogen added to the first detennination. This gave a means of judging whether the reaction had been completed in the time allowed. Because of the extra tim.e and labor required to ob- tain rehable results, this method of determining the amino nitrogen proved less practical than that of Sorensen. The original method of Van Slyke (1911, 1912) was adliered to as closely as possible, and the calculations of amino nitrogen made by reference to the tables. EXPERIMENTAL DATA Early in the investigation the course of autolysis was followed \vith the biuret and Sorensen tests. The results appeared to indicate a direct relationship between proteolytic acti\'ity and gelatin-liquefying power of bacteria, on the one hand, and autol- ysis on the other. Thus, B. subtilis suspensions underwent complete autolysis witliin two to three days, while Bad. coli showed no changes. Ps. pyocyanea, while slower to undergo self digestion than B. subtilis, gave biuret figures wliich ap- proached 0. With both of these organisms the Sorensen figures rapidly increased. In the study of Bad. coli material no changes could be observed in the biuret or Sorensen values (see chart 1). Following these earUer observations, the plan of study was considerably enlarged, and efforts were made to acquire as BACTERIAL AUTOLYSIS 559 complete evidence as possible concerninp; so-called "autolysis" in certain organisms selected for tlie purpose. As the work concerned entirely changes going on in the complex nitrogenous substances of the dead bacterial cells, the following methods were employed and found in a large measure to supplement one another and to act as checks on each other: 1. The quantitative biuret test as used and advocated by Vernon. 2. The Sorensen method for the determination of amino nitrogen. 3. The Van Slykc method for the determination of amino nitrogen. 4. The conductivity test. 5. Staining and microscopic examination of the bacterial cells. 8. suBriLia B. COL/ ^ B ProCTANLUS @ S RAMOSUS lO /S ■ ^O 2S Title m OATS Chart 1. B. subtilis, B. mycoides, Bact. coli, and Ps. pyocyanea The biuret readings are expressed in percentage of Witte's peptone which will give a color of equal intensity. The Sorensen figures are expressed in the number of cubic centimeters of ^ NaOH required to neutralize the amino acids in 100 cc. of the solution. Erythrobacillus prodigiosus This organism was chosen because it readily undergoes autol- ysis and can be easily obtained in large quantities. The ac- companying charts (charts 2 and 3) show the course of changes 560 WILLIAM S. STURGES AND LEO F. RETTGER observed in five different experiments by the conductivity, bi'uret and Van Slyke methods. Number 4 was a physiological saline solution suspension of the agar surface growths, while the others were made up in distilled water. Two strains of E. prodigiosus were used in these tests. The rate of autolysis was most rapid within the first twenty- four hours, but required from ten to twenty days to be practically completed. The biuret figures continued slightly VAN SLYKE DETERMINATIONS T t^ IS 20 2S <*3 3S to Tin€ IN DAYS Chart 2. Erythrohacillus prodigiosus Electrical conductivity is expressed in reciprocal ohms. 0 J lO IS ^O 2S JO JS 40 riMC IN DATS Chart 3. Eryihrobacillus prodigiosus The Van Slyke determinations are recorded in the number of milligrams of primary amino acids in 100 cc. of the bacterial suspension. in the downward course even after the thirtieth day, whereas the Van Slyke values attained their maximum by the tenth day in two experiments, and by the twentieth day in a third. The other two experiments were interrupted early. The biuret and Van Slyke figures are in fairly close agreement; and both are in a large measure supported by the results of the conduc- tivity tests. The conductivity rose during the first seven to BACTERIAL AUTOLYSIS 561 eight days, but remained stationarj' after this period. Its increase was not proportional to the amino nitrogen increase. Bacterium coli In the first study 14 experiments were conducted with six different strains of Bad. coli. No decrease in the intensity of the biuret reaction could be observed except in three instances in which only a slight change was apparent. In four of the tests only a small increase in the Sorensen figures was shown. The biuret tests in this series were less delicate than those which followed. In the preparation of standards 0.25 per cent TABLE 1 Shoimng amino nitrogen changes during incubation of suspensions of Bad. coli. NTMBER OF EXPEBIUENT AGE OF CDLTCRE ORIGINAL AMINO NITROGEN CONTENT DURATION OF INCUBATION FINAL AMINO NITROGEN /tours Tngma. per ce. mgm. per cc. 1 36 0.19 3 days 0.38 2 14 0.10 4 days 0.11 3 24 0.13 3 days 0.12 24 0.13 3 months 0.11 4 48 0.08 8 hours 0.165 5 48 0.08 15 days 0.19 6 48 0.31 1 day 0.85 7 16 0.17 4 months 0.12 8 12 0.038 3 months 0.078 9 48 0.061 3 months 0.105 solutions of Witte's peptone were used, these giving a series of standards with 0.25 mgm. intervals between successive tubes. Each interval represents 0.025 on the ordinate of the biuret graph. Closer readings were not attempted. Table 1 gives the results of Van Slyke determinations of amino nitrogen on 9 different suspensions. While they are not de- cisive, they do indicate that an increase in amino nitrogen may take place. In the three instances where no increase was shown the bacterial suspensions were prepared from cultures only twenty-four hours old or less, and the appropriate enzjTues may not have had time to develop. The amino nitrogen content 562 WILLIAM S. STURGES AND LEO F. RETTGER was SO small in these suspensions that the total gas volume seldom reached 1 cc, and hence considerable error may have been introduced into the determination, although all of the tests were run in duplicate. In the following experiments on Bad. coli suspensions the biuret method was somewhat modified so as to make it more delicate. Instead of the 0.25 per cent basal solutions of Witte's peptone, 0.1 per cent strengths were employed, which furnished standards with 0.1 mgm. intervals, each interval being repre- sented on the ordinate by 0.01. 0.0 BIURET '^^l_ © 5- /O /S 20 ^S 30 JS TIM€ IN DATS Chart 4. Bad. coli AMIN0-NITR06EN /S 30 £S .30 3S Tinc IN OAYS f f^ ^ ELECTRICfd. CONDUCTIVITY CAPfXSSEO IN fKCIPmCAL OMfH S 10 IS iC !: TiMC IN curs Chart 5. Bact. coli Charts 4 and 5, barring curve Via, show only a sUght decrease in biuret and a corresponding increase in amino nitrogen. The biuret change in Via was far greater than in any of the others, and is suggestive of action by enzjTiies other than those of Bact. coli, that is the presence of contaminating bacteria. It was impossible to demonstrate any contamination, though it may have existed, owing to the toluol which was added as an anti- septic, and to the inconclusive results of microscopic study. Curves Vlb and Vic represent control suspensions which were heated to 60° and 75°, respectively, for thirty minutes, before BACTERIAL AUTOLYSIS 663 incubation. Both show a sli,u;ht decrease in biuret, antl a very abrupt and pronounced decrease in amino nitrogen. This sharp drop in amino nitrogen is an anomaly, and no explanation is offered. The conductivity of the suspensions of Bad. colt remained remarkably constant, except that of Via which corresponded to the decrease in biuret. Of the three constant suspensions, however, two were heated controls. The hydrogen ion concentrations of the different bacterial suspensions ranged from pH 7 to pH 9. Vb was brought to pH 9 bj^ the addition of sodium carbonate, while Va remained of its own accord at about pH 8. Number \T had a hydrogen ion concentration of pH 7. The results recorded here indicate that Bad. coli undergoes but slight autolysis, at the best, under conditions which are most favorable for this sort of enzymatic action. Staphylococcus aureus and Streptococcus pyogenes These two organisms are discussed together because they showed practically the same autolytic changes, and because the curves are given on the same chart (6). The results are of considerable interest in that the staphylococcus, which is an active gelatin liquefier, underwent but partial self digestion, and the streptococcus, which possessed no gelatin-liquefying property, whatever, elaborated enzymes which not only reduced the biuret appreciably, but brought about a marked increase in the amino nitrogen. Four experiments were carried out with the staphylococcus, two different strains being employed. Curve 4 represents one strain, and 1, 2, and 3 the other. Curves I and II represent two experiments on one strain of streptococcus. All of these strains showed autolytic changes, although one of the strains of Staphylococcus aureus failed to give any evidence of this in one of the experiments (1). The suspensions were made in distilled water. BIURET iaA._^^ 9 OS JO fs ec es 30 3S ^dng decrease in amino nitrogen of about one-eighth. No satisfactory explana- tion can be offered for these apparent discrepancies. That these curves represent, in the main, the actual course of the reaction is supported by the general agreement of the many readings, not all of which have been plotted on this chart. It would seem that the first reading on the amino nitrogen curve for Id is plotted a little too high. This point was obtained from the average of two out of three detemiinations which gave gas volumes of 1.60, 1.74 and 1.80 cc. of nitrogen, respectively. Conductivity determinations were also made from time to time of samples taken from the test suspensions. The results are expressed in the insert in chart 11. A slight but uniform increase in conductivity was noted throughout the experhnent. To explain the anomalous behaviour of suspensions Ic and Id the hypothesis was advanced that we are concerned here with some vigorous deamidizing action whereby the amino acids lost their identity as fast as they were formed. It was too late, however, to study changes in ammonia content, but no evidence could be adduced that there had been any considerable forma- 570 WILLIAM S. STURGES AND LEO F. RETTGER tion of ammonia. Analysis by the Folin micro-method revealed less than 2 mgm. of ammonia nitrogen per 100 cc. of the sus- pension, while there were 20 mgm. of amino nitrogen to be ac- counted for. The deamidizing action of the autolysate was tested on a weak solution of glycocoU, but little, if any, increase in ammonia nitrogen could be detected. The long incubation of the suspension may, however, have brought about a loss of deamidizing action by this time. The hydrogen ion concentra- tion was, at the end of the incubation period, approximately pH 7.0, which does not suggest the liberation of any considerable amount of ammonia. It may be of some interest to note (chart 11) the close agree- ment of the initial determinations of the three saline suspensions. In preparing these it was attempted to give lb a concentration twice as strong, and Ic 6 times as strong as that of la. WTiile this involved careful portioning of the pasty bacterial material, which must have varied more or less in concentration in its different parts, the first determinations of both biuret and amino nitrogen show that the 3 suspensions bore the concentration relationship of 1, 2 and 6. A liver extract agar was employed for the massive growth of meningococcus required in the following meningococcus ex- periment. A very luxuriant growth developed on this medium (employed in 1-Uter Blake bottles) which was removed by scrap- ing across the surface with a heavy wire bent into a loop at the end. Thirty of the Blake bottles yielded enough growth to make 11 grams of moist scrapings. This material was trans- ferred to a sterile glass-stoppered bottle and kept in a freezing mixture for three days, or until the experiments could be started. Five grams of the viscid material were suspended in 100 cc. of sterile distilled water (le). The remainder was divided and one part made into a dilute (1 gram in 30 cc.) and the other into a concentrated suspension in (2 grams in 30 cc.) physiological saline solution. For the purpose of obser\-ing the effects of incubation at 37° versus low temperature, each of these two suspensions was divided into 2 parts. Five different suspensions were used, therefore, namely; BACTERIAL AUTOLYSIS 571 la, dilute (1 gram in 30 cc. of saline), at SVC. lb, dilute (1 gram in 30 cc. of saline), at ice box temperature Ic, concentrated (3 grams in 30 cc. saline), at 37°C. Id, concentrated (3 grams in 30 cc. saline), at ice box temperature le, 5 grams of the concentrated bacterial material suspended in 100 cc. of distilled water, and incubated at 37°. The results are shown in charts 12 and 13. The biuret curves indicate an extensive autolysis of two of the incubated samples, /o /s io ^s T7MC //v o/t-rs Chart 12. Meningococctjs 5/« .^ / I © /O /J- 20 ^jr JQ j^ T/MS /A' OArS ELECTRICAL CONDUCT/I//TY £''PK£iSED'lN flCCIPf;OCAL. OHMS O S- /O ^S £0 es JO 3S T//^C IN OATS Chart 13. Meningococcus Ic and le, and but little or no change in the others. In the dis- tilled water .sample (le) the reduction in biuret was very shght during the first five daj^s but subsequently progressed rapidly so that at the end of twelve days the biuret-giving substances were reduced to one-third of the original. Owing to lack of sufficient material, duplicate amino acid determinations were made of le only (chart 13). All of the amino nitrogen curves have a few points of similarity, and all attain their maximum level within ten to fifteen days. Beyond this the values fail off somewhat^ barring Ic of course, and, as 572 WILLIAM S. STURGES AND LEO F. RETTGER was observed in the pneumococcus experiments (chart 11), there appears to be some evidence of deamidization. No proof of such a process could be established, however, by further study. There is the possibility, of course, that such a deamidizing agent may have been lost from the suspension during the long and continued incubation. Amino nitrogen curve le shows an apparent discrepancy. As compared with la and lb, and with the biuret curve for this same suspension, the two and the five day points should be practically on the same level. The conductivity readings for suspensions le show very sUght changes. GENERAL DTSCTTSSION The rapid and profound autolysis of the liquefying bacteria, particularly Erythrobacillus prodigiosus, which could be readily demonstrated by the methods employed in this investigation, is to be expected, but that such well-known non-liquefying or- ganisms as Bad. coli, the gonococcus, meningococcus and pneu- mococcus, should undergo a similar process, though in a much lesser degree, evoked considerable surprise in the authors. In order to satisfy themselves as to the ability of the pathogenic cocci to attack gelatin under the most favorable conditions of environment and temperature, several fruitless attempts were made to induce gelatin liquefaction in enriched media. The enzymes invoived in the autolysis of the above-mentioned non-Uquefying organisms cannot but have a specific or at least a limited sphere of action. According to Rosenow (1912) the pneumococcus is able to attack the proteins of ascitic fluid, blood serum and meat extract, but we have been unable to demonstrate any proteolysis of albumin, casein or gelatin by the pneumococcus, with the use of the more delicate and specific methods employed in tliis work. It has been assumed throughout the investigation that en- zymes, when present, would assert themselves in a neutral or weakly alkaline medium. These experiments have as a rule been conducted, therefore, in hydrogen ion concentrations rang- BACTERIAL AUTOLYSIS 573 ing from pll 7.0 to 9.0. A series of tests made upon E. prodig- iosus in different hydrogen ion concentrations showed that, although slow digestion took place at pH 4.0, the most rapid digestion occurred at pH 7.0. Rosenow claimed that the en- zyme action demonstrated by him in the pneumococcus is favored in a +0.5 medium (phenolphthalcin titration). Dernby (1917 and 1918), held, on the other hand, that a pH of 6.0 was the most favorable for the autolysis of animal organs. The very slight changes which were observed in the Bad. coli suspensions are of particular interest and soUcit further in- quiry. In the first set of experiments very little enzymatic action of any sort could be demonstrated, and then in 4 of the suspensions only. In the second series (Charts 4 and 5) a slight reduction in biuret (1 per cent to 6 per cent) and an increase in amino acids were shown in every instance barring experiment Via from which the possibihty of contamination cannot be excluded. The chemical changes in these suspensions were so slight as not to be indicated at all by the conductivity tests. As the 2 heated controls showed almost the same change as the unheated test samples the "autolysis" manifestations become all the more insignificant. It is possible, however, that the heating at 60° and 75°C. may not have been sufficient to destroy the enzj-mes. Abbott and Gildersleeve (1903) found that heating at 100° for fifteen to thirty minutes did not always destroy certain bacterial enzymes. Meyer (1911) reported that the proteases of Pseudomonas pyocyanea are little affected by 15 minutes of exposure to 100°C. Wells and Corper (1912) describe Upolytic enzjTnes in Mycobacterium tuberculosis which are not entirely destroyed in thirty minutes at 100°. On the other hand, Fermi (1894) found 55 to 70'' to be sufficient to destroy the proteolytic enzjmies of bacteria. The possibihty of shght decreases in biuret and increases in amino nitrogen from causes other than autolysis must not be ignored. If, in the process of harvesting the bacterial growths any appreciable amount of peptone with its various biuret-giv- ing substabces should be removed from the agar medium, and if the suspended organisms elaborate enzymes which act on any 574 WILLIAM S. STURGES AND LEO F. RETTGER of these biuret-containing substances the results of such action would be revealed in subsequent tests. In fact, it is more than possible that the bacterial cells contain relatively simple biuret- giving polypeptides which have been absorbed from the medium or which are intermediate synthetic products occurring in the natural course of bacterial metabolism. Berman and Rettger (1918) showed that Bad. coU is unable to utilize the more com- plex or proteose fraction of Witte's peptone, but that some of the simpler biuret-containing substances in the commercial peptone, probably polypeptides, are attacked by pure culture of this organism and broken down into the simpler amino acids. Furthermore, it is doubtful whether very slight increases in amino nitrogen need necessarily indicate any hydrotysis. We may be dealing with more or less masking of amino nitrogen by the protein, which does not allow all of the amino nitrogen to react in the Van Slyke procedure, during the early part of the incubation period. Such a protective action may to a certain extent be removed by such purely phj^sical processes as diffusion, partial disintegration from osmotic disturbance, toluol extrac- tion, etc. Changes of this type would render the amino acids more and more available during the course of incubation for the Van Slyke determination. Such influences scarcely seem suflficient, however, to explain the more pronounced increases in amino nitrogen noted in some of the experiments, as for ex- ample in chart 6 where the increases are shown to range from 4 to 60 per cent. The failure to demonstrate marked changes in the suspensions of Bad. coli by the biuret, Sorensen, Van Slyke and conduc- tivity methods is fully supported bj^ direct microscopic examina- tions of the different suspension during the periods of observa- tion. The individual cells at all times took the ordinary stains well and with a high degree of uniformity, and there was no evidence of morphological change, except in Via of the second series of experiments which also differed markedly from all others in its reaction to the biuret, Van Slyke and conductivity tests, and which must be excluded from serious consideration on account of the probabiUty of bacterial contamination. BACTERIAL AUTOLYSIS 675 As SO few tests were made with the Staphj'lococcus and Strep- tococcus no further comment should be necessary than to state in review that both of these organisms showed distinct evidence of elaborating; enzymes which have appreciable autolytic action, especially the staphj-lococcus. The results obtained with the pneumococcus, gonococcus and meningococcus indicate that these organisms undergo autolytic changes as the result of their own enzyme action, and particularlj' the meningococcus and the gonococcus. This is further sub- stantiated in the case of the two last mentioned organisms by the microscopic examination of stained slides in which decreased staining ability and change in cell morphology are shown during the course of the autolysis, as had been observed previously by Flexner (1907) for the meningococcus. StTMMARY AND CONCLUSIONS It has been possible to follow in a quantitative way the nitrog- enous changes taking place in autolyzing bacterial suspensions. In this studjr the quantitative biuret test of Vernon and the Sorensen titration have proven of the greatest value. The Van Slyke determination of amino nitrogen has also been em- ployed to advantage, although the material to be analyzed was not particularly suited to this method. Electrical conductivity has been found to increase during autolysis, but its increase has not been proportional to the amount of amino nitrogen formed by the hydrolysis of protein, as Bay liss (1907-8) claims to have found to be the case in the tryptic digestion of caseinogen and gelatin. By the application of such methods it has been possible to show that: 1. Proteolytic bacteria of the type of Erythrobacillis prodig- iosus, Pseudomonas pyocyanea, and B. subtilis autolyze rapidly. 2. Bad. coli undergoes slight changes which may be autolj'- tic in nature, but which at best involve only a small part of the complex nitrogenous constituent or constituents of the cells. 3. The pathogenic cocci, pneumococcus, gonococcus and men- ingococcus undergo an actual autolysis with a breaking down of the protein or protein-like substances of the bacterial cells. 576 WILLIAM S. STUKGES AND LEO F. EETTGER REFERENCES Abbott, A. C, AND GiLDERSLEEVE, N. 1903 A study of the proteols^tic enzymes and of the so-called hemolysins of some of the common saprophytic bacteria. Jour. Med. Res., 10, 42-62. Alilaihe, E. 1913 Experiences sur I'autolyse du coli-bacille. Annal de I'lnst. Pasteur, 22, 118-121. Batliss, W. M. 1904 The kinetics of tryptic action. Arch des Science Biol., 11, 261. Supplement Pavlov Jubilee Vol.; 1907-190S Researches on the nature of enzyme action. Jour. Physiol., 36, 221-252. Benson, R. L.; and Wells, H. G. The study of autolysis by physiological methods. Jour. Biol. Chem., 8, 61-76. Berman, N. and Rettger, L. F. 1918 Bacterial nutrition. Jour. Bact., 3, 367-388. Brieger, L. and Mater, M. 1903 Weitere Versuche zur Darstellung specifischer Substanzen aus Bacteriern. Deut. med. Woch., 29, 309-310, also 30, 980-982. BtJRGERS, Schermann AND ScHREiBER. 1911 Ucber Auflosungserscheinungen von Bakterien. Zeitschr. Hyg., 70, 119-134. Cacace, E. 1901 Ueberdas proteolytische VermogenderBakerien. Centralbl. Bakt., 30, 244-248. Conradi, E. 1903 Ueber losliche, durch aseptische Autolyse erhaltene Gift- stoffe von Ruhr-und Typhusbazillen. Deut. med. Woch., 29, 26-28. CoRPER, H. J., AND Sweeney, H. C. 1918 The enzymes of the tubercle bacillus. Jour. Bact., 3, 129-151. Dernby, K. G. 1917 Studien iiber die proteolytischen Enzyme der Hefe und ihre Beziehung zu der Autolyse. Biochem. Zeitschr., 81, 107-208. Dernby, K. G. 1918 A study on autolysis of animal tissues. Jour. Biol. Chem., 35, 179-219. ElJKMANN, C. 1901 Ueber Enzyme bei Bakterien und Schiemmelpilzen. Centralbl. Bakt., 29, 841-848. Fermi, C. 1890, 1891 and 1894 Ueber die Enzyme. Arch. Hyg., (1890) 11, (1891) 12, 238. Centralbl. Bakt., (1894) 15, 229, 303 and 722; and 16, 830. Flexner, S. 1907 Biology of Dip. intracellularis. Jour. Exp. Med., 9, 105-141. Hahn, M. 1898 Das proteol.ytische Enzyme des Hefepressaftes. Ber. d. deut. chem. Gesellsch., 31, 200-205, and 2335. Hahn, M. 1900 Ueber das Hefeendotrypsin. Zeitschr. Biol., 40, 117-172. Hall, I. C. 1916 Testicular infusion agar; a sterilizable culture medium for the gonococcus. Jour. Bact., 1, 34.3-351. Jacoby, M. 1900 Ueber die fermentative Eiweissspaltungund Ammoniakbild- ung in der Leber. Zeitschr. physiol. Chem., 30, 149-173. Levy, E., and Phersdorfp, F. 1902 Ueber die Gewinnung der schwer zugang- lichen, in der Leibsubstanz erhaltenen Stoffwechselprodukte der Bakterien. Deut. med. Woch., 28, 879-880. McGlintock, C. T., and Clark, L. T. 1909 Autolysis of the gonococcus. Jour. Inf. Dis. 6, 217. McFadyen, a. 1892 A research into the nature and action of the enzymes produced by the bacteria. Jour. Anat. and Physiol^ 26, 409-429. BACTERIAL AUTOLYSIS 677 Meter, K. 1911 Zur Kcntniss dcr Baktcricnproteasen. Biochcm. Zeitschr., 32, 274-279. Neisseb, N., and Shiga, K. 1903 Ueber freie Receptorcn von Typhus- und Dysenteriebazillen und ueber die Dysenterietoxin. Deut. med. Woch., 29, Gl-62. NiCOLLE, N. 1913 L'autolyse. Annal. de I'lnst. Pasteur, 27, 97-118. Rettoeh, L. F. 1904 On the autolysis of yeasts and bacteria. Jour. Med. Res., 8, 79-92. Rettoer, L. F., Herman, N. and Sturoe.s, W. S. 1916 Jour. Bact., 1, 15-33. Rosenow, E. C. 1912 Toxic substances obtainable from pneumococci. Jour. Inf. Dis. 11, 94-108. See also 286-293. Rosenow, E. C. 1912 Production of anaphylatoxic substances by autolysis of bacteria and thoir relation to endotoxins. Idem, 10, 1 13-128. Salimbeni, a. T. 1913 Preparation de solutions toxiques a I'aide de l'autolyse. Annal. de I'lnst .Pasteur, 27, 122-129. Saleowsei, E. 1891 Ueber Autodigestion der Organe. Zeitschr. klin. Med., 17, Supplem. 77. SjOqvist, J. 189.'5 Skand. Arch. Physiol., 6, 277-368. SoRENSEN, S. P. L. 1907 Enzymstudien. Biochem. Z. 7, 45-101 . Van Sltke, D. D. 1911,1912 A method for the quantitative determination of aliphatic amino groups. (1911) Jour. Biol. Chem., 9, 185-204, 12, 275-284. (1912) Vernon, H. M. 1903 The peptone-splitting ferments of the pancreas and intestine. Jour. Physiol., 30, 330-370. VioNAL, 1896 Fliigge's Microorganismen (Leipzig) Aufl. 3, p. 207. Warden, C. C. 1913, 1915 Studies on the gonococcus. Jour. Inf. Dis., (1913) 12, 93-105: 13, 124-135, and 21, (1915), 426-440. Warden, C. C. 1917 The physico-chemistry of the gonococcus. Jour. Am. Med. Assoc. ,68, 432-436. Wells, H. G., and Corper, H. J. 1912 The lipase of B. tuberculosis and other bacteria. Jour. Inf. Dis., 11, 388-392. WoLLSTEiN, M. 1907 Biological relationship of Diplococcus intracellularis and gonococcus. Jour. Exp. Med., 9, 588-605. Wood, C. Rep. Lab. Col. Phys. (Edinburg), 11, 253. BACILLUS HEMOGLOBINOPHILUS CANIS (FRIEDBERGER) (HEMOPHILUS CANIS EMEND.) T. M. RIVERS From the Department of Pathology and Bacteriology Johns Hopkins University Received for publication February 18, 1922 Friedberger in (1903), working in Pfeifler's clinic reported finding in the preputial secretions of a dog a small, Gram- negative, non-motile, hemoglobinophilic bacillus which he called B. hemoglobinophilus canis. From his description it is difficult to see how he differentiated it from H. influenzae except that he isolated it from a dog instead of a human being. Odaira (1912), also working with Pfeiffer, coinparcd H. pertussis, H. influenzae and H. canis, and found that by agglutination tests they were different. When complement fixation tests were used less specificity was shown. Since the last pandemic of influenza much interest has been taken in the so-called hemophilic bacilli in regard to their rela- tion to disease, their growth requirements and their biological reactions. It seemed, as advance had been made in differentiat- ing many of these organisms by biological reactions, that H. canis might also be differentiated culturally from closely allied organisms. During the past year six strains have been isolated from the preputial secretions of dogs. JMost of the male dogs examined had a certain amount of pus in the preputial secre- tions. Direct smears of this pus often showed numerous small Gram-negative bacilli and when cultured H. canis was usually found to be the predominating organism. Up to the present this organism has been isolated from dogs only, but it is con- cei^^able that laboratory workers and people who handle dogs might show this bacillus accidentally in throat or nose cultures. 579 580 T. M. RIVERS Morphology, staining, motility. It is a small pleomorphic, Gram-negative, non-motile rod which looks very much like H. influenzae. Type of growth. On 2 per cent rabbit-blood meat-infusion agar young colonies are non-hem oly tic, round, mth a small granular area on top. At this time they are indistinguishable from colonies of H. influenzae but as they grow older the former become distinctly more opaque than the latter. Old cultures on blood agar slants show a luxuriant opaque growth that resembles H. pertussis more than H. influenzae. A diffuse growth occurs in the proper Uquid medium. Grou'th requirements. H. canis grows well on 5 per cent human blood agar. This is a distinct difference between H. canis and H. influenzae as the latter does not grow well on media contain- ing fresh unheated human blood (Rivers, 1919). H. influenzae (Rivers and Poole, 1921) requires two food accessory factors, one autoclave labile, the other autoclave stable. H. canis, how- ever, requires the addition of only the autoclave stable sub- stance as an accessory factor. It will not hve more than one or two generations on meat infusion agar, meat infusion agar plus ascitic fluid or 2 per cent peptone agar plus yeast extract. Suc- cessful transplants can be made indefinitely on meat infusion agar plus hematin or 2 per cent peptone agar plus hematin. It is aerobic. Indole production. All the strains produced indole. Nitrate reduction. All the strains reduced nitrates to nitrites. Reaction in milk. Very Uttle change was noticed in blood milk mixtures when inoculated with H. canis. Sugar fermentations. The medium used for fermentation tests was the same as described in a previous paper. (Rivers and Kohn, 1921). All the strains formed acid without gas from glucose, fructose, galactose, mannitol, sucrose, and xj'lose. Neither acid nor gas was formed from maltose, lactose, dextrin, arabinose and glycerol. Hemolysis. Red blood cells were not hemolyzed in solid or in liquid media by any of the strains. BACILLUS HEMOGLOBINOPHILUS CANIS 581 Pathogenicity. One cuhir centimeter of a twenty-four-hour culture in blood broth did not kill a white mouse when given intraperitoneally; two cubic centimeters given intraperitoneally did not kill a small pn-iinea pig; one cubic centimeter intravenously did not kill a small rabbit. From table 1 it can be seen that H. canis can be differentiated from H. pertussis by indole production, nitrate reduction and sugar fermentations; from H. influenzae by growth accessory factors and mannitol fermentation. TABLE 1 Most important differential cultural characteristics of H. pertussis, H. influenzae and H. canis BACTERI GM H. per- tussis H. influ- enzae H. canis ACCESSORY OROvrrn factors Old cultures require neither autoclave la- bile nor autoclave stable factor Autoclave labile and autoclave stable fac- tors required Only autoclave stable factor required INDOLE PRODCCriON Never pro- duced May or may not be produced Always produced NITRATE REDUCTION Never re- duced Always reduced Always reduced SUGAR FERMENTATIONS No sugar fermented Different sugars fer- mented ; never mannitol Different sugars fermented, in- cluding mannitol CONCLUSIONS B. hemoglobinophilus canis of Friedberger has been further described and differentiated from H. pertussis and H. influenzae by cultural methods. According to the new classification this bacillus should be called Hemophilus canis. REFERENCES Friedberger, E. 1903 Centralbl. f. Bakteriol., I; Orig., 33, 401. Odair,\ 1912 Centralbl. f. Bakeriol., Orig., 61, 289. Rivers, T. M. 1919 Bull. Johns Hopkins Hosp., 30, 129. Rivers, T.M., AND KoHN.L. A. 1921 Jour. Exper.Med., 34, 477. Rivers, T. M., and Poole, A. K. 1921 Bull. Johns Hopkins Hosp., 32, 202. SALT EFFECTS IN BACTERIAL GROWTH III. SALT EFFECTS IN RELATION TO THE LAG PERIOD AND VELOCITY OF GROWTH' J. M. SHERMAN, G. E. HOLM and W. R. ALBUS From the Research Lahoralories of the Dairy Division, United States Department of Agriculture, Washington, D. C. Received for publication February IS, 1922 In a previous paper of this series (Holm and Sherman, 1921), it has been shown that certain neutral salts, in proper concen- trations, accelerate the growth of Bad. coli, as measured by the time required to produce visible turbidity, the time required to reduce methylene blue, and the rate of acid production in the presence of a fermentable carbohydrate. In the present work we have extended these experiments by the use of the plate count in an effort to throw more light on the mechainism of the salt action. Since our previous experi- ments have shown that the salt effect is magnified on the acid side of the region of optimum growth (Sherman and Holm, 1922), we have used in the present work media adjusted to a reaction of pH 5.4. All of the counts here reported represent the average of tripUcate plates on extract-pepton agar incubated for three days at 33°C. Further details of the experiments are given in the appendix. From figures 1 and 2, plotted from the data obtained in ex- periments 1 and 2 in which the growth of Bad. coli in 1 per cent pepton, 1 per cent pepton plus 0.2 M NaCl, and 1 per cent pepton plus 0.1 j\I Na2S04 was measured, it is seen that the accelerating action of the salts is due to an increase in the velocity of growth of the organisms. In other words, the period of logarithmic growth is shortened since the number of bacteria present in the ' P^iblished with the permission of the Secretar>' of Agriculture. 583 e7 o CD Ik .'/ ", sT'j -i J 4 7 H S 1^ II li- 1% li 'y '■' Ht-A at 57" C ■ n"' iT'-'i'-i ■"'W' -l-^e_picrt ~s Z 7 S ■ iL ■•n--id—i'f -■W~--i> be entirely lacking, as is indicated by the results of experi- ment 2. The effect of NaCl upon the period of lag was therefore ex- tended in experiments 3 to 6 in which plate counts were made at hourly intervals. The results of these additional experiments again show that while NaCl may or may not decrease the latent period, it increases in every case the velocity of growth during the period of active multiplication. It was thought that it might be possible to magnify the effect of NaCl by incubation at a temperature which allowed only a THE JOURNAL OF BACTERIOLOOY, VOL. VII, NO. I 586 J. M. SHERMAN, G. E. HOLM AND "W. E. ALBUS slow multiplication of the organisms. This was done in tests in which cultures were incubated at a temperature of 12°C. At this temperature the increments of growth when measured daily are about the same as those taken at hourly intervals at 37°C. The results obtained in these tests (experiments 7 and 8) show the same characteristic increase in the velocity of growth with NaCl, and also a well defined shortening of the lag period. The data from experiment 7 are plotted in figure 3. SUMMARY It has been shown that the accelerating effect of certain salts upon the growth of Bad. coli is due primarily to an increase in the velocity of growth of the organism during the period of maxi- mum multiplication. The same salts usually also increase the accelerating effect by decreasing the duration of the preliminary latent period. REFERENCES Holm, G.E., AND Sherman, J. M. 1921 Jour.Bact., 6, 511. Sherman, J. M., and Holm, G. E. 1922 Jour. Bact., 7, 465. APPENDIX The organism used in all of these experiments was a laboratory strain of Bact. coli. Inoculations were made from cultures one or more days old in 1 per cent pepton. The media used for determining the growth rates were put up in 100-cc. amounts; all contained 1 per cent pepton, with the indicated amount of salt, and were adjusted to a reaction of pH 5.4. In experiments 1 to 6, incubations were at 37°C. and plate counts were made at hourly intervals. Experiments 7 and 8 were conducted at 12°C. and counts made at daily intervals. Standard extract-pepton agar was used for plating, and the plates were incubated for three days at 33°C. before counting. TripUcate plates of each dilution were made in every case. SALT EFFECTS IN BACTERIAL GROWTH 587 Experiment 1 NnUOEB or BACTERIA PEB CUDIC CEKTIUETEB BOCB8 Pcpton 0.20MNaCI 0.10 M NaiSO. M NaiSOi 0 39,000 41,000 26,000 1 29, 000 28,000 24,000 2 17,000 23, 300 33,700 3 16, 000 48, 000 86,000 4 17, 000 163, 000 473, 000 6 33,000 020, 000 2, 630, 000 6 47, 000 4, 700, 000 12, 300, 000 7 66, 000 20, 300, 000 76, 000, 000 8 117,000 135,000,000 153, 000, 000 0 210, 000 164, 000, 000 149,000,000 10 410, 000 183,000,000 11 780, 000 188,000,000 198,000,000 12 1,560,000 208,000,000 183,000,000 13 4, 300, 000 231,000,000 220,000,000 14 12, 100, 000 16 67, 000, 000 18 104,000,000 20 121, 000, 000 NUMBER OF BACTERIA PER CtTBIC CENTIMETER HODBS Pepton 0.20 M NaCl 0.10 M Na^SOi M NajSOi 0 99, 000 95, 000 103, 000 1 58, 000 79, 000 78,000 2 77, 000 106,000 125,000 3 300, 000 181,000 471,000 4 766, 000 1, 750, 000 2,630,000 6 3,730,000 8, 500, 000 16, 300, 000 6 17,900,000 74, 000, 000 74,000,000 7 49, 500, 000 172,000,000 168, 000, 000 8 109, 000, 000 181,000,000 191,000,000 9 149, 000, 000 169,000,000 201,000,000 10 148, 000, 000 190,000,000 209,000,000 11 162, 000, 000 207,000,000 214, 000, 000 12 175,000,000 214, 000, 000 215, 000, 000 Experiment S NUMBER OF BACTERIA PER CUBIC CENTIM ETEB HOUBS Pepton 0.20 M NaCl 0 54,000 56,000 1 52, 000 52, 000 2 49, 000 76, 000 3 86, 000 129,000 4 152, 000 480,000 5 360, 000 780, 000 6 790, 000 14 100,000 7 3, 200, 000 53 000, 000 Experiment 4 NUMBER OF BACTERIA HOURS PER CUBIC CENTIMETER Pepton 0.20MNaCl 0 56,000 57,000 1 51,000 54,000 2 59, 000 59,000 3 94, 000 57,000 4 148, 000 139, 000 5 273, 000 1,120,000 6 790, 000 20,600,000 7 23,700,000 105, 000, 000 Experiment S NUMBER OP BACTERIA HOURS PER CUBIC CENTIMETER Pepton 0.20 M NaCl 0 35, 000 44, 000 1 41,000 42,000 2 87,000 83, 000 3 310,000 203,000 4 1,090,000 423, 000 5 8, 100, 000 23, 800, 000 6 23,500,000 64, 000, 000 7 29, 200, 000 63, 000, 000 Experimeid 6 NUMBER OF BACTERIA BOUK8 PER CUBIC CENTIMETER Pepton 0. 20 M NaCl 0 126,000 119,000 1 128, 000 132,000 2 175, 000 188,000 3 676, 000 940, 000 4 2,660,000 4, 440, 000 5 17,800,000 85, 000, 000 6 24, 500, 000 135, 000, 000 7 45,200,000 205, 000, 000 Experiment 7 NU.MBER OF BACTERIA PER CUBIC CENTIMETER DATS Pepton 0.20 M NaCl 0 28, 200 25, 300 1 29,000 37,500 2 38,000 69, 000 3 38,600 161,000 4 47, 000 240, 000 5 46, 000 523, 000 6 129, 000 1,630,000 7 360,000 14, 750, 000 8 834, 000 54, 000, 000 9 3, 100, 000 96, 000, 000 10 9, 300, 000 150, 000, 000 11 45, 000, 000 185, 000, 000 12 73, 000, 000 148, 000, 000 13 156,000,000 280, 000, 000 14 260, 000, 000 290, 000, 000 Experiment 8 NUMBER OF BACTERIA PER CUBIC CENTIMETER DATS Pepton 0.20 M NaCl 0 26,800 30,000 1 29,500 36,000 2 41,000 57, 000 3 36, 000 182, 000 4 52, 000 257, 000 5 53, 000 340, 000 6 186, 000 1,480,000 7 800, 000 16, 600, 000 8 2, 020, 000 54,000.000 9 7, 300, 000 91,000,000 10 35, 000, 000 100, 000, 000 11 56, 000, 000 160,000,000 12 88,000,000 186, 000, 000 13 165,000,000 250, 000, 000 14 250, 000, 000 270, 000, 000 588 FURTHER OBSERVATIONS ON "COLOR STANDARDS" FOR THE COLORLMETRIC DETER.AIINATION OF H-ION CONCENTRATION LEON S. MEDALIA From the Research Laboratories Department of Biology and Public Health, Massachusetts Institute of Technology Received for publication March 23, 1922 The following additional observations on color standards for the colorimetric determination of H-ion concentration have been made by the author since the publication of his first article on this subject (INIedaUa, 1920). It will be recalled that in the first article just referred to a method was described of preparing "color standards" for the principal indicators of Clark and Lubs. This was done by using varying amounts of indicator in acid and alkaUne solutions choosing such solutions as would bring out the acid and alkaline colors of the double colored indicator. Seven pairs of tubes were used; each pair containing 0.8 cc. of indicator solution in the ratio of 1:7, 2:6, 3:5, etc., increasing by 0.1 cc. in the alkaline solution up to 0.7 cc. and decreasing by 0.1 cc. in the acid solution down to 0.1 cc. These pairs of tubes were viewed by looking through them preferably in a comparator block, when placed one behind the other, i.e., superimposing the acid and alkaline colors of the indicator in the varjnng strengths — similar to the superimposing method used by Salm (1904) to determine the half transformation point, and by Barnett and Chapman (1918) to prepare the color standards for phenol red. The pH repre- sented by the color of each pair of "color standards," was caU- brated by using buffer mixtures of known H. I. C, prepared according to Clark and Lubs (1917). The possibility suggested itself of making use, in addition to the seven pairs of "color standards" just referred to, of an 589 590 LEON S. MEDALIA acid and an alkaline tube at either end of each set of the color standards. This would add two more colors to the range of each indicator and thus greatly enhance the usefulness of each set of color standards, more especially as regards the two end pairs. Such tubes would contain each 10 cc. of the acid and alkaline solutions, respectively, as used for the particular indicator, to each of which would be added in turn 0.8 cc. of the "indicator watery solution." They would give the full acid and full alkaline colors of the particular indicator and would thus be of value in determining whether a given unknown falls within the visual or "virage" range of the particular indicator or not. There were two questions, however, that had to be deter- mined experimentally with reference to this : 1. Whether there would be a difference in color sufficiently marked to distinguish between such a tube, on the acid side, when compared with the first pair of the "color standards;" and the tube, on the alkaline end, when compared with the seventh pair of the set. 2. Whether any definite pH could be attributed to such tubes when studied by calibrating them with buffer mixtures of known H-ion concentration. The first question can be answered in the affirmative: There is, according to the findings, a definite and distinct difference between the tubes on either side of the scale as compared with the respective pairs nearest to them. The solution in the tube on the acid end of the scale is the same as that described for the acid tubes of the pairs in each of the sets of "color standards" for each indicator. Similarl}^ the tube on the alkahne end of the scale contains the solution used for the alkaline tubes of the pairs in each set. Thus with thymol blue (acid range) , the acid tube of this indi- cator is made with 0.5 volume per cent of concentrated HCl, the alkaline tube of tliis indicator contains 0.001 per cent con- centrated HCl and to each of these is added 0.8 cc. of 0.02 per cent of the indicator watery solution. DETERMINATION OF II-ION CONCENTRATION 591 With brom thymol blue: Tlic acid tube contains 10 co. of 0.1 per cent concentrated HCl and the alkahne tube N/200 NaOH solution. The acid and alkaline tubes of methyl red, brom cresol piirpk, brom thymol blue, and cresol red, are made with 0.1 per cent concentrated HCl and N/20 NaOH, respectively. To each, of course is added 0.8 cc. of the respective indicator solution, in dilutions given in the previous article for the preparation of the color standards. The acid and alkaline tubes of phenol red contain 0.1 per cent HCl, and N/100 NaOH respectively,' and those of thymol blue (alkaline range) were made with 0.001 per cent concentrated HCl and N/20 NaOH respectively. As to the second question: whether these tubes at the acid and alkaline ends of each set could be shown to represent a definite H-ion exponent, the tubes were matched against buffer mixtures of known H-ion concentration and showed the results indicated in the appropriate column of table 1. With methyl red: The "alkaline tube" practically matched a pH 6.0 buffer mixture, the latter showing a very slightly more reddish tinge. With phenol red: The "acid tube" practically matched a pH 6.8 buffer mixture, approaching a pH 6.6 buffer mixture (there seems to be only a very slight difference in color between pH 6.6 and pH 6 8 with this indicator while, pH 6.4 of buffer and pH 6.6 are alike in color). With cresol red: The "acid tube" is very sUghtly clearer or brighter in color than a pH 7.2 buffer mixture, being practically the same otherwise. With thymol blue (alkaline range): The "acid tube" is slightly more yellowish in color than a pH 8 buffer mixture approaching pH 7.8 of the buffer. There is a marked difference in color be- tween this tube and pair no. 1 of the set. The "alkaline tube" matched the pH 9.6 buffer mixture. There is a difference in color between this tube and pair no. 7, of the set. The value of these "acid" and "alkaline" tubes, at the end of each set of "color standards" is not so much as to the actual pH which they represent, but rather in the fact that they allow a ' The N/100 NaOH was found to be the better strength to use for the alkaline tubes for the color standards of this indicator rather than N/20 NaOH as recom- mended in the previous article. 592 LEON S. MEDALIA w P n iJ a ill e. odcdoDcooociciooi 0 d So d w Yellow through green to blue 0 lid K g jj. (N-*coajoc}t^t:~t^00CO00 S ►J 0 is s n 0 a n -^ C^TJHCOOOOC^-^COOO 0 d d w .c 0 b. 3 a g 0 tu 0 ^ ^ M X5 A CO CO <© CD t^ t^' t--' t^ t^ s ii 0 a. a n a: (M■*O0ir-lC^3lO^~00 d 1x1 Izilz; ^ 0 d K 0 ^ c t: "S '5 ■£. S, >< s ^ IC 10 iO »0 CO CD CO CO CO « w E mt-Oi-HiM-^oooo d d w M a) 0 fe •^ J5 0 0 P 0 0 S, -!)<' T)i ■*■ 10 lO 10 10 10 CO "5> g is 0 m tC (NtJlatini/.ed asbestos upon oxygen and hydrogen when they are brought in contact. The apparatus consists of a heavy walled Xo. 80.>1K Mus'^um .far, through the glass cover of which have been ilrilled two holes in which are inserted rubber sto|)pcrs carrying l)ent glass stop-cocks. One stop-cock has attached to its inner end a perforated glass bulb filled with platinized asbestos; to the end of the other is attached a piece of rubber tubing extend- ing to the bottom of the jar. The cover is held in place by the usual clamp, a rubber gas- ket being used to make a gas tight joint. (For a complete description of the method, see Smillic's article in The .lournal of Experimental Medicine. Vol. XXVI, No. 1, for .luly, 1017, l)age .50, '"Xew Anaerobic Methods." The same outfit will be found useful for either gas, vacuiim or pyrogallate methods, anil is particularly valuable because of the number of Petri dishes or tubes which may be placed in it. Diameter of jar inside, 5 inches; depth inside, 12 inches. Complete as illustrated 11. .50 Cfe^jTrfi^JlySGlENTIFIG COMPANi^§' 46o E.Obio Si. u.y?A JOURNAL OF BACTERIOLOGY INTERNATIONAL CENTRIFUGES Latest model, self -balancing cen- trifuge with brake and brush re- lease, and removable top adapted to use with perforated basket. International Equipment Co. 352 Western Avenue (Brighton) Boston 35, Mass. Catalogue Cy tent on Requttt CULTURE MEDIA IN HERMETICALLY SEALED TUBES .These Culture -Media in hermctifally sealed tubes, 14x150 mm, are cotton plugged so that when a tube is broken at the scored mark a cotton plugged, sterile tube of media is ready for use. Securely packed in cartons and pack- aged in cartons of twelve tubes. The slants are excellent for making pre- liminary inoculations, bu may be em- ployed equally well in laboratory rou- tine. Made according to latest methods of a 6sed concentration or reaction. These tubes possess the ad\antage that the media cannot dry out or spoil, since evaporation or contamination cannot take place as in ordinary tubes. Being securely packed no special pre- cautions as to storage need be taken. Tubes ordinarily packed to contain approximately 10 re of culture media. Price $1.50 per dozen. Special dis- counts on larger quantities. Any culture media in amounts up to 12 cc can be sup- plied at same price. WilTTorporation PpoducU for Every Laboratory Ouarantecd Mlhout Reservation Rochester . ^ - Y BIOLOGICAL STAINS A Complete List for Bacteriology, Pathology, Zoology, Josinl Botany, Etc. CHEMICAL INDICATORS Indicators for Every Purpose, including the Hydrogen-Ion Indicators (Clark and Lub's List, Sorensen's List, Etc.) FINE ORGANIC CHEMICALS Especially Prepared for Laboratory Use. Write for our NEW CATALOGUE giving the entire list, together with Methods for Staining, Solubilities, Indicator Chart, Etc. Coleman & Bell Products may be secured from Laboratory Supply Houses through- out the United Slates and in Foreign Countries, or may be ordered direct. The Coleman & Bell Company, Norwood, Ohio, U. S. A. Formerly National Stain & Reagent Co. JOURNAL OF BACTERIOLOGY "National" Biological Dyes These are a few of the National Specialties which laboratory workers will find to their ad- vantage to use: Acid Fuchsine Basic Fuchsine Fuchsinefor Endo Agar Brilliant Cresyl Blue Brilliant Vital Red {For blood 'vottnue estimation Cresyl Echt Violet Janus Green B Fluorescein, U. S. P. Nile Blue Sulfate Special Absolute Pure Methyl Alcohol, Acetone Free { Manufactureil especially far the preparation of blood stainitiis solutions Send for our catalog using coupon below for convenience. Pharmiiceuticul Di\'ision National Aniline & Chemical Company, Inc. Ne\v York HP^Txomrmm Pharm:iceiitical Division National Aniline Ut(!y LaMotte Standards tJUAKANTKKI) Section 1. — Standardized Indicator dyes. Covering a wide range of Il-ion con- centration. Supplied in dry form and in sterile atock solutionR. Each indicator is standardized in strict accordance with the specifications f>f Clark fi Liihs. (Jr. lUct., Vol. II, Noa. 1,2, 3, 1917. Common Nam* Color (bat-ge Pb V.1u. Thymol Blue (acid ranc*) red-yellow I.2-J.8 Methyl Orange Rroinphrnol Blue red-yellow 2»-I.O vellow-blue 30-«.e Reaorcin Illiie pink-blue 4 0-7.2 Methyl I{eY RKVIKW, which has a really international collaboration IT IS THK ONXY Kl-yil'AV, of absolutely world-wide circulation IT IS TUK ONLY KKVIIiAV, occupyinc itself with the sj-nthesis and unification of knowledge, which d als with the fundamental questions of all the sciences: history of the sciences, mathematics, astronomy, geology, physics, chcmi5tr\', biology, psyschology and sociology. IT IS TIIK ONLY REVIEW, which by means of enquiries among the most eminent scientists and writers on: I he philosophical principles of the various sciences; The most fundamental astronomical and physical questions of current interest; The contribution given by the various countries to the different branches of knowledge; The question of vitalism; The social question; The great international ques- tions raised l)y the world war, makes .i study of the most important questions interesting scientific and intellectual circles throughout the world. Abbot - Arrhcnius - Ashley - Bayliss - Beicbmann - Benes • Bigounbn - Boblin - Bohn - Bonncsen - Ilorcl - Bottazzi - linuty Bragg ■ Btillouin - Bruni • Burdick - Carracido - Carver - Castelouovo - Caullcry - Chambcrlin - Charlicr - Ciamician - ClaparMe ■ Clark ■ Coslanlin - Crommelin - Crowther - Darwin - DelaKC - De Martonne - De Vries - Durkhcim - Eddington • Edgeworth " Emery • Enriaues - Fabry - Eindlay - Fisher - Foi - Fowler - Frcdericq ■ Galeotti - Go!gi - Gregory - Guignebert - Harper - Hartog ■ Hfibcrg - Hinks - Hopkins - Ifiiguez - Innes - Janet - Jespersen - Kapteyn - Karpi ski - Kaye - Kidd Knibbs - Langevin - Lebedcw Lloyd Morgan - Lodge - Loisy • Lorentz - Loria - Lowell - MacBride Matruchot ■ Maunder - Meillel - Moret - Muir • Pareto - Piano • Pearl - Perrin - Picard - Pigou - Plans - Poincari - Puiseux - Ribaud - ReuterskjSld - Rcy Pastor - Righi - Rignano • Rus sell • Rutherford - Sagnac - Sarto i - Sayce - Schiaparclli - Scott - See - ScUcman - Sliapley - Sherrington ■ Soddy - Starling - Stojan- o\ ich - Struycken - Svedberg - Tannery - Teixeira - Thalbitzer • Thomson - Thorndike - Turner - Vinogradoff - Volterra - Von Zeipel - U ebb - Weiss - WcstermarcK - WickscU - U'illcy - Xenopol - Zecman - Zeuthen and more than a hundred others. *• Scicntio "publishes itsarticles in the language of its author, and joins to the principal text a supplement containing the French translalio ■ of all articles that are not in French. {Write /or a Specimen Number to the General Secretary of "Scientia, Milan, sendinf:. simply to defray postal and other expenses, 2 francs in stamps of your country. ANNUAL SUBSCRIPTION: $10. post free. OFFICE: 43 Foro Bonaparte, Milan (Italy) General Secretary: Doct. Paolo BoKETTi. .\ray be ordered from the authorized agents in the United States: WILLIAMS & WILKINS COMPANY Publishers of Scientific .lournals and Books MOUNT ROYAL and GUILFORD AVENUES, BALTIMORE (Maryland, U. S. A.) 12 JOURNAL OF BACTERIOLOGY JOURNAL OF BACTERIOLOGY Contents SEPTEMBER, 1922— Vol. VII— No. 5 S. Henry Ayers and Courtland S. Mudge. The Relation of Vitaniines to the Growth of a Streptococcus James M. Sherman and George E. Holm. Salt Effects in Bacterial Growth. II. The Growth on Bact. coli in Relation to H-Ion Concentration. Ivan E. Wallin. A Note on the Morphology of Bacteria Symbiotic in the Tissues of Higher Organisms. WiLBURT C. Davison. Observations on the Properties of Bacteriolysants (D'Her- elle's Phenomenon, Bacteriophage, Bacteriolytic Agent, etc.). Part I. WiLBURT C. Davison. Observations on the Nature of Bacterioly.sants (D'Her- elle's Phenomenon, Bacteriophage, Bacteriolytic Agent, etc.). Part II. Geor(;e F. Reddish and Leo F. Rettger. Clostridium putrificum ( B. putrificus Bienstock), a Distinct Species. .1. Howard Brown and Paul E. Howe, Transparent Milk as a Bacteriological Medium. G. J. Hucker and \A'. a. Wall. The Use of Agar Slants in Detecting Ammonia Production and its Relation to the Reduction of Nitrates. H. J. Conn, K. N. Atkins, H. J. Brown, F. Eberson, G. E. Harmon, G. J. Hucker, F. W. Tanner and S. A. Waksman. Methods of Pure Culture Study. Report of Committee on Bacteriological Technic. H. .1. Conn. An Investigation of American Gentian Violets. Report of Committee on Bacteriological Technic. JULY, 1922— Vol. VII— No. 4 L. D. BusHNELL. Quantitative Determinations of Some of the Biochemical Changes Produced by a Saprophytic Anaerobe. G. S. Wilson. The Proportion of Viable Bacteria in Young Cultures with Especial Reference to the Technique Employed in Counting. H. ,1. Conn. A Method of Detecting Rennet Production by Bacteria. MAY, 1922— VOL. VII— No. 3 J. Howard Mueller. Studies on Cultural Requirements of Bacteria. I. J. Howard Mueller. Studies on Cultural Requirements of Bacteria. II. Selman a. Waksman. A Method for Counting the Number of Fungi in the Soil. Lethe E. Morrison and Fred W. Tanner. Studies on Thermophilic Bacteria. I. Aerobic Thermophilic Bacteria from Water. Robert G. Green. An Apparatus for the Rapid Measurement of Surface Tension. MARCH, 1922— Vol. VII— No. 2 F. C. Harrison. Our Society. Victor Burke. Notes on the Gram Stain with Description of a New Method. Barnett Cohen. Disinfection Studies. The Effects of Temperature and Hydrogen Ion Concentration upon the Viability of Bact. coli and Bact. typhosum in Water. Selman A. Waksman. Microorganisms Concerned in the Oxidation of Sulfur in the Soil. I. Introductory. Selman A. Waksman and J. S. Jofpe. Microorganisms Concerned in the Oxida- tion of Sulfur in the Soil. II. Thiobacillus Thiooxidans, a New Sulfur-oxidiz- ing Organism Isolated from the Soil. E. B. Fred and W. H. Peterson. The Production of Pink Sauerkraut by Yeasts. Constantino Gorini. Studies on the Biology of Lactic Acid Bacteria: A Summarj- of Personal Investigations. L. D. BusHNELL. A Method for the Cultivation of Anaerobes. L. D. Bushnell. Influence of Vacuum upon Growth of Some Aerobic Spore-Bear- ing Bacteria. H. R. Baker. Substitution of Brom-Thymol-Blue for Litmus in Routine Laboratory Work. W. A. Wall and A. H. Robertson. The Use of Domestic Methylene Blue in Staining Milk by the Breed Method. ORDER BLANK Williams & Wilkins Company Publishers of Scientific Journals and Books Baltimore, Maryland, U. S. A. Kindly enter my subscription for the Journal of Bacteriology Volume V'll, 1922, for which I enclose $5.00 (Canada, $5.25; foreign, $5.50), or, send me a state- ment and I will remit. Signed... Address. INFORMATION FOR CONTRIBUTORS Although there are numerous joumala in the United States that deal with varioua ■pecial phaaea of bacteriology (as applied to Medicine, Sanitary Science, Agriculture and the like , Thb Journal of Bactesioloot is the only journal in the English language to represent the science as a whole. The Society of American Bactoriologistfl has established the Jouenal of BACTEitiouxiT as its official organ and as a medium for the discussion of the more general problems of the science — the structure and physiology of the microbes, the inter-relationships of mi- crobic types, the eflecto of physical and chemical agents upon microbic life, the mutual interactions of microbes growing together in varioua media, the nutritional needs and products of metabolic activity of various microbes, and new methods of laboratory tech- nique— and similar advances in knowledge which are so fundamental as to be of vital interest to workers in all parts of this great field. The Journal of Bacterioloot includes in its scope not only the bacteria but other related micro-organismR, yeasts, molds, pro- tosoa, etc. While it is planned to make the Journal in particular an or^an far the more fundamental and general aspects of bacteriology, it will necessarily include many papers whose interest is mainly technical, particularly in those fields of bacteriology which have now no satisfactory organ of publication at their disposal. Manuscripts should be sent to Prof. C.-E. A. Winslow, Yale Medical School, New Haven, Conn. All other communications pertaining to editorial work should be addressed to Major A. Parker Hitchens, M. C, Army Medical School, Washington, D. C. Twenty-five reprints, without covers, of articles will be furnished gratia to contributors when ordered in advance. A table showing cost, with an order slip, is sent with proof. INFORMATION FOR SUBSCRIBERS The Journal of Bacteriologt is issued bi-monthly, in January, March, May, July, September, and November. A volume consists of approximately 600 pages. Subscriptions to the Journal are taken for the volume only. Under the present plans, one volume is issued a yeu. Subscription Price: 1922 volume, Vol. VII, $5.00, United States, Mexico, Cuba; $5.25, Canada; $5.50, elsewhere, net postpaid. Foreign subscriptions to the current volume accepted at par of exchange. Back Volumes are supplied on orders for Vols. I to VI, inclusive. Price, $36.00, United States, Mexico, Cuba; $37.50, Canada; $39.00, elsewhere. Extra single volumes will be supplied on request when in stock at $5.00, United States, Mexico, Cuba; $5.25, Canada; $5.50, elsewhere. Prices are net postpaid. Single numbers can be furnished, in many in- stances, a-t $1.00 a copy. Foreign subscriptions to the back volumes accepted at the current rate of exchange. Correspondence concerning business matters should be addressed to the Williams & Wilkins Company, Publishers of Scientific Journals and Books, Mount Royal and Guilford Avenue!!, Baltimore, U. S. A. Subscriptions are received: For Argentina and Uruguay : Beutelspacher & Cia., Sarmiento 815, Buenos Aires, Argentina- For Australia: Stirling & Co., 317 Collins Street, Melbourne. For Belgium: Henri Lamertin, 58 Rue Coudenberg, Bru.xelles. For the British Empire, except Australia and Canada: Bailliere, Tindall & Cox; Arthur F. Bird; Wm. Dawson & Sons, Ltd.; Dulaii & Co., Ltd.; H. K. Lewis & Co.; David Nutt; J. Poole & Co.; Wheldon k Wesley; Williams & Norgate, London; James Thin, Edinburgh. For Canada: Wm. Dawson & Sons, Ltd., 87 Queen Street East, Toronto, Canada. For Denmark: H. Hagerup's Boghandel, Gothersgade 30, Kobenhavn. For France: Emile Bougault, 48 Rue des Ecoles, Paris. For Germany: R. FriedlSnder